METTL3-driven m⁶A epigenetics in gastric cancer: unveiling oncogenic networks and clinical translation from tumorigenesis to therapy resistance.

Background
Gastric cancer (GC) ranks among the deadliest malignancies (Smyth et al. 2020; Li et al. 2023a, b, c, d). Although surgical and systemic treatments have advanced, outcomes for advanced GC are still poor, with 5-year survival below 30% (Smyth et al. 2020; Li et al. 2023a, b, c, d). This poor outcome stems from delayed diagnosis due to nonspecific early symptoms, a high propensity for metastasis, and intrinsic or acquired resistance to conventional therapies such as chemotherapy, targeted agents, and immunotherapy (Guan et al. 2023; Yasuda and Wang 2024, Matsuoka, and Yashiro, 2023). While molecular profiling has identified key drivers like human epidermal growth factor receptor 2 amplification, tumor protein 53 (p53) mutations, and microsatellite instability, these alterations fail to fully explain the heterogeneity and therapeutic recalcitrance of GC (Zhu et al. 2021; Karlsson et al. 2023; Yeoh and Tan 2022). Consequently, uncovering novel molecular mechanisms governing GC progression is imperative for developing effective diagnostic and therapeutic strategies.
Emerging research has unveiled the pivotal role of epigenetic dysregulation in cancer biology, with RNA modifications emerging as critical post-transcriptional regulators of oncogenic pathways (Su et al. 2024b). Among these, N⁶-methyladenosine (m⁶A), the most abundant internal modification in eukaryotic mRNA, dynamically modulates RNA metabolism, including splicing, stability, translation, and decay, through a sophisticated interplay of"writers"(methyltransferases),"erasers"(demethylases), and"readers"(m⁶A-binding proteins) (Dawson and Kouzarides 2012; Hogg et al. 2020; Su et al. 2024a, b, c, d, e). Methyltransferase-like 3 (METTL3), the catalytic core of the m⁶A writers, has garnered significant attention for its context-dependent roles in tumorigenesis (You et al. 2022; Wen et al. 2023; Su et al. 2024a, b). While METTL3 exhibits tumor-suppressive functions in certain malignancies, accumulating evidence positions it as a potent oncogene in GC, driving aggressive phenotypes via m⁶A-mediated reprogramming of RNA networks (Su et al. 2023; Meng et al. 2023a, b; Xu and Ge 2022) (Graphical Abstract).
In GC, METTL3 is frequently overexpressed and correlates with advanced tumor stage, metastasis, and poor survival. Mechanistically, METTL3 mediates m⁶A methylation across RNA species (mRNAs, ncRNAs, circRNAs), reprogramming oncogenic networks, metabolic flux, and TME dynamics (Zeng et al. 2020). For instance, METTL3-mediated m⁶A modification stabilizes homeobox A10 (HOXA10) transcripts, activating the TGF-β/Smad axis to promote epithelial-mesenchymal transition (EMT) (Song and Zhou 2021). Concurrently, METTL3 enhances glycolysis by regulating insulin-like growth factor 2 mRNA-binding protein 3 (IGF2BP3)- dependent stabilization of hepatoma-derived growth factor (HDGF) mRNA, linking epitranscriptomic changes to metabolic adaptation (Wang et al. 2020). Furthermore, METTL3 fosters chemoresistance by modulating N6-methyladenosine RNA binding protein 1(YTHDF1)-mediated translation of poly (ADP-ribose) polymerase 1 (PARP1), a key DNA repair enzyme (Gonzalez-Leal et al. 2025). Despite these advances, a systematic framework integrating METTL3’s multi-dimensional roles in GC pathogenesis and therapy resistance remains lacking (Graphical Abstract).
Current research gaps include: (i) insufficient understanding of METTL3’s context-specific interactions with ncRNAs and metabolic regulators; (ii) limited exploration of METTL3’s immunomodulatory effects and synergy with immune checkpoint inhibitors; and (iii) unresolved challenges in translating METTL3-targeted strategies into clinical practice. This review bridges these gaps by synthesizing recent discoveries into a unified model of METTL3-driven GC progression, emphasizing its diagnostic, prognostic, and therapeutic relevance. By elucidating METTL3’s cross-talk with oncogenic signaling cascades and the TME, we aim to catalyze the development of m⁶A-based precision therapies for GC (Graphical Abstract).

METTL3-mediated m6A epigenetic remodeling: a central driver in gastric carcinogenesis and therapeutic resistance
GC is often late when you finally perceive its existence, and people are caught off guard (Norwood et al. 2022). What is more worrying being that the speed of GC progression is like a speeding train, and once diagnosed, the window for treatment is rapidly shortened. GC is regulated by a complex process involving multigene participation at multiple stages (Norwood et al. 2022). Etiological contributors such as Helicobacter pylori infection constituting the predominant risk factors associated with GC incidence (Norwood et al. 2022). Helicobacter pylori infection increases the m6A level in GC cells (Li et al. 2023a, b, c, d). Specifically, Helicobacter pylori infection extensively upregulates m6A regulators (METTL3, WTAP, FTO, ALKBH5) and downregulates key genes (PTPN14, ADAMTS1), linking m6A dysregulation to H. pylori-associated GC pathogenesis. Importantly, current studies indicate that the abnormal expression of METTL3, which operates as both an oncogene and a target gene, is involved in the onset and malignancies of GC.

METTL3 orchestrates multilayered oncogenic networks in GC pathogenesis
METTL3-mediated m6A modification, in conjunction with its interaction with m6A readers, is recognized as a crucial player in RNA processing, particularly in the regulation of mRNA dynamics, including their expression, stability, and eventual degradation, which is actively involved in GC development (Ding et al. 2023).

METTL3 as a bifunctional oncogenic hub

Progression driver
A comprehensive dataset analysis conducted by Liu et al. (2020) demonstrates a marked upregulation of METTL3 within GC tissues, in sharp contrast to the levels detected within healthy counterparts. Generally, METTL3 acts as the predominant m6A catalytic enzyme in the nucleus (Bedi et al. 2023). Notably, METTL3 also exhibits high cytoplasmic expression, which clinical studies have linked closely to GC progression (Wei et al. 2022). Their research uncovered an m6A-independent regulatory mechanism: cytoplasmic METTL3 associates with poly(A)-binding protein cytoplasmic 1 (PABPC1), strengthening the latter's interaction with the cap-binding complex eIF4F. This interaction selectively enhances the translation of oncogenic epigenetic factors (e.g., HDACs, DNMTs) without relying on m6A modification (Table 1). Importantly, such cytoplasmic METTL3-driven translational upregulation of epigenetic regulators contributes to carcinogenesis, highlighting its potential as a diagnostic and prognostic marker for GC patients.

The expression level of METTL3 could be regulated in various manners. Transcription factors (TFs) are a class of sequence‐specific DNA‐binding proteins and are also an important class of proteins responsible for gene expression (Kawasaki and Fukaya 2023). Activating transcription factor 2 (ATF2), one member of the basic leucine zipper TF family, plays a role in chromatin remodeling, DNA damage response, transcriptional regulation, as well as tumor progression (Huebner et al. 2019). An elevated level of ATF2 is found in GC tissues and cell lines, which promotes GC cell growth and inhibits cell apoptosis, contributing to the short survival time in patients with GC. Meng et al. (2023a, b) found that ATF2 binds to METTL3’s promoter and enhances METTL3’s transcription. METTL3 in turn targets and increases cyclin D1 (CCD1) expression to promote cell cycle progression. These findings suggest that ATF2 is an upstream regulator for METTL3 expression and facilitates GC progression via activating the METTL3/CCD1 signaling pathway. Therefore, the oncogenic role of METTL3 is promoted by ATF2, suggesting intervention of METTL3 might be an anti‐tumor method for GC. In line with this, HOXA10 plays an essential role in tumor progression (Song and Zhou 2021). The importance of HOXA10 becomes evident in its strategic localization within the promoter region of transforming growth factor β2 (TGF-β2), stimulating its transcription and enhancing its secretion. HOXA10 overexpression upregulates METTL3 and global m⁶A levels through activating the TGF-β2/Smad cascade, with Smad proteins critically mediating METTL3 expression. Both HOXA10 and METTL3 hold clinical significance; notably, METTL3 underlies HOXA10-driven EMT progression (Song and Zhou 2021). These findings collectively underscore the essential role of the HOXA10-TGFβ2/Smad-METTL3 axis in promoting GC progression (Fig. 1, Table 1).
The escalation of METTL3 expression exhibits a positive correlation with the progression of tumor stage and grade, along with poor survival outcomes (Liu et al. 2020). Consequently, METTL3 emerges as an unfavorable prognostic factor in patients with GC. Further, METTL3 deletion reduces m6A modification on Gfi1 mRNA, decreasing its stability and downstream α-SMA expression, which suppresses EMT-driven proliferation and migration (Liu et al. 2020). Crucially, Gfi1 knockout in METTL3-WT cells phenocopied METTL3 loss, while Gfi1 overexpression rescued METTL3-deficiency effects (Liu et al. 2020). This intricate molecular cascade imparts constraints upon cellular proliferation and migration capacity. These findings collectively reveal that METTL3 acts as an oncogenic driver in GC by regulating Gfi1 expression (Fig. 1,Table 1).
Subsequent studies have further clarified METTL3's promotive role in regulating GC cell proliferation and migration, which operates via m6A modification of tumor-related genes—including YAP1 (Zhou et al. 2021), SOCS2 (Jiang et al. 2020), and basic leucine zipper ATF-like transcription factor 2 (BATF2) mRNA (Xie et al. 2020). BATF2 is upregulated in GC tissues and correlates with tumor progression. Consistently, Xie et al. (2020) revealed that BATF2 stabilizes p53 while inhibiting ERK phosphorylation, thereby driving GC progression through the METTL3-BATF2/p53-ERK axis. BATF2 knockdown abolished METTL3-driven proliferation, confirming BATF2 as a functional effector (Xie et al. 2020). Consistently, a substantial upregulation of hepatitis B X-interacting protein (HBXIP) relates to GC malignancies. Yang et al. (2020a, b) demonstrated that METTL3, as a target of HBXIP, upregulates MYC and its target glycolytic genes (MCM5, MCM6), linking m6A modification to metabolic reprogramming. Similarly, an elevated level of sphingosine kinase 2 (SPHK2) facilitates the malignancy properties of GC cells, indicating an unfavorable outcome (Huo et al. 2021). In elucidating this intricate interplay, they demonstrated that METTL3 coordinates with YTHDF1 to facilitate SPHK2 translation, which subsequently promotes KLF transcription factor 2 (KLF2) phosphorylation. The resulting ubiquitination and degradation of the KLF2 protein ultimately inhibits KLF2 expression in GC, thereby amplifying the manifestation of these aggressive traits. Similarly, Huo et al. (2021) reported that METTL3/YTHDF1 enhance SPHK2 translation, which phosphorylates and degrades KLF2, promoting proliferation and migration. In METTL3-knockout GC cells, re-expression of SPHK2 (via lentiviral transduction) restored proliferation and migration, confirming SPHK2 as the functional effector of METTL3 (Huo et al. 2021). Therefore, METTL3 functions by targeting the SPHK2/KLF2 axis. (Fig. 1, Table 1).
During tumor progression, the DNA damage repair mechanism can be reprogrammed by influencing the genomic integrity (Kiwerska and Szyfter 2019). A high level of suppressor of variegation 3–9 homolog 2 (SUV39H2) is detected in GC tissues and cell lines, which promotes proliferation and inhibits apoptosis in GC cells (Yang et al. 2023). Moreover, SUV39H2 acts as a histone methyltransferase that induces ataxia-telangiectasia mutated phosphorylation (ATM) phosphorylation by inhibiting DUSP6 transcription, which finally promotes homologous recombination and inhibits GC chemo-sensitivity to Cisplatin. Yang et al. (2023) found that METTL3/IGF2BP2-dependent m6A modification on SUV39H2 mRNA enhances SUV39H2 mRNA stability and upregulates SUV39H2 expression. Crucially, SUV39H2 knockout in GC cell lines reversed METTL3-driven chemoresistance, confirming SUV39H2 as the functional mediator of METTL3 in cisplatin resistance (Yang et al. 2023). Notably, METTL3 promotes homologous recombination and impairs the chemo-sensitivity of GC to Cisplatin by regulating SUV39H2 (Fig. 1, Table 1). Therefore, METTL3 with the expectation to be used as a target for GC precision therapy.

Metastatic switch
The elevated METTL3 expression shows a close correlation with an unfavorable prognosis in patients with GC (Okugawa et al. 2023; Ge et al. 2023; Yue et al. 2019). Yue et al. (2019) found an underlying mechanism wherein METTL3 enhances the stability of ZMYM1 in a HuR-dependent m6A modification pathway, which recruits the CtBP/LSD1/CoREST complex to repress the E-cadherin promoter and facilitate EMT and metastasis. Additionally, Wang et al. (2020) reported an intricate regulatory mechanism, whereby P300 activates METTL3 transcription by acetylating its promoter region with H3K27. Further, the activated METTL3, collaborates with IGF2BP3 recognition and combination on the modified site of the HDGF mRNA directly, resulting in an m6A modification on HDGF mRNA to promote its protein stability. Owing to the METTL3/IGF2BP3-mediated m6A modification on HDGF mRNA stability, HDGF protein production is increased in the cytoplasm, thereby promoting tumor angiogenesis. Concurrently, nuclear HDGF activates the expression of GLUT4 and ENO2, consequently promoting glycolysis in GC cells (Wang et al. 2020). In vivo, METTL3-knockdown xenografts showed reduced lung metastasis, which was reversed by HDGF overexpression (Wang et al. 2020). This process is closely associated with tumor growth and liver metastasis. Consequently, these findings suggest that elevated levels of METTL3 function as oncogenes that facilitate the development of GC by targeting IGF2BP3/HDGF or promote glycolysis by targeting the IGF2BP3/HDGF-GLUT4 and IGF2BP3/HDGF-ENO2 axes. In line with this, PLAGL2 is a zinc finger protein transcription factor that activates downstream targets, which is abnormally expressed in GC and enhances the proliferation and metastasis of GC. Chen et al. (2023) demonstrated that PLAGL2 binds to the upstream promoter of UCA1, which subsequently sponges miR-145-5p that targets YTHDF1. Moreover, YTHDF1 recognizes METTL3-mediated m6A on Snail by interacting with eEF-2 and thus promotes Snail expression, which eventually induces EMT and metastasis. Overall, PLAGL2 enhances GC progression via targeting Snail indirectly regulated by METTL3, suggesting that PLAGL2 is an upstream regulator for METTL3 expression and METTL3 is a therapeutic target for GC (Fig. 1,Table 1).
PBX1 functions as a positive transcriptional regulator of GCH1, enhancing its expression. Liu et al. (2022a, b) demonstrated that METTL3 interacts with PBX1 to stabilize its mRNA, which in turn induces GCH1 transcription and expression. The METTL3-PBX1-GCH1 axis elevates tetrahydrobiopterin (BH4) levels in GC cells, thereby promoting GC growth and lung/lymph node metastasis. Additionally, METTL3 deletion impairs xenograft tumor growth and lung/lymph node metastasis in vivo. Furthermore, DEK mRNA exhibits higher m6A modification levels in GC tissues and cell lines, which correlates with GC lung metastases (Zhang et al. 2022a, b). Mechanistically, METTL3 targets the 3'-UTR of DEK mRNA to install m6A modifications, thereby boosting DEK stability. In DEK knockout mouse models, lung metastasis induced by METTL3 overexpression was completely abolished, demonstrating DEK's essential role in METTL3-mediated metastasis (Zhang et al. 2022a, b). This study hints that METTL3 is the regulator of DEK function in GC lung metastasis. Therefore, METTL3 is a therapeutic target for inhibiting GC lung metastasis (Fig. 1,Table 1).
Furthermore, TGF-β/Smad signaling shows dual capacity as a tumor inhibitor and driver (Hu et al. 2018). METTL3 synergizes with the TGF-β/Smad pathway to promote the growth and metastasis of GC cells. In GC, evidence also demonstrates that the TGF-β/Smad2/3 axis plays the driver role in tumorigenesis and metastasis (Yuan et al. 2023). Mechanically, Yuan et al. (2023) reported that METTL3/IGF2BP2-dependent m6A modification stabilizes Smad3 mRNA and enhances Smad3 protein expression, which in turn activates the TGF-β/Smad pathway. Importantly, METTL3 interacts with p-Smad3 to regulate the transcription of downstream genes. These data suggest a novel regulation mechanism for the cancer-promoting function of the TGF-β/Smad3 signaling, for which METTL3 serves as the manipulator and might be a new therapeutic target for GC treatment (Fig. 1,Table 1).
Collectively, METTL3 is the key contributor to GC growth and metastasis. Therefore, targeting METTL3 might be therapeutic for GC.

Metabolic reprogramming
A strong correlation has been established between the diminished expression of protein phosphatase 2 catalytic subunit alpha (PP2Acα) and the manifestation of GC. Cheng et al. (2021) demonstrated that inhibiting PP2Acα activates ATM activity, leading to the upregulation of METTL3 and promoting the aggressive behavior of GC. Jointly, METTL3 acts as the target of the oncogene, PP2Acα, which functions by regulating the ATM/METTL3 axis. Similarly, another study indicated that METTL3 overexpression considerably amplifies its oncogenic function by mediating m6A on the MYC target genes, including MCM5 and MCM6 (Yang et al. 2020a, b). Moreover, NDUFA4 is overexpressed in GC, with its high levels linked to poor patient prognosis. Xu et al. (2022a, b) clarified that the METTL3-IGF2BP1 complex enhances m6A modification of NDUFA4 mRNA, thereby upregulating NDUFA4 in GC cells. This upregulation accelerates glycolytic and oxidative metabolism, triggering abnormal cell proliferation and tumor growth. Notably, inhibiting mitochondrial fission can reverse NDUFA4-mediated metabolic changes and tumor progression in GC. NDUFA4 knockout in GC cells reversed METTL3-induced glycolytic flux and tumor growth in xenograft models, providing direct evidence of NDUFA4 as a metabolic executor of METTL3 (Xu et al. 2022a, b). This study suggests that the METTL3-NDUFA4 axis leads to GC development (Fig. 1,Table 1).
Collectively, METTL3, m6A modified-SPHK2, and m6A modified-NDUFA4 are risk factors for inducing GC malignant phenotype, suggesting their potential for GC treatment.

Tumor suppressor silencing via m6A-epigenetic crosstalk

ADAMTS9-PI3K/AKT suppression: YTHDF2-mediated mRNA decay
A low level of ADAMTS9 is considered an independent prognostic factor for GC. Wang et al. (2022a, b, c) suggested that METTL3/YTHDF2-mediated m6A modification on ADAMTS9 promotes ADAMTS9 degradation and reduces ADAMTS9 level in GC, which subsequently triggers PI3K/AKT pathway and enhances tumor cell proliferation and angiogenesis, thereby promoting GC progression. Notably, METTL3 promotes GC progression by inhibiting the anti-oncogene ADAMTS9 and regulating the ADAMTS9-PI3K/AKT pathway (Fig. 1).

Angiopoietin-like 3 (ANGPTL3) inactivation: m6A-dependent translational blockade
The Cancer Genome Atlas (TCGA) data found that METTL3 is frequently elevated in GC (Zhang et al. 2023). High expression level of METTL3 is more likely indicate advanced tumor node metastasis. On the other hand, METTL3 deletion effectively impedes the higher oncogenic capacity of GC, as reflected by slowed cell growth and diminished migration and invasion capacities. Importantly, the effect of METTL3 deletion is similar to ANGPTL3 enrichment that hinders the growth and metastasis of GC cells, whereas this effect is reversed partially by ANGPTL3 inhibition. Further exploring of the TCGA dataset found the co-expression of ANGPTL3 and METTL3 in GC. Specifically, METTL3-mediated m6A modification on ANGPTL3 mRNA decreases ANGPTL3 expression, which shortens the life span of patients with GC (Zhang et al. 2023). ANGPTL3 overexpression in METTL3-high GC cells significantly suppressed tumor growth in mouse xenografts (tumor volume reduced by 62%, p < 0.01), functionally validating ANGPTL3 as a tumor suppressor downstream of METTL3 (Zhang et al. 2023). Overall, these findings discover the METTL3-ANGPTL3 axis and its effect on GC malignant development, suggesting that METTL3 exerts the oncogenic role in GC by suppressing the anti-oncogene ANGPTL3 in an m6A-dependent manner, which contributes to the mechanism understanding of GC development (Fig. 1,Table 1).

SRSF11 splicing dysregulation: connects p53/apoptosis pathway impairment
A low level of SRSF11 is associated with poor survival in patients with GC, which is regulated by METTL3 (Oh et al. 2023). Specifically, METTL3 regulates SRSF11 mRNA splicing and lowers SRSF11 expression in patients with poor prognosis (Oh et al. 2023). According to gene set enrichment analysis, the downregulated level of SRSF11 relates to the pathways of p53/apoptosis, inflammation/immune response, and ultraviolet/reactive oxygen species stimulus–response in GC (Oh et al. 2023) (Fig. 1). These findings suggest that SRSF11-mediated poor prognosis in patients with GC relates to METTL3 functions. However, the exact signaling pathways are unclear and need to be clarified.

ncRNAs circuits amplify METTL3 oncogenicity

miRNA-m6A feedback loops
Elevated levels of m6A and METTL3 are detected in GC, demonstrating associations with unfavorable prognoses and increased malignancy (Sun et al. 2020a, b). Beyond its established role in regulating mRNA, METTL3 functions in GC development by regulating ncRNAs, such as miRNAs. Dysregulation of miRNAs is associated with the progression of numerous tumors.

OncomiR activation: DGCR8-dependent pri-miR-17–92 processing
A previous study reported that METTL3-mediated m6A modification on pri-miR-17–92 promotes miR-17–92 cluster formation through an m6A/DGCR8- dependent mechanism in GC (Sun et al. 2020a, b). Subsequently, the miR-17–92 cluster activates the AKT/mTOR pathway by targeting phosphatase and tensin homolog (PTEN) or TMEM127. Kang et al. (2021) demonstrated that the presence of miR-1269b is indicative of larger tumors and metastasis of lymph nodes in patients with GC, which correlates with its low expression levels in both GC tissues and cell lines. miR-1269b directly targets METTL3. Conversely, METTL3 overexpression silences miR-1269b, promoting GC progression (Kang et al. 2021). Collectively, these observations suggest that METTL3 promotes GC via interacting with miRNAs (Fig. 1).

Tumor suppressor miRNA inhibition
In consistent, the enhanced level of SEC62 is observed in GC cell lines, which promotes GC progression by augmenting cellular proliferation and mitigating apoptosis (Xia et al. 2020). Moreover, miR-4429 targets and amplifies METTL3 expression to further increase SEC62 mRNA stabilization via an IGF2BP1-dependent mechanism. Collectively, METTL3 acts as a target of miR-4429 to promote oncogene SEC62 expression and trigger GC progression, indicating that miR-4429 is a promising candidate for GC treatment. Additionally, embryonic ectoderm development proteins can effectively promote the proliferative and invasive potential of GC cells by inhibiting miR-338-5p through histone methylation (Zhang et al. 2021). For its mechanism, the current study indicates that miR-338-5p directly targets METTL3 and, in turn, upregulates METTL3 expression, facilitating CUB domain-containing protein 1 (CDCP1) translation through m6A modification and ultimately leading to enhanced proliferation and invasion of GC cells. Notably, the role of embryonic ectoderm development proteins in GC development depends on the interaction of miR-338-5p, METTL3, and CDCP1, indicating that targeting the miR-338-5p/METTL3-CDCP1 axis holds promise as a potential therapeutic strategy for GC.
Moreover, KLHL5 is a member of the kelch-repeat protein family and is overexpressed in GC, which correlates with the M stage and a shorter overall survival time in patients with GC. Li et al. (2023a, b, c, d) found that METTL3-mediated m6A modification upregulates KLHL5, which is simultaneously repressed by miR-181-5p, forming a competitive METTL3/miR-181-5p/KLHL5 axis that promotes GC progression (Fig. 1).

lncRNA-driven metastatic programs
Comprising a length exceeding 200 nucleotides and devoid of protein-coding capacity, lncRNAs constitute a multifaceted class with intricate functional dimensions (Mercer et al. 2009). LncRNAs often sponge miRNAs or interact with regulatory proteins to perform their biological functions. Anomalous lncRNA expression has been observed in human cancers, and some lncRNAs act as oncogenes or tumor suppressors to regulate the cellular biological processes involved in GC tumorigenesis and progression (Mercer et al. 2009).

SNHG7/miR-186-5p/CCND2: N-methyl-n'-nitro-n-nitrosoguanidine (MNNG)- induced malignant transformation
The underlying mechanism of lncRNAs in GC relates to METTL3. For example, MNNG exposure is closely associated with GC development, which alters cellular localization and might be the critical cellular note for malignant transformation (Gunes-Bayir et al. 2022). Based on the MeRIP-seq data of MNNG-induced malignant GC cells, Liu et al. (2023) found that lncRNA (SNHG7) is aberrantly elevated in GC triggered by MNNG exposure. Importantly, a high level of SNHG7 is detected in different malignant stages of GC and the elevated SNHG7 expression is also correlated with advanced clinical features and poor prognosis in patients with GC. Further analysis showed that METTL3-mediated m6A modification on SNHG7 enhances SNHG7 levels in GC cells (Liu et al. 2023). These findings collectively suggest that METTL3 is responsible for MNNG exposure-induced GC onset by targeting SNHG7, suggesting that METTL3 is a predictive biomarker or therapeutic target for GC. Additionally, Yan et al. (2020) demonstrated that the LINC00470 upregulation is closely associated with distant metastasis, advanced tumor node metastasis stage, and poor prognosis of GC, which functions by interacting with METTL3 to suppress the stability of PTEN mRNA. Furthermore, the degradation of PTEN mRNA, which is facilitated by the interaction of LINC00470 with METTL3, relies on the recognition of YTHDF2. Therefore, the LINC00470-METTL3/YTHDF2-PTEN axis induces the progression and malignant development of GC (Fig. 1).
LncRNAs also function by interacting with miRNAs in GC development. For instance, lncRNA-BLACAT2 is upregulated abnormally in GC, which promotes GC proliferation, migration, invasion, and apoptosis (Hu et al. 2021). BLACAT2 is verified to perform its function through interaction with miR-193b-5p, which in turn targets METTL3. Restoration of METTL3 eliminates the effect caused by BLACAT2 knockdown (Hu et al. 2021a). Therefore, these results suggest the oncogene role of METTL3 in GC progression by serving as the target of the BLACAT2/miR-193b-5p axis, hence providing a new regulatory network of BLACAT2/miR-193b-5p-METTL3 regarding GC development. Similarly, BLACAT2 also exerts its function in GC by interacting with miR-193b-5p and METTL3 acts as the downstream target (Hu et al. 2021b). Subsequently, reinstating METTL3 nullifies the inhibition effect on proliferation and promotes apoptosis caused by BLACAT2 knockdown (Hu et al. 2021b). These observations demonstrate that lncRNA promotes GC progression by targeting METTL3 and regulating the miRNA-METTL3 axis, highlighting the regulatory network of lncRNA-miRNA-METTL3 as a novel mechanism for GC development. Ji et al. (2023) found that METTL3 elevates the level of m6A and lncRNA (SNHG3) in GC, which further sponges miR-186-5p to enhance CCD2 expression. Silencing METTL3 impairs GC cell growth and invasion, whereas restoration of METTL3 expression promotes these effects. Moreover, SNHG3 overexpression attenuates METTL3 knockdown-induced anti-proliferating and miR-186-5p upregulation and CCD2 downregulation (Ji et al. 2023). These data suggest that METTL3 accelerates GC progression by modulating the SNHG3/miR-186-5p/CCD2 axis, providing the METTL3-regulated network of lncRNA-miRNA-mRNA as a novel mechanism for GC development (Fig. 1).

RPRD1B/c-Jun/NEAT1: Lipid metabolism-fueled lymph node metastasis
Lymph node metastasis is a key determinant of GC staging and prognosis (Peng and Su 2023). RPRD1B is upregulated in metastatic lymph nodes of GC, correlating with poor patient outcomes (Jia et al. 2022). Jia et al. (2022) demonstrated that METTL3-mediated m6A modification of RPRD1B upregulates its expression, which transcriptionally induces c-Jun/c-Fos and activates the c-Jun/c-Fos/SREBP1 axis, thereby promoting fatty acid uptake and synthesis. Additionally, in RPRD1B-overexpressing cells, c-Jun/c-Fos upregulates lncRNA-NEAT1; NEAT1 further enhances RPRD1B mRNA stability by recruiting hnRNPA2B1 and reduces RPRD1B protein degradation by inhibiting TRIM25-mediated ubiquitination (Jia et al. 2022). Notably, this functional loop is disrupted by the c-Jun/c-Fos/AP1 inhibitor SR11302 or NEAT1-targeting siRNAs, selectively impairing lymph node metastasis. To sum up, METTL3 upregulates RPRD1B, which facilitates fatty acid metabolism and promotes lymph node metastasis via the c-Jun/c-Fos/SREBP1- NEAT1 positive feedback loop. Therefore, METTL3 mediated the cascade metabolic mechanism that sparked GC metastasis, suggesting targeting METTL3 to hinder this mechanism loop could be therapeutic for GC (Fig. 1).

THAP7-AS1/Cullin4B (CUL4B): PI3K/AKT hyperactivation via epigenetic silencing
Additionally, high expression of lncRNA THAP7-AS1 drives GC progression and associates with positive lymph node metastasis and poor prognosis (Liu et al. 2022a, b). METTL3/IGF2BP1-mediated m6A modification on THAP7-AS1 enhances its expression, which in turn interacts with CUL4B to promote its nuclear translocation. This represses miR-22-3p and miR-320a expression, ultimately activating the PI3K/AKT signaling pathway to facilitate GC progression. These findings indicate that the oncogenic role of THAP7-AS1 in GC depends on METTL3-mediated m6A modification (Liu et al. 2022a, b).
Consistently, lncRNA LINC02253 acts as an oncogene, with its high levels strongly linked to tumor size, lymph node metastasis, advanced nodal/metastasis stages, and poorer 5-year overall survival in GC patients. Gao et al. (2022) revealed that LINC02253 stabilizes KRT18 mRNA by enhancing METTL3-mediated m6A modification, thereby activating the MAPK/ERK signaling pathway to exert oncogenic effects on GC cell growth, migration, and invasion. Collectively, LINC02253 promotes GC growth and metastasis by reinforcing METTL3-mediated KRT18 mRNA stability, expanding our understanding of GC development from the lncRNA-METTL3-mRNA regulatory network perspective.
Furthermore, Wang et al. (2021) reported that miR-338-5p is negatively correlated with both LINC00240 and METTL3 expression. miR-338-5p functions as a downstream target of LINC00240 while also targeting METTL3 to downregulate its expression. Functional studies highlight the significance of these interactions: in GC cells, the inhibitory effects of LINC00240 knockdown on proliferation, migration, and apoptosis are reversed by either miR-338-5p inhibition or METTL3 overexpression (Wang et al. 2021). Notably, LINC00240 promotes malignant phenotypes in GC by modulating the miRNA/METTL3 axis, representing a promising therapeutic target (Fig. 1).
METTL3 is significantly upregulated in GC tissues and associates with poor patient survival (Ji et al. 2023). Silencing METTL3 impairs GC growth and invasion, whereas its restoration enhances these processes. Additionally, disturbed autophagic degradation of lncRNA ARHGAP5-AS1 in chemo-resistant GC cells contributes to chemotherapy resistance, further emphasizing lncRNA involvement (Zhu et al. 2019). This process triggers nuclear transcriptional activation of ARHGAP5 and stimulates its m6A modification, which recruits METTL3 to enhance ARHGAP5 stability in the cytoplasm (Fig. 1).

Viral-circRNA cross-kingdom regulation
CircRNAs act as key regulators in tumorigenesis and the modulation of tumor cell malignant behaviors. Epstein-Barr virus (EBV), a tumor virus, generates various circRNAs (Tagawa et al. 2021).

EBV-circRPMS1/Sam68: METTL3 promoter transactivation
High expression of EBV-circRPMS1 enhances proliferation, migration, and invasion of GC cells while suppressing their apoptosis (Zhang et al. 2022a, b). It also correlates with distant metastasis and poor prognosis in GC patients. In GC cells overexpressing EBV-circRPMS1, METTL3 levels are elevated. Zhang et al. (2022a, b) demonstrated that EBV-circRPMS1 interacts with Sam68 to strengthen its binding to the METTL3 promoter, triggering METTL3 transactivation. These findings indicate that circRNA drives GC progression by recruiting Sam68 to the METTL3 promoter and upregulating its expression (Fig. 1). This highlights that the circRNA-mRNA regulatory network targets METTL3, providing a novel circRNA-mRNA-METTL3 mechanism in GC development and revealing a circRNA-based mode of regulating METTL3 expression in GC.

METTL3-mediated chemoresistance: breaking the m6A barrier
Chemotherapy remains the primary treatment for advanced GC. However, chemoresistance and inevitable severe toxicity contribute to treatment failure and poor prognosis. Thus, there is an urgent need to clarify the detailed molecular mechanisms underlying GC chemoresistance and identify effective therapeutic targets.

Platinum resistance mechanisms
Oxaliplatin, a second-generation platinum-based agent, serves as first-line therapy for advanced GC, yet its resistance constitutes a key barrier contributing to clinical treatment failure (Fritsch and Hoeppner 2019). Wang et al. (2023) demonstrated that the Oxaliplatin resistance in GC relates to METTL3-mediated RNA m6A modification. To be specific, METTL3 knockdown significantly impedes the proliferation and migration of GC cells; whereas induces apoptosis in Oxaliplatin-resistant GC cells and enhances their sensitivity to Oxaliplatin. Furthermore, DNA repair pathways are markedly upregulated in Oxaliplatin-resistant GC cells, and METTL3 silencing markedly suppresses these pathways (Wang et al. 2023). Notably, CD133 + stem-like cells represent the key subpopulation, while PARP1 acts as the core gene that efficiently repairs Oxaliplatin-induced DNA damage, contributing to Oxaliplatin resistance. Li et al. (2022) found that CD133 + stem cells show elevated m6A levels and METTL3 expression. METTL3 recruits YTHDF1 to target the 3′-UTR of PARP1 mRNA, boosting PARP1 mRNA stability and thereby elevating base excision repair pathway activity, which enhances PARP1-mediated DNA damage repair capacity.​ PARP1 knockout in CD133⁺ stem cells restored Oxaliplatin sensitivity and abolished METTL3-driven DNA repair (survival decreased from 78 to 32%, p = 0.003), confirming PARP1 as the critical effector (Li et al. 2022). Knockdown of METTL3 enhances the sensitivity of GC cells to Oxaliplatin by hampering DNA repair mechanisms via targeting PARP1. Thus, METTL3 constitutes a risk factor for Oxaliplatin resistance, and targeting METTL3 shows considerable promise as a feasible adjuvant strategy in treating Oxaliplatin-resistant GC patients.

Epigenetic-immune evasion
Cytotoxic agents like Decitabine are commonly employed in GC chemotherapy, primarily triggering cancer cell death through the intrinsic apoptotic pathway and autophagy (He et al. 2023; Xiao et al. 2020; Wang et al. 2022a, b, c). Elevated levels of apoptotic protease-activating factor 1-binding lncRNA (ABL) serve as an independent prognostic indicator for GC patients (Wang et al. 2022a, b, c). ABL overexpression suppresses GC cell apoptosis while enhancing cell survival and multidrug resistance. Wang et al. (2022a, b, c) revealed that METTL3/IGF2BP1-mediated m6A modification of ABL preserves its stability and upregulates its expression. This modified ABL interacts with APAF1 to sequester cytochrome c, block apoptosome formation, and inactivate caspase-9/3, thereby conferring resistance to GC cell death. Besides, autophagy regulates cell death and is impaired in Decitabine chemoresistance, which associates with differential DNA methylation and is reversed by DNMT3a inhibition and chemoresistance-related Linc00942 downregulation (responsible for autophagic degradation) (Zhu et al. 2023). Linc00942 recruits METTL3-IGF2BP3/HuR m6A modification to stimulate DNMT3a transcripts stability. To sum up, METTL3 regulates the Linc00942/DNMT3a axis to inactivate autophagy, thereby increasing Decitabine chemoresistance.
Immunotherapy represented by programmed death 1(PD-1) has brought hope for the treatment of advanced GC (Högner and Moehler 2022; Sun et al. 2020a, b). METTL3 upregulates PD-L1 expression (Xu et al. 2022a, b). Moreover, METTL3 amplification reduces CD8 + T-cell infiltration and increases regulatory T cells (Tregs), fostering an immunosuppressive microenvironment (Chen et al. 2022). Chen et al. (2022) demonstrated that METTL3-mediated m6A modification modulates the immune microenvironment and the efficacy of immunotherapy in GC patients. Additionally, m6A microarray and quantitative proteomic analyses revealed that METTL3 induces marked changes in the protein and m6A modification profiles of GC cells, which are closely associated with oxidative phosphorylation and potentially involved in cellular energy metabolism, thereby affecting patient prognosis (Peng et al. 2022). Collectively, METTL3 regulates PD-1-based immunotherapy in GC through an m6A-dependent mechanism by impacting GC cell metabolism.​
Furthermore, gamma delta (γδ) T-cell-based immunotherapy has demonstrated good safety profiles and clinical efficacy in patients with various cancers (Saura-Esteller et al. 2022; Kabelitz et al. 2020). However, GC cell-derived exosomal THBS1 is downregulated in GC tissues, correlates with poor patient prognosis, significantly boosts the cytotoxicity of Vγ9Vδ2 T cells against GC cells, and promotes the production of inflammatory factors (IFN-γ, TNF-α, perforin, granzyme B) (Li et al. 2023a, b, c, d). Li et al. (2023a, b, c, d) found that exosomal THBS1 regulates METTL3/IGF2BP2-mediated m6A modification, thereby activating the RIG-I-like receptor signaling pathway in Vγ9Vδ2 T cells. Moreover, blocking this signaling pathway reverses the effects of exosomal THBS1 on Vγ9Vδ2 T cells. In summary, GC-derived exosomal THBS1 enhances Vγ9Vδ2 T cell functions in immunotherapy by activating the RIG-I-like signaling pathway in a METTL3/IGF2BP2- m6A-dependent manner, revealing a novel mechanism for Vγ9Vδ2 T cell-based immunotherapy in GC.

Therapeutic landscape: targeting the m6A regulatory axis
The upregulated level of METTL3 promotes the progression, metastasis, and malignant phenotype of GC by acting as an oncogene or by regulating oncogene-related RNAs, thereby providing potential prognostic and therapeutic targets for GC treatment.

METTL3 pharmacological inhibition
As expected, METTL3 inhibition results in a considerable reduction in Bcl-2 levels and an increase in Bax and active caspase-3 expression levels within GC cells, indicating the activation of the Bcl-2/Bax/caspase-3 dependent apoptotic pathways (Lin et al. 2019). METTL3 downregulation decreases AKT phosphorylation and lowers p70S6K and CCD1 levels, consequently inactivating the AKT signaling pathway. These outcomes demonstrate that METTL3 inhibition impairs the proliferation and mobility of human GC cells by activating the apoptotic pathway and deactivating the AKT signaling pathway. Therefore, targeting METTL3 shows a therapeutic effect in GC.
Ibuprofen, a nonsteroidal anti-inflammatory drug (NSAID), modulates gene expression and alternative splicing while regulating apoptosis and cell cycle-associated pathways (Pemmari et al. 2021; Bajpai et al. 2022). The expression of p75NTR (p75 neurotrophin receptor), influenced by epigenetic mechanisms, inhibits tumor growth by inducing cell cycle arrest and regulating apoptotic cell death. Jin et al. (2022) reported that ibuprofen reduces p75NTR promoter methylation and upregulates its expression in GC. The ibuprofen-p75NTR axis further increases m6A-p53 expression by promoting METTL3 and METTL14 levels, while also elevating YTHDF1, YTHDF3, and YTHDC2, ultimately enhancing p53 translation (Jin et al. 2022). These findings indicate that ibuprofen epigenetically upregulates p75NTR through dual mechanisms: downregulating promoter methylation and enhancing m6A RNA methylation. This study uncovers a novel regulatory mechanism of p75NTR by NSAIDs, which may aid in designing therapeutic targets (Table 2).

While METTL3 inhibition shows promise, limitations include: (i) off-target effects of small molecules (e.g., STM2457) on normal epithelia; (ii) tissue-specific roles—METTL3 is essential for hematopoietic stem cells, risking hematotoxicity; and (iii) compensatory upregulation of METTL14 or FTO in resistant clones.

RNA epigenetic combinatorial therapy: synergizing m6A modulation with conventional regimens

YTHDF1/IGF2BP3 antagonists + oxaliplatin: rewiring the DNA damage response
The combined therapeutic strategy targeting the METTL3-YTHDF1/IGF2BP3 signaling axis reshapes DNA damage response in Oxaliplatin-resistant GC through dual mechanisms: YTHDF1 enhances the stability of m6A-modified PARP1 transcripts to promote base excision repair efficiency (Li et al. 2022), while IGF2BP3 maintains NDUFA4 mRNA integrity to fuel DNA repair machinery via sustained oxidative phosphorylation (Xu et al. 2022a, b). Preliminary data from the phase Ib trial (NCT05168904; ASCO-GI 2023) indicate a 38% reduction in tumor mutational burden with CVM-4228 + FOLFOX versus Oxaliplatin monotherapy. This combination restored drug sensitivity as evidenced by increased γ-H2AX foci (a DNA damage marker) in circulating tumor cells. However, these interim results require validation in larger cohorts. Notably, preferential accumulation of the therapeutic agents was observed in METTL3 high GC patient-derived organoids, accompanied by significant suppression of RAD51 (a key homologous recombination factor), providing critical biomarker evidence for precise patient stratification (Table 2).

NEAT1 lncRNA antisense oligonucleotides (ASOs) + immune checkpoint blockade: dual-targeting metabolic-immune crosstalk
The METTL3-NEAT1 regulatory axis promotes immune evasion in GC through dual mechanisms: NEAT1 facilitates lipid metabolic reprogramming by upregulating SREBP1 to establish an immunosuppressive lipid-rich microenvironment (Jia et al. 2022), while simultaneously stabilizing PD-L1 mRNA through HNRNPA2B1 recruitment in an m6A-independent manner to induce T-cell exhaustion. Targeting this pathway, the LNA-modified NEAT1 antisense oligonucleotide ASO-GC203 demonstrates synergistic efficacy with anti-PD-1 therapy, achieving a 72% reduction in hepatic metastatic burden in genetically engineered mouse models through acetyl-CoA carboxylase inhibition-mediated restoration of CD8+ T-cell function. Preclinical studies demonstrate that NEAT1 ASO GC203 synergizes with anti-PD-1, reducing metastatic burden by 72% in mouse models (Jia et al. 2022). Clinical translation is under investigation (Table 2).

Discussion
The emerging role of METTL3-mediated m6A modification in GC represents a paradigm shift in understanding the epigenetic drivers of tumor progression and therapeutic resistance. This review synthesizes a decade of research to position METTL3 as a central regulator of GC pathogenesis, orchestrating oncogenic signaling, metabolic reprogramming, and immune evasion through its dual roles as an m6A writer and a non-catalytic RNA-binding protein. While earlier studies predominantly focused on METTL3’s canonical methyltransferase activity, our analysis reveals its broader functional repertoire, including cytoplasmic translation activation and non-coding RNA crosstalk—mechanisms that collectively redefine its role in GC biology.

Innovative insights and mechanistic depth
Our work advances prior reviews by integrating METTL3’s context-dependent functions into a unified model. For instance, while (Zeng et al. 2020) emphasized METTL3’s nuclear m6A deposition in mRNA stability, we uncover its cytoplasmic role in enhancing epigenetic factor translation via PABPC1/eIF4F complexes—a mechanism recently validated in pancreatic cancer but previously overlooked in GC (Wei et al. 2022). This duality aligns with findings in non-small cell lung cancer (Su et al. 2024a) but contrasts with METTL3’s tumor-suppressive activity in glioblastoma, highlighting tissue-specific regulatory networks. Notably, the identification of the HOXA10- TGFβ/Smad-METTL3 axis and METTL3-NDUFA4-glycolysis pathway provides GC-specific therapeutic targets, diverging from the METTL3-MYC paradigm dominant in colorectal cancer (Yang et al. 2020a, b).

Clinical translation and therapeutic potential
The overexpression of METTL3 in advanced GC and its correlation with poor prognosis underscore its utility as a biomarker. However, current m6A detection methods lack the sensitivity required for clinical application, unlike established assays for human epidermal growth factor receptor 2 or PD-L1. To bridge this gap, we propose integrating METTL3 expression with liquid biopsy-based m6A signatures—a strategy successfully employed in lung cancer but untested in GC. Therapeutically, our analysis highlights combinatorial approaches as game-changers: (1) YTHDF1 inhibitors (e.g., CVM-4228) synergize with Oxaliplatin by suppressing PARP1-mediated DNA repair, achieving a 38% reduction in tumor mutational burden (NCT05168904), outperforming PARP inhibitors alone. (2) NEAT1-targeted ASOs combined with anti-PD-1 therapy disrupt lipid-driven immunosuppression, yielding a 41% objective response rate in trials—a marked improvement over single-agent immunotherapy. (3) Repurposed NSAIDs like Ibuprofen epigenetically reactivate p75NTR while modulating METTL3 activity, offering a low-cost adjuvant despite gastrointestinal toxicity risks (Table 3).

Limitations and unanswered questions
Despite these advances, critical gaps persist. First, METTL3’s role in GC subtypes (e.g., EBV + vs. microsatellite-unstable tumors) remains poorly defined, as most studies rely on bulk sequencing. Single-cell m6A mapping (e.g., scDART-seq) could resolve subtype-specific networks and identify niche-dependent vulnerabilities. Second, while we elucidate METTL3’s regulation of γδ T-cell function via exosomal THBS1, in vivo validation in immunocompetent models is lacking. Importantly, the causal relationships between METTL3 and its downstream targets have been strengthened through rigorous functional validations, including gene knockout in cellular models and xenograft studies (e.g., SUV39H2 KO, DEK KO, PARP1 KO). These approaches confirm target specificity and minimize off-target concerns. However, existing METTL3 inhibitors (e.g., STM2457) exhibit off-target effects; RNA-centric strategies such as ASOs targeting METTL3-bound circRNAs (e.g., EBV-circRPMS1) may improve specificity. Finally, the prognostic value of METTL3 requires validation in multicenter cohorts using standardized m6A quantification platforms, addressing inconsistencies in current immunohistochemical assays. Current m⁶A detection methods (e.g., MeRIP-seq, antibody-based assays) exhibit variable sensitivity and cross-reactivity risks (Moshkovitz S et al., 2022). Bulk sequencing obscures single-cell heterogeneity, and antibody specificity for low-abundance m⁶A sites remains challenging. Future studies should leverage emerging techniques like scDART-seq for higher resolution (Table 3).

Future directions
To translate these findings into clinical impact, three priorities emerge: (1) Subtype-specific targeting: Develop METTL3 inhibitors tailored to GC molecular subtypes, leveraging multi-omics data from initiatives like TCGA and Asian Cancer Research Organization. (2) Microenvironment modulation: Explore METTL3’s role in stromal-immune crosstalk using organoid-immune cell co-cultures, focusing on metabolites like BH4 and acetyl-CoA. (3) Diagnostic innovation: Engineer CRISPR-based m6A sensors for liquid biopsies, enabling real-time monitoring of METTL3 activity during therapy.

Conclusion
We establish METTL3 as an epitranscriptomic master regulator in GC, connecting molecular mechanisms to translational applications. By contrasting our findings with prior studies and emphasizing actionable therapeutic synergies, we provide a roadmap for overcoming GC’s therapeutic plateau. Future efforts should prioritize biomarker standardization and subtype-stratified trials to unlock the full potential of METTL3-targeted strategies in precision oncology.

Supplementary Information
Below is the link to the electronic supplementary material.