Acetylcholine From Solitary Chemosensory Cell, Not Neuron, Regulates Basal Cell Fate Driving Eosinophilic Chronic Rhinosinusitis With Nasal Polyps

ABSTRACT

Background

Basal cells (BCs) play a crucial role in epithelial remodeling, a hallmark of eosinophilic chronic rhinosinusitis with nasal polyps (eCRSwNP). Single‐cell sequencing has revealed an increased number of solitary chemosensory cells (SCCs) alongside BC hyperplasia in eCRSwNP, yet the underlying mechanism of BC hyperplasia in eCRSwNP remains unclear. This study aimed to investigate the role of SCC‐derived acetylcholine (Ach) in determining BC fate.

Methods

Tissue samples from healthy individuals, patients with eCRSwNP, and those with non‐eCRSwNP (neCRSwNP) were analyzed to investigate BC proliferation, differentiation abnormalities, and the prevalence of SCCs. The relationship between SCC‐derived Ach, BC dysfunction, and disease severity was examined. Ach sources and receptor expression patterns were characterized. In vitro studies using submerged cell cultures and air–liquid interface cultures, along with in vivo murine models, were employed to elucidate the mechanisms by which Ach influences BC fate. The inhibitory effects of tiotropium bromide (TB) on Ach‐driven processes were also evaluated.

Results

Our results indicated that SCC‐derived Ach, rather than by parasympathetic nerves, contributed to abnormal BC proliferation and differentiation through muscarinic acetylcholine receptors (mAChRs) and had potential impact on the development of eCRSwNP. These effects were associated with the activation of YAP and could be partially reversed both in vitro and in vivo by blocking mAChRs with TB.

Conclusion

These results demonstrate that SCC‐derived Ach plays a critical role in eCRSwNP by regulating BC fate. This provides a potential translational framework for developing prevention and treatment strategies targeting the cholinergic pathway.

This study identifies that Ach levels are elevated in eCRSwNP and primarily released from SCCs rather than neurons. Ach induces epithelial remodeling in eCRSwNP by binding to M1R/M3R and activating YAP signaling in basal cells, driving basal cell hyperplasia and abnormal differentiation. Blocking M1R/M3R with TB inhibits eCRSwNP progression. Ach, acetylcholine; DB, denatonium benzoate; eCRSwNP, eosinophilic chronic rhinosinusitis with nasal polyps; M1R, muscarinic acetylcholine receptor M1; M3R, muscarinic acetylcholine receptor M3; neCRSwNP, non‐eCRSwNP; NP, nasal polyp; OVA, ovalbumin; PBS, phosphate buffered saline; SCC, solitary chemosensory cell; SEB, Staphylococcus aureus enterotoxin B; TB, tiotropium bromide; YAP, yes‐associated protein. The workflow graphs were generated using BioRender.

1. Introduction

Chronic rhinosinusitis (CRS), which affects over 10% of adults globally [1], is a chronic inflammatory disorder of the sinonasal mucosa. CRS is dichotomized into two phenotypes: CRS with nasal polyps (CRSwNP) and CRS without nasal polyps (CRSsNP). With a recurrence rate of over 90% after surgery, eosinophilic CRSwNP (eCRSwNP) is considered the “difficult to treat” type [2, 3, 4, 5, 6]. In our previous studies, we observed profound epithelial remodeling in eCRSwNP [7, 8] that several other studies attributed to the imbalance between epithelial barrier damage and repair. Currently, the high recurrence rate of eCRSwNP is attributed to this imbalance [4, 5, 9]. Thus, it is urgent to identify the underlying mechanism responsible for causing epithelial cells to lose their ability to maintain a repair balance.

Located at the base of the pseudostratified airway epithelium, basal cells (BCs) are essential progenitors in epithelial barrier repair. Upon barrier damage, BCs can differentiate into ciliated cells (CCs), secretory cells, and other types of epithelial cells to restore the epithelium and protect against pathogens and microbial invasion [10, 11, 12]. In eCRSwNP, several studies reported that type 2 (T2) inflammation alters the fate of BCs, promoting BC proliferation ability while impairing differentiation ability. This BC fate alteration drives epithelial remodeling, compromises the epithelial barrier, and increases susceptibility to pathogen exposure and bacterial invasion from [6, 13, 14, 15, 16]. However, clinical trials targeting T2 inflammation with anti‐IL‐4Rα or anti‐IL‐5 monoclonal antibodies have shown that approximately 30%–50% of patients with severe eCRSwNP achieved an ≥ 2‐point improvement in their endoscopic nasal polyp (NP) score, indicating the existence of additional mechanisms beyond T2 inflammation in determining BC ability differentiation from [2, 6, 17, 18].

Acetylcholine (Ach), the classic neurotransmitter of cholinergic neurons, is known to drive airway epithelial remodeling in lower airway diseases such as asthma and chronic obstructive pulmonary disease (COPD), primarily through smooth muscle contraction [19, 20, 21]. However, the upper airway lacks smooth muscle and primarily consists of epithelial cells [22]. Recent studies have reported that Ach can directly regulate intestinal epithelial stem cell proliferation and differentiation through muscarinic acetylcholine receptors (mAchRs) [23, 24, 25]. Given the abundant expression of mAchRs in the nasal mucosa [26, 27], it is plausible that Ach might regulate BC fate in eCRSwNP, contributing to epithelial remodeling. Furthermore, NPs have traditionally been considered devoid of neural innervation [28, 29, 30]. Thus, the source of this Ach becomes a critical question for investigation.

Solitary chemosensory cells (SCCs), also known as tuft cells, are rare secretory epithelial cells that play a crucial role in coordinating immune and neural functions within the mucosal epithelium [31, 32, 33]. In CRSwNP, SCCs increase within the epithelium, and this increase is thought to be linked to disease progression [34, 35]. Under allergen stimulation, altered BC fate appears to be associated with SCCs rather than T2 inflammatory factors [36]. Recent studies have found that choline acetyltransferase (ChAT), the key enzyme in Ach synthesis, is also the protein marker of SCCs [35, 37, 38, 39, 40], indicating that SCCs are likely the primary source of Ach in eCRSwNP. Thus, investigating the role of SCC‐derived Ach in eCRSwNP, particularly its paracrine effects on epithelial remodeling, is crucial for understanding its impact on disease progression.

In this study, we explored the mechanism of Ach derived from SCC in regulating BC fate in eCRSwNP. Our findings indicate that SCC‐derived Ach induces BC hyperplasia—reflected by an increased BC frequency—and prosoplasia through mAChR by promoting YAP nuclear translocation, thus contributing to epithelial remodeling in eCRSwNP. These results enrich the mechanism of non‐neurogenic cholinergic transmitters in epithelial remodeling and provide the clinical translational basis for the prevention and treatment of eCRSwNP by targeting the cholinergic pathway.

2. Materials and Methods

2.1. Single‐Cell RNA Sequencing Data Analysis

The single‐cell RNA sequencing (scRNA‐seq) data was sourced from the HRA000772 dataset [41] and processed through the 10× Genomics Cell Ranger platform to form a corresponding matrix, which was further analyzed using the Seurat package (version 4.4.0). Cells were identified as low‐quality and removed when their mitochondrial percentage was no less than 15% and nfeatures was less than 200 or more than 6000. The data was normalized with NormalizeData and further integrated with Harmony. The FindVariableFeatures function was applied to identify cell markers, and the FindNeighbors and FindClusters functions were performed to cluster a total of 104,949 cells with a resolution of 0.5. Cell‐type signatures were established based on marker lists from published single‐cell sequencing studies [12, 42, 43, 44]. Differentially expressed genes (DEGs) analysis of BCs was performed across the three groups with the edgeR package, followed by GSEA and KEGG pathway analysis. Subsequently, eCRSwNP samples were divided into high (ChAT^high) and low (ChAT^low) expression groups based on ChAT levels. The DEGs analysis between these two groups identified 160 upregulated and 393 downregulated genes. GO and KEGG pathway analyses on the differentially expressed BC genes were then conducted. Interactions with p < 0.05 were considered significant. The analysis and data mining were performed on Dr. Tom Multi‐omics Data mining system (https://biosys.bgi.com).

2.2. Bulk RNA Sequencing for Human Biospecimens

For whole transcriptome sequencing, the BGl Optimal Dual‐Module mRNA Library Preparation Kit (BGI, Shenzhen, China) was used. Following the manufacturer's protocol, a specified amount of an RNA sample was heat‐denatured to open secondary structures, and the mRNA was enriched using oligo (dT) magnetic beads. The enriched mRNA was fragmented under the optimal temperature, then reverse‐transcribed into first‐strand cDNA using random hexamer primers, followed by the synthesis of the second cDNA strand. Next, the reaction mixture was prepared, the ends of the double‐stranded cDNA were repaired, and an A base was added to the 3′ end. Adapter ligation was conducted, followed by product amplification and quality control. The PCR product was denatured into a single strand, and a circularization reaction was established to obtain single‐strand circular products, with any remaining linear DNA molecules digested to yield the final library. This library was amplified via phi29 to create DNA nanoballs (DNBs), which were loaded onto a patterned nanoarray and sequenced on a G400/T7/T10 platform (BGI) using PE100/PE150 reads with combinatorial probe‐anchor synthesis (CPAS) technology. The sequencing data were filtered using SOAPnuke [45], and subsequent analysis was performed using the Dr. Tom Multi‐omics Data Mining System (https://biosys.bgi.com). Fold change (FC) values were calculated to assess differences in gene expression levels.

2.3. Study Population

Nasal tissue samples were collected with informed consent and Ethics Committee approval from the First Affiliated Hospital of Chongqing Medical University (2022‐K301). Biopsy specimens were obtained from 71 patients with eCRSwNP, 29 patients with non‐eCRSwNP (neCRSwNP), and 34 healthy controls, all from the hospital's Department of Otolaryngology. The patient characteristics are presented in Table S1. CRSwNP was diagnosed according to the 2020 European Position Paper on Rhinosinusitis and Nasal Polyps (EPOS) criteria. None of the CRSwNP patients or controls had concurrent upper respiratory infections, systemic diseases, or a history of nasal surgery, and none had taken macrolide antibiotics, corticosteroids, antibiotics, or T2 biologics in the 3 months preceding the study. Nasal polyp tissues were collected during endoscopic sinus surgery, while uncinate mucosa samples were obtained from control subjects undergoing nasal septoplasty. eCRSwNP was defined as an average histological eosinophil count of ≥ 10 per high‐power field (×400) on hematoxylin–eosin (H&E) staining and a JESREC score ≥ 11 [46].

2.4. Epithelial Cell Submerged Culture

After the resected polyp tissues from the patients with eCRSwNP were digested, the digested samples were transferred into 6 cm dishes and cultured in PneumaCult‐Ex medium (05040, Stemcell Technologies, Vancouver, BC, Canada), with medium changes every 2 days. After the cells reached 70%–80% confluence, they were stimulated for 24 h with Ach chloride (5 mM, HY‐P70445, MCE, Monmouth Junction, NJ, USA) and/or tiotropium bromide (TB, 100 nM, HY‐17360, MCE, Monmouth Junction, NJ, USA), following a 24‐h incubation period in medium without supplements or steroids to ensure cell starvation.

2.5. Air‐Liquid Interface Cultures

To create Air‐liquid interface (ALI) cultures, 100,000 cultured epithelial cells per well were added to 0.4 μm pore 24‐well polyester membrane inserts (3412, Corning, Corning, NY, USA) with Pneumacult‐Ex media (05008, Stemcell Technologies, Vancouver, BC, Canada) on both sides of the membrane. Apical media was removed completely, and basal media was changed to Pneumacult‐ALI (05001, Stemcell Technologies, Vancouver, BC, Canada) supplemented with 5 mL 100× penicillin–streptomycin (PB180120, Pricella, Wuhan, China), 1 mL 0.2% heparin sodium salt (07980, Stemcell Technologies, Vancouver, BC, Canada) in D‐PBS (37,350, Stemcell Technologies, Vancouver, BC, Canada), and 2.5 mL 200× hydrocortisone stock solution (07925, Stemcell Technologies, Vancouver, BC, Canada). Depending on the experimental condition, the cultures were additionally treated with PBS, 10 μM α‐NETA (HY‐138097, MCE, Monmouth Junction, NJ, USA), 5 ng/mL IL‐4 (HY‐P70445, MCE) with or without 10 μM α‐NETA, 5 ng/mL IL‐5 (HY‐P70612, MCE) with or without 10 μM α‐NETA, 5 ng/mL IL‐13 (HY‐P70568, MCE) with or without 10 μM α‐NETA, or 5 mM acetylcholine chloride (HY‐P70445, MCE) with or without 100 nM tiotropium bromide (TB, HY‐17360, MCE). Basal media was changed every 2 days for 21 days. The morphology of the cultures was assessed using light microscopy. Subsequently, transwell inserts were embedded in paraffin according to the manufacturer's instructions.

2.6. Murine NP Model

All animal experimentation protocols received approval from the Institutional Animal Care and Use Committee (IACUC‐CQMU‐2023‐0249) and adhered to both governmental and international guidelines for animal research. Wild‐type 4‐week‐old BALB/c mice were obtained from the Experimental Animal Center at Chongqing Medical University, where they were kept under specific pathogen‐free conditions. The NP murine model was constructed using ovalbumin (OVA) (A5503, Sigma‐Aldrich, St. Louis, MO, USA) and Staphylococcus aureus enterotoxin B (SEB) (HY‐P1900, MCE, Monmouth Junction, NJ, USA) [47]. Briefly, on days 0 and 5, mice were injected intraperitoneally with 25 μg of OVA dissolved in 200 μL of PBS containing 2 mg of aluminum hydroxide. From day 12 to 19, 40 μL of 6% OVA dissolved in PBS was instilled daily by nasal infusion. Thereafter, mice received 40 μL of 6% OVA via nasal infusion three times per week for 4 weeks. Subsequently, mice were challenged three times per week with 10 ng of SEB in addition to a nasal infusion of 40 μL of 6% OVA three times a week for 8 weeks. The control group was treated with PBS alone via nasal infusion. The experimental groups were treated with 6% OVA, SEB, and PBS or 10 mM of denatonium benzoate (DB) or 0.1 mM of TB in PBS three times a week on days 47–103 via nasal infusion. The detailed schedule is shown in Figure S2C and Figure 6A. Twenty‐four hours after the last administration, the mice were sacrificed by intraperitoneal injection of sodium pentobarbital. After partial tracheal resection under deep anesthesia, a micropipette was inserted into the choana through the tracheal opening in the direction of the upper airway. Each nasal cavity was gently perfused with 200 μL of PBS, and the fluid from the nostril was collected and centrifuged. The supernatants were stored at −80°C. The tissue was dehydrated, processed according to standard paraffin‐embedding procedures, and cut into coronal sections.

FIGURE 6.

TB inhibits NP formation in vivo. (A) A schematic diagram of the murine modeling process is shown. (B) H&E staining images of the murine nasal cavity, highlighting classic polypoid lesions (asterisks) with enlarged areas (black box). (C) Quantification of polyp‐like lesions across murine groups (n = 6 per group). (D, E) Representative sirius red staining images of eosinophils (D, orange box) and quantification of eosinophilic infiltration (E, n = 6 per group) across murine groups. (F, G) H&E staining of the murine nasal mucosal epithelium (blue box) was performed (F), and quantitative analysis of murine nasal cavity epithelial thickness (G, n = 6 per group). (H) PAS staining showing in each murine group (red box). (I) Quantification of GC (n = 6 per group) in each murine group. (J) IF staining for MUC5AC (red) in each murine group. (K, L) The percentage area (K) and average OD (L) of MUC5AC‐positive staining in each murine group (n = 6 per group). (M) IF staining for α‐tubulin (red) in each murine group. (N‐O) The percentage area (N) and average OD (O) covered by α‐tubulin‐positive staining in each murine group (n = 6 per group). (P) IF staining for P63 (red) and Ki67 (green) in each murine group. (Q) Quantification of P63⁺ cells in each murine group (n = 6 per group). (R) The percentage of P63⁺Ki67⁺ cells among P63⁺ cells in each murine group (n = 6 per group). (S) IF staining of P63 (red) and YAP (green) in each murine group. (T) The percentage of P63⁺YAP⁺ cells among P63⁺ cells in each murine group (n = 6 per group). Data are presented as mean ± SEM; *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001; ns, no significance.

2.7. Ach Assessment

The collected cell supernatants, basolateral fluids from ALI cultures, and mice nasal lavage fluids were analyzed for Ach (SMK1525A, Sumeike, Jiangsu, China; SMK1525B, Sumeike, Jiangsu, China).

2.8. Histopathologic Analysis

Sections of mice mucosal tissues and transwell inserts were analyzed histopathologically. H&E staining was performed to assess polypoid lesions and epithelial thickness in both types of samples. Sirius red staining (G3632, Solarbio, Beijing, China) was performed to evaluate eosinophils in mice mucosal tissues, and periodic acid‐Schiff (PAS) staining (Servicebio, Hubei, China) was performed to determine the number of GCs in both sections. Polypoid lesions were identified by observation of distinct mucosal bulges, eosinophilic infiltration, and microcavity formation [47]. To ensure accuracy and verify the presence of these lesions, two consecutive sections were examined to minimize processing errors. Quantification of epithelial thickness and PAS‐positive goblet cells was conducted in five randomly selected fields per slide at ×400 magnification.

2.9. Cell Proliferation Assay

Cell proliferation was assessed by performing CCK‐8 analysis (C0038, Beyotime, Jiangsu, China). Human nasal epithelial cells were collected and seeded in 96‐well plates at a density of 5000 cells per well. After a 24 h starvation period in an unsupplemented medium, the cells were stimulated for 24 h with Ach chloride (5 mM, HY‐P70445, MCE, Monmouth Junction, NJ, USA) and/or TB (100 nM, HY‐17360, MCE, Monmouth Junction, NJ, USA). Following stimulation, CCK‐8 solution was added at 24, 48, and 72 h. The cells were then incubated for 1 h, and absorbance was measured at a wavelength of 450 nm.

2.10. Immunohistochemistry

Nasal biopsy specimens were fixed in formalin, embedded in paraffin, and sectioned to a thickness of 6 μm using a Leica microtome. The paraffin sections were then deparaffinized and hydrated through a series of graded alcohols. For antigen retrieval, the sections were heated in a microwave oven at 100°C for 10 min in a pH 6 retrieval buffer (BL604A, Biosharp, Hefei, China). Endogenous peroxidase activity was inhibited using Endogenous Peroxidase Blocking Buffer (PV‐9000, Labgic, Beijing, China) at room temperature for 10 min. Non‐specific immunoreactivity was blocked with 10% goat serum (C0265, Beyotime, Shanghai, China) for 30 min at room temperature. The slides were then incubated overnight at 4°C with an anti‐ChAT primary antibody (ab181023, Abcam, Cambridge, UK). The following day, the sections were washed thoroughly with PBS, and detection was conducted using a horseradish peroxidase‐conjugated secondary antibody. Finally, the sections were counterstained with hematoxylin, dehydrated with ethanol and xylene, mounted with neutral resin, and observed under a light microscope. Image analysis was performed using ImageJ software. Immunohistochemical quantification was conducted across five fields per slide at ×400 magnification.

2.11. Immunofluorescence Staining

After the paraffin sections of nasal biopsy specimens, ALI cultures, and mice tissues were deparaffinized and rehydrated, they underwent antigen retrieval as outlined in the immunohistochemistry section. Submerged cell cultures were seeded onto confocal dishes and permeabilized with 0.1% TritonX‐100 for 10 min at room temperature. After treatment, the paraffin sections, ALI culture sections, and cells cultured on 0.5 mL confocal dishes (BS‐15‐GJM‐B, Biosharp, Hefei, China) were blocked with 10% goat serum (C0265, Beyotime, Shanghai, China) for 30 min at room temperature. The sections and cells were then incubated overnight at 4°C with primary antibodies: anti‐Ki67 (AF02778, AiFang Biological, Hunan, China), anti‐P63 (ab124762, Abcam), anti‐β3 tubulin (ab18207, Abcam), anti‐CHAT (ab181023, Abcam), anti‐GNAT3 (LS‐C166554, LSBio, Lynnwood, WA, USA; PA5‐50670, Thermo Fisher Scientific, Waltham, MA, USA), anti‐acetyl‐α‐tubulin (5335 T, CST, Danvers, MA, USA), anti‐MUC5AC (AF08487, AiFang Biological, Hunan, China), FOXJ1 (14–9965‐82, Thermo Fisher Scientific, Waltham, MA, USA), Involucrin (PA5‐32454, Thermo Fisher Scientific, Waltham, MA, USA) and anti‐YAP (14,074T, CST, Danvers, MA, USA). The following day, the sections and cells were thoroughly washed with PBS and detected using a horseradish peroxidase‐conjugated secondary antibody. Lastly, tyramide signal amplification (TAS) dual‐label triple‐color multiplex fluorescent staining kits (AFIHC023, AiFang Biological, Hunan, China) were employed per the manufacturer's instructions, and the slides and cells were mounted using antifade reagent with DAPI (C1006, Beyotime) for fluorescence imaging. Image analysis was performed using ImageJ software. Immunofluorescence quantification was performed in five randomly selected fields per slide at ×400 magnification.

2.12. RNA Extraction and Real‐Time Quantitative PCR

Following the manufacturer's instructions, RNA was extracted from tissue samples using the SteadyPure RNA Extraction Kit (AG21024, AG, Hunan, China). The RNA was then converted to cDNA with the RT Master Mix for qPCR II (HY‐K0511A, MCE, Monmouth Junction, NJ, USA). The mRNA levels of all markers were quantified using SYBR Green gene expression assays (HY‐K0523, MCE, Monmouth Junction, NJ, USA). Relative gene expression was analyzed with the CFX Connect Real‐Time PCR Detection System (Bio‐Rad, Hercules, CA, USA) employing the $2^{-\mathrm{\Delta }{C}_{\mathrm{T}}}$ method and using GAPDH as the internal control. Detailed primer sequences for quantitative reverse transcription polymerase chain reaction (qRT‐PCR) can be found in Table S2.

2.13. Western Blot Analysis

Cells were treated with radioimmunoprecipitation assay lysis solution containing protease inhibitors (1:100) and incubated on ice for 30 min. After centrifugation at 4°C (12,000 rpm) for 10 min, the supernatant was collected for concentration determination. Denatured protein samples were loaded onto a 10% polyacrylamide gel (PG112, Epizime, Cambridge, MA, USA), and SDS‐PAGE was conducted at 90 and 120 V for 30 and 60 min, respectively. Following electrophoresis, the gel was cut, and proteins were transferred to a polyvinylidene fluoride membrane (Merck Millipore, Burlington, MA, USA) at 250 mA for 70 min. The membrane was then blocked with 5% skim milk buffer for 40 min and incubated overnight at 4°C with anti‐YAP primary antibody (14,074 T, CST, Danvers, MA, USA). The next day, the membrane was incubated with secondary antibodies (SA00001‐2, Proteintech, Rosemont, IL, USA) at room temperature for 1 h. Membranes were washed three times with PBS, and blots were visualized using enhanced chemiluminescence western blotting reagent. Images of the blots were captured, and band densities were analyzed using ImageJ software, with the ratio of each band to GAPDH considered the expression level of the target protein.

2.14. Statistical Analyses

All data were analyzed using GraphPad Prism 9 and presented as the mean ± standard error of the mean (SEM). Normality and chi‐square tests were conducted for all data. For normally distributed data with equal variances, one‐way analysis of variance (ANOVA) was applied for group comparisons. Otherwise, the Kruskal–Wallis H‐test was used to evaluate relationships among variables, with p < 0.05 deemed statistically significant.

3. Results

3.1. BC Hyperplasia Contributes to Epithelial Remodeling in eCRSwNP

To explore the specific characteristics of epithelial remodeling in eCRSwNP, we re‐analyzed the scRNA‐seq dataset reported by Wang et al. [dataset] [41]. After integrating data from healthy controls (n = 5) and patients with neCRSwNP (n = 5) or eCRSwNP (n = 6), we identified a total of 104,949 cells, including 59,983 immune cells and 44,966 non‐immune cells (Figure S1A,B and Tables S3 and S4). Similarly, we also found a significant shifting of cell types, increased numbers of immune cells with decreased numbers of epithelial cells, in patients with eCRSwNP compared with healthy controls (Figure S1C and Table S3). Given that epithelial remodeling played a vital role in eCRSwNP, we further focused on epithelial cell alterations. A total of 31,488 epithelial cells (EPCAM and KRT19) were identified and then subclustered into 22 subsets in 8 major groups, according to cell marker expression, which was previously defined by Wang et al. (Figure 1A and Table S5). We found a significant alteration in epithelial cell composition, a decreased number of CCs, and an increased number of BCs and goblet cells (GCs) in the eCRSwNP group compared with the control group (Figure 1B,C and Table S6). Previous studies suggested that BC hyperplasia decreases the differentiation ability of BCs and contributes to epithelial remodeling in NPs [12, 41].

FIGURE 1.

Epithelial remodeling is a prominent feature of eCRSwNP. (A) T‐sne plot of 31,488 single epithelial cells from control, neCRSwNP and eCRSwNP groups through scRNA‐seq. (B) The fraction of epithelial cells in control, neCRSwNP and eCRSwNP group through scRNA‐seq data. (C) Histogram showing relative mRNA expression of basal cell (BC) marker TP63 and KRT5 in total cell. (D) GSEA analysis of the differentially expressed genes. (E) Histogram showing Ki67 expression in BC through scRNA‐seq data. (F) IF staining of BC marker P63 (green) in nasal mucosa from healthy control and nasal polyp from patients with eCRSwNP. (G) The numbers of P63⁺ cells (n = 8 per group). (H) IF staining of CC marker α‐tubulin (red) in nasal mucosa from healthy control and nasal polyp from patients with neCRSwNP and eCRSwNP. (I) The numbers of CCs (n = 5 per group). (J) IF staining of GC marker MUC5AC (red) in nasal mucosa from healthy control and nasal polyp from patients with neCRSwNP and eCRSwNP. (K) The numbers of GCs (n = 5 per group). (L) IF staining of P63 (red) and Ki67 (green) in nasal mucosa from healthy control and nasal polyp from patients with neCRSwNP and eCRSwNP. (M) The percentage of P63⁺Ki67⁺ cells among total P63⁺ cells (n = 8 per group). Data are presented as mean ± SEM; *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001; ns, no significance.

When we performed differentially expressed genes (DEGs) analysis of BCs in the three groups to examine this suggestion, we identified significant upregulation of proliferation‐related transcription factors, including JUN, JUNB, FOSB, FOSL1, FOSL2, and MYC in the eCRSwNP group (Figure S1D and Table S7). Gene ontology (GO) enrichment and gene set enrichment analysis (GSEA) demonstrated that these DEGs were significantly associated with the epithelial cell proliferation pathway and the cell cycle (Figure S1E and Figure 1D). More importantly, Ki67, a cell proliferation marker, was significantly higher in BCs from the eCRSwNP group compared with the normal control group (Figure 1E).

Given that scRNA‐seq [dataset] was insufficient to construct the landscape of eCRSwNP, in addition, squamous metaplasia has been reported as a prominent feature of epithelial remodeling in CRSwNP [48, 49]. We performed immunofluorescence (IF) staining to support our observations. We applied P63, α‐tubulin, MUC5AC, and involucrin (IVL) to track BCs, CCs, GCs, and squamous epithelial cells (SCs), respectively. We identified a significantly increased number of BCs (Figure 1F,G), GCs (Figure 1H,I), and SCs (Figure S6A,B), and a decreased number of CCs (Figure 1J,K) in the eCRSwNP group.

To examine whether this epithelium composition alteration was induced by BC hyperplasia, we stained the colocalization of BC marker P63 (red) and Ki67 (green). We observed a notably higher percentage of Ki67⁺BC in the eCRSwNP group compared with the neCRSwNP and control groups (Figure 1L,M). These results confirm the existence of remarkable epithelial remodeling in eCRSwNP that is significantly correlated with BC hyperplasia. Thus, it is important to investigate the underlying mechanism of BC fate regulation.

3.2. SCC Is Correlated With BC Hyperplasia and Promotes eCRSwNP

Previous studies have reported that SCCs, rare secretory cells important in mediating the host–microbe interaction, are expanded in NPs [31, 32, 34, 35]. Our bulk RNA‐seq and scRNA‐seq [dataset] results revealed elevated expressions of SCC markers (POU2F3, TRPM5, and GNAT3) in the eCRSwNP group compared with the neCRSwNP and control groups (Figure 2A,B and Figure S2A,B). When we performed qRT‐PCR in tissues collected from the three groups to validate this finding, we found upregulated mRNA expressions of GNAT3 in the eCRSwNP group (Figure 2C). When we further labeled SCCs with GNAT3 in nasal epithelium by IF staining, similar to the results of qRT‐PCR, we identified SCC hyperplasia in the eCRSwNP group compared with both the neCRSwNP and control groups (Figure 2D,E). Given the critical role of BCs in eCRSwNP, we detected the spatial and quantitative relationship between SCCs and BCs. Tissue IF revealed localization of SCCs near BCs and a positive correlation between the number of SCCs and adjacent BCs (Figure 2F,G). These findings highlight the potential role of SCCs in BC fate.

FIGURE 2.

SCC exacerbates epithelial remodeling and promotes eCRSwNP progression. (A) Differentially expressed gene analysis across the three groups in bulk RNA‐seq. (B) Histogram showing relative mRNA expression of SCC markers‐POU2F3, TRPM5, and GNAT3 in bulk RNA sequencing. (C) Histogram showing relative mRNA expression of GNAT3 in tissues of each group through qRT‐PCR analysis (n = 6 per group). (D) IF staining of GNAT3 expression in each group. (E) Histogram showing SCC numbers (n = 8 per group). (F) IF staining of GNAT3 (green) and P63 (red) in the eCRSwNP epithelium. (G) Line chart showing the correlation between SCCs and adjacent BCs (n = 15). (H) Line chart showing the correlation between SCCs and NP endoscopy scores (n = 15). (I) H&E staining images of the murine nasal cavity, highlighting classic polypoid lesions (asterisks) with enlarged areas (black box). (J) Histogram showing number of polyp‐like lesions in the murine model (n = 6 per group). (K) H&E staining of the nasal mucosal epithelium of mice in each group (blue box). (L) Histogram showing murine nasal mucosal epithelium thickness (n = 6 per group). (M) PAS staining in each murine group (red box). (N) Histogram showing number of GC in each murine group (n = 6 per group). Data are presented as mean ± SEM; *p < 0.05, ***p < 0.001; ns, not significant.

Given that BC hyperplasia is a key feature of eCRSwNP [16, 50] and that we identified a positive correlation between SCC and BC number, we further assessed whether SCCs contribute to eCRSwNP. First, we analyzed SCC number and nasal polyp score in patients with eCRSwNP. Intriguingly, SCC number was positively correlated with nasal polyp score (Figure 2H), indicating the potential promotional role of SCCs in eCRSwNP. To confirm this role, we applied DB, a bitter receptor agonist that had activated SCCs [51] in a murine NP model (Figure S2C), which is regarded as a CRS model with NPs and T2 inflammation [52, 53, 54]. H&E and PAS staining revealed that the DB treatment group had larger NPs (Figure 2I,J), increased epithelial thickness (Figure 2K,L), and profound GC hyperplasia (Figure 2M,N).

In summary, these results indicate that SCCs and hyperplasia in eCRSwNP significantly contribute to the progression of eCRSwNP, likely through regulating BC fate.

3.3. SCC‐Derived Ach, Not Neurons, May Mediate SCC‐BC Crosstalk in eCRSwNP

SCCs are known to secrete Ach [35, 37, 38], which is essential for regulating the proliferation and differentiation of epithelial stem cells [24, 55]. We identified significantly elevated mRNA expression of ChAT, the key enzyme for Ach synthesis, and vesicular acetylcholine transporter (VAChT), responsible for Ach release, in the eCRSwNP group compared with both the control and neCRSwNP groups through bulk RNA‐seq (Figure 3A). This finding was further validated by qRT‐PCR analysis of tissue samples (Figure 3B). When we performed immunohistochemical analysis, we found increased ChAT expression in patients with eCRSwNP (Figure 3C–E). In addition, enzyme‐linked immunosorbent assay (ELISA) analysis revealed a significantly higher level of Ach in the supernatant of submerged cell cultures from patients with eCRSwNP (Figure S3A). These results suggested elevated Ach levels in eCRSwNP, prompting us to investigate whether Ach serves as a vital mediator between SCCs and BCs.

FIGURE 3.

Increased Ach levels in eCRSwNP, predominantly derived from SCC. (A) Histogram showing relative mRNA expression of ChAT and VAchT in bulk RNA‐seq. (B) Histogram showing relative mRNA expression of ChAT and VAchT in control, neCRSwNP, and eCRSwNP tissues through qRT‐PCR analysis (n = 6 per group). (C) Immunohistochemical staining showing ChAT expression in tissues from control, neCRSwNP, and eCRSwNP groups. (D‐E) Histogram showing ChAT immunoreactivity and positive staining areas in each group of tissues (n = 8 per group). (F) IF staining of ChAT (red) and the neuronal marker βIII Tublin (green) in control and eCRSwNP tissues. (G) Histogram showing CHAT⁺βIII Tublin⁺ cell numbers in subepithelium (n = 8 per group). (H) IF staining of ChAT (red) and the SCC marker GNAT3 (green) in control and eCRSwNP tissues. (I) Histogram showing CHAT⁺GNAT3⁺ cell numbers (n = 8 per group). (J) IF staining showing co‐localization of ChAT (red) with CC marker FOXJ1 (green), GC marker MUC5AC (red), BC marker P63 (green), and SCC marker GNAT3 (green) in ALI cultures. (K) Histogram showing Ach concentrations in the basolateral medium of ALI cultures under the following conditions: Untreated, α‐NETA‐treated, and IL‐4/IL‐5/IL‐13 treatment with or without α‐NETA, as measured by ELISA (n = 3 per group). Data are presented as mean ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Given that Ach is a classical neurotransmitter, we then examined whether increased Ach in eCRSwNP originates from neurons. When we co‐stained ChAT with the neuron marker β3‐tubulin, we found that in patients with eCRSwNP, β3‐tubulin expression decreased sharply in the subepithelium area, resulting in less co‐localization of β3‐tubulin and ChAT, while co‐localization of β3‐tubulin and ChAT in the lamina propria did not change compared with that in healthy controls (Figure 3F,G). Furthermore, in the eCRSwNP group, ChAT expression was highly located in the epithelial layer compared with the other two groups (Figure 3C–E). Since Ach is rapidly inactivated in the body and primarily functions in autocrine or paracrine signaling [56, 57, 58], these intriguing findings indicate that neuron‐derived Ach has limited effects on the epithelium in eCRSwNP; rather, other epithelial cells secrete higher levels of Ach in eCRSwNP.

Recent studies have suggested that SCCs release Ach in the intestinal tract [37, 38]. The scRNA‐seq [dataset] results confirmed higher expression of ChAT in SCCs, with nearly no expression of other types of epithelial cells (Figure S3B and Table S8). When we then co‐stained the samples with ChAT and the SCC marker GNAT3 (Figure 3H,I), we observed that ChAT was predominantly co‐localized with GNAT3 in patients with eCRSwNP compared with healthy controls. Collectively, these findings suggest that elevated Ach in eCRSwNP may be derived primarily from SCCs rather than neurons and might be the key transmitter between SCCs and BCs.

Additionally, we verified the cellular source of Ach and investigated how type 2 inflammatory cytokines influence Ach release using ALI culture models established from eCRSwNP tissues. IF staining was performed on ALI cultures using antibodies targeting ChAT alongside several epithelial cell markers, including FOXJ1 (CC), MUC5AC (GC), P63 (BC), and GNAT3 (SCC). Our results demonstrated that ChAT expression was exclusively co‐localized with GNAT3‐positive cells, with no detectable overlap with other cell‐type markers. This finding confirmed that within the ALI culture model (Figure 3J), ChAT is specifically expressed by SCCs regardless of the presence of a type 2 inflammatory environment, which was consistent with our tissue‐based observations and aligned with previous reports [36, 37, 38]. To further examine the regulation of Ach release, we re‐established ALI cultures and cultured them with IL‐4, IL‐5, or IL‐13 (Figure 3K). Basolateral culture media were collected and analyzed by ELISA. The results revealed that IL‐4 and IL‐13 significantly elevated Ach levels, while IL‐5 showed no significant effect. To elucidate the underlying regulatory mechanisms, we employed the ChAT inhibitor α‐NETA (Figure 3K). Treatment with α‐NETA had no significant impact on Ach levels in untreated controls or IL‐5‐treated cultures. However, it effectively suppressed the IL‐4‐ and IL‐13‐induced elevation in Ach secretion. These findings suggested that IL‐4 and IL‐13 serve as key activators of SCCs, promoting Ach production in eCRSwNP, which is consistent with previous reports [59, 60].

3.4. Ach/mAchR Induce Similar Epithelial Remodeling In Vitro

Ach has been reported to promote the proliferation and disrupt the differentiation of epithelial stem cells in the intestinal tract [23, 24, 25, 55], but its role remains unclear in rhinosinusitis. mAchRs are key mediators in Ach‐regulated epithelial proliferation and differentiation [23, 25, 61]. Prior studies using immunohistochemistry have indicated that M1, M2, and M3 receptors (M1R, M2R, and M3R, respectively) are the most prevalent mAchRs in the nasal epithelium [26, 27].

To confirm these findings, we performed IF staining for M1R, M2R, and M3R on nasal tissues, including control nasal mucosa, neCRSwNP, and eCRSwNP. In the control group, our results aligned with previous reports. However, in the eCRSwNP group, M1R and M3R were more widely distributed among BCs within the epithelium compared to controls (Figure 4A,B,E,F). In contrast, M2R showed more limited distribution in the eCRSwNP epithelium (Figure 4C,D).

FIGURE 4.

Ach induces epithelial remodeling by promoting BC hyperplasia via mAchR. (A) IF staining of M1R (green) and P63⁺ (red) in each group. (B) The percentage of P63⁺M1R⁺ cells among P63⁺ cells (n = 4 per group). (C) IF staining of M2R (green) and P63⁺ (red) in each group. (D) The percentage of P63⁺M2R⁺ cells among P63⁺ cells (n = 4 per group). (E) IF staining of M3R (green) and P63 (red) in each group. (F) The percentage of P63⁺M3R⁺ cells among P63⁺ cells (n = 4 per group). (G) Cells in submerged cultures were divided into indicated groups and proliferation ability was determined through CCK8 assay (n = 3 per group). (H) Schematic of the ALI culture model. (I) H&E staining in ALI cultures. (J) Quantification of epithelial thickness in ALI cultures (n = 3 per group). (K) IF staining of basal cell marker P63 (red) in ALI cultures. (L) Quantification of P63⁺ cells in ALI cultures (n = 3 per group). (M) PAS staining images in ALI cultures. (N) Quantification of GCs in ALI cultures (n = 3 per group). (O) IF staining of MUC5AC (red) in ALI cultures. (P) Quantification of MUC5AC⁺ cells in ALI cultures (n = 3 per group). (Q) IF staining of α‐tubulin (red) in ALI cultures. (R) Quantification of CCs in ALI cultures (n = 3 per group). (S) IF staining of P63 (red) and Ki67 (green) in ALI cultures. (T) The percentage of P63⁺Ki67⁺ cells among P63⁺ cells in ALI cultures (n = 3 per group). Data are presented as mean ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

To further investigate the role of Ach in BC regulation, we employed submerged cell cultures and ALI culture models. TB, a selective long‐acting mAchR antagonist targeting M1R and M3R, was used to assess the impact of Ach on BCs [19, 62]. IF staining of submerged cultures revealed that most cells highly expressed P63, confirming that the epithelial composition primarily consisted of BCs (Figure S4A).

Using a Cell Counting Kit‐8 (CCK‐8) assay, we examined BC proliferation in response to Ach. After 24 h, cells treated with Ach displayed higher optical density (OD) values compared to PBS‐treated cells, a difference that persisted for 72 h. These findings indicate that Ach promotes BC proliferation. Moreover, blocking M1R and M3R with TB significantly reduced BC proliferation within the first 24 h, with this effect sustained over 72 h (Figure 4G).

We further applied ALI culture models to describe the effect of Ach on BC proliferation and differentiation by seeding BCs in a transwell chamber to form a single‐layer structure in the first 4 days before adding Ach +/− TB to the differentiation medium. When we collected the differentiated structures after 21 days for further analysis (Figure 4C), H&E staining revealed a significant increase in epithelial thickness in the Ach‐treated group compared to the PBS group, whereas TB reversed this effect (Figure 4I,J). Moreover, IF staining showed that Ach treatment significantly increased the total number of BCs and the percentage of Ki67⁺ BCs. These effects were markedly reduced by TB treatment (Figure 4K,L,S,T).

Previous studies have shown that BC hyperplasia can lead to abnormal differentiation [63, 64, 65, 66]. Thus, we stained CC marker (α‐tubulin), GC marker (MUC5AC), and SC marker (IVL) in the ALI model to investigate whether Ach could regulate BC fate. PAS staining showed Ach increased the number of GCs compared with the PBS group (Figure 4M,N). Furthermore, IF staining showed both the GC and SC numbers significantly increased (Figure 4O,P and Figure S6C,D) while the CC number sharply decreased (Figure 4Q,R). Notably, TB treatment reversed these effects, reducing GC and SC numbers and restoring CC numbers (Figure 4M–R and Figure S6C,D).

These findings indicate that Ach alters BC fate via M1R and/or M3R. However, TB effectively attenuates Ach‐induced changes.

3.5. Ach Promotes the Expression of YAP in BC Nuclei

To understand the mechanisms underlying Ach‐induced BC hyperplasia, we compared DEGs in BCs among the control, neCRSwNP, and eCRSwNP groups. We identified 654 genes that were differentially expressed exclusively between the eCRSwNP and control groups (Figure 5A and Tables S9–S12). Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis revealed that these genes were significantly enriched in pathways such as the Hippo, Wnt, and PI3K/AKT signaling pathways (Figure 5B). After further stratifying the eCRSwNP group into ChAT^High and ChAT^Low groups based on ChAT expression, we identified 160 upregulated and 393 downregulated BC DEGs (Figure 5C and Table S13). KEGG pathway analysis of these genes showed significant enrichment in the Hippo, MAPK, and TGF‐β signaling pathways (Figure 5D).

FIGURE 5.

Ach promotes YAP expression in BC nuclei. (A, B) Venn diagrams (A) and KEGG analysis (B) of 654 genes with differential expression unique to eCRSwNP through scRNA‐seq. (C, D) Volcano plot (C) and KEGG analysis (D) of differentially expressed genes between ChAT^High and ChAT^Low BC groups in eCRSwNP tissues through scRNA‐seq. (E‐F) Western blot analysis of YAP expression in submerged cell cultures with or without Ach and/or TB treatment (n = 3 per group). (G) IF staining of YAP in submerged cell cultures. (H) Histogram showing the percentage of P63⁺YAP⁺ cells among the total cell population in submerged cultures (n = 3 per group). (I) IF staining of P63 (red) and YAP (green) in ALI cultures. (J) Histogram showing the percentage of P63⁺YAP⁺ cells among the total P63⁺ cell population in ALI cultures (n = 3 per group). Data are presented as mean ± SEM. Data are presented as mean ± SEM; *p < 0.05, **p < 0.01, ***p < 0.001.

Given that the Hippo pathway was the common significantly enriched pathway, we further explored the correlation between Ach and YAP, a key effector molecule of the Hippo pathway [67]. Western blot assay showed Ach could significantly upregulate YAP expression, while TB could reverse this upregulation (Figure 5E,F). As nuclear translocation of YAP is known as a key mechanism of avoiding YAP degradation [68], we detected the nuclear expression of YAP in a submerged cell culture model derived from the polyp tissues of patients with eCRSwNP. IF staining showed that Ach could upregulate YAP expression, which was mainly located in the cell nucleus, compared with the control group. Cells treated with TB and Ach had a remarkably decreased expression of YAP in not only the nucleus but also the cytoplasm compared with cells treated with Ach (Figure 5G,H).

We next applied ALI models derived from the polyp tissues of patients with eCRSwNP to detect whether YAP expression is also important in BC differentiation. Intriguingly, YAP expression was mainly located in BCs, and Ach could upregulate YAP expression in not only BCs but also other types of epithelial cells, while TB could reverse this upregulation. Furthermore, Ach mainly promoted YAP translocation to the nucleus in BCs, and TB downregulated the nuclear expression of YAP (Figure 5I,J). These results suggest that Ach regulates BC proliferation and differentiation in eCRSwNP by promoting nuclear translocation of YAP.

3.6. MAchR Inhibitors Suppress NP Formation Through YAP In Vivo

To confirm the impact of Ach on epithelial remodeling in eCRSwNP, we examined whether blocking mAchR with TB could suppress NP formation in a murine NP model (model establishment and intervention methods are shown in Figure 6A). We first detected whether Ach was upregulated in a positive control group (OVA+SEB treatment) compared with a negative control group (PBS treatment). ELISA analysis of nasal lavage fluid revealed significantly higher levels of Ach in the OVA+SEB group compared with the control group, which confirmed the vital role of Ach in eCRSwNP formation (Figure S5A). We next validated whether TB could reduce NP formation and eosinophil infiltration. H&E staining showed that the OVA+SEB group experienced significant polyp formation compared with the PBS group, while treatment with TB remarkably decreased the polyp number in the OVA+SEB + TB group compared with the OVA+SEB group (Figure 6B,C). Moreover, sirius red staining showed that OVA+SEB treatment successfully upregulated eosinophil number compared with PBS treatment, and treatment with TB significantly decreased eosinophil infiltration in the OVA+SEB + TB group compared with the OVA+SEB group (Figure 6D,E).

Given that epithelial remodeling is the key process in eosinophilic NP formation, we next detected the epithelial composition in vivo. H&E staining showed a remarkably increased epithelial thickness in the OVA+SEB group that could be reversed by TB treatment (Figure 6F,G). PAS staining further confirmed that TB treatment reduced GC metaplasia compared with the OVA+SEB group (Figure 6H,I). IF staining of epithelial markers to determine whether TB could reverse the epithelial remodeling revealed higher expression of MUC5AC, lower expression of α‐tubulin, and elevated expression of IVL in the OVA+SEB group compared with the PBS group (Figure 6J–O and Figure S6E,F). However, TB treatment significantly downregulated MUC5AC expression, upregulated α‐tubulin expression, and decreased IVL expression (Figure 6J–O and Figure S6E,F), indicating TB could reverse the mucous hyperplasia, ciliary defect, and suppress squamous metaplasia in eCRSwNP.

As BC hyperplasia was proven to be the key process in epithelial remodeling, we examined whether TB treatment could inhibit this hyperplasia. BC number (tracked by P63) was significantly increased in the OVA+SEB group, while this number was decreased in the OVA+SEB + TB group. Moreover, TB treatment could reduce both BC number and colocalization of P63 and Ki67 in the OVA+SEB + TB group compared with the OVA+SEB group (Figure 6P–R). Given YAP was identified as having a key role in BC hyperplasia, we also detected YAP expression in a murine NP model. Nuclear expression of YAP in BCs was upregulated in the OVA+SEB group, and TB treatment downregulated the YAP expression. In addition, an increased number and percentage of YAP⁺ BCs were identified in the OVA+SEB group, but TB treatment could downregulate this increase (Figure 6S,T). Collectively, these results indicate that TB inhibits eCRSwNP through suppressing YAP translocation in BCs.

4. Discussion

An imbalance between epithelial injury and repair, leading to epithelial remodeling and barrier dysfunction, is a key pathological feature of eCRSwNP [9, 69, 70]. Several recent studies have confirmed that fate alteration of BCs, airway epithelial stem cells, leads to epithelial remodeling in eCRSwNP [12, 41] and that T2 inflammation is one of the most significant contributors to BC hyperplasia induction [60, 69, 71]. However, clinical trials targeting IL‐5 reported that 21.8% of patients experienced worsening symptoms and that 27.7% experienced no improvement. Moreover, trials targeting IL‐4Rα showed residual nasal polyp scores of 3.4 after 16 weeks of treatment [2, 17]. Collectively, these results indicate that factors beyond T2 inflammation drive eCRSwNP. Here, we observed that the number of SCCs increased within the epithelium of patients with eCRSwNP, consistent with findings of Kohanski et al. [34], and we further identified that SCC expansion positively correlated with BC hyperplasia and the severity of eCRSwNP. Moreover, we discovered that elevated levels of Ach in eCRSwNP, primarily derived from SCCs rather than neurons, played a crucial role in regulating BC fate through the Hippo‐YAP pathway. Finally, we demonstrated that the mAchR antagonist TB, the commonly used airway drug, effectively blocked the effects of Ach in epithelial remodeling. Above all, these results suggest the potential role of SCC‐derived Ach in eCRSwNP and provide new insights for eCRSwNP treatment.

Our results are in line with several studies that revealed BC hyperplasia in eCRSwNP [12, 16]. We observed marked hyperplasia of both BCs and GCs, a sharp reduction in CCs, and prominent squamous metaplasia together suggesting that alterations in BC fate may be a key pathological mechanism in eCRSwNP. SCCs are rare secretory epithelial cells dispersed across most mucosal surfaces [31, 32]. Previous studies have shown an increase in SCC number within the epithelium of CRSwNP [34, 35]. We also found an increase in SCC number within the epithelium of CRSwNP, and we further observed SCCs in close proximity to BCs and a positive correlation between the number of SCCs and both BC number and disease severity. These results are suggestive of the vital role of SCCs in eCRSwNP.

Previous studies of SCCs focused on IL‐25, an inflammation mediator secreted by SCCs [72, 73]. A recent study on SCCs in the olfactory epithelium of WT, Il7r−/−, Stat6−/−, and Pou2f3−/− mice found that under allergen stimulation, SCCs can activate BCs independently of the T2 inflammatory pathway [36]. However, the authors did not identify which mediators contributed to this activation through a non‐IL‐25 mechanism. Recent studies identified Ach as one of the mediators secreted by SCCs, and we found that ChAT, a key enzyme involved in Ach synthesis, is primarily located in SCCs. These results indicate that Ach could be released from SCCs.

Ach is commonly recognized as a major classical neurotransmitter released by the parasympathetic nervous system. However, NPs are generally considered to lack neural innervation [28, 29, 30], and our IF staining results confirmed sharply decreased expression of a neuron marker (β3‐tublin) in the subepithelium. By contrast, cholinergic nerves are predominantly localized to the stromal layer at the base of the polyps, making it unlikely for Ach released from these nerves to significantly affect most of the epithelial tissue. Thus, we proved that elevated Ach in eCRSwNP predominantly originated from SCCs rather than neurons. These findings provide evidence that neurotransmitters of non‐neural origin contribute to eCRSwNP beyond T2 inflammation.

Ach has been shown to regulate epithelial stem cells in various tissues through mAchRs, promoting epithelial repair and regeneration [24, 61, 74, 75]. However, little is known about the role of Ach in NPs. In this study, we found Ach plays a crucial role in regulating BC fate in both in vivo and in vitro models. Our results accord with studies in salivary gland epithelial progenitor cells, which are functionally similar to airway BCs [61]. Our IF staining study showed that mAchR subtypes M1R and M3R rather than M2R are widely expressed on BCs in the eCRSwNP epithelium. Furthermore, selective inhibition of M1R and M3R with TB effectively suppressed Ach‐induced effects both in vitro and in vivo. Notably, previous studies have shown that persistent activation of nicotinic acetylcholine receptors (nAchRs) can lead to squamous metaplasia [76]. In our study, TB treatment was able to suppress Ach‐induced squamous metaplasia, suggesting that Ach may also mediate this process through activation of M1R and/or M3R on BCs. Collectively, these findings not only provide new insight into the cholinergic regulation of BC fate in eCRSwNP but also underscore the therapeutic potential of TB, a widely used airway medication, as a promising treatment option for this condition.

The Hippo signaling pathway is a critical regulator of epithelial progenitor cell proliferation and differentiation [77, 78, 79]. YAP, the core effector of the Hippo pathway, has been implicated in epithelial remodeling in NPs, contributing to epithelial thickening, ciliary loss, GC metaplasia, and BC hyperplasia [67, 80, 81, 82]. Additionally, YAP activation has been linked to the induction of squamous metaplasia in epithelial cells [83]. Ach is also known as an important activator of YAP, typically through mAchRs [84]. Analysis of [dataset] scRNA‐seq data reported by Wang et al. [41] revealed a strong association between ChAT and the Hippo signaling pathway in BCs. Our in vitro and in vivo experiments further support the role of Ach in activating YAP in BCs. Moreover, treatment with TB was able to reverse this process. These findings suggest that Ach may regulate BC fate through mAchR activation of the YAP pathway, providing deeper insights into the mechanisms by which Ach influences BC fate in eCRSwNP.

This study has several limitations. First, we did not identify which specific mAchR subtype primarily mediates YAP activation and epithelial remodeling in eCRSwNP. Second, although M1 and M3 receptors are predominantly expressed on BCs, they are also present, albeit to a lesser extent, on other epithelial cell types. After treating differentiating ALI cultures with Ach, we observed increased nuclear expression of YAP in these other epithelial subtypes, an effect that was also inhibited by TB. However, we did not conduct a detailed analysis of these subtypes. Additionally, the pathogenesis of eCRSwNP involved multiple interacting factors, and the NP murine model, induced by allergen exposure, was not totally identical to the human disease. Although we observed a correlation between SCC activity and eCRSwNP, further studies are required to determine whether SCC activation represents the initiating event in pathogenesis.

Author Contributions

Yucheng Yang, Tao Lu, Yang Shen, Bowen Zheng, and Shuman Li conceived and designed the study. Yucheng Yang, Tao Lu, Bowen Zheng, Yu Jiang, Yijun Liu, and Longlan Shu contributed to experimental design. Bowen Zheng, Yalan Zhang, Ji Wang, Yuan Chen, Yin Zhou, Yue Gu, Dayu Guan, and Chenxi Li performed the experiments, data collection, and analysis. Tao Lu, Ji Wang, Panhui Xiong, and Jieyu Zeng analyzed the sequencing data. Bowen Zheng wrote the manuscript. Yucheng Yang, Tao Lu, Yang Shen, Lei Zhao, and Shujin Wang reviewed and revised the manuscript. Yucheng Yang and Yang Shen provided funding support. All authors read and approved the manuscript.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Appendix S1: all70048‐sup‐0001‐AppendixS1.zip.

Zheng B., Lu T., Wang J., et al., “Acetylcholine From Solitary Chemosensory Cell, Not Neuron, Regulates Basal Cell Fate Driving Eosinophilic Chronic Rhinosinusitis With Nasal Polyps,” Allergy 81, no. 5 (2026): 1617–1634, 10.1111/all.70048.

Data Availability Statement

All data are available in the main text or Appendix S1. The accession number for the scRNA‐seq data from human samples in this paper is GSA (Genome Sequence Archive) HRA000772. GRCh38 human genome was used as the reference genome (https://cf.10xgenomics.com/supp/cell‐exp/refdata‐gex‐GRCh38‐2020‐A.tar.gz) [39]. The Bulk RNA‐seq data can be obtained from SRA PRJNA1205537 (https://www.ncbi.nlm.nih.gov/bioproject/PRJNA1205537).