VAV2 drives glycolytic reprogramming in esophageal squamous cell carcinoma via EIF3F-mediated MTA1 deubiquitination

Abstract

Despite therapeutic advances, esophageal squamous cell carcinoma (ESCC) remains lethal due to metabolic adaptation via aerobic glycolysis. Here we report that VAV2 functions as an oncoprotein promoting this metabolic switch in ESCC. VAV2 overexpression enhances glycolytic flux—evidenced by increased glucose uptake, lactate production, and ECAR—while reducing oxygen consumption. Mechanistically, VAV2 upregulates EIF3F expression, enabling EIF3F to deubiquitinate and stabilize MTA1, which subsequently activates HIF-1α and its downstream targets GLUT1 and LDHA. EIF3F knockdown attenuates VAV2-driven oncogenic phenotypes and tumor progression in xenograft models, validating this axis as a functional dependency. This study uncovers a “signaling-deubiquitination-metabolism” regulatory nexus in ESCC and provides rationale for targeting VAV2-EIF3F interaction to overcome glycolysis-dependent therapy resistance.

Graphical Abstract

Supplementary Information

The online version contains supplementary material available at 10.1186/s40170-026-00447-0.

Introduction

Esophageal cancer is one of the most lethal malignancies worldwide, ranking seventh in incidence and sixth in cancer-related mortality globally [1, 2]. The two main histological subtypes, esophageal squamous cell carcinoma (ESCC) and esophageal adenocarcinoma (EAC), exhibit distinct epidemiological and pathological characteristics, with ESCC accounting for approximately 90% of cases in high-incidence regions including East Asia [3]. Despite advances in surgical techniques, chemoradiotherapy, and targeted therapy, the prognosis for esophageal cancer patients remains dismal, with a five-year survival rate below 20% [4]. The high mortality is primarily attributed to late diagnosis, early metastasis, and therapeutic resistance [5]. Therefore, elucidating the molecular mechanisms underlying esophageal cancer progression is urgently needed to identify novel diagnostic biomarkers and therapeutic targets.

Reprogramming of energy metabolism, particularly aerobic glycolysis (the Warburg effect), has been recognized as a hallmark of cancer that provides both ATP and biosynthetic precursors for rapid proliferation [6, 7]. Hypoxia-inducible factor 1α (HIF-1α) plays a central role in this metabolic switch by transcriptionally activating glycolytic genes including GLUT1 and LDHA [8, 9]. HIF-1α stability is tightly regulated by ubiquitination and proteasomal degradation [10], and its dysregulation contributes to aberrant glycolysis in various cancers [11]. However, the molecular events governing HIF-1α stability and subsequent glycolytic reprogramming in esophageal cancer remain largely unexplored.

VAV2, a member of the VAV family of guanine nucleotide exchange factors, functions as a signaling hub regulating cytoskeletal reorganization, cell migration, and proliferation [12, 13]. Its overexpression has been implicated in several malignancies including breast cancer and head and neck squamous cell carcinoma (hnSCC) [14, 15]. We previously reported that VAV2 is frequently overexpressed in esophageal squamous cell carcinoma and plays a critical role in DNA repair and radioresistance [16]. However, whether VAV2 contributes to metabolic reprogramming in ESCC remains unknown. MTA1, a core component of the NuRD complex, serves as a master regulator of cancer progression and has been shown to stabilize HIF-1α under normoxic conditions [17, 18]. Notably, MTA1 itself is regulated by ubiquitination and proteasomal degradation, but the upstream factors controlling its stability have not been identified. EIF3F, a subunit of the translation initiation factor eIF3 complex with intrinsic deubiquitinase activity [19], has been implicated in tumorigenesis through non-canonical functions beyond translation regulation [19, 20]. Based on these observations, we hypothesized that VAV2 may promote glycolytic reprogramming in ESCC by activating the EIF3F-MTA1-HIF-1α axis. Specifically, we propose that VAV2 upregulates EIF3F expression, enabling EIF3F to deubiquitinate and stabilize MTA1, which in turn activates HIF-1α and its downstream glycolytic targets.

In this study, we identify a previously unrecognized role of VAV2 in promoting glycolytic reprogramming and tumor growth in ESCC. Mechanistically, VAV2 interacts with EIF3F and upregulates its expression; EIF3F in turn binds to MTA1 and facilitates its deubiquitination, leading to MTA1 stabilization. Stabilized MTA1 subsequently activates HIF-1α signaling and its downstream glycolytic targets. Our findings reveal a novel VAV2-EIF3F-MTA1 regulatory axis linking ubiquitination control to metabolic reprogramming, providing potential therapeutic targets for ESCC intervention.

Materials and methods

Cell lines and culture conditions

Human esophageal cancer cell lines KYSE150(Cell Bank of the Chinese Academy of Sciences, TCHu150) and KYSE450(Cell Bank of the Chinese Academy of Sciences, TCHu450) were obtained from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China). Human embryonic kidney 293T (HEK293T) cells were obtained from the American Type Culture Collection (ATCC, CRL-3216). All cell lines were authenticated by short tandem repeat (STR) profiling and tested negative for mycoplasma contamination. KYSE150 and KYSE450 cells were cultured in RPMI-1640 medium (Gibco, 11875093) supplemented with 10% fetal bovine serum (FBS, Gibco, 10099141) and 1% penicillin-streptomycin (Gibco, 15140122). HEK293T cells were cultured in DMEM (Gibco, Catalog No. 11965092) supplemented with 10% FBS and 1% penicillin-streptomycin. All cells were maintained at 37 °C in a humidified atmosphere containing 5% CO₂.

Antibodies and reagents

The following primary antibodies were used: anti-VAV2 (Biodragon, RM4685, 1:1000 dilution), anti-EIF3F (Abcam, ab176853, 1:3000 dilution), anti-MTA1 (Proteintech, 30545-1-AP, 1:1000 dilution), anti-HIF-1α (Proteintech, 66730-1-Ig, 1:2000 dilution), anti-GLUT1 (Abcam, ab115730, 1:1000 dilution), anti-LDHA (Abcam, ab52488, 1:1000 dilution), anti-Myc tag (Yeasen, 30601ES60, 1:1000 dilution), anti-Flag tag (Yeasen, 30505ES60, 1:1000 dilution), anti-HA tag (Yeasen, 30704ES60, 1:1000 dilution), anti-β-Actin (Proteintech, 66009-1-Ig, 1:5000 dilution), and anti-Ki67 (Abcam, ab15580, 1:1000 dilution). Horseradish peroxidase (HRP)-conjugated secondary antibodies, including goat anti-rabbit IgG (Abcam, ab6721, 1:5000 dilution) and goat anti-mouse IgG (Abcam, ab205719, 1:5000 dilution), were used. Cycloheximide (CHX, Sigma-Aldrich, C4859, 100 µg/mL) and MG132 (Sigma-Aldrich, C2211, 20 µM) were obtained from Sigma-Aldrich.

Plasmid construction and transfection

Full-length human VAV2 cDNA was cloned into the pCDH-EF1-MCS-T2A-Puro vector with a Flag tag. Full-length human EIF3F cDNA was cloned into the pCDH-EF1-MCS-T2A-Puro vector. Full-length human MTA1 cDNA was cloned into the pcDNA3.1-Myc vector with a Myc tag. Short hairpin RNAs (shRNAs) targeting VAV2, EIF3F, and MTA1, as well as non-targeting control shRNA (shNC), were designed and cloned into the pLKO.1 vector. For CRISPR-Cas9 mediated knockout, sgRNAs targeting VAV2 were cloned into the pUC19-U6-sgRNA plasmid and co-transfected with pCAG-Cas9-EGFP into KYSE150 and KYSE450 cells as previously described [16]. All constructs were verified by colony PCR screening, restriction enzyme digestion, and Sanger sequencing using vector-specific primers (CMV-F for pcDNA3.1 vectors; EF1α-F or pCDH-F for pCDH vectors; U6-Fwd for pUC19-U6-sgRNA vectors; pLKO.1-F for pLKO.1 vectors). The targeting sequences for shRNAs are listed in Supplementary Table S1. Plasmid transfections were performed using Lipofectamine 3000 (Invitrogen, L3000015) according to the manufacturer’s instructions.

Lentivirus production and stable cell line establishment

Lentiviruses were produced by co-transfecting HEK293T cells with the lentiviral expression plasmid and packaging plasmids (psPAX2 and pMD2.G) using Lipofectamine 3000. Viral supernatants were collected 48 and 72 h post-transfection, filtered through 0.45 μm filters, and used to infect target cells in the presence of 8 µg/mL polybrene. Infected cells were selected with puromycin (2 µg/mL) for 14 days to establish stable cell lines.

RT-qPCR

Total RNA was extracted from cultured cells using TRIzol reagent (Invitrogen, 15596018) according to the manufacturer’s protocol. RNA concentration and purity were measured using a NanoDrop spectrophotometer. Complementary DNA (cDNA) was synthesized from 1 µg of total RNA using PrimeScript RT reagent Kit (Takara, RR037A). Quantitative real-time PCR was performed using SYBR Green Premix Ex Taq (Takara, RR420A) on a StepOnePlus Real-Time PCR System (Applied Biosystems). The primer sequences used for qRT-PCR are listed in Supplementary Table S2. Relative gene expression was calculated using the 2⁻^ΔΔCt method with ACTB as the internal control.

Western blot analysis

Cells were lysed in RIPA buffer (Beyotime, P0013B) supplemented with protease and phosphatase inhibitor cocktails. Protein concentrations were determined using a BCA protein assay kit (Thermo Scientific, 23225). Equal amounts of protein (20–40 µg) were separated by SDS-PAGE and transferred onto PVDF membranes (Millipore, IPVH00010). Membranes were blocked with 5% skim milk in TBST for 1 h at room temperature, followed by incubation with primary antibodies overnight at 4 °C. After washing with TBST, membranes were incubated with HRP-conjugated secondary antibodies for 1 h at room temperature. Protein bands were visualized using enhanced chemiluminescence (ECL) substrate (Millipore, WBKLS0500) and detected using a ChemiDoc MP Imaging System (Bio-Rad). β-Actin was used as a loading control.

Co-Immunoprecipitation (Co-IP) assay

For exogenous Co-IP, cells were transfected with indicated plasmids for 48 h, then lysed with IP lysis buffer (Beyotime, P0013) containing protease inhibitors. Cell lysates were incubated with specific antibodies or control IgG overnight at 4 °C, followed by incubation with Protein A/G agarose beads (Santa Cruz Biotechnology, sc-2003) for 4 h at 4 °C. For endogenous Co-IP, cells were directly lysed and incubated with antibodies against endogenous proteins. After washing with lysis buffer, immunoprecipitated proteins were eluted by boiling in SDS sample buffer and analyzed by Western blot.

Immunofluorescence (IF) staining

Cells cultured on glass coverslips were fixed with 4% paraformaldehyde for 15 min, permeabilized with 0.2% Triton X-100 for 10 min, and blocked with 5% bovine serum albumin (BSA) for 1 h at room temperature. Cells were then incubated with primary antibodies overnight at 4 °C, followed by incubation with Alexa Fluor-conjugated secondary antibodies (Invitrogen, A11008) for 1 h at room temperature. Nuclei were counterstained with DAPI. Coverslips were mounted with anti-fade mounting medium and visualized under a confocal laser scanning microscope (Olympus, FV3000). For colocalization analysis, images were acquired with identical settings. Pearson’s correlation coefficients (PCC) were calculated using the Coloc 2 plugin in Fiji from individual cells after background subtraction and Costes threshold regression. Scatterplots were generated for each representative image, and PCC values were annotated on the plots.

Ubiquitination assay

For in vivo ubiquitination assays, cells were co-transfected with the indicated plasmids together with HA-ubiquitin. Forty-eight hours post-transfection, cells were treated with MG132 (20 µM) for 8 h before harvesting. Cells were lysed with denaturing lysis buffer (1% SDS, 50 mM Tris-HCl pH 7.5, 5 mM EDTA) and boiled for 10 min. Lysates were diluted with regular lysis buffer to reduce SDS concentration to 0.1%, then subjected to immunoprecipitation with anti-Myc or anti-MTA1 antibodies. Ubiquitinated proteins were detected by Western blot using anti-HA antibody.

CHX chase assay

Cells were seeded in 6-well plates and treated with cycloheximide (CHX) at a concentration of 100 µg/mL for the indicated time points (0, 2, 4, 8 h). Cells were harvested at each time point, and protein levels were analyzed by Western blot. Protein band intensities were quantified using ImageJ software and normalized to the 0-hour time point.

MG132 rescue assay

Cells were treated with MG132 (20 µM) or DMSO vehicle control for 8 h before harvesting. Protein levels were analyzed by Western blot.

Glucose uptake and lactate production assays

Glucose uptake and lactate production were measured using commercial kits according to the manufacturer’s instructions. Briefly, cells were seeded in 6-well plates at equal densities. Culture medium was collected, and glucose consumption and lactate production were normalized to cell number or protein concentration. Glucose uptake was measured using the Glucose Uptake Assay Kit (Beyotime, S0556S). Lactate production was measured using the Lactate Assay Kit (Beyotime, S0208S).

Seahorse metabolic analysis

Extracellular acidification rate (ECAR) and oxygen consumption rate (OCR) were measured using a Seahorse XFe96 Analyzer (Agilent Technologies). Cells were seeded in XFe96 cell culture microplates at a density of 1.0-1.2 × 10⁴ cells per well and incubated overnight. For ECAR measurement, cells were incubated with assay medium supplemented with 2 mM L-glutamine (Seahorse XF DMEM, pH 7.4, Agilent, 103575-100). For OCR measurement, cells were incubated with assay medium supplemented with 10 mM glucose, 1 mM sodium pyruvate, and 2 mM L-glutamine. ECAR and OCR were measured under basal conditions and following sequential injections of 10 mM glucose, 1.0 µM oligomycin, and 50 mM 2-DG (2-deoxy-D-glucose) for ECAR; 1.5 µM oligomycin, 1.0 µM FCCP, and 0.5 µM rotenone/antimycin A for OCR. Data were normalized to cell number or protein content and analyzed using Wave software.

Xenograft tumor model

All animal experiments were approved by the Animal Care and Use Committee of Hubei Provincial Center for Disease Control and Prevention (Animal Ethics No: 202610059). Female BALB/c nude mice (4–6 weeks old) were housed under specific pathogen-free conditions. For tumor implantation, mice were anesthetized by inhalation of isoflurane (induction at 3–5%, maintenance at 1.5–2.5% in 100% oxygen) prior to subcutaneous injection. KYSE150 cells stably transfected with the indicated constructs (Vector+shNC, OE-VAV2 + shNC, Vector + sh-EIF3F, OE-VAV2 + sh-EIF3F) were harvested and resuspended in PBS. A total of 5 × 10⁶ cells in 100 µL PBS were subcutaneously injected into the right flank of each mouse. Each experimental group contained 6 mice (n = 6 per group). Tumor volumes were measured every three days using a caliper and calculated using the formula: volume = (length × width²)/2. Mice were euthanized 4 weeks after injection, and tumors were excised, photographed, and weighed. Tumor tissues were then processed for Western blot and immunohistochemistry analysis.

Immunohistochemistry (IHC) staining

Xenograft tumor tissues were fixed in 4% paraformaldehyde, embedded in paraffin, and sectioned at 4 μm thickness. Sections were deparaffinized, rehydrated, and subjected to antigen retrieval by boiling in citrate buffer (pH 6.0). Endogenous peroxidase activity was blocked with 3% H₂O₂. Sections were then blocked with 5% BSA and incubated with primary antibodies against VAV2, EIF3F, MTA1, and Ki67 overnight at 4 °C. After washing, sections were incubated with HRP-conjugated secondary antibodies, followed by DAB substrate staining (Dako Liquid DAB+ Substrate Chromogen System, Catalog No. K3468) and hematoxylin counterstaining. Images were captured using a Nikon Eclipse E400 microscope or Leica DM4000 microscope. Scale bars represent 100 μm for 200× magnification and 50 μm for 400× magnification.

RNA sequencing and data analysis

The transcriptome sequencing data of VAV2-knockdown and control KYSE150 cells were obtained from our previously published study [16]. Briefly, total RNA was extracted using TRIzol reagent, and library preparation and sequencing were performed on an Illumina platform with paired-end 150 bp reads. Raw sequencing data were processed as described previously [16]. For the present study, differentially expressed genes (DEGs) were identified with criteria of |fold change| > 1.5 and adjusted P-value < 0.05. Gene Ontology Biological Process (GOBP) analysis was performed using the clusterProfiler package in R. Single-sample Gene Set Enrichment Analysis (ssGSEA) was performed using the GSVA package in R to evaluate the enrichment scores of glycolytic process and HIF-1α pathway-related gene sets.

DIA proteomics and data analysis

The DIA proteomics data of VAV2-knockout and control KYSE150 cells were derived from our previous study [16]. In the current analysis, differentially expressed proteins were identified with criteria of |fold change| > 1.5 and adjusted P-value < 0.05. Gene Ontology Molecular Function (GOMF) analysis was performed using the clusterProfiler package in R.

IP-MS analysis

The immunoprecipitation-mass spectrometry data for VAV2-interacting proteins were obtained from our previously published work [16]. Proteins identified in the VAV2-IP samples but not in the IgG control were considered candidate VAV2-interacting proteins. For the present study, we re-analyzed this dataset to identify potential interactions relevant to the VAV2-EIF3F-MTA1 axis.

Molecular docking

Molecular docking was performed to predict the binding modes among VAV2, EIF3F, and MTA1. Protein structures were obtained from the RCSB Protein Data Bank: VAV2 (PDB ID: 7WFY, chain C), EIF3F (PDB ID: 6YBD, chain 4), and MTA1 (PDB ID: 4BKX, chain A). Docking simulations were performed using the HDOCK server, with sequential docking of VAV2-EIF3F, MTA1-EIF3F, and the ternary complex. The docking models with the lowest binding energy were selected for visualization using PyMOL software.

Bioinformatics analysis of public datasets

Single-cell RNA sequencing data (GSE203067) were downloaded from the Gene Expression Omnibus (GEO) database. UMAP visualization, feature plots, and violin plots were generated using the Seurat package in R. TCGA esophageal carcinoma data were accessed through the GEPIA2 database (http://gepia2.cancer-pku.cn). VAV2 expression levels in normal and tumor tissues were analyzed and visualized as box plots.

Statistical analysis

All experiments were performed independently at least three times. Data are presented as mean ± standard deviation (SD). Statistical analyses were performed using GraphPad Prism 10.1.2 software. Prior to parametric analysis, the normality of data distribution was assessed using the Shapiro-Wilk test, and the homogeneity of variances was evaluated using Levene’s test. Comparisons between two groups were analyzed using two-tailed Student’s t-test when normality and equal variance assumptions were met; otherwise, the Mann-Whitney U test was used. Comparisons among multiple groups were analyzed using one-way or two-way ANOVA followed by appropriate post-hoc tests. Specifically, Tukey’s honestly significant difference (HSD) test was used as the post-hoc analysis following one-way ANOVA, and Šidák’s multiple comparisons test was applied following two-way ANOVA. For data that violated the assumptions of normality or homoscedasticity, the Kruskal-Wallis test followed by Dunn’s post-hoc test was used for multiple-group comparisons. P-values < 0.05 were considered statistically significant. Significance levels are denoted as *p < 0.05, **p < 0.01, ***p < 0.001; ns, not significant.

Results

VAV2 expression is upregulated in ESCC

To investigate the potential role of VAV2 in ESCC, we first analyzed its expression pattern using publicly available single-cell and bulk transcriptome datasets. Single-cell RNA sequencing data from the GSE203067 cohort revealed distinct cell clusters in ESCC tissues (Fig. 1A), with canonical marker genes confirming the identity of each cluster (Fig. 1B). UMAP feature plots revealed that VAV2 expression was markedly higher in ESCC tumor epithelial cells compared to normal epithelial cells (Fig. 1C-D). Furthermore, VAV2 mRNA levels were significantly higher in tumor tissues compared to adjacent non-cancer tissues in the GSE203067 dataset (Fig. 1E). Consistently, analysis of TCGA esophageal carcinoma cohort demonstrated elevated VAV2 expression in primary tumor tissues (n = 182) relative to paired normal tissues (n = 286) (Fig. 1F). This upregulation was further validated in TCGA esophageal carcinoma samples, where primary tumors (n = 184) exhibited markedly higher VAV2 expression than normal tissues (n = 11) (Fig. 1G).

Fig. 1.

VAV2 Expression Is Upregulated in ESCC. (A) UMAP visualization of single-cell transcriptomes from ESCC tissues (GSE203067), colored by cell clusters. (B) Dot plot showing expression of canonical marker genes across identified cell clusters. Dot size represents percentage of cells expressing each gene, and color intensity indicates average expression level. (C) UMAP feature plot showing VAV2 expression in normal epithelial cells. (D) UMAP feature plot showing VAV2 expression in ESCC tumor epithelial cells. (E) Violin plot comparing VAV2 mRNA levels between tumor and adjacent non-cancer tissues in GSE203067 dataset. (F) Box plot showing VAV2 expression levels in peritumoral normal tissues (N = 286) and esophageal carcinoma (ESCA) tumor tissues (T = 182) from the GEPIA2 database. (G) VAV2 expression in normal (n = 11) and primary tumor (n = 184) samples from TCGA esophageal carcinoma cohort. (H) ROC curve analysis of VAV2 expression for distinguishing esophageal carcinoma tissues from normal tissues in the TCGA cohort (AUC = 0.791). (I) VAV2 expression levels stratified by pathological stage (Stage I vs. Stage II–IV) in the TCGA esophageal carcinoma cohort. Data are presented as mean ± SD. Statistical significance was determined by two-tailed Student’s t-test. *p < 0.05, **p < 0.01

We further assessed the clinical significance of VAV2 upregulation. ROC curve analysis showed that VAV2 expression distinguished esophageal carcinoma tissues from normal tissues with an AUC of 0.791 (Fig. 1H), indicating moderate diagnostic accuracy. Additionally, VAV2 expression levels were significantly higher in Stage II–IV tumors compared to Stage I tumors (Fig. 1I), suggesting that elevated VAV2 correlates with advanced disease progression. Collectively, these results indicate that VAV2 is significantly upregulated in ESCC and may contribute to tumor progression.

VAV2 promotes glycolytic reprogramming in ESCC Cells

To explore the functional significance of VAV2 in ESCC, we performed transcriptome sequencing on VAV2-knockdown KYSE150 cells. GOBP analysis revealed that differentially expressed genes were significantly enriched in metabolic processes (Fig. 2A). Notably, ssGSEA demonstrated that VAV2 knockdown markedly reduced the enrichment score of the glycolytic process gene set (Fig. 2B). We then established VAV2-overexpressing and VAV2-knockout KYSE150 cells, as confirmed by Western blot analysis (Fig. 2C, D). Functional assays showed that VAV2 overexpression significantly enhanced glucose uptake by approximately 2.1-fold(Fig. 2E) and lactate production by approximately 2.3-fold(Fig. 2G), while VAV2 knockout reduced glucose uptake by approximately 45% (Fig. 2F) and lactate production by approximately 55% (Fig. 2H). Moreover, Seahorse analysis revealed that VAV2 overexpression increased ECAR (Fig. 2I) and decreased OCR (Fig. 2K), whereas VAV2 knockout decreased ECAR (Fig. 2J) and increased OCR (Fig. 2L), indicating that VAV2 promotes a metabolic shift toward glycolysis. Similar results were observed in KYSE450 cells (Supplementary Fig. S1A-J). These findings indicate that VAV2 promotes glycolytic reprogramming in ESCC cells.

Fig. 2.

VAV2 Promotes Glycolytic Reprogramming in ESCC Cells. (A) Gene Ontology Biological Process (GOBP) analysis of differentially expressed genes (DEGs) based on transcriptome sequencing data from VAV2-knockdown KYSE150 cells compared to control cells. (B) Single-sample Gene Set Enrichment Analysis (ssGSEA) scores for the GOBP glycolytic process in VAV2-knockdown and control KYSE150 cells, derived from the transcriptome data. (C) Western blot analysis confirming VAV2 overexpression efficiency in KYSE150 cells. (D) Western blot analysis confirming VAV2 knockout efficiency in KYSE150 cells. (E) Glucose uptake in KYSE150 cells upon VAV2 overexpression. (F) Glucose uptake in KYSE150 cells upon VAV2 knockout. (G) Lactate production in KYSE150 cells upon VAV2 overexpression. (H) Lactate production in KYSE150 cells upon VAV2 knockout. (I) Extracellular acidification rate (ECAR) in KYSE150 cells upon VAV2 overexpression. (J) Extracellular acidification rate (ECAR) in KYSE150 cells upon VAV2 knockout. (K) Oxygen consumption rate (OCR) in KYSE150 cells upon VAV2 overexpression. (L) Oxygen consumption rate (OCR) in KYSE150 cells upon VAV2 knockout. Data are presented as mean ± SD of at least three independent experiments. For (E-L), statistical significance was determined by two-tailed Student’s t-test. The p-values are denoted by asterisks as follows: * p < 0.05; ** p < 0.01

VAV2 regulates HIF-1α pathway and glycolysis via MTA1

To elucidate the mechanism by which VAV2 regulates glycolysis, we analyzed the overlap between VAV2 knockdown-associated differentially expressed proteins (from DIA proteomics) and glycolytic gene sets, identifying MTA1 as a potential downstream effector (Fig. 3A). ssGSEA further revealed that VAV2 knockdown significantly correlated with suppression of HIF-1α pathway-related gene sets (Fig. 3B-D). qPCR analysis showed that VAV2 overexpression did not affect MTA1 mRNA levels (Fig. 3E, F), whereas Western blot analysis demonstrated that VAV2 overexpression increased MTA1 protein expression (Fig. 3G), and VAV2 knockdown decreased it (Fig. 3H). Consistently, VAV2 overexpression upregulated HIF-1α and its target LDHA (Fig. 3I) while VAV2 knockdown reduced VAV2, HIF-1α, GLUT1, and LDHA expression (Fig. 3J). Rescue experiments demonstrated that MTA1 knockdown partially reversed VAV2 overexpression-induced upregulation of HIF-1α, GLUT1 and LDHA (Fig. 3K), as well as the enhanced glucose uptake (reversed to approximately 0.49-fold of VAV2 overexpression alone), lactate production (reversed to approximately 0.46-fold of VAV2 overexpression alone), ECAR, and reduced OCR (Fig. 3L-O). Similar results were observed in KYSE450 cells (Supplementary Fig. S2A-K).

Fig. 3.

VAV2 Regulates HIF-1α Pathway and Glycolysis via MTA1 in ESCC Cells. (A) Venn diagram showing the overlap of differentially expressed proteins identified from DIA proteomics analysis (VAV2-knockdown vs. control KYSE150 cells) and genes associated with glycolytic process gene sets. (B-D) ssGSEA scores for HIF-1α pathway-related gene sets based on transcriptome data from VAV2-knockdown and control KYSE150 cells, showing significant correlation with VAV2 knockdown. (E, F) qPCR analysis of VAV2 (E) and MTA1 (F) mRNA levels in KYSE150 cells with VAV2 overexpression. (G) Western blot analysis of VAV2 and MTA1 protein levels in KYSE150 cells upon VAV2 overexpression. (H) Western blot analysis of VAV2 and MTA1 protein levels in KYSE150 cells upon VAV2 knockout. (I) Western blot analysis of HIF-1α, MTA1, and LDHA levels in KYSE150 cells following VAV2 overexpression. (J) Western blot analysis of VAV2, HIF-1α, MTA1, GLUT1, and LDHA levels in KYSE150 cells following VAV2 knockdown. (K) Western blot analysis of VAV2, HIF-1α, MTA1, GLUT1, and LDHA protein levels in KYSE150 cells co-transfected with indicated combinations of OE-VAV2 (or Vector) and sh-MTA1 (or sh-NC). (L-O) Glucose uptake (L), lactate production (M), extracellular acidification rate (ECAR; N), and oxygen consumption rate (OCR; O) in KYSE150 cells transfected with the indicated combinations of OE-VAV2 (or Vector) and shMTA1 (or shNC). Data are presented as mean ± SD of at least three independent experiments. For (B-F), statistical significance was determined by two-tailed Student’s t-test. For (L-O), one-way ANOVA with Tukey’s post-hoc test was used for multiple-group comparisons. Significant differences were only observed for Glycolytic Capacity and Non-mitochondrial; no significant differences were found for Basal Glycolysis, Glycolysis, ATP Production, or Maximal Respiration among groups. The p-values are denoted by asterisks as follows: * p < 0.05; ** p < 0.01; *** p < 0.001; ns, not significant

Having established that MTA1 mediates VAV2-driven HIF-1α upregulation, we proceeded to investigate the direct mechanism. Co-immunoprecipitation assays revealed that endogenous MTA1 interacts with endogenous HIF-1α in KYSE150 cells, which was further confirmed using exogenous MTA1-Myc and HIF-1α-Flag (Supplementary Fig. S3A). CHX chase assays showed that MTA1 overexpression prolonged the half-life of HIF-1α protein, whereas MTA1 knockdown accelerated its degradation (Supplementary Fig. S3B). Treatment with the proteasome inhibitor MG132 rescued MTA1 knockdown-induced reduction of HIF-1α protein levels (Supplementary Fig. S3C), suggesting that MTA1 stabilizes HIF-1α by inhibiting its proteasomal degradation. Indeed, ubiquitination assays demonstrated that MTA1 overexpression significantly reduced the ubiquitination level of HIF-1α (Supplementary Fig. S3D). Collectively, these results indicate that MTA1 directly binds to HIF-1α and protects it from ubiquitin-proteasome-dependent degradation under normoxic conditions.

VAV2 stabilizes MTA1 protein by inhibiting Its Ubiquitination

Given that VAV2 regulated MTA1 at the protein level without affecting its mRNA, we investigated the underlying post-translational mechanism. GOMF analysis of DIA data from VAV2-knockout KYSE150 cells revealed enrichment of ubiquitination-related molecular functions (Fig. 4A), and GSEA confirmed a negative correlation between VAV2 knockout and ubiquitination gene sets (Fig. 4B). CHX chase assays showed that VAV2 overexpression prolonged the half-life of MTA1 protein in both KYSE150 and KYSE450 cells (Fig. 4C, D). Furthermore, treatment with the proteasome inhibitor MG132 abolished the VAV2 knockdown-induced decrease in MTA1 protein levels (Fig. 4E, F), suggesting involvement of the ubiquitin-proteasome pathway. Ubiquitination assays revealed that VAV2 overexpression significantly reduced MTA1 ubiquitination (Fig. 4G). These results demonstrate that VAV2 stabilizes MTA1 by inhibiting its proteasomal degradation through deubiquitination.

Fig. 4.

VAV2 Stabilizes MTA1 Protein by Inhibiting Its Ubiquitination. (A) Gene Ontology Molecular Function (GOMF) analysis of differentially expressed proteins based on DIA data from VAV2-knockout KYSE150 cells compared to control cells. (B) Gene Set Enrichment Analysis (GSEA) showing negative correlation between VAV2 knockout and ubiquitination-related gene sets. (C, D) CHX (100 µg/ml) chase assay assessing MTA1 stability in KYSE150 (C) and KYSE450 (D) cells with OE-VAV2 or vector. (E, F) MG132 (20 µM) assay in KYSE150 (E) and KYSE450 (F) cells with OE-VAV2 or vector, treated with or without MG132. (G) Ubiquitination assay of MTA1 in KYSE150 cells co-transfected with indicated plasmids

VAV2 promotes EIF3F-mediated deubiquitination of MTA1

To identify the mechanism underlying MTA1 stabilization, we integrated DIA proteomics, IP-MS, and ubiquitination datasets, revealing EIF3F as a candidate protein potentially interacting with both VAV2 and MTA1 (Fig. 5A, B). Molecular docking predicted a direct binding interface between VAV2 and EIF3F (Fig. 5C), prompting us to validate this interaction experimentally. We first confirmed that VAV2 regulates EIF3F expression: VAV2 overexpression increased, while VAV2 knockout decreased, EIF3F protein levels in KYSE150 cells (Fig. 5D, E) and KYSE450 cells (Supplementary Fig. S4A, B). Co-immunoprecipitation assays confirmed that endogenous VAV2 interacts with endogenous EIF3F, and this interaction was further verified using exogenous VAV2-Flag in KYSE150 cells (Fig. 5F); similar results were obtained in KYSE450 cells (Supplementary Fig. S4C). Immunofluorescence analysis revealed their colocalization in the cytoplasm of both cell lines (Fig. 5G and Supplementary Fig. S4D).

Fig. 5.

VAV2 Promotes EIF3F-Mediated Deubiquitination of MTA1. (A, B) Venn diagram (A) and heatmap (B) showing the overlapping gene identified from DIA data, IP-MS, and ubiquitination assays. (C) Molecular docking model depicting the interaction between VAV2 and EIF3F. (D, E) Western blot analysis of VAV2 and EIF3F protein levels in KYSE150 cells upon VAV2 overexpression (D) or knockout (E). (F) Co-immunoprecipitation (Co-IP) assays confirming the interaction between endogenous VAV2 and EIF3F (left) and between exogenous VAV2-Flag and endogenous EIF3F (right) in KYSE150 cells. (G) Immunofluorescence (IF) analysis showing colocalization of VAV2 and EIF3F in KYSE150 cells. Scale bar, 50 μm (for 400×). (H) Molecular docking model depicting the interaction between MTA1 and EIF3F. (I, J) Western blot analysis of MTA1 protein levels in KYSE150 cells upon EIF3F overexpression (I) or knockdown (J). (K) Co-IP assays confirming the interaction between endogenous EIF3F and MTA1 (left) and between exogenous MTA1-Myc and endogenous EIF3F (right) in KYSE150 cells. (L) Immunofluorescence analysis showing colocalization of EIF3F and MTA1 in KYSE150 cells with or without EIF3F overexpression. Scale bar, 50 μm (for 400×). (M) Ubiquitination assay showing MTA1 ubiquitination levels in 293T cells transfected with EIF3F or vector control. (N) Ubiquitination assay showing MTA1 ubiquitination levels in 293T cells co-transfected with VAV2-Flag and sh-EIF3F or control. (O) Western blot analysis of MTA1 in VAV2-knockdown KYSE150 cells reconstituted with wild-type EIF3F or EIF3F-ΔMPN (DUB-dead) mutant. (P) Ubiquitination assays showing that EIF3F overexpression specifically reduces K48-linked but not K63-linked polyubiquitination of MTA1 in 293T cells

We then examined the relationship between EIF3F and MTA1. Molecular docking predicted a direct interaction between MTA1 and EIF3F (Fig. 5H). Notably, molecular docking revealed that EIF3F possesses distinct binding interfaces for VAV2 and MTA1 (Fig. 5C and H; Supplementary Fig. S4G), suggesting that EIF3F can functionally couple VAV2-mediated signaling to MTA1 stabilization without direct competition between the two interactions. Functionally, EIF3F overexpression increased MTA1 protein levels, while EIF3F knockdown decreased them in KYSE150 cells (Fig. 5I, J); consistent results were observed in KYSE450 cells (Supplementary Fig. S4E, F). Co-IP assays confirmed that endogenous EIF3F binds to endogenous MTA1, and this interaction was further validated with exogenous MTA1-Myc in KYSE150 cells (Fig. 5K), with parallel findings in KYSE450 cells (Supplementary Fig. S4H). Immunofluorescence showed their colocalization in both cell lines, which was enhanced upon EIF3F overexpression (Fig. 5L and Supplementary Fig. S4I).

To determine whether EIF3F mediates MTA1 deubiquitination, we determined whether EIF3F mediates MTA1 deubiquitination. Ubiquitination assays revealed that EIF3F overexpression significantly reduced MTA1 ubiquitination in 293T cells (Fig. 5M). Furthermore, VAV2 overexpression-induced reduction of MTA1 ubiquitination was largely attenuated by EIF3F knockdown (Fig. 5N).

To investigate how VAV2 upregulates EIF3F, we examined whether this regulation occurs at the transcriptional level. RT-qPCR analysis revealed that VAV2 overexpression significantly increased EIF3F mRNA levels in both KYSE150 and KYSE450 cells (Supplementary Fig. 5A–D). Consistently, dual-luciferase reporter assays demonstrated that VAV2 overexpression markedly enhanced the transcriptional activity of the EIF3F promoter (Supplementary Fig. 5E–F). Moreover, correlation analysis using the TCGA dataset revealed a significant positive correlation between VAV2 and EIF3F mRNA expression in esophageal carcinoma tissues (Supplementary Fig. 5G). Additionally, bioinformatic analysis identified potential c-Myc binding motifs within the EIF3F promoter region (Supplementary Fig. 5H), and the VAV2-Rac1 signaling axis is known to activate c-Myc, providing a plausible mechanistic link. Collectively, these data indicate that VAV2 upregulates EIF3F primarily at the transcriptional level, likely through activating downstream transcription factors, rather than by inhibiting its protein degradation.

We further generated a DUB-dead mutant of EIF3F (EIF3F-ΔMPN) and found that it failed to rescue MTA1 stability in VAV2-knockdown cells (Fig. 5O). These results confirm that the deubiquitinase activity of EIF3F is essential for MTA1 regulation. To determine the type of ubiquitin linkage removed by EIF3F, we performed ubiquitination assays using ubiquitin mutants that preserve only K48 or K63 linkages. EIF3F overexpression significantly reduced K48-linked polyubiquitination of MTA1 but had no effect on K63-linked polyubiquitination (Fig. 5P). These findings indicate that EIF3F specifically cleaves K48-linked ubiquitin chains from MTA1, which are canonical signals for proteasomal degradation. Taken together, these findings demonstrate that VAV2 upregulates EIF3F, which directly deubiquitinates MTA1.

VAV2 promotes glycolysis and tumor growth via EIF3F in Vitro and in Vivo

We next performed rescue experiments to determine whether VAV2 promotes glycolysis and tumor growth through EIF3F. In vitro, VAV2 overexpression in KYSE150 cells increased the expression of VAV2, EIF3F, MTA1, HIF-1α, GLUT1, and LDHA, and these effects were partially reversed by EIF3F knockdown (Fig. 6A). Consistently, EIF3F knockdown attenuated VAV2 overexpression-induced enhancement of lactate production (to approximately 0.66-fold of VAV2 overexpression alone) (Fig. 6B), glucose uptake (to approximately 0.58-fold of VAV2 overexpression alone) (Fig. 6C), ECAR (Fig. 6D), and suppression of OCR (Fig. 6E). Similar results were observed in KYSE450 cells (Supplementary Fig. S6A-E). To further confirm that EIF3F acts downstream of VAV2, we performed reciprocal rescue experiments by overexpressing EIF3F in VAV2-deficient KYSE150 cells. VAV2 knockout significantly reduced the expression of VAV2, EIF3F, MTA1, HIF-1α, GLUT1, and LDHA, along with decreased glucose uptake, lactate production, and ECAR, and increased OCR. EIF3F overexpression in VAV2-deficient cells restored the expression of these downstream markers and reversed the glycolytic defects to levels comparable to control cells (Supplementary Fig. S6F–J).

Fig. 6.

VAV2 Promotes Glycolysis and Tumor Growth via EIF3F In Vitro and Vivo. (A) Western blot analysis of VAV2, EIF3F, MTA1, HIF-1α, GLUT1, and LDHA protein levels in KYSE150 cells transfected with the indicated combinations of OE-VAV2 and sh-EIF3F or controls. (B) Lactate production in KYSE150 cells with the indicated transfections. (C) Glucose uptake levels in KYSE150 cells with the indicated transfections. (D) Extracellular acidification rate (ECAR) in KYSE150 cells with the indicated transfections. (E) Oxygen consumption rate (OCR) in KYSE150 cells with the indicated transfections. (F) Representative images of xenograft tumors from nude mice injected with KYSE150 cells stably transfected with the indicated constructs. (G) Tumor volume growth curves over time for each experimental group. (H) Tumor weights at endpoint for each experimental group. (I) Western blot analysis of VAV2, EIF3F, MTA1, HIF-1α, GLUT1, and LDHA protein levels in xenograft tumor tissues from each group. (J) Immunohistochemistry (IHC) staining of VAV2, EIF3F, MTA1, and Ki67 in xenograft tumor tissues from each group. Scale bar,100 μm (for 200×) and 50 μm (for 400×). Each experimental group contained 6 mice (n = 6 per group). Data are presented as mean ± SD of at least three independent experiments. For (B-E, G-H), two-way ANOVA was used for tumor growth curve analysis. Significant differences were only observed for Glycolytic Capacity and Non-mitochondrial; no significant differences were found for Basal Glycolysis, Glycolysis, ATP Production, or Maximal Respiration among groups. The p-values are denoted by asterisks as follows: * p < 0.05; ** p < 0.01; *** p < 0.001

In vivo, we established xenograft models using KYSE150 cells with indicated transfections. Tumors derived from VAV2-overexpressing cells exhibited increased growth rates (tumor volume increased by approximately 1.6-fold at day 28), larger volumes, and greater weights compared to controls, which were partially reversed by EIF3F knockdown (tumor volume and weight reduced by approximately 43%) (Fig. 6F-H). Western blot analysis of tumor tissues confirmed that EIF3F knockdown attenuated VAV2 overexpression-induced upregulation of MTA1, HIF-1α, GLUT1, and LDHA (Fig. 6I). Immunohistochemical staining further demonstrated that EIF3F knockdown reduced VAV2 overexpression-induced expression of VAV2, EIF3F, MTA1, and the proliferation marker Ki67 (Fig. 6J). These results collectively demonstrate that VAV2 promotes glycolysis and tumor growth through EIF3F-mediated MTA1 deubiquitination.

Discussion

Reprogramming of energy metabolism has been increasingly recognized as a critical driver of tumor progression, yet the molecular mechanisms underlying this process in ESCC remain incompletely understood. In the present study, we identified a previously unrecognized role of VAV2 in promoting glycolytic reprogramming and tumor growth through the EIF3F-MTA1-HIF-1α axis. Our findings demonstrate that VAV2 is significantly upregulated in ESCC tissues and correlates with enhanced glycolysis. Mechanistically, VAV2 interacts with EIF3F and promotes its expression, which in turn binds to MTA1 and facilitates its deubiquitination, leading to MTA1 stabilization. Stabilized MTA1 subsequently activates HIF-1α signaling and its downstream glycolytic targets GLUT1 and LDHA. These findings were validated through rescue experiments both in vitro and in vivo, demonstrating that EIF3F knockdown largely abrogates VAV2 overexpression-induced oncogenic effects. These findings reveal a novel VAV2-EIF3F-MTA1 regulatory axis that links ubiquitination control to metabolic reprogramming in ESCC.

VAV2, as a guanine nucleotide exchange factor, has been primarily studied in the context of cytoskeletal dynamics and cell migration through activation of Rho GTPases [21]. Accumulating evidence has implicated VAV2 in tumor progression, with its overexpression reported in breast cancer, prostate cancer, and hnSCC [12, 14, 15]. However, its potential involvement in metabolic regulation has not been previously explored. Our study uncovers a non-canonical function of VAV2 in promoting aerobic glycolysis, expanding the understanding of how this signaling hub contributes to tumorigenesis beyond its classical GEF activity. Rather than functioning solely through Rho GTPase activation, VAV2 operates as a molecular scaffold that physically recruits EIF3F and enhances its expression, thereby coupling upstream signaling events to downstream metabolic reprogramming. The VAV2-EIF3F interaction occurs through a specific binding interface that does not compete with EIF3F’s simultaneous engagement of MTA1, enabling the formation of a functional VAV2-EIF3F-MTA1 ternary complex. These context-dependent mechanisms suggest that VAV2 may exert pro-tumorigenic functions through multiple modalities depending on cellular milieu and interacting partners. Additionally, our prior study established that VAV2 facilitates DNA repair and confers radioresistance through Ku70/Ku80 interaction [16]. The metabolic function uncovered here thus complements its known role in DNA repair, positioning VAV2 as an integrator of metabolic adaptation and genomic stability in ESCC.

MTA1 is well-established as a master regulator of cancer metastasis and progression, functioning primarily as a component of the NuRD complex to modulate gene expression [18, 22]. Recent studies have revealed that MTA1 can stabilize HIF-1α under normoxic conditions, thereby promoting glycolytic metabolism [23–25]. However, the regulatory mechanisms controlling MTA1 stability itself have remained elusive. Here, we demonstrate that the VAV2-EIF3F-MTA1 axis governs MTA1 protein stability through ubiquitin-proteasome pathway inhibition. This mechanism explains how MTA1 is frequently upregulated in cancers without corresponding transcriptional changes and identifies EIF3F as a previously uncharacterized DUB responsible for MTA1 deubiquitination, expanding the substrate repertoire of this eIF3 subunit.

EIF3F is conventionally known as a subunit of the eIF3 translation initiation complex, with its canonical function in protein synthesis [26]. However, emerging evidence indicates that individual eIF3 subunits exert non-canonical functions beyond translation regulation [19, 20, 27]. Notably, EIF3F contains an intrinsic JAMM/MPN domain with deubiquitinase activity [19, 20]. Our study uncovers a novel function of EIF3F as a critical link between VAV2 and MTA1 stability, expanding its known deubiquitinase activity to the regulation of metabolic reprogramming in ESCC. Through molecular docking and co-immunoprecipitation assays, we demonstrated that EIF3F directly interacts with both VAV2 and MTA1, and that these interactions occur at distinct domains on EIF3F, enabling EIF3F to directly deubiquitinate MTA1. We further validated this by generating a DUB-dead mutant of EIF3F (EIF3F-ΔMPN), which failed to rescue MTA1 stability in VAV2-knockdown cells, confirming that the deubiquitinase activity (not translation initiation) is responsible for MTA1 regulation. Mechanistically, VAV2 regulates this axis through two synergistic modes: direct scaffolding by binding EIF3F and facilitating its interaction with MTA1, and expression amplification by upregulating EIF3F protein levels to increase the available DUB pool for MTA1 stabilization. These dual mechanisms collectively ensure robust deubiquitination of MTA1 and subsequent activation of HIF-1α-driven glycolysis. Importantly, EIF3F overexpression reduced MTA1 ubiquitination, and its knockdown attenuated VAV2-induced deubiquitination of MTA1. These findings establish EIF3F as a critical link between VAV2 signaling and MTA1 stability, revealing a previously unrecognized role of this translation initiation factor in regulating protein stability.

The HIF-1α pathway plays a central role in metabolic reprogramming by transcriptionally activating glycolytic genes [28–31]. Our results indicate that MTA1 stabilization by the VAV2-EIF3F axis converges on HIF-1α activation, consistent with previous reports [32–34], but uniquely define the upstream ubiquitination-controlled mechanism governing this process. The functional consequences of this metabolic shift were substantiated by complementary assays measuring glucose uptake, lactate production, ECAR, and OCR, together with xenograft validation.

Several limitations of this study should be acknowledged. First, while we demonstrated that VAV2 positively regulates EIF3F protein expression, the underlying mechanism, whether through transcriptional activation, mRNA stabilization, or translational enhancement, remains to be elucidated. Second, although our cellular and in vivo data strongly support that EIF3F stabilizes MTA1 through deubiquitination, we did not perform in vitro reconstitution assays with purified proteins. Given that EIF3F is a JAMM-domain DUB that may require cofactors or the intact eIF3 complex for optimal substrate recognition and catalytic activity [35], future structural and biochemical studies are warranted to elucidate the precise catalytic mechanism by which EIF3F cleaves ubiquitin chains from MTA1. Third, whether the VAV2-EIF3F-MTA1 axis operates in other cancer types beyond ESCC warrants further investigation. Additionally, the clinical significance of this axis, particularly its potential as a prognostic biomarker or therapeutic target, needs to be evaluated in larger patient cohorts. While the subcutaneous xenograft model provided robust evidence for the oncogenic function of this axis in glycolytic reprogramming, we acknowledge that this model does not fully recapitulate the organ-specific tumor microenvironment of ESCC. Future studies employing orthotopic or patient-derived xenograft models will be essential to validate the therapeutic potential of targeting this axis in a more clinically relevant context.

Conclusion

In conclusion, our study reveals a novel mechanism by which VAV2 promotes glycolytic reprogramming and tumor growth in ESCC through the VAV2-EIF3F-MTA1 axis. This regulatory circuit links protein ubiquitination control to metabolic adaptation via HIF-1α activation, representing a previously unrecognized targetable vulnerability in ESCC. Targeting components of this axis, particularly the VAV2-EIF3F interaction, may offer new strategies for treating ESCC patients with aberrant glycolysis.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

CHX

Cycloheximide

Co-IP

Co-Immunoprecipitation

DEGs

Differentially Expressed Genes

DIA

Data-Independent Acquisition

EAC

Esophageal Adenocarcinoma

ECAR

Extracellular Acidification Rate

EIF3F

Eukaryotic Translation Initiation Factor 3 Subunit F

ESCA

Esophageal Cancer

ESCC

Esophageal Squamous Cell Carcinoma

GEO

Gene Expression Omnibus

GLUT1

Glucose Transporter 1

GOBP

Gene Ontology Biological Process

GOMF

Gene Ontology Molecular Function

GSEA

Gene Set Enrichment Analysis

HIF−1α

Hypoxia-Inducible Factor 1α

hnSCC

Head and Neck Squamous Cell Carcinoma

IHC

Immunohistochemistry

IP-MS

Immunoprecipitation-Mass Spectrometry

LDHA

Lactate Dehydrogenase A

MTA1

Metastasis-Associated Protein 1

NuRD

Nucleosome Remodeling and Deacetylase

OCR

Oxygen Consumption Rate

shRNA

Short Hairpin RNA

ssGSEA

Single-sample Gene Set Enrichment Analysis

TCGA

The Cancer Genome Atlas

UMAP

Uniform Manifold Approximation and Projection

VAV2

Vav Guanine Nucleotide Exchange Factor 2

Author contributions

Weiling Liu: Conceptualization, Methodology, Investigation, Formal analysis, Writing-original draft, Funding acquisition. Yangyang Hou: Methodology, Investigation, Formal analysis, Writing-original draft. Bo Wang: Methodology, Investigation, Validation. Fengna Liu: Investigation, Data curation. Lu Zheng: Investigation, Validation. Hailing Wang: Investigation, Data curation. Shaomei Li: Resources, Investigation. Xiaowan Zhou: Methodology, Formal analysis. Mengting Zhang: Methodology, Software. Shujun Yang: Conceptualization, Supervision, Writing review & editing, Project administration. Yan Zhao: Conceptualization, Supervision, Writing-review & editing, Project administration, Funding acquisition. All authors read and approved the final manuscript.

Funding

This work was supported by the National Natural Science Foundation of China (Grant No. 82303958), the Henan Province Clinical Medical Scientist Program (Grant No. HNCMS202407), the Henan Provincial Medical Science and Technology Tackling Key Problems Program Project (Grant No. SBGJ202302017 and SBGJ202402020).

Data availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Declarations

Competing interests

The authors declare no competing interests.

Ethics approval

All animal experiments were approved by the Animal Care and Use Committee of Hubei Provincial Center for Disease Control and Prevention (Approval No. 202610059) and conducted in accordance with the guidelines for the care and use of laboratory animals.This study did not involve multicentre research or independent review board approval.