The collection and use of human samples was carried out strictly following The Declaration of Helsinki (7th revision, 2008). The present study obtained approval from the Ethics Committee of the Sixth Hospital of Shanxi Medical University (approval no. 201310012; Taiyuan, China). The pancreatic cancer tissue and adjacent non-tumorous tissue samples were collected from 37 patients (21–78 years old; 21 males and 16 females) who had undergone surgery at the Sixth Hospital of Shanxi Medical University from January 2014 to June 2018. Patients who had no other digestive system tumors or tumor history, had never received any chemotherapy or radiotherapy, and were confirmed as pancreatic cancer by pathology and genetics were included. Patients with serious injury to the heart, liver, kidney and other important organs, history of autoimmune diseases, and chronic or acute infectious diseases were excluded. All of the patients had provided a written informed consent prior to inclusion in the study.

The expression pattern of FGD5-AS1 in pancreatic cancer tissues and normal tissues was predicted using the Gene Expression Profiling Interactive Analysis (GEPIA) database (http://gepia.cancer-pku.cn/). The potential downstream target miRNAs of FGD5-AS1 were predicted by the starBase database (http://starbase.sysu.edu.cn/panCancer.php) and LncBase Predicted v.2 database (http://carolina.imis.athena-innovation.gr) (18).

Normal pancreatic ductal epithelial cells (HPDE6-C7 cells) and human pancreatic cancer cell lines (PANC-1, BXPC-3, CAPAN-1 and SW1990) were obtained from Shanghai Institute of Biochemistry and Cell Biology (Shanghai, China). Cells were cultured in Dulbecco's modified Eagle's medium (DMEM; Invitrogen; Thermo Fisher Scientific, Inc.) containing 0.1 mg/ml streptomycin, 100 U/ml penicillin and 10% fetal bovine serum (FBS; all from Hyclone; Cytiva) in an incubator at 37°C in 5% CO₂.

Control siRNA (si-NC), FGD5-AS1 siRNA (si-FGD5-AS1), miR-577 mimics (miR-577), miR-577 inhibitors (miR-577 in), microRNA mimics control (miR-con), and microRNA inhibitor control (miR-con in) were obtained from Shanghai GenePharma Co., Ltd. Cells were cultured in 6-well plates, and when cell confluence reached 80–90%, the pancreatic cancer cells were cultured in the fresh serum-free and antibiotic-free medium for 8 h. Subsequently, 100 nM siRNAs or 50 nM miRNAs were transfected with pancreatic cancer cells using Lipofectamine^® 2000 (Invitrogen, Thermo Fisher Scientific, Inc.) at room temperature according to the supplier's protocol. A total of 12 h following transfection, the medium was replaced by complete medium, and the cell culture was continued for 24 h. Subsequently, the transfection efficiency was evaluated by reverse transcription-quantitative polymerase chain reaction (RT-qPCR). The sequences of the oligonucleotides were as follows: FGD5-AS1 siRNA1: 5′-UUGGUCGUUGUCAACUUCCCA-3′; FGD5-AS1 siRNA2: 5′-UAUUGUAUGAAUACACUGCUA-3′; si-NC: 5′-UUCUCCGAACGUGUCACGUTT-3′; miR-577 mimics: 3′-UAGAUAAAAUAUUGGUACCUG-5′; microRNA mimics control: 5′- UAAGUAGUCUGAAAUAGUUAC-3′; miR-577 inhibitors: 3′-CAGGUACCAAUAUUUUAUCUA-5′. microRNA inhibitor control: 5′-AUCAGUUCAAUCAUGUAUCAU-3′.

Total RNA was extracted using TRIzol reagent (Invitrogen, Thermo Fisher Scientific, Inc.) following the manufacturer's protocol. To remove the DNA, 2 µg of RNA was treated with DNase and then reverse-transcribed into cDNA using SuperScript First Strand cDNA System (Invitrogen; Thermo Fisher Scientific, Inc.) according to the manufacturer's protocol. Subsequently, with cDNA as the template, RT-qPCR was performed using an ABI 7500 Fast Real-Time PCR System with SYBR^® PremixExTaq™ kit (Takara Biotechnology Co., Ltd.). The qPCR thermocycling conditions were as follows: pre-denaturation at 95°C for 10 min; followed by 40 cycles of denaturation at 95°C for 15 sec, and annealing and extension at 60°C for 60 sec. The primers were designed and synthesized by BGI (Shenzhen, China). The expression levels of FGD5-AS1, miR-577, low-density lipoprotein receptor-related protein 6 (LRP6) and β-catenin were determined with 2^−∆∆Cq method (19). The levels of U6 were used to normalize the miR-577 expression, and the levels of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) were used to normalize the expression levels of lncRNA and mRNA. The sequences of the primers are listed in Table I.

An MTT assay was utilized to determine cell proliferation rates. Cells in each group were transferred into 96-well plates (5×10⁴ cells/well), and routinely cultured. Following 12, 24, 48 and 72 h of culture, the cells were incubated with MTT reagent (Beyotime Institute of Biotechnology) for 4 h at 37°C. Subsequently, the formazan crystals were dissolved in dimethyl sulfoxide. Finally, a spectrophotometer was utilized to measure the absorbance of the cells at 570 nm.

In the migration assay, SW1990 cells were harvested and resuspended with serum-free medium, and then 200 µl of cell suspension (containing ~5×10⁴ cells) was added into the upper compartment of each Transwell chamber (8-µm pore size; BD Biosciences) in 24-well plates, and complete DMEM (containing glucose, L-glutamine, and sodium bicarbonate) with 10% FBS (600 µl per well) were added in the lower compartment. The cells were cultured for 24 h at 37°C, and then the cells on the upper surface of the membrane were gently wiped off with cotton swabs. The cells on the lower surface were fixed in 10% formalin for 20 min at room temperature and then stained with 0.1% crystal violet solution for 20 min at room temperature. Subsequently, the cells were counted under a light microscope. In the invasion assay, the procedures were the same as the migration assay using Transwell chambers covered with a layer of Matrigel (BD Biosciences) at 4°C overnight.

To determine the binding association between FGD5-AS1 and miR-577, the wild-type (WT) or mutant (MUT) FGD5-AS1 sequence was subcloned into the psi-CHECK2 reporter vector (Promega Corporation). The cells were subsequently transferred to 24-well plates. A total of 24 h later, miR-577 mimics or miR-NC were co-transfected into SW1990 cells with WT or MUT FGD5-AS1 reporter vectors, respectively, using Lipofectamine 2000 (Invitrogen; Thermo Fisher Scientific, Inc.). Luciferase activity was determined 48 h following transfection. Luciferase activity was determined by a Dual-Luciferase^® Reporter Assay system (Promega Corporation). Renilla luciferase activity was used as a control for firefly luciferase activity. Similarly, the targeting relationship between miR-577 and LRP6 (or β-catenin) was determined. To determine the activity of Wnt/β-catenin pathway, the TOPflash and FOPflash system (Upstate Biotechnology, Inc.) was used. This system was used to evaluate β-catenin-dependent signaling which drives the expression of T-cell factor (TCF). TOPflash is a TCF reporter plasmid containing WT TCF binding sites driven by the thymidine kinase minimal prompter and an upstream luciferase reporter gene. FOPflash contains mutated TCF binding sites driven by the same thymidine kinase promoter and upstream luciferase reporter gene. FOPflash was used as a control for TOPflash activity.

The interaction between FGD5-AS1 and miR-577 was analyzed using the EZ-Magna RNA binding protein immunoprecipitation kit (cat. no. 17–700; EMD Millipore). Briefly, following transfection, 1×10⁶ SW1990 cells were lysed in the RIP lysis buffer (included in the kit), and then 100 µl of cell lysate was mixed with 50 µl of magnetic beads coupled with 5 µg of anti-Ago2 antibody (cat. no. MABE253; EMD Millipore) or anti-IgG antibody (cat. no. AP101; EMD Millipore) in RIP buffer. Following incubation at 4°C for 8 h, the sample was centrifuged at 12,000 × g for 30 sec at 4°C, and the supernatant was discarded, and then the immunoprecipitate was obtained and incubated at 55°C for 30 min with proteinase K (EMD Millipore) to remove the protein. The RNA was isolated using TRIzol reagent according to the manufacturer's protocol. Subsequently, the RNA was reversely transcribed into cDNA (the procedure was the same as aforementioned in RT-qPCR section). Finally, the level of FGD5-AS1 in the immunoprecipitate was analyzed by RT-qPCR.

Biotin-labeled miR-577 (Bio-miR-577) and the control miRNAs (Bio-miR-NC) were incubated with 500 µl cell lysates at 4°C overnight. Then, streptavidin-coated 400 µl of magnetic beads (Thermo Fisher Scientific, Inc.) were added in the cell lysates, and incubated at 4°C overnight, to enrich the complex containing the biotin. Next, the complex was eluted, and then the RNA in the complex was extracted using TRIzol reagent as aforementioned. Subsequently, the FGD5-AS1 enrichment in the complex was evaluated via qPCR.

Cells were lysed in RIPA buffer (Pierce; Thermo Fisher Scientific, Inc.) containing protease inhibitors. Then the lysates were centrifuged at 12,000 × g for 15 min at 4°C and the supernatant was collected. The supernatant was added with loading buffer and boiled, and the protein samples were prepared. Next, 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis was employed to separate the protein samples (20 µg per lane), and then the proteins were electrotransferred onto polyvinylidene fluoride (PVDF) membranes (EMD Millipore). Subsequently, the proteins were blocked with 5% skimmed milk for 2 h at room temperature, and then incubated overnight with primary antibodies anti-Axin2 (product code ab109307; 1:1,000), anti-c-Myc (product code ab32072; 1:1,000), anti-cyclin D1 (product code ab16663; 1:1,000), β-actin (product code ab8227; 1:1,000; all from Abcam) at 4°C. After being rinsed three times with Tris-buffered saline and 0.05 % Tween-20 (TBST), the proteins were incubated with horseradish peroxidase-coupled secondary antibodies goat anti-rabbit IgG H&L (HRP) (ab205718; 1:5,000; Abcam) at room temperature for 30 min. Then, TBST was employed again, to wash the membranes 3 times. Ultimately, a hypersensitive ECL kit (Beyotime Institute of Biotechnology) was utilized to develop the protein bands on an X-ray film.

A total of 20 nude mice (male; 4 weeks old; weighing ~16 g) were obtained commercially from the National Laboratory Animal Center (Beijing, China). Mice were housed and maintained at ~20°C in a relative humidity of 40–70% with a 12-h light/dark cycle and received food and water ad libitum. All the mice were randomly divided into two groups (si-NC group vs. si-FGD5-AS1 group, 10 mice per group). SW1990 cells were transfected with si-NC or FGD5-AS1 siRNA1. Then the cells were suspended in sterile phosphate-buffered saline and injected into the tail vein of each 4-week-old mouse (1×10⁷ cells/mouse) (20). The animal health and behaviour were monitored weekly. A total of 3 weeks later, all 20 mice were anesthetized by intraperitoneal injection with 10% chloral hydrate (300 mg/kg), and then sacrificed by decapitation. The lung tissues were fixed in 4% paraformaldehyde at 4°C for 24 h and embedded in paraffin; after which, 4-µm sections were cut and stained with hematoxylin for 5 min and eosin for 2 min at room temperature. Metastatic nodules in the lungs were evaluated by pathological examination under a light microscope (magnification, ×200). No mouse exhibited signs of peritonitis, pain or discomfort following the administration of 10% chloral hydrate. The animal experiments were approved (approval no. 8217113963) by the Institutional Animal Care and Use Committee of Chinese PLA General Hospital (Beijing, China).

Data analysis was carried out using the SPSS 20.0 software (IBM Corp.). The data was presented as the mean ± standard deviation (SD). Pearson's correlation analysis was used to determine the correlation between the expression levels of FGD5-AS1 and miR-577 in pancreatic cancer tissues. The Kolmogorov-Smirnov test was used for normality and equal variance of the data. Unpaired Student's t-tests were utilized to perform comparisons. For data that had skewed distribution, comparisons between two groups were performed by Wilcoxon signed-rank test. P<0.05 was considered to indicate a statistically significant difference.

GEPIA database (http://gepia.cancer-pku.cn/) revealed that FGD5-AS1 was differentially expressed in pancreatic cancer tissues and normal tissues, and FGD5-AS1 expression in cancerous tissues was higher than that in normal pancreatic tissues (Fig. 1A). Consistently, RT-qPCR revealed that FGD5-AS1 expression was significantly elevated in pancreatic cancer tissues in comparison to the adjacent non-tumorous tissues (Fig. 1B). Subsequently, RT-qPCR was utilized for detecting FGD5-AS1 expression in the pancreatic cancer cell lines (SW1990, PANC-1, BXPC-3 and CAPAN-1) and pancreatic ductal epithelial cell line (HPDE6-C7), and it was revealed that compared with HPDE6-C7 cells, FGD5-AS1 expression was increased in the pancreatic cancer cell lines (Fig. 1C).

To determine the functions of FGD5-AS1 in pancreatic cancer progression, two siRNAs for FGD5-AS1 were designed. Then, FGD5-AS1 siRNA or control siRNA was transfected into SW1990 cells, and the knockdown efficiency was verified by RT-qPCR (Fig. 2A). Next, the more effective siRNA (si-FGD5-AS1 1) was selected for subsequent functional experiments. MTT and Transwell assays were utilized to detect proliferation, migration and invasion of SW1990 cells. It was revealed that FGD5-AS1 knockdown significantly suppressed viability, migration and invasion of SW1990 cells in contrast to the si-NC group (Fig. 2B and C).

To clarify the mechanism of FGD5-AS1 in pancreatic cancer progression, the potential downstream target miRNAs of FGD5-AS1 were predicted by the starBase v2.0 database (http://starbase.sysu.edu.cn/panCancer.php) and LncBase Predicted v.2 database (http://carolina.imis.athena-innovation.gr) (Tables SI and SII). Notably, a potential binding site between miR-577 and FGD5-AS1 was predicted (Figs. 3A and S1). Recent research has revealed that miR-577 is downregulated in pancreatic ductal adenocarcinoma, and the low expression of miR-577 is significantly associated with poor prognosis (21). Therefore, the role of miR-577 in pancreatic cancer was investigated. To verify the targeting relationship between miR-577 and FGD5-AS1, MUT FGD5-AS1 luciferase reporter vector (FGD5-AS1-MUT) or WT FGD5-AS1 luciferase reporter vector (FGD5-AS1-WT) was co-transfected into SW1990 cells with miR-577 mimics or control miRNA, and it was revealed that miR-577 could only suppress the luciferase activity of FGD5-AS-WT (Figs. 3B and S2). Subsequently, Pearson's correlation analysis revealed that FGD5-AS1 and miR-577 expression levels were negatively correlated in pancreatic cancer tissues (Fig. 3C). Additionally, RNA pull-down and RIP assays were performed, and it was revealed that ectogenic miR-577 enriched FGD5-AS1, and miR-577 and FGD5-AS1 were enriched in Ago2-containing microribonucleoproteins (Fig. 3D and E). All of the aforementioned findings indicated that miR-577 is a direct downstream target of FGD5-AS1.

The expression of FGD5-AS1 in pancreatic cancer was then investigated. RT-qPCR revealed that miR-577 expression was significantly reduced in pancreatic cancer tissues in comparison with para-tumorous tissues (Fig. 4A), and compared with HPDE6-C7 cells, miR-577 expression was significantly decreased in pancreatic cancer cell lines (Fig. 4B). Next, miR-577 inhibitors or miR-con inhibitors were transfected into SW1990 cells (Fig. 4C). Subsequently, MTT and Transwell assays were used to detect proliferation, migration and invasion of SW1990 cells. As revealed in Fig. 4D and E, miR-577 inhibition significantly increased proliferation, migration and invasion of SW1990 cells. These results suggested that miR-577 is a tumor suppressor in pancreatic cancer.

A previous study reported that miR-577 inhibits the growth of glioblastoma multiforme via modulating the Wnt/β-catenin pathway; both β-catenin and LRP6, two crucial components of the Wnt/β-catenin signaling, are directly targeted by miR-577 (22). Therefore, in the present study, it was hypothesized that miR-577 was also associated with the Wnt/β-catenin signaling in pancreatic cancer. Based on the previous study (22), WT luciferase reporter vectors (LRP6-WT and β-catenin-WT) and MUT luciferase reporter vectors (LRP6-MUT, β-catenin-MUT1, β-catenin-MUT2 and β-catenin-MUT1&2) were constructed. Subsequently, reporter vectors and miR-577 mimics were co-transfected into SW1990 cells, and the dual-luciferase reporter assays were performed. As revealed in Fig. 5A, miR-577 mimics significantly reduced the luciferase activity of LRP6-WT, yet exerted no significant effect on that of LRP6-MUT. miR-577 significantly suppressed the luciferase activity of β-catenin-WT, β-catenin-MUT1 and β-catenin-MUT2, yet did not suppress that of β-catenin-MUT1&2 (Fig. 5B). The aforementioned data suggested that miR-577 could bind to LRP6 3′ untranslated region (UTR) and β-catenin 3′UTR in pancreatic cancer cells. As expected, following transfection of miR-577 mimics into SW1990 cells, RT-qPCR revealed that the expression levels of β-catenin mRNA and LRP6 mRNA were significantly reduced (Fig. 5C). Subsequently, miR-577 mimics (or control miRNA) and the Wnt/β-catenin signaling reporter vector TOPflash were co-transfected into SW1990 cells, and the luciferase activity was detected 2 days following transfection. It was revealed that miR-577 mimics notably decreased the luciferase activity of pancreatic cancer cells, while control miRNA exerted no significant effect, which suggested that miR-577 indeed suppressed the activity of the Wnt/β-catenin signaling (Fig. 5D). Furthermore, Pearson's correlation analysis revealed that FGD5-AS1 expression was positively correlated with LRP6 expression and β-catenin expression in pancreatic cancer tissues (Fig. 5E and F), whereas miR-577 expression was negatively correlated with them (Fig. 5G and H). These results indicated that LRP6 and β-catenin were directly negatively regulated by miR-577, and FGD5-AS1 could probably indirectly modulate the activity of the Wnt/β-catenin pathway.

Subsequently, miR-577 inhibitors were transfected into SW1990 cells with FGD5-AS1 knockdown (Fig. 6A). The MTT assay revealed that miR-577 inhibition partially reversed the inhibitory effect of FGD5-AS1 knockdown on SW1990 cell proliferation (Fig. 6B). The Transwell assay indicated that miR-577 inhibition partially reversed the reduction caused by FGD5-AS1 knockdown in the migration and invasion of SW1990 cells (Fig. 6C). Additionally, to delve deeper into the association between the Wnt/β-catenin signaling pathway and the FGD5-AS1/miR-577 axis, western blotting was conducted to detect the expression levels of Axin2, cyclin D1 and c-Myc, which are considered as Wnt/β-catenin pathway-related proteins. It was revealed that FGD5-AS1 knockdown significantly decreased Axin2, cyclin D1 and c-Myc expression levels in SW1990 cells, and miR-577 inhibition partially counteracted these effects of FGD5-AS1 knockdown (Fig. 6D). The aforementioned findings suggested that FGD5-AS1 could regulate the activity of the Wnt/β-catenin signaling via modulating miR-577.

To confirm that FGD5-AS1 was indeed implicated in the metastasis of pancreatic cancer, a lung metastasis model was constructed with nude mice. After SW1990 cells were injected into the mice via the caudal vein, the metastatic nodules in the lung tissues were detected. It was revealed that SW1990 cells with FGD5-AS1 knockdown reduced metastatic potential (Fig. S3 and Table SIII), forming less metastatic nodules. This result further supported that FGD5-AS1 participates in the progression of pancreatic cancer.