Data collection and processing

The TCGA-THCA RNAseq data were obtained from the UCSC database in FPKM format. The UCSC Xena browser (http://xena.ucsc.edu/) was employed to evaluate RNF115 expression across various tumor types. This study included a total of 615 samples from three categories: normal solid tissues, primary tumor samples, and metastatic tumors. The GEPIA tool (http://gepia.cancer-pku.cn/) was used to examine RNF115 expression in paired and unpaired normal and tumor samples. The cutoff value for RNF115 expression was set at high (75%) and low (25%) in GEPIA. Log2(FPKM + 1) processing was applied to the corresponding numbers of adjacent and cancer samples. Visualization of the data was achieved using the ggplot2 package.

Unpaired sample plots were assessed via the Mann–Whitney U test, while paired sample plots were analyzed using a t-test.

Samples

Tissue specimens were collected from THCA patients (n = 113) undergoing primary surgical resection at the First Affiliated Hospital of Xi'an Jiaotong University. Expression of RNF115 was analyzed in the collected tumor samples and their corresponding non-tumorous thyroid tissues. The procurement of tissue specimens received the endorsement of the Ethics Committee at the First Affiliated Hospital of Xi'an Jiaotong University (Approval no. 2021–1407). Informed consent for sample collection was acquired from each patient.

Cell culture and treatment

Human thyroid regular cells (HTori-3) and THCA-specific cell strains (TPC-1, KTC-1, CAL62, and SNU-790) were sourced from Beina Biology (China). They were cultured in a DEME/F12 medium (Hyclone, UT, USA) supplemented with 10% fetal bovine serum (FBS) and a 1% antibiotic combination of penicillin/streptomycin, maintained at 37 °C in an atmosphere containing 5% CO2. TPC-1 and SNU-790 cells were transfected with vectors carrying RNF115 overexpression (Ad-RNF115), RNF115 knockdown (Ad-shRNA1 and Ad-shRNA2), negative controls (Ad-NC and Ad-shNC), and/or CDK10 overexpression, using the Lipofectamine™2000 transfection kit (Invitrogen, CA, USA). The designated sequences for shRNA include: shRNA-1: 5′-GCCGUGGCUAGAACUGCAUTT-3′ and shRNA-2: 5′-CGTCTGAATAGAATTAATT-3′. A fraction of the transfected cells was treated with cycloheximide (CHX, a protein synthesis inhibitor) for various durations (0, 4, 8, and 12 h). Meanwhile, another subset received a 10 μM dose of MG132, a proteasome inhibitor, for a duration of 8 h.

Methylthiazoltetrazolium (MTT) assay

TPC-1 and SNU-790 cells, with a concentration of 1.5 × 104/mL, were allocated into 96-well plates and then incubated for a period of 48 h. Post this incubation phase, 20 μL of 3-(4, 5-dimethylthiazol-2-yl)-2, 5-diphenyltetrazolium bromide was introduced to every well, followed by a further 4-h incubation period. Subsequently, the medium was aspirated, and the resulting formazan crystals were solubilized in 150 μL of DMSO per well. The viability of the cells was ascertained by gauging the absorbance at 490 nm with the aid of a spectrophotometer.

Colony formation assay

The colony formation capacity of TPC-1 and SNU-790 cells was evaluated using a previously established method (Wei et al. 2019). Cells were dispensed into a 6-cm culture dish at a density of 1 × 103 cells per dish. They were then allowed to grow for two weeks, with the culture medium being refreshed every three days. Subsequently, the cells were rinsed using phosphate-buffered saline (PBS), underwent fixation with a 4% paraformaldehyde solution, and were subsequently stained with a 0.4% crystal violet solution for a duration of 15 min. Following this, the colony count was determined.

in vivo tumor model

The First Affiliated Hospital Ethics Committee of Xi'an Jiaotong University granted approval for all procedures involving animals (Approval no. 2021–1407). Mice (sourced from Charles River, Beijing, China), aged 4–6 weeks, were randomly assigned to one of four groups: Ad-NC, Ad-RNF115, Ad-shNC, and Ad-shRNA1, with six mice in each group. TPC-1 cells that underwent stable transfection were administered into mice via the lateral tail vein injection. Tumor volume was documented at seven-day intervals. Thirty days post-injection, in vivo imaging of mice was conducted utilizing the IVIS Spectrum In Vivo Imaging System to monitor lung metastasis. Following imaging, mice were humanely euthanized, and tumor tissues were collected for further investigations. All animal procedures complied with the National Institutes of Health Laboratory Animal Care and Used Guidelines.

Wound-healing assay

Cells that underwent transfection were placed in six-well culture plates and grown in a serum-deprived medium until they reached full confluence. Subsequently, a defined wound area was established by introducing a scratch into the cell monolayer using a 10-µL pipette tip. The state of the scratch wounds was documented at 0 and 24 h post-initiation through microscopic observation (Olympus, Japan).

Transwell invasion assay

An invasion assay was executed utilizing 24-well transwell plates (Corning, NY, USA). Cells, with a concentration of 1 × 106, were suspended in serum-free medium and introduced into the Matrigel-pre-coated upper chambers (BD Biosciences, NJ, USA). The bottom chamber was supplemented with 0.6 mL of DEME/F12 medium containing 10% FBS. After incubating for 24 h, cells were fixed with 4% paraformaldehyde and then stained with crystal violet. The number of cells that invaded was then quantified using a microscope.

Immunofluorescence assay

An immunofluorescence assay was executed based on the methodology delineated by Sáenz et al. (2019). Initially, cells were fixed using 4% paraformaldehyde and permeabilized with 0.3% Triton X-100. After a blocking step with 5% FBS, the cells were exposed to primary antibodies, specifically anti-E-cadherin, anti-N-cadherin, anti-Flag, and anti-HA, and incubated at 4 °C overnight. Thereafter, they were subjected to secondary antibodies. To demarcate the nuclei, 4, 6-diamidino-2-phenylindole (DAPI; procured from Yeasen, China) was employed. Observations were made utilizing a confocal laser scanning microscope (Model LSM510; procured from Zeiss, Germany).

Hematoxylin and eosin (HE) staining

Tumor tissue morphology was assessed via HE staining. Following the processes of deparaffinization and rehydration, tissue slices underwent staining procedures using hematoxylin and eosin dyes. The stained sections were subsequently examined under a microscope.

Enrichment analysis

RNF115-associated genes were predicted via the BioGRID (https://thebiogrid.org/) and underwent enrichment analysis using Metascape (https://metascape.org/) and the R programming language. Enrichment analysis, incorporating both Gene Ontology (GO) and the Kyoto Encyclopedia of Genes and Genomes (KEGG), was conducted using a suite of ontology databases: KEGG Pathway, GO Biological Processes, Reactome Gene Sets, Canonical Pathways, CORUM, WikiPathways, and PANTHER pathways. Enriched terms were chosen based on the following criteria: a p-value below 0.01, an observed count of at least 3, and an enrichment factor greater than 1.5, with the enrichment factor delineating the ratio of observed counts relative to those anticipated by chance.

Identification of critical genes interacted with RNF115.

The top 25 RNF115-interacting genes were identified via Pathway Commons (https://www.pathwaycommons.org/). Genes regulated by RNF115 were analyzed for their correlation with RNF115 in THCA. Data for THCA samples were obtained from the TCGA database via cBioPortal. We probed the expression of genes regulated by RNF115 via GEPIA, an online tool accessible at http://gepia.cancer-pku.cn/, leveraging data from the TCGA database.

RT-qPCR

RNA from both tissues and cells was isolated utilizing the TRIzol reagent (Invitrogen, CA, USA). RT-qPCR was performed following a previously described protocol (Pan et al. 2019). PCR was performed using primers detailed in Supplementary Table S1.

Western blotting

Protein extraction from tumor tissues and cells was performed using RIPA buffer, with the subsequent western blotting procedure carried out as detailed in earlier methods (Bacopulos et al. 2012). The primary antibodies used for western blotting were anti-RNF115 (0.4 µg/mL), anti-E-cadherin (1:10,000), anti-N-cadherin (1 µg/mL), anti-CDK10 (1:2,000), anti-SFN (2 µg/mL), anti-MYC (2 µg/mL), anti-p-Raf-1 (1:1,000), anti-Raf-1 (1:2,000), anti-p-MEK1/2 (1;1,000), anti-MEK1/2 (1:1,000), anti-p-ERK1/2 (1:1,000), anti-ERK1/2 (1 µg/mL), anti-CyclinD1 (1 µg/mL), anti-CDK4 (2 µg/mL), anti-Bax (1:100), anti-Cleaved caspase-3 (1:2,000), and anti-β-actin (1:5,000) (Invitrogen, CA, USA).

Immunohistochemistry (IHC)

The IHC staining procedure was adopted. In a summarized protocol, paraffin-embedded tissue samples were sectioned into 4-μm slices. Following deparaffinization and rehydration, the sections underwent antigen retrieval using 1 mM EDTA (pH 8.0). To neutralize the tissue's inherent peroxidase activity, sections were treated with 0.3% hydrogen peroxide. Blocking was achieved using 5% goat serum. This was followed by an incubation phase with primary antibodies: anti-RNF115 (1:200), anti-CDK10 (1:200), anti-Ki67 (1:200), anti-E-cadherin (10 µg/mL), and anti-N-cadherin (1:100) sourced from Invitrogen, CA, USA. Subsequently, sections were incubated with biotinylated secondary antibodies and observed under a microscope (Olympus, Japan).

Co-immunoprecipitation (CoIP)

Cells of the HEK-293 T line, numbered at 1 × 106, underwent co-transfection utilizing constructs that harbored Flag-tagged RNF115 alongside HA-tagged CDK10. Following a 24-h period, cells were analyzed through western blot and immunohistochemistry (IHC) to assess the expression and localization of RNF115 and CDK10, utilizing primary antibodies against Flag and HA.

Statistical analysis

In vitro studies were repeated on three separate occasions, whereas in vivo assays involved six biological repeats and three technical repetitions. Data analysis was executed using SPSS 27.0 software (IBM, IL, USA), with results expressed as the mean ± standard deviation. For comparisons involving multiple groups, one-way ANOVA was utilized. For pairwise comparisons, Tukey’s test was employed. A statistically significant result is indicated when p < 0.05.