Hypoxia-regulated lncRNAs play pivotal roles in the development of various cancers, including HCC, gastric cancer, and pancreatic cancer, by regulating cellular proliferation, invasion, metastasis, metabolism, and autophagy[14]. As an example, Zhao et al. found that lncRNA-MALAT1 is significantly overexpressed in HCC cells under hypoxic conditions, whereas knock-down of MALAT1 weakened the promotive effect of hypoxia on cellular proliferation, migration, and invasion[10]. Zhang et al. reported that lncRNA-PCGEM1 is induced in GC cells under hypoxic conditions and acts as an oncogenic factor[15], while up-regulation of lncRNA-BX111 in response to hypoxia promotes metastasis and progression of pancreatic cancer[9]. In the present study, we confirmed the response of lncRNA-NEAT1 to hypoxia and demonstrated that lncRNA-NEAT1 is transcriptionally regulated by HIF-1α in HCC cells. Integrated analysis of public HCC-related datasets was performed to select a group of HCC-associated miRNA-mRNA pairs that could potentially be modulated by lncRNA-NEAT1 in a ceRNA-related manner. Moreover, the regulatory effects of lncRNA-NEAT1 on the miR-199a-3p/UCK2 axis in HCC were validated both in vitro and in vivo.
LncRNA-NEAT1 has been established as a target for the diagnosis and treatment of various solid tumors. Elevated expression of lncRNA-NEAT1 drives tumor initiation and progression by regulating cellular growth, migration, invasiveness, epithelial-to-mesenchymal transition, and stemness. Furthermore, lncRNA-NEAT1 has been reported to affect the sensitivity to sorafenib and radiotherapy, as well as immune escape in HCC in vitro. Consistent with this role, aberrant expression of lncRNA-NEAT1 has been demonstrated in HCC and is associated with poor survival of HCC patients [11, 16–18]. Liu et al. demonstrated that high expression of lncRNA-NEAT1 in a Chinese population is an independent risk factor for poor survival of patients with HCC[19]. Furthermore, Ling et al. evaluated the expression level and clinical relevance of lncRNA-NEAT1 in HCC based on data from TCGA-LIHC and other HCC datasets from the Oncomine database, and found that lncRNA-NEAT1 is consistently up-regulated in HCC tumor tissues, though in the TCGA-LIHC dataset, lncRNA-NEAT1 was not significantly associated with overall patient survival but was significantly correlated to distant metastasis. Combined with data from in vitro experiments showing that knock-down of lncRNA-NEAT1 inhibits proliferation and induces apoptosis, Ling et al. suggested that lncRNA-NEAT1 promotes deterioration in HCC[20]. In the present study, we focused more on the up- and downstream regulatory mechanisms of lncRNA-NEAT1 in HCC rather, than on its tumor-promoting role under conventional conditions, and our results, therefore, may be increase the mechanistic understanding of lncRNA-NEAT1 in each of these prior investigations.
In general, the abnormal expression of lncRNA-NEAT1 in cancer cells is known to be caused by genetic alterations, transcription factors, DNA methylation, miRNAs, and RNA-binding proteins[21]. Emerging evidence indicates that hypoxia can modulate the expression of lncRNAs, including lncRNA-NEAT1. Up-regulation of lncRNA-NEAT1 has been found in breast cancer cells under hypoxic conditions[13] and in cardiomyocytes[22], further indicating that lncRNA-NEAT1 may be a hypoxia-responsive lncRNA. However, the response of lncRNA-NEAT1 to hypoxia in HCC has not been well elucidated. Therefore, we investigated changes in lncRNA-NEAT1 expression levels in HCC cells under hypoxic conditions (1% O2) or treatment with the hypoxia mimetic CoCl2. Our data verify that lncRNA-NEAT1 is induced by hypoxia. HIF-1α has been established as a predominant transcriptional regulator in response to hypoxia with the ability to binding to HREs and enhance expression of target genes, including lncRNAs[23]. By accessing the JASPAR database, we identified a potential putative HIF-1α-related HRE (5’-ACGTGC-3’) located in the promoter of lncRNA-NEAT1. Knock-down of HIF-1α eliminated the response of lncRNA-NEAT1 to hypoxia. Furthermore, the results of the ChIP and luciferase reporter assays supported the binding of HIF-1α to the promoter of lncRNA-NEAT1, suggesting that lncRNA-NEAT1 is transcriptionally induced by HIF-1α. These results provide evidence for the role of HIF-1α.and the HRE in the lncRNA-NEAT1 promoter as a mechanism that regulates lncRNA-NEAT1 expression under conditions of hypoxia.
We further demonstrated that overexpression of lncRNA-NEAT1 does not promote proliferation of HCC cells under normoxic conditions, possibly because the overactive proliferative properties of tumor cells under normal conditions may obscure the effect of lncRNA-NEAT1 overexpression by the “ceiling effect”. On the other hand, the overall proliferative rate is lower under conditions of hypoxia, and though lncRNA-NEAT1 expression is higher, exogenous overexpression in hypoxia conditions increased the proliferation rate. As expected, knock-down of lncRNA-NEAT1 inhibited the growth of HCC cells, which was evident both under normoxic and hypoxic conditions. Moreover, acute hypoxia inhibited growth and promoted apoptosis and cell cycle arrest of HCC cells, and HCC cells overexpressing lncRNA-NEAT1 grew relatively faster with less apoptosis and G1 phase arrest. Therefore, these results suggest that increased lncRNA-NEAT1 levels sustain the growth of HCC cells under hypoxic conditions.
LncRNAs regulate gene expression in cancers through distinct mechanisms. For instance, lncRNAs may regulate target genes by specific recruitment of transcriptional activators or suppressors; by acting as decoys that bind to and block transcription factors from target genes; or by recruiting chromatin-remodeling complexes as scaffolding proteins, thereby affecting target genes [24]. In tumor biology, lncRNAs primarily serve as ceRNAs that sponge tumor-promotive or tumor-suppressive miRNAs. Sponged miRNAs lose their regulatory effect on target mRNAs, which ultimately influences tumor progression[25]. Thus, we employed microarray analysis and a series of advanced online bioinformatics tools to identify potential miRNA-mRNA pairs that may interact with lncRNA-NEAT1 according to a ceRNA mechanism. The candidate miRNA-mRNA pairs were filtered according to their reported suppressive functions and predicted interactions with tumor-related genes, as well as by their down-regulated expression patterns in HCC. To obtain more evidence for the roles of candidate miRNAs in HCC, we performed pathway annotation analysis of predicted miRNA targets based on the KEGG database. The candidate miRNAs were determined to be involved in multiple tumor-related pathways. To filter HCC-related mRNA targets of lncRNA-NEAT1-miRNAs, 7 independent datasets from HCC patients with different backgrounds were used for integrated analysis using the RRA method, which strengthened the evidence. Eight candidate mRNAs were up-regulated in HCC tissues in the 7 datasets. To better explore the potential biological functions of the candidate mRNAs, GO enrichment analysis was performed. Enriched GO-terms, such as ‘regulation of cell growth’ and ‘positive regulation of cell cycle’, provided a potential explain of how lncRNA-NEAT1 may sustain growth of HCC cells by regulating the candidate mRNAs identified in our study. Though we could not perform a detailed analysis of all candidate miRNA-mRNA pairs in the present study, we selected UCK2, a hub target gene and impressive prognosis risk factor of HCC, for confirmation. The miRNAs that potentially regulate UCK2, including miR-199a-3p, miR-483-3p, miR-486-5p, miR-582-5p, and miR-129-5p, were considered. However, definite interactions were verified only for the lncRNA-NEAT1-miR-199a-3p-UCK2 axis. The results of ChIP and luciferase analyses confirmed the binding of lncRNA-NEAT1/miR-199a-3p and miR-199a-3p/UCK2, thus providing a downstream mechanism that may regulate lncRNA-NEAT1 function.
Previous studies have reported that miR-199a-3p acts as a tumor suppressor via various mechanisms in HCC. For instance, miR-199a-3p inhibits tumor growth in an animal model of HCC by modulating the mTOR pathway[26]. Giovannini et al. suggested that miR-199a-3p down-regulation is a common characteristic of HCC and that miR-199a-3p regulates E-cadherin expression through Notch1[27], and Guan et al. reported that miR-199a-3p represses tumorigenesis in HCC by targeting the ZHX1/PUMA pathway[28]. In the present study, the suppressive role of miR-199a-3p was confirmed and expression changes of lncRNA-NEAT1 were shown to induce alteration of miR-199a-3p in HCC cells. Furthermore, luciferase reporter and RIP assay demonstrated that lncRNA-NEAT1 sponges miR-199a-3p, which is consistent with our other bioinformatics and experimental data.
Functionally, UCK2 is a pyrimidine ribonucleotide kinase that catalyzes phosphorylation of uridine to uridine monophosphate and cytidine to cytidine monophosphate. Overexpression of UCK2 is regarded as an indicator of unfavorable prognosis in various cancers, including HCC, pancreatic cancer, breast cancer, and lung cancer[29–32]. However, few studies have revealed the detailed mechanisms underlying the regulation of UCK2. Zhou et al. found that UCK2 promotes metastasis via up-regulation of MMP2/9 expression and activation of STAT3 signaling[33]. The upstream mechanisms of UCK2, especially those involved with lncRNA/miRNA, had not been clarified prior to this study. Therefore, we confirmed the growth-promotive effect of UCK2 in HCC cells and demonstrated that UCK2 is regulated by lncRNA-NEAT1/miR-199a-3p. Most importantly, lncRNA-NEAT1 was shown to function under hypoxic conditions partly through miR-199a-3p/UCK2. Moreover, an animal model was used to further explore the role and regulatory relationship of lncRNA-NEAT1/miR-199a-3p/UCK2. As a limitation of this study, some miRNAs and mRNAs that may be also controlled by lncRNA-NEAT1 were not validated so that we could focus our efforts on validating the lncRNA-NEAT1/miR-199a-3p/UCK2 axis. These miRNAs or mRNAs should be investigated in future studies.