Background
Hepatocellular carcinoma (HCC) is now the second leading cause of cancer death worldwide [
1]. Aggression and postoperative recurrence are significant characteristics of HCC [
2], and treatment of the advanced HCC remains as a challenge [
3]. Liver transplantation (LT) is a common curative approach for HCC treatment. However, about 8–16% of HCC patients undergoing LT developed tumor recurrence [
4‐
6]. During HCC recurrence, the potentially residual tumor cells in the micrometastasis or in circulation can migrate to the graft liver or extrahepatic organs, locate, and finally form a recurrent tumor, indicating an intense progression of these cells. The overall survival time after HCC recurrence is only 12.97 months [
4]. Therefore, the analysis of recurrent HCCs after LT can help to explore new mechanisms of advanced hepatocarcinogenesis in order to gain insights into tumor progression.
The factors that drive HCC cells to become more aggressive are the alteration of molecular characteristics, such as cancer genome and gene regulation [
7,
8]. The evolution of molecular expression and phenotype of HCC cells can be regarded as tumor heterogeneity [
7]. Non-coding RNAs (ncRNAs), which refer to the RNA transcripts that lack significant protein-coding potential, are found to mediate various cellular processes in cancers [
9]. Long non-coding RNAs (lncRNAs) are ncRNA transcripts that are longer than 200 nucleotides (nt). By interacting with proteins, DNA and other RNA molecules, lncRNAs widely carry out their biological roles at epigenetic, transcriptional, and post-transcriptional levels [
10‐
13]. Some lncRNAs have been identified to be related to HCC progression, such as tumor recurrence after LT [
14‐
16]. MicroRNAs (miRNAs, miR) are another class of small ncRNAs that bind with other RNA transcripts, either mRNAs or other ncRNAs, to inhibit protein translation or lead to degradation of target RNAs [
17]. High-throughput screening discovered several miRNAs that serve as biomarkers of tumor recurrence risk after LT [
18,
19]. The levels of some miRNAs in circulation or in serum exosomes are also indicative of HCC recurrence risk [
20,
21]. However, the reported ncRNAs were mostly regarded as biomarkers in HCC advance and recurrence, and the underlying molecular pathway has not been adequately elucidated.
In this study, we found that two lncRNAs, HERH-1 and HERH-4, were highly expressed in the advanced HCC cells. These two lncRNAs promoted HCC cell cycle progression by facilitating CCNA2 expression, which may sequentially accelerate HCC progression.
Methods
Clinical tissue samples
Twelve pairs of human HCC tissue samples, including HCC primary and recurrent tissues from patients undergoing liver transplantation (Table S
1), were obtained from the Biological Sample Resource Sharing Center (BSRSC) of the Tianjin First Central Hospital with the patients’ informed consent. After surgical resection or biopsy, the tissue samples were flash-frozen in liquid nitrogen and then stored at − 80 °C until use. Two pairs of the tissues were applied in microarray analysis, and the other ten pairs were used in the following quantitative reverse transcription PCR (qRT-PCR) for validation of the dysregulated lncRNAs. This study was performed in accordance with the Declaration of Helsinki and was approved by the Ethics Committee of the Tianjin First Central Hospital.
HCC cell lines and transfection
An immortalized human benign hepatocyte cell line HL-7702, and four human HCC cell lines HepG2, QGY-7703, SMMC-7721 and Huh-7 were obtained from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China) with confirmed identities of these cell lines. These cell lines were maintained in DMEM (Solarbio, Beijing, China; for HL-7702, HepG2 and Huh-7) or RPMI-1640 medium (Solarbio; for QGY-7703 and SMMC-7721) supplemented with 10% fetal bovine serum (FBS; Biowest, Nuaillé, France) at 37 °C in a humidified chamber supplemented with 5% CO2. Transfection of plasmids or oligonucleotides was performed using Lipofectamine 2000 (ThermoFisher, Waltham, MA, USA).
RNA was extracted from HCC tissues and cell lines using the mirVana™ miRNA Isolation Kit (ThermoFisher), according to the manufacturer’s instructions. Long (> 200 nt) and short (< 200 nt) RNAs were isolated and purified. The separation of cytoplasmic and nucleic RNAs was achieved using a Cytoplasmic and Nuclear RNA Purification Kit (Norgen, Thorold, ON, Canada).
Microarray analysis
Complementary DNA (cDNA) labeled with Cy3-dCTP (for primary HCC) or Cy5-dCTP (for recurrent HCC) was produced by Eberwine’s linear RNA amplification method and subsequent enzymatic reaction using a cRNA Amplification and Labeling Kit (CapitalBio, Beijing, China) [
22]. The labeled cDNA was hybridized with CapitalBio Technology Human LncRNA Array V4 containing probes inspecting about 41,000 human lncRNAs and approximately 34,000 human mRNAs. Microarray data were analyzed using the GeneSpring software V13.0 (Agilent, Santa Clara, CA, USA). Genes with an absolute fold change value ≥2 and a Benjamini-Hochberg corrected
P- value ≤0.05 were treated as differentially expressed genes. Hierarchical clustering analysis was performed using Cluster 3.0 software (Stanford University, CA, USA).
qRT-PCR
For quantification of lncRNAs and protein-coding genes, 5 μg of long RNA was reverse transcribed into cDNA using oligo dT (for mRNAs) or Random 6 (for lncRNAs) primers using a PrimeScript II 1st Strand cDNA Synthesis Kit (TaKaRa, Otsu, Shiga, Japan). The cDNA was then used for amplification of the target RNA. β-actin was used as an endogenous control. Quantification of miRNAs and endogenous control U6 snRNA was performed using the stem-loop RT-PCR method [
23]. All the qRT-PCR tests were carried out in two independent experiments with each PCR reaction performed in duplicate. The sequence of all primers and oligonucleotides used in this study are provided in Table S
2.
All the real-time quantitative PCRs were performed using TB Green Premix Ex Taq II (TaKaRa) on a LightCycler 96 Real-Time PCR System (Roche, Basel, Switzerland). Gene expression was analyzed using LightCycler 96 software (V1.1, Roche).
Northern blot assay
The lncRNA length was evaluated by Northern blot assay using a NorthernMax Kit (ThermoFisher), following the manufacturer’s instructions. The blot images were captured using a ChemiDoc XRS+ imaging system (Bio-Rad, Hercules, CA, USA).
Fluorescent in situ hybridization (FISH) assay
The subcellular localization of lncRNA was evaluated by FISH assay using a Ribo Fluorescent In Situ Hybridization Kit (Ribobio, Guangzhou, China). Briefly, HCC cells were planted into 24-well plate at 8 × 104 (HepG2) or 4 × 104 (QGY-7703) per well. The transfected HCC cells were fixed in 4% paraformaldehyde and incubated in 0.5% Triton X-100 for higher membrane permeability. After prehybridization, the cells were incubated with Ribo lncRNA FISH Probe Mix (Ribobio) at 37 °C overnight. After stringency washing and DAPI staining, fluorescence was observed under an IX71 fluorescence microscope (Olympus, Shinjuku, Japan).
Artificial alteration of lncRNAs, miRNAs, and protein-coding genes in HCC cells
Overexpression of lncRNAs in HCC cells was achieved by the eukaryotic expression plasmid pcDNA3.1(+) (ThermoFisher). The exon fragments were amplified by PCR using human genomic DNA as a template. The fragments were then cloned into the pcDNA3.1(+) plasmid (Fig. S
1). Suppression of endogenous lncRNAs was achieved by transfecting synthesized siRNA into HCC cells.
A double-strand RNA fragment was synthesized to serve as mimic of the miR-29b/c. A 2′-O-methyl modified single-strand RNA fragment that was inversely complementary to mature miR-29b/c (antisense oligonucleotide, ASO) was synthesized to serve as miR-29b/c inhibitor. These oligonucleotides were induced into HCC cells to artificially change miR-29b/c levels.
Overexpression of protein-coding genes was also achieved using the pcDNA3.1(+) vector. The full-length coding sequence was amplified by PCR using a human cDNA library as a template. The fragment was then cloned into the pcDNA3.1(+) plasmid.
Cell counting kit-8 (CCK-8) cell viability assay
HCC cells were planted into 24-well plate at 8 × 104 (HepG2) or 4 × 104 (QGY-7703) per well and transfected on the next day. At 24 h after transfection, the cells were dissociated and planted into 96-well plate at 1 × 104 (HepG2) or 5 × 103 (QGY-7703) per well with three duplicates for each group. All the CCK-8 tests were carried out in three independent experiments. HCC cell viability was detected using the CCK-8 reagent (Dojindo, Tokyo, Japan). The absorbance at 450 nm (A450) was measured using an EnSpire Multimode Plate Reader (PerkinElmer, Waltham, MA, USA).
Cell doubling time detection
The HCC cells were planted and treated similarly as the CCK-8 assay. At 24 h and 48 h after planting, the number of cells in each well of the 96-well plate was accurately counted. The doubling time was calculated using the formula: doubling time = Δt(lg2/(lgNt-lgN0)), in which Δt is the interval between the cell counting (24 h in this experiment), and N0 and Nt are the cell numbers at 24 h and 48 h after planting, respectively. All the cell doubling time tests were carried out in three independent experiments with each group in triplicate.
5-Ethynyl-2′-deoxyuridine (EdU) cell proliferation assays
As a kind of thymidine analog, EdU incorporates into genome DNA in the S phase during DNA synthesis. The quantity of the incorporated EdU reflects cell proliferation activity. EdU staining of HCC cells was performed using the Cell-Light EdU Apollo488 In Vitro Flow Cytometry Kit or Cell-Light EdU Apollo488 In Vitro Kit (Ribobio). For the flow cytometry analysis, HCC cells were planted into 6-well plate at 3 × 105 (HepG2) or 1.5 × 105 (QGY-7703) per well. The cells were transfected, stained and analyzed using an Accuri C6 Plus flow cytometer (BD, San Jose, CA, USA). For imaging, the HCC cells were planted and treated similarly as the CCK-8 assay. The cells in 96-well plate were stained and captured under an IX71 fluorescence microscope (Olympus). Fluorescence intensity was quantified using ImageJ software. The EdU cytometry and imaging tests were performed in two independent experiments with each group in duplicate.
Propidium iodide (PI) staining cell cycle analysis
HCC cells were planted into 6-well plate at 3 × 105 (HepG2) or 1.5 × 105 (QGY-7703) per well. Approximately 106 transfected HCC cells were fixed in 70% ethanol for at least 2 h. After washing with PBS, cells were stained in 1 mL PI staining solution containing 10 μg/mL of PI, 100 μg/mL of RNase A, and 0.1% Triton X-100 dissolved in PBS, for 30 min in the dark. Cell cycle distribution was analyzed by flow cytometry using an Accuri C6 Plus flow cytometer (BD). The cell cycle tests were performed in two independent experiments with each group in duplicate.
The predicted CCNA2 promoter region was amplified by PCR and cloned into the pGL3/Enhancer vector (Promega, Madison, WI, USA). In addition, a CCNA2 promoter with deleted CREB1 binding sites was also amplified and cloned into the reporter vector. HCC cells were planted into 24-well plate at 8 × 104 (HepG2) or 4 × 104 (QGY-7703) per well. The reporter plasmids were transfected into HCC cells and the promoter efficiency was evaluated by measuring the luciferase activity using a Luciferase Assay System (Promega). Chemiluminescence was measured using an EnSpire Multimode Plate Reader (PerkinElmer). The gene promoter efficiency tests were performed in three independent experiments with each group in triplicate.
Chromatin immunoprecipitation (ChIP) assay
The interaction between TF and gene promoter was confirmed by ChIP assay using ab500 ChIP Kit (Abcam, Cambridge, UK) according to the manufacturer’s instructions. Approximately 1 × 106 HCC cells were collected for each ChIP assay. The TF conjugated DNA fragments were purified and the target DNA was identified using quantitative PCR.
RNA pull-down assay
The interaction between lncRNA and TF was detected by RNA pull-down assay using a Pierce Magnetic RNA-Protein Pull-Down Kit (ThermoFisher) according to the manufacturer’s instructions. Approximately 6 × 106 HCC cells were collected for each RNA pull-down assay. The potential RNA-binding proteins were eluted and purified for the further Western blot analysis.
Western blot assay
Protein samples were resolved on an SDS denaturing polyacrylamide gel and transferred onto a nitrocellulose membrane (Boster, Wuhan, China). The membrane was incubated with the primary antibody overnight at 4 °C. The membrane was then washed and incubated with horseradish peroxidase (HRP)-conjugated secondary antibody. After chemiluminescence, the bands were captured using a ChemiDoc XRS+ imaging system (Bio-Rad). The band intensity was quantified using AlphaView SA software V3.4.0 (ProteinSimple, San Jose, CA, USA).
RNA immunoprecipitation (RIP) assay
RNA targets of RNA-binding proteins were identified using a Magna RIP RNA-Binding Protein Immunoprecipitation Kit (Millipore, Billerica, MA, USA), following the manufacturer’s instructions. Approximately 2 × 107 HCC cells were collected for each RIP assay. RNA in the RBP immunoprecipitation was purified and the target RNA was analyzed via RT-PCR and agarose gel electrophoresis.
miRNA-target fluorescent reporter assay
HCC cells were planted into 24-well plate at 8 × 104 (HepG2) or 4 × 104 (QGY-7703) per well. After 24 h, the cells were transfected with a GFP reporter vector along with associated plasmids or oligonucleotides. At 48 h after transfection, the cells were lysed with RIPA lysis buffer (Solarbio) and GFP intensity was measured using an EnSpire Multimode Plate Reader (PerkinElmer). The fluorescent reporter assays were performed in three individual experiments with each group in triplicate.
Statistical analysis
All numerical values were recorded as mean ± standard deviation (SD). The hypothesis test for significance between two groups utilized the Student’s t test. For three or more groups, one-way analysis of variance (ANOVA) was applied, followed by the Student-Newman-Keuls q test for comparing two of these groups. Statistical significance was set at P ≤ 0.05. Data processing and figure drawing were performed using a GraphPad Prism 7 software (GraphPad Software, La Jolla, CA, USA).
Discussion
The high aggression rate has been a significant obstacle for HCC treatment [
4‐
6]. The recurrent HCC after LT is characterized as significantly enhanced progression, making it a typical model of the advanced HCC cells. Recurrence of HCC is indicative for HCC advancing and metastasis, which indicates worse prognosis of the patients. Therefore, the study indented to illustrate the underlying molecular mechanism in HCC recurring.
Interestingly, more than 60% of the human genome is transcribed, and the protein-coding genes account for only less than 2% of the genome [
33,
34]. The non-protein-coding transcripts are extensively involved in many cellular pathways and processes, including oncogenic signaling [
34]. In this study, we obtained the following messages through microarray data. First, both mRNA and lncRNA profiles in the recurrent tumor cells exhibited characteristic changes compared with those in the HCC primary tissues. Second, we identified that a significant biological feature of the recurrent tumor cells was acceleration of the cell cycle progression. The retinoblastoma (RB) family plays a pivotal role in the negative control of the cell cycle and in tumor progression [
30]. The advanced HCC cells exhibited a similar mRNA profile as the cells in which the RB gene family is downregulated, indicating a promoted cell cycle in these cells. Third, two of the obviously changed lncRNAs, named HERH-1 and HERH-4, were selected for mechanistic studies.
lncRNAs can interact with RNA-binding proteins (RBPs) and regulate their function [
35]. These lncRNAs are required for the correct localization of the TFs to genome DNA [
36]. In HCC cells, transcription of the cell cycle regulator CCNA2 was accelerated by CREB1, and this regulation process was HERH-1-dependent. This hypothesis was verified by the following experiments. First, CCNA2 transcription was accelerated by CREB1 protein, and this regulation depended on the three CREB1 binding motifs within the CCNA2 promoter. Second, HERH-1 directly interacted with CREB1 in a sequence-specific manner. Third, ectopic expression of CREB1 improved CCNA2 levels, promoted proliferation, and accelerated the cell cycle of HCC cells. Importantly, HERH-1 was essential in the CREB1-CCNA2 axis-mediated cell cycle acceleration. These data demonstrated that HERH-1 positively regulates CCNA2 expression and the HCC cell cycle at the transcriptional level.
lncRNA also acts as a negative regulator of miRNA [
12,
13]. RNA transcripts can sequester a limited pool of special miRNAs and prevent other RNA molecules from being inhibited by these miRNAs, known as competing endogenous RNA (ceRNA) [
31,
37,
38]. In HCC cells, the lncRNA HERH-4 acts as a natural miRNA decoy to promote CCNA2 expression at the post-transcriptional level. First, both HERH-4 and CCNA2 mRNA possess functional miR-29b/c binding sites. The miRNAs and their targets were all recruited to AGO2 protein, a key factor of RNA-induced silencing complex (RISC), in which the miRNA-targeted RNA molecules are degraded [
39,
40]. Second, the MRE fragment within the HERH-4 sequence promoted CCNA2 gene expression, which was further restored by miR-29b/c mimics. Third, miR-29b/c and their two targets had comparable levels in HCC cells, which ensured an appropriate molecular environment for ceRNA cross-talk [
32].
Cyclin is a classical protein family that controls cell cycle progression by activating cyclin-dependent kinase (CDK) enzymes or other cell cycle-associated factors. As a widely expressed cyclin A subtype, cyclin A2 (CCNA2) binds and activates CDK2 to control the G1/S transition [
41]. Our research demonstrated two novel CCNA2 upstream regulation pathways involved in HCC progression. Another key regulator of G1/S transition cyclin E2 (CCNE2) was also upregulated in the advanced HCC (Table S
9). The possible regulation of CCNE2 by HERH-1/4 are of importance to be explored in the future.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.