Regulatory role of IRF2 in proliferation and invasion of nasopharyngeal carcinoma and its underlying mechanism Authors: Chaofen Zhao, Jiayu Song, Xunyan Luo, Yue Chen, Feng Jin Published: 2026 ABSTRACT SummaryIRF2 is an interferon regulatory factor with context-dependent roles in cancer. We examined IRF2 expression and function in nasopharyngeal carcinoma (NPC) using immunohistochemistry, immunofluorescence, western blot, and functional assays in cell lines and a xenograft model. IRF2 was upregulated and predominantly nuclear in NPC, correlating with advanced T stage and higher EBV DNA load. ROC analysis indicated diagnostic value (AUC = 0.837). IRF2 overexpression promoted proliferation, migration, invasion, epithelial-mesenchymal transition (EMT) and suppressed apoptosis, whereas knockdown inhibited these phenotypes and reduced tumor growth in vivo. Mechanistically, IRF2 activated the Wnt/β-catenin pathway by upregulating β-catenin and facilitating its nuclear translocation. These findings establish IRF2 as an oncogenic driver in NPC and suggest the IRF2/Wnt/β-catenin axis as a potential therapeutic target. Introduction Nasopharyngeal carcinoma (NPC) is an epithelial carcinoma predominantly originating from the pharyngeal recess of the nasopharynx, with high incidence in Southeast Asia and Southern China. According to the international agency for research on cancer, approximately 120,000 new NPC cases were reported globally in 2022.1 Studies have reported that 20%–30% of NPC patients develop distant metastases within 5 years after treatment,2 while 10%–20% experience local recurrence.3,4 The 5-year survival rate for patients receiving re-irradiation after recurrence is less than 40%.5 Local recurrence and progression remain major challenges in NPC management. Therefore, in-depth investigation of the molecular mechanisms underlying tumor cell proliferation and invasion, and the identification of potential therapeutic targets based on these mechanisms, have become crucial research directions in NPC treatment. Tumor development involves dysregulation of cellular signaling pathways, microenvironment reprogramming, immune surveillance escape, uncontrolled cell proliferation, invasion, and metastasis, which are regulated by a series of genes and signaling molecules.6 Interferon regulatory factors (IRFs) are a family of pleiotropic transcriptional regulatory proteins comprising nine members (IRF1-IRF9), playing significant roles in cell growth regulation and tumorigenesis.7 For instance, IRF1 inhibits tumor formation by binding to p300 and stimulating p53-dependent transcription to enhance p21 gene expression.8 IRF3 suppresses colorectal cancer development by specifically antagonizing β-catenin nuclear translocation.9 Members such as IRF1, IRF3, IRF5, and IRF8 exert tumor-suppressive effects by regulating the cell cycle, inducing apoptosis, and inhibiting tumor angiogenesis.10,11 Interferon regulatory factor 2 (IRF2) is a key transcription factor within the IRF family, acting as a hub in various signaling pathways. It regulates interferon (IFN) and immune-related gene expression, and participates in innate immune responses, cell cycle checkpoint control, and the coordination of programmed cell death signaling pathways.12,13,14 Unlike the tumor-suppressive IRF1, IRF2 exhibits tissue-specific dual regulatory roles in solid tumors. Currently, Qi et al.15 have confirmed IRF2 overexpression in NPC via immunohistochemistry (IHC) and western blot (WB), identifying it as an oncogene, but its downstream effectors and signaling network remain incompletely elucidated. Tumor cell proliferation, invasion, and migration are core biological processes in cancer progression and metastasis, involving aberrant activation of multiple signaling pathways and dynamic regulation of the cellular microenvironment.16,17 Tumor development is complex, with epithelial-mesenchymal transition (EMT) being a critical initial step. EMT is a cellular process driven by inducing signals, transcriptional regulators, and downstream effectors under pathological and physiological conditions, wherein epithelial cells lose epithelial characteristics and acquire a mesenchymal phenotype and behavior, leading to enhanced cell migration and invasion. This process plays key roles in embryonic development, tissue repair, and cancer metastasis.18 During EMT, tumor cells gain invasive and metastatic capabilities, accompanied by loss of cell polarity, disruption of intercellular junctions, and cytoskeletal reorganization.19 Molecularly, EMT is characterized by reduced expression of the epithelial marker E-cadherin and increased expression of mesenchymal markers such as ZEB1, ZEB2, N-cadherin, vimentin, snail, and twist.20 EMT enhances tumor cell invasiveness, stemness, chemotherapy resistance, and distant metastasis.21,22 Studies report that increased IRF2 expression can inhibit tumor growth and improve patient prognosis in gastric cancer by directly regulating AMER1 transcription and suppressing the Wnt/β-catenin pathway.23 Additionally, IRF2 regulates EMT in tumors. In glioma, IRF2 enhances EMT-mediated migration and invasion.24 While IRF2 is established as an oncogene in NPC, its role in regulating EMT remains unclear. The Wnt/β-catenin signaling pathway plays crucial roles in embryonic development, tissue homeostasis, and various diseases. It regulates gene expression, influencing cell proliferation, differentiation, and migration. Key molecules in this pathway include Wnt proteins, frizzled receptors, LRP5/6 co-receptors, and β-catenin.25 β-catenin, a multifunctional protein, acts as a key signaling hub in the Wnt pathway. Its nucleocytoplasmic translocation is closely associated with tumorigenesis, making it a potential therapeutic target.26 The stability of β-catenin, a key downstream effector of Wnt signaling, is tightly regulated. Upon binding of extracellular Wnt ligands to frizzled receptors, disheveled (Dvl) protein is activated, inhibiting glycogen synthase kinase-3 beta (GSK-3β) activity, thereby preventing the phosphorylation-dependent degradation of β-catenin and promoting its accumulation in the cytoplasm. When cytoplasmic β-catenin concentration exceeds a threshold, it translocates to the nucleus, forms complexes with other proteins, and initiates target gene expression. In tumor cells, β-catenin accumulation promotes nuclear translocation, where it interacts with T cell factor (TCF)/lymphoid enhancer factor (LEF) transcription factors, upregulating genes like snail and slug, promoting EMT, and driving metastasis.27 The Wnt/β-catenin pathway is a key regulator of EMT, and its aberrant activation is closely linked to invasion and metastasis in various cancers.28,29,30,31 In gastric cancer cells, IRF2 inhibits EMT by regulating the Wnt/β-catenin pathway.23 In liver cancer, IRF2 increases lenvatinib mesylate resistance by regulating β-catenin expression.32 IRF2-mediated regulation of the Wnt/β-catenin pathway exhibits differential effects on EMT across different tumors. However, whether IRF2 influences EMT in NPC via the Wnt/β-catenin pathway remains unreported. Current evidence suggests that IRF2 can co-regulate EMT and Wnt/β-catenin pathway activation. Nevertheless, the role of the “IRF2-Wnt-EMT” regulatory network in NPC pathogenesis remains a research gap. This study aimed to investigate the clinical significance of IRF2, its impact on the malignant behaviors of NPC cells, and the potential underlying signaling pathways, with the goal of offering new theoretical insights that may inform targeted therapy for NPC. Results Subcellular localization of IRF2 in NPC and its prognostic potential Comparative proteomic analysis revealed that the expression level of IRF2 in head and neck squamous cell carcinoma is significantly higher than in most solid tumors. IRF2 has been demonstrated to play roles in various cancers (Figure 1A). However, its precise subcellular localization in NPC cells remains unclear. To determine the subcellular localization of IRF2, we first queried the Human Protein Atlas (HPA) database. It was found that IRF2 is primarily localized to the nucleus, with minor expression observed at cell-cell junction structures and within the cytoplasmic matrix (Figure 1B). To preliminarily assess the prognostic value of IRF2, we performed a Kaplan-Meier survival analysis using the TCGA head and neck squamous cell carcinoma (HNSC) dataset via the GEPIA2 platform. As shown in Figure 1C, no significant association was observed between IRF2 expression level and overall survival of HNSC patients (log rank p = 0.83; HR for high expression = 0.97, p(HR) = 0.82). This result suggests that the prognostic significance of IRF2 may be limited or context-dependent, warranting further investigation in dedicated NPC cohorts (Figure 1C). Subsequently, we experimentally validated the localization of IRF2 in NPC cell lines (CNE1, HONE1, 58F and 6-10B) using immunofluorescence (IF) staining. The experimental results confirmed that IRF2 is predominantly localized in the nucleus, with only a minor fraction distributed in the cytoplasm, which is consistent with the database prediction (Figure 2).Figure 1Localization, expression, and prognostic analysis of IRF2 based on public databases(A) Protein expression of IRF2 in a variety of cancer tissues in the HPA database.(B) Visualization of IRF2 subcellular localization (HPA database, https://www.proteinatlas.org/), green and red fluorescence represent IRF2 protein and microtubules, respectively. Micrograph is supplied with a scale bar, 20 μm.(C) Kaplan-Meier survival curve of overall survival for head and neck squamous cell carcinoma patients stratified by IRF2 expression level based on TCGA data (analyzed using GEPIA2). The high-expression group (red) and low-expression group (blue) were divided by the median expression level.Figure 2Immunofluorescence detection of IRF2 localization in NPC cell lines (800×)Representative images are shown with a scale bar, 25 μm (indicated in the merged image). Localization, expression, and prognostic analysis of IRF2 based on public databases (A) Protein expression of IRF2 in a variety of cancer tissues in the HPA database. (B) Visualization of IRF2 subcellular localization (HPA database, https://www.proteinatlas.org/), green and red fluorescence represent IRF2 protein and microtubules, respectively. Micrograph is supplied with a scale bar, 20 μm. (C) Kaplan-Meier survival curve of overall survival for head and neck squamous cell carcinoma patients stratified by IRF2 expression level based on TCGA data (analyzed using GEPIA2). The high-expression group (red) and low-expression group (blue) were divided by the median expression level. Immunofluorescence detection of IRF2 localization in NPC cell lines (800×) Representative images are shown with a scale bar, 25 μm (indicated in the merged image). IRF2 expression in NPC versus benign tissues and cell lines Consistent with previous findings that IRF2 is highly expressed in NPC,15 our IHC and WB analyses confirmed its significant upregulation in NPC tissues and cell lines compared to chronic inflammatory tissues and the normal NP69 cells (p < 0.05, Figure 3).Figure 3IRF2 exhibits elevated expression levels in both nasopharyngeal carcinoma tissues and corresponding cell lines(A) Immunohistochemical analysis of IRF2 expression in nasopharyngeal carcinoma tissues. Representative micrographs are supplied with a scale bar, 100 μm (200×) or 50 μm (400×).(B) Statistical representation of IRF2 immunohistochemical scoring. Data are represented as mean ± SEM.(C) Western blot analysis of IRF2 expression levels in nasopharyngeal carcinoma cell lines and nasopharyngeal epithelial cells.(D) Quantitative analysis of IRF2 protein expression across different cell groups. Data are represented as mean ± SEM. Statistical significance was determined by two-tailed Student’s t test. (∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001). IRF2 exhibits elevated expression levels in both nasopharyngeal carcinoma tissues and corresponding cell lines (A) Immunohistochemical analysis of IRF2 expression in nasopharyngeal carcinoma tissues. Representative micrographs are supplied with a scale bar, 100 μm (200×) or 50 μm (400×). (B) Statistical representation of IRF2 immunohistochemical scoring. Data are represented as mean ± SEM.(C) Western blot analysis of IRF2 expression levels in nasopharyngeal carcinoma cell lines and nasopharyngeal epithelial cells. (D) Quantitative analysis of IRF2 protein expression across different cell groups. Data are represented as mean ± SEM. Statistical significance was determined by two-tailed Student’s t test. (∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001). Predictive potential of IRF2 for NPC diagnosis To preliminarily explore the diagnostic potential of IRF2 for NPC, we performed ROC curve analysis based on IHC scores from chronic inflammatory tissues and NPC tissues. The area under the ROC curve (AUC) for IRF2 in discriminating between chronic inflammation and NPC was 0.837 (95% CI: 0.760–0.915). At a cutoff value of 168.17, the sensitivity was 94.44% and the specificity was 58.73% (Figure 4). These findings suggest that IRF2 may serve as a candidate biomarker for NPC; however, its diagnostic utility requires further validation in larger, independent cohorts. Subsequently, the 63 NPC patients were stratified into high- and low-expression groups based on this cut-off. The expression level of IRF2 (high vs. low) showed significant correlations with T stage and plasma Epstein-Barr virus (EBV) DNA load (p < 0.05), but not with smoking history, N stage, or M stage (p > 0.05) (Table 1).Figure 4The receiver operating characteristic curve was used to assess the predictive value of IRF2 in nasopharyngeal carcinomaTable 1Correlation analysis of IRF2 expression levels in nasopharyngeal carcinoma tissues with clinical characteristicsClinical featuresIRF2χ2PHigh expressionLow expressionAll cases3726––Age0.7450.388 ≥609 (24.32%)4 (15.38%)–– <6028 (75.68%)22 (84.62%)––Sex0.6740.412 Male29 (78.38%)18 (69.23%)–– Female8 (21.62%)8 (30.77%)––EBV DNA3.9880.046 >500 copies31 (83.78%)16 (61.54%)–– <500 copies6 (16.22%)10 (38.46%)––Smoking1.8930.169 Yes24 (64.86%)21 (80.77%)–– No13 (35.14%)5 (19.23%)––T stage4.3010.038 T1-22 (5.41%)6 (23.08%)–– T3-435 (94.59%)20 (76.92%)––N stage0.7920.373 N0-19 (24.32%)9 (34.62%)–– N2-328 (75.68%)17 (65.38%)––M stage0.2220.637 M023 (62.16%)17 (65.38%)–– M114 (37.84%)9 (34.62%)–– The receiver operating characteristic curve was used to assess the predictive value of IRF2 in nasopharyngeal carcinoma Table 1Correlation analysis of IRF2 expression levels in nasopharyngeal carcinoma tissues with clinical characteristicsClinical featuresIRF2χ2PHigh expressionLow expressionAll cases3726––Age0.7450.388 ≥609 (24.32%)4 (15.38%)–– <6028 (75.68%)22 (84.62%)––Sex0.6740.412 Male29 (78.38%)18 (69.23%)–– Female8 (21.62%)8 (30.77%)––EBV DNA3.9880.046 >500 copies31 (83.78%)16 (61.54%)–– <500 copies6 (16.22%)10 (38.46%)––Smoking1.8930.169 Yes24 (64.86%)21 (80.77%)–– No13 (35.14%)5 (19.23%)––T stage4.3010.038 T1-22 (5.41%)6 (23.08%)–– T3-435 (94.59%)20 (76.92%)––N stage0.7920.373 N0-19 (24.32%)9 (34.62%)–– N2-328 (75.68%)17 (65.38%)––M stage0.2220.637 M023 (62.16%)17 (65.38%)–– M114 (37.84%)9 (34.62%)–– Correlation analysis of IRF2 expression levels in nasopharyngeal carcinoma tissues with clinical characteristics Establishment of stable IRF2 overexpression and knockdown NPC cell lines To further investigate whether IRF2 regulates the malignant biological behavior of NPC, we established IRF2 overexpression and knockdown cell models by selecting cell lines with high endogenous IRF2 expression for knockdown and those with low expression for overexpression, in order to study its effects on the invasion, migration, and proliferation of NPC. Based on the findings from the first part of this study, we selected the NPC cell lines 58F and 6-10B for IRF2 overexpression, and the cell lines CNE1 and HONE1 for IRF2 knockdown. The results showed that in the 58F and 6-10B cell lines, IRF2 levels were significantly higher in the overexpression groups compared to the empty vector control groups (p < 0.05), indicating the successful construction of stable IRF2-overexpressing NPC cell lines. Similarly, in the CNE1 and HONE1 cell lines, IRF2 expression was significantly lower in the knockdown groups compared to the control groups (p < 0.05), with the sh2-IRF2 and sh3-IRF2 constructs showing the most pronounced reduction (Figure 5). These two knockdown cell lines were selected for subsequent experimental validation.Figure 5Constructing IRF2 knockdown and overexpression cell models(A) Gray scale graph of overexpression of IRF2 protein expression in 58F nasopharyngeal carcinoma cells detected by western blot.(B) Statistical representation of IRF2 protein overexpression in 58F nasopharyngeal carcinoma cell line. Data are represented as mean ± SEM.(C) Gray scale graph of overexpression of IRF2 protein expression in 6-10B nasopharyngeal carcinoma cells detected by western blot.(D) Statistical graph of overexpression of IRF2 protein expression in 6-10B nasopharyngeal carcinoma cells. Data are represented as mean ± SEM.(E) Western blot analysis grayscale image depicting the knockdown of IRF2 protein expression in CNE1 nasopharyngeal carcinoma cells.(F) Statistical graph of knockdown IRF2 protein expression in CNE1 nasopharyngeal carcinoma cells. Data are represented as mean ± SEM.(G) Western blot grayscale image showing the knockdown of IRF2 protein expression in HONE1 nasopharyngeal carcinoma cells.(H) Statistical graph of knockdown IRF2 protein expression in HONE1 nasopharyngeal carcinoma cells. Data are represented as mean ± SEM. Statistical significance was determined by two-tailed Student’s t test. (nsp > 0.05, ∗p < 0.05, ∗∗p < 0.01). Constructing IRF2 knockdown and overexpression cell models (A) Gray scale graph of overexpression of IRF2 protein expression in 58F nasopharyngeal carcinoma cells detected by western blot. (B) Statistical representation of IRF2 protein overexpression in 58F nasopharyngeal carcinoma cell line. Data are represented as mean ± SEM. (C) Gray scale graph of overexpression of IRF2 protein expression in 6-10B nasopharyngeal carcinoma cells detected by western blot. (D) Statistical graph of overexpression of IRF2 protein expression in 6-10B nasopharyngeal carcinoma cells. Data are represented as mean ± SEM. (E) Western blot analysis grayscale image depicting the knockdown of IRF2 protein expression in CNE1 nasopharyngeal carcinoma cells. (F) Statistical graph of knockdown IRF2 protein expression in CNE1 nasopharyngeal carcinoma cells. Data are represented as mean ± SEM. (G) Western blot grayscale image showing the knockdown of IRF2 protein expression in HONE1 nasopharyngeal carcinoma cells. (H) Statistical graph of knockdown IRF2 protein expression in HONE1 nasopharyngeal carcinoma cells. Data are represented as mean ± SEM. Statistical significance was determined by two-tailed Student’s t test. (nsp > 0.05, ∗p < 0.05, ∗∗p < 0.01). Effect of IRF2 on the proliferative capacity of NPC cells To further investigate the role of IRF2 in the proliferation of NPC cells, we employed the successfully constructed IRF2-overexpressing and knockdown NPC cell models. The CCK-8 assay and colony formation assay were utilized to systematically evaluate the regulatory effect of IRF2 on the proliferative potential of NPC cells. The results indicated that IRF2 overexpression significantly enhanced the proliferative activity of NPC cells (p < 0.05) (Figure 6). Conversely, IRF2 knockdown markedly inhibited the proliferative capacity of NPC cells (p < 0.05) (Figure 6).Figure 6IRF2 overexpression promotes proliferation of nasopharyngeal carcinoma cells, while knockdown of IRF2 exerts an inhibitory effect on proliferation(A, B, and D) Plate cloning assay to detect the clone formation ability of 58F and 6-10B nasopharyngeal carcinoma cells after overexpression of IRF2.(C and E) CCK-8 assay to detect the proliferation ability of 58F and 6-10B nasopharyngeal carcinoma cells after overexpression of IRF2. Data are represented as mean ± SEM (n = 3 independent experiments).(F, G, and I) The impact of IRF2 knockdown on the clonogenic ability of CNE1 and HONE1 nasopharyngeal carcinoma cells was evaluated using a colony formation assay.(H and J) The influence of IRF2 knockdown on the proliferative potential of CNE1 and HONE1 nasopharyngeal carcinoma cells was analyzed via CCK-8 assay. Data are represented as mean ± SEM (n = 3 independent experiments). Statistical significance was determined by two-tailed Student’s t test. (∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001). IRF2 overexpression promotes proliferation of nasopharyngeal carcinoma cells, while knockdown of IRF2 exerts an inhibitory effect on proliferation (A, B, and D) Plate cloning assay to detect the clone formation ability of 58F and 6-10B nasopharyngeal carcinoma cells after overexpression of IRF2. (C and E) CCK-8 assay to detect the proliferation ability of 58F and 6-10B nasopharyngeal carcinoma cells after overexpression of IRF2. Data are represented as mean ± SEM (n = 3 independent experiments). (F, G, and I) The impact of IRF2 knockdown on the clonogenic ability of CNE1 and HONE1 nasopharyngeal carcinoma cells was evaluated using a colony formation assay. (H and J) The influence of IRF2 knockdown on the proliferative potential of CNE1 and HONE1 nasopharyngeal carcinoma cells was analyzed via CCK-8 assay. Data are represented as mean ± SEM (n = 3 independent experiments). Statistical significance was determined by two-tailed Student’s t test. (∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001). Effect of IRF2 on the migration and invasion abilities of NPC cells To further investigate the regulatory role of IRF2 in the migration and invasion of NPC cells, this study utilized successfully established NPC cell models with IRF2 overexpression and knockdown. The effects of IRF2 on the migratory and invasive abilities of NPC cells were systematically evaluated through wound healing assays, transwell migration assays and Matrigel invasion assays. The experimental data revealed that IRF2 overexpression significantly enhanced the migration and invasion capabilities of NPC cells as assessed by transwell assays (p < 0.05, Figure 7). In wound healing assays, IRF2 overexpression accelerated wound closure. However, wound closure can be influenced by both cell proliferation and apoptosis, which are also modulated by IRF2. Therefore, the wound healing results should be interpreted with caution and considered together with the transwell migration data. Conversely, IRF2 knockdown markedly inhibited cell migration and invasion in transwell assays (p < 0.05, Figure 7), and also reduced wound closure, although the latter may partly reflect changes in proliferation and apoptosis.Figure 7Bidirectional regulation of migration and invasion by IRF2 in nasopharyngeal carcinoma cells: enhancement by overexpression in 58F/6-10B and suppression by knockdown in CNE1/HONE1(A) Transwell assay was used to evaluate the migration and invasion abilities of 58F and 6-10B nasopharyngeal carcinoma cells (40×). Representative micrographs are supplied with a scale bar, 100 μm.(B) Wound healing assay was performed to detect the migration ability of 58F and 6-10B nasopharyngeal carcinoma cells (40×). Representative micrographs are supplied with a scale bar, 500 μm.(C–F) Statistical analysis of the transwell assay and wound healing assay demonstrated that IRF2 overexpression significantly enhanced the migration and invasion capabilities of nasopharyngeal carcinoma cells. Data are represented as mean ± SEM.(G) The migration and invasion abilities of CNE1 and HONE1 nasopharyngeal carcinoma cells were evaluated by transwell assay (40×). Representative micrographs are supplied with a scale bar, 100 μm.(H) The migration ability of CNE1 and HONE1 nasopharyngeal carcinoma cells was detected using wound healing assay (40×). Representative micrographs are supplied with a scale bar, 500 μm.(I–L) Statistical analysis of the transwell assay and wound healing assay demonstrated that IRF2 knockdown significantly inhibited the migration and invasion capabilities of nasopharyngeal carcinoma cells. Data are represented as mean ± SEM. Statistical significance was determined by two-tailed Student’s t test. (∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001). Bidirectional regulation of migration and invasion by IRF2 in nasopharyngeal carcinoma cells: enhancement by overexpression in 58F/6-10B and suppression by knockdown in CNE1/HONE1 (A) Transwell assay was used to evaluate the migration and invasion abilities of 58F and 6-10B nasopharyngeal carcinoma cells (40×). Representative micrographs are supplied with a scale bar, 100 μm. (B) Wound healing assay was performed to detect the migration ability of 58F and 6-10B nasopharyngeal carcinoma cells (40×). Representative micrographs are supplied with a scale bar, 500 μm. (C–F) Statistical analysis of the transwell assay and wound healing assay demonstrated that IRF2 overexpression significantly enhanced the migration and invasion capabilities of nasopharyngeal carcinoma cells. Data are represented as mean ± SEM. (G) The migration and invasion abilities of CNE1 and HONE1 nasopharyngeal carcinoma cells were evaluated by transwell assay (40×). Representative micrographs are supplied with a scale bar, 100 μm. (H) The migration ability of CNE1 and HONE1 nasopharyngeal carcinoma cells was detected using wound healing assay (40×). Representative micrographs are supplied with a scale bar, 500 μm. (I–L) Statistical analysis of the transwell assay and wound healing assay demonstrated that IRF2 knockdown significantly inhibited the migration and invasion capabilities of nasopharyngeal carcinoma cells. Data are represented as mean ± SEM. Statistical significance was determined by two-tailed Student’s t test. (∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001). Effect of IRF2 on apoptosis in NPC cells Based on the aforementioned experimental results, IRF2 significantly influences the migratory, proliferative, and invasive capabilities of NPC cells. To further investigate whether IRF2 regulates the apoptotic process in NPC cells, we selected stably transfected NPC cell lines with either overexpression or knockdown of IRF2 and performed apoptosis analysis using flow cytometry. The experimental data indicated that IRF2 overexpression significantly suppressed the apoptosis of NPC cells (p < 0.05). Conversely, IRF2 knockdown markedly promoted apoptosis in NPC cells, and this difference was statistically significant (p < 0.05, Figure 8).Figure 8Effects of IRF2 on apoptosis in nasopharyngeal carcinoma detected by flow cytometry(A and C) Effects of overexpression of IRF2 on apoptosis in 58F and 6-10B nasopharyngeal carcinoma cells analyzed by flow cytometry.(B and D) Quantitative statistical analysis of apoptosis rate in 58F and 6-10B nasopharyngeal carcinoma cells. Data are represented as mean ± SEM.(E and G) Effects of knockdown of IRF2 on apoptosis in CNE1 and HONE1 nasopharyngeal carcinoma cells detected by flow cytometry. apoptosis.(F and H) Quantitative statistical analysis of apoptosis rate of CNE1 and HONE1 nasopharyngeal carcinoma cells. Data are represented as mean ± SEM. Statistical significance was determined by two-tailed Student’s t test. (∗p < 0.05, ∗∗p < 0.01). Effects of IRF2 on apoptosis in nasopharyngeal carcinoma detected by flow cytometry (A and C) Effects of overexpression of IRF2 on apoptosis in 58F and 6-10B nasopharyngeal carcinoma cells analyzed by flow cytometry. (B and D) Quantitative statistical analysis of apoptosis rate in 58F and 6-10B nasopharyngeal carcinoma cells. Data are represented as mean ± SEM. (E and G) Effects of knockdown of IRF2 on apoptosis in CNE1 and HONE1 nasopharyngeal carcinoma cells detected by flow cytometry. apoptosis. (F and H) Quantitative statistical analysis of apoptosis rate of CNE1 and HONE1 nasopharyngeal carcinoma cells. Data are represented as mean ± SEM. Statistical significance was determined by two-tailed Student’s t test. (∗p < 0.05, ∗∗p < 0.01). Effect of IRF2 on the subcutaneous tumorigenicity of NPC cells Based on the aforementioned in vitro experimental results, we found that IRF2 overexpression significantly promoted the proliferation of NPC cells, while its knockdown inhibited proliferative capacity (p < 0.01). To establish correlative evidence linking in vitro and in vivo experiments, this study constructed a subcutaneous xenograft tumor model in nude mice to evaluate the regulatory role of IRF2 on NPC proliferation within the in vivo microenvironment. The experimental results demonstrated that IRF2 knockdown significantly inhibited the in vivo proliferative ability of NPC cells (p < 0.05). Further pathological analysis of the subcutaneous xenograft tumor tissues via hematoxylin and eosin (HE) staining confirmed the successful establishment of the xenograft model. IHC staining results revealed that compared to the control group, the expression levels of both IRF2 and the cell proliferation marker Ki-67 were significantly reduced in the IRF2 knockdown group (p < 0.05, Figure 9). These findings indicate that IRF2 plays an important role in regulating the in vivo proliferation of NPC cells. However, it should be noted that the subcutaneous xenograft model in this study was used primarily to assess tumor growth. The proposed involvement of the Wnt/β-catenin pathway and EMT markers in vivo was not directly examined in these tumor tissues, representing a limitation of the current animal study.Figure 9In vivo validation of the effect of IRF2 on the proliferation of nasopharyngeal carcinoma cells in nude mice subcutaneous xenograft models(A and C) Subcutaneous xenograft assay to evaluate the regulatory effect of IRF2 knockdown on the proliferative capacity of nasopharyngeal carcinoma cells in vivo.(B) Growth curves of subcutaneous xenograft tumor volumes. Data are represented as mean ± SEM (n = 5 mice per group).(D) Statistical analysis of subcutaneous xenograft tumor weights. Data are represented as mean ± SEM.(E) Histological characteristics of subcutaneous xenograft tumors from each group were analyzed by HE staining, and the expression levels of IRF2 and Ki-67 proteins in xenograft tissues were detected by IHC. Representative micrographs are supplied with a scale bar, 100 μm.(F) Semi-quantitative scoring analysis of IHC staining results for IRF2 and Ki-67 proteins in nude mice subcutaneous xenograft tumors. Data are represented as mean ± SEM. Statistical significance was determined by two-tailed Student’s t test. (∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001). In vivo validation of the effect of IRF2 on the proliferation of nasopharyngeal carcinoma cells in nude mice subcutaneous xenograft models (A and C) Subcutaneous xenograft assay to evaluate the regulatory effect of IRF2 knockdown on the proliferative capacity of nasopharyngeal carcinoma cells in vivo. (B) Growth curves of subcutaneous xenograft tumor volumes. Data are represented as mean ± SEM (n = 5 mice per group). (D) Statistical analysis of subcutaneous xenograft tumor weights. Data are represented as mean ± SEM. (E) Histological characteristics of subcutaneous xenograft tumors from each group were analyzed by HE staining, and the expression levels of IRF2 and Ki-67 proteins in xenograft tissues were detected by IHC. Representative micrographs are supplied with a scale bar, 100 μm. (F) Semi-quantitative scoring analysis of IHC staining results for IRF2 and Ki-67 proteins in nude mice subcutaneous xenograft tumors. Data are represented as mean ± SEM. Statistical significance was determined by two-tailed Student’s t test. (∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001). Effect of IRF2 on epithelial-mesenchymal transition in NPC cells Based on the aforementioned findings, the expression level of IRF2 was significantly positively correlated with the invasion, migration, and proliferation abilities of NPC, suggesting that IRF2 may function as an oncogene involved in regulating malignant progression. To further elucidate its molecular mechanism, and considering the crucial role of EMT in tumor metastasis, we hypothesized whether IRF2 might drive the malignant phenotype of NPC by regulating the EMT process. We examined the levels of key EMT-related molecules following IRF2 overexpression and knockdown via WB analysis. The results demonstrated that after IRF2 overexpression, the expression of the epithelial marker E-cadherin was significantly downregulated, while the expression levels of the mesenchymal markers N-cadherin and vimentin, as well as the EMT transcriptional regulators snail1, snail2 (slug) and ZEB1 were markedly upregulated (p < 0.05). Conversely, in the IRF2 knockdown group, E-cadherin expression significantly increased (p < 0.05), whereas the expression levels of the aforementioned mesenchymal markers and regulatory factors were significantly reduced (Figure 10). These results confirm that IRF2 may participate in the formation of the malignant phenotype in NPC by regulating EMT.Figure 10Effects of altered IRF2 expression levels on epithelial-mesenchymal transition (EMT)-related markers in nasopharyngeal carcinoma cells(A) Western blot analysis of the impact of IRF2 overexpression on EMT-related protein expression in 58F nasopharyngeal carcinoma cells.(B) Quantitative analysis of EMT marker expression levels following IRF2 overexpression. Data are represented as mean ± SEM.(C) Western blot analysis of the effect of IRF2 knockdown on EMT-related protein expression in CNE1 nasopharyngeal carcinoma cells.(D) Quantitative statistical analysis of the impact of IRF2 knockdown on the Wnt/β-catenin signaling pathway and EMT marker expression levels. Data are represented as mean ± SEM. Statistical significance was determined by two-tailed Student’s t test. (∗p < 0.05, ∗∗p < 0.01, ∗∗∗∗p < 0.0001). Effects of altered IRF2 expression levels on epithelial-mesenchymal transition (EMT)-related markers in nasopharyngeal carcinoma cells (A) Western blot analysis of the impact of IRF2 overexpression on EMT-related protein expression in 58F nasopharyngeal carcinoma cells. (B) Quantitative analysis of EMT marker expression levels following IRF2 overexpression. Data are represented as mean ± SEM. (C) Western blot analysis of the effect of IRF2 knockdown on EMT-related protein expression in CNE1 nasopharyngeal carcinoma cells. (D) Quantitative statistical analysis of the impact of IRF2 knockdown on the Wnt/β-catenin signaling pathway and EMT marker expression levels. Data are represented as mean ± SEM. Statistical significance was determined by two-tailed Student’s t test. (∗p < 0.05, ∗∗p < 0.01, ∗∗∗∗p < 0.0001). IRF2 regulates EMT in NPC cells in association with the Wnt/β-catenin signaling pathway Previous studies have demonstrated that IRF2 overexpression significantly promotes EMT in NPC cells, while its knockdown effectively inhibits this process. However, the specific molecular mechanism by which IRF2 regulates EMT remains unclear. This section aims to explore the potential signaling pathways and regulatory networks that may be involved in IRF2-mediated EMT in NPC, while acknowledging that the current data support associative rather than definitive causal relationships. Given the limited availability of public genomic databases specifically for NPC and its shared biological characteristics with HNSC, we initially utilized the GEPIA database (http://GEPIA.cancer-pku.cn/) to analyze the correlation between IRF2 and β-catenin gene expression within the HNSC dataset. The analysis yielded a p value of 0 and a correlation coefficient (R) of 0.41, suggesting a significant correlation between IRF2 and β-catenin (Figure 11). This prompted the investigation of whether IRF2 potentially regulates β-catenin in NPC. Subsequent validation via WB analysis revealed that upregulating IRF2 led to decreased expression of GSK3β, a key protein in the Wnt/β-catenin pathway, and increased expression of β-catenin (p < 0.05). Conversely, IRF2 knockdown produced the opposite effect (p < 0.05, Figure 11). These results were consistent with the possibility that IRF2 regulates EMT in NPC cells through the Wnt/β-catenin signaling pathway.Figure 11Changes of Wnt/β-catenin signaling pathway related protein indexes after overexpression or knockdown of IRF2(A) Relationship between IRF2 and β-catenin.(B) Gray scale graph of the effect of overexpression of IRF2 Wnt/β-catenin signaling pathway proteins in 58F nasopharyngeal carcinoma cells by western blot assay.(C) Statistical graph of the effect of overexpression of IRF2 on Wnt/β-catenin pathway proteins. Data are represented as mean ± SEM.(D) WB assay gray scale graph of the effect of knockdown of IRF2 Wnt/β-catenin signaling pathway protein in CNE1 nasopharyngeal carcinoma cells.(E) Statistical graph of the effect of knockdown of IRF2 on Wnt/β-catenin pathway protein. Data are represented as mean ± SEM. Statistical significance was determined by two-tailed Student’s t test. (∗p < 0.05, ∗∗p < 0.01). Changes of Wnt/β-catenin signaling pathway related protein indexes after overexpression or knockdown of IRF2 (A) Relationship between IRF2 and β-catenin. (B) Gray scale graph of the effect of overexpression of IRF2 Wnt/β-catenin signaling pathway proteins in 58F nasopharyngeal carcinoma cells by western blot assay. (C) Statistical graph of the effect of overexpression of IRF2 on Wnt/β-catenin pathway proteins. Data are represented as mean ± SEM. (D) WB assay gray scale graph of the effect of knockdown of IRF2 Wnt/β-catenin signaling pathway protein in CNE1 nasopharyngeal carcinoma cells. (E) Statistical graph of the effect of knockdown of IRF2 on Wnt/β-catenin pathway protein. Data are represented as mean ± SEM. Statistical significance was determined by two-tailed Student’s t test. (∗p < 0.05, ∗∗p < 0.01). IRF2 regulates EMT in NPC cells through modulation of β-catenin nuclear translocation The above findings indicate that IRF2 may influence EMT in NPC via the Wnt/β-catenin signaling pathway. As the central signaling hub of this pathway, β-catenin plays a pivotal connecting role. To further verify whether IRF2 regulates the nuclear translocation of β-catenin, we selected two NPC cell lines (58F and CNE1) and again constructed knockdown and overexpression models. Stable IRF2 expression was confirmed by WB analysis (Figure 12). Subsequently, nuclear and cytoplasmic proteins were fractionated, and the expression levels of β-catenin in these compartments were assessed by WB using lamin A/C and GAPDH as nuclear and cytoplasmic loading controls respectively. The results demonstrated that in both CNE1 and 58F cells, IRF2 knockdown reduced nuclear accumulation of β-catenin (p < 0.05), whereas IRF2 overexpression increased it (p < 0.05, Figure 13). Furthermore, immunofluorescence staining confirmed that IRF2 overexpression enhanced β-catenin nuclear localization in 58F cells (p < 0.05), while IRF2 knockdown decreased it in CNE1 cells (p < 0.05) (Figure 14).Figure 12Constructing IRF2 knockdown and overexpression nasopharyngeal carcinoma cell models(A) Gray scale graph of overexpression of IRF2 protein expression in CNE1 nasopharyngeal carcinoma cells detected by western blot.(B) Statistical graph of overexpression of IRF2 protein expression in CNE1 nasopharyngeal carcinoma cells. Data are represented as mean ± SEM.(C) Gray scale graph of knockdown of IRF2 protein expression in 58F nasopharyngeal carcinoma cells detected by western blot.(D) Statistical graph of knockdown of IRF2 protein expression in 58F nasopharyngeal carcinoma cells. Data are represented as mean ± SEM. Statistical significance was determined by two-tailed Student’s t test. (∗p < 0.05).Figure 13IRF2 regulates the nuclear translocation of β-catenin(A and C) Grayscale images showing β-catenin protein expression in the nucleus and cytoplasm of 58F nasopharyngeal carcinoma cells following overexpression or knockdown of IRF2.(B and D) Quantitative analysis of β-catenin protein expression in the nucleus and cytoplasm of 58F nasopharyngeal carcinoma cells after IRF2 overexpression or knockdown. Data are represented as mean ± SEM.(E and G) Quantitative analysis of β-catenin protein expression in the nucleus and cytoplasm of CNE1 nasopharyngeal carcinoma cells after IRF2 overexpression or knockdown.(F and H) Quantitative analysis of β-catenin protein expression in the nucleus and cytoplasm of CNE1 nasopharyngeal carcinoma cells following IRF2 overexpression or knockdown. Data are represented as mean ± SEM. Statistical significance was determined by two-tailed Student’s t test. (∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001).Figure 14Immunofluorescence assay of the effect of IRF2 on β-catenin nuclear translocation(A) Immunofluorescence analysis of β-catenin expression in the nucleus and cytoplasm of CNE1 nasopharyngeal carcinoma cells following IRF2 knockdown (400×).(B) Immunofluorescence analysis of β-catenin expression in the nucleus and cytoplasm of 58F nasopharyngeal carcinoma cells following IRF2 overexpression (400×). Representative images from each condition are shown with a scale bar, 25 μm. Constructing IRF2 knockdown and overexpression nasopharyngeal carcinoma cell models (A) Gray scale graph of overexpression of IRF2 protein expression in CNE1 nasopharyngeal carcinoma cells detected by western blot. (B) Statistical graph of overexpression of IRF2 protein expression in CNE1 nasopharyngeal carcinoma cells. Data are represented as mean ± SEM. (C) Gray scale graph of knockdown of IRF2 protein expression in 58F nasopharyngeal carcinoma cells detected by western blot. (D) Statistical graph of knockdown of IRF2 protein expression in 58F nasopharyngeal carcinoma cells. Data are represented as mean ± SEM. Statistical significance was determined by two-tailed Student’s t test. (∗p < 0.05). IRF2 regulates the nuclear translocation of β-catenin (A and C) Grayscale images showing β-catenin protein expression in the nucleus and cytoplasm of 58F nasopharyngeal carcinoma cells following overexpression or knockdown of IRF2. (B and D) Quantitative analysis of β-catenin protein expression in the nucleus and cytoplasm of 58F nasopharyngeal carcinoma cells after IRF2 overexpression or knockdown. Data are represented as mean ± SEM. (E and G) Quantitative analysis of β-catenin protein expression in the nucleus and cytoplasm of CNE1 nasopharyngeal carcinoma cells after IRF2 overexpression or knockdown. (F and H) Quantitative analysis of β-catenin protein expression in the nucleus and cytoplasm of CNE1 nasopharyngeal carcinoma cells following IRF2 overexpression or knockdown. Data are represented as mean ± SEM. Statistical significance was determined by two-tailed Student’s t test. (∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001). Immunofluorescence assay of the effect of IRF2 on β-catenin nuclear translocation (A) Immunofluorescence analysis of β-catenin expression in the nucleus and cytoplasm of CNE1 nasopharyngeal carcinoma cells following IRF2 knockdown (400×). (B) Immunofluorescence analysis of β-catenin expression in the nucleus and cytoplasm of 58F nasopharyngeal carcinoma cells following IRF2 overexpression (400×). Representative images from each condition are shown with a scale bar, 25 μm. Discussion Expression of IRF2 in NPC and its clinical significance NPC is a malignant tumor with a distinct geographical distribution, showing a particularly high incidence in Southeast Asia and southern China.33 Current treatment for NPC primarily relies on a comprehensive approach involving radiotherapy, chemotherapy, immunotherapy, and targeted therapy. Recent clinical studies and meta-analyses indicate that induction chemotherapy followed by concurrent chemoradiotherapy significantly improves the overall survival (OS), progression-free survival (PFS), and distant metastasis-free survival (DMFS) of NPC patients.34,35 However, despite advancements in therapeutic strategies, some patients still face challenges such as treatment resistance, recurrence, and metastasis. Therefore, exploring effective therapeutic targets, establishing personalized precision treatment models and optimizing therapeutic strategies have become urgent needs to overcome current treatment bottlenecks and improve long-term patient prognosis. Previous studies have confirmed that IRF2 plays diverse roles in various cancers and acts as an oncogene in NPC tissues.36 Nevertheless, the precise mechanisms of IRF2 in NPC development and signaling transduction remain incompletely elucidated. Based on this, our study aims to further investigate the intrinsic molecular mechanisms of IRF2 in NPC pathogenesis, intending to provide new theoretical foundations and potential targets for the precision treatment of NPC. Firstly, this study revealed the expression profile of IRF2 in NPC and its potential clinical significance through multi-level experimental analyses. Preliminary immunohistochemistry results demonstrated that the expression of IRF2 was significantly higher in NPC tissues and cell lines compared to chronic nasopharyngeal inflammatory tissues and the normal nasopharyngeal epithelial cell line NP69, with statistical significance. This finding is consistent with previous reports,36 further supporting the role of IRF2 as an oncogene in NPC. The clinical significance of IRF2 varies across different cancer types. For instance, high IRF2 expression is associated with poorer overall survival in pancreatic cancer37,38; in liver cancer, IRF2 affects patient prognosis by regulating the SOCS family39; in cervical cancer, IRF2 is considered a potential diagnostic marker40; in childhood acute myeloid leukemia, IRF2 can predict disease prognosis and relapse41; in glioma, high IRF2 levels correlate with more advanced tumor grades and worse outcomes, potentially serving as a prognostic biomarker and indicator of immune status42; IRF2 has been used to construct clinical prognostic models for predicting outcomes in breast cancer.43 ROC curve analysis suggested that IRF2 might have potential utility in distinguishing chronic nasopharyngeal inflammation from NPC, with an AUC of 0.837 (95% CI: 0.760–0.915). At a cutoff value of 168.17, the sensitivity was 94.44% and the specificity was 58.73%. Although the high sensitivity hints at a possible role for preliminary screening, the relatively low specificity and the lack of an independent validation cohort indicate that IRF2 should be considered a candidate biomarker at this stage. Its clinical applicability requires confirmation through larger, multi-center prospective studies with long-term follow-up. Our results support further investigation of IRF2 as a potential auxiliary diagnostic marker for NPC, rather than a definitive clinical tool. Furthermore, based on the IRF2 expression levels from 64 NPC patients categorized into high and low expression groups according to the cut-off value, we found that IRF2 expression was significantly correlated with patients’ T stage and plasma EBV DNA load, but not with smoking history, N stage or M stage. This suggests that IRF2 may play an important role in the local invasiveness of NPC. EBV is closely associated with NPC, and elevated viral DNA levels typically indicate active viral replication. IRF2 may promote NPC development by facilitating EBV replication.44,45 Although this study preliminarily confirmed that IRF2 is highly expressed in NPC and correlated with T stage and EBV DNA load, the sample size is limited, and the study is a single-center retrospective analysis. The oncogenic mechanisms of IRF2 require further in-depth investigation. Future studies should expand the cohort and conduct functional experiments to validate the causal role of IRF2. Additionally, does IRF2 influence tumor progression by regulating the EBV life cycle or interferon signaling pathways? In summary, IRF2 holds promise as a potential target for individualized NPC therapy, but its clinical translation necessitates further exploration of its molecular mechanisms and optimization of diagnostic models. Investigation into the regulatory role and mechanism of IRF2 in NPC proliferation and invasion The interferon regulatory factor family, a class of evolutionarily conserved transcriptional regulatory proteins, participates in various pathophysiological regulations by coordinating processes such as innate immune responses and cell fate determination. Among them, IRF2 has attracted significant attention due to its unique “double-edged sword” characteristic. IRF2 can function as either an oncogene or a tumor suppressor in human tumorigenesis. For example, IRF2 acts as an oncogene in liver cancer32 and rectal cancer,46 while serving as a tumor suppressor in gastric cancer,23 lung cancer,47 and osteosarcoma.36 This functional heterogeneity suggests that the oncogenic tendency of IRF2 may be finely regulated by the tissue microenvironment, co-regulatory factor networks, or epigenetic modifications. This study focuses on the role of IRF2 in NPC and, for the first time, reveals its clear pro-malignant phenotype in this context. A series of functional experiments confirmed that overexpression of IRF2 significantly enhanced cell proliferation activity and maintained a survival advantage by inhibiting apoptotic pathways. Furthermore, transwell assays demonstrated that IRF2 overexpression increased the number of migrated and invasive cells, while IRF2 knockdown decreased these capabilities. In wound healing assays, IRF2 overexpression also accelerated wound closure. However, the observed changes in wound closure could be partially attributable to these processes rather than solely to enhanced cell migration, because IRF2 simultaneously promotes proliferation and suppresses apoptosis. Thus, the wound healing results should be interpreted as supporting evidence rather than definitive proof of altered migratory capacity. Nevertheless, the consistent trends observed across multiple functional assays (including transwell migration, which directly measures cell movement) support a role for IRF2 in regulating NPC cell migration. This study further discovered through subcutaneous tumor formation experiments in nude mice that knocking down IRF2 significantly inhibited tumor growth, suggesting that the pro-metastatic property of IRF2 might be related to its ability to remodel the microenvironment. Does IRF2 promote the invasion and migration of NPC cells by regulating the EMT process? We further explored the underlying molecular mechanisms via WB experiments. Further studies indicated that IRF2 overexpression significantly induced downregulation of E-cadherin and upregulation of N-cadherin, snail1, snail2, and ZEB1, while knockdown produced the opposite effect. These findings demonstrate the oncogenic role of IRF2 in NPC from both functional phenotypic and molecular mechanistic perspectives. Existing literature suggests that β-catenin, the core effector molecule of the Wnt signaling pathway, can directly drive the EMT process by activating transcription factors such as snail and twist upon its nuclear accumulation.27 In cancer, nuclear localization of β-catenin is closely associated with tumor invasion, metastasis, and chemoresistance. Given the limited public genomic database resources for NPC and its shared biological characteristics with HNSC, we performed a correlation analysis between IRF2 and β-catenin gene expression in the HNSC dataset using the GEPIA database. The results indicated a high correlation between IRF2 and β-catenin. Therefore, we asked whether IRF2 regulates EMT in NPC cells in association with the Wnt/β-catenin pathway. To explore this possibility, we first validated changes in the expression levels of key Wnt/β-catenin pathway proteins, GSK-3β and β-catenin, following IRF2 knockdown or overexpression using WB. Subsequently, nuclear and cytoplasmic protein fractionation experiments were conducted to extract cytoplasmic and nuclear proteins. WB analysis detected changes in β-catenin localization and IF staining assessed β-catenin distribution in the cytoplasm and nucleus upon IRF2 overexpression. The results indicated that IRF2 overexpression downregulated GSK-3β, increased β-catenin expression levels, and promoted its nuclear translocation. Conversely, IRF2 knockdown yielded opposite results. This discovery expands the understanding of the IRF family proteins’ functions in tumor invasion and proliferation, providing a new perspective on the molecular mechanisms of EMT in NPC. This study suggests that IRF2 may be an upstream driver of this process. The positive correlation between IRF2 expression levels and β-catenin nuclear localization indicates that IRF2 overexpression may promote malignant progression by activating the Wnt/β-catenin pathway. We consider several potential mechanisms underpinning this interaction: (1) IRF2 may influence EMT by regulating the subcellular distribution of β-catenin. As a transcription factor, the nuclear localization signal (NLS) of IRF2 might assist β-catenin in crossing the nuclear membrane. We plan to further investigate potential direct interaction between IRF2 and β-catenin through co-immunoprecipitation (Co-IP) or fluorescence resonance energy transfer (FRET) experiments. (2) IRF2 may affect the stability or transcriptional activity of β-catenin by transcriptionally regulating genes such as Axin, APC, or TCF4. For instance, the APC protein can inhibit β-catenin nuclear accumulation by promoting its degradation.48 If IRF2 downregulates APC expression, it might indirectly lead to β-catenin accumulation. (3) β-catenin is typically anchored to the cell membrane by binding to E-cadherin, and the loss of E-cadherin during EMT may release β-catenin for nuclear translocation.49 In this study, IRF2 overexpression led to E-cadherin downregulation, suggesting that IRF2 might disrupt the E-cadherin/β-catenin complex, promoting the migration of free β-catenin into the nucleus. In summary, IRF2 is predominantly localized in the nucleus of NPC cells and may function as an oncogene potentially involved in local invasion and EBV replication. IRF2 appears to drive the progression of NPC by promoting cell proliferation, migration, invasion, and EMT, while simultaneously inhibiting apoptosis. These effects are potentially associated with the activation of the Wnt/β-catenin signaling pathway, though the causal relationship requires further investigation. Additionally, the in vivo validation of this proposed mechanism was not performed in the current xenograft model, representing an important direction for future research. Finally, the potential clinical utility of IRF2 as a diagnostic or prognostic biomarker in NPC warrants further evaluation in larger, independent cohorts. Limitations of the study Several limitations of the present study should be acknowledged. First, although our data demonstrate a strong correlation between IRF2 expression, Wnt/β-catenin pathway activation, and EMT phenotypes, direct causal evidence linking IRF2-mediated Wnt/β-catenin activation to these functional outcomes is lacking. Rescue experiments (e.g., simultaneous β-catenin knockdown or Wnt pathway inhibition in IRF2-overexpressing cells, and Wnt activation in IRF2-knockdown cells) are needed to establish causality. Second, the precise molecular mechanism by which IRF2 promotes β-catenin nuclear translocation remains unexplored—whether through direct protein-protein interaction, transcriptional regulation of β-catenin or its regulators (e.g., APC, Axin), or disruption of the E-cadherin/β-catenin complex. Third, the clinical sample size is relatively modest (n = 64) and derived from a single center, with no external validation cohort or long-term follow-up survival data; the diagnostic specificity was also limited (58.73%). Therefore, larger, multi-center prospective studies with comprehensive follow-up are required to validate the diagnostic and prognostic value of IRF2 in NPC. Fourth, although the subcutaneous xenograft experiments demonstrated that IRF2 knockdown significantly inhibited tumor growth, we did not perform IHC or WB analyses to assess Wnt/β-catenin pathway activity or EMT marker expression within the harvested tumor tissues. Consequently, whether the proposed IRF2-Wnt/β-catenin-EMT regulatory axis operates in the in vivo tumor microenvironment remains to be directly validated in future studies. Additionally, combining clinical NPC samples to systematically examine the correlation between IRF2 protein expression and β-catenin nuclear translocation, along with the spatial co-localization of EMT markers, is necessary. Finally, based on clinical follow-up data, assessing the prognostic and predictive value of the co-expression patterns of these molecular features for patient outcomes will be crucial. Resource availability Lead contact Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Feng Jin (tjzlk2023@163.com). Materials availability All unique reagents generated in this study are available from the lead contact with a completed Materials Transfer Agreement. Data and code availability •Data: This study did not generate or deposit any new large-scale datasets or public repositories. All data supporting the findings of this study are presented within the manuscript and its supplementary materials. No accession codes are applicable.•Code: No custom software, algorithms, or code were generated or used for this study beyond standard commercial and open-source tools listed in the key resources table.•Other items: Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request. •Data: This study did not generate or deposit any new large-scale datasets or public repositories. All data supporting the findings of this study are presented within the manuscript and its supplementary materials. No accession codes are applicable. Data: This study did not generate or deposit any new large-scale datasets or public repositories. All data supporting the findings of this study are presented within the manuscript and its supplementary materials. No accession codes are applicable. •Code: No custom software, algorithms, or code were generated or used for this study beyond standard commercial and open-source tools listed in the key resources table. Code: No custom software, algorithms, or code were generated or used for this study beyond standard commercial and open-source tools listed in the key resources table. •Other items: Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request. Other items: Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request. Acknowledgments This research is supported by the 10.13039/501100001809National Natural Science Foundation of China (grant no. 82060556) and the 10.13039/501100018555Science and Technology Program of Guizhou Province (grant no. Qiankehe Foundation—Z.K.[2024] General 189). Meanwhile, we are also grateful to Ding’an Zhou from the Affiliated Hospital of Guizhou Medical University and Kai Shang from the Chongqing University Cancer Hospital for their assistance with this study. Author contributions Conceptualization, C.Z. and J.S.; data curation, J.S. and X.L.; formal analysis, C.Z., J.S., and X.L.; funding acquisition, F.J. and C.Z.; investigation, Y.C.; methodology, C.Z. and J.S.; project administration, F.J.; supervision, F.J.; validation, C.Z., J.S., and X.L.; visualization, J.S., X.L., and Y.C.; writing – original draft, J.S. and X.L.; writing – review & editing, C.Z., J.S., X.L., and F.J. The published version of the manuscript has been read and agreed by all authors. Declaration of interests The authors declare no competing interests. STAR★Methods Key resources table REAGENT or RESOURCESOURCEIDENTIFIERAntibodiesIRF2 Polyclonal antibodyProteintech, Wuhan, ChinaCat#12525-1-APE-cadherin Polyclonal antibodyProteintech, Wuhan, ChinaCat#20874-1-APN-cadherin Polyclonal antibodyProteintech, Wuhan, ChinaCat#22018-1-APVimentin Polyclonal antibodyProteintech, Wuhan, ChinaCat#10366-1-APSnail1 Polyclonal antibodyProteintech, Wuhan, ChinaCat#13099-1-APSnail2/Slug Polyclonal antibodyProteintech, Wuhan, ChinaCat#12129-1-APZEB1 Polyclonal antibodyProteintech, Wuhan, ChinaCat#21544-1-APGSK3β Polyclonal antibodyProteintech, Wuhan, ChinaCat#51065-1-APBeta-catenin Polyclonal antibodyProteintech, Wuhan, ChinaCat#51067-1-APLamin A/C Polyclonal antibodyProteintech, Wuhan, ChinaCat#10298-1-APGAPDH Polyclonal antibodyProteintech, Wuhan, ChinaCat#10494-1-APKi-67 Polyclonal antibodyProteintech, Wuhan, ChinaCat#27309-1-APHRP-Goat Anti-Rabbit IgG (H + L)Proteintech, Wuhan, ChinaCat#RGAR001Bacterial and virus strainsLentivirus for IRF2 overexpression (GV492 vector)GeneChem, Shanghai, ChinaCat#GV492Lentivirus for IRF2 knockdown (GV493 vector, shRNA)GeneChem, Shanghai, ChinaCat#GV493Software and algorithmsGraphPad Prism 8.2GraphPad Software, San Diego, CA, USARRID: SCR_002798SPSS 27.0IBM, Armonk, NY, USARRID: SCR_002865ImageJNational Institutes of Health, Bethesda, MD, USARRID: SCR_003070FlowJo 10.0BD Biosciences, Ashland, OR, USARRID: SCR_008520GEPIA2Zhang Lab, Peking Universityhttp://gepia2.cancer-pku.cnHuman Protein Atlas (HPA)HPA Consortiumhttps://www.proteinatlas.org REAGENT or RESOURCESOURCEIDENTIFIERAntibodiesIRF2 Polyclonal antibodyProteintech, Wuhan, ChinaCat#12525-1-APE-cadherin Polyclonal antibodyProteintech, Wuhan, ChinaCat#20874-1-APN-cadherin Polyclonal antibodyProteintech, Wuhan, ChinaCat#22018-1-APVimentin Polyclonal antibodyProteintech, Wuhan, ChinaCat#10366-1-APSnail1 Polyclonal antibodyProteintech, Wuhan, ChinaCat#13099-1-APSnail2/Slug Polyclonal antibodyProteintech, Wuhan, ChinaCat#12129-1-APZEB1 Polyclonal antibodyProteintech, Wuhan, ChinaCat#21544-1-APGSK3β Polyclonal antibodyProteintech, Wuhan, ChinaCat#51065-1-APBeta-catenin Polyclonal antibodyProteintech, Wuhan, ChinaCat#51067-1-APLamin A/C Polyclonal antibodyProteintech, Wuhan, ChinaCat#10298-1-APGAPDH Polyclonal antibodyProteintech, Wuhan, ChinaCat#10494-1-APKi-67 Polyclonal antibodyProteintech, Wuhan, ChinaCat#27309-1-APHRP-Goat Anti-Rabbit IgG (H + L)Proteintech, Wuhan, ChinaCat#RGAR001Bacterial and virus strainsLentivirus for IRF2 overexpression (GV492 vector)GeneChem, Shanghai, ChinaCat#GV492Lentivirus for IRF2 knockdown (GV493 vector, shRNA)GeneChem, Shanghai, ChinaCat#GV493Software and algorithmsGraphPad Prism 8.2GraphPad Software, San Diego, CA, USARRID: SCR_002798SPSS 27.0IBM, Armonk, NY, USARRID: SCR_002865ImageJNational Institutes of Health, Bethesda, MD, USARRID: SCR_003070FlowJo 10.0BD Biosciences, Ashland, OR, USARRID: SCR_008520GEPIA2Zhang Lab, Peking Universityhttp://gepia2.cancer-pku.cnHuman Protein Atlas (HPA)HPA Consortiumhttps://www.proteinatlas.org Experimental model and study participant details Human subjects Formalin-fixed, paraffin-embedded tissue samples were obtained from patients with nasopharyngeal carcinoma (NPC; diagnosed between February 2018 and June 2020) and chronic nasopharyngitis (diagnosed between January 2013 and April 2014) at the Department of Pathology, Affiliated Cancer Hospital of Guizhou Medical University (Guiyang, China). A total of 63 NPC patients (47 males, 16 females; age range not specified) and 63 chronic nasopharyngitis patients were included. The sex distribution and age information are reported in Table 1. The influence of sex on IRF2 expression was analyzed (p = 0.412, no significant association). Age of subjects is reported as ≥60 vs. <60 years. The study was approved by the Ethics Committee of the Affiliated Hospital of Guizhou Medical University (approval number: 2023-575). Written informed consent was obtained from all patients prior to sample collection and the informed consent process strictly followed the ethical guidelines of the Declaration of Helsinki. Cell lines The immortalized human nasopharyngeal epithelial cell line NP69 (gifted by the Li Ka Shing Faculty of Medicine, The University of Hong Kong, Hong Kong, China) and NPC cell lines including CNE1, HONE1, HK-1, 58F and 6-10B were used. All NPC cell lines were gifted by the Sun Yat-sen University Cancer Center (Guangzhou, China). The sex of these cell lines is not specified by the original providers. Cells were maintained under standard culture conditions (37°C, 5% CO2) as described in method details. All cell lines were authenticated by short tandem repeat (STR) profiling and routinely tested negative for mycoplasma contamination. Animal models Immunodeficient male nude mice (4–5 weeks old, weighing 14–15 g) were purchased from Vital River Laboratory Animal Technology (Beijing, China). Only male mice were used. Therefore, the influence of sex on the results could not be assessed. Mice were maintained under specific pathogen-free (SPF) conditions with controlled temperature (22°C–25°C) and humidity (40–50%), a 12-h light/dark cycle, and free access to food and water. All animal experiments were approved by the Animal Care Welfare Committee of Guizhou Medical University (approval number: 2304694). Method details Cell culture Cells were cultured following standard protocols. Cryopreserved cells were rapidly thawed in a 37°C water bath, resuspended in complete medium (RPMI-1640 or DMEM supplemented with 10% fetal bovine serum and 1% penicillin/streptomycin), and centrifuged. The pellet was seeded into T25 flasks and maintained in a humidified incubator at 37°C with 5% CO2. For subculturing, cells at 70–80% confluence were washed with PBS, detached using trypsin, and neutralized with complete medium. After centrifugation, the cell pellet was resuspended and passaged at a 1:3 to 1:4 split ratio. Cell concentration was determined using an automated cell counter. For cryopreservation, harvested cells were resuspended in serum-free cryopreservation medium, stored at −80°C, and subsequently transferred to liquid nitrogen. Vector construction and stable transfection Lentiviral vectors for IRF2 overexpression (GV492) and knockdown (GV493) were constructed and packaged by GeneChem (Shanghai, China). Viral titers were 5×108 TU/mL (overexpression) and 1–3×109 TU/mL (knockdown). For transfection, cells were seeded in 6-well plates to 70–80% confluence. Medium was replaced with 1 mL fresh medium containing 40 μL enhancer solution, and lentivirus was added at appropriate MOI. After 16 h, viral supernatant was replaced with complete medium. GFP expression was monitored at 48 h; when transfection efficiency exceeded 90%, cells were selected with puromycin for 14 days to establish stable lines. IRF2 expression was confirmed by WB. Nuclear and cytoplasmic protein extraction Nuclear and cytoplasmic proteins were extracted using a commercial subcellular fractionation kit (according to manufacturer’s instructions). Cells in logarithmic growth phase were harvested, washed with PBS, and resuspended in pre-cooled cytoplasmic extraction reagent A containing 1 mM PMSF. After vortexing and ice incubation, reagent B was added. The lysate was centrifuged (12,000–16,000 × g, 5 min, 4°C) to obtain supernatant (cytoplasmic fraction). The nuclear pellet was washed and lysed in nuclear extraction buffer with 1 mM PMSF, incubated on ice for 30 min with intermittent vortexing, and centrifuged (12,000–16,000 × g, 10 min, 4°C) to obtain nuclear fraction. Both fractions were mixed with 5× loading buffer, boiled, and stored at −80°C. WB Proteins were extracted using ice-cold RIPA buffer supplemented with protease inhibitors, followed by ultrasonication and centrifugation. Protein concentration was determined by BCA assay. Equal amounts of protein were separated by SDS-PAGE and transferred onto methanol-activated PVDF membranes. Membranes were blocked with 5% skim milk, incubated overnight at 4°C with primary antibodies (see key resources table), then with HRP-conjugated secondary antibodies. Bands were visualized using enhanced chemiluminescence (ECL) and analyzed with ImageJ software. Immunofluorescence staining Cells seeded on coverslips were fixed with 4% paraformaldehyde, permeabilized with 0.2% Triton X-100, and blocked with 5% BSA. Primary antibodies (anti-IRF2, anti-β-catenin) were applied overnight at 4°C, followed by fluorophore-conjugated secondary antibodies (in the dark). Nuclei were counterstained with DAPI. Coverslips were mounted with 90% glycerol and sealed before imaging. Wound healing assay Cells were seeded in 6-well plates pre-marked with parallel reference lines. At >90% confluence, cells were serum-starved for 12 h, then a uniform scratch wound was created using a sterile pipette tip. Detached cells were washed away with PBS, and serum-free medium was added. Images were taken at 0 h and 24 h under an inverted phase-contrast microscope. Wound closure was quantified by measuring the remaining cell-free area using ImageJ. Experiments were performed in triplicate and repeated three times. Colony formation assay Cells (500–1000 per well) were seeded in 6-well plates and cultured for 10–14 days with periodic medium changes. Visible colonies were fixed with 4% paraformaldehyde (30 min) and stained with crystal violet (10 min). Colonies were imaged and counted using ImageJ. Transwell migration and invasion assay Transwell chambers (Corning) were used. For invasion assays, the upper chamber was pre-coated with Matrigel (1:8 dilution in serum-free medium) and polymerized overnight. Serum-starved cells (3–7 × 104 in 200 μL serum-free medium) were seeded into the upper chamber; the lower chamber contained 700 μL complete medium with 20% FBS as chemoattractant. After 24–48 h, non-migrated/invaded cells on the upper membrane surface were removed with a cotton swab. Cells that traversed the membrane were fixed, stained with crystal violet, and imaged. The number of migrated/invaded cells was quantified from multiple random fields using ImageJ. CCK-8 assay Cells (1–3 × 103 per well in 200 μL) were seeded in 96-well plates. At 0, 24, 48 and 72 h, medium was replaced with 100 μL fresh medium containing 10% CCK-8 reagent, incubated at 37°C for 2 h in the dark. Absorbance at 450 nm was measured using a microplate reader. All experiments had at least three replicate wells. Data were analyzed using GraphPad Prism. Apoptosis detection by flow cytometry Apoptosis was assessed using an Annexin V-APC/7-AAD staining kit. Cells (1–5 × 105) were resuspended in 500 μL 1× binding buffer, stained with 5 μL Annexin V-APC and 10 μL 7-AAD, and incubated for 5 min at room temperature in the dark. Flow cytometry was performed within 1 h. Data were analyzed with FlowJo 10.0 and SPSS 27.0. HE staining Tumor tissues were fixed in 4% paraformaldehyde, dehydrated, cleared, and paraffin-embedded. Sections (3–5 μm) were stained with hematoxylin (5 min), differentiated in 1% acid alcohol, counterstained with eosin (3 min), dehydrated, cleared, and mounted with neutral resin. IHC IHC was performed on 3–5 μm FFPE sections. After baking, dewaxing, and rehydration, antigen retrieval was done by microwave heating. Endogenous peroxidase was blocked with 3% H2O2, followed by goat serum blocking. Sections were incubated overnight at 4°C with anti-IRF2 primary antibody at optimal dilution. After washing, secondary antibody was applied and signal amplified. Visualization used DAB substrate, followed by hematoxylin counterstaining. Staining intensity was scored 0–3 by two independent pathologists. Immunoreactivity score (0–300) was calculated as: (1 × % weak) + (2 × % moderate) + (3 × % strong). Xenograft animal model Male nude mice (4–5 weeks old, 14–15 g) were randomly allocated to shControl (n = 5) or shIRF2 (n = 5) groups. Exponentially growing CNE1 cells (stable shControl or shIRF2) were harvested, resuspended in pre-cooled PBS at 1×107 cells/mL. Two hundred microliters (2×106 cells) were injected subcutaneously into the right shoulder. Tumor size was measured every two days with a digital caliper; volume = (length × width2)/2. After three weeks, mice were euthanized by cervical dislocation, tumors excised, weighed, and either fixed in 4% paraformaldehyde or snap-frozen in liquid nitrogen. Quantification and statistical analysis Statistical analyses were performed using SPSS 27.0 and GraphPad Prism 8.2. Continuous variables are expressed as mean ± SD. For normally distributed data, independent Student’s t test was used. Otherwise, Mann-Whitney U test. Categorical variables were compared by Chi-square test or Fisher’s exact test as appropriate. Receiver operating characteristic (ROC) curve analysis evaluated diagnostic performance. A two-sided p < 0.05 was considered statistically significant. Exact n values, what n represents (e.g., number of independent experiments, number of animals, number of cells), and the statistical test used for each experiment are indicated in the figure legends. For bar graphs, error bars represent SEM unless otherwise specified. Asterisks in figures denote significance levels (∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001) as defined in each legend.