A Positive Feed-Forward Loop between LncRNA-CYTOR and Wnt/β-Catenin Signaling Promotes Metastasis of Colon Cancer

Abstract

We previously demonstrated that long non-coding RNA cytoskeleton regulator RNA (CYTOR), also known as Linc00152, was significantly overexpressed in colon cancer and conferred resistance to oxaliplatin-induced apoptosis. At the same time, elevated CYTOR expression was also reported in gastric cancer and exerted influences on epithelial-mesenchymal transition (EMT) markers. However, the precise mechanism by which CYTOR promotes the EMT phenotype and cancer metastasis remains poorly understood. Here, we showed that loss of epithelial characteristics and simultaneous gain of mesenchymal features correlated with CYTOR expression. Knockdown of CYTOR attenuated colon cancer cell migration and invasion. Conversely, ectopic expression of CYTOR induced an EMT program and enhanced metastatic properties of colon cancer cells. Mechanistically, the binding of CYTOR to cytoplasmic β-catenin impeded casein kinase 1 (CK1)-induced β-catenin phosphorylation that enabled it to accumulate and translocate to the nucleus. Reciprocally, β-catenin/TCF complex enhanced the transcription activity of CYTOR in nucleus, thus forming a positive feed-forward circuit. Moreover, elevated CYTOR, alone or combined with overexpression of nuclear β-catenin, was predictive of poor prognosis. Our findings suggest that CYTOR promotes colon cancer EMT and metastasis by interacting with β-catenin, and the positive feed-forward circuit of CYTOR-β-catenin might be a useful therapeutic target in antimetastatic strategy.

Keywords: CYTOR; colon cancer; feed-forward loop; metastasis; β-catenin.

Figure 1

Loss of Epithelial Features and Simultaneous Gain of Mesenchymal Characteristics Correlate with CYTOR Expression in Colon Cancer Cell Lines (A) Western blot analysis of EMT markers in different cell lines. (B) CYTOR and EMT marker levels were evaluated by qRT-PCR in three colon cancer cells; colon mucosal epithelial cell line NCM460 was used as control. (C and D) The shRNA-mediated CYTOR repression was confirmed by qRT-PCR after lentivirus infection in both SW620 (C) and HCT8 (D) cells, and the mRNA levels of EMT markers E-cadherin, N-cadherin, and Vimentin were also assessed by qRT-PCR. (E) Overexpression of CYTOR was confirmed by qRT-PCR after lentivirus infection in Hct116 cells; related EMT markers mRNA levels were also assessed. (F and G) CYTOR knockdown or overexpression affected the protein levels of EMT markers in both western blot analysis (F) and confocal immunofluorescent assay (G). Error bars, ±SD. *p < 0.05.

Figure 2

CYTOR Promotes Colon Cancer Cell Metastasis In Vitro and In Vivo (A and B) Stable HCT8 (A) and Hct116 (B) cells were subjected to wound healing assay. The wound space was photographed at 0 and 48 hr. (C) The invasive ability of HCT8 cells was suppressed in response to knockdown of CYTOR. (D) Ectopic expression of CYTOR significantly enhanced Hct116 cells invasion into Matrigel-coated transwell membranes. (E and F) Representative mice injected with modified CYTOR expressing HCT8 (E) or Hct116 cells (F) (n = 5 per group). The luciferase signal intensity from days 7 to 35 is on equivalent scales in lung metastasis model. (G and H) Representative results of H&E staining of the metastatic nodules in the lung. The stained metastatic nodules are indicated by arrows, and the number of metastatic nodules for each mouse were counted under the microscope. Error bars, ±SD. **p < 0.01, ***p < 0.001.

Figure 3

CYTOR Activates the Wnt/β-Catenin Signaling by Blocking CK1-Mediated β-Catenin Phosphorylation (A) Signaling pathway reporter array was used for seeking the relative pathway associated with CYTOR on Hct116 cells. (B) Knockdown of CYTOR caused a decreased expression of total β-catenin, nuclear β-catenin, c-myc, and cyclin D1, whereas overexpression of CYTOR significantly increased their expression. (C) The role of CYTOR in promoting the redistribution of cytoplasmic β-catenin to nuclear localization was confirmed through immunofluorescence assays. Depletion of CYTOR resulted in a redistribution of nuclear β-catenin to the cytoplasmic localization, whereas overexpression of CYOTR caused substantial nuclear accumulation of β-catenin. (D) TOPflash and FOPflash luciferase reporter analyses revealed that the transactivation of TCF reporter was inhibited by the depletion of CYTOR and enhanced via overexpression of CYTOR. (E and F) RIP assay (E) and RNA pull-down assay (F) revealed that non-phospho-β-catenin^Ser45, rather than other active non-phospho sites, was required for β-catenin interaction with CYTOR. (G) Coimmunoprecipitation of CK1 and β-catenin in lysates of colon cancer cells with modified CYTOR expression. (H) The role of CYTOR in protecting β-catenin from phosphorylation by CK1 was confirmed via an in vitro phosphorylation assay. (I) HCT8 cells were treated with 10 nmol/L PF670462 for 24 hr, inhibiting CK1 activity by PF670462 abolished the effects of CYTOR-shRNA on β-catenin^Ser45 phosphorylation and Wnt/β-catenin signaling activity. Error bars, ±SD. ***p < 0.001.

Figure 4

β-Catenin/TCF4 Complex Promotes CYTOR Transcription in Colon Cancer Cells (A) CYTOR expression was repressed in HCT8 cells treated with 15 μmol/L XAV939 for 24 hr. (B) CYTOR expression was markedly increased in Hct116 cells treated with 8 μmol/L CHIR99021 for 24 hr. (C and D) Treatment with 1.6 μmol/L LF3 for 24 hr significantly decreased CYTOR transcription in both Hct8 (C) and HCT116 (D) cells. (E) The enrichment of β-catenin on two TBEs in the CYTOR promoter was confirmed. (F and G) The enrichment was significantly downregulated in HCT8 cells when Wnt signaling was blocked (F); an inverse result was observed in Hct116 cells after activating the Wnt signaling (G). (H and I) The co-occupancy of β-catenin and TCF4 on CYTOR promoter was identified via ChIP-re-ChIP assay. Error bars, ±SD. **p < 0.01, ***p < 0.001.

Figure 5

Wnt/β-Catenin Signaling Is Responsible for CYTOR-Mediated EMT and Metastasis (A) Stable HCT8 cells were treated with 8 μmol/L CHIR99021 for 24 hr; indicated protein levels were assayed by western blotting. (B) Stable Hct116 cells were treated with 15 μmol/L XAV939 for 24 hr; indicated protein levels were assayed by western blotting. (C and D) The inhibition of TCF/LEF activity caused by CYTOR knockdown was significantly rescued by CHIR99021 in HCT8 cells (C); on the other hand, ectopically expressed CYTOR enhanced the TCF/LEF activity, which was suppressed by XAV939 in Hct116 cells (D). (E and F) The transwell assay showed different cell invasive capacities in stable HCT8 (E) and Hct116 (F) cells treated with CHIR99021 (8 μmol/L) or XAV939 (15 μmol/L) for 24 hr. (G) β-catenin, c-myc, and cyclin D1 were reduced with decreased lung metastasis burden and upregulated with more lung colonization caused by modified CYTOR expression according to the immunohistochemistry. Error bars, ±SD. ***p < 0.001.

Figure 6

The Prognostic Value of Combining CYTOR and β-Catenin for Colon Cancer Patients (A) Elevated expression of CYTOR was observed by using qRT-PCR in colon cancer tissues when compared with adjacent normal mucosa. (B) The levels of CYTOR were positively correlated with the levels of Wnt/β-catenin target genes c-myc and cyclin D1. (C) Immunochemistry staining was performed to measure the β-catenin expression in colon cancer tissues. (D) Relative percentages of nuclear β-catenin in high (72.0%) and low (34.0%) CYTOR groups. (E) Kaplan-Meier analysis of OS (log-rank test, p < 0.001) or DFS (log-rank test, p < 0.001) with cytoplasmic or nuclear β-catenin in colon cancer patients. (F) Kaplan-Meier analysis of OS (log-rank test, β-catenin^negative versus β-catenin^nuc, p = 0.001; β-catenin^cyto versus β-catenin^nuc, p = 0.001) or DFS (log-rank test, β-catenin^negative versus β-catenin^nuc, p = 0.002; β-catenin^cyto versus β-catenin^nuc, p = 0.001) with negative or cytoplasmic or nuclear β-catenin in colon cancer patients. (G) Kaplan-Meier analysis of OS (log-rank test, CYTOR^lowβ-catenin^cyto versus CYTOR^lowβ-catenin^nuc, p = 0.041; CYTOR^highβ-catenin^cyto versus CYTOR^highβ-catenin^nuc, p = 0.046) or DFS (log-rank test, CYTOR^lowβ-catenin^cyto versus CYTOR^lowβ-catenin^nuc, p = 0.033; CYTOR^highβ-catenin^cyto versus CYTOR^highβ-catenin^nuc, p = 0.066) with high or low CYTOR and cytoplasmic or nuclear β-catenin in colon cancer patients.

Figure 7

A Schematic Model of CYTOR-β-Catenin Signaling Circuit in Colon Cancer Cells