METTL3 and METTL14 drive pancreatic adenocarcinoma progression via m6A-dependent alternative splicing in PANC-1 cells

Abstract

METTL3 and METTL14, key components of the m6A methyltransferase complex, have been extensively studied in the context of various cancers. However, their roles in regulating alternative splicing in pancreatic cancer remain largely unexplored. In this study, we analyzed high-throughput RNA-seq data (GSE146806), alongside CLIP-seq datasets (GSE132306 and SRP163326), to investigate the molecular mechanisms underlying METTL3_14-mediated regulation. The results of differentially expressed genes (DEG) showed that that METTL3_14 knockdown significantly altered the expression of genes associated with canonical tumor-related pathways. Alternative splicing analysis identified METTL3_14-regulated alternatively spliced genes (RASGs) enriched in pathways such as protein processing in the endoplasmic reticulum, the spliceosome, and HIF-1 signaling, which are closely related to the progression of pancreatic cancer. Reanalysis of CLIP-seq data showed that METTL3_14 preferentially binds to the 3′UTR and coding sequences (CDS) regions. A total of 17 genes overlapped between DEGs and m6A-modified genes, while 59 genes overlapped with RASGs. Validation by qPCR confirmed significant regulatory effects on seven genes, especially EIF4A2, whose splicing was directly modulated by METTL3_14. These findings indicate that METTL3_14 may regulate gene expression and alternative splicing via m6A modifications in pancreatic cancer, particularly in pathways involved in transcription and DNA damage repair. This study provides novel mechanistic insights and potential targets for clinical diagnosis and treatment of pancreatic cancer.

Supplementary Information

The online version contains supplementary material available at 10.1007/s12672-025-03393-3.

Introduction

Pancreatic cancer is one of the most malignant digestive tract tumors, with an overall five-year survival rate of only 11% [1]. Early symptoms of pancreatic cancer, such as jaundice and weight loss, are often nonspecific and may progress indolently, contributing to delayed clinical recognition [2]. By the time symptoms become diagnostically actionable, the disease is frequently at an advanced stage, leaving a narrow therapeutic window for curative surgical resection [3]. Moreover, pancreatic cancer demonstrates limited sensitivity to chemotherapy and poor responsiveness to both targeted therapies and immunotherapy [4]. Therefore, in-depth exploration of the mechanisms of pancreatic cancer and the search for therapeutic targets are of great significance for improving the prognosis of pancreatic cancer patients.

In recent years, post-transcriptional regulatory events, such as RNA methylation, have attracted increasing attention from researchers and have become a new mechanism for regulating tumorigenesis. N6-methyladenosine (m6A) is the most common internal RNA modification in mammalian messenger RNA [5, 6]. m6A is transferred from the sixth nitrogen atom on adenine by the methyltransferase complex METTL3, METTL14, and WTAP, and may also be demethylated by the oxidases FTO and ALKBH5 [7]. m6A-modified RNA is specifically recognized by YTHDF1, YTHDF2, and YTHDF3 proteins in cells, achieving regulation of downstream RNA functions [8]. In mammals, N6-methyladenosine modification participates in multiple physiological processes, such as regulating the self-renewal of mouse embryonic stem cells and the circadian clock [9]. Studies have shown that N6-methyladenosine can regulate processes such as mRNA splicing, nuclear export, protein translation, and degradation [9]. At present, research on the role of m6A in the development and progression of pancreatic cancer [10, 11], as well as its significance in pancreatic cancer treatment, is still in its infancy and needs further exploration.

Additionally, m6A variation not only affects gene expression by altering RNA stability and mRNA export but also regulates alternative splicing (AS) and 3’ end processing [12]. AS is a dynamic process coupled with transcription, involved in many physiological functions and disease pathogenesis [13]. METTL3 and METTL14, as m6A writers, mediate RNA methylation. Numerous studies have shown that METTL3 and METTL14 are highly expressed in pancreatic cancer and promote tumor growth and invasion [14–18]. Accumulating evidence suggests that METTL3 deficiency can alleviate lipopolysaccharide-induced inflammatory responses by altering the expression pattern of MyD88 isoforms [19]. METTL3 regulates cancer-associated alternative splicing switches including breast cancer [20], prostate cancer [21], glioma [22] and gastrointestinal cancer [23]. However, how m6A methylation-related AS regulated by METTL3 and METTL14 affects pancreatic cancer remains unclear.

In this study, we identified AS features associated with pancreatic cancer by knocking out METTL3 and METTL14 in pancreatic cancer cells, reducing m6A levels. Clarifying how m6A methylation-related AS events function in pancreatic cancer provides new potential targets for future treatment and intervention.

Materials and methods

Cell culture

PANC-1 cell line was obtained from Chinese Academy of Sciences Kunming cell bank. Cells were maintained in RPMI-1640 medium (Invitrogen, 11875093) supplemented with 10% FBS (Invitrogen, 10091148) and 1% PS. All cells were incubated at 37 °C and 5% CO2.

SiRNA transfection, RNA extraction, and polya RNA isolation

The siRNA against METTL3 and METTL14 (Invitrogen) or control was transfected into PANC-1 cells using 7.5 µL RNAiMAX reagent (Invitrogen, 13778150). Total RNA was extracted with a Pure-link mini kit (Invitrogen, 12183018 A) 48 h after transfection. RNA concentration was measured via a Nanodrop 2000, and RNA integrity (RIN) was detected using an RNA 6000 nanochip (Agilent Technologies) (Requirement: RIN > 8.8). PolyA RNA was isolated from 75 µg of total RNA using a Dynabeads™ kit.

Real-time quantitative PCR and western blotting

cDNA transcription was performed using a SuperScript III synthesis kit (Invitrogen) and then used for qPCR based on SYBR (Applied Biosystems). The primer were METTL3, F: CTTGCATGGATTCTGAGGCC and R: GTCAGCCATCACAACTGCAA; GAPDH, F: TCAAGAAGGTGGTGAAGCAGG and R: TCAAAGGTGGAGGAGTGGGT siRNA pool against METTL3 and METTL14 were designed. The amplification profile was as follows: denaturing at 95 °C for 5 min, followed by 40 cycles of 95 °C for 15 s and 60 °C for 60 s. All samples were run in duplicate, including blank controls without cDNA. The relative expression levels of genes were calculated using the ΔΔCt method and normalized with GAPDH (Glyceraldehyde-3-phosphate dehydrogenase). For qPCR validation, Total RNA were extracted, reverse transcription and qPCR following above protocol. The primer sequences can be found in the supplementary Table 1. We observed that the knockdown efficiency of siMETTL3_14 was more than 80% for further analyses.

The siRNA-transfected PANC-1 cells were lysed in RIPA lysis buffer and Western blotting was conducted as previously. Then, the intensity of METTL3 (1:500 dilution, Proteintech, 15073-1-AP) and GAPDH (1:2000 dilution, Santa Cruz, sc-25778) bands was quantified using ImageJ software and GAPDH was used as the housekeeping marker.

RNA-Seq data processing

Paired-end sequencing (2 × 150 bp) of siNC- and siMETTL3-14-transfected cells was performed on a HiSeq 4000 sequencing platform (Illumina). The raw reads obtained after library sequencing were processed for alignment. Quality check (QC) was performed on raw reads (paired end fastq files) using NGS QC Toolkit. In brief, adaptor contaminated reads were removed and high quality reads with 70% high quality bases of more than or equal to 20 phred score were filtered by discarding low quality reads. Clean reads were aligned to the human hg38 genome (UCSC) using HISAT2 (v2.2.1) software. The average read coverage per sample after alignment was more than 40 million reads, with > 97% mapped. Genes with counts per million (CPM) < 1 in more than half the samples were excluded from downstream differential expression analysis to ensure robustness. PCA of three replicates was executed. Differential gene expression (DEG) analysis was conducted using DESeq2. Gene Ontology (GO) and KEGG enrichment analyses were performed using the clusterProfiler R package.

Data processing and peak calling for m6A sequencing

Data was obtained from GEO under the accession number GSE132306 and SRP 163,326. m6A peak calling was performed using R package exomePeak. Homer v4.9.1 was used to search for the enriched motif in the m6A peak region where the random peaks of 200 bp on human transcriptome were used as background sequence for motif discovery. m6A peak distribution on the metagene was plotted by R package Guitar. We adjusted for multiple testing using the Benjamini-Hochberg FDR procedure.

Differential expression analysis

The input library of m6A sequencing is essentially an mRNA sequencing library. Thus, we performed gene level differential expression analysis using the input libraries. Aligned sequencing reads were quantified by Rsubread, to obtain a count matrix of gene counts per sample. R package DESeq2 was used to test for differential expression where sequencing batch, sex, and age were included as covariates. We used a less stringent cut-off P value < 0.01 to select differential genes for pathway and gene ontology enrichment analysis. ConsensusPathDB was used to perform enrichment analysis. For METTL3 and METTL14 knockdown in PANC-1 cell line, significantly differentially expressed genes were selected at an adjusted P value (FDR) cut-off of 0.10.

AS analysis

The seven AS events were alternate acceptors (AA), alternate donors (AD), exon skipping (ES), retained intron (RI), alternate promoters (AP), alternate terminators (AT), and mutually exclusive exons (ME). Each AS event detected by the PSI value in siMETTL3_14 vs. siNC-treated cell was identified and counted using the software R-4.0.2. The corresponding p values were obtained using the t-test in the generalized linear model, and the difference multiplication relations were calculated to screen the AS with significant differences. Finally, an adjusted p value of < 0.05 was set as a criterion to screen the significantly differentially expressed AS (DEAS).

Gene Ontology (GO) analysis Gene ontology (GO) analysis was performed using the web tool The Database for Annotation, Visualization and Integrated Discovery (DAVID) (https://david.ncifcrf.gov/).

Motif analysis of spliced genes

The regions of differential intron retention and exon skipping events were extracted for motif analysis using HOMER (http://homer.ucsd.edu/homer/motif/) software. The resulting motifs were visualized using the ggseqlogo R package.

Statistical analysis

Differences in PSI values of differential intron retention and exon skipping events between groups were assessed using the Wilcoxon rank-sum test. The significance of overlaps between m6A sites and DEGs, spliced genes, or junction regions was determined using a hypergeometric test by R software (Version 4.0.2). The significance of candidate genes in qPCR experiments was determined by independent Student’s t-test. All P-values were two-sided, and values < 0.05 were considered statistically significant.

Results

Transcriptome-wide analysis revealed abnormal gene expression after knockdown METTL3-14 in PANC-1 cell line

To elucidate the molecular mechanisms of METTL3-14 in pancreatic cancer, we performed RNA-Seq on PANC-1 cells transfected with METTL3_14. First, qPCR and Western blotting confirmed that the knockdown efficiency of METTL3-14 exceeded 80% (Supplementary Fig. 1A-B). Differential gene expression analysis using DESeq2 revealed 942 differentially expressed genes (DEGs) between the siMETTL3-14 and control groups, including 736 upregulated and 206 downregulated genes (adjusted p < 0.05; Fig. 1A–D). GO term enrichment analysis, demonstrated 736 upregulated DEGs were significantly enriched in intracellular signal transduction, extracellular matrix organization, negative regulation of canonical Wnt receptor signaling pathway, homophlic cell adhesion and canonical Wnt receptor signaling pathway (Fig. 1E). 206 downregulated DEGs was significantly enriched in extracellular matrix organization, wound healing, integrin-mediated signaling pathway, cell adhesion, signal transduction and leukocyte migration (Fig. 1F). KEGG pathway analysis indicated that upregulated DEGs were predominantly enriched in pathways such as basal cell carcinoma and melanogenesis, whereas downregulated DEGs were most significantly associated with the NF-κB signaling pathway and ECM-receptor interaction (Fig. 1G–H).

Fig. 1.

Transcriptome-wide analysis revealed abnormal gene expression after knockdown METTL3-14 in PANC-1 cell line. (A) Hierarchical clustering heat map showing correlation between METTL3-14-knockdown (siMETTL3_14 rep1/2/3) and siC (siC rep1/2/3) samples based on FPKM value of all expression genes. (B) Volcano plot showing all differential expressed genes (DEGs) between METTL3_14-knockdown and siC samples using DESeq2. FDR ≤ 0.05 and FC (fold change) ≥ 2 or ≤ 0.5. (C) Principal component analysis (PCA) of siMETTL3_14 and siC samples in PANC-1 cell line based on FPKM value of all Differentially expressed genes (DEGs). The ellipse for each group is the confidence ellipse. (D) Heatmap showing expression profile of all significant DEGs between METTL3_14-knockdown and siC samples. (E) Bar plot showing the top10 most enriched GO biological process terms of the up-regulated DEGs. (F) Bar plot showing the top10 most enriched GO biological process terms of the down-regulated DEGs. (G) Bar plot exhibited the most enriched KEGG pathways results of the up-regulated DEGs. (H) Bar plot exhibited the most enriched KEGG pathways results of the down-regulated DEGs

Regulated alternative splicing events (RASEs) and genes expressed in METTL3-14-knockdown PANC-1 cell line

We analyzed all RASEs between siNC and siMETTL3-14 PANC-1 cell line (Fig. 2 and Supplementary Fig. 2). Among 2105 RASEs found between siNC and siMETTL3-14 PANC-1 cell line, A5SS (alternative 5’ splice site) and A3SS (alternative 3’ splice site) were the most frequently reported (Fig. 2A). A PCA of the 3 siMETTL3-14 samples and 3 siNC samples based on the percent spliced in (PSI) value of all differential nonintron retention (NIR) events shows that the siMETTL3_14 group can be separated from the siNC group (Fig. 2B). A PSI heatmap showing all NIR-RAS events among the siMETTL3-14 samples and siNC samples showed obvious differences (Fig. 2C). Compared with the siNC, GO analysis demonstrated that all regulated alternative splicing genes (RASGs) were mainly enriched in DNA damage and repair pathways, such as transcription, DNA-dependent, DNA repair, chromatin modification, mitotic cell cycle, response to DNA damage stimulus (Fig. 2D). We conducted functional enrichment analysis on genes associated with significantly altered non-intron retention (NIR) alternative splicing events, and found that these genes were predominantly enriched in pathways related to pyrimidine metabolism, adherens junction, ubiquitin mediated proteolysis, and citrate cycle (Fig. 2E). It suggests that alternative splicing of genes may play a role in the progression of pancreatic cancer.

Fig. 2.

The occurrence of variable splicing events with significant differences suggested the regulatory role of METTL3_14 in PANC-1 cell line. (A) The bar plot showing the number of all significant regulated alternative splicing events (RASEs). X-axis: RASE number. Y-axis: the different types of AS events. (B) Principal component analysis (PCA) of siMETTL3_14 and siC samples in PANC-1 cell line based on FPKM value of all significant NIR RASEs based on PSI. AS filtered should have detectable splice junctions in all samples and at least 80% samples should have > = 10 splice junction supporting reads. The ellipse for each group is the confidence ellipse. (C) Hierarchical clustering heatmap of all significant NIR RASEs based on PSI. AS filtered should have detectable splice junctions in all samples and at least 80% samples should have > = 10 splice junction supporting reads. (D) Bar plot exhibited the most enriched GO biological process results of the NIR regulated alternative splicing genes (RASGs). (E) Venn diagram showing the overlap genes number of NIR RASGs and DEGs

Peak analysis of m6A methylation in METTL3 knockout

Given that METTL3 and METTL14 function as core components of the m⁶A methyltransferase complex, we investigated whether the identified RASEs (regulated alternative splicing events) are associated with m⁶A modifications. We analyzed the sequencing data of CLIP-seq from GSE132306 and SRP 163,326. Peak calling identified 2,587 and 907 METTL3/METTL14-binding sites in the respective datasets. Upon merging the overlapping peaks, we identified 765 consensus binding sites detected in both datasets (Fig. 3A). Genomic distribution analysis of the overlapped peaks revealed they were mainly in CDS regions, 3’UTR and intron (Fig. 3B). Enrichment analysis of m6A-overlap Peak and input reads around translation start and stop codons and TSS and TTS (Fig. 3C). Motif analysis of the overlapped peaks revealed they were enriched in CU-rich elements (Fig. 3D). Functional enrichment analysis of the overlapped peak genes showed the top 10 terms/pathways in Fig. 3E, including gene expression, cellular protein metabolic process, nuclear mRNA splicing, via spliceosome, RNA splicing, translation (Fig. 3E). Furthermore, KEGG pathway analysis showed significant enrichment in pathways related to protein processing in the endoplasmic reticulum, spliceosome assembly, RNA transport, and the HIF-1 signaling pathway (Fig. 3F).

Fig. 3.

Peak analysis of m6A methylation in METTL3 knockout. (A) Venn diagram shows the overlap of two sets of data m6A peak. (B) Barplot shows the distribution proportion of overlap peak. (C) Enrichment analysis of m 6 A -overlap Peak and input reads around translation start and stop codons and TSS and TTS.Reads count were normalized in a ± 1000 bp window around translation start and stop codons and TSS and TTS. (D) Motif analysis showing the top 3 preferred bound motifs of METTL3_14 using HOMER software. (E) Bar plot exhibited the most enriched GO biological process results of METTL3_14 overlap Peak genes. (F) Bar plot exhibited the most enriched KEGG pathways process results of the overlap peak genes

METTL3_14 regulated AS of pancreatic cancer related genes by binding directly to RNA

Based the hypothesis that METTL3_14 regulates alternative splicing by directly binding to the RNA transcripts, we made an extensive analysis between siMETTL3_14 RNA-seq and CLIP-seq datasets. We first used non-IR region to analyze the overlapped genes between METTL3_14-regulated DEGs and RASGs in RNA-seq, and obtained 115 such genes, showing significant interaction between METTL3_14-bound genes and RASGs (Fig. 4A). Functional enrichment analysis of these co-regulated genes revealed significant enrichment in cell junction assembly, cellular component movement, gene expression, RNA splicing and other pathways (Fig. 4B, top 10 pathways by significance). Additionally, using region-level analysis, we identified 27 splicing events were enriched in apoptotic process (Fig. 4C-D). Notably, strong METTL3_14 binding signals were observed within the genomic loci of EIF4A2, RFC4, and SNORA genomic locus, particularly within exon regions (Fig. 4E, green rectangular frame). Alternative splicing analysis revealed an increased intron retention (IR) event in METTL3_14-KD PANC-1 cells (Fig. 4E). It suggests that METTL3_14 may modulate alternative splicing via direct RNA binding and m6A-dependent mechanisms.

Fig. 4.

METTL3_14 regulated AS of pancreatic cancer related genes by binding directly to RNA. (A) Venn diagram showing the overlap of Fig. 3 overlap peaks and METTL3_14-regulated alternative splicing events (use non-IR region). (B) Bar plot exhibited the most enriched GO biological process results of genes which the overlap peaks (showed in A) located (use Symbol). (C) Venn diagram showing the overlap of Fig. 3 overlap peaks and METTL3_14-regulated NIR RASG (use Region). (D) Bar plot exhibited the most enriched GO biological process results of the overlapped genes in C. (E) IGV-sashimi plot showed the METTL3_14-regulated alternative splicing events across mRNA of EIF4A2. The transcript of the gene was plotted at the bottom of the graph

Overlap genes regulated by METTL3_14 in PANC-1 cells

We next investigated the relationship between previously identified DEGs and RASGs. A total of 17 genes overlapped between DEGs and m6A-modified genes (Fig. 5A and Table S2). KEGG pathway analysis showed that these co-regulated genes were enriched in vitamin digestion and absorption, fat digestion and absorption, ABC transporters, bile secretion pathway (Fig. 5B and supplementary Table 2). Furthermore, we observed 59 out of 942 (6.26%) alternatively spliced genes were identified in PANC-1 cells (Fig. 5C), with enrichment in apoptosis-multiple sepecies, apoptosis, glycosaminoglycan degradation, transcriptional misregulation in cancer, and fat digestion and absorption pathway (Fig. 5D). SCARB1 were among the identified genes. Notably, several genes, including FBXO7, ENTPD6, EIF4A2, DENND5A, RNH1, and PLD3, exhibited marked alternative splicing changes and are potential direct targets of METTL3. These observations suggest that METTL3 may regulate alternative splicing of these transcripts through direct binding and m6A methylation. Collectively, these data support a regulatory role for METTL3 in modulating gene expression and RNA splicing via m6A modifications in PANC-1 cells.

Fig. 5.

Overlap genes between DEG, RASG, and m6A related genes regulated by METTL3_14 in pancreatic cancer. (A) Venn diagram showing the overlap of Fig. 3 overlap peaks and DEG. (B) Bar plot exhibited the most enriched KEGG pathways process results of the overlap genes. (C) Venn diagram showing the overlap of RASG and DEG. (D) Bar plot exhibited the most enriched KEGG pathways process results of the overlap genes

qPCR validation on overlap genes regulated by METTL3_14 in PANC-1 cells

To validate the METTL3_14-regulated differentially expressed genes (DEGs) and alternatively spliced (AS) genes identified in our analysis, we conducted quantitative PCR (qPCR). Seven candidate genes involved in apoptosis and DNA damage repair pathways—namely EIF4A2, DENND5A, RNH1, ENTPD6, OGDH, NONO, and FBXO7—were selected for evaluation. The results showed EIF4A2 and RNH1 were significantly downregulated in siMETTL3_14 cells, whereas OGDH, NONO, and FBXO7 were significantly upregulated. No significant changes were observed in DENND5A and ENTPD6 expression between siNC and siMETTL3_14 groups (Fig. 6). These findings highlight a potential regulatory role of METTL3_14 in pancreatic cancer pathophysiology.

Fig. 6.

qPCR validation on genes regulated by METTL3_14 in PANC-1 cells. The shNC and three shMETTL3-14 were transfected into PANC-1 cell line. qPCR were performed to examine the mRNA levels of genes from apoptosis and DNA damage and repair pathway, including EIF4A2, DENND5A, RNH1, ENTPD6, OGDH, NONO, and FBXO7. * p < 0.05 vs. shNC

Discussion

Although the roles of METTL3 and METTL14 have been extensively studied in various cancers, their roles on alternative splicing (AS) transitions in pancreatic cancer remains less understood. In this study, we compared gene expression (GSE146806) and splicing events in PANC-1 cells before and after knocking out METTL3 and METTL14. We identified 942 differentially expressed genes, mainly enriched in cell adhesion and DNA damage repair. Our splicing analysis revealed that knockout of METTL3 and METTL14 increased cassette splicing events and exon skipping.

At the transcriptional level, METTL3_14 depletion increased cassette exon events and promoted exon skipping. RT-qPCR validation of seven randomly selected genes confirmed these gene expression changes, like EIF4A2 and RNH1—known for their roles in inhibiting mRNA translation and association with cancer stem cell properties [24, 25]. In KEGG analysis, METTL3_14 knockout altered cancer-associated pathways, including the Rap1, Hippo, Wnt, MAPK, and NF-kappa B pathways, consistent with prior studies in other cancers [26–30]. These findings indicates METTL3_14 as critical modulators of precursor mRNA splicing in PDAC.

Recent studies have provided important insights into the regulatory mechanisms of METTL3 and METTL14 in different cellular contexts that may also be relevant to our findings. For example, Kong et al. [31] reported that METTL3 mediates osteoblast apoptosis by modulating endoplasmic reticulum (ER) stress during LPS-induced inflammation. Similarly, Li et al. [32] demonstrated that METTL3 activates a PERK-eIF2α-dependent apoptotic pathway by targeting the ER degradation-related protein SEL1L in echinoderms. In another study, Cao et al. [33] showed that METTL14-mediated m6A modification is essential for maintaining ER homeostasis during liver regeneration. In preeclampsia, Chen et al. [34] found that METTL3 exacerbates ER stress through a YTHDF2-dependent mechanism by targeting TMBIM6. These studies collectively reveal that METTL3 and METTL14 regulate key processes—ranging from apoptosis and ER stress to tissue regeneration—by modulating mRNA stability and processing through m6A modifications. Although our study primarily focuses on how METTL3_14 influence alternative splicing in pancreatic cancer cells, these mechanisms align with our findings that m6A regulators exert multi-level control over gene expression. At the alternative splicing level, our genome-wide analysis of METTL3_14-regulated AS events in PANC-1 cells indicate that they primarily modulate splicing switches related to the cell cycle, transcription, and DNA damage repair. These pathways may be driven by cellular stress responses. These results suggest that METTL3 or METTL14 regulates endoplasmic reticulum-related AS events in pancreatic cancer.

Further investigation revealed that only 9–15% of AS-affected genes overlap with the DEGs, and most of these are involved in isoform switching. This suggests that while AS dysregulation might not broadly change overall gene expression, it does modify functionally tailored transcriptome. Interestingly, the pathways enriched among AS genes (such as DNA repair and response to DNA damage) overlap with those identified in the DEG analysis. This implies that METTL3 and METTL14 co-regulate pancreatic cancer progression through multiple levels of gene regulation. Our study also showed that METTL3_14 knockdown resulted in increased intron retention and exon skipping. The impact of m6A on AS appears to vary by cell type and transcript, as some studies report that m6A promotes exon inclusion while others show that depletion of METTL3 leads to exon skipping [20, 35, 36]. Specifically, METTL3_14 was found to mainly affect the AS of genes related to translation initiation, such as EIF4A2 and RFC4. EIF4A, as translation initiation factor, has been implicated in malignant transformation. Pro-oncogenic mRNAs typically have longer and more structured 5’-UTRs, which require high levels of EIF4A helicase activity for efficient translation [37]. Wu et al. reported an alternative splicing isoform of EIF4H plays a significant role in carcinogenesis [38]. In our study, we observed that a cassette exon in EIF4A2 was deleted. Research conducted by Yu et al. delved into the relationship between methylation, alternative splicing and RFC4 expression, and they identified differential methylation of RFC4 in various cancers [39]. These results indicate that METTL3-14 could simultaneously regulate expression and AS of genes involved in DNA damage/repair pathways.

There are limitations to our study. Short-read RNA-seq may overlook low-abundance or repetitive-region isoforms; long-read sequencing (Oxford Nanopore/PacBio) would enhance AS detection accuracy. The modest overlap between AS-affected genes and DEGs suggests that while AS regulation affects specific transcript variants, its overall impact on gene expression may be limited. Additionally, the functional roles of many dysregulated genes and AS events in pancreatic cancer remain unconfirmed. Future work should focus on elucidating the detailed molecular mechanisms by which METTL3_14 regulate AS and contribute to pancreatic cancer progression. For instance, investigating the interplay between m6A modifications and specific splicing factors could reveal novel therapeutic targets. Expanding these works to clinical samples and animal models will be crucial for understand the complex regulatory networks governed by METTL3_14 in pancreatic cancer.

Conclusion

In conclusion, our study, through integrated data analysis of METTL3_14-knockdown RNA-seq and METTL3_14 CLIP-seq, suggests that METTL3_14 may be involved in the molecular mechanisms of pancreatic cancer not only by affecting the expression of genes involved in pancreatic cancer progression but also by binding and regulating the alternative splicing of genes related to transcription and DNA damage repair in pancreatic cancer cells. These findings provide new insights into RNA modifications in pancreatic cancer.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

Qiuhong Wang, Dongyun Cun and Tao Wu designed this study. Bo Tang, Lianmin Wang, Dong Wei and Tao Wang performed experiments. Renchao Zou and Kun Su analyzed the data. Dong Wei, Bo Tang and Qiuhong Wang wrote the first draft. Qiuhong Wang, Dongyun Cun and Tao Wu revised and finalized the manuscript. All authors have given the final approval of the version to be published and agreement to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

Data availability

The data is available from the corresponding authors upon reasonable request. The datasets generated and/or analysed during the current study are available in the GEO database repository, GSE146806, GSE132306 and SRP 163326.

Declarations

Ethics approval and consent to participate

This study does not involve ethical practices.

Consent for publication

All authors have read and agreed to the published version of the manuscript.

Competing interests

The authors declare no competing interests.