Long noncoding RNA DIO3OS interacts with miR-122 to promote proliferation and invasion of pancreatic cancer cells through upregulating ALDOA

Human PC cell lines (AsPC-1, MIA PaCa-2, PANC-1 and BxPC-1) and human pancreatic duct epithelial cell line HPDE6-C7 were purchased from the American Type Culture Collection (Rockville, USA). The cells were grown in RPMI1640 or DMEM (Gibco, USA) supplemented with 10% fetal bovine serum (FBS). Plasmids containing DIO3OS and ALDOA, or an empty vector pcDNA3.1 were obtained from Genepharma (Shanghai, China). DIO3OS siRNA, control siRNA miR-122 mimic, control mimic, miR-122 inhibitor and control inhibitor were purchased from IGEbio (Guangzhou, China). Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA) was used for cell transfection following the manufacturer’s instructions.

Total RNA was isolated from cells using TRIzol reagent (Invitrogen, Carlsbad, CA, USA) and then was converted to cDNA using an M-MLV Reverse Transcriptase Kit (Invitrogen, Carlsbad, CA, USA). Real-time PCR analysis was carried out using the SYBR-Green-quantitative real-time PCR Master Mix kit (Toyobo, Osaka, Japan). The primers for DIO3OS and GAPDH have been reported [5]. GAPDH served as the endogenous control. For detecting miRNA expression, the mirVanaTM qRT-PCR microRNA Detection Kit (Ambion Inc., Austin, TX, USA) was used according to the manufacturer’s instructions. MiR-122 expression was normalized to U6.

Total protein from cells was extracted using RIPA buffer (Beyotime). An equal amount of each protein sample was separated on a 10% SDS-PAGE gel and transferred to a PVDF membrane (Millipore, Bedford, MA, USA). The membranes were blocked with 5% nonfat milk at room temperature for 1 h and incubated with specific primary antibody, ALDOA (1:2000, Abcam, Cambridge, UK) and GAPDH (1:5000, Santa Cruz, CA, USA) overnight, followed by incubation with HRP-conjugated secondary antibodies (Santa Cruz, CA, USA). The protein bands were detected using ECL western blotting kit (Amersham Biosciences, Buckinghamshire, UK). GAPDH was used as the loading control.

Cell proliferation was measured by performing the CCK-8 assay (Beyotime Institute of Biotechnology, Jiangsu, China) according to the manufacturer’s instructions. 5000 cells were seeded into a 96-well plate and were transfected as indicated. Then, cell proliferation was measured 72 h after transfection. The absorbance was measured at 450 nm by a microplate reader (Bio-Rad, Hercules, CA, USA).

A total of 1000 cells were seeded in 24-well plates. After culturing for 14 days, colonies were fixed with 100% methanol for 15 min and then stained with 0.5% crystal violet for 20 min. Colonies with > 100 cells were counted and analyzed.

Transwell invasion assay was performed according to the method described previously [6, 7]. Briefly, 2 × 10⁴ cells in serum-free medium were plated into the upper chamber. The medium containing 10% FBS was added to the lower chamber. After culturing for 24 h, Cell that had invaded were fixed with 75% methanol and stained with crystal violet. Evaluation of invasive capacity was performed by counting invaded cells under a microscope, and five random fields of view were analyzed for each chamber.

A luciferase reporter assay was performed as previously described [8]. The fragment from wild-type DIO3OS (DIO3OS-WT) containing the predicted miR-122-binding site, mutant DIO3OS (DIO3OS-MUT), the 3′-UTR fragment from wild-type ALDOA 3′-UTR (ALDOA-WT) containing the potential miR-122-binding site and mutant ALDOA 3′-UTR (ALDOA-MUT) were amplified using PCR and sub-cloned into a pMIR-GLO Luciferase vector (Promega, Madison, WI, USA). PC cells were co-transfected with the above luciferase reporter vectors containing DIO3OS (WT or MUT) or ALDOA 3′-UTR (WT or MUT) together with miR-122 mimic, miR-122 inhibitor or their negative controls using Lipofectamine 2000 (Invitrogen). The relative luciferase activity was measured with the Dual-Luciferase Reporter Assay System (Promega, China) after 48 h.

RNA immunoprecipitation assay was conducted using the Magna RIP RNA-Binding Protein Immunoprecipitation Kit (Millipore) as previously described [9, 10]. Briefly, anti-Argonaute2 (Ago2) antibody (Millipore, Bedford, MA, USA) or normal mouse IgG (Millipore) as a negative control were conjugated to magnetic beads and were incubated with the cell extract in RIP buffer. The immunoprecipitated RNAs were isolated and were subjected to qRT-PCR analysis of DIO3OS and miR-122 expression.

All animal procedures were approved by the Institutional Animal Care and Use Committee of First Affiliated Hospital of Zhengzhou University. Female BALB/c nude mice (4 weeks old) were purchased from Beijing HFK Bioscience (Beijing, China) and maintained under pathogen-free conditions.

To evaluate the in vivo tumorigenic effects, MIA PaCa-2 and AsPC-1 cells (2 × 10⁶) were inoculated subcutaneously in the right flank of the nude mice (n = 4 per group). After implantation for 6 days, tumor volume measurement began and was performed every 3 days, using the following formula: volume = length (mm) × width² (mm²)/2. After 3 weeks, the mice were sacrificed and the tumors were collected for immunohistochemistry assay. The xenografts tissues were formalin-fixed/paraffin-embedded and cut into 4 μm slides. The primary antibody used was anti-Ki-67 (1:1000, Abcam, Cambridge, UK). The secondary streptavidin–horseradish peroxidase-conjugated antibody staining was performed at room temperature, visualized in 3, 3′-diaminobenzidine (ZLI9018, ZSGBBIO, China).

First, we analyzed the TCGA expression data of DIO3OS using the MethHC database (http://​methhc.​mbc.​nctu.​edu.​tw/​php/​index.​php). DIO3OS was highly overexpressed in a variety of cancer types (including PC) as compared to the corresponding normal tissues (Fig. 1a). Then, we performed the meta-analysis on PC samples and non-tumor samples retrieved from the TCGA database using the UALCAN web server [11]. We found that the DIO3OS transcript was remarkably increased in PC tissues compared with normal samples (Fig. 1b). Furthermore, we measured the levels of DIO3OS in four PC cell lines (AsPC-1, MIA PaCa-2, PANC-1 and BxPC-1 cells) and a normal pancreatic duct epithelial cell line HPDE6-C7. Our qRT-PCR experiments suggested that DIO3OS expression was significantly elevated in PC cells compared to HPDE6-C7 cells (Fig. 1c). Furthermore, we surveyed the association between DIO3OS levels and overall survival in TCGA PC dataset using the web portal of UALCAN (http://​ualcan.​path.​uab.​edu/​index.​html). A total of 177 PC patients with RNA sequencing data and overall outcome data were used for Kaplan–Meier survival analysis. The patients with higher level of DIO3OS displayed shorter overall survival period (Fig. 1d). These data implied that increased expression of DIO3OS may be associated with tumorigenesis or progression of PC.

The effect of DIO3OS on the ability of PC cells to proliferate and invade after transfection with DIO3OS siRNA or control siRNA was then characterized using CCK-8 assay, clone formation assays and Matrigel invasion assays. CCK-8 assays, clone formation assays and Matrigel invasion assays revealed that the knockdown of DIO3OS significantly decreased cell proliferation and invasion in MIA PaCa-2 cells (Fig. 2a–d). To further explore the oncogenic roles of DIO3OS in PC, we overexpressed DIO3OS in AsPC-1 cells by transfecting AsPC-1 cells with the DIO3OS expression vector (Fig. 2a). CCK-8 assays, clone formation assays and invasion assays showed that the overexpression of DIO3OS can promote proliferation and invasion in AsPC-1 cells (Fig. 2b, c and e), suggesting a critical role for DIO3OS in promoting growth and invasiveness of PC cells.

Several studies have identified that lncRNAs might act as a sponge for miRNAs [12]. To examine whether DIO3OS regulates the expression of miRNAs in such a manner, interactions between DIO3OS and miRNAs were predicted using starBase v2.0 [13]. Interestingly, this bioinformatics analysis of miRNAs target sequences on DIO3OS revealed that miR-122 was complementary to DIO3OS sequence (Fig. 3a). To test whether DIO3OS can regulate miR-122 expression, we investigated the effect of DIO3OS knockdown or overexpression on miR-122 expression in PC cells. Our qRT-PCR analysis suggested that miR-122 expression was upregulated in MIA PaCa-2 cells when DIO3OS was knockdown (Fig. 3b). However, the overexpression of DIO3OS downregulated the expression of miR-122 in AsPC-1 cells (Fig. 3b). Then, miR-122 expression levels in two PC cell lines and the normal renal cell line HPDE6-C7 were investigated. Our qRT-PCR results showed an inverse correlation between DIO3OS and miR-122 expression in PC cell lines (Fig. 1c, e).

To confirm whether DIO3OS can interact with miR-122, the fragment of DIO3OS was inserted downstream of the luciferase gene and a dual-luciferase reporter assay was performed to explore the relationship between DIO3OS and miR-122. When miR-122 was overexpressed, the luciferase activity of wild-type DIO3OS was significantly reduced (Fig. 3c). To further determine whether there is a direct interaction between miR-122 and DIO3OS putative binding site, we mutated the miR-122 binding site by site-directed mutagenesis, and found that mutations in the DIO3OS abrogated the repressive effect of miR-122 (Fig. 3c). As expected, the transfection with miR-122 inhibitor significantly increased in the luciferase activity of wild-type DIO3OS in AsPC-1 cells, while the luciferase activity of mutant DIO3OS was unaffected (Fig. 3c).

To investigate whether DIO3OS and miR-122 are part of the RNA-induced silencing (RISC) complex, RIP experiment was conducted on MIA PaCa-2 cells lysates using an antibody against Ago2, a key component of the RISC complex. When DIO3OS and miR-122 levels were quantified in the immunoprecipitates using qRT-PCR analysis, they were found to be enriched in Ago2 immunoprecipitates compared to control IgG immunoprecipitates (Fig. 3d).

To examine if miR-122 was capable of counteracting the DIO3OS-mediated promotion of cell proliferation and invasion, MIA PaCa-2 cells were transfected with DIO3OS siRNA in combination with miR-122 inhibitor, and AsPC-1 cells were transfected with the DIO3OS vector in combination with miR-122 mimic. The inhibition of miR-455 rescued the effects of DIO3OS silencing on MIA PaCa-2 cell proliferation and invasion (Fig. 3e). Moreover, the overexpression of miR-455 overcame the effects of DIO3OS overexpression in AsPC-1 cells (Fig. 3f). Taken together, these results provide evidence that miR-122 was involved in DIO3OS-mediated regulation of PC cell proliferation and invasion.

We explored the genes that were potentially regulated by miR-122. Our bioinformatic-based target prediction analysis using TargetScan (http://​www.​targetscan.​org) indicated that ALDOA was predicted to contain the binding sequence of miR-122 (Fig. 4a). Consistently, the protein expression of ALDOA was elevated in AsPC-1 and MIA PaCa-2 cells compared to the normal pancreatic cell line HPDE6-C7 (Fig. 1f), supporting a negative correlation between miR-122 and ALDOA expression.

As shown in Fig. 4b, luciferase assays showed the ectopic expression of miR-122 could significantly decrease the luciferase activity of the wild-type ALDOA 3′-UTR, but not that of the mutant ALDOA 3′-UTR. In contrast, the knockdown of miR-122 significantly enhanced the luciferase activity of the wild-type ALDOA 3′-UTR, but did not affect the luciferase activity of the mutant ALDOA 3′-UTR (Fig. 4b). Western blot analysis demonstrated that overexpression of miR-122 could reduce the levels of ALDOA. However, the inhibition of miR-122 increased the protein expression of ALDOA in PC cells (Fig. 4c). These data suggested that miR-122 directly binds to ALDOA 3′-UTR and represses ALDOA expression level in PC cells.

Then, we asked whether DIO3OS might regulate the level of ALDOA through inhibiting miR-122 expression. Western blot analysis revealed that decreased expression of ALDOA by DIO3OS knockdown could be partially rescued by miR-122 inhibitor, and the induction of ALDOA expression caused by DIO3OS overexpression could be largely reduced by miR-122 mimic in AsPC-1 cells (Fig. 4d), suggesting that DIO3OS enhances ALDOA expression through negative modulation of miR-122.

We investigated whether the functions of DIO3OS and miR-122 in MIA PaCa-2 cells were dependent on the expression of ALDOA. CCK-8 and invasion assays showed that the suppression of cell proliferation and invasion by DIO3OS knockdown or miR-122 overexpression could be restored by the ectopic expression of ALDOA (Fig. 5a–d). On the other hand, the increased proliferation and invasion of AsPC-1 cells by DIO3OS overexpression or miR-122 inhibition was suppressed by the knockdown of ALDOA (Fig. 5e–h).

Importantly, to further confirm the existence of the DIO3OS/miR-122/ALDOA axis in PC tissues, we detected the levels of miR-122 and ALDOA in PC samples and non-tumor samples obtained from the TCGA database using the MethHC database and UALCAN web server, respectively. We found that the level of miR-122 was downregulated, while the expression of ALDOA was upregulated in PC tissues compared with normal samples (Fig. 1g, h), indicating a negative correlation between DIO3OS and miR-122 expression (Fig. 1b, g), and a negative association between miR-122 and ALDOA expression in PC samples (Fig. 1g, h). These data suggested that ALDOA mediates the effects of DIO3OS and miR-122 in PC cells.

To further investigate whether DIO3OS promotes PC growth in vivo, we knocked down the expression of DIO3OS in MIA PaCa-2 cells or overexpressed DIO3OS in AsPC-1 cells, and injected these cells into the flanks of nude mice to establish subcutaneous PC xenografts. As shown in Fig. 6a, b, the depletion of DIO3OS could suppress the tumor growth, and the overexpression of DIO3OS significantly increased tumor growth. Furthermore, the knockdown of DIO3OS downregulated the expression of Ki-67, a cell proliferation marker (Fig. 6c), and the overexpression of DIO3OS increased the levels of Ki-67 in vivo (Fig. 6d). Collectively, these results showed that DIO3OS promotes PC cell growth in vivo. Taken together, our results support the notion that DIO3OS promotes PC cell growth and invasion via regulating the miR-122/ALDOA axis (Fig. 6e).