TNFSF11/TNFRSF11A Axis Amplifies HDM-Induced Airway Remodeling by Strengthening TGFβ1/STAT3 Action

Reagents

HDM was obtained from Greer Laboratories (Lenoir, NC, USA). Denosumab was purchased from Selleck (Shanghai, China), and 4,6-diamino-2-phenylindole dihydrochloride (DAPI, 28718-90-3) was acquired from Sigma-Aldrich (St. Louis, MO, USA). Anti-vimentin (ET1610-39), anti-fibronectin (ET1702-25), anti-N-cadherin (ET1607-37), anti-EPCAM (EM1111), anti-glyceraldehyde 3-phosphate dehydrogenase (GAPDH) (ET1601-4), anti-TGFβ1 (ER31210), anti-TGFβR1 (ER1917-65), anti-interleukin (IL)-4 (ER1706-55), anti-IL-5 (HA500081), horseradish peroxidase (HRP)-conjugated goat anti-mouse immunoglobulin (Ig)G (HA1006), and HRP-conjugated goat anti-rabbit IgG (HA1001) were obtained from HUABIO (Zhejiang, China). Anti-collagen I (ab260043), anti-STAT3 (ab68153), anti-p-STAT3 (ab76315), anti-p-STAT5 (ab32364), anti-p-STAT6 (ab32520), and anti-p-JAK2 (ab32101) were purchased from Abcam (Cambridge, UK). Recombinant human TNFSF11 protein (rhTNFSF11, AF-310-01) was obtained from Thermo Fisher Scientific (Waltham, MA, USA). Recombinant mouse TNFSF11 protein (rmTNFSF11, 462-TEC) was provided from RD-Biotech (Franche-Comte, France). Bicinchoninic acid (BCA) protein assay (PC0020), hematoxylin–eosin (HE) staining (G1120), periodic acid–Schiff (PAS) staining (G1280), Masson staining (G1346), enzyme-linked immunosorbent assay (ELISA) IL-4 (SEKM-0005), ELISA IL-33 (SEKM-0028), and broad-spectrum SP (SP0041) kits were bought from Solarbio (Beijing, China). Anti-TNFSF11 (sc-52950), anti-eosinophil major basic protein (EMBP, sc-365701), and anti-TNFRSF11A (sc-374360) were obtained from Santa Cruz Biotechnology (Dallas, TX, US). Lentiviruses of human TNFSF11 shRNA (sc-29464-V), human TNFRSF11A shRNA (sc-42960-V), human TGFβ1 shRNA (sc-270322-V), and human TGFβR1 shRNA (sc-40222-V) were purchased from Santa Cruz Biotechnology. Dylight 488 goat anti-mouse IgG (A23210), Dylight 488 goat anti-rabbit IgG (A23220), and Dylight 549 goat anti-rabbit IgG (A23320) were sourced from Abbkine (Wuhan, China). CY5 goat anti-mouse IgG (GB27301) was obtained from Servicebio (Wuhan, China). ELISA kits for human TNFSF11 (EK0842) and mouse TNFSF11 (EK0843) were provided by Boster (Wuhan, China).

Human sample collection

The criteria for asthma diagnosis were based on the 2021 Global Strategy for Asthma Management and Prevention.13 Serum samples were collected from the respiratory outpatient department and physical examination center of the First Affiliated Hospital of Shandong First Medical University. Lung specimens were collected from patients with asthma and patients with suspected tumors who were finally diagnosed as normal using the respiratory endoscope of the First Affiliated Hospital of Shandong First Medical University. The clinical characteristics of the patients are shown in Supplementary Table S1. This study was approved by the Human Ethics Review Committee (S132) of the First Affiliated Hospital of Shandong First Medical University.

Inflammation-related biomarkers

The Olink inflammatory biomarker panel was used to analyze inflammatory markers in serum from patients with asthma (n = 11) and normal patients (n = 10).14 The panel requires only 1 μL of serum to simultaneously analyze 92 different inflammatory factors. Those 92 different target proteins were bound by a pair of oligonucleotide-labeled antibody probes, which were close to one another, and the oligonucleotides hybridized in pairs. This process resulted in unique polymerase chain reaction (PCR) target sequences under DNA polymerase action. The target sequences were obtained using microfluidic real-time PCR. Data were expressed as normalized protein expression (NPX) after quality control and log2 normalization. The median value corresponded to protein expression.

Bioinformatics analysis

t-test was performed between two groups, and P < 0.05 indicated that a screened protein meets the threshold for Olink data and is a differentially expressed protein (DEP). A differential protein heatmap was used for clustering analysis based on the similarity of protein expression profiles in serum. A volcano plot was employed to visualize DEP distribution. The DEPs were then subjected to Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) functional analyses. Finally, the STRING protein-protein interaction (PPI) was employed for interaction analysis on the identified proteins. Bioinformatics analysis was completed by LC-Bio Technologies Co., Ltd. (Hangzhou, China).

Animal grouping and treatment

The animal study was approved by the Animal Ethics Review Committee (S131) of the First Affiliated Hospital of Shandong First Medical University. The animal operation conformed to the Guidelines for the Care and Use of Experimental Animals of the National Institute of Health.

Female C57BL/6 mice were purchased from Jinan Pengyue Animal Feeding Co., Ltd. (Jinan, China). The mice (n = 6, 7–8 weeks, 18–20 g) were stimulated by HDM or rmTNFSF11.

Tnfsf11 is located on chromosome 14, and Tnfrsf11a is located on chromosome 1 in mice. An sgRNA was designed using CRISPR/Cas9 technology for the high-throughput electrotransfer of fertilized eggs to obtain mice with Tnfsf11 and Tnfrsf11a genes. However, Tnfsf11^−/− or Tnfrsf11a ^−/− caused the death of mice. Thus, Tnfsf11^+/− and Tnfrsf11a^+/− mice were sourced from Saiye Biology for experimental studies. The gene identification report of Tnfsf11^+/− and Tnfrsf11a^+/− mice can be found in Supplementary Fig. S1.

NOG-EXL mice are transgenic mouse models constructed by transferring human cytokine granulocyte-macrophage colony-stimulating factor (GM-CSF) and IL-3 on the basis of NOG. NOG-EXL mice transplanted with human cord blood-derived hematopoietic stem cells (HuCD34+huHSC) nearly emulate the human immune system including mature granulocytes, dendritic cells, mast cells, monocytes, macrophages, B cells, and T cells.15 Humanized HSC-NOG-EXL mice are the most complete animal model of human immune system reconstruction to date. Therefore, humanized HSC-NOG-EXL mice effectively simulate myeloid cells that are mainly involved in mediating inflammatory reactions, such as allergies and asthma. A humanized mouse model was constructed to study asthmatic airway remodeling and Muc-5ac secretion via human IL-33.15 Humanized HSC-NOG-EXL mice were exposed to HDM to simulate the pathogenesis of human asthma, and denosumab (full human monoclonal antibody, anti-TNFSF11) was given to further explore the role of TNFSF11.16

Establishment of the air-liquid interface (ALI) of primary HBE

This study was approved by the Human Ethics Review Committee (S132) of the First Affiliated Hospital of Shandong First Medical University and conducted after the participants signed informed consent forms. All the participants underwent electronic bronchoscopy, and grades 2–3 bronchial epithelia were collected via lightly brushing with a sterile brush. The obtained HBE cells were placed in a T-25 cm² cell culture flask coated with collagen. An appropriate amount of PneumaCult™-Ex medium (STEMCELL Technologies, Vancouver, Canada) was added for culture in the liquid phase, and the changes in cell status were observed. The HBE cells grew outward in clusters on the 4th–6th days and reached 80% confluence on the 10th–12th days. The HBE cells in liquid phase were digested using 0.25% trypsin/ethylenediaminetetraacetic acid (EDTA) and transferred to the air–liquid phase chamber for ALI culture. PneumaCult™-Ex medium was added to the upper and lower Transwell chambers (12 mm and 0.4 μm pore polyester membrane inserts; Corning Inc., Corning, NY, USA). The media in the upper and lower chambers were removed and washed with phosphate-buffered saline (PBS), PneumaCult™-ALI medium (STEMCELL Technologies) was added to the lower chamber, and the HBE cells were cultured in the ALI for more than 20 days. The multilayer cells of the ALI were observed by HE staining under an inverted light lens.17

HBE cells in exponential growth state were digested with 0.25% trypsin/EDTA, and a cell suspension was prepared with PneumaCult™-Ex complete culture medium. The cell suspension was inoculated in 6-well plates at an appropriate density, and lentiviral transfection was performed when the cell growth and fusion reached approximately 50%–70%. The suspension was divided into lentivirus-TNFSF11 or TNFRSF11A or TGFβ1 or TGFβR1 or negative control shRNA groups. Following the lentiviral transfection procedure, a lentiviral vector was mixed with PneumaCult™-EX culture medium and added to a cell flask with HBE cells for transfection. Changes in HBE cells were observed after 12–15 hours of transfection, and the culture medium containing the lentivirus was replaced with PneumaCult™-EX complete culture medium. Puromycin was used to remove the HBE cells that were not transfected with the lentivirus. The cell medium was then changed to PneumaCult after the cells were washed with PBS for the preparation of HBE with low expression. After the cells with the ALI grew to more than 80% of fusion, sterile HDM (50 μg/mL) or rhTNFSF11 protein (50, 100, and 200 ng/mL) was added for 24 hours of stimulation.18, 19 The HBE cells were pretreated with stattic (10 μM/L) for 2 h and then stimulated with rhTNFSF11 (50 ng/mL) for 24 hours.20 The upper chamber cells were collected for protein analysis. The longitudinal section of the upper chamber cells was collected and fixed with 4% paraformaldehyde, and differentiated HBE sections were prepared by optimum cutting temperature (OCT) embedding for immunofluorescence detection.

Flow cytometry

After the HSC-NOG-EXL mice were anesthetized with isoflurane, their peripheral blood (100 μL) was collected and added to a flow tube. The blood was added with 1 μL of FITC anti-human CD45 (#304005; BioLegend, San Diego, CA, USA) and APC anti-mouse CD45 (#147707; BioLegend). Red blood cells were removed with ammonium–chloride–potassium lysis solution. The mixtures were then centrifuged at 500 ×g for 5 minutes, then the supernatant was discarded. Finally, the cells were suspended with 300 μL of PBS and placed on a FACS LSRII flow cytometer (BD Biosciences, San Jose, CA, USA). Data were analyzed using FlowJo software (Tree Star, Ashland, OR, USA).

Cytokine detection

The mouse peripheral blood was centrifuged at 4°C and 1,000 ×g for 15 minutes, and the supernatant was packaged and stored. The serum TNFSF11 and IgE levels were detected in accordance with the operating procedure of the kit.

The mice were given anesthesia using isoflurane. Their chests were opened, the left main bronchus was ligated and 0.5 mL of sterile PBS was injected into the trachea to collect unilateral bronchoalveolar lavage fluid (BALF). All BALF samples were then centrifuged at 1,000 ×g and 4°C for 20 minutes. The TNFSF11, IL-4, and IL-33 levels in the BALF supernatant were then measured using ELISA kits.

mRNA sequencing

Total RNA of lung tissues was isolated using TRIzol reagent. The samples were enriched for mRNA with polyA structures by using Oligo (dT) magnetic beads. The mRNA was fragmented by divalent cations under elevated temperature in an Illumina proprietary fragmentation buffer. First-strand cDNA was first synthesized using random oligonucleotides and SuperScript II, followed by second-strand cDNA. After the adenylation of the 3′ ends of the DNA fragments, Illumina PE adapter oligonucleotides were ligated to prepare for hybridization. cDNA fragments with a length of 400–500 bp were selected and then purified. DNA fragments with ligated adaptor molecules on both ends were amplified using Illumina PCR primer cocktail in a 15-cycle PCR reaction. Sequencing library construction and RNA-sequencing were carried out by Shanghai Bioprofile Technology Company, Ltd. (Shanghai, China).

Immunohistochemistry

The mouse lung tissues were fixed in 10% neutral formalin, paraffin-embedded, and cut into 5 μm-thick sections. The paraffin sections were antigenically repaired, and endogenous peroxidase was removed and then blocked with 5% bovine serum albumin for 1 hour at 37°C. The slides were incubated with anti-N-cadherin overnight at 4°C and then processed with the corresponding secondary horseradish peroxidase-conjugated antibodies for 30 minutes at room temperature. The sections were stained with 3,3′-diaminobenzidine and hematoxylin dehydrated with a gradient alcohol series, cleared in xylene, and sealed with neutral gum.

Western blot

The cells or lung tissues were fully lysed with radioimmunoprecipitation assay and phenylmethylsulfonyl fluoride lysate. A BCA kit was used to detect protein concentrations in the samples. The proteins were separated through 10% sodium dodecyl sulfate polyacrylamide gel electrophoresis and transferred to polyvinylidene fluoride (PVDF) membranes at low temperatures. The nonspecific binding sites of the PVDF membranes were blocked with 5% bovine milk powder for 1 hour. The PVDF membranes were washed with Tris-buffered saline Tween 20 (TBST) and then incubated with primary antibodies (anti-TNFSF11, anti-TNFRSF11A, anti-N-cadherin, anti-fibronectin, anti-vimentin, anti-collagen I, anti-TGFβ1, anti-p-STAT3, anti-STAT3, anti-p-STAT5, anti-p-STAT6, anti-p-JAK2, IL-4, IL-5, anti-EMBP or anti-GAPDH) at 4°C overnight. On the 2nd day, the PVDF membrane was washed with TBST and incubated with goat anti-rabbit or goat anti-mouse antibody at room temperature for 1 hour. Finally, the PVDF membranes were washed with TBST and exposed with a protein strip. Their proteins were performed with Image J software.

HE staining of HBE in the ALI

The differentiated HBE cells were washed with PBS and fixed with 4% paraformaldehyde for 30 minutes. The cells were cut along the longitudinal section with a sharp blade, embedded in an embedding box with OCT embedding agent, and frozen in a refrigerator at −20°C. The frozen samples were cut into 5 μm-thick pieces with a freezing microtome, and the morphological changes of the cells were observed through HE staining.

Immunofluorescence staining of HBE in the ALI

The 5 μm-thick sections were incubated with goat serum working solution at room temperature for 30 minutes after the cell sections were washed with PBS. A filter paper used to suck the goat serum working fluid, and an antibody (anti-fibronectin, anti-vimentin, anti-collagen I, anti-p-STAT3, anti-TGFβR1) was added to the cell sections, which was then incubated at 4°C overnight. On the 2nd day, the HBE cells were washed with PBS and incubated with a fluorescent secondary antibody. Finally, the sections were stained with DAPI, incubated for 7 minutes, and observed and photographed under a fluorescence microscope.

Lung histology and immunofluorescence

The lungs were embedded in an embedding box with paraffin wax and cut into 5 μm-thick sections. HE and Masson staining were performed as described previously.4 The lung sections were dewaxed with xylene, dehydrated with gradient concentrations of alcohol, immersed in 0.01 M sodium citrate buffer, and heated to boiling for the heat repair antigen. Goat serum was incubated in the lungs for 40 minutes after liquid cooling. The excess serum working fluid was removed, and the lung tissues were incubated at 4°C for 24 hours with antibodies (anti-TNFSF11, anti-EPCAM, and anti-p-STAT3, anti-TGFβ1). On the 2nd day, the lung tissues were washed with PBS and incubated with fluorescent secondary antibody and DAPI for 30 minutes. Finally, the lung tissues were observed and photographed under a fluorescence microscope.

Statistical analysis

GraphPad Prism 8.4 and SPSS17.0 were used for statistical tests, and the results were expressed as means ± standard deviation of at least 3 biological duplications. P value < 0.05 indicated statistically significant difference. The t-test (normally distributed parameters) or Mann–Whitney U test (non-normally distributed parameters) was used to evaluate statistical significance for the two groups. The normally distributed parameters of one-way analysis of variance (ANOVA) and Welch ANOVA and non-normally distributed parameters of Kruskal–Wallis test were used in evaluating statistical significance for multiple comparisons. Correlations were determined by Spearman analysis.

Elevated TNFSF11 expression in asthmatic serum as determined by Olink inflammation-related biomarker analysis

Olink inflammatory biomarker was used to evaluate the expression levels of inflammatory-associated proteins (n = 92) in the asthmatic and normal groups. DEPs were identified between the asthmatic and normal groups; among which, 11 proteins in the asthma group were up-regulated, and TNFSF11expression significantly increased (Fig. 1A). Interprotein correlation analysis was performed to explore interactions between different proteins. TNFSF11, IL-6, and TNF were highly rated in the PPI network (Fig. 1B). IL-6 and TNF play an important role in asthma progression.4, 21 GO and KEGG enrichment analyses were performed on the background of DEPs. The results showed that the functions of TNFSF11 were mainly concentrated in extracellular space and region, immune response, and signaling molecule interaction (Fig. 1C and D). Therefore, TNFSF11 plays an important role in asthma.

Fig. 1
DEPs among inflammation-related biomarkers between the asthmatic (n = 11) and normal (n = 10) groups. (A) Volcano plot of DEPs, including TNFSF11. (B) PPI network analysis of TNFSF11. (C) GO enrichment analysis with scatter plot. (D) KEGG enrichment analysis with scatter plot.
DEP, differentially expressed protein; TNFSF11, TNF ligand superfamily member 11; PPI, protein-protein interaction; GO, Gene Ontology; KEGG, Kyoto Encyclopedia of Genes and Genomes.

• Click for larger image
• Download as PowerPoint slide

Elevated TNFSF11 expression in bronchial epithelium and serum from asthmatic subjects

Endoscopic specimens were collected from healthy individuals and patients with asthma. The asthmatic subjects had collagen deposition around the airway compared with the normal group (Fig. 2A and C). Immunohistochemical results indicated that TNFSF11 protein levels significantly increased in the bronchial epithelia of the patients with asthma relative to those in healthy individuals (Fig. 2B and D). Subsequently, this study also analyzed TNFSF11 in serum from asthmatic and normal subjects using ELISA. This study verified that serum TNFSF11 level in asthmatic subjects was significantly higher after the number of samples was increased (Fig. 2E). The correlation of eosinophil cells (Eos) and lung function with TNFSF11 was then analyzed. The TNFSF11 level in serum was positively correlated with Eos (Fig. 2F) and negatively correlated with forced vital capacity (FVC), forced expiratory volume in 1 second (FEV1) % and FEV1/FVC (Fig. 2G-I). In summary, TNFSF11 is associated with asthma.

Fig. 2
Elevated TNFSF11 expression in human specimens of airway and serum. (A) Collagen deposition measured by Masson staining. (B) TNFSF11 expression measured by immunofluorescent staining. (C) Quantification of collagen volume fraction. (D) Quantification of TNFSF11 intensity. (E) serum TNFSF11 in the asthmatic (n = 37) and normal (n = 22) groups measured by ELISA. (F) Correlation analysis between serum TNFSF11 and Eos. (G) Correlation analysis between serum TNFSF11 and FVC. (H) Correlation analysis between serum TNFSF11 and FEV1%. (I) Correlation analysis between serum TNFSF11 and FEV1/FVC. Magnification 200×, scale bar: 50 μm. Data are expressed as means ± standard deviation of 3 independent experiments.
TNFSF11, TNF ligand superfamily member 11; ELISA, enzyme-linked immunosorbent assay; Eos, eosinophil cells; FVC, forced vital capacity; FEV1, forced expiratory volume in 1 second.

^*P < 0.05, ^**P < 0.01, ^***P < 0.001, ^****P < 0.0001.

• Click for larger image
• Download as PowerPoint slide

Elevated TNFSF11 expression in the BALF, sera, and bronchial epithelia of mice with HDM-induced asthma

The animal models are shown in Fig. 3A.22 Serum IgE, Eos number and percentage, and BALF IL-33 level were significantly aggravated in the mice with HDM-induced asthma (Fig. 3B-E). Inflammatory cell infiltration and collagen deposition were elevated (Fig. 3F and G), and Tnfsf11 mRNA (Fig. 3H) and protein levels were significantly elevated (Fig. 3I and J). We further examined TNFSF11 expression in the sera and found higher levels in the asthmatic model than in normal mice (Fig. 3K). Immunofluorescence double staining confirmed that TNFSF11expression was elevated in the airway epithelia of the asthmatic mice compared with that in the normal mice (Fig. 3L and M). Finally, BALF TNFSF11 level in the asthmatic model was higher than that in the normal mice (Fig. 3N). In conclusion, TNFSF11 level is associated with HDM-induced asthma in mice.

Fig. 3
Increased TNFSF11 expression in mice with HDM-induced asthma. (A) Animal schedule of HDM-induced asthma model (n = 6). (B) Serum IgE detected by ELISA. (C, D) Number of Eos measured by hematologic cell analysis. (E) IL-33 levels in BALF detected by ELISA. (F) HE staining, (G) Masson-stained lung tissues. Magnification 200×, scale bar: 50 μm. (H) Tnfsf11mRNA level in mice with HDM-induced asthma. (I) TNFSF11 level in mice with HDM-induced asthma evaluated by Western blot. (J) TNFSF11/GAPDH ratio. (K) Mouse serum TNFSF11 detected by ELISA. (L) Immunofluorescence staining of TNFSF11 and EPCAM from mouse lungs. (M) TNFSF11 intensity analysis. (N) BALF TNFSF11 detected by ELISA. Data are expressed as means ± standard deviation of 3 independent experiments.
TNFSF11, TNF ligand superfamily member 11; HDM, house dust mite; Ig, immunoglobulin; ELISA, enzyme-linked immunosorbent assay; Eos, eosinophil cells; IL, interleukin; BALF, bronchoalveolar lavage fluid; HE, hematoxylin–eosin; GAPDH, glyceraldehyde 3-phosphate dehydrogenase; PBS, phosphate-buffered saline.

^*P < 0.05, ^**P < 0.01, ^***P < 0.001, ^****P < 0.0001.

• Click for larger image
• Download as PowerPoint slide

Elevated TNFSF11 protein level promotes airway remodeling

Next, we intranasally administered rmTNFSF11 active protein to the mice (Fig. 4A). HE and PAS staining results indicated that inflammatory cells aggregated around the airway and epithelial mucus secretion was upregulated after rmTNFSF11 protein stimulation (Fig. 4B, C, F, and G). Masson staining results indicated that a large amount of collagen was deposited around the airway after the intranasal administration of rmTNFSF11 protein (Fig. 4D and H). N-cadherin histochemistry and western blot results further confirmed that rmTNFSF11 protein induced mouse airway remodeling (Fig. 4E, I, J-M). In conclusion, TNFSF11 protein promotes mucus secretion, accelerating airway remodeling.

Fig. 4
TNFSF11 promoted airway inflammatory responses and remodeling. (A) Animal experimental schedule (n = 6). (B) HE, (C) PAS, (D) Masson, and (E) N-cadherin staining. Magnification 100×, scale bar: 100 μm. Magnification 200×, scale bar: 50 μm. (F-I) Intensity analysis of (B-E). (J) Changes of fibronectin, vimentin, and collagen I in rmTNFSF11-induced mice evaluated by Western blot. (K-M) Relative protein level of (J). Data are expressed as the means ± standard deviation of 3 independent experiments.
TNFSF11, TNF ligand superfamily member 11; HE, hematoxylin–eosin; PAS, periodic acid–Schiff; WT, wild type; PBS, phosphate-buffered saline; GAPDH, glyceraldehyde 3-phosphate dehydrogenase.

^*P < 0.05, ^**P < 0.01, ^***P < 0.001, ^****P < 0.0001.

• Click for larger image
• Download as PowerPoint slide

TNFSF11/TNFRSF11A axis deficiency alleviates HDM-induced airway remodeling

The effect of TNFSF11/TNFRSF11A axis in the Tnfsf11^+/− and Tnfrsf11a^+/− mice with HDM-induced asthma was further verified (Figs. 5A and 6A). Western blot results confirmed that TNFSF11/TNFRSF11A axis deficiency reduced the expression levels of N-cadherin, vimentin, and collagen I in mice with HDM-induced asthma (Figs. 5B-5F and 6B-6F). In particular, BALF IL-4 and serum IgE were significantly lower in the Tnfsf11^+/− and Tnfrsf11a^+/− mice with HDM-induced asthma than in the wild type (WT) mice with HDM-induced asthma (Figs. 5G, 5H, 6G, and 6H). Finally, HE and Masson staining results showed that pathological injury and collagen deposition decreased in the Tnfsf11^+/− and Tnfrsf11a^+/− mice with HDM-induced asthma than in the WT mice with HDM-induced asthma (Figs. 5I, 5J, 6I, and 6J).

Fig. 5
Tnfsf11^+/− alleviated HDM-induced asthma. (A) Strategy for HDM-induced asthma in Tnfsf11^+/− mice (n = 6). (B) Effects of the Tnfsf11^+/− on HDM-induced collagen I, vimentin, and N-cadherin verified by Western blot. (C-F) Relative protein level analysis. (G) BALF IL-4 and (H) Serum IgE detected by ELISA. (I) Effects of the Tnfsf11^+/− on HDM-induced inflammation detected by HE staining. (J) Effects of the Tnfsf11^+/− on HDM-induced airway remodeling detected by Masson staining. Magnification 100×, scale bar: 100 μm. Data are expressed as the means ± standard deviation of 3 independent experiments.
TNFSF11, TNF ligand superfamily member 11; HDM, house dust mite; BALF, bronchoalveolar lavage fluid; IL, interleukin; Ig, immunoglobulin; ELISA, enzyme-linked immunosorbent assay; HE, hematoxylin–eosin; WT, wild type; PBS, phosphate-buffered saline; GAPDH, glyceraldehyde 3-phosphate dehydrogenase.

^*P < 0.05, ^**P < 0.01, ^***P < 0.001, ^****P < 0.0001.

• Click for larger image
• Download as PowerPoint slide

Fig. 6
Tnfrsf11a^+/− alleviated HDM-induced asthma. (A) Strategy for HDM-induced asthma in Tnfrsf11a^+/− mice (n = 6). (B) Effects of the Tnfrsf11a^+/− on HDM-induced collagen I, vimentin, and N-cadherin verified by Western blot. (C-F) Relative protein level of (B). (G) BALF IL-4 and (H) Serum IgE detected by ELISA. (I) Effects of the Tnfrsf11a^+/− on HDM-induced inflammation detected by HE staining. (J) Effects of the Tnfrsf11a^+/− on HDM-induced airway remodeling detected by Masson staining. Magnification 100×, scale bar: 100 μm. Data are expressed as the means ± standard deviation of 3 independent experiments.
TNFRSF11A, TNF receptor superfamily 11A; HDM, house dust mite; BALF, bronchoalveolar lavage fluid; IL, interleukin; Ig, immunoglobulin; ELISA, enzyme-linked immunosorbent assay; HE, hematoxylin–eosin.

*P < 0.05, ^**P < 0.01, ^***P < 0.001, ^****P < 0.0001.

• Click for larger image
• Download as PowerPoint slide

TNFSF11/TNFRSF11A axis activates STAT3 signaling in mice with HDM-induced asthma

Messenger RNA transcription was analyzed in the WT and HDM-induced Tnfsf11^+/− mice with HDM-induced asthma. Volcanic maps showed the expression levels of some differential genes (Fig. 7A). Functional enrichment of differential genes showed that TNFSF11 played an important role in the extracellular region, and the function that might be related to the JAK-STAT signaling pathways (Fig. 7B). Western blot results showed that STAT3 phosphorylation levels significantly decreased in the Tnfsf11^+/− and Tnfrsf11a^+/− mice with HDM-induced asthma than in the WT mice with HDM-induced asthma (Fig. 7C-F). However, the phosphorylation levels of STAT5, STAT6, and JAK showed no change in the Tnfsf11^+/− mice with HDM-induced asthma than in the WT mice with HDM-induced asthma (Fig. 7C). Finally, immunofluorescence co-localization results revealed a significant reduction in STAT3 phosphorylation in the airway epithelium of the Tnfsf11^+/− mice with HDM-induced asthma than in the WT mice with HDM-induced asthma (Fig. 7G and H). In conclusion, TNFSF11/TNFRSF11A deficiency ameliorates HDM-induced asthma via the p-STAT3 signaling pathway.

Fig. 7
STAT3 signaling involved in HDM-induced asthma in Tnfsf11^+/− or Tnfrsf11a^+/− mice (n = 6). (A) Volcano plots for screening differential expression proteins with thresholds of P < 0.05 and fold change > 2. (B) KEGG enrichment after HDM-induced Tnfsf11^+/− vs. HDM-induced WT. (C, D) Enriched signaling pathway of JAK-STAT validated by Western blot. (E) p-STAT3/STAT3 ratio in mice with HDM-induced Tnfsf11^+/− vs. HDM-induced WT. (F) p-STAT3/STAT3 ratio in mice with HDM-induced Tnfrsf11a^+/− vs. HDM-induced WT. (G) The p-STAT3 and EPCAM around the airway epithelium in Tnfsf11^+/− mice with HDM-induced asthma measured by double-labeling immunofluorescence. Magnification 200×, scale bar: 50 μm. (H) p-STAT3 intensity analysis. Data are expressed as the means ± standard deviation of 3 independent experiments.
STAT3, signal transducer and activator of transcription 3; HDM, house dust mite; TNFSF11, TNF ligand superfamily member 11; TNFRSF11A, TNF receptor superfamily 11A; KEGG, Kyoto Encyclopedia of Genes and Genomes; WT, wild type.

^*P < 0.05, ^**P < 0.01, ^***P < 0.001, ^****P < 0.0001.

• Click for larger image
• Download as PowerPoint slide

TNFSF11/TNFRSF11A axis activates STAT3 signaling in HDM-stimulated HBE

TNFSF11 binding with TNFRSF11A activates downstream signaling. Therefore, TNFSF11/TNFRSF11A axis in HBE were interrupted by TNFSF11 or TNFRSF11A shRNA (Fig. 8A). The decrease in HDM-induced STAT3 phosphorylation was larger in HDM+TNFSF11 shRNA group than in the HDM group (Fig. 8B and C). TNFSF11 shRNA decreased the HDM-induced expression of vimentin and collagen I (Fig. 8B, D, and E). In addition, the STAT3 phosphorylation level in the HDM+TNFRSF11A shRNA group was lower than that in the HDM group (Fig. 8F and G). TNFRSF11A shRNA also decreased the HDM-induced expression of fibronectin and collagen I (Fig. 8F, H, and I).

Fig. 8
Intervention on TNFSF11/TNFRSF11A axis alleviated HDM-induced remodeling proteins with HBE cells by STAT3 phosphorylation. (A) Strategy for cell models. (B-E) Effects of TNFSF11 shRNA on HDM-stimulated HE, p-STAT3, collagen I, and vimentin. (F-I) Further analysis on the effects of TNFRSF11A shRNA on HDM-stimulated HE, p-STAT3, fibronectin, and collagen I. Magnification 200×, scale bar: 50 μm. Data are expressed as the means ± standard deviation of 3 independent experiments.
TNFSF11, TNF ligand superfamily member 11; TNFRSF11A, TNF receptor superfamily 11A; HDM, house dust mite; HBE, human bronchial epithelial; STAT3, signal transducer and activator of transcription 3; shRNA, short hairpin RNA; HE, hematoxylin–eosin; LV-NC, negative control lentivirus; NS, not significant.

^*P < 0.05, ^**P < 0.01, ^***P < 0.001, ^****P < 0.0001.

• Click for larger image
• Download as PowerPoint slide

TNFSF11/TNFRSF11A axis aggravated airway remodeling by TGFβ1/STAT3 action

We further added different concentrations of TNFSF11 (50, 100, and 200 ng/mL) to stimulate HBE cells. Western blot results showed that TGFβ1, fibronectin, vimentin, and collagen I levels significantly increased under TNFSF11 (50 ng/mL) stimulation (Fig. 9A-E). HBE cells were pretreated with stattic20 and then stimulated with rhTNFSF11. Stattic decreased the levels of collagen I and vimentin by inhibiting STAT3 phosphorylation (Fig. 9F-I).

Fig. 9
TNFSF11 promoted airway remodeling by TGFβ1/STAT3 signaling. (A-E) Remodeling proteins and TGFβ1 in rhTNFSF11-stimulated HBE cells evaluated by Western blot. rhTNFSF11 (50 ng/mL) selected to stimulate HBE cells. (F-I) Intervention with STAT3 phosphorylation used by stattic to further explore the role of STAT3 phosphorylation in remodeling proteins. (J-N) Intervention with TGFβ1 used by lentivirus to further explore the role of TGFβ1 signaling in STAT3 phosphorylation and remodeling proteins. (O) HBE staining of HE, TGFβR1, p-STAT3, collagen I, vimentin, and fibronectin in the rhTNFSF11-stimulated HBE cells with TGFβR1 shRNA. (P-T) Fluorescence intensity analysis. Magnification 200×, scale bar: 50 μm.
TNFSF11, TNF ligand superfamily member 11; TGFβ1, transforming growth factor β1; STAT3, signal transducer and activator of transcription 3; HBE, human bronchial epithelial; HE, hematoxylin–eosin; GAPDH, glyceraldehyde 3-phosphate dehydrogenase; NS, not significant.

^*P < 0.05, ^**P < 0.01, ^***P < 0.001, ^****P < 0.0001.

• Click for larger image
• Download as PowerPoint slide

Next, we constructed TGFβ1 or TGFβR1 shRNA to further verify the role of TGFβ1/TGFβR1 in STAT3 phosphorylation and remodeling induced by TNFSF11. The TGFβ1 expression decreased in the TNFSF11-stimulated HBE cells with TGFβ1 shRNA compared with that in the TNFSF11-stimulated HBE cells (Fig. 9J and K). The levels of STAT3 phosphorylation, vimentin, and collagen I significantly decreased in the TNFSF11-stimulated HBE cells with TGFβ1 shRNA (Fig. 9J, L-N). Finally, HBE cells were stained to locate TGFβR1 expression, STAT3 phosphorylation, remodeling proteins (Fig. 9O). TGFβR1 expression and STAT3 phosphorylation decreased in the TNFSF11-stimulated HBE cells with TGFβR1 shRNA compared with that in the TNFSF11-stimulated HBE cells (Fig. 9P and Q). The levels of fibronectin, vimentin, and collagen I significantly decreased in the TNFSF11-stimulated HBE cells with TGFβR1 shRNA compared with those in the HDM-stimulated HBE cells (Fig. 9R-T).

Denosumab improves HDM-induced asthma in humanized HSC-NOG-EXL mice by inhibiting TGFβ1/STAT3 action

Flow cytometry results showed that human CD45 was more than 50% after 7 weeks of human CD34+huHSC transplantation (Supplementary Fig. S2). The effect of denosumab was further verified in the humanized HSC-NOG-EXL mice with HDM-induced asthma (Fig. 10A). Western blot analysis showed that denosumab significantly inhibited or neutralized HDM-induced TNFSF11 expression, but had no effect on HDM-induced TNFRSF11A expression in the HSC-NOG-EXL mice (Fig. 10B-D). IL-4 cytokine level significantly decreased in the HDM + denosumab group compared with those in the HDM group (Fig. 10B and E). IL-5 cytokine level showed no change between the HDM + denosumab and HDM groups (Fig. 10B and F). The EMBP expression decreased in the HDM + denosumab group compared with that in the HDM group (Fig. 10B and G).

Fig. 10
Effects of denosumab (anti-human TNFSF11) on HDM-induced humanized HSC-NOG-EXL murine. (A) Strategy for HDM-induced asthma in HSC-NOG-EXL murine (n = 6). (B) Western blot analysis of TNFSF11, TNFRSF11A, IL-4, IL-5, and EMBP. (C) TNFSF11/GAPDH ratio. (D) TNFRSF11A/GAPDH ratio. (E) IL-4/GAPDH ratio. (F) IL-5/GAPDH ratio. (G) EMBP/GAPDH ratio. (H) HE, (I) PAS, (J) Masson staining to examine the effects of denosumab on HSC-NOG-EXL mice with HDM-induced asthma. (K) Mouse lungs stained with TGFβ1 by immunofluorescence. (L) Mouse lungs stained with p-STAT3 by immunofluorescence. (M-Q) Intensity analysis of (H-L). Magnification 100×, scale bar: 100 μm. Magnification 200×, scale bar: 50 μm. Data are expressed as the means ± standard deviation of 3 independent experiments.
TNFSF11, TNF ligand superfamily member 11; HDM, house dust mite; TNFRSF11A, TNF receptor superfamily 11A; IL, interleukin; EMBP, eosinophil major basic protein; GAPDH, glyceraldehyde 3-phosphate dehydrogenase; HE, hematoxylin–eosin; PAS, periodic acid–Schiff; TGFβ1, transforming growth factor β1; STAT3, signal transducer and activator of transcription 3; NS, not significant.

^*P < 0.05, ^**P < 0.01, ^***P < 0.001, ^****P < 0.0001.

• Click for larger image
• Download as PowerPoint slide

HE and PAS results showed that denosumab significantly reduced the accumulation of HDM-induced inflammatory cells into the airway and the airway secretion of mucus (Fig. 10H, I, M, and N). Masson staining further demonstrated that denosumab significantly inhibited HDM-induced airway collagen deposition (Fig. 10J and O), TGFβ1 level, and STAT3 phosphorylation in the HSC-NOG-EXL mice (Fig. 10K, L, P, and Q). Overall, denosumab decreases TGFβ1 level and STAT3 phosphorylation in the HSC-NOG-EXL mice with HDM-induced asthma.