Clinical samples

A total of 88 paired tumor and nontumor tissues were collected from well informed HCC patients in the Eastern Hepatobiliary Surgery Hospital (Naval Medical University, Shanghai, China). Fetal liver tissues were collected from well informed patients who receipt pregnancy termination in the Changzheng Hospital (Naval Medical University, Shanghai, China). Ethical consent was granted from the Committee on Ethics of Biomedicine, Naval Medical University.

LncRNA microarray analysis

Biological triplicates of both fetal liver and adult liver tissues were used for lncRNA profiling by microarray analysis. In brief, total RNA was extracted, labeled cRNA was prepared, and RNA hybridization was conducted using Human LncRNA 3.0 (8 × 60k, Arraystar, USA) by Aksomics (Shanghai, China). Array images were acquired by Agilent Feature Extraction software and further analyzed using GeneSpring GX v11.5.1 software package (Agilent Technologies, USA).

Cell culture

Human hepatoma cell line Huh7 (TCHu182) and HepG2 (TCHu 72) and human embryonic kidney cell line 293T (SCSP-502) were purchased from National Collection of Authenticated Cell Cultures, Chinese Academy of Sciences. All the cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum (FBS) and were maintained in a humidified atmosphere of 5% CO2 at 37 °C.

RNA extraction, reverse transcription and quantitative polymerase chain reaction (qPCR)

Total RNA was extracted from clinical samples and cells by TRIzol (Thermo, USA). The first-strand cDNA was generated by PrimeScript™ RT Master Mix (Takara Bio, China). TB Green® Premix Ex Taq™ (Takara Bio, China) was used to perform qPCR reactions according to the manufacturer’s instructions. β-actin was used as endogenous control to normalize the total RNA in each sample. miRNA qPCR was conducted using miRCURY LNA SYBR® Green PCR kit (Qiagen, Germany), and miRNA levels were normalized to U6 levels. The results were calculated using 2−△△CT method. The sequences of all the primers used were listed in Additional file 2: Table S1.

Lentivirus infection and transient cell transfection

MIR4435-2HG overexpression and control lentiviruses were purchased from GenePharma (Shanghai, China). To obtain MIR4435-2HG overexpression cell lines, cells were cultured in medium added with 5 μg/mL Puromycin after infection. Cells were collected for further experiments at 2 weeks post infection.

All the small interfering RNAs (siRNAs) used in this study were purchased from GenePharma (Shanghai, China). Lipofectamine 3000 (Thermo, USA) was used to transfect siRNA according to the manufacturer’s instruction. The sequences of all the siRNAs used were listed in Additional file 2: Table S1.

Western blot

Protein samples were extracted by RIPA (Beyotime, China), separated by SDS-PAGE and transferred to NC membranes (Millipore, Germany). After blocked with 5% nonfat milk, the membranes were incubated with primary antibodies against METTL3 (64 kDa, 1:1000, ab195352, Abcam, USA), NOP58 (60 kDa, 1:1000, NBP1-46846, Novus, USA), IGF2BP1 (63 kDa, 1:1000, ab184305, Abcam, USA) or GAPDH (36 kDa, 1:5000, 60004-1-Ig, Proteintech, USA) at 4 °C overnight. After washing for three times, the membranes were incubated with IRdye 700-conjugated goat anti-mouse IgG and IRdye 800-conjugated goat anti-rabbit IgG for 40 min at room temperature. The signals were detected by Odyssey infrared scanner (Li-Cor Biosciences, USA). GAPDH was applied as endogenous control. Raw images from gels or western blots were provided in Additional file 3.

Immunohistochemistry staining

Immunohistochemistry staining was performed as previously described with a specific anti-NOP58 (1:200, NBP1-46846, Novus) antibody, anti-Ki67 antibody (1:200, 28074-1-AP, Proteintech), or an anti-PCNA (1:200, 10205-2-AP, Proteintech) [14]. Quantification of NOP58 was performed by blindly by pathologist.

Cell counting kit-8 (CCK-8) assay

3 × 103 cells were seeded into each well of 96-well plates with 100 μL normal culture medium. The experiments were performed using CCK-8 reagent (Dojindo Laboratories, Japan) following the procedures of manufacturer. The value of OD450 was normalized to plot the cell growth curve. All of the experiments were performed in triplicate.

Colony formation and spheroids formation assay

3 × 103 cells were seeded into each well of 6-well plates with 2 mL normal culture medium and incubated for 10 days. Then fix the cells with 4% polyformaldehyde for 30 min and stain the cells with crystal violet (Beyotime) for 30 min. The colonies in the dishes were photographed and counted.

3 × 103 cells were seeded into each well of 6-well ultra-low attachment culture dishes with 2 mL normal culture medium and incubated for 10 days. Spheroid formation was photographed and assessed by visual inspection under microscope.

EdU cell proliferation assay

The assay was conducted with EdU kit (RiboBio Co. Ltd, China) according to the instruction. Briefly, incubate the cells with 50 μM EdU A medium at culture environment and fix with 4% polyformaldehyde for 30 min. Then, the cells were stained with 1 × Apollo® for 30 min and the nuclei of cells were stained by DAPI. The total cells (identified by DAPI staining) and the EdU-positive cells (identified by Apollo® fluorescence) were photographed and counted under Olympus BX51 Fluorescence photomicroscope (Olympus, Japan). Each experiment was performed in triplicate.

Transwell migration and invasion assays

Transwell migration and invasion assays were performed with Millicell hanging Biocoat Matrigel and control chambers from BD Biosciences (24-wellinsert, 8-lm pore size) as previously described [15]. Briefly, 4 × 104 cells in 200 μL serum free medium were loaded into the upper chambers, while lower chambers were filled with 500 μL custom medium. In migration assays, the cells on the underside of the membrane were stained 24 h later. In invasion assays, the cells on the underside of the membrane were stained 48 h later. Then, the cells were counted under a microscope.

Wound healing assay

Wound healing assays were performed as previously described [16]. Briefly, cells were plated in 6-well plates and incubated at 37 °C. With the cells were complete attached, we scraped the middle of the plate to form a wound and replaced the medium with serum-free medium. After 48 h, the coverage of the line was measured.

Xenograft tumor model

Five-week-old male BALB/c athymic nude mice were purchased and maintained as previously described [17]. Ethical consent was granted from the Committee on Ethics of Biomedicine, Naval Medical University. Each mouse was injected with 5 × 106 MIR4435-2HG overexpressed Huh7 cells on the left armpit and control on the right armpit (six mice per group). The growth of subcutaneous tumor was recorded each week and the tumor volume was calculated as MIN(a)2 × MAX(b) × 0.5. After 6 weeks of observation, the mice were sacrificed and the tumors were obtained, measured, photographed and stored at − 80 °C.

In vitro limiting dilution assay

The experiments were conducted as previously described with slightly modification [17]. A total of 40, 80, 100, 400, 600, 800 HCC cells with MIR4435-2HG overexpression or control were seeded into 96-well ultra-low attachment culture dishes separately and incubated for 10 days (n = 15). Spheroid formation was observed under microscope. The proportion of spheroid-initiation cells was calculated by L-Calc software program (Stem Cell Technologies, Canada) and Poisson’s distribution statistics.

In vivo limiting dilution assay

A total of 1 × 103, 1 × 104, 1 × 105 and 1 × 106 HCC cells with MIR4435-2HG overexpression or control were injected subcutaneously into male nude mice (i = 5). After 8 weeks’ observation, the frequency of tumor-initiating cell was determined by L-Calc software program (Stem Cell Technologies) and Poisson’s distribution statistics.

RNA pulldown assay

RNA pulldown assay was performed as previously described [13]. Briefly, the biotin labeled MIR4435-2HG, truncated MIR4435-2HG and its antisense RNA were in vitro transcribed from vector pSPT19-MIR4435-2HG by T7 RNA polymerase (Roche, Switzerland) in vitro. Then, incubate the biotin-labeled transcript with Huh7 cell extracts and isolate the RNA–protein complex with Dynabeads Myone Streptavidin T1 beads (Invitrogen, USA). The proteins pulled down were separated by SDS-PAGE, detected by mass spectrum and verified by Western blot.

Co-Immunoprecipitation (co-IP) assay

RNA pulldown assay was performed as previously described [15]. Briefly, co-IP was performed in Huh7 cells using Pierce™ Co-IP Kit (ThermoFisher Scientific). Immunoprecipitations of NOP58 (NBP1-46846, Novus) or IGF2BP1 (ab184305, Abcam) were performed using an anti-NOP58 or IGF2BP1 antibody overnight at 4 ℃.

RNA immunoprecipitation (RIP) analysis and methylated RNA immunoprecipitation (MeRIP) analysis

The EZ-Magna RNA-Binding Protein Immunoprecipitation Kit (Millipore) was used to perform RIP and MeRIP assays according to the manufacturer’s instructions. The enriched RNAs were detected by qPCR. Primary antibody against NOP58 (NBP1-46846, Novus), IGF2BP1 (63 ab184305, Abcam) or m6A (A-1801, Epigentek, USA) was used in this study. The sequences of primers were separately listed in Additional file 2: Table S1. These experiments were performed in triplicate.

Analysis of NOP58 protein stability by CHX and MG132 treatment

Different groups of cells were cultured in 6-well plate and treated with CHX at a concentration of 25 μg/mL. Cell extracts were obtained at specific time points and were further detected by Western blot analysis.

Reverse transcription at low dNTP concentrations followed by PCR (RTL-P)

The RTL-P approach was conducted as previously described [18]. The key principle is that the 2′-O-methylated nucleotide could impede reverse transcription with low dNTP concentration. The total RNA was extracted and reversely transcribed with specific primers upstream to a methylation site in either a low level (1 μM) or a high level (1 mM) of dNTP (Thermo). The cDNAs were detected by qPCR and the methylation ratios was calculated following the function 2(CTlow−CThigh). The sequences of primers were listed in Additional file 2: Table S1.

OP-Puro (OPP) incorporation assay

The protein expression rates were detected by Click-iT plus OPP protein synthesis assay kit (Thermo). For initiation, Click-iT OPP was added to culture medium to a final concentration of 20 μM for 30 min. Then, wash the cells with PBS and fix with 4% formaldehyde in PBS for 15 min. After fixation, incubate the cells with 0.5% Triton X-100 in PBS for 15 min. Prepare Click-iT plus OPP reaction cocktail and incubate the cells for 30 min at room temperature. After removing the rinse buffer, stain DNA with DAPI for 30 min. Finally, wash cells twice with PBS, remove the wash solution and proceed to imaging.

Dual-luciferase assay

The pmirGLO Dual-Luciferase Vector was purchased from GenePharma (Shanghai, China). The wide type 1–1000 bp of MIR4435-2HG and the corresponding mutated were subcloned into the Vector. The pcDNA3-RLuc-PolIRES-FLuc plasmid was kindly provided by Nahum Soneberg [19]. The IRES sequences of insulin like growth factor 1 receptor (IGF1R) and MYC were cloned from Huh7 cells. Dual-luciferase assays were performed following the procedures of manufacturer. The above experiments were conducted in triplicate.

Statistical analysis

All statistical analyses were performed with SPSS 16.0, Graphpad Prism 8.0 and L-Calc™. Two-tailed Student t test, one-way analysis of variance, and chi-square test were performed for statistical comparison when appropriate. Spearman’s or Pearson’s correlation coefficient was used for statistical correlation. Survival curves were assessed by Kaplan–Meier’s method and log-rank test. P values were two-side and a P value < 0.05 was considered to be statistically significant. All data were shown as mean ± standard deviation (SD).