A circRNA promotes esophageal squamous cell carcinoma progression by inhibiting TRIM25-mediated degradation of IGF2BP family members

Abstract

Background: Esophageal squamous cell carcinoma (ESCC) is a highly aggressive malignancy with poor prognosis and limited treatment options. While circular RNAs (circRNAs) are frequently dysregulated in ESCC, their functional roles and molecular mechanisms in tumor progression remain largely unexplored.

Methods: We characterized circRNA using RT-PCR, Sanger sequencing, and fluorescent in situ hybridization. Gain- and loss-of-function studies in vitro and in vivo were performed to assess circRNA function. Molecular interactions were investigated via RNA pull-down assays coupled with mass spectrometry, electrophoretic mobility shift assays, co-immunoprecipitation, and immunoblot analysis. Clinical relevance was evaluated by RT-qPCR and immunohistochemistry in patient specimens.

Results: We identified circLNF, a novel circRNA derived from the long non-coding RNA FIRRE gene, which is significantly upregulated in ESCC. Functional assays demonstrated that circLNF promotes proliferative and migratory capacities in cultured cells and accelerates tumor progression and metastasis in animal models. Mechanistically, circLNF directly interacts with IGF2BP family proteins through their "CAUC" motifs, protecting them from TRIM25-mediated ubiquitination and proteasomal degradation. Importantly, circLNF expression positively correlated with IGF2BP1 protein levels in ESCC patient tissues, underscoring the clinical relevance of the circLNF-IGF2BP axis in ESCC progression.

Conclusions: Our findings reveal an oncogenic role of circLNF in ESCC by inhibiting TRIM25-mediated proteasomal degradation of IGF2BP proteins and highlight circLNF as a potential therapeutic target for ESCC.

Keywords: Esophageal squamous cell carcinoma; IGF2BPs; TRIM25; circRNA.

Fig. 1

Characterization of circLNF in ESCC. A Schematic illustration of circLNF derived from exons 5–10 of FIRRE gene. The sequence crossing the junction site was amplified by divergent primers (DP) and validated by Sanger sequencing. Convergent primers (CP) and quantitative PCR primers (F1/R1) are indicated. B circLNF was validated in Eca109 and KYSE410 cells by RT-PCR. Divergent primers successfully amplified circLNF from cDNA, but not genomic DNA (gDNA). FIRRE linear transcripts and GAPDH mRNA served as amplification controls. C Detection of circLNF (divergent primers) versus linear transcripts (convergent primers) in Eca109 and KYSE410 cells by RT-PCR assays. D Stability analysis of circLNF and FIRRE linear transcripts in Eca109 and KYSE410 cells with or without Actinomycin D treatment. RNA levels were normalized to U6 snRNA. Data are presented as mean ± SEM. E Subcellular localization of circLNF by RT-qPCR analyses following cytoplasmic/nuclear fractionation of Eca109 and KYSE410 cells. β-actin mRNA and U6 snRNA served as cytoplasmic and nuclear RNA markers, respectively. Immunoblot analysis of β-actin (cytoplasmic) and Lamin A/C (nuclear) confirmed fractionation efficiency. Data are shown as mean ± SEM. F Fluorescence in situ hybridization (FISH) detecting circLNF in Eca109 cells. 18 S rRNA and U6 snRNA were used as cytoplasmic and nuclear RNA references, respectively. Nuclei were counterstained with DAPI (blue). G, H Comparative circLNF expression in ESCC tumors and normal tissues (G), and between non-metastatic (NM) and metastatic (M) ESCC tumors (H). *p < 0.05, ****p < 0.0001, unpaired two-tailed Mann Whitney U test. I Kaplan-Meier survival analysis of ESCC patients with high versus low circLNF expression. The p-value was determined by the Log-rank test

Fig. 2

Ectopic circLNF expression promotes ESCC malignant progression. (A-C) Effects of circLNF overexpression on cell proliferation (A), migration and invasion (B), and motility (C) of KYSE30 and KYSE150 cells. (D-F) circLNF overexpression promotes tumor growth in KYSE30 subcutaneous xenografts (n = 6 mice/group). (G, H) Representative bioluminescence imaging of lung metastasis (G) and H&E staining of lung metastatic lesions (H) of mice (n = 5 mice/group) after tail-vein injection of KYSE30 and KYSE150 cells. Data are indicated as mean ± SEM. p-values were determined with the two-way ANOVA (D) and unpaired two-tailed Mann Whitney U test (F-H), *p < 0.05, **p < 0.01

Fig. 3

CircLNF knockdown inhibits tumor growth and metastasis. A-C circLNF knockdown inhibits cell proliferation (A), migration and invasion (B), and motility (C) of KYSE410 and Eca109 cells. (D-F) Effects of circLNF knockdown in KYSE410 cells on tumor growth. Tumor growth curves (D), endpoint tumor images (E), and terminal tumor weights (F) of nude mice (n = 6 mice/group). (G, H) circLNF knockdown in KYSE410 and Eca109 cells decreased lung metastatic burden using tail-vein injection model. Representative bioluminescence imaging of lung metastasis (G) and H&E staining of lung metastatic lesions (H) of mice (n = 5 mice/group). Data are displayed as mean ± SEM. p-values were identified by two-way ANOVA (D) and unpaired two-tailed Mann Whitney U test (F-H). *p < 0.05, **p < 0.01, and ****p < 0.0001

Fig. 4

CircLNF interacts with IGF2BP proteins via the “CAUC” motif. A, B RNA pull-down assays with Coomassie blue staining of circLNF-interacting proteins after incubating biotin-labeled sense (S) or antisense (AS) DNA probes crossing the junction site of circLNF with whole-cell lysates (A) or cytoplasmic fractions (B) of Eca109 and KYSE410 cells. C, D Immunoblot validation of IGF2BP proteins pulled down from KYSE410 cell lysates by AS DNA probe (C) or in vitro transcribed and circularized circLNF (D), with GAPDH mRNA and SF3B1 protein as negative controls. E RIP and semi-quantitative RT-PCR assays demonstrate endogenous circLNF association with IGF2BPs in KYSE410 cells using control IgG or anti-IGF2BP antibodies. F Schematic illustration showing the locations and sequences of five potential IGF2BPs-binding regions within circLNF containing either “CAUC” or “UGGAC” motifs, with wild-type (WT), mutant (mut), and scramble control (NC) probes indicated. G, H Endogenous IGF2BP proteins were pulled down from KYSE410 cell extracts by WT p1 and p2 probes (containing “CAUC” motifs), rather than other mutant probes. I, J RNA pull-down (I) and RNA-EMSA assays (J) using purified IGF2BP proteins and RNA probes showed direct interaction between IGF2BPs and p1/p2 WT probes

Fig. 5

CircLNF/IGF2BP interaction contributes to ESCC progression. A Association of circLNF with IGF2BPs by RNA pull-down assays using in vitro transcribed and circularized WT circLNF, “CAUC”-mutant circLNF (mut), or antisense oligonucleotides (AS) of circLNF. B Agarose gel electrophoresis of semi-quantitative RT-PCR analyses to detect WT or mutant circLNF expression in stable KYSE150 and KYSE30 cells using specific primers complementary to wild-type or mutant sequence. M, DNA markers (C, D) IGF2BP binding to WT circLNF in KYSE30 stable cells confirmed by RNA pull-down assays using AS DNA probe (C) and RIP assays using anti-IGF2BP antibodies (D). E-G Clone formation (E), cell migration and invasion (F) and wound healing assays (G) of KYSE150 and KYSE30 stable cells. H-J Tumor volumes (H), tumor images (I) and tumor weights (J) of nude mice (n = 6 mice/group) after subcutaneous injection of KYSE150 stable cells. K, L Representative bioluminescence images of lung metastasis (K) and H&E staining of lung metastatic lesions (L) of mice (n = 5 mice/group) after tail-vein injection of KYSE150 stable cells. Data represent mean ± SEM. p-values were determined by two-way ANOVA (H) and unpaired two-tailed Mann Whitney U test (J-L). *p < 0.05, **p < 0.01, ***p < 0.001; n.s., not significant

Fig. 6

CircLNF stabilizes IGF2BP proteins by inhibiting ubiquitin-proteasomal degradation. A, C Immunoblot analysis of IGF2BP expression in different stable cells overexpressing WT circLNF, circLNF-mut, or with circLNF knockdown (KD). B, D Representative immunohistochemistry staining of lung metastatic lesions of mice after tail-vein injection of stable cells with circLNF knockdown (B), WT circLNF or circLNF-mut overexpression (D). Histochemistry scores (H scores) are shown as mean score ± SEM for five mice in each group. **P < 0.01; n.s., not significant; unpaired two-tailed Mann Whitney U test. E, F Effects of circLNF on IGF2BP protein stability. ESCC cells with circLNF overexpression (E) or knockdown (F) were treated with cycloheximide (50 µg/ml) for indicated times and IGF2BP levels were then detected by immunoblot analysis. G Eca109 cells with or without circLNF knockdown were treated with MG132 (20 µM) and IGF2BP protein levels were then measured by immunoblot analysis. H, I Ubiquitination assays showing polyubiquitination status of IGF2BPs immunoprecipitated from different stable cells

Fig. 7

CircLNF inhibits TRIM25-mediated degradation of IGF2BP proteins. A, B Effects of TRIM25 overexpression (A) or knockdown (B) on IGF2BP expression in KYSE410 cells. C Co-IP assays using anti-IGF2BP antibodies confirmed TRIM25-IGF2BP interaction in KYSE410 cells. D, E Co-IP assays using anti-TRIM25 (D) or anti-IGF2BP antibodies (E) indicated that WT circLNF, but not the IGF2BP-binding-deficient mutant (mut), reduced the association of TRIM25-IGF2BP complex in KYSE30 cells. F Positive correlation between circLNF expression (RT-qPCR) and IGF2BP1 protein levels (immunohistochemistry) in ESCC microarray specimens