GES-1 cell line was supplied by the Cell Bank of Chinese Academy of Sciences. BGC-823 cell line was purchased from Beyotime. They were cultured in RPMI-1640 medium (Hyclone), supplemented with 10% fetal bovine serum (Gibco), 0.1 mg/mL streptomycin, and 100 U/mL penicillin. All these cells were incubated at 37°C with a 5% CO₂ concentration.

The full-length sequence of circ-malat1 was amplified by PCR and inserted into the circular RNA expression vector pcD25-ciR. The 3′UTR (wild type or mutant) sequence of the CCND2 gene and the sequence of circ-malat1 (wild type or mutant) were also amplified and inserted into the dual-luciferase reporter vector psi-check2, respectively. Circ-malat1 siRNA, miR-154-5p mimic, and inhibitor were synthesized by Genepharma. After the vector construction is completed or the small RNA is synthesized, the cells are transfected using Lipofectamine 2000 (Invitrogen), following the procedure outlined in the instructions.

Total cellular RNA was extracted using Trizol reagent (Vazyme) following the manufacturer’s instructions. Subsequently, 1 μg of total RNA, either treated with RNase R (Epicenter Technologies) or untreated, was reverse transcribed into cDNA using the PrimeScript™ FAST RT reagent Kit with gDNA Eraser (Takara), following the manufacturer’s instructions. qRT-PCR was performed using SYBR Green (Vazyme) on an ABI qPCR system, and fold changes were determined using the relative quantification 2^−ΔΔCT method. Additionally, the nuclear and cytoplasmic fractions of cells were separated using the PARIS™ Kit (Invitrogen™) following the manufacturer’s instructions. Expression ratios of specific RNA molecules between the nuclear and cytoplasmic fractions were subsequently assessed by qRT-PCR. The primer sequences used are listed in Supplementary Table S1.

RIPA lysis buffer containing PMSF and protease inhibitors was added to the cells, and the protein concentration was measured using a BCA kit (Beyotime). Equal amounts of total protein were loaded into each well for SDS-PAGE gel electrophoresis, followed by transfer to a methanol-activated PVDF membrane. After blocking with skim milk at room temperature for 1 hour, the membrane was incubated with the corresponding primary antibody overnight at a low temperature. The membrane was then incubated with the appropriate secondary antibody for 1 hour at room temperature and developed using an ECL luminescent solution (Beyotime). The antibody information is listed in Supplementary Table S2.

RNA immunoprecipitation was performed according to the protocol provided with the Geneseed kit. Briefly, cells were lysed using a solution containing protease inhibitors and RNase inhibitors. The lysate was incubated for 10 min, followed by low-temperature centrifugation to obtain the supernatant. A 50 μL aliquot of the supernatant was reserved as the input group, mixed with loading buffer, boiled, and stored at low temperature for subsequent protein immunoblotting. An additional 100 μL of the supernatant was taken directly for RNA extraction. Next, 200 μL of protein A + G beads were washed three times with buffer, thoroughly mixed with fresh buffer, then divided into two equal parts, with Ago2 antibody added to one part and IgG antibody added to the other. The mixtures were rotated at low temperatures for 2 h, followed by multiple washes, discarding the supernatant each time. Afterward, the Ago2 and IgG antibody-bound beads were incubated with fresh lysis solution and rotated slowly at low temperature overnight. Using a magnetic stand, the supernatant was discarded, and the beads were washed three times with washing solution. In the final wash, 100 μL of the magnetic bead complex was collected, placed on the magnetic stand, and the supernatant was discarded. Loading buffer was added to the beads, boiled, and placed back on the magnetic stand, and the supernatant was collected for protein immunoblotting analysis. The remaining magnetic bead complex was used for RNA extraction, followed by qRT-PCR detection.

The growth status of BGC-823 cells was measured by a CCK-8 kit (Solarbio) following the manufacturer’s instructions. In brief, the cells were seeded on the 96 well plates. After transfection for about 24 h, add 10 μL CCK-8 reagent to each well and continue to incubate for 4 h. To detect the absorbance value at 450 nm by using a microplate reader (Tecan).

Seed 3,000 cells in the growth phase per well into a 96-well plate. Twenty hours after transfection, perform the following operations according to the EDU kit instructions (RiboBio). Briefly, add EdU-containing medium (1:1,000) and continue culturing for 4 hours. Then, fix the cells with a solution containing 4% paraformaldehyde for 30 min. Stain the cells with Apollo staining solution for 30 min, and after washing, stain the nuclei with Hoechst 33,342 staining solution for 10 min. Finally, observe the cells under a fluorescence microscope.

Inoculate the counted cells into a 60 mm culture dish. Once the cells have adhered and grown stably, further process them according to experimental requirements. When the cell density reaches approximately 80%, collect and fix the cells using a PBS solution containing 70% ethanol. The cell cycle is then analyzed by flow cytometry (BD Biosciences) according to the instructions provided with the cell cycle kit (MedChemExpress, HY-K1071).

The counted cells were seeded into a 6-well plate. After the cells had grown stably, they were processed according to the experimental requirements. The adherent cells were then digested with EDTA-free trypsin and washed with PBS. Subsequently, follow the instructions of the apoptosis kit (MultiSciences, AP101) to analyze cell apoptosis using flow cytometry.

Wild-type or mutant circ-malat1 (or CCND2) dual-luciferase reporter vectors were co-transfected with miR-154-5p mimics in HEK293T cells, respectively. Twenty-four hours after transfection, luciferase activity was analyzed using a dual-luciferase reporter assay kit (Beyotime), following the manufacturer’s instructions. Firefly luciferase activity served as the control.

The data in the experiment are presented as means ± SEM. Significance was analyzed using t-tests with GraphPad Prism software (*p < 0.05; **p < 0.01).

In this study, we compared the expression levels of circ-malat1 in gastric cancer cells and normal cells. We found that the expression level of circ-malat1 was significantly higher in gastric cancer cells compared to normal cells (Figure 1A). To identify endogenous circ-malat1, we designed convergent primers located within the circRNA sequence and a divergent primer pair spanning the back-splice junction of the circRNA. This design ensures the specific amplification of the circular form of the RNA, distinguishing it from the linear counterpart (Figure 1B). RNA was isolated and extracted from the nucleoplasm to detect circRNA expression, and circ-malat1 was found to be mainly expressed in the cytoplasm (Figure 1C). After treating the RNA with RNase R, the expression of linear RNA was significantly reduced, while circ-malat1 remained stably expressed (Figure 1D). This result suggests that circ-malat1 may play an important role in gastric cancer.

Research has shown that the expression level of circ-malat1 is significantly higher in gastric cancer cells compared to normal gastric epithelial cells. To further investigate the role of circ-malat1 in the growth of gastric cancer cells, several functional experiments were conducted. After transfection with the circ-malat1 overexpression vector or siRNA for 24 h, the efficiency of circ-malat1 expression was assessed using qRT-PCR (Figures 2A,B). Using CCK-8, it was found that overexpressing circ-malat1 exhibited a significantly faster proliferation rate of gastric cancer cells (Figure 2C). In contrast, knocking down circ-malat1 expression significantly reduced the proliferation capacity of gastric cancer cells (Figure 2D). EdU results demonstrated that overexpression of circ-malat1 increased the number of S phase cells (Figure 2E), while knockdown of circ-malat1 reduced the number of S phase cells (Figure 2F). The cell cycle assay showed that overexpression of circ-malat1 decreased the number of G1 phase cells (Figure 2G) whereas knockdown of circ-malat1 increased the number of G1 phase cells and reduced the S phase population (Figure 2H). Flow cytometry analysis of apoptosis rates revealed that overexpression of circ-malat1 has no significant effect on the apoptosis of gastric cancer cells (Figure 2I), while knockdown of circ-malat1 promoted apoptosis (Figure 2J). This finding underscores the crucial role of circ-malat1 in enhancing gastric cancer cell survival.

The expression pattern of circ-malat1 in cells was analyzed using nuclear-cytoplasmic fractionation, revealing its predominant localization in the cytoplasm (Figure 1C). The hypothesis suggests that circ-malat1 might function as a competitive endogenous RNA (ceRNA) based on the mechanism of circRNA action. The bioinformatics predictions dataindicates the presence of 81 miRNAs binding sites in circ-malat1 (Supplementary Table S3). To further validate that circ-malat1 functions as a ceRNA, RNA immunoprecipitation (RIP) revealed enrichment of circ-malat1 by the AGO2 protein compared to the IgG group (Figure 3A). To confirm the direct binding and sequestration of specific miRNAs by circ-malat1, we constructed luciferase reporter vectors containing either the wild-type or mutant circ-malat1 sequences, in which the miR-154-5p seed sequence binding sites were mutated(Figure 3B). Based on the candidate miRNA function in gastric cancer, we selected miR-154-5p for further research (Song et al., 2018). miR-154-5p mimic and luciferase reporter vectors were co-transfected into cells, and the relative luciferase activity data are presented in Figure 3C. Compared with the control group, miR-154-5p overexpression significantly decreased wild-type luciferase activity, while it had no noticeable effect on the activity of the mutant-type luciferase. These findings collectively suggest that circ-malat1 functions as a ceRNA by competitively binding to miR-154-5p.

Bioinformatics tool (TargetScan) was used to predict the binding sites of miR-154-5p. The prediction results indicate that there are binding sites for miR-154-5p in the 3′UTR of CCND2 mRNA (Figure 4A). Previous studies have demonstrated that disrupting CCND2 expression can inhibit the proliferation of gastric cancer cells, which heightens the significance and relevance of our research (Zhang et al., 2013). To validate the direct interaction between miR-154-5p and the 3′UTR of CCND2, the 3′UTR of CCND2 containing the predicted miR-154-5p binding sites (wild type or mutant type) were inserted into a luciferase reporter plasmid, respectively (Figure 4B). miR-154-5p mimic and luciferase reporter plasmid were co-transfected into gastric cancer cells. The results showed that co-transfected with miR-154-5p mimic and wild-type CCND2 3′UTR reduced luciferase activity, but not with the mutant 3′UTR (Figure 4C). To further validate the relationship miR-154-5p with CCND2, CCND2 expression at mRNA and protein levels was detected after transfection of miR-154-5p mimic or inhibitor. The results showed that overexpression of miR-154-5p decreased the CCND2 protein level (Figure 4D). However, the knockdown of miR-154-5p increased the CCND2 protein level (Figure 4E). The relationship between miR-154-5p and CCND2 in gastric cancer was further elucidated.

To validate whether circ-malat1 regulates gastric cancer development through the miR-154-5p/CCND2 axis, overexpression or knockdown of circ-malat1 was performed, followed by detection of miR-154-5p or CCND2 expression using qRT-PCR. Overexpression of circ-malat1 does not affect the RNA level of miR-154-5p but promotes the mRNA level of CCND2 (Figure 5A). Conversely, the knockdown of circ-malat1 does not affect the RNA level of miR-154-5p while inhibiting the mRNA level of CCND2 (Figure 5B). To further validate the relationship between circ-malat1 and miR-154-5p/CCND2 in the regulation of gastric carcinogenesis. Compared to the circ-malat1 overexpression group, co-overexpression of circ-malat1 and miR-154-5p revealed that the proliferation ability of gastric cancer cells was inhibited, as shown by the CCK-8 assay (Figure 5C). Flow cytometry results indicated G₁ phase cell cycle arrest (Figure 5D) and an increase in the number of apoptotic cells (Figure 5E). Furthermore, co-overexpression of Circ-malat1 and interference with CCND2 siRNA inhibited the proliferation of gastric cancer cells compared to the circ-malat1 overexpression group (Figure 5C). There was an increase in G₁ phase cells (Figure 5D) and an increase in the number of apoptotic cells (Figure 5E). These results indicate that circ-malat1 regulates gastric carcinogenesis through the miR-154-5p/CCND2 axis.