Patients and tissue species

Twenty-five people diagnosed with PDAC at Changhai Hospital provided tumor tissues and adjacent non-cancerous pancreatic tissues for analysis. The participants in the study were granted written informed consent. The features of the patients were recorded in Supplementary Table 1. The diagnosis of PDAC in all patients was conclusively confirmed through surgical procedures and pathological examinations. Before the surgical procedure, none of the patients had received radiotherapy or chemotherapy treatments. Upon receiving the tissue samples from patients, they were promptly frozen in liquid nitrogen and stored at -80°C.

Extraction of RNA and performing real-time PCR

RNA was extracted from tissue and cell cultures with TRIzol® reagent (Invitrogen, MA, USA). RNA concentrations and quality were assessed using UV spectrophotometry. RNAs were reverse transcribed using the HiScript® III 1st Strand cDNA Synthesis Kit from Vazyme in Nanjing, China, following the provided guidelines. The quantitative real-time PCR (qPCR) assay was conducted using the SYBR-green mastermix (Vazyme, Nanjing, China) and done on LightCycler 480II equipment (Roche, Mannheim, Germany).In both mRNA and lncRNA, endogenous control was utilized via b-actin. The qPCR temperature procedure includes starting with a denaturation step at 95°C for 5 minutes, then proceeding with 40 cycles of temperature fluctuation between 95°C and 65°C for 45 seconds each.cDNA synthesis and qPCR were used to assess the expression levels of microR-185-5p and microR-148b-3p, with the miScript RT kit (Qiagen GmbH) and miScript SYBR Green PCR kit (Qiagen GmbH) being employed, respectively. PCR primers can be found in Supplementary Table 4.

Cell culture and treatment

Cell lines from the American Type Culture Collection, including HPNE, PANC-1, AsPC-1, CaPAN-2, SW1990, BxPC-3, and MIA PaCa-2, were acquired for the study. Cells were maintained in Dulbecco's modified Eagle's medium (DMEM) with 10% fetal bovine serum (Invitrogen, Carlsbad, CA, USA) and 1% penicillin/streptomycin. Cells were cultured at 37 °C in the atmosphere with 5% carbon dioxide.

In the hypoxic culture experiment, the cells were subjected to an incubation environment consisting of 94% nitrogen (N2), 1% oxygen (O2), and 5% carbon dioxide (CO2). The treatment of 250μM cobalt chloride (CoCl2) for 24 hours induced hypoxia.

Subcellular RNA extraction was performed according to the manufacturer's instructions using a PARISTM Kit (Ambion). Quantitative reverse transcription polymerase chain reaction (qRT-PCR) was used to further analyze RNA in the cytoplasm and nucleus. β-actin and U6 served as controls for cytoplasmic and nuclear compartments, respectively.

Oligonucleotides and transfection

In Shanghai, China, GenePharma created vectors for overexpression, shRNA, and siRNA that target HIF-1α, MKLN1-AS, and TEAD1. A lentivirus vector carrying these sequences (GenePharma, Shanghai, China) was also constructed containing matching controls. MiR-185-5p mimics, miR-185-5p inhibitor, and their respective negative controls were supplied by GenePharma in Shanghai, China. PDAC cells were transfected with the oligonucleotides using Lipofectamine 3000 (Invitrogen, USA), followed by confirmation of transfection efficiency through qRT-PCR analysis.

Western blot analysis

The cell lysate was created by treating the cells with RIPA buffer from Epizyme in Shanghai, China, and the amount of protein was measured using the BCA Protein Quantification Kit. A 30 μg protein sample was subjected to electrophoretic separation on a 12% SDS-polyacrylamide gel. After electrophoresis, the proteins were moved to PVDF membranes. Following incubation with a 5% skim milk solution in TBS-T for four hours at room temperature, primary antibodies against TEAD1 (diluted 1:2000, purchased from Abcam, catalog number ab133533, UK) and HIF-1α (diluted 1:2000, acquired from Abcam, catalog number ab79546, UK) were left overnight at 4°C. To ensure equal loading of protein samples, an anti-β-actin antibody (diluted at 1:5000, sourced from Abcam, catalog number ab6275, United Kingdom) was employed. After washing the membranes three times with TBS-T for 15 minutes each, they were then incubated with secondary antibodies (diluted at 1:10,000; Westang) conjugated to horseradish peroxidase for 45 minutes at room temperature. Protein bands were visualized using a developing solution (Vazyme, Nanjing, China) and exposed to X-ray film.

Bioinformatic analysis

Data on clinical characteristics and gene expression in relation to PDAC were acquired from The Cancer Genome Atlas (TCGA) database and Gene Expression Omnibus (GEO) database, available at https://portal.gdc.cancer.gov/ and https://www.ncbi.nlm.nih.gov/geo/ respectively. All data included 171 PDAC patients from the TCGA database and 63 PDAC patients from the GEO database, and the quality controls were analyzed by R language. The details of data from the public database are exhibited in Supplementary Tables 2, and 3. Levels of gene expression were standardized and then divided into groups based on a predetermined threshold to distinguish between high and low expression. Correlation maps involving two genes were generated using the ggstatsplot software in R, while heatmaps displaying correlations among multiple genes were visualized using the R software package. Molecular interactions and relevant clinical data were analyzed using the R programming language.

Shared microRNAs (miRNAs) between MKLN1-AS and TEAD1 were identified using the miRcode and TargetScan databases, available at http://www.mircode.org/index.php and http://www.targetscan.org/vert_72/, respectively. The downstream miRNA and binding site sequences of MKLN1-AS were obtained from the JEFFERSON database, accessible at https://cm.jefferson.edu/rna22/Interactive/.

Cell counting kit-8 (CCK-8) assay

Cell counting kit-8 (CCK-8) assay was performed to count viable cells. A controlled climate incubation environment at 37°C for 2 hours was used for incubation of cells in 96-well microtiter plates. Cells were placed in 96-well microtiter plates at a concentration of 2×103 cells per well and kept at 37°C for 2 hours in a regulated incubation setting. After a 24-hour incubation period, every well was treated with 10 microliters of the Cell Counting Kit-8 (CCK-8, Dojindo, Japan). Following incubation, the plates were washed twice with phosphate-buffered saline (PBS). Afterward, absorbance at 450 nm was measured using a microplate reader following the guidelines provided by the manufacturer.

Colony formation assay

Pancreatic cancer cells were placed in 6-well dishes with a concentration of 1×103 cells per well. Colonies with a cell count exceeding 200 cells per colony were subjected to a 10-day incubation period, followed by fixation for 15 minutes using a 4% paraformaldehyde solution. Afterward, the colonies were treated with a 0.1% crystal violet solution for a duration of 15 minutes. The number of colonies in each well was quantified using Image J analysis software (version 1.48), and the mean colony area (mm2) was computed to facilitate comparisons of size. Three replicates of each experiment were conducted.

EdU assay

The experiment involving EdU (5-ethynyl-2'-deoxyuridine) proliferation was carried out following the guidelines provided by the manufacturer, utilizing the Cell-Light EdU Apollo 567 In Vitro Imaging Kit from Ribobio. Briefly, cells were processed according to the suggested procedures and then placed in 96-well plates at a density of around 5×103 cells per well. Each well received 100 microliters of medium containing 50 micromolar (μM) EdU, 24 hours after seeding. Afterward, the cells were placed in an incubator at a temperature of 37 degrees Celsius for a duration of 2 hours. After the cells were incubated, they were treated with a 4% paraformaldehyde solution and then stained with a mixture of Hoechst and Apollo reagents. Images were acquired using fluorescence microscopy and subsequently merged with Image J analysis software (version 1.48). Relative Edu incorporation was used for quantification analysis, followed by determining the total cell count and the count of EdU-positive cells in each field.

Cell migration/invasion assay

The test was performed in Boyden chambers with 24 wells, each holding 12 Millipore inserts. These inserts had an 8 mm pore size and were coated with a layer of basement membrane matrix on top, or left uncoated. Tests were performed in Boyden chambers using 24-well tissue culture plates, each equipped with 12 Millipore inserts. These inserts had a polycarbonate membrane with 8μm pores and were coated with either an ECMatrix layer (for invasion experiments) or left uncoated (for migration experiments). The bottom compartments were filled with a solution that included 10% fetal bovine serum, which acted as a chemoattractant.300 μL of serum-free medium contained 5 × 105 cells in the chambers above. Cells that infiltrated the ECMatrix or moved through the polycarbonate membrane were observed and documented under a microscope following 48 hours of incubation at 37°C.

Scratch wound healing assays

Pancreatic cancer cells were grown in 6-well dishes until they covered the surface, then manually scraped using a 20μl pipette tip. Next, a wound was induced by the addition of phosphate-buffered saline (PBS). A wound was created by manually scratching a cell monolayer with a 20-μl pipette tip, then rinsing with phosphate-buffered saline (PBS). Afterward, the cells were cultured for at least 12 hours in a medium containing 1% FBS, a level that does not impact the growth of pancreatic cancer cells. We established specific time points for capturing images within the cell growth cycle. Images of the wound healing process were captured using a phase contrast microscope. Image J analysis software (version 1.48) was utilized to measure the regions where cells migrated.

In vivo assay

Male BALB/C athymic nude mice, five weeks old, were maintained in an environment devoid of pathogens. The Animal Research Ethics Committee at Second Military Medical University approved the animal study in advance. To monitor tumor growth progression, we administered 0.1 ml of Hank's balanced salt solution into the flank of every mouse, utilizing 1 × 106 MIA PaCa-2 and SW1990 cells. Tumor volumes were measured weekly using the formula length × width2 × 0.5. Following a period of four weeks, the mice were put to death, and the tumors were removed.An extra set of mice received injections of MIA PaCa-2 and SW1990 cells (1× 106 each) via their ileocolic veins four weeks later for the metastasis test. All mice were euthanized on the 28th-day post-injection or upon reaching a pre-mortal state. Liver resection was performed, and the number of surface metastatic lesions on the livers was quantified.

Construction of MKLN1-AS promoter reporter plasmids

The MKLN1-AS segment, spanning from -891 to -4 in relation to the start of transcription, was inserted into a pGL3-basic vector (Promega), resulting in a 1000-bp DNA piece. This facilitated the generation of the reporter plasmid, designated as pGL3-891, which was purposefully engineered to encompass multiple hypoxia response elements (HREs). The nomenclature for the promoter-reporter plasmids was systematically based on the specific initiation site of the HREs. Subsequently, deletion mutation reporters were meticulously crafted for the plasmids, namely pGL3-681 and pGL3-30.To validate the integrity of these constructs, comprehensive sequencing was conducted on both the inserted sequences and the flanking regions of the plasmids.

RNA immunoprecipitation (RIP)

The RIP method was carried out using the Magna RIP™ RNA-Binding Protein Immunoprecipitation Kit from Millipore in Germany, following the instructions given, and abiding by the guidelines provided by the producer. Roughly 10 million SW1990 and MIA PaCa-2 cells were lysed using RIP lysis buffer. Following centrifugation at 10,000 times gravity and 4 degrees Celsius for 5 minutes, the PDAC cells were exposed to magnetic beads linked with human Ago2 antibodies obtained from Abcam in Cambridge, UK, or control IgG antibodies from Millipore in the USA. This overnight incubation occurred at 4°C. Subsequently, the samples were treated with Proteinase K while agitating to facilitate protein digestion. The immunoprecipitated RNAs were then extracted and subjected to qRT-PCR analysis.

Chromatin immunoprecipitation assay

The chromatin immunoprecipitation assay kit (Millipore) was used to prepare 2×106 PDAC cells as specified in the manufacturer's instructions for the chromatin immunoprecipitation (ChIP) assay. DNA samples were first treated to induce precipitation before undergoing PCR amplification aimed at specific regions of the MKLN1-AS promoter. The resulting PCR products were then separated by electrophoresis on a 2% agarose gel and visualized with ethidium bromide staining.

Promoter construction and Dual-luciferase reporter assay

The promoter sequence of MKLN1-AS was identified using the UCSC gene browser, accessible at https://genome.ucsc.edu/. To determine the binding site sequence responsible for interacting with a transcription factor and facilitating MKLN1-AS transcription, a search was conducted on the JASPAR database (https://jaspar.genereg.net). Chip-seq data for transcription factors were acquired from Cistrome Data Browser at http://cistrome.org/db.

Recombinant luciferase plasmids were created by inserting the target fragments of MKLN1-AS or TEAD1 3’-UTR sequence, with wild-type (WT) or mutant-type (MT) miR-185-5p binding sites, into the pGL3-Basic luciferase vector from Promega in the USA. Transfection of cells was performed with the wild-type (WT) or mutant (MUT) luciferase reporter plasmid, along with miR-185-5p mimics, inhibitors, or negative control (NC), using Lipofectamine 3000 (Life Technologies Corporation, Carlsbad, CA, USA). After incubating for 48 hours, the Promega Dual-Luciferase assay was employed to measure the firefly and Renilla enzyme activity. Creating vectors for overexpressing or silencing the HIF-1α transcription factor, as well as wild-type or mutant MKLN1-AS sequences, is essential for conducting the combination test. The procedure remains unchanged. The relative luciferase activities were determined by subtracting the Renilla luciferase activities from the firefly luciferase activities. The analysis was conducted in triplicate for each group.

Statistical analysis

The mean ± SME of the data obtained from three triplicated independent experiments was reported. Data analysis was performed with the assistance of GraphPad Prism software (Intuitive programmer for Science, San Diego, CA) and SPSS 26.0 (SPSS Inc., Chicago, IL, USA). The comparison between two normal distribution groups was conducted using the Two-tailed Student's t-test. The disparity between the two abnormal distribution groups was analyzed by nonparametric tests. For the analysis of numerous groups, a one-way ANOVA was undertaken. Survival curves were compared using log-rank tests and depicted using the Kaplan-Meier method. Genes were analyzed for correlation using Pearson's correlation coefficient. Statistical significance was determined by using P values below 0.05 in all conducted tests.