Dabigatran ameliorates airway smooth muscle remodeling in asthma by modulating Yes‐associated protein

Abstract Accumulating evidence indicates that thrombin, the major effector of the coagulation cascade, plays an important role in the pathogenesis of asthma. Interestingly, dabigatran, a drug used in clinical anticoagulation, directly inhibits thrombin activity. The aim of this study was to investigate the effects and mechanisms of dabigatran on airway smooth muscle remodeling in vivo and in vitro. Here, we found that dabigatran attenuated inflammatory pathology, mucus production, and collagen deposition in the lungs of asthmatic mice. Additionally, dabigatran suppressed Yes‐associated protein (YAP) activation in airway smooth muscle of asthmatic mice. In human airway smooth muscle cells (HASMCs), dabigatran not only alleviated thrombin‐induced proliferation, migration and up‐regulation of collagen I, α‐SMA, CTGF and cyclin D1, but also inhibited thrombin‐induced YAP activation, while YAP activation mediated thrombin‐induced HASMCs remodeling. Mechanistically, thrombin promoted actin stress fibre polymerization through the PAR1/RhoA/ROCK/MLC2 axis to activate YAP and then interacted with SMAD2 in the nucleus to induce downstream target genes, ultimately aggravating HASMCs remodeling. Our study provides experimental evidence that dabigatran ameliorates airway smooth muscle remodeling in asthma by inhibiting YAP signalling, and dabigatran may have therapeutic potential for the treatment of asthma.

characterized by abnormal proliferation, migration and excessive extracellular matrix (ECM) and muscle protein expression of airway smooth muscle cells. 5 Recent years, several studies showed that the coagulation system was activated in the asthmatic airway, suggesting that it may be related to the pathogenesis of asthma. 6,7 Thrombin, the major effector of the coagulation cascade, was detected in elevated level of induced sputum and bronchoalveolar lavage fluid (BALF) in asthmatic patients. 8,9 In vitro, thrombin augmented inflammation of airway cells via cleavage of protease-activated receptors 10 and stimulated contraction of human bronchial rings. 11 In addition, thrombin is also a potent inducer of cell proliferation, pro-inflammatory chemokines and ECM proteins of lung fibroblasts. 12 Our previous study reported that thrombin promoted HASMCs proliferation, and for the first time, thrombin was found to stimulate the production of ECM proteins in HASMCs. 13 Dabigatran, a novel oral thrombin inhibitor for clinical anticoagulation, was shown to exert anti-inflammatory and anti-fibrotic effects in a mouse model of bleomycin-induced pulmonary fibrosis, 14 and dabigatran also impaired thrombin-induced cell proliferation, α-SMA expression and organization, and production of collagen and connective tissue growth factor (CTGF) in lung fibroblasts. 15 However, the effects of dabigatran on airway remodeling in OVA-evoked asthmatic mice and on thrombin-induced HASMCs remodeling have not been reported.
Yes-associated protein, a transcriptional coregulator, is a key effector of the Hippo pathway that regulates organ size and tissue homeostasis by affecting cell proliferation and differentiation. 16 When the Hippo pathway is OFF, YAP is dephosphorylated and accumulates in the nucleus, where they bind to corresponding transcription factors to induce the transcription of downstream target genes such as CTGF and cyclin D1. 17 Recent studies indicated that YAP was up-regulated in the bronchial airway smooth muscle of OVA-evoked asthmatic mice, suggesting that YAP could be associated with airway smooth muscle remodeling in asthma. 18 It was shown that S1P stimulated proliferation, migration and contraction of rat airway smooth muscle cell by activating YAP, 19 and Jung-Soon Mo et al 20 21 In addition, We previously confirmed that thrombin induced actin stress fibre polymerization in HASMCs, 13 while changes in actin cytoskeleton represent a central mechanism for controlling YAP activity. 22,23 However, the role of YAP in asthma, whether thrombin activates YAP in HASMCs and its underlying mechanisms remain unclear.
Based on these findings, this study aimed to characterize the role of dabigatran in airway smooth muscle remodeling in asthma and explore the underlying mechanisms.

| Establishment of OVA-evoked asthma mouse model
Male BALB/c mice, weighing 25 g, aged 6-8 weeks, were purchased from Laboratory Animal Center of Southern Medical University, and the animal experimental procedures were approved by the Animal Research Ethics Committee of Southern Medical University. All mice were housed in the SPF facility, and the mice were fed with sterile water and radiation-irradiated feed for a 12-hour light/dark cycle. Mice were treated as shown in Figure 1A. Mice were randomly divided into 6 groups accord-

| Lung histology and immunohistochemistry analysis
The lungs were dissected from the chest cavity after the lavage.
The right lung was immediately fixed in 4% paraformaldehyde and embedded in paraffin, and tissue sections (5 μm) were prepared.
To assess airway remodeling, HE staining was performed to assess airway thickness and inflammatory infiltration, PAS to quantify airway goblet cells and mucus production, and Masson's trichrome to visualize collagen deposition, as described previously. 24 For immunohistochemical analysis, the sections were initially incubated with anti-YAP rabbit monoclonal antibody at 4°C overnight and then were incubated with HRP-conjugated goat anti-rabbit for 30 minutes at 37°C. Positive staining was detected with HRPconjugated streptavidin, visualized with 3, 3′diaminobenzidine.
Finally, sections were counterstained with haematoxylin and observed under a light microscope.

| Enzyme-linked immunosorbent assay
The levels of thrombin (Abcam), thrombin-antithrombin complexes (TAT), interleukin-4 (IL-4), IL-5, IL-13 in BALF and total serum IgE (Elabscience) were measured by Enzyme-linked immunosorbent F I G U R E 1 Dabigatran reduced OVA-induced elevated levels of thrombin and inflammatory cytokines in the airway of mice. A, The construction of OVA-evoked asthma mouse model. Mice were sensitized by OVA/Al(OH) 3 on day 1 and day 7, while from day 14 to day 16 and day 21 to 23, mice were challenged with OVA alone by intranasal (i.n.). B-G, Levels of thrombin, TAT, IL-4, IL-5, IL-13 in BALF and total IgE in serum were measured by ELISA. Data are presented as mean ± SEM. n = 6 mice per group; **P < .01, ***P < .001, compared with the control group; # P < .05, ## P < .01, ### P < .001, compared with the OVA group assay (ELISA) kits, respectively. The protocols were followed according to the manufacturer's instructions. Briefly, the standard working solution and the sample to be tested were separately added to a 96well plate at 100 μL per well. One hundred microliter of biotin-conjugated antibody was added to each well and incubated at 37°C for 1 hour. Next, 100 μL of the enzyme conjugate working solution was added per well and incubated at 37°C for 30 minutes. Subsequently, a substrate solution (TMB) was added to each well at 90 μL per well and incubated at 37°C for 15 minutes in the dark. Finally, 50 μL of the stop solution was added to each well to terminate the reaction.
The optical density (OD) of each well was measured using a microplate reader set to 450 nm. A standard curve of the average of the OD repeat readings is created to calculate the concentration to be tested.

| Transfection with small interfering RNA
To knock down the expression of YAP, siRNA (RiboBio) was transfected into HASMCs with Lipofectamine™ 3000 reagent

| RNA extraction and quantitative real-time PCR
Total RNA was extracted from HASMCs using RNAiso Plus (Takara) in accordance with the manufacture's protocol. The mRNA was quantified by a NanoDrop Spectrophotometer (NanoDrop Tech), and cDNA was synthesized with the PrimeScript™ RT Master Mix (Takara). The gene-specific primers (listed in Table 1) were obtained from Sangon Biotech. Real-time quantitative PCR (qPCR) was performed on a LightCycler 480 II (Roche) with a SYBR Premix Ex Taq kit (Takara). The level of GAPDH mRNA expression was used as the internal reference. Data are presented as the fold change over the control group, as determined using the 2 −△△Ct method.

| Western blot analysis and quantification
Proteins from cultured cells or animal lung tissues were extracted using radioimmunoprecipitation assay (RIPA) lysis buffer (FDbio Science), containing a protease inhibitor cocktail and phosphatase inhibitors (Sigma). Protein concentration was determined by the BCA Protein Assay Kit (FDbio science). For immunoblotting, equal amounts of proteins (20 μg/lane) were subjected to electrophoresis on a 10% or 12% SDS-polyacrylamide gel and electroblotted to PVDF membrane (EMD Millipore). The membranes were blocked for 1 hour with 5% non-fat milk in Tris-Buffered Saline Tween-20 (TBST). Blots were then incubated over night with specific primary antibodies at 4°C and washed, followed by a 1-hour incubation with secondary antibodies conjugated to HRP (FDbio science) at room temperature. Target proteins were recorded by using enhanced chemiluminescence (ECL) reagents (FDbio science) and quantified using the ImageJ software. Reverse: 5′-ACGTAGGCACCATAGAGGTTG-3′

| Immunofluorescence staining
Human airway smooth muscle cells were fixed with 4% paraformaldehyde for 15 minutes and then permeabilized with 0.3% Triton X-100 for 10 minutes. After blocking for 1 hour in 5% goat serum, cells were incubated with primary antibodies overnight at 4°C and followed by fluorescent-conjugated antihuman secondary antibody for 1 hour at room temperature in the next day.
Actin-Tracker Green (Beyotime Biotech) was used to stain actin filaments, and DAPI (Beyotime Biotech) was used for cell nuclei.
Photographs were taken by a fluorescence inverted/laser scanning confocal microscope (Leica Imaging Systems).

| Scratch migration assay
Human airway smooth muscle cells were seeded into 6-well plates and grown to a confluent density. After serum starvation and pre-

| Co-immunoprecipitation
The

| Statistical analysis
All values are presented as mean ± SEM. Differences between two mean values were analysed using Student's t test, and multiple mean values were compared using one-way ANOVA followed by the Tukey post hoc test. Statistical analysis was performed in Prism 8 software (GraphPad). Values of P < .05 were considered statistically significant.

| Dabigatran reduced OVA-induced elevated level of thrombin in the airway of mice
The levels of thrombin and TAT in BALF of mice were determined by ELISA. As shown in Figure 1B,C, compared with the control mice, the levels of thrombin and TAT were greater in the OVA-evoked asthmatic mice, which were significantly decreased by two different concentrations of dabigatran treatment (5 and 10 mg/g).

| Dabigatran or verteporfin attenuated OVAinduced airway inflammation, remodeling and YAP activation in the lungs of mice
As shown in Figure 1D Compared with the control mice, the OVA-evoked asthmatic mice exhibited thicker airway wall, more inflammatory cell infiltration, goblet cells and mucus production, and collagen deposition.
However, these histopathological changes were notably alleviated in the presence of dabigatran and verteporfin compared with the OVA-evoked asthmatic mice (Figure 2A-G).
Then, immunohistochemistry and Western blot of mice lung tissue were performed to explore the expression and distribution of YAP and the effect of dabigatran on YAP. Figure 2H depicted that YAP expression was apparently up-regulated in the OVA-evoked asthmatic mice and mainly distributed in airway smooth muscle.
However, treatment with dabigatran or verteporfin in asthmatic mice reduced YAP expression compared with the OVA-evoked asthmatic mice. As shown in Figure 2I and J, compared with the control mice, the level of phosphorylated YAP (p-YAP) was decreased in the OVA-evoked asthmatic mice. Treatment with dabigatran or verteporfin prevented alterations in the level of p-YAP induced by OVA challenge.
In summary, the above results indicated that YAP was activated in airway smooth muscle of asthmatic mice, while dabigatran and verteporfin not only inhibited the activation of YAP, but also attenuated the lung pathology of asthma models.

| Dabigatran suppressed thrombin-induced HASMCs remodeling and YAP activation
We further determined the effects of dabigatran on the cellular biological functions associated with HASMCs remodeling and on YAP activity in response to thrombin. As shown in Figure 3A

| YAP mediated the effects of thrombin on HASMCs remodeling
To investigate the role of YAP in thrombin-induced proliferation, migration, and up-regulation of remodeling-related proteins, siYAP and verteporfin were utilized. The efficiency of YAP knockdown confirmed by immunoblotting, and the best efficiency of three sequences was siYAP#1 ( Figure 4A,B). As shown in Figure 4C

| Thrombin promoted actin stress fibre polymerization through the PAR1/RhoA/ROCK/ MLC2 axis to activate YAP
The mRNA levels of PAR1-PAR4 were analysed by qPCR. It was shown that all four PARs were expressed in HASMCs, and PAR1 was the most abundant ( Figure 5A) Based on these observations, we concluded that thrombin induced actin stress fibres by activating PAR1/RhoA/ROCK/MLC2 signalling to regulate YAP dephosphorylation and subcellular localization.

| Dabigatran blocked thrombin-induced interaction between YAP and SMAD2
Yes-associated protein, a transcriptional coactivator, exerts its regulatory functions by binding to transcription factors in the nucleus. We detected that thrombin stimulated SMAD2 phosphorylation in a time-dependent manner and partially recovered after 1 hour ( Figure 6A,C). However, treatment with dabigatran decreased thrombin-induced SMAD2 phosphorylation ( Figure 6B,D).
Therefore, we aimed to investigate whether thrombin induced YAP interaction with SMAD2 and whether dabigatran influenced their interaction. YAP and SMAD2 nucleo-cytoplasmic localization was studied in parallel by immunofluorescence double staining, and we found a significant increase in nuclear accumulation of YAP and SMAD2 after thrombin treatment, but dabigatran reduced thrombin-induced nuclear accumulation of YAP and SMAD2 ( Figure 6E). Furthermore, we performed CoIP to explore the binding of YAP to SMAD2. Figure 6F demonstrated that thrombin promoted the binding of YAP to SMAD2. Nevertheless, the presence of dabigatran reversed thrombin-induced the binding of YAP to SMAD2 ( Figure 6G).

| D ISCUSS I ON
The role of airway smooth muscle cells in airway remodeling is being increasingly recognized, with hyperplasia and hypertrophy of this cell type contributing to wall thickness and airway hyper-responsiveness and, in addition, airway smooth muscle cells being a source of extracellular matrix, and of growth factors and cytokines that promote inflammation and airway remodeling itself. 27  The hippo pathway is an evolutionarily conserved signalling cascade that controls organ size by regulating cell proliferation, apoptosis, differentiation and stem/progenitor cell fate, which involves a series of kinases and adaptors that leads to inactivation of transcriptional coactivator with YAP. 30 Increased YAP expression has been detected in airway smooth muscle cells and bronchial epithelial cells, suggesting that YAP may participate in the pathogenesis of asthma. 18,[31][32][33] Our experiments confirmed that both YAP knockdown and YAP inhibition were effective in alleviating airway remodeling in asthmatic mice and thrombin-induced proliferation, migration, and up-regulation of remodeling-related proteins in HASMCs. However, the upstream F I G U R E 3 Dabigatran suppressed thrombin-induced HASMCs remodeling and YAP activation. A, HASMCs were treated with different concentrations of dabigatran (0-5000 ng/mL) and thrombin (1 U/mL) for 24 h, and cell proliferation was measured by CCK-8 assay. B, HASMCs were exposed to dabigatran (3000 ng/mL) and thrombin (1 U/mL) for indicated times, and cell proliferation was measured by CCK-8 assay. C, HASMCs were pre-treated with dabigatran (3000 ng/mL) for 1 h before stimulation with thrombin (1 U/mL) for 24 h, HASMCs proliferation was measured by EdU incorporation assay. HASMCs were stained with EdU (red) and counterstained with DAPI (blue) to visualize the nuclei (scale bar, 100 μm). D, HASMCs migration was assessed by scratch migration assay (scale bar, 100 μm). G, The mRNA levels of collagen I, α-SMA, CTGF and cyclin D1 were detected by qPCR. H, The protein levels of collagen I, α-SMA, CTGF and Cyclin D1 were analysed by Western blot. J, HASMCs were treated with thrombin (1 U/mL) for indicated times, and Western blot was performed to detect the protein levels of p-YAP and YAP. L, HASMCs were pre-treated with dabigatran (3000 ng/mL) for 1 h before stimulation with thrombin (1 U/mL) for 1 h, and Western blot was performed to detect the protein levels of p-YAP and YAP. N, HASMCs were subjected to immunofluorescence staining for YAP (red) to determine its subcellular localization and counterstained with DAPI (blue) to visualize the nuclei under a confocal microscope (scale bar, 50 μm). to Gαs, such as glucagon and epinephrine, suppresses YAP activity. [34][35][36] In addition, dynamic modulation of the F-actin cytoskeleton is required for hippo pathway regulation by GPCR ligands, 37,38 and it was reported that a key factor in stress fibre formation was Rho-GTPase-mediated phosphorylation of MLC2. 39 It is worth mentioning that Jung-Soon Mo et al confirmed that PAR1, a receptor for thrombin, regulated YAP dephosphorylation by Gα 12/13 in MDA-MB-231 F I G U R E 4 YAP mediated the effects of thrombin on HASMCs remodeling. A, HASMCs were pre-transfected with YAP siRNA (siYAP) or non-targeting siRNA (siNC) for 48 h, and the silencing effect was assessed by Western blot. HASMCs were pre-transfected with siYAP for 48 h or pre-treated with verteporfin for 1 h and followed by thrombin (1 U/mL) stimulation for 24 h. C, HASMCs proliferation was measured by EdU incorporation assay. HASMCs were stained with EdU (red) and counterstained with DAPI (blue) to visualize the nuclei (scale bar, 100 μm). D, HASMCs migration was assessed by scratch migration assay (scale bar, 100 μm). G, The mRNA levels of collagen I, α-SMA, CTGF and cyclin D1 were detected by qPCR. H, The protein levels of collagen I, α-SMA, CTGF and cyclin D1 were analysed by Western blot. (B), (E), (F) and (I) are the corresponding quantification histograms of (A), (C), (D) and (H), respectively. Data are presented as mean ± SEM of at least three independent experiments. *P < .05, **P < .01, ***P < .001, compared with the untreated cells or the cells treated with siNC; # P < .05, ## P < .01, ### P < .001, compared with the cells treated with thrombin F I G U R E 6 Dabigatran blocked thrombin-induced interaction between YAP and SMAD2. A, HASMCs were treated with thrombin (1 U/ mL) for indicated times, and Western blot was performed to detect the protein levels of p-SMAD2 and SMAD2. B, HASMCs were pretreated with dabigatran (3000 ng/mL) for 1 h before stimulation with thrombin (1 U/mL) for 1 h, and Western blot was performed to detect the protein levels p-SMAD2 and SMAD2. C and D, are the corresponding quantification histograms of (A) and (B). E, Subcellular localization of YAP and SMAD2 was determined by immunofluorescence. F and G, Co-immunoprecipitation of YAP with SMAD2. Cell lysates were subjected to immunoprecipitation with YAP antibody or control IgG. The co-immunoprecipitated SMAD2 was detected by Western blot. Scale bar, 50 μm. All data are presented as mean ± SEM of at least three independent experiments. **P < .01, compared with the untreated cells; # P < .05, compared with cells treated with thrombin | 8191 DENG Et al.
cells. 20 In this study, we first confirmed that thrombin induced YAP activation in HASMCs. Afterwards, PAR1 was identified to be the most abundant of the four PARs in HASMCs, and the inhibition of All data are presented as mean ± SEM of at least three independent experiments. *P < .05, **P < .01, ***P < .001, compared with untreated cells; # P < .05, ## P < .01, ### P < .001, compared with cells treated with thrombin F I G U R E 7 Proposed therapeutic mechanism of dabigatran in alleviating asthmatic airway smooth muscle remodeling applied separately, and thrombin-mediated YAP activation was significantly inhibited. It is intriguing that YAP activity is regulated by the conformation of the F-actin cytoskeleton, which in turn primarily depends on the substrate adhesion and intercellular tension, 40  accumulation of YAP and SMAD2/3. Interestingly, we detected that thrombin also induced rapid phosphorylation of SMAD2. Therefore, immunofluorescence was performed to detect colocalization of YAP with SMAD2, and CoIP to further verify their interaction. Our results demonstrated that YAP and SMAD2 binded to each other and traslocated to the nucleus after thrombin stimulation, while dabigatran reversed thrombin-induced these changes.
In conclusion, this study is the first to demonstrate that dabigatran reverses airway smooth muscle remodeling effectively in vivo and in vitro. Mechanistically, the protective effect of dabigatran might be associated with the suppression of thrombin-induced YAP activation. The proposed scheme is displayed in Figure 7.
Currently, the use of anti-inflammatory drugs and bronchodilators is the principle of conventional drug therapy for asthma.
Although these drugs alleviate the symptoms, they do not reverse the established changes in airway structure, especially airway remodeling. Our study may provide some evidence that dabigatran has the potential to relieve airway smooth muscle remodeling in asthmatic patients.

CO N FLI C T O F I NTE R E S T
The authors declare no conflicts of interest.

AUTH O R CO NTR I B UTI O N S
Yuanxiong Cheng was responsible for the study hypothesis and de-

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the first author upon reasonable request.