HCT116, SW480, LOVO, and HEK293T were provided as a kind gift from Professor Dajun Deng at Peking Union Medical College Hospital, and were cultured in DMEM (Corning, VA, USA) containing 10% fetal bovine serum (Gibco, Australia) and 1% penicillin/streptomycin (Gibco, NY, USA) in a 5% CO2 incubator at 37 °C.

The PCDH-CMV-EF1a-copGFP-T2A-Puro lentiviral vector was used to generate a P14AS expression construct by Syngentech Co., Ltd. (Beijing, China). The psiCHECK2 vector was obtained from Sangon Biotech Co., Ltd (Shanghai, China). Lipofectamine 3000 (Invitrogen, CA, USA) was used to transfect cells with appropriate miRNAs, siRNAs, or other constructs based on provided instructions. Colon cancer cells were transfected with miRNA mimics (final concentration 100nM) or miRNA inhibitors (final concentration 100nM) or siRNA (final concentration 100nM) for 48-72 h. Post-transfection, cells were collected for CCK8 assay or WB assay or qRT-PCR. Stably transfected HCT116, SW480, and LOVO cells were obtained through culture in media containing puromycin (1 µg/mL, InvivoGen, CA, USA).

A TransDetect Cell Counting Kit-8 (CCK-8, TransGen Biotech, Beijing, China) was used based on the provided directions to assess transfected cell viability. Briefly, 100 µL of media containing either 20,000 or 80,000 cells/mL (for proliferation and cytotoxicity assays, respectively) was added to each well of a 96-well plate. In cytotoxicity assays, after allowing cells to adhere to the plate, media was exchanged for media containing a range of BAY 11-7085 (Selleck, Shanghai, China) concentrations (0, 2, 4, 8, 12, 16, 32µM) or DMSO. At appropriate time points, CCK-8 solution (10 µL) was added per well, followed by an additional 3 h incubation at 37 °C. Absorbance was then measured at 450 and 630 nm with a BioRad microplate reader. Proliferation was assessed once daily on 5 sequential days, and average absorbance values were measured.

Trizol (TransGen Biotech, Beijing, China) was used to extract cellular RNA, after which a TransScript First-Strand cDNA Synthesis SuperMix (Roche, IN, USA) was used to prepare cDNA. Then, FastStart Universal SYBR Green Master (ROX) (TransGen Biotech, Beijing, China) was used to conduct qPCR analyses with an ABI-7500 Fast system (Applied Biosystems) with the following settings: 94 °C for 30 s; 40 cycles of 94 °C for 5 s, 60 °C for 15 s, 72 °C for 34 s. GAPDH was used as a normalization control. Primers used for this study are listed in Supplementary Table S1.

NP-40 buffer (Solarbio, Beijing, China) supplemented with protease inhibitors (LabLead, Beijing, China) was used to extract protein from cells, and these samples were then separated via SDS-PAGE and transferred onto PVDF membranes (Merck Millipore). Following a 1 h room temperature blocking step using 5% skim milk, the membranes were probed with appropriate primary antibodies overnight at 4 °C followed by probing for 1 h at room temperature with secondary anti-mouse or anti-rabbit IgG (ZSGB Biotech, Beijing, China). A chemiluminescence detection system (Cytiva, Amersham ImageQuant 800, Japan) was then used for protein detection. The primary antibodies used were specific for AUF1 (Abcam, ab61193, UK), CDCP1 (Abcam, ab252947), ADAM10 (Abcam, ab124695), UBE2D3 (Abcam, ab176568), IκBa (Abcam, ab32518), Ago2 (Abcam, ab186733), YY1 (Santa Cruz, sc-7341, TX, USA), HNF3a (Santa Cruz, sc-514,695), GAPDH (Protein Tech, 50430-2-AP; China), β-tubulin (Abcam, ab6046). Antibodies were used at dilutions ranging from 1:1000-1:5000.

Biotinylated P14AS probes (#1-#6) and corresponding control probes (#1-#2) that had been synthesized in vitro by RiboBio (Guangzhou, China) were incubated with SW480 cell lysates, after which an RNA pull-down assay was performed based on protocols published previously(Ma et al. 2020).

An RNA-Binding Protein Immunoprecipitation Kit (Cat# 17–701, EZ Magna, Millipore, USA) was used based on the directions provided. Total RNA bound to AUF1 or Ago2 was precipitated and isolated with anti-AUF1 (Abcam, ab61193, Cambridge, UK) or anti-Ago2 (Abcam, ab186733).

HindIII was used to insert the full-length P14AS sequence containing wild-type or mutated binding sites for target miRNAs (miR-23a-5p, miR-6855-5p, and miR-6842-5p) into the pGL3-control vector (Promega, WI, USA). The 3’-UTR sequences for target genes or mutant isoforms were inserted into the psiCHECK2 vector using XhoIand NotI(Sangon Biotech, Shanghai, China). For further details regarding the binding sites used to generate these plasmids, see Fig. 3E. After adhering overnight to 12-well plates, cells were co-transfected with WT or mutated luciferase reporter vectors together with miRNA mimics (final concentration 100nM), inhibitors (final concentration 100nM), or control vectors and incubated for 48 h. A dual luciferase reporter assay kit (Promega, WI, USA) was used based on the directions provided.

Supernatants were collected from cells that had been stably transfected, and the concentrations IL-6 (EH004-96) and IL-8 (EH005-96) therein were measured using appropriate ELISA kits (ExCell, Shanghai, China) based on the manufacturer’s instructions.

A ChIP kit (Beyotime Biotech, Beijing, China) was used based on the provided directions, after which the fragments of DNA precipitated using YY1 (Santa Cruz, sc-7341, TX, USA), HNF3a (Santa Cruz, sc-514,695), and control IgG were analyzed via qPCR.

The transcriptomes of the stably P14AS-overexpressed HCT116 cells were sequenced by RiboBio Co., Ltd. (Guangzhou, China). The data sets were deposited in the Gene Expression Omnibus (GEO) database with accession number GSE127905. KEGG pathway analysis for differentially expression genes was performed using KOBAS3.0 software (http://www.genome.jp/kegg). P14AS-binding miRNAs and miRNA-target proteins were analyzed using the miRanda, Pita, RNAhybrid, and TarPmiR databases.

Data were compared using two-sample Student’s t-tests and presented using GraphPad Prism 7.0 software (Dotmaticus, USA). Results are reported as means ± SD. Pearson correlation analyses were used to assess relationships between variables. Two-sided statistical tests were used for all analyses. *P < 0.05, **P < 0.01, N.S: not significant.