Developmentally endothelial locus-1 facilitates intestinal inflammation resolution by suppressing the Cmpk2-cGAS-STING pathway and promoting reparatory macrophage transition

Graphical abstract

DEL-1 attenuates intestinal inflammation and promotes inflammation resolution through several mechanisms: DEL-1 inhibits the Cmpk2-dependent mtDNA synthesis and consequently cGAS-STING signaling activation. DEL-1 also promotes reparative macrophage transition. Furthermore, the regulatory effect of DEL-1 is mediated through promoting the ubiquitin–proteasome-dependent degradation of transcription factor Spi1, and affecting its transcriptional regulatory capacity of Cmpk2 and Il10. In parallel, DEL-1 inhibits neutrophil recruitment, repairs the intestinal barrier, and improves intestinal microbiota dysbiosis to promote intestinal inflammation resolution.

Highlights

• •DEL-1 was downregulated in IBD patients, and resurged as inflammation resolution.
• •DEL-1 plays a dual role of anti-inflammatory and pro-resolution in IBD.
• •DEL-1 suppresses Cmpk2-cGAS-STING pathway to alleviate intestinal inflammation.
• •DEL-1 promotes reparative macrophage transition in the repair model of colitis.
• •DEL-1 regulates transcription of Cmpk2 and Il10 by promoting Spi1 degradation.

Abstract

Introduction

Abnormalities in inflammation resolution function are intimately linked to chronic inflammation, and proresolution therapies may offer novel opportunities for IBD treatment. Developmental endothelial locus 1 (DEL-1), a natural modulator of tissue immunity and inflammation resolution, has not been studied in IBD.

Objectives

We aimed to investigate the expression and functions of DEL-1 in IBD.

Methods

Assessment of DEL-1 expression in patients, murine models, and cellular levels. To explore the effects of DEL-1 in the acute and recovery phases of inflammation, overexpression plasmids, adeno-associated viruses for DEL-1 knockdown, and DEL-1-Fc fusion proteins were administered to cells and mice. Additionally, the potential mechanism of DEL-1 in IBD was demonstrated using flow cytometry, RNA-Seq, ChIP, dual-luciferase reporter assays and 16S rRNA.

Results

DEL-1 levels were significantly reduced in IBD patients, colitis mice and macrophages, while the levels increased with inflammation to resolve. Transfection with DEL-1 overexpression plasmid or DEL-1-Fc intervention reduces levels of inflammatory cytokines in both phases and upregulates reparative gene levels in the recovery phase. DEL-1 knockdown inhibits inflammation resolution of colitis. Mechanistically, we demonstrated that DEL-1 inhibits Cmpk2-dependent mtDNA synthesis, thereby inhibiting the cGAS-STING pathway to ameliorate intestinal inflammation. Moreover, DEL-1 promotes reparative macrophage transition in the repair model of colitis. Spi1 was identified as a transcription factor that regulates Cmpk2 and the reparative gene Il10. Intervention with overexpression plasmid of Spi1 or Cmpk2 or the STING agonist DMXAA reverses the effects of DEL-1. In parallel, DEL-1 also inhibits neutrophil recruitment, repairs the intestinal barrier, and improves intestinal microbiota dysbiosis.

Conclusion

We report the first demonstration that DEL-1 significantly ameliorates colonic inflammation in colitis mice. Our findings elucidate a novel mechanism wherein DEL-1 exerts its protective effects by suppressing the Cmpk2-cGAS-STING pathway and promoting reparative macrophage transition. These results collectively position DEL-1 as a promising therapeutic avenue for IBD.

Introduction

Inflammatory bowel disease (IBD) is a recurring chronic gastrointestinal disease, consisting of ulcerative colitis (UC) and crohn’s disease (CD). It is characterized by an excessive mucosal immune response triggered by genetic susceptibility, environmental stimuli, intestinal microbiota dysbiosis, and impaired epithelial barrier [1]. The worldwide incidence and prevalence of IBD are exhibiting a concerning upward trajectory, imposing a substantial burden on public health. The significant impact of IBD and its associated complications on patient quality of life underscores the critical importance of inducing and maintaining remission. Current treatment strategies for IBD are primarily anti-inflammatory, such as aminosalicylic acid, steroids, immunosuppressive agents, and biological agents targeting cytokines and inflammatory pathways. However, these treatments may have reached the therapeutic ceiling, inducing mucosal healing in less than 50 % of IBD patients with moderate to severe, as well as a high risk of opportunistic infections and high costs [2]. Therefore, it is important to explore the process of inflammation and mucosal homeostasis restoration in IBD and find new therapeutic targets.

Inflammation resolution is a precisely orchestrated process involving a cascade of cells and molecules that actively terminate the inflammatory response and initiate tissue repair, ultimately restoring homeostasis [3]. Abnormalities in inflammation resolution led to an uncontrolled inflammatory response, which are hallmarks of multiple chronic inflammatory diseases such as multiple sclerosis, atherosclerosis, SLE, and IBD [[4], [5], [6], [7]]. Intestinal macrophages are highly plastic cells within the mucosal immune system that not only advance inflammation, but also maintain gut homeostasis and induce resolution of inflammation [8]. The polarization of macrophages towards a proresolving phenotype is instrumental in mitigating excessive inflammation, encompassing the suppression of TH1 and TH17 immune responses and the promotion of epithelial barrier repair [9]. Macrophage phagocytosis and clearance of apoptotic cells (efferocytosis) facilitate macrophage repolarization towards a reparative phenotype, leading to the production of specialized pro-resolving mediators (SPMs) and anti-inflammatory cytokines like Il10, which contribute to tissue homeostasis restoration [10]. Chimeric efferocytic receptors promoting efferocytosis significantly relieve colitis, and SPMs induce the proliferation of intestinal epithelial cell, alleviate intestinal inflammation and facilitate mucosal healing [11,12].

Developmentally endothelial locus 1 (DEL-1), also referred to as EGF-like repeats and discoidin I-like domains 3 (EDIL3), is a multi-domain protein secreted by tissue-resident cells, including macrophages, stromal and endothelial cells [13]. It plays a pivotal role in regulating immune plasticity and inflammation resolution. Endothelially derived DEL-1 inhibits leukocyte-endothelial adhesion by interacting with LFA-1 (αLβ2 integrin), thereby inhibiting inflammatory cell recruitment [14]. Macrophage-derived DEL-1 acts as a bridge molecule, C-terminus recognizing phosphatidylserine on apoptotic cells and N-terminus binding to β3 integrins on macrophages, promoting macrophage efferocytosis and accelerating resolution of inflammation [15]. Accumulating evidence shows that DEL-1 protects against inflammatory disorders, such as periodontitis [15], inflammatory arthritis [16], asthma [17], and experimental autoimmune encephalomyelitis [18]. In both human periodontal inflammation and mouse ligature-induced periodontitis model, DEL-1 expression decreases during active inflammation and resurges significantly during the recovery phase [15]. However, there is a lack of studies on the effect and mechanism of DEL-1 in IBD, with only one study reporting elevated levels of DEL-1 in the rectal mucosa of UC-Ca patients than in UC-NonCa patients [19].

Given the anti-inflammatory and pro-resolving properties of DEL-1, we explored whether DEL-1 has a protective effect on IBD and its specific mechanism. By using the acute and repair model, and DEL-1 overexpression, DEL-1 knockdown and pharmacologic approaches, we found that DEL-1 attenuates colonic inflammation, accelerates resolution of inflammation and tissues repair. The underlying regulatory mechanism is that DEL-1 inhibits activation of the Cmpk2-cGAS-STING pathway and induces reparative macrophage transformation.

Materials and methods

Patients.

In this study, we collected a total of 98 colon biopsy samples of patients (39 UC, 41 CD, and 18 controls) and 161 blood samples of patients (49 UC, 56 CD, and 56 controls). IBD patients were diagnosed based on clinical, radiological, endoscopic and histological findings, and the control group comprised individuals with hemorrhoids or polyps, with the normal endoscopic examination. The assessment of clinical disease activity in UC and CD patients was based on the Modified Mayo score and SES-CD score, respectively [20,21]. The clinical characteristics are shown in Table S1.

Proteins.

DEL-1 was used as a fusion protein consisting of the full-length DEL-1 protein sequence and the Fc region of human IgG1. DEL-1-Fc was expressed and purified by GenScript. Fc protein control was purchased from Sino Biological.

DSS-induced acute and repair colitis mouse model.

The adeno-associated virus (AAV) specific for DEL-1 knockdown was obtained from Cyagen Biosciences. Mice were injected intraperitoneally with AAV-DEL-1 (5*10^11vg/mouse) or the control viruses. Four weeks following the virus injection, the mice were subjected to the repair model. Specifically, mice were initially administered 2.5 % dextran sulfate sodium (DSS, MP Biomedicals) for 7 days, and then DSS was withdrawn and replaced with sterile water for 6 subsequent days (Fig. 6A).

Fig. 6.

DEL-1knockdown inhibits inflammation resolution of DSS-induced colitis in the recovery phase. (A) Diagram of the modeling and treatment strategy for the induced repair model. Briefly, mice were treated with AAV-DEL-1 for 4 weeks, followed by DSS feeding for 7 days. Subsequently, DSS was withdrawn and replaced with sterile water for 6 days (n = 5). (B) Body weight loss was calculated as the percent change relative to day 0. (C) Disease activity index (DAI) scores. (D) Representative images of colons. (E) Colonic length. (F) Representative images of hematoxylin and eosin (H&E) staining. (G) Histological score. (H) RT-qPCR of cytokines (Il1β, Il6, Tnfα, Il10, Arg1, Ifnα, and Ifnβ), normalized to β-actin. (I, J) Western blot of SPI1, CMPK2 and cGAS-STING pathway related protein expression. The intensity ratio of the target protein to corresponding controls quantified using densitometric analysis, including SPI1/GAPDH, CMPK2/GAPDH, CGAS/GAPDH, p-STING/STING, p-TBK1/TBK1, and p-IRF3/IRF3. Statistical analysis was calculated by student’s t tests. ns (not significant), p > 0.05; *p < 0.05; **p < 0.01; ***p < 0.001.

To further elucidate the timing of DEL-1 action, we constructed DEL-1-Fc fusion proteins and designed an intervention study. With reference to previous studies, the dosage of DEL-1-Fc and the induction methods for modeling colitis in different periods were determined [8,[22], [23], [24]]. Fig. 7F, Fig. S6A and Fig. S7A illustrate the intervention strategies for DSS, DEL-1-Fc/Fc, and DMXAA. Briefly, mice were treated with 2.5 % DSS for 7 days to induce colitis. DSS was then withdrawn and replaced with sterile water, and simultaneously, the mice were intraperitoneally injected with DEL-1-Fc (1 ug/mouse), or Fc (1 ug/mouse), or the STING pathway agonist DMXAA (70 ug/mouse, Topscience) for 6 consecutive days during the recovery phase. For the acute model, mice were treated with 2.5 % DSS in their drinking water for 7 days. Simultaneously, mice were given intraperitoneally with DEL-1-Fc (1 ug/mouse) or Fc (1 ug/mouse) daily. After 7 days, mice were euthanized, and colons were obtained for further analysis. The disease activity index (DAI) was calculated by body weight loss scores, fecal consistency scores, and stool bleeding scores [25].

Fig. 7.

DEL-1 promotes intestinal inflammation resolution dependent on cGAS-STING pathway in the repair model. Mice were treated with 2.5 % DSS for 7 days, and then DSS was withdrawn and replaced with sterile water and intraperitoneally injected with DEL-1-Fc (1 ug/mouse) for 6 days (n = 4–6) (A-E). (A, B) Flow cytometry analysis of CD11b + Ly-6G + neutrophils and their percentage in colonic tissues. (C) Assessment of F4/80 (green) and ARG1 (red) expression and colocalization in the colon tissues of the repair model by immunofluorescence. (D, E) Western blot of iNOS, CD206, and ARG1 in colonic tissues, and densitometric analysis quantified the intensity ratio of the target protein to GAPDH. (F-M) Mice were treated with 2.5 % DSS for 7 days, and then DSS was withdrawn and replaced with sterile water and intraperitoneally injected with DEL-1-Fc (1 ug/mouse) and DMXAA (70 ug/mouse) for 6 days (n = 10). (F) Diagram of the modeling and treatment strategy for the induced repair model. (G) Survival rates. (H) Representative images of colons. (I) Colonic length. (J) Representative images of H&E staining. (K) Histological score. (L) RT-qPCR of cytokines (Il1β, Il6, Tnfα, Il10, Arg1, Ifnα, and Ifnβ), normalized to β-actin. (N, M) Western blot of iNOS, CD206, and ARG1 in colonic tissues, and densitometric analysis quantified the intensity ratio of the target protein to GAPDH. Statistical analysis was calculated by one-way-analysis of variance (ANOVA). ns (not significant), p > 0.05; *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

Cell cultures and treatment.

The mouse macrophage cell line (RAW264.7) and human embryonic kidney (HEK293T) cells were cultured in DMEM (Gibco) containing 10 % FBS (Procell) and 1 % penicillin–streptomycin (Gibco) in 5 % CO2 and 37◦C. Mouse bone marrow-derived macrophages (BMDMs) were isolated and cultured as follows: femoral and tibial bone marrow from male C57BL/6 mice were collected, washed with PBS, and erythrocytes were lysed by ACK (Gibco). BMDMs were subsequently induced with DMEM medium (supplemented with 10 % FBS, 1 % penicillin–streptomycin and 20 ng/ml M−CSF (Genscript)) for 7 days before subsequent experiments.

Macrophages were pulsed with 1 μg/ml Lipopolysaccharide (LPS, biological source Escherichia coli (O55:B5), Sigma) for 4 h to induce an acute inflammation model, while cells were treated with 1 μg/ml DEL-1-Fc or Fc, with the dosage based on previously reported studies [22,24]. For the recovery phase, macrophages were pulsed with LPS for 4 h, and then replaced with fresh media and intervened with DEL-1-Fc or Fc for 24 h. Fig. 2A shows the LPS interventions in both models, with reference to previously reported [8]. In both models, DMXAA (1 μg/ml) was intervened at the time of LPS stimulation and after LPS withdrawal, respectively. In addition, cycloheximide (CHX), MG132, and chloroquine were dosed at 60 μg/ml, 20 uM, and 50 uM, respectively.

Fig. 2.

DEL-1 attenuates macrophages inflammation and promotes inflammation resolution. (A) The intervention methods of LPS induced the acute and repair model. Macrophages were pulsed with LPS (1 μg/ml) for 4 h (the acute inflammation model), and then replaced with fresh media after washing away the LPS for 24 h (the repair model) (n = 3–4). (B, E) Del-1 mRNA expression was measured by RT-qPCR after pulsed with LPS and hours post LPS withdrawal in RAW264.7 macrophages (B) and BMDMs (E). Results were normalized by β-actin gene. (C, D, F, G) Expression of DEL-1 protein levels were detected using western blot after pulsed with LPS and 24 h post LPS withdrawal in RAW264.7 cells (C, D) and BMDMs (F, G). Densitometry analysis quantified the intensity ratio of DEL-1 to GAPDH. (H) RAW264.7 macrophages were transfected with DEL-1 overexpression plasmids and corresponding controls for 24–36 h, and then subjected to LPS stimulation detailed in (A). The mRNA expression of Il1β, Il6, Tnfα, Il10, and Arg1 were measured by RT-qPCR in the acute and repair model, normalized to β-actin. (I) BMDMs were pulsed with LPS (1 μg/ml) and simultaneously treated with DEL-1-Fc (1 μg/ml) for 4 h (the acute inflammation model), or withdrawn LPS stimulation and replaced with fresh media and treated with DEL-1-Fc (1 μg/ml) for 24 h (the repair model). The mRNA expression of Il1β, Il6, Tnfα, Il10, and Arg1 were measured by RT-qPCR in the acute and repair model. Results were normalized by β-actin gene. (J-M) RNA-seq analysis of the NC + LPS and OE-DEL-1 + LPS groups of RAW264.7 macrophages 24 h post LPS withdrawal. (J) Heatmaps of DEGs with adjusted p value < 0.05 and |fold change| > 1.2. (K) Gene set enrichment analysis (GSEA). (L) Gene ontology (GO) enrichment analysis. (N) Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis. (M) Volcano plot of DEGs. Statistical analysis was calculated by student’s t tests or one-way-analysis of variance (ANOVA). ns (not significant), p > 0.05; *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

Transfection of plasmids DNA-mediated gene overexpression

DEL-1, Spi1, Cmpk2 overexpression plasmids and corresponding controls were purchased from Genechem. Macrophages were transfected with plasmid DNA using Lipofect5000 transfection reagent (BIOG). After 24–36 h of transfection, cells were harvested for further experiment. Next, cells were pulsed with LPS for 4 h, or the LPS stimulation was withdrawn after 4 h to continue incubation for 24 h.

RNA-seq analysis.

The total RNA of macrophages with LPS-stimulated and DEL-1 overexpression in the resolution phase was extracted using Trizol. RNA integrity was assessed using the Agilent 2100 Bioanalyzer, and the Illumina NovaSeq 6000 Sequencing platform was used for sequencing. Transcriptome sequencing and analysis were performed by Smart Genomics. Differentially expressed genes (DEGs) were performed using DESeq2, and significant DEGs were defined as genes with adjusted p value < 0.05 and |fold change| > 1.2. The heatmap, Volcano plot, GO, KEGG, and GSEA were mapped and analyzed by the OECloud tools (https://cloud.oebiotech.com).

Histological analysis.

The distal colon tissues were fixed in paraformaldehyde, embedded in paraffin, and sectioned serially to a thickness of 4 µm. Histological Analysis was conducted on HE stained paraffin sections of mice. Histological scores were assigned based on the extent of tissue damage and inflammatory cell infiltration within the lamina propria, as described previously [26].

Immunohistochemistry staining.

Paraffin sections were dewaxed, antigen heat-retrieved, endogenous peroxidase quenched, blocked with donkey serum, incubated with DEL-1 mouse monoclonal antibody (1:100, Santa Cruz Biotechnology) at 4◦C overnight, incubated with secondary antibody and then stained with DAB. Images were observed and recorded using microscope with bright field (Olympus).

Immunofluorescence staining.

For the immunofluorescence staining in colon tissue, sections were incubated overnight at 4 °C with primary antibody to DEL-1, F4/80 (1:100, Abcam), ARG1 (1:200, Proteintech), Ki-67 (1:300, Huabio), ZO-1 (1:200, Abclonal), or E-cadherin (1:200, Cell Signaling Technology). Subsequently, slides were incubated with indicated fluorescent secondary antibodies (1:150, Alexa Fluor 488 or Alexa Fluor 594, AntGene Biotechnology), and then counterstained with DAPI (AntGene Biotechnology). Images were visualized and captured by a microscope (Olympus) or confocal laser-scanning microscope (Nikon), and fluorescence intensity was quantified by ImageJ software.

For the immunofluorescence staining in macrophages, cells were incubated with 200 nM MitoTracker Red CMXRos (Yeasen Biotechnology) for 30 min. Following a PBS wash, cells were fixed, permeabilized, blocked, and incubated with anti-dsDNA antibodies (1:100, Santa Cruz Biotechnology) overnight at 4 °C. Subsequently, slides were stained with fluorescent secondary antibody (1:150, Alexa Fluor 647, AntGene Biotechnology) and counterstained with DAPI.

RT-qPCR assay.

Total RNA was extracted using Trizol (Vazyme) and cDNA was synthesized with HiScript III RT SuperMix (Vazyme). RT-qPCR was conducted using SYBR Green Master Mix (Vazyme) and the LightCycler® 480 System (Roche). The internal reference was human GAPDH or mouse β-action, and the relative gene expression levels were calculated using the 2-ΔΔCt method. The primer sequences are listed in Table S2.

ELISA.

The levels of serum DEL-1 were detected using an anti-human DEL-1 ELISA kit (R&D systems) according to the manufacturer’s method. The concentrations were calculated by a standard curve.

Western blot.

Colon tissues and cells were lysed with RIPA buffer (Applygen) containing protease and phosphatase inhibitors (Roche). Total protein concentrations were quantified using a BCA kit (Vazyme). Nuclear and cytoplasm proteins were extracted with a nuclear protein extraction kit (Solarbio). Equivalent protein was separated by SDS–PAGE (Vazyme), transferred to PVDF membranes (Millipore), blocked with rapid closure solution (EpiZyme Biotechnology), and then incubated with primary antibodies overnight at 4 °C. Subsequently, menbranes were incubated with appropriate secondary antibodies (Cell Signaling Technology) and then visualized using the electrochemical luminescence (ECL) kit (Biology) and chemiluminescence detection system (Bio-Rad). Images were quantified with ImageJ software and normalized with GAPDH or Histone H3. The following primary antibodies were used: cGAS, STING, p-STING, TBK1, p-TBK1, IRF-3, p-IRF-3 (Cell Signaling Technology), Cmpk2, iNOS, CD206, ARG1 (Proteintech), DEL-1, Spi1 (Santa Cruz Biotechnology), ZO-1, Occludin (Abclonal), GAPDH, Histone H3 (AntGene Biotechnology), and E-cadherin (Genetex).

Measurement of total mtDNA.

Total DNA was extracted using the genomic DNA extraction kit (Solarbio). RT-qPCR was performed to quantify mtDNA levels, including mitochondrial D-loop, Cytochrome b (Cytb), and NADH: Ubiquinone Oxidoreductase Core Subunit 2 (Nd2). Relative mtDNA content was standardized to nuclear DNA encoding Tert. Specific primer sequences are shown in Table S2.

Isolation of colonic infiltrating immune cells and flow cytometry.

The whole colon was isolated and washed with ice-cold PBS. Colon tissues were sectioned to 0.5–1 cm segments and incubated in 20 ml of HBSS (Servicebio) containing 2 mM EDTA (Gibco) and 1 mM DTT (Gibco) with rotation for 15 min at 37 °C. The colons were then treated with 10 ml of DMEM containing 1 mg/ml collagenase D (Sigma-Aldrich) and 0.02 % deoxyribonuclease I (Sigma-Aldrich) for 15 min at 37 °C under rotation. The supernatants were collected through a 70-um filter and centrifuged at 1500 r.p.m. Cells were blocked with anti-mouse CD16/32 (BioLegend) for 10 min, and then incubated with CD45-APC-Cy7 (BD Pharmingen), CD11b-FITC, Ly-6G-PE-Cy7, and Zombie-BV510 (BioLegend) at 4 °C for 30 min in the dark. The results were analyzed with flow cytometer (BD Biosciences) and FlowJo software.

ChIP assays.

Using the ChIP kit (Beyotime), macrophages were cross-linked in 1 % formaldehyde, fragmented by ultrasound, and protein-DNA complexes were incubated with Spi1 antibody (Abcam) or IgG (Cell Signaling Technology) overnight at 4 °C with gentle agitation. The Complexes were incubated with magnetic beads for 1 h at 4 °C under rotation, and then the DNA was eluted, reverse-crosslinked, and purified. qPCR was performed using primers flanking the Spi1 binging site on the Cmpk2 or Il10 gene promoter, and the sequences are listed in Table S2. Moreover, PCR amplification was conducted using M−PCR OPTI™ Mix (Selleck) and detected by agarose gel electrophoresis.

Luciferase reporter assay.

The promoter sequences of the Cmpk2 and Il10 genes were subcloned into the pGL3 promoter luciferase vector (General Biol). HEK293T cells were transfected with empty vector, reporter plasmids, Renilla luciferase vector, and Spi1 plasmid using Lipofectamine 2000 (Invitrogen). The dual-luciferase activity was quantified in a dual luciferase assay kit (Beyotime), and standardized with Renilla luciferase activities.

Statistical analysis.

Statistical analysis was performed using GraphPad Prism and SPSS. Data are presented as the mean ± standard error (SEM). Associations between serum DEL-1 expression and CRP and ESR were evaluated by Spearman’s correlation test. The diagnostic test of serum DEL-1 levels was determined by ROC curve analysis. Student's t-test or ANOVA was utilized to compare group differences between two or multiple groups, respectively. Statistical significance was considered at a p-value of < 0.05.

Results

DEL-1 expression was downregulated in the acute phase and reversed in the recovery phase of IBD.

Previous studies have reported that promoting resolution of inflammation and repair is a novel strategy for the treatment of chronic inflammatory diseases [2,3]. To identify a potential key gene inducing IBD inflammation resolution, we combined the GEO dataset of colon tissues from IBD patients (GSE179285) and inflammation resolution related gene from the GeneCards website. Venn diagram identified 258 intersecting genes (Fig. S1B), and the heatmap shows the expression of these intersecting genes in UC and CD (Fig. S1C and D). Gene enrichment analysis indicated 258 intersecting genes associated with inflammatory response, interferon-gamma-mediated signaling pathway, IL-17 signaling pathway, and TNF signaling pathway (Fig. S1E and F). Among the significantly down-regulated intersecting genes, we identified the homeostatic molecule DEL-1 with promotion inflammation resolution effect (Fig. S1G and H) [13,15]. Based on the same GEO dataset, we found that DEL-1 was significantly reduced in the inflamed colon of patients with UC and CD compared to controls and uninflamed colon (Fig. 1A). We examined DEL-1 expression levels in patients with different disease activity in a new GEO dataset (GSE75214), and found that a significant elevation of colonic DEL-1 levels in patients with UC in remission, while ileal DEL-1 levels remained unchanged in patients with CD in an inactive state (Fig. 1A and Fig. S1I).

Fig. 1.

DEL-1 down-regulated in the acute phase and up-regulated in the recovery phase in patients with IBD and DSS-induced colitis mice. (A) Comparison of DEL-1 expression in the colon tissues between controls (Ctrl), UC, and CD (GSE179285, GSE75214). (B) DEL-1 mRNA expression in the colon tissues from controls and IBD patients were measured by RT-qPCR, normalized to GAPDH. (C) Serum DEL-1 concentrations of controls and IBD patients were determined by ELISA. (D) Receiver operating characteristic (ROC) curve of serum DEL-1 in identifying patients with active IBD from the control group. (E) Correlation between serum DEL-1 and CRP or ESR levels were determined by Spearman’s correlation analysis. (F) Immunohistochemical staining of colonic tissues from controls, UC, and CD patients. (G) Comparison of DEL-1 expression in the colon tissues between controls (Ctrl), and colitis in the acute and recovery phase (GSE42768). (H) Relative mRNA levels of DEL-1 in different organs of wild-type mice were detected by RT-qPCR. The Y axis indicate the ratio of the cycle threshold (CT) value of β-actin to DEL-1 (n = 5). (I and J) DEL-1 mRNA (I) and protein (J) expression in the colon tissues of control mice and colitis mice in the acute and recovery phase was measured by RT-qPCR and western blot. (K) The intensity ratio of DEL-1 to GAPDH was quantified by densitometry analysis. (L, M) Assessment of F4/80 (green) and DEL-1 (red) expression and localization in the colon tissues from control mice and colitis in the acute and repair model by immunofluorescence. (N) The DEL-1 fluorescence intensity of control mice and colitis mice in the acute and recovery phase. Statistical analysis was calculated by student’s t tests or one-way-analysis of variance (ANOVA). ns (not significant), p > 0.05; *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

We further evaluated DEL-1 expression in IBD patients and colitis mice. As shown in Fig. 1B, the mRNA level of DEL-1 in the inflamed colonic mucosa of UC and CD were significantly lower than those in the uninflamed mucosa of patients and normal mucosa of controls, and colonic DEL-1 levels were significantly higher in inactive UC and CD patients. Serum DEL-1 levels were dramatically decreased in UC and CD patients compared with the control group, while DEL-1 levels were upregulated during the recovery phase (Fig. 1C). The ROC curve revealed that when the serum DEL-1 cutoff value was taken as 34.16 pg/ml, the AUC and p value for identifying patients with active IBD from the control group was 0.817 and < 0.001, respectively (Fig. 1D). A statistically significant negative correlation was observed between serum DEL-1 levels and CRP or ESR (Fig. 1E). Immunohistochemical analysis showed that DEL-1 was mainly expressed in colonic mucosa lamina propria and reduced in UC and CD patients (Fig. 1F). In addition, the GSE16879 dataset showed that DEL-1 levels correlated with treatment response to infliximab in UC patients (Fig. S1J). The GEO dataset of colon tissues from colitis mice (GSE42768; GSE34874; GSE64932) also showed down-regulation of DEL-1 expression relative to controls in the acute phase and a rebound in the recovery phase (Fig. 1G and Fig. S1K). As shown in Fig. 1H, DEL-1 was expressed in various organs, with higher expression levels in the colon. RT-qPCR and western blot results confirmed that DEL-1 levels were decreased during the acute phase and increased during the recovery phase in mice with colitis (Fig. 1I-K). Immunofluorescence staining showed that differential expression of DEL-1 in different inflammatory states, and partially co-localized with macrophages (F4/80, a mouse mature macrophage − specific marker) (Fig. 1L-N). As reported, we speculate that DEL-1 is expressed not only in resident stromal cells and macrophages, but also partially in recruited circulating monocyte-derived mature macrophages [13]. Our findings collectively suggest that DEL-1 may serve as a crucial biomarker and potential therapeutic target for IBD.

DEL-1 attenuates macrophage inflammation and promotes inflammation resolution

Inflammation resolution is an active process of reduction of inflammatory effectors and restoration of tissue structure after the withdrawal of inflammatory stimuli [3]. Given that DEL-1 is partially derived from macrophages and that macrophages play an important role in inflammation resolution, we next investigated the expression of DEL-1 in macrophages under different inflammatory states. We pulsed macrophages with LPS for 4 h to induce an acute model, or induced a repair model by replacing with fresh media after washing away the LPS (Fig. 2A). The results showed that the mRNA level of DEL-1 was down-regulated after pulsed with LPS in RAW264.7 macrophages, while DEL-1 levels gradually increased with prolonged LPS withdrawal time, with a consistent trend of protein level changes (Fig. 2B-D). This phenomenon was also observed in BMDM cells (Fig. 2E-G). To dissect the role of DEL-1 in inflammation, RAW264.7 macrophages were transfected with DEL-1 overexpression plasmid, in which DEL-1 mRNA and protein expression levels were significantly up-regulated (Fig. S2A-C). Immediately following the LPS pulse in RAW264.7 macrophages, the levels of inflammatory cytokines Il1β, Il6, and Tnfα were significantly lower in the DEL-1 overexpression group than in the transfected control plasmid group (Fig. 2H). Following the withdrawal of LPS, the production of inflammatory cytokines was reduced, and the between groups trend was consistent with the acute model, but the DEL-1 overexpression group significantly up-regulated the expression of reparative genes Il10 and Arg1 (Fig. 2H). To further elucidate the timing of DEL-1 action in the two models, we constructed DEL-1-Fc fusion proteins and administered DEL-1 intervention at different times in RAW264.7 macrophages. The results showed that treatment with DEL-1-Fc reduced the levels of inflammatory cytokines in the two models, and increased the expression of reparative genes in the repair model (Fig. S3A-E). This suggests that DEL-1 induced inflammation resolution during the recovery phase is not merely a continuation of anti-inflammatory effect of the acute phase. We also used fusion proteins to intervene in BMDMs and showed that DEL-1-Fc intervention significantly reduced inflammatory cytokines expression in both models, and up-regulated the levels of reparative genes Il10 in the repair model (Fig. 2I).

To explore the underlying mechanisms of DEL-1 in inflammation resolution, we performed RNA-seq analysis on RAW264.7 macrophages with and without DEL-1 overexpression in a repair model. The heatmap displayed 616 genes up-regulated and 719 genes down-regulated in DEGs (Fig. 2J). GSEA analysis revealed that inflammation-related pathways were significantly inhibited in the DEL-1 overexpression group, including the NF-kappa B signaling pathway, TNF signaling pathway, chemokine signaling pathway, and IL-17 signaling pathway (Fig. 2K). The GO and KEGG analysis showed that DEGs concentrated in positive regulation of interferon-beta production, mitochondrion, and cytosolic DNA-sensing pathway (Fig. 2L and N). Cytosolic DNA is sensed by cGAS and activates STING, triggering type I interferon and inflammatory response. cGAS-STING activation is implicated in the development of IBD [27]. The volcano plot demonstrated a significantly decreased expression of Cmpk2 in the macrophage with transfection DEL-1 overexpression plasmid (Fig. 2M). Cmpk2, a mitochondrial nucleotide kinase, is a key regulator of the mitochondrial DNA (mtDNA) salvage pathway [28]. Recently reported Cmpk2-dependent mtDNA synthesis is strongly associated with the activation of the cGAS-STING pathway [29,30]. Thus, DEL-1 attenuates inflammation and induces inflammation resolution, and the mechanisms may involve the Cmpk2 and cGAS-STING pathway.

DEL-1 inhibits the Cmpk2-dependent mtDNA synthesis and cGAS-STING pathway in macrophages.

RT-qPCR and western blot were used to determine the expression of Cmpk2 and cGAS-STING pathway in macrophages after LPS pulse or 24 h post LPS. Our results showed a dramatically reduction in the mRNA expression of Cmpk2 and Ifnβ in the OE-DEL-1 + LPS group compared to the LPS group in both models in RAW264.7 cells (Fig. 3A). The similar decreased expression of Cmpk2 and Ifnβ were observed in BMDMs and RAW264.7 cells treated with DEL-1-Fc in the acute and repair model (Fig. 3D, and Fig. S4A and B). The protein levels of Cmpk2 and cGAS were increased after LPS pulse and remained higher than those of controls even 24 h post LPS withdrawal, and DEL-1 overexpression inhibited their elevated levels (Fig. 3B, C). In addition, we observed that the obviously decreased phosphorylation levels of STING, TBK1, and IRF3 in the DEL-1 overexpression group compared to the LPS group in both models (Fig. 3B, C). The protein levels of the Cmpk2 and cGAS-STING pathway showed a consistent trend in both interventions (DEL-1 overexpression and DEL-1-Fc treatment) and across the two cell types (RAW264.7 and BMDMs) (Fig. 3E, F and Fig. S4C-E). Given that Cmpk2 is required for stimulation of mtDNA synthesis, and mtDNA further mediates activation of the cGAS-STING pathway [31,32], we hypothesized that DEL-1 inhibits Cmpk2 expression, which in turn reduces mtDNA levels and inhibits cGAS-STING activation. As illustrated in Fig. 3G and H, DEL-1 overexpression significantly reduced mtDNA synthesis after LPS challenge in both models in RAW264.7 macrophages and BMDMs. Moreover, DEL-1 overexpression strongly inhibited the LPS-stimulated increase in extranuclear dsDNA in the acute and repair model in RAW264.7 cells (Fig. 3I and J). DEL-1-Fc treatment also reversed the reduced mitochondrial membrane potential after LPS challenge in both models in RAW264.7 macrophages, suggesting that DEL-1 could ameliorate mitochondrial functional impairment after LPS stimulation (Fig. S4F-H). Together, these results suggest that DEL-1 inhibits Cmpk2-dependent mtDNA synthesis and activation of the cGAS-STING pathway.

Fig. 3.

DEL-1 inhibits the Cmpk2-dependent mtDNA synthesis and the activation of cGAS-STING pathway in macrophages. RAW264.7 macrophages were transfected with DEL-1 overexpression plasmids and corresponding controls for 24–36 h, and then pulsed with LPS (1 μg/ml) for 4 h or withdrawal of LPS stimulation for 24 h to induce acute and repair models (n = 3–4). BMDMs were pulsed with LPS (1 μg/ml) and simultaneously treated with DEL-1-Fc (1 μg/ml) for 4 h (the acute inflammation model), or withdrawn LPS stimulation and replaced with fresh media and treated with DEL-1-Fc (1 μg/ml) for 24 h (the repair model) (n = 3–4). (A, D) The mRNA expression of Cmpk2, Ifnα, and Ifnβ were measured by RT-qPCR in the acute and repair model in RAW264.7 macrophages (A) and BMDMs (D), normalized to β-actin. (B, C, E, F) The expression of CMPK2 and cGAS-STING pathway related protein were determined by western blot in RAW264.7 macrophages (B, C) and BMDMs (E, F), and densitometric analysis quantified the intensity ratio of the target protein to corresponding controls: CMPK2/GAPDH, CGAS/GAPDH, p-STING/STING, p-TBK1/TBK1, and p-IRF3/IRF3. (G, H) Relative total mtDNA amount in macrophages pulsed with LPS and 24 h post LPS withdrawal was detected using RT-qPCR in RAW264.7 macrophages (G) and BMDMs (H). Results were standardized by Tert nuclear (n) DNA. (I, J) Representative images of immunofluorescence staining for dsDNA (green) and mitochondria (red) in the acute and repair model in RAW264.7 cells. Statistical analysis was calculated by one-way-analysis of variance (ANOVA). ns (not significant), p > 0.05; *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

DEL-1 regulates transcription of Cmpk2 and reparative gene Il10 by modulating the ubiquitin–proteasome-dependent degradation of transcription factor Spi1.

Since DEL-1 regulates Cmpk2 mRNA levels, we further explored possible mechanisms of transcriptional regulation. We predicted transcription factors likely to bind the Cmpk2 promoter regions using the UCSC Genome Browser and the JASPAR databases. Among the candidate transcriptional factors, we identified the differentiation of macrophage lineage determining transcription factor Spi1. Previous studies have reported that targeting Spi1 inhibits the inflammatory response in macrophages and that Spi1 null mice rapidly repair skin wounds [33,34]. This suggests that Spi1 may be a key transcription factor in the regulation of inflammation and repair. The JASPAR database also identified the transcription binding sites of Spi1 in the promoter region of Il10. We detected the expression of Spi1 in the nuclear and cytoplasm in RAW264.7 cells. The results showed DEL-1 overexpression significantly reversed LPS-stimulated elevated expression of Spi1 in the nuclear and cytoplasm in both models (Fig. 4A and B). In BMDMs, DEL-1-Fc treatment also decreased the overall level of Spi1 after LPS challenge in both models (Fig. 4C and D). The overexpression of Spi1 in RAW264.7 macrophages via the transfection of plasmid was confirmed by RT-qPCR and western blot (Fig. S2D-F). RT-qPCR showed that Spi1 overexpression significantly increased the expression of Cmpk2 and decreased the expression of Il10 (Fig. 4E and F). We further investigated the regulatory role of the transcription factor Spi1 on target genes in RAW264.7 and HEK293T cells using ChIP and dual-luciferase reporter assays. ChIP analysis confirmed that Spi1 was directly associated with the Cmpk2 promoter and the IL-10 promoter in the normal environment (Fig. 4G and J). Moreover, the interaction of Spi1 with the Cmpk2 promoter and the IL-10 promoter was down-regulated by DEL-1 overexpression after LPS challenge in both models (Fig. 4H, I, K, L). Dual-luciferase reporter assays showed that the luciferase activity was elevated in the pGL3 luciferase reporter vector containing the Cmpk2 promoter sequence and decreased in the reporter vector containing the Il10 promoter sequence upon overexpression of Spi1 in HEK293T cells (Fig. 4M and N).

Fig. 4.

DEL-1 regulates transcription of Cmpk2 and reparative gene Il10 though transcription factor Spi1. (A, B) The expression of nuclear SPI1 and cytoplasm SPI1 were measured by western blot in DEL-1 overexpressed RAW264.7 macrophages pulsed with LPS and 24 h post LPS withdrawal, and the intensity ratio of the target protein to HISTONE 3 or GAPDH quantified using densitometric analysis (n = 3–4). (C, D) The expression of SPI1 were measured by western blot in BMDMs in the acute and repair model, and densitometric analysis quantified the intensity ratio of the SPI1 to GAPDH (n = 3–4). (E, F) The mRNA expression of Cmpk2 (E) and Il10 (F) were measured by RT-qPCR in RAW264.7 macrophages with Spi1 overexpression, normalized to β-actin. (G-N) Chromatin immunoprecipitation (ChIP) and dual-luciferase reporter assays analyzed the regulatory role of the transcription factor Spi1 on target genes in RAW264.7 and HEK293T cells. (G) ChIP analyzed the association between Spi1 and the promoter of Cmpk2 gene without intervention. Agarose gel electrophoresis of products (top), and %input calculated from CT values of RT-qPCR (bottom). (H, I) ChIP analyzed the association between Spi1 and the promoter of Cmpk2 gene in the acute and repair model. (J) ChIP analyzed the association between Spi1 and the promoter of Il10 gene without intervention. Agarose gel electrophoresis of products (top), and %input calculated from CT values of RT-qPCR (bottom). (K, L) ChIP analyzed the association between Spi1 and the promoter of Il10 gene in the acute and repair model. (M) The luciferase activity of the Cmpk2 promoter with Spi1 binding sites in HEK293T cells. (N) The luciferase activity of the Il10 promoter with Spi1 binding sites in HEK293T cells. Statistical analysis was calculated by student’s t tests or one-way-analysis of variance (ANOVA). ns (not significant), p > 0.05; *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

To determine the regulatory relationships, we overexpressed Spi1, Cmpk2 or intervened with the STING pathway agonist DMXAA in both models in RAW264.7 cells (Fig. S2D-F and Fig. 5A-C). We observed significant reversal of Spi1 overexpression, Cmpk2 overexpression, or DMXAA intervention for DEL-1 overexpression induced reduction of the levels of Cmpk2 and cGAS after LPS challenge in both models (Fig. 5D-F). Moreover, the significantly down-regulated phosphorylated STING, TBK1, and IRF3 in the LPS + OE-DEL-1 group of the acute and repair model, which was abolished by overexpression of Spi1, Cmpk2, or DMXAA intervention (Fig. 5D-F). We explored the mechanism of the regulating Spi1 expression of DEL-1, and RT-qPCR assay showed the mRNA levels of Spi1 without significant change after DEL-1 overexpression in both models in RAW264.7 macrophages (Fig. 5G). Thus, we analyzed whether DEL-1 could influence the protein stability of Spi1. Macrophages were treated with CHX, proteasome inhibitor MG132, and lysosomal inhibitor chloroquine to block new protein synthesis, the ubiquitin–proteasome pathway, and autophagy-lysosome pathway, respectively. The results indicated DEL-1 overexpression markedly accelerated the degradation of Spi1 proteins after LPS challenge in both models, which was reversed MG132 but not chloroquine (Fig. 5H and I). Therefore, our data demonstrated that DEL-1 regulates the transcription of Cmpk2 and reparative gene Il10 by repressing the protein stability of transcription factor Spi1, thereby influencing the cGAS-STING pathway activation and reparatory macrophage transition.

Fig. 5.

DEL-1 induces the ubiquitin–proteasome-dependent degradation of transcription factor Spi1, and further inhibits the Cmpk2-cGAS-STING pathway in macrophages. (A-C) Cmpk2 mRNA (A) and protein (B, C) expression in RAW264.7 macrophages with transfection Cmpk2 overexpression plasmid was determined using RT-qPCR and western blot (n = 4). (D-F) RAW264.7 macrophages were transfected with DEL-1, Spi1, and Cmpk2 overexpression plasmids and corresponding controls for 24–36 h. Cells were pulsed with LPS (1 μg/ml) and the STING pathway agonist DMXAA (1 μg/ml) for 4 h in the acute phase, or withdrawn LPS stimulation and treated with DMXAA for 24 h in the recovery phase (n = 4). The expression of CMPK2 and cGAS-STING pathway related protein were measured by western blot, and densitometric analysis quantified the intensity ratio of the target protein to relevant controls: CMPK2/GAPDH, CGAS/GAPDH, p-STING/STING, p-TBK1/TBK1, and p-IRF3/IRF3. (G) The mRNA expression of Spi1 were determined by RT-qPCR in DEL-1 overexpressed macrophages pulsed with LPS and 24 h post LPS withdrawal. (H, I) DEL-1 overexpressed macrophages treated with cycloheximide (CHX, 60 μg/ml), MG132 (20 uM), and chloroquine (50 uM) in the acute and repair model, and the expression of SPI1 was measured using western blot. Statistical analysis was calculated by student’s t tests or one-way-analysis of variance (ANOVA). ns (not significant), p > 0.05; *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

DEL-1 alleviated intestinal inflammation and induced tissue repair.

To investigate the role of DEL-1 in the resolution of intestinal inflammation, we delivered AAV-DEL-1 to knockdown DEL-1 expression four weeks prior to establishing the repair model (Fig. 6A). Administration of AAV-DEL-1 treatment reduced the expression of DEL-1 in colon (Fig. S5A-C). As exhibited in Fig. 6B-E, AAV-DEL-1 intervention significantly exacerbated body weight loss, DAI scores, and colon shortening compared to the DSS mice in the repair model. Histological analysis showed that DSS mice suffered from structural disruption and inflammatory cell infiltration, which were worsened by DEL-1 knockdown (Fig. 6F, G). RT-qPCR showed that DEL-1 knockdown significantly increased proinflammatory cytokine levels (Il1β, Il6, Tnfα), reduced reparative genes (Il10 and Arg1) and upregulated STING pathway associated cytokines (Ifnα and Ifnβ) in colitis mice of repair model (Fig. 6H). Compared with the DSS + AAV-scramble group, the DSS + AAV-DEL-1 group enhanced Spi1-Cmpk2-cGAS-STING pathway activation, as evidenced by increased levels of Spi1, Cmpk2 and cGAS, as well as elevated phosphorylation of STING, TBK1, and IRF3 (Fig. 6I, J).

Additionally, we treated mice with DEL-1-Fc fusion protein in the acute and repair model of colitis (Fig. S6A and Fig. S7A). Our data revealed that DEL-1 ameliorated acute intestinal inflammation and induced tissue repair, indicated by improved of body weight, DAI, colon length, spleen weight, and histological score in both models (Fig. S6B-H and Fig. S7B-H). The mRNA levels of proinflammatory and STING pathway associated cytokines were significantly reduced by DEL-1-Fc treatment in both models, while the levels of reparative genes only upregulated in the repair model (Fig. S6I and Fig. S7I). This suggests that reparative genes are altered during the recovery phase. Western blot showed that DEL-1 inhibits Spi1-Cmpk2-cGAS-STING signaling activation in DSS mice of both models (Fig. S6J, K and Fig. S7J, K). Collectively, these results suggested that DEL-1 ameliorates intestinal inflammation, and promotes resolution of inflammation and repair of tissue damage.

DEL-1 reduced neutrophils infiltration, promoted reparatory macrophage transition, induced intestinal epithelial barrier repair, and improved intestinal microbiota of DSS-induced colitis in the repair model.

To further explore the pro-resolving function of DEL-1, we analyzed DSS mice treated with DEL-1 in the repair model. Previously reported that DEL-1 inhibits leukocyte recruitment and promotes apoptotic neutrophil clearance [15]. Flow cytometry analysis showed a significant reduction in colonic infiltrating neutrophils in DSS mice treated with DEL-1-Fc relative to those in only DSS mice (Fig. 7A and B). Reparative macrophages play a crucial contribution in the process of inflammation resolution. We found that DEL-1 intervention induced macrophage phenotype switching from pro-inflammatory to reparative phenotype (Fig. 7C-E). Mice with colitis were concurrently administered DEL-1-Fc and DMXAA to verify whether DEL-1 promotes the resolution of intestinal inflammation via the cGAS-STING pathway (Fig. 7F). DSS-induced mortality is attenuated by DEL-1-Fc intervention, but this attenuation is reversed by DMXAA (Fig. 7G). Moreover, DMXAA abolished the beneficial effects of DEL-1 on colon length restoration, histological improvement, and reduction of proinflammatory and STING pathway associated cytokines in the repair model of colitis in mice (Fig. 7H-L). The upregulated reparative genes in colitis mice administered DEL-1-Fc were withdrawn by DMXAA (Fig. 7L). We also found that DMXAA significantly reversed DEL-1-mediated polarization of macrophages towards a reparative phenotype (Fig. 7N, M). Our findings indicate that DEL-1 contributes to the resolution of intestinal inflammation by inhibiting the activation of the cGAS-STING pathway. Furthermore, the cGAS-STING pathway is also involved in regulating the reparative transition of macrophages.

Considering the crucial role of restored intestinal epithelial homeostasis and integrity in inflammation resolution, we evaluated the proliferation and apoptosis of intestinal epithelial cell, and intestinal barrier function using TUNEL staining, western blot and immunofluorescence. The numbers of apoptotic cells were increased after DSS challenge, which was reversed by treating with DEL-1-Fc (Fig. S8A). We also identified increased proliferation in the DEL-1-Fc-treated group, as evidenced by increased Ki-67 positive cells in the intestinal crypt compared with the DSS group (Fig. S8B). The expression of ZO-1, occludin and E-cadherin were significantly up-regulated in mice with colitis after DEL-1-Fc treatment (Fig. S8C-F). Then, we detected the intestinal barrier function after DEL-1-Fc intervention in vitro using HT-29 cells. DEL-1 promoted the synthesis of intestinal barrier-associated proteins after LPS challenge, as evidenced by western blot and immunofluorescence staining (Fig. S9B-E). We also found that DEL-1 induced HT-29 cells proliferation in dose and time dependent manner (Fig. S9A). Furthermore, DEL-1-Fc treatment reversed the activation of the cGAS-STING pathway induced by DMXAA in HT-29 cells (Fig. S9F-H).

Given that gut microbiota dysbiosis is associated with aberrant immune activation in IBD patients, we investigated whether DEL-1 altered microbiome using 16S rRNA. Venn diagram showed the number of shared and unique OTUs between groups, and a lower number of OTUs was observed in the DSS group (Fig. S10A). Compared to the DSS group, DEL-1-Fc treatment significantly increased the microbial richness and diversity, as evidenced by alpha diversity analysis (chao1, simpson, and shannon_2 indexes) (Fig. S10B). We further examined microbial community composition using NMDS and PCA of beta diversity. The result showed that there were significant differences in microbial community structure between the DSS group and the control group, while the DEL-1-Fc intervention restored the composition of microbial to be similar to those of normal mice (Fig. S10C). The taxonomic composition of intestinal microbiota was assessed at the phylum, family and genus level (Fig. S10D). At the phylum level, the results indicated that DSS mice treated with DEL-1-Fc exhibited a higher abundance of Firmicutes and a lower abundance of both Proteobacteria and Verrucomicrobia compared to DSS mice. We also identified significantly different species using LEfSe analysis, and found differences in microbial community dominance among the three groups at multiple phylogenetic levels (Fig. S10E and F). According to the LDA score, genus Allobaculum was the most predominant microbiota in the Ctrl group. In the DSS + Fc group, class Verrucomicrobiae was the significantly different strain, while in the DSS + DEL-1-Fc group, family Rhodobacteraceae was the notably distinct strain. These results provide evidence that DEL-1 contributes to resolution of inflammation by improving intestinal microbiota, reducing neutrophil infiltration, promoting reparatory macrophage transition, and repairing intestinal barrier damage.

Discussion

To the best of our knowledge, this is the first study of DEL-1 expression and mechanism in IBD. We found that DEL-1 expression was downregulated during the acute phase of inflammation, and resurged as inflammation resolution in IBD patients, colitis mice, and LPS-stimulated macrophages. We confirmed that DEL-1 attenuates intestinal inflammation by being anti-inflammatory and promoting inflammation resolution. Most notably, our study demonstrates that DEL-1 inhibits the Cmpk2-cGAS-STING pathway and induces reparatory macrophage by regulating Spi1 transcriptional modulation. Furthermore, the cGAS-STING pathway is involved in regulating reparatory macrophage transition. In parallel, DEL-1 also plays an important role in the regulation of neutrophils infiltration, intestinal microbiota, and intestinal barrier.

The proposal of resolution pharmacology offers new opportunities for the treatment of inflammatory diseases [35]. Resolution-based therapeutic strategy that not only terminates the inflammatory response but also induces endogenous repair pathways to promote healing. However, therapeutic strategies using SPM mimetics and analogs to induce resolution are limited by synthetic complexity and pharmacokinetics. In this study, we targeted and explored the multifunctionality and homeostatic properties of the modular structural protein DEL-1 in IBD in conjunction with GEO database analysis and previous studies [13]. DEL-1 has been previously reported to inhibit the initiation of inflammation and induce inflammation resolution by mechanisms involving inhibition of inflammatory cell recruitment, promotion of efferocytosis, and modulation of SPM expression and Treg cell responses [14,15,36,37]. Consistent with previous studies of inflammation-related diseases, the expression of DEL-1 showed a trend of down-regulation followed by up-regulation from the initiation of inflammation to its resolution in IBD patients, colitis mice, and macrophages [15,18]. Moreover, DEL-1 was predominantly expressed in intestinal macrophages and stromal cells.

In our study, we found that DEL-1 overexpression significantly attenuated macrophage inflammation in both models, and induced reparative gene expression in the recovery phase of inflammation. Consistent with these findings, DEL-1-Fc intervention yielded similar results. Collectively, the results provide compelling evidence for the anti-inflammatory properties of DEL-1, and demonstrate a specific role for DEL-1 in promoting inflammation resolution. Pro-resolution strategies have been reported to target endogenous pathways leading inflammation to resolve during the recovery phase of inflammation [2,3]. Next, we performed RNA-seq analysis of LPS-stimulated macrophages with and without DEL-1 overexpression at the recovery stage of inflammation and showed significant changes in the cytosolic DNA-sensing pathway and Cmpk2 gene. Double-stranded DNA (dsDNA) is recognized by cGAS, and activated cGAS induces the second messenger cGAMP, which in turn activates STING and initiates the phosphorylation of TBK1 and IRF3, leading to the expression of type I interferon and autoimmune disease [38]. Cmpk2, a member of the nucleotide kinase family in mitochondria, catalyzes the synthesis of dCDP and then converts to dCTP to provide the necessary deoxyribonucleotides for mtDNA synthesis, which is associated with the cGAS-STING pathway activation [29,30,32]. Negative regulation of immune signaling is indispensable for resolving immune responses and avoiding excessive inflammation. We demonstrate that DEL-1 reduces Cmpk2-dependent mtDNA synthesis, thereby inhibiting activation of the cGAS-STING pathway in both acute and repair models.

Our study has identified that DEL-1 reduced the mRNA expression of Cmpk2, and speculated that the transcription factor Spi1 plays a transcriptional regulatory role in it. Spi1 is a lineage-determining transcription factor involved in macrophage differentiation and activation that drives NF-κB activation and inflammatory cytokine expression [[39], [40], [41]]. Whole-blood transcriptomic analysis of IBD patients found that Spi1 was a central transcription factor in IBD, where it can regulate Th9 immunity and macrophage polarization [42,43]. Increased Spi1 expression levels were identified in IBD, and Pterostilbene inhibited the DNA-binding ability of Spi1, which in turn promoted DC-induced Tregs differentiation [44,45]. It has been reported that Spi1 was associated with reparative function, and Spi1 null mice rapidly repair skin wounds [34]. In addition to regulating Cmpk2 and subsequent activation of the cGAS-STING pathway, DEL-1 also increased the reparative gene Il10 mRNA expression, which was also demonstrated in our study to be a direct target of Spi1. Previous studies have reported the development of spontaneous chronic enterocolitis in IL-10-deficient mice [46]. Anti-TNF treatment increased Il10 production from macrophages, which induced macrophages transition toward CD206 + phenotype, whereas blockade of IL-10 signaling by anti-IL-10Rα decreased the efficacy of anti-TNF therapy in IBD [47]. EPRS1 coordinates AKT signaling and then increases the production of anti-inflammatory cytokine Il10, leading to inflammation resolution and mucosal homeostasis restoration [48]. Notably, we found that DEL-1 inhibited the stability of Spi1 and reduced its expression in the nucleus and cytoplasm in both models by inducing a ubiquitin–proteasome-dependent degradation pathway. Thus, Del-1 negatively regulates the immune pathway on the one hand, and induces the expression of reparative gene on the other hand, thereby promoting reparatory macrophage transition while exerting an anti-inflammatory effect. This two-pronged mechanism of action plays an important role in inducing the inflammation to resolve.

Although it has been shown that DEL-1 reduces inflammation and accelerates its resolution in different mouse models, such as periodontitis [15], inflammatory arthritis [16], asthma [17], and experimental autoimmune encephalomyelitis [18], there are no studies that have explored the role of DEL-1 in models of colitis. In this study, we further investigated the therapeutic effect of the two-pronged mechanism of action of DEL-1 in colitis. Our results showed that DEL-1-Fc intervention significantly alleviated intestinal inflammation whether administered at the initiation of the inflammation or during the recovery phase. In addition, DEL-1 inhibited the Spi1-Cmpk2-cGAS-STING signaling in both acute and repair models, and upregulated the expression of reparative genes and induced reparatory macrophage in the repair model. Notably, DEL-1 knockdown mice exhibited opposite trends in these effects. The cGAS-STING pathway plays an important role in macrophage polarization [49,50]. Furthermore, we discovered that the reparative phenotype switch of macrophages induced by DEL-1 was reversed by DMXAA. Recent studies have shown accumulating evidence that DEL-1 could prevent the recruitment of proinflammatory cells to target tissues and induce macrophage reprogramming to a proresolving phenotype [15,16,23,51]. As reported, we found that the increased neutrophil recruitment after DSS stimulation was reversed by DEL-1-Fc treatment. Collectively, we confirmed here that DEL-1, in parallel, not only promoted reparative macrophage transition during the recovery phase of intestinal inflammation but also inhibited neutrophil recruitment. Furthermore, we noted that cGAS-STING activation blocked this transition to a reparative macrophage phenotype.

An important part of inflammation resolution in IBD is intestinal epithelial repair, where rapid epithelial restoration prevents the mucosal immune system from being exposed to bacteria and bacterial antigens, avoiding an imbalanced immune reaction and immune cell recruitment [2,7,52]. Intestinal epithelium is the fastest self-renewing tissue in adult mammals and is essential for maintaining intestinal barrier function. The coordination of proliferation and apoptosis of intestinal epithelial cell is involved in the regulation of intestinal epithelial cell renewal and is important for the maintenance of mucosal homeostasis [53,54]. DEL-1 has been reported to facilitate apoptotic neutrophil clearance and hematopoietic stem cell proliferation [15,55]. Here, we reported that DEL-1 treatment significantly improved intestinal epithelial barrier function by promoting enterocyte proliferation and inhibiting intestinal epithelial cell apoptosis in vivo and in vitro. Furthermore, DEL-1 was also found to inhibit the cGAS-STING pathway in intestinal epithelial cells, the underlying mechanisms still require further investigation.

Additionally, we showed that DEL-1 contributes to resolution of intestinal inflammation by reversing gut microbiota dysbiosis using 16S rRNA sequencing. The dysregulation of intestinal microbiota is an important pathogenesis of IBD, and remolding microbiome plays an important role in restoring homeostasis and inducing inflammation to resolve [56,57]. At the taxonomic level, we observed changes in the abundance of specific phyla in response to DEL-1-Fc treatment. DSS mice treated with DEL-1-Fc exhibited an increase in Firmicutes and a decrease in Proteobacteria and Verrucomicrobia compared to DSS mice. These changes are of particular interest as Firmicutes are often dominant in a healthy gut microbiota, while Proteobacteria have been implicated in inflammatory conditions [58,59]. Furthermore, LEfSe analysis identified significantly different species among the three groups at multiple phylogenetic levels. Most notably, in the DSS + DEL-1-Fc group, family Rhodobacteraceae was identified as the notably distinct strain. Rhodobacteraceae are associated with tryptophan metabolism of the kynurenine pathway [60], and the tryptophan metabolites xanthurenic and kynurenic acids reduce the severity of colitis by affecting intestinal epithelial cells and T cells [61], although the direct involvement of Rhodobacteraceae in producing these specific metabolites remains to be elucidated.

However, our study is subject to certain limitations. Firstly, we used AAV to knockdown DEL-1, but did not employ DEL-1 knockout mice or mice with conditional knockout specifically in intestinal macrophages. Furthermore, despite observing changes in the gut microbiota following DEL-1 intervention, we did not conduct further validation using germ-free mice and microbiota transplantation experiments.

Conclusion

In summary, our results show that DEL-1 expression was strongly associated with disease activity, decreasing during the acute phase of inflammation and resurging after inflammation to resolve. Furthermore, we found that DEL-1 alleviates intestinal inflammation and triggers inflammation resolution by inhibiting the activation of Cmpk2-cGAS-STING signaling and promoting reparative macrophage transition, in which Spi1 plays an important transcriptional regulatory role. The cGAS-STING pathway is also involved in regulating the transition of macrophages towards a reparative phenotype. Taken together, our findings provide a new target for proresolution therapies in IBD by targeting DEL to promote resolution of chronic inflammation and restore intestinal homeostasis.

Compliance with ethics requirements

All Institutional and National Guidelines for the care and use of animals (fisheries) were followed. The collection of samples from patients was approved by Tongji Medical College, Huazhong University of Science and Technology, and adhered to the principles of Helsinki Declaration (S051). The protocol of mouse experiment was approved by the Animal Care and Use Committee of Huazhong University of Science and Technology (IACUC Number 3938).

CRediT authorship contribution statement

Meihui Tao: Data curation, Formal analysis, Investigation, Methodology, Project administration, Software, Validation, Writing – original draft, Writing – review & editing. Li Wang: Data curation, Methodology, Supervision, Resources. Chaoyue Chen: Data curation, Methodology, Project administration, Writing – review & editing. Mengfan Tang: Investigation, Methodology. Yanping Wang: Investigation, Validation. Jingyue Zhang: Investigation, Software. Xi Zhao: Investigation, Visualization. Qinyu Feng: Investigation, Software. Junfa Chen: Investigation, Validation. Wei Yan: Funding acquisition, Resources, Supervision, Writing – review & editing. Rong Lin: Resources, Supervision, Writing – review & editing. Yu Fu: Conceptualization, Data curation, Funding acquisition, Supervision, Writing – review & editing.

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: