2.1 Dataset selection

The GSE227541 dataset in the Gene Expression Omnibus (GEO) database of Pubmed (https://www.ncbi.nlm.nih.gov/) was selected for analysis. This dataset was uploaded and made public by Zhang P et al. in December 2023, comprising a total of 14 samples (7 NPC tissues and 7 sh-RNA tissues). The analysis platform was GPL16791: Illumina HiSeq 2500 (Homo sapiens).

2.2 DEG screening

The GEO2R interactive web tool was used to perform quality control, merging, functional analysis, and other standardized steps on the GSE227541 dataset. According to the random effects model for heterogeneous transcriptome datasets applicable to different batches and platform files, the screening parameters were set to missing values ≤ 10%, |logFC|≥ 5, and P < 0.05.

2.3 Functional enrichment analysis

For genomic functional enrichment analysis, we utilized the gene GO annotation and KEGG rest API in the R software (Nanjing University) package org.Hs.eg.db (version 3.1.0) for enrichment analysis.

2.4 Protein–protein interaction (PPI) network construction

After the construction of the PPI network (version 12.0) (https://cn.string-db.org/) of these DEGs by the PPI network analysis platform STRING, the CytoHubba plugin in Cytoscape (3.5.0) was utilized for PPI network model screening. The top 20 core target genes were visually presented using the Maximal Clique Centrality (MCC) algorithm.

2.5 Expression and prognosis analysis

The expression of DEGs in NPC and other malignant tumor diseases was screened in the GEPIA database (http://gepia.cancer-pku.cn/index.html), and a boxplot of expression was drawn. Based on the survival analysis platform Kaplan–Meier Plotter (https://kmplot.com/analysis/index.php?p=background), the core target genes screened in the previous step were grouped according to high (high expression group) and low expression (low expression group). Subsequently, the impact of the core target genes on patients’ overall survival (OS) was explored by the Kaplan–Meier method.

2.6 Immune microenvironment analysis

The CIBERSORT algorithm was used to estimate the infiltration abundance of immune cells in each sample. Using the TIMER database (https://cistrome.shinyapps.io/timer/), which analyzes 32 types of human cancers and 10,897 samples from the TCGA database, the relationship between mucin 5B, oligomeric mucus/gel-forming (MUC5B) expression and the infiltration of 6 major immune cells (CD8+ T cells, CD4+ T cells, B cells, dendritic cells, macrophages, and neutrophils) in the TIME was analyzed.

2.7 Clinical data

Cancer tissues and adjacent counterparts were collected from 30 NPC patients admitted to our hospital from October 2023 to March 2024 for subsequent analysis. The mean age of these patients was (61.26 ± 6.84) years, 19 were males and 11 were females. Pathologic staging was stage I in 14 cases and stage II in 16 cases. Inclusion criteria: Age 18–60 years; conforming to the clinical manifestations of NPC [20], with stage 0–II NPC diagnosed by pathological biopsy; successful collection of cancer tissues and adjacent counterparts (more than 5 cm away from the cancer tissues). Exclusion criteria: Multiple tumors, immunodeficiency diseases, or hematological disorders.This study involving human subjects complied with the Declaration of Helsinki. The Human Ethics Committees of The First People's Hospital of Jiande approved this study (No. 20240823001).All participants provided written informed consent.

2.8 Western blot

The tissue samples were placed in a centrifuge tube, added with 1 mL of PBS (10010001, Thermo Scientific, Shanghai, China) for thorough grinding, and centrifuged (12,000 rpm/min) for 10 min to discard the supernatant. The total protein was measured using bicinchoninic acid (BCA) (23235, Thermo Scientific, Shanghai, China) following tissue lysis with a radioimmunoprecipitation assay (RIPA) lysis buffer (Cat: R0010, Beijing Solarbio Technology Co., Ltd., Beijing, China). After the preparation of the electrophoresis gel, the protein was loaded and electrophoresed at 80 V for 30 min, followed by 45 min of electrophoresis at 120 V. The sample was then transferred to a polyvinylidene fluoride (PVDF) membrane (IPFL00010, Merck, Darmstadt, Germany) at a constant current of 200 mA, followed by blocking with a 5% blocking solution and the addition of a MUC5B (ab77995), Glutathione peroxidase 4 (GPX4) (ab262509), Recombinant solute carrier family 7, member 11 (SLC7A11) (ab300667), p53 (ab32509), Recombinant acyl coenzyme a synthetase long chain family, member 4 (ACSL4) (ab155282), β-actin (ab179467) primary antibody (1:1000, Abcam, Cambridge, UK) for a night-long incubation at 4 °C. The primary antibody was recovered the next day, the membrane was washed with TBST for 5 min (3 times), and then the secondary antibody (1:2000, ab8245, Abcam, Cambridge, UK) was added for incubation for 1 h. Exposure and development were performed after the addition of the enhanced chemiluminescence (ECL) (HY-K1005, MedChemExpress, New Jersey, USA) developer, and the relative MUC5B expression relative to β-actin was calculated.

2.9 Cell data

The human NPC cell CNE-2Z, ordered from Shanghai Honsun Biological Technology Co., Ltd., was cultured in the supporting RPMI-1640 medium in a constant temperature (37 °C) incubator with 5% CO2. Passage was performed every 2 days, strictly following the instructions for cell use. Cells in the logarithmic growth phase of 3–5 generations were collected for subsequent experiments. The cells used in this study have completed mycoplasma and STR testing to confirm that they are not contaminated.

2.10 Cell transfection

Fubio (Suzhou) Biomedical Technology Co., Ltd. was entrusted to design the MUC5B abnormal expression vectors. Logarithmic-growth-phase CNE-2Z was seeded in 6-well plates at 1.5 × 105/well. When the cell fusion degree reached 60–80%, the MUC5B abnormal expression vectors were transfected into the cells as per the instructions of the LipofectamineTM2000 kit (11668027, Thermo Scientific, Shanghai, China). Among them, the cells transfected with the MUC5B overexpression vector were labeled as the Overexpression group, those transfected with the MUC5B short hairpin RNA were regarded as the sh-RNA group, and the cells transfected with the MUC5B empty vector were labeled as the blank group. The expression of MUC5B was detected by referring to the method described in the Western blot section to verify the effect of interference expression.

2.11 (4,5)-dimethylthiahiazo (-z-y1)-3,5-di-phenytetrazoliumromide (MTT)

Cells were seeded into 96-well plates at a density of 5 × 103 cells per well, and the supernatant was discarded at 24, 48, 72, and 96 h of cultivation, respectively, followed by the addition of MTT solution (298-93-1, Shanghai Yuanye Bio-Technology Co, Ltd, Shanghai, China). The supernatant was removed after 4 h of incubation, and dimethyl sulfoxide was added. The plate was then subjected to low-speed oscillation on a shaker for 10 min. The absorbance (A) value at 490 nm was detected using a microplate reader (JC-1200 MB, Qingdao Jingcheng Instrumentation Co, Ltd, China.), and the cell growth curve was plotted.

2.12 EdU (5-ethynyl-2ʹ-deoxyuridine)

Cells were seeded into 96-well plates at a density of 2 × 104 per well and cultured for 24 h. Then, 50 mmol/L of EdU culture medium (E6032S, Suzhou UElandy Biotechnology Co., Ltd., Jiangsu, China) was added for 2 h. After fixing with 4% paraformaldehyde for 30 min, staining with 1% crystal violet for 30 min, washing with PBS, and staining with Hoechst 33342 (23491-52-3, MedChemExpress, New Jersey, USA), the cells were observed under a fluorescence microscope. Five random fields of view were selected for photography.

2.13 Transwell

The Matrigel glue (HY-K6001, MedChemExpress, New Jersey, USA) was diluted and spread on the micro-membrane of the Transwell chamber for later use. Cells were seeded at 2 × 104/mL in the upper chamber, while 500 μL of medium containing 10% fetal bovine serum was added to the lower chamber. After 48 h of culture, the upper chamber was removed to be washed with PBS, and the transmembrane cells were fixed with formaldehyde and stained with crystal violet. The invasive cell status of each chamber was observed under a microscope.

2.14 Cell scratch test

Cells were seeded into 6-well plates at 4 × 105 per well, and the intermediate cells were scraped off with a 200 μL pipette tip when the cells grew to confluence. Then, the culture was continued for 24 h, and the cell migration distance after 24 h was observed. Cell migration rate = (0 h width—24 h width)/0 h width × 100%.

2.15 Fluorescence staining

After cell blocking, the diluted rabbit anti-human CD8 (1:500, ab237709, Abcam, Cambridge, UK), as well as rat-derived anti-human Programmed death-1 (PD-1) and Programmed cell death 1 ligand 1 (PD-L1) (1:500, ab269812, Abcam, Cambridge, UK) antibody solutions, were added for incubation overnight. After washing, the cells were incubated with the secondary antibody (1:2000, ab237728, Abcam, Cambridge, UK) in the dark overnight, and then incubated with a DAPI staining solution (28718-91-4, MedChemExpress, New Jersey, USA). After washing and mounting, the staining results were observed under the fluorescence microscope.

2.16 Animal information

Twenty-eight 5-week-old specific-pathogen-free (SPF) male BALB/c nude mice (18–20 g) were purchased from Nanjing GenScript Biotech Corporation [SYXK (Su) 2023-0017]. The mice were raised under standard light/dark (12 h/12 h) alternation conditions with free access to food and clean drinking water. This study received approval from Animal Ethics Committee of our institution (No. 20240823001).We confirmed that all experiments in this study were performed in accordance with the relevant guidelines and regulations.All the procedure of the study is followed by the ARRIVE guidelines.

2.17 Construction of tumor-bearing mice

The treated CNE-2Z cell suspension (1.0 × 107/mL, 0.2 mL per mouse) was subcutaneously injected into the back of each BALB/c nude mouse to establish a nude mouse NPC model. The animals were then normally raised, and the tumor growth was observed after 14 days. Subsequently, mice were anesthetized by intraperitoneal injection of sodium pentobarbital (30 mg/kg), mice were executed by cervical dislocation, and intact tumor tissues were isolated and measured and weighed. The auditing criteria for the size of tumors in rats with loaded tumors in the First People’s Hospital of Jiande states that the weight of tumors in adult rats should not exceed 10% of the weight of the animal in any direction. The size of the subcutaneous tumors constructed in this study complied with the requirements of the Ethics Committee of the First People’s Hospital of Jiande.

2.18 Flow cytometry

The tumor tissues were made into single-cell suspensions by collagenase digestion. CD8+ cells were labeled, and the cells were counted using the FCS Express version 6 flow cytometer (Guangzhou Jiyuan Biotechnology Co., Ltd., Guangdong, China).

2.19 Hematoxylin–eosin (HE) and immunohistochemistry (IHC) staining

The cells were fixed with 10% formaldehyde fixative, routinely dehydrated, transparentized, embedded, sectioned, dewaxed, and dehydrated. HE staining was performed, followed by rinsing, dehydration, transparency, and mounting in neutral balsam. The pathological morphology of the decidual tissue was observed under a light microscope. In addition, another set of tissue sections was incubated with PD-1 (ab52587), PD-L1 (ab228462), and Ki-67 (ab15580) primary antibodies overnight (1:500, Abcam, Cambridge, UK) after deparaffinization, heat-induced antigen retrieval, and serum blocking. After the dropwise addition of a secondary antibody (Evension polymer, 1:1000), DAB (Delf-15395, Hefei Wanke Biotechnology Co., Ltd., Anhui, China) was used for color development, hematoxylin was utilized for counterstaining the nuclei, and neutral balsam was adopted for mounting. Tumor cells with stained cell membranes were considered positive cells.

2.20 Intervention of pathway expression

Logarithmic-growth-phase CNE-2Z cells were assigned to blank, pathway inhibition (PI), and combined intervention (CI) groups. Cells in the blank group were cultured normally, those in the PI group were cultured in a medium containing the Wnt/β-catenin pathway inhibitor MSAB (5 μM, 16 h, 173436-66-3, MedChemExpress, New Jersey, USA), and cells in the CI group were transfected with the MUC5B overexpression vector on the basis of MSAB. According to the above-mentioned methods, the biological behavior of NPC cells and the growth of tumors in vivo were detected to verify that MUC5B regulates the progression of NPC through the Wnt/β-catenin signaling pathway.

2.21 Statistical analysis

Statistical analysis was conducted using SPSS 24.0 (IBM, New York, USA). All experiments in this study were repeated 3 times. The Shapiro–Wilk test was used to test the normality of the data. For normally distributed data expressed as (\(\overline\)± s), the inter-group and multi-group differences were identified by independent sample t-tests and the repeated measures ANOVA plus LSD intra-group tests, respectively. Non-normally distributed data, expressed as [median (quartile range)], were analyzed by the non-parametric Mann–Whitney U test (between groups) and the Kruskal–Wallis H test (multiple groups). P < 0.05 was considered statistically significant.