m6A-Modified circRAPGEF1 Interaction with IGF2BP3 Promotes Hepatocellular Carcinoma Progression via Reprogramming Aspartate Metabolism

Abstract

Hepatocellular carcinoma (HCC) progression and therapy sensitivity are critically fueled by liver cancer stem cells (LCSCs), yet the regulatory mechanisms of circular RNAs (circRNAs) on LCSCs remain elusive. Here, through circRNA microarray analysis of LCSCs and non-stem HCC cells, circRAPGEF1 is identified as a LCSC-enriched circRNA upregulated in HCC tissues and predictive of poor patient survival. Functionally, circRAPGEF1 promoted the stemness properties, proliferation, and tumorigenicity of HCC cells. Mechanistically, the METTL3-mediated N⁶-methyladenosine (m⁶A) modification of circRAPGEF1 facilitated KH domain-dependent binding of IGF2BP3 to its UGGAC motif, which conferring stability to circRAPGEF1 while competitively disrupting the IGF2BP3/ASS1 mRNA interaction. This process led to the degradation of ASS1 mRNA, triggering aspartate accumulation and activation of the S6K/CAD signaling pathway. Crucially, circRAPGEF1 overexpression reduced the sorafenib sensitivity, whereas targeting circRAPGEF1 using nanoparticles-mediated systematic siRNAs delivery effectively sensitized HCC cells to sorafenib. Collectively, these findings unveil a METTL3/circRAPGEF1/IGF2BP3/ASS1 regulatory axis that drives aspartate metabolic reprogramming to fuel HCC stemness properties, positioning circRAPGEF1 as a dual prognostic biomarker and therapeutic target to enhance sorafenib efficacy in HCC.

Keywords: aspartate metabolism; circular RNAs; hepatocellular carcinoma; liver cancer stem cells; m6A modification.

Figure 1

circRAPGEF1 is upregulated and associated with poor prognosis in HCC patients. A) Schematic diagram of circRNAs screening in CD133‐positive LCSCs. B) qRT‐PCR analysis of differentially expressed circRNAs expression in adherent monolayer cells and cell spheres of HCC cells. C) Agarose gel electrophoresis of PCR products with divergent and convergent primers in cDNA and gDNA amplification. D) Schematic diagram of the back‐splicing structure of circRAPGEF1 and Sanger sequencing results of PCR products. E) qRT‐PCR analysis of circRAPGEF1 and host gene RAPGEF1 mRNA expression in reverse transcription with oligo (dT) or random primers. F) qRT‐PCR analysis of the circRAPGEF1 and RAPGEF1 mRNA expression post‐RNase R treatment. G) qRT‐PCR analysis of circRAPGEF1 and RAPGEF1 mRNA expression over time with Actinomycin D treatment (5 µg mL⁻¹) in HCC cells. H) qRT‐PCR analysis of circRAPGEF1 expression in nucleus and cytoplasm of HCC cells. I) Representative FISH images of circRAPGEF1 in HCC cells. Scale bar: 10 µm. J) qRT‐PCR analysis of circRAPGEF1 expression in HCC tissues versus NATs (n = 30). K) Kaplan–Meier analysis of the recurrence‐free survival and overall survival of HCC patients with low versus high circRAPGEF1 expression (n = 74). The cutoff is the median. Data are presented as the means ± SD and analyzed by Student's t‐test. * p < 0.05; ** p < 0.01; *** p < 0.001; **** p < 0.0001; ns: not significant.

Figure 2

circRAPGEF1 enhances stemness properties and proliferation in HCC. A) qRT‐PCR analysis of circRAPGEF1 and RAPGEF1 expression in HuH‐7 cells transfected with control or circRAPGEF1‐targeting siRNAs and Vector or circRAPGEF1 overexpression lentivirus. B,C) Representative images and quantification of tumor sphere formation assays in HCC cells with the indicated treatments. Scale bar: 100 µm. D) Western blot of CD133 expression in HCC cells with the indicated treatments. E) Image of HCC tumors formed after subcutaneous implantation of different numbers of stably overexpressed circRAPGEF1 and control HuH‐7 cells in BALB/c nude mice (n = 5). F) Extreme limited dilution analysis for LCSC proportions. G) Weights of subcutaneous HCC tumors in the 8 × 10⁵ HuH‐7 cell implantation group. H,I) Representative images and quantification of IHC staining for H&E, CD133, and Ki‐67 in tumor specimens. Scale bar: 100 µm. J–L) Image, volumes, and weights of PDX tumors in mice with the indicated treatments (n = 5). M,N) Representative images and quantifications of IHC staining of H&E, Ki‐67, and CD133 in PDX specimens. Scale bar: 100 µm. Data are presented as mean ± SD and analyzed by Student's t‐test or one‐way ANOVA with Tukey's multiple comparison test. * p < 0.05; ** p < 0.01; *** p < 0.001; ns: not significant.

Figure 3

circRAPGEF1 interacts with IGF2BP3 in HCC cells. A) Silver staining image of RNA pull‐down assay with circRAPGEF1 and negative control (NC) probes. B) Venn diagram of potential RBPs binding circRAPGEF1 in mass spectrometry analysis and databases. C) Mass spectrometry analysis of IGF2BP3 peptides pulled down by circRAPGEF1 probe. D) Western blot analysis of the interaction between IGF2BP3 and circRAPGEF1 in HCC cells post RNA pull‐down assays. E) qRT‐PCR analysis of circRAPGEF1 enrichment in IGF2BP3 RIP assays. F) Representative images of circRAPGEF1 (Red) and IGF2BP3 (Green) co‐localization in HCC cells. Scale bar: 20 µm. G) Diagram predicting IGF2BP3 and circRAPGEF1 interaction via catRAPID. H) Schematic diagram of IGF2BP3 structural domain truncated constructs. I) Western blot analysis of Flag‐labeled IGF2BP3 truncations enrichment post RNA pull‐down assays in HepG2 cells. J) Western Blot analysis of IGF2BP3 expression in HCC cells with circRAPGEF1 silencing or overexpression. K) qRT‐PCR analysis of IGF2BP3 and circRAPGEF1 expression in HCC cells transfected with IGF2BP3‐targeting siRNAs. Data are presented as mean ± SD and analyzed by Student's t‐test or one‐way ANOVA with Tukey's multiple comparison test. ** p < 0.01; *** p < 0.001; ns: not significant.

Figure 4

METTL3/IGF2BP3 enhances the stability of m⁶A‐modified circRAPGEF1. A) qRT‐PCR analysis of circRAPGEF1 enrichment in m⁶A RIP assays in HCC cells. B) qRT‐PCR analysis of the circRAPGEF1 expression over time in HCC cells transfected with IGF2BP3‐targeting siRNAs. C) Schematic diagram of circRAPGEF1 sequence labeled with IGF2BP3‐recognizing m⁶A motif. D) Relative luciferase activity of HEK‐293T cells with the indicated treatments. E–G) qRT‐PCR and Western Blot analysis of IGF2BP3 and METTL3 expression in adherent cells and sphere cells of HCC cells. H) qRT‐PCR analysis of circRAPGEF1 enrichment in IGF2BP3 RIP assays in HepG2 cells with METTL3 silencing or STM2457 treatment (40 µm). I,J) qRT‐PCR analysis of circRAPGEF1 expression in HCC cells with the indicated treatments. Data are presented as mean ± SD and analyzed by Student's t‐test or one‐way ANOVA with Tukey's multiple comparison test. * p < 0.05; ** p < 0.01; *** p < 0.001; ns: not significant.

Figure 5

circRAPGEF1 decreases the stability of ASS1 mRNA in HCC. A) Heatmap of DEGs in HepG2 cells with control or circRAPGEF1 overexpression. B) A Venn diagram showing the co‐regulated genes by circRAPGEF1 and IGF2BP3 in GSE90684 and ENCORI database. C–F) qRT‐PCR analysis of ASS1 and OCC1 expression in HCC cells with circRAPGEF1 silencing (C,D) and overexpression (E,F). G) Western blot analysis of ASS1 expression in HCC cells with the indicated treatments. H) qRT‐PCR analysis of ASS1 mRNA enrichment in IGF2BP3 RIP assay in HepG2 cells. I) qRT‐PCR analysis of ASS1 mRNA expression over time in HCC cells with the indicated treatments. J) Western blot analysis of ASS1 expression in HCC cells with the indicated treatments. Data are presented as mean ± SD and analyzed by Student's t‐test or one‐way ANOVA with Tukey's multiple comparison test. * p < 0.05; ** p < 0.01; ns: not significant.

Figure 6

circRAPGEF1 downregulates ASS1 to enhance stemness properties in HCC. A–C) Boxplots illustrating the ASS1 expression (A,B) and K‐M curve analysis of OS (C) in TCGA‐LIHC cohort. D) Chord diagram illustrating the correlation between ASS1 and stemness markers. E–G) Representative images and quantification of tumor sphere formation assays in HCC cells with ASS1 silencing or overexpression. Scale bar: 100 µm. H) Western blot analysis of CD133 and ASS1 expression of HCC cells with the indicated treatments. I,J) Representative images and quantification of tumor sphere formation assays in HCC cells co‐transfected with vector/circRAPGEF1 overexpression lentivirus and empty vector (EV)/ASS1 overexpression plasmid. Scale bar: 100 µm. K) Western blot analysis of CD133 and ASS1 expression in rescue assay. L) Image of HCC tumors formed after subcutaneous implantation of different numbers of HCC cells with the indicated treatments in BALB/c nude mice (n = 5). M) Extreme limited dilution analysis for LCSC proportions. N–P) Representative images and quantifications of IHC staining of H&E, CD133, ASS1, and Ki‐67, and weights of subcutaneous HCC tumors in the 8 × 10⁵ HuH‐7 cell implantation group. Data are presented as mean ± SD and analyzed by Student's t‐test or one‐way ANOVA with Tukey's multiple comparison test. * p < 0.05; ** p < 0.01; *** p < 0.001; ns: not significant.

Figure 7

circRAPGEF1 upregulates aspartate levels to enhance stemness properties in HCC cells. A) Schematic diagram of aspartate metabolism in cells. B) GSEA of amino acid metabolism‐related genes in HepG2 cells with vector and circRAPGEF1 overexpression. C) Intracellular aspartate levels of HCC cells co‐transfected with vector/circRAPGEF1 overexpression lentivirus and EV/ASS1 overexpression plasmid. D) Western blot analysis of p‐S6K, S6K, p‐CAD, and CAD expression of HCC cells with the indicated treatments. E) qRT‐PCR analysis of Citrin expression in HCC cells co‐transfected with vector/circRAPGEF1 overexpression lentivirus and negative control/Citrin‐targeting siRNAs. F) Intracellular aspartate levels of HCC cells with the indicated treatments. G,H) Representative images and quantification of tumor sphere formation assays in HCC cells with the indicated treatments. Scale bar: 100 µm. I) Western blot analysis of p‐S6K, S6K, p‐CAD, CAD, CD133, and ASS1 expression in HCC cells with the indicated treatments. Data are presented as mean ± SD and analyzed by one‐way ANOVA with Tukey's multiple comparison test. * p < 0.05; ** p < 0.01; *** p < 0.001; **** p < 0.0001; ns: not significant.

Figure 8

NPs‐mediated circRAPGEF1 silencing enhances sorafenib efficacy in HCC. A) CCK‐8 assay determined the inhibitory efficacy of sorafenib in HCC cells co‐transfected with vector/circRAPGEF1 overexpression lentivirus and EV/ASS1 overexpression plasmid. B,C) CCK‐8 assay and colony formation assay determined the inhibitory efficacy of sorafenib in HCC cells transfected with control or circRAPGEF1‐targeting siRNAs. D) Size distribution of NPs‐siCirc. E) qRT‐PCR analysis of circRAPGEF1 expression in HuH‐7 cells with control or NPs‐siCirc treatment. F) Schematic diagram for the in vivo validation of the combined treatment of NPs‐siCirc and sorafenib. G–I) Image, volumes, and weights of subcutaneous tumors of mice treated with PBS, NPs‐siNC, NPs‐siCirc, NPs‐siNC/Sorafenib (SOR), and NPs‐siCirc/SOR (n = 5). J, K) Representative images and quantification of IHC staining for ASS1, p‐CAD, CD133, and Ki‐67 in tumor specimens. Scale bar: 100 µm. Data are presented as mean ± SD and analyzed by one‐way ANOVA with Tukey's multiple comparison test. * p < 0.05; ** p < 0.01; *** p < 0.001; **** p < 0.0001; ns: not significant.