The study was approved by the committee of West China Hospital of Sichuan University, and all subjects had signed a written consent form. Animal treatments were processed in coincidence with Animal Care and Use Guidelines of Animal Ethic Committee in West China Hospital of Sichuan University.
Ninety-six PC patients were enrolled, and cancer tissue and normal tissues were surgically obtained. Patients were confirmed as PC through the WHO standard histopathological diagnosis. No radiotherapy or chemotherapy was performed before and after the operation. After the operation, the patients were followed up every six months by outpatient reviews, and the median follow-up was 60 months.
PC cell lines AsPC-1 (CRL-1682), BxPC-3 (CRL-1687), PANC-1 (CRL-1469), MIA PaCa-2 (CRM-CRL-1420) and SW1990 (CRL-2172), and human pancreatic duct epithelial cell line hTERT-HPNE (CRL-4023) were obtained from ATCC (CA, USA). AsPC-1 and BxPC-3 cells were cultured with Roswell Park Memorial Institute 1640 medium (A4192301), PANC-1 and MIA PaCa-2 cells in Dulbecco’s Modified Eagle Medium (DMEM; A4192101) while SW1990 cells in L-15 medium (11415114). The above medium was added with 10% fetal bovine serum (FBS; 16140063). hTERT-HPNE cells were placed in complete growth medium. All media were purchased from Gibco (CA, USA) .
PANC-1 cells with 40-50% confluence were subjected to transfection with plasmids (GenePharma, Shanghai, China) through Lipofectamine® 2000 (11668019; Invitrogen, CA, USA). The plasmids were transfected alone or in combination as following: small interfering RNA (si)-negative control (NC), si-MIR4435-2HG, mimic-NC, mimic-miR-582-5p, short hairpin RNA (sh)-NC, sh-GPX8, si-MIR4435-2HG + inhibitor-NC, si-MIR4435-2HG + inhibitor-miR-582-5p, si-MIR4435-2HG + overexpressed (oe)-NC, si-MIR4435-2HG + oe-GPX8. At 48 h post transfection, cells were amassed .
Cell counting kit (CCK)-8 assay
Using the CCK-8 kit (CK04; Dojindo, Kumamoto, Japan), cell proliferation was tested. Cells (4000 cells/well) were added with CCK-8 reagent (200 µL) at 0, 24, 48, 72, and 96 h, respectively, and incubated to detect absorbance on a microplate reader at 450 nm .
Colony formation assay
Cells were cultured in 10% FBS-DMEM and cultured for 2 w at 1 × 103 cells/well. Then, cell colonies that were fixed in 10% formaldehyde (252549; Sigma-Aldrich, CA, USA) were stained with 0.1% crystal violet (C0775; Sigma-Aldrich) and counted by a microscope (Olympus, Tokyo, Japan) .
Cells with 80-90% confluence were wounded by a 10 µL pipette tip. At 0 and 48 h, the linear wounds were observed under the microscope and quantified by Image J software .
A 24-well Transwell chamber (354480; Corning, USA) with a 8-µm PET membrane was prepared, of which the lower chamber was loaded with medium containing 10% FBS (800 µL) and the upper chamber coated with Matrigel was covered with 10 × 104 cells. After 24 h, invaded cells were stained with crystal violet and microscopically counted in 5 fields of view .
Cells were rinsed with phosphate-buffered saline (P4417; Sigma-Aldrich) and resuspended in binding buffer at 1 × 106 cells/mL. After staining with Annexin V-fluorescein isothiocyanate and propidium iodide, apoptosis rate was detected by a FACS Calibur flow cytometer (BD Biosciences, NJ, USA) .
Detection of glycolysis
Cells (1 × 105 cells/well) were cultured in 6-well plates overnight and incubated under normoxia for 48 h. Glucose consumption was measured using glucose measurement reagent (G3293; Supelco, USA) and lactate production was determined using lactic acid measurement kit (MAK064; Sigma-Aldrich) .
Detection of lactate dehydrogenase (LDH)
NK cell cytotoxicity was analyzed by the LDH kit (C0016; Beyotime, Shanghai, China). Cells (1 × 104) were allowed to grow in 96-well plates for 4 h and the release of LDH was measured .
Tumor xenografts in nude mice
SPF-free BALB/c male nude mice (3-5 weeks old, 20 ± 5g) were reared at 25 ± 3°C with a humidity of 45-50%. The mice were randomly divided into 10 groups, each with 6 mice. Cells were resuspended in 50% Matrigel (354234; Corning, NY, USA) to 1 × 107 cells/mL, and the cell suspension (0.5 mL) was subcutaneously into the left armpit of mice which had been anesthetized with pentobarbital sodium (50 mg/kg). Tumor size was measured with a vernier caliper every 5 days, a total of 6 times. After 30 d, the mice were anesthetized with pentobarbital sodium (50 mg/kg) and euthanized to take tumors. Tumors were weighed and tumor volume (mm3) was measured: (length × width2)/2 .
Dual luciferase reporter gene assay
MIR4435-2HG wild-type (MIR4435-2HG-WT), MIR4435-2HG mutant (MIR4435-2HG-MUT), GPX8 wild-type (GPX8-WT) and GPX8 mutant (GPX8-MUT) fragments containing miR-582-5p binding sites were generated by GenePharma. Renilla luciferase vector pRL-TK (Promega, WI, USA), sequence fragment, and miR-582-5p mimic or miR-582-5p NC were co-transfected into PANC-1 cells for 48 h. The luciferase activity was detected by the dual luciferase system (Promega) .
RNA immunoprecipitation (RIP) assay
MIR4435-2HG and miR-582-5p, or miR-582-5p and GPX8 were co-transfected into PANC-1 cells, which were subsequently lysed by radio-immunoprecipitation assay lysis buffer (89901; Thermo Fisher Scientific, MA, USA) and combined with protein G agar magnetic beads containing anti-Ago2 (1:50; ab186733; Abcam, USA) or anti-IgG (1:100; sc-2025; Santa Cruz Biotechnology, Shanghai, China). The immunoprecipitates were treated with DNase I (18047019; Invitrogen) and proteinase K (4333793; Invitrogen), and RNA was recovered for reverse transcription quantitative polymerase chain reaction (RT-qPCR) .
RNA extraction was performed with Trizol reagent (15596018; Invitrogen). For MIR4435-2HG, GPX8 and MICA/B, reverse transcription was conducted using PrimeScript RT-PCR kit (RR014B; Takara, Kyoto, Japan), while for miR-582-5p, cDNA was assessed after treatment using TaqManTM Advanced miRNA kit (A28007; Applied Biosystems, CA, USA). Quantitative PCR was implemented with miScript SYBR Green PCR kit (218076; Qiagen, Germany) in the ABI 7500 system. Data assessment was conducted with 2−ΔΔCq method, employing U6 and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) as internal references. The primer sequences are shown in Table 1.
Western blot assay
Tissues or cells were lysed by 1% protease and phosphatase inhibitor, and protein concentration was detected with PierceTM BCA kit (23225; Thermo Fisher Scientific). The separated protein by sodium dodecyl sulphate polyacrylamide gel electrophoresis was transferred to a polyvinylidene fluoride membrane (FFP24; Beyotime), blocked with 5% skim milk, incubated with primary antibodies GAPDH (1:500; ab8245), GPX8 (1:500; ab183664), MICA/B (1:1000; ab222098, all from Abcam) and with the secondary antibody IgG (1:1000; ab222098). Finally, the bands were observed by enhanced chemiluminescence reagent (GE Healthcare, USA) .
SPSS 22.0 software (IBM, NY, USA) was applied for statistical analysis. Each experiment was repeated 3 times. Data were displayed as mean ± standard deviation and assessed by t-test or one-way analysis of variance (ANOVA) and Tukey's post-hoc test. Chi-square test and Kaplan-Meier analysis were adopted to determine the relationship between MIR4435-2HG expression and patients’ pathological features and survival, respectively. Pearson test was allowed to evaluate the correlation of MIR4435-2HG, miR-582-5p and GPX8. P < 0.05 presented statistical significance.