Clinical samples
The study analysed a group of 59 squamous cell carcinoma tissue samples and 20 paracancerous tissue samples from patients treated at the Department of Oral and Maxillofacial Surgery, Stomatological Hospital Affiliated with Southwest Medical University, between January 2018 and December 2020. The follow-up deadline was January 2024. None of the patients had undergone preoperative radiation or chemotherapy, and all tissue samples were obtained with appropriate informed consent and approval from the Institutional Review Board of the Stomatological Hospital Affiliated with Southwest Medical University.
IHC staining
All the tissue sections were deparaffinized, rehydrated, subjected to antigen retrieval, blocked and incubated with an anti-YTHDF2 antibody (1:2000, Abcam) overnight at 4°C. This was followed by incubation with secondary antibodies and detection via a diaminobenzidine (DAB) kit (Gene Tech, China), followed by haematoxylin counterstaining. The images were captured via an upright microscope system (Nikon, Japan), and the YTHDF2 immunostaining score was calculated by multiplying the score for the proportion of positively stained tumour cells (categories: 1 (< 25%), 2 (25–50%), 3 (50–75%), and 4 (> 75%) by the score for staining intensity (categories: 1-weak staining, 2-moderate staining, and 3-strong staining). The staining score was assessed by two blinded pathologists, and patients were categorized into low-staining (1–6) and high-staining (8–12) groups. Detailed patient information can be found in Table S1.
Cell culture and transfection
SCC9 and NOK cells were kindly provided by Professor Wang.[14] SCC9 cells were cultured in DMEM/F12 (Gibco, USA), and NOK cells were cultured in high-glucose DMEM (Gibco, USA). Both growth media were supplemented with 10% foetal bovine serum (FBS) (Gibco, USA) and 1% penicillin and streptomycin (Beyotime Biotech, China). The cells were kept in a 5% carbon dioxide incubator, with regular medium changed 2–3 times every week. The cells were passaged when they reached 80% confluence.
Plasmid transfection was conducted via a siRNA transfection kit (Ribo, China) following the manufacturer’s instructions, with three YTHDF2 siRNA sequences designed: siYTHDF2-1, GACCAAGAATGGCATTGCA; siYTHDF2-2, GCACAGAAGTTGCAAGCAA; and siYTHDF2-3, GGTAGCGGGTCCATTACTA.
Quantitative real-time PCR (qRT‒PCR)
Total RNA was extracted from cells via the Steady Pure Rapid RNA Extraction Kit according to the manufacturer’s instructions. This was followed by cDNA synthesis with ReverTra Ace qPCR RT Master Mix (Toyobo, Japan). PCR was then conducted via the SYBR Green Master Mix Kit (Takara, Japan) on a StepOne Plus Real-time PCR System (Applied Biosystems, USA), with GAPDH as the internal control. Each sample was analysed in triplicate, with data analysis based on the 2−ΔΔCT method. The primers used are listed in Table S2.
Western blot
Western blot
Cellular proteins were extracted by lysing cells in RIPA buffer supplemented with protease inhibitors (Sigmar, USA). The extracted protein concentrations were detected via a BCA kit. Subsequently, 20 µg of protein per sample was separated via 10% sodium dodecyl sulfate‒polyacrylamide gel electrophoresis (SDS‒PAGE) and transferred onto a PVDF membrane (Millipore, USA). The samples were incubated overnight at 4°C with primary antibodies against YTHDF2 (1:1000, Abcam), JAK2 (1:5000, Abcam), STAT3 (1:1000, Abcam), p-JAK2 (1:5000, Abcam), p-STAT3 (1:1000, Abcam) and GAPDH (1:10000, affinity), followed by the application of an HRP-labelled secondary antibody (1:1000, Beyotime). The protein‒antibody complexes were visualized with enhanced chemiluminescence (ECL) detection reagents and analysed via Image Lab software.
Proliferative assay
Cell proliferation was assessed through CCK-8 and colony formation assays. For the CCK-8 assay, cells were plated in 96-well plates at a density of 2000 cells per well. The absorbance was measured at 0, 24, 48, 72, and 96 hours at a wavelength of 450 nm postseeding via the CCK-8 system (Dojindo, Japan). For the colony formation assay, cells were seeded in 6-well plates at a density of 1000 cells per well and cultured in DMEM/F12 medium supplemented with 10% FBS for approximately 2 weeks. Colonies were visualized and quantified after fixation with methanol and 0.1% crystal violet staining for 30 minutes.
Scratch wound assay
The cells were seeded into 6-well plates and allowed to grow until they reached confluence. A scratch of approximately 1 mm in the confluent cell layer was made via a P200 pipette tip, followed by continued cell culture. Images were captured at 0, 24, 48, and 72 hours, and the reduction in the wound area was analysed via ImageJ software.
Cell migration and invasion assays
Cell migration and invasion abilities were assessed using chambers with or without Matrigel. A cell suspension was prepared in serum-free DMEM/F12 medium at a concentration of 1🞨105; 500 µL of the suspension was placed in the upper chamber; and 1 ml of DMEM/F12 medium containing 10% FBS was added to the lower chamber. After incubation for 24 hours, the cells on the upper surface of the filter were removed, and those that had migrated through the filter and adhered to the lower surface were stained with 0.1% crustal violet. Five random fields of view were selected under a microscope, images were captured, and the cells were quantified.
Transcriptome sequencing
Total RNA was extracted from three groups of SCC9 cells transfected with siYTHDF2 or control plasmid and sent to Lianchuan Biotech Biotechnology Co., Ltd. (HangZhou, China) for transcriptome sequencing. The RNA was initially tested for quantity, purity, and integrity. cDNAoligo(dT) magnetic beads and high-temperature fragmentation technology were subsequently used to capture and fragment mRNAs with a PolyA tail. This was followed by reverse transcription for conversion into cDNA and second-strand synthesis, with a fragment size of approximately 300 bp ± 50 bp, for library construction. Paired-end sequencing was then conducted via an Illumina Novaseq™ 6000 in PE150 mode. The sequencing data underwent quality control procedures, including the removal of adapters and the filtering of low-quality and repetitive sequences. Alignment and transcript assembly were performed via HISAT2 and StringTie software programs. FPKM quantification was carried out via the ballgown package, and significant differences between samples were analysed via the R package edgeR. Genes with a fold change (FC) greater than 2 or less than 0.5 and a p value less than 0.05 were considered differential genes, confirmed by GO and KEGG enrichment analysis.
Bioinformatic analysis
Pancancer RNA expression data were obtained from the TCGA (https://portal.gdc.cancer.gov/) database in three standardized forms: TPM, FPKM, and RPM. OSCC data were extracted from various oral cancer sites within TCGA-HNSC, consisting of 362 OSCC samples and 32 adjacent cancer samples. Clinical information and RNA-seq expression profiles were downloaded for subsequent analysis.
The “limma” package in R (version 4.0.3) was utilized for data integration and analysis, with a focus on differentially expressed genes, through extracting gene names, p values, false discovery rates (FDRs), and log2-fold change values. Genes with incomplete information were excluded, and those meeting the criteria of FDR < 0.05 and |1log2FC| > 3 were subjected to further analysis. Differential gene visualization was carried out via the “ggplot 2” package. Kaplan‒Meier overall survival curves were constructed with the “survial” and “survminer” packages, while ROC analysis was conducted with the “pROC” package, and ROC curves were visualized via “ggplot2”. The area under the ROC curve served as an indicator of accuracy in diagnosis and prognosis prediction.
Statistical analysis
Data analysis and statistics were conducted via R and SPSS 23.0. Student's t test was utilized to compare two subgroups, whereas one-way ANOVA was used for comparisons among more than two subgroups. Kaplan‒Meier survival analysis was conducted to assess the differences in overall survival (OS) based on different YTHDF2 expression levels. A p value less than 0.05 was considered statistically significant.