M6A-Methylated circRAPGEF5 drives lung adenocarcinoma progression and metastasis via IGF2BP2/NUP160-mediated autophagy suppression

Abstract

Background

Lung adenocarcinoma (LUAD), the predominant histological subtype of non-small cell lung cancer, demonstrates critical regulatory involvement of RNA-binding proteins (RBPs) and circular RNAs (circRNAs) in tumorigenic processes. Emerging evidence highlights the circRNA-autophagy regulatory axis as a crucial modulator of cancer progression. This study systematically investigates the functional interplay within the RBP-circRNA-autophagy network in LUAD pathogenesis.

Methods

Employing RNA pull down, mass spectrometry and RNA immunoprecipitation facilitated the exploration of the circRAPGEF5 binding protein. M6A methylation RNA immunoprecipitation-PCR was utilized for m6A analysis. Immunofluorescence (IF) and fluorescence in situ hybridization (FISH) assays were conducted to ascertain the subcellular localization of target genes. Employing mRFP-GFP-LC3 fluorescent lentivirus labelling facilitated the monitoring of autophagy flow levels. Xenografts in mice were instrumental in affirming the role of circRAPGEF5.

Results

Through comprehensive molecular profiling, we identified elevated circRAPGEF5 expression in LUAD cells, which significantly suppressed autophagic flux while promoting malignant phenotypes including enhanced proliferation, migration, and invasion. Mechanistic investigations revealed that circRAPGEF5 directly interacts with the KH3-4 functional domain of Insulin-like Growth Factor 2 mRNA-Binding Protein 2 (IGF2BP2), an m6A reader protein. This interaction facilitated IGF2BP2-mediated stabilization of NUP160 mRNA, a nuclear pore complex component. Genetic ablation of NUP160 through RNA interference effectively restored autophagic activity, thereby attenuating the aggressive biological behaviors of LUAD cells. In vivo validation using xenograft models demonstrated that the circRAPGEF5/IGF2BP2/NUP160 signaling axis promotes tumor growth and metastatic dissemination through autophagy suppression.

Conclusion

Our findings reveal a novel epigenetic regulatory mechanism wherein m6A-modified circRAPGEF5 orchestrates autophagy inhibition via IGF2BP2-dependent stabilization of NUP160 transcripts, ultimately driving LUAD progression and metastasis. These results establish the circRAPGEF5/IGF2BP2/NUP160 axis as a potential therapeutic target for LUAD intervention.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12943-025-02399-3.

Background

Over the past five decades, there has been a notable increase in the incidence and mortality rates of lung cancer in numerous countries [1]. Pathologically staged, lung cancer is categorized into small-cell lung cancer and non-small-cell lung cancer (NSCLC). Lung adenocarcinoma emerges as the predominant subtype, constituting approximately 55% of NSCLC cases. It is distinguished by high heterogeneity, intricate clinical responses and limited healing [2]. Despite the advancements in lung cancer treatments, encompassing surgery, chemotherapy, radiotherapy and molecular targeting, epidemiological evidence underscores that the 5-year survival rate for lung cancer patients remains below 15.9% [3]. Furthermore, recurrence and metastasis are prevalent even in early-stage T1a/b/c and N0/N1 lung cancer cases, resulting in mortality. Currently, aberrant expression of genes is thought to underlie the occurrence and development of LUAD. Aberrantly expressed genes often regulate cell proliferation, survival as well as resistance to radiotherapy and chemotherapy [3, 4].

Autophagy is ubiquitously present in nearly all eukaryotes and plays a crucial role in the homeostasis of cancer cells [5]. It exerts a dual impact on tumor cells, manifesting as protective autophagy under adverse stimuli such as hypoxia, thereby suggesting its potential benefit to the survival of cancer cells [6]. Moreover, functioning as a programmed intracellular death process, autophagy can impede the growth of tumor cells by degrading proteins and eliminating damaged organelles [7]. Inducing autophagic death in tumor cells may emerge as a viable strategy for tumor treatment [8]. Currently, studies have shown that abnormally expressed genes often regulate autophagy in LUAD cells through the MAPK signaling pathway of PI3K / AKT / mTOR pathway, and then regulate the proliferation, metastasis and drug resistance of LUAD [9, 10]. Numerous eukaryotic organisms encompass the nucleoporin 160 (namely NPHS19) [11, 12]. At present, most studies on NUP160 have focused on diabetic nephropathy, steroid-resistant nephrotic syndrome [13, 14]. It has been shown that the inhibition of NUP160 alleviates diabetic nephropathy by activating autophagy [12]. But whether NUP160 in LUAD mediated cell malignant behavior through autophagy is still unknown.

Circular RNA lackslacking 3’ and 5’ end-capping structures, exhibits resistance to degradation by exonucleases and displays high conservation, stability, and tissue specificity [15]. Recent investigations have unveiled the substantial role of circRNAs in modulating microRNA (miRNA) expression [16], gene transcription [17] and protein translation [18]. These molecules are implicated in a myriad of biological processes, including tumorigenesis and hold potential therapeutic value for various diseases [15, 18–20]. It was revealed that circROBO1 fosters carcinogenesis and liver metastasis in breast cancer through the circROBO1/KLF5/FUS feedback loop. Additionally, it impedes the autophagy of afadin by suppressing BECN1 transcriptional inhibition [21]. CircMRPS35 regulates the progression and autophagy of osteosarcoma cells by recruiting KT6B to modulate FoxO3 [22]. In summary, the circRNA-autophagy axis plays a pivotal role in tumor cell proliferation, metastasis, invasion, and resistance to drugs or radiation [23]. However, it is currently unknown whether circRAPGEF5 mediates autophagy and regulates the progression and metastasis of LUAD.

Methylation at the sixth position of adenosine (m6A) constitutes a prevalent and crucial chemical modification of RNA, influencing various RNA types such as mRNA, rRNA, tRNA, long non-coding RNA (lncRNA) and miRNA [24]. The regulatory mechanism of m6A methylation modification is a dynamic process involving catalysis by m6A methyltransferases, clearance by demethylases, recognition by m6A methylation recognition proteins and binding to m6A methylation modification sites. This process governs diverse cellular processes and has been implicated in cancer progression [25]. IGF2BP2, a member of the m6A reader family, primarily functions by recruiting cofactor proteins for the stabilization of m6A-modified transcripts. Previous studies have indicated high expression of IGF2BP2 in LUAD patients, although its mechanism remains unclear [26].

RBPs assume significant roles in gene transcription and translation, and their interactions with circRNAs are functionally crucial components [27]. RBPs influence all stages of the circRNA life cycle, including biogenesis, localization, function, and degradation. Studies have shown that circRAPGEF5 interacts with RBFOX2 to confer ferroptosis resistance by modulating alternative splicing of TFRC in endometrial cancer [28]. Our earlier findings revealed that circRAPGEF5 is highly expressed in LUAD patients, promoting LUAD proliferation, migration, invasion and epithelial-mesenchymal transition (EMT) through the miRNA-126-3p/ZEB1 axis. Serum exosomal circRAPGEF5 may serve as a marker for identifying LUAD tumors [29]. Building upon these discoveries, we noted a robust correlation between circRAPGEF5 and autophagy in LUAD cells, suggesting a potential mechanism involving RBPs. Therefore, this study aims to delve deeper into the molecular components governing circRAPGEF5-mediated autophagy in LUAD and elucidate their mechanisms of action.

Materials and methods

Tissues samples

Ninety-three samples of LUAD and corresponding precancerous tissues were obtained from 51 male and 42 female patients at the First Affiliated Hospital of Wenzhou Medical University between August 2019 and August 2022. The clinical data of the patients are detailed in Table S1-S3. None of the patients had undergone chemotherapy or radiotherapy before sample collection. All samples underwent confirmation through histopathological examination and informed consent was obtained from all patients/participants. The Institutional Ethical Review Committee of the First Affiliated Hospital of Wenzhou Medical University approved this study (YS2023298).

Cell culture

A549, H1299 and HCC827 cells were procured from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China), and normal human bronchial epithelial cells (BEAS-2B). A549, H1299 and HCC827 cells were sustained in RPMI-1640 medium (Gibco, MA, USA) supplemented with 10% fetal bovine serum (FBS) (Gibco), while BEAS-2B cells were cultured in 10% FBS Dulbecco’s Modified Eagle Medium (Gibco). The cells were cultivated in an incubator at 37 °C with 5% CO₂.

Transfection

Lentiviral vectors containing circRAPGEF5 overexpression sequences (OE-circRAPGEF5), IGF2BP2 overexpression (OE-IGF2BP2), IGF2BP2 knockdown (sh-IGF2BP2), and NUP160 knockdown (sh-NUP160) were procured from Genechem (Shanghai, China). A549, H1299 and HCC827 cells were cultured in containing the lentiviral vector and 10% FBS RPMI-1640. The complete medium was replaced 12 h post-infection. After 48 h, fluorescence was observed in all transfected lentiviral LUAD cells. Stable cell lines were screened using puromycin (Solarbio, Beijing, China) at a concentration of 2 µg/mL. CircRAPGEF5 knockdown (kd-circRAPGEF5) was performed using antisense oligonucleotides from Ribo Bio (Guangzhou, China), with a matching kd-NC at a final concentration of 50 nM for transfection with lipofectamine 3000 (Shanghai, China). Transfection efficiency was verified by reverse transcription quantitative real-time polymerase chain reaction (RT-qPCR). The sequences of small Antisense Oligonucleotide and short hairpin RNA are presented in Table 1.

Table 1.

The sequences of small interfering RNAs in this study

Name	Sequence (5’-3’)
stB0010212A genOFFTM st-h-NUP160_001	GTGTCAGGATCATAAACTA
stB0010212B genOFFTM st-h-NUP160_002	GCCTTAACTTCCACGGATA
stB0010212C genOFFTM st-h-NUP160_003	GGATTGAATTGGCCTGAAA
stB0001811A genOFFTM st-h-IGF2BP2_001	CATGCCGCATGATTCTTGA
stB0001811B genOFFTM st-h-IGF2BP2_002	GAACGAACTGCAGAACTTA
stB0001811C genOFFTM st-h-IGF2BP2_003	AACAGGGACCAAGATAACA
lnc6201009014444ASO-h-hsa_circ_0001681_002	TCCTTTCGTCCAGTGCAGAT

Total RNA extraction and RT-qPCR

Total RNA was extracted from frozen tissue samples or cell lines utilizing TRIzol reagent (Thermo Fisher Scientific, USA). The quality and concentration of RNA were evaluated using a NanoDrop 2000 (Thermo Fisher, Shanghai, China). Subsequently, RNA underwent reverse transcription into cDNA using the PrimeScript RT kit (TaKaRa, Dalian, China) in accordance with the manufacturer’s protocol. Amplification was conducted using the SYBR GREEN PCR kit (TaKaRa), and analysis was executed employing an ABI 7500 real-time PCR system (Thermo Fisher Scientific, USA). The primer sequences are detailed in Table 2.

Table 2.

Primer sequences for qRT-PCR in this study

Gene name	Sequence (5’-3’)
IGF2BP2-F	ACCCTCTCGGGTAAAGTGGA
IGF2BP2-R	CTGTGTCTGTGTTGACTTGTTCC
NUP160-F	GGCCCTGTTTGCCTTACCAT
NUP160-R	GCGACTGGTCACCCCTGATA
METTL3-F	TTGTCTCCAAACCTTCCTAGT
METTL3-R	CCAGATCAGAGAGGTGGTGTAG
circRAPGEF5-F	GCCTCTCATTCCTGCCAGAG
circRAPGEF5-R	TCTTGATAGAGTCGCAGATGTTAGA

Measurement of colony formation

Cells were seeded into either 6-well or 12-well plates supplemented with 10% FBS in RPMI-1640 medium. Subsequently, they were incubated in a CO2 incubator at 37 °C for 7–14 days. Following this incubation period, the cells were fixed with 4% paraformaldehyde at 22 °C for 15 min, washed twice with phosphate-buffered saline (PBS), stained with 0.1% crystal violet for 15 min and then washed twice with PBS. The colonies were captured using a digital camera and quantified using ImageJ.

Cell counting kit-8 assay

An inoculation of 2000 cells per well was performed in 96-well plates, followed by a 24-hour incubation period. The original medium was then replaced with 10% Cell Counting Kit-8 (CCK-8) reagent (Tong Ren, Japan). The cells were incubated at 37 °C with 5% CO₂ for 1 h, followed by absorbance detection at 450 nm using an enzyme marker (Eq. 800, USA). This process was repeated daily for four consecutive days.

Transwell cell migration invasion assay

For the invasion assay, cells were washed once with PBS, and the cells were resuspended in RPMI-1640 medium. Cells concentration was adjusted to 50,000 cells/mL. Matrigel gel (BD Biosciences) was diluted in RPMI-1640 medium at a ratio of 1:8. Subsequently, 30 µL of diluted Matrigel gel was added to the small chamber and incubated for 2 h until solidification. A 200-µL cells suspension was introduced into the upper chamber (Millipore), followed by the addition of 600 µL of RPMI-1640 medium containing 20% FBS to the lower chamber. The cells were incubated for 48 h, fixed with 4% paraformaldehyde at room temperature for 15 min and stained with 0.1% crystal violet for 15 min. Migration experiments followed a similar procedure to the invasion experiments, except that matrigel gel was not added to the transwell inserts.

EdU proliferation experiment

On the first day, cells were seeded onto 24-well plates at approximately 50% confluence. On the second day, the medium was replaced with 1× EdU-1640 complete medium and incubated at 37 °C with 5% CO₂ for two hours. The medium was then discarded, and the cells were fixed with 4% paraformaldehyde at room temperature for 15 min, followed by manipulation according to the instructions provided by the EdU kit (Beyotime, Shanghai, China). Finally, the slices were sealed with an anti-quenching agent (Solabio, Beijing, China) and observed under an ortho fluorescence microscope.

Methylated RNA Immunoprecipitation and RNA Immunoprecipitation

Methylated RNA immunoprecipitation (MeRIP) and RNA immunoprecipitation (RIP) assays were conducted using an RNA immunoprecipitation kit (BersinBio, Guangzhou, China). The RIP procedure involved lysing cells in RIP lysis buffer, removing DNA and subsequently incubating the entire cell extract with anti-IGF2BP2 (Proteintech) antibody and immunoglobulin IgG antibody at 4 °C for 16 h for immunoprecipitation. This was followed by incubation with protein A/G magnetic beads for 1 h. RNA was eluted from the magnetic beads and extracted through co-precipitation with an RNA extraction solution (phenol/chloroform/isoamyl alcohol 25:54:1, Solabio). The extracted RNA underwent RT-qPCR using circRAPGE5 and NUP160 primers. To detect m6A RNA, a methylated RNA immunoprecipitation kit (BersinBio) was utilized. Cellular RNA was initially extracted using the TRIzol reagent and subsequently fragmented to approximately 300 bp using ultrasound. Fragmented RNA was incubated with an anti-m6A antibody in a vertical mixer for 4 h, followed by incubation with protein A/G magnetic beads for 1 h. Purified RNA was extracted using an RNA extraction solution and verified by qRT-PCR using circRAPGEF5 and NUP160 primers.

Immunofluorescence and fluorescence in situ hybridization

For immunofluorescence analysis, LUAD cells were fixed in 4% paraformaldehyde for 15 min on the day after inoculation in 24-well plates. The cells were permeabilized with 0.5% Triton X-100 (Solabio) for 15 min at room temperature and blocked with 5% bovine serum albumin (Beyotime) for 1 h. Primary antibodies against NUP160 (Proteintech 1:100) and IGF2BP2 (Proteintech 1:200) were added and incubated overnight at 4 °C. Fluorescein isothiocyanate (FITC)-coupled goat anti-rabbit IgG (H + L) (Proteintech 1:500) served as the secondary antibody. Nuclear staining was performed using DAPI (Ribo Biotech, Guangzhou, China).

For fluorescence in situ hybridization analysis, a Cy3-labelled circRAPGEF5 FISH probe was synthesized by Reebok Biotech and probed using the Reebok FISH kit (C10910, Reebok Biotech). LUAD cells were seeded in 24-well plates and fixed the following day, followed by overnight incubation at 37 °C in a hybridization buffer containing 4 nM circRAPGEF5 FISH probe. Nuclear staining was performed using DAPI. The images were captured using an upright fluorescence microscope.

Generation of circRAPGEF5 knock-out and knock-in mouse model

The preparation of transgenic mice with elevated circRAPGEF5 gene expression involved several key steps. Initially, gRNA and Cas9 plasmids, along with a donor vector, were constructed. Subsequently, in vitro transcription of the gRNA and Cas9 plasmid was performed. The resulting in vitro transcribed mRNA and donor vector were co-microinjected into zygotes from fertilized eggs and then implanted into multiple surrogate mice. Following birth, a small sample from the toe or tail of the mice was collected for PCR analysis to identify positive knock-in mice. The first generation of positive mice was bred with wild-type mice, and the offspring were again subjected to PCR analysis to confirm the presence of positive knock-in F1 generation mice. Seven days postnatally, the mice were categorized based on their genic phenotype into two groups: (1) wild-type mice and (2) circRAPGEF5 transgenic mice. Fourteen days after birth, a 1% diethylnitrosamine (DEN) was administered at a dosage of 28 mg/kg once weekly, culminating in a total DEN dose of 5.88 mg.

This procedure aims to generate lung-specific circRAPGEF5 gene knockout homozygous mice. To achieve this, circRAPGEF5-/- mice were crossed with wild-type C57BL/6 mice to produce the first generation of circRAPGEF5+/- lung-specific knockout heterozygous mice. The circRAPGEF5-/- mice were bred with wild-type C57B/L6 mice to produce the first generation of circRAPGEF5+/- lung-specific knockout heterozygous mice. These first-generation heterozygous mice (circRAPGEF5+/-) were then interbred. Seven days postnatally, the offspring were categorized into three distinct groups based on their genetic phenotypes: circRAPGEF5-/- lung-specific knockout homozygous mice, circRAPGEF5+/- lung-specific knockout heterozygous mice, and WT control group.

RNA pull down experiments and mass spectrometry

The fundamental approach involved the initial design of sense and antisense strand PCR primers, enabling the formation of the sense and antisense strands of circRAPGEF5 via PCR to generate the DNA template with absorptive capability. Subsequently, circRAPGEF5 was biotin-labeled, forming complexes with some of its counterparts to create circRAPGEF5-RBPs complexes for the extraction of RBPs. The RNA pull-down kit (Thermo Fisher Scientific, Inc, MA, USA) facilitated the adsorption of proteins interacting with circRAPGEF5. ProbecircRAPGEF5 was labelled to collect the eluent containing specifically bound proteins, and the enriched proteins were analyzed through sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) followed by silver staining. Differential proteins were identified using mass spectrometry.

Actinomycin D experiment

Actinomycin D, obtained from Solabio, was introduced to cells with stable knockdown and overexpression of IGF2BP2, initially seeded in 12-well plates. After cell morphology returned to normal on the second day, a 5 µg/mL actinomycin D solution was applied. RNA was subsequently extracted at 2, 4, and 6 h post-actinomycin D treatment, and RT-qPCR was employed to assess NUP160 mRNA expression. The baseline reference value was established using NUP160 mRNA expression at 2 h.

Truncated body experiments

The plasmid used for transient transfection was acquired from Ribo Bio. Specifically, one wild-type plasmid and six structural domain IGF2BP2 plasmids (including del-RM, del-KH1-2, del-KH3-4, del-KH2-4, del-KH1-4, del-KH1-3) with distinct fragment deletions were constructed. Transient transfection of circRAPGEF5 H1299 cells was carried out at a concentration of 800 ng/mL, and cells were harvested 24 h post-transfection for RIP experiments. The expression of circRAPGEF5 was evaluated using RT-qPCR.

mRFP-GFP-LC3 fluorescence detection

The mRFP-GFP-LC3 lentivirus, obtained from Hanheng Biological (Shanghai, China), was utilized for real-time monitoring of autophagic flow. A549, H1299 and HCC827 cells were cultured in complete medium containing the lentiviral vector and 10% FBS RPMI-1640 using an infection enhancement solution. The volume of complete medium was doubled at 4 h post-infection and replaced with complete medium at 24 h post-infection. Stable cells were screened using puromycin (Solarbio) and employed in subsequent experiments. After transfection of the target gene for 48 h, the medium was discarded and the cells were fixed with 4% paraformaldehyde for 15 min at room temperature, and autophagosome formation was detected by fluorescence microscopy.

Western blot

Total cellular protein was extracted using RIPA lysis buffer (Beyotime) and phenylmethanesulfonylfluoride fluoride (PMSF) at a ratio of 1:99. The protein concentration was standardized to 2 µg/µL and 10 µL of each sample was utilized. Proteins were electrophoretically separated and transferred to 0.45 μm and 0.2 μm PVDF membranes (Millipore, MA, USA). Following the protein bands were incubated overnight at 4 °C with corresponding primary antibodies: E-cadherin (Proteintech 1:5000), N-cadherin (Proteintech 1:2000), vimentin (Proteintech 1:2000), IGF2BP2 (Proteintech 1:2000), NUP160 (Proteintech 1:5000), P62 (Proteintech 1:5000), LC3-B (Cell Signaling Technology 1:1000), and β-actin (Cell Signaling Technology 1:1000). The strips were then incubated with horseradish peroxidase-labelled secondary antibodies for 1 h at room temperature (Beyotime 1:2000). Protein expression was detected using Alpha View software and quantified using ImageJ software. β-actin served as an endogenous reference.

Xenografts experiment in nude mice

Fifteen BALB/c nude mice (Zhejiang Weitong Lihua Laboratory Animal Technology Co., Ltd., Hangzhou, China) were divided into two groups: the A549 OE-NC group (n = 7) and the A549 OE-circRAPGEF5 group (n = 8). The cell suspension was adjusted to 10^7 cells/mL and subcutaneously injected into the right axilla of the nude mice to establish a subcutaneous tumor model. Each nude mouse was injected with 200 µL cell suspension, after which tumor formation was monitored every 2 days and the body weight and tumor volume of the mice were recorded. Tumor volume was calculated using the formula: ([L×W] ²)/2. After 24 days, the mice were euthanized under deep anesthesia, and the tumors were excised, photographed and weighed.

Statistical analysis

Statistical analysis of the entire dataset was conducted using SPSS software (version 26.0) and GraphPad Prism 8.0.1. First, the samples undergo Shapiro-Wilk normality and F-test homogeneity of variance tests. Under conditions of normal distribution and homogeneity of variance, non-paired samples are compared using t-tests, while paired samples use paired t-tests. For non-normal distributions, Mann-Whitney U tests or Wilcoxon tests are used. Data from three or more groups that follow a normal distribution are analyzed using One-way analysis of variance (ANOVA); otherwise, Kruskal-Wallis tests are conducted. The outcomes were depicted as bar charts or line graphs. Statistical significance was set at P < 0.05.

Results

High expression of circRAPGEF5 inhibits autophagy in LUAD cells

We previously demonstrated the elevated expression of circRAPGEF5 in LUAD patients, promoting LUAD proliferation, migration, invasion and EMT through the miRNA-126-3p/ZEB1 axis. Moreover, serum exosomal circRAPGEF5 has potential as a marker for identifying LUAD tumors [30]. Building on this foundation, we found that circRAPGEF5 is closely related to autophagy. CircRAPGEF5 knockdown increased autophagic lysosomes and autophagosomes, while circRAPGEF5 overexpression decreased these structures (Fig. 1 A). Autophagy-related protein analysis through Western blot (WB) experiments revealed that circRAPGEF5 knockdown decreased P62 and increased LC3-II/I, while circRAPGEF5 overexpression had the opposite effect (Fig. 1B). Autophagy inhibition by 3-methyladenine (3-MA) at 5 µM and induction by 50 nM rapamycin (RAPA) were validated in LUAD cells (Figure S1A, B). Consistent with previous studies [26], the autophagy inhibitor 3-MA facilitated EMT, migration, and invasion, whereas RAPA inhibited these processes in LUAD cells (Figure S1C-F). 3-MA reversed the inhibitory effects of circRAPGEF5 knockdown on proliferation, migration, invasion, and clone formation. Conversely, RAPA reversed the promoting effect of circRAPGEF5 overexpression on these cellular processes (Fig. 1 C–E). Our findings suggest that circRAPGEF5 suppresses autophagy in LUAD cells, thereby enhancing their proliferation, migration, invasion, and colony formation abilities.

Fig. 1.

CircRAPGEF5 inhibition of autophagy in LUAD cells. (A) Analysis of autophagy flow levels after circRAPGEF5 knockdown or overexpression. (B) Protein expression of P62 and LC3-II/I when circRAPGEF5 knockdown or overexpression. (C) Proliferation ability measurement of LUAD cells after circRAPGEF5 knockdown or overexpression, followed by the addition of 3-MA or RAPA. (D) Determination of colony formation ability after adding 3-MA or RAPA post circRAPGEF5 knockdown or overexpression. (E) Analysis of migration and invasion ability after adding 3-MA or RAPA following circRAPGEF5 knockdown or overexpression. Two data sets were compared with unpaired t-tests, data from three groups were compared using ANOVA. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001

M6A methylation-modified circRAPGEF5 binds to IGF2BP2 protein and is regulated by m6A methylation modification and IGF2BP2

To elucidate the molecular mechanism underlying circRAPGEF5-mediated autophagy, we conducted RNA pull down experiments. The circRAPGEF5 probe was prepared and its efficiency was verified through RNA pull down experiments (Fig. 2 A). Silver nitrate staining revealed distinct protein bands in the circRAPGEF5 probe group (Fig. 2B). Mass spectrometry and WB experiments demonstrated a significant binding of the circRAPGEF5 probe group to IGF2BP2 (Fig. 2 C). Concurrently, RIP experiments exhibited a notable increase in circRAPGEF5 enrichment in the IGF2BP2 antibody group compared to the control IgG group, affirming the interaction between IGF2BP2 and circRAPGEF5 (Fig. 2D, E). IF and FISH illustrated the predominant co-localization of IGF2BP2 with circRAPGEF5 in the cytoplasm (Fig. 2 F).

Fig. 2.

CircRAPGEF5 binds to the KH 3–4 region of IGF2BP2, m6A methylation modifies its expression. (A) Measurement of circRAPGEF5 probe enrichment using RT-qPCR. (B) Silver staining pattern after SDS-PAGE. (C) Detection of IGF2BP2 enrichment through Western blot. (D, E) RT-qPCR analysis for IGF2BP2 enrichment in circRAPGEF5 probe sets. (F) In situ fluorescence hybridization experiments and immunofluorescence co-localization of circRAPGEF5 and IGF2BP2 predominantly located in the cytoplasm. (G) MeRIP analysis for methylation levels in BEAS-2B cells versus H1299 cells. (H) MeRIP analysis for circRAPGEF5 methylation levels in A549 cells and H1299 cells. (I) CircRAPGEF5 expression levels after METTL3 knockdown. (J) RT-qPCR detection of mRNA expression of circRAPGEF5 after overexpression and knockdown of IGF2BP2. (K) After knockdown METTL3, RIP was used to detect the binding of IGF2BP2 and circRAPGEF5. (L) Truncated RIP performed in the H1299 overexpressing circRAPGEF5 cell line. (M) The expression of circRAPGEF5 in each group was detected by RT-qPCR. Two data sets were compared with unpaired t-tests, data from three or more groups were compared using ANOVA. ns: nonsense, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001

Utilizing the SRAMP tool (http://www.cuilab.cn/sramp) for predicting circRAPGEF5 methylation abundance unveiled numerous methylation sites in circRAPGEF5 (Figure S2A), possible m6A modification sites of circRAPGEF5 (Figure S2B), with higher overall m6A levels in H1299 cells compared to BEAS-2B cells (Fig. 2G). The meRIP assay detected a higher level of circRAPGEF5 methylation in A549 and H1299 cells, particularly in the anti-m6A group, compared to the IgG group (Fig. 2 H). RT-qPCR demonstrated that METTL3 knockdown decreased circRAPGEF5 levels (Fig. 2I). Moreover, knockdown or overexpression of IGF2BP2 inhibited or promoted circRAPGEF5 expression respectively (Fig. 2 J). Knockdown of METTL3 weakened the binding of IGF2BP2 and circRAPGEF5 (Fig. 2 K). To identify the binding site of IGF2BP2 and circRAPGEF5, six plasmids with different fragments of IGF2BP2 were constructed (Figure S2C), RIP experiments were conducted in the H1299 cell line overexpressing circRAPGEF5. The results indicated that IGF2BP2 is predominantly bound to circRAPGEF5 in the KH 3–4 region (Fig. 2 L). In order to verify the biological function of IGF2BP2 in KH3-4, we transfected WT and del KH3-4 IGF2BP2 plasmids after knocking down circRAPGEF5. Compared with WT-IGF2BP2, del KH3-4 plasmid weakened the response to circRAPGEFF5(Fig. 2 M). Cytological experiments also showed that the deletion of the KH3-4 region of IGF2BP2 plasmid weakened the proliferation and metastasis of LUAD cells (Figure S2D-G). These findings collectively suggest that circRAPGEF5 binds to IGF2BP2, and its expression is intricately regulated by m6A modification and IGF2BP2.

High IGF2BP2 expression promotes LUAD metastasis and progression. Knockdown of IGF2BP2 induces autophagy in LUAD cells

We observed elevated expression of IGF2BP2 in clinical paired tissue samples from the TCGA database, as illustrated in Figure S3A. Simultaneously, patients with high IGF2BP2 expression exhibited shorter overall survival, as depicted in Figure S3B. A549 cells displayed low IGF2BP2 levels, while H1299 and HCC827 cells exhibited high expression (Fig. 3 A). To investigate IGF2BP2’s role in LUAD, we constructed IGF2BP2 knockdown cell lines in H1299 and HCC827 cells and IGF2BP2 overexpression cell lines in A549 and HCC827 cells. Successful generation of IGF2BP2 overexpression cell lines in A549 and HCC827 cells is presented in Fig. 3B. Sh1-IGF2BP2 significantly reduced IGF2BP2 levels and it was utilized for subsequent experiments to knock down IGF2BP2 expression (Fig. 3 C).

Fig. 3.

IGF2BP2 promotes LUAD metastasis and progression. (A) RT-qPCR and western blot analysis of mRNA and protein expression of IGF2BP2 in normal human bronchial epithelial cells versus LUAD cells. (B) RT-qPCR and western blot validation of IGF2BP2 overexpression efficiency in A549 and HCC827 cells. (C) RT-qPCR and WB validation of the efficiency of knockdown of IGF2BP2 in H1299 and HCC827 cells. (D, F) EdU and CCK-8 proliferation assays after knockdown of IGF2BP2 in LUAD cells. (E) Colony formation analysis to determine colony formation ability after knockdown of IGF2BP2 in LUAD cells. (G, H) Transwell assay analysis of migration and invasion ability after IGF2BP2 knockdown in LUAD cells. (I) Western blot detection of epithelial-mesenchymal transition marker proteins after IGF2BP2 knockdown in LUAD cells. Two data sets were compared with unpaired t-tests, ns: nonsense, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001

CCK-8 and EdU assays revealed a significant increase in proliferation in both A549 and HCC827 cells after IGF2BP2 overexpression (Figure S3C, D). Conversely, the knockdown of IGF2BP2 resulted in a significant inhibition of proliferation in H1299 and HCC827 cells (Fig. 3D, F). Migration and invasion assays showed that IGF2BP2 overexpression significantly promoted migration and invasion of A549 and HCC827 cells (Figure S3F), while knockdown of IGF2BP2 inhibited migration and invasion of H1299 and HCC827 cells (Fig. 3G, H). Colony-formation assays demonstrated that IGF2BP2 overexpression increased colony formation (Figure S3E), knockdown of IGF2BP2 inhibited colony formation in H1299 and HCC827 cells (Fig. 3E). Western blot experiments indicated that IGF2BP2 overexpression promoted N-cadherin and vimentin expression and decreased E-cadherin expression in A549 and HCC827 cells (Figure S3G). In contrast, the knockdown of IGF2BP2 inhibited N-cadherin and vimentin expression, upregulated E-cadherin expression in H1299 and HCC827 cells (Fig. 3I).

It has been reported that IGF2BP2 facilitates NSCLC proliferation by stabilizing the lncRNA MALAT1 and increasing ATG12 [31]. Additionally, in colorectal cancer, IGF2BP2 promotes resistance to cisplatin by stabilizing the lncRNA TUC1 to regulate autophagy [32]. To investigate the potential association between IGF2BP2, a binding protein of circRAPGEF5, and autophagy in LUAD cells, we analyzed autophagy-related proteins and autophagic flow levels. These findings indicated that a reduction in IGF2BP2 results in a decrease in P62 and an increase in LC3-II/I (Fig. 4 A), and that knockdown of IGF2BP2 leads to elevated levels of autophagy lysosomes and autophagosomes (Fig. 4B). Notably, 3-MA reversed the inhibitory effects on cell proliferation, migration, and invasion after knockdown of IGF2BP2, whereas RAPA inhibits these processes in LUAD cells after overexpression of IGF2BP2 (Fig. 4 C-E). The experimental findings indicated that elevated expression of IGF2BP2 facilitates the proliferation, migration, invasion, and EMT of LUAD cells, whereas suppression of IGF2BP2 triggers autophagy in vitro.

Fig. 4.

IGF2BP2 inhibition of autophagy in LUAD cells. (A) Western blot analysis of P62 and LC3-II/I protein expression after IGF2BP2 knockdown. (B) Analysis of autophagy flow levels following IGF2BP2 knockdown. (C) Proliferation ability measurement of LUAD cells after IGF2BP2 knockdown followed by the addition of 3-MA. (D) Analysis of migration or invasion ability after adding 3-MA post IGF2BP2 knockdown. (E) Determination of colony formation ability after adding 3-MA post IGF2BP2 knockdown. Two data sets were compared with unpaired t-tests, data from three groups were compared using ANOVA. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001

IGF2BP2 synergizes with circRAPGEF5 to promote LUAD metastasis and progression

To substantiate the association between circRAPGEF5 and IGF2BP2, we conducted subsequent experiments. Specifically, two cell lines were established in HCC827 cells, one with the knockdown of IGF2BP2 after overexpression of circRAPGEF5 and the other with the knockdown of circRAPGEF5 in the overexpression IGF2BP2 cell line. CCK-8 and EdU assays (Fig. 5 A, B), migration and invasion assays (Fig. 5 C), colony formation assay (Fig. 5D) and WB assay (Fig. 5E) confirmed that the knockdown of IGF2BP2 reverses the effects of overexpression of circRAPGEF5 on LUAD cell proliferation, migration, invasion, colony formation, and EMT promotion. Conversely, overexpression of IGF2BP2 reversed the inhibitory effects of the knockdown of circRAPGEF5 on these processes in LUAD cells. These results suggest that IGF2BP2 and circRAPGEF5 synergistically promote the metastasis and progression of LUAD.

Fig. 5.

Rescue experiment to confirm that circRAPGEF5 and IGF2BP2 cooperate to promote LUAD progression. (A, B) EdU and CCK-8 assays to test the proliferative capacity of HCC827 cells with IGF2BP2 knockdown after circRAPGEF5 overexpression and IGF2BP2 overexpression after circRAPGEF5 knockdown. (C) Determination of migration and invasion abilities of HCC827 cells by Transwell assay. (D) Colony formation assay of the colony formation ability of HCC827 cells. (E) Western blot detection of EMT proteins in HCC827 cells. Data from three groups were compared using ANOVA. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001

IGF2BP2 interacts with NUP160 and stabilizes NUP160 mRNA. Knockdown of NUP160 induces autophagy in LUAD cells

To explore the target genes of IGF2BP2, we utilized the StarBase (https://starbase.sysu.edu.cn/) database for prediction. The results indicated that NUP160 may be a crucial molecule for IGF2BP2 action and is positively correlated with IGF2BP2 (Fig. 6 A). NUP160 was highly expressed in LUAD patient tissues (Figure S4A, C), immunostaining revealed a negative association between NUP160 expression and immune infiltration (Figure S4B). FISH and IF double-staining showed that NUP160 co-localized predominantly in the cytoplasm with circRAPGEF5 (Fig. 6B). RIP experiments demonstrated a significant increase in NUP160 mRNA enrichment in the IGF2BP2 antibody group compared with the control IgG group, indicating an interaction between IGF2BP2 and NUP160 (Fig. 6 C). NUP160 mRNA was expressed at low levels in A549 cells and at high levels in H1299 and HCC827 cells (Fig. 6D), therefore, H1299 and HCC827 cells were selected for the next study. We found that sh1-NUP160 effectively reduced NUP160 levels in H1299 and HCC827 cells, it was used for the knockdown of NUP160 expression (Fig. 6E). Subsequently, NUP160 knockdown significantly inhibited LUAD cell proliferation, migration, invasion, EMT and colony-forming ability (Fig. 6 F-J). The Co-IP experiment ruled out the possibility of interaction between IGF2BP2 and NUP160 protein (Figure S4D). Additionally, an actinomycin D assay revealed that IGF2BP2 plays a role in stabilizing NUP160 mRNA expression in the cells (Fig. 6 K).

Fig. 6.

NUP160 is a downstream target of IGF2BP2, NUP160 promotes LUAD metastasis and progression. (A) Positive correlation between IGF2BP2 and NUP160 in LUAD. (B) FISH and immunofluorescence staining revealing co-localization of NUP160 and circRAPGEF5, mainly in the cytoplasm. (C) RIP of circRAPGEF5 with NUP160 antibody enrichment in H1299 and HCC827 cells. (D) NUP160 mRNA expression in BEAS-2B cells and LUAD cells. (E) RT-qPCR verification of NUP160 knockdown efficiency in H1299 and HCC827 cells. (F, G) EdU and CCK-8 assays of cell proliferation after NUP160 knockdown. (H) Transwell assay of cell invasion and migration abilities after NUP160 knockdown. (I) Colony formation assay after knockdown of NUP160. (J) Western blot detection of EMT marker proteins in HCC827 cells. (K) Measurement of NUP160 mRNA stability using actinomycin D. Two data sets were compared with unpaired t-tests, data from three or more groups were compared using ANOVA.*P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001

Previous studies have demonstrated that NUP160 knockdown inhibits cell proliferation and induces apoptosis and autophagy [11]. To further investigate this phenomenon, we examined the autophagy-related proteins P62 and LC3-II/I. The results showed that P62 decreased and LC3-II/I increased after NUP160 knockdown (Fig. 7 A). The number of autophagosomes and lysosomes increased after NUP160 knockdown (Fig. 7B). Prior investigations into autophagy have revealed that the Akt/mTOR pathway is a crucial pathway that suppresses autophagy [33, 34]. Knocking down NUP160 inhibited the PI3K/AKT/mTOR pathway(Figure S4E). The autophagy inhibitor 3-MA reversed the inhibitory effects on cell proliferation, migration, and invasion after the knockdown of NUP160 (Fig. 7 C-E). These findings collectively suggest that IGF2BP2 combines with NUP160 and stabilizes NUP160 mRNA expression, and that NUP160 knockdown induces autophagy in LUAD cells in vitro.

Fig. 7.

Knockdown of NUP160 promotes autophagy in LUAD cells. (A) Western blot analysis of P62 and LC3-II/I protein expression after NUP160 knockdown. (B) Analysis of autophagy flow levels following NUP160 knockdown. (C) Proliferation ability measurement of LUAD cells after NUP160 knockdown, followed by the addition of 3-MA. (D) Analysis of migration or invasion ability after adding 3-MA post NUP160 knockdown. (E) Determination of colony formation ability after adding 3-MA post NUP160 knockdown. Two data sets were compared with unpaired t-tests, data from three groups were compared using ANOVA. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001

Knocking down NUP160 can reverse the pro-autophagy effects of circRAPGEF5, and the role of overexpression of IGF2BP2 in promoting LUAD metastasis and progression

We have confirmed that knocking down NUP160 can reverse the pro-autophagy effects of circRAPGEF5 (Fig. S5A, B). Meanwhile, reverse experiments have confirmed that knocking down NUP160 can reverse the function of IGF2BP2 in promoting cancer progression. Cell function assays were performed in H1299 and HCC827 cells after NUP160 knockdown and IGF2BP2 overexpression. CCK-8 and EdU assays revealed that the knockdown of NUP160 reversed the proliferation-promoting effect of IGF2BP2 overexpression in LUAD cells (Fig. 8 A, B). Migration and invasion assays revealed that the knockdown of NUP160 reversed the promoting migration and invasion abilities of IGF2BP2 overexpression in LUAD cells (Fig. 8 C, D). The colony formation assay revealed that NUP160 knockdown reversed the promoting colony formation of IGF2BP2 overexpression in LUAD cells (Fig. 8E).

Fig. 8.

Rescue experiment to confirm that IGF2BP2 cooperates with NUP160 to promote LUAD progression. (A, B) CCK-8 and EdU assays revealed that the knockdown of NUP160 reversed the proliferation-promoting effect of IGF2BP2 overexpression in H1299 and HCC827 cells. (C, D) Transwell assay to analyze the migration and invasion abilities of H1299 and HCC827 cells after NUP160 knockdown following IGF2BP2 overexpression. (E) Colony formation assay of H1299 and HCC827 cells after NUP160 knockdown following IGF2BP2 overexpression. Data from three groups were compared using ANOVA. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001

circRAPGEF5/IGF2BP2/NUP160 inhibit LUAD cell autophagy in vivo

We previously established tumor transplantation models of OE-NC and OE-circRAPGEF5 in nude mice to investigate the circRAPGEF5 ceRNA mechanism. Subsequently, we extracted RNA and proteins from the tumor tissues of these mice and conducted RT-qPCR and WB experiments. The findings indicated that in vivo overexpression of circRAPGEF5 leads to an increase in the expression of IGF2BP2 and NUP160 mRNA (Fig. 9 A). WB blotting showed elevated levels of P62 and IGF2BP2, whereas the LC3-II/I ratio decreased in the circRAPGEF5 overexpression group (Fig. 9B). Additionally, the N-cadherin and VIMENTIN proteins were upregulated (Fig. 9B). We collected 93 LUAD samples from the First Affiliated Hospital of Wenzhou Medical University, the basic information of the patients is shown in Table S1-S3; RT-qPCR results showed upregulation in the expression of circRAPGEF5/IGF2BP2/NUP160 (Fig. 9 C-E). Furthermore, we also analyzed the correlation between circRAPGEF5 / IGF2BP2 / NUP160 (Fig. 9 F-H). Patients with high expression of circRAPGEF5, IGF2BP2 and NUP160 had shorter survival (Figure S5C-E). Figure I is the HE stains plot of a genetically engineered mouse primary tumor model (Figure.9I). This schematic diagram illustrates the potential molecular mechanism of circRAPGEF5 in LUAD regarding the regulation of tumor progressiveness (Fig. 9 J).

Fig. 9.

In vivo experiments confirm that circRAPGEF5/IGF2BP2/NUP160 promotes LUAD metastasis and progression by inhibiting autophagy. (A) RT-qPCR analysis of RNA expression in tumor tissues of nude mice. (B) Western blot of IGF2BP2, NUP160, P62, N-cadherin, vimentin, and LC3-II/I expression in the tumor tissues of nude mice. (C-E) RT-qPCR detection of circRAPGEF5 (C), IGF2BP2 (D) and NUP160 (E) expression in 93 LUAD tissues compared with paired paracancerous tissues. (F) Correlation analysis between circRAPGEF5 and IGF2BP2. (G) Correlation analysis between circRAPGEF5 and NUP160. (H) Correlation analysis between NUP160 and IGF2BP2. (I) The HE staining of genetically engineered mice. (J) Schematic representation of the potential molecular mechanism of circRAPGEF5 in LUAD regarding the regulation of tumor progressiveness. The expression of genes and proteins in tumor tissues of nude mice was compared by unpaired t-tests, and the differences of clinical LUAD samples were analyzed by Kruskal-Wallis test. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001

Discussion

Autophagy, a ubiquitous process observed in nearly all eukaryotic organisms, plays a crucial role in maintaining cellular material homeostasis, especially in cancer cells. Previous research has highlighted the involvement of various circular RNAs (circRNAs) in modulating autophagy-related processes in different cancers. For instance, circ_0010235 acts as an miR-433-3p sponge, regulating TIPRL levels and influencing cell proliferation, autophagy, and migration in lung cancer [35]. Similarly, circ_100565 knockdown impedes autophagy and reduces cisplatin resistance in lung cancer by acting as an miR-377-3p sponge and upregulating ADAM28 levels [36]. In laryngeal squamous cell carcinoma (LSCC), circPARD3 inhibits autophagy via the PRKCI-Akt-mTOR pathway, thereby promoting LSCC progression and chemoresistance [37]. Therefore, the circRNA-autophagy axis plays a critical role in the proliferation, metastasis, invasion, and drug or radiation resistance of various tumour cells [23, 35, 36]. In our study, we found that high circRAPGEF5 expression inhibited autophagy in LUAD cells, prompting further exploration.

Previously, we demonstrated that circRAPGEF5 promotes LUAD progression through a ceRNA mechanism. In this study, our focus was primarily on exploring the RBPs of circRAPGEF5. RNA pulldown assays and immunoprecipitation experiments revealed a binding interaction between circRAPGEF5 and IGF2BP2. Previous studies have implicated IGF2BP2 in cancer metastasis. In colorectal cancer, IGF2BP2 enhances HMGA1 mRNA stability and protein expression, facilitating colorectal cancer cell proliferation [38]. Similarly, circCD44 promotes triple-negative breast cancer proliferation, migration, invasion, and tumorigenesis, partly by sponging miR-502-5p and interacting with IGF2BP2 [39]. Additionally, IGF2BP2 functions as a “reading protein” for m6A methylation modification by binding to target RNA [40]. We hypothesized that circRAPGEF5 might also bind to IGF2BP2 for m6A modification, leading us to investigate the m6A modification of circRAPGEF5. Our m6A site prediction analysis revealed hypermethylated sites in circRAPGEF5, confirmed by MeRIP experiments. Exploring the truncator region identified the primary binding site, where deletion of the IGF2BP2 KH 3–4 region significantly reduced the binding capacity to circRAPGEF5 compared with the RM 1–2 and KH 1–2 regions. This suggests that the IGF2BP2 KH 3–4 region is the primary binding site for both proteins. Investigating the role of circRAPGEF5 in autophagy in LUAD cells, our findings revealed that circRAPGEF5 overexpression impeded autophagy, whereas circRAPGEF5 knockdown facilitated autophagy in LUAD cells, as evidenced by WB analysis and autophagic flux monitoring. Simultaneous, autophagy agonist and inhibitor construct reversion experiments confirmed the association of circRAPGEF5 with autophagy and characterized the role of autophagy in LUAD cell function.

By binding to RNA, RBPs typically modulate the splicing of circRNAs outside the nucleus and stabilize RNA expression [41–43]. We experimentally found that IGF2BP2 governs circRAPGEF5 expression in vitro and hypothesized that IGF2BP2 plays a role as an RBP, directly engaging in the reverse splicing process of circRAPGEF5, stabilizing circRAPGEF5, regulating its expression. To investigate the role of IGF2BP2 in LUAD, a series of in vitro cytological experiments were conducted. These results indicated that IGF2BP2 suppresses autophagy in LUAD cells and facilitates their proliferation, migration, invasion and EMT. Rescue experiments validated the correlation between circRAPGEF5 and IGF2BP2 expression. Subsequently, we investigated the downstream targets of circRAPGEF5-IGF2BP2. Using the StarBase database, we predicted the mRNA associated with IGF2BP2 and discovered a positive correlation between IGF2BP2 expression and NUP160. The binding of IGF2BP2 to NUP160 was confirmed by RIP and IF. Previous research has indicated that in diabetic nephropathy mice, NUP160 expression is upregulated, leading to the inhibition of autophagy and exacerbation of the inflammatory response in rat renal tubular epithelial cells. Conversely, NUP160 knockdown inactivates the NF-κB pathway, thereby inducing autophagy [12]. Our study validated the finding that knockdown of NUP160 leads to an increase in autophagy and a decrease in the proliferation, migration, and invasion abilities of LUAD cells. This finding is consistent with that of Wang P [11]. Typically, the expression level of P62 is regulated by autophagy. The upregulation of P62 results from the inhibition of autophagy, while its downregulation is due to the activation of autophagy. Conversely, autophagy can also regulate P62. The ZZ domain of P62 selectively recognizes arginine substrates and induces autophagy [44, 45]. P62 could be used as a biomarker for the evaluation of mTOR-dependent autophagy defects [46]. The depletion of NUP160 inhibited the PI3K/AKT/mTOR signaling pathway, induced autophagy in LUAD cells. Additionally, our rescue experiments demonstrated that the knockdown of NUP160 counteracts the promoting effect of IGF2BP2 overexpression in LUAD cells. The IGF2BP family stabilizes downstream target mRNAs through m6A modifications [31, 38, 46]. To further investigate this mechanism, we conducted actinomycin D stabilization experiments, revealing that IGF2BP2 stabilizes NUP160 mRNA expression.

Our study provides evidence for the binding of IGF2BP2 to circRAPGEF5. This indicates that circRAPGEF5 promotes LUAD metastasis through IGF2BP2/NUP160-mediated autophagy. Furthermore, our investigation using a nude mouse tumour transplantation model revealed the upregulation of IGF2BP2 and NUP160 expression in the tumor tissues of mice overexpressing circRAPGEF5. Expression of the autophagy marker P62 was observed to decrease, while the LC3-II/I ratio increased. The experimental results validated that circRAPGEF5/IGF2BP2/NUP160 hinders autophagy in mouse LUAD cells, promoting the invasion and progression of LUAD.

Conclusions

In conclusion, our research results indicate that m6A-methylated circRAPGEF5 binds to IGF2BP2, inhibiting autophagy in LUAD cells through the IGF2BP2/NUP160 signaling axis. This process promotes LUAD cell proliferation, invasion, and migration, ultimately contributing to LUAD metastasis and progression. These findings suggest that circRAPGEF5 could serve as a potential target for the development of novel therapeutic interventions for LUAD.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Abbreviations

3-MA

3-methyladenine

RAPA

Rapamycin

CCK-8

Cell counting kit-8

CeRNA

competing endogenous RNA

CircRNA

Circular RNA

EMT

Epithelial-mesenchymal transition

FBS

Fetal bovine serum

FISH

Fluorescence in situ hybridization

IF

Immunofluorescence

Ig

Immunoglobulin

Kd

Knockdown

LncRNA

Long non-coding RNA

LSCC

Laryngeal squamous cell carcinoma

LUAD

Lung adenocarcinoma

m6A

Methylation at the sixth position of adenosine

MeRIP

Methylated RNA immunoprecipitation

MiRNA

microRNA

NSCLC

Non-small-cell lung cancer

OE

Overexpression

PBS

Phosphate-buffered saline

qRT-PCR

Quantitative real-time reverse transcription polymerase chain reaction

RBPs

RNA-binding proteins

RIP

RNA immunoprecipitation

WB

Western blot

Author contributions

L.L., T.H., and C.K. contributed to the conception and design of the study; Y.D., L.H., and Y.C. contributed to the acquisition and analysis of the data; Z.H., K.H., J.C. and Y.W. contributed to drafting a significant portion of the manuscript and figures.

Funding

This study was financially supported by the Zhejiang Provincial Natural Science Foundation (ZCLY24H2001), the Discipline Cluster of Oncology, Wenzhou Medical University, China (NO.z1-2023006), the Zhejiang Provincial Health Planning Commission (2023KY902), the Key Laboratory of Clinical Laboratory Diagnosis and Translational Research of Zhejiang Province (2022E10022), and the Zhejiang Provincial Research Center for Cancer Intelligent Diagnosis and Molecular Technology (JBZX-202003).

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

All animal procedures were approved by the Experimental Animal Ethics Committee of the First Hospital of Wenzhou Medical University (WYYY-AEC-2021-310).

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.