ACOD-itaconate in macrophage attenuates oxidative stress and inflammation in benign airway stenosis by upregulating and transferring FTH1

Abstract

The oxidative stress of macrophage plays pivotal roles of acute and chronic inflammation and chronic fibrotic phases, in which the metabolic mechanism needs to be further explored. In our research, multi-omics analyses of human and murine during Benign airway stenosis (BAS) biopsy identified ACOD1 as a hallmark of immunometabolic regulation during acute inflammation stage. ACOD1 knockout aggravated both acute and chronic inflammation, which increased the granulation tissue formation. The ACOD1-itaconate axis, along with its derivative, 4-octyl itaconate (4-OI), orchestrated acute and chronic inflammation, which attenuated the fibrosis of BAS. 4-OI upregulated FTH1 expression in macrophages by activating NRF2, which effectively suppressed oxidative stress and acute inflammation. Furthermore, 4-OI promoted the packaging of FTH1 into macrophage-derived exosomes, which were transferred to fibroblasts in a SCARA5-dependent manner, inducing fibroblast ferroptosis and alleviating chronic fibrosis. In sum, this study illustrates that the ACOD1-itaconate metabolic axis decreases oxidative stress and inflammation in macrophage, which attenuates fibrosis by inducing FTH1 transfer, offering a therapeutic target for fibrotic airway diseases.

Highlights

• •scRNA sequence reveals that ACOD1 is significantly upregulated during BAS.
• •ACOD1 deficiency aggravates acute and chronic inflammation of BAS.
• •Itaconate derivative 4-OI attenuates acute and chronic inflammation of BAS.
• •4-OI-induced high expression of FTH1 in macrophages restricts oxidative stress and acute inflammation.
• •Macrophage exosomal FTH1 transferred via SCARA5 induces fibroblast ferroptosis to attenuate BAS inflammation and fibrosis.

1. Introduction

Benign airway stenosis (BAS), which commonly develops as a result of endotracheal intubation or tracheostomy, presents a complex clinical condition marked by recurrent dyspnea and potentially life-threatening airway obstruction [1]. The high incidence of recurrence, despite considerable advances in endoscopic interventions [2], underscores a pressing need for new therapeutic strategies that directly address the underlying pathophysiology. Crucially, the pathogenesis of BAS is characterized by two interconnected yet mechanistically distinct phases: early acute inflammatory stage dominated by immune cell infiltration and cytokine storms, followed by chronic inflammatory stage featuring excessive extracellular matrix deposition and fibroblast proliferation [3,4]. Macrophages play a crucial role in both stages, dynamically transitioning from a pro-inflammatory role during the acute inflammation phase to an anti-fibrotic function in the chronic inflammation phase [5,6]. However, the intrinsic metabolic drivers regulating this functional shift in BAS remain poorly understood.

Recent breakthroughs have highlighted immunometabolic reprogramming as a central regulator of macrophage function [7]. The metabolite itaconate, derived from the tricarboxylic acid (TCA) cycle through the enzyme aconitate decarboxylase 1 (ACOD1), has emerged as a potent anti-inflammatory agent in models of sepsis and autoimmune diseases [8,9]. Itaconate exert their effects by activating the master antioxidant regulator nuclear factor erythroid 2-related factor 2 (NRF2), promoting its nuclear translocation and subsequent transcription of antioxidant response element (ARE)-driven genes, thereby inhibits the production of inflammatory cytokines [10,11]. 4-octyl itaconate (4-OI) is a derivative of itaconate that offers improved cellular permeability and can degrade into itaconate, thereby exerting effects similar to those of endogenous itaconate [12]. Despite its established role in systemic inflammation, the function of the ACOD1/itaconate axis in airway pathologies, particularly its impact on the critical transition from inflammation to fibrosis in BAS, remains completely unexplored. Furthermore, the potential for itaconate or 4-OI to modulate the crucial macrophage-fibroblast crosstalk is unknown, representing a major therapeutic opportunity.

In this study, we have identified the ACOD1/itaconate pathway as a major regulator that governs phase-specific immune responses in the pathogenesis of BAS. Our findings indicate that this pathway mediates protective mechanism in BAS: During the acute inflammation phase, the ACOD1/itaconate pathway and exogenous 4-OI, upregulates the ferritin heavy chain 1 (FTH1) in macrophages, suppressing both inflammation and oxidative stress. In the chronic inflammation phase, 4-OI stimulates macrophages to package FTH1 into exosomes and secrete it. These exosomal FTH1 are taken up by fibroblasts via SCARA5 receptors, leading to iron overload and subsequent ferroptosis. These findings highlight the ACOD1/itaconate pathway as a potential therapeutic target and support further investigation into its role in precision-based strategies for BAS management.

2. Result

2.1. Increased ACOD1 expression in BAS patients and mice

BAS progresses through distinct inflammatory and fibrotic phases, yet early molecular drivers remain poorly defined. Single-cell analysis was constructed by integrating data from three normal tracheal samples and four BAS granulation tissue samples (Fig. 1A). KEGG enrichment analysis revealed a significant enrichment of “metabolic pathways” within the granulation tissue (Fig. 1B). Among numerous differentially expressed genes in metabolic pathways, the expression of immunometabolic gene ACOD1 was majorly in the granulation tissue (Fig. 1C). In human specimens, immunohistochemistry and immunofluorescence revealed markedly increased expression of ACOD1 in BAS granulation tissues compare to normal tracheae (Fig. 1D–G).

Fig. 1.

ACOD1 expression is significantly upregulated in human airway stenosis granulation tissue and early inflammatory stage murine BAS models.

A. UMAP visualization of Normal group and Granulation group. B. TOP 10 of enriched KEGG terms of metabolism in Granulation group compared with Normal group. C. Heatmap displaying the DEGs in metabolic pathway identified between Normal group and Granulation group, and Violin plot showing the expression distribution of ACOD1 across all cells from normal (blue) and granuloma (red) tissues. D, E. Representative images of ACOD1 immunohistochemical staining and quantitative of immunoreactive area in normal trachea group and granulation group. Scale bars indicate 200 μm and 50 μm (n = 3 sample per group). F, G. Representative tracheal immunofluorescence images and mean fluorescence intensity of ACOD1(red) in normal trachea group and granulation group (n = 3 sample per group). Scale bars indicate 200 μm and 50 μm. H. Schematic diagram illustrating the mouse tracheal stenosis model establishment and sampling timeline for transcriptome sequencing at day 1 post-injury. I. Volcano plot of differentially expressed genes (DEGs) in mouse tracheal tissue at day 1 post-stenosis compared to sham controls (n = 3 mice per group). J. Heatmap displaying the top 15 upregulated DEGs identified in (F). ACOD1 expression is highlighted. K, L. Western blot analysis showing the levels of ACOD1 in the trachea of BAS group and control group (n = 3 mice per group) at 24 h. M. qPCR analysis confirms elevated ACOD1 mRNA levels in BAS group compare to control group (n = 5 mice per group). N, O. Representative tracheal immunofluorescence images and mean fluorescence intensity of ACOD1(green) in BAS group and control group (n = 5 mice per group). Scale bars indicate 200 μm and 50 μm. P, Q. Representative FISH images and mean fluorescence intensity of ACOD1(yellow) in BAS group and control group (n = 5 mice per group). Scale bars indicate 200 μm and 50 μm. R, S. Representative images of ACOD1 immunohistochemical staining and quantitative of immunoreactive area in control group and BAS group at day1,4,7. Scale bars indicate 200 μm and 50 μm (n = 4 mice per group). Data are presented as the mean ± SEM. Two-sided student's T-test were used in E, G, L, M, O and Q. One-way ANOVA analysis followed by Tukey post hoc multi-comparison test was used in S.

Next, BAS mouse model was established by inducing airway injury through mechanical scraping of the mouse trachea with a brush, and on the day1 after trachea injury, tracheal tissues were collected from the mice for transcriptome sequencing (Fig. 1H). RNA sequencing of day 1 post-injury tracheae revealed 3127 differentially expressed genes (1534 upregulated, 1593 downregulated), with ACOD1 exhibiting significant upregulation in volcano plot analysis (Fig. 1I). Consistent with this, ACOD1 ranked among the top 15 most upregulated transcripts (Fig. 1J). Western blot analysis revealed that expression of ACOD1 were elevated at day1 post-tracheal injury compared with the control group (Fig. 1K and L). qPCR also showed high expression of ACOD1 on the day1 after tracheal injury (Fig. 1M). Immunofluorescence and FISH analysis revealed high ACOD1 expression in the tracheal tissues, specifically localized to subepithelial inflammatory lesions (Fig. 1N–Q). Subsequently, immunohistochemical analysis was performed on tracheal tissues collected from BAS mice at different time points, and temporal IHC further demonstrated peak expression at day1 after tracheal injury (Fig. 1R and S). Collectively, these results indicate that ACOD1 as a signature molecule of early BAS inflammation, prompting investigation of its functional role.

2.2. ACOD1 knockdown exacerbates acute and chronic inflammation in BAS mice

BAS progresses from an early acute inflammatory phase to a late fibrotic phase characterized by tissue repair and extracellular matrix deposition. To further investigate the role of the ACOD1 gene in BAS, we conducted experiments using ACOD1 knockdown mice and collected samples during the inflammatory(day1 after tracheal injury) and fibrotic phases(day7 after tracheal injury). In the acute inflammatory stage, HE staining results of mouse trachea showed that ACOD1 knockout mice had more severe inflammatory edema and inflammatory cell infiltration in the trachea after modeling compared to WT mice (Fig. 2A and B). In the chronic inflammatory stage, ACOD1 knockdown significantly exacerbated airway stenosis following tracheal injury (Fig. 2C and D). Consistent with this structural remodeling, micro-CT imaging revealed a pronounced increase in tracheal stenosis in ACOD1 KO mice (Fig. 2E and F), indicating enhanced fibrotic obstruction. Additionally, immunofluorescence analysis confirmed that the expression of fibrotic markers α-SMA and COL1 was substantially upregulated in ACOD1 KO BAS mice compare to WT BAS mice(Fig. 2G and H). Thus, the ACOD1 plays a critical protective role in BAS. Its deficiency exacerbates acute and chronic inflammation.

Fig. 2.

ACOD1 Deficiency Exacerbates acute and chronic inflammation in BAS mice.

A. Representative images of H&E staining in control group (WT mice), BAS group (WT mice) and BAS group (ACOD1-KO mice) in day1 after modeling. Scale bars indicate 200 μm and 50 μm. B. Quantitative comparison of inflammatory area among three groups(n = 5 mice per group). C. Representative images of H&E staining in control group (WT mice), BAS group (WT mice) and BAS group (ACOD1-KO mice) in day7 after modeling. Scale bars indicate 200 μm and 50 μm. D. Quantitative comparison of tracheal stenosis degree among three groups (n = 5 mice per group). E. Micro CT scans of different groups in Horizontal, Coronal and Sagittal plane with red arrows indicating the stenosis site. F. Quantitative comparison of the area of tracheal stenosis measured by SYNPASE 3D (n = 3 mice per group). G, H. Representative immunofluorescence images and mean fluorescence intensity of COL1(green) and αSMA (red) in control group (WT mice), BAS group (WT mice) and BAS group (ACOD1-KO mice). Scale bars indicate 200 μm and 50 μm (n = 5 mice per group). Data are presented as the mean ± SEM. One-way ANOVA analysis followed by Tukey post hoc multi-comparison test was used in B,D,F,H.

2.3. 4-OI attenuates acute and chronic inflammation in BAS mice

Itaconate is the classical downstream product of ACOD1, we then assessed functional consequences of itaconate in early and lately BAS pathology. Liquid chromatography-tandem mass spectrometry (LC-MS) was used to measure the content of itaconate, and result show that significantly elevated itaconate levels in day1 post-injury tracheae compare to controls (Fig. 3A). To evaluate therapeutic potential, 4-octyl itaconate (4-OI) was administered intraperitoneally 12 h prior to injury induction and immediately following injury establishment (within 5 min) with tissues harvested at 24h (Fig. 3B). HE analysis revealed that vehicle-treated mice exhibited severe subepithelial edema and inflammatory cell infiltration, whereas 4-OI treatment markedly reduced the area of inflammatory cell infiltration (Fig. 3C and D). Consistent with these findings, qPCR demonstrated 4-OI mediated suppression of pro-inflammatory cytokines, with IL6 and IL1β mRNA levels decreased (Fig. 3E and F). Immunofluorescence quantification further confirmed reduced levels of IL-6 and IL-1β (Fig. 3G–J) and the oxidative stress marker 4-HNE (Fig. 3K and L) in BAS mice treated with 4-OI compared to vehicle controls. Next, we further investigated whether supplementing exogenous itaconate could attenuate fibrotic progression. Using a therapeutic administration protocol of 4-OI (Fig. 3M), We observed that 4-OI treatment markedly suppressed injury-induced granulomatous hyperplasia in BAS mice at day7 after modeling, correlating with reduced airway stenosis (Fig. 3N and O). Masson's trichrome staining further demonstrated that elevated collagen deposition in BAS mice was significantly reduced by 4-OI (Fig. 3P and Q).Taken together, these results showed that 4-OI could attenuates acute and chronic inflammation in BAS mice.

Fig. 3.

4-OI attenuates acute and chronic inflammation in the BAS mice.

A. Itaconate levels are significantly elevated in BAS group at day 1 post-stenosis compared to sham controls (n = 4 mice per group). B. Schematic diagram of the experimental design for 4-OI treatment. Mice received exogenous 4-OI (or vehicle control) and tracheal tissues were harvested at 24 h post-stenosis induction. C,D. Representative images of H&E staining and quantitative comparison of inflammatory area in BAS group and control group at day1 after modeling (n = 5 mice per group). Scale bars indicate 200 μm and 50 μm. E,F Quantitative of qPCR analysis of IL6 and IL1β mRNA levels in control group, BAS group and BAS group treated with 4-OI at day1 after modeling (n = 5 mice per group). G-J. Representative tracheal immunofluorescence images and mean fluorescence intensity of IL6(red), IL1β(green) in control group, BAS group and BAS group treated with 4-OI at day1 after modeling (n = 5 mice per group). K,L. Representative tracheal immunofluorescence images and mean fluorescence intensity of 4-HNE in control group, BAS group and BAS group treated with 4-OI at day1 after modeling (n = 5 mice per group).Scale bars indicate 200 μm and 50 μm. M. Schematic diagram of the experimental design for 4-OI treatment. N,O. Representative images of H&E staining and quantitative comparison of tracheal stenosis degree in control group, BAS group and BAS+4-OI group at day7 after modeling. Scale bars indicate 200 μm and 50 μm. P,Q. Representative images of Masson staining and relative collagen content in control group, BAS group and BAS+4-OI group at day7 after modeling. Scale bars indicate 200 μm and 50 μm (n = 5 mice per group). Data are presented as the mean ± SEM. Two-sided student's T-test were used in A. One-way ANOVA analysis followed by Tukey post hoc multi-comparison test was used in D,E,F,H,J,L,O,Q.

2.4. ACOD1-itaconate axis decreases inflammatory responses in macrophages

To further investigate the cellular localization of ACOD1, we analyzed the single-cell data from BAS patients. Based on the marker genes of different cell populations, we identified and categorized 14 distinct cell clusters (Supplementary Fig. 1A and B). Single-cell transcriptomics revealed that ACOD1 was predominantly enriched within macrophage clusters (Fig. 4A). Subsequently, we performed additional subclustering specifically on the macrophage population (Supplementary Fig. 1C and D), and high ACOD1 level was enriched in Mc4 cluster (Fig. 4B). Gene Ontology analysis confirmed Mc4 cluster enrichment in inflammatory response (Fig. 4C). Immunofluorescence co-staining of granulation tissues from BAS patients revealed co-localization of ACOD1 with the macrophage marker CD68, which was rarely observed in normal tracheal tissue (Fig. 4D). Similarly, tracheal sections from BAS murine models at day 1 post-injury showed the presence of ACOD1⁺CD68⁺ macrophages (Fig. 4E). To investigate the relationship between ACOD1 expression in macrophages and inflammation, we stimulated the macrophages with LPS and found that ACOD1 expression exhibited a time-dependent increase (Fig. 4F and G). Immunofluorescence confirmed enhanced ACOD1 protein levels following LPS treated (Fig. 4H and I), while quantitative PCR demonstrated significant elevation of ACOD1 transcripts at this timepoint (Fig. 4J). Next, Transfection with ACOD1 overexpression plasmids (Supplementary Fig. 2A and B) significantly suppressed the expression of inflammation-related factors induced by LPS (Fig. 4K), while transfection with ACOD1 siRNA (Supplementary Fig. 2C and D) significantly increased the expression of inflammatory factors induced by LPS (Fig. 4L). Moreover, in terms of oxidative stress, overexpression of ACOD1 effectively reduced ROS levels, while knockdown of ACOD1 increased ROS levels (Fig. 4M and N). Under inflammatory conditions, inflammatory stimuli potently augment macrophage-derived itaconate synthesis (Fig. 4O). Crucially, exogenous 4-OI treatment effectively decreased inflammatory cytokine transcription (Fig. 4P) and significantly mitigated oxidative damage (Fig. 4Q and R). Collectively, these findings establish the ACOD1-itaconate pathway as a central regulatory mechanism governing macrophage-mediated inflammation resolution.

Fig. 4.

ACOD1 is predominantly expressed in pro-inflammatory macrophages and regulates inflammation and oxidative stress via itaconate.

A. Single-cell RNA sequencing (scRNA-seq) analysis reveals enriched AcoD1 expression specifically within macrophage clusters. B. the highest ACOD1 expression in the Mc4 subpopulation. C. Gene Ontology (GO) analysis of the Mc4 macrophage subcluster shows significant enrichment for biological processes related to inflammatory response. D. Representative images of ACOD1(green) and CD68(red) immunofluorescence staining in normal trachea group and granulation group. Scale bars indicate 200 μm and 50 μm. E. Representative images of ACOD1(green) and F4/80(red) immunofluorescence staining in normal group and BAS group. Scale bars indicate 200 μm and 50 μm. F, G. Western blot analysis showing the levels of ACOD1 in macrophage stimulated with LPS at different time, n = 3 independent experiments. H, I. Representative immunofluorescence images and mean fluorescence intensity of ACOD1(green) in macrophage stimulated by LPS or not (n = 3 independent experiments). Scale bars indicate 10 μm. J. qPCR analysis confirms elevated ACOD1 mRNA levels in macrophage stimulated by LPS compare to control group (n = 3 independent experiments). K. qPCR analysis shows that overexpression of ACOD1 (ACOD1-OE) in macrophages significantly reduces LPS-induced mRNA expression of IL6 and iNOS (n = 3 independent experiments). L. qPCR analysis shows that knockdown of ACOD1 in macrophages exacerbates LPS-induced mRNA expression of IL6 and iNOS (n = 3 independent experiments). M, N. Representative immunofluorescence images and mean fluorescence intensity of ROS (red) in four macrophage group: control, LPS, LPS + ACOD1-OE and LPS + ACOD1 konckdown (n = 3 independent experiments). Scale bars indicate 50 μm. O. Itaconate levels are significantly elevated in macrophage stimulated by LPS compared to controls (n = 3 independent experiments). P. qPCR analysis demonstrates that treatment with 4-OI suppresses LPS-induced mRNA expression of IL6 and IL1β in macrophages compared to LPS-stimulated vehicle control (n = 3 independent experiments). Q, R. Representative immunofluorescence images and mean fluorescence intensity of ROS (red) in four macrophage group: control, 4-OI, LPS and LPS+4-OI (n = 3 independent experiments). Scale bars indicate 50 μm. Data are presented as the mean ± SEM. Two-sided student's T-test were used in I,J,O. One-way ANOVA analysis followed by Tukey post hoc multi-comparison test was used in J, K, L, N, P, R.

2.5. The ACOD1-itaconate-NRF2 axis upregulates FTH1 expression to execute restricts oxidant stress and acute inflammation

In this study, the Mc4 cluster of macrophages was found to exhibit high expression of ACOD1. Subsequently, this cluster was further divided into two groups: one cluster expressing ACOD1(ACOD1⁺ cluster) and another not expressing it (ACOD1^- cluster) (Fig. 5A). With KEGG pathway analysis demonstrating strong enrichment of necroptosis signaling specifically in ACOD1⁺ cluster (Fig. 5B) and within necroptosis-related genes, FTH1 exhibited the most pronounced differential expression (Fig. 5C). Single-cell data indicate that the expression of FTH1 is higher in the granulation tissue of patients with BAS compared to normal tracheal tissue (Fig. 5D). Immunohistochemistry confirms that FTH1 expression is higher in granulation tissue than in normal tracheal tissue (Fig. 5E and F). To further investigate the relationship between ACOD1/itaconate and FTH1, mice were treated with 4-OI. WB analysis revealed that FTH1 expression was elevated following tracheal injury and further increased upon administration of 4-OI (Fig. 5G and H). Immunofluorescence analysis also confirmed the upregulation of FTH1 expression in the tracheal tissue of BAS model mice after 4-OI treatment (Fig. 5I and J). Critically, multiplex immunofluorescence identified co-expressing ACOD1⁺FTH1⁺ macrophages infiltrating in the granulation tissue of BAS patient (Fig. 5K). Similarly, in the BAS mouse model, macrophages exhibiting high expression of FTH1 and ACOD1 were present following tracheal injury (Fig. 5L). In vitro experiments, we further established the relationship between FTH1 and itaconate in macrophages. 4-OI failed to alter FTH1 basally but rescued LPS-suppressed FTH1 expression in macrophages (Fig. 5M and N). Building on established roles of FTH1 in oxidative stress regulation [13,14], we demonstrated that FTH1 knock down in macrophages (Supplementary Fig. 3A and B) elevated intracellular ROS levels, indicating its essential function in mitigating oxidative damage (Supplementary Fig. 3C and D). Immunofluorescence quantification demonstrated that 4-OI attenuated oxidative damage, but this protective effect was abolished when FTH1 was knocked down in macrophages, indicating that FTH1 as the essential executor of ACOD1-itaconate-mediated antioxidant responses (Fig. 5O and P).

Fig. 5.

ACOD1/itaconate axis upregulates FTH1 to mitigate oxidative stress in macrophages.

A. Re-clustering of the Mc4 macrophage subpopulation based on AcoD1 expression identifies ACOD1⁺ and ACOD1^- subsets. B. KEGG pathway enrichment analysis shows the necroptosis pathway is most significantly enriched in ACOD1⁺ macrophages compared to ACOD1^- subsets. Top 5 enriched pathways shown. C. Differential gene expression analysis within the necroptosis pathway identifies FTH1 as the most upregulated gene in ACOD1⁺ macrophages. D. Quantitative comparison of FTH1 expression levels in scRNA-seq data between normal trachea group and granulation group. E, F. Representative images of FTH1 immunohistochemical staining and quantitative of immunoreactive area in normal trachea group and granulation group. Scale bars indicate 200 μm and 50 μm (n = 3 sample per group). G, H. Western blot analysis showing the levels of ACOD1 in the trachea of control group, BAS group and BAS+4-OI group (n = 3 mice per group) at 24 h. I, J. Representative tracheal immunofluorescence images and mean fluorescence intensity of FTH1(green) in control group, BAS group and BAS+4-OI group (n = 5 mice per group) at 24 h. Scale bars indicate 200 μm and 50 μm. K. Representative immunofluorescence images of CD68(red), FTH1(green) and ACOD1(yellow)in normal trachea group and granulation group. Scale bars indicate 200 μm and 50 μm. L. Representative tracheal immunofluorescence images of F4/80(red), FTH1(green) and ACOD1(yellow)in control group and BAS group. Scale bars indicate 200 μm and 50 μm. M, N. Western blot analysis in macrophages reveals that 4-OI rescues LPS-induced FTH1 downregulation (n = 3 independent experiments). O, P. Representative immunofluorescence images and mean fluorescence intensity of ROS (red) in 4-OI reduces LPS-induced 4-HNE in controls, but not in FTH1-knockdown macrophages. (n = 3 independent experiments). Scale bars indicate 50 μm. Q, R. Western blot and quantitative analysis of FTH1 in macrophages after transfection with siNRF2 and treated or not with 4-OI, LPS (n = 3 independent experiments). Data are presented as the mean ± SEM. Two-sided student's T-test were used in D,F. One-way ANOVA analysis followed by Tukey post hoc multi-comparison test was used in H, J. Two-way ANOVA analysis followed by Tukey post hoc multi-comparison test was used in N, P, R.

Next, we investigate how ACOD1 regulates the expression of FTH1. ACOD1/Itaconic acid axis exerts immune regulatory effects through multiple pathways, among which the downstream NRF2 pathway represents its typical anti-inflammatory and antioxidant mechanisms [15,16]. RNA-seq and immunohistochemistry revealed significant activation of the NRF2 pathway in the trachea after injury (Supplementary Fig. 4A–D). Functional studies using the NRF2 inhibitor ML385 demonstrated that the protective effects of 4-OI, the attenuation of submucosal edema, inflammatory cell infiltration, and the reduction of pro-inflammatory cytokines (IL6, IL1β) and oxidative stress (4-HNE)were dependent on NRF2, as ML385 co-treatment abrogated these benefits (Supplementary Fig. 4E–J). Immunofluorescence confirmed NRF2 expression in infiltrating macrophages during inflammation (Supplementary. Fig. 5A). Then, we use siRNA knock down the ACOD1 gene in macrophages(Supplementary. Fig. 5B–C), and in vitro mechanistic experiments established that both ACOD1 overexpression and 4-OI treatment promoted NRF2 nuclear translocation in macrophages(Supplementary Fig. 5D–G) and that NRF2 knockdown abolished their anti-inflammatory and antioxidant effects(Supplementary Fig. 5H and I), conclusively positioning NRF2 as the critical downstream mediator of the ACOD1-itaconate axis in BAS mice. Building on this, we further investigated the relationship between NRF2 and FTH1. WB analysis revealed that under inflammatory conditions, 4-OI promoted FTH1 expression in macrophages. However, this induction was abolished upon NRF2 knockdown (Fig. 5Q and R). In a complementary rescue experiment, transfection with an siRNA-resistant NRF2 plasmid construct restored FTH1 upregulation in NRF2-silenced macrophages following 4-OI treatment (Supplementary Fig. 6A and B).These results demonstrate that the effects of the ACOD1-itaconate axis are mediated by NRF2 activation, which subsequently upregulates FTH1 expression.

2.6. 4-OI induces fibroblast ferroptosis via exosomal transfer of macrophage FTH1

To investigate whether 4-OI enhances FTH1 release from macrophages, we performed ELISA assays on culture supernatants. While 4-OI alone did not alter basal FTH1 secretion, it synergistically increased FTH1 release under LPS-induced inflammatory conditions (Fig. 6A). We next examined whether FTH1 is secreted via exosomes. Exosomes were isolated from the supernatants of macrophages treated with LPS+4-OI and characterized by TEM, which revealed typical cup-shaped vesicles (Fig. 6B). NTA confirmed a size distribution consistent with exosomes(Fig. 6C). Western blot analysis detected the exosomal markers TSG101, CD63, and CD81, while the endoplasmic reticulum marker calnexin was absent, confirming exosome purity (Fig. 6D). To determine whether fibroblasts internalize macrophage-derived exosomes, we labeled exosomes with the fluorescent dye DiD and added them to fibroblast cultures (Fig. 6E). Immunofluorescence analysis revealed punctate DiD fluorescence within fibroblasts, confirming active uptake of macrophage-derived exosomes(Fig. 6F). Western blot analysis of exosomal content showed that FTH1 was enriched in exosomes from LPS+4-OI-treated macrophages compared to controls (Fig. 6G). To assess the functional role of exosome-mediated FTH1 transfer, we used the exosome secretion inhibitor GW4869. Pretreatment of macrophages with GW4869 significantly reduced FTH1 levels in recipient fibroblasts following co-culture, demonstrating that exosomal release is required for efficient FTH1 transfer (Fig. 6H and I).

Fig. 6.

4-OI promotes exosomal FTH1 secretion from macrophages and induces fibroblast ferroptosis.

A. ELISA detects elevated FTH1 secretion in supernatants of macrophages treated with LPS + 4-OI, but not 4-OI alone (n = 3 independent experiments). B. Representative TEM images of exosomes isolated from LPS+4-OI treated macrophages. Scale bars indicate 500 nm and 200 nm. C. Representative NTA of exosomes isolated from LPS+4-OI treated macrophages. D. Representative immunoblot shows exosome markers (TSG101, CD63, and CD81) and endoplasmic reticulum marker (Calnexin). E. Schematic diagram of exosome uptake experiment. F. Representative immunofluorescence images of DiD (red) and F-actin (green) in two fibroblasts group. Scale bars indicate 20 μm. G. Representative immunoblot shows FTH1 and CD63 in two EXO group. H,I. Representative immunofluorescence images and mean fluorescence intensity of FTH1 (red) in three fibroblasts group (n = 3 independent experiments). Scale bars indicate 50 μm. J. Experimental scheme: Macrophage-fibroblast co-culture system. K,L. Representative immunofluorescence images and mean fluorescence show co-culture with macrophages treated with LPS+4-OI significantly enhanced FTH1 internalization in fibroblasts. This effect was abolished when macrophages were subjected to FTH1 knockdown (LPS+4-OI + siFTH1), and was restored upon addition of recombinant FTH1 (LPS+4-OI + siFTH1+rFTH1) (n = 3 independent experiments). Scale bars indicate 50 μm. M,N. Representative immunofluorescence images and mean fluorescence of PGSK in fibroblasts after co-culture with the indicated macrophage groups: control, LPS+4-OI, LPS+4-OI + siFTH1, and LPS+4-OI + siFTH1+rFTH1 (n = 3 independent experiments). Scale bars indicate 100 μm. O,P. Representative immunofluorescence images and Mean fluorescence intensity of BODIPY C11(red) and oxidized form(green) in two group(n = 3 independent experiments) Scale bars indicate 50 μm. Q. Representative TEM images of differently treated fibroblasts group. Scale bars indicate 2 μm and 500 nm. R,S. Representative immunofluorescence images and mean fluorescence intensity of 4HNE (red) in three fibroblasts group (n = 3 independent experiments). Scale bars indicate 50 μm. Data are presented as the mean ± SEM. Two-sided student's T-test were used in P. One-way ANOVA analysis followed by Tukey post hoc multi-comparison test was used in A,I,L,N,S.

To establish causality, we performed knockdown and rescue experiments (Fig. 6J). Immunofluorescence analysis revealed that FTH1 levels increased in fibroblasts co-cultured with LPS+4-OI-treated macrophages, but this effect was abolished when macrophages were transfected with FTH1 siRNA (Fig. 6K and L). Importantly, supplementing exogenous FTH1 protein to co-cultures containing FTH1-knockdown macrophages restored FTH1 expression in fibroblasts(Fig. 6K and L). Consistent with these findings, intracellular labile iron levels in fibroblasts were elevated following co-culture with LPS+4-OI-stimulated macrophages, reduced upon macrophage FTH1 knockdown, and restored by exogenous FTH1 supplementation (Fig. 6M and N). To determine whether FTH1 transfer leads to ferroptotic cell death, we assessed key ferroptosis hallmarks in fibroblasts. Using the BODIPY 581/591 C11 probe, we observed a marked increase in oxidized (green) fluorescence in fibroblasts co-cultured with LPS+4-OI-treated macrophages compared to controls, indicating elevated lipid peroxidation (Fig. 6O and P). TEM revealed characteristic ferroptotic features in these fibroblasts, including shrunken mitochondria with increased membrane density and diminished cristae (Fig. 6Q). Immunofluorescence staining for 4-HNE showed enhanced signals in fibroblasts following co-culture with LPS+4-OI-treated macrophages (Fig. 6R and S). Notably, this increase was effectively attenuated by adding the iron chelator deferoxamine (DFO) to the culture medium, confirming iron dependency (Fig. 6R and S).

To validate the translational relevance of our findings, we established a co-culture system using THP-1-derived human macrophages and human tracheal fibroblasts (Supplementary Fig. 7A). Immunofluorescence analysis revealed that fibroblasts co-cultured with LPS+4-OI-treated macrophages exhibited markedly increased intracellular FTH1 expression compared to controls (Supplementary Fig. 7B and C). This increase was accompanied by a significant elevation in labile iron (Fe²⁺), as detected by the PGSK probe (Supplementary Fig. 7D and E), and enhanced lipid peroxidation, indicated by 4-hydroxynonenal (4-HNE) staining (Supplementary Fig. 7F and G). Collectively, these results demonstrate that under inflammatory conditions, 4-OI promotes the release of FTH1 via exosomes from macrophages. These exosomes are taken up by fibroblasts, leading to FTH1 accumulation, iron overload, and subsequent ferroptosis.

2.7. Macrophage-derived FTH1 transfers to fibroblasts via SCARA5 to alleviate fibrosis in BAS

Having demonstrated that 4-OI promotes the packaging of FTH1 into macrophage-derived exosomes and that these exosomes are taken up by fibroblasts, we next investigated the molecular mechanism underlying this uptake. Single-cell RNA sequencing analysis of tracheal tissues revealed that fibroblasts could be classified into five distinct transcriptional clusters (clusters 1–5) (Supplementary Fig. 8A), which exhibited clear separation in UMAP space(Fig. 7A). Pseudotime trajectory analysis identified cluster 2 as the terminal differentiation state of fibrogenic fibroblasts (Fig. 7B). Consistently, KEGG pathway analysis revealed that cluster 2 was significantly enriched in apoptosis and ferroptosis pathways (Supplementary Fig. 8B).

Fig. 7.

Macrophage-derived FTH1 binds fibroblast SCARA5 to inhibit fibrosis.

A. Re-clustering of fibroblasts from single-cell RNA-seq data identifies distinct subpopulations in human BAS granulation tissue. B. Pseudotime trajectory analysis reveals fibroblast differentiation converging on pro-fibrotic Cluster 2. C. CellChat interaction analysis predicts FTH1 (from ACOD1⁺ macrophages) binding to SCARA5 (on Cluster 2 fibroblasts) as the top ligand-receptor pair. D. Representative immunofluorescence images of FTH1(green), SCARA5(yellow) and Vimentin (red) in normal trachea group and granulation group. Scale bars indicate 50 μm and 20 μm. E. Co-immunoprecipitation (Co-IP) confirms FTH1 binding to SCARA5 in fibroblasts co-cultured with LPS+4-OI-treated macrophages. F,G. Representative tracheal immunofluorescence images of FTH1 (green), GPX4(yellow), SCARA5(red) and VIMENTIN(pink) in mouse trachea and quantitative of mean fluorescence of FTH1 and GPX4 in the control, BAS and BAS+4-OI group(n = 3 mice per group). Scale bars indicate 200 μm and 50 μm. H. Schematic diagram of the experimental design for 4-OI and Anti-SCARA5 treatment. I,J. Representative immunofluorescence images and mean fluorescence intensity of αSMA (red) and COL1(green) in four group: Control, BAS,BAS+4-OI,BAS+4-OI + Anti-SCARA5 (n = 3 mice per group). Scale bars indicate 200 μm and 50 μm. Data are presented as the mean ± SEM. One-way ANOVA analysis followed by Tukey post hoc multi-comparison test was used in G, J.

To explore potential intercellular communication between ACOD1⁺ macrophages and cluster 2 fibroblasts, we performed ligand-receptor interaction analysis. This revealed FTH1-SCARA5 as the top-scoring interaction pair, suggesting that macrophage-derived FTH1 may engage SCARA5 on fibroblasts (Fig. 7C). Immunofluorescence staining of granulation tissue from BAS patients confirmed high SCARA5 expression on fibroblasts, which colocalized with FTH1 (Fig. 7D). To examine the molecular basis of this interaction, we performed protein-protein docking using PyMol2.6 software, which predicted direct binding between FTH1 and SCARA5 (Supplementary Fig. 9). Co-immunoprecipitation (Co-IP) assays in co-culture systems further confirmed physical association between fibroblast SCARA5 and macrophage-derived FTH1 when macrophages were stimulated with LPS+4-OI (Fig. 7E). To determine whether SCARA5 is required for FTH1 uptake by fibroblasts, we knocked down SCARA5 using siRNA (Supplementary Fig. 10A and B). Exogenous FTH1 protein readily increased intracellular FTH1 levels in control fibroblasts, but this effect was completely abrogated in SCARA5-knockdown cells (Supplementary Fig. 10C and D), indicating that SCARA5 mediates FTH1 internalization.

In vivo, immunofluorescence staining of injured tracheal tissue demonstrated colocalization of the fibroblast marker vimentin with FTH1, SCARA5, and the ferroptosis regulator GPX4. Notably, 4-OI treatment significantly increased FTH1 expression while decreasing GPX4 signal in fibroblasts, suggesting that FTH1 uptake coincides with the induction of ferroptosis (Fig. 7F and G). To assess the clinical relevance of SCARA5, we performed multiplex immunofluorescence staining for SCARA5, α-SMA, and COL1 on human tracheal tissue sections. Compared to normal trachea, granulation tissue exhibited pronounced co-localization of α-SMA and COL1 in fibrotic areas, where SCARA5 expression was significantly upregulated and spatially associated with these fibrotic markers (Supplementary Fig. 11).Finally, to functionally validate the requirement of the FTH1-SCARA5 axis for the therapeutic effect of 4-OI in vivo, we administered a neutralizing antibody against SCARA5 (Fig. 7H). While 4-OI treatment significantly attenuated airway fibrosis in the injury-induced stenosis model, co-administration of the anti-SCARA5 antibody completely abolished this protective effect (Fig. 7I and J). These results demonstrate that the anti-fibrotic action of 4-OI is critically dependent on FTH1-SCARA5-mediated intercellular communication.

3. Discussion

This study reveals that the ACOD1-itaconate axis serves as a crucial regulator of phase-specific immune responses BAS, offering new insights into the disease's pathogenesis and potential therapeutic strategies. In the early stages of BAS, we identified ACOD1 as a key gene expressed predominantly in pro-inflammatory macrophages within the Mc4 cluster. While itaconate's roles in inflammatory diseases are established [17,18], our work extends its function to acute inflammation after tracheal injury. Due to the potential limitations itaconate faces in crossing the cell membrane to enter the cytoplasm, researchers have synthesized various itaconate derivatives, such as 4-OI. In vivo, 4-OI can efficiently hydrolyze into itaconate, exhibiting effects equivalent to those of endogenous itaconate [19].The administration of 4-OI effectively reduced submucosal edema and cytokine storms, demonstrated by a significant decrease in IL6 and IL1β levels, and activated the NRF2-FTH1 pathway, resulting in reduction of reactive oxygen species. These findings suggest that itaconate could act as a tissue-specific inflammation resolver.

The activation of NRF2 signaling by ACOD1/itaconate represents a well-characterized cytoprotective mechanism [20]. Under oxidative stress or inflammatory conditions, the Keap1-NRF2 complex dissociates, allowing NRF2 to accumulate and translocate to the nucleus, where it binds antioxidant response elements (AREs) and drives expression of antioxidant and anti-inflammatory genes [21,22]. Itaconate acts as an endogenous NRF2 inducer that couples metabolic regulation with anti-inflammatory gene expression [23]. FTH1, the catalytic core of the ferritin complex, is essential for intracellular iron storage and homeostasis [24]. By sequestering labile iron, FTH1 prevents the Fenton reaction-driven generation of ROS, acting as a key defense against oxidative damage [25]. Its expression is regulated by NRF2, which binds to AREs in the FTH1 promoter under stress conditions [26]. Inflammatory diseases with dysregulated FTH1 expression are associated with increased tissue injury, underlining its critical role as a cytoprotective effector. As we demonstrate, within the ACOD1-itaconate-NRF2 axis, FTH1 serves as the crucial cytoprotective effector in BAS, with our findings expanding its role beyond traditional iron storage. FTH1 knockdown in macrophages abolished the antioxidant effects of 4-OI, establishing its indispensable role in mediating itaconate bioactivity. Mechanistically, FTH1 sequesters labile iron to prevent ROS generation through the Fenton reaction, and reduces oxidative damage by decreasing intracellular iron availability, directly quenching lipid peroxidation and mitochondrial ROS. Therapeutically, 4-OI reverses LPS suppressed FTH1 expression, allowing for feedback amplification of antioxidant defenses.

The repair process following tracheal injury shares similar characteristics with the healing of other soft tissues and organs, encompassing distinct stages: hemostasis, inflammation, proliferation, and remodeling [27]. Controlling early inflammation levels in BAS effectively inhibits fibroblast activation and delays fibrosis progression. Dysregulated immune responses mediated by immune cells and inflammatory cytokines represent a core pathological feature during BAS advancement [28], and the inflammatory phase serves as a critical window for fibrotic development, driving the transformation of granulation tissue into scar tissue through synergistic interactions among inflammatory factors and signaling cascades [29]. Early control of inflammation can effectively alleviate the fibrosis process of the trachea [30]. In this study, administration of 4-OI significantly reduced granulomatous hyperplasia and fibrotic severity after tracheal injury. The ACOD1-itaconate axis exerted protective effects during the acute phase through anti-inflammatory and antioxidant activities, contributing to attenuated fibrogenesis. Conversely, ACOD1 deficiency exacerbated granulomatous hyperplasia, underscoring the importance of endogenous itaconate in limiting fibrotic progression. Beyond its intracellular role, we uncovered a new function for FTH1 as a mediator of fibrosis resolution through transcellular signaling during the chronic inflammatory stage of BAS. 4-OI promoted the packaging and release of macrophage FTH1 via exosomes. These exosomal FTH1 were subsequently taken up by fibroblasts through SCARA5, a membrane receptor expressed on fibroblasts that bound FTH1 and mediated iron delivery. This interaction led to iron overload and ferroptosis in fibrogenic fibroblasts, suggesting a new therapeutic avenue for targeting pathologic fibroblasts.

Therapeutically, our findings support a dual-phase approach for targeting the ACOD1-itaconate axis. In the acute phase, 4-OI helps prevent inflammatory escalation via NRF2 and FTH1, while in the chronic phase, the FTH1-SCARA5 interaction promotes the death of fibrogenic fibroblasts, significantly reducing fibrosis. Currently, itaconate is employed in disease treatment, such as promoting bone repair and reducing fibrosis following myocardial infarction [31,32]. Notably, 4-OI's efficacy in combating established fibrosis suggests its potential application in treating recurrent BAS.

There are limitations to this study that need to be addressed in future research. The mechanical brush injury model used in this study effectively recapitulates key features of post-traumatic or iatrogenic BAS, including the acute inflammatory response and subsequent fibroproliferative phase central to our investigation. However, we acknowledge that this model does not encompass all BAS etiologies. Specifically, it does not directly model infection-driven stenosis, which may involve distinct, persistent immune activation patterns, or ischemia-driven stenosis, where hypoxic signaling and HIF-1α likely play predominant roles. Despite these different initiating stimuli, the final common pathway of fibroblast activation and persistence may share convergent mechanisms. The FTH1-SCARA5-dependent ferroptosis pathway identified here represents a fundamental insight into macrophage-fibroblast crosstalk that may contribute to fibrotic progression across contexts. Future studies employing models of infectious or ischemic airway injury will be essential to test the broader applicability of this axis and to define the boundaries of its therapeutic relevance.

Additional limitations warrant consideration. The precise mechanisms linking SCARA5-mediated FTH1 uptake to downstream iron overload and ferroptosis remain to be fully elucidated. Although our single-cell RNA sequencing analysis incorporated human samples, clinical trials with 4-OI in BAS patients are needed to confirm therapeutic potential. Furthermore, the functional heterogeneity of ACOD1⁺ macrophage subsets, particularly the Mc4 cluster, requires deeper investigation through single-cell functional assays.

In conclusion, we delineate a unified pathway in which ACOD1-itaconate modulates macrophage metabolism, suppressing early inflammation through NRF2-FTH1 and resolving late fibrosis via exosomal FTH1 transfer to fibroblasts through SCARA5, leading to ferroptosis (Fig. 8). This work highlights immunometabolic modulation as a promising strategy for treating fibrotic airway diseases.

Fig. 8.

Schematic diagram of mechanism of the ACOD1/itaconate axis in regulating airway inflammation and fibrosis.

During the inflammatory stage, macrophages upregulate ACOD1 expression, catalyzing itaconate production. Itaconate and its derivative 4-OI activate the NRF2 signaling pathway by promoting NRF2 nuclear translocation. Nuclear NRF2 induces FTH1 transcription, leading to intracellular FTH1 protein accumulation to suppressing inflammation and attenuating oxidative stress. In fibrotic repair, FTH1 packaged into exosomes and secreted. These exosomes are taken up by fibroblasts via bind to SCARA5, leading to FTH1 delivery and subsequent induction of ferroptosis.

4. Methods

4.1. Patients and samples

This study was approved by the Ethical Committee of The First Affiliated Hospital of Naval Medical University (Approval No: CHEC2021-043) and conducted in accordance with the Declaration of Helsinki. Written informed consent was obtained from all participants prior to inclusion. Participant characteristics are detailed in Supplementary Table 1. Granulation tissues from benign airway stenosis patients were collected via bronchoscopic electrosurgical snaring, while normal tracheal tissues were obtained from lobectomy patients. All samples were procured at The First Affiliated Hospital of Naval Medical University (Shanghai, China).

4.2. Animals

ACOD1^−/− C57BL/6 J mice were kindly provided by the National Key Laboratory of Immunity and Inflammation (Naval Medical University) [33]. Female and male C57BL/6J mice (8 weeks old) were obtained from SPF (Beijing) Biotechnology Co., Ltd. All animals were housed in specific pathogen-free (SPF) facilities at the First Hospital of the Naval Medical University under a 12-h light/dark cycle, with ad libitum access to food and water. Ambient temperature was maintained at 20–26 °C, and relative humidity at 40–70%. All procedures complied with the NIH Guide for the Care and Use of Laboratory Animals and were approved by the Animal Care and Utilization Committee of Shanghai Changhai Hospital (Ethical Approval No. CHEC2021-006).

4.3. Modeling of BAS in mice

To establish a benign airway stenosis model [34], tracheal injury was induced in anesthetized mice (intraperitoneal ketamine/xylazine: 100/10 mg/kg) using a small-brush scraping technique. Following cervical skin sterilization (75% ethanol), a 1.5 cm midline incision exposed the trachea via blunt dissection. A transverse incision (two-thirds tracheal circumference) was made below the cricoid cartilage between two cartilaginous rings. A 1-mm diameter brush was inserted 6 mm into the tracheal lumen via a 20G safety IV catheter cannula (B|BRAUN Medical). The cannula was then withdrawn using an attached suture, leaving the brush in situ to perform three full-length scrapes of the tracheal wall. The trachea was closed with 6-0 sutures, and subcutaneous tissue/skin with 3-0 sutures. Post-procedurally, mice were euthanized by cervical dislocation under 3% isoflurane anesthesia.Sham-operated controls underwent identical surgical exposure and tracheal incision, but without brush insertion or scraping.

4.4. Drug administration

To supplement exogenous itaconate in mice, the mice were given 25 mg/kg 4-OI (#HY-112675, Med-Chem Express, USA) dissolved in 10% DMSO (# HY-Y0320, Med-Chem Express, USA) and 90% corn oil (#HY-Y1888, Med-Chem Express, USA) by intraperitoneal injection on 12 h before modeling, immediately following injury establishment (within 5 min),days 2, 4 and 6 after modeling.

For inhibit NRF2 in mice, the mice were given 30 mg/kg ML385(# HY-100523, Med-Chem Express, USA) dissolved in 10% DMSO + 40% PEG300 + 5% Tween-80 + 45% saline by intraperitoneal injection on 24 h before modeling and immediately following injury establishment (within 5 min).

To inhibit SCARA5 in vivo, mice were administered an anti-SCARA5 neutralizing antibody (#AF4754, Bio-Techne) via intratracheal injection on following injury establishment and day4 (3 μL per administration).

4.5. Cell culture and treatment

The mouse macrophage cell (Raw 264.7, #CL-0190) line, mouse tracheal fibroblasts (#CP-M008), human macrophage cell (THP-1, # CL-0233) and human tracheal Fibroblasts( # CP-H247) were purchased from Procell Biotech, Ltd. Cells were maintained in DMEM (#SH30243.01, Hyclone, USA) supplemented with 10% fetal bovine serum (#SH30406.05, Hyclone, USA) and 1% penicillin and streptomycin (#SV30010, Hyclone, USA) at 37 °C in a humidified atmosphere containing 5% CO₂ atmosphere.

To investigate the effect of 4-OI,cells were seeded in 12-well plates at a density of 10⁵ cells per well, and the following day, treated with 1 μg/ml LPS(#HY-D1056, Med-Chem Express, USA)and/or 125 μm 4-OI(#HY-112675, Med-Chem Express, USA)for 24 h, with these parameters selected based on prior experimental data [35].

To investigate the relationship between macrophage and fibroblasts, a Transwell coculture system was employed. Fibroblasts were plated in the lower chambers of 6-well plates, while macrophage were seeded in the upper inserts. Before coculture, macrophage were treated with 1 μg/ml LPS+125 μm 4-OI and/or transfected with siFTH1. The pretreated macrophage was subsequently cocultured with fibroblasts. For add exogenous FTH1, fibroblasts were treated with recombinant FTH1 protein(#CSB-EP009030MO, cusabio).

For exosome-mediated transfer, macrophages were pretreated with the exosome secretion inhibitor 10 μM GW4869 (#HY-19363, Med-Chem Express, USA) for 12 h before and during LPS+4-OI stimulation. After pretreatment, macrophages were co-cultured with fibroblasts, and the inhibitor was maintained during co-culture.

For ferroptosis inhibition, fibroblasts were pretreated with 5 μM deferoxamine (DFO) (HY–B1625, Med-Chem Express, USA) for 2 h prior to co-culture with LPS+4-OI-treated macrophages and DFO was maintained throughout the co-culture period.

4.6. Plasmids, RNA interference and transfection

siRNAs were purchased from HANBIO(ShangHai, China). A scrambled siRNA was used as an NC. The target sequences of the siRNAs and siNC are presented in Supplementary Table 2.

Plasmids were purchased from HANBIO (ShangHai, China). For ACOD1 overexpression, the full-length mouse Acod1 coding sequence (NM_008392.1) was PCR-amplified from a mouse cDNA library. The pHBLV-CMV-MCS-3FLAG-EF1-ZsGreen-T2A-PURO lentiviral vector was digested, gel-purified, and the Acod1 fragment was inserted using In-Fusion HD cloning (HB infusion). The ligation mix was transformed into DH5α competent cells, plated on ampicillin-containing LB agar, and incubated overnight (37 °C). Positive colonies were screened by colony PCR and confirmed by DNA sequencing. Validated plasmids were extracted and QC-checked for cell transfection. The overexpression efficiency of these constructs in target cells was validated by Western blotting.

An siRNA-resistant NRF2 overexpression plasmid was constructed using site-directed mutagenesis. Silent mutations were introduced into the siRNA-targeting region of the mouse NRF2 coding sequence (NM_010902.5) by PCR with specifically designed primers (Supplementary Table 3). The mutated fragment was cloned into the pcDNA3.1-EF1a-mcs-3flag-CMV-EGFP vector using the HB Infusion one-step cloning system. The construct was confirmed by DNA sequencing, ensuring that the amino acid sequence remained unchanged while conferring resistance to NRF2-targeting siRNA.

For transfection, cells were seeded plate and cultured overnight before transfection with Lipo3000(#L3000001, Invitrogen) according to the manufacturer's instructions.

4.7. qRT-PCR

Total RNA was extracted from tracheal tissues and cells using the RNeasy extraction kit (#RC112, Vazyme). cDNA was synthesized from the extracted RNA using the One-step iScript cDNA Synthesis Kit (#R323, Vazyme). Quantitative real-time PCR (qRT-PCR) was then performed using SYBR Green Master Mix (#Q711, Vazyme) on a QuantStudio™ 3 System (Thermo Fisher Scientific). GAPDH served as the reference gene. Relative gene expression was calculated using the 2-ΔΔCt method. Primer sequences are detailed in Supplementary Table 4.

4.8. Histology

After sampling, tracheal samples were fixed in 4% buffered formaldehyde for 24 h, paraffin-embedded, and sectioned axially at 3-μm thickness. Consecutive sections were stained with hematoxylin-eosin (H&E) and Masson's trichrome. Inflammatory infiltration area,tracheal stenosis degree Collagen deposition area was quantified from Masson-stained sections using ImageJ software. When measuring airway stenosis degree, we utilize the following formula: (1 - s/S) * 100%, where s represents the cross-sectional area of the lumen within the trachea, and S denotes the curved area formed by the tracheal cartilage [36].

4.9. Immunohistochemistry

For immunohistochemistry, the antigen was incubated in antigen repair solution at 95 °C for 30 min, and the slides were immunostained overnight at 4 °C with primary antibody. The combined antibody was detected by biotinylated goat anti-rabbit secondary antibody and incubated at room temperature for 1 h. The slices were then developed with DAB solution and counterstained with hematoxylin. The concentration of antibody used in Immunohistochemistry is shown in Supplementary Table 5.

4.10. Immunofluorescence staining

Human and mouse sections (5 μm) were deparaffinized, hydrated, and PBS-washed (3 × 5 min). Cultured cells on coverslips were fixed with 4% PFA (15 min) and PBS-washed (3 × 3 min). Both preparations were blocked with 10% goat serum (37 °C, 1 h), then immunolabeled with primary antibodies (4 °C overnight) and corresponding secondary antibodies (37 °C, 1 h). After final PBS washes, samples were mounted with DAPI. The concentration of antibody used in immunofluorescence staining is shown in Supplementary Table 6.

4.11. Western blotting

Proteins were extracted from mouse tracheal tissues and cells using a lysis buffer containing RIPA buffer supplemented with protease and phosphatase inhibitors. Protein concentration was determined by the BCA method. Nuclear protein extraction was performed using a Nuclear and Cytoplasmic Protein Extraction Kit (#EX1470, Solarbio), following the manufacturer's instructions for use.

Samples were mixed with loading buffer, boiled at 100 °C for 10 min, and separated by SDS-polyacrylamide gel electrophoresis (SDS-PAGE). Equal amounts of protein and a protein ladder (Epizyme #WJ103) were loaded onto the gel and electrophoresed at 80-120 V for 60 min. Proteins were then transferred onto a 0.45 μm PVDF membrane using a methanol-based transfer buffer. After blocking, the membrane was incubated overnight at 4 °C with primary antibodies, followed by incubation with HRP-conjugated goat anti-rabbit IgG secondary antibody (1:8000, Proteintech #SA00001-2). Protein bands were visualized using the Tanon 5200 chemiluminescent imaging system (Shanghai) and quantified with ImageJ software. The concentration of antibody used in Western blotting is shown in Supplementary Table 7.

4.12. Co-immunoprecipitation

Cell pellets were lysed and protein concentration was determined using the BCA assay, and SCARA5/FTH1 expression levels were detected. Subsequently, 200 μl of Protein A/G agarose beads were added to each group. The IP group received FTH1 antibody (#4393, Cell Signaling Technology, 1:50), while the IgG control group received rabbit IgG antibody. Samples were incubated at 4 °C for 2 h. Cell lysate was then added to each group and rotated overnight at 4 °C. Beads were washed three times with PBS, followed by elution using SDS loading buffer. Eluted proteins were subsequently analyzed by standard Western blotting.

4.13. Micro CT and measurement method of tracheal stenosis in imaging

Mouse airways were visualized using microCT (90 kV, 0.065 mA, PINGSENG Healthcare). Cross-sectional areas were quantified within the adipose tissue window. Tracheal stenosis was assessed at the narrowest point according to the formula: Stenosis (%) = (1 − s/S) × 100%, where s denotes the lumen area at the most constricted site, and S represents the mean lumen area measured 3 mm proximal and distal to the stenosis. All tracheal measurements and three-dimensional reconstructions were performed using SYNAPSE 3D software (v4.4, Fujifilm Medical, Tokyo, Japan).

4.14. Liquid chromatography mass spectrometry analysis

Samples were separated using an Agilent 1290 Infinity UHPLC system. Standards were stored in a 4 °C autosampler. The column temperature was maintained at 50 °C with a mobile phase flow rate of 400 μL/min and an injection volume of 5 μL. The mobile phase composition was as follows: Phase A was water with 0.1% formic acid; Phase B was acetonitrile with 0.1% formic acid. The gradient program was as follows: 0–2 min, 15% phase B; 2–9 min, linear increase to 98% B; 9–11 min, hold at 98% B; 11–11.5 min, linear decrease to 15% B; 11.5–14 min, equilibration at 15% B. MS analysis was performed in positive ion mode on a 5500 QTRAP mass spectrometer (AB SCIEX). Target ion pairs were detected via MRM (Multiple Reaction Monitoring), and quantification was conducted using MultQuant or Analyst software. This experiment was performed by the Sinopharm Health Industry Research Institute (Shanghai), China State Institute of Pharmaceutical Industry.

4.15. Fluorescence in situ hybridization

The fluorescence in situ hybridization (FISH) protocol was as follows: Specific probes were designed based on target sequences from the NCBI gene database (Supplementary Table 8). Tissues were immediately fixed in DEPC-treated water-based fixative, dehydrated through graded alcohols, embedded in paraffin, and sectioned, followed by baking at 62 °C. After deparaffinization with xylene, rehydration through graded alcohols, and rinsing with DEPC water, sections were boiled in retrieval solution for antigen retrieval. Following natural cooling, sections were digested with proteinase K (20 μg/ml), rinsed with distilled water, and washed with PBS. Prehybridization was performed at 37 °C, after which the prehybridization solution was removed and replaced with probe-containing hybridization buffer for overnight hybridization at 37 °C. Post-hybridization washes included sequential incubation in 2 × SSC (37 °C), 1 × SSC (37 °C), and 0.5 × SSC (room temperature), with formamide washes added as needed. Nuclei were counterstained with DAPI in the dark and mounted with anti-fade medium.

4.16. Enzyme-linked immunosorbent assay (ELISA)

To detect the content of FTH1 in cell supernatant, a commercially available enzyme-linked immunosorbent assay kit (#CSB-EL009030MO, cusabio) was used according to the manufacturer's instructions.

4.17. Measurement of ROS

Cells were seeded in culture plates and incubated overnight. After treatment, total ROS levels were measured using the Fluorometric Intracellular ROS Kit (#MAK142, Sigma, USA).

4.18. Iron assay

The quantification of chelatable intracellular iron was conducted using the fluorescent indicator Phen Green SK (#P14312, Thermo Fisher Scientific). Cells were first seeded and treated accordingly, after which the culture medium was discarded. The slides were rinsed three times with PBS. Subsequently, a 20 μM solution of the Phen Green SK probe was added, and the cells were incubated at 37 °C for 20 min. This was followed by washing the slides three more times with PBS. Finally, images were captured using a fluorescence microscope.

4.19. Isolation and identification of exosomes

Exosomes were purified from culture media using sequential centrifugation. Conditioned media were centrifuged at 300×g for 10 min and 2000×g for 20 min at 4 °C to remove cells and debris, respectively. Exosomes were pelleted by ultracentrifugation at 100,000×g for 120 min at 4 °C, washed once with PBS, and re-pelleted. The final exosome pellet was resuspended in PBS or appropriate buffer. The exosomal markers were confirmed by Western blot for TSG101, CD63, CD81 and calnexin. The size distribution and morphology of exosomes were characterized by nanoparticle tracking analysis (NTA) and transmission electron microscopy (TEM), respectively. The NTA and TEM were performed at WEKNOW Bio-TECH Co., Ltd. (Shanghai, China).

4.20. DiD labeling of exosomes

Exosomes were labeled with 5 μM DiD (#C1039, Beyotime)(30 min, 37 °C) and ultracentrifuged to remove free dye. Labeled exosomes were incubated with fibroblasts for 6 h. Cells were fixed, permeabilized, and stained with phalloidin (F-actin, green) and DAPI (nuclei, blue). Exosome uptake was visualized as punctate red fluorescence within the cytoskeleton-framed cytoplasm.

4.21. C11-BODIPY staining for lipid peroxidation detection

Lipid peroxidation in fibroblasts was assessed using the C11-BODIPY probe(#D3861, Invitrogen). After co-culture, fibroblasts were incubated with 2 μM C11-BODIPY in serum-free medium for 30 min at 37 °C in the dark. Cells were then washed with PBS and immediately visualized under a fluorescence microscope. The oxidation of the probe was detected by the shift from red to green fluorescence.

4.22. Transmission electron microscopy (TEM) for ferroptosis detection

Fibroblasts were fixed with electron microscope fixative for 2–4 h, embedded in 1% agarose, dehydrated, and sectioned (60–80 nm) using a Leica UC7 microtome. Sections were stained with uranyl acetate and lead citrate, then examined under a Hitachi transmission electron microscope. Ferroptosis was characterized by shrunken mitochondria with increased membrane density and reduced or vanished cristae, with intact nuclear morphology. These analyses were performed at WEKNOW Bio-TECH Co., Ltd. (Shanghai, China).

4.23. Bulk-RNA sequencing analysis

Raw and processed data from the common database (PRJNA1169631). Differential gene expression analysis was performed between the two groups using DESeq2 software. Additionally, edgeR was used for comparison between specific sample pairs. Genes with a false discovery rate (FDR) < 0.05 and an absolute fold change ≥1 were defined as differentially expressed genes (DEGs). Volcano plot and Heat maps were generated using the R package (v4.1.3). To elucidate the functional implications of the DEGs and highlight differences at the gene functional level, Gene Ontology (GO) enrichment analysis and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis were conducted using the clusterProfiler R package (v4.1.3). GO terms and KEGG pathways with a P-value <0.05 were considered significantly enriched. Gene Set Enrichment Analysis (GSEA) was using software GSEA (v4.1.0) with the MSigDB (v7.5.1). All RNA sequencing data analysis was performed using the OmicStudio tools created by LC-BIO Co., Ltd (HangZhou, China) at https://www.omicstudio.cn/cell.

4.24. Single-cell data analysis

Raw and processed data from the common database (PRJNA1182252). scRNA-seq data were processed and analyzed using Seurat (version 3.1.1). A total of 58,586 cells passed quality control filters based on the following criteria:Genes detected in fewer than 3 cells (default parameter: 1 cell) were excluded; Cells expressing fewer than 100 genes were removed (lower bound); Cells with UMI counts below 500 were excluded; Cells where the percentage of reads mapping to mitochondrial genes exceeded 25% were removed.Dimensionality Reduction, Clustering, and Visualization From the filtered dataset, 18,339 cells were analyzed further. The workflow comprised the following steps: Gene expression levels were normalized using the “LogNormalize” method within Seurat, Principal Component Analysis (PCA) was performed on the normalized data. The top 10 principal components (PCs) were selected for subsequent clustering and visualization; Cell clusters were identified using the shared nearest neighbor (SNN) modularity optimization-based clustering algorithm within Seurat. For each cluster, marker genes were identified using the FindAllMarkers function in Seurat, employing the Wilcoxon rank-sum test (default parameters). Genes were considered markers if they were expressed in >10% of cells within the cluster and exhibited an average log (fold change) > 0.25 compared to all other clusters. Cells were visualized in two dimensions using Uniform Manifold Approximation and Projection (UMAP) based on the top 10 PCs.

For macrophage and fibroblast reclustering, macrophage and fibroblast clusters were subset and unsupervised clustering was used to identify macrophage and fibroblast subpopulations. macrophage and fibroblast subsets were analyzed with FindMarkers, to characterize cluster identities and DGE between treatment conditions.

To investigate the potential functions, cluster Profiler (v4.1.32) R software package was used for Gene Ontology (GO) analysis and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis. Pathways with p value less than 0.05 were considered as significantly enriched. Pseudotime analysis of fibroblasts was performed using monocle3 (v1.3.7). CellChat cell-cell interaction network analysis was used to characterize ACOD1+ macrophage-fibroblast cluster2 interactions by Cellphonedb (v5.0.0).

All SingleCell Data Analysis was performed using the OmicStudio tools created by LC-BIO Co., Ltd (HangZhou, China) at https://www.omicstudio.cn/cell.

4.25. Statistical analysis

All dates are presented as mean ± SEM. GraphPad Prism 9 was used for statistical analysis (GraphPad software, San Diego, California, USA). Differences between two groups were identified using the two-sided Student's test. Multiple groups using One- ANOVA analysis followed by Kruskal-Wallis's test. P values of less than 0.05 were considered statistically significant.

CRediT authorship contribution statement

YiLin Chen: Investigation, Visualization, Writing – original draft, Writing – review & editing. ChengFei Xu: Investigation, Writing – original draft. Tao Luo: Investigation. DongChen Shi: Investigation. XinYi Guo: Validation. ChengCheng Yang: Validation. YuChao Dong: Resources. HaiDong Huang: Resources. YiFei Zhang: Resources. Zhe Zong: Validation. XiaoMin Wang: Validation. ZhiRu Xu: Resources. Yue Shi: Resources. ChaoFeng Han: Supervision, Writing – review & editing. Hui Shi: Funding acquisition. Chong Bai: Conceptualization, Funding acquisition, Validation.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following are the Supplementary data to this article: