Introduction

Hepatocellular carcinoma (HCC) is the most prevalent malignancy of the liver in adults, and the 3rd most frequent cause of cancer-related deaths around the world [1]. HCC cases are mainly associated with hepatic cirrhosis, which is typically induced by chronic hepatitis B (HBV) or hepatitis C virus (HCV) infection [2]. HCC staging is critical to clinical prognosis or optimal therapy selection, which includes evaluation of tumor size, clinical status, liver function, and portal pressure [3]. HCC occurrence is strongly associated with chronic liver disease attributable to viral hepatitis, nonalcoholic steatohepatitis and alcohol abuse [4]. Commonly applied treatment modalities include drugs (sorafenib), surgical (liver resection, liver transplantation), and ablative (transarterial chemoembolization, and radiofrequency ablation) [5]. Currently, clinical prognosis of HCC is based on factors such as tumor extension, tumor histologic subtype, degree of liver dysfunction, and systemic status of patients [6]. However, only ~30% patients remain successfully treatable [7]. Therefore, there exists a need for development of novel and effective therapeutic modalities for HCC, which may improve the quality of life of HCC patients.

Long noncoding RNAs (lncRNAs) have a length greater than 200 nucleotides without protein-coding potential and play key roles in tumor biology, including tumorigenesis, cell differentiation, and cancer metastasis [8]. Dysregulation of lncRNAs may lead to aberrant gene expression and contribute to tumor progression of HCC [9]. The lncRNA LINC01419 was found to be upregulated in HCC according to the expression profiles (GSE40367 and GSE45267). Zhang et al. observed that LINC01419 was significantly overexpressed in HBV and HCV-related HCC tissue as compared to matched normal hepatic tissues [10]. In our present study, ZIC1 was determined to be the target gene of LINC01419 and involved in HCC cell-cycle distribution and metastasis through the phosphoinositide 3-kinase (PI3K)/protein kinase B (Akt) using the Multi Experiment Matrix (MEM) website (https://biit.cs.ut.ee/mem/index.cgi). There is mounting evidence that Zic1, a zinc-finger transcription factor, has anti-tumor activity in many cancers such as colon cancer, endometrial cancer, and breast cancer [11]. Former evidence revealed that genes including Twist1, Midkine, and ZIC1, showed a specific expression profile in the regenerating liver with the highest expression at 4, 24, and 6 h respectively [12]. It has been suggested that hypermethylation may contribute to promoter silencing of ZIC1 mRNA, leading to poor survival in HCC patients [13]. ZIC1 regulates cell-cycle distribution and cell migration by inactivation of the sonic hedgehog, PI3K, and MAPK signaling pathways in the development and progression of gastric cancer [14]. Inhibition of the PI3K/Akt/mTOR pathway has been exhibited in HCC carcinogenesis [15], suggesting a basis for a putative role of ZIC1 and its upstream regulators. In this study, we explored the effects of LINC01419 on cell-cycle distribution and cell migration in HCC by targeting ZIC1 via the PI3K/Akt signaling pathway.

Materials and methods

Bioinformatics analysis

The Gene Expression Omnibus database (http://www.ncbi.nlm.nih.gov/geo) was retrieved, and expression profiles of HCC-related microarrays (GSE40367 and GSE45267) and the associated annotation files were obtained. Affymetrix Human Genome U133 Plus 2.0 Array and Agilent-062918 OE Human lncRNA Microarray V4.0 028004 microarray were used to determine the two aforementioned profiles. Background calibration and normalization of microarray data were carried out using the “Affy” package in the R software [16]. A nonspecific filtration of expression profiles was conducted using the linear model-empirical Bayes method in the “Limma” package and t test, and lncRNAs with differential expression were selected for further experimentation [17]. Information related to HCC prognosis and associated gene expression was collected from The Cancer Genome Atlas (TCGA) database (http://cancergenome.nih.gov/). The relationship between LINC01419 and HCC prognosis was determined and a differential analysis was performed for transcriptome profiling data using the “edgeR” package in R software [18]. False positive discovery (FDR) correction was applied to the p value using the R package “multitest”, and an FDR < 0.05 with |log2 (fold change)| > 1 was set as the threshold to obtain significant differentially expressed genes. Kaplan–Meier survival curves were used to evaluate the overall survival of patients with predicted high- or low-expression groups. The survival variances between high-expression group and low-expression group were verified using a two-sided log-rank test, using the R package “survival” [19], and Hazard ratio (HR) and 95% confidence intervals (CI) were estimated by the Cox proportional hazards regression model. The MEM (http://biit.cs.ut.ee/mem/) website was utilized to predict lncRNAs with differential expression. LncATLAS (http://lncatlas.crg.eu/) was used to determine the subcellular localization of the target lncRNA. Kyoto Encyclopedia of Genes and Genomes analysis of the target gene was performed using The Database for Annotation, Visualization and Integrated Discovery (DAVID; https://david.ncifcrf.gov/) in order to identify co-expressed genes. Putative target gene and lncRNA were compared using BLAST searches.

Tissue collection

Specimens (tumor and adjacent normal tissues) were obtained from 76 patients with pathologically confirmed HCC who underwent surgical resection at Sichuan Provincial People’s Hospital between July 2012 and July 2014. The included patients comprised of 43 males and 33 females, with a mean age of 45.7 ± 9.4 years (age range: 23–63 years). All included patients had not received any treatment for HCC before the surgery. Each specimen was diagnosed independently by 2 experienced physicians, respectively. Tumor grading was performed according to the tumor–node–metastasis staging standard of the Union Internationale Contre le Cancer 2002 and the Edmondeon pathological grading standard [20]. Tumor and adjacent normal tissues (at least 2 cm from the edge of the tumor) were collected, and then part of the specimens were frozen in liquid nitrogen, and transferred to a refrigerator at −80 °C for subsequent experiments.

Cell culture

Normal hepatocytes (Shanghai Haling Biotechnology Co., Ltd, Shanghai, China) and HCC cells (SMMC7721 and SK-Hep-1; purchased from iCell Bioscience Inc, Shanghai, China) were cultured at 37 °C in a humidified incubator containing 5% CO2 using Roswell Park Memorial Institute 1640 medium (Hyclone, South Logan, UT, USA) supplemented with 10% fetal bovine serum (FBS) (Hyclone, South Logan, UT, USA). The medium was changed every 24 h when the cells were found adherent to the walls of the container. Cells were passaged for 3–4 days and the cells at the logarithmic phase of growth were used for subsequent experiments. Each experiment was repeated three times.

Construction of lentiviral vector

HCC cell lines with stable overexpression and low expression of LINC01419 were each constructed using lentiviral vector packaging. LINC01419 cDNA and zinc finger of the cerebellum (ZIC1) sequence (Fig. 1A, B) (Ensembl database) with the Xhol and Notl enzyme cleavage sites, and shRNA sequence (designed by thermo fisher website) were synthesized. T4 ligase was used to connect the sequence with the lentiviral expression vector PLVX-IRES-neo (Fig. 1C) (Clontech, Mountain View, CA, USA), and transferred into competent Escherichia coli. After 24 h culture at 37 °C, the plasmids were extracted. The successful construction of LINC01419, ZIC1, and shRNA eukaryotic expression vectors was verified by sequencing. The shRNA sequence of LINC01419 was: 5′-GCAGTAAGTCCACTTGAAATT-3′, and the shRNA sequence of ZIC1 was: 5′-GCACTTCACTGAGGACTTTGC-3′. Each experiment was repeated three times.

Fig. 1: Construction of lentiviral vector using a pLVX-IRES-Neo plasmid.
figure 1

A cDNA sequence of LINC01419; B cDNA sequence of ZIC1; C structure of lentivirus expression vector.

Full size image

LINC01419-cDNA plasmid, LINC01419-shRNA plasmid, ZIC1-cDNA plasmid, ZIC1-shRNA, blank plasmid, and blank vector were transfected into the human embryonic kidney cell line 293T (Microbiological Culture Collection Center, Institute of Biochemistry and Cell Biology, Chinese Academy of Sciences, Shanghai, China) using Lipofectamin 2000 (Invitrogen, Carlsbad, CA, USA). The cells were passaged for no longer than 6 months. After 6–8 h of culture, the original medium was replaced with the complete medium. The cell supernatant containing lentivirus was obtained after 48–72 h after transfection, and centrifuged at 2000 rpm for 20 min. The cell debris was filtered (pore size: 0.45 μm) to obtain the high concentration of virus liquid, which was stored at −80 °C. The experiments were repeated at least in triplicate.

Cell grouping and transfection

SMMC7721 cells at the logarithmic phase of growth were assigned to the following ten groups: the blank group (cells without any treatment), the LINC01419-cDNA group (cells transfected with LINC01419-cDNA lentiviral vector), the LINC01419-shRNA group (cells transfected with LINC01419-shRNA lentiviral vector), the negative control (NC) group (cells transfected with NC virus vector), the ZIC1-cDNA group (cells transfected with ZIC1-cDNA lentiviral vector), the ZIC1-shRNA group (cells transfected with ZIC1-shRNA lentiviral vector), the LINC01419-cDNA + ZIC1-cDNA group (cells co-transfected with LINC01419-cDNA and ZIC1-cDNA lentiviral vectors), the LINC01419-shRNA + ZIC1-shRNA group (cells co-transfected with LINC01419-shRNA and ZIC1-shRNA lentiviral vectors), the ZIC1-cDNA + IGF-1 group (cells treated with IGF-1, a PI3K/Akt signaling pathway activator [R&D Systems Inc., Minneapolis, MN, USA, dissolved in PBS at pH 7.4, working concentration 10 ng/mL] and transfected with ZIC1-cDNA lentiviral vector), and the ZIC1-shRNA + Wortmannin group (cells treated with a PI3K/Akt signaling pathway inhibitor Wortmannin [LC Sciences, Houston, TX, USA, dissolved in DMSO, working concentration 2.5 μM] and transfected with ZIC1-shRNA lentiviral vector). All experiments were repeated in three times.

The SMMC7721 cells were seeded in a 6-well plate, and infected with LINC01419-cDNA lentiviral vector, LINC01419-shRNA lentiviral vector and empty vector at different titers, respectively, when the cells had reached 70–80% confluence. The original medium was renewed with fresh medium containing G418 (500 μg/mL, GIBCO BRL, Grand Island, NY, USA) after 24 h. The medium was renewed every 2–3 days in accordance with cell conditions. Infection was carried out for 12–15 days. Subsequently, the screening medium containing 500 μg/mL G418 was used for selection every 4–5 days until stably transfected cells with drug-resistance were obtained. Each experiment was repeated three times.

Nucleus/cytoplasm extraction of HCC cells

The nucleus and cytoplasm of SMMC7721 cells were separated according to the instructions of the nucleus/cytoplasm extraction kit (BioVision, Exton, PA, USA), and RNA was extracted. The SMMC7221 cells (2 × 106) were lysed with lysis buffer containing 10 mM Tris-HCl (pH = 7.4), 100 mM MgCl2, and 40 mg/mL digitonin buffer solution for 10 min. Next, the cells were centrifuged at 45,803 × g for 5 min, and the supernatant was transferred to a precooled eppendorf (EP) tube. Precooled Radio Immunoprecipitation Assay (RIPA) buffer solution (100 μL) was added to resuspend the precipitate. The mixture was shaken for 15 s (4 times, with an interval of 10 min), and centrifuged at 16,000 rpm for 5 min, and the supernatant was transferred to another precooled EP tube. All operations were carried out on ice, and the temperature was maintained at 4 °C during centrifugations. Reverse transcription was carried out using the RNA of nucleus and cytoplasm. U6 and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) were regarded as the internal controls. Reverse transcription quantitative polymerase chain reaction (RT-qPCR) was employed to determine the content of LINC01419 in the nucleus and cytoplasm, and the subcellular localization of LINC01419 was analyzed. Each experiment was repeated three times.

Immunohistochemistry

Paraffin sections were dewaxed with xylene and dehydrated with gradient alcohol. A total of 100 μL protein kinase K solution (0.2 mg/mL, Sigma-Aldrich, St. Louis, MO, USA) was added for antigen repair for 10 min. Next, the protein expression of ZIC1 in HCC tissues and the adjacent normal tissues was evaluated using a two-step method (PV-900). In the NC group, the primary antibody was replaced with antibody dilution. The sections were blocked with 3% peroxidase blocking solution for 10 min at room temperature, and incubated with 50 μL nonimmune goat serum for 30 min at room temperature. The primary antibody, rabbit polyclonal antibody against ZIC1 (1:100, ab134951, Abcam, Cambridge, UK) was added and incubated at 4 °C overnight. Following incubation with polymerase auxiliary agent at room temperature for 20 min, the sections were probed with the secondary antibody labeled with horseradish peroxidase (HRP) (ab191866) at room temperature for 30 min and developed with diaminobenzidine chromogen. All antibodies were procured from Abcam (Cambridge, UK). Counterstaining and mounting were performed. A positive result was determined when cells showed cytoplasmic membrane staining in brown. Five fields of each slice were selected and observed under an optical microscope. The NIS-ELEMENTS (Nikon Instruments Inc., Melville, NY, USA) image software was used to assess the intensity and range for immunohistochemistry as follows: positive cell range 0% = 0, range < 25% = 1, range of 25–50% = 2, range of 50–75% = 3, and range of 75–100% = 4. For the staining intensity, weakly positive = 1, moderately positive = 2, and strongly positive = 3. The final value of immunohistochemistry = staining intensity value × positive cell range value, ranging from 0 to 12 [21]. Each experiment was repeated three times.

Dual-luciferase reporter gene assay

Based on the bioinformatics analysis, it was determined that ZIC1 was a target gene of the lncRNA LINC01419. A dual-luciferase reporter gene assay was carried out to verify if ZIC1 was a direct target of LINC01419. The 3′ untranslated regions (3′UTR) of ZIC1 were obtained using PCR. Restriction enzyme digestion with Xhol I and Not I was used to clone the target fragment to the downstream of dual-luciferase reporter gene pmirGLO (3577193, Promega, Madison, WI, USA). The vector was named ZIC1-wild type (Wt). After colony selection and sequencing, the plasmids were extracted. Site-directed mutation of the binding sites between lncRNA LINC01419 and ZIC1 was performed and the mutant (Mut) vector ZIC1-Mut was constructed. Two reporter plasmids were co-transfected with LINC01419-cDNA, LINC01419-shRNA, and NC to HEK-293T cell (cell culture center, Institute of cell biology, Chinese Academy of Sciences, Shanghai, China,). Dual-Luciferase Reporter Assay System (Promega, Madison, WI, USA) was then used to determine the luciferase activity. The ratio of firefly luciferase activity to renilla luciferase activity was regarded as the relative luciferase activity. Each experiment was repeated three times.

Methylation-specific PCR (MSP)

DNA from HCC tissues and adjacent normal tissue specimens was treated with bisulfite. The Wizard resin (Promega Corporation, Madison, WI, USA) was utilized to purify the DNA samples. Samples were precipitated with ethanol after NaOH treatment, and resuspended in water. PCR primers were designed according to the Herman method [22, 23]. The primer sequences are described in Table 1. ZIC1 MSP primers (M) and unmethylation-specific primers (U) were purchased from Invitrogen (Carlsbad, CA, USA). The MSP amplification protocol was as follows: denaturation at 94 °C for 5 min, followed by 44 cycles at 94 °C for 15 s, at 60 °C for 10 s, and at 72 °C for 9 s. The unmethylation-specific amplification protocol was as follows: denaturation at 94 °C for 5 min, followed by 36 cycles at 94 °C for 15 s, at 60 °C for 10 s, and at 72 °C for 8 s. PCR products were subjected to 1.5% agarose gel electrophoresis (Shanghai Sangon Biological Engineering Technology & Services Co., Ltd., Shanghai, China) and stained with Gelred (Biotium Inc., Hayward, CA, USA). The product bands were observed under Ultraviolet illumination (Bio-Rad, Hercules, CA, USA). Each experiment was repeated three times.

Table 1 Primers for MSP and sulfide sequencing PCR.
Full size table

DNA was extracted from human peripheral blood lymphocytes using a reagent (Tianjin Haoyang Biotech Co., Ltd, Tianjin, China). The DNA (50 μg) was dissolved in 150 μL water and incubated with 2.5 μL 32 mM SAM (New England Biolabs, Beverly, MA, USA), 25 μL buffer solution (10×, TaKaRa, Tokyo, Japan), 12.5 μL soluble starch synthase (SSS) I (New England Biolabs, Beverly, MA, USA), and 60 μL ddH2O at 37 °C for at least 4 h. Next, the mixture was incubated with 5 μL 32 mM SAM and 6 μL SSS I (at 37 °C for at least 4 h. Subsequently, the product was centrifuged for 10 min at 13,000 rpm with 260 μL phenol–chloroform (25:25). The upper aqueous phase was transferred to a fresh centrifuge tube and allowed to stand overnight at −20 °C with 350 μL 7.5 M ammonium acetate, 2 μL glycogen, and absolute ethanol (3×), followed by centrifugation for 20 min at 13,000 rpm. Next, 70% ethanol was used to wash the precipitate, followed by another 5 min centrifugation at 13,000 rpm. After drying at room temperature, the DNA was diluted with 100 μL ddH2O, and its concentration in the sample was measured. Each experiment was repeated three times.

RNA immunoprecipitation (RIP)

The experiment was conducted according to the instructions of Magna RIP RNA-Binding Protein Immunoprecipitation kit (Millipore Corp., Billerica, MA, USA). Cells in the blank, pLVX-IRES-neo-LINC01419-cDNA and pLVX-IRES-neo-LINC01419-shRNA groups were lysed with 100 μL lysis buffer solution containing proteinase inhibitors and ribonuclease inhibitors on ice for 30 min, centrifuged at 4 °C for 3 min at 12,000 rpm. A small amount of supernatant was selected as the input positive control. The remaining supernatant was probed with 1 μg rabbit antibodies to DNMT1 (ab13537), DNMT3a (ab2850), or DNMT3b (ab2851), and 10–50 μL Protein A/G-beads overnight at 4 °C and centrifuged at 3000 rpm at 4 °C for 5 min. The protein A/G-beads precipitate was rinsed 3–4 times with 1 mL lysis buffer solution, followed by a 1 min centrifugation at 1000 rpm at 4 °C after each rinsing. Following the addition of 15 μL 2× sodium dodecyl sulfate buffer for a 10 min boiling, RNA was obtained from precipitate using the RNA extraction method. Interactions among DNMT1, DNMT3a, DNMT3b, and lncRNA LINC01419 were confirmed using RT-qPCR with specific primers of LINC01419 (Table 2). Each experiment was repeated three times.

Table 2 The primer sequences for reverse transcription quantitative polymerase chain reaction.
Full size table

Chromatin immunoprecipitation (ChIP)

The ChIP kit (Millipore Corp., Billerica, MA, USA) was utilized to investigate the enrichment of DNMT1, DNMT3a, and DNMT3b in the ZIC1 gene promoter region. When cells reached 70–80% confluence, the cells were fixed with 1% formaldehyde at room temperature for 10 min to obtain intracellular DNA and protein cross-linking. The ultrasonic breaker was set to 10 s per ultrasonic cycle with 10 s intervals for 15 cycles in order to break the chromatin. Centrifugation was conducted at 13,000 rpm at 4 °C. The supernatant was probed with positive control antibody (RNA polymerase II), NC antibody (normal mouse antibody to IgG), and specific rabbit antibodies to target proteins, DNMT1 (ab13537), DNMT3a (ab2850), and DNMT3b (ab2851) at 4 °C overnight. All antibodies were procured from Abcam (Cambridge, UK). Protein Agarose/Sepharose was used to precipitate the endogenous DNA–protein complex. The cross-linking was reversed overnight at 65 °C. DNA fragments were extracted and purified with phenol/chloroform, and the specific primers of ZIC1 (Table 1) were used to examine the binding of DNMT1, DNMT3a, and DNMT3b to the ZIC1 promoter region via RT-qPCR. Each experiment was repeated three times.

Immunofluorescence assay

Accordance with instructions of the immunofluorescence staining kit (Shanghai Beyotime Biotechnology Co., Ltd, Shanghai, China), the expression and localization of ZIC1 protein in HCC were analyzed. Briefly, HCC cells were seeded in the 24-well plate, fixed with a fixative, and blocked with immunostaining blocking buffer for 60 min. Subsequently, the cells were probed with diluted the primary antibody, rabbit antibody to ZIC1 (ab134951, 1:100–1:250, Abcam, Cambridge, UK), overnight at 4 °C, and the secondary antibody, fluorescein isothiocyanate (FITC)-labeled goat anti-rabbit antibody (1:1000–1:5000, ab6717, Abcam, Cambridge, UK), at room temperature for 1 h. All antibodies were procured from Abcam (Cambridge, UK). The nucleus was stained with 4′,6-diamidino-2-phenylindole (DAPI) for 5 min, and mounted immediately. The protein expression of ZIC1 in HCC cells from the Blank, LINC01419-cDNA, and LINC01419-shRNA groups were each observed under a laser scanning confocal microscope. Each experiment was repeated three times.

RT-qPCR

Total RNA was extracted from tissues and cells using Trizol (15596026, Invitrogen, Carlsbad, CA, USA). The RNA concentration and purity were evaluated using an ultraviolet spectrophotometer (Beckman Coulter, Miami, FL, USA). cDNA was synthesized using RNA via reverse transcription, performed in accordance with the instructions of the PrimeScript RT reagent Kit (RR047A, Takara, Tokyo, Japan). RT-qPCR was conducted using the ABI 7500 PCR machine (Applied Biosystems, Foster City, CA, USA) using the SYBR Premix EX Taq kit (RR420A, Takara, Tokyo, Japan). Then three replicates of each group were performed. Primers (Table 2) were synthesized by Genechem (Shanghai Genechem Co., Ltd, Shanghai, China). The threshold cycle (Ct) value was obtained with U6 and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) serving as the internal controls. Relative expression of products was calculated using the 2−ΔΔCt method; ΔΔCt = [Ct target gene − Ct reference gene] experiment group − [Ct target gene – Ct reference gene] control group [24]. Each experiment was repeated three times.

Western blot

Cells were seeded in a 12-well plate, lysed using RIPA lysis buffer on ice and resolved on 4% stacking gel and 10% running gel. Thereafter, the protein on the gel was transferred to a polyvinylidene fluoride (PVDF) membrane and blocked for 2 h. The blots were probed with the following primary antibodies (rabbit antihuman antibodies) overnight: ZIC1 (ab134951, 1: 1000-1: 10,000), Akt (ab8805, 1: 500), mTOR (ab2732, 1: 1200), P-Akt (ab38449, 1: 500-1: 1000), P-mTOR (ab109268, 1: 1000-1: 10,000), CypA (ab41684, 1 µg/mL), CXCR4 (ab124824, 1: 100), E-cadherin (ab40772, 1: 10,000-1: 50,000), caspase-3 (ab13847, 1: 500), bcl-2 (ab59348, 1: 500-1: 1000) and GAPDH (ab9485, 1: 2500). Next, the secondary antibody labeled with HRP (ab191866) was added and incubated at room temperature for 2 h. All antibodies were procured from Abcam (Cambridge, UK). The blots were developed using chemiluminescence (Invitrogen, Carlsbad, CA, USA) with a Bio-rad Gel Dol EZ imager (Bio-rad, Hercules, CA, USA). Each experiment was repeated three times.

Flow cytometry

Cell apoptosis was evaluated using the Annexin V-FITC/propidium iodide (PI) double staining kit (556547, Shanghai Surejbio Biotechnology Co., Ltd, Shanghai, China). Briefly, transiently transfected cells were suspended in prechilled 1× phosphate buffered saline (PBS), and 5 µL Annexin V-FITC and PI working solution were added in sequence. The stained cells were then analyzed using a flow cytometer (Cube6, Partec GmbH, Münster, Germany) at 530 nm and >575 nm, respectively.

Cell-cycle distribution was assessed using PI single staining. The SMMC7721 cells were dispersed into a single-cell suspension, and seeded in a 10 mm plate at a cell density of 106 cells/well. Cellular DNA was stained with 500 µL prepared PI solution. The cells then were sorted by a fluorescent activated cell sorter calibur flow cytometer (BD Bioscience, San Diego, CA, USA).

SMMC7721 cells at the logarithmic growth phase were detached using 4 °C 0.25% trypsin (GIBCO BRL, Grand Island, NY, USA), centrifuged at 800 rpm for 5 min with Dulbecco’s modified Eagle’s medium (DMEM; Hyclone, South Logan, UT, USA), and resuspended using DMEM. A single-cell suspension was obtained after an incubation for 30 min to 1 h and centrifuged at 800 rpm for 5 min. A total of 1 × 107 cells were incubated with flow sorting antibody (PE-CD133, MiltenyiBiotec GmbH, BergischGladbach, Germany) in dark conditions for 30 min. Flow cytometry was used for analysis, where the cell tube without any fluorescent antibody served as the NC. Each experiment was repeated three times.

3-(4,5-dimethylthiazol-2-yl)-5(3-carboxymethonyphenol)-2-(4-sulfophenyl)-2H-tetrazolium (MTS) calorimetric assay

Cells were seeded in a 96-well plate at a density of 1000 cells/well with five replicates per group. Cell growth was observed at regular intervals (0, 24, 72, and 120 h) after cell adherence. After incubation with 100 and 10 μL MTS reagent for 1 h, the optical density value was measured at 490 nm in accordance with the instructions of the MTS kit and the cell growth curve was plotted. Each experiment was repeated three times.

Scratch test

SMMC7721 cells in the logarithmic growth phase were seeded in a 6-well plate at a density of 106 cells/well. Lines were drawn evenly at the back of the 6-well plate with a Marker pen, as the location for recording. When cells reached 100% confluence, the original culture medium was replaced with culture medium containing 1% FBS, and the cells were starved for 12 h. Scratches were created on the 6-well plate using a 200 μL pipette tip held vertically with a ruler. After three rinses with 2 mL PBS, the cells were incubated in a humidified incubator containing 5% CO2 at 37 °C for 24 h. Photos were taken at 0 and 24 h time points. A total of 15 evenly distributed lines were marked on the images, the scratch width was measured and the average scratch width was calculated. Cell migration (%) = (1 − scratch width/original scratch width) × 100%. Each experiment was repeated three times.

Transwell assay

Starved SMMC7721 cells were placed into the apical chamber of the cabinet containing Matrigel (Becton Dickinson, Franklin Lakes, NJ, USA). Thereafter, culture medium containing 10% FBS was added to the basolateral chamber. The cells in the apical chamber were removed after a 24 h incubation, fixed with 4% paraformaldehyde for 15 min, and stained with crystal violet. A total of five fields were selected, and cells were counted using an inverted microscope. Each experiment was repeated three times.

Xenograft tumor in nude mice

Thirty-two BALB/c nude mice (aged 4–6 weeks; weighing 17–21 g; irrespective of gender, purchased from Hunan SJA Laboratory Animal Co., Ltd, Changsha, Hunan, China) were maintained in a specific-pathogen-free environment. The nude mice were assigned into 8 groups (n = 4). Briefly, stably transfected 1 × 106 SMMC7721 cells were dissolved in 200 μL saline, and injected subcutaneously into the nude mice. Tumor formation and growth conditions were observed and recorded regularly. When the tumor was visible, the length and width of the tumor were measured using a Vernier caliper every 2 days. The volume of the tumor was calculated according to the following formula: length × width2 × 0.5. Accordingly, a tumor growth curve was plotted.

Tumor xenograft model in nude mice liver was established. The nude mice were euthanized by anesthesia injection when the diameter of the subcutaneous tumor reached 10 mm. The tumor tissues were excised and sliced into 1 mm3 pieces. The nude mice were anesthetized using pentobarbital sodium intraperitoneal injections at 30 g/L. A 1–2 cm incision was made under the left costal margin. The left lobe of liver was exposed, and three pieces of tumor tissues were implanted into the liver of each mouse (n = 3). The abdominal wall was sutured, and the mice were kept in a sterile environment. Nude mice were euthanized 42 days after implantation. Liver tumor invasion of adjacent organs and tissues was observed. All lung tissues were removed and embedded in paraffin wax. A total of five coronal sections (4–5 μm) were prepared from the two lung paraffin blocks at 0.8 mm intervals. Hematoxylin–eosin staining was performed. Metastatic lung nodules were counted and compared under a microscope. Alpha-Fetoprotein (AFP) immunohistochemical staining was performed using the streptomycin avidin-peroxidase (SP) method, and metastasis was observed.

Statistical analysis

Statistical analyses were performed using SPSS 21.0 (IBM Corp, Armonk, NY, USA). The enumeration data were presented as percentage and analyzed using Fisher’s exact test. Measurement data were expressed as mean ± standard deviation. The paired t test was used for comparison between the experimental data of HCC tissues and adjacent normal tissues, whereas an independent-sample t test was used for other pairwise comparisons. One-way analysis of variance (ANOVA) was carried out for comparisons between multiple groups followed by Tukey’s post hoc test. Comparison of data at different time points was performed using repeated-measures ANOVA followed by Dunnett’s post hoc test. The relationship between LINC01419 and the prognosis of HCC was analyzed by Kaplan–Meier method with log-rank test. p < 0.05 was considered to be statistically significant.

Results

LncRNA LINC01419 participates in the progression of HCC

As shown in Fig. 2A, B, lncRNA LINC01419 was found to be the most significantly upregulated lncRNA in HCC through differential analysis of DNA microarray gene expression data (GSE40367 and GSE45267). Utilizing the TCGA database we further substantiated the upregulation of LINC01419 expression in HCC, which was responsible for the poor prognosis of HCC (Fig. 2C, D). Thus, LINC01419 was selected for the current study. The LncATLAS website predicted that the subcellular location of the lncRNA LINC01419 was in the nucleus (Fig. 2E). The mRNA of GAPDH was chiefly in cytoplasmic fraction, while U6 was mainly found located in the nuclear fraction. Consequently, the nucleus and cytoplasm were fully separated. The result of the lncRNA LINC01419 localization showed a similar trend as U6, with more than 80% of lncRNA LINC01419 existing in the nucleus (Fig. 2F). This finding indicated that the lncRNA LINC01419 functioned primarily in the nucleus, consistent with the prediction made by the lncATLAS tool. Through the MEM website, the ZIC1 gene was confirmed to be a target gene of LINC01419 and involved in the cell-cycle distribution and metastasis regulation of HCC through the PI3K/Akt signaling pathway (Fig. 2G). BLAST results confirmed base pairing between LINC01419 and the ZIC1 gene promoter region (Fig. 2H). The dual-luciferase reporter gene assay showed that luciferase activity in the WT LINC01419-cDNA group was decreased, whereas it was markedly increased in the LINC01419-shRNA group (p < 0.05). There were no significant differences in the luciferase activity of the mutant plasmids (p > 0.05; Fig. 2I). These findings provide evidence for the speculation that lncRNA LINC01419 could bind to ZIC1.

Fig. 2: LncRNA LINC01419 participates in HCC progression.
figure 2

A The expression of LINC01419 in HCC profile GSE40367; B the expression of LINC01419 in HCC profile GSE45267. C Predicted expression of LINC01419 in HCC from the TCGA database; D relationship between LINC01419 expression and HCC prognosis; E subcellular localization of LINC01419 predicted by the lncATLAS website; F percentage of lncRNA LINC01419 in the nucleus or cytoplasm; G the potential role of ZIC1 in HCC progression according to analysis from the MEM website; H base pairing between LINC01419 and ZIC1 gene promoter region according to BLAST analysis; I Luciferase activity of ZIC1-Wt and ZIC1-Mut by dual-luciferase reporter gene assay. Measurement data were expressed as mean ± standard deviation. Independent-sample t test was used for pairwise comparisons. One-way ANOVA was carried out for comparisons among multiple groups followed by Tukey’s post hoc test. The relationship between LINC01419 and the prognosis of HCC was analyzed by Kaplan–Meier method with log-rank test. The experiment was independently repeated three times. *p < 0.05 compared with the NC group.

Full size image

LncRNA LINC01419 modulates HCC progression by promoting HCC cell viability, migration, and invasion

After revealing the involvement of LINC01419 through bioinformatic analysis, we then managed to validate it and explore the underlying mechanisms. According to the results of RT-qPCR, LncRNA LINC01419 expression presented a higher level in HCC tissues than in adjacent normal tissues (p < 0.05), and a higher level in SMMC7721 cells than in normal human hepatocytes (p < 0.05; Fig. 3A, B). A significant difference in viable cell numbers was observed in different groups at different time points (p < 0.05) using MTS assay. The number of viable cells was observed to change in the same group at different times, and at the same time point in different groups except for that at 24. Moreover, LINC01419 overexpression resulted in enhancement of cell viability and LINC01419 interference resulted in repression (p > 0.05; Fig. 3C, Supplementary Fig. 1A). PI staining indicated an increase in the number of cells in the G2 and S phase in the LINC01419-cDNA group (p < 0.05), while cells in the G0–G1 phase were decreased (p < 0.05). Cells in the G2 and S phase in the LINC01419-shRNA group were fewer (p < 0.05), and those in the G0–G1 phase were increased (p < 0.05). There were no significant differences between the NC and Blank groups (p > 0.05; Fig. 3D, Supplementary Fig. 1B). Then, the number of CD133+ liver cancer stem cell, as detected by flow cytometry, was increased in response to LINC01419 overexpression and decreased in response to LINC01419 interference (Fig. 3E). As shown in the Fig. 3F, G, and Supplementary Fig. 1C, D, cell migration was verified by Scratch test and invasion by Transwell assay. Compared with the Blank group, cell migration and invasion were significantly enhanced in the LINC01419-cDNA group (p < 0.05), whereas in the LINC01419-shRNA group, cell migration and invasion were significantly reduced (p < 0.05), and no significant differences were observed between the NC and Blank groups (p > 0.05). Taken together, LINC01419 promotes HCC cell-cycle entry from G1 phase to the S phase, and enhances cell viability, self-renewal, migration, and invasion, whereby participating in the progression of HCC.

Fig. 3: LncRNA LINC01419 is highly expressed in HCC tissues and SMMC7721 cells and participates in HCC development.
figure 3

SMMC7721 cells were treated with LINC01419-shRNA or LINC01419-cDNA. A LINC01419 expression in HCC and adjacent normal tissues (n = 76); *p < 0.05 compared with adjacent normal tissues; B LINC01419 expression in HCC cells as compared to that in human normal hepatocytes; C proliferation of SMMC7721 cells determined by MTS assay; D SMMC7721 cell-cycle evaluated by Annexin V-FITC/PI double staining; E CD133+ HCC stem cell ratio in SMMC7721 cells sorted by flow cytometry; F Scratch assay to observe the SMMC7721 cell migration in each group; G Transwell assay to examine the invasion of SMMC7721 cells. Measurement data were expressed as mean ± standard deviation. The paired t test was used for comparison between the experimental data of HCC tissues and adjacent normal tissues. One-way ANOVA was carried out for comparisons among multiple groups followed by Tukey’s post hoc test. Comparison of data at different time points was analyzed by repeated-measures ANOVA followed by Dunnett’s post hoc test. The experiment was independently repeated three times. *p < 0.05 compared with the blank group; #p < 0.05 compared with normal hepatocytes.

Full size image

ZIC1 is implicated in HCC progression through the PI3K/Akt signaling pathway

Immunohistochemical staining of HCC and matched adjacent normal tissues revealed that the positive expression of ZIC1 was mainly located in the nucleus with a small degree of expression in the cytoplasm. Statistical analyses showed the positive expression of ZIC1 was significantly higher in adjacent normal tissues than in the HCC tissues (p < 0.05; Fig. 4A). RT-qPCR and Western blot analyses identified the expression of ZIC1 gene and PI3K/Akt signaling pathway-related genes in HCC tissues and adjacent normal tissues and as shown in Fig. 4B, C, PI3K/Akt signaling pathway-related genes exhibited higher expression in HCC tissues than in the adjacent normal tissues (p < 0.05), whereas ZIC1 was poorly expressed in HCC tissues compared to the adjacent normal tissues (p < 0.05). In contrast to the blank group, ZIC1 expression was decreased in the ZIC1-shRNA and ZIC1-shRNA + Wortmannin groups (p < 0.05). ZIC1 expression was found increased in the ZIC1-cDNA and ZIC1-cDNA + IGF-1 groups (p < 0.05). Further, we measured the expression of and apoptosis-related genes (bcl-2 and Caspase-3) as well as the upstream regulators of PI3K/AKT signaling pathway, CypA, CXCR4, E-cadherin, which have been identified by previous studies [25,26,27]. According to the results (Fig. 4D, E), the levels of CypA, CXCR4, E-cadherin, and bcl-2 were upregulated whereas Caspase-3 level was downregulated in response to ZIC1 silencing or the combination of ZIC1 overexpression and IGF-1, suggesting the PI3K/Akt signaling pathway was activated and the apoptosis was suppressed; in contrast, ZIC1 overexpression or the combination of ZIC1 silencing and Wortmannin led to the opposite results. Moreover, no obvious difference was found in the expression of the aforementioned factors and ZIC1 between tissues subjected to ZIC1 silencing alone and its combination with Wortmannin, or between tissues subjected to ZIC1 overexpression alone and its combination with IGF-1 (p > 0.05). Annexin V-FITC/PI double staining and MTS assay revealed that the cell apoptosis rate was much lower in the ZIC1-shRNA and ZIC1-cDNA + IGF-1 groups than in the blank group (p < 0.05), with a significantly higher proliferation rate (p < 0.05). The cell apoptosis rate was found to be elevated in the ZIC1-cDNA and ZIC1-shRNA + Wortmannin groups (p < 0.05), along with a retarded cell proliferation rate (p < 0.05). No significant differences in cell apoptosis rate and the proliferation rate were observed between the ZIC1-cDNA and ZIC1-shRNA + Wortmannin groups, and between the ZIC1-shRNA and ZIC1-cDNA + IGF-1 groups (p > 0.05), as shown in Fig. 4F, G and Supplementary Fig. 2. These results indicated that the ZIC1 gene could regulate HCC progression by inhibiting the PI3K/Akt signaling pathway.

Fig. 4: ZIC1 gene, downregulated in SMMC7721 cells, is involved in the regulation of HCC development through the PI3K/Akt signaling pathway.
figure 4

A immunohistochemical staining of ZIC1 expression in the adjacent normal tissues and HCC tissues and the staining scoring of ZIC1 expression in HCC tissues and adjacent normal tissues; B mRNA expression of ZIC1 in HCC tissues and adjacent normal tissues (n = 76), normalized to GAPDH; C The expression of ZIC1 protein and PI3K pathway-related proteins in 76 HCC tissues and adjacent normal tissues (T cancer tissue, P adjacent normal tissue, No. 1–5 5 randomly selected pairs); *p < 0.05 compared with adjacent normal tissues; D mRNA levels of ZIC1 gene and apoptosis and proliferation-related genes in different groups; E protein levels of ZIC1 gene and apoptosis and proliferation-related genes in different treatments; F quantitative flow cytometry detection of SMMC7721 cell apoptosis; G SMMC7721 cell proliferation evaluated by MTS assay. Measurement data were expressed as mean ± standard deviation. The paired t test was used for comparison between the experimental data of HCC tissues and adjacent normal tissues. One-way ANOVA was carried out for comparisons among multiple groups followed by Tukey’s post hoc test. Comparison of data at different time points was analyzed by repeated-measures ANOVA followed by Dunnett’s post hoc test. The experiment was independently repeated three times. *p < 0.05 compared with the blank group.

Full size image

LncRNA LINC01419 epigenetically silenced ZIC1 through methylation of ZIC1 promoter region

Methylation of the ZIC1 gene in 76 cases of HCC and adjacent normal tissues was evaluated using the MSP assay and the methylation rate of ZIC1 in 76 cases of HCC tissues was found as 56.58% (43/76), while no methylation was detected in the adjacent normal tissues. As shown in Fig. 5A, the methylation of ZIC1 gene appeared to occur more frequently in HCC tissues, whereas no methylation was observed in normal tissues. Moreover, when was ZIC1 activated, there were significantly fewer methylation cases (p < 0.05). When ZIC1 expression was reduced or lost, an opposite trend was observed. As shown in the Fig. 5B, the methylation of ZIC1 gene led to decreased or blocked expression of ZIC1. The MethPrimer software was used to input a 3000 bp nucleotide sequence near the promoter region to analyze the CpG islands. It revealed that CpG islands existed both in the upstream and downstream of the ZIC1 gene promoter region (Fig. 5C), and the expression of the ZIC1 gene was affected by methylation of the promoter region. MSP results demonstrated that methylation of ZIC1 gene occurred in the blank and LINC01419-cDNA groups. The ZIC1 gene was partially methylated in the Blank group whereas it was completely methylated in the LINC01419-cDNA group. However, there was no methylation in the LINC01419-shRNA group (Fig. 5D). To further substantiate the interaction between lncRNA LINC01419 and ZIC1 methylation, RIP and ChIP were used to determine the recruitment of lncRNA LINC01419 and ZIC1 gene to DNMT1, DNMT3a, and DNMT3b in the Blank, LINC01419-cDNA, and LINC01419-shRNA groups, respectively. As shown in Fig. 5E, versus the IgG group, lncRNA LINC01419 and methyl transferase were significantly enriched in the Blank, LINC01419-cDNA, and LINC01419-shRNA groups, and additionally, ZIC1 gene and methyl transferase were also enriched. Compared to the Blank group, the binding of ZIC1 gene with DNMT1, DNMT3a, and DNMT3b increased significantly in the LINC01419-cDNA group (p < 0.05), but decreased remarkably in the LINC01419-shDNA group (p < 0.05; Fig. 5F). The immunofluorescence assay showed that, compared with the Blank group, the protein expression of ZIC1 in the LINC01419-cDNA group exhibited a lower level, while a higher level was observed in the LINC01419-shRNA group (Fig. 5G). These results suggest that the protein expression of ZIC1 is downregulated by lncRNA LINC01419, and LINC01419 epigenetically silences ZIC1 by methylation of ZIC1 gene in the promoter region.

Fig. 5: LncRNA LINC01419 epigenetically silences ZIC1 by methylation in the promoter region of ZIC1 gene.
figure 5

A Methylation of ZIC1 gene in HCC tissues and adjacent normal tissues (H2O double negative control, IVD in vitro methylated DNA, positive control, NL normal lymphocytes from healthy individuals, No. 1–6 six pairs of HCC tissues and matched adjacent normal tissues which were selected randomly (U unmethylation, M methylation); B correlation analysis of ZIC1 methylation and its expression in HCC tissues and matched adjacent normal tissues (U unmethylation, M methylation); *p < 0.05 compared with HCC tissues/ZIC1 expression; C CpG island distribution of ZIC1 gene in the promoter region (light blue area: CpG island; red thick line: inputted sequences); D methylation of ZIC1 gene in the Blank, LINC01419-cDNA and LINC01419-shRNA groups (H2O double negative control, IVD in vitro methylated DNA, positive control, NL normal lymphocytes from healthy individuals, U unmethylation, M methylation); E the recruitment of DNMTs of LINC01419 and ZIC1; F histogram representing the results of ChIP assessing the binding of LINC01419-cDNA and ZIC1 to DNMT1, DNMT3a, and DNMT3b; G immunofluorescence staining of SMMC7721 cells transfected with LINC01419-cDNA and LINC01419-shRNA (×400). Enumeration data were presented as percentage and analyzed by Fisher’s exact test. Measurement data were expressed as mean ± standard deviation. One-way ANOVA was carried out for comparisons among multiple groups followed by Tukey’s post hoc test. The experiment was independently repeated three times. *p < 0.05 compared with the blank group.

Full size image

LncRNA LINC01419 modulates the biological functions of HCC cells through ZIC1 gene

In order to determine that silencing ZIC1 can reverse the effect of LINC01419 knockdown in HCC cells, sh-LINC01419 NC + sh-ZIC1 NC, sh-LINC01419 + sh-ZIC1 NC, and sh-LINC01419 + sh-ZIC1 groups were set up to observe the changes in HCC cell proliferation, apoptosis, migration, and invasion. The results showed that the proliferation of HCC cells in the sh-LINC01419 NC + sh-ZIC1 NC and sh-LINC01419 + sh-ZIC1 NC groups was significantly decreased, and the apoptosis rate was increased, accompanied by diminished cell migration and invasion (Fig. 6, Supplementary Fig. 3). When LINC01419 and ZIC1 were knocked down at the same time, the cell proliferation, migration, and invasion significantly increased and the apoptosis rate dropped sharply relative to silencing of LINC01419 in the presence of sh-ZIC1 NC, which indicated that ZIC1 silencing could reverse the effect of sh-LINC01419 on cell proliferation, apoptosis, migration, and invasion.

Fig. 6: LncRNA LINC01419 modulates the biological function of SMMC7721 cells by downregulating ZIC1 gene.
figure 6

SMMC7721 cells were transfected with LINC01419-shRNA in the presence of ZIC1-shRNA NC or ZIC1-shRNA. A Proliferation of SMMC7721 cells determined by MTS assay; B SMMC7721 cell apoptosis evaluated by PI staining; C SMMC7721 cell migration evaluated by scratch test; D SMMC7721 cell invasion measured by Transwell assay. Measurement data were expressed as mean ± standard deviation. One-way ANOVA was carried out for comparisons among multiple groups followed by Tukey’s post hoc test. The experiment was independently repeated three times. *p < 0.05 compared with the LINC01419-shRNA NC + ZIC1-shRNA NC group; #p < 0.05 compared with the LINC01419-shRNA + ZIC1-shRNA NC group.

Full size image

LncRNA LINC01419 induces activation of the PI3K/Akt signaling pathway by repressing the expression of ZIC1

RT-qPCR and western blot analysis showed (Fig. 7) that poor expression of lncRNA LINC01419 or overexpression of ZIC1 could reduce the expression of PI3K signaling pathway-related genes and proteins (p < 0.05), while the upregulation of lncRNA LINC01419 and decline of ZIC1 could elevate their expression (p < 0.05). ZIC1 gene silencing combined with LINC01419-shRNA resulted in elevated expression of PI3K signaling pathway-related genes and proteins as compared to LINC01419-shRNA alone (p < 0.05) and overexpression of LINC01419 and ZIC1 gene at the same time resulted in their lowered expression as compared to the overexpression of LINC01419 alone (p < 0.05). These results together suggest that the lncRNA LINC01419 could activate the PI3K/Akt signaling pathway via suppression of the ZIC1 gene.

Fig. 7: LncRNA LINC01419 stimulates the activation of the PI3K/Akt signaling pathway by inhibiting the expression of ZIC1 gene.
figure 7

SMMC7721 cells were treated with LINC01419-shRNA and/or ZIC1-shRNA, LINC01419-cDNA and/or ZIC1-cDNA. A LINC01419 expression and mRNA expression of ZIC1 gene determined by RT-qPCR; B The expression of ZIC1, p-Akt/Akt, and p-mTOR/mTOR determined by western blot analysis. Measurement data were expressed as mean ± standard deviation. One-way ANOVA was carried out for comparisons among multiple groups followed by Tukey’s post hoc test. The experiment was independently repeated three times. *p < 0.05 compared with the blank group; #p < 0.05 compared with the LINC01419-cDNA group; &p < 0.05 compared with the LINC01419-shDNA group.

Full size image

LncRNA LINC01419 promotes HCC cell growth and metastasis in vivo by downregulating the expression of ZIC1

Following the aforementioned in vitro experiments, we then moved to in vivo substantiation of the HCC-promoting effect of LINC01419. No accidental mouse deaths occurred throughout the duration of the experiment. Subcutaneous tumors (Supplementary Fig. 4) in mice appeared to grow rapidly with thick and red tumor vessels on the surface of tumor. The tumor tissue was fish-meat like, crisp, and easily bleeding. Necrotic tissues were observed at the center of the tumor. In addition, tumor and deep tissues exhibited invasive growth. Forty-two days later, failure in mice with situ tumor formation was observed. The nude mice were euthanized, and the abdominal cavity was subsequently exposed. Tumor blocks with nodules and rich vessels were found on the liver surface. Most of the liver seemed to be invaded by the tumor, which even appeared to reach the lungs, mesentery, and other organs. Lungs appeared dark red in the macro-examination. The microscopic examination showed metastatic foci distributed around the lungs with a majority of interstitial micro-metastasis and intravascular tumor thrombi. Cancer cell volume was found increased profoundly, with larger nucleus, decreased cytoplasm, and pathological karyokinesis. Relative indexes of subcutaneous tumor formation and in situ tumor formation were evaluated, as shown in the Fig. 8A–D. The downregulation of lncRNA LINC01419 and overexpression of ZIC1 inhibited HCC cell growth and metastasis in vivo (p < 0.05), while high expression of lncRNA LINC01419 and poor expression of ZIC1 enhanced HCC cell growth and metastasis in vivo (p < 0.05). Silencing of both LINC01419 and ZIC1 led to a downregulated inhibitory effect as compared to the low lncRNA LINC01419 expression alone (p < 0.05). When lncRNA LINC01419 and ZIC1 were expressed at a high level together, the tumor promoting effect was found reduced as compared to low lncRNA LINC01419 expression alone (p < 0.05). The aforementioned tendencies of levels of LINC01419, ZIC1, Akt, and p-Akt were also validated in xenografted tumors of different groups (Fig. 8E). These results indicated that LINC01419 could accelerate the formation and metastasis of HCC in vivo by repressing ZIC1 expression.

Fig. 8: LncRNA LINC01419 promotes HCC tumor growth and metastasis in vivo by reducing the expression of ZIC1.
figure 8

Nude mice were treated with stably transfected SMMC7721 cells treated with LINC01419-shRNA and/or ZIC1-shRNA, LINC01419-cDNA and/or ZIC1-cDNA. A The volume of subcutaneous tumors of nude mice; B staining for tumors (×200) (a HE staining of HCC, b HE staining of lung metastasis lesions, c HE staining of pulmonary intravascular tumor thrombus, d AFP immunohistochemical staining of lung metastasis); C the number of lung metastasis lesions (n = 3). D The expression of LINC01419 in xenografted tumors of different groups; E the expression of ZIC1, Akt, and p-Akt in xenografted tumors different groups. Measurement data were expressed as mean ± standard deviation. One-way ANOVA was carried out for comparisons among multiple groups followed by Tukey’s post hoc test. Comparison of data at different time points was analyzed by repeated-measures ANOVA followed by Dunnett’s post hoc test. The experiment was independently repeated three times. *p < 0.05 compared with the blank group; #p < 0.05 compared with the LINC01419-cDNA group; &p < 0.05 compared with the LINC01419-shRNA group.

Full size image

Discussion

HCC, a prevalent malignancy, represents the fourth-leading cause of cancer deaths and accounts for more than 75% of primary liver cancer cases all over the globe [28]. Nowadays, the incidence of HCC is still on the rise due to the continuous increase in the incidence of obesity and diabetes [29]. The pathogenesis of HCC is a complicated process involving progressive accumulation of molecular alterations, different cellular targeting, and a variety of specific molecular events [30]. Further, HCC is generally diagnosed at the advanced stage and characterized by a poor prognosis since currently available therapies are mainly of palliative nature [31]. In this sense, the discovery of novel molecular mechanisms underlying HCC progression is of significance for the development of HCC treatment. Recent studies have found dysregulated lncRNAs in several cancers including HCC, as important molecular regulators in HCC progression [32, 33]. In the present study, we elucidated that LncRNA LINC01419 may modulate the proliferative, migratory, invasive, and metastatic potential of HCC cells as well as HCC tumorigenesis through the LINC01419/ZIC1/PI3K/Akt signaling pathway axis.

Our initial finding of the upregulated expression of LINC01419 in HCC-related gene expression microarrays indicated the potential participation of LINC01419 in HCC. This finding corroborates a previous study revealing the aberrantly elevated level of LINC01419 expression in HBV-related and HCV-related HCC [10]. Besides, our bioinformatic analysis validated that the subcellular localization of LINC01419 was restricted to the nucleus, and further in vitro experiments suggested that LncRNA LINC01419 may enhance the viability, progression, migration, and invasion of HCC cells. Following this, we then moved to the exploration of the downstream regulatory network of LINC01419, and ZIC1 was identified to be a target gene of LINC01419.

The ZIC family genes, consisting of five members including ZIC1, are vertebrate homologues of the Drosophila odd-paired gene and encode zinc-finger transcription factors, known to be involved in several diseases, including gastric cancer and breast cancer [34, 35]. At present, a detailed understanding of the expression pattern and significance of Z1C1 in HCC appears to be limited. In our study, complementary base pairing between the promoter region of LINC01419 and ZIC1 was observed using BLAST analysis. ZIC1 gene expression has been documented earlier in the regenerating liver at an early phase, suggesting its function in hepatic tissue growth [12]. In addition, it has been reported that hypermethylation could result in promoter silencing of ZIC1 mRNA and poor survival in HCC [13]. Notably, Chen et al. has proven that LINC01419 could provoke the esophageal squamous cell carcinoma progression by promoting the DNA methylation of glutathione S-transferase pi 1 [36]. In relation to these previous reports, our data substantiated that LncRNA LINC01419 epigenetically silenced ZIC1 through methylation of ZIC1 promoter region.

Furthermore, through a series of gain- and loss-of-function experiments we uncovered that ZIC1 acted as a regulator of HCC cell-cycle distribution and metastasis via the PI3K/Akt signaling pathway. Consistent with our findings in HCC, the restoration of ZIC1 expression has been noted to significantly restrict cell proliferation, colony formation, as well as migration and induce cell-cycle apoptosis by inhibiting the PI3K/Akt signaling pathway in thyroid cancer cells [37]. Accumulating evidence has illuminated that an important family of lipid kinase enzymes consisting of PI3Ks can control cellular processes through the regulation of a network of signal transduction pathways, and play a key role in cell proliferation in other human cancers, such as lung cancer [38, 39]. It has also been suggested that miR-7 could function as a tumor suppressor and play a key role in repressing the tumorigenesis and reversing HCC metastasis through the PI3K/Akt signaling pathway [40]. The current study observed that LINC01419 was involved in HCC through silencing ZIC1 via the induction of PI3K/Akt signaling pathway. Additionally, Huang et al. found that the neuronal precursor cell-expressed developmentally downregulated protein 4 plays a crucial role in enhancing HCC proliferation and metastasis via activation of the PI3K/Akt signaling pathway [41].

In conclusion, the current study suggests that the lncRNA LINC01419 activates the PI3K/Akt signaling pathway by silencing ZIC1 epigenetically, and thus promotes HCC cell proliferation, self-renewal, and migration, as well as tumor formation. The results of this study may shed light on potential molecular targets and therapeutic strategies for the treatment of HCC. Nonetheless, although we elucidated that methylation of ZIC1 promoted HCC development, it is also notable that our study failed to identify any association between ZIC1 expression and clinicopathologic features or prognosis of HCC. Further efforts are warranted in order to confirm these findings and verify the specific molecular mechanisms involved.