A positive feedback circuit driven by m6A-modified circular RNA facilitates colorectal cancer liver metastasis

Abstract

Background: Liver metastasis is the leading cause of death in patients with colorectal cancer (CRC). Emerge evidence suggests that circular RNA (circRNA) is a pivotal player in cancer progression. However, its role in CRC liver metastasis remains largely unknown.

Methods: Circ-YAP expression was detected by qRT-PCR and in situ hybridization. The function of circ-YAP was tested by wound healing, transwell and CCK-8 assays. RNA immunoprecipitation, pull-down, luciferase reporter, chromatin immunoprecipitation assays were used to investigate the mechanism underlying circ-YAP promoting CRC liver metastasis. CRC liver metastasis animal model was established to assess the effect of circ-YAP in vivo.

Results: Circ-YAP was notably upregulated in CRC with liver metastasis, which was associated with dismal prognosis. Circ-YAP promoted CRC cell migration and invasion in vitro, and facilitated liver metastasis in patient-derived xenografts (PDX) models in vivo. Mechanistically, circ-YAP encoded a novel truncated protein containing 220 amino acids, termed as YAP-220aa, which competitively bound to LATS1, resulting in YAP dephosphorylation and nuclear translocation, thereby activating a cohort of metastasis-promoting genes. Importantly, N⁶-methyladenosine (m⁶A) modification orchestrated efficient initiation of circ-YAP translation, requiring m⁶A reader YTHDF3 and eIF4G2 translation initiation complex. Intriguingly, circ-YAP was transcriptionally enhanced by YAP/TEAD complex, thus forming a positive regulatory feed-forward loop.

Conclusions: Our findings reveal a previously uncharacterized oncoprotein encoded by circ-YAP, implying a promising biomarker and therapeutic target for CRC patients with liver metastasis.

Keywords: Biomarker; Circular RNA; Hippo signaling; N6-methyladenosine; Translation.

Fig. 1

Identification of circ-YAP in CRC liver metastasis tissues. A, B. qRT-PCR analysis of circ-YAP expression in CRC tissues and cell lines. C. Sanger sequencing verifying the junction site of circ-YAP. D, E. Cells were treated with 3U/μg of RNase R or 5 μg/ml Actinomycin D, followed by qRT-PCR analysis of circ-YAP and YAP mRNA levels. F. FISH assay detecting the location of circ-YAP, DAPI was used to stain cell nucleus. Scale bar, 25 μm. G-I. ISH staining detecting circ-YAP expression in paraffin embedded tissues (G, H), followed by ROC curve analysis of the predictive accuracy (I). The dark purple denotes positive staining of circ-YAP. Scale bar, 50 μm. J. The survival curve of CRC patients with low and high circ-YAP levels. ***P < 0.001. Data (B, D, E) are the mean ± SD of three independent experiments carried out in triplicate

Fig. 2

Silencing of circ-YAP alleviates CRC liver metastasis burden. A. qRT-PCR verifying the knockdown efficiency of the designed sgRNAs. B, C. Wound healing assay testing cell migration in circ-YAP-silenced LoVo cells. D, E. Transwell assay testing cell invasion in circ-YAP-silenced CRC cells. Scale bar, 100 μm. F. qRT-PCR verifying the overexpression efficiency of circ-YAP. G, H. Cell migration and invasion were tested in DLD1 and SW480 cells after circ-YAP overexpression. I. The sketch showing the establishment of the spontaneous liver metastasis model. J. The representative images of CRC liver metastasis in the indicated groups. Scale bar, 100 μm. K. The incidence of CRC liver metastasis in each group. ***P < 0.001. Data (A, C, E, F, G, H) are the mean ± SD of three independent experiments carried out in triplicate

Fig. 3

Circ-YAP encodes YAP-220aa mediated by m⁶A. A, B. The sketch and full-length sequence of circ-YAP. C. The sketch showing the construction of the indicated vectors. D. Western blot analysis of Flag and YAP-220aa protein levels in 293T cells transfected with the above vectors. E. Coomassie blue staining of protein samples from control and circ-YAP-overexpressing cells, followed by mass spectrometry of the indicated gels. F. Western blot analysis of YAP-220aa protein levels in circ-YAP-overexpressing cells. G. IF staining of Flag and YAP-220aa in 293T cells. Scale bar, 25 μm. H. Western blot analysis of YAP-220aa protein levels in CRC cell lines. I. Western blot testing the effect of METTL3 knockdown on YAP-220aa expression. J. meRIP assay testing the m⁶A levels on circ-YAP in CRC cells. K, L. Cell migration and invasion in HT29 and SW480 cells transfected with the indicated vectors. M. The experimental liver metastasis model testing the effect of circ-YAP or circ-YAP-m⁶A-mutation on CRC liver metastasis (n = 5 per group). ***P < 0.001. Data (J, K, L) are the mean ± SD of three independent experiments carried out in triplicate. The uncropped western blot data are provided as a Original Blot Image file

Fig. 4

YTHDF3 recruits eIF4G2 complex to driver circ-YAP translation. A. Western blot testing the levels of the indicated proteins in circ-YAP-overexpressing 293T cells after knockdown of YTHDF1/2/3. B, C. Western blot testing the effect of YTHDF3 knockout on YAP-220 level. D. The indicated vectors were transfected into LoVo and SW620 cells, followed by western blot analysis of YAP-220 level. E. RIP assay using anti-YTHDF3 antibody, followed by qRT-PCR analysis of circ-YAP enrichment. F, G. The in vivo and in vitro pull-down assays using biotin-labeled circ-YAP probes, followed by western blot analysis of YTHDF3 level. H. Co-IP assay testing the interaction between YTHDF3 and eIF4G2 proteins. I. Western blot testing the effect of eIF4G2 silencing on YAP-220aa expression. J, K. RNA pull-down assays using biotin-labeled circ-YAP probes in YTHDF3^−/− SW620 and LoVo cells, followed by western blot analysis of the indicated protein levels. L, M. Transwell assay testing cell invasion in circ-YAP-overexpressing SW480 and HT29 cells after YTHDF3 knockout or eIF4G2 silencing. *P < 0.05, ***P < 0.001. Data (E, L, M) are the mean ± SD of three independent experiments carried out in triplicate. The uncropped western blot data are provided as a Original Blot Image file

Fig. 5

YAP-220aa increases YAP activity. A-C. Co-IP assay testing the binding of YAP-220aa to LATS1. D, E. Co-IP assay testing the interaction between YAP and LATS1/14-3-3 after circ-YAP knockdown. F, G. Western blot analysis of the indicated protein levels in circ-YAP-silenced LoVo and SW620 cells transfected with circ-YAP or circ-YAP-m⁶A-mutation. H, I. IF staining and western blot testing the location of YAP in circ-YAP-overexpressing cells. Scale bar, 25 μm. J. Luciferase reporter assay testing the effect of circ-YAP knockdown on YAP transcription activity in CRC cells. K. qRT-PCR analysis of the pro-metastasis gene expression downstream of YAP after circ-YAP knockdown. L, M. Cell migration and invasion in HT29 and DLD1 cells after YAP knockdown or verteporfin treatment. N. The representative images and burden of spontaneous CRC liver metastases (n = 5 per group). Scale bar, 100 μm. ***P < 0.001. Data (J, L, M) are the mean ± SD of three independent experiments carried out in triplicate. The uncropped western blot data are provided as a Original Blot Image file

Fig. 6

Circ-YAP is transcriptionally activated by YAP. A, B. qRT-PCR analysis of total and nascent circ-YAP levels in YAP-silenced or verteporfin-treated cells. C. Luciferase reporter assay testing the effect of YAP activation on circ-YAP promoter activity. D. Two putative YAP/TEAD binding motifs on circ-YAP promoter. E, F. The promoter activity of circ-YAP was tested in HT29 and SW480 cells co-transfected with p-YAP-5SA and wild-type or mutant luciferase reporter vectors. G, H. qRT-PCR analysis of circ-YAP expression in YAP-activated cells transfected with the indicated siRNAs. I-K. ChIP assay using the indicated antibodies, followed by qPCR analysis of the enrichment of YAP, PoI II and p-PoI II on circ-YAP promoter. L, M. ChIP-re-ChIP assay using anti-YAP/TEAD1 antibody, followed by qPCR analysis of the enrichment of YAP and TEAD1 on circ-YAP promoter. N. DNA pull-down assay using the wild-type or mutant biotinylated circ-YAP promoter probe, followed by western blot analysis of YAP and TEAD1 protein levels in CRC cells. *P < 0.05, **P < 0.01, ***P < 0.001. Data are the mean ± SD of three independent experiments carried out in triplicate. The uncropped western blot data are provided as a Original Blot Image file

Fig. 7

Verification of the circ-YAP/YAP-220aa/YAP regulatory axis in clinical samples. A-D. The correlations between circ-YAP expression and CYR61, CTGF, TWIST1 or FOXM1 in 100 CRC tissues. E-G. The representative ISH and IHC images of circ-YAP, YAP-220aa and YAP in CRC tissues (E), followed by statistical analysis (F, G). Scale bar, 100 μm. H. The protein expression of YAP-220aa analyzed by IHC staining in CRC tissues with liver metastasis, followed by ROC curve analysis. J. The survival curve of CRC patients with low and high circ-YAP&YAP-220aa expression

Fig. 8

The cartoon showing the promoting effect of circ-YAP on CRC liver metastasis through translating into a novel YAP isoform, YAP-220aa, in an m⁶A-dependent manner; subsequently, YAP-220aa directly binds to LATS1 and blocks the interaction between LATS1 and YAP, resulting in YAP nuclear translocation and transcription activation of pro-metastasis genes and circ-YAP, thus forming a positive regulatory loop