ENO2 Promotes Colorectal Cancer Metastasis by Interacting with the LncRNA CYTOR and Activating YAP1-Induced EMT

Abstract

The glycolytic enzyme enolase 2 (ENO2) is dysregulated in many types of cancer. However, the roles and detailed molecular mechanism of ENO2 in colorectal cancer (CRC) metastasis remain unclear. Here, we performed a comprehensive analysis of ENO2 expression in 184 local CRC samples and samples from the TCGA and GEO databases and found that ENO2 upregulation in CRC samples was negatively associated with prognosis. By knocking down and overexpressing ENO2, we found that ENO2 promoted CRC cell migration and invasion, which is dependent on its interaction with the long noncoding RNA (lncRNA) CYTOR, but did not depend on glycolysis regulation. Furthermore, CYTOR mediated ENO2 binding to large tumor suppressor 1 (LATS1) and competitively inhibited the phosphorylation of Yes-associated protein 1 (YAP1), which ultimately triggered epithelial-mesenchymal transition (EMT). Collectively, these findings highlight the molecular mechanism of the ENO2-CYTOR interaction, and ENO2 could be considered a potential therapeutic target for CRC.

Keywords: CYTOR; ENO2; YAP1; colorectal cancer; metastasis.

Figure 1

ENO2 was upregulated in CRC tissues and negatively correlated with poor prognosis. (A): Representative immunohistochemical staining of the ENO2 protein in CRC samples and in the corresponding adjacent normal tissue. (B): Overall survival analysis in 184 CRC patients with low (n  =  66), middle (n = 65) or high (n  =  53) levels of ENO2 by Kaplan–Meier analysis. (C): GEO analysis (GSE49355 and GSE14297) (http://www.ncbi.nlm.nih.gov/geo/ (accessed on 6 April 2020). * p  <  0.05, *** p  <  0.001.

Figure 2

ENO2 promoted CRC migration and invasion in vitro and metastasis in vivo. (A): Expression of ENO2 RNA (up) and protein (down) in CRC cell lines normalized to β-actin. (B): Identification of ENO2 knockdown and its effects on RKO and SW480 migration and invasion. Shown from left to right are mRNA as measured by qPCR, protein levels as measured by Western blot, representative transwell images and statistical results from the transwell migration and invasion assays. (C): The migration and invasion of DLD1 cells overexpressing ENO2. (D): Representative transwell results of the phenotypic rescue of ENO2KO cells. ENO2 was reintroduced into ENO2-silenced RKO cells. Characterization of ENO2 re-expression and the corresponding statistical analysis of migration and invasion. (E): Metastatic foci in a mouse model of liver metastasis. The red rectangles indicate the mouse liver region, n = 2. Values are presented as the means  ±  SDs, * p  <  0.05, ** p  <  0.01, *** p  <  0.001, **** p < 0.0001.

Figure 3

ENO2 promoted metastasis independently of altering the glycolytic rate in CRC. (A): qPCR and Western blot analysis confirmed ENO2 knockdown in RKO cells. (B–D): ENO2 knockdown did not significantly inhibit cell proliferation, glucose consumption or lactate production in RKO cells. (E): The ECAR of the indicated cells was detected by using a Seahorse XFe96 Extracellular Flux Analyzer, and the maximum glycolytic rates are summarized. PKM2 was knocked down as a positive control. n  =  3 per group. (F): The total enolase activity of RKO ENO2 knockdown cells was detected by performing a relative fluorescence unit (RFU) assay. Enoblock was used as a positive control. (G–J): Seahorse and total enolase activity of SW480 ENO2 knockdown cells and DLD1 ENO2 overexpressing cells. (K): Characterization of ENO2 with mutations in the substrate binding region. (L): Mutations in the substrate binding region of ENO2 did not suppress its metastatic effect. Data are shown as the mean ± SD. * p < 0.05, ** p < 0.01, *** p < 0.001. ns: non significance.

Figure 4

ENO2 induced EMT through a lncRNA CYTOR-dependent pathway. (A–C): ENO2 induced the expression of mesenchymal genes. Western blotting was used to analyze the expression of EMT-related genes in stable cell lines. (D): RNA pull-down assay results showing that ENO2 was a CYTOR-binding protein. (E,F): The RIP experiment detected the binding of ENO2 to CYTOR and several EMT-related lncRNAs. (G): The OCLN level in RKO cells with ENO2 knockdown and CYTOR KO. (H): The migration and invasion of RKO cells with ENO2 knockdown and CYTOR KO. (I): Identification of full-length ENO2 and different truncated ENO2 sequences reintroduced into RKO ENO2 KO cells. (J): The functions of the ENO2 truncation mutants depicted in (I). (K): The function of single point mutations to amino acid residues 374–380 of ENO2. (L): RIP assay detected the binding of ENO2 single point mutation to CYTOR. Data are shown as the mean ± SD. * p < 0.05, ** p < 0.01, *** p < 0.001. ns: non significance.

Figure 5

ENO2 induced YAP1 activation by suppressing its phosphorylation. (A): ENO2 stimulates the Hippo pathway in SW480 cells. (B–D): The phosphorylation of YAP1 Ser127 and the degradation of YAP1 were detected by Western blot after ENO2 knockdown. (E,F): IF assay showing ENO2 and YAP1 expression. Scale bar = 10 μm. YAP1 is stained red, ENO2 is stained green, and the nucleus is stained blue (DAPI). (G,H): Identification of YAP1 and ENO2 knockdown in RKO cells (G). Migration and invasion of RKO cells with YAP1 and ENO2 knockdown (H). (I,J): Identification of YAP1 and ENO2 knockdown in SW480 cells (I). Migration and invasion of SW480 cells with YAP1 and ENO2 knockdown (J). (K): The migration of DLD1 cells overexpressing ENO2 with YAP1 knockdown. Data are shown as the mean ± SD. * p < 0.05, ** p < 0.01, *** p < 0.001. ns: non significance.

Figure 6

CYTOR mediated ENO2 binding to LATS1 and inhibited the degradation of YAP1. (A): Co-IP assay of anti-LATS1 in RKO cells. (B): RIP assay of CYTOR with anti-ENO2, anti-LATS1 and anti-YAP1 in RKO cells. (C): Co-IP assay of anti-LATS1 in RKO mock/RKO CYTOR KO cells. (D,E): Co-IP assay of anti-ENO2 (left) and anti-LATS1 (right) in SW480 cells. (F): KO of ENO2 significantly enhanced the binding between LATS1 and YAP1. (G): Immunohistochemical staining revealed the differences in ENO2 and YAP1 expression in implanted orthotopic splenic tumors from the two groups (magnification, 200×). (H): Schematic representation of the ENO2 signal transduction pathway in CRC. Data are shown as the mean ± SD. * p < 0.05, ** p < 0.01.