SF3B4 regulates proliferation and apoptosis in hepatocellular carcinoma via alternative splicing and interaction with TRIM28 and SETD5

Abstract

Background

SF3B4 encodes a core subunit of the U2-type spliceosome and is implicated in abnormal cell growth and tumorigenesis. However, its role in regulating gene expression and alternative splicing in hepatocellular carcinoma (HCC) remains inadequately understood.

Methods

SF3B4 expression was downregulated in HCC cells, followed by high-throughput transcriptome sequencing to capture the transcriptomic changes induced by SF3B4. This approach facilitated the identification of potential targets regulated by SF3B4 at both the transcriptional and alternative splicing levels in HCC cells. Additionally, SF3B4-binding RNAs in Huh7 cells were identified through iRIP-seq. The RNA-seq data were subsequently analyzed to elucidate the molecular mechanisms by which SF3B4 affects gene expression and alternative splicing in HCC.

Results

Downregulation of SF3B4 promoted apoptosis and inhibited cell proliferation. In the RNA-seq data, the number of regulated alternative splicing events (RASE) significantly outnumbered the differentially expressed genes (DEGs), consistent with SF3B4’s role as a splicing factor that regulates a wide array of pre-mRNA splicing events. Furthermore, a total of 252 common RNA targets bound by SF3B4 were identified. Correlation analysis with RNA-seq data suggested that SF3B4 may bind to TRIM28, potentially modulating its mRNA expression levels. Additionally, SF3B4 may influence pre-mRNA alternative splicing by interacting with SETD5.

Conclusion

SF3B4 contributes to HCC progression by directly binding mRNAs and interacting with proteins, thereby regulating gene expression and alternative splicing.

Introduction

Hepatocellular carcinoma (HCC) ranks as one of the most prevalent malignancies globally, with a high mortality rate, placing it third in terms of cancer-related deaths [1]. Over 679,000 new cases are diagnosed annually, and more than 622,000 individuals succumb to the disease worldwide. Current treatment options for HCC include surgical resection, radiotherapy, chemotherapy, and systemic therapies [2]. Early diagnosis and intervention can significantly improve patient survival; however, many patients present at advanced stages, resulting in poor prognoses. Despite undergoing surgical treatment, approximately 70% of patients experience relapse within 5–10 years post-surgery [3]. Therefore, further investigation into the molecular mechanisms underlying HCC is essential to enhance diagnostic accuracy, refine treatment strategies, and improve patient outcomes.

Splicing factor 3b subunit 4 (SF3B4), a spliceosome-associated protein, is a key component of the SF3B complex. As a classical RNA-binding protein (RBP), SF3B4 encodes one of the four subunits of splicing factor 3B, playing pivotal roles in transcriptional regulation and pre-mRNA splicing, while also interacting with various other proteins [4]. Studies have indicated that SF3B4 is overexpressed in several malignancies, positioning it as a potential therapeutic target [5]. In esophageal squamous cell carcinoma, high SF3B4 expression correlates with poor lymphatic invasion and prognosis, underscoring its role in lymphatic progression [6]. SF3B4 is similarly overexpressed in cervical cancer, where it promotes cell proliferation and invasion, significantly contributing to the disease’s progression [7]. In HCC, SF3B4 is also highly expressed, with its expression negatively regulated by miRNA-133b, further advancing the progression of HCC. SF3B4 stabilizes GPAA1 mRNA, promoting proliferation, invasion, and migration in HCC cells [8]. Additionally, SF3B4 knockdown induces G1/S cell cycle arrest by restoring p27kip1 levels while simultaneously inhibiting cyclins and CDKs in HCC cells [9]. These findings suggest that SF3B4 could serve as a pivotal biomarker in the early stages of HCC and may present a promising target for therapeutic intervention.

A critical step in transcriptional regulation is splicing, during which intronic sequences are removed from pre-mRNA, exonic sequences are ligated, and mature mRNA is formed. In malignant tumors, alternative splicing can lead to defects that generate aberrant splicing variants, promoting tumor progression [10]. Studies have revealed that elevated expression of SF3B4 results in the mislocalization of the tumor suppressor gene Kruppel-like factor 4 into nonfunctional transcripts in cancer cells, thereby facilitating tumorigenesis in HCC [11]. SF3B4 has been shown to enhance cell proliferation, inhibit apoptosis, and promote migration and invasion of HCC cells. Dysregulation of the Notch signaling pathway is known to be closely linked to the initiation and progression of various malignancies, including breast and colorectal cancers. Similarly, extensive evidence supports the critical role of the Notch signaling pathway in HCC progression. SF3B4 regulates the expression of ENAH, which in turn promotes HCC proliferation, invasion, and migration by activating Notch signaling, ultimately leading to a poor prognosis [12].

Based on these findings, SF3B4 plays a pivotal role in HCC development. This study aims to investigate the impact of SF3B4 expression on HCC and to demonstrate its association with regulated alternative splicing events (RASE), positioning SF3B4 as a potential diagnostic and therapeutic target for HCC.

Methods

Cloning and plasmid construction

All siRNAs were purchased from Gemma (Suzhou, China). Non-targeting control siRNA (siNegative): 5’-UUCUCCGAACGUGUCACGUTT-3’ (sense). siRNA targeting SF3B4 (siSF3B4): 5’-GCACCAAGGCUAUGGCUUUTT-3’ (sense).

Cell culture and transfections

Huh7 cells (Procell Life Science & Technology Co., Ltd., China) were cultured at 37 °C with 5% CO₂ in DMEM supplemented with 10% fetal bovine serum (FBS), 100 µg/mL streptomycin, and 100 U/mL penicillin. siRNA transfection of Huh7 cells was carried out using Lipofectamine™ RNAiMAX Transfection Reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s instructions. Infected cells were harvested after 48 h for RT-qPCR and Western blot analyses.

Assessment of gene expression

GAPDH (glyceraldehyde-3-phosphate dehydrogenase) served as a control gene for assessing the efficiency of SF3B4 knockdown. cDNA synthesis was performed according to standard protocols, and RT-qPCR was conducted using the Bio-Rad S1000 Thermal Cycler with SYBR^® Green Master Mix (Low Rox Plus; YEASEN, China). The concentration of each transcript was normalized to the GAPDH mRNA level using the 2-ΔΔCT method. Statistical comparisons were performed using a paired Student’s t-test with GraphPad Prism software (San Diego, CA).

Western blot

For Western blotting, Huh7 cells were lysed in ice-cold Wash Buffer (1× PBS, 0.1% SDS, 0.5% NP-40, and 0.5% sodium deoxycholate) supplemented with a protease inhibitor cocktail (Roche, ) and incubated on ice for 30 min. The samples were then boiled for 10 min in 1X SDS sample buffer and separated by 10% SDS-PAGE. Membranes were incubated with TBST buffer (20 mM Tris-buffered saline, 0.1% Tween-20) containing 5% non-fat milk for 1 h at room temperature. Primary antibodies, including SF3B4 antibody (1:1,000, A17608, Abclonal) and GAPDH (1:5,000, ATPA00013Rb, AtaGenix), were applied, followed by HRP-conjugated secondary antibodies. Bound secondary antibodies (anti-mouse or anti-rabbit, 1:10,000, Abcam) were detected using an enhanced chemiluminescence (ECL) reagent (Bio-Rad, 170506).

Cell proliferation assay

Cell proliferation was assessed using the Cell Counting Kit-8 (CCK-8; Dojindo, Co., Shanghai, China). Briefly, Huh7 cells were seeded at 6,000 cells/well in 96-well plates. Cells treated with an equal volume of phosphate-buffered saline served as controls, while wells without cells were used as blank controls. After incubation for 24, 48, and 72 h, 20 µL of CCK-8 solution was added to the culture medium and incubated for an additional 3 h. The optical density of the cells was measured at 450 nm using a PerkinElmer EnVision microplate reader. The cell proliferation rate was calculated using the following formula: Proliferation rate = [(experimental OD value − blank OD value) / (control OD value − blank OD value)] × 100%.

Annexin V apoptosis assay

To detect tumor cell apoptosis, an Annexin V-fluorescein isothiocyanate (FITC)/propidium iodide (PI) apoptosis detection kit (YEASEN, China) was used according to the manufacturer’s instructions. Briefly, Huh7 cells were seeded into six-well plates, cultured for 48 h, and transfected with siRNA targeting SF3B4. Cells treated with an equal volume of PBS served as the control. After treatment, cells were stained with 5 µL Annexin V-FITC and 10 µL of 20 µg/mL PI reagents, and incubated in the dark at room temperature for 10–15 min. After adding 400 µL PBS, cell apoptosis was assessed by flow cytometry. Cells that were positive for Annexin V-FITC but negative for PI fluorescence were considered apoptotic.

Co-immunoprecipitation and library preparation

For RNA-related assays, Huh7 cells were irradiated once at 400 mJ/cm2 and subsequently lysed in ice-cold wash buffer (1× PBS, 0.1% SDS, 0.5% NP-40, and 0.5% sodium deoxycholate) supplemented with 400 U/mL RNase inhibitor (2311 A, Takara, Japan) and protease inhibitor cocktail (B14001, Bimake, China). After a 30-minute incubation on ice, the lysate was clarified by centrifugation at 10,000 rpm for 10 min at 4 °C. To digest RNA, RQ I (Promega, 1 U/µL) was added to a final concentration of 0.1 U/µL, and the mixture was incubated at 37 °C for 30 min. The solution was then vortexed and centrifuged at 13,000 × g for 15 min at 4 °C to remove cell debris. RNA digestion was performed using MNase and RNase T1 (60140-101, Thermo Scientific, America), and digestion was stopped with 0.5 M EDTA. For immunoprecipitation, the supernatant was incubated overnight at 4 °C with 10 µg SF3B4 antibody (10482-1-AP, Proteintech, America) or control IgG antibody (AC005). The immunoprecipitate was then incubated with protein A/G Dynabeads (10009D, Thermo Scientific, America) for 2 h at 4 °C. After applying a magnetic separator and removing the supernatant, the beads were washed sequentially with lysis buffer, high-salt buffer (250 mM Tris 7.4, 750 mM NaCl, 10 mM EDTA, 0.1% SDS, 0.5% NP-40, and 0.5% sodium deoxycholate), and PNK buffer (50 mM Tris, 20 mM EGTA, and 0.5% NP-40), each for two washes. The beads were resuspended in Elution buffer (50 mM Tris 8.0, 10 mM EDTA, and 1% SDS) and incubated at 70 °C for 30 min to release the immunoprecipitated RBPs with crosslinked RNA, followed by vortexing. After removing the magnetic beads, the supernatant was transferred to a clean 1.5 mL microfuge tube. Proteinase K (A600451, Sangon Biotech, China) was added at a final concentration of 1.2 mg/mL to both the 10% input (non-immunoprecipitated) and the immunoprecipitated samples, and incubation was performed for 120 min at 55 °C. RNA was purified using Phenol: Chloroform: Isopentyl Alcohol (25:24:1, pH < 5) reagent.

For high-throughput sequencing, cDNA libraries were prepared using the KAPA RNA Hyper Prep Kit (KK8541, KAPA) following the manufacturer’s protocol. Libraries were sequenced using the Illumina NovaSeq 6000 system with 150nt paired-end sequencing.

Retrieval and process of public data

SF3B4-bound peaks were downloaded from the ENCODE project (https://www.encodeproject.org/) (ENCSR267OLV and ENCSR279UJF).

Data analysis

After aligning the reads to the genome using HISAT2 [13], the unique genome alignments were retained, and PCR duplicates were removed. Peak calling was then performed using two software tools: Piranha and ABLIRC. Piranha, as previously described [14], was utilized for peak identification, while ABLIRC was employed for identifying the binding regions on the GRCm39 genome, as previously described [15]. The peak calling process involved scanning the entire genome with a 5 bp window and 5 bp step size from the start of each chromosome. A peak was identified if the depth of the first window was at least 2.5 times greater than the depth of eight consecutive windows or if the mean depth was greater than 50. If the depth of these eight consecutive windows was less than 4% of the maximum depth of the peak, the peak was considered to have ended. To ensure robustness, the reads from each gene were randomly distributed 500 times, and the peak depth frequency was analyzed for statistical significance (P < 0.05) or for peaks with a maximum depth of at least 10. Using input samples as the control, the abundance of peaks was analyzed, and peaks with IP abundance greater than 4 times the input abundance (adjustable parameter) were considered as the final binding peaks. The target genes of the IP were determined by analyzing the peak locations, and the binding motifs of the IP protein were identified using HOMER software.

Functional enrichment analysis

To explore the functional categories of the target genes associated with these peaks, Gene Ontology (GO) terms and KEGG pathways were analyzed using the KOBAS 2.0 server. A hypergeometric test and the Benjamini-Hochberg false discovery rate (FDR) controlling procedure were employed to assess the enrichment of each term.

Reverse transcription qPCR validation

Finally, total RNA from RNA-seq and iRIP-seq library preparation was used for RT-qPCR. RNA was reverse transcribed into cDNA using M-MLV Reverse Transcriptase (R021-01, Vazyme, China), and real-time PCR was performed using the StepOne Real-Time PCR System with HieffTM qPCR SYBR^® Green Master Mix (Low Rox Plus; YEASEN, China). PCR conditions consisted of an initial denaturation at 95 °C for 5 min, followed by 40 cycles of denaturation at 95 °C for 15 s and annealing/extension at 60 °C for 30 s. Each PCR reaction was performed in triplicate for each sample.

Results

SF3B4 promote HCC proliferation

The TCGA database analysis revealed a significant upregulation of SF3B4 expression in HCC compared to normal tissues (Fig. 1A). Kaplan-Meier survival analysis indicated that elevated SF3B4 expression was associated with poorer overall survival (Fig. 1B). To investigate the functional impact of SF3B4 knockdown, three siRNAs targeting SF3B4 were synthesized. Among them, siSF3B4-2 demonstrated the highest silencing efficiency, which was statistically significant (P < 0.05) (Fig. 1C-D). Additionally, silencing SF3B4 expression resulted in a marked reduction in HCC cell proliferation and an increase in apoptosis (P < 0.05) (Fig. 1E).

Fig. 1

Knockdown of SF3B4 inhibits cellular proliferation and promotes apoptosis in HCC. (A) SF3B4 expression in TCGA Liver hepatocellular carcinoma (LIHC) dataset accessed through GEPIA2. (B) Prognostic analysis of SF3B4 expression in TCGA HCC cohort in GEPIA2. (C) Histogram illustrating the expression levels of SF3B4 siRNAs. ***P < 0.001. (D) Relative protein expression of SF3B4 following siSF3B4 treatment in Huh-7 cells. (E) Proliferation and apoptosis analysis in Huh7 cells post-SF3B4 knockdown

SF3B4 related gene expression in HCC

Subsequently, RNA-seq quantitative analysis and PCA were performed on SF3B4 (P < 0.001) (Fig. 2A-B). si-SF3B4 treatment in Huh7 cells led to the identification of 713 differentially expressed genes (DEGs), including 430 upregulated and 283 downregulated genes (Fig. 2C). The expression levels of these DEGs were visualized via hierarchical clustering heat maps (Fig. 2D). GO analysis of biological processes revealed that upregulated DEGs were primarily associated with NMDA receptor activity regulation, cell adhesion, and transcriptional regulation by RNA polymerase II. Downregulated genes were enriched in nucleosome assembly, lipid metabolism, and transmembrane transport (Fig. 2E). Further correlation analysis revealed that CCN1 and SETD7 showed a positive correlation with SF3B4, while NOLC1, CCN1, EMC6, SETD7, and CDK14 exhibited a negative correlation with SF3B4 (Fig. 2F).

Fig. 2

Interacting genes of SF3B4. (A) Quantification of SF3B4 expression by RNA-seq. (B) PCA based on FPKM values of all detected genes following SF3B4 knockdown. (C) Volcano plot showing all DEGs between siSF3B4 and NC samples. (D) Hierarchical clustering heat map depicting expression levels of DEGs. (E) Scatter plot illustrating the most enriched GO biological processes for upregulated and downregulated DEGs. (F) Bar plot representing expression patterns and statistical significance of selected DEGs from RNA sequencing and RT-qPCR validation. ***P < 0.001, **P < 0.01, *P < 0.05

SF3B4 regulates gene alternative splicing in HCC

Additionally, RASEs were identified in the si-SF3B4 RNA-seq dataset (P < 0.05) (Fig. 3A). These RASE were regulated in accordance with PSI values of SF3B4, as determined by principal component analysis (PCA) at varying expression levels (Fig. 3B). A heatmap was generated to display the RASE regulated by SF3B4 (Fig. 3C). Following SF3B4 knockdown in Huh7 cells, GO functional enrichment analysis of the alternative splicing genes (RASGs) revealed their involvement in multiple biological processes, including mRNA processing, RNA splicing, and chromatin organization (Fig. 3D). Furthermore, significant differences in the splicing ratios of TIA1, KTN1, and FOXP2 were identified, which were selected based on their significance (Fig. 3E).

Fig. 3

SF3B4 is involved in gene-specific splicing in Huh7 cells. (A) Bar plot displaying SF3B4-regulated RASE. (B) PCA based on PSI values of SF3B4-regulated RASE at various expression levels. (C) Hierarchical clustering heatmap of SF3B4-regulated RASE based on splicing ratio values. (D) Scatter plot illustrating the most enriched GO biological processes associated with RASGs. (E) Schematic diagrams showing the structures of ASEs. RNA-seq and RT-qPCR validation of ASEs are shown at the bottom of the right panel. ***P < 0.001, **P < 0.01, *P < 0.05

SF3B4 binds to mRNAs associated with HCC

Subsequently, iRIP-seq was performed using an anti-SF3B4 antibody, followed by Western blot analysis, resulting in high-throughput sequencing data (Fig. 4A). The heatmap analysis revealed a strong correlation between SF3B4_IP_1 and SF3B4_IP_2, while a low correlation was observed between SF3B4_IP and Input samples, indicating distinct characteristics between the two groups (Fig. 4B). Pie charts were utilized to illustrate the regional distribution of transcript expression in the two IP samples. In the IP_1 sample, the majority of reads were concentrated in the CDS and Nc_exon regions, whereas in the IP_2 sample, the reads were more abundant in intron and Nc_exon regions (Fig. 4C). Motif analysis identified the top 5 enriched motifs (Fig. 4D). Furthermore, the read density map revealed that the enriched regions in the IP samples aligned with the transcript structure of the FTH1 gene, suggesting that SF3B4 may interact with the FTH1 gene in these regions. This enrichment was validated by qPCR, supporting the hypothesis that SF3B4 protein may have binding or regulatory effects on the FTH1 gene (Fig. 4E).

Fig. 4

SF3B4 regulates mRNA in HCC. (A) Western blot analysis of SF3B4 immunoprecipitation using anti-SF3B4 antibody. (B) Heatmap displaying the hierarchically clustered Pearson correlation matrix comparing transcript expression values for IP and Input samples. (C) Bar plots illustrating the distribution of SF3B4 iRIP-seq samples across different genomic regions. (D) Motif analysis results showing the enriched motifs from SF3B4-bound peaks across the two biological replicates. (E) SF3B4 binding peaks in the FTH1 gene. The read density map of SF3B4-binding peaks across FTH1 is shown on the left. ***P < 0.001, **P < 0.01, *P < 0.05

Identification of SF3B4 binding RNA targets

A total of 252 genes were identified through SF3B4 in eCLIP-seq data from K562 and HepG2 cells (Fig. 5A). The genomic distribution of RNA binding protein peaks, shown in a histogram, revealed that the majority of peaks (81.48%) were located within coding sequences (CDS), while a smaller portion (18.52%) overlapped with three prime UTRs (Fig. 5B). Motif analysis, based on P-values, identified the top motif as CAUCGUC (P-value 1e-9), with additional motifs including UCGUGGGC and AAACGGC (Fig. 5C). GO enrichment analysis highlighted biological processes significantly associated with translation initiation, translation, and SRP-dependent co-translational protein targeting to membranes. Furthermore, KEGG pathway analysis revealed significant enrichment in metabolic and disease-related pathways, including ribosome function, biosynthesis of amino acids, and carbon metabolism (Fig. 5D).

Fig. 5

Exploration of RNA targets binding to SF3B4. (A) Venn diagram illustrating the overlap of peak-associated genes between iRIP-seq and eCLIP-seq. (B) Distribution of oligo peaks across the reference genome. (C) Motif analysis results showing enriched motifs from oligo peaks identified in both biological replicates. (D) Bar plot depicting the most enriched GO biological processes and KEGG pathways associated with the overlapping peak genes from iRIP-seq and eCLIP-seq

SF3B4 binds to RNA and regulates its expression

Subsequent overlap analysis between RNA-seq data and the 252 peak-associated genes revealed six genes with differential expression that are potentially regulated by SF3B4 (Fig. 6A). Analysis of TRIM28 expression in NC and siSF3B4-treated cells showed a significant reduction in TRIM28 levels following SF3B4 knockdown. These results were validated by qPCR, which confirmed a lower relative expression of TRIM28 in the siSF3B4 group compared to the NC (Fig. 6B). Genome browser visualization of the TRIM28 gene locus (chr19) indicated multiple transcript variants, suggesting potential direct or indirect binding of SF3B4 to these regions. RIP-qPCR further confirmed the enrichment of TRIM28 at the binding sites of SF3B4 (Fig. 6C).

Fig. 6

SF3B4 binds to TRIM28 and regulates its expression. (A) Venn diagram showing the overlap between peak-associated genes and DEGs. (B) Bar plot depicting the expression pattern and statistical significance of DEGs for TRIM28 from RNA-seq and RT-qPCR validation. (C) SF3B4 binding peaks on the TRIM28 gene. The reads density landscape of SF3B4 binding peaks across TRIM28 (left) and quantification of TRIM28 expression by qRT-PCR from iRIP-seq data (right). ***P < 0.001, **P < 0.01, * P < 0.05

SF3B4 binds to mRNA and regulates its alternative splicing

A Venn diagram illustrates 63 overlapping regions between peak-associated genes and RASGs (Fig. 7A). Additionally, SETD5 gene transcripts, SF3B4 protein binding across different gene regions, and SETD5 gene expression were analyzed in various samples (Fig. 7B). The analysis also included the examination of alternative splicing events of SETD5 under different conditions, highlighting the distinct splice variants of SETD5 in samples such as siSF3B4 and NC. Furthermore, RNA sequencing data revealed a decrease in the variable splicing ratio of SETD5 following siSF3B4 treatment. This observation was validated by qPCR, which confirmed significant alterations in the splicing ratio between different sample groups (Fig. 7C).

Fig. 7

SF3B4 binds to SETD5 and regulates its alternative splicing. (A) Venn diagram depicting the overlap of genes between peak-associated genes and RASGs. (B) SF3B4 binding to SETD5. The reads density landscape of SF3B4-binding peaks across SETD5 (left). Quantification of SETD5 expression by qRT-PCR using iRIP-seq data (right). (C) SF3B4 regulates alternative splicing of SETD5. Left panel: IGV-sashimi plot illustrating regulated alternative splicing events and binding sites across the SETD5 mRNA. Reads distribution of RASE is plotted in the upper panel, with gene transcripts displayed below. Right panel: Schematic representation of the structures of ASEs. RNA-seq and RT-qPCR validation of ASEs are shown at the bottom of the right panel. * P < 0.05

The SF3B4 regulated transcriptome in LIHC of TCGA database

To identify SF3B4-regulated genes, a volcano plot was generated to illustrate differential gene expression, revealing 5,537 down-regulated genes and 4,017 up-regulated genes (Fig. 8A). Heat maps displayed the expression levels of DEGs in the two groups (Fig. 8B). A Venn diagram of DEGs from RNA-seq data and TCGA revealed 72 up-regulated and 57 down-regulated overlapping DEGs (Fig. 8C). GO analysis of the up-regulated DEGs highlighted processes such as response to organic cyclic compounds, response to drugs, and positive regulation of RNA polymerase II transcription (Fig. 8D). Conversely, GO analysis of the down-regulated DEGs indicated enrichment in processes like positive regulation of transcription, DNA-templated, and signal transduction (Fig. 8E). Expression levels of TRIM28 and SETD7 were further examined using FPKM values (Fig. 8F). A Venn diagram of RASGs from TCGA and RNA-seq data revealed overlapping genes (Fig. 8G), with GO analysis indicating correlations with mRNA processing, mRNA splicing, and mitochondrial respiratory chain complex I (Fig. 8H). SF3B4 was shown to regulate the alternative splicing of KTN1 and SETD5 (Fig. 8I).

Fig. 8

The SF3B4-regulated transcriptome in LIHC from the TCGA database. (A) Identification of SF3B4-regulated genes, with up-regulated genes in red and down-regulated genes in blue in the bar plot. (B) Hierarchical clustering of DEGs in high and low SF3B4 expression samples, with FPKM values log2-transformed and median-centered by gene. (C) Venn diagram showing the overlap of up- and down-regulated DEGs between LIHC and RNA-seq data. (D-E) The top 10 GO biological processes of overlapping DEGs in LIHC and RNA-seq. (F) Gene expression levels of SF3B4-regulated genes in LIHC from the TCGA database. (G) Venn diagram showing the overlap of RASGs between LIHC and RNA-seq. (H) The top 10 GO biological processes of overlapping RASGs in LIHC and RNA-seq. (I) SF3B4 regulates alternative splicing of genes. RNA-seq quantification of alternative splicing events (ASEs) is shown in the bottom right panel. The altered ratio of AS events was calculated using the formula: alternative splice junction reads / (alternative splice junction reads + model splice junction reads). * P < 0.05, ** P < 0.01

Discussion

The progression of HCC is strongly linked to the aberrant splicing of numerous tumor-related genes. Identifying tumor-specific splicing variants that promote HCC proliferation could provide deeper insights into the underlying pathogenesis of the disease [16]. While the precise mechanisms remain unclear, dysregulation of splicing is widely regarded as a key driver of HCC progression [17]. SF3B4, an RNA-SF3b splicing factor, is involved in splicing, translation, and transcription regulation, all of which contribute to tumor development. Downregulation of SF3B4 expression in HCC led to a reduction in cell proliferation and an increase in apoptosis, findings that are consistent with the results of Liu et al. [8] SF3B4 overexpression, driven by increased DNA copy number, is associated with HCC progression, intrahepatic metastasis, and poor prognosis [18].

Following SF3B4 knockdown, GO functional enrichment analysis revealed alterations in various biological processes, including mRNA processing, RNA splicing, and chromatin organization. SF3B4 regulates cancer cell proliferation and senescence by modulating p21 mRNA stability through the NMD pathway. Its depletion disrupts NMD factor recruitment, stabilizing p21 mRNA and promoting senescence [19]. Meanwhile, SF3B4 plays a crucial role in regulating cellular senescence, with its knockdown inducing p21-dependent senescence in fibroblasts and lung cancer cells. Overexpression of SF3B4 mitigates therapy-induced senescence, highlighting its potential as a therapeutic target in cancer treatment [20]. This analysis identified the RASEs of TIA1, KTN1, and FOXP2, all of which are subject to SF3B4-regulated alternative splicing in HCC, highlighting their potential relevance for further investigation. As demonstrated by Pan et al., KTN1 plays a pivotal role in the initiation and progression of HCC by regulating cell survival, migration, invasion, cell cycle progression, and apoptosis inhibition [21]. Additionally, FTH1, a gene binding to SF3B4, enhances iron storage in HCC cells, thereby conferring resistance to ferroptosis [22]. These findings underscore the significance of SF3B4 in HCC progression and its potential as a therapeutic target in the diagnosis and treatment of the disease. Moreover, SF3B4 binds to TRIM28 and regulates its expression. The role of TRIM28 in cancer has been a subject of debate for over 15 years due to its complex, dual functions [23]. High expression levels of TRIM28 have been implicated in cancer progression by promoting epithelial-mesenchymal transition, metabolic reprogramming, and suppression of p53, as well as by inhibiting pro-apoptotic genes, thus supporting cancer cell survival [24]. In the context of alternative splicing, SF3B4 also binds to SETD5 and regulates its splicing events. SETD5 has been shown to drive resistance to MEK1/2 inhibitors in pancreatic ductal adenocarcinoma (PDAC), and its deletion restores sensitivity to MEK1/2 inhibition. SETD5 forms a co-repressor complex that regulates drug resistance pathways, and co-targeting MEK1/2, HDAC3, and G9a effectively suppresses PDAC growth in vivo [25]. Given these insights, the interaction between SF3B4 and SETD5 warrants further investigation.

Despite the critical role SF3B4 plays in HCC progression, research on its function remains incomplete. This study further elucidates the role of SF3B4 in HCC through analysis of variable splicing and RNA-seq data, suggesting that SF3B4 may serve as an important target for diagnostic and therapeutic strategies in HCC. These findings highlight SF3B4 as a key regulator of gene expression and alternative splicing in HCC, providing potential therapeutic insights. However, this study is limited to in vitro experiments, and further in vivo validation is necessary to confirm the clinical relevance of SF3B4 in HCC progression.

Conclusion

The progression of HCC is closely associated with the aberrant splicing of tumor-related genes, with SF3B4 emerging as a critical factor in promoting cell proliferation and survival. Downregulation of SF3B4 leads to reduced proliferation and increased apoptosis in HCC cells, emphasizing its pivotal role in regulating the alternative splicing of genes such as TIA1, KTN1, and FOXP2. This study underscores the potential of SF3B4, through its regulation of TRIM28 and SETD5, as a promising target for both diagnostic and therapeutic approaches in HCC. Nevertheless, some gaps remain in our understanding. Further studies are needed to elucidate the precise molecular mechanisms underlying SF3B4-mediated regulation, conduct comparative analyses across different cancer types, and, most importantly, perform in vivo validation to confirm its physiological and clinical relevance.