Fig. 3: The association of IGF2BP2 protein with circMAN1A2(2,3,4,5).

a Schematic of in-cell RNA pull-down assay using EGFP and ivcMAN1A2(2,3,4,5), followed by mass spectrum (MS) analysis (n = 2 independent experiments). b Volcano plot of enriched protein candidates via in-cell ivcMAN1A2(2,3,4,5) RNA pull-down assay followed by MS (n = 2 independent experiments). EGFP RNA was a negative control. IG2FBP2 is highlighted in red. c IvcMAN1A2(2,3,4,5) and linear MAN1A2(2,3,4,5) interacted with IGF2BP2, as revealed by in-cell (top panel) and in vitro (bottom panel) RNA pull-down assay followed by Western Blot (WB). d IGF2BP2 interacted with ivcMAN1A2(2,3,4,5) and linear MAN1A2(2,3,4,5), as revealed by EMSA. EGFP protein was a negative control. e IGF2BP2 enriched more endogenous circMAN1A2(2,3,4,5) and linear MAN1A2, as detected by native RIP assay. IgG was performed as a negative control; GAPDH mRNA and U6 RNA were detected as negative controls. f Identified IGF2BP2-binding sequence in circMAN1A2(2,3,4,5). Top, genomics locus and diagram of linear MAN1A2 and circMAN1A2(2,3,4,5) (shown as cylinders in magenta). Bottom, IGF2BP2 binding peaks and wiggle-tracks from eCLIP-seq data in K562 cells (GEO, GSE91445) revealed that IGF2BP2 proteins could binding at the circMAN1A2(2,3,4,5). LNAs pairing to circMAN1A2(2,3,4,5) are shown as violet lines. g LNA-blocked in vitro RNA pull-down assay revealed the specific binding sites of IGF2BP2 on circMAN1A2(2,3,4,5). h KD of IGF2BP2 showed a negligible effect on circMAN1A2(2,3,4,5) expression. Left panel: KD efficiency of IGF2BP2 in DLD-1 cells. Right panel: circMAN1A2(2,3,4,5) expression level. i KD of circMAN1A2(2,3,4,5) barely affected IGF2BP2 expression in DLD-1 cells. Images are representative of three independent experiments (c, d, g, i). RNA levels were detected by RT-qPCR (e, h), normalized to ACTB (h) and presented as mean ± s.d., with biologically individual data points shown (e, h). P values were determined by two-tailed Welch’s t-tests without correction for multiple comparisons (b), ordinary two-way ANOVA with Dunnett’s post-hoc comparisons versus scramble control at each row factor level (adjusted P values) (e) and ordinary one-way ANOVA test with Dunnett’s post-hoc comparisons versus scramble control (FDR-adjusted P values) (h). Source data are provided as a Source Data file.