Verteporfin, an inhibitor of nuclear YAP, improved multi-ciliated cell differentiation in the airway epithelium Authors: Ryosuke Nakamura, Yo Kishimoto, Tomoko Kita, Susumu Yoshie, Shinji Kaba, Yukiko Okuno, Koichi Omori Published: 2025 ABSTRACT BackgroundImpaired multi-ciliated cell (MCC) differentiation, observed in various airway diseases, causes prolonged inflammation and further airway complications due to malfunction of the airway defense mechanism. YAP/TAZ-TEAD signaling mediates multiple mechanisms associated with MCC differentiation. However, YAP reportedly has both negative and positive roles in MCC differentiation, putatively due to its dual functions in the cytoplasm and nucleus. We hypothesize inhibition of nuclear YAP, while preserving cytoplasmic YAP functions, restores MCCs in various airway disorders. Here, we investigated the relationship between YAP nuclear localization and the presence of MCCs across patients. In addition, we identified compounds that promote MCC differentiation based on the inhibitory function of YAP nuclear localization.MethodTracheal specimens were collected from patients with different conditions (smoking, irradiation, tracheotomy, dysphagia), and YAP nuclear localization in MCCs was assessed by immunohistochemistry at areas with varying MCC compositions. To identify compounds that efficiently inhibit nuclear localization of YAP, airway epithelial cells were treated with multiple YAP inhibitors and YAP nuclear localization was assessed by immunofluorescence-based image analysis. The effects of the compounds on MCC differentiation were evaluated using air-liquid interface (ALI) culture and rat airway squamous metaplasia models.ResultsIn the human trachea, YAP nuclear localization in MCCs was negatively correlated with the proportion of MCCs in the epithelium, regardless of patient condition. Among known YAP inhibitors, verteporfin and atorvastatin efficiently inhibited YAP nuclear localization in cultured airway epithelial cells. Additionally, verteporfin promoted MCC differentiation in ALI culture and in vivo squamous metaplasia models.ConclusionOur findings suggest YAP nuclear localization in human airway MCCs is broadly associated with decreased MCCs regardless of patient condition, and verteporfin improves airway MCC differentiation.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12967-025-07250-3. Background Mucociliary clearance is a defense mechanism essential to maintain airway homeostasis [1, 2]. Multiciliated cells (MCCs), predominantly covering the airway epithelial surface, play a central role in this mechanism. Rapid ciliary strokes create mucus streams and excrete inhaled microorganisms. Impairment of this defense mechanism elevates the risk of infection in the respiratory tract [3, 4]. Malformation of cilia and the lack of MCCs are broadly associated with respiratory complications ranging from chronic inflammation to more severe diseases, such as airway stenosis and cystic fibrosis [5, 6]. Additionally, mucus congestion-induced dyspnea reduces quality of life, which is particularly problematic in the elderly [7]. Although the loss of MCCs associated with inflammatory damage is typically repaired by immediate regenerative responses [8], repetitive or severe damage compromises regeneration by the persistence of abnormal signaling events. Numerous signaling molecules reportedly disrupt MCC differentiation. Inflammatory cytokines, such as tumor necrosis factor-α (TNF-α) and interleukin 1b (IL1b), are associated with decreased MCCs [9, 10]. Goblet cell hyperplasia is induced by asthma as a result of prolonged Th2 immune response, most likely via IL4 and IL13 [11, 12]. In addition, smoking stimulates epidermal growth factor (EGF) and transforming growth factor-β (TGF-β) expression, and is associated with squamous metaplasia and goblet cell hyperplasia [13–16]. Wnt/β-catenin signaling is putatively activated in the airway of patients with chronic obstructive pulmonary disease (COPD) and idiopathic pulmonary fibrosis (IPF), which both are associated with decreased MCCs and impaired ciliogenesis [14, 17, 18]. Iatrogenic injuries also impair MCC differentiation [19]. Physical damage induces inflammatory genes; intubation and tracheostomy lead to squamous metaplasia [19–21]. In addition, tracheal reconstruction by end-to-end anastomosis and implantation of tracheal grafts result in sustained loss of MCCs at the anastomosis sites and in the middle of implants [22, 23]. Reduced MCCs increase the risk of infection, driving the cycle of inflammation and impaired mucociliary functions. Furthermore, prolonged or severe inflammation can lead to life-threatening airway stenosis, rejection of tracheal graft, and dissociation of anastomosis sites. Treatments to restore MCCs are likely to prevent further complications [1, 6, 18]. Given the variety of signaling events involved in impaired MCC differentiation, a therapeutic that prevents all anti-MCC differentiation signals would be ideal. The Hippo signaling pathway, initially considered as a key pathway involved in restricting organ growth, has recently emerged as an integrator of various cell signaling events triggered by different stimuli [24, 25]. The effector Yes-associated protein (YAP) and its paralogue transcriptional co-activator with PDZ binding motif (TAZ) fulfill a central role in the Hippo pathway. Activated YAP/TAZ enters the nucleus and mediates transcription primarily via transcriptional enhanced associate domain (TEAD) transcription factors. However, beyond the role in TEAD-dependent transcription, YAP/TAZ modulates various signaling pathways by cooperating with other transcription factors [25, 26]. For example, Notch- and SMAD2/3-mediated signaling pathways require concurrent YAP activation to induce trophectoderm lineage specification and myofibroblast differentiation [24, 27]. Interestingly, various signaling molecules known to disrupt MCC differentiation (e.g. EGF, TGF-β, TNF-α, β-catenin) activate YAP in non-airway tissue cells [15, 17, 26–31]. Therefore, YAP may be broadly involved in impaired MCC differentiation via mediating multiple signals across airway complications. Recent reports on airway epithelial YAP have suggested its negative role in MCC differentiation. Constitutive YAP activation induced hyper-stratification of airway epithelial cells and inhibited MCC differentiation [32, 33]. In addition, undifferentiated epithelial cells positive for nuclear YAP stratify in the airway of COPD patients [34]. Nuclear YAP is frequently induced in squamous epithelial cells of nasal polyps [35]. However, gene knockout and knockdown of YAP inhibited MCC differentiation in the developing murine airway and in vitro, suggesting YAP expression is indispensable for MCC differentiation [36]. These apparently contradictory findings may be explained by the positive and negative roles of cytoplasmic and nuclear YAP in regulating epithelial differentiation. Cytoplasmic YAP is transcriptionally inactive; however, the recruitment of YAP to cell junctions is associated with differentiation and maturation of epithelial cells [26]. To that end, we hypothesize inhibition of YAP nuclear localization restores MCCs across the spectrum of airway disorders. The current study aimed to evaluate a MCC restoration therapy based on pharmacological inhibition of nuclear YAP activity. First, we assessed the relationship between airway MCC differentiation and YAP nuclear localization in patients with different conditions: Smoking, dysphagia, irradiation, and tracheotomy. Subsequently, we identified compounds that efficiently inhibit YAP nuclear localization in airway epithelial cells and evaluated their effects on differentiation in vitro. Finally, we employed a rat tracheal reconstruction model to confirm improved MCC differentiation and mucociliary function with the compounds. This study provides foundational data regarding intervening on nuclear YAP as a strategy to restore MCCs and mucociliary clearance. Materials and methods Human samples Pieces of cervical tracheae were obtained from 20 patients who underwent laryngectomy for the treatment of malignant tumors in the head-and-neck region (the pharynx, larynx, esophagus, and tongue), dysphagia, and subglottic stenosis in 2017 and 2018. The ages of the patients ranged from 34 to 85 years. Before surgery, six and seven patients received cervical radiation within 11 years and tracheotomy within 3 years, respectively. Five patients reported dysphagia. Thirteen patients had a history of smoking, but a patient who quit smoking approximately 30 years before the laryngectomy was excluded from the smoker group; 12 patients were classified as smokers. Written informed consent was obtained from all patients prior to the surgery. This study was conducted under authorization by the Ethics Committee of the Kyoto University Graduate School and Faculty of Medicine (R1204-2). All procedures were performed in accordance with the Kyoto University Hospital Institutional Guidelines. Inhibitory compounds Forskolin, atorvastatin calcium trihydrate, pitavastatin calcium, pravastatin sodium, fluvastatin sodium (n-hydrate), and lovastatin were purchased from FUJIFILM Wako Pure Chemical (Osaka, Japan). Mevastatin, simvastatin, ivermectin, verteporfin, dobutamine, and epinephrine were obtained from Millipore-Sigma (Burlington, MA, USA). Dasatinib was obtained from Cell Signaling Technology (Danvers, MA, USA). Dimethyl sulfoxide was used as a vehicle. Powder and solutions of verteporfin, as well as cells cultured in verteporfin-containing media, were maintained in the dark with the following exceptions: Solutions containing verteporfin were exposed to ~ 77 lx light for ~ 37 s during the procedure of intra-tracheal administration, and to ~ 3.7 lx light during the preparation of verteporfin solutions and replacement of cell culture media. Cell lines An immortalized rat airway epithelial cell line, EGV-4T, was provided by the Japanese Collection of Research Bioresources (Tokyo, Japan). Cells were maintained in a 1:1 mixture of Dulbecco’s modified Eagle’s medium and Ham’s F-12 nutrient mixture (DMEM/F-12; Gibco, Waltham, MA, USA) supplemented with GlutaMAX (Gibco), 10% fetal bovine serum (FBS; Cytiva, Marlborough, MA), and antibiotics (Nacalai Tesque, Kyoto, Japan) at 37 °C, in a 5% CO2 environment. EGV-4T cells expanded to P22 were employed for immunofluorescence analysis and luciferase reporter assay, as described below. Normal human airway epithelial cells from three donors were purchased from Lonza (Basel, Switzerland). The cells were seeded onto dishes coated with 500 µg/mL type I collagen (Cellmatrix Type I-A; Nitta Gelatin, Osaka, Japan), and maintained in an airway epithelial cell growth medium (AEC-GM): DMEM/F-12 supplemented with GlutaMAX, 10 µg/mL recombinant human insulin (Nacalai Tesque), 5 µg/mL transferrin (FUJIFILM Wako Pure Chemical), 0.5 µM o-phosphorylethanolamine (Nacalai Tesque), 0.5 µM monoethanolamine (Nacalai Tesque), 0.2 µM hydrocortisone (Millipore-Sigma, Burlington, MA), 30 µg/mL bovine pituitary extract (Gibco), 20 ng/ml recombinant human EGF (Gibco), 50 nM retinoic acid (Wako Pure Chemical), 10 nM 3,3′,5-triiodo-L-thyronine sodium salt (Millipore-Sigma), and 1% FBS. Normal human airway epithelial cells were employed to analyze gene expression related to YAP/TAZ-TEAD signaling and evaluate the effects of reportedly YAP-inhibitory drugs on cell death and differentiation as described below. Animals Male Sprague-Dawley rats, aged 9 weeks, were obtained from Shimizu Laboratory Supplies (Kyoto, Japan). All animal experimental protocols were pre-approved by the Animal Care and Use Committee of Kyoto University (Med Kyo 19131), and conducted in accordance with the Institutional Guidelines for Animal Experiments, the Japanese Government Law Concerning the Protection and Control of Animals, and the Japanese Government Notification on Feeding and Safekeeping of Animals. Animal care was provided by the Institute of Laboratory Animals of Kyoto University. All experiments were performed at Kyoto University. At least 5 days of acclimation was implemented before surgical procedures. A total of 123 rats were used in this study. Tracheal reconstruction was performed on all 123 rats: Compounds or PBS were administered to 111 rats 7 days post-reconstruction; nothing was administered to 12 rats, as detailed below. Groups of 8 rats each were treated with PBS; 0.3, 1, 3, or 10 µM atorvastatin; or 3, 10, 30, or 100 µM verteporfin, and euthanized 7 days post-administration. Regeneration of the ciliated epithelium was observed via scanning electron microscopy. Additionally, groups of 13 rats each received PBS, 3 µM atorvastatin, or 100 µM verteporfin and were euthanized 7 days later. Five rats from each group were assessed via immunohistochemistry, and the remaining rats were subjected to analysis of mucociliary transport. In the untreated group, three rats each were euthanized at 1, 3, 7, or 14 days post-reconstruction and assessed via immunohistochemistry. Group allocation was managed by RN, and the administration order of PBS and compounds at varying concentrations was randomized daily. Reverse-transcriptase polymerase chain reaction (RT-PCR) Total RNA was isolated from human airway epithelial cells using TRIzol Reagent (Invitrogen, Waltham, MA, USA), and reverse-transcribed using the PrimeScrip High Fidelity RT-PCR Kit (Takara Bio, Shiga, Japan). Specific DNA segments within YAP1, WWTR1, TEAD1‒4, and GAPDH cDNAs were amplified by polymerase chain reaction (PCR) with Ex Taq DNA Polymerase (Takara Bio); primers are listed in Supplemental Table S1. Following 35 cycles of amplification, PCR products were fractionated through agarose gel electrophoresis, and stained with ethidium bromide. ChemiDoc Gel Imaging System (Bio-Rad Laboratories, Hercules, CA) was employed to detect the amplified DNA segments (see Table 1). Table 1Characteristics of donors of tracheal samples for immunohistochemistryParticipants20Age (mean ± SD)68.6 ± 14.1Female7 (35%)Smoker12 (60%)History of tracheotomy7 (35%)History of irradiation6 (30%)Dysphagia5 (25%) Table 1Characteristics of donors of tracheal samples for immunohistochemistryParticipants20Age (mean ± SD)68.6 ± 14.1Female7 (35%)Smoker12 (60%)History of tracheotomy7 (35%)History of irradiation6 (30%)Dysphagia5 (25%) Characteristics of donors of tracheal samples for immunohistochemistry Immunohistochemistry of human and rat trachea Human and rat tracheal specimens were fixed with 4% paraformaldehyde. Following dehydration using ethanol, the samples were embedded in paraffin and sectioned at 5 μm. Tissue sections were deparaffinized using xylene and ethanol. Following antigen retrieval, specimens were incubated with phosphate-buffered saline (PBS, pH 7.4) containing Tween-20 (TPBS) and 1% bovine serum albumin for 1 h to block non-specific antibody binding. Primary and secondary antibodies listed in Table S2 were employed for immunostaining. To detect horse radish peroxidase (HRP)-conjugated secondary antibodies, specimens were incubated with 3,3′-diaminobenzidine or fluorescein-labeled tyramide (PerkinElmer, Waltham, MA, USA) in the presence of hydrogen peroxide. Cell nuclei were counterstained with Mayer’s hematoxylin solution or 4′,6-diamidino-2-phenylindole (DAPI; PromoCell, Heidelberg, Germany). BX-51 (Olympus, Tokyo, Japan) and BZ-X710 (Keyence, Osaka, Japan) were employed to obtain bright-field and fluorescent images. To analyze fluorescently labeled sections of human tracheal samples, images of the epithelium were obtained in as many microscopic fields as possible. The numbers of p63+, FOXJ1+, and MUC5AC+ cells with or without nuclear YAP staining (NucYAP), and total epithelial cells (tECs) were counted in each microscopic field. The ratios of NucYAP+FOXJ1+/total FOXJ1+, NucYAP+MUC5AC+/total MUC5AC+, NucYAP+p63+/total p63+ cells were calculated. The rank correlation of these ratios vs. the ratios of FOXJ1+/tECs, MUC5AC+/tECs, and p63+/tECs was analyzed by Spearman’s method. Luciferase reporter assay EGV-4T cells were transfected with the 8xGTIIC-luciferase plasmid vector (Addgene, Watertown, MA, USA) using Lipofectamine3000 (Invitrogen) [37], and seeded onto 96-well plates. Cells were cultured in DMEM/F-12 supplemented with 1% FBS overnight, and then treated with the experimental compounds for 24 h. Cells were rinsed with ice-cold PBS. Luciferase activity was measured using Luciferase Reporter Assay Kit I (PromoCell) and a multimode plate reader, SPARK (Tecan, Männedorf, Switzerland). Immunofluorescence image analysis of YAP nuclear localization EGV-4T cells were cultured in 96-well plates. Following overnight serum starvation with 1% FBS-containing DMEM/F-12, cells were exposed to YAP inhibitors for 24 h. Cells were fixed with 4% paraformaldehyde for 10 min and subsequently permeabilized using 1% Triton X-100 for 10 min. After blocking with TPBS containing 1% bovine serum albumin, cells were reacted with an anti-YAP antibody and subsequently a secondary antibody (Table S2). Cells were rinsed with TPBS after each antibody reaction. Nuclei were counterstained using DAPI. Images were captured and analyzed using a high-content image analysis system, ArrayScan VTI (Thermo Fisher Scientific, Waltham, MA, USA). Nuclear and peri-nuclear regions were respectively defined as a DAPI-stained area and a line two pixels away from the outline of the DAPI-stained area. The difference in average fluorescence intensities between the nuclear and peri-nuclear regions was measured in individual cells to evaluate the nuclear localization of YAP. Air-liquid interface culture and analysis of airway epithelial cell differentiation Human airway epithelial cells were expanded in cell culture inserts (0.4-µm pore size; Corning, Corning, NY, USA) until confluence. ALI culture was employed to induce MCC differentiation, with AEC-GM supplied to the bottom of wells. A half volume of medium was replaced every day. YAP inhibitors were supplemented 10 days following the initiation of ALI culture, and the media containing YAP inhibitors were replaced with simple AEC-GM by a half volume every day. Concentrations of YAP inhibitors theoretically decreased to 2−X of the initial concentrations at day X after the supplementation of YAP inhibitors. Cells were fixed with 4% paraformaldehyde 20 days after the initiation of ALI culture. MCCs and goblet cells were detected by immunofluorescence using primary antibodies against acetylated tubulin and MUC5AC, and secondary antibodies (Table S2). Microscopic observation and image analysis were performed using a BZ-X710 microscope (Keyence, Osaka, Japan). Acetylated tubulin- and MUC5AC-positive cells in a 3,768 × 2,826 μm area were counted to evaluate MCC and goblet cell differentiation. Cytotoxicity assay Conditioned media were obtained from the ALI culture of human airway epithelial cells 24 h after incubation with drugs. The activity of lactate dehydrogenase (LDH) released into the media was measured using the LDH cytotoxicity detection kit (Takara Bio, Shiga, Japan). Media were also obtained from cells incubated with 1% Triton X-100 as a positive control. LDH activity was expressed as a percentage relative to the positive control. Tracheal reconstruction and compound administration All surgical procedures were performed under general anesthesia induced via intraperitoneal injection of butorphanol tartrate (2.4 mg/kg; Meiji Seika Pharma, Tokyo, Japan), midazolam (1.9 mg/kg; Astellas Pharma, Tokyo, Japan), and medetomidine (0.14 mg/kg; Nippon Zenyaku Kogyo, Fukushima, Japan). A polypropylene mesh covered with a spongious collagen scaffold was implanted in rat tracheal defects as described previously [23]. Briefly, a partial tracheal defect 2 mm in width and spanning 2 tracheal rings was created using a scalpel, and covered with the tracheal graft. One week after implantation, rats were anesthetized and the tracheae were exposed. Fifty µL PBS +/- compounds was administered to the luminal side of the reconstructed area using a curved 27-gauge needle. Rats were euthanized 1 week after drug administration, via exposure to CO2 gas. Scanning electron microscopy Regenerated epithelium in the implant was observed by scanning electron microscopy and ciliated regions were quantified as described previously [23]. Briefly, tracheal samples were fixed with 2.5% glutaraldehyde and subsequently with 1% osmium tetroxide (Wako). Following dehydration with ethanol and immersion in tert-butyl alcohol, samples were freeze-dried and coated with platinum ions. The epithelium was observed using a field emission scanning electron microscope (SU8220; Hitachi High-Technologies, Tokyo, Japan). Areas covered by cilia were quantified based on dark appearances of cilia in low-magnification images. The region of interest (ROI) was set at 600 × 600 μm in the center of the reconstructed area. Areas with darker appearances were determined as ciliated regions using the auto-thresholding function of ImageJ. Analysis of mucociliary transport Mucociliary transport was analyzed as described previously with some modifications [38]. Briefly, fluorescent microspheres (Fluoresbrite YO Carboxylate Microspheres 0.20 μm, Polysciences, Warrington, PA, USA) were spread onto the tracheal samples. Movement of microspheres was recorded at 50 frames/s using a high-speed camera (FASTCAM mini UX50; Photron, Tokyo, Japan) and an upright microscope (BX51; Olympus, Tokyo, Japan). Video footage was captured in a 220 × 170 μm region, 240-µm proximal from the center of the reconstructed part. Movement of microspheres in 6 ROIs (54 × 42 μm) set within the captured video was analyzed using the DIAtrack software (downloaded from http://www.diatrack.org/), and mean microsphere speed was estimated from the 6-ROI data for each sample. Data analysis Spearman’s rank correlation coefficients were determined for immunofluorescence data of human tracheal samples using RStudio software (https://github.com/rstudio/rstudio). Other statistical analyses, described below, were conducted using SPSS (IBM, Armonk, NY). Dunnett’s test was employed to analyze differences in YAP nuclear localization between EGV-4T cells treated with the aforementioned concentrations of drugs and untreated cells. Student’s t-test was used to compare luciferase reporter assay data. Kruskal-Wallis test and post-hoc pairwise comparison with Bonferroni correction were performed for the data obtained from ALI culture and animal experiments. Differences were considered statistically significant at p < 0.05. Results Expression of YAP/TAZ and TEADs in human airway epithelial cells We initially assessed expression of YAP/TAZ and TEADs in human airway epithelial cells by conventional RT-PCR and immunohistochemistry. Human airway epithelial cells from three donors (provided by Lonza; Basel, Switzerland) were cultured for RT-PCR. All three donors were positive for YAP/TAZ and TEAD1-4 gene expression (Fig. 1a). Figure 1b shows representative immunohistochemical images of tracheal specimens from 20 patients who underwent tracheotomy at Kyoto University Hospital. Airway epithelial cells were positive for commercially available antibodies against YAP/TAZ, and TEAD1, 3 and 4 in the epithelium with the typical morphology of a pseudostratified columnar ciliated epithelium (Fig. 1b). TEAD2 expression was unclear based on the immunohistochemical examination. Fig. 1YAP/TAZ and TEAD expression in human airway epithelial cells. Gene expression of YAP (encoded by YAP1), TAZ (encoded by WWTR1), and TEAD1–4 in cultured normal human airway epithelial cells were assessed by RT-PCR (a). Results of three donors (#1–3) are shown. YAP, TAZ, and TEAD1–4 in the pseudostratified ciliated epithelium were detected on paraffin-embedded sections of human tracheae (b). Scale bar: 50 μm. Representative images are shown YAP/TAZ and TEAD expression in human airway epithelial cells. Gene expression of YAP (encoded by YAP1), TAZ (encoded by WWTR1), and TEAD1–4 in cultured normal human airway epithelial cells were assessed by RT-PCR (a). Results of three donors (#1–3) are shown. YAP, TAZ, and TEAD1–4 in the pseudostratified ciliated epithelium were detected on paraffin-embedded sections of human tracheae (b). Scale bar: 50 μm. Representative images are shown YAP localization negatively correlates with the quantity of MCCs in human airway epithelium To assess the relationship between YAP nuclear localization and complicated airway epithelial differentiation, we interrogated the differences in YAP distribution across areas with differential epithelial cell compositions. Three major populations in the airway epithelium were identified based on the expression of the following markers: FOXJ1, an initiator of early-stage differentiation of MCCs and continuously expressed in mature MCCs; MUC5AC, a component of mucus expressed by goblet cells; p63, a marker of undifferentiated stem/progenitor cells [39–41]. Proportions of FOXJ1+ MCCs and MUC5AC+ goblet cells in the epithelium varied across areas within each patient, and some FOXJ1+ MCCs and MUC5AC+ goblet cells were nuclear YAP-positive (NucYAP+) (Fig. 2a, b). Most p63+ undifferentiated cells were NucYAP+ in the epithelia with a pseudostratified morphology (Fig. 2c) and squamous metaplasia (Fig. 2d). We measured the proportion of NucYAP+ cells among FOXJ1+, MUC5AC+ and p63+ cells in individual microscopic fields in the tracheal specimens of 20 patients. Initially, differences related to underlying conditions were disregarded to focus on YAP distribution in areas with varied epithelial cell populations. The ratio of NucYAP+FOXJ1+ cells to total FOXJ1+ cells considerably varied across the observation areas (Fig. 2e). Interestingly, the NucYAP+FOXJ1+ cell ratio to total FOXJ1+ cells was higher in areas where the ratio of FOXJ1+/total epithelial cells (tECs) was low compared to areas where the FOXJ1+/tEC ratio was high (Fig. 2f). A negative correlation was observed between the NucYAP+FOXJ1+/total FOXJ1+ cell and FOXJ1+/tEC ratios. The ratios of NucYAP+MUC5AC+/total MUC5AC+ and NucYAP+p63+/total p63+ cells were similar at low and high levels across multiple microscopic fields, regardless of the ratios of MUC5AC+ and p63+ cells to tECs (Fig. 2e, g, h). Fig. 2YAP nuclear localization in different cell populations in human airway epithelium. Paraffin-embedded sections of the human tracheal epithelium were immunostained to detect FOXJ1, MUC5AC, p63, and YAP (a–d). Representative images of epithelia with abundant and sparse FOXJ1 + cells (a); sparse and abundant MUC5AC+ cells (b); pseudostratified and stratified squamous epithelia (c, d). Arrows indicate nuclear YAP in FOXJ1+ and MUC5AC+ cells (a, b). Scale bar: 50 μm. The violin plot and scatter plots illustrate the distributions of the ratios of nuclear YAP-positive (NucYAP+) cells (%) in FOXJ1+, MUC5AC+, and p63+ cells (counts obtained from multiple microscopic fields of specimens from 20 patients) (e–l). Ratios of NucYAP+FOXJ1+/FOXJ1+ cells, FOXJ1+/total epithelial cells, NucYAP+MUC5AC+/MUC5AC+ cells, MUC5AC+/total epithelial cells, NucYAP+p63+/p63+ cells, and p63+/total epithelial cells were measured in each microscopic field. In the scatter plots, the x-axis represents the ratios of FOXJ1+/total epithelial cells (f, i–l), MUC5AC+/total epithelial cells (g), and p63+ cells/total epithelial cells (h). Data from different cohorts divided based on the conditions of irradiation, smoking, tracheotomy, and dysphagia were plotted as different colors (i–l). Spearman’s rank correlation coefficient (Rs) is noted in the scatter plots (*p < 0.05) YAP nuclear localization in different cell populations in human airway epithelium. Paraffin-embedded sections of the human tracheal epithelium were immunostained to detect FOXJ1, MUC5AC, p63, and YAP (a–d). Representative images of epithelia with abundant and sparse FOXJ1 + cells (a); sparse and abundant MUC5AC+ cells (b); pseudostratified and stratified squamous epithelia (c, d). Arrows indicate nuclear YAP in FOXJ1+ and MUC5AC+ cells (a, b). Scale bar: 50 μm. The violin plot and scatter plots illustrate the distributions of the ratios of nuclear YAP-positive (NucYAP+) cells (%) in FOXJ1+, MUC5AC+, and p63+ cells (counts obtained from multiple microscopic fields of specimens from 20 patients) (e–l). Ratios of NucYAP+FOXJ1+/FOXJ1+ cells, FOXJ1+/total epithelial cells, NucYAP+MUC5AC+/MUC5AC+ cells, MUC5AC+/total epithelial cells, NucYAP+p63+/p63+ cells, and p63+/total epithelial cells were measured in each microscopic field. In the scatter plots, the x-axis represents the ratios of FOXJ1+/total epithelial cells (f, i–l), MUC5AC+/total epithelial cells (g), and p63+ cells/total epithelial cells (h). Data from different cohorts divided based on the conditions of irradiation, smoking, tracheotomy, and dysphagia were plotted as different colors (i–l). Spearman’s rank correlation coefficient (Rs) is noted in the scatter plots (*p < 0.05) Subsequently, the same analysis was performed for cohorts based on underlying conditions (irradiation, smoking, tracheotomy, and dysphagia). Similar to data obtained from all patients, a negative correlation was observed between the ratios of NucYAP+FOXJ1+/total FOXJ1+ cells and FOXJ1+/tEC across cohorts (Fig. 2i–l); this negative correlation is unlikely affected by those conditions. In addition, low and high proportions of NucYAP+ cells among MUC5AC+ and p63+ cells were consistently observed in all cohorts (Supplemental Figure S1). Based on these observations, increased nuclear localization of YAP in MCCs may be a broadly shared phenomenon related to decreased MCCs. Evaluation of YAP inhibitors Various pharmaceutical and natural bioactive compounds are known to inhibit YAP. Most of these compounds indirectly inhibit YAP by targeting signaling pathways associated with YAP activity. Agonists of adrenergic receptors (epinephrine and dobutamine) [42, 43], activators of adenyl cyclase (forskolin) [42], anti-cancer drugs with Src-inhibitory function (dasatinib) [44, 45], and inhibitors of HMG-CoA reductase (statins) [46], indirectly inhibit nuclear localization of YAP. Regarding mechanisms of forskolin and statins, these drugs inhibit nuclear YAP activity by attenuating Rho signaling downstream of their effects on primary targets [42, 46, 47]. Contrastingly, verteporfin interferes with YAP-TEAD interaction by directly binding to YAP [48]. To identify drugs that efficiently inhibit nuclear localization of YAP in airway epithelial cells, the distribution of YAP in immortalized airway epithelial cells was evaluated by immunofluorescence-based image analysis after 24-hour incubation with these drugs. Among the drugs assessed, forskolin, statins, and verteporfin reduced nuclear localization of YAP in a concentration-dependent manner (Fig. 3a, b). Dasatinib led to the detachment of airway epithelial cells from the cell culture substrate (Supplemental Figure S2). Epinephrine and dobutamine did not clearly reduce nuclear YAP in airway epithelial cells (Supplemental Figure S2). Fig. 3Inhibitory effects of forskolin, statins, and verteporfin on YAP nuclear localization in airway epithelial cells. EGV-4T cells were incubated with forskolin, atorvastatin, fluvastatin, lovastatin, or verteporfin for 24 h. YAP and nuclei in the cells were visualized using the anti-YAP antibody and DAPI, and fluorescent images were analyzed using ArrayScan VTI (a, b). Representative images of cells treated with 20 µM forskolin, 10 µM atorvastatin, 2 µg/mL fluvastatin sodium n-hydrate, 5 µM lovastatin, and 5 µM verteporfin are shown (a). Scale bar: 100 μm. Levels of nuclear YAP in cells treated with different concentrations of the drugs were compared with untreated control cells, and relative nuclear YAP levels are shown as bar graphs (b; mean ± SD; n = 5). *P < 0.05 (vs. control). Luciferase activities in EGV-4T cells expressing 8xGTIIC-luciferase were measured following 24-hour exposure to 20 µM forskolin, 10 µM atorvastatin, 2 µg/mL fluvastatin sodium n-hydrate, 5 µM lovastatin, or 5 µM verteporfin (c). Relative luciferase activities, with the activity level in control cells determined as 1, are shown as bar graphs (c; mean ± SD; n = 5) *P < 0.05 (vs. control). Representative data from independently performed triplicate experiments are shown (b, c) Inhibitory effects of forskolin, statins, and verteporfin on YAP nuclear localization in airway epithelial cells. EGV-4T cells were incubated with forskolin, atorvastatin, fluvastatin, lovastatin, or verteporfin for 24 h. YAP and nuclei in the cells were visualized using the anti-YAP antibody and DAPI, and fluorescent images were analyzed using ArrayScan VTI (a, b). Representative images of cells treated with 20 µM forskolin, 10 µM atorvastatin, 2 µg/mL fluvastatin sodium n-hydrate, 5 µM lovastatin, and 5 µM verteporfin are shown (a). Scale bar: 100 μm. Levels of nuclear YAP in cells treated with different concentrations of the drugs were compared with untreated control cells, and relative nuclear YAP levels are shown as bar graphs (b; mean ± SD; n = 5). *P < 0.05 (vs. control). Luciferase activities in EGV-4T cells expressing 8xGTIIC-luciferase were measured following 24-hour exposure to 20 µM forskolin, 10 µM atorvastatin, 2 µg/mL fluvastatin sodium n-hydrate, 5 µM lovastatin, or 5 µM verteporfin (c). Relative luciferase activities, with the activity level in control cells determined as 1, are shown as bar graphs (c; mean ± SD; n = 5) *P < 0.05 (vs. control). Representative data from independently performed triplicate experiments are shown (b, c) The inhibitory effects of forskolin, statins, and verteporin on transcription via TEAD-binding elements were assessed by luciferase reporter assay. Consistent with immunofluorescence, forskolin, statins, and verteporfin reduced luciferase activity (Fig. 3c). Among the drugs employed in this study, verteporfin most efficiently inhibited nuclear YAP localization and transcription of the reporter gene in these assay systems. Forskolin, atorvastatin, and verteporfin promoted MCC differentiation in vitro Human airway epithelial cells were treated with forskolin, atorvastatin, and verteporfin in ALI culture, to evaluate their cytotoxicity and effects on MCC differentiation in vitro. To establish a challenging condition for MCC differentiation, we added EGF to the cell culture medium. EGF signaling is known to activate YAP in hepatocellular carcinoma and renal epithelial cells and inhibit airway MCC differentiation [15, 30, 49]. The drugs were supplemented to the medium 10 days after the initiation of ALI culture. Half of the medium was replaced daily with a fresh medium, to gradually reduce drug concentrations. Cytotoxicity was not observed in the ALI-cultured cells that underwent 24-hour treatment with ~ 30 µM forskolin, ~ 10 µM atorvastatin, or ~ 10 µM verteporfin (Fig. 4a). However, forskolin and atorvastatin, which inhibit Rho signaling [42, 46, 47], induced ‘pores’ in the cell layer at 30 and 3 µM (Fig. 4b), making it difficult to maintain ALI culture. A similar effect has been reported with 5 µM atorvastatin [50]. With regard to MCC differentiation, the three drugs all increased acetylated α-tubulin-positive MCCs at certain concentrations 10 days after administration (Fig. 4c, d). However, concentrations of forskolin and atorvastatin required to increase MCCs were ~ 3 and ~ 1 µM; at these concentrations, forskolin and atorvastatin showed only weak inhibitory effects on YAP nuclear localization in the immunofluorescence assay (Fig. 3b). The discrepancy in the effective concentrations for YAP inhibition and MCC differentiation may reflect the involvement of the primary functions of forskolin and atorvastatin. In contrast, the concentration of verteporfin required to increase MCCs was 3 µM, close to the concentration that most efficiently inhibited YAP nuclear localization (Fig. 3b). Verteporfin and atorvastatin did not significantly promote goblet cell differentiation, likely beneficial for averting goblet cell hyperplasia (Fig. 4e). Fig. 4Cytotoxicity and pro-differentiation effects of forskolin, atorvastatin, and verteporfin in human airway epithelial cells. Normal human airway epithelial cells in ALI culture were incubated with 1–30 µM forskolin, 0.3–10 µM atorvastatin, or 0.3–10 µM verteporfin for 24 h. The activity of LDH released into the media was measured. LDH activity in cells killed by exposure to 1% Triton X-100 was determined as 100% cell death. The rates of cell death are shown as box plots (a; n = 6). Representative data from three donors. No statistical difference was observed among the groups. The phase contrast microscopic images show untreated control, 30 µM forskolin-treated, and 3 µM atorvastatin-treated cells in ALI culture (b). Pores were observed in the cell layer within 2 days of 30 µM forskolin and 3 µM atorvastatin treatment. Cells at day 10 of ALI culture were treated with gradually decreasing concentrations of forskolin, atorvastatin, or verteporfin for 10 days, and subsequently immunostained with anti-acetylated tubulin and anti-MUC5AC antibodies (c–e). Representative images of cells treated with the drugs (c); initial concentrations of forskolin, atorvastatin, and verteporfin were 3, 1, and 3 µM, respectively. Green: acetylated tubulin. Red: MUC5AC. Blue: DAPI. Scale bar: 1 mm. Numbers of acetylated tubulin-positive MCCs (d) and MUC5AC-positive goblet cells (e) in a 3,768 × 2,826 μm area are shown as box plots. N = 6. *P < 0.05 (vs. control). Representative data from three donors Cytotoxicity and pro-differentiation effects of forskolin, atorvastatin, and verteporfin in human airway epithelial cells. Normal human airway epithelial cells in ALI culture were incubated with 1–30 µM forskolin, 0.3–10 µM atorvastatin, or 0.3–10 µM verteporfin for 24 h. The activity of LDH released into the media was measured. LDH activity in cells killed by exposure to 1% Triton X-100 was determined as 100% cell death. The rates of cell death are shown as box plots (a; n = 6). Representative data from three donors. No statistical difference was observed among the groups. The phase contrast microscopic images show untreated control, 30 µM forskolin-treated, and 3 µM atorvastatin-treated cells in ALI culture (b). Pores were observed in the cell layer within 2 days of 30 µM forskolin and 3 µM atorvastatin treatment. Cells at day 10 of ALI culture were treated with gradually decreasing concentrations of forskolin, atorvastatin, or verteporfin for 10 days, and subsequently immunostained with anti-acetylated tubulin and anti-MUC5AC antibodies (c–e). Representative images of cells treated with the drugs (c); initial concentrations of forskolin, atorvastatin, and verteporfin were 3, 1, and 3 µM, respectively. Green: acetylated tubulin. Red: MUC5AC. Blue: DAPI. Scale bar: 1 mm. Numbers of acetylated tubulin-positive MCCs (d) and MUC5AC-positive goblet cells (e) in a 3,768 × 2,826 μm area are shown as box plots. N = 6. *P < 0.05 (vs. control). Representative data from three donors YAP activation during airway epithelial regeneration We employed a rat tracheal reconstruction model with collagen graft implantation to induce squamous metaplasia as a pilot preclinical animal study [23]. Before evaluating the effects of drugs, we assessed changes in the distribution of YAP during airway epithelial regeneration in this model. In the normal rat trachea, faint nuclear and cytoplasmic YAP-positive staining was observed in some epithelial cells (Fig. 5). Different from the human trachea, intense nuclear YAP-positive staining was rarely observed in basal cells of the rat trachea. After tracheal reconstruction using the collagen implant, epithelial cells penetrated the implant to regenerate the epithelium and fully covered the implant within 7 days. Intense YAP-positive staining was observed in the nuclei and cytoplasm of epithelial cells with a flattened morphology 1–7 days after implantation. Phosphorylation at Ser112 of YAP was reduced in the penetrating flattened epithelial cells at days 1 and 3, likely reflecting the partial inhibitory effect of Ser112-phosphorylation on YAP nuclear localization [51]. Following formation of flattened immature epithelium, MCC differentiation occurred by day 14. However, the epithelium in the middle of the implant was only partially ciliated at day 14, and still showed relatively intense YAP-positive staining in both the nucleus and cytoplasm compared with the normal region. Fig. 5Immunolocalization of YAP in the regeneration process of rat airway epithelium after tracheal reconstruction. A collagen-based tracheal graft was implanted to reconstruct partial tracheal defects in rats. Paraffin-embedded sections were immunostained with antibodies against pan-cytokeratin, YAP, and Ser112-phosphorylated YAP (S112-pYAP) at days 1, 3, 7, and 14 after implantation. Antibody reaction was detected with DAB staining, and nuclei were counterstained with a hematoxylin solution. Scale bar = 100 μm Immunolocalization of YAP in the regeneration process of rat airway epithelium after tracheal reconstruction. A collagen-based tracheal graft was implanted to reconstruct partial tracheal defects in rats. Paraffin-embedded sections were immunostained with antibodies against pan-cytokeratin, YAP, and Ser112-phosphorylated YAP (S112-pYAP) at days 1, 3, 7, and 14 after implantation. Antibody reaction was detected with DAB staining, and nuclei were counterstained with a hematoxylin solution. Scale bar = 100 μm Atorvastatin and verteporfin increased MCCs in reconstructed rat trachea Verteporfin and atorvastatin were administered onto the luminal surface of the implant at day 7 to evaluate their efficacy in increasing MCCs through the differentiation process from immature epithelial coverage. In the control group, a large part of the graft was covered with the squamous epithelium 7 days after vehicle administration (Fig. 6, Supplemental Figure S3). Verteporfin significantly increased ciliated areas at 30 and 100 µM (Fig. 6a, b). Atorvastatin only increased ciliated areas at 3 µM, implying a narrow therapeutic window. As previously reported in this animal model [52], no MUC5AC+ goblet cells were observed in the middle of the graft in the control group 14 days after implantation (Fig. 6c). Neither atorvastatin nor verteporfin increased MUC5AC+ cells. Mucociliary transportation, assessed by tracking fluorescent microspheres, was improved with 3 µM atorvastatin and 100 µM verteporfin, confirming the formation of functional cilia (Fig. 7, Movie 1). Fig. 6Increased MCCs in reconstructed rat trachea following atorvastatin and verteporfin treatment. Defected rat tracheae were reconstructed by implanting a collagen-based tracheal graft. Atorvastatin and verteporfin were administered to the luminal surface 7 days after implantation. The epithelium was inspected by scanning electron microscopy 7 days after drug administration (a). High magnification images are also shown for red rectangle areas. Bars = 500 μm (low magnification), 50 μm (middle magnification), 10 μm (high magnification). The proportions of ciliated areas in the images were quantified, with 600 × 600 μm ROI being set in the middle of the implant, and are represented as box plots (b). N = 8. *P < 0.05 (vs. control). MUC5AC-positive goblet cells (green) and acetylated tubulin-positive MCCs (red) were detected on longitudinal cryosections obtained from the implants and normal trachea, by immunofluorescence microscopy 7 days after drug administration (c). Scale bar: 100 μm Increased MCCs in reconstructed rat trachea following atorvastatin and verteporfin treatment. Defected rat tracheae were reconstructed by implanting a collagen-based tracheal graft. Atorvastatin and verteporfin were administered to the luminal surface 7 days after implantation. The epithelium was inspected by scanning electron microscopy 7 days after drug administration (a). High magnification images are also shown for red rectangle areas. Bars = 500 μm (low magnification), 50 μm (middle magnification), 10 μm (high magnification). The proportions of ciliated areas in the images were quantified, with 600 × 600 μm ROI being set in the middle of the implant, and are represented as box plots (b). N = 8. *P < 0.05 (vs. control). MUC5AC-positive goblet cells (green) and acetylated tubulin-positive MCCs (red) were detected on longitudinal cryosections obtained from the implants and normal trachea, by immunofluorescence microscopy 7 days after drug administration (c). Scale bar: 100 μm Fig. 7Mucociliary transport of airway epithelium regenerated in tracheal graft. Tracheal grafts were collected 7 days after drug administration. Fluorescent microspheres were spread to the epithelium, and the movement of microspheres was tracked using the DIA track software in combination with video microscopy. Trajectories of the microspheres are shown (a, white lines). Scale bar: 20 μm. The speeds of microsphere transport in control, 3 µM atorvastatin, and 100 µM verteporfin treatment groups are represented in box plots (b). N = 8. *P < 0.05 (vs. control) Mucociliary transport of airway epithelium regenerated in tracheal graft. Tracheal grafts were collected 7 days after drug administration. Fluorescent microspheres were spread to the epithelium, and the movement of microspheres was tracked using the DIA track software in combination with video microscopy. Trajectories of the microspheres are shown (a, white lines). Scale bar: 20 μm. The speeds of microsphere transport in control, 3 µM atorvastatin, and 100 µM verteporfin treatment groups are represented in box plots (b). N = 8. *P < 0.05 (vs. control) Discussion The airway is continuously exposed to stress at the interface between the external environment and body. The lack of MCCs drives a cycle of inflammation and additionally reduced MCCs, and is associated with chronicity and severity of airway complications. In the current study, we sought to identify drugs that restore MCCs based on the hypothesis that inhibition of nuclear YAP improves MCC differentiation in a spectrum of airway complications. Our data implies verteporfin is an efficient tool to inhibit nuclear YAP in airway epithelial cells and increases MCCs. The role of nuclear YAP in undifferentiated airway epithelial cells has been investigated extensively. Stratification of nuclear YAP-positive epithelial cells is observed in squamous metaplasia in nasal polyps and the airways of patients with COPD [34, 35]. Nuclear YAP-positive cells increase in naphthalene-injured murine airway epithelia, and rapidly form epithelial coverage [34]. Additionally, constitutive activation of YAP nuclear localization via gene modification leads to hyper-stratification of undifferentiated airway epithelial cells and YAP null mutation significantly reduces basal cells in the airway epithelium [32]. Therefore, YAP nuclear localization is fundamental for the proliferation of airway epithelial stem/progenitor cells and retention of their undifferentiated state. Consistently, p63+ cells in our clinical samples had a high positivity rate for nuclear YAP regardless of patient condition. The collagen-implanted region in our animal model was covered with nuclear YAP-positive undifferentiated cells during the early phase of epithelial recovery. Regarding airway MCCs, YAP distribution reportedly shifts from the nucleus to the cytoplasm and cell junctions during early-stage differentiation in well-controlled experimental models [53]. Contrastingly, the environment in the human airway varies across patients. Various signaling molecules are upregulated in airway complications [15, 17, 26–30]. In our human samples, nuclear YAP was not exclusively observed in undifferentiated airway epithelial cells. In addition, correlation between nuclear YAP-positivity and decreased MCCs was observed across all patients. These findings likely support our hypothesis that intervening on nuclear YAP is a reasonable strategy to restore MCC populations for a spectrum of airway diseases. The importance of cytoplasmic YAP has indirectly been suggested by the inhibitory effect of nuclear YAP on MCC differentiation. For example, genetically replacing YAP by 5SA-YAP (constitutively active YAP) in the developing airway epithelium significantly impaired MCC differentiation [33]. YAP activation driven by depletion of MST1/2 led to decreased MCCs [54]. Mechanistically, YAP sequestered in the cytoplasm and junctional protein complexes are thought to be prevented from nuclear translocation [26, 55]. Beyond this passive role, cytoplasmic YAP inhibited Wnt/β-catenin signaling as a component of the β-catenin destruction complex in HEK293 cells [56]. A recent study using EpH4 cells suggested YAP incorporated in the apical Crumbs complex sequester SMAD2/3 and is involved in tolerance to TGF-β signaling [57]. Since aberrant activation of canonical Wnt and TGF-β signaling is associated with impaired MCC differentiation [17, 31, 58], hypothetical mechanistic roles of non-active YAP in MCC differentiation would be mutual restriction of such signal transduction molecules and YAP. However, actual involvement of these mechanisms in differentiation and maintenance of MCCs is unknown. Moreover, gene transcription mediated by nuclear YAP in FOXJ1+ airway epithelial cells and its role in differentiation and maintenance of MCCs remain unclear. It should also be noted that nuclear YAP co-condensed with transcription factors can inhibit transcription via negative feedback [59]. Because of the dynamics of YAP-mediated gene transcription, the impact of YAP on gene transcription is not necessarily related to nuclear localization of YAP. Further investigations are warranted to clarify the roles of nuclear YAP in the differentiation and maintenance of airway MCCs. Various pharmaceutical and natural compounds have been found to exert indirect inhibitory effects on YAP, because of the multifaceted interplay of YAP/TAZ-TEAD with other signaling pathways. Among the indirect YAP inhibitors employed in the current study, statins and forskolin inhibited nuclear localization of YAP in airway epithelial cells. This finding implies the involvement of the mevalonate pathway and cAMP signaling in the regulation of YAP in airway epithelial cells, through downstream effects on Rho signaling. However, other known indirect YAP inhibitors did not clearly reduce nuclear YAP localization in airway epithelial cells, likely due to differential physiology across cell types. Dasatinib, of which YAP-inhibiting effect had been identified in cancer cell lines [44, 45], was cytotoxic to airway epithelial cells. Dobutamine and epinephrine did not clearly inhibit nuclear YAP localization in the current study, implying adrenergic receptor signaling do not contribute to YAP inhibition in airway epithelial cells [42, 43]. Verteporfin, a direct inhibitor of the YAP-TEAD interaction, most efficiently inhibited nuclear localization of YAP in our assay system. Concentrations of verteporfin required for the YAP inhibitory effect and to promote in vitro MCC differentiation were similar. However, concentrations of atorvastatin and forskolin required to efficiently inhibit YAP nuclear localization and promote MCC differentiation were quite different. These findings suggest the effects of these indirect YAP inhibitors on MCC differentiation rely mainly on YAP-independent mechanisms. In addition, forskolin had an unpreferable effect regarding goblet cell differentiation in the ALI culture. Differences in primary target molecules likely altered the outcomes of goblet cell differentiation. Statins have controversial roles in ciliogenesis and MCC differentiation. Their anti-inflammatory actions potentially support MCC differentiation, whereas cholesterol depletion disrupts cilia formation [50, 60]. The narrow-range concentration for atorvastatin to increase MCCs, observed in our in vivo experiment, might reflect difficulty balancing the mixed effects of statins. Increasing nuclear YAP/TAZ via gene depletion of MST1/2 and LATS1/2 decreases goblet cells [54, 61], suggesting nuclear YAP/TAZ inhibits not only MCC differentiation, but also goblet cell differentiation. However, depletion of YAP/TAZ dramatically increases goblet cells [61]. Null mutation of YAP increases goblet cells, but decreases MCCs in the developing airway [36]. Based on these reports, cytoplasmic YAP/TAZ likely regulate goblet cell and MCC differentiation differently. Our findings using ALI culture might reflect this putative differential regulation of YAP/TAZ, since the MCC increase via verteporfin treatment did not concur with increased goblet cell differentiation. Regarding such a biased effect on differentiation, Notch signaling has been reported to selectively induce goblet cell differentiation and Notch inhibition is a promising therapeutic strategy to prevent goblet cell hyperplasia [62]. YAP/TAZ-TEAD signaling, due to its role in integrating other signaling, is capable of modulating Notch signaling. For example, Cdx2 transcription requires co-operation between the Notch intracellular domain and a YAP-TEAD complex in the nuclei of trophectoderm cells [63, 64]. However, further studies are required to determine the contribution of YAP/TAZ-TEAD pathway to Notch signaling in the differentiation of airway epithelial cells. Our in vivo data confirmed the effect of verteporfin on increasing MCCs in the reconstructed rat trachea. In this animal model, the epithelium is physically removed and squamous metaplasia occurs in the reconstructed area [23]. Our findings imply verteporfin is a promising drug to restore MCCs in squamous metaplasia. However, data from only one animal model have limited implications. Further studies using different models are required to confirm whether inhibition of YAP nuclear localization increases MCCs in a broad range of conditions. A significant limitation of the current study is related to the route of drug administration, in addition to the aforementioned limitations (i.e., the necessity of clarifying mechanistic roles of YAP and validating the effect of YAP inhibition using multiple models). Although the local administration in the current animal study helped understand the effects of drugs by directly delivering drugs to sites lacking MCCs, this method is clinically infeasible. Drugs likely need to be formulated as inhalants or for nebulization. Conclusion Nuclear localization of YAP in human airway MCCs is associated with decreased MCCs. This finding was consistent across patients with conditions of smoking, dysphagia, irradiation, and tracheotomy. Verteporfin promoted MCC differentiation without significantly increasing goblet cell differentiation in both in vitro and in vivo models. These preliminary data are foundational for YAP-based therapeutic approaches to restore MCCs. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1 Supplementary Material 2 Supplementary Material 2 Supplementary Material 3 Supplementary Material 3 Supplementary Material 4 Supplementary Material 4