Hsa_circ_0060467 promotes breast cancer liver metastasis by complexing with eIF4A3 and sponging miR-1205

Abstract

Breast cancer (BC) is the most common cancer and the top cause of female mortality worldwide. The prognosis for patients with breast cancer liver metastasis (BCLM) remains poor. Emerging studies suggest that circular RNAs (circRNAs) are associated with the progression of BC. Exploration of circRNAs presents a promising avenue for identifying metastasis-targeting agents and improving the prognosis of patients with BCLM. Microarray and bioinformatic analyses were used to analyze differentially expressed circRNAs between three pairs of BCLM and primary BC. The roles of hsa_circ_0060467 (circMYBL2) and its target gene E2F1 in BC cells were explored by multiple functional experiments. And xenograft mouse models and hepatic metastases of BC hemi-spleen models were used to illustrate the function of circMYBL2 in vivo. The intrinsic molecular mechanism involving circMYBL2 was confirmed by bioinformatics analyses, RIP assays, CHIRP assays, luciferase reporter assays, and rescue experiments. CircMYBL2 was overexpressed in BCLM tissues and BC cells. Functionally, circMYBL2 can facilitate the proliferation and liver metastasis of BC. Mechanistically, circMYBL2 upregulated the transcription factor E2F1 by sponging miR-1205 and complexing with eukaryotic translation initiation factor 4A3 (eIF4A3) and then facilitated the epithelial-mesenchymal transition (EMT) process in BC cells. Our findings showed that circMYBL2 promoted the tumorigenesis and aggressiveness of BC through the circMYBL2/miR-1205/E2F1 and circMYBL2/eIF4A3/E2F1 axes, which may provide a novel targeted therapy for patients with BCLM.

Fig. 1. BCLM tissues and BC cells showed high circMYBL2 expression.

A Using microarray analysis, the expression profile of circRNAs in BCLM tissues was compared to that of primary BC tissues, and visualized through a volcano plot. B The heatmap illustrates the top 30 upregulated circRNAs in BCLM tissues compared with primary BC tissues. C qRT‒PCR detected the expression of circMYBL2 in four pairs of BCLM and primary BC tissues. The expression of circMYBL2 was normalized to that of GAPDH. D qRT‒PCR detected the expression of circMYBL2 in BC cell lines. E Schematic illustration showing the structure of circMYBL2 by circularizing exons of the MYBL2 gene. The back-splice site of circMYBL2 was confirmed by Sanger sequencing. F PCR assay with divergent and convergent primers showed the amplification of circRNAs from cDNA or gDNA in BC cells in agarose gel electrophoresis. G qRT‒PCR detected the expression of circMYBL2 and MYBL2 mRNA in BC cells after RNase R treatment. H qRT‒PCR detected the expression of circMYBL2 and MYBL2 mRNA in BC cells after actinomycin D treatment at the indicated times. I qRT‒PCR detected the expression of circMYBL2, U6, and GAPDH in the cytoplasm and nucleus of BC cells. CircMYBL2 was normalized to U6 in the nucleus and GAPDH in the cytoplasm. J RNA FISH detected the subcellular localization of circMYBL2 in BC cells. Nuclei were stained with DAPI. All data were expressed as the mean ± SD (three independent experiments). *p < 0.05; **p < 0.01; ***p < 0.001.

Fig. 2. CircMYBL2 promoted BC proliferation and migration.

BC cells were transfected with si-circMYBL2 or si-NC and transfected with LV-circMYBL2 or LV-vector via lentivirus plasmids. A–D For evaluation of the proliferative ability, CCK-8 and colony formation assays were used. E–G Wound healing and transwell assays were used to evaluate migratory capabilities. The scale bar in wound-healing assays indicates 50 μm; the scale bar in transwell assays indicates 100 μm. All data were expressed as the mean ± SD (three independent experiments). *p < 0.05; **p < 0.01; ***p < 0.001.

Fig. 3. CircMYBL2 promoted BC growth and liver metastasis.

MDA-MB-231 cells with stable luciferase expression were infected with LV-circMYBL2 or LV-vector and then injected into BALB/c nude mice. A In vivo optical imaging system was used to observe xenograft tumors in the indicated groups. B, C Representative images of xenograft tumors in the indicated groups. D Xenograft tumor weight in the indicated groups. E Xenograft tumor volumes were measured and calculated on the indicated days. Tumor volume was measured as (width²/2 × length). F The number of metastatic nodules in the livers of mice was calculated. G Hepatic metastases of BC hemi-spleen models in the indicated groups were collected and monitored by an in vivo optical imaging system. H The livers in the indicated groups were stained with H&E. All data were expressed as the mean ± SD. *p < 0.05; **p < 0.01; ***p < 0.001.

Fig. 4. CircMYBL2 served as a miR-1205 sponge to regulate E2F1 expression.

A Venn diagram of the miRNAs of circMYBL2 based on circBank and CircInteractome. B qRT‒PCR detected the expression of miR-1205 in the BC cells transfected with si-circMYBL2 or LV-circMYBL2. C Luciferase reporter assays detected the luciferase activities in the BC cells cotransfected with circMYBL2-WT/miR-1205 mimics or circMYBL2-MUT/miR-1205 mimics. D The binding sequence between miR-1205 and E2F1 via TargetScan databases. E, F qRT‒PCR and western blot detected the expression of E2F1 in the BC cells transfected with miR-1205 mimic or inhibitor. G, H qRT‒PCR and western blot detected the expression of E2F1 in BC cells transfected with si-circMYBL2 or LV-circMYBL2. I Western blot analysis of the expression of E2F1 in the BC cells cotransfected with LV-circMYBL2/miR-1205 mimic or si-MYBL2/miR-1205 inhibitor. All data were expressed as the mean ± SD (three independent experiments). *p < 0.05; **p < 0.01; ***p < 0.001.

Fig. 5. CircMYBL2 recruited eIF4A3 to stabilize E2F1 mRNA.

A Potential binding site between circMYBL2 and eIF4A3 in CircInteractome. B Western blot was used to detect the expression of eIF4A3 pulled down by the circMYBL2 probe or NC via a CHIRP assay. C, D qRT‒PCR and nucleic acid electrophoresis were used to detect the expression of circMYBL2 pulled down by anti-eIF4A3 or anti-IgG via RIP assay. E, F Western blot and qRT‒PCR showed the expression of eIF4A3 in the BC cells transfected with si-circMYBL2 or LV-circMYBL2. G Pearson’s correlation analysis identified correlations between eIF4A3 and E2F1 in TCGA databases. H, I qRT‒PCR and western blot showed the expression of E2F1 in the BC cells transfected with si-eIF4A3 or si-NC. J, K qRT‒PCR and nucleic acid electrophoresis were used to detect the expression of E2F1 mRNA pulled down by anti-eIF4A3 or anti-IgG via RIP assays. L qRT‒PCR was used to detect the stability of E2F1 mRNA in BC cells after administration of actinomycin D. All data were expressed as the mean ± SD (three independent experiments). *p < 0.05; **p < 0.01; ***p < 0.001.

Fig. 6. Knocking down E2F1 rescued the promotive effect induced by circMYBL2 overexpression in BC.

BC cells were transfected with LV-vector/si-NC, LV-vector/si-E2F1, LV-circMYBL2/si-NC, and LV-circMYBL2/si-E2F1. A, B For evaluation of the proliferative ability, CCK-8 and colony formation assays were used. C, D Wound healing and transwell assays were used to evaluate the migratory capabilities of BC cells. The scale bar in wound-healing assays indicates 50 μm; the scale bar in transwell assays indicates 100 μm. E Western blot showed the expression of Vimentin, E-Cadherin, and N-Cadherin in the BC cells transfected with si-circMYBL2 or LV-circMYBL2. F Western blot showed the expression of Vimentin, E-Cadherin, and N-Cadherin in the BC cells transfected with LV-vector/LV-circMYBL2 and si-NC/si-E2F1. All data were expressed as the mean ± SD (three independent experiments). *p < 0.05; **p < 0.01; ***p < 0.001.

Fig. 7. A schematic illustration of the molecular mechanism of circMYBL2 in promoting the development of BC.

CircMYBL2 regulates E2F1 expression in BC cells by complexing with eIF4A3 and sponging miR-1205, and eventually activates the process of EMT.