Aryl hydrocarbon receptor impairs HK2-controlled flux of the hexosamine biosynthesis pathway to suppress NETosis in an N-glycosylation-dependent manner

Graphical abstract

Aryl hydrocarbon receptor (AhR) targets at hexokinase 2 (HK2) to promote its ubiquitination degradation at K48 sites, impairs the flux of hexosamine-biosynthesis pathway and subsequent N-glycosylation of alpha-1 antitrypsin (AAT) and alpha-2-macroglobulin (A2M), reducing the activity of neutrophil elastase (NE) to abolish NETosis.

Highlights

• •AhR activation has a negative relationship with NETosis in ulcerative colitis (UC).
• •AhR activation suppresses the NETosis via reducing NE activity.
• •AhR activation regulates NE activity by inhibiting N-glycosylation of AAT and A2M.
• •AhR binds with HK2 to restrict flux of HBP and decrease N-glycosylation modification.
• •Inhibiting NETosis is an important approach for targeting AhR to alleviate UC.

Abstract

Introduction

The aryl hydrocarbon receptor (AhR) is a promising therapeutic target for ulcerative colitis (UC) and plays a role in regulating neutrophil function.

Objective

We aimed to investigate the effects and mechanisms of AhR on NETosis, a neutrophil-driven process that disrupts intestinal epithelial homeostasis, to support the development of anti-UC therapies.

Methods

A dextran sulphate sodium (DSS)-induced colitis mouse model was established, and bioinformatics analyses combined with multiple molecular biology techniques were used to assess changes in NETosis and signalling pathway activation.

Results

Data from the Gene Expression Omnibus database and DSS-induced colitis mice confirmed an inverse correlation between AhR activation and NETosis in UC. In vitro experiments, including assays for double-stranded DNA release, co-localisation of myeloperoxidase with DNA and neutrophil elastase (NE)/citrullinated histone H3, histone H4 degradation, and chromatin decondensation, demonstrated that AhR activation directly inhibits NETosis. Further investigations using gene knockdown plasmids, enzyme substrate assays, and flow cytometry revealed that AhR activation reduced the NE activity—independent of pcDNA-peptidyl arginine deiminase 4—through an N-glycosylation-dependent mechanism involving the physiological NE inhibitors alpha-1 antitrypsin (AAT) and alpha-2-macroglobulin (A2M). Mechanistically, AhR functioned as an E3 ligase that bound to hexokinase 2 (HK2), promoting its K48-linked ubiquitination and degradation, thereby impairing the flux of the hexosamine biosynthesis pathway (HBP) and reducing the availability of the glycosylation precursor UDP-GlcNAc.

Conclusion

AhR activation suppresses NETosis by modulating HK2-mediated HBP flux and the subsequent N-glycosylation of AAT and A2M, thereby decreasing NE activity.

Introduction

Ulcerative colitis (UC) is an autoimmune intestinal disease that affects the mucosa and submucosa of the intestinal tract and has a rapidly increasing global incidence. Current therapeutic approaches for UC focus on suppressing aberrant intestinal inflammation and excessive immune responses by blocking inflammatory cytokines such as tumour necrosis factor alpha and interleukin 23, downregulating the activity of the non-receptor tyrosine kinase Janus Kinase (JAK), and inhibiting leukocyte migration via anti-α4β7 integrin therapy [[1], [2], [3]]. Although molecularly targeted drugs have reduced the major side effects of conventional treatments and exhibit strong therapeutic efficacy, approximately 30–55 % of patients with UC do not benefit from these therapies due to variations in the type and quantity of pro-inflammatory substances during disease progression [4]. Therefore, elucidating the pathogenesis of UC and identifying novel regulatory strategies remain crucial.

Neutrophils are the most abundant immune cells, constituting approximately 70 % of circulating leukocytes in humans, and are well known for their rapid recruitment to sites of infection or tissue damage [5]. In UC, persistent activation and excessive recruitment of neutrophils in the intestinal tract are prominent, contributing to inflammation and tissue damage through the release of neutrophil extracellular traps (NETs) via NETosis [6,7]. NETs are web-like structures composed of decondensed chromatin, DNA, and antimicrobial peptides that protrude from the membranes of activated neutrophils. Studies have reported that medium from ex vivo cultures of inflamed mucosa from patients with UC induces increased spontaneous NETs formation, and NETs accumulation is elevated in the inflamed intestinal mucosa, stool, and blood of individuals with inflammatory bowel disease. Moreover, NETs abundance positively correlates with disease severity [8]. In contrast, inhibiting NETs formation enhances the expression of barrier-associated proteins in colonic tissues and restores tight junction integrity in colitis models. These findings suggest that targeted inhibition of NETosis may serve as a potential therapeutic strategy for UC.

The aryl hydrocarbon receptor (AhR), a receptor that functions as an environmental sensor, plays multiple roles in immune regulation. In 2019, benvitimod, an AhR agonist, was approved for the treatment of mild-to-moderate psoriasis and atopic dermatitis [9]. Notably, AhR is highly expressed in the intestinal tract, and its deletion exacerbates UC severity by disrupting the intestinal epithelial barrier [10,11]. In addition, AhR regulates neutrophil phenotypes, including infiltration, chemokine expression, and aggregation [12,13]. As reported, microbiota metabolites (indole alkaloid) antagonize the function of P2X Purinoceptor 1 (P2RX1) via the “AhR-IL-22” axis, thereby decreasing the neutrophil infiltration at lesion sites in colitis-prone mice [14]; the endogenous AhR agonist 6-Formylindolo [3, 2-b] carbazole (FICZ) significantly inhibited the aggregation of neutrophils in lungs of mice with pneumonia [15]; the infiltration of neutrophils in colonic tissues was significantly increased in the colorectal cancer model established with Ah receptor nuclear translocator protein (ARNT) gene knockout mice [16]. Given these findings, this study aimed to investigate whether AhR could inhibit NETosis and elucidate the underlying mechanisms.

Materials and methods

The Supplementary Materials include an expanded Supplementary Materials and Methods section.

Animals

Female C57BL/6 mice (6–8 weeks old, 18–22 g) were obtained from the Comparative Medical Center of Yangzhou University. The experiments were approved by the Animal Ethics Committee of China Pharmaceutical University (approval nos. 2023-02-001, 2023-10-004, and 2023-10-005) and conducted in accordance with the National Institute of Health guidelines on the ethical use of animals. The mice were housed under a 12-h light/dark cycle at approximately 25 °C, with ad libitum access to pathogen-free food and potable water, and were allowed to acclimatise for three days before experimentation.

Colitis model induction and drug administration

The mice were randomly assigned to groups based on body weights. (a) To explore the relationship between AhR and NETosis in UC: normal and model groups; (b) To verify the mechanism for AhR to inhibit NETosis: normal, model, pcDNA-neutrophil elastase (NE; 10 μg/mouse), pcDNA-hexokinase 2 (HK2; 10 μg/mouse), FICZ (1 μg/mouse), 3, 3′-diindolylmethane (DIM; 10 mg/kg), FICZ + pcDNA-NE, FICZ + pcDNA-HK2, DIM + pcDNA-NE, DIM + pcDNA-HK2, and sivelestat (50 mg/kg) groups. Except for the normal group, all mice received 2.5 % dextran sulfate sodium (DSS) in drinking water for seven consecutive days, followed by distilled water for three days. FICZ, DIM, and sivelestat were administered intraperitoneally or orally daily from day 1 to day 10. pcDNA-NE or pcDNA-HK2 was mixed with an equal volume of Entranster™-in vivo transfection reagent and administered rectally on days 1 and 5.

During the experiment, weight loss, stool consistency, and faecal blood were monitored to calculate the disease activity index (DAI), following previously reported scoring criteria [17]. After the final administration, the mice were euthanised, colonic tissues were collected and photographed, and colon lengths were measured.

Intestinal permeability

Following the final drug administration, fluorescein isothiocyanate (FITC)-dextran 4000 (60 mg/100 g; Sigma-Aldrich, St. Louis, MO, USA) was administered intragastrically. After 4 h, whole blood was collected, and serum was prepared. FITC-dextran 4000 concentrations in the serum were measured at excitation and emission wavelengths of 490 and 520 nm, respectively, using a microplate reader (Varioskan Flash; Thermo Electron Corporation, Vantaa, Finland).

Myeloperoxidase (MPO) activity

Colonic tissues from mice with colitis were homogenised, and MPO activity was measured using a commercial kit (Jiancheng Bioengineering Institute, Nanjing, China) according to the manufacturer’s instructions.

Cell culture

Neutrophils were isolated from the peripheral blood of participants, as approved by the Ethics Committee of the Affiliated Hospital of Integrated Traditional Chinese and Western Medicine (Approval No. 2022-LWKY-035), using Polymorphprep™ (Serumwerk Bernburg AG, Germany) according to the manufacturer’s instructions.

The human promyelocytic leukaemia cell line HL60 was obtained from Procell Life Science & Technology Co., Ltd. (Wuhan, China) and cultured in Iscove’s Modified Dulbecco’s Medium supplemented with 20 % fetal bovine serum at 37 °C. HL60 cells were differentiated into neutrophil-like cells (dHL60) by incubation with 1.25 % dimethyl sulfoxide (DMSO) and evaluated by flow cytometry using a PE-anti-CD11b antibody or Wright-Giemsa staining.

dsDNA release assay

Neutrophils and dHL60 cells were incubated with 1 μM SYTOX Green dye, and fluorescence intensity at 502/525 nm was measured using a microplate reader. Fluorescent and confocal laser scanning microscopy images were also acquired.

NE activity assay

NE activity was determined using the hydrolysable substrate MeOSU-Ala-Ala-Pro-Val-PNA. Briefly, cells were lysed in pre-cooled 0.3 % Triton X-100 on ice via ultrasonication (2-mm ultrasonic probe, 300 W power, 5 s pulse, 5 s interval, total duration: 3 min). Colonic tissues were ground to prepare samples. MeOSU-Ala-Ala-Pro-Val-PNA was then added to each sample to a final concentration of 20 μM, and absorbance at 380/460 nm was measured using a microplate reader. Recombinant NE protein was directly incubated with FICZ, DIM, or CH223191, and NE activity was determined using the same method.

Metabolomics analysis

Neutrophils were frozen, ground in liquid nitrogen, and extracted with chloroform/methanol to prepare samples. Spectra were acquired using high-resolution nuclear magnetic resonance (HRNMR). The nuclear overhauser effect spectroscopy (NOESY) pulse train and one-dimensional spectra were obtained under the following conditions: spectral width of 10 kHz, acquisition time of 1.64 s, relaxation delay of 2.00 s, and temperature of 299 K. The free induction decay signal was collected, and data points were acquired. The collected nuclear magnetic resonance (NMR) spectra were analysed using Chenomx NMR Suite software, and statistical analysis with data visualisation was performed using the MetaboAnalyst platform.

Statistical analysis

Statistical analyses were conducted using IBM SPSS Statistics for Windows, version 19.0 (IBM Corp., Armonk, NY, USA). Data are presented as mean ± S.E.M. All experiments were performed with at least three to five independent repeats or six to ten mice per group. In all cases, n refers to the number of independent experiments rather than technical replicates. Student’s t-test was used for comparisons between two groups, while one-way ANOVA was used for comparisons among multiple groups. If variances were equal, the LSD test was applied; otherwise, the Games-Howell test was used. Pearson correlation analysis was applied to assess the correlation between variables in the two groups. Statistical significance was set at P < 0.05.

Results

AhR activation is inversely associated with NETs formation in UC

To determine the relationship between AhR and NETosis, the datasets GSE53306 and GSE67577, derived from mucosal biopsies of patients with UC and colons of mice with DSS-induced colitis, respectively, were selected from the Gene Expression Omnibus (GEO) database for gene set enrichment analysis (GSEA). As shown in Fig. S1, AhR-related signalling was insufficient under UC conditions. In contrast, except the cytokine-cytokine receptor interaction and JAK-signal transducer and activator of transcription (STAT) signals, the upstream signals regulating NETs formation were significantly enriched and exhibited an opposing trend, which were negative correlation to AhR activation. Then, a classic colitis mouse model was established, and disease severity was assessed based on DAI scores, colon length, and histopathological changes in the colons. The quantitative polymerase chain reaction (Q-PCR) analysis was performed to quantify mRNA levels of AhR and its target genes, CYP1A1 and CYP1B1, in the colons. NETs formation was evaluated by staining for the co-localisation of MPO, DNA, and NE/citrullinated histone H3 (CitH3). As shown in Fig. 1a–e, 2.5 % DSS induced a significant increase in colitis-related symptoms in mice. The mRNA levels of CYP1A1 and CYP1B1, but not AhR, were concomitantly reduced in the colons of colitis mice, whereas NETs formation was markedly increased. Notably, among neutrophils, bone marrow cells, monocytes, and lymphocytes, neutrophils exhibited the lowest degree of AhR activation in colitis mice (Fig. S2). Pearson correlation analysis further indicated that AhR activation in the colons was inversely correlated with colitis severity and NETs formation (Fig. 1f). These findings suggest that AhR activation is negatively associated with NETosis in UC and that NETosis inhibition may be feasible.

Fig. 1.

The negative relationship between aryl hydrocarbon receptor (AhR) activation and the formation of neutrophil extracellular traps (NETs) in ulcerative colitis (UC). (a-e) The mice were fed with 2.5 % dextran sulfate sodium salt (DSS) to establish colitis model. The disease activity index (DAI) scores (a), colon length (b), histological changes of colons (scale bar = 50 μm) (c), and mRNA levels of AhR, CYP1A1 and CYP1B1 in colons of mice were calculated or detected (d). The colons were stained for NE/CitH3 (red), MPO (green) and DNA (blue) to determine the formation of NETs (scale bar = 50 μm) (e). (f) The correlations between the mRNA levels of CYP1A1 and CYP1B1 with disease indicators and NETs formation were analyzed using pearson correlation. The data were presented as the means ± S.E.M. of ten mice in each group. ^#p < 0.05, ^##p < 0.01 vs. Normal group. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

AhR activation suppresses NETosis

NETosis involves the formation and release of NETs, which consist of a complex network of dsDNA as the structural backbone, integrated with granule proteins, histones, cytoplasmic proteins, and key components such as NE, MPO, and CitH3 [1]. The inhibitory effect of AhR activation on NETosis was assessed using the endogenous and plant-derived AhR agonists FICZ and DIM, along with the specific antagonist CH223191, by evaluating dsDNA release, network formation, and the co-localisation of MPO, DNA, and NE/CitH3. SYTOX Green fluorescence staining demonstrated that lipopolysaccharide (LPS; 10 μg/mL) and phorbol-12-myristate-13-acetate (PMA; 50 nM) significantly enhanced the double-stranded DNA (dsDNA) release in neutrophils, with maximal effects observed at 3.5 h. Concurrently, a distinct network structure indicative of fully formed NETs was observed (Fig. S3). As shown in Fig. 2a-c, FICZ (100 nM) and DIM (10 μM) markedly inhibited LPS (10 μg/mL)- and PMA (50 nM)-induced dsDNA release, whereas CH223191 (10 μM) exhibited a promotive effect. Under the influence of FICZ (100 nM) and DIM (10 μM), the co-localisation of MPO, DNA, and NE/CitH3 was reduced, suggesting that AhR activation suppresses NETs formation (Fig. 2d).

Fig. 2.

The aryl hydrocarbon receptor (AhR) activation directly suppresses the NETosis. (a-f) The neutrophils were treated with FICZ (100 nM), DIM (10 μM) or CH223191 (10 μM) in the present or absence of LPS (10 μg/mL) or PMA (50 nM). The release of dsDNA was stained, and measured using fluorescence microscope (scale bar: 50 μm) (a) and microplate reader (b, c). The formation of NETs were detected by staining for NE/CitH3 (red), MPO (green) and DNA (blue) (scale bar: 20 μm) (d). The levels of histone H4 (e) as well as the nucleus morphology (f) was detected (scale bar = 5 μm). (g) The HL60 cells were pre-transfected with shAhR, incubated with 1.25 % dimethyl sulfoxide, and treated with FICZ (100 nM) or DIM (10 μM) at the present or absence of LPS (10 μg/mL) or PMA (50 nM). The release of dsDNA was stained (scale bar: 50 μm). The data were presented as the means ± S.E.M. of three or five independent experiments. ^##p < 0.01 vs. Control group; ^**p < 0.01 vs. LPS or PMA group. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

To further confirm the inhibitory role of AhR activation, key steps in NETosis, including histone H4 degradation and chromatin decondensation [18,19], were investigated. As shown in Fig. 2e and f, LPS (10 μg/mL), PMA (50 nM), and CH223191 (10 μM) accelerated histone H4 degradation and chromatin decondensation in neutrophils, whereas these effects were blocked by FICZ (100 nM) and DIM (10 μM). Given previous reports indicating that reduced neutrophil activity or apoptosis may shift the mode of cell death to suppress NETosis [20,21], the effects of AhR activation on neutrophil activity, proliferation, and apoptosis were examined. As shown in Fig. S4, FICZ (100 nM) and DIM (10 μM) had minimal effects on these processes, and neither the apoptosis inhibitor Z-VAD-FMK (10 μM) nor the necroptosis inhibitor necrostatin-1 (50 μM) altered the inhibitory effects of FICZ (100 nM) and DIM (10 μM) on dsDNA release.

AhR dependence was further validated using AhR knockdown plasmid (shAhR) in combination with FICZ (100 nM) and DIM (10 μM). Because neutrophils have a short lifespan in vitro and are unsuitable for transfection, dHL60 cells, which exhibit neutrophil-like properties, including NETosis, were used. These cells were characterised by CD11b expression and morphological features (Fig. S5a and b). As shown in Fig. S5c–e and Fig. 2g, FICZ (100 nM) and DIM (10 μM) did not affect the viability, integrity, or apoptosis of dHL60 cells, whereas shAhR abrogated their inhibitory effects on dsDNA release, confirming AhR-dependent suppression of NETosis.

AhR activation downregulates NE activity to abrogate NETosis

During NETosis, NE and pcDNA-peptidyl arginine deiminase 4 (PAD4) are critical upstream molecules that mediate the degradation of histone H4 and CitH3, thereby promoting chromatin decondensation and NETs formation [7,22]. The inhibitory effect of FICZ (100 nM) and DIM (10 μM) on dsDNA release was nearly abolished when co-administered with pcDNA-NE but not pcDNA-PAD4, highlighting the importance of NE in this process (Fig. 3a). Previous studies indicate that NE transcription and translation occur in promyelocytes, the precursor cells of neutrophils, followed by maturation mediated by cathepsin C (CTSC) [[23], [24], [25]]. Neither FICZ (100 nM) nor DIM (10 μM) affected NE mRNA expression at 0, 1, 3, and 5 days during neutrophil differentiation, nor did they influence CTSC protein levels (Fig. 3b–d). However, NE activity (but not its protein levels) was significantly downregulated in neutrophils treated with FICZ (100 nM) and DIM (10 μM), as determined using the hydrolysed NE substrate MeOSU-Ala-Ala-Pro-Val-PNA (Fig. 3e and f).

Fig. 3.

The aryl hydrocarbon receptor (AhR) activation inhibits NETosis with the help of neutrophil elastase (NE). (a) The HL60 cells were pre-transfected with pcDNA-NE or pcDNA-PAD4, differentiated for 5 days by incubation of 1.25 % dimethyl sulfoxide (DMSO), and treated with FICZ (100 nM) or DIM (10 μM) at the present or absence of lipopolysaccharide (LPS, 10 μg/mL) or phorbol-12-myristate-13-acetate (PMA, 50 nM). The release of dsDNA was stained (scale bar: 50 μm). (b-d) The HL60 cells were differentiated for 0, 1, 3 and 5 days, and incubated with 1.25 % DMSO in the presence of FICZ (100 nM), DIM (10 μM) or CH223191 (10 μM). The mRNA level of NE (b) and protein level of CTSC (c, d) was detected. (e-h) The neutrophils were treated with FICZ (100 nM), DIM (10 μM) or CH223191 (10 μM) at the present or absence of LPS (10 μg/mL) or PMA (50 nM), and the protein level (e) and activity (f) of NE was detected. The NE (red) and DNA (blue) was stained for detecting the nuclear localisation of NE (scale bar = 20 μm) (g). The NE (green) and F-actin (red) was stained for detecting the cutting capacity of NE (scale bar: 5 μm) (h). (i) The HL60 cells were pre-transfected with shAhR, incubated with 1.25 % DMSO, and treated with FICZ (100 nM) and DIM (10 μM) at the present or absence of LPS (10 μg/mL) or PMA (50 nM). The NE activity was determined. The data were presented as the means ± S.E.M. of three or five independent experiments. ^##p < 0.01 vs. Control group; ^**p < 0.01 vs. LPS or PMA group; ^$$p < 0.01 vs. LPS/PMA + FICZ group; ⁺⁺p < 0.01 vs. LPS/PMA + DIM group. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Increased NE activity enhances its ability to cleave actin and translocate into the nucleus [26]. The impact of AhR activation on NE activity was therefore examined. As shown in Fig. 3g, stimulation with LPS (10 μg/mL) and PMA (50 nM) increased NE nuclear localisation and fluorescence intensity in neutrophils. Treatment with FICZ (100 nM) and DIM (10 μM) significantly reduced NE nuclear localisation, whereas CH223191 (10 μM), an AhR antagonist, enhanced it. Similarly, FICZ (100 nM) and DIM (10 μM) prevented LPS (10 μg/mL)- and PMA (50 nM)-induced actin depolymerisation in neutrophils (Fig. 3h). When co-administered with shAhR, the downregulatory effect of FICZ (100 nM) and DIM (10 μM) on NE activity was abolished (Fig. 3i). These findings indicate that AhR activation inhibits NETosis by downregulating NE activity.

N-glycosylation of alpha-1 antitrypsin (AAT) and alpha-2-macroglobulin (A2M) mediates the downregulation of NE activity via AhR activation

To determine whether NE activity was directly or indirectly regulated by AhR activation, recombinant NE protein was incubated with FICZ (100 nM) and DIM (10 μM) or with cytoplasmic proteins isolated from neutrophils treated with these compounds. The results revealed that FICZ (100 nM) and DIM (10 μM) did not directly affect NE activity in a cell-free system, whereas cytoplasmic proteins from treated neutrophils exhibited inhibitory effects (Fig. 4a and b). These findings suggest an indirect mechanism, prompting an investigation into the role of AAT, A2M, and secretory leukocyte peptidase inhibito (SLPI)—physiological inhibitors of NE. The Co-immunopreciptation (Co-IP) assays demonstrated the enhanced AAT and A2M binding to NE in neutrophils treated with FICZ (100 nM) and DIM (10 μM), whereas SLPI binding remained unchanged (Fig. 4c). Moreover, pre-transfection with shAAT or shA2M significantly reversed the inhibitory effects of FICZ (100 nM) and DIM (10 μM) on NE activity (Fig. 4d). However, AAT and A2M protein levels were not affected (Fig. 4e).

Fig. 4.

The aryl hydrocarbon receptor (AhR) activation prevents the N-glycosylation of alpha-1 antitrypsin (AAT) and alpha-2-macroglobulin (A2M) to down-regulate the neutrophil elastase (NE) activity. (a) The recombinant protein of NE was incubated with FICZ (100 nM), DIM (10 μM) or CH223191 (10 μM), and NE activity was determined. (b, c) The neutrophils were treated with FICZ (100 nM), DIM (10 μM) or CH223191 (10 μM) at the present or absence of lipopolysaccharide (LPS, 10 μg/mL) or phorbol-12-myristate-13-acetate (PMA, 50 nM). The cytoplasmic proteins in each group were isolated and co-incubated with the recombinant protein of NE, and NE activity was determined (b). The association of AAT, A2M or secretory leukocyte protease inhibitor (SLPI) with NE was detected (c). (d) The HL60 cells were pre-transfected with shAAT or shA2M, incubated with 1.25 % dimethyl sulfoxide, treated with FICZ (100 nM) and DIM (10 μM) at the present or absence of LPS (10 μg/mL) or PMA (50 nM), and NE activity was determined. (e-h) The neutrophils were treated with FICZ (100 nM), DIM (10 μM) or CH223191 (10 μM) at the present or absence of LPS (10 μg/mL) or PMA (50 nM). The levels of AAT, A2M, SLPI (e), ROS (f), N-glycosylation (g) and neutral lipid content (h) were detected. (i, j) The neutrophils were treated with FICZ (100 nM), DIM (10 μM) or CH223191 (10 μM), and tunicamycin (TM; 1 μM) or MK-8719 (21 nM) were jointly given at the present or absence of LPS (10 μg/mL) or PMA (50 nM). The association of AAT or A2M with NE (i) and release of dsDNA (scale bar: 50 μm) (j) were detected. The data were presented as the means ± S.E.M. of three or five independent experiments. ^#p < 0.05, ^##p < 0.01 vs. Control group; *p < 0.05, ^**p < 0.01 vs. LPS group; ^$$p < 0.01 vs. LPS + FICZ group; ⁺⁺p < 0.01 vs. LPS + DIM or CH223191 group.

Intracellular reactive oxygen species (ROS), neutral lipids, and N-glycosylation are key regulators of AAT and A2M, modulating their ability to interfere with NE activity [27,28]. As shown in Fig. 4f–h, FICZ (100 nM) and DIM (10 μM) significantly reduced N-glycosylation in neutrophils, whereas intracellular ROS and neutral lipid levels remained largely unchanged. Further analysis revealed that tunicamycin (TM; 1 μM), an N-glycosylation inhibitor, but not MK-8719 (21 nM), an O-glycosylation inhibitor, significantly suppressed LPS (10 μg/mL)-induced dsDNA release. TM also acted synergistically with FICZ (100 nM) and DIM (10 μM) while abolishing the effects of CH223191 (10 μM) on AAT/A2M binding to NE and dsDNA release (Fig. 4i and j). These findings indicate that AhR activation inhibits the N-glycosylation of AAT and A2M, thereby promoting their association with NE and downregulating NE activity.

Inhibition of glucose metabolism is the main pathway for AhR activation to control N-glycosylation

Magts play a pivotal role in facilitating the transfer of glycan groups to substrate proteins by encoding N-glycosyltransferases. The synthesis rate of uridine 5′-diphospho-N-acetylglucosamine (UDP-GlcNAc), a key substrate for N-glycosylation, is regulated by glutamine-fructose-6-phosphate transaminase (GFPT) and fructose-6-phosphate (F-6-P). F-6-P serves as an intermediate metabolite in glucose metabolism, and GFPT converts F-6-P to UDP-GlcNAc via the hexosamine biosynthesis pathway (HBP) (Fig. 5a) [[29], [30], [31]]. As shown in Fig. 5b–e, FICZ (100 nM), DIM (10 μM), and CH223191 (10 μM) did not affect the expression levels of magnesium transporter protein 1 (Magt1), Magt2, Magt4, and Magt5, nor did they alter the protein levels or activity of GFPT. Therefore, a metabolomic analysis of neutrophils before and after FICZ treatment (100 nM) was performed using NMR spectroscopy. As depicted in Fig. 5f–j, FICZ (100 nM) significantly altered the metabolic profile of LPS (10 μg/mL)-stimulated neutrophils. The levels of glucose-6-phosphate (G-6-P), UDP-glucose (UDP-Glc), UDP-galactose (UDP-Gal), and other glycosylation-related metabolites were downregulated, while those of lactic acid and pyruvate were significantly decreased. In conjunction with the results of differential metabolite enrichment analysis, these findings suggest that AhR activation substantially disrupts glucose metabolism-related pathways, including glycolysis and HBP.

Fig. 5.

Glucose metabolism is the main pathway for aryl hydrocarbon receptor (AhR) activation to down-regulate the neutrophil elastase (NE) activity. (a) The diagram of glucose metabolism. (b-e) The neutrophils were treated with FICZ (100 nM), DIM (10 μM) or CH223191 (10 μM) at the present or absence of lipopolysaccharide (LPS, 10 μg/mL) or phorbol-12-myristate-13-acetate (PMA, 50 nM). The levels of Magt1, Magt 2, Magt 4, Magt5 and glutamine-fructose-6-phosphate transaminase (GFPT) were detected. (f-j) The neutrophils were treated with FICZ (100 nM) at the present or absence of LPS (10 μg/mL), and the intracellular metabolites were detected using a high-resolution nuclear magnetic resonance (NMR). The results were analyzed by PCA (f), PLS (g), OPLS (h), KEGG enrichment analysis (i) and cluster analysis (j). (k, l) The neutrophils were treated with FICZ (100 nM), DIM (10 μM) or CH223191 (10 μM) at the present or absence of LPS (10 μg/mL) or PMA (50 nM). The relative glucose uptake (k) and intracellular pH values (l) were determined. (m, n) The neutrophils were treated with FICZ (100 nM) or DIM (10 μM) alone or in combination with L-sodium lactate (20 mM), ATP (100 μM) or GlcNAc (20 mM) at the present or absence of LPS (10 μg/mL) or PMA (50 nM), and the release of dsDNA (m) and NE activity (n) was determined. The data were presented as the means ± S.E.M. of three or five independent experiments. ^##p < 0.01 vs. Control group; *p < 0.05, ^**p < 0.01 vs. LPS group; ^$$p < 0.01 vs. LPS + FICZ group; ⁺⁺p < 0.01 vs. LPS + DIM group.

To verify these findings, glucose uptake and intracellular pH were measured using flow cytometry, revealing significant downregulation in response to FICZ (100 nM) and DIM (10 μM) (Fig. 5k and l). However, as shown in Fig. 5m, n, N-acetylglucosamine (GlcNAc; 20 nM), but not lactic acid (20 mM) or adenosine triphosphate (ATP; 100 μM), reversed the inhibition of dsDNA release induced by FICZ (100 nM) and DIM (10 μM). Furthermore, GlcNAc (20 nM) also prevented FICZ (100 nM) and DIM (10 μM) from reducing NE activity. These results indicate that AhR activation inhibits HBP flux, thereby downregulating N-glycosylation and NE activity by reducing UDP-GlcNAc levels.

AhR functions as an E3 ligase to promote HK2 ubiquitination and restrict HBP flux

In glucose metabolism, the levels of F-6-P are primarily regulated by the rate-limiting enzymes HK2 and phosphofructokinase (PFK), which are positioned upstream and downstream, respectively. These enzymes ensure the stability of glycolytic flux, and either a reduction in HK2 activity/expression or an increase in PFK function leads to elevated F-6-P levels [30,31]. As shown in Fig. 6a, LPS (10 μg/mL) significantly upregulated HK2 and PFK activity in neutrophils, whereas FICZ (100 nM), DIM (10 μM), and CH223191 (10 μM) had no significant effect. However, LPS (10 μg/mL) and CH223191 (10 μM) dramatically increased HK2 protein levels, an effect that was abolished by FICZ (100 nM) and DIM (10 μM) (Fig. 6b and c). Transfection with pcDNA-HK2 reversed the suppressive effects of FICZ (100 nM) and DIM (10 μM) on NE activity and UDP-GlcNAc levels, underscoring the crucial role of HK2 (Fig. 6d and e). In the presence of the protein synthesis inhibitor cycloheximide (15 μg/mL), FICZ (100 nM) and DIM (10 μM) accelerated LPS (10 μg/mL)-blocked HK2 protein degradation but did not affect its transcription (Fig. 6f, g). Further experiments demonstrated that only the proteasome inhibitor MG132 (25 μg/mL) reversed the downregulation of HK2 protein levels induced by FICZ (100 nM) and DIM (10 μM), and both FICZ (100 nM) and DIM (10 μM) really enhanced ubiquitin-mediated degradation of HK2 (Fig. 6h, i). Moreover, increased co-localisation of HK2 with the proteasomal 26S subunit (PSMD2), an ATPase component of the proteasome [18], corroborated these findings (Fig. 6j).

Fig. 6.

The aryl hydrocarbon receptor (AhR) acts as an E3 ligase to promote the ubiquitination of hexokinase-2 (HK2) and restrain glucose metabolism. (a-c) The neutrophils were treated with FICZ (100 nM), DIM (10 μM) or CH223191 (10 μM) at the present or absence of lipopolysaccharide (LPS, 10 μg/mL) or phorbol-12-myristate-13-acetate (PMA, 50 nM). The relative activity (a) as well as protein levels (b, c) of hexokinase 2 (HK2) and phosphofructokinase (PFK) were detected. (d-f) The HL60 cells were pre-transfected with pcDNA-HK2, incubated with 1.25 % dimethyl sulfoxide (DMSO), and treated with FICZ (100 nM) and DIM (10 μM) at the present or absence of LPS (10 μg/mL) or PMA (50 nM). The activity of NE (d) and levels of UDP-GlcNAc (e), HK2, PFK (f) were detected. (g) The neutrophils were treated with FICZ (100 nM) or DIM (10 μM) before receiving LPS (10 μg/mL) and CHX (15 μg/mL) for 0, 0.5, 1, 1.5 h, and the protein level of HK2 was detected. (h, i) The neutrophils were treated with FICZ (100 nM), DIM (10 μM) or CH223191 (10 μM), and MG132 (25 μg/mL) or hydroxychloroquine (HCQ; 10 μM) were jointly given at the present or absence of LPS (10 μg/mL) or PMA (50 nM). The protein level (h) and ubiquitination (i) of HK2 was detected. (j, k) The neutrophils were treated with FICZ (100 nM), DIM (10 μM) or CH223191 (10 μM) at the present or absence of LPS (10 μg/mL). The co-localization of HK2 and PSMD2 was stained (scale bar: 5 μm) (j). The mRNA levels of CYP1A1 and CYP1B1 was detected (k). (l) The HL60 cells were pre-transfected with shARNT, incubated with 1.25 % DMSO, treated with FICZ (100 nM) or DIM (10 μM) at the present or absence of LPS (10 μg/mL), and release of dsDNA was stained (scale bar = 50 μm). (m, n) The association of AhR and HK2 was simulated using molecular docking and the association of AhR and HK2 was detected. (o) The neutrophils were treated with FICZ (100 nM), DIM (10 μM) or CH223191 (10 μM) at the present or absence of LPS (10 μg/mL), and the association of ubiquitination at Lys 48 and Lys 63 of HK2 protein was detected. The data were presented as the means ± S.E.M. of three or five independent experiments. ^#p < 0.05, ^##p < 0.01 vs. Control group; *p < 0.05, ^**p < 0.01 vs. LPS group; ^$p < 0.05, ^$$p < 0.01 vs. LPS + FICZ group; ⁺p < 0.01, ⁺⁺p < 0.01 vs. LPS + DIM group.

AhR, as a ligand-dependent receptor, regulates downstream target gene transcription (e.g., CYP1A1 and CYP1B1) in cooperation with the ARNT and also facilitates ubiquitination-mediated protein degradation—referred to as genomic and non-genomic actions, respectively [32]. As shown in Fig. 6k, FICZ (100 nM), DIM (10 μM), and CH223191 (10 μM) upregulated and downregulated CYP1A1 and CYP1B1 levels, thereby activating and antagonising AhR, respectively. Unlike shAhR, ARNT knockdown (shARNT) failed to reverse the inhibitory effects of FICZ (100 nM) and DIM (10 μM) on dsDNA release (Fig. 6l), highlighting the significance of the non-genomic mechanism. Computational docking analysis confirmed the stable interaction between AhR and HK2 (Fig. 6m). Furthermore, FICZ (100 nM) and DIM (10 μM) promoted HK2–AhR binding and enhanced ubiquitination at lysine 48 (Lys 48) but not lysine 63 (Lys 63) of HK2 (Fig. 6n, o). These results suggest that AhR activation facilitates HK2 ubiquitination and degradation, thereby suppressing HBP flux.

In vivo validation of the inhibitory effect and mechanism of AhR activation on NETosis

Because of the close association between NETs and sepsis, a model of LPS-induced sepsis was established in mice to verify the inhibitory effect and mechanism of AhR activation on NETosis. Mice with sepsis exhibited severe lung injury, but pathological severity was significantly alleviated by FICZ (1 μg/mouse) and DIM (10 mg/kg) (Fig. 7a). Assessment of dsDNA release and the co-localisation of MPO, NE, and DNA revealed that AhR activation substantially reduced neutrophil capacity to generate NETs (Fig. 7b, c). Additionally, AhR activation significantly reduced NE activity and decreased UDP-GlcNAc and HK2 protein levels in neutrophils (Fig. 7d–f).

Fig. 7.

The in vivo validation of the inhibitory effect and mechanism of aryl hydrocarbon receptor (AhR) activation on NETosis. (a-f) The mice received intraperitoneal injections of lipopolysaccharide (LPS; 10 mg/kg) to establish sepsis model. The histological changes of lungs (scale bar = 50 μm) (a) and dsDNA release of neutrophils originated from bone marrow, spleen and peripheral blood (scale bar = 50 μm) (b) were detected. The NETs formation (scale bar: 20 μm) (c), NE activity (d), UDP-GlcNAc level (e) and HK2 level (f) were determined. (g-n) The mice were fed with 2.5 % dextran sulfate sodium salt (DSS) to establish colitis model. The disease activity index (DAI) scores (g), colon length (h), histological changes of colons (scale bar = 50 μm) (i), myeloperoxidase (MPO) activity in colons (j) and intestinal permeability (k) was measured. (l) The NE activity in colons were assessed. (m) The colons were stained for NE (red), MPO (green) and DNA (blue) (scale bar = 50 μm). (n) The protein level of HK2 in colons were assessed. The data were presented as the means ± S.E.M. of six mice in each group. ^##P < 0.01 vs. Normal group; *P < 0.05, ^**P < 0.01 vs. Model group; ^$p < 0.05, ^$$p < 0.01 vs. FICZ group; ⁺P < 0.05, ⁺⁺P < 0.01 vs. DIM group. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

To confirm the inhibitory effect and mechanism of AhR activation on NETosis in the UC state, a combination of pcDNA-NE or pcDNA-HK2 with FICZ and DIM was used. As shown in Fig. 7g, FICZ (1 μg/mouse), DIM (10 mg/kg), and the NE-active inhibitor sivelestat (50 mg/kg) led to a marked reduction in DAI scores from day 5 to day 10. In addition, excessive MPO activity, histopathological changes, intestinal permeability, and colon shortening were alleviated (Fig. 7h–k). However, the pcDNA-NE or pcDNA-HK2 (10 μg/mouse) reversed the effects of FICZ (1 μg/mouse) and DIM (10 mg/kg). Furthermore, NE and MPO levels in the colons of colitis mice were significantly increased, with evident co-localisation, suggesting NETs formation. As shown in Fig. 7l and m, FICZ (1 μg/mouse), DIM (10 mg/kg), and sivelestat (50 mg/kg) downregulated NE activity and decreased NETs formation in colons, which was significantly weakened by pcDNA-HK2 and/or pcDNA-NE (10 μg/mouse). In addition, the HK2 protein levels in the colons were also significantly reduced by FICZ (1 μg/mouse) and DIM (10 mg/kg) (Fig. 7n). Therefore, the in vitro inhibitory effect and mechanism of AhR activation on NETosis were validated.

Discussion

AhR has gained increasing attention as a promising drug target with therapeutic potential against UC. Notably, various intestinal microorganisms and their metabolites can activate AhR to mitigate epithelial barrier damage. Furthermore, deletion of ARNT, the primary molecular chaperone of AhR, has been reported to increase NETs levels in the colonic tissues of colorectal cancer mice [14,16]. In the present study, GSEA results and findings from DSS-induced colitis mice revealed a negative correlation between AhR activation and NETs formation. In vitro experiments further confirmed that AhR activation prevents NETs formation and release, as demonstrated by reduced dsDNA release and diminished network formation and co-localisation of MPO, NE, and DNA. Consistent with these findings, degradation of histone H4 and subsequent chromatin decondensation were also downregulated, as expected. Moreover, given the sensitivity of neutrophils to stimulation, viability, injury, and apoptosis were excluded as potential reasons for the AhR activation-mediated inhibition of NETosis. These findings indicate that AhR activation directly inhibits NETosis.

NE and PAD4 are two critical upstream regulators of histone H4 and CitH3 degradation in NETosis, promoting chromatin decondensation and NETs formation [22,33]. In the present study, pcDNA-NE, but not pcDNA-PAD4, counteracted the inhibitory effect of AhR activation on dsDNA release, highlighting the importance of NE in this process. As previously reported, neutrophils differentiate from promyelocytes, where the transcription and translation of NE are completed. NE is then stored in precursor azurophilic granules (primary granules) as an inactive zymogen [23]. CTSC subsequently removes the N-terminal dipeptides of pro-NE, enabling its maturation and activation. Loss of CTSC function results in a lack of NE activity [24,25]. Under AhR activation, the transcription and protein levels of NE and CTSC remained unchanged; however, NE activity was significantly reduced. Concurrently, reduced nuclear translocation and actin cytoskeleton depolymerisation were observed following AhR activation-mediated NE inhibition. Unlike NE localisation in the cytoplasm of neutrophils, FICZ and DIM did not directly downregulate the activity of recombinant NE proteins in cell-free conditions. These findings suggest that AhR activation indirectly downregulates NE activity by modifying the cytoplasmic environment. It is worth noting that sivelestat, the NE activity inhibitor, can directly down-regulate the NE activity in NETs that have been secreted and released by neutrophils, or enter into cells to prevent the formation of NETs, and has been approved for listing and sale.

Under physiological conditions, NE activity is tightly regulated by intracellular natural substrates. Among these, SLPI, AAT, and A2M exert opposing regulatory effects on NE [34]. In the present study, knockdown of AAT or A2M, but not SLPI, abolished the inhibitory effect of AhR activation on NETs release. This was accompanied by a significant increase in NE binding to AAT and A2M. Further investigation revealed that AhR activation downregulated NE activity by inhibiting the N-glycosylation modification of AAT and A2M, without altering intracellular ROS or neutral lipid levels. Consistent with these findings, N-glycosylation has been shown to alter the folding conformations of AAT and A2M, reducing their NE-binding affinity and thereby enhancing NE activity [35,36]. Consequently, inhibition of N-glycosylation in AAT and A2M has become an important topic for future research. The primary pathways impairing N-glycosylation modification involve either the deletion of N-glycosyltransferases encoded by Magt1, Magt2, Magt4, and Magt5, which reduces glycan transfer to substrate proteins, or the inhibition of UDP-GlcNAc synthesis, which serves as the donor substrate for N-glycosylation. The Q-PCR analysis ruled out the involvement of Magt genes, suggesting a role for glucose metabolism in this process.

Evidence indicates that glucose metabolism is significantly upregulated during NETosis, and this process can be suppressed by the glycolytic inhibitor 2-deoxy-d-glucose, which reduces metabolic flux, or oligomycin, which inhibits intracellular ATP synthesis [37,38]. Interestingly, AhR regulates glucose metabolism in various cell types, including macrophages and T cells [39]. Therefore, metabolomic analysis was employed to assess the impact of AhR activation on energy metabolism in LPS-activated neutrophils. Comparisons with other metabolic pathways confirmed the involvement of glucose metabolism. The expression and activity levels of GFPT, HK2, and PFK were assessed, revealing that AhR activation reduced HK2 protein levels but not transcription levels. This inhibition of HK2 blocked the HPB flux, ultimately downregulating N-glycosylation modification, reducing NE activity, and preventing NETosis.

The mechanism underlying AhR activation-mediated HK2 suppression was then investigated. HK2 undergoes ubiquitination at the K63 site, leading to lysosomal degradation via the autophagy receptor sequestosome 2. In contrast, ubiquitination at the K48 site targets HK2 for proteasomal degradation [40]. The present findings showed that AhR activation promoted K48-linked ubiquitination of HK2, enhancing its co-localisation with the proteasome and accelerating proteasome-mediated degradation. Intriguingly, AhR functions as a ligand-dependent nuclear receptor, forming a heterodimer with ARNT in the nucleus to bind to dioxin response elements or xenobiotic response elements, thereby promoting transcription of target genes. Additionally, AhR exerts non-genomic effects by acting as an E3 ubiquitin ligase, facilitating substrate protein ubiquitination [14,41,42]. In this study, shAhR, but not shARNT, reversed the inhibitory effect of AhR activation on dsDNA release, confirming its role in non-genomic regulation. Moreover, molecular docking and co-immunoprecipitation experiments demonstrated that AhR interacts with HK2, promoting its K48-linked ubiquitination but not K63-linked ubiquitination. Finally, in vivo experiments using sepsis and colitis models fully validated the function and mechanism of AhR activation in regulating NETosis.

Conclusion

AhR directly associates with HK2 to promote its ubiquitination by the proteasome, impairing the flux of the HBP and subsequent N-glycosylation modification of AAT and A2M. This downregulation of NE activity prevents NETosis and mitigates UC. The above-mentioned results indicate that AhR has a good inhibitory effect on NETosis, and acts as a highly potential therapeutic target for UC. Meanwhile, a variety of natural AhR agonists with relatively low toxicity and side effects are discovered. For example, the endogenous AhR agonist FICZ and the naturally sourced DIM, used in this study, own good prospects for development and application. In 2019, benvitimod is discovered to be able to treat psoriasis by stimulating AhR and approved for marketing. Furthermore, the Ⅱa phase of clinical trial for colitis and psoriasis of oral AhR regulator NTI-528 has been completed. However, when AhR is excited, it may acquire potential pro-tumor effect, which is related to its immunosuppressive efficacy. To avoid this risk, tissue-targeting strategies can be employed for local intestinal delivery of AhR agonists to limit their systemic exposure. Apart from NE, it remains unclear whether AhR can also exert an inhibitory effect on NETosis by regulating other proteases such as Proteinase 3, etc, and further exploration should be continued. In addition, genetically modified animals provide better data for the exploration and verification of molecular mechanisms, and is suggested to be used in the future.

Compliance with ethics requirements

All Institutional and National Guidelines for the care and use of animals were followed.

All procedures followed were in accordance with the ethical standards of the responsible committee on human experimentation (institutional and national) and with the Helsinki Declaration of 1975, as revised in 2008 (5).

Funding

This work was supported by the Qing Lan Project of Jiangsu Province (2019), the “Double First-Class” University Project (CPU2022QZ31), and partially supported by the Priority Academic Program Development of Jiangsu Higher Education Institutions.

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: