Recombinant human interleukin-6 (IL-6) was purchased from R&D and was reconstituted in Dulbecco's phosphate-buffered saline. IL-6 was used at concentration of 10 ng/ml in cell culture experiments as described in later section. A Janus protein tyrosine kinase (JAK) inhibitor, ruxolitinib (23,24), was obtained from Seleck and used at the concentration of 1 µM.

Human OSCC cell line; SAS was obtained from RIKEN BioResource Center Cell Bank (RCB1974; RRID:CVCL1675). The SAS cells with fluorescent ubiquitination-based cell cycle indicators (Fucci), termed ‘SAS-Fucci’, were established as previously described (25) by introducing Fucci plasmids (CFII-EF mKO2-hCdt1 [30/120] and CFII-EF-mAG-hGeminin [1/110]), a kind gift from Dr. Atsushi Miyawaki (26). SAS-Fucci cells were maintained in Dulbecco's Modified Eagle's Medium (DMEM; Nacalai Tesque) supplemented with 10% fetal bovine serum (FBS; Sigma or Biowest), 100 units/ml penicillin, and 100 µg/ml streptomycin (Nacalai Tesque) in a humidified incubator at 37°C with 5% CO₂. 293FT (Thermo Fisher Scientific, R70007; RRID:CVCL6911) were maintained in DMEM supplemented with 10% FBS, 1% non-essential amino acid solution (Nacalai Tesque), 100 units/ml penicillin, and 100 µg/ml streptomycin in a humidified incubator at 37°C with 5% CO₂.

Depending on the experiment, cells were treated either with IL-6 or ruxolitinib and subjected to reverse transcription-quantitative PCR (RT-qPCR) or cell migration assay as described in later sections,

Gene Expression Profiling Interactive Analysis (GEPIA; http://gepia.cancer-pku.cn/index.html) (27), a web server for assessing RNA expression data from The Cancer Genome Atlas along with the ‘Genotype-Tissue Expression’ projects, was used to investigate the top 500 genes of enhanced expression in head and neck squamous cell carcinoma (HNSCC) database. Expressions of candidate genes in HNSCC tumors vs. their expressions in normal tissues were assessed using differential gene expression analysis with GEPIA web server. The significance was estimated with utility in GEPIA employing one-way ANOVA, using disease state (Tumor or Normal) as variable for calculating differential expression. The correlation between gene expression and stage progression of HNSCC was investigated using ‘Stage Plot’ utility within GEPIA. The pathological stage was considered as variable for calculating differential expression; ‘gene expression ~ pathological stage’. Significance of differential gene expression between stages was estimated using one-way ANOVA. For differential analysis the expression data presented in transcripts per million (TPM) values were log₂ transformed after adding 1 followed by log₂FC defined as median(Tumor)-median(Normal).

Survival of patients HNSCC was determined using KM-plotter (https://kmplot.com/analysis/) (28). Overall survival of patients was assessed for 500 candidate genes splitting patients at ‘auto best select cut-off’. This analysis was based on the RNA sequence dataset of 500 patients with HNSCC available in the KM-plotter. The P-value was computed with utility available in the KM-plotter using the Cox-Mantel (log rank) test.

The experimental procedures were approved by Genetically Modified Organisms Safety Committee of Institute of Science Tokyo (approval numbers: G2019-026C9 and G2024-007C5). Human UBE2T cDNA (GenBank Accession No. NM_001310326.2) was amplified using KOD Plus Neo DNA polymerase (TOYOBO) with gene specific primers. The forward primer contained a BamHI restriction site and Kozak sequence, while the reverse primer contained an XhoI restriction site. The sequences of the primers are as follows (UBE2T-specific sequences are underlined): forward primer: 5′-AAAGGATCCGCCACCATGCAGAGAGCTTCAC−3′ and reverse primer: 5′-AAACTCGAGAACATCAGGATGAAATTTCTTTTCTATG−3′. The amplified UBE2T cDNA was then digested with BamHI/XhoI and inserted into vector (Invitrogen) to construct pENTR201-UBE2T-FLAG vector. The Gateway Technology (Invitrogen) was then used to transfer UBE2T cDNA into the pCSII-EF-RfA (RIKEN, RDB04387), a gift from Dr. Hiroyuki Miyoshi, RIKEN (29) to generate lentiviral expression vector.

Lentivirus particles were produced by transfecting 293FT cells with the expression vectors (pCSII-EF-RfA-EMPTY, pCSII-EF-UBE2T) and packaging vectors (pCMV–VSV-G-RSV-Rev [RIKEN, RDB04393] and pCAG-HIVgp [RIKEN, RDB04394]) as previously reported (30) to produce empty lentivirus (negative control) and lentivirus expressing UBE2T. The media were refreshed 24 h post-transfection, and the supernatants containing viral particles were collected after 48 h. The viral particles were concentrated for 7 days at 4°C using a Lenti-X Concentrator (Takara). The concentrated viral particles were then used to infect the SAS-Fucci cells as described previously (31).

Cell lysates from the SAS-Fucci cells infected with pCSII-EF-RfA-EMPTY or pCSII-EF-UBE2T lentivirus were prepared using 50 mM Tris-HCl buffer pH 7.5, 150 mM NaCl, 1% NP-40 containing Protease Inhibitor Cocktail (Nacalai Tesque). The protein concentration in obtained lysates was determined by using BCA Protein Assay Kit (Thermo Fisher Scientific). Cell lysates (47 µg of total proteins/lane) were then separated on 10% SDS-PAGE, followed by transfer onto polyvinylidene fluoride membranes (Merck). The membranes were blocked with 5% skim milk (FUJIFILM Wako) for 30 min at room temperature and incubated with mouse monoclonal anti-FLAG M2 antibody (dilution 1:5,000, cat. no. F1804, Sigma-Aldrich) and rabbit polyclonal anti-α-tubulin (dilution 1:10,000, cat. no. ab4074, Abcam) overnight at 4°C. The membranes were then incubated with anti-mouse IgG HRP-linked antibody (goat, dilution 1:5,000, cat. no. 7074S, Cell Signaling Technology) or anti-rabbit IgG HRP-linked antibody (goat, dilution 1:5,000, cat. no. 7074S, Cell Signaling Technology) for 1 h at room temperature. Target proteins were detected with Chemiluminescence Kit (ECL detection reagent; Cytiva) and visualized with a Fusion Solo S Imaging System (Vilber).

The SAS-Fucci cells infected with pCSII-EF-RfA-EMPTY or pCSII-EF-UBE2T lentivirus were seeded onto the upper chamber of a 24-well culture insert with 8-µm pore filter (BD Bioscience) and allowed to migrate for 48 h. The filters were then stained with DiffQuik (Sysmex), and non-migrated cells were removed with a cotton swab. Migrated cells on the bottom side of the chamber were photographed using an All-in-One microscope (BZ-710, Keyence) and counted with ImageJ software (National Institute of Health) (32).

To examine the effect of IL-6 on OSCC cell migration, SAS-Fucci cells infected with pCSII-EF-RfA-EMPTY lentivirus were seeded on 6-well plate, and incubated for 18 h followed by treatment without or with IL-6 (final concentration: 10 ng/ml) for 72 h, collected and seeded into migration chamber with 8-µm pore filter (Corning) in 24-well plate. Cells were then allowed to migrate for 48 h. The concentration of IL-6 in medium was maintained during migration. Staining and image acquisition were done as described above.

To examine the effect of ruxolitinib on OSCC cell migration, SAS-Fucci cells infected with pCSII-EF-UBE2T lentivirus were seeded on 6-well plate, and incubated for 18 h followed by treatment without or with ruxolitinib (final concentration: 1 µM) for 72 h. The cells were then collected, seeded into migration chamber with 8-µm pore filter (Corning) in 24-well plate and allowed to migrate for 48 h. The concentration of ruxolitinib in the medium was maintained during migration. Staining and image acquisition were done as described above.

The brightness of representative images of migrated cells were adjusted with Adobe Photoshop 25 (Adobe) and used for figure preparation. The experiments were performed in triplicates and repeated twice.

Total RNA was extracted using Sepasol-RNA I Super G (Nacalai Tesque) or RNeasy Plus Mini Kit (Qiagen). The RNA was then reverse transcribed using ReverTraAce qPCR RT Master Mix (TOYOBO Life Science) according to the manufacturer's protocol. The RT-qPCR was performed with the Step One Plus Real-Time PCR System (Applied Biosystems) or QuantStudio 3 (Applied Biosystems) using gene-specific primers PowerUp SYBR Green Master Mix (Applied Biosystems). The relative-standard curve method was used to determine the relative expression of target genes (33). All expression data were normalized to the expression of β-actin. The genes and corresponding primer sequences are listed in Table SI. The experiments were performed in triplicates and repeated twice.

Total RNA from SAS-Fucci cells was prepared using an RNeasy Plus Mini kit (Qiagen). The quality of RNA was evaluated using an Agilent 4200 TapeStation (Agilent). A cDNA library was prepared using TruSeq Stranded mRNA Kit (Illumina). The quality of the library was then evaluated with KAPA Library Quantification Kit (Roche) followed by sequencing NextSeq 500/550 High Output Kit v2.5 (Illumina, 75 cycles pair-end, 40/40 cycles). The average read size of 58 million. FASTQ files were processed with bcl2fastq software (Illumina Inc.) and reads were aligned against the human genome (GRCh38/hg38) using CLC Genomics Workbench version 12.0.2 (Qiagen). Normalized gene expression data, the raw TPM values were applied to GeneString version 14.9.1 (Agilent). For data analysis, TPM values were log₂-transformed after adding 1, and z-scores were calculated.

Gene set enrichment analysis (GSEA) was performed using GSEA version 4.2.3 and previously curated gene sets (34,35). Gene Ontology (GO) analysis was performed using ‘Database for Annotation, Visualization, and Integrated Discovery’ (36). The raw RNA sequencing data have been deposited with links to BioProject accession number PRJDB20207 in the DNA Data Bank of Japan (DDBJ) BioProject database.

The protein-protein interaction network of differentially expressed genes (DEGs) was constructed using ‘Search Tool for the Retrieval of Interacting Genes/Proteins’ (STRING; http://string-db.org) version 12.0 (37). The interaction was considered statistically significant at an interaction score >0.4.

Significant differences between means were determined by two-tailed unpaired Student t-test using Prism software version 10.3.1 (GraphPad). Values are presented as mean ± SD. P<0.05 was considered to indicate a statistically significant difference. The significances of gene sets identified by GSEA were based on normalized enrichment scores (NES) and nominal (NOM) P-values.

To identify factors critical for the progression of head and neck cancer, we performed an analysis using the GEPIA web server. We focused on the top 500 genes with elevated expression in head and neck cancer tissues compared with those in normal tissues and evaluated their prognostic significance using the KM-plotter database for patients with HNSCC. We identified 64 genes whose expression was significantly associated with a poor prognosis and had hazard ratios greater than 1.5 (Table SII). Further analysis of these 64 poor prognostic factors using the GEPIA Stage Plot revealed that UBE2T had the strongest correlation (F-value=5.02) with head and neck cancer stage progression (Table SIII). The results of this unbiased screening, focusing on UBE2T, are summarized in Fig. 1. UBE2T expression was significantly higher in HNSCC tumor tissues than in normal tissues (Fig. 1A). There was also a significant association between high UBE2T expression levels and poor prognosis in patients with HNSCC, as revealed by KM-plotter analysis (Fig. 1B), as well as the strongest correlation between UBE2T expression and the pathological stage of HNSCC (Fig. 1C). Therefore, we concluded that UBE2T is an unfavorable prognostic factor whose expression is elevated during the development and progression of head and neck cancer and can be considered a potential target for the treatment of HNSCC.

We next investigated the functional role of UBE2T in OSCC using the SAS-Fucci cell line. We established SAS-Fucci oral cancer cells expressing UBE2T (hereafter termed ‘UBE2T SAS cells’) (Fig. 2A) and performed chamber migration assay. UBE2T SAS cells exhibited significantly enhanced motility compared with SAS-Fucci oral cancer cells infected with empty lentivirus (hereafter termed ‘Control SAS cells’) (Fig. 2B and C), suggesting that UBE2T promotes the motility of oral cancer cells.

To examine whether UBE2T induces EMT in OSCC cells, we performed RT-qPCR to analyze the expression of EMT marker genes in UBE2T SAS cells. Compared with Control SAS cells, UBE2T SAS cells showed decreased expression of the epithelial cell marker E-cadherin (Fig. 3A), accompanied by increased expression of the mesenchymal cell marker vimentin (Fig. 3B).

To further explore the molecular changes induced by UBE2T, we performed RNA sequencing on Control and UBE2T SAS cells, followed by GSEA (Table SIV). Increased expression of UBE2T resulted in enrichment of genes associated with the ‘HALLMARK_EMT’ (Fig. 3C; Table SIV), indicating that UBE2T induces EMT in OSCC cells. Collectively, these findings suggest that UBE2T induces EMT, which contributes to the enhanced motility of OSCC cells.

To determine the factors and biological processes regulated by UBE2T in oral cancer cells, we focused on DEGs whose expression was upregulated more than threefold in UBE2T SAS cells compared with Control SAS cells and identified 62 DEGs upregulated in UBE2T SAS cells (Table SV). Further GO analysis of these 62 DEGs revealed the enrichment in ‘positive regulation of cell motility’ and ‘intermediate filament organization’ terms which were among the top-ranked categories (Fig. 4A). We have also examined the expression of already reported, several motility- and EMT-related factors, including ankyrin repeat domain 1 (ANKRD1) (38), endothelin-1 (39), matrix metalloproteinase-9 (MMP-9) (40), and plasminogen activator, urokinase (PLAU) (41). RT-qPCR analysis revealed increased expression levels of ANKRD1 (Fig. 4B), endothelin-1 (Fig. 4C), MMP-9 (Fig. 4D), and PLAU (Fig. 4E) in UBE2T SAS cells compared with Control SAS cells. These results suggest that UBE2T promotes the expression of various factors associated with motility and EMT in OSCC cells.

To identify the key factors mediating UBE2T-induced motility and EMT, we performed STRING analysis on the 62 DEGs whose expression was upregulated by more than threefold in UBE2T SAS cells compared with Control SAS cells. Among the 62 DEGs upregulated in UBE2T SAS cells, IL-6 was identified as the central hub (Fig. 5A). Additionally, the protein-protein network revealed the interaction of IL-6 with mesenchymal cell marker genes, such as vimentin, endothelin-1, MMP-9, and PLAU (Fig. 5A), whose expression was upregulated in UBE2T SAS cells (Figs. 3B and 4C-E).

IL-6/JAK/signal transducer and activator of transcription 3 (STAT3) signaling is known for its role in cancer progression, including OSCC (42,43). GSEA revealed the possibility of IL-6/JAK/STAT3 signaling pathway activation in UBE2T SAS cells, ranking sixth among the enriched gene sets (Fig. 5B; Table SIV). Furthermore, GO analysis showed enrichment in terms related to ‘positive regulation of tyrosine phosphorylation of STAT protein’ (Fig. 4A), supporting GSEA results. Additionally, RT-qPCR revealed that IL-6 expression was upregulated in UBE2T SAS cells (Fig. 5C), suggesting that UBE2T regulates the IL-6/JAK/STAT3 signaling pathway in OSCC cells.

To examine the role of IL-6 in the induction of EMT, Control SAS cells were cultured in the absence or presence of recombinant IL-6, and EMT-related gene expression levels were analyzed. RT-qPCR revealed that IL-6 treatment decreased the expression of the epithelial cell marker E-cadherin (Fig. 6A) and increased the expression of the mesenchymal cell marker vimentin (Fig. 6B), suggesting that IL-6 induces EMT in oral cancer cells. We also examined the migration of oral cancer cells after treatment with IL-6. The chamber migration assay demonstrated that IL-6 enhanced the motility of Control SAS cells (Fig. 6C and D), revealing that IL-6 not only induced EMT but also promoted motility in OSCC cells. Collectively, these results indicate that IL-6, which is upregulated by increased levels of UBE2T, is critical for maintaining mesenchymal traits and motility in OSCC cells.

As our RNA sequencing data suggested that IL-6/JAK/STAT3 signaling was activated by elevated UBE2T expression (Fig. 5C), we examined whether inhibition of the JAK/STAT3 pathway affected UBE2T-induced EMT. UBE2T SAS cells were treated with ruxolitinib, a selective JAK inhibitor followed by RT-qPCR analysis of EMT marker gene expression. Treatment of UBE2T SAS cells with ruxolitinib induced increased expression of an epithelial cell marker E-cadherin (Fig. 7A). Additionally, ruxolitinib decreased expression of the mesenchymal cell marker vimentin (Fig. 7B), suggesting that cells underwent mesenchymal-epithelial transition (MET), a process reversal to EMT. Moreover, ruxolitinib treatment inhibited the migration of UBE2T SAS cells compared with that of the untreated control (Fig. 7C and D). Taken together, our data suggest that UBE2T induces EMT and enhances cell motility in OSCC, potentially through regulation of the IL-6/JAK/STAT3 signaling pathway (Fig. S1).