MicroRNA485-3p negatively regulates the transcriptional co-repressor CtBP1 to control the oncogenic process in osteosarcoma cells

Abstract

Carboxyl-terminal binding protein 1 (CtBP1), a well-known transcriptional co-repressor, is highly expressed in a number of cancer types. However, it is still absent in osteosarcoma cells. Here, we found that CtBP1, but not CtBP2, is overexpressed in invasive osteosarcoma tissues and cells. The overexpressed CtBP1 in turn represses its downstream targets, such as the pro-apoptotic regulators Bax, Bim and p53 upregulated modulator of apoptosis (PUMA), cell adhesion molecule E-cadherin, and the cell cycle regulators p16, p21 and phosphatase and tensin homolog (PTEN). To explore the molecular mechanism of CtBP1 overexpression, we subjected three independent clinical samples to miRNA microarray analysis and found that miR-485-3p could specifically bind to the 3'-untranslated region (3'-UTR) of CtBP1, thereby negatively controlling CtBP1 expression. The overexpression of miR-485-3p in osteosarcoma cells significantly repressed CtBP1 levels and inhibited cell proliferation, colony formation, cell migration and sphere formation. Further analysis indicated that DNA hypermethylation in the promoter region of miR-485-3p caused the downregulation of miR-485-3p. Treatment with the DNA methylation inhibitor 5-aza-2'-deoxycytidine (AZA) resulted in the upregulation of miR-485-3p and the downregulation of CtBP1 as well as inhibited osteosarcoma cell growth. This study provides evidence that CtBP1 is also overexpressed in osteosarcoma cells and demonstrates the underlying mechanism regarding its overexpression. Thus, therapeutically targeting CtBP1 may represent an effective strategy for osteosarcoma therapy.

Keywords: CtBP1; CtBP2; DNA methylation; miR-485-3p; osteosarcoma.

Figure 1

CtBP1 is overexpressed in osteosarcoma tissues. (A and B) The relative mRNA levels of CtBP1/2 in noncancerous (HC, n=24) and cancerous (OS, n=30) tissues determined by qRT-PCR. **P<0.001. (C) The protein levels of CtBP1/2 in noncancerous (HC, n=4) and cancerous (OS, n=4) tissues. GAPDH was used as a control. (D) Statistical analysis of the protein levels in (C). **P<0.001.

Figure 2

CtBP1 is overexpressed in osteosarcoma cells. (A) The relative mRNA levels of CtBP1/2 in hFOB1.19, U2OS, MG63, Saos-2 and HOS cells determined by qRT-PCR. **P<0.001. (B) The relative mRNA levels of CtBP1 downstream target genes (Bax, Bim, PUMA, p16, p21, E-cadherin and PTEN) in hFOB1.19, U2OS, MG63, Saos-2 and HOS cells determined by qRT-PCR. *P<0.05, **P<0.001. (C) The protein levels of CtBP1 downstream targets (Bax, Bim, PUMA, p16, p21, E-cadherin and PTEN) in hFOB1.19, U2OS, MG63, Saos-2 and HOS cells. (D) Statistical analysis of the protein levels in (C). *P<0.05, **P<0.001.

Figure 3

The miRNA expression profile in osteosarcoma tissues. The heat maps of the consistently downregulated and upregulated miRNAs in osteosarcoma tissues. MicroRNAs from three noncancerous tissues (HC-1, -2 and -3) and three cancerous tissues (OS-1, -2 and -3) were subjected to microarray analysis. The heat maps indicated high (red) or low (green) levels of miRNA expression. (B-G) Verification of miRNA levels in osteosarcoma tissues by qRT-PCR. The qRT-PCR was performed to verify the expression of three downregulated miRNAs, including miR-19b (B), miR-200a (C), and miR-485-3p (D), and three upregulated miRNAs, including miR-22-3p (E), miR-198 (F), and miR-600 (G). ** P < 0.001.

Figure 4

CtBP1 is a direct target of miR-485-3p. The relative miR-485-3p level in hFOB1.19, U2OS, MG63, Saos-2 and HOS cells determined by qRT-PCR. **P<0.001. (B) Schematic representation of CtBP1 3′-UTR contained a putative miR-485-3p binding site. The binding position of miR-485-3p in the 3'-UTR of CtBP1 is shown by the red arrow. The seed location of miR-485-3 is indicated with the red font. The wild-type (WT) and mutant (Mut) 3′-UTRs of CtBP1 were shown. The following combinations of vectors: (a) no transfection, (b) miR-NC (negative control), (c) miR-NC + pCDNA3-CtBP1-3′-UTR^WT, (d) miR-NC + pCDNA3-CtBP1-3′-UTR^Mut, (e) miR-485-3p-mimic, (f) miR-485-3p-mimic + pCDNA3-CtBP1-3′-UTR^WT, (g) miR-485-3p-mimic + pCDNA3-CtBP1-3′-UTR^Mut, (h) anti-miR-485-3p, (i) anti-miR-485-3p + pCDNA3-CtBP1-3′-UTR^WT, and (j) anti-miR-485-3p + pCDNA3-CtBP1-3′-UTR^Mut, were transfected into U2OS cells, respectively. Then, qRT-PCR was performed to examine the miR-485-3p level (C) and CtBP1 mRNA level (D). **P<0.001. (E) Western blot was performed to detect the protein levels of CtBP1 and CtBP2 in cells used in (C). GAPDH was used as a loading control. (F) The miR-485-3p failed to bind the mutated 3′-UTR of CtBP1. The following combinations of plasmids were transfected into U2OS cells, respectively. (a) pGL3-CtBP1-3′-UTR^WT + pRL-TK-Renilla + miR-NC; (b) pGL3-CtBP1-3′-UTR^WT + pRL-TK-Renilla + miR-485-3p-mimic; (c) pGL3-CtBP1-3′-UTR^WT + pRL-TK-Renilla + anti-miR-485-3p; (d) pGL3-CtBP1-3′-UTR^Mut + pRL-TK-Renilla + miR-NC; (e) pGL3-CtBP1-3′-UTR^Mut + pRL-TK-Renilla + miR-485-3p-mimic; and (f) pGL3-CtBP1-3′-UTR^Mut + pRL-TK-Renilla + anti-miR-485-3p. The luciferase activity was measured using a Dual-Luciferase Reporter Assay System. ** P < 0.001.

Figure 5

The miR-485-3p-mimic upregulated the expression of CtBP1 target genes. Cells used in Figure 4C were subjected to examine the mRNA levels (A) and protein levels (B and C) of CtBP1 downstream targets, including Bax, Bim, PUMA, p16, p21, E-cadherin and PTEN were measured by qRT-PCR and Western blot, respectively. ** P < 0.001.

Figure 6

Overexpression of miR-485-3p inhibited oncogenic phenotypes. Cells used in Figure 4C were subjected to a series of studies to evaluate oncogenic phenotypes. (A) Cell proliferation assay assessed by MTT assay for 5 days. ** P < 0.001. (B) Colony formation assay was performed in RPMI1640 medium with a density of 100 cells per well. After incubation for 14 days, colonies were stained by crystal violet. (C) Colony numbers in (B) were quantified. **P < 0.001. (D) Sphere formation assay was carried out in FBS free DMEM-F12 medium supplemented with B27, EGF, bFGF and heparin for 10 days. Bars=50 μm. (E) Sphere numbers in (D) were quantified. **P < 0.001. (F) Cell migration assessed by scratch assay after treatment with mitomycin C for 2 h. The quantification results of the cell migration gap distance represented three biological replications. ** P < 0.001.

Figure 7

Downregulation of miR-485-3p in osteosarcoma cells was caused by the DNA hypermethylation of CpG islands. (A) The adjacent upstream of miR-485-3p had a CpG island. The genomic locus of the CpG island and miR-485-3p were shown. (B) AZA treatment significantly upregulated miR-485-3p levels. The hFOB1.19, U2OS and MG63 cells were treated with DMSO, 1 μM of AZA and 300 nM of TSA, respectively. The qRT-PCR was performed to measure miR-485-3p levels. **P < 0.001. (C) AZA treatment dramatically decreased DNA methylation of CpG island in U2OS and MG63 cells. The qMSP was performed to determine DNA methylation of CpG island in hFOB1.19, U2OS and MG63 cells. **P < 0.001. (D) AZA treatment dramatically decreased the CtBP1 level in U2OS and MG63 cells. The qRT-PCR was performed to determine CtBP1 levels in hFOB1.19, U2OS and MG63 cells. **P < 0.001. (E) AZA treatment significantly upregulated the CtBP1 target gene in U2OS and MG63 cells. The qRT-PCR was performed to determine CtBP1 targets, including Bax, Bim, PUMA, p16, p21, E-cadherin and PTEN in hFOB1.19, U2OS and MG63 cells. **P < 0.001.

Figure 8

Schematic model of miR-485-3p in osteosarcoma cells. (A) The working model of miR-485-3p in normal cells. In normal cells, the CpG methylation level in the miR-485-3p promoter region is very low, which cannot effectively inhibit the expression of miR-485-3p, resulting in the upregulation of miR-485-3p, downregulation of CtBP1 expression, and upregulation of CtBP1 downstream targets. (B) The working model of miR-485-3p in osteosarcoma cells. In osteosarcoma cells, the CpG methylation level in the miR-485-3p promoter region is very high, which significantly inhibits the expression of miR-485-3p, resulting in the downregulation of miR-485-3p, upregulation of CtBP1 expression, and downregulation of CtBP1 downstream targets, which eventually leads to carcinogenesis.