The study population included 45 patients with HCC who underwent surgical treatment between September 2021 and September 2023 at the Department of Hepatobiliary Surgery, Guizhou Medical University (Guiyang, China). To ensure the accuracy of the study population, the following inclusion criteria were followed: i) Preoperative imaging diagnosis of HCC; ii) proposed laparoscopic resection or open surgery; and iii) postoperative pathological diagnosis of HCC. The exclusion criteria were as follows: i) Postoperative pathological diagnosis of benign tumors or malignant tumors that were not HCC; and ii) metastatic carcinoma from other tumors that metastasized to the liver. The age range of the patients with HCC was 13-88 years (median age, 56 years) and 85.1% of the patients were male. The collection and use of clinical specimens were approved by the Ethics Committee of Guizhou Medical University (approval no. 2021220; Guiyang, China), and each patient provided written informed consent.

HepG2 and Hep3B liver cancer cells were purchased from The Cell Bank of Type Culture Collection of The Chinese Academy of Sciences. The cell lines were genetically confirmed by short tandem repeat analysis and were routinely tested for mycoplasma contamination. HepG2 and Hep3B cells were incubated in high-glucose DMEM and MEM (both from Gibco; Thermo Fisher Scientific, Inc.), respectively, at 37°C and 5% CO₂. In addition, 1% penicillin/streptomycin and 10% fetal bovine serum (FBS) (Gibco; Thermo Fisher Scientific, Inc.) were added to culture medium.

To investigate the expression pattern of ZNF740 in liver cancer, the online tool GEPIA (https://gepia.cancer-pku.cn/) was employed, which facilitates comprehensive analysis of RNA-seq data from TCGA project (https://www.cancer.gov/ccg/research/genome-sequencing/tcga). The GEPIA platform was accessed at the official GEPIA website, and LIHC; (Liver Hepatocellular Carcinoma) was specified as the cancer type of interest. The gene of focus, ZNF740, was entered into the search query to retrieve expression data. Utilizing GEPIA, a differential expression analysis was conducted between tumor and normal samples. The platform automatically adjusts for multiple testing and provides adjusted P-values to calculate the false discovery rate (FDR). Upon completion of the analysis, the graphical output and the underlying numerical data were exported for further statistical processing and integration into the current research findings.

The cultivation of 293T cells, procured from Pricella Biotechnology Co., Ltd. In the context of the second-generation triple plasmid packaging system, the stoichiometric ratios of the plasmids were calibrated as follows: pMD2.G/packaging support plasmid/target plasmid (pHBLV-CMV-MCS-IRES-Puro), 2:3:4 (Addgene, Inc.). Concurrently, the mass ratio of plasmid DNA (in μg) to the volume of Lipofectamine^® 2000 reagent (in μl) (Invitrogen; Thermo Fisher Scientific, Inc.) was meticulously adjusted to 8:10 to optimize the transfection efficiency. Following a 48-h incubation period at 37°C in a CO₂-controlled incubator, the culture medium was harvested, with the supernatant fraction being reserved. Transfection was performed utilizing Lipofectamine^® 2000 reagent (Invitrogen; Thermo Fisher Scientific, Inc.), in conjunction with Opti-MEM medium (Gibco; Thermo Fisher Scientific, Inc.). The transfection protocol involved the introduction of 10 μg of the designated plasmids into the 293T cells. Post-transfection, the cells were subjected to a 12-h incubation phase with fresh medium, after which they are harvested 48 h post-transfection for downstream application.

To achieve stable ZNF740 knockdown, METTL3 knockdown, ZNF740 overexpression and METTL3 overexpression, recombinant lentiviruses expressing short hairpin RNA (shRNA/sh)-ZNF740, sh-METTL3, METTL3 and ZNF740 were prepared by Guangzhou RiboBio Co., Ltd. In addition, sh-control was used as a negative control for knockdown and empty vector was used as a negative control for overexpression. For cell culture, 5×10⁴ cells/well were equally distributed into 6-well plates. The degree of cell confluence was regulated at ~30% after the cells had fully attached to the well after ~10 h. The necessary virus titer was determined using the following formula: Viral titer (TU/ml)=MOI value x number of cells/viral volume (ml). A MOI of 10 was employed to infect the cells. For subsequent culture, lentivirus-containing culture media was used in lieu of the initial culture medium. After 12 h, the medium was replaced with fresh medium and the cells were placed in a constant temperature incubator at a 37°C. Under a fluorescence microscope fitted with the appropriate filter sets for GFP detection, the expression of the EGFP gene could be detected in HepG2 and Hep3B liver cancer cells 72 h after infection, since the lentiviruses included the EGFP gene sequence. Furthermore, because the lentiviruses included a puromycin resistance gene sequence, cells were screened using 2 μg/ml puromycin (Dalian Meilun Biology Technology Co., Ltd.) to further reduce the impact of wild-type cells on subsequent investigations. Antibiotic resistance was mostly seen in the cell population following more than three rounds of subculturing. The protein and mRNA expression levels were assessed using reverse transcription-quantitative PCR (RT-qPCR) and western blotting, respectively. The shRNA sequences are as follows: sh-control, forward 5′-CCGGCAACAAGATGAAGAGCACCAACTCGAGTTGGTGCTCTTCATCTTGTTGTTTTTG-3′, reverse 5′-AATTCAAAAACAACAAGATGAAGAGCACCAACTCGAGTTGGTGCTCTTCATCTTGTTG-3′; sh-ZNF740#1, forward 5′-CCGGGCTGCTCAAGAAACAAAGGAACTCGAGTTCCTTTGTTTCTTGAGCAGCTTTTTG-3′, reverse 5′-AATTCAAAAAGCTGCTCAAGAAACAAAGGAACTCGAGTTCCTTTGTTTCTTGAGCAGC-3′; sh-ZNF740#2, forward 5′-CCGGGCGATATATGTGATATGCGTTCTCGAGAACGCATATCACATATATCGCTTTTTG-3′; reverse, 5′-AATTCAAAAAGCGATATATGTGATATGCGTTCTCGAGAACGCATATCACATATATCGC-3′; sh-METTL3#1, forward, 5′-CCGGGCCTTAACATTGCCCACTGATCTCGAGATCAGTGGGCAATGTTAAGGCTTTTTG-3′, reverse 5′-AATTCAAAAAGCCTTAACATTGCCCACTGATCTCGAGATCAGTGGGCAATGTTAAGGC-3′; sh-METTL3#2, forward 5′-CCGGGCAAGTATGTTCACTATGAAACTCGAGTTTCATAGTGAACATACTTGCTTTTTG-3′, reverse 5′-AATTCAAAAAGCAAGTATGTTCACTATGAAACTCGAGTTTCATAGTGAACATACTTGC-3′.

Total RNA was extracted from liver cancer cells using TRIzol^® reagent (Invitrogen; Thermo Fisher Scientific, Inc.) and the quality of the RNA was evaluated with a NanoDrop spectrophotometer (Thermo Fisher Scientific, Inc.). Subsequently, cDNA libraries were created with the TruSeq Stranded mRNA LT Sample Prep Kit (VAHTS Universal V6 RNA-seq Library Prep Kit for Illumina; cat. no. NR604-01; Vazyme, Inc.), according to the manufacturer's guidelines. The sequencing and analysis of the transcriptome were conducted on the Illumina HiSeq X Ten platform (Illumina, Inc.) using the HiSeq X Reagent Kit (cat. no. FC-501-2501; Illumina, Inc.). Nucleotide reads range from 75 bp using paired end sequencing mode. The BioAnalyzer 2100 (Agilent Technologies, Inc.) and the Ion Torrent™ General Library Quantification Kit (cat. no. A26217; Thermo Fisher Scientific, Inc.) were utilized for quality control and quantitative detection of the sequence libraries; the 10 pM sequence libraries were mutated into single-stranded DNA according to Illumina instructions, captured on an Illumina flow cell, clustered in situ, and clustered in the Illumina NovaSeq 6000 sequencer for 150 cycle sequencing. The raw data were subjected to quality control processing using Trimmomatic (24), which included the removal of low-quality reads. HISAT2 (25) was utilized to align the high-quality reads to the human genome (GRCh38.p14), and counts were obtained for each gene. The DESeq2 package (version 1.40.2) in R (26) was utilized for conducting differential expression analysis. Statistical significance was assessed using a threshold of P<0.05 and |log2Fold change|>2. The clusterProfiler package (version 3.17) in R (27) was utilized to perform functional enrichment analysis on the genes that were differentially expressed. Sequencing data have been uploaded to the National Center for Biotechnology Information (NCBI) Gene Expression Omnibus (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE268479) under the accession number GSE268479.

Total RNA was extracted from liver cancer cells and human liver tissues using TRIzol reagent. Follow the manufacturer's operating instructions, the PrimeScript™ RT reagent kit and gDNA Eraser kit (Takara Biotechnology Co., Ltd.) were used to perform RT, and qPCR was performed using the SYBR^® Premix Ex Taq™ kit on a LightCycler 480 instrument (Takara Biotechnology Co., Ltd.). Briefly, qPCR was performed as follows: Initial denaturation was 95°C for 5 min; followed by 40 cycles of denaturation (95°C, 10 sec), annealing and extension (60°C, 30 sec). All reactions were conducted in a 20-μl reaction volume in triplicate. The relative mRNA expression levels were normalized to GAPDH levels. The comparative Cq (2^−ΔΔCq) method was used to determine the relative expression levels of mRNA (28). The RT-qPCR primers used were acquired from Sangon Biotech, Co., Ltd., as follows: GAPDH, forward 5′-GGAGCGAGATCCCTCCAAAAT-3′, reverse 5′-GGCTGTTGTCATACTTCTCATGG-3′; ZNF740, forward 5′-GCAGGTGTGAGTTTGGTTCC-3′, reverse 5′-CCTCAGCACATCAGGGCTAC-3′; METTL3, forward 5′-TTGTCTCCAACCTTCCGTAGT-3′, reverse 5′-CCAGATCAGAGAGGTGGTGTAG-3′; hypoxia-inducible factor-1α (HIF1A), forward 5′-GAACGTCGAAAAGAAAAGTCTCG-3′, reverse 5′-CCTTATCAAGATGCGAACTCACA-3′.

Cells and tissues were lysed using RIPA lysis buffer (Beyotime Institute of Biotechnology), which prevents protein degradation and the loss of phosphorylation. The BCA Protein Assay Kit (Beyotime Institute of Biotechnology) was used to determine the concentration of total proteins in accordance with the manufacturer's instructions. Equal quantities of protein extracts (20 μg) were separated by SDS-PAGE on 15% gels and were then transferred onto a PVDF membrane (MilliporeSigma). Subsequently, the PVDF membrane was blocked with skim milk for 1 h at 4°C and was incubated with primary antibodies overnight at 4°C. After washing with TBS-0.1% Tween three times, the membranes were incubated with a secondary antibody conjugated to HRP (1:1,000; cat. no. SA00001-2; Wuhan Sanying Biotechnology) for 60 min at 4°C. The presence of immunoreactivity was observed by employing an enhanced chemiluminescence system (Shanghai Yeasen Biotechnology Co., Ltd.), with GAPDH (1:50,000; cat. no. 60004-1-Ig; Wuhan Sanying Biotechnology) serving as a loading control. The following antibodies were used: Anti-ZNF740 (1:500; cat. no. 25411-1-AP), anti-METTL3 (1:500; cat. no. 15073-1-AP), anti-YTHDF1 (1:1,000; cat. no. 17479-1-AP), anti-YTHDF2 (1:2,000; cat. no. 24744-1-AP), anti-YTHDF3 (1:5,000; cat. no. 25537-1-AP) and anti-HIF1A (1:2,000; cat. no. 20960-1-AP) (all from Wuhan Sanying Biotechnology).

According to the manufacturer's instructions, cells were transfected with small interfering (si)RNAs using Lipofectamine RNAiMAX transfection reagent (cat. no. 13778-075; Invitrogen; Thermo Fisher Scientific, Inc.). The cells were cultured in 6-well plates until they achieved 60% confluence. To create the transfection mix, 150 μl Opti-MEM (Gibco; Thermo Fisher Scientific, Inc.) was diluted with 7.5 μl Lipofectamine RNAiMAX, and 150 μl Opti-MEM was used to dilute 25 pmol siRNA. To create the siRNA-lipid combination, the diluted siRNA was mixed 1:1 with the Lipofectamine RNAiMAX that had previously been diluted. This compound was incubated at room temperature for 15 min to form a stable siRNA-lipid combination, which was used to transfect the cells for 6 h at 37°C. Subsequently, the cells were placed in medium containing 5% FBS and incubated at 37°C in a constant temperature incubator. The cells underwent RT-qPCR and western blot analysis 48 h after transfection, according to the aforementioned protocols. The siRNAs were purchased from Guangzhou RiboBio Co., Ltd. The si-control was used as a negative control and the siRNA sequences are as follows: si-control, forward 5′-CAACAAGAUGAAGAGCACCAAU-3′, reverse 5′-UUGGUGCUCUUCAUCUUGUUG-3′; si-METTL3, forward 5′-UCUAACUCAGGAUCUGUAGCU-3′, reverse 5′-CUACAGAUCCUGAGUUAGAGA-3′; si-YTH N6-methyladenosine RNA binding protein (YTHDF)1, forward 5′-UUAUCUUGUCCUUUUGUUCUC-3′, reverse 5′-GAACAAAAGGACAAGAUAAUA-3′.

Hep3B cells exposed to sh-control, sh-ZNF740#1 and sh-ZNF740#2 lentiviruses for were cultured in MEM supplemented with 10% FBS and 0.1 mM non-essential amino acids (Nanjing KeyGen Biotech Co., Ltd.). When cells reached the logarithmic phase of growth, they were collected and counted to establish a xenograft mouse model. A total of 18 BALB/c nude mice were allocated into three groups: sh-control, sh-ZNF740#1 and sh-ZNF740#2 (n=6/group). Briefly, to induce tumor growth, 0.1 ml PBS (Gibco; Thermo Fisher Scientific, Inc.) containing 2×10⁶ Hep3B cells was injected under the skin on the right side of the back of each BALB/c nude mouse (age, 4-5 weeks; weight, 18-20 g) were purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd. and were bred under pathogen-free conditions (temperature, 18-22°C; humidity, 50-60%; 12-h light/dark cycle). The bedding, feed and water were replaced every 2 days. The formula used to measure tumor volume was volume=(width)² x length/2, with width and length representing the shortest and longest diameters of the tumor, respectively. The Guizhou Medical University Animal Care Committee approved the present study (approval no. 2100616). The mice were sacrificed after 4 weeks of observation, or when they reached humane endpoints, such as a weight loss of 20% or behaving in a depressed manner. To reduce any possible pain or discomfort during the sacrifice, the mice were profoundly anesthetized with 5% isoflurane prior to cervical dislocation, and 3% isoflurane was administered for maintenance. Following the collection of the tumors, the tumor volume curve was created.

The tumor tissues collected from patients with liver cancer and from a mouse xenograft model were fixed in 4% paraformaldehyde at 4°C for 24 h and then embedded in paraffin. Sections of paraffin-embedded tissue (2-3 μm) were dried for 60 min at 60°C and then subjected to two rounds of dewaxing in xylene for 30 min each. Subsequently, a decreasing gradient of alcohol concentrations was used for hydration. For antigen retrieval, the sections were heated in citrate buffer at 100°C for 10 min. The sections were then blocked with 5% bovine serum albumin (Gibco; Thermo Fisher Scientific, Inc.) for 1 h at room temperature, and 3% H₂O₂ was used to block endogenous peroxidase/phosphatase activity for 10 min at room temperature. Subsequently, the sections were incubated overnight at 4°C with the following primary antibodies: Anti-ZNF740 (1:1,000; cat. no. 25411-1-AP), anti-Ki67 (1:2,000; cat. no. 27309-1-AP), anti-METTL3 (1:750; cat. no.15073-1-AP) and anti-HIF1A (1:50; cat. no. 20960-1-AP) (all from Wuhan Sanying Biotechnology). The sections were then treated with secondary antibodies labeled with biotin (1:1,000; cat. no. SA00001-2; Wuhan Sanying Biotechnology) for 1 h at room temperature and incubated with an avidin solution labeled with horseradish peroxidase (HRP) (Wuhan Sanying Biotechnology) for 10 min at room temperature. DAB was used to identify positive staining, followed by a 3-min application of hematoxylin as a counterstain. Finally, the sections were dried by progressively adding ethanol, as well as xylene, and were subsequently sealed using neutral balsam. Representative images were obtained using a fluorescence microscope (Olympus Corporation) at a magnification of ×200. Cells were categorized into grades based on the intensity of staining at the membrane, nucleus or cytoplasmic location, and then the percentage of the number of cells in each grade was counted: Grade 1 (0-10% intensity) was scored as 1 point; grade 2 (11-40% intensity) was scored as 2 points; grade 3 (41-70% intensity) was scored as 3 points; and grade 4 (71-100% intensity) was scored as 4 points. Finally, all of the scores were added up to obtain the total score.

In each well of a 6-well plate, 1×10⁷ HepG2 or Hep3B cells were cultured in DMEM or MEM supplemented with 10% FBS. When the cell confluence reached 90%, a scratch was created using a 200-μl pipette tip. Subsequently, the cells were incubated in FBS-free medium for 24 h at room temperature. Representative images were obtained under a fluorescence microscope (Olympus Corporation) at 0 and 24 h. Mobility was assessed using the following formula: Cell migration rate (%)=(0 h scratch width-scratch width after incubation)/0 h scratch width ×100.

HepG2 or Hep3B cells (1×10⁵/well) were seeded and cultured in the upper chamber of a 24-well Transwell plate (pore size, 8 μm) in FBS-free medium. In addition, 600 μl medium supplemented with 10% FBS was added to the lower chamber of the Transwell plate. Subsequently, the dish was placed in a humidified environment with 5% CO₂ and was maintained at 37°C for 48 h. After incubation, the cells that had not migrated to the lower chamber were removed from the upper chamber, and the migrated cells were fixed with 4% formaldehyde for 10 min at room temperature and stained with 0.1% crystal violet for 10 min at room temperature. To determine the quantity of cells that had migrated, five arbitrary fields were chosen for cell counting under a fluorescence microscope. In addition, 100 μl Matrigel in FBS-free media was added to the upper chamber and incubated at 37°C for 4 h to conduct the Transwell invasion assay. The procedures were conducted in a similar fashion as aforementioned for the migration assay.

HepG2 or Hep3B cells transfected with siRNAs or infected with lentiviruses were seeded in 6-well plates at a density of 1×10³ cells/well. After being cultured for 2 weeks, the cells were fixed with 4% paraformaldehyde for 15 min and then subjected to staining with 0.5% crystal violet for 10 min at room temperature. Colonies consisting of >100 cells were counted manually and included in the analysis to quantify colony formation.

The CCK-8 assay (Shanghai Yeasen Biotechnology Co., Ltd.) was conducted in accordance with the manufacturer's guidelines. Briefly, 96-well plates were seeded with HepG2 or Hep3B cells at a density of ~4×10³ cells/well 24 h after transfection or infection. The cells were incubated for 2 h at 37°C in the dark with 10 μl CCK-8 reagent. The optical density was then measured using a microplate reader (Tecan Group, Ltd.) at 450 nm.

METTL3 mutants were constructed, and a dual-luciferase reporter assay was performed to identify the interaction sites between METTL3 and ZNF740. The mutation sites and wild-type sequences are as follows: METTL3 binding sequences -813 to -804: 5′-CCCTCCCCAC-3′; -758 to -749: 5′-CCGCCCCCTC-3′; -296 to -287: 5′-AAACCCCCAA-3′. The corresponding mutant sequences were as follows: P1: 5′-AAAGAAAACA-3′; P2: 5′-AATAAAAAGA-3′; P3: 5′-CCCAAAAACC-3′. Lipofectamine 2000 was used to transfect HepG2 and Hep3B cells when they reached ~70% confluence. To conduct luciferase reporter assays, 100 ng METTL3-promoter luciferase plasmid and 2 ng Renilla luciferase control reporter plasmid, which was used to ensure accurate transfection efficiency, were transfected into HepG2 and Hep3B cells in 96-well plates. The day after transfection, the Dual-Luciferase Reporter Assay System (Promega Corporation) was used to measure the fluorescence intensities. Briefly, HepG2 and Hep3B cells transfected with the luciferase reporter plasmids were lysed in 20 μl 1X lysis solution and agitated at room temperature for 20 min. Then, 10 μl cellular lysate was mixed with 50 μl Luciferase Assay Reagent II. Subsequently, the activity of firefly luciferase was measured for 10 sec using a luminometer, after which, 50 μl Stop & Glo Reagent was added to the mixed solution, and the Renilla luciferase activity was measured for 10 sec Firefly luciferase activity was normalized to Renilla luciferase activity. Each assay was performed at least in triplicate to ensure statistical robustness.

Actinomycin D (ActD; MilliporeSigma) is a commonly used transcription inhibitor that intercalates into the DNA double helix, blocking the activity of RNA polymerase, thus inhibiting the synthesis of new mRNA. After treating cells with ActD, the half-life of mRNA can be calculated by quantitatively measuring the decrease of mRNA over time, thereby assessing its stability. HepG2 and Hep3B cells from different groups were placed in 6-well plates and exposed to 10 μg/ml ActD. At the onset of the experiment (i.e., the 0-h time point), cell samples were collected to serve as the experimental control group. Subsequently, cell samples were collected again at the 2, 4, 6 and 8-h time points at 37°C. The collection of samples at these various time points allowed for comparative analysis to observe and quantify the degradation rate and half-life of mRNA under the influence of ActD. By measuring the levels of mRNA at different time points, the stability of mRNA could be inferred. RT-qPCR was performed according to previously described methods (24).

According to the manufacturer's protocol, a ChIP assay kit (cat. no. ab500; Abcam) was used to perform the ChIP assay. After HepG2 and Hep3B cells were crosslinked, lysed and sonicated, immunoprecipitation was performed using the anti-ZNF740 antibody (1:500; cat. no. 25411-1-AP; Wuhan Sanying Biotechnology) or IgG (1:2,000; cat. no. 30000-0-AP; Wuhan Sanying Biotechnology). Subsequently, the precipitated DNA was extracted and subjected to RT-qPCR as aforementioned.

The 45 patients with HCC, from whom clinical samples were collected, were followed up to assess the survival time of the patients. Survival data were plotted using Kaplan-Meier survival curves and the log-rank test was used to determine the P-value. For survival analysis, a threshold can be set to define samples with expression above the threshold as a high expression group and samples below the threshold as a low expression group. In the present study, the median value of ZNF740 expression level was set as the cutoff value for distinguishing the high (n=23) and low (n=22) expression groups.

In the present study, ROC curves were analyzed using ZNF740 expression levels from data collected from the 45 clinical samples. The area under the curve (AUC) reflects the magnitude of the value of the diagnostic test; the larger the area, the closer it is to 1.0, the higher the diagnostic trueness; the closer it is to 0.5, the lower the diagnostic trueness; and when it equals 0.5, there is no diagnostic value. AUC=0.5 to 0.7, lower accuracy; AUC=0.7 to 0.9, some accuracy; AUC >0.9, high accuracy.

Raw sequencing reads from liver cancer cells were preprocessed using FastQC (Babraham Bioinformatics) and Trimmomatic (24). Differential expression analysis was performed using the DESeq2 R package (26) with a P-value cutoff of 0.05 and FDR correction. KEGG enrichment analysis was conducted with the clusterProfiler R package (29), identifying pathways significantly enriched with differentially expressed genes. Visualization of the enrichment results was achieved using ggplot2 (27). All analyses were conducted using R version 3.6.1 (29).

The reference sequence of the METTL3 gene was obtained from the NCBI GenBank database (Gene ID: 56339; https://www.ncbi.nlm.nih.gov/gene/56339). The promoter region 2,000-3,000 base pairs upstream of the transcription start site was predicted using FGENESH software (version number 7.2.2; Softberry Inc.). Potential transcription factor binding sites within promoter regions were predicted using ALIEN software (http://www.genome.jp/tools/alien/) and the TRANSFAC database (BioBase GmbH). MatInspector (version number 2.2; Genomatix Software Suite) was used to further analyze these sites.

Cells were lysed in RIPA Buffer (cat. no. 89901; Cell Signaling Technology, Inc.), which contained a mixture of protease inhibitors (Roche Diagnostics) to inhibit proteolytic degradation. The cell lysate was then incubated with the primary antibodies anti-YTHDF1 (1:1,000; cat. no. 17479-1-AP), anti-YTHDF2 (1:2,000; cat. no. 24744-1-AP), anti-YTHDF3 (1:5,000; cat. no. 25537-1-AP) and IgG (1:2,000; cat. no. 30000-0-AP) (all from Wuhan Sanying Biotechnology), at a concentration of 1:250 total protein, overnight at 4°C with constant rotation. Protein A/G Plus-Agarose beads used for immunocomplex capture were purchased from Santa Cruz Biotechnology, Inc. These beads were pre-washed and pre-incubated with non-specific mouse IgG (Jackson ImmunoResearch Laboratories, Inc.) to reduce non-specific binding. The antibody-protein complex was isolated by adding the Protein A/G agarose beads to the lysate and incubating for 2 h at 4°C with rotation. The agarose beads with bound immune complexes were pelleted by centrifugation at 1,000 x g for 5 min at 4°C and washed five times with ice-cold PBS (pH 7.4). The immune complexes were eluted by boiling in 2X Laemmli buffer (Bio-Rad Laboratories, Inc.) for 5 min at 95°C. The eluted proteins were resolved by SDS-PAGE on 15% gels using a Mini-PROTEAN^® TGX gel (Bio-Rad Laboratories, Inc.) and were transferred onto a PVDF membrane. Subsequently, western blotting was performed as aforementioned.

GSEA of the RNA-seq results comparing METTL3-overexpressing cells and control cells was performed in the present study. GSEA was performed using GSEA software version 4.3.3 (Broad Institute; https://www.gsea-msigdb.org/gsea/downloads.jsp) (26). For GSEA, the Molecular Signature Database v6.2 was used, which is integrated into the GSEA software suite and provides a comprehensive collection of annotated gene sets for GSEA use (26). The expression data consists of a matrix of gene expression values for different samples, which has been prepared in supported file formats such as txt. Each row in the dataset represents a gene, each column represents a sample, and the expression values indicate the level of gene expression measured under specific conditions. The GSEA software calculates an enrichment score (ES) for each gene set. The ES was then statistically evaluated to determine its significance and the FDR was calculated to correct for multiple hypothesis testing. The results of the GSEA were interpreted by examining the ES, normalized ES (NES), P-value and FDR q-value for each gene set.

Statistical analyses were conducted using SPSS 23.0 (IBM Corporation) or GraphPad Prism 7 (Dotmatics) software. Each experiment was repeated three times. A paired or unpaired Student's t-test was used to compare the two groups. One-way ANOVA with Tukey's multiple comparisons test was used to compare more than two groups. The Kaplan-Meier method and log-rank test were used to generate and compare differences between survival curves. For statistical analysis of IHC-related data the Wilcoxon signed-rank test for paired data. Correlation analysis was performed using the Pearson correlation coefficient, -1 indicates a perfect negative correlation, 0 indicates no correlation, and 1 indicates a perfect positive correlation. The χ² test was used to analyze whether there was a significant association between categorical variables. P<0.05 was considered to indicate a statistically significant difference.

To examine the involvement of ZNF740 in HCC progression, RNA-seq data acquired from TCGA database were investigated; of which 369 cases were tumor tissues and 160 cases were normal tissues. ZNF740 expression was significantly greater in HCC tissues than in adjacent healthy tissues (Fig. 1A and B). To confirm these results, RT-qPCR analysis of ZNF740 mRNA expression levels was conducted in a cohort of 45 patients with HCC to verify the increased expression of ZNF740 in HCC tissues (Fig. 1C and D). Analysis of this cohort of 45 patients revealed that ZNF740 protein expression was significantly increased in 42.1% of patients with HCC compared with the median value of ZNF740 expression in HCC samples. Another 45.6% of patients with HCC had upregulated ZNF740 expression levels in tumor tissues compared with in normal tissues, but the difference was not significant. In addition, 12.3% of patients had lower ZNF740 expression levels in tumor tissues than in normal tissues. Overall, the protein expression level of ZNF740 was upregulated in liver cancer tumor tissues (Fig. 1E). These findings were further confirmed by performing IHC and western blotting, which both revealed elevated expression of ZNF740 in HCC samples compared with in nontumor samples (Fig. 1F-H). Increased ZNF740 expression was associated with a poor prognosis in patients with HCC, as shown by the overall survival curve plotted on the basis of the collected samples (Fig. 1I). To validate the prognostic accuracy of ZNF740, a ROC/AUC analysis was performed and exhibited good performance, with an AUC value of 0.944, suggesting that ZNF740 has high accuracy as a biological marker for HCC (Fig. 1J). These findings indicated that ZNF740 may be upregulated in HCC and that elevated levels of ZNF740 could be linked to an unfavorable prognosis among patients with HCC.

To clarify the biological functions of ZNF740 in liver cancer, lentiviral vectors were used to achieve stable overexpression or knockdown of ZNF740 in HepG2 and Hep3B cells. The knockdown efficiency of sh-ZNF740 was detected using RT-qPCR and western blotting; sh-ZNF740 effectively knocked down the expression levels of ZNF740 in liver cancer cells (Fig. S1A). ZNF740 overexpression lentivirus effectively upregulated the expression levels of ZNF740 in liver cancer cells (Fig. S1B). A wound healing assay demonstrated that ZNF740 overexpression significantly improved the migratory capacity of liver cancer cells (Fig. 2A). By contrast, specific knockdown of ZNF740 decreased the migration of liver cancer cells (Fig. 2B). In addition, the migration and invasion of HepG2 and Hep3B cells were significantly enhanced by ZNF740 overexpression, but were significantly inhibited by ZNF740 knockdown (Fig. 2C and D). The impact of ZNF740 on liver cancer cell proliferation was also assessed by conducting a colony formation assay and analyzing proliferation curves. Colony formation and proliferation were significantly increased in ZNF740-overexpressing liver cancer cells, whereas colony formation and proliferation were significantly reduced in ZNF740-knockdown cells (Fig. 2E-H). These results indicated that ZNF740 may serve a vital role in enhancing liver cancer cell invasion, migration and proliferation.

To determine the effects of ZNF740 on the proliferation of tumors in vivo, a mouse xenograft model of Hep3B cells was generated. In line with the in vitro findings, specific knockdown of ZNF740 in Hep3B cells led to the development of smaller tumor masses and decreased tumor weight in comparison to those in the control group (Fig. 3A and B). Moreover, the volume of tumor generated from Hep3B cells with targeted ZNF740 knockdown was significantly reduced after 4 weeks of xenograft transplantation (Fig. 3C). Furthermore, the results of IHC demonstrated a decrease in the ratio of Ki-67-positive cells in the ZNF740 knockdown group, along with a decrease in the expression of METTL3 and HIF1A (Fig. 3D). Collectively, these results provided convincing evidence that ZNF740 may have a vital role in enhancing the proliferation of HCC cells in a mouse xenograft model.

To examine the possible molecular mechanisms underlying the effects of ZNF740 on liver cancer, RNA-seq analysis was performed on HepG2 cells in the ZNF740-specific knockdown (sh-ZNF740#2) and control groups. The findings indicated notable differences in gene expression between the ZNF740-specific knockdown group and the control group. In the ZNF740-specific knockout group, METTL3 expression was notably reduced (Fig. 4A and B). Furthermore, KEGG enrichment analysis revealed a significant alteration in the 'RNA polymerase' (gene counts: 21; P=2.8×10⁻²), 'mRNA surveillance pathway' (gene counts: 18; P=6.4×10⁻²), 'RNA degradation' (gene counts: 20; P=1.9×10⁻³), 'ATP-dependent chromatin remodeling' (gene counts: 12; P=1.1×10⁻⁵), 'Spliceosome' (gene counts: 11; P=2.3×10⁻²), 'Ubiquitin mediated proteolysis' (gene counts: 13; P=9.6×10⁻³), 'Nucleocytoplasmic transport' (gene counts: 10; P=1.7×10⁻²), 'Lysine degradation' (gene counts: 12; P=2.0×10⁻²) and 'Thermogenesis' (gene counts: 17; P= 6.4×10⁻²) in the ZNF740-specific knockdown group (Fig. 4C). These results indicated that ZNF740 could have an essential function in controlling METTL3 expression and regulating RNA modification in liver cancer cells. METTL3 is a methyltransferase mainly involved in N6-methyladenosine (m6A) modifications of RNA. These modifications are one of the most common epigenetic modifications on eukaryotic mRNAs, and have an important impact on several processes, such as RNA stability, splicing, nucleation, degradation and translation (30). Therefore, it was hypothesized that a regulatory relationship may exist between ZNF740 and METTL3. To confirm the regulatory connection between ZNF740 and METTL3, the effect of stable overexpression or knockdown of ZNF740 on METTL3 mRNA expression was examined. The findings indicated that ZNF740 overexpression markedly elevated the mRNA expression levels of METTL3, whereas ZNF740 knockdown significantly decreased METTL3 mRNA expression levels (Fig. 4D and E). Western blotting of HepG2 and Hep3B lysates supported this conclusion (Fig. 4F). To further investigate the linkage between ZNF740 and the METTL3 promoter, the sequence of the binding motifs of ZNF740 was examined and a schematic representation of the different motifs (P1-P3) in the METTL3 promoter region is shown in Fig. 4G and H. Dual luciferase reporter assays were performed to determine the activities of the different motifs of the METTL3 promoter, and the results showed that the P1 motif showed the most differential changes in METTL3 promoter activity after ZNF740 overexpression; thus, the P1 motif may be considered the binding site of ZNF740 to the METTL3 promoter (Fig. 4I). Subsequently, a ChIP assay was performed and it was demonstrated that ZNF740 could directly bind to the P1 motif on the promoter of METTL3 (Fig. 4J). Considering these findings, a regulatory connection between ZNF740 and METTL3 was revealed, and ZNF740 was identified as a novel TF of METTL3.

To further explore the vital role of ZNF740 in enhancing the invasive and proliferative abilities of liver cancer by controlling the transcription of METTL3, ZNF740 was stably overexpressed in the HepG2 and Hep3B cell lines. Moreover, siRNAs were utilized to suppress METTL3 expression. The knockdown efficiency of si-METTL3 was detected using RT-qPCR and western blotting (Fig. S1C). After long-term observation, the migration and invasion of HepG2 and Hep3B cells were significantly increased by ZNF740 overexpression, whereas no such effects were observed in cells also transfected with si-METTL3 (Fig. 5A and B). The findings from the wound healing assay led to similar conclusions (Fig. 5C). Similarly, the proliferation of HepG2 and Hep3B cells was significantly increased by ZNF740 overexpression, whereas ZNF740 overexpression had no impact on the colony formation or proliferation of si-METTL3-transfected liver cancer cells (Fig. 5D and E). Rescue assays using si-METTL3 successfully inhibited the tumor-promoting effects of ZNF740 on the proliferation of liver cancer cells. These findings indicated that the effects of ZNF740 may rely on its ability to control METTL3 transcription.

Since METTL3 is an important m6A writer (30), the present study further examined whether METTL3 exerted regulatory effects on key oncogenic signaling pathways via m6A modification in liver cancer. Infection with the METTL3 overexpression lentivirus effectively increased the expression levels of METTL3 in liver cancer cells. The comparison of gene sets between HepG2 cells overexpressing METTL3 and control cells using GSEA revealed significant activation of the HIF1A signaling pathway (Fig. 6A). The knockdown efficiency of sh-METTL3 was detected using RT-qPCR and western blotting (Fig. S1D). Subsequently, RT-qPCR and western blotting were conducted to confirm the regulatory effects of METTL3 on HIF1A. According to the RT-qPCR analysis findings, the mRNA expression levels of HIF1A were significantly greater in HepG2 and Hep3B cells with METTL3 overexpression, whereas the expression levels were significantly decreased upon METTL3 knockdown (Fig. 6B and C). Infection with the METTL3 overexpression lentivirus effectively increased the expression levels of METTL3 in liver cancer cells (Fig. S1E). Similarly, the findings from western blotting indicated a notable increase in the protein levels of HIF1A in HepG2 and Hep3B cells with METTL3 overexpression, whereas a marked decrease was observed upon METTL3 knockdown (Fig. 6D). In addition, HIF1A mRNA stability was investigated by inhibiting fresh RNA production in HepG2 and Hep3B cells using ActD. The degradation efficiency of HIF1A mRNA was detected in ActD-treated HepG2 and Hep3B cells 8 h after METTL3 overexpression, and the results suggested that METTL3 overexpression slowed the degradation efficiency of HIF1A mRNA (Fig. 6E). IP is a laboratory technique used to isolate and enrich specific proteins from cell lysates; this method is commonly used to study protein interactions. To identify the m6A readers involved in this process, IP assays were performed with YTHDF1, YTHDF2, YTHDF3 and METTL3. The results showed a direct link between YTHDF1 and METTL3, but not between YTHDF2 or YTHDF3, thus suggesting that YTHDF1 can function in conjunction with METTL3 (Fig. 6F). The knockdown efficiency of si-YTHDF2 was detected using RT-qPCR and western blotting (Fig. S1F). Subsequently, a HIF1A mRNA stability assay was performed and demonstrated that YTHDF1 gene silencing could reverse the biological effects of METTL3 overexpression (Fig. 6G). Similarly, the RT-qPCR and western blotting results indicated a notable increase in relative mRNA and protein expression levels of HIF1A due to METTL3 overexpression; however, the expression levels were reduced in the METTL3 + si-YTHDF1 group (Fig. 6H-I). These findings indicated that METTL3 may enhance the proliferation of liver cancer cells by controlling the level of HIF1A expression in an m6A modification-dependent manner, with YTHDF1 acting as the m6A reader.

To further validate the clinical relevance of ZNF740, METTL3 and HIF1A, correlation analysis was performed on the 45-patient and TGCA cohorts. The r-values were all greater than 0. TCGA database analysis showed that the correlation analysis between ZNF740 and METTL3 had an r-value of 0.73, and the correlation analysis between ZNF740 and VEGFA had an r-value of 0.51. Due to the limited number of samples collected, the correlation of ZNF740 with METTL3, HIF1A and VEGFA detected in the clinical samples showed a positive correlation, but it was very weak (r<0.3) (Fig. 7A and B); therefore, further analysis in subsequent studies is required. The results were additionally validated through IHC staining of ZNF740, METTL3 and HIF1A. Distinguishing patients with high and low expression based on median immunohistochemistry score revealed that the levels of METTL3 and HIF1A were notably decreased in samples with low ZNF740 expression (Fig. 7C). The clinical significance of ZNF740, METTL3 and HIF1A was confirmed by these findings, indicating the notable therapeutic efficacy and potential utilization of ZNF740.