MAZ-LINC00645-GP73 Axis Promotes Hepatocellular Carcinoma Proliferation and Metastasis

BACKGROUND: Long non-coding RNAs (lncRNA) have been shown to play important roles in the development and progression of hepatocellular carcinoma (HCC). In this report, we examined the role of lncRNA LINC00645 in HCC. MATERIAL AND METHODS: Based on public databases and integrating bioinformatics analyses, the overexpression of LINC00645 in HCC tissues was detected and further validated in a cohort of liver tissues. A series of in vitro and in vivo functional experiments were executed to investigate the role of LINC00645 in the carcinogenesis and development of HCC. Comprehensive transcriptional analysis, chromatin immunoprecipitation (ChIP) assay, dual-luciferase reporter assay and western blot etc. were performed to explore the molecular mechanisms underlying the functions of LINC00645. RESULTS: LINC00645 was signicantly upregulated in HCC cell lines and HCC tissues, which was correlated with poor prognosis in HCC patients. LINC00645 knockdown remarkably suppressed tumor growth in vitro and in vivo. Mechanistically, LINC00645 could competitively bind with miR-141-3p to prevent the degradation of its target gene GP73, which acts as a tumor-promoter in HCC. Furthermore, the ChIP assay showed that the transcription factor MAZ could bind to the LINC00645 promoter and increase its transcription. CONCLUSIONS: Collectively, this study demonstrated that LINC00645 plays a critical regulatory role in hepatocellular carcinoma cells and LINC00645 may serve as a potential diagnostic biomarker and therapeutic target of HCC. Thus, targeting MAZ/LINC00645/miR-141-3p/GP73 signaling axis may prevent the progression of HCC. Cell proliferation was measured by the CCK-8 method and colony formation method. Briey, in the CCK-8 experiment, 1 × 10 3 cells were cultured in 96-well plates at 37˚C. 96 well plates were incubated with 10 µl of CCK-8 solution per well for 1 hour. Cell proliferation curves were drawn by measuring the absorbance at 450 nm at each indicated time point. The cell proliferation curves were plotted by measuring the 450 nm absorbance at each indicated time point. Experiments were performed in triplicate. For the colony formation assay, cells were exposed to the indicated treatments, were seeded in 6-well plates and were cultured for 2 weeks. Cell colonies were washed with phosphate-buffered saline (PBS), xed with 4% paraformaldehyde, stained with 0.1% crystal violet and imaged using an optical microscope. expression in cells after transfection with sh-NC or sh-LINC00645 detected by RT-PCR.


Introduction
Hepatocellular carcinoma (HCC) is one of the most common malignancies worldwide and ranks as the third most common cause of cancer-related death [1]. Recently, great advances have been gained for the pathogenesis, diagnosis and treatment of HCC. However, the molecular mechanisms underlying HCC pathogenesis have not been fully understood and the survival of patients remains poor [2,3]. Hence, much hope is placed in understanding the pathogenesis and exploring a novel strategy for the treatment of HCC.
Large amounts of studies report that dysregulation of oncogenes and tumor suppressor genes contribute to HCC tumorigenesis and progression, but most of them focus on protein-coding genes [4]. Only 2% of the human genome accounts for protein coding genes, while about 70% of the genome is identi ed as non-coding RNAs (ncRNAs) due to the great progressions of genome and transcriptome sequencing [5]. Long noncoding RNAs (lncRNAs), transcripts longer than 200nt that lack an extended open reading frame and thus do not code for proteins, have emerged as major regulators of a wide range of cellular processes [6]. In human cancer, the aberrant expression of lncRNAs has been associated with tumor development and progression. Moreover, lncRNAs regulate malignant behaviors of cancer cells [7], such as proliferation, apoptosis resistance, migration, invasion and drug resistance. Aberrant expression of lncRNAs has been frequently observed in cancers. Mechanistically, lncRNAs may in uence the function of transcriptional complexes, modulate chromatin structures by serving as scaffolds between proteins, or act as microRNA sponges [8]. Long intergenic non-protein coding RNA 645 (LINC00645) was rst annotated as a long intergenic noncoding RNA (lincRNA) on chromosome 14q12. Recent studies have reported that LINC00645 plays an oncogenic role in glioma and it may serve as a prognostic biomarker and a potential therapeutic target for the treatment of glioma in humans [9]. However, the function and mechanism of LINC00645 as ceRNAs in HCC remain unclear.
Golgi protein 73 (GP73), a Golgi glycoprotein, is mainly expressed in bile duct epithelial cells, whereas it is rarely or seldom expressed in hepatocytes. [10,11]. The serum GP73 levels increase in patients with viral hepatitis B or C, or other chronic liver diseases, which are superior to those of AFP [12,13]. GP73 is a serum biomarker of liver brosis [14]and a potential biomarker for HCC [15]. However, the upstream regulatory factors of GP73 in HCC progression and metastasis have not been fully identi ed, this study was planned to ll this gap in literature.
Here, this study demonstrated that LINC00645, which was upregulated in HCC tissues compared with adjacent non-tumor tissues and that elevated LINC00645 levels were associated with poor prognosis in HCC patients. Moreover, the knockdown of LINC00645 signi cantly inhibited the malignant proliferation, invasion and metastasis of HCC cells. Mechanistically, LINC00645 promotes tumor proliferation by sponging miR-141-3p to regulate GP73. Consequently, establishing a new regulatory axis of the "MAZ-LINC00645-miR-141-3p -GP37" could better explore the cancer-promoting mechanism of LINC0645 in HCC.

Materials And Methods
Clinical Sample and Tissue Specimen Acquisition Hepatocellular carcinoma specimens and adjacent noncancerous tissues were obtained from Harbin Medical University Cancer Hospital (HMUCH), and patients with a histological diagnosis of hepatocellular carcinoma who had received neither chemotherapy nor radiotherapy before surgical resection were recruited for the present study between 2009 and 2019. This study conformed to clinical research guidelines and was approved by the research ethics committee of Harbin Medical University Cancer Hospital. We obtained written informed consent from all patients.

Cell Culture
The human immortalized normal hepatocyte cell line LO2 and HCC cell lines HepG2, Hep3B, Huh7 and SMMC-7721 were obtained from the Chinese Academy of Sciences Cell Bank and Cellbio (China) and were cultured according to the suppliers' instructions. All cell lines were cultured according to instructions.

Plasmids and Transfection
LINC00645 and sh-LINC00645 and controls were constructed by Genechem (Shanghai, China).
Concentrated viruses were used to infect 5 × 10 5 cells in a 6-well plate with 4-6 µg/ml Polybrene. The infected cells were then subjected to selection with 1 µg/ml puromycin (Cat#540411, Calbiochem, USA) for two weeks. Stable over-expression cell lines or knockdown cell lines were identi ed using qRT-PCR or western blotting. The miR-141-3p mimic and inhibitor were purchased from Ribobio (Guangzhou, China For the quanti cation of gene expression, we used the 2 −ΔΔCt method. GAPDH or U1 expression was used for normalization. The primer sequences were synthesized by Genepharma (Shanghai, China). All the primer sequences are available in Table S2.

Subcellular Fractionation
Nuclear/cytoplasmic fractionation was performed by NE-PER™ Nuclear and Cytoplasmic Extraction Reagents (Cat#78835, Thermo Fisher) according to the manufacturer's protocols. U1 was used as a nuclear control, while GAPDH was used as a cytoplasmic control.

Cell Proliferation and Colony Formation Assay
Cell proliferation was measured by the CCK-8 method and colony formation method. Brie y, in the CCK-8 experiment, 1 × 10 3 cells were cultured in 96-well plates at 37˚C. 96 well plates were incubated with 10 µl of CCK-8 solution per well for 1 hour. Cell proliferation curves were drawn by measuring the absorbance at 450 nm at each indicated time point. The cell proliferation curves were plotted by measuring the 450 nm absorbance at each indicated time point. Experiments were performed in triplicate. For the colony formation assay, cells were exposed to the indicated treatments, were seeded in 6-well plates and were cultured for 2 weeks. Cell colonies were washed with phosphate-buffered saline (PBS), xed with 4% paraformaldehyde, stained with 0.1% crystal violet and imaged using an optical microscope.

Ethynyl Deoxyuridine (EdU) Incorporation Assay
A Cell-Light™ EdU Apollo567 In Vitro Kit (Catalogue Number C10310-1, RiboBio, China) was used to perform the EdU proliferation assay according to the manufacturer's instructions as previously described.

Cell Migration and Invasion Assay
Cell migration and invasion assays were performed as described previously. In a wound healing assay, cells were seeded in 6-well plates to form a con uent monolayer. Then, a scratch wound was induced by a pipette tip. Photographs of cells migrating to the scratched area were taken and the data were shown as a percentage of the control group. In the transwell invasion and migration assay, cells (5 × 10 4 cells per well) were seeded in the upper chambers of the transwell plate and placed in FBS-free medium with or without matrix gel. After 24 hours of incubation, the cells that invaded/migrated to the lower surface of the membrane were xed, stained with crystal violet and observed using an inverted microscope.
Chromatin Immunoprecipitation (ChIP) ChIP assays were performed using a commercially available kit (Beyotime) according to the manufacturer's protocol. Brie y, cells were cross-linked with 1% formaldehyde and were sonicated on ice to create 200-500 bp fragments. Stained chromatin was cultured overnight with an anti-MAZ antibody (Novus Biologicals, NB100-86984, 1:50) or IgG (Cell Signaling Technology, Cat#3900)as an isotype control. The precipitated chromatin DNA was recovered and analyzed by qRT-PCR. The primer sequences are shown in Table S2 RNA Immunoprecipitation (RIP) RIP assay was performed using the Magna RIP RNA-Binding Protein Immunoprecipitation (RIP) Kit (EMD Millipore). HepG2 cells lysis solution (Sigma Aldrich Chemical Company, USA) was added to 3 mg of cells and left to incubate at 4 ˚C for 1 h. Cells were centrifuged at 12,000 g at 4 ˚C for 10 min in order to collect the supernatant, which was employed for RIP experiments using an anti-AGO2 antibody (Abcam, Cambridge, MA, USA) according to the manufacturer's instructions.. The RNA fraction isolated by RIP was subjected to qRT-PCR analysis to identify the direct binding between linc00645 and miR-141-3p. These experiments were repeated three times.

Immunohistochemical (IHC) Analysis
The para n-embedded sections were dewaxed in xylene and rehydrated in alcohol. Endogenous peroxidase was blocked by 3% H 2 O 2 , and microwave heating was performed for antigen retrieval. After blocking nonspeci c antigen binding with 5% BSA at 37 °C for 1 h, the sections were incubated with a speci c primary antibody against Ki67 (Abcam, ab15580, 1:1000), at 4 °C overnight. After incubating with the corresponding secondary antibodies ( Abcam, ab205718) at 37 °C for 1 h, the sections were stained with diaminobenzidine and counterstained with hematoxylin. Representative images were taken using an Olympus light microscope Luciferase Reporter Assays HepG2 cells were seeded at 5 × 10 4 cells/well in 24-well plates and were cultured overnight. On the next day, the cells were cotransfected with pmirGLO-LINC00645-WT, pmirGLO-LINC00645-MUT, pmirGLO-GP73-3'UTRWT, pmirGLO-GP73-3'UTR-MUT reporter plasmids (Genechem, Shanghai), NC mimics or miR-141-3p mimics. Twenty-four hours posttransfection, cells were lysed using passive lysis buffer (Promega), and the luciferase activity was measured by a GloMax20/20 Luminometer (Promega) using the Dual-Luciferase Reporter Assay System (Promega) and was normalized to the Renilla luciferase activity.

Tumor Xenograft Model
The animal study protocol was approved by the Institutional Animal Care and Use Committee of Harbin Medical University Cancer Hospital and was performed in accordance with the Guide for the Care and Use of Laboratory Animals (Institute of Laboratory Animal Resources, Commission on Life Sciences, National Research Council). Mouse xenograft models were established using 4-week-old BALB/c nude female mice. HepG2 cells (5 × 10 5 per injection) that were transfected with sh-LINC00645 and sh-control, respectively, were implanted into the mice via subcutaneous injection. Tumor volumes were measured every 3 days after being apparently observed and calculated with the following formula: Volume = (length × width2)/2. After 4 weeks, all mice were sacri ced under anesthesia.

Statistical Analysis
Data are expressed as mean ± SEM. And all data represent at least three independent experiments.
Statistical analysis was performed using unpaired Student t-test or 1-way ANOVA followed by Tukey post hoc analysis. P < 0.05 was considered statistically different and indicated by *P < 0.05, ** P < 0.01, ***P < 0.001.

LINC00645 Expression Is Up-regulated in HCC Cell lines, HCC Tissues and Associated With Poor Prognosis
To investigate the expression of LINC00645 in HCC, we search TCGA data from starBase V3.0. The results showed that LINC00645 in HCC tissues was signi cantly higher than that in normal liver tissues ( Figure.1A, p < 0.01). To further validate this result, we investigated the expression of LINC00645 in 40 paired HCC tissues and adjacent non-tumor tissues from Harbin Medical University Cancer Hospital (HMUCH). These results showed that the expression of LINC00645 was markedly increased in HCC tissues compared with adjacent non-tumor tissues ( Figure.1B, p < 0.001). Meanwhile, we observed that the expression of LINC00645 also observed in HCC cell lines (HepG2, Hep3B, SMMC-7721 and Huh7) compared to LO2 cells ( Figure.1C, p < 0.001). Moreover, the patients were divided into high (n = 23) or low (n = 17) LINC00645 level group according to the mean level of LINC00645 in cancer tissues. The patients with high LINC00645 level displayed poor overall survival (OS) compared with those in low expression group (p = 0.0389) (Figuer.1D). To further explore the clinical characteristics connected with LINC00645 in HCC, we examined the correlation of LINC00645 expression with patients' clinicopathological characteristics in HCC. As illustrated in Table 1, patients with high LINC00645 expression exhibited a dramatically association with tumor size (P = 0.014), TNM stage (P = 0.012), venous invasion (P = 0.08), and lymph node metastasis (P = 0.017), while there were no signi cantly association between LINC00645 expression with Gender, age, liver cirrhosis, and alpha fetoprotein (AFP) level (P > 0.1). The results of nuclear/cytoplasmic RNA fractionation from the subcellular distribution assay con rmed that LINC00645 was mainly located in the cytoplasm (Figuer1E and F). Collectively, these ndings revealed that high expression of LINC00645 may be involved in tumor cell proliferation, invasion and LINC00645 may be an oncogene in HCC. To determine the biological function of LINC00645 in HCC cells, short interference shRNAs against human LINC00645 (Sh-1 and Sh-2) were applied to knockdown LINC00645 expression, whereas fulllength recombinant plasmid (Lv-LINC00645) with LINC00645 was used to increase LINC00645 expression. The knockdown and over-expression e ciency were con rmed by RT-PCR ( Fig. 2A, p < 0.001 and Fig. 2D, p < 0.001). CCK-8 and colony formation assays revealed that depletion of LINC00645 inhibited the growth and proliferation of HepG2 and Hep3B cells ( Fig. 2B-C, p < 0.001), while LINC00645 over-expression promoted cell growth and proliferation in HepG2 cell lines ( Fig. 2E and F, p < 0.001). EdU incorporation assays also indicated that knocking down LINC00645 prominently suppressed the growth of HepG2 and Hep3B cells (Fig. 2G, p < 0.01). Furthermore, we then investigated the role of LINC00645 in the motility of HCC cells. The results showed that LINC00645 knockdown signi cantly impaired the migration and invasion of HCC cells (Fig. 2H, p < 0.001). Given that epithelial-mesenchymal transition (EMT) is one of the major mechanisms for cancer metastasis, we further evaluated the effect of LINC00645 on EMT-related markers [9]. Western blot analysis showed that LINC00645 knockdown could increase the expression of epithelial markers (E-cadherin) and decrease the expression of mesenchymal markers (N-cadherin) (Fig. 2I), indicating that LINC00645 could regulate the EMT process to modulate HCC progression. Taken together, these data show that LINC00645 promotes HCC cell growth and invasion of HCC in vitro.

LINC00645 Acted as a Sponge For MiR-141-3p in HCC Cells
Recently, many lncRNAs have been reported to function as competing endogenous RNAs (ceRNAs) in modulating the expression and biological functions of miRNAs [16,17]. Since LINC00645 was distributed predominantly in the HCC cell cytoplasm, we hypothesized that LINC00645 might act as a miRNA sponge to prevent miRNAs from binding with their target mRNAs. Through prediction of the bioinformatics database microrna.org (http:// microrna.org/microrna/getMirnaForm.do) and LncBook (https://bigd.big.ac.cn/lncbook), we found that there were potential binding sites between LINC00645 and miR-141-3p. To validate the above theory, we subcloned the wild-type (LINC00645-WT) and mutated (LINC00645-MUT) miR-141-3p binding sites into dual-luciferase reporters (Fig. 3A). The luciferase assay showed that transfection of miR-141-3p mimics signi cantly reduced the relative luciferase activity of LINC00645-WT-treated HCC cells, but did not affect that of LINC00645-MUT-treated HCC cells (Fig. 3B). The AGO2 immunoprecipitation assay showed that the AGO2 antibody was able to pull down both endogenous LINC00645 and miR-141-3p ( Fig. 3C and D).To determine the relationship between LINC00645 and miR-141-3p, we used RT-qPCR assay to evaluate miR-141-3p expression in HCC patients. Interestingly, the level of miR-141-3p signi cantly decreased in HCC tissues compared with non-tumor liver tissues (Fig. 4E, p < 0.01). Analysis of our HCC database revealed that LINC00645 was inversely correlated with the expression of miR-141-3p in the HCC tissues (Fig. 3F). In contrast to LINC00645, miR-141-3p expression in HCC cell lines was much lower than that in LO2 cells (Fig. 3G). In addition, miR-141-3p was upregulated when the HCC cells were transfected with LINC00645 shRNAs (Fig. 3H). However, LINC00645 was down-regulated when the HCC cells were transfected with miR-141-3p mimics (Fig. 3I).
Taken together, the above data suggested that LINC00645 acts as a molecular sponge for miR-141-3p in HCC cells.

GP73 is the Direct Target of MiR-141-3p in HCC Cells
By using starBase V3.0, miRWalk, miRPathDB, and TargetScan databases, we found that Golgi protein 73 (GP73) was a potential target of miR-141-3p (Fig. 4A). Recent studies reported that GP73, a golgi glycoprotein, was reported to be oncogenes in HCC [18,19]. According to this inference, we performed the luciferase reporter assays and con rmed that the repression of luciferase activity was diminished in HepG2 and Hep3B cells contransfected with the miR-141-3p mimics and GP73-WT 3'-UTR (Fig. 4B).
Meanwhile, we examined a set of HCC tumors and paired adjacent normal tissues from HCC patients. qRT-PCR results showed that GP73 expression was signi cantly increased in tumors compared to adjacent non-tumor tissues (Fig. 4C). Then, we detected the mRNA levels of GP73 after miR-141-3p overexpression in HepG2 and Hep3B cells. We found that miR-141-3p over-expression reduced the level of GP73 mRNA (Fig. 4D-E). Furthermore, GP73 protein expression was signi cantly down-regulated by miR-141-3p mimics in both HepG2 and Hep3B cells (Fig. 4F). In short, these data implied that GP73 was a direct target of miR-141-3p in HCC cells.

LINC00645-miR-141-3p-GP73 Axis Promotes HCC cells Proliferation and Metastases
In light of the above ndings, we hypothesized that LINC00645/miR-141-3p/GP73 axis might play a role in the progression of HCC. Restored experiments were performed introducing the miR-141-3p inhibitor into cells of LINC00645 knockdown. Remarkably, a reduction in the GP73 mRNA (Fig. 5H, p < 0.01) and protein amount (Fig. 5I, p < 0.01) in HepG2 and Hep3B cells, as a result of LINC00645 knockdown, were countered when the miR-141-3p inhibitor was co-transfected. The recovery of GP73 by the mediation of miR-141-3p inhibitor inhibited the effects of the LINC00645 silencing on HepG2 and Hep3B cells proliferation ( Fig. 5A and B, p < 0.01), colony formation ( Fig. 5C and D, p < 0.001), migration and invasion ( Fig. 5E and F, p < 0.001). Western blot analysis showed that the expression of E-cadherin was decreased and the Ncadherin was increased (Fig. 5G), Taken together, the LINC00645 could play an oncogenic role by miR-141-3p/GP73 axis in HCC cells.

The Downregulation of LINC00645 Inhibited HCC Tumor Growth In Vivo
To elucidate the biological roles of LINC00645 in HCC tumorigenesis in vivo, we inoculated nude mice with HepG2 cells that stably expressed lentiviral shLINC00645 to suppress LINC00645 expression (Fig. 6A). Tumor xenografts with down-regulated LINC00645 showed markedly reduced volumes and weights compared to control xenografts ( Fig. 6B and E). Tumor tissues were harvested for qRT-PCR analysis of LINC00645. We con rmed that lower expression of LINC00645 was detected in tumor tissues arising from the LINC00645 knockdown group compared to the control group (Fig. 6F). Moreover, immunohistochemistry (IHC) assays con rmed that LINC00645 knockdown caused increased Ki67 expression (Fig. 6G), indicating reduced cell proliferation. Besides, GP73 and miR-141-3p expression in the tumor tissues were detected by the qRT-PCR analysis. GP73 expression was observed to be reduced (Fig. 6I), whereas miR-141-3p expression was enlarged (Fig. 6H). Similarly, GP73 protein was sharply down-regulated in the sh-LINC00645 group compared with the sh-NC group (Fig. 6J) by Western blot analysis. Together, these results suggested that knockdown LINC00645 signi cantly suppressed HCC tumorigenesis in vivo.

MAZ Transcriptionalyl Regulates LINC00645 Expression in HCC Cells
To examine the possible transcription factor-binding sites in promoter loci of LINC00645 in hepatocellular cancer, we searched the TRAN SFAC (http://gene-regulation.com/) and JASPAR (http://jaspar.genereg.net/) databases to identify transcription factors that may regulate LINC00645. The transcription factor Myc-associated zinc nger protein (MAZ) was predicted by both the TRAN SFAC and JASPAR databases with high scores (Supporting Information, Table S3). The predicted binding sites of MAZ in the LINC00645 promoter sequence are illustrated in Fig. 7A. To determine the differences between the expression levels of MAZ in HCC and normal tissues, the mRNA levels of MAZ in HCC and normal tissues were analyzed based on TCGA and GTEX data in the GEPIA platform and HMUCH (Harbin Medical University Cancer Hospital) database respectively. The results from different databases showed to be similar results from each other. The mRNA expression levels of MAZ were up-regulated in patients with HCC in GEPIA and HMUCH database, respectively ( Fig. 7B and C). To explore whether MAZ regulated LINC00645 expression, we knocked down MAZ by transfecting with siRNA in HepG2 cells, which led to a signi cant decrease in LINC00645 expression ( Fig. 7D and E). Furthermore, the over-expression of MAZ signi cantly elevated LINC00645 expression in HepG2 cells (Fig. 7Fand G). Moreover, ChIP assays showed that the LINC00645 promoter was speci cally pulled down by a MAZ-speci c antibody but not the control antibody (Fig. 7H). Taken together, these ndings suggested that MAZ is a bona de transcriptional activator of LINC00645.

Discussion
Emerging evidence indicates key regulatory roles of lncRNAs in HCC [20,21]. However, evidence of lncRNAs with clinical prognostic value is still limited. Ideally, in addition to exhibiting HCC-speci c expression patterns, lncRNAs should be demonstrated to regulate clear mechanistic pathways that drive the growth of HCC to support their potential use as a therapeutic target during clinical treatment [22]. By using clinical specimens, in vitro hepatoma cell lines, and xenograft/orthotopic mice models, we demonstrated that a novel lncRNA, LINC00645, plays a key role in HCC growth. Our data showed that over-expression of LINC00645 could be used to predict the prognostic outcome of HCC patients' overall survival ratios. Alteration of endogenous cellular LINC00645 expression in uenced the sensitivity of both hepatoma cells and HCC tumors. In addition, LINC00645 promoted HCC growth partly via binding to miR-141-3p to accelerate GP73 expression in HCC progression. Therefore, our study provides clinical and mechanistic data to support the role of this lncRNA in HCC.
Previously de ned mechanisms of lncRNAs in cancer growth include regulation of viability, proliferation, immortality, mobility and angiogenesis [23][24][25]. Additionally, HCC-related lncRNAs that have previously been characterized as target mRNAs [26], promoter regions [27,28], or proteins [26] to exert their regulatory effects on HCC cells. Our study provides a novel lncRNA LINC00645 to regulate cancer growth and to serve as a potential prognostic marker.
MiRNAs are usually acted as the regulatory targets of lncRNAs that involved in the development of human cancers. MiRNAs regulate target genes in two different ways: 1) At the transcriptional level, miRNAs can activate or suppress target genes by binding to their promoters [29,30]; 2) At the posttranscriptional level, miRNAs can repress target gene translation or cause target mRNA degradation by binding to their 3'UTR regions [31,32]. In addition, miR-141-3p has been veri ed to suppress the growth and metastasis of tumors in multiple diseases in hepatocellular carcinoma [33] and its expression was substantially reduced in HCC cells and tissue samples [34]. In this study, LINC00645 primarily localized to the cytoplasm of HCC cells, suggesting that LINC00645 might exert the functions of ceRNA. Next, LncBook (https://bigd.big.ac.cn/lncbook) was employed as a means of predicting LINC00645 targeting miRNAs, with the identi ed miR-141-3p being of particular interest based on its biological function in HCC. MiR-141-3p was down-regulated in HCC tissues, serving as a tumor suppressor that targets several genes to inhibit proliferation and invasion in these cancer cells. [35][36][37] Luciferase reporter and RIP assays further con rmed miR-141-3p was a LINC00645 target in HCC. miR-141-3p expression was substantially decreased in HCC tissues, and its expression was negatively correlated with LINC00645. Moreover, changes to cell proliferation, cycle arrest, migration and invasion upon LINC00645 knockdown could be partially reversed by miR-141-3p inhibitors in HCC cells. These results suggested that LINC00645 promoted HCC progression via modulating miR-141-3p.
In this study, we chose GP73, a speci c serum marker for HCC diagnosis and prognosis, as a target gene to identify upstream miRNAs that potentially regulates HCC progression. GP73 plays important roles in regulating HCC cellular and molecular processes. It has been reported that GP73 is a promising serum marker for diagnosis of HCC with high sensitivity (76%), speci city (86%), and diagnostic odds ratio (18.59) [38]. GP73-SphK1sR-Ad5 serves as a novel OA and can inhibit HCC progression with high speci city and e cacy [39]. Some studies have indicated that highly expressed GP73 promotes the migration and invasion of HCC [19], but the molecular processes are far more complete. In the previous study, it has been found that miR-141-3p over-expression up-regulated E-cadherin protein expression (epithelial marker) [40], and down-regulated N-cadherin and vimentin expression (mesenchymal markers), whereas over-expression of GP73 reversed expression levels of these markers. We found that miR-141-3p has the greatest potential to regulate GP73 expression and affect HCC progression. Our luciferase assay showed that miR-141-3p mimics signi cantly inhibited GP73 wild type 3'UTR activity, whereas the mimics had no effects on mutant 3′ UTRs, suggesting that miR-141-3p regulates GP73 expression at the posttranscriptional level.
In conclusion, LINC00645 was up-regulated in HCC tissues and correlated with the poor prognosis of HCC patients. Silencing the expression of LINC00645 inhibited the proliferation, migration and invasion, and promoted the apoptosis of HCC cells via targeting miR-141-3p. Down-regulated expression of LINC00645 also repressed the xenograft tumor growth in rats. Moreover, we provided evidence that MAZ has potential biding sites with LINC00645 and down-regulated expression of MAZ repressed the LINC00645 expression in HCC cells, which suggested that MAZ transcriptionally regulates LINC00645 expression in hepatocellular carcinoma cells. Thus, MAZ and LINC00645 may serve as potential prognostic biomarkers and therapeutic targets for HCC.

Declarations
Con ict of Interest The authors declare that they have no con ict of interests.

Funding
The research and manuscript preparation are funded by Haitao Xu.