Cadherin-mediated cell adhesion networks drive tumor growth and therapeutic responses in head and neck cancer

Abstract

Background

Cadherin family genes (CDH1, CDH2, and CDH3) regulate cell–cell adhesion and epithelial integrity and have emerging roles in tumorigenesis. However, their molecular and functional relevance in Head and Neck Squamous Cell Carcinoma (HNSC) remains incompletely defined.

Methods

Cadherin expression was evaluated in HNSC and normal oral epithelial cell lines using RT-qPCR. Multi-omics analyses of transcriptomic, methylation, mutation, and copy number variation (CNV) data were performed using TCGA datasets through GSCA, OncoDB, cBioPortal, and UALCAN. Prognostic significance was assessed using Kaplan–Meier and meta-analysis approaches. miRNA–mRNA interactions were predicted using miRNet and validated experimentally. Immune associations, drug sensitivity, and protein–protein interactions were explored using TISIDB, GSCA, STRING, and DAVID. Functional effects of CDH1 and CDH2 were evaluated by siRNA knockdown in FaDu and SCC9 cells, followed by proliferation, colony formation, wound healing, and xenograft assays.

Results

CDH1, CDH2, and CDH3 were significantly upregulated in HNSC cell lines and TCGA tumor samples, with CDH3 showing the highest diagnostic accuracy (AUC = 0.90). Hypomethylation and CNV amplification contributed to cadherin dysregulation, particularly for CDH2 and CDH3. Elevated cadherin expression correlated with poor overall survival. miR-200a-3p and miR-200c-3p were identified as shared regulators. Cadherin expression also showed associations with immune modulators and drug response. Functionally, CDH1 and CDH2 knockdown suppressed proliferation, migration, and tumor growth.

Conclusion

Cadherin genes, particularly CDH3, show diagnostic and prognostic relevance in HNSC, while CDH1 and CDH2 contribute to tumor cell growth and migration. Further studies are required to determine their therapeutic potential.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12957-026-04427-y.

Introduction

Head and Neck Squamous Cell Carcinoma (HNSC) represents a heterogeneous group of epithelial malignancies arising from the mucosal linings of the oral cavity, pharynx, and larynx [1, 2]. Despite advancements in surgical techniques, radiotherapy, and chemotherapy, the prognosis for HNSC patients remains poor, with a five-year survival rate of less than 50% in advanced stages [3, 4]. This high mortality is primarily due to late diagnosis, aggressive tumor behaviour, and resistance to standard therapies [5, 6]. Therefore, there is a critical need to identify reliable molecular biomarkers that can facilitate early detection, predict clinical outcomes, and guide the development of more effective therapeutic strategies.

Cadherins (CDHs), a superfamily of calcium-dependent transmembrane glycoproteins, are central regulators of cell–cell adhesion, tissue integrity, and cellular signalling [7, 8]. These molecules form adherens junctions by linking with the actin cytoskeleton through intracellular catenins, thereby maintaining epithelial homeostasis and preventing inappropriate cell migration and proliferation [8, 9]. Among them, CDH1 (E-cadherin), CDH2 (N-cadherin), and CDH3 (P-cadherin) are the most widely studied and are known to play pivotal roles in tumorigenesis and metastasis [9]. In various cancers, dysregulation of cadherin expression contributes to disease progression via the epithelial-to-mesenchymal transition (EMT)—a cellular program that endows epithelial cells with mesenchymal properties, including enhanced motility, invasiveness, and resistance to apoptosis [10, 11]. Loss of CDH1, a hallmark of EMT, has been extensively documented in gastric, breast, colorectal, and prostate cancers, where it facilitates tumor invasion and dissemination [12]. Conversely, upregulation of CDH2, often referred to as “cadherin switching,” promotes mesenchymal behaviour and is associated with poor prognosis in glioblastoma, pancreatic, and non-small cell lung cancers [13, 14]. CDH3 overexpression, meanwhile, has been implicated in enhanced cell proliferation and migration in breast, ovarian, and bladder cancers, with evidence suggesting its role in tumor aggressiveness and resistance to therapy [15, 16]. Previous studies have reported altered expression and functional roles of cadherin family members in HNSC progression, invasion, and metastasis [17–22]. However, most of these investigations have focused on individual cadherin genes or specific biological aspects, and a systematic integrative analysis combining multi-omics datasets with experimental validation remains limited. Therefore, the comprehensive regulatory landscape, clinical relevance, and functional implications of multiple cadherin family members in HNSC are not yet fully understood.

To address this gap, the present study investigates the expression patterns, clinical relevance, immune associations, and therapeutic implications of CDH family genes in HNSC using a comprehensive approach. We employed large-scale in silico analyses based on public datasets to examine transcriptional alterations, survival correlations, mutation burden, and interactions with immune modulators. Additionally, in vitro experiments using HNSC cell lines and in vivo assays in xenograft models were conducted to validate the functional roles of selected CDH genes in cancer cell behaviour, including proliferation, invasion, and response to therapeutic agents.

Our findings aim to clarify the multifaceted role of cadherin family members in HNSC pathobiology and to propose novel cadherin-based strategies for cancer diagnosis, prognosis, and treatment. By integrating computational and experimental evidence, this study contributes to the ongoing effort to identify clinically actionable molecular targets in HNSC.

Methodology

Cell lines and culture conditions

A panel of human HNSC cell lines was obtained directly from the American Type Culture Collection (ATCC, Manassas, VA, USA). These included FaDu (hypopharyngeal squamous cell carcinoma; ATCC^® HTB-43™), SCC-9 (tongue squamous cell carcinoma; ATCC^® CRL-1629™), SCC-15 (tongue squamous cell carcinoma; ATCC^® CRL-1623™), SCC-25 (tongue squamous cell carcinoma; ATCC^® CRL-1628™), CAL 27 (tongue squamous cell carcinoma; ATCC^® CRL-2095™), and Detroit 562 (pharyngeal squamous cell carcinoma; ATCC^® CCL-138™). Three control cell lines includes Human Oral Keratinocytes (HOK; ScienCell, Cat. #2610) and HGEPp (CELLnTEC), as well as TR146 cell line (Sigma-Aldrich, ECACC 10032305). All cell lines were authenticated by short tandem repeat (STR) profiling and routinely tested for mycoplasma contamination prior to experimentation. HNSC cell lines (FaDu, SCC-9, SCC-15, SCC-25, CAL 27, and Detroit 562) were cultured according to ATCC recommendations in Dulbecco’s Modified Eagle Medium (DMEM) or DMEM/F-12, supplemented with 10% fetal bovine serum (FBS; Thermo Fisher Scientific) and 1% penicillin–streptomycin, depending on the specific cell line requirements. Primary HOK cells were maintained in Oral Keratinocyte Medium (OKM; ScienCell) supplemented with the manufacturer-provided growth supplement. HGEPp cells were cultured in CnT-Prime epithelial culture medium (CELLnTEC) under serum-free conditions, as recommended by the supplier. The TR146 cell line was maintained in DMEM supplemented with 10% FBS and 1% penicillin–streptomycin, in accordance with ECACC guidelines. All cells were cultured at 37 °C in a humidified incubator with 5% CO₂ and were passaged at approximately 70–80% confluence using 0.25% trypsin–EDTA.

RNA extraction and quantitative real-time PCR (RT-qPCR)

Total RNA was extracted from cultured cell lines using the PureLink™ RNA Mini Kit (Thermo Fisher Scientific), according to the manufacturer’s instructions [23]. First-strand cDNA synthesis was carried out using the High-Capacity cDNA Reverse Transcription Kit (Thermo Fisher Scientific) with 1 µg of total RNA in a final volume of 20 µL. The synthesized cDNA was diluted 1:5 with nuclease-free water and used for real-time PCR analysis. RT-qPCR was performed on a QuantStudio™ 5 Real-Time PCR System (Thermo Fisher Scientific) using PowerUp™ SYBR™ Green Master Mix (Thermo Fisher Scientific) [24]. Gene-specific primers were used to quantify the expression levels of CDH1, CDH2, and CDH3 (Table 1). GAPDH was used as the internal reference gene to normalize the relative expression levels. Relative gene expression was calculated using the 2^{^–ΔΔCt} method [25].

Table 1.

Primer sequences used for RT-qPCR analysis

Gene	Primer Direction	Sequence (5′→3′)
GAPDH	Forward (GAPDH-F)	ACCCACTCCTCCACCTTTGAC
GAPDH	Reverse (GAPDH-R)	CTGTTGCTGTAGCCAAATTCG
CDH1	Forward (CDH1-F)	GCCTCCTGAAAAGAGAGTGGAAG
CDH1	Reverse (CDH1-R)	TGGCAGTGTCTCTCCAAATCCG
CDH2	Forward (CDH2-F)	CCTCCAGAGTTTACTGCCATGAC
CDH2	Reverse (CDH2-R)	GTAGGATCTCCGCCACTGATTC
CDH3	Forward (CDH3-F)	CAGGTGCTGAACATCACGGACA
CDH3	Reverse (CDH3-R)	CTTCAGGGACAAGACCACTGTG

TCGA-based expression validation analysis

Transcriptomic data of CDH1, CDH2, and CDH3 in HNSC were analyzed using the GSCA platform (http://bioinfo.life.hust.edu.cn/GSCA/) [26] based on TCGA-HNSC RNA-seq (FPKM) datasets. Expression was further compared across HPV subgroups (positive vs. negative) and pathological stages (I–IV). Functional relevance was evaluated using Gene Set Enrichment Analysis (GSEA).

Methylation and pathway activity analysis

Promoter methylation profiles of CDH1, CDH2, and CDH3 in HNSC were assessed using TCGA-HNSC data via the OncoDB database (https://oncodb.org) [27]. Beta values (ranging from 0 to 1) were used to compare methylation levels between tumor and normal tissues. To determine the regulatory impact of methylation on gene expression, Spearman correlation analysis was conducted using the GSCA platform [26], and FDR < 0.05 was considered significant. Pathway activity scores were also obtained from GSCA, evaluating the association of cadherin gene expression with ten key oncogenic pathways (e.g., EMT, RTK, hormone receptor, DNA damage response). Pathway regulation was defined based on the direction and strength of gene–pathway correlations.

Mutation and copy number variation (CNV) analysis

Somatic mutation and CNV data for CDH1, CDH2, and CDH3 were analyzed using the cBioPortal for Cancer Genomics (https://www.cbioportal.org) [28] based on the TCGA-HNSC dataset. Mutation types, frequencies, and single-nucleotide variant (SNV) classifications were extracted and visualized. Somatic mutation rates were calculated and mapped to functional domains. CNV profiles were assessed to identify gene-level amplifications and deletions, and results were visualized using on-platform graphical outputs.

Survival and meta-analysis

The prognostic relevance of CDH1, CDH2, and CDH3 expression in HNSC was assessed using Kaplan–Meier Plotter (https://kmplot.com) [29] and GENT2 (http://gent2.appex.kr) [30]. Overall survival (OS) curves were generated for high versus low gene expression groups, and hazard ratios (HR) with 95% confidence intervals and log-rank p-values were calculated. For cross-validation, a meta-analysis was performed using GENT2, incorporating multiple HNSC gene expression datasets. Pooled HRs were estimated using a random-effects model, and forest plots were used to visualize the prognostic impact of cadherin genes.

miRNA–mRNA interaction and expression validation

Putative miRNAs targeting CDH1, CDH2, and CDH3 were identified using miRNet 2.0 (https://www.mirnet.ca) [31], and interaction networks were constructed. Expression levels of key miRNAs (hsa-miR-200a-3p and hsa-miR-200c-3p) were analyzed using UALCAN (https://ualcan.path.uab.edu/) based on TCGA-HNSC miRNA datasets [32]. Survival analysis was conducted using the Kaplan–Meier Plotter module of UALCAN. For experimental validation, total RNA including small RNAs, was extracted from cell lines. Expression of candidate miRNAs was quantified by RT-qPCR using TaqMan Advanced miRNA assays (Thermo Fisher Scientific). Relative expression was calculated using the 2^{^–ΔΔCt} method, with U6 snRNA as internal control [33].

Dual-luciferase reporter assay

To validate the direct interaction between hsa-miR-200a-3p, hsa-miR-200c-3p, and cadherin genes, dual-luciferase reporter assays were performed in HNSC cell lines. Wild-type (WT) 3′ untranslated region (3′UTR) fragments of CDH1, CDH2, and CDH3 containing the predicted miRNA binding sites were cloned downstream of the firefly luciferase gene in a luciferase reporter vector. Corresponding mutant (MUT) constructs were generated by introducing point mutations in the miRNA seed-match regions. HNSC cells were co-transfected with WT or MUT reporter plasmids together with hsa-miR-200a-3p mimic, hsa-miR-200c-3p mimic, or negative control mimic using Lipofectamine™ 3000 (Thermo Fisher Scientific), according to the manufacturer’s instructions. A Renilla luciferase plasmid was included for normalization. After 24–48 h, luciferase activities were measured using a dual-luciferase reporter assay system (Promega), and firefly luciferase activity was normalized to Renilla luciferase activity. Relative luciferase activity was expressed as fold change compared with control mimic.

Immunogenomic and drug sensitivity analysis

To investigate immune associations, expression levels of CDH1, CDH2, and CDH3 were analyzed across five immune subtypes using the TISIDB database (http://cis.hku.hk/TISIDB/) [34]. Correlations with immune inhibitor and stimulator genes were visualized using heatmaps. Immune cell infiltration levels were assessed using the GSCA database [26], where Spearman correlation coefficients were computed between cadherin gene expression and infiltration of major immune cell types. For drug sensitivity profiling, gene–drug sensitivity correlations were explored using GSCA based on GDSC (Genomics of Drug Sensitivity in Cancer) datasets. Pearson correlation analysis was performed between cadherin expression and half-maximal inhibitory concentration (IC50) values across cancer cell lines to infer potential drug resistance patterns.

Protein–protein interaction (PPI) and functional enrichment analysis

PPI networks for CDH1, CDH2, and CDH3 were constructed using the GeneMANIA (https://genemania.org/) [35] and STRING (https://string-db.org/) [36] databases to identify key interacting partners. The common interactors across both networks were determined using a Venn diagram tool (http://bioinformatics.psb.ugent.be/webtools/Venn/) [37]. To investigate the biological functions and pathway involvement of the cadherin genes and their interactors, gene enrichment analysis was performed using the DAVID database (https://david.ncifcrf.gov/) [38]. Enrichment analyses covered Gene Ontology categories, including biological processes, molecular functions, and cellular components, along with KEGG pathway analysis [39, 40] to identify relevant signaling pathways.

CDH1 and CDH2 knockdown and functional validation assays

To investigate the functional relevance of CDH1 and CDH2 in HNSC, loss-of-function experiments were conducted using FaDu and SCC9 cell lines. For gene silencing, cells were transfected with 20 nM siRNA targeting CDH1 and CDH2 (Silencer Select, Thermo Fisher Scientific, USA) using Lipofectamine™ RNAiMAX transfection reagent (Thermo Fisher Scientific) according to the manufacturer’s protocol. A scrambled non-targeting siRNA was used as the negative control. After 48 h, RNA was extracted using the PureLink™ RNA Mini Kit, and protein was isolated using RIPA lysis buffer supplemented with protease inhibitors. The knockdown efficiency was validated using RT-qPCR and Western blotting [41, 42].

To assess the impact on cell proliferation, a Cell Counting Kit-8 (CCK-8; Dojindo, Japan) assay was performed at 24-, 48-, and 72-hours post-transfection. For colony formation assays, 500 cells were seeded into 6-well plates and cultured for 10–14 days. Colonies were fixed with methanol, stained with 0.5% crystal violet, and counted manually. Wound healing assays were performed by creating a scratch using a sterile pipette tip in confluent monolayers. Wound closure was photographed at 0 and 24 h and quantified using ImageJ software.

FaDu cells transfected with si-CDH1, si-CDH2, or control siRNA were harvested, washed, and resuspended in sterile PBS. Approximately 5 × 10⁶ cells in 100 µL PBS were subcutaneously injected into the flanks of 6-week-old female BALB/c nude mice (n = 5 per group). Tumor volume was measured every 3 days using digital calipers, and volumes were calculated using the formula: (length × width²)/2. Body weights were recorded throughout the study to monitor systemic toxicity. Tumor growth was monitored for approximately 35–40 days until tumors reached the experimental endpoint criteria. Mice were then anesthetized with isoflurane (2–3% for induction, 1–2% for maintenance) delivered via a precision vaporizer, and euthanasia was performed via cervical dislocation following deep anesthesia, in accordance with AVMA Guidelines for the Euthanasia of Animals [43]. Tumors were excised, photographed, and weighed. All procedures involving animals were approved by the Institutional Animal Care and Use Committee (IACUC) of Cholistan University of Animal and veterinary Sciences and conducted in accordance with institutional and national ethical guidelines. The experimental protocol and reporting follow the ARRIVE 2.0 guidelines [44] to ensure transparency, reproducibility, and high standards of animal welfare in biomedical research.

Statistical analysis

All experiments were performed in at least three independent biological replicates, and data are presented as mean ± standard deviation (SD) unless otherwise indicated. Comparisons between two groups (e.g., siRNA knockdown vs. control) were conducted using an unpaired two-tailed Student’s t-test. For comparisons involving multiple groups, one-way ANOVA followed by Tukey’s post hoc test was used. The Shapiro-Wilk test was applied to assess the normality of data distribution prior to parametric testing. For RT-qPCR data, relative expression levels were calculated using the 2^{^–ΔΔCt} method, and significance was assessed based on log2-transformed fold-change values. ROC curves and area under the curve (AUC) values were computed using the pROC package in R to evaluate diagnostic accuracy. Kaplan–Meier survival analysis was performed using the log-rank (Mantel–Cox) test, and hazard ratios (HRs) with 95% confidence intervals (CIs) were derived from Cox proportional hazards models. Meta-analysis statistics were generated using a random-effects model, and heterogeneity was quantified with I² statistics.

Results

Expression analysis of cadherin genes via RT-qPCR

The expression patterns of the cadherin gene family (CDH1, CDH2, and CDH3) were analyzed in six HNSC cell lines and 3 control cell lines using RT-qPCR. As shown in Fig. 1A, all three cadherin genes exhibited significantly elevated expression in HNSC cell lines compared to normal controls. To assess the diagnostic potential of these genes, ROC curve analyses were performed (Fig. 1B). The area under the curve (AUC) values were 0.80 for CDH1, 0.75 for CDH2, and 0.90 for CDH3, suggesting that particularly CDH3 possesses strong discriminatory power to distinguish between cancerous and non-cancerous samples.

Fig. 1.

Expression analysis and diagnostic potential of cadherin genes in HNSC cell lines. A RT-qPCR analysis showing significantly elevated expression of CDH1, CDH2, and CDH3 in six HNSC cell lines compared to three control cell lines. B Receiver operating characteristic (ROC) curve analysis illustrating the diagnostic accuracy of cadherin genes. P*-value < 0.05

Validation of cadherin gene expression in HNSC using TCGA data

The expression patterns of CDH1, CDH2, and CDH3 were further validated in HNSC patients using TCGA transcriptomic data via the GSCA platform. All three cadherin genes were significantly upregulated in tumor tissues compared to normal samples, indicating their consistent dysregulation in HNSC (Fig. 2A). Notably, stratification of HNSC patients based on molecular subtypes revealed that CDH1, CDH2, and CDH3 mRNA levels were significantly higher in the HPV-positive subgroup than in the HPV-negative subgroup, suggesting a potential association between HPV status and cadherin gene expression (Fig. 2B). Analysis across pathological stages demonstrated persistently high expression levels of CDH1 and CDH3 from stage I through IV, with a slight decreasing trend, while CDH2 displayed a less consistent pattern across stages (Fig. 2C-D). Furthermore, GSEA showed positive enrichment scores for all three genes in HNSC, suggesting their functional involvement in HNSC carcinogenesis (Fig. 2E).

Fig. 2.

Validation of cadherin gene expression and clinical correlations in HNSC using TCGA data. A Transcriptomic expression profiles of CDH1, CDH2, and CDH3 in HNSC tumor tissues versus normal tissues from the TCGA dataset, analyzed via the GSCA database. B Subtype-specific expression analysis of cadherin genes. C–D Expression levels of CDH1, CDH2, and CDH3 across different pathological stages (I–IV) of HNSC. E Gene Set Enrichment Analysis (GSEA) analysis of cadherin genes. P-value < 0.05

Aberrant methylation and oncogenic pathway regulation of cadherin genes in HNSC

To explore the potential epigenetic involvement of cadherin genes in HNSC, promoter methylation profiles of CDH1, CDH2, and CDH3 were analyzed using TCGA-HNSC data accessed through the OncoDB database. As shown in Fig. 3A, beta-value distributions across promoter-associated CpG sites revealed altered methylation patterns between tumor and normal tissues for all three genes, although the magnitude and direction of change varied across individual probes. To further examine the relationship between DNA methylation and gene expression, Spearman correlation analysis was performed using the GSCA platform. No significant correlation was observed between CDH1 promoter methylation and mRNA expression (cor = − 0.01, FDR = 0.87), suggesting that promoter methylation is unlikely to be a major determinant of CDH1 expression levels in HNSC (Fig. 3B). In contrast, CDH2 and CDH3 exhibited moderate but statistically significant inverse correlations between promoter methylation and transcript abundance (CDH2: cor = − 0.33, FDR = 7.7 × 10⁻¹⁵; CDH3: cor = − 0.27, FDR = 4.8 × 10⁻¹⁰), indicating that higher methylation levels are associated with lower expression of these genes in HNSC (Fig. 3B). Collectively, these findings suggest that while promoter methylation may be associated with transcriptional variability of CDH2 and CDH3 in HNSC, CDH1 expression appears to be regulated predominantly by methylation-independent mechanisms. Given that these observations are derived from in silico correlation analyses, further experimental validation is required to establish direct causal relationships between DNA methylation and cadherin gene regulation. Furthermore, pathway activity analysis using the GSCA database revealed distinct roles of cadherin genes in oncogenic signaling. CDH1 and CDH3 predominantly inhibited EMT, DNA damage response, hormone receptor, and RTK pathways, while CDH2 was strongly associated with activation of the EMT pathway, aligning with its known role in tumor progression (Fig. 3C).

Fig. 3.

Methylation status and oncogenic pathway involvement of cadherin genes in HNSC. A Promoter methylation analysis of CDH1, CDH2, and CDH3 in HNSC tissues using TCGA data via the OncoDB platform. B Spearman correlation analysis between promoter methylation and gene expression levels. C Oncogenic pathway activity analysis of cadherin genes using GSCA. P-value < 0.05

Mutational and CNV landscape of cadherin genes in HNSC

To characterize genomic alterations in cadherin genes in HNSC, we analyzed somatic mutations and CNVs of CDH1, CDH2, and CDH3 using the cBioPortal database. As shown in Fig. 4A, CDH3 was most frequently mutated (44%), followed by CDH2 (33%) and CDH1 (22%) (Fig. 4A) across the HNSC samples. knockdown and functional validation larly in CDH3 and CDH2, while CDH1 also harbored in-frame deletions and a smaller proportion of nonsense mutations (Fig. 4A). SNV classification revealed that C > T transitions were the most common single-nucleotide variant (SNV) class (Fig. 4A). Somatic mutation rate analysis further confirmed that CDH3 had the highest mutation frequency (1.57%), followed by CDH2 (1.18%) and CDH1 (0.79%), with mutational hotspots primarily located within cadherin repeat domains and calcium-binding regions (Fig. 4B). CNV profiling revealed a diverse pattern of genomic alterations, with both heterozygous deletions and amplifications detected across all three genes (Fig. 4C).

Fig. 4.

Mutational spectrum and copy number variations of cadherin genes in HNSC. A Genomic alteration frequencies of CDH1, CDH2, and CDH3 across 18 HNSC samples from the cBioPortal database. B Somatic mutation rates and distribution of mutations in functional domains, with CDH3 showing the highest mutation frequency (1.57%), followed by CDH2 (1.18%) and CDH1 (0.79%). C Copy number variation (CNV) profiles showing both amplifications and heterozygous deletions across all three cadherin genes in HNSC

Prognostic significance of cadherin gene expression in HNSC

To evaluate the prognostic relevance of cadherin genes in HNSC, survival analyses were conducted using KM Plotter and GENT2 databases. As shown in Fig. 5A, Kaplan–Meier survival curves revealed that high expression of CDH3 was significantly associated with worse overall survival (HR = 1.39, P = 0.015), followed by CDH2 (HR = 1.26, P = 0.01) and CDH1 (HR = 1.18, P = 0.028), suggesting that elevated expression of these cadherin family members may correlate with poor patient outcomes (Fig. 5A). To further validate these findings, a meta-analysis was performed using GENT2, which incorporated multiple gene expression datasets. The pooled hazard ratio from the random-effects model demonstrated a significant association of high CDH3 expression with poor prognosis (HR = 1.43, 95% CI: 0.52–3.92), and a similar trend was observed for CDH2 (HR = 1.24, 95% CI: 0.7–2.20), although with broader confidence intervals, indicating variability among studies (Fig. 5B). For CDH1, the meta-analysis showed an HR of 1.29 (95% CI: 0.31–5.31), reflecting a consistent, though not statistically robust, trend toward unfavorable prognosis (Fig. 5B). Notably, while elevated cadherin expression was consistently associated with adverse survival trends, the observed hazard ratios were modest and confidence intervals in the meta-analysis were relatively wide, indicating inter-study variability and underscoring the need for cautious interpretation of prognostic strength.

Fig. 5.

Prognostic significance of cadherin gene expression in HNSC. A Kaplan–Meier survival curves showing overall survival in HNSC patients stratified by expression levels of CDH1, CDH2, and CDH3 using the KM Plotter database. B Forest plots from a meta-analysis using the GENT2 database. P-value < 0.05

Identification and validation of miRNAs targeting cadherin family genes in HNSC

To investigate potential post-transcriptional regulatory mechanisms involving cadherin genes in HNSC, a miRNA–mRNA interaction network was constructed using the miRNet database. Network analysis identified hsa-miR-200a-3p and hsa-miR-200c-3p as common predicted regulators of CDH1, CDH2, and CDH3, suggesting that these miRNAs may be broadly associated with cadherin-related regulatory networks in HNSC (Fig. 6A). Expression analysis based on TCGA-HNSC datasets using the UALCAN platform revealed that both hsa-miR-200a-3p and hsa-miR-200c-3p were significantly upregulated in primary tumor tissues compared with normal samples (Fig. 6B), indicating a tumor-associated expression pattern. However, Kaplan–Meier survival analysis demonstrated no significant association with overall survival for either miRNA (p = 0.4 for hsa-miR-200c-3p and p = 0.45 for hsa-miR-200a-3p), suggesting that their expression levels alone may not serve as independent prognostic indicators in HNSC (Fig. 6C). To experimentally validate these observations, RT-qPCR analysis was performed in HNSC cell lines and non-malignant epithelial control cells. Consistent with TCGA data, both hsa-miR-200c-3p and hsa-miR-200a-3p exhibited significantly higher expression in HNSC cell lines compared with controls (p = 6.6 × 10⁻⁷ and p = 1.2 × 10⁻⁸, respectively; Fig. 6D). Furthermore, receiver operating characteristic (ROC) curve analysis demonstrated a strong discriminatory ability between tumor and normal samples, with an AUC of 1.0 for hsa-miR-200c-3p and 0.96 for hsa-miR-200a-3p (Fig. 6E), indicating their potential utility as diagnostic biomarkers for HNSC. These findings are exploratory and warrant further validation in independent clinical cohorts. Finally, dual-luciferase reporter assays showed that transfection of hsa-miR-200a-3p or hsa-miR-200c-3p mimics significantly reduced normalized luciferase activity from WT CDH1/CDH2/CDH3 3′UTR reporters compared with control mimic, whereas mutation of the predicted seed-match sites substantially attenuated these effects (Fig. 6F).

Fig. 6.

Identification and validation of miRNAs targeting cadherin genes in HNSC. A Predicted miRNA–mRNA interaction network generated using the miRNet database showing hsa-miR-200a-3p and hsa-miR-200c-3p as common putative regulators of CDH1, CDH2, and CDH3, suggesting shared post-transcriptional regulatory associations. B Differential expression analysis of hsa-miR-200a-3p and hsa-miR-200c-3p between HNSC tumor and normal tissues based on TCGA-HNSC data obtained from the UALCAN platform. C Kaplan–Meier survival analysis of HNSC patients stratified by high and low miRNA expression levels, demonstrating no significant association with overall survival. D RT-qPCR validation of hsa-miR-200a-3p and hsa-miR-200c-3p expression in six HNSC cell lines and three non-malignant epithelial control cell lines, confirming significantly higher expression in cancer cells. E Receiver operating characteristic (ROC) curve analysis evaluating the diagnostic performance of hsa-miR-200a-3p and hsa-miR-200c-3p in distinguishing HNSC samples from normal controls. F Dual-luciferase reporter assays showing that overexpression of hsa-miR-200a-3p and hsa-miR-200c-3p resulted in a modest but consistent reduction in normalized luciferase activity from wild-type CDH1, CDH2, and CDH3 3′UTR reporters across HNSC cell lines, whereas mutation of the predicted miRNA binding sites attenuated these effects, indicating sequence-specific interactions. P-value < 0.05

Immunogenomic and drug sensitivity analysis of cadherin family genes in HNSC

To explore potential immunological associations of cadherin genes in HNSC, expression patterns of CDH1, CDH2, and CDH3 were examined across five immune subtypes using the TISIDB database. Subtype-specific expression differences were observed, with CDH1 and CDH3 exhibiting relatively higher expression in the C1 (wound healing) and C2 (IFN-γ dominant) subtypes, whereas CDH2 displayed more variable expression across immune subtypes (Fig. 7A). Correlation analysis further revealed positive associations between cadherin gene expression and several immune inhibitory molecules, including correlations between CDH3 and TGFB1, LAG3, and CSF1R, as well as between CDH1 and TGFB1 and KDR (Fig. 7B). In contrast, negative correlations were observed between cadherin expression and selected immune stimulatory genes, such as ULBP1, TNFSF4, and CXCR4 for CDH3, and TNFSF4 and CD40 for CDH1 (Fig. 7C). Immune cell infiltration analysis using the GSCA platform demonstrated inverse correlations between CDH1 expression and infiltration levels of multiple immune cell populations, including macrophages, natural killer (NK) cells, and Th1 cells, while CDH3 expression showed negative associations with B cell and Th2 cell infiltration (Fig. 7D). These findings suggest that cadherin gene expression may be associated with distinct immune microenvironmental features in HNSC, although no direct immunomodulatory effects can be inferred from these analyses. Additionally, drug sensitivity analysis based on GSCA–GDSC datasets revealed positive correlations between cadherin gene expression and higher half-maximal inhibitory concentration (IC50) values for several anticancer agents. Specifically, CDH1 expression correlated with reduced sensitivity to cisplatin, docetaxel, and afatinib, while CDH3 expression showed similar associations with lapatinib, gefitinib, and erlotinib (Fig. 7E). These correlations indicate that elevated cadherin expression may be associated with altered drug response profiles in cancer cell lines. However, as these observations are derived from in silico correlation analyses, they should be interpreted as hypothesis-generating, and further experimental validation is required to establish any causal role of cadherin genes in immune modulation or therapeutic resistance.

Fig. 7.

Immunogenomic and pharmacogenomic associations of cadherin genes in HNSC. A Expression of CDH1, CDH2, and CDH3 across five immune subtypes of HNSC (C1: wound healing; C2: IFN-γ dominant; C3: inflammatory; C4: lymphocyte depleted; C6: TGF-β dominant), based on TISIDB database analysis. B Correlation heatmap between cadherin genes and immune inhibitory markers. C Correlation heatmap showing associations between cadherin genes and immune stimulators. D Immune cell infiltration analysis via GSCA database. E Drug sensitivity analysis of cadherin genes. P-value < 0.05

PPI networks and functional enrichment analysis of cadherin family genes

In this section of the study, the interaction networks and functional roles of cadherin family genes in HNSC were analyzed. Firstly, the PPI networks of cadherin genes (CDH1, CDH2, and CDH3) were constructed using the GeneMANIA and STRING databases (Fig. 8A–B), and common interacting genes were identified through a Venn diagram analysis (Fig. 8C). The PPI networks and shared interacting partners revealed that cadherin proteins interact with other cadherins and members of the catenin family, which are key regulators of cell adhesion and intracellular signaling. This finding is consistent with previous studies highlighting the importance of cadherin-cadherin and cadherin-catenin interactions in cancer progression, particularly in the regulation of EMT and metastasis [45, 46]. Importantly, the identification of common genes between the two networks was critical to understanding which signaling pathways may be disrupted due to cadherin gene dysregulation. Subsequently, we performed gene enrichment analysis using the DAVID tool to explore the biological processes, cellular components, molecular functions, and pathways associated with these genes (Fig. 8D-F). Gene enrichment analysis (Figs. 8D-G) provided further insights into the biological significance of these cadherin genes. Cellular component enrichment (Fig. 8D) highlighted their association with structures such as the “zonula adherens, catenin complex, plasma membrane, and adherens junctions, emphasizing their role in maintaining cell-cell adhesion and integrity.” Molecular function enrichment (Fig. 8E) revealed interactions with proteins involved in “gamma-catenin binding, beta-catenin binding, cadherin binding, and cytoskeletal protein binding, indicating their involvement in intracellular signaling and cytoskeletal organization.” Biological process enrichment (Fig. 8F) identified key roles for these genes in “cell junction assembly, cell adhesion, and the regulation of the canonical Wnt signaling pathway,” which is known to contribute to cancer progression and metastasis. Additionally, pathway enrichment analysis (Fig. 8G) demonstrated significant associations with the “adherens junction pathway, leukocyte transendothelial migration, the Hippo signaling pathway, and pathways related to gastric and endometrial cancer,” further supporting the role of cadherin genes in tumorigenesis and metastasis. However, as a next step, future mechanistic studies are essential to investigate how these cadherins modulate the aforementioned pathways at the molecular and cellular levels.

Fig. 8.

PPI networks and enrichment analysis of cadherin genes in HNSC. A–B Protein–protein interaction (PPI) networks of cadherin genes (CDH1, CDH2, and CDH3) constructed using GeneMANIA and STRING databases. C Venn diagram identifying common interacting genes shared across both networks. D Gene ontology (GO) cellular component enrichment analysis. E Molecular function enrichment analysis. F Biological process enrichment analysis. G KEGG pathway analysis (data was obtained from the KEGG database: www.kegg.jp/kegg/kegg1.html). P-value < 0.05

Silencing of CDH1 and CDH2 impairs proliferation, colony formation, migration, and invasion in HNSC in vitro and in vivo models

To elucidate the functional roles of CDH1 and CDH2 in HNSC, we performed loss-of-function studies in two distinct HNSC cell lines, FaDu and SCC9, followed by comprehensive phenotypic characterization. RT-qPCR and Western blot analyses confirmed effective silencing of CDH1 and CDH2 at both transcript and protein levels in si-CDH1 and si-CDH2 groups compared to controls (Figs. 9A–B and 10A–B, and Supplementary data Fig. 1). Functional assays revealed that knockdown of either gene significantly reduced cellular proliferation in both FaDu (Fig. 9C) and SCC9 cells (Fig. 10C). Consistently, colony formation assays demonstrated a marked decrease in the number of colonies upon CDH1 or CDH2 silencing in both cell lines (Figs. 9D–E and 10D–E), indicating a critical role of these cadherins in sustaining clonogenic potential. Furthermore, wound healing assays showed impaired migratory capacity in si-CDH1 and si-CDH2 groups after 24 h in both FaDu and SCC9 cells (Figs. 9F–G and 10F–G). In vivo xenograft assays using FaDu cells demonstrated that tumors in the si-CDH1-FaDu and si-CDH2-FaDu groups grew significantly more slowly than control tumors, as shown by representative tumor images and reduced tumor volumes over time (Fig. 10H–I). Body weight was monitored throughout the study (Fig. 10J), and although differences among groups were observed, all animals remained within acceptable physiological ranges with no signs of overt distress. As body weight changes may reflect multiple factors beyond tumor burden, further experiments are needed to confirm these outcomes. Finally, a schematic model summarises the proposed mechanism, where disruption of cadherin-mediated cell adhesion and architecture upon CDH1/2 knockdown may inhibit oncogenic signaling, ultimately attenuating proliferation, survival, and migration (Fig. 10K).

Fig. 9.

Functional characterization of CDH1 and CDH2 knockdown in FaDu cells. A–B RT-qPCR and Western blot analysis confirming successful knockdown of CDH1 and CDH2 in FaDu cells. C Cell proliferation assays demonstrating significantly reduced growth in si-CDH1 and si-CDH2 groups. D–E Colony formation assays showing decreased clonogenic potential following knockdown. F–G Wound healing assays indicating impaired migration capacity in CDH1- and CDH2-silenced FaDu cells. P***-value < 0.001

Fig. 10.

Functional validation of CDH1 and CDH2 knockdown in SCC9 cells in vitro, and FaDu-derived in vivo xenograft models. A–B Validation of CDH1 and CDH2 knockdown in SCC9 cells via RT-qPCR and Western blotting. C Proliferation assays in SCC9 cells indicating suppressed cell growth upon knockdown. D–E Colony formation assays showing significant reduction in colony numbers in si-CDH1 and si-CDH2 SCC9 cells. F–G Wound healing assays displaying impaired migratory ability post-knock down. H–I In vivo tumor growth assays using FaDu xenografts: reduced tumor volumes and sizes in CDH1- and CDH2-silenced groups. J Body weight measurements over time showing physiological impact of gene silencing. K Graphical model summarizing the proposed mechanism by which CDH1 and CDH2 promote tumor proliferation and migration in HNSC. P***-value < 0.001

Discussion

This study provides an integrated analysis of CDH1, CDH2, and CDH3 in HNSC, highlighting their diagnostic associations, prognostic correlations, and functional relevance in experimental systems [47, 48]. Importantly, these dimensions represent distinct levels of translational evidence. While CDH3 demonstrates strong diagnostic discrimination and cadherin expression correlates with patient outcomes, the prognostic effect sizes are modest, and causal therapeutic benefit cannot be inferred from the present data.

Our expression profiling demonstrated significantly elevated levels of CDH1, CDH2, and CDH3 in HNSC cell lines and TCGA tumor tissues, consistent with previous reports in other epithelial cancers, including gastric, breast, and prostate carcinomas, where aberrant cadherin expression promotes tumor progression by disrupting normal tissue architecture [49, 50]. Notably, CDH3 showed the highest diagnostic accuracy with an AUC of 0.90, supporting its potential as a robust biomarker. While CDH1 and CDH2 have been widely studied, our data reinforce recent observations from Liu et al. suggesting CDH3 may play a more dominant role in squamous tumors [51]. Interestingly, the elevated expression of all three genes in HPV-positive HNSC patients suggests a possible link between viral oncogenesis and cadherin gene regulation [52]. The situation also raises important mechanistic questions regarding viral oncogene–mediated regulation of cadherin genes. High-risk HPV infection, particularly through the viral oncoproteins E6 and E7, is known to alter host transcriptional programs by targeting tumor suppressors such as p53 and RB and by modulating multiple signaling pathways involved in epithelial differentiation and cell adhesion [53, 54]. It is therefore plausible that HPV-driven transcriptional reprogramming may influence cadherin gene expression in HPV-positive tumors. However, the present study relies on transcriptomic correlations and does not directly address whether HPV oncogenes regulate cadherin transcription. Future studies using HPV E6/E7 overexpression or knockdown models, together with promoter activity assays and chromatin-level analyses, would be required to determine whether cadherin genes are direct or indirect targets of HPV-mediated transcriptional control in HNSC.

Moreover, while bulk transcriptomic analyses robustly validate the overexpression of cadherin genes in HNSC, these approaches cannot resolve intratumoral heterogeneity at the cellular level. Future studies incorporating single-cell RNA sequencing (scRNA-seq) could help determine whether CDH1, CDH2, and CDH3 dysregulation occurs uniformly across tumor cells or is enriched within specific cellular subpopulations, such as EMT-progressing or stem-like tumor subsets. Such analyses may further refine the clinical utility of CDH3, which demonstrated strong diagnostic performance (AUC = 0.90), by identifying high-risk cellular subclones associated with aggressive tumor behavior and therapeutic resistance.

An important conceptual consideration in this study is the observation that CDH1 upregulation in HNSC tissues and cell lines is associated with unfavorable clinical outcomes, which contrasts with the classical view of CDH1 loss as a hallmark of epithelial–mesenchymal transition (EMT) and tumor invasion. Increasing evidence indicates that the role of CDH1 in cancer is highly context dependent and may vary according to tumor type, differentiation state, and microenvironmental conditions. In HNSC, which frequently retains epithelial characteristics even at advanced stages, elevated CDH1 expression may reflect aberrant or compensatory adhesion programs rather than intact tumor-suppressive function. Importantly, increased CDH1 transcript levels do not necessarily correspond to functional membranous E-cadherin, as mislocalization to the cytoplasm, disruption of adherens junction integrity, or impaired interaction with catenin proteins may render E-cadherin functionally ineffective despite high expression. Tumor heterogeneity may further contribute to this phenomenon, with subpopulations of cells exhibiting partial EMT states or co-expression of epithelial and mesenchymal markers. A limitation of the present study is the lack of protein-level and subcellular localization analyses. Therefore, future studies incorporating immunohistochemical evaluation of membranous versus cytoplasmic E-cadherin localization, together with biochemical assays examining E-cadherin interactions with β-catenin and associated junctional complexes, will be essential to clarify whether CDH1 overexpression in HNSC reflects functional adhesion, compensatory signaling, or dysfunctional cadherin complexes contributing to tumor progression.

Epigenetically, we observed lower promoter methylation in CDH1, CDH2, and CDH3 in HNSC samples, with inverse correlations to mRNA levels particularly evident for CDH2 and CDH3. This is in agreement with earlier studies showing that cadherin gene expression is frequently regulated via promoter demethylation in carcinomas [55, 56]. Our data support the notion that hypomethylation-driven activation of CDH2 and CDH3 may contribute to their oncogenic overexpression. Furthermore, pathway analysis revealed CDH1 and CDH3 as inhibitors of EMT and DNA repair pathways, while CDH2 activation was associated with enhanced EMT, corroborating the dualistic and sometimes opposing roles cadherins play depending on cellular context [57].

Our mutational profiling revealed CDH3 as the most frequently altered cadherin gene, with prominent hotspot mutations in cadherin repeat and calcium-binding domains. This is aligned with studies in breast and colorectal cancers, where CDH3 mutations were shown to compromise adhesion and increase cellular invasiveness [51, 58]. Copy number variations were also frequent, suggesting that structural alterations contribute to cadherin gene dysregulation. These findings offer new insights into how both genetic and epigenetic mechanisms converge to modulate cadherin gene expression in HNSC [59]. In survival analyses, elevated expression of CDH3, followed by CDH2 and CDH1, was significantly associated with poor overall survival, consistent with meta-analyses from other epithelial malignancies [60, 61]. These results extend the prognostic relevance of cadherins to HNSC and suggest that their overexpression may reflect aggressive tumor biology [62, 63].

At the post-transcriptional level, miRNA analysis identified hsa-miR-200a-3p and hsa-miR-200c-3p as potential upstream regulators targeting all three cadherins. Upregulation of these miRNAs in HNSC aligns with studies suggesting that members of the miR-200 family modulate EMT by targeting ZEB1/2 and E-cadherin [64, 65]. Despite the concurrent elevation of miR-200a-3p/miR-200c-3p and cadherin transcripts, dual-luciferase reporter assays confirmed a direct and sequence-specific interaction between these miRNAs and the 3′UTRs of CDH1, CDH2, and CDH3, as evidenced by reduced luciferase activity in wild-type constructs but not in seed-region mutants. These findings support a role for miR-200 family members in the post-transcriptional regulation of cadherin genes. Although the present study demonstrates direct binding of miR-200a-3p and miR-200c-3p to the 3′UTRs of CDH1, CDH2, and CDH3 through luciferase reporter assays, the regulatory dynamics underlying their concurrent upregulation in HNSC remain incompletely understood. Future investigations distinguishing precursor and mature forms of miR-200 family members could help clarify whether altered transcriptional activity or post-transcriptional processing contributes to their elevated levels in tumors. In addition, transcriptome-wide approaches such as RNA immunoprecipitation sequencing (RIP-seq) or AGO2-associated RNA profiling may help identify competing endogenous RNA (ceRNA) networks that modulate miR-200 availability and thereby influence cadherin expression. Such studies would provide a more comprehensive understanding of the post-transcriptional regulatory landscape governing cadherin signaling in HNSC.

The apparent coexistence of elevated miRNA and target gene expression suggests that miR-200-mediated regulation of cadherins in HNSC is likely modulatory rather than fully suppressive. Several non-mutually exclusive mechanisms may explain this phenomenon. First, miRNA-mediated repression is often partial and context-dependent, allowing fine-tuning of protein output rather than complete silencing, particularly in highly expressed genes such as cadherins. Second, transcriptional upregulation, copy number gains, or alternative regulatory inputs may outweigh post-transcriptional repression, resulting in net increases in cadherin mRNA levels despite active miRNA binding. Third, competitive endogenous RNA (ceRNA) effects, RNA-binding proteins, or altered miRNA loading efficiency in cancer cells may further attenuate miRNA-induced repression. Together, these mechanisms support a model in which miR-200a-3p and miR-200c-3p act as homeostatic regulators, constraining excessive cadherin expression rather than functioning as dominant on/off switches. Importantly, survival analyses revealed no significant association between miR-200a-3p or miR-200c-3p expression and overall survival, suggesting that while these miRNAs participate in early epithelial–mesenchymal regulatory circuits and tumor initiation, they may not independently determine long-term clinical outcomes. This observation is consistent with prior studies indicating that miR-200 family members primarily influence phenotypic plasticity, epithelial stability, and context-specific signaling, rather than serving as robust prognostic biomarkers. Collectively, these findings indicate that miR-200a-3p and miR-200c-3p contribute to cadherin-associated regulatory networks in HNSC through direct but partial post-transcriptional repression, highlighting a nuanced regulatory axis that warrants further investigation at the protein and functional signaling levels.

Immunogenomic analysis revealed context-dependent associations between cadherin genes and immune modulators. CDH1 and CDH3 correlated positively with immune suppressive genes such as TGFB1 and LAG3, and inversely with immune stimulatory markers like TNFSF4, suggesting a potential role in shaping an immunosuppressive tumor microenvironment. These observations echo recent findings in gastric and pancreatic cancers where cadherins influence immune cell recruitment and function [66, 67]. Additionally, cadherin expression correlated with resistance to chemotherapy drugs including cisplatin and gefitinib, implying a possible role in therapy evasion, which complements studies linking EMT with drug resistance mechanisms [68, 69].

Mechanistically, PPI and enrichment analyses demonstrated that cadherin genes interact extensively with the catenin family and Wnt pathway components. Our enrichment for adherens junction, Hippo, and Wnt signaling pathways aligns well with previous reports establishing cadherins as critical regulators of contact inhibition and signal transduction [70]. These findings strengthen the biological plausibility that cadherin dysregulation directly contributes to oncogenesis and metastasis.

Finally, functional validation through siRNA-mediated silencing of CDH1 and CDH2 in SCC9 and FaDu cell lines significantly impaired proliferation, colony formation, and wound healing. The observed phenotypes were corroborated by in vivo xenograft assays, where tumor growth was substantially reduced in CDH1/CDH2-silenced groups. These results align with previous functional studies where E- and N-cadherin loss suppressed tumorigenicity and EMT in squamous cell and colorectal carcinomas [71–73]. Importantly, our schematic model proposes a mechanistic basis where cadherin knockdown disrupts adhesion-mediated signaling and Wnt activation, ultimately impairing cancer cell proliferation and survival.

From a translational perspective, several limitations should be considered. Although CDH3 showed strong diagnostic performance, the prognostic hazard ratios for cadherin genes were moderate, and meta-analysis confidence intervals were wide, suggesting variability across cohorts. Moreover, while siRNA-mediated knockdown experiments demonstrate functional importance of CDH1 and CDH2 in tumor cell behavior, no pharmacologic or antibody-based cadherin-targeting strategies were evaluated in this study. As cadherins are widely expressed in normal epithelial tissues, their direct therapeutic targeting may present challenges related to specificity and toxicity. Therefore, at present, cadherin genes should be regarded primarily as biomarker-associated and biologically relevant molecules, rather than validated therapeutic targets. Future studies incorporating clinically relevant targeting approaches, combination therapies, and patient-derived models will be required to establish therapeutic tractability.

Conclusion

In conclusion, this study delineates the diagnostic associations, prognostic correlations, and functional relevance of cadherin family genes in HNSC. CDH3 emerges as a promising diagnostic biomarker, while CDH1 and CDH2 influence tumor cell proliferation and migration in experimental models. However, the prognostic impact of cadherins is modest, and their therapeutic tractability remains unproven. These findings support cadherins as biologically informative markers and contributors to tumor behavior, rather than established therapeutic targets. Continued mechanistic and translational studies are required before clinical application can be considered.

Authors’ contributions

Mostafa A. Abdel-Maksoud contributed to conceptualization, study design, data interpretation, and critical revision of the manuscript. Abdulaziz Alamri participated in data analysis, methodology development, and manuscript editing. Aljawharah Fahad contributed to project administration, resource provision, and manuscript review. Salman Alrokayan assisted in experimental design, data curation, and validation. Mohammed Aufy contributed to formal analysis, visualization, and statistical interpretation. Bushra Hafeez Kiani participated in literature review, data preprocessing, and drafting of the manuscript. Hanyong Zhang contributed to methodology optimization, experimental supervision, and data verification. Hamza Hameed performed bioinformatics analysis, laboratory-based experiments, contributed to result interpretation, and participated in writing the original draft. Yasir Hameed carried out laboratory experiments, functional assays, and contributed to manuscript writing and revision. Saeedah Almutairi provided supervision, secured funding, and approved the final version of the manuscript. All authors reviewed and approved the final manuscript and agree to be responsible for its content.

Funding

The authors extend their appreciation to the ongoing Research Funding Program (ORF-2026-470) King Saud University, Riyadh, Saudi Arabia.

Data availability

Any type of the data will be provided by the corresponding author.

Declarations

Ethics approval and consent to participate

All procedures involving animals were approved by the Institutional Animal Care and Use Committee (IACUC) of Cholistan University of Veterinary and Animal Sciens and conducted in accordance with institutional and national ethical guidelines. The experimental protocol and reporting follow the ARRIVE 2.0 guidelines [44] to ensure transparency, reproducibility, and high standards of animal welfare in biomedical research.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.