Long non-coding RNA UFC1 promotes gastric cancer progression by regulating miR-498/Lin28b

A total of 79 paired gastric cancer and adjacent non-cancerous tissues (5 cm away from the tumor edge), 60 serum samples from GC patients, 35 serum samples from gastritis patients, and 40 serum samples from healthy donors were obtained from Department of General Surgery, the Affiliated People’s Hospital of Jiangsu University between April 2015 and September 2016. Written informed consent were obtained from all the patients and this study was approved by the Institutional Ethical Committee of Jiangsu University. All of the tissues were frozen in liquid nitrogen and then stored at − 80 °C for further use. The intravenous blood was centrifuged at 3000 g for 10 min and the serum were stored at − 80 °C until RNA extraction. The patients included in this study had not received any preoperative therapies.

Human GC cell lines MGC-803, BGC-823, SGC-7901, HGC-27, and MKN-45 were purchased from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China). Human normal gastric mucosa epithelial cell line GES-1 was obtained from Gefan Biological Technology (Shanghai, China). MGC-803 and HGC-27 cells were cultured in high glucose-DMEM with 10% fetal bovine serum (FBS; Invitrogen, Shanghai, China). GES-1, BGC-823, SGC-7901, and MKN-45cells were cultured in RPMI 1640 medium (Invitrogen) containing 10% FBS. All the cells were cultured in a 37 °C incubator with 5% CO₂ atmosphere.

Total RNA was isolated from tissue samples and cells by using Trizol reagent (Invitrogen) according to the manufacturer’s procedures. Total RNA in serum was purified using miRNeasy Serum/Plasma kit according to the manufacturer’s instructions (Qiagen, Shanghai, China). The reverse transcription (RT) for mRNA, miRNA and lncRNA was carried out by using the miScript II RT Kit (Qiagen). Quantitative real time polymerase chain reaction was conducted with UltraSYBR Mixture (Cwbio, Beijing, China) on a real time PCR Detection System (CFX96, Bio-Rad, Shanghai, China). The target genes were normalized to U6 to obtain the relative expression level. The sequences of primers were provided in Additional file 1: Table S1.

Cells were lysed with RIPA buffer (Beyotime, Shanghai, China) containing protease inhibitors (Roche, CA, USA). Equal amounts of proteins were separated by SDS-polyacrylamide gel electrophoresis (SDS-PAGE) on a 12% polyacrylamide gel. The proteins were transferred electrophoretically onto 0.22 μm PVDF membranes (Millipore), blocked in 5% non-fat milk, and then incubated with primary antibodies against cyclin D1, Bcl2, Bax, Slug, Twist, Snail, E-cadherin, N-cadherin, and Vimentin (Cell Signaling Technology, Shanghai, China). After incubation with HRP-linked secondary antibody, the protein bands were visualized by using chemiluminescence (Millipore, Shanghai, China). GAPDH was used as the loading control.

Cells were seeded in 6-well plates at a density of 2 × 10⁵/well and cultured in 37 °C incubator overnight. The overexpressing plasmid and knockdown shRNA (Hanbio, Shanghai, China) were transfected into the cells by using LipoFiter transfection reagent (Hanbio) in serum-free medium. Cells were changed to complete medium at 6 h after transfection and cultured for another 30 h. The target sequences of shRNAs were provided in Additional file 1: Table S2.

MGC-803 cells were co-transfected with miR-498 mimics (or miR-498 antago) and the luciferase reporter vector containing wild type (WT) or mutant (MUT) 3’-UTR of UFC1 (or Lin28b) as indicated. At 36 h after transfection, the cells were lysed and the luciferase activity was detected by using the dual luciferase assay kit (Promega, Madison, WI, USA).

RNA immunoprecipitation (RIP) assay was performed by using a Magna RIP kit (Millipore) according to the instruction of the manufacturer. Whole cell lysate was incubated with RIP buffer containing magnetic beads which had been conjugated with human anti-Ago2 antibody, or normal mouse IgG as negative control. The immunoprecipitated RNAs were extracted, purified, and analyzed by using qRT- PCR to detect the binding of target RNAs.

Cell cycle analysis was conducted with a cell cycle detection kit (Fcmacs, Jiangsu, China). The transfected cells were collected and fixed in 95% ethanol overnight. Afterwards, the cells were stained with 50 μg/ml propidium iodide (PI) for 30 min in dark. The cell cycle distribution was analyzed on a flow cytometer (BD, FACS Calibur) by using CellQuest software.

The Annexin V-Alexa Fluor 647/PI apoptosis detection kit (Fcmacs, Jiangsu, China) was used to detect cell apoptosis. After transfection for 24 h, the cells were digested with collagenase. The cells were collected and resuspended in binding buffer. Subsequently, the cells were stained with Annexin V-Alexa Fluor 647 and PI and then incubated for 15 min at room temperature. The apoptotic rate was analyzed by using flow cytometry.

The transfected cells were seeded in 24-well plates (1 × 10⁴ per well) and were counted for 6 days. For cell colony formation assay, the transfected cells were seeded in 6-well plates (1 × 10³/well) and cultured for 10 days. The medium was changed every 3 days. The cells were fixed with 4% paraformaldehyde and stained with crystal violet. The number of colonies was calculated under a microscope.

Cell migration assay were carried out by using transwell chambers with inserts of 8-μm pore size (Chemicon, Temecula, CA, USA). A total of 2 × 10⁴ cells were seeded into the top chamber in 200 μL serum-free media and 600 μL complete medium was added into the bottom chamber. After 24 h, the cells were fixed with 4% paraformaldehyde and stained with crystal violet. The number of migrated cells was counted from five randomly selected field and then averaged.

After matrigel (BD Biosciences, Shanghai, China) was added on the transwell chamber and clotted, a total of 1 × 10⁵ cells were seeded into the top chamber in 200 μL serum-free media. The bottom well was added with 600 μL complete medium and cells were allowed to invade for 36 h. The matrigel and the cells on the top chamber were removed with cotton swab. The cells invaded through the pore were fixed with 4% paraformaldehyde and stained with crystal violet. The number of invaded cells were counted from five randomly selected fields and averaged.

BALB/c nude mice aged 4–6 weeks were purchased from the Slac Laboratory Animal Center (Shanghai, China) and maintained in accordance with the institutional policies. Sh-UFC1 or sh-control stably transfected cells were collected in PBS and subcutaneously injected into the mice (2 × 10⁶ cells/mice, n = 6). Tumor size was assessed every 3 days and tumor volumes were calculated using the formula: V = 0.5 × D × d² where V represents volume, D represents longitudinal diameter and d represents latitudinal diameter. The protocol was approved by the Animal Use and Care Committee of Jiangsu University.

For immunohistochemical analyses, 4% paraformaldehyde fixed tissues were embedded in paraffin and cut into 4 μm-thick sections. The sections were incubated with primary monoclonal antibody against Ki-67 (Cell Signaling Technology) followed by incubation with the secondary antibody for 30 min at room temperature. After being incubated with 3, 3′-Diaminobenzidine (3, 3’-DAB, Maxim, Fuzhou, China) for 5 min, the sections were counterstained with hematoxylin for 30 s. Finally, the sections were photographed under a TE2000 microscope (Nikon, Tokyo, Japan).

We first detected the relative expression levels of UFC1 in 79 paired gastric cancer tissues and adjacent non-tumor tissues. The results showed that 64.6% (51/79) of GC tissues exhibited at least two-fold increase in UFC1 expression level compared to the paired non-cancerous tissues (Fig. 1a, P < 0.001). We then explored the correlation between UFC1 expression level and clinicopathological parameters. The expression level of UFC1 was positively correlated with tumor size, TNM stage and lymphatic metastasis (Additional file 1: Table S3). Furthermore, the patients who had high levels of UFC1 came out with a notably poorer prognosis than those who had low levels of UFC1 (Fig. 1b, P < 0.05).We next examined the expression level of UFC1 in the serum of patients with gastric cancer. The serum levels of UFC1 were elevated in GC patients compared to that in gastritis patients and healthy controls (Fig. 1c, P < 0.001). We then determined the correlation between the expression levels of UFC1 in serum and the clinicpathological parameters. The results of correlation analyses showed that the serum levels of UFC1 were positively associated with TNM stage and lymphatic metastasis (Additional file 1: Table S4). We further isolated exosomes from the serum samples and detected the expression levels of UFC1 in exosomes. We found that the expression level of exosomal UFC1 was increased in GC patients compared to that in healthy controls (Fig. 1d). The levels of exosomal UFC1 were also positively associated with TNM stage and lymphatic metastasis (Additional file 1: Table S5). The area under the receiver operating characteristic (ROC) curve for exosomal UFC1 was 0.860 (95% CI, 0.780 to 0.900, Fig. 1e). Finally, we assessed the expression levels of UFC1 in GC cell lines and normal gastric mucosa epithelial cell line. The results showed that UFC1 expression was higher in GC cell lines (including HGC-27, MGC-803, BGC-823 and SGC-7901) than that in normal gastric mucosa epithelial cell line GES-1 (Fig. 1f).

The increased expression of UFC1 in GC led us to hypothesize that UFC1 might function as an oncogene in GC. To this end, we knocked down UFC1 expression in GC cells by using shRNA. The knockdown efficiency was verified by using qRT-PCR (Fig. 2a). The results of cell counting assay showed that UFC1 knockdown inhibited gastric cancer cell proliferation (Fig. 2b), which was further confirmed by the results of colony formation assays (Fig. 2c). GC cells transfected with UFC1 shRNA had decreased S phase and increased apoptosis (Fig. 2d, e). The number of migrated and invaded cells was decreased in sh-UFC1 group compared to that in control group (Fig. 2f and g). The inhibitory effects of UFC1 knockdown on GC cell proliferation, migration and invasion were further confirmed by using another UFC1-targeting shRNA (Additional file 2: Figure S1). Moreover, the expression levels of Bcl-2 and cyclin D1 was decreased in shRNA transfected cells (Fig. 2h and i). The expression of epithelial marker E-cadherin was increased while that of mesenchymal marker N-cadherin and Vimentin was decreased in UFC1 knockdown cells. In addition, the expression of EMT transcription factors including Slug, Snail, and Twist was significantly downregulated in sh-UFC1 transfected cells (Fig. 2h, i). In contrast to that observed for UFC1 knockdown, UFC1 overexpression promoted GC cell proliferation, migration and invasion (Additional file 2: Figure S2).

LncRNAs could function as ceRNAs to regulate target gene expression by binding to miRNA. We predicted the potential targets of UFC1 by searching the bioinformatic database miRDB and identified miR-498 as a candidate binding miRNA (Fig. 3a and Additional file 2: Figure S3). Thus, we attempted to analyze the regulation of UFC1 by miR-498. QRT-PCR results showed that UFC1 expression was downregulated in miR-498 transfected GC cells (Fig. 3b). Luciferase reporter assay results showed that the luciferase activity of UFC1–3’UTR was significantly decreased in miR-498 transfected group. On the contrary, the luciferase activity of UFC1–3’UTR was increased in miR-498 antago transfected cells (Fig. 3c). However, no significant difference was found in the luciferase activities of mutant UFC1–3’UTR between miR-498 (or miR-498 antago) group and control group (Fig. 3c). In addition, the half-life of UFC1 was remarkably shortened in GC cells transfected with miR-498 (Fig. 3d). RIP assay results showed that UFC1 and miR-498 were co-immunoprecipitated by the anti-Ago2 antibody but not the IgG antibody (Fig. 3e). The expression level of miR-498 was significantly lower in tumor tissues than that in adjacent non-tumor tissues (Additional file 2: Figure S4A). In support of this finding, miR-498 expression level was lower in gastric cancer cell lines than that in normal gastric mucosa epithelial cell line (Additional file 2: Figure S4B). Moreover, there was a negative association between the expression level of UFC1 and that of miR-498 in tumor tissues (r = − 0.52, P < 0.001, Fig. 3f). GC patients with high levels of miR-498 had a better prognosis that those with low levels of miR-498 (Fig. 3g, P < 0.05). Taken together, these findings suggest that UFC1 function as a miR-498 sponge in gastric cancer cells.

We next determined the biological roles of miR-498 in GC. The results of cell counting and colony formation assays showed that miR-498 overexpression inhibited GC cell proliferation (Fig. 4a, b, c). MiR-498 overexpression increased the percentage of cells at G1 phase while decreased that of cells at S phase (Fig. 4d). The rate of apoptotic cells increased when miR-498 was overexpressed in GC cells (Fig. 4e). In addition, miR-498 overexpression inhibited the migration and invasion of GC cells (Fig. 4f, g). In consistent with that observed for UFC1 knockdown, miR-498 overexpression also led to the downregulation of Bcl-2, cyclin D1 and EMT transcription factors including Slug, Snail, and Twist in GC cells (Fig. 4h, i). However, UFC1 co-transfection reversed the suppressive roles of miR-498 in GC cell proliferation, migration and invasion (Additional file 2: Figure S5).Taken together, these findings suggest that miR-498 plays a tumor suppressive role in gastric cancer.

We then predicted the potential target genes of miR-498 by using online miRNA target prediction tools including Targetscan, DIANA, and miRDB and identified Lin28b as a potential downstream gene of miR-498 (Fig. 5a and Additional file 2: Figure S3). Lin28b expression was downregulated in miR-498 transfected GC cells (Fig. 5b). There was a negative association between the expression level of Lin28b and that of miR-498 in tumor tissues (r = − 0.414, P < 0.01, Fig. 5c). Moreover, the expression level of Lin28b was significantly upregulated in gastric cancer tissues and gastric cancer cell lines compared to non-tumor tissues and normal gastric mucosa epithelial cell line, respectively (Additional file 2: Figure S4C, D). MiR-498 overexpression decreased while miR-498 antago increased the luciferase activity of wild type Lin28b-3’UTR (Fig. 5d). However, no significant difference was found in the luciferase activity of mutant Lin28b-3’UTR between miR-498 (or miR-98 antago) and control groups.

We further explored the biological roles of Lin28b in GC. We found that Lin28b knockdown inhibited GC cell proliferation (Additional file 2: Figure S6A, B, C). Lin28b knockdown induced cell cycle arrest and apoptosis in GC cells (Additional file 2: Figure S6D, E). Lin28b knockdown also inhibited the migration and invasion of GC cells (Additional file 2: Figure S6F, G). The expression of Bcl-2, Cyclin D1, and EMT transcription factors including Slug, Snail, and Twist was downregulated in Lin28b knockdown GC cells (Additional file 2: Figure S6H, I). On the contrary, Lin28b overexpression led to the opposite effects (Additional file 2: Figure S7).

We conducted rescue experiments to determine the importance of Lin28b in miR-498-regulated tumor suppression in GC. The co-transfection of Lin28b rescued the inhibitory roles of miR-498 in the proliferation, migration and invasion of GC cells (Fig. 5e, h, i). Lin28b co-transfection also reversed the inducing effects of miR-498 on cell cycle arrest and apoptosis in GC cells (Fig. 5f, g). Therefore, these data suggest that Lin28b is an important downstream target of miR-498.

Lin28b expression was downregulated in UFC1 knockdown GC cells (Fig. 6a). The luciferase activity of Lin28b-3’UTR was increased in UFC1-overexpressing group but decreased in UFC1 knockdown group (Fig. 6b). There was no significant difference in the luciferase activity of mutant Lin28b-3’UTR between UFC1 (or sh-UFC1) and control groups (Fig. 6b). On the contrary, a positive association between the expression level of UFC1 and that of Lin28b was observed in gastric cancer tissues (r = 0.604, P < 0.001, Fig. 6c).

We conducted rescue experiments to determine the importance of Lin28b in UFC1-regulated tumor progression in GC. The co-transfection of Lin28b rescued the inhibitory roles of UFC1 knockdown in the proliferation, migration and invasion of GC cells (Fig. 6d, g, h). Lin28b co-transfection also reversed the inducing effects of UFC1 knockdown on cell cycle arrest and apoptosis in GC cells (Fig. 6e, f). On the contrary, the promoting roles of UFC1 overexpression in GC cell proliferation, migration and invasion were reversed by Lin28b knockdown (Additional file 2: Figure S8).

To further confirm the significance of Lin28b regulation by UFC1 in tumor growth, UFC1 knockdown cells co-transfected with or without Lin28b were subcutaneously injected into nude mice. The mean tumour size and tumor weight in sh-UFC1 group were remarkably lower than that in control group (Fig. 6i). However, the mean tumor size and tumor weight in Lin28b co-transfection group were remarkably higher than that in sh-UFC1 alone group (Fig. 6i). In addition, the results of immunohistochemical analyses showed that the tumors tissues from UFC1 knockdown group displayed weaker Ki-67 staining than those from the control group (Fig. 6j). However, the tumors tissues from Lin28b co-transfection group displayed stronger Ki-67 staining than those from sh-UFC1 alone group (Fig. 6j). Taken together, these findings suggest that UFC1 promotes GC growth through the upregulation of Lin28b.