Uncinate process samples from the nasal cavity, and polyp tissues were obtained from 25 patients (7 males and 18 females). Normal uncinate process tissues were harvested during rhinoplastic surgery. Nasal polyp tissues were obtained from the region of the middle meatus at the beginning of the endoscopic surgical procedure of CRS patients. The CRS was diagnosed based on historical, endoscopic, and radiographic criteria and CT findings of sinuses according to the 2012 European position paper on rhinosinusitis and nasal polyps (EPOS) guidelines. All subjects were recruited from the Department of Otorhinolaryngology, Korea University Medical Center, Korea. All tissues were derived from patients with no signs of inflammation, allergies, asthma, or aspirin sensitivity. No patients had a history of consuming oral steroids, non-steroidal anti-inflammatory drugs, antihistamines, or antibiotics for at least past 2 months. Informed consent was collected according to the Declaration of Helsinki. This study was approved by the Korea University Medical Center Institutional Review Board, which also authorized the research. It was carried out in accordance with the guidelines of the Human Ethics Committee of the Korea University Guro Hospital (2020GR0308). Clinical characteristics of patients are summarized in Table 1 .

GSEA was performed to determine whether inflammation and ECM regulation are involved in CRS. GSEA software v4.1.0 was used to analyze the NGS data and GSE36830 data. Gene sets were analyzed using 10 hallmarks involving the inflammatory response and NABA ECM regulators. The number of permutations was set as 1000, the type was gene_set, and the chip platform was human_NCBI_Gene_ID_MSigDB.v7.4.chip.

Total RNA was extracted using the Trizol reagent (Invitrogen). The cDNA was synthesized using the Maxime RT PreMix kit (Intron Biotechnology, Korea) according to the manufacturer’s protocol. Amplification reaction was performed with the following steps: an initial 2 min denaturation step at 94°C; 40 cycles of 94°C for 5 seconds, 60°C for 10 seconds, 72°C for 20 seconds. All reactions were performed in 20 μL volume. Real-time PCR was performed on the QuantStudio 3 system (Applied Biosystems, Foster City, CA) with 100 ng cDNA template, 400 nM of each primer, and 10 mL Power SYBR Green PCR Master Mix (Applied Biosystems) in 20 μL. Analysis of relative gene expression was conducted using the 2^-ΔΔCT method. Each experiment was repeated at least three times, and GAPDH was used as an internal control.

Flow cytometry was performed to confirm that EOL-1 cells differentiated into eosinophils. EOL-1 were seeded as 1x10⁶ cells in a 100mm culture dish and cells were treated with or without 0.5 mM BA and cells were harvested after 3-5 days. Cells were incubated with CCR3-PE (10 µL per 1x10⁶ cells, R&D Systems) and IL-5Ra-FITC (10 µL per 1x10⁶ cells, R&D Systems) conjugated antibodies for 30 minutes at 4°C and analyzed by FACS. All groups were analyzed with 10,000 cells. Flow cytometry analysis was performed using a Flow Cytometer Analyzer (LSR Fortessa X-20).

To evaluate the protein expression and localization, immunocytochemical staining was performed. Before using the cell culture slides, they were coated with poly-L-lysine for 4 h, after which they were exposed to UV light for 6 h. Fibroblasts and EOL-1 cells were seeded on the poly-L-lysine coated slides. Subsequently, cells were fixed with 4% paraformaldehyde and permeabilized with 0.01% Triton X-100 in 1% bovine serum albumin for 10 minutes. Then, cells were blocked with 3% bovine serum albumin for 1 h. Fibroblasts were incubated with primary antibodies; anti-Fibronectin (1:200), anti-SMA (1:200), anti-collagen type I (1:200). Eol-1 were incubated with primary antibodies; anti-CCR3 (1:200), anti-IL-5Ra (1:200) and then incubated with secondary antibodies; anti-mouse Alexa 488 (Invitrogen) or anti-rabbit Alexa 555 (Invitrogen) secondary antibodies. Finally, cells were counterstained with 4’-6-diamidino-2-phenylindole for 10 min and stained cells were visualized using a confocal laser scanning microscope (LSM700, Zeiss, Oberkochen, Germany).

Nasal polyp and uncinate process tissue sections were deparaffinized and rehydrated. Sections were stained with primary antibodies, anti-OPN (1:1,000, Abcam) and anti-IL5Ra (1:1,000, Abcam) overnight at 4°C. Subsequently, they were incubated with anti-mouse Alexa 488 (Invitrogen) or anti-rabbit Alexa 555 (Invitrogen) secondary antibodies or 3,3’- diaminobenzidine (DAB) substrate (Zytomed systems). Sections were counterstained with hematoxylin or 4’-6-diamidino-2-phenylindole (Sigma-Aldrich).

First, co-culture of fibroblasts and EOL-1 was performed. In 60mm culture dish, fibroblasts were seeded with 3x10⁵ cells, and EOL-1 were seeded with 6x10⁵ and 1.2x10⁶, respectively. After 3 days of co-culture, the levels of IL-6, IL-8, and VEGF were measured using ELISA (Biolegend, San Diego, CA, USA). In the second experiment, the expression levels of IL-6, IL-8 and VEGF were checked after 3 days of treatment with OPN (0.5-4ug/ml) in fibroblasts. The sample of ELISA was carried out using the supernatant from which cells were removed.

Briefly, each well was blocked with blocking buffer for 2 h and washed using a wash buffer. Antibodies against IL-6, IL-8, or VEGF were added to the media and incubated for 2 h. A substrate solution and stop solution were added sequentially, and the optical density of each well was determined within 30 minutes using a microplate reader (Bio-Rad).

A cytokine and chemokine array (Proteome Profiler™ Human XL Cytokine Array Kit, R&D Systems, Minneapolis, MN, USA) was used to survey the changes in the expression levels of 102 cytokines and chemokines in EOL-1 cells after treatment with or without BA. Cell culture supernatant was collected, diluted, and incubated overnight with the contents of the array kit. The membrane was washed to remove unbound material followed by incubation with a cocktail of biotinylated detection antibodies. The membrane was incubated with streptavidin-horse radish peroxidase and chemiluminescent detection reagents, and a signal was generated at each capture spot corresponding to the amount of bound protein. The membrane was exposed to X-ray film for 10 minutes, and profiles of mean spot pixel density were analyzed using Quantity One software (Bio-Rad).

Nasal inferior turbinate and polyp tissues were cut into 2 to 3 mm³ pieces, washed three times with phosphate-buffered saline, and cultured in DMEM supplemented with 2% FBS (Invitrogen) 1% 10,000 units/mL penicillin, and 10,000 μg/mL streptomycin (Invitrogen).

Results were obtained from at least three independent experiments. Statistical significance of differences between control and experimental data was analyzed using the unpaired t-test or one-way analysis of variance followed by Tukey’s test (GraphPad Prism, version 5, GraphPad Software, San Diego, CA). Significance was established at a 95% confidence level and P < 0.05 indicated statistical significance.

We performed GSEA to determine the degree of inflammation and the amount of ECM produced in ECRS along with the effect of eosinophil on these aspects. The gene set database used Hallmark_inflammatory_response and NABA_ECM_Regulators. We observed that the correlation between inflammation and the expression of ECM regulators was higher in the CRS patient group than in the healthy control group, and a significant correlation was also found with the ECRS patient group. Among them, genes such as those encoding pro-inflammatory mediators IL-6 and IL-8 and ECM markers such as a-SMA and fibronectin showed a high normalized enrichment score (NES). ( Figure 1A ). When GSEA was performed using various genesets, high NES values ​​were derived for inflammation, tissue remodeling such as ECM and EMT, and protein secretion ( Figure 1B ). To examine whether eosinophils correlated with the expression of pro-inflammatory mediators and ECM components, correlation analysis was performed with the markers CD11b, CD62L, siglec-8, siglec-F, and CCR3. A partially positive correlation was confirmed ( Figure 1C ).

We used BA to differentiate EOL-1 cells into eosinophils. Incucyte-based cytotoxicity assay and WST-1 assay were performed to confirm the effect of BA on the viability of EOL-1 cells. When the cells were treated with BA at a concentration of 0 – 0.25 mM, EOL-1 cells showed a survival rate of 96% to 100%, and approximately 78% when treated with 0.5-1 mM. In contrast, when treated with 2 mM and 4 mM concentrations, EOL-1 cells showed 50.3% and 23.05% survival rates, respectively ( Figure 2A, B ). When not stimulated with BA, EOL-1 cells showed a doubling time of approximately 16.74 h. Treatment with BA inhibited cell proliferation. Treatment with BA at a concentration of 0.5 mM showed a doubling time of approximately 51.42 h, and the proliferation of EOL-1 cells was reduced by 68% compared to EOL-1 cells cultured under normal conditions. Treatment with 4 mM BA completely inhibited the proliferation of EOL-1 cells ( Figure 2C ). In WST-1 assay, concentrations of BA over 1 mM exerted cytotoxicity ( Figure 2D ). To determine whether BA induced differentiation of EOL-1 cells into eosinophils, we examined the morphology and eosinophil markers in EOL-1 cells. EOL-1 cells were treated with 0.5 mM BA for 3 and 5 days and the cell morphology was confirmed by using an inverted microscope and H&E staining. EOL-1 cells, when not stimulated with BA, proliferated during the cell growth phase and formed colonies. However, the induction of differentiation in EOL-1 cells by BA decreased their ability to form colonies and proliferate ( Figure 2E ). H&E staining results showed that EOL-1 cells without BA stimulation were approximately 15-20 µm in size, present the form of a myeloblast and were circular or oval shaped. EOL-1 cells treated with BA exhibited a size of 9-17 µm and were slightly smaller than normal EOL-1 cells, with the two lobes of the nucleus being clearly visible ( Figure 2F ). CCR3 is a receptor for the C-C type chemokine and is highly expressed in eosinophils. IL-5Ra is a subunit of the IL-5 receptor. It primarily binds to IL-3, colony-stimulating factor-2, and IL-5, and affects the activation of eosinophils. We confirmed the expression of the eosinophil marker in untreated EOL-1 cells and EOL-1 cells treated with BA using flow cytometry. In EOL-1 cells without BA stimulation, approximately 40-50% of cells expressed IL-5Ra but did not express CCR3. In contrast, EOL-1 cells treated with BA for three days showed 32.9% of all cells, and 43.4% of EOL-1 cells treated for 5 days were double-positive CCR3 and IL-5Ra ( Figure 2G ).

To confirm the effect of EOL-1 cells on fibroblasts, we co-cultured fibroblasts and EOL-1 cells at a ratio of 1:2 and 1:4 ( Figure 3A ). After 72 h of co-culture, EOL-1 cells were removed and the expression of tissue remodeling markers, namely, a-SMA, fibronectin, and collagen type I in fibroblasts was analyzed, followed by confirmation of the expression by immunohistochemistry. In fibroblasts co-cultured with differentiated EOL-1 cells, the expression of a-SMA, fibronectin, and collagen type I was significantly increased; moreover, the expression of ECM components was further upregulated in the group cultured with EOL-1 cells at a ratio of 1:4 ( Figure 3B ). Fibroblasts are traditionally recognized for their role in the production of the extracellular matrix. However, recent studies have indicated that fibroblasts produce pro-inflammatory mediators in addition to tissue remodeling factors. Hence, we examined the effect of the co-culture of EOL-1 cells and fibroblasts on the expression of pro-inflammatory mediators and VEGF. EOL-1 cells also upregulated the expression of pro-inflammatory mediators and VEGF in fibroblast. When fibroblasts were co-cultured with EOL-1 cells at a ratio of 1:2, 1:4, the secretion of IL-6, IL-8 and VEGF was increased. When co-culture was performed using differentiated EOL-1, higher production of IL-6, IL-8 and VEGF was observed ( Figure 3C ).

In co-culture conditions, RNA levels of ECM components (a-SMA, Fibronectin and collagen type I) of fibroblasts were partially elevated. a-SMA and collagen type 1 showed the greatest increase in RNA levels under co-culture conditions with differentiated EOL-1. However, Fibronectin did not show a significant increase in gene expression ( Figure 3D ). Total soluble collagen was measured using the Sircol assay. EOL-1 cells activated the fibroblast collagen production. Differentiated EOL-1 cells led to more collagen production compared to EOL-1 cells without differentiation ( Figure 3E ).

Figure 3 shows changes in the expression of IL-6, IL-8, VEGF, and ECM components in response to direct co-culture of EOL-1 cells and fibroblasts. However, we did not know which cells released those mediators. Hence, we physically separated EOL-1 cells and fibroblasts using transwell and evaluated whether the mediators secreted from EOL-1 cells affect the production of ECM components and cytokines in fibroblasts. After inserting the transwell into a 12-well plate, fibroblasts were cultured in a 12-well plate and EOL-1 cells were cultured in the transwell, respectively. At first, EOL-1 cells were treated with BA for 3 days before co-culturing with fibroblasts. Subsequently, differentiated EOL-1 cells were transferred in a transwell. IL-6, IL-8, and VEGF were mainly secreted from fibroblasts, not EOL-1 cells, and the secretion increased in fibroblasts co-cultured with differentiated EOL-1 cells ( Figure 4A ). EOL-1 cells also induced the expression of genes encoding ECM components. The differentiated EOL-1 cells further stimulated fibroblasts to increase the expression of ECM components (a-SMA, fibronectin, COL1A1) ( Figure 4B ). In addition, EOL-1 cells significantly increased the production of soluble collagen in fibroblasts ( Figure 4C ). Based on these results, we confirmed that EOL-1 cells stimulated fibroblast without direct interaction, leading to the production of inflammatory mediators, VEGF, and ECM components.

To confirm that the mediators secreted by EOL-1 cells play an important role in fibroblast stimulation, we used EOL-1 cell-conditioned media to treat the fibroblasts. After culturing EOL-1 cells with or without BA for 3 days, EOL-1 cells were removed by centrifugation, and the media was used for fibroblast culture. EOL-1 cell-conditioned media stimulated fibroblasts to increase the expression of IL-6, IL-8, and VEGF. When the EOL-1 cell ratio was higher and when differentiated EOL-1 cells were used, the expression of inflammatory mediators and VEGF was increased ( Figure 4D ). The expression of ECM components (a-SMA, fibronectin, and collagen type I) and soluble collagen showed similar results ( Figure 4E, F ). Therefore, we confirmed that cytokine and chemokine secreted from EOL-1 cells play an important role in fibroblast stimulation.

We used a cytokine array to identify the key cytokine secreted from differentiated EOL-1 cells that alters the secretory capacity of fibroblasts After treatment of EOL-1 cells with 0.5 mM butyric acid for 3 days, the differences in cytokine expression compared to that from undifferentiated EOL-1 cells was analyzed using a human XL cytokine array kit. The expressions of serpin E1, osteopontin, and sex hormone binding receptor (SHBR) increased 11-fold, 6.7-fold, and 5.2-fold, respectively, in EOL-1 cells differentiated with butyric acid compared to those in undifferentiated EOL-1 cells ( Figure 5A ). Immunocytochemical analysis showed increased expression of CCR3 and IL-5Ra in differentiated EOL-1 cells, and overexpression of osteopontin (OPN) was also observed ( Figure 5B ).

To evaluate the effects of OPN on nasal fibroblasts, we directly treated fibroblasts with osteopontin, a cytokine overexpressed in differentiated EOL-1 cells. Fibroblasts were treated with osteopontin at concentrations ranging from 0.5-4 µg/mL for 48 h and 72 h, respectively, and RNA and protein levels were measured using real-time PCR and ELISA. Recombinant OPN significantly increased the expression of IL-6, IL-8, and VEGF in fibroblasts in a dose-dependent manner ( Figure 6A, B ). RNA levels of a-SMA were significantly increased only upon treatment with OPN at a concentration of 4 µg/mL or more. The expressions of fibronectin and collagen type I were also elevated upon treatment with OPN at 4 µg/mL ( Figure 6C ). Immunocytochemical analysis showed that OPN significantly increased the protein expression of ECM components (a-SMA, fibronectin, collagen type I) in fibroblasts. All the ECM components were expressed in the fibroblast cytoplasm, with the most significant increase observed upon treatment with 4 µg/mL OPN ( Figure 6D ).

To evaluate the effect of OPN on CRS, we examined the expression of OPN in the nasal tissues. The expression of OPN was higher in CRS tissues than in healthy control tissues. Especially, a significant increase was observed in the ECRS. IL-5Ra, an eosinophil marker, was highly expressed in the ECRS tissue, and the expression of OPN and IL-5Ra showed a positive correlation. The expression level of OPN showed a significant correlation with the severity of CRS (CT score) ( Figure 7A ). IHC was performed to examine the expression of OPN in tissues. OPN was primarily overexpressed on the surface of nasal epithelial cells, endothelial cells in blood vessels, and eosinophils. The expression level was increased in ECRS tissue ( Figure 7B ). Results of IHC showed high expression of OPN in ECRS (Green), and co-staining was observed in IL-5Ra (Red) + cells; eosinophils ( Figure 7C ).