UBE2A expression across human cancer types was analyzed using RNA-sequencing (RNA-seq) data obtained from The Cancer Genome Atlas (TCGA; http://portal.gdc.cancer/gov/). Transcript per million (TPM)-format expression matrices processed using the STAR pipeline were downloaded and normalized by log₂ (TPM+1) transformation. Differential expression analyses between tumor and normal tissues were performed in R software (version 4.2.1; Posit Software, PBC) using the Wilcoxon rank-sum test and data visualization was conducted using the ‘ggplot2’ package (version 3.4.4) (23). For tumor types lacking sufficient normal tissue controls in TCGA, the pan-cancer differential expression module of SangerBox (http://sangerbox.com/) was additionally used, which integrates TCGA and Genotype-Tissue Expression (GTEx; http://www.gtexportal.org/home/) datasets (24). To validate the expression pattern of UBE2A, two independent breast cancer cohorts [GSE29044 (25) and GSE65194 (26)] were downloaded from the Gene Expression Omnibus (GEO; http://www.ncbi.nlm.nih.gov/geo/) and differential expression analyses were performed between tumor and normal tissues using the same statistical methods. In addition, the associations between UBE2A expression and immune or molecular subtypes across cancers were explored using the Tumor-Immune System Interaction Database (http://cis.hku.hk/TISIDB/) (27).

To investigate the association between UBE2A expression and clinicopathological characteristics across cancers, normalized pan-cancer transcriptomic data (TCGA Pan-Cancer; n=10,535) were downloaded from the UCSC Xena database (https://xenabrowser.net/). UBE2A expression data were extracted from the dataset and samples derived from primary tumors were included for further analysis. Expression values were transformed using log₂ (x + 0.001) normalization prior to statistical analysis. Differences in UBE2A expression among different clinicopathological stages, including T stage, N stage and pathological stage, were analyzed using R software. Unpaired Student's t-tests were used for pairwise comparisons, while one-way ANOVAs were performed for comparisons among multiple groups.

Diagnostic performance of UBE2A expression in distinguishing tumor tissues from normal tissues was evaluated using receiver operating characteristic (ROC) curve analysis. RNA-seq expression data in TPM format and corresponding clinical information were downloaded from TCGA. Gene expression values were transformed using log₂ (TPM + 1) normalization prior to analysis. ROC analyses were performed in R software using the ‘pROC’ package (version 1.18.0) (28) and visualization was generated using the ‘ggplot2’ package (version 3.4.4) (23). Tumor samples were defined as the positive group, whereas normal tissues served as the reference group for binary classification analysis. The ‘pROC’ package automatically determined the direction of comparison to ensure an upward ROC curve orientation. The area under the ROC curve (AUC), sensitivity, specificity and optimal cut-off values were calculated to assess diagnostic efficacy. The optimal cut-off threshold was determined based on the maximum Youden index. An AUC value closer to 1.0 indicated superior discriminative performance. Pan-cancer analyses were performed across TCGA datasets and cancer types with significant diagnostic performance are presented in the results.

Prognostic value of UBE2A was evaluated using the GEPIA2 (http://gepia2.cancer-pku.cn/#analysis) and Kaplan-Meier Plotter (http://kmplot.com/analysis/) databases. GEPIA2, which integrates transcriptomic and clinical data from TCGA, was used to assess the associations between UBE2A expression and overall survival (OS) as well as disease-free survival (DFS) using the log-rank test (29). Kaplan-Meier Plotter, which incorporates survival data from TCGA and GEO cohorts, was used to further validate the prognostic importance of UBE2A in breast cancer, including OS, distant metastasis-free survival (DMFS) and relapse-free survival (RFS). Patients were stratified into high- and low-expression groups using the default cut-off settings of each platform. Hazard ratios (HRs), 95% CIs and log-rank P-values were used to evaluate prognostic importance (30).

Genetic alteration data for UBE2A were obtained from the cBioPortal platform, which provides access to TCGA-based cancer genomics datasets (31). Alteration frequency, mutation type, mutation distribution and copy number variation of UBE2A were assessed across multiple cancer types. In addition, the associations between UBE2A genetic alterations and clinical outcomes were evaluated.

Genes co-expressed with UBE2A were identified using the LinkedOmics database, which contains multi-omics data from TCGA cohorts (32). Pearson correlation analysis was performed to determine genes significantly associated with UBE2A expression. The top 500 positively and negatively correlated genes were subsequently subjected to Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses using the ‘clusterProfiler’ R package (version 4.14.6) (33–35). In addition, GSEA was performed in the TCGA-BRCA cohort by comparing samples with high and low UBE2A expression (median cut-off). Genes were ranked according to differential expression and significantly enriched pathways were visualized using the ‘enrichplot’ package (version 1.26.6) (34).

A PPI network centered on UBE2A was constructed using the GeneMANIA platform (http://genemania.org) (36). Genes functionally associated with UBE2A were identified based on numerous evidence sources, including co-expression and physical interactions, to explore the potential molecular functions of UBE2A.

Epigenetic modifications, including DNA methylation and RNA methylation, serve key roles in cancer development and progression (37,38). Promoter DNA methylation levels of UBE2A across 33 TCGA cancer types were evaluated using TCGA HumanMethylation450 BeadChip data accessed through the UALCAN platform (http://ualcan.path.uab.edu) (39). In addition, correlations between UBE2A expression and RNA modification-associated regulators, including genes involved in m6A, m5C and m1A modification pathways, were analyzed to explore potential epigenetic regulatory mechanisms.

Immune cell infiltration levels were estimated using the Cell-type Identification by Estimating Relative Subsets of RNA Transcripts algorithm (40) based on TCGA transcriptomic data. The relative proportions of 22 immune cell subsets were quantified and Spearman's correlation analysis was performed to evaluate the associations between UBE2A expression and immune cell infiltration levels across cancer types. The results are visualized as heatmaps. To improve the robustness of immune infiltration analyses, multiple immune deconvolution algorithms available in the SangerBox platform (http://www.sangerbox.com) were applied, including Tumor Immune Estimation Resource (TIMER) (41), Estimating the Proportions of Immune and Cancer cells (EPIC) (42), Quantification of the Tumor Immune Contexture from Human RNA-seq Data (quanTIseq) (43) and xCell (44). Correlations between UBE2A expression and immune cell infiltration were evaluated using these algorithms.

TCGA transcriptomic data, TMB data and MSI data were integrated for pan-cancer analysis. Spearman's correlation analysis was performed to evaluate the associations between UBE2A expression and TMB or MSI across different cancer types. Correlation coefficients and corresponding P-values were calculated using R software and the results are visualized as radar plots.

Single-cell transcriptomic analysis was performed using the Tumor Immune Single-Cell Hub 2 (TISCH2) database, a comprehensive single-cell RNA-seq (scRNA-seq) resource for tumor microenvironment analysis (45). TISCH2 provides uniformly processed scRNA-seq datasets with standardized quality control, clustering, batch-effect correction and hierarchical cell-type annotation using the Model-based Analyses of Single-cell Transcriptome and Regulome pipeline (45). In the present study, four breast cancer datasets, including BRCA_GSE114727 (46), BRCA_GSE110686 (47), BRCA_GSE148673 (48) and BRCA_GSE176078 (49), were analyzed to investigate the single-cell expression pattern of UBE2A in different cell populations within the tumor microenvironment. Cell annotations curated by TISCH2 were used at the malignancy, major-lineage and minor-lineage levels to ensure consistency and comparability across datasets. Gene expression distributions and cell-type-specific expression patterns were directly obtained and visualized through the TISCH2 platform.

MCF-10A, the human mammary epithelial cell line and breast cancer cell lines MCF-7, MDA-MB-231, MDA-MB-468 and Hs578T were purchased from Procell Life Science & Technology Co., Ltd. MCF-10A cells were cultured in DMEM/F12 medium, whereas breast cancer cell lines were maintained in DMEM medium. All media were supplemented with 10% FBS (Biological Industries; Sartorius AG) and 1% penicillin-streptomycin (Beijing Solarbio Science & Technology Co., Ltd.). Cells were cultured at 37°C in a humidified incubator containing 5% CO₂.

Total RNA was extracted from cultured breast epithelial and breast cancer cells lines, including MCF-10A, MCF-7, MDA-MB-231, MDA-MB-468 and Hs578T, using the Cell/Tissue Total RNA Kit (Suzhou Xinsaimei Biotechnology Co., Ltd.) following the manufacturer's protocol. RT was performed using HiScript III RT SuperMix (Vazyme Biotech Co., Ltd.) according to the manufacturer's protocol. RT-qPCR was carried out on the qTOWER3G system (Analytik Jena GmbH + Co. KG) with ChamQ Universal SYBR qPCR Master Mix (Vazyme Biotech Co., Ltd.) under the following thermocycling conditions: Initial denaturation at 95°C for 30 sec, followed by 40 cycles of denaturation at 95°C for 10 sec and annealing/extension at 60°C for 30 sec. β-actin was used as the internal reference gene for normalization and relative expression levels were calculated using the 2^−ΔΔCq method (50). The primer sequences are listed in Table SI.

Targeted silencing of UBE2A was achieved using specific siRNA and corresponding scrambled negative control siRNA, all synthesized by Hanbio Biotechnology Co., Ltd. For siRNA transfection, MCF-7 and MDA-MB-231 cells in 6-well plates were transfected with 100 pmol siRNA per well using Lipofectamine™ 3000 (Invitrogen; Thermo Fisher Scientific, Inc.) according to the manufacturer's protocol. Following 48 h of transfection, knockdown efficiency was validated at both mRNA and protein levels through qPCR and western blotting, respectively. The siRNA sequences are detailed in Table SII.

MCF-7 and MDA-MB-231 cells were seeded in 96-well plates at a density of 3,000 cells/well (100 µl medium/well) with 5 technical replicates per group. After cell attachment (0 h), the medium was replaced with 100 µl fresh complete medium containing 10% CCK-8 reagent (Vazyme Biotech Co., Ltd.) at 0, 24, 48 and 72 h, followed by incubation under standard culture conditions for 2 h. Absorbance was measured at 450 nm using a microplate reader (BioTek; Agilent Technologies, Inc.). Cell viability was calculated and growth curves were plotted based on optical density values.

To evaluate cell migration capacity, pre-marked reference lines were drawn on the bottom of 6-well plates prior to cell seeding. MCF-7 and MDA-MB-231 cells were cultured in complete medium until reaching 95–100% confluence. A wound was created by scratching the monolayer with a 200 µl sterile pipette tip perpendicular to the reference lines. After washing with PBS to remove detached cells, fresh serum-free medium was added. Wound areas were photographed at 0 and 24 h post-scratching using an inverted phase-contrast microscope (Olympus-IX71; Olympus Corporation). ImageJ (National Institutes of Health; version 1.53a) was used to calculate the scratch area.

For migration assays, MCF-7 and MDA-MB-231 cells (5×10⁴ cells per well) were suspended in 100 µl serum-free medium and seeded into the upper chamber of Transwell inserts (Corning, Inc.). The lower chamber was filled with 800 µl complete medium containing 10% FBS (Biological Industries; Sartorius AG) as a chemoattractant. For invasion assays, transwell inserts were pre-coated with Matrigel (Shanghai Yeasen Biotechnology Co., Ltd.) and incubated at 37°C for 2 h before cell seeding. Cells were then incubated at 37°C in a humidified atmosphere containing 5% CO₂ for 24 h. After incubation, non-migrated cells were removed from the upper surface of the membrane using cotton swabs. Cells that had migrated or invaded to the lower surface were fixed with fixed with 4% paraformaldehyde (Beijing Solarbio Science & Technology Co., Ltd.) at room temperature for 20 min and stained with 0.1% crystal violet (Beyotime Biotechnology) at room temperature for 20 min. Images were captured using an inverted microscope (Olympus IX73; Olympus Corporation) and cell numbers were quantified using ImageJ software.

Total cellular proteins were extracted from MCF-7 and MDA-MB-231 cells using a column-based total protein extraction kit (Suzhou Xinsaimei Biotechnology Co., Ltd.) and protein concentrations were determined by a BCA assay (Beytome Biotechnology). Equal quantities of protein (30 µg per lane) were separated by 10% SDS-PAGE and subsequently transferred onto PVDF membranes (MilliporeSigma; Merck KGaA). Following transfer, membranes were blocked with 5% non-fat milk at room temperature for 1 h and then incubated with the indicated primary antibodies at 4°C overnight. After washing, membranes were probed with HRP-conjugated secondary antibodies for 1 h at room temperature. Immunoreactive signals were detected using an ECL detection system (Thermo Fisher Scientific, Inc.) and captured with an imaging system from e-BLOT Life Science (Shanghai) Co., Ltd. The antibodies used included anti-UBE2A (1:1,000; cat. no. sc-365507; Santa Cruz Biotechnology, Inc.), anti-β-actin (1:20,000; cat. no. 66009-1-Ig; Proteintech Group, Inc.), HRP-conjugated goat anti-rabbit IgG (1:10,000; cat. no. SA00001-2; Proteintech Group, Inc.) and HRP-conjugated goat anti-mouse IgG (1:10,000 cat. no. SA00001-1; Proteintech Group, Inc.).

Statistical analyses were performed using GraphPad Prism (version 10.1.2; Dotmatics) and R software. Differences between groups were evaluated using unpaired Student's t-test or one-way ANOVA, as appropriate. Unless otherwise specified, two-sided P<0.05 was considered to indicate a statistically significant difference. For bioinformatic analyses, statistical testing and multiple-testing correction were performed according to the corresponding analytical pipelines and false discovery rate-adjusted P-values were applied in enrichment analyses.

To characterize the expression profile of UBE2A across human cancers, TCGA transcriptomic datasets were analysed and UBE2A expression was compared between tumor tissues and corresponding normal tissues. As shown in Fig. 2A, UBE2A expression was found to be significantly elevated in bladder urothelial carcinoma (BLCA), BRCA, cervical squamous cell carcinoma and endocervical adenocarcinoma (CESC), cholangiocarcinoma (CHOL), esophageal carcinoma (ESCA), glioblastoma multiforme (GBM), head and neck squamous cell carcinoma (HNSC), liver hepatocellular carcinoma (LIHC), lung adenocarcinoma (LUAD), lung squamous cell carcinoma (LUSC), stomach adenocarcinoma (STAD) and thyroid carcinoma (THCA) compared with corresponding normal tissues. By contrast, kidney chromophobe (KICH) exhibited significantly lower UBE2A expression. To further validate these findings, an integrated analysis combining TCGA tumor samples and GTEx normal tissues was performed. Consistent with the TCGA-based analysis, UBE2A remained significantly upregulated in BLCA, BRCA, CESC, CHOL, ESCA, GBM, HNSC, LIHC, LUAD, LUSC, STAD and THCA. In addition, further cancer types with significant dysregulation were identified after incorporating GTEx normal tissues, resulting in elevated UBE2A expression in 25 cancer types and reduced expression in five cancer types overall (Fig. 2B). In addition, paired tumor-normal analyses within TCGA cohorts further demonstrated significantly elevated UBE2A expression in BLCA, BRCA, CHOL, ESCA, HNSC, LIHC, LUAD, LUSC, STAD and THCA, while significantly reduced expression was observed in KICH (Fig. 2C). By contrast, no significant differences were detected in CESC, COAD, KIRC, KIRP, PAAD, PCPG, PRAD, READ or UCEC.

To further assess the robustness of these observations, two independent breast cancer cohorts from the GEO database (GSE29044 and GSE65194) were analyzed. Consistent with the TCGA results, UBE2A expression was found to be significantly increased in breast cancer tissues compared with normal breast tissues in both datasets (Fig. 2D and E). Collectively, these findings demonstrate that UBE2A is aberrantly upregulated in a broad spectrum of human cancer types, particularly in breast cancer, suggesting that UBE2A may serve an important role in tumor development and progression.

To investigate the clinical relevance of UBE2A, its association with clinicopathological characteristics was evaluated across numerous cancer types. Analysis of tumor invasion status revealed significant differences in UBE2A expression among T-stage subgroups in ESCA, STES, KIRP, pan-kidney cohort, PRAD and LIHC (Fig. 3A). With regard to lymph node metastasis, UBE2A expression was found to be significantly associated with N stage in BRCA, STES, KIRP, PRAD and HNSC (Fig. 3B). Furthermore, significant correlations between UBE2A expression and overall pathological stage were observed in KIRP and LIHC (Fig. 3C). These findings suggest that UBE2A expression is associated with tumor progression in numerous malignancies. Notably, the significant association between UBE2A expression and N stage in BRCA indicates a potential role for UBE2A in breast cancer progression and metastasis.

Diagnostic values of UBE2A expression were evaluated across a number of cancer types using ROC curve analysis. UBE2A exhibited good discriminatory ability between tumor and normal tissues in 15 cancer types (Fig. 4A-O). Notably, excellent diagnostic performance was observed in CHOL, esophageal adenocarcinoma (ESAD), ESCA, GBM, KICH and LIHC, with AUC values exceeding 0.90. These findings indicate that UBE2A expression may serve as a potential diagnostic biomarker in selected malignancies.

To evaluate the prognostic relevance of UBE2A across cancers, survival analyses were performed using OS and DFS endpoints. Elevated UBE2A expression was significantly associated with shorter OS in BRCA, low-grade glioma (LGG), LIHC, mesothelioma, sarcoma (SARC) and uveal melanoma (UVM; Fig. 5A-G). Similarly, patients with high UBE2A expression exhibited significantly worse DFS in LGG, LIHC and UVM (Fig. 5H-K). To further validate these findings, independent BRCA cohorts were analyzed using the Kaplan-Meier Plotter platform. Consistent with the pan-cancer findings, elevated UBE2A expression was associated with unfavorable OS, RFS and DMFS (Fig. 5L-N). Collectively, these results suggest that UBE2A may represent a prognostic indicator in a number of malignancies, with reproducible associations observed across different survival endpoints and analytical platforms.

Genomic alteration landscapes of UBE2A were next investigated across human cancers. UBE2A alterations were detected in ~1.1% of all analyzed tumors, although notable heterogeneity was observed among cancer types (Fig. 6A). The highest alteration frequency occurred in mature B-cell neoplasms (>10%), followed by endometrial carcinoma and non-small cell lung cancer, each exhibiting alteration frequencies of ~2% (Fig. 6B). Analysis of mutation patterns revealed that missense mutations were the predominant alteration type, with recurrent mutations affecting the R140 residue identified as hotspot events (Fig. 6C). Copy number gain represented the most common copy number alteration (Fig. 6D). Furthermore, tumors harboring UBE2A alterations exhibited significantly higher mutation frequencies of cystatin SN, SLC51 subunit β, ATPase inhibitory factor 1 and a number of additional genes compared with UBE2A wild-type tumors (Fig. 6E), suggesting that UBE2A alterations may co-occur with specific genomic events during tumor evolution.

To gain insight into the biological functions associated with UBE2A, genes correlated with UBE2A expression were identified and analyzed. The top 50 positively and negatively correlated genes are visualized as heatmaps (Fig. 7A and B). Functional enrichment analyses based on the top 500 UBE2A-associated genes revealed significant enrichment in ‘chromosome segregation’, ‘mitotic cell cycle phase transition’ and other cell cycle-associated biological processes (Fig. 7C). Consistently, KEGG pathway analysis identified significant enrichment of ‘proteasome’ and ‘cell cycle’ pathways (Fig. 7D).

To further characterize biological pathways associated with UBE2A expression, GSEA was performed in BRCA. GO-based GSEA identified enrichment of ‘GOMG_OLFACTORY_RECEPTOR_ACTIVITY’, amongst other terms, whereas KEGG-based GSEA demonstrated significant enrichment of ‘KEGG_ANTIGEN_PROCESSING_AND_PRESENTATION’ as well as ‘KEGG_CYTOSOLIC_DNA_SENSING_PATHWAY’ in the UBE2A high-expression group (Fig. 7E). In addition, a PPI network centered on UBE2A revealed numerous closely connected interacting proteins and functional modules (Fig. 7F), supporting the potential involvement of UBE2A in cell cycle regulation and proteostasis-associated processes.

To investigate potential epigenetic mechanisms underlying UBE2A dysregulation, promoter methylation patterns were analyzed across multiple cancer types. Compared with normal tissues, a significantly lower promoter methylation levels of UBE2A was observed in BLCA, LIHC, BRCA and STAD (Fig. 8A-D). By contrast, significantly increased promoter methylation was detected in PRAD (Fig. 8E). The association between UBE2A expression and RNA modification regulators was further examined. UBE2A expression exhibited significant correlations with numerous RNA modification-associated genes across cancer types, particularly a number m6A regulators, including methyltransferase-like 14 and Wilms tumor 1-associating protein (Fig. 8F). Collectively, these findings suggest that epigenetic mechanisms may contribute to the dysregulated expression of UBE2A in human cancers.

Whether UBE2A expression differed among distinct molecular and immune subtypes across cancers was next examined. Significant differences in UBE2A expression among molecular subtypes were observed in BRCA, COAD, HNSC, KIRP, LGG, LUSC, ovarian cancer (OV) and pheochromocytoma and paraganglioma (Fig. 9A-H). Similarly, UBE2A expression varied markedly among immune subtypes in 12 cancer types (Fig. 9I-T). These observations indicated that UBE2A expression may be associated with tumor heterogeneity at both molecular and immune levels.

To explore potential associations between UBE2A expression and the tumor immune microenvironment, correlations between UBE2A expression and immune cell infiltration were evaluated across cancers. UBE2A expression exhibited significant correlations with numerous immune cell populations in different tumor types, including resting memory CD4⁺ T cells, activated memory CD4⁺ T cells, activated NK cells and M2 macrophages (Fig. 10A). Similar trends were observed when additional immune deconvolution algorithms, including EPIC, QUANTISEQ, TIMER and xCELL, were applied (Fig. S1A-D). The associations between UBE2A expression and immune regulatory genes were further examined. UBE2A expression exhibited widespread correlations with numerous immunomodulatory molecules, including C-X-C motif chemokine ligand 10 and IL-10, across cancer types (Fig. S2). In addition, significant associations were observed between UBE2A expression and a number of immune checkpoint-associated genes, including IL-12A and TGF-β1 (Fig. S3).

Given the clinical relevance of genomic instability in immunotherapy, correlations between UBE2A expression and TMB or MSI were subsequently analyzed (51,52). UBE2A expression was found to be positively associated with TMB in BLCA, BRCA, LGG, OV, PRAD, SARC and STAD, whereas negative correlations were observed in COAD, KIRP and thymoma (Fig. 10B). For MSI, positive correlations were detected in adrenocortical carcinoma, HNSC, PRAD, SARC, STAD, tenosynovial giant cell tumor and THCA, while negative associations were observed in COAD, diffuse large B-cell lymphoma, LUAD, LUSC and OV (Fig. 10C). Collectively, these findings suggest that UBE2A expression may be associated with distinct immune and genomic features in a cancer type-dependent manner.

Single-cell omics technologies enable high-resolution profiling of tumor heterogeneity by capturing cell-to-cell variability that is masked by traditional bulk sequencing, thereby advancing precision oncology and improving our understanding of tumor biology (53,54). To further characterize the cellular distribution of UBE2A within the tumor microenvironment, single-cell RNA-seq datasets from four independent BRCA cohorts were analyzed. Across all datasets, UBE2A expression was found to be predominantly enriched in the proliferative T-cell population (Fig. 11A-E). Consistent expression patterns across independent datasets indicated a potential association between UBE2A expression and proliferative T-cell states within the breast tumor microenvironment. Notably, these observations were derived from transcriptomic datasets and therefore should be considered hypothesis-generating findings that require further experimental validation.

Based on the pan-cancer analysis results, breast cancer was selected for functional validation as UBE2A exhibited significant upregulation and prognostic relevance in BRCA. In addition, the present bioinformatic analyses suggested a potential involvement of UBE2A in tumor progression and tumor microenvironment regulation in breast cancer. Therefore, the biological functions of UBE2A in breast cancer cell lines were further investigated. qPCR analysis demonstrated differential UBE2A expression among breast cancer cell lines. Compared with the normal breast epithelial cell line MCF-10A, UBE2A expression was significantly elevated in MCF-7, MDA-MB-231 and MDA-MB-468 cells (Fig. 12A). Based on their relatively high expression levels, MCF-7 and MDA-MB-231 cells were selected for subsequent functional experiments.

A total of three siRNAs targeting UBE2A were designed and transfected into MCF-7 and MDA-MB-231 cells. qPCR and western blotting analyses demonstrated that all three siRNAs significantly reduced UBE2A expression (Fig. 12B-G), with siUBE2A#1 and siUBE2A#2 exhibiting the strongest knockdown efficiency; thus, these two sequences were chosen for subsequent experiments. CCK-8 assays revealed that UBE2A silencing significantly inhibited cell proliferation in both lines, compared with the negative control (Fig. 12H and I). Wound healing assays further indicated a significant reduction in migration capacity upon UBE2A knockdown, compared with the negative control (Fig. 13A-D). Lastly, Transwell assays demonstrated that UBE2A knockdown significantly impaired both migration and invasion abilities in MCF-7 and MDA-MB-231 cells (Fig. 13E-J).