AR-induced ZEB1-AS1 represents poor prognosis in cholangiocarcinoma and facilitates tumor stemness, proliferation and invasion through mediating miR-133b/HOXB8

Abstract

Zinc finger E-box binding homeobox 1 antisense 1 (ZEB1-AS1) has displayed vital regulatory function in various tumors. However, the biological function of ZEB1-AS1 in cholangiocarcinoma (CCA) remains unclear. In this study, we confirmed that ZEB1-AS1 expression was increased in CCA tissues and cells, respectively. Upregulated ZEB1-AS1 was related to lymph node invasion, advanced TNM stage and poor survival of CCA patients. ZEB1-AS1 exhibited high sensitivity and specificity to be an independent poor prognostic factor of patients with CCA. Functionally, knocking down ZEB1-AS1 attenuated tumor cell stemness, restrained cellular viability in vitro and in vivo, and inhibited CCA cell migration and invasion by reversing epithelial-mesenchymal transition. For the mechanism, androgen receptor (AR) directly promoted ZEB1-AS1 expression, and further ZEB1-AS1 increased oncogene homeobox B8 (HOXB8) by sponging miR-133b. In addition, malignant phenotypes of CCA promoted by ZEB1-AS1 dysregulation were rescued separately through interfering miR-133b and HOXB8, suggesting AR/ZEB1-AS1/miR-133b/HOXB8 exerted crucial functions in tumorigenesis and progression of CCA.

Keywords: HOXB8; ZEB1-AS1; cholangiocarcinoma; miR-133b; prognosis.

Figure 1

The expression of ZEB1-AS1 and its correlation with clinicopathological characteristics and prognosis. (A) ZEB1-AS1 expression in CCA tissues and paired adjacent nontumor bile duct tissues was detected by qRT-PCR. (B) CCA patients were divided into two groups according to average value of ZEB1-AS1 expression. Overall survival was evaluated between high and low ZEB1-AS1 expression groups by using Kaplan-Meier method and log-rank test. (C) The correlation between relative ZEB1-AS1 expression and survival time of CCA patients was assessed by Pearson correlation analysis. (D) The sensitivity and specificity of ZEB1-AS1 as a prognostic marker were analyzed by ROC curve. ^***P < 0.001.

Figure 2

Increased ZEB1-AS1 facilitated the viability and stemness of CCA cells. (A) The expression of ZEB1-AS1 in CCA QBC939, CCLP-1, RBE, TFK-1 cells and normal HIBEC. (B) The knockdown efficiencies of si-ZEB1-AS1-1 and si-ZEB1-AS1-2 as well as amplification efficiency of pcDNA3.1-ZEB1-AS1 were monitored through qRT-PCR. (C) CCK-8 proliferation curves were drawn to show the effect of ZEB1-AS1 on cellular proliferation. (D) The red stains representing proliferative activity were reduced in QBC939 and CCLP-1 cells transfected with si-ZEB1-AS1-1 and si-ZEB1-AS1-2. (E) The cell colonies were decreased in si-ZEB1-AS1 cells revealed by colony formation assays. (F) Spheroid forming abilities of QBC939 and CCLP-1 cells transfected with si-ZEB1-AS1-1 and si-ZEB1-AS1-2 were restrained. ^*P < 0.05, ^**P < 0.01, ^***P < 0.001.

Figure 3

Upregulated ZEB1-AS1 promoted cellular migration and invasion through promoted EMT process. (A) The wound closure of QBC939 and CCLP-1 cells transfected with si-ZEB1-AS1-1 and si-ZEB1-AS1-2 was delayed verified by wound healing assays. (B) The numbers of migrating cells were decreased in si-ZEB1-AS1 cells. (C) Transwell assays displayed that knocking down ZEB1-AS1 inhibited invasion of QBC939 and CCLP-1 cells compared with controls. (D) EMT-related proteins including epithelial marker (E-cadherin) and mesenchymal markers (snail and vimentin) were measured via western blot. ^*P < 0.05, ^**P < 0.01, ^***P < 0.001.

Figure 4

ZEB1-AS1 was induced by transcription factor AR. (A) AR sequence and binding sites (E1, E2 and E3) to ZEB1-AS1 promoter region were predicted by using JASPAR database (http://jaspar.genereg.net/). (B) The pcDNA3.1-AR amplified AR mRNA expression in QBC939 and CCLP-1 cells compared with empty vector. (C) Upregulated AR facilitated ZEB1-AS1 expression in QBC939 and CCLP-1 cells corroborated by qRT-PCR. (D) ChIP assays were performed to confirm the direct binding of AR to ZEB1-AS1 promoter in QBC939 and CCLP-1 cells. (E) Luciferase reporter plasmids were constructed with TFBS E2, including wild type and mutant type. (F) The luciferase activity of TFBS E2 WT was markedly promoted by pcDNA3.1-AR cotransfection compared with controls in QBC939 and CCLP-1 cells. ^***P < 0.001.

Figure 5

ZEB1-AS1 functioned as a sponge for miR-133b in CCA cells. (A) Subcellular localization of ZEB1-AS1 was tested by subcellular fractionation assays. GAPDH and U6 were used as endogenous controls for cytoplasm and nucleus, respectively. (B) The expression levels of predicted miRNAs were detected after knocking down ZEB1-AS1 in QBC939 and CCLP-1 cells. (C) The expression of miR-133b in CCA tissues and paired adjacent nontumor bile duct tissues. (D) The correlation between relative ZEB1-AS1 expression and relative miR-133b expression in CCA tissues. (E) The miR-133b expression in CCA cells (QBC939, CCLP-1, RBE, TFK-1) and normal HIBEC. (F) Luciferase reporter plasmids were constructed with miR-133b-binding site region of ZEB1-AS1 sequence, including wild type and mutant type. (G) Luciferase reporter assays showed that cotransfected miR-133b mimics significantly inhibited luciferase activity of ZEB1-AS1 wild type. (H) AGO2 RIP assays further suggested the binding of miR-133b to ZEB1-AS1. ^*P < 0.05, ^**P < 0.01, ^***P < 0.001.

Figure 6

MiR-133b was a direct regulator of HOXB8 in CCA cells. (A) MiR-133b restrained HOXB8 mRNA expression confirmed by qRT-PCR. (B) MiR-133b refrained HOXB8 protein expression testified via western blot in QBC939 and CCLP-1 cells. (C) The expression of HOXB8 mRNA in CCA tissues and paired adjacent nontumor bile duct tissues. (D) The correlation between relative HOXB8 mRNA expression and relative miR-133b expression in CCA tissues. (E) The HOXB8 mRNA expression in QBC939, CCLP-1, RBE, TFK-1 and normal HIBEC. (F) The HOXB8 protein expression in CCA cells (QBC939, CCLP-1, RBE, TFK-1) and normal HIBEC. (G) Luciferase reporter plasmids were constructed with miR-133b-binding site region of HOXB8 sequence, including wild type and mutant type. (H) The luciferase activity of HOXB8 wild type was inhibited by miR-133b mimics cotransfection. (I) AGO2 RIP assays were conducted to further demonstrate the binding of miR-133b to 3’UTR of HOXB8. ^*P < 0.05, ^**P < 0.01, ^***P < 0.001.

Figure 7

ZEB1-AS1 promoted malignant progression of CCA through mediating miR-133b/HOXB8. (A, B) HOXB8 downexpression caused by ZEB1-AS1 knockdown was saved by silencing miR-133b. (C–E) Rescue assays of EdU, spheroid formation and transwell confirmed that inhibition of proliferation, stemness and invasion induced by knocking down ZEB1-AS1 was saved through silencing miR-133b, respectively. (F–H) CCLP-1 cells cotransfected with pcDNA3.1-ZEB1-AS1 and sh-HOXB8 were used to carry out EdU, spheroid formation and transwell assays, respectively. (I–K) Restoration of HOXB8 expression rescued the promotion of proliferation, stemness and invasion generated through miR-133b knockdown in EdU, spheroid formation and transwell assays, respectively. ^*P < 0.05, ^**P < 0.01, ^***P < 0.001.

Figure 8

ZEB1-AS1/miR-133b/HOXB8 contributed to CCA tumorigenesis in vivo. (A) Xenograft tumors were resected on the 21^st day after injection. (B) Tumor volumes were calculated every 3 days throughout the course of tumor growth. (C) CCLP-1 cells cotransfected with sh-ZEB1-AS1 and antagomir-133b were subcutaneously injected into the posterior flanks of mice. (D) Tumor weights were measured after excision. (E, F) HOXB8 expression in xenograft tumors of the three groups (sh-NC/antagomir-NC, sh-ZEB1-AS1/antagomir-NC, sh-ZEB1-AS1/antagomir-133b). ^*P < 0.05, ^**P < 0.01, ^***P < 0.001.