Tissue samples
Total 125 pairs of HCC tissue samples and adjacent non-tumor tissue samples, which were histopathologically confirmed, were collected from HCC patients stood hepatectomy in the First Affiliated Hospital of Xi’an Jiaotong University. Neither chemotherapy nor radiotherapy was administered before surgery to any of the patients. A temperature of -80°C was used to store all samples. Our study got approval from the Ethics Committees of the First Affiliated Hospital of Xi’an Jiaotong University, all patients provided informed consent.
Cell culture
The human normal liver cell line (MIHA), six cell lines (HepG2, Huh7, Hep3B, MHCC97H and SK-Hep-1) and human embryonic kidney (HEK) 293T cells were purchased from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China). All of the cells were maintained in incubator (37℃, 5% CO2), and cultured in DMEM (Gibco; Thermo Fisher Scientific, Inc., Waltham, MA, USA) supplemented with 10% FBS (Gibco, Grand Island, NY, USA) and 1% penicillin-streptomycin (Invitrogen, CA, USA). For hypoxia treatment, physical hypoxic condition (1%O2) was generated by Forma Series II 3130 incubator (Thermo Scientific).
Plasmids and cell transfection
The small hairpin RNAs (shRNAs) targeting SZT2-AS1 (shSZT2-AS1#1, shSZT2-AS1#2) were obtained from GeneCreate Biological Engineering Co., Ltd. (Wuhan, China). HIF-1α, HIF-2α shRNAs and scrambled shRNA (shNTC) were purchased from GeneCopoeia (Guangzhou, China). The human HIF-1α or SMYD2 ORF cDNA clone (ov-HIF-1α, ov-SMYD2), and control empty vector were purchased from GeneCopoeia, Inc. All sequences were verified by DNA Sanger sequencing. Lentiviral production was achieved by transfecting target plasmids with psPAX2 packaging plasmid and pMD2.G envelope plasmids into HEK293T cells. Viral supernatant added to HCC cells with 8 µg/ml polybrene (Beyotime Biotech Inc., Shanghai, China), The transfected cells were treated with 3 µg/ml puromycin and 100 µg/ml Ampicillin to select knockdown and overexpressed single clone cells. Cell transfections were performed by using Lipofectamine 3000 reagent (Invitrogen, CA, USA) according to the manufacturer’s instructions.
Quantitative real-time PCR (RT-qPCR)
By using Trizol (Thermo Fisher Scientific) after completing the designated intervention, the RNA from HCC cells and tissues was extracted. A reverse transcription kit was used to reverse-transcribe total RNA into cDNA (Invitrogen, CA, USA). Real-time qPCR analysis was performed using SYBR Green Premix PCR Master Mix (Roche Diagnostics, Mannheim, Germany). Normalizing the relative expression level to 18S and the expression was calculated by 2−ΔΔCt methods. The sequences of the primers used are listed in Supplementary Table 1.
Western blot and immunoprecipitation (IP) assays
RIPA buffer was used to isolate total protein from cells which supplemented with proteinase inhibitors and phosphatase inhibitors (Beyotime, Hangzhou, China). According to the manufacturer's instructions, BCA Protein assay kits (ZHHC Biotech Inc., Shaanxi, China) were used to determine protein concentration. Protein was separated by 10% or 8% or 15% concentration SDS-PAGE gels, then transferred to 0.22µm PVDF membranes (Millipore, Billerica, MA, USA). Following 3 hours of blocking by 10% nonfat milk, primary antibodies were used to incubate membranes at 4℃ temperature overnight. Afterwards, secondary antibodies conjugated to HRP were incubated on the membranes at room temperature for 1 hour. The blots were detected using enhanced chemiluminescence reagent (Millipore, Billerica, MA, USA). For immunoprecipitation, equal amounts of WCLs (500 µg) were incubated with primary antibody HIF-1α (2 µg) in the presence of protein G-Sepharose beads (Amersham Biosciences) at 4°C overnight, and the immunoprecipitates were subjected to SDS-PAGE and immunoblot assays.
Transwell migration and invasion assays
As we previously reported[19], after completing the specified experimental processing, cell migration and invasion ability were investigated using Transwell migration and invasion assays following established protocols from prior research.
Wound healing assay
Wound healing assay was used to detect cell migration ability. After transfected HCC cells with different plasmid or virus, the experiment was conducted according to the protocols as we previously reported[24].
MTT assay and EdU assay
Like MTT assay, specified transfected cells were added into 96-well plates at a density of 1 × 104 cells/ well. Then with 0, 24, 48, and 72h after seeding, 20µL MTT solution (Sigma, USA) was added to each well and continued incubate for 4h at 37℃. After removed the supernatants, added 100 µl DMSO to each well. Each absorbance was measured at 490 nm by a microplate reader (Bio-Rad, Richmond, CA). For EdU assay, Cell-Light™ EdU Apollo®567 (RiboBio Co., Ltd. Guangzhou, China) was used to evaluate of cell proliferation. Briefly, transfected HCC cells were cultured in 96-well plates. Then complete the experiment according to the protocol supply by manufacturer. The percentage of EdU positive cells was calculated using ImageJ software. EdU positive rate is calculated by counting at least five random fields.
Tube formation assay
The different treatment HCC cell culture medium was changed to serum-free DMEM medium for 48 h and then was collected, centrifuged and filtered to obtain tumor-conditioned medium (TCM). Prepare the dissolved Matrigel and precool the 24-well plate and pipets at -20℃. The precool 24-well plate was laid with 200µl Matrigel matrix (Corning Inc., Corning, NY, USA) incubated 37℃ for 30min. HUVEC (5×104) cells were added to each well with 200µl TCM which came from HCC cells and supplemented with 10% FPS, and then incubated at 37°C in 5% CO2 for 8h. Pictures were taken under a bright-field microscope and the capillary tubes were quantified by counting branch number and total tube length with Image J.
Dual-luciferase reporter assay
The SZT2-AS1 promoter region sequence containing wild-type (WT) or mutated (MUT) sequences of hypoxia response elements (HREs) were embeded into pGL3-based vectors (Promega, USA). Luciferase reporter plasmid pGL3-based vectors expressing SZT2-AS1-WT or MUT were co-transfected into Hep3B and MHCC97H cells in 96-well plates with empty vector or pcDNA3.1/HIF-1α. After 24h, cells lysates were collected, the dual-luciferase reporter system kit (Beyotime, Shanghai, China) were used to measure Renilla and firefly luciferase activities on a microplate reader according to the manufacturer’s advices and protocols. Normalized luciferase activity according to The Renilla luciferase internal control.
Bimolecular fluorescence complementation assay
The pCMV-NRluc and pCMV-CRluc plasmids were constructed. The human bHLH-PAS domain of HIF-1α (amino acid residues 12–396) was prepared by PCR amplification and the PCR product was cloned downstream of the N-terminal segment (residues 1-229) of Rluc, in pCMV-NRluc-HIF-1α12−396. Similarly, the bHLH-PAS domain of HIF-1β (residues 11–510) was amplified and inserted upstream of the C-terminal segment (residues 230–311) of Rluc, in pCMV-HIF-1β11−510-CRluc. MHCC97H and Hep3B cells were seeded at 2x105 cells per well of a 24-well plate and incubated for 24h. Cells were co-transfected with 300 ng of NRluc-HIF-1α12−396, 300 ng of HIF-1β11−510-CRluc, and 80 ng of pGL2-promoter, using Fugene-6 (Roche) according to the manufacturer’s instructions. Following 7h incubation, cells were treated with indicated shNTC or shZST2-AS1 for 24 h. Cells were then lysed and analyzed for the ratio of Rluc/Fluc using the Dual Luciferase Assay System (Promega).
RNA-seq analysis
Hep3B cells were seeded into six-well plates for 24h, and then incubated in 20%O2 or 1%O2 for another 24h. Total RNA was isolated from the cells using TRIzol (Invitrogen) and treated with deoxyribonuclease (Qiagen). Library preparation and sequencing using the NovaSeq 5000 platform (Illumina) were performed. The FASTQ files were subjected to quality check and analyzed by Genialis Inc. (https://www.genialis.com). Differential expression results with a false discovery rate of < 0.05 and mRNA fold change of > 1.5 were used as acut off for further downstream analysis.
RNA immunoprecipitation (RIP)
RIP assay was conducted by utilizing the EZ-Magna-RIP kit (MilliporeSigma, Birlington, MA) according to the the manufacturer's instructions. Cells were lysed by lysis buffer which added with protease inhibitor cocktail (Thermo Fisher Scientific, Waltham, MA) and RNase inhibitors (Millipore Sigma). Then, the cell lysate was pre-washed with recombinant protein A/G agarose (Thermo Fisher Scientific) to reduce non-specific binding for 30 min at 4°C. One to twenty percent of the cell lysate were used as input. Then the specific antibody (anti-HIF-1α, anti-HIF-1β, anti-HIF-2α or IgG) and protein A/G magnetic beads were placed in equal amount of cell lysates for overnight at 4℃, negative control is used IgG. At the next day, RNA was eluted from the precipitated complex by using Trizol and transcribed into cDNA. RT-qPCR assay was performed to detect binding of RNA (SZT2-AS1) to proteins or antibody. The primary antibodies were listed in Supplementary Table 2.
RNA pull-down assay
The interaction between SZT2-AS1 and SMYD2 or HIF-1α or HIF-1β was predicted with RNA-Protein Interaction Prediction (RPISeq). Biotin labelled Sense and antisense of SZT2-AS1 RNA were in vitro transcribed with AmpliScribe T7-Flash Biotin-RNA Transcription Kit (Epicentre), treated with RNase-free DNase I and purified with a RNeasy Mini Kit (Qiagen). To establish the appropriate secondary structure, biotinylated SZT2-AS1 RNA supplied with RNA structure buffer (10mM Tris pH7, 0.1M KCl and 10 mM MgCl2) was first heated up to 90°C for 2 min, then incubated on ice for 2 min and last transferred to room temperature (RT) for 20 min. The RNA was then mixed with hypoxic Hep3B and MHCC97H cells extract or purified proteins and incubated at RT for 1 h, followed by incubating with Streptavidin Mag Sepharose (GE Healthcare) at RT for 1 h. After follow-up wash, extract the pull-down complexes were analysed by standard western blot technique.
RNA Subcellular fraction
To determine the cellular localization of SZT2-AS1, cytoplasmic and nuclear fractions were isolated and collected with the PARIS Kit (Life Technologies, Inc., Carlsbad, CA, USA) according to the manufacturer’s instructions. Thereafter, collections from HCC cells both cytoplasm and nucleus were extracted out total RNA and cDNA was synthesized for the evaluation of SZT2-AS1. Briefly, we collected 1×107 cells and washed in PBS three times, and then 300µl Cell Fractionation Buffer resuspended cells and incubated at 4°C for 10 min. After 12000rpm centrifugation, aspiration of supernatant containing cytoplasmic components and 300µl Cell Disruption Buffer were resuspended collection of centrifugal precipitation containing nuclear fragments. The manufacturer’s instructions were used to extract RNA from the buffer containing cytoplasmic/nuclear fraction, following RT-PCR analysis of the levels of nuclear control transcript (U6), cytoplasmic control transcript (GAPDH) and SZT2-AS1.
Protein isolation and analysis
Nuclear and Cytoplasmic Protein Extraction Kit (Beyotime Biotechnology, Shanghai, China) was using to protein isolation according to the manufacturer’s instructions. Briefly, cells that have been treated in advance were harvested and dissociated in 200µl Reagent mixture containing 1 mM PMSF (Key Gen BioTech, Nanjing, China). The dissociated cell was incubated at 4°C for 15 min. Then, adding 10µl Reagent B and vortex the mixture for 5 s with 1 min ice bath, after that centrifugation at 16000 g, for 5 min. The supernatant we collected was cytoplasmic protein and 50µl nuclear protein extraction reagent containing 1 mM PMSF were further resuspend the precipitation. After being vortexed and 4°C in turn for 30 min, the mixture was centrifugated 5 min at 16000 g for 4°C, and the supernatant was stored as nuclear protein. The subcellular fractions were determined by using BCA protein assay kit, and then subjected to immunoblotting.
Purification of GST-HIF-1α and In vitro RNA-binding assay
Luria–Bertani medium with ampicillin (50µg/ml) used to culture Escherichia coli host BL21(DE3) harbouring the expression vector pGEX-6p-1-HIF-1a, and induced by 0.3 mM IPTG at 30°C for 16 h. Affinity purification of Recombinant protein by Pierce Glutathione Super Flow Agarose (Pierce) following the manufacturer’s instructions. SZT2-AS1 RNA was synthesized in vitro and added to RNA structural buffer (10 mM Tris pH 7, 0.1 M KCl and 10 mM MgCl2), mixture first heated to 90°C for 2 min, follow incubated on ice for 2 min and then transfer to RT for 20 min to form the proper secondary RNA structure. GST fusion proteins on 20µl glutathione Sepharose beads were incubated with 2 mg SZT2-AS1 RNA synthesized in vitro which included in 50µl of RNA-binding buffer (0.1% NP-40, 100 mM KCl, 2 mM MgCl2, 50 mM Tris-HCl, pH 7.4, 1 mM dithiothreitol and ribonuclease inhibitor) for 30 min at 4°C. Follow up, the glutathione Sepharose beads were washed with RNA-binding buffer three times to remove non-attached RNAs. Trizol reagent were used to extract the RNA samples retained on the beads and detected expression by RT–qPCR. Calculate the relative retention value of the input RNA level.
Chromatin immunoprecipitation (ChIP)
MHCC97H and Hep3B cells were incubated at 20 or 1% O2 for 16 hours and harvested for ChIP assay, cells cross-linked in 3.7% formaldehyde for 15 min, neutralized in 0.125M glycine for 5 min, and then used SDS lysis buffer to lysed HCC cells. Chromatin was sheared by sonication to an average length of 200–1000 bp, and salmon sperm DNA/protein A agarose slurry (Millipore) precleared cells lysates for 1 hour, protein A–agarose beads incubated with antibody against HIF-1α, HIF-2α, SMYD2, HIF-1β, H3K4me3, H3K36me3, H3 overnight at 4°C. Follow by serial washing of the agarose beads with low-salt, high-salt, and LiCl buffers, 1% SDS with 0.1 M NaHCO3 elute DNA from beads, and reverse cross-links by addition of 0.2M NaCl. Purification of DNA by phenol-chloroform extraction and ethanol precipitation and amplified by RT-qPCR using primers listed in Supplementary Table 1. And antibodies were listed in Supplementary Table 2.
RNA fluorescent in situ hybridization (FISH)
FISH kit (RiboBio, Guangzhou, China) was used to detected the Subcellular localization of SZT2-AS1 according to the manufacturer’s procedure. In brief, MHCC97H and Hep3B cells were cultured on glass coverslips into 24-well plates. After 4% paraformaldehyde was utilized to fix HCC cells and washed with PBS, then subjected to permeabilization (0.5%Triton-X100 PBS). HCC cells were incubated with prehybridization solution and hybridized with hybridization solution, and then incubated with hybridization solution contain Cy3-labeled SZT2-AS1 oligonucleotide probe overnight. HCC cells nuclei were visualized with DAPI. All images were captured and recorded under a Zeiss fluorescence photomicroscope (Carl Zeiss AG).
Animal experiments
The growth and metastasis ability of cells in vivo were assessed by the orthotopic HCC model, subcutaneous xenograft model and lung metastasis model in mice. 4 weeks old male BALB/C nude mice purchased from the Centre of Laboratory Animals at The Medical College of Xi’an Jiao tong University and the mice were randomly grouped (n = 5 per group). All the animal experiments were approved by the Research Ethics Committee of Xi’an Jiao tong University.
For the construction of orthotopic HCC model, 1×107 Hep3B-shSZT2-AS1 or MHCC97H-shSZT2-AS1 or control subclones were dissolved in 0.1 mL of DMEM culture medium. Mice were anesthetized with 3% pentobarbital sodium, and the liver was exposed by open surgery. The cells were injected into the liver of the nude mice, and the wound was sutured with 5 − 0 silk thread. 4 weeks later, the mice were sacrificed in accordance with ethical procedures to observe the tumor formation, and the liver tissues were used for H&E staining.
In vivo subcutaneously tumor growth assay, 1×107 Hep3B-shSZT2-AS1 or MHCC97H-shSZT2-AS1 subclones and the same numbers corresponding control subclones were transplanted into the flank of 4-week-old BALB/c nude mice via subcutaneous injection. After injection, we measured tumor size calculated 0.5 × length × width × width every 3 days. The nude mice were killed after 21 days in different groups, and the tumor specimens were weighed, fixed and harvested for IHC experiments.
In vivo lung metastasis model, we intravenously injected 1×106 cells into the lateral tail vein of nude mice. After 5 weeks killed the mice and collect the lung tissues. Thereafter, the lungs were fixed, photographed, preserved, and stained with H&E to analyze the presence of metastatic nodules.
Immunohistochemistry (IHC)
For immunohistochemistry, xenograft tumors from subcutaneous xenograft models were fixed with paraformaldehyde and paraffin-embedded. Slice the sample and install it on the slide. The sectioned slides were dewaxed in xylene, rehydrated with ethanol of decreasing concentration and subsequently microwave boiled in the antigen repair solution to expose the antigen, after that the slides were incubated by antibody against Ki-67, α-SMA or CD31 at 4°C overnight and corresponding horseradish peroxidase coupling with secondary antibodies for 10 min at room temperature. Next, diaminobenzidine was reacted under horseradish peroxidase catalyzation, brown pigments are formed at the site. Then, Nuclear staining with hematoxylin and examined under a microscope. The results of Ki-67 staining were analyzed by the positive staining cell. The IHC scores used to assess the results of Ki-67, α-SMA and CD31 staining, defined as percentage score (0 for < 5%; 1 for 5–25%; 2 for 25–50%; 3 for > 50%) × staining intensity score (none scored 0; weak scored 1; moderate scored 2; strong scored 3). The antibodies were listed in Supplementary Table 2.
Statistical analysis
GraphPad Prism software version 8.0 (GraphPad Software, Inc., San Diego, CA, USA) and SPSS 20.0 software (SPSS, Inc., Chicago, IL, USA) were used for statistical analysis. All data of the study are presented as mean ± S.D. Statistical methods in this study included Student’s t test, one-way ANOVA, Chi-square test, Kaplan–Meier method, log-rank test and Pearson's correlation coefficient analysis and so on. Difference with P<0.05 was deemed to indicate statistically significant.