Female BALB/c mice (Orient, Seongnam, Korea), 6 weeks of age, were used in all experiments. The mice were randomly allocated to the following groups: (1) control (CON); (2) ovalbumin (OVA) challenge; (3) OVA challenge + miRNA negative control (NC) as a vehicle control (OVA + miRNA NC); and (4) OVA challenge + miR-21 inhibitor (OVA + miR-21 inhibitor). All procedures were in accordance with the Laboratory Animals Welfare Act, Guide for the Care and Use of Laboratory Animals, and Guidelines and Policies for Rodent Experiments provided by the Institutional Animal Care and Use Committee (IACUC) of the College of Medicine of The Catholic University of Korea (No. CUMS-2016-0166-02). The IACUC and Department of Laboratory Animals of the Catholic University of Korea, Songeui Campus, was accredited as a Korea Excellence Animal Laboratory Facility by the Korea Food and Drug Administration in 2017 and acquired full accreditation by the Association for Assessment and Accreditation of Laboratory Animal Care International in 2018.

The mice were immunized with 25 μg of OVA (chicken egg albumin, grade V, from Sigma-Aldrich, St. Louis, MO, USA) with 1 mg of aluminium hydroxide (alum) in 200 μL of phosphate-buffered saline (PBS). The mice were immunized by subcutaneous injection on days 0, 7, 14, and 21, and then challenged with intranasal OVA (20 μg/50 μL PBS) on day 33, 35, and 37 and then again twice a week for 3 months while anaesthetized with isoflurane (Vedco, St. Joseph, MO, USA). CON mice were treated with PBS in the same manner. Mice were killed 24 hours after the final OVA challenge, and bronchoalveolar lavage fluid (BALF) and lung tissues were obtained.

We used the mirVana miR-21 inhibitor (Ambion, Applied Biosystems, Foster City, CA, USA) to inhibit the expression of miR-21 in this OVA-induced chronic asthma mouse model [19]. The mirVana miRNA inhibitor was used for the NC. The miR-21 inhibitor or miRNA NC (50 μg/50 μL in PBS) was given intranasally once a week for sensitization and then twice a week for 3 months, as shown in Fig. 1.

AHR was measured as airway resistance (resistance of the respiratory system [RrS]) after methacholine administration and was recorded using the FlexiVent system (SCIREQ, Montreal, QC, Canada) [20]. Mice were anaesthetized with an intraperitoneal injection of a mixture of Rompun and Zoletil (1:4). The trachea was exposed and cannulated, and the animal was connected to a computer-controlled small-animal ventilator and ventilated with the following settings to achieve a mean lung volume close to that during spontaneous breathing: tidal volume of 10 mL/kg, frequency of 150 breath/minute, and positive end-expiratory pressure of 2 cmH₂O. Mice were first exposed to nebulized PBS for 3 minutes to establish the baseline Rrs values and then to increasing concentrations of nebulized methacholine (6.25, 12.5, 25, and 50 mg/mL; Sigma-Aldrich) in PBS using an ultrasonic nebulizer (Aerosonic, DeVilbiss, Somerset, PA, USA). Recordings were taken for 3 minutes following each exposure. The Rrs values measured during each 3 minutes interval were averaged and are presented for each methacholine concentration.

Bronchoalveolar lavage was performed immediately after the measurement of AHR. The exposed trachea was cannulated with silicone tubing attached to a 23-gauge needle on a 1 mL tuberculin syringe. BALF was collected after instillation of 0.8 mL of sterile PBS through the trachea into the lungs. The total cell counts in BALF were obtained using a LUNA Automated Cell Counter (Logos Biosystems, Inc., Annandale, VA, USA). BALF was centrifuged onto microscope slides at 2,000 rpm for 7 minutes in a Cytospin centrifuge (Thermo Fisher Scientific, Waltham, MA, USA) and stained with Diff-Quick (Sysmex, Kobe, Japan). The percentages of macrophages, eosinophils, lymphocytes, and neutrophils in BALF were obtained by counting 500 leukocytes in randomly selected fields under light microscopy.

The concentrations of IL-4, IL-5, IL-13, and interferon γ (IFN-γ) in BALF were measured using enzyme-linked immunosorbent assay (ELISA) kits (R&D Systems, Minneapolis, MN, USA) according to the manufacturer’s instructions. Active TGF-β1 concentration was measured after pretreating the samples with 10 μL of 1 N hydrochloric acid (HCl) for 10 minutes at room temperature to convert the active TGF-β1 and then neutralizing the samples with 1.2 N NaOH/0.5 M hydroxyethyl piperazine ethane sulfonicacid (HEPES) according to the manufacturer’s instructions (R&D Systems).

Lung samples were fixed in 4% paraformaldehyde for 24 hours and then embedded in paraffin wax. Sections were cut at a thickness of 4 μm using a microtome, and deparaffinized tissue sections were stained with H&E for the examination of airway inflammation. For assessment of airway inflammation, the slides were numbered randomly and evaluated independently by two blinded investigators. The quantity of perivascular or peribronchial inflammation was assessed as described previously [21]. The following 5-point scoring system (grades 0 to 4) was used: no inflammation (grade 0), occasional cuffing with inflammatory cells (grade 1), most bronchi or vessels surrounded by a thin layer (grade 2), moderate layer (grade 3), or thick layer (grade 4) of inflammatory cells. Periodic acid-Schiff (PAS) staining was used to identify goblet cell hyperplasia and submucosal gland hypertrophy. Hyperplasia of goblet cells in the epithelial lining is expressed as the percentage of the goblet cells in the epithelial cells. A 5-point scoring system (grades 0 to 4) was used to quantify the percentages of goblet cells exhibiting hyperplasia as follows: no goblet cells (grade 0), < 25% (grade 1), 25% to 50% (grade 2), 51% to 75% (grade 3), and > 75% (grade 4).

Six-micrometre-thick sections of lung tissue from each paraffin block were deparaffinized with xylene and hydrated in ethanol. For immunohistochemical detection of collagen III, collagen V, and α-SMA, the lung sections were incubated overnight at 4°C with anti-collagen III antibody, anti-collagen V antibody (both from Abcam, Cambridge, MA, USA), and anti-α-SMA antibody (Dako, High Wycombe, UK). Immunoreactivity was detected by sequential incubation of lung sections with a biotinylated secondary antibody followed by peroxidase reagent (Vector Laboratories, Burlingame, CA, USA) and diaminobenzidine chromogen (Invitrogen, Carlsbad, CA, USA). The results are expressed as the area of immunostaining per micrometre length of the basement membrane of bronchioles with an internal diameter of 150 to 200 μm. At least 10 bronchioles were counted in each slide using a slide scanner (Pannoramic MIDI, 3DHISTECH Ltd., Budapest, Hungary).

Hydroxyproline content was quantified in 60 mg of lung tissue from each mouse to determine the total amount of collagen. A hydroxyproline colorimetric assay kit (BioVision, Milpitas, CA, USA) was used according to the manufacturer’s instructions.

hBSMCs were obtained from Lonza (Walkersville, MD, USA) and were maintained in smooth muscle basal medium supplemented with 5% (v/v) foetal bovine serum, 5 ng/mL insulin, 2 ng/mL human basic fibroblast growth factor, 50 ng/mL human epidermal growth factor, 50 μg/mL gentamicin, and 50 ng/mL amphotericin B at 37°C in 5% CO₂. Cells between passages 3 and 8 were used for all experiments. We used the mirVana miR-21 inhibitor to inhibit the expression of miR-21 in hBSMCs. The Vana miRNA negative inhibitor was used as the NC. hBSMCs were grown in 100 mm culture dishes overnight, and cells at 60% to 70% confluence were transfected with 30 pmol of the miR-21 inhibitor or negative inhibitor using Lipofectamine 2000 (Invitrogen). After 72 hours of incubation, the cells were cultured under serum-free conditions for 16 hours and then treated with recombinant human TGF-β1 (R&D Systems) for the indicated times.

Total RNA was isolated from lung homogenates using TRIzol (Invitrogen) and reverse-transcribed. Real-time polymerase chain reaction (PCR) was performed using a CFX96 Real-Time PCR Detection System (Bio-Rad Laboratories, Hercules, CA, USA). Reactions were amplified using the selective primers described in Table 1, and an iQ SYBR gene expression assay (Bio-Rad Laboratories), according to the manufacturer’s instructions.

The lungs were sonicated in radioimmunoprecipitation assay (RIPA) buffer with ceramic beads at 5,000 rpm for 20 seconds. The lung homogenates were separated using 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and blotted with monoclonal antibodies to the specific proteins, as indicated. Antibodies to Smad7 (#ab190987) and β-actin (#ab8227) were purchased from Abcam.

All data are presented as the mean ± SEM. Results were compared between groups using one-way analysis of variance (ANOVA) with the Tukey post hoc test or two-way ANOVA with the Bonferroni test. GraphPad Prism statistical software package (San Diego, CA, USA) was used for the analyses. A p value of < 0.05 was considered to be significant.

To investigate the effects of miR-21 inhibition on chronic allergic asthma, we used a well-established OVA-induced asthma mouse model [22]. Mice were sensitized to and challenged with OVA with and without the miR-21 inhibitor or miRNA NC, as described in Fig. 1 We first used TaqMan PCR analysis to evaluate the expression of miR-21 in this OVA-induced chronic allergic asthma mouse model. Compared with OVA + miRNA NC mice, miR-21 expression was increased by 10 to 15-fold in OVA-sensitized and -challenged mice and in miRNA NC-treated OVA (OVA + miRNA NC) mice. By contrast, miR-21 expression was decreased by threefold in OVA-administered mice treated with the miR-21 inhibitor (OVA + miR-21 inhibitor) (in Taq-Man RT-qPCR analysis section of Supplementary methods and Supplementary Fig. 1A).

A previous study showed that IL-12p35 is an miR-21 mRNA target [23] and is downregulated in asthma [6]. We next evaluated the expression of IL-12p35 in the OVA-induced chronic allergic asthma mouse model. As shown in Supplementary Fig. 1B, the mRNA levels of IL-12p35 were markedly reduced following OVA administration but were markedly increased following treatment with the miR-21 inhibitor. Collectively, these data indicate that the miR-21 inhibitor was effective in OVA-induced chronic allergic asthma mouse model.

To determine the role of miR-21 inhibition on airway inflammation, we counted the total number of cells and the percentages of macrophages, lymphocytes, neutrophils, and eosinophils in BALF. The total number of cells and percentages of macrophages and eosinophils were significantly higher in the OVA and OVA + miRNA NC groups than in the CON group (p < 0.001). By contrast, treatment with miR-21 inhibitor significantly decreased the total cell count and percentage of eosinophils in BALF (p < 0.001) compared with the OVA + miRNA NC group (Fig. 2A). H&E-stained lung sections from OVA-administered and OVA + miRNA NC mice showed an increase in the infiltration of inflammatory cells in both the peribronchial and perivascular regions. However, miR-21 inhibitor treatment markedly reduced the infiltration of inflammatory cells in the airway, as evidenced by the lower absolute cell count in BALF (Fig. 2B and 2C).

Rrs was analysed to determine the effect of miR-21 inhibition on AHR. Rrs was elevated in OVA and OVA + miRNA NC mice but was markedly lower (p < 0.001) in the OVA-treated groups treated with the miR-21 inhibitor; this decrease exhibited a methacholine dose-dependent pattern (Fig. 2D).

To investigate the effects of miR-21 inhibition on inflammatory cytokine expression, we measured the levels of T_H2 and T_H1 cytokines in the lungs of the four groups. OVA + miR-21 inhibitor mice exhibited significantly lower levels of the T_H2 cytokines IL-4 (p < 0.001), IL-5 (p < 0.05), and IL-13 (p < 0.001) than OVA + miRNA NC mice (Fig. 3A). OVA + miR-21 inhibitor-treated mice also had significantly higher levels of IFN-γ (p < 0.001) than OVA + miRNA NC mice (Fig. 3B). Taken together, these findings suggest that miR-21 inhibition reduced airway inflammation and AHR in this OVA-induced chronic allergic asthma model.

Airway remodelling is a critical structural alteration associated with chronic allergic asthma and is manifested as goblet cell hyperplasia, peribronchial subepithelial fibrosis, and smooth muscle cell hypertrophy. To examine the effects of miR-21 inhibition on these pathological components in OVA-induced chronic allergic asthma, we first examined mucus production in the lungs of these mice following OVA sensitization and challenge. OVA and OVA + miRNA NC mice exhibited more mucus-secreting cells than CON mice, as determined by PAS staining (Fig. 4A) and the hyperplasia score. miR-21 inhibitor-treated mice had significantly (p < 0.001) fewer hyperplastic goblet cells than OVA + miRNA NC mice (Fig. 4B). As shown in Fig. 4I, the mRNA levels of Muc5ac increased markedly in the OVA group but were much lower in the lung tissue following treatment with the miR-21 inhibitor.

To investigate the effects of miR-21 inhibition on peribronchial subepithelial fibrosis, we measured the levels of collagen III and V. Collagen III and V immunostaining of the peribronchial and perivascular regions was significantly greater in OVA and OVA + miRNA NC mice than in CON mice. Collagen immunostaining was significantly lower in samples from mice treated with the miR-21 inhibitor than OVA + miRNA NC mice (Fig. 4C–4F).

We measured hydroxyproline content to determine the total collagen level in the lungs of mice in the four groups. Hydroxyproline content was significantly (p < 0.05) lower in the miR-21 inhibitor-treated mice than in the OVA + miRNA NC mice (Fig. 4J). Lastly, we examined whether the miR-21 inhibitor acted on smooth muscle hyperplasia, one of the most important aspects of airway remodelling. OVA and OVA + miRNA NC mice showed increased peribronchial α-SMA immunostaining. By contrast, the expression of α-SMA was inhibited by the miR-21 inhibitor (Fig. 4G). This pattern was consistent with the quantification of the area of α-SMA staining (Fig. 4H). Collectively, these data indicate that the miR-21 inhibitor was effective in preventing or limiting airway remodelling.

TGF-β1 is directly involved in airway remodelling in asthma [14,24]. We wondered whether TGF-β1 might also be involved in airway remodelling in this OVA-induced chronic asthma mouse model. The TGF-β1 level was much higher in lung homogenates from OVA-challenged mice (p < 0.05) than from the CON mice. The TGF-β1 level was significantly lower in lung homogenates from mice treated with the miR-21 inhibitor (p < 0.05) than from OVA-challenged mice (Fig. 5A).

Smad7 has been shown to inhibit the TGF-β signalling pathway [15,25] and to be a direct target of miR-21 in various diseases [26,27]. Therefore, to examine the suppression of Smad7 by miR-21, we used TargetScan and miRanda to predict the paired target regions of the Smad7 3′untranslated region and the miR-21 seed sequence (Fig. 5B). We first checked to determine whether miR-21 inhibitor treatment induced Smad7 expression in this model of OVA-induced chronic allergic asthma. expression was significantly reduced in OVA-challenged mice (p < 0.001) and was increased by miR-21 inhibitor treatment compared with OVA + miRNA NC treatment (Fig. 5C). Smad7 protein expression was increased by 2 to 3-fold in mice treated with the miR-21 inhibitor compared with the OVA + miRNA NC mice (Fig. 5D and 5E). These results suggest that miR-21 inhibition can inhibit TGF-β1 expression and upregulate Smad7 gene and protein expression.

To investigate whether miR-21 inhibition influences Smad7 expression, we examined the expression of Smad7 protein in TGF-β1-activated hBSMCs. The content of Smad7 protein was lower after TGF-β1 stimulation than with no treatment. The miR-21 inhibitor increased Smad7 protein level (Fig. 6A). Altered extracellular matrix deposition is an important contributor to the remodelling of airways in chronic allergic asthma. We next examined whether miR-21 inhibition affects collagen production in hBSMCs. TGF-β1 stimulation increased the production of fibrotic matrix components such as collagen types I (COL1A1), IV (COL4A1), and VII (COL7A1) in hBSMCs. These changes were attenuated substantially by treatment with the miR-21 inhibitor (Fig. 6B). Expression of COL3A1 was not changed.

These findings prompted us to explore further the mechanisms underlying the anti-fibrotic effects of miR-21 inhibition in hBSMCs. The mRNA levels of plasminogen activator inhibitor-1 (PAI-1) and connective tissue growth factor (CTGF) increased markedly after treatment with TGF-β1 and decreased after treatment with the miR-21 inhibitor compared with miRNA NC treatment (both comparisons, p < 0.001) (Fig. 6C). As expected, treatment of hBSMCs with the miR-21 inhibitor reversed the TGF-β1-induced profibrotic gene expression and collagen production. We further examined the role of miR-21 inhibition in the expression of α-SMA (ACTA2) as a marker of bronchial smooth muscle hypertrophy in hBSMCs. Consistent with the above results, the miR-21 inhibitor also reduced ACTA2 mRNA expression (Fig. 6D). Taken together, these data provide in vitro evidence that miR-21 inhibition decreases airway remodelling by modulating TGF-β1–Smad7 signalling.