Introduction

Oral squamous cell carcinoma (OSCC) remains one of the most aggressive malignancies globally, with a stagnant 5-year survival rate of 55–60% [1], significantly lower than those of common cancers such as breast cancer (89%) and colorectal cancer (65%). According to global cancer statistics, 377,713 new OSCC cases were reported worldwide in 2023, accounting for 2.1% of all malignancies [2]. With the accelerating aging population, the disease burden is projected to exceed 500,000 cases by 2035. Current standard treatment protocols for OSCC still rely on surgical resection combined with radiotherapy and chemotherapy. However, even with reconstructive techniques like free tissue flap transplantation, postoperative patients frequently suffer from severe complications, including masticatory dysfunction (47–68% incidence), speech impairment (35–52%), and osteoradionecrosis (5–15%) [3]. Although targeted immunotherapies against EGFR and PD-1/PD-L1 pathways have shown preliminary efficacy, only 15–20% of patients benefit from these interventions [4]. Therefore, there is an urgent need to identify highly specific molecular markers, elucidate their underlying mechanisms, and explore targeted therapeutic strategies to advance precision medicine for OSCC.

V-ATPase is a multisubunit complex composed of a cytosolic catalytic domain (V1) and a transmembrane proton channel domain (V0) [5]. The V0 domain contains four a-subunit isoforms (a1–a4), which regulate the localization of the enzyme complex in distinct cellular membrane systems through differential targeting mechanisms [6]. The gene encoding the a4 isoform, ATP6V0A4, exhibits tissue-specific expression patterns, with high levels primarily detected in the kidney and epididymis. Its encoded product mediates the polarized localization of V-ATPase in the plasma membrane of renal α-intercalated cells, contributing to active H⁺ secretion in the renal tubule lumen [7]. Recent studies have revealed that this subunit not only regulates the acidification of acidic organelles such as lysosomes and endosomes [8] but also plays critical roles in fundamental biological processes, including vesicular transport, protein processing and degradation, and cotransport of small molecules, as well as physiological functions such as urinary acidification and bone metabolism regulation, by maintaining intracellular and extracellular pH homeostasis [9].

Emerging evidence highlights the dual functional properties of ATP6V0A4 in tumor progression. In breast cancer models, high expression of this subunit was observed in invasive MDA-MB-231 cells, and siRNA-mediated gene silencing significantly suppressed there in vitro invasive capacity [7]. Consistently, treatment with specific V-ATPase inhibitors (bafilomycin A1 / concanamycin A) also potently inhibited the invasive phenotype of these cells [10]. Mechanistic studies in highly metastatic HCCLM3 hepatocellular carcinoma cells demonstrated that downregulation of V-ATPase suppressed proton pump activity and extracellular acidification rate (ECAR), thereby inhibiting matrix metalloproteinase-2 (MMP-2) activation and gelatinolytic function, ultimately leading to dual suppression of tumor growth and metastasis [11]. Notably, downregulation of ATP6V0A4 has been observed in renal cell carcinoma, suggesting heterogeneous regulatory patterns of this subunit across different tumor types. These functional discrepancies may be attributed to multiple factors, including pH homeostasis regulation in the tumor microenvironment, activation of invasion-related signaling pathways, and reprogramming of cellular energy metabolism. However, the expression pattern, functional role, and molecular mechanism of ATP6V0A4 in OSCC remain poorly understood.

This study aims to clarify the role of ATP6V0A4 in OSCC and investigate its underlying mechanisms, thereby providing novel therapeutic targets for this disease.

Methods

Bioinformatics analysis

Data download

The corresponding clinicopathological characteristics and prognostic information of OSCC patients, including gender, age, and stage, were downloaded from the The Cancer Genome Atlas (TCGA) Genomic Data Commons (GDC) database (https://portal.gdc.cancer.gov). Samples with the primary sites of “lip”, “hard palate”, “gingiva”, “base of tongue”, “floor of mouth”, “other and ill - defined sites in lip, oral cavity and pharynx”, “other and ill - defined sites of tongue”, and “other and unspecified sites in oral cavity” were selected for OSCC analysis. Among them, there were 32 normal samples and 397 cancer samples. The OSCC sequencing data were downloaded from the Gene Expression Omnibus (GEO) database (https://www.ncbi.nlm.nih.gov/geo), series GSE25099. The platform of GSE25099 is GPL5175, which contains a total of 79 OSCC samples, including 22 normal samples and 57 cancer tissues [12, 13].

Differential expression analysis

The R package “limma” was used to perform differentially expressed gene (DEG) analysis on the OSCC data from TCGA and GSE25099.|log₂FC| ≥ 1.5 and P value < 0.05 were set as the thresholds for DEGs. Genes were considered as up regulated DEGs when log₂FC > 1.5 and P value < 0.05, while genes were considered as down - regulated DEGs when log₂FC < − 1.5 and P value < 0.05. Subsequently, a volcano plot was applied to visualize the relevant significant DEGs. The volcano plot was described using the R package “ggplot2”.

Selection of differentially expressed genes and construction of prediction models

The Venn Diagram R package [14] was used to create a Venn diagram to display the differentially expressed genes. Based on survival time, survival status, and gene expression data, a univariate Cox proportional hazards regression model was employed for prognostic analysis. The hazard ratio (HR) of each differentially expressed gene was calculated using this model. Protective genes and risk genes were defined as those with HR < 1 and HR > 1, respectively (p < 0.05). To avoid overfitting, the “glmnet” package was used to conduct least absolute shrinkage and selection operator (LASSO) regression analysis to identify characteristic genes [15]. Finally, the survival random forest (RF) algorithm was applied to determine the final core genes.

Survival analysis

The “surv_cutpoint” function in the R package “survminer” was used to determine the optimal cut - off value for survival analysis of the target genes [16]. The Kaplan - Meier (KM) product - limit method was used to compare the probabilities of progression - free survival (PFS) among different subgroups, and the log - rank test was used to determine the P - value.

Pan - cancer analysis, the human protein atlas, and enrichment analysis

The expression differences of the ATP6V0A4 protein across pan - cancers were visualized through the TIMER database (http://timer.cistrome.org/) [17]. The Human Protein Atlas (HPA) database was established in Sweden in 2003 and is dedicated to using various omics technologies, including antibody - based imaging, mass spectrometry - based proteomics, systems biology, and transcriptomics. In this study, we utilized the HPA database to validate the intracellular localization of the ATP6V0A4 protein and compared its protein expression in normal and tumor tissues. To verify the potential functions of ATP6V0A4, gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analyses were performed using the “org.Hs.eg.db”, “ggplot2”, “clusterProfiler”, and “rich - plot” packages in R software to uncover the biological processes and signaling pathways associated with the ATP6V0A4 gene. The results of gene functional enrichment analysis were presented through bar charts and bubble plots.

Clinical correlation analysis

To explore the correlation between the ATP6V0A4 gene and clinical characteristics, a group analysis was conducted on the TCGA dataset, including age, gender, T stage, M stage, N stage, and clinical stage.

Immune infiltration analysis and immunotherapy response

To investigate the association between the expression of the ATP6V0A4 gene and the immune microenvironment, the CIBERSORT algorithm [18] was first used to quantitatively evaluate the relative abundances of 22 types of immune cell types. The high and low expression groups of ATP6V0A4 were determined based on the median value of ATP6V0A4 expression levels in the analyzed patient cohort. The Spearman correlation test was employed to analyze the correlation between the expression level of ATP6V0A4 and the infiltration proportion of each immune cell, and the corresponding correlation coefficients were calculated. In addition, the immune phenotype score (IPS) was calculated based on the tumor immune dysfunction and exclusion model to predict the response potential of different immune subsets to immunotherapy.

Drug sensitivity analysis

The drug sensitivity data were obtained from the Genomics of Drug Sensitivity in Cancer database (GDSC, https://www.cancerrxgene.org/) [19]. The R language oncoPredict package was used to predict the half - maximal inhibitory concentration of each drug. The high and low expression groups of ATP6V0A4 were determined based on the median value of ATP6V0A4 expression levels in the analyzed patient cohort. Subsequently, the correlation between drug sensitivity and the expression level of the target gene was analyzed, and the differences in drug sensitivity among different patient subgroups were calculated.

In vitro cell experiments

Cell line acquisition and culture

The human tongue squamous cell carcinoma cell line CAL27 (ATCC® CRL − 2095™) was purchased from the American Type Culture Collection (ATCC), and the HN6, SAS, HSC3 cell lines and the normal oral epithelial keratinocyte cell line HOK were purchased from AULU Company in Guangdong, China. All cells were cultured in Dulbecco’s Modified Eagle Medium (DMEM, Gibco, Waltham, MA, USA). The complete medium contained 10% fetal bovine serum (FBS, Gibco) and 1% penicillin/streptomycin (Gibco). The serum - free medium used for subsequent experiments (such as scratch assay, cell migration and invasion assays) was DMEM containing 1% penicillin/streptomycin. Cells were cultured in a constant - temperature and constant - humidity incubator at 37 °C with 5% CO₂.

Plasmid transfection

The empty vector control (NC) and ATP6V0A4 overexpression plasmids (OE - NC and OE - ATP6V0A4) were purchased from Shanghai Kunkun Biotechnology Co., Ltd. When the confluence of oral squamous cell carcinoma cell lines (OSCC) reached 60%, 50 nM of the plasmid was transfected using Lipofectamine 3000 (Invitrogen) according to the manufacturer’s instructions. The overexpression efficiency of ATP6V0A4 was verified by real - time fluorescent quantitative polymerase chain reaction (RT - qPCR) 48 h after transfection.

Real - time fluorescent quantitative reverse transcription polymerase chain reaction

Total cellular RNA was extracted using an RNA extraction kit (19221ES50, Shanghai Yeasen Biotechnology). The RNA quality was evaluated by detecting the A260/A280 ratio using a spectrophotometer (NanoDrop 2000, Thermo Fisher Scientific). 2 µg of total RNA was reverse - transcribed to synthesize cDNA (11141ES60, Shanghai Yeasen Biotechnology), and qRT-PCR was performed using a SYBR Green qPCR kit (11199ES08, Shanghai Yeasen Biotechnology). GAPDH was used as an internal reference, and the relative expression level of ATP6V0A4 was calculated by the 2⁻ΔΔCt method. The experiment was repeated more than 3 times. The primer sequences are listed in Table 1.

Table 1 RT qPCR primers used in this study
Full size table

Cell migration assay

SAS and HSC3 cells were seeded at a density of 8 × 10⁴ in the upper chamber of a Transwell insert (pore size 8 μm, Corning) with serum - free DMEM, and the lower chamber was filled with complete medium containing 10% FBS. After 24 h of culture for HSC3 cells and 30 h for SAS cells, the cells in the upper chamber were wiped off. The cells were fixed with 4% paraformaldehyde for 15 min and stained with crystal violet for 15 min. Five fields of view were randomly selected for each group for photography, and the experiment was repeated 3 times.

Cell invasion assay

The Transwell insert was coated with 60 µL of Matrigel (200 mg/mL, Corning) and incubated at 37 °C for 2 h. SAS and HSC3 cells were seeded at a density of 2 × 10⁵ in the upper chamber with serum - free DMEM, and the lower chamber was filled with complete medium containing 10% FBS. After 24 h of culture for SAS cells and 30 h for HSC3 cells, the cells in the upper chamber were wiped off. The fixation and staining methods were the same as those in the migration assay. Five fields of view were randomly selected for each group for photography, and the experiment was repeated 3 times.

Scratch wound - healing assay

The transfected SAS and HSC3 cells were cultured until they formed a confluent monolayer, and then a scratch was made with a pipette tip. After washing with PBS, the cells were cultured in serum - free medium. Photographs were taken at 0 h and 9 h (for HSC3) or 24 h (for SAS) after scratching. The wound - healing rate was calculated using ImageJ software (Wound - healing rate % = [(Area at 0 h - Area at 1 h)/Area at 0 h] × 100%). The experiment was repeated more than 3 times.

Cell proliferation assay

OSCC cells (2000 cells/well) were seeded in a 96 - well plate, and cell proliferation was detected using the Cell Counting Kit − 8 (Dojindo, Kumamoto, Japan) on the 1st, 3rd, and 5th days of culture. The experiment was repeated more than 3 times.

Western blot analysis

Briefly, cells were lysed using a cell lysis buffer (Cell Signaling Technology, Danvers, MA, USA) with protease and phosphatase inhibitors (Beyotime). The BCA protein assay kit (Beyotime) was used to isolate and measure the total protein content. Equal amounts of protein extracts (20–30 µg) were then separated using sodium dodecyl sulfate-polyacrylamide gel electrophoresis (PG112, Epizyme, China) and transferred onto polyvinylidene fluoride membranes (0.45 μm, Millipore, Burlington, MA, USA). The target protein was visualized using an enhanced chemiluminescence kit (P001AS, Beyotime) after incubation with primary and secondary antibodies. The following antibodies were used in the western blot assay: ATP6V0A4 (21570-1-AP, Proteintech).

Statistical analysis

R software (version 4.4.2) and its corresponding packages were used. Student’s t - test or Wilcoxon rank - sum test was used for comparisons between groups. Spearman’s correlation test was used to determine the interaction between variables. The chi - square test was used to analyze the correlation between ATP6V0A4 and clinicopathological characteristics. Kaplan - Meier curves and log - rank tests were used to compare overall survival (OS) among different ATP6V0A4 expression levels. Univariate Cox regression analysis was used to evaluate the independent prognostic value of ATP6V0A4 for OS. SPSS Statistics 16.0 software was used for data analysis, and the results were expressed as mean ± standard deviation. Student’s t - test was used for comparisons between groups with normally distributed data, and the Mann–Whitney U test was used for non - normally distributed data. A p - value < 0.05 was considered statistically significant.

Results

Identification of ATP6V0A4 as a critical prognostic biomarker in OSCC

First, differential expression analysis was performed on TCGA and GSE25099 datasets using R software (Fig. 1A). The results showed 2,273 and 413 differentially expressed genes (DEGs) in the TCGA and GSE25099 datasets, respectively. (Fig. 1B, C). By intersecting upregulated and downregulated gene sets, 217 common DEGs were identified, including 96 upregulated and 121 downregulated genes (Fig. 1D). Univariate analysis was then applied to screen core genes closely associated with the prognosis of oral squamous cell carcinoma (OSCC) patients, yielding 24 core genes (Fig. 1E). LASSO regression and random forest analyses further confirmed that ATP6V0A4 was the most critical gene (Fig. 1F-H). Subsequent investigations revealed that low ATP6V0A4 expression was significantly correlated with poor patient outcomes (Fig. 1I). Specifically, OSCC patients with lower ATP6V0A4 expression in tumor tissues exhibited worse treatment responses and survival outcomes. Collectively, these findings establish ATP6V0A4 as a key prognostic biomarker for OSCC.

Fig. 1
Fig. 1
Full size image

Identification of ATP6V0A4 as a critical prognostic biomarker in OSCC. (A) Schematic diagram of the bioinformatics analysis pipeline; (B, C) Volcano plot analysis based on TCGA and GSE25099 datasets respectively revealed differentially expressed genes between normal and tumor tissues; (D) Venn diagram displayed overlapping genes from the two datasets; (E) Univariate regression analysis preliminarily screened candidate prognosis-related genes; (F, G) Feature gene dimensionality reduction was performed using LASSO algorithm; (H) Random forest model (Fig. H) further validated the predictive performance of key genes; (I) Kaplan-Meier survival curves (Fig. I) showed significant correlation between ATP6V0A4 expression and overall survival in OSCC patients. Statistical significance was defined as P < 0.05

Expression and function of ATP6V0A4 in OSCC

To evaluate ATP6V0A4 expression, transcriptomic data from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) were comprehensively analyzed. Results showed that ATP6V0A4 expression was significantly downregulated in OSCC tumor tissues compared to normal tissues (Supplementary Fig. 1A, B). Comparative analysis across multiple cancer types revealed reduced ATP6V0A4 expression in kidney renal clear cell carcinoma (KIRC), kidney renal papillary cell carcinoma (KIRP), lung squamous cell carcinoma (LUSC), and prostate adenocarcinoma (PRAD), suggesting a shared feature across these malignancies. Conversely, ATP6V0A4 was upregulated in kidney chromophobe (KICH) and thyroid carcinoma (THCA), indicating context-dependent roles (Fig. 2A). Immunohistochemical analysis further confirmed reduced ATP6V0A4 protein expression in OSCC tissues, corroborating transcriptomic results (Fig. 2B). Gene Ontology (GO) enrichment analysis indicated that ATP6V0A4 may participate in biological processes such as synaptic vesicle lumen acidification and vacuolar proton-transporting V-type ATPase-related functions (Fig. 2C). Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis linked ATP6V0A4 to collecting duct acid secretion, lysosome function, and oxidative phosphorylation, suggesting roles in cellular acid-base balance, metabolism, and energy production (Supplementary Fig. 1C).

Fig. 2
Fig. 2
Full size image

Expression and function of ATP6V0A4 in OSCC. (A) TIMER analysis identified differential ATP6V0A4 expression across various cancer types and normal specimens; (B) Expression levels of ATP6V0A4 in normal and tumor tissues from the Human Protein Atlas (HPA) database; (C) Gene Ontology (GO) enrichment results for ATP6V0A4, including biological process (BP), cellular component (CC), and molecular function (MF) analyses. *p < 0.05, **p < 0.01, ***p < 0.001

Clinical pathological correlation of ATP6V0A4 in OSCC

Logistic regression analysis using clinical datasets revealed a significant association between ATP6V0A4 expression and T stage (tumor size and extent) (Fig. 3A, B). No correlations were observed with age, clinical stage, M stage, or N stage (Fig. 3C-G).

Fig. 3
Fig. 3
Full size image

ATP6V0A4 expression correlates with clinicopathological features in OSCC (A) Heatmap showing the distribution of clinical features between high and low ATP6V0A4 expression groups. Differential expression of ATP6V0A4 in OSCC patients according to clinical characteristics: (B) T stage, (C) Gender, (D) M stage, (E) N stage, (F) TNM stage, (G) Age. *p < 0.05, **p < 0.01

Association of ATP6V0A4 expression with immune infiltration

CIBERSORT algorithm analysis evaluated the correlation between ATP6V0A4 expression and 22 immune cell types, including B cells, T cells, natural killer (NK) cells, and macrophages. Pearson correlation analysis showed positive correlations between ATP6V0A4 expression and infiltration of T follicular helper cells, regulatory T cells, naive B cells, and resting dendritic cells, suggesting roles in immune cell recruitment or activation. Conversely, negative correlation was observed with resting NK cells (Fig. 4A-G). Immunotherapy response analysis revealed no significant differences in ATP6V0A4 expression between responders and non-responders to CTLA4 or PD1 inhibitors (Fig. 4H), indicating limited predictive value for these therapies under current conditions.

Fig. 4
Fig. 4
Full size image

Association of ATP6V0A4 expression with immune infiltration. (A) Comparison of immune cell infiltration between high- and low-expression subgroups of ATP6V0A4; (B-G) Correlation analysis between immune cells and ATP6V0A4 gene expression; (H) Immunotherapy response comparison between high- and low-expression subgroups of ATP6V0A4. *p < 0.05, **p < 0.01, ***p < 0.001

Correlation of ATP6V0A4 expression with drug sensitivity

Currently, radiotherapy, chemotherapy, and immunotherapy remain important treatment methods for OSCC. In order to improve treatment strategies and achieve personalized treatment, we conducted a predictive analysis of the drug sensitivity of OSCC. We calculated the correlation between the expression of ATP6V0A4 and drug sensitivity. Our results show that patients with low ATP6V0A4 expression are more sensitive to GDC0810, GSK591, and MK8776, while showing resistance to doramapimod, WIKI4, and PF-4,708,671 (Fig. 5A-F). Figure 5G-L describes the correlation between the IC50 values of the drugs and the expression level of ATP6V0A4. These results provide insights into the potential drug responses in OSCC.

Fig. 5
Fig. 5
Full size image

Drug sensitivity analysis. (A-F) IC50 values of various drugs in high- and low-expression groups of ATP6V0A4; (G-L) Correlation analysis between the IC50 values of drugs and the expression of ATP6V0A4 gene. *p < 0.05, **p < 0.01, ***p < 0.001

Overexpression of ATP6V0A4 inhibits OSCC proliferation, migration, and invasion

qRT-PCR confirmed that the expression of ATP6V0A4 was decreased in OSCC cell lines (SAS, HSC3, HN6, CAL27) compared with that in normal human oral keratinocyte (HOK) cells (Fig. 6A). Overexpression of ATP6V0A4 in SAS and HSC3 cells (via plasmid transfection, Fig. 6B-D) significantly inhibited cell migration and invasion (Fig. 6E-G), as demonstrated by scratch wound-healing and Transwell assays. CCK-8 proliferation assays showed reduced cell growth in ATP6V0A4-overexpressing cells (Fig. 6I). These results collectively indicate that ATP6V0A4 suppresses OSCC progression.

Fig. 6
Fig. 6
Full size image

Over expression of ATP6V0A4 inhibits malignant behaviors in OSCC cell lines. (A) Relative mRNA expression of ATP6V0A4 in OSCC and HOK cell lines. (B) Relative mRNA expression of ATP6V0A4 in SAS and HSC3 cells after plasmid transfection. (C, D) The protein expression situation and quantitative analysis of ATP6V0A4 in SAS cells and HSC3 cells after plasmid transfection. (E, F) Representative images and quantitative analysis of the migration and invasion of SAS cells and HSC3 cells in each group after plasmid transfection. Scale bar: 250 μm. (G, H) Representative images of wound-healing assays of transfected OSCC cells at 0 and 24 h and statistical results from each group. Scale bar:500 μm. (I) CCK-8 assay of transfected OSCC cells. Error bars represent the standard deviation. n = 3. *p < 0.05, **p < 0.01, ***p < 0.001

Discussion

Oral squamous cell carcinoma (OSCC), the most prevalent subtype of head and neck squamous cell carcinoma, has seen a rising global incidence, particularly in Asian and South American regions [20]. Its pathogenesis is closely linked to smoking, alcohol consumption, and HPV infection, characterized by high invasiveness, early metastasis, and poor prognosis [21]. Despite recent advancements, including the development of prognostic models based on clinicopathological features, mechanistic insights into EGFR/PI3K/MAPK [22] and Wnt/β-catenin signaling pathways [23], and explorations of immunotherapy for HPV-positive patients, the 5-year survival rate remains below 50%, compounded by escalating chemoradiotherapy resistance [24]. The complexity of tumor heterogeneity and immune evasion further complicates treatment strategies. Thus, there is an urgent need to identify highly specific biomarkers, unravel molecular mechanisms, and advance precision combinatorial therapies.

This study integrated TCGA and GSE25099 gene expression datasets through systematic bioinformatics analysis. Applying stringent criteria (|log2 fold change| ≥1.5 and p < 0.05) and cross-validation, 217 DEGs were identified. To construct a precise prognostic model for OSCC, a hierarchical statistical approach was employed: univariate Cox proportional hazards modeling first screened candidate genes significantly associated with overall survival (p < 0.05), followed by feature importance ranking and model optimization using Lasso regression and random forest algorithms. Notably, among the machine learning-validated signature genes, ATP6V0A4 emerged as the strongest prognostic predictor. Survival analysis revealed that reduced ATP6V0A4 expression correlated significantly with adverse outcomes, suggesting its potential as a critical biomarker for OSCC prognosis.

Previous studies have reported elevated ATP6V0A4 expression in highly invasive breast cancer and glioma [5, 7]. Contrastingly, this study observed marked downregulation of ATP6V0A4 in OSCC tissues and cell lines. Additionally, pan-cancer analysis revealed that the expression changes of ATP6V0A4 exhibit diverse characteristics in different cancer types: its expression levels are significantly reduced in renal cell carcinoma, KIRP, LUSC, and PRAD, while the gene is upregulated in THCA. These findings indicate the heterogeneity of ATP6V0A4 in various tumors, which may be closely associated with the diversity of tumor microenvironment characteristics, cellular functional regulatory networks, and gene regulatory mechanisms [25,26,27]. In vitro cell experiments demonstrated that ATP6V0A4 overexpression suppressed OSCC cell proliferation, migration, and invasion. Functional enrichment analyses implicated ATP6V0A4 in lysosomal function, oxidative phosphorylation, and acid secretion pathways. As a key subunit of the V-ATPase complex, its low expression may disrupt intracellular acidification homeostasis, impairing lysosomal function and promoting tumor invasion and metastasis [28]. Additionally, ATP6V0A4 is specifically expressed in the basal cell layer of normal oral mucosa, where its encoded proton pump subunit maintains epithelial barrier integrity by regulating cell polarity and extracellular pH [29]. During carcinogenesis, progressive loss of ATP6V0A4 expression coincides with epithelial dedifferentiation, leading to disrupted intercellular junctions and aberrant activation of matrix metalloproteinases. Collectively, these findings highlight ATP6V0A4 as a tumor suppressor in OSCC, with its downregulation may drive malignant progression through epithelial differentiation dysfunction and acidic microenvironment formation.

The immune microenvironment plays a pivotal role in tumor development. Correlation analysis revealed positive associations between ATP6V0A4 and follicular helper T cells, regulatory T cells, and resting dendritic cells. Follicular helper T cell infiltration is often linked to B cell activation and antibody responses, whereas regulatory T cell enrichment suggests an immunosuppressive microenvironment, partially explaining OSCC immune evasion [30]. The positive correlation with resting dendritic cells may reflect ATP6V0A4-mediated immune cell recruitment, though the precise mechanism warrants further investigation. Conversely, negative correlations with resting NK cells imply potential inhibition of innate immune surveillance. No significant differences in ATP6V0A4 expression were observed between CTLA4- and PD1-responsive groups, suggesting limited predictive value for these checkpoint inhibitors under current conditions. However, given the complexity of tumor immunity, ATP6V0A4 may hold predictive potential in other immunotherapeutic contexts or specific tumor subtypes, necessitating further exploration. ATP6V0A4 expression was significantly associated with clinical T staging, underscoring its potential as a biomarker for early diagnosis and personalized therapy.

In addition, this study proposed personalized drug treatment recommendations based on the expression level of ATP6V0A4. Our research results indicate that patients with low ATP6V0A4 expression may be suitable for treatment with GDC0810, GSK591, and MK8776. GDC0810 has been extensively reported to have potential in the treatment of breast cancer [31, 32]. As a PRMT5 inhibitor, GSK591, when used in combination with a PD-L1 blocking agent, can effectively inhibit the growth of liver cancer and improve the overall survival of mice, and it has also demonstrated good therapeutic effects in the treatment of B-cell lymphoma and colorectal cancer [33,34,35]. MK-8776 is a selective checkpoint kinase 1 (CHK1) inhibitor that can effectively inhibit the growth of neuroblastoma and ovarian cancer [36, 37]. ATP6V0A4, as a key subunit of the proton pump, plays a crucial role in maintaining the intracellular pH homeostasis, regulating vesicular transport, and promoting membrane fusion. We speculate that ATP6V0A4 may regulate the cellular sensitivity to drugs by controlling these biological processes. For example, ATP6V0A4 may be involved in maintaining the acidity of the tumor microenvironment, which can affect the stability of drugs and their trans-membrane transport efficiency; alternatively, it may regulate the intracellular vesicular transport system, thereby altering the targeted delivery of drugs within the cell and their sites of action, ultimately influencing the therapeutic effect. An in-depth exploration of the relationship between ATP6V0A4 and drug sensitivity holds significant theoretical and practical implications for optimizing individualized tumor treatment regimens and improving the precision of clinical medication.

This study has several limitations: mechanistic insights rely primarily on bioinformatics and require in vivo validation; immune therapy analysis is limited to PD1/CTLA4; and clinical sample size is small, necessitating multicenter validation. Future research should focus on dissecting ATP6V0A4-mediated tumor microenvironment regulation, developing ATP6V0A4-targeted antibody-drug conjugates (ADCs), and exploring interactions with HPV infection status to optimize stratified treatment strategies.

Conclusion

This study identifies ATP6V0A4 may serve as a novel tumor suppressor in OSCC, with downregulation associated with poor prognosis. These findings provide a theoretical basis for developing ATP6V0A4-targeted therapies and highlight the need to explore its role in epithelial differentiation and immune modulation to advance precision medicine for OSCC.