Targeting Long Non-Coding RNAs in Hepatocellular Carcinoma: Progress and Prospects

Hepatocellular carcinoma is the fifth-ranked cancer worldwide with a relatively low five-year survival rate. Long non-coding RNAs are a group of RNAs with remarkable aberrant expression which could act on multiple bioprocesses and ultimately impact upon tumor proliferation, invasion, migration, metastasis, apoptosis, and therapy resistance in cancer cells including hepatocellular carcinoma cells. In recent years, long non-coding RNAs have been reported to be indispensable targets in clinical target therapy to stop the growth of cancer and prolong the lifespan of patients with hepatocellular carcinoma. In this review, we enumerate the signaling pathways and life activities affected by long non-coding RNAs in hepatocellular carcinoma cells to illustrate the role of long non-coding RNAs in the development and therapy resistance of hepatocellular carcinoma.


INTRODUCTION
Liver cancers comprise diverse, histologically distinct primary hepatic neoplasms, which include hepatocellular carcinoma (HCC), intrahepatic bile duct carcinoma (holangio carcinoma), hepatoblastoma, bile duct cystadenocarcinoma, hemangiosarcoma, and epitheliod hemangioendothelioma (1). Among them, HCC is one of the most prevalent liver cancers worldwide (2,3). The main risk factors for HCC vary from region to region. Chronic hepatitis B virus (HBV) infection and AFB 1 exposure are major risk factors in most high-incidence regions like Asia and Africa. In contrast, HCV infection, excessive drinking, and diabetes/obesity/metabolic syndrome play a more important role in low-incidence areas, with the primary risk factor being HCV infection. In addition, some hereditary metabolic disorders such as hemochromatosis, a 1 antitrypsin deficiency, tyrosinemia, and several porphyrias also increase the risk (4)(5)(6)(7)(8). Till now, lots of treatments for HCC such as resection, ablation, transplantation, chemoembolization, and targeted drugs like sorafenib and lenvatinib are put into clinical practice (9)(10)(11). Despite all those treatments, the current five-year survival rate provided by the National Cancer Institute is still not satisfactory, so we still devote a lot of effort to create more effective methods (12). LncRNAs which are aberrant expressions are proved to be associated with tumorigenesis and poor prognosis of HCC. Given that, lncRNAs are promising potential biomarkers or targets of HCC.

LncRNAs and miRNAs
LncRNAs Acting as miRNA Sponges or miRNA Inhibitors MALAT1 (LncRNA metastasis-associated lung adenocarcinoma transcript 1) binds to and inhibits miR-143-3p expression to decrease ZEB1 (zinc finger E-box binding home box 1) (21). Also, MALAT1 acts as molecular sponge of miR-146b-5p and miR-204 to facilitate HCC cells (22,23). The UCA1 gene belongs to the HERV-H family. It contains the gag region, the proteasepolymerase region, but no envelope region. In three regions, the UCA1 full-length cDNA consists of an unusual number of stop codons that transcribe non-coding RNA (45,46). UCA1 facilitates FGFR1-ERK pathway by inhibiting expression of miR-216b (76). CCAT1 (The long non-coding RNA Colon Cancer Associated Transcript 1) contains two predicted let-7 targeting sites. It is reported that let-7 decreases tumor proliferation and induces apoptosis. Let-7 binds to CCAT1 but does not induce degradation of CCAT1. In other words, CCAT1 is physically related to let-7 and serves as a miRNA sponge for let-7. At the same time, CCAT1 regulates HMGA2 and c-Myc by competitively binding to let-7 (49). HOTAIR (Hox transcript antisense intergenic RNA), which is overexpressed in HCC tissues, enhances EMT by inhibiting miR-23b-3p, leading to malignant tumors of HCC and increased tumor metastasis (28)(29)(30). TUG1 is proved to act as a molecular sponge of miR-144.

LncRNAs Acting as Competing Endogenous RNAs of miRNAs
LncRNA-ATB (activated by TGF-b) which is up-regulated in HCC is activated by TGF-b and up-regulates ZEB1 and ZEB2 by competitive binding to the miR-200 family (77)(78)(79). ZEB1 gene encodes a zinc finger transcription factor that plays an important role in normal embryonic development which induce EMT (24). EMT converts cancerous epithelial cells into mesenchymal-like cells, confers migration and invasion properties, enables primary tumor cells to move, settles distant organs, and forms secondary tumors (18,19,80). As a result, the silence of ZEB1 hinders the metastasis and invasion of HCC through EMT inhibition. TUG1 promotes HCC development by competing with miR-132 to combine sonic hedgehog (Shh) as well as Kruppel-like factor 2 (KLF2) to combine polycomb repressive complex 2 (PRC2) (58, 59). Overexpression of MEG3 competitively inhibits miR-664, thereby releasing the inhibitory effect of miR-664 on ADH4 and promoting the expression of ADH4 (81).

MiRNAs Acting as lncRNA Inhibitors
HOTTIP (HOXA transcript at the distal tip) is mainly present in the nucleus, binds to AGO2 in the nucleus, and is regulated by some miRNAs. The level of HOTTIP is significantly decreased in cells when miR-192, miR-204, and miR-125b are overexpressed (82). MiR-192 and miR-204 inhibit the activity of HOTTIP through their target molecules such as DHFR, ZEB2, BCL2, and so on, thereby achieving the purpose of inhibiting the activity of HCC cells (55).
HULC reduces Sirt1 and COX2 degradation by elevating the expression of USP22 (83). TUC338 (Transcribed ncRNA encoding uc.338) post-transcriptionally regulates plasminogen activator inhibitor-1 RNA binding protein (PAI-RBP1), occupying a genomic region rich in unique or known motifs homologous to the tumor suppressor Pax6 and p53 (48). C-Myc directly binds to the promoter of CCAT1 and promotes HCC development (50). Activation of TCF/LEF by b-catenin is one of the most common molecular changes in HCC cells as well as a general regulator of stem cell self-renewal, tumorigenicity and tumor progression. DANCR (Differentiation antagonizing nonprotein coding RNA) regulates the stability of above molecular changes to affect tumor proliferation (51,84). Linc00152 (Long intergenic non-protein coding RNA 00152) prevents E-cadherin expression by interacting with EZH2 and promotes EMT in HCC cells (61). MEG3 (Maternally expressed gene 3) promotes p53 expression and inhibits MDM2 expression, and increasing p53 also inhibits the expression of MDM2. So that the ubiquitination of P53 by MDM2 is prevented, promoting apoptosis and inhibiting tumor proliferation (62,63,85,86).

Interactions Within lncRNAs
It is reported that MALAT1/HULC is positively correlated with the expression of TRF2 in human hepatocellular carcinoma tissues. MALAT1 and TRF2 are highly expressed in HCC tissues and are positively correlated. The increased TRF2 binds to HULC and MALAT1 to form a complex, which is loaded into the telomere region of the chromosome. Therefore, the telomeres are greatly extended, leading to the rapid growth of HCC stem cells (25). Overexpression of HULC (highly up-regulated in liver cancer) prevents PTEN and miR15a, which leads to high expression of LC3I and LC3II (autophagy marker) and more autophagy in hepatoma cells. HULC inhibits PTEN by autophagy and P62-mediated ubiquitin-protein system and finally activates the AKT-PI3K-mTOR pathway to promote cell growth, colonyforming ability, and cell growth in vivo (41).

LncRNAs as Tumor-Suppressive Genes in HCC
MEG3 expression is down-regulated in both HCC cell lines and tissues. Re-expression of MEG3 in HCC cells significantly reduces anchorage-dependent and independent cell growth and induces apoptosis (64,87). Also, adenosine can resist HCC through up-regulating the expression of MEG3 (88).
Phosphatase and tensin homolog pseudogene 1 (PTENP1) is a pseudogene of the tumor suppressor gene phosphatase and tensin homolog deleted on chromosome ten (PTEN) (67,68). It is shown that PTENP1 and PTEN are down-regulated in HCC cells. Over-expression of PTENP1 and PTEN in HCC cells can inhibit the oncogenic PI3K/AKT pathway, inhibit cell proliferation, migration, invasion, and induces autophagy, apoptosis, and inhibition of angiogenesis.

LncRNAs as Oncogenes in HCC
lncRNA-ATB promotes HCC colonization by inducing autocrine induction of IL-11 and activating STAT3 signaling (20). Plasma Linc00152 can be used as a potential non-invasive biomarker to predict the diagnosis of HCC (42). MALAT1 is up-regulated in HCC and plays an oncogenic role through activating the Wnt pathway and inducting oncogenic splicing factor SRSF1 to activate the mTOR pathway and resist apoptosis (26). HOTAIR with a length of 2,158 bp is remarkably associated with poor prognosis of HCC. It affects the histone H3 tri-methylated at lysine 27 (H3K27me3) by recruiting poly bulking inhibitor complex 2 (PRC2 complex) at the 5′ end, so that LSD1 (lysinespecific demethylase 1)/CoREST (RE1-silencing transcription factor co-repressor)/REST (RE1-silent transcription factor) complex promotes histone H3 Lysine 4 demethylation, eventually leading to gene silencing (37)(38)(39). UCA1 (urothelial cancer associated 1) is overexpressed in HCC, making it a potential biomarker to detect progression and prognosis in patients with HCC (89). Decreasing the expression level of UCA1 inhibits the proliferation, migration, and invasion of HCC cells and induces apoptosis. Moreover, bioinformatics analysis indicates that UCA1 may significantly disrupt the hippo signal pathway (47). TUC338 is a super-conservative lncRNA that contributes to the growth of transformed cells in HCC (HCC). TUC338 functions in a manner similar to transcription factors to regulate cell proliferation and transform cell growth in HCC. Overexpression of CCAT1 in hepatoma cells promotes proliferation, invasion, and metastasis of tumor cells (50). As a newly discovered cancer-associated lncRNA, HOTTIP is located at the 5′ end of the HOXA gene. Overexpression of HOTTIP could promote cell proliferation, migration, and invasion of HCC cells (54,56,(90)(91)(92). Linc00152 is up-regulated in the human HCC cell line. Overexpressed Linc00152 in HCC cells increased cell proliferation and invasion. Knocking out Linc00152 inhibits the mTOR signaling pathway. The underlying mechanism is that expression of Linc00152 increases EpCAM levels, leading to activation of the mTOR signaling pathway and causing proliferation of HCC (49). DANCR positively regulates proliferation in cells by regulating miR-634, miR-496 miR-33a-5p, CDKN1A (cell cycle inhibitor p21), etc, indicating that it may be a carcinogenic lncRNA, and plays potential roles as an adenocarcinoma (ADC) biomarker and therapeutic target (52,53). The combination of sorafenibinduced enhanced tumor growth inhibition and overexpression of RASAL1 in tumor xenografts suggests that the TUC338/ RASAL1 axis may be a potential therapeutic target for current HCC treatment (93).

LncRNAs and Therapeutic Sensitivity in HCC
LncRNAs not only influence HCC proliferation, invasion, and migration through specific cell signaling pathways and molecules, but also induce HCC therapy resistance, which covers chemotherapy resistance and radiotherapy resistance. It has been found that lncRNA ROR reduces the sensitivity of HCC to radiotherapy (94). For chemotherapy, the most commonly used chemotherapy drugs for HCC include sorafenib, oxaliplatin, 5-fluorouracil, cisplatin, etc (95). Figure 1 is a schematic diagram of the relationship between lncRNAs and chemotherapeutic resistance. Sorafenib SNHG1, SNHG3, and SNHG16 in the small nucleolar RNA host gene (SNHG) family are related to sorafenib resistance in HCC cells (96)(97)(98). SNHG1 serves miR-21 as a mediator, as well as SNHG3 to produce positive feedback on the downstream Akt signaling pathway, and ultimately induces sorafenib resistance (99). Coincidentally, nuclear-enriched abundant transcript 1 (NEAT1) also acts on the downstream Akt signaling pathway through the miR-335/c-Met axis to induce sorafenib resistance (100). Other lncRNAs related to sorafenib resistance in HCC cells include linc00160, FOXD2-AS1, MALAT1, H19, ROR, etc (101). Linc00160 promotes phosphoinositide-3-kinase regulatory subunit 3 (PIK3R3) to induce sorafenib resistance by inhibiting miR-132 (102). FOXD2-AS1 acts as a competitive endogenous RNA of miR-150-5p to increase sensitivity to drugs (103). MALAT1 and H19 have similar mechanisms when inducing HCC resistance, which both promote EMT through intermediary molecules to induce resistance (104-106).

DISCUSSION
LncRNA-related clinical trials have not been conducted on hepatocellular carcinoma, but a clinical trial from Wuhan Union Hospital on lncRNA as a potential target for lung cancer diagnosis is underway. This clinical trial is mainly based on the identification of early lung-cancer-specific exosomal lncRNA biomarkers to improve the diagnosis rate of early lung cancer (113). Also, a clinical trial which is related to HOTAIR and thyroid cancer is being processed (114).
In this review, we discussed a variety of lncRNAs that are proved to involve in HCC (Figures 2-4). In the past few years, a series of studies have shown the essential role of lncRNAs on cell proliferation, invasion, migration, and therapy resistance through diverse signaling pathways and molecules. Based on current studies, lncRNAs are expected to be a marker for tumor diagnosis, prognosis, and expected therapeutic effects, while lncRNA-targeted drugs still have a long way to go. Especially, FIGURE 1 | Diagram of the mechanism of lncRNA mediating chemotherapy resistance. Part of the mechanism is related to microRNAs, and part is related to Akt pathway. The promotion arrow in the panel refers to the promotion of drug resistance to the corresponding chemotherapeutic drug, and the inhibition arrow refers to the increase of the sensitivity to the chemotherapy drug. considering that lncRNAs are related to the sensitivity of HCC radiotherapy and chemotherapy, lncRNAs can be used as a molecular marker to predict the clinical treatment effect of therapy treatment, and it can also be used as a target in conjunction with chemotherapy or radiotherapy to increase treatment sensitivity. Our ideal expectation for lncRNA-targeted drugs is that lncRNA-targeted drugs will serve as independent targeted drugs to treat HCC, or they will be used as adjuvant drugs to increase the efficacy of existing chemotherapeutic drugs. First, given that more than a single lncRNA is confirmed to function in HCC, to make sure which type plays a leading role in development of HCC is fatal for further research. Not only in HCC, in tumors of various tissue sources, the tissue specificity of lncRNAs is also indispensable. If future research studies prove that it is unfeasible to aim at a single lncRNA to achieve better effect, perhaps using drug that has a targeted effect on several major lncRNAs with a certain commonality based on clinical research and individual differences in patients, or employ these drugs as adjuvants can be the solutions. Also, existing studies have shown that almost each lncRNA found is involved in a variety of cell signaling pathways, but these studies are far from sufficient. Moreover, the functions of lncRNAs remained undetected, which is also a burning question. Finally, due to the structural particularity of lncRNA molecules and the characteristics of being easily degraded, how to target lncRNA molecules is also an urgent problem to be solved. With further studies, the clinical transformation of lncRNA is bound to be more mature, bringing more effective treatments to patients suffering from HCC, and contributing to the fight against cancer in humans.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/supplementary material. Further inquiries can be directed to the corresponding author.

AUTHOR CONTRIBUTIONS
XL and XX wrote the review article. RW reviewed the manuscript. All authors contributed to the article and approved the submitted version.