Cancer tissues (n=92) and their corresponding cancer-adjacent tissues were collected from patients at Fujian Provincial Hospital (Fuzhou, China). Patients with LUAD were diagnosed histopathologically without chemotherapy or radiotherapy. This study was approved by the Research Ethics Committee of Fujian Provincial Hospital (K-2021-040-04) and conform to the ethical standards of the 1964 Declaration of Helsinki.

Human LUAD cell lines A549 and PC9 and normal human lung epithelial cell line BEAS-2B were obtained from the American Type Culture Collection. Both of them were cultured in PRIM-1640 medium (Hyclone, USA) with 10% fetal bovine serum (Gibco, USA) and incubated in a 37°C atmosphere with 5% CO₂. A549 cells were seeded in 6-well plates at a density of 5 × 10⁴ cells/well and cultured for 24 h. METTL3 control, METTL3 inhibitor, METTL3 mimics, and 0.5 µg/mL polybrene (Genepharma, China), were mixed in PRIM-1640 medium according to the manufacturer’s instruction. This mixture was used to infect A549 cells. A total of 4 groups were transfected, including blank group, unloaded virus group(Control), METTL3 knockdown group (METTL3-KD)and METTL3 overexpression group(METTL3-OE). siFGF2 (Genepharma, China) was used to infect A549 METTL3-KD cells. Three groups were infected including Control-siNC, METTL3_KD+siNC (transient siNC plasmid transfer in METTTL3-KD stable cell line), and METTL3_KD+siFGF2 (transient FGF2 plasmid silencing in METTTL3-KD stable cell line).

RNA was extracted using TRIzol Reagent (Invitrogen, USA). A PrimeScript RT Reagent Kit (TaKaRa, Japan) was used for reverse transcription. TB Green® Premix Ex Taq™ II (TaKaRa, Japan) was used for qPCR. Experimental manipulations and reaction temperature conditions were performed according to the manufacturer’s instructions. GAPDH was used as the internal standard. The primers are following (SunYa, China): METTL3-Forward: CTGCAACGCATCATTCGGAC, METTL3-Reverse: AGACCCTGGTTGAAGCCTTG; GAPDH-Forward: GGTGTGAACCATGAGAAGTATGA, GAPDH-Reverse: GAGTCCTTCCA CGATACCAAAG.

MeRIP experiments were performed using the EpiQuik CUT&RUN m6A RNA Enrichment Kit (Epigentek, New York), according to the manufacturer’s protocols. RNA was calculated using RT-qPCR description like 2.3. The primers are following (SunYa, China): FGF2-Forward: GCACTGAAACGAACTGGGCA, FGF2-Reverse: GAAATGAGATTAGATGT GGC.

CCK8 was used to test cell viability (Beyotime, China). Three thousand cells per well were plated in 96-well plates and incubated for 0, 24, 48, and 72 h. The old medium was discarded, and a 10% solution of CCK8 was added. The plates were incubated at 37°C for an additional hour. Absorbance was measured at 450 nm using a Multiskan spectrophotometer. During detection, the liquid was shaken, and the bubbles were removed.

About 1×10⁶ cells were seeded to each well of a 6-well plate. Cells were scraped using a sterile pipette tip once they reached over 95% confluency for the wound-healing assay. The detached cells were washed with PBS and incubated at 37°C for 48 h. Photographs were taken at the start (0 h) and after 48 h to determine the percentage of wound closure. The scratch area was calculated using ImageJ software.

Approximately 5×10⁴ cells were seeded in the upper chamber of a polycarbonate-coated trans well filter coated with Matrigel. The cells were then cultured at 37°C for 48 h. Next, the medium in the upper and lower chambers was discarded, and 4% paraformaldehyde solution was added to the lower chamber to fix the cells for 30 min. Cells that did not pass through the membrane were removed using a cotton swab. A 1% crystal violet solution was added to stain the cells, and the mixture was incubated for 15 min. Images were captured under a microscope, and the average value of five random ×400 fields per well was calculated. This process was repeated five times for each group. The results are presented as the mean ± standard deviation.

Gene Expression Profiling Interactive Analysis (GEPIA; http://gepia.cancer-pku.cn) is a web server for analyzing RNA sequencing expression data from The Cancer Genome Atlas Program (TCGA) and GTEx projects. Using GEPIA, we explored the expression of METTL3 in LUAD. A Kaplan–Meier plotter (KM, http://kmplot.com/analysis) was used to assess the correlation between METTL3 and tumor survival (16). SRAMP (http://www.cuilab.cn/) is used to prediction FGF2 methylation modification sites (17).

Cancerous tissue, para-cancerous tissue, and cells were harvested and lysed with RIPA buffer supplemented with protease and phosphatase inhibitors on ice for 30 min. Protein quantities were determined using enhanced BCA protein assay kit (Beyotime Biotechnology), ensuring consistent loading (20ug total). Equal amounts of protein were electrophoretically separated on 10% polyacrylamide gels before being transferred to polyvinylidene fluoride membranes. The membranes underwent with 5% skim milk for blocking. Antibodies targeting METTL3, p‐PI3K, PI3K, p‐AKT, AKT, mTOR, p-mTOR and GAPDH (Cell Signaling Technology; diluted 1:1000) were incubated overnight at 4°C. Subsequently, the membranes were washed three times, followed by incubation with HRP-conjugated affinipure goat anti-rabbit IgG (H + L) (1:10000, SA00001-2, proteintech) secondary antibodies for 1.5 h at room temperature. The membranes underwent a triple wash using TBST. Visualization of protein bands was achieved using an ECL kit (ThermoFisher Scientific).

To examine the regulation of related factors in A549 cells by METTL3, we performed transcriptome sequencing of virus-transfected cells, completed by Shanghai Megi Biological. Data were analyzed using the online platform Majorbio Cloud Platform (www.majorbio.com).

To detect the RNA stability in cells, actinomycinD (Act-D; 5 mg/mL; Sigma, USA) was administrated to cells. At the indicated time point, RNA was isolated and analyzed by RT-qPCR.

Figures were generated using GraphPad Prism version 6.02. An unpaired Student’s t-test or paired t-test was used for difference analysis. Spearman’s correlation analysis was used to describe the correlation between quantitative variables without a normal distribution. Significance was set at P < 0.05. Statistical analysis was performed using SPSS 22.0 and R 4.1.3.

The GEPIA results showed that METTL3 was expressed at lower levels in tumor tissues ( Figure 1A ). Subsequently, we validated the expression levels of METTL3 mRNA in 92 pairs of tumor and adjacent normal tissues ( Figure 1B ). The results indicated that METTL3 was generally expressed at lower levels in tumor tissues. To verify METTL3 protein expression levels, we performed western blot analysis on 6 pairs of tissues. The results showed that METTL3 protein was expressed at lower levels in tumor tissues ( Figure 1C ). We further assessed METTL3 mRNA and protein expression at the cellular level. The results indicated that METTL3 mRNA was expressed at lower levels in cells A549 and PC9 compared to BEAS-2B ( Figure 1D ). Western blot results also indicated that METTL3 was downregulated in A549 ( Figure 1E ). Additionally, we analyzed the impact of METTL3 on the survival time of LUAD patients. The results suggested that lower METTL3 expression levels were associated with poorer prognosis ( Figure 1F ). In summary, these analyses indicate that METTL3 is downregulated in tumors and suggest a poor prognosis.

To further validate the role of METTL3 in LUAD, we constructed A549 cell lines with METTL3 knockdown and overexpression. The RT-qPCR results ( Figure 2A ) of indicated that we have a successful infection. Subsequently, we conducted a cell scratch assay ( Figure 2B ). The scratch assay results showed that the scratch healing area decreased after METTL3 knockdown, suggesting that downregulation of METTL3 promotes cell migration. Using the CCK8 assay ( Figure 2C ), we observed that METTL3-KD increased tumor cell viability, while overexpression of METTL3 decreased tumor cell viability, indicating that METTL3 inhibits tumor cell proliferation. To explore the effect of METTL3 on tumor migration, we performed transwell assays. The results showed that the number group of METTL3-KD cells crossing the transwell subventricular membrane was greater more than that of the control group, and while the METTL3-OE group crossing was less than that of the control group (as shown in Figure 2D ). Collectively, it indicated that knockdown of METTL3 promoted the proliferation and migration of A549 cells, and overexpression of METTL3 inhibited the proliferation and migration of A549 cells.

As shown in Figure 3A , RNA-seq analysis revealed that compared to the control, METTL3 knockdown (METTL3-KD) resulted in changes in the expression of 2,468 genes, with 710 genes downregulated and 1,758 genes upregulated. In contrast, only 27 genes were differentially expressed in METTL3 overexpression (OE) compared to the control. Consequently, we chose METTL3-KD for further analysis. Figure 3B displays the top ten hub genes identified, all of which play significant roles in tumors. According to Gene Ontology (GO) analysis ( Figure 3C ), the differentially expressed genes were primarily enriched in processes such as endodermal cell differentiation, positive regulation of apoptotic cell clearance, and regulation of epithelial to mesenchymal transition. The Kyoto Encyclopedia of Genes and Genomes (KEGG) is a comprehensive knowledge base for systematically analyzing gene functions and linking genomic information with functional information. KEGG analysis ( Figure 3D ) revealed that the differentially expressed genes were associated with cancer-related pathways, including the PI3K/AKT signaling pathway and the TGF-β signaling pathway. These pathways are closely related to tumor cell growth and differentiation. Our findings suggest that downregulating METTL3 induces alterations in the cellular function and downstream pathways of A549 cells, which are significantly correlated with tumor progression. This implies that the dysregulation of METTL3 plays a pivotal role in A549.

To investigate whether the regulation of FGF2 by METTL3 is related to m6A modification, we predicted the m6A modification sites of FGF2. As shown in Figure 4A , there are abundant m6A modification sites on the FGF2 sequence. Subsequently, we confirmed the expression levels of FGF2 in METTL3-KD and METTL3-OE cell lines. As shown in Figure 4B , FGF2 expression was upregulated in METTL3-KD cell lines, indicating that silencing METTL3 induced an increase in FGF2. Using MeRIP-qPCR ( Figure 4C ), We found that the methylation level of FGF2 was significantly downregulated after METTL3 knockdown, suggesting that FGF2 is subject to METTL3-mediated methylation modification. To determine whether METTL3 affects the stability of FGF2 RNA, we conducted a half-life assay for FGF2. The results ( Figure 4D ) indicated that the half-life of FGF2 was prolonged in METTL3-KD cell lines, suggesting that METTL3 promotes the degradation of FGF2 through methylation modification.

To understand whether METTL3 plays a role in LUAD by regulating FGF2, we transfected siFGF2 into the METTL3-KD cell line. The infection efficiency was verified by RT-qPCR and western blot, and the results are shown in Figures 5A, B . In the constructed cell line, we performed CCK8 cell viability assay and transwell experiment. Figure 5C shows that the tumor cell viability in METTL3_KD+siFGF2 was significantly lower than that in METTL3_KD+siNC, indicating that silencing FGF2 can reduce the cell proliferation activity of METTL3-KD. Figure 5D shows that the migration ability in METTL3_KD+siFGF2 was lower than that in METTL3_KD+siNC, indicating that silencing FGF2 can reduce the cell migration ability of METTL3-KD. In short, FGF2 alleviated the malignant progression of tumors caused by METTL3 in LUAD cells. In addition, compared with Control+siNC, the cell viability and invasion ability in METTL3_KD+siNC were improved, further verifying our previous conclusions.

To elucidate the downstream pathways impacted by METTL3-mediated FGF2 upregulation in LUAD cells, we suppressed FGF2 expression in METTL3 knockdown cell lines. Our RNA-seq analysis revealed that differentially expressed genes in LUAD following METTL3 knockdown were significantly enriched in the PI3K/AKT pathway. Thus, we sought to investigate whether METTL3/FGF2 influenced this pathway's activity level. The results ( Figure 6 ) demonstrated that silencing led to inhibition of the PI3K/AKT/mTOR pathway activation level in METTL3 knockdown LUAD cells. In summary, our findings unveil a signaling network involving METTL3/FGF2/PI3K/AKT/mTOR in LUAD and shed light on the molecular mechanism underlying tumor malignant progression while holding great significance for targeted therapy in research on LUAD.