Long non-coding RNA LINC00467 regulates hepatocellular carcinoma progression by modulating miR-9-5p/PPARA expression.

The aim of this study was to analyse the expression pattern and elucidate the mechanistic involvement of long non-coding RNA LINC00467 in hepatocellular carcinoma (HCC). The relative expression of LINC00467 and microRNA (miR)-9-5p was determined by real-time polymerase chain reaction. Cell viability was measured using the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide assay. The cell proliferation was analysed by cell counting. Cell migration and invasion were monitored by Transwell assay. The luciferase reporter assay was employed to investigate the regulatory effect of miR-9-5p on LINC00467 and peroxisome proliferator-activated receptor alpha (PPARA). The endogenous PPARA protein was quantified by western blotting. It was found that LINC00467 was aberrantly decreased in HCC. The ectopic expression of LINC00467 significantly suppressed cell viability, proliferation, migration and invasion. LINC00467 functioned as a sponge for miR-9-5a and negatively regulated miR-9-5p expression. We also identified PPARA as the direct target of miR-9-5p. siRNA-mediated knockdown of PPARA in LINC00467-proficient cells promoted cell viability, migration and invasion. Our data indicate the critical involvement of LINC00467/miR-9-5p/PPARA signalling in the incidence and progression of HCC.

Introduction

Hepatocellular carcinoma (HCC) is one of the most common human malignancies worldwide and the most common cause of death in patients with cirrhosis [1]. It is estimated that 662 000 deaths occur as a result of this disease globally per year and the vast majority of HCC deaths occur in Asia and sub-Saharan Africa. The incidence of HCC is intimately associated with chronic liver inflammation, especially links to chronic viral hepatitis infection (hepatitis B or C) or frequent exposure to toxins such as alcohol and aflatoxin [2]. The other complications including haemochromatosis, alpha 1-antitrypsin deficiency, metabolic syndrome and non-alcoholic steatohepatitis are also recognized as risk factors for HCC tumorigenesis [3]. Notably, the predisposition to HCC in type 2 diabetics is greater and depends on the duration of the diabetes and treatments [4]. Molecularly, HCC features in epigenetic alterations and mutations affecting the cellular machinery and has a higher replicative rate, which was radically driven by hepatitis infections [5]. Clinical diagnosis of this disease evolved with the advance of medical imaging technology, which is conclusive enough without biopsy of the tumour [6]. Based on the stage of tumour progression, the treatments for HCC vary and include surgery, liver transplantation, targeted therapy and palliation [7]. Despite the advances in medical care, the morbidity and mortality of HCC remain high, and HCC urgently requires in-depth understanding of this disease to identify novel therapeutic targets.

With completion of the human genome and transcriptome project, approximately 80% of transcripts are characterized without protein-coding potential. The family of RNA molecules longer than 200 bp was categorized into long non-coding RNAs (lncRNA) [8]. According to the LncRNAdb (http://www.lncrnadb.org) database, more than 300 lncRNAs with recognized physiological function have been curated and annotated. The diverse biological functions of lncRNAs are increasingly uncovered to involve multiple aspects of the gene regulatory network [9]. LncRNAs are identified to target a number of cis- and trans-components during gene transcription processing in eukaryotes, which include activators, repressors, polymerase and DNA strands per se [10]. Moreover, it has been reported that lncRNAs are involved in mRNA processing, such as splicing, translocation, stability and translation, which mechanistically resembles microRNAs (miRs) [11]. LncRNAs have also played a fundamental role in the epigenetics reprogramming underlying sex chromosome imprinting and telomere maintenance [12]. LINC00467 is a novel lncRNA with an unknown physiological function. The sole investigation into LINC00467 has demonstrated the oncogenic N-Myc suppressed expression of this lncRNA in neuroblastoma, which in turn induced Dickkopf-related protein 1 expression and consequently contributed to the malignancy [13]. However, the expression status and underlying function of LINC00467 in HCC remains elusive. Here, we set out to characterize the alteration with respect to LINC00467 expression in clinical tumour samples, and further elucidate the underlying molecular mechanism.

Material and methods

Cell culture

The human HCC cell lines SMMC-7721 and HepG2 were obtained and authenticated by the American Type Culture Collection (Manassas, VA, USA). Cells were verified by short tandem repeat analysis before use. Both cell lines were cultured in Roswell Park Memorial Institute (RPMI) medium (Gibco, Grand Island, NY, USA) supplemented with 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin/glutamine. The exponential-phase cells were maintained in a humidified incubator under 5% CO2.

Transfection

The indicated cells were seeded into six well plates and transfection was performed with Lipofectamine 2000 on the following day with 1 µg of the indicated plasmid. The procedure for cell transfection was in strict accordance with the manual provided by the manufacturer. The green fluorescent protein was employed in the parallel assay to evaluate transfection efficiency.

Real-time PCR

The RNA was extracted from either HCC or SMMC-7721/HepG2 cells using Trizol reagent (Invitrogen, Waltham, MA, USA) following the provider's guide. The quantification and integrity of RNA was checked by BioAnalyzer 2100 (Agilent, Santa Clara, CA, USA). An equal amount of RNA was reversely transcribed using a PrimeScript First Strand cDNA Synthesis Kit (Clontech, Dalian, China) for lncRNA quantification and a miScript Reverse Transcription Kit (Qiagen, Valencia, CA, USA) for miRNA quantification in strict accordance with the manufacturer's instructions. A quantitative polymerase chain reaction (PCR) was performed with TaqMan miRNA assays on an ABI Prism 7600 HT. All primer sequence information is available upon request. The primers used were as follows: LINC00467 (F:GCCAGAGCAAGACTCTGTCTAC, R: GATGGGATACACATTCAATCAT); GAPDH (F:GAAGCTCACTGGCATGGCCTTC, R: GACCACCTGGTGCTCAGTGTAG); miR-9-5p (F:GCCGCAACCCGTAGATCCGATC, R:GTGCAGGGTCCGAGGT); U6 (F:GCTTCGGCAGCACATATACTAAAAT, R:CGCTTCACGAATTTGCGTGTCAT).

3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT)

The cell viability was assayed with a commercial Vybrant MTT Cell Viability Assay Kit (ThermoFisher, Waltham, MA, USA) following the manufacturer's recommendations. The exponential-phase cells were cultured in 96-well plates in triplicate and subjected to the indicated treatments. MTT stock solution (10 µl) was added into each well after 24 h and allowed to incubate for 4 h at 37°C. The formazan dye was dissolved in 100 µl of the SDS–HCl solution at 37°C for another 4 h. The absorbance at 570 nm was recorded using a microplate reader (Molecular Devices, Sunnyvale, CA, USA).

Cell counting

The single-cell suspension was prepared via trypsin digestion, which was followed by serum neutralization. The cell number was counted under a light microscope with a haemocytometer counting chamber.

Transwell assay

Migration and invasion were assessed by Transwell chamber (BD Biosciences, Franklin Lakes, NJ, USA). The subject cells were starved in serum-free medium for 24 h, then a single-cell suspension was prepared and added onto the top of a polycarbonate Transwell filter (for invasion investigation, the filter was precoated with 0.1% Matrigel (BD Biosciences)). The complete culture medium with 10% FBS was supplied in the lower compartment for convenience. Twelve hours later, the unattached cells in the upper chamber were carefully removed with cotton Q-tips. The migrated/invaded cells were fixed with 4% paraformaldehyde for 15 min at room temperature and stained with 0.025% crystal violet for 15 min. The representative images were captured under the optical microscope and the average cell number was calculated from three random fields to determine the Transwell capacity.

Luciferase reporter assay

The potential impact of miR-9-5p on expression of either LINC00467 or peroxisome proliferator-activated receptor alpha (PPARA) was interrogated by luciferase reporter assay. We cloned the complete LINC00467 sequence and PPARA 3′UTR region, and these were then fused into the luciferase plasmid via double digestion and ligation. The putative binding sites were scrambled by mutagenesis PCR. The luciferase plasmids and miR-9-5p were co-transfected into 293T cells. After 48 h regular culture, the relative luciferase expression was determined by a Bright-Glo luciferase assay system (Promega, Madison, WI, USA) following the manufacturer's instructions.

Western blotting

The protein was extracted in radioimmunoprecipitation buffer for 15 min on ice and the concentration was quantified by comparing with the standard curve with the bicinchoninic acid quantification kit. The protein samples were resolved by sodium dodecyl sulfate–polyacrylamide gel electrophoresis and blots were transferred onto a methanol-activated polyvinylidene fluoride membrane. After briefly blocking with 5% milk and incubation with primary antibody (mouse anti-β-actin, CST#3700, 1 : 1000; mouse anti-PPARA, CST#2435, 1 : 1000; Cell Signaling Technology, MA, USA) at 4°C overnight, and after washing and hybridization with secondary antibody (anti-mouse, CST#7074; Cell Signaling Technology), the protein blots were visualized using the enhanced chemiluminiscence (ECL, Millipore, MO, USA) method. The intensity of the protein blots was quantified via densitometry scanning.

Tissue specimens

A total of 65 HCC tissues and matched adjacent non-cancerous tissues were collected from the Affiliated Second Hospital, Mudanjiang Medical University, and frozen and stored in liquid nitrogen. The study was approved by the Ethics Committee of Mudanjiang Medical University and written informed consent was obtained from all patients.

Statistical analysis

All assays were repeated at least three times unless specified. Data were processed by SPSS 23.0 software and expressed as the mean ± standard deviation (s.d.). The statistical comparison was performed with one- or two-way analysis of variance followed by Turkey's test. The p-values were calculated and p < 0.05 was accepted with statistical significance.

Results

Overexpression of LINC00467 in hepatocellular carcinoma

We first set out to analyse the expression of LINC00467 in clinical HCC samples. The abundance of LINC00467 was interrogated in 65 tumours in comparison with 31 normal tissue samples. As shown in figure 1a, the relative expression of LINC00467 was significantly decreased in the malignant counterparts (0.58 versus 1.25 mean value, p < 0.01). Further analysis demonstrated that the suppressed expression of LINC00467 was intimately associated with hepatocarcinoma metastasis, wherein the LINC00467 contents were much lower than those in the primary tumour (figure 1b). Our results demonstrated the aberrant low expression of LINC00467 in HCC, especially in metastases, which suggested the potential tumour suppressor role of LINC00467 in this disease.

Figure 1.

The expression of LINC00467 in HCC clinical samples was analysed using quantitative reverse transcription polymerase chain reaction (qRT-PCR). (a) LINC00467 levels were measured via qRT-PCR and normalized to the level of GAPDH in HCC tissues and adjacent non-cancerous tissues. (b) LINC00467 levels were measured via qRT-PCR and normalized to the level of GAPDH in HCC tissues with or without lymphatic metastasis. (**p < 0.01.)

LINC00467 suppressed cell viability and proliferation in vitro

Based on our preliminary results acquired from clinical samples, next we sought to investigate the potential contribution of LINC00467 to the viability and proliferation of HCC cells. To this purpose, here we ectopically overexpressed LINC00467 in both SMMC-7721 and HepG2 cells. As shown in figure 2a, a significant increase in LINC00467 expression was achieved in both cell lines. The cell viability in response to LINC00467 overexpression was determined by MTT assay, and our results demonstrated remarkably suppressive effects elicited by LINC00467 (figure 2b). We monitored cell growth consecutively as well by counting. As shown in figure 2c,d, the ectopic expression of LINC00467 significantly inhibited cell proliferation, which was consistent with our in vivo finding. Therefore, we consolidated the phenotype that LINC00467 functioned as a tumour suppressor in HCC both in vivo and in vitro.

Figure 2.

LINC00467 suppresses cell viability and proliferation in HCC cells. (a) LINC00467 levels were measured via qRT-PCR and normalized to the level of GAPDH in SMMC-7721 and HepG2 cells. (b) SMMC-7721 and HepG2 cells subjected to MTT assay after plating for 48 h. (c) SMMC-7721 and (d) HepG2 cells were subjected to cell number assay every 24 h. (**p < 0.01.)

LINC00467 suppressed hepatocellular carcinoma cell migration and invasion

Our previous data suggested that a low level of LINC00467 is associated with metastasis, which hinted that a decrease in LINC00467 might relate to a metastatic phenotype as well and prompted us to clarify whether overexpression of LINC00467 had an impact on the migratory and invasive behaviours of HCC cells. As shown in figure 3a,b, the forced expression of LINC00467 significantly compromised both migration and invasion capacity of SMMC-7721 cells. The comparable effects were also observed in HepG2 cells (figure 3c,d). Therefore, high expression of LINC00467 contributed to the metastatic process in addition to cell proliferation in HCC.

Figure 3.

LINC00467 suppresses cell migration and invasion in HCC cells. SMMC-7721 cells were subjected to Transwell migration assay (a) and invasion assay (b). HepG2 cells were subjected to Transwell migration assay (c) and invasion assay (d). Data are the mean ± s.d. of three independent experiments and each is measured in triplicate (**p < 0.01).

LINC00467 was a molecular sponge for miR-9-5p

Mechanistically, we attempted to understand whether LINC00467 functioned as a microRNA sponge. With the aid of bioinformatics tools (www.mircode.org), we predicted that miR-9-5p might be the candidate target of LINC00467 and illustrate the alignment in figure 4a. To experimentally confirm our predictions, here, we constructed a LINC00467 luciferase reporter plasmid. Co-transfection with miR-9-5p markedly suppressed luciferase activity in both SMMC-7721 and HepG2 cells (figure 4b,c), whereas the mutation in the core region of the LINC00467 binding sequence completely abolished this effect. We further examined the endogenous level of miR-9-5p upon ectopic expression of LINC00467. As shown in figure 4d, the forced expression of LINC00467 significantly suppressed the miR-9-5p level in both SMMC-7721 and HepG2 cells. Likewise, the introduction of miR-9-5p into HCC cells elicited a dramatic decrease in LINC00467 (figure 4e). Our results indicated the reciprocal inhibitory effects between LINC00467 and miR-9-5p.

Figure 4.

LINC00467 is a molecular sponge for miR-9-5p. (a) Schematic showing the putative miR-9-5p target site in LINC00467. (b,c) Dual-luciferase assays showing the repression of LINC00467 following the transfection of synthetic miRNA precursor. Renilla luciferase activity was normalized to firefly activity and is presented as relative luciferase activity. (d) miRNA levels were measured via qRT-PCR and normalized to the level of U6. (e) LINC00467 levels were measured via qRT-PCR and normalized to the level of GAPDH following the transfection of the synthetic miRNA precursor. Data are the mean ± s.d. of three independent experiments and each is measured in triplicate (*p < 0.05, **p < 0.01).

LINC00467 relieved the inhibitory effect of miR-9-5p on PPARA

Next, we sought to identify the target gene of miR-9-5p in this setting. Using bioinformatics tools (www.targetscan.org), we found that PPARA might be the candidate downstream gene of miR-9-5p. The alignment between PPARA 3′UTR regions with miR-9-5p is illustrated in figure 5a. Similarly, we constructed luciferase reporter plasmids to interrogate the possible regulator action of miR-9-5p on 3′UTR of PPARA. As shown in figure 5b,c, co-expression of miR-9-5p imposed obvious suppression on luciferase activity, which was readily abolished by the mutations introduced into the putative binding region. We further confirmed this phenotype at the protein level via western blotting assay. The ectopic expression of miR-9-5p tremendously inhibited PPARA expression, while LINC00467 displayed the opposite effect with a significant increase in PPARA instead (figure 5d,e). Therefore, we identified PPARA as the target gene of miR-9-5p, which might consequently be involved in the biological effects of LINC00467.

Figure 5.

LINC00467 relieves the inhibitory effect of miRNA on PPARA. (a) The potential miR-9-5p binding site at the 3′UTR of PPARA. The target site mutations are underlined. (b,c) Dual-luciferase assays showing the repression of 3′UTR following the transfection of synthetic miRNA precursor. (d) Western blot analysis of PPARA expression following the transfection of synthetic miRNA precursor or NC. β-actin serves as an internal control. (e) Western blot analysis of PPARA expression in endometrial carcinoma cells which stably transfected pcDNA3.1-vec or pcDNA3.1-LINC00467. β-actin serves as an internal control. Data are the mean ± s.d. of three independent experiments and each is measured in triplicate (**p < 0.01).

PPARA was critical for the function of LINC00467 in hepatocellular carcinoma

Our previous data suggested that PPARA might directly and predominantly mediate the tumour suppressor role of LINC00467 in HCC, so next we sought to experimentally validate this possibility. As shown in figure 6a, the deficiency of PPARA readily restored the cell viability which was compromised by the ectopic introduction of LINC00467 in both SMMC-7721 and HepG2 cells. Likewise, the suppressed cell growth curve in LINC00467-positive cells was relieved by PPARA knockdown (figure 6b,c). We further consolidated the predominant role of PPARA with respect to cell migration and invasion, which was consistently induced by PPARA deficiency in the LINC00467-expressing cells (figure 6d,e). Therefore, we provide evidence that PPARA is directly involved in the tumour suppression effect of LINC00467.

Figure 6.

PPARA is critical for the function of LINC00467 in HCC. (a) SMMC-7721 and HepG2 cells subjected to MTT assay after plating for 48 h. (b) SMMC-7721 and (c) HepG2 cells were subjected to cell number assay every 24 h. (d) SMMC-7721 and HepG2 cells were subjected to Transwell migration assay. (e) SMMC-7721 and HepG2 cells were subjected to Transwell invasion assay. Data are the mean ± s.d. of three independent experiments and each is measured in triplicate (*p < 0.05, **p < 0.01).

Discussion

The many biological roles of cryptic lncRNAs are increasingly being discovered and the accumulating evidence suggests the critical involvement of lncRNAs in tumorigenesis and the progression of diversity in human malignancies [14]. LINC00467 was one of the novel lncRNAs with unaddressed physiological and pathological functions. Recently, an elegant study by Atmadibrata et al. [13] demonstrated that the knockdown of LINC00467 with specific siRNA in neuroblastoma cells reduced the number of viable cells and increased the percentage of apoptotic cells, which suggested that LINC00467 might function as an oncogene in this disease. However, in our study, we have characterized the aberrant overexpression of LINC00467 in HCC, which is also closely associated with metastasis, and have hinted at the tumour suppressor role of LINC00467 in this malignancy. We experimentally validated this phenotype in the cell culture system, wherein overexpression of LINC00467 significantly suppressed cell viability and proliferation in both SMMC-7721 and HepG2. Furthermore, we demonstrated that the ectopic overexpression of LINC00467 remarkably blocked cell migration and invasion as well. Mechanistically, we identified that LINC00467 functioned as a sponge for miR-9-5p. LINC00467 negatively regulated endogenous miR-9-5p expression and vice versa. We also predicted and validated that PPARA was the direct target of miR-9-5p. Via endogenous competition with miR-9-5p, LINC00467 increased the intracellular PPARA level, which consequently contributed to the tumour suppressive effects. SiRNA-mediated knockdown of PPARA stimulated cell proliferation, migration and invasion in the presence of LINC00467, which unambiguously highlighted the predominant role of PPARA in mediating the tumour suppressor role of LINC00467. Taking these results together, our study elucidated the critical role of the LINC00467/miR-9-5p/PPARA signalling axis in the tumorigenesis and progression of HCC, which might hold invaluable promise for future diagnostic and therapeutic exploitations.

Of note, the previous investigation on the biological function of LINC00467 indicated an oncogenic role for LINC00467 in neuroblastoma, wherein the knockdown of LINC00467 decreased cell viability and induced remarkable apoptosis. In sharp contrast, our in vivo and in vitro data supported a tumour suppressor function of this lncRNA, which might indicate that LINC00467 plays different roles in a cell context-dependent manner.

Assembling data disclosed the multifaceted actions of miR-9-5p in the range of human tumours. For instance, Miguel et al. [15] suggested a protective role for miR-9-5p in the fibrogenic transformation of human dermal fibroblasts. Xie et al. [16] demonstrated that lncRNA TUG1 contributed to tumorigenesis of human osteosarcoma by sponging miR-9-5p and regulating POU2F1 expression. On the contrary, Li et al. [17] showed that miR-9-5p promoted mesenchymal stem cell migration by activating the β-catenin signalling pathway. Li et al. [18] suggested that miR-9-5p promoted cell growth and metastasis in non-small cell lung cancer through the repression of TGFBR2. In line with the notion of the oncogenic role of miR-9-5p, we demonstrated here, for the first time, that LINC00467 inhibited miR-9-5p via an endogenous competition mechanism, which fundamentally contributed to the tumour suppressive effects of LINC00467.

Multiple target genes of miR-9-5p have been identified in diverse human diseases. For example, Fierro-Fernandez et al. [19] reported that miR-9-5p suppressed pro-fibrogenic transformation of fibroblasts and prevented organ fibrosis by targeting NOX4 and TGFBR2. Along with miR-124-3p and miR-132-3p, miR-9-5p was shown to regulate BCL2L11 in tuberous sclerosis complex angiomyolipoma [20]. Yu et al. [21] demonstrated that epigenetically regulated miR-9-5p suppressed the activation of hepatic stellate cells via modulation of TGFBR1 and TGFBR2. D'Amore et al. [22] identified miR-9-5p as the direct regulator of ABCA1 and HDL-driven reverse cholesterol transport in circulating CD144 + cells of patients with metabolic syndrome. In addition to the recognized targets, in this study, we identified PPARA as a direct substrate gene via bioinformatics prediction and experimental examination. Our data definitely contributed to the better understanding of the biological function of miR-9-5p.

The close link between PPARA and human malignancies has been increasingly acknowledged. Several studies have identified that polymorphisms of PPARA were associated with an increased risk of breast cancer in different populations [23,24]. Leclerc et al. [25] further showed that differential gene expression and methylation in the retinoid/PPARA pathway and in tumour suppressors might modify intestinal tumorigenesis induced by low folate in mice. In agreement with the tumour suppressor role of PPARA, here we demonstrated that LINC00467 positively regulated PPARA via competing with miR-9-5a, and a deficiency in PPARA significantly promoted cell proliferation, migration and invasion behaviours.

It is worth noting that there are several limitations in the current study. First, although two human HCC cell lines were employed, it would be more convincing to see similar regulation of LINC00467 in other tumour cells. Second, only in vitro studies were conducted in this research; evidence from in vivo animals would strengthen the conclusions. Last, a larger number of clinical samples should be further verified for the correlation between the expression of LINC00467 and HCC.

In summary, here we elucidated that low expression of LINC00467 is associated with the incidence and progression of HCC and highlighted the critical role of the LINC00467/miR-9-5a/PPARA signalling axis in this disease.