A novel lncRNA MCM3AP-AS1 promotes the growth of hepatocellular carcinoma by targeting miR-194-5p/FOXA1 axis

Background Hepatocellular carcinoma (HCC) is the most common malignant liver tumor with poor clinical outcomes. Increasing amount of long non-coding RNAs (lncRNAs) have been revealed to be implicated in the carcinogenesis and progression of HCC. However, the expressions, clinical significances, and roles of most lncRNAs in HCC are still unknown. Methods The expression of lncRNA MCM3AP antisense RNA 1 (MCM3AP-AS1) in HCC tissues and cell lines was detected by qRT-PCR and fluorescence in situ hybridization. Immunoblotting, CCK-8, EdU, colony formation and flow cytometry were performed to investigate the role of MCM3AP-AS1 in HCC cell proliferation, cell cycle and apoptosis in vitro. A subcutaneous tumor mouse model was constructed to analyze in vivo growth of HCC cells after MCM3AP-AS1 knockdown. The interactions among MCM3AP-AS1, miR-194-5p and FOXA1 were measured by RNA pull-down, RNA immunoprecipitation and luciferase reporter assay. Results We revealed a novel oncogenic lncRNA MCM3AP-AS1, which is overexpressed in HCC and positively correlated with large tumor size, high tumor grade, advanced tumor stage and poor prognosis of HCC patients. MCM3AP-AS1 knockdown suppressed HCC cell proliferation, colony formation and cell cycle progression, and induced apoptosis in vitro, and depletion of MCM3AP-AS1 inhibited tumor growth of HCC in vivo. Mechanistically, MCM3AP-AS1 directly bound to miR-194-5p and acted as competing endogenous RNA (ceRNA), and subsequently facilitated miR-194-5p’s target gene forkhead box A1 (FOXA1) expression in HCC cells. Interestingly, FOXA1 restoration rescued MCM3AP-AS1 knockdown induced proliferation inhibition, G1 arrest and apoptosis of HCC cells. Conclusions Our results recognized MCM3AP-AS1 as a novel oncogenic lncRNA, which indicated poor clinical outcomes in patients with HCC. MCM3AP-AS1 exerted an oncogenic role in HCC via targeting miR-194-5p and subsequently promoted FOXA1 expression. Our findings suggested that MCM3AP-AS1 could be a potential prognostic biomarker and therapeutic target for HCC. Electronic supplementary material The online version of this article (10.1186/s12943-019-0957-7) contains supplementary material, which is available to authorized users.


Background
According to the 2018 Global Cancer Statistics, about 841,000 newly diagnosed liver cancer cases and 782,000 liver cancer deaths were appeared worldwide, with China alone accounting for about 50% of the total number of cases and deaths [1,2]. Hepatocellular carcinoma (HCC) accounts for 75-80% of all liver cancer cases [1]. Although a significantly decreasing incidence and mortality trend for HCC was observed in China, a huge population base and rapid population growth still led to a large and rising number of new HCC cases [3]. Thus, it is worth to better understand the molecular mechanisms underlying HCC tumorigenesis and progression, and develop more efficient targeted therapies for HCC.
Large amount 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 identified as non-coding RNAs (ncRNAs) due to the great progressions of genome and transcriptome sequencing [5]. ncRNAs are further grouped into long ncRNAs (lncRNA) and small ncRNAs depending on their transcript size [6]. LncRNAs, defined as a form of ncRNAs greater than 200 nt in length, are found to exert their gene transcription regulatory function by epigenetic regulatory mechanism [7]. Increasing evidences indicate that lncRNAs are implicated in several pathophysiological processes including human cancers [8][9][10][11][12]. Aberrant expression of lncRNAs has been frequently observed in cancers [13][14][15][16]. Moreover, lncRNAs regulate malignant behaviors of cancer cells, such as proliferation, apoptosis resistance, migration, invasion and drug resistance [17][18][19][20]. For instance, lncRNA miR503HG expression is found to be down-regulated in HCC and represses HCC metastasis via regulating the heterogeneous nuclear ribonucleoprotein A2/B1 (HNRNPA2B1)/nuclear factor κB (NF-κB) signaling [21]. High expression of lncRNA linc00210 is detected in liver cancer and contributes to tumor progression by driving the activation of Wnt/β-catenin pathway in a catenin beta interacting protein 1 (CT NNBIP1)-dependent manner [22]. Forkhead box A2 (FOXA2)-induced lncRNA-NEF is frequently down-regulated in HCC, and suppresses epithelial-mesenchymal transition (EMT) and tumor metastasis by antagonizing Wnt/β-catenin pathway [23]. Moreover, lncRNA-MUF is found to be highly expressed in HCC and facilitates hepatocarcinogenesis via directly regulating Annexin A2 (ANXA2)/Wnt/β-catenin signaling and miR-34a/Snail1/ EMT axis [24]. In our previous study, we find that lncRNA TUSC7 is down-regulated in HCC and indicates poor prognosis of patients, and it inhibits EMT and HCC metastasis by acting as miR-10a sponge and subsequently leads to Eph tyrosine kinase receptor A4 (EphA4) upregulation [25]. Furthermore, we investigate the expression and function of lncRNA CASC2 in HCC and reveal that CASC2 exerts an anti-metastatic role by targeting miR-367/F-box and WD repeat domain containing 7 (FBXW7) axis [26]. Although several lncRNAs have been reported to participate in the tumorigenesis and progression of HCC, the expressions and roles of most lncRNAs in HCC are still unclear.
In this study, we analyzed differentially expressed lncRNAs in HCC compared to normal liver tissues based on the microarray data from National Center for Biotechnology Information (NCBI) Gene Expression Omnibus (GEO) dataset (GSE65485) and identified a novel highly expressed lncRNA MCM3AP antisense RNA 1 (MCM3AP-AS1) in HCC. Next, we investigated the expression, clinical significance, functional role and underlying mechanisms of MCM3AP-AS1 in HCC.

Clinical specimens
A total of 80 pairs of HCC and tumor-adjacent tissues were collected from patients who underwent hepatectomy at the First Affiliated Hospital of Xi'an Jiaotong University (Xi'an, China). None of HCC patients received any pre-operative treatments, such as radiofrequency ablation (RFA), transcatheter arterial chemoembolization (TACE), immunotherapy and targeted therapy. The tissue samples were confirmed by two histopathologists. All samples were immediately snap-frozen in liquid nitrogen and subsequently stored at − 80°C until RNA extraction and protein isolation. The demographic and clinicopathological features for HCC patients were described in Table 1.
Quantitative real-time polymerase chain reaction (qRT-PCR) TRIzol reagent (Invitrogen) was used for total RNA isolation from HCC tissues and cultured cells. Total RNA was reverse transcribed into cDNA using a RevertAid First Strand cDNA Synthesis Kit (Thermo-Fisher Scientific). qRT-PCR analyses were performed using SYBR® Premix Ex Taq™ II (Takara, Dalian, China) and Taqman UniversalMaster Mix II (Life Technologies Corporation, Carlsbad, CA, United States) on an ABI PRISM 7300 Sequence Detection system (Applied Biosystems, Foster City, CA, USA) in accordance with the manufacturers' instructions. The 2 -ΔΔCt method was used to calculate the relative gene expression normalized by GAPDH and U6. The sequences of the primers were listed in Table 2.

Colony formation assay
Forty-eight hours after transfection, HCC cells (1 × 10 3 per well) were seeded in a 6-well plate and cultured with complete medium for 2 weeks. Cell colonies were fixed with 4% paraformaldehyde for 30 min and stained with 0.5% crystal violet for 30 min at room temperature.

Ethynyl deoxyuridine (EdU) incorporation assay
EdU incorporation assay was performed with the EdU kit (Roche, Indianapolis, IN, USA) in accordance with the manufacturer's instruction. Results were acquired using the Zeiss fluorescence photomicroscope (Carl Zeiss, Oberkochen, Germany) and quantified via counting at least five random fields.

Flow cytometry assay
For cell cycle analysis, cells were collected and fixed using 70% ethanol. After washing with PBS and subsequently washing with stain buffer, 1 × 10 6 cells were resuspended in 0.5 mL of PI/RNase Staining Buffer (BD biosciences, San Jose, CA, USA), and cells were incubated for 15 min at room temperature (RT), protected from light. A FACSCanto II flow cytometer (BD biosciences) was used to analyze cell cycle distribution. For apoptosis assay, the PE Annexin V Apoptosis Detection Kit I (BD biosciences) was used following the manufacturer's protocols. Cells were harvested and washed with pre-cold PBS buffer twice. Then, 5 μl of PE Annexin V and 5 μl of 7-AAD solution were added to each sample, and cells were incubated for 15 min at RT. FACSCanto II flow cytometer was used to measure the cell apoptosis.
Ltd. (Shanghai, China) and randomly divided into two groups (n = 6 per group). Hep3B cells (1 × 10 6 per injection) that were transfected with sh-MCM3AP-AS1 and sh-control, respectively, were implanted into the right flank of the mice via subcutaneous injection. Tumor volumes were measured every 3 days after being apparently observed and calculated with the following formula: Volume = (length × width 2 )/2. After 3 weeks, all mice were sacrificed under anesthesia. Tumor tissues were harvested and subjected to immunohistochemistry for Ki-67 staining [28]. The animal experiments were approved by the Animal Care and Use Committee of Xi'an Jiaotong University.

RNA immunoprecipitation (RIP)
RIP assay was performed with the EZ-Magna RIP Kit (Millipore, Bedford, MA, USA) and a AGO2 antibody (Millipore) as previously described [26]. qRT-PCR was carried out to detect co-precipitated RNAs.

Luciferase reporter assay
The sequence of 3′-UTR of FOXA1 or MCM3AP-AS1 was amplified from human genomic DNA. Then these sequences were respectively subcloned into pGL3 luciferase reporter vector (Promega, Madison, WI, USA). The potential miR-194-5p binding sites were mutated by the Quick-change site-directed mutagenesis kit (Agilent Technologies, Santa Clara, CA, USA). The wt (mt) 3′-UTR of FOXA1 vector or wt (mt) MCM3AP-AS1 vector and control mimics or miR-194-5p mimics were co-transfected into HepG2 and SMMC-7721 cells. The luciferase activity was measured and normalized as previously described [26].
Quantitation of MCM3AP-AS1 and miR-194-5p expression levels The exact copy numbers of MCM3AP-AS1 and miR-194-5p transcripts per Hep3B and HepG2 cell were quantified by using quantitative real-time RT-PCR assay.
In this assay, serially diluted RT-PCR products of MCM3AP-AS1 and miR-194-5p were used as templates to formulate standard curves, and then, the exact copies of MCM3AP-AS1 and miR-194-5p per cell were calculated accordingly.

Statistical analysis
Data were analyzed using GraphPad Prism 6.0 Software (GraphPad Inc., San Diego, CA, USA). The Student's t-test was used to analyze differences between two groups, and two-way ANOVA was used when more than two groups were compared. The correlations between MCM3AP-AS1 and miR-194-5p expression were analyzed using the Pearson correlation test. Overall survival curves were protracted using the Kaplan-Meier method and estimated by the log-rank test. Differences were defined as statistically significant if P < 0.05.

A novel lncRNA MCM3AP-AS1 is overexpressed in HCC
To further disclose differentially expressed lncRNAs in HCC, we analyzed a microarray data comparing the expression of lncRNAs in 50 HCC tissues and 5 adjacent non-tumor tissues from the GEO database with the accession number GSE65485. Thirty-seven lncRNAs were differentially expressed in HCC tissues compared to adjacent non-tumor tissues (Table 3). A novel lncRNA MCM3AP-AS1 (also known as MCM3APAS), which was 2.84-fold higher in HCC tissues than that in adjacent non-tumor tissues, caught our attention. Next, we search for the expression pattern of MCM3AP-AS1 in HCC based on TCGA data from starBase V3.0 [31].
Moreover, based on two other GEO datasets (GSE45436 and GSE54236) from R2: Genomics Analysis and Visualization Platform (http://r2.amc.nl), we found that MCM3AP-AS1 expression was prominently higher in HCC tissues compared to normal liver tissues (P < 0.0001, Fig. 1c and d). Thus, these results indicated that MCM3AP-AS1 upregulation was a frequent event in HCC.
High level of MCM3AP-AS1 correlates with poor prognosis of HCC patients Next, we aimed to reveal the clinical significance of MCM3AP-AS1 in HCC. TCGA data from R2: Genomics Analysis and Visualization Platform (http://r2.amc.nl) revealed that MCM3AP-AS1 was more highly expressed in HCC with high tumor grades (G3 + G4) than that in HCC with low tumor grades (G1 + G2) (P = 0.0032, Fig. 2a). Furthermore, MCM3AP-AS1 was also more highly expressed in HCC with advanced tumor stages (III-IV) than that in HCC with early tumor stages (I-II) (P = 0.0013, Fig. 2b). We divided HCC patients into tow subgroups (low/high MCM3AP-AS1 level) by using the median of the cohort as a cut-off value. As shown in Table 1, the correlation analysis between MCM3AP-AS1 expression and clinicopathologic characteristics of these 80 HCC patients indicated that high expression of MCM3AP-AS1 was positively correlated with large tumor size (P = 0.006), high tumor grade (P = 0.039), and advanced TNM stages (P = 0.004). Kaplan-Meier survival analysis showed that HCC patients with high MCM3AP-AS1 expression had a significant poorer overall survival than those with low MCM3AP-AS1 expression (P = 0.0054, Fig. 2c). Furthermore, TCGA data from OncoLnc (http://www.oncolnc.org/) further demonstrated that high MCM3AP-AS1 expression also indicated poor survival of HCC patients (P = 0.0112, Fig.  2d). Collectively, our data showed that high MCM3AP-AS1 expression was associated with poor clinical outcomes of HCC patients.

Depletion of MCM3AP-AS1 suppresses cell growth and induces apoptosis of HCC cells
Since MCM3AP-AS1 expression was associated with tumor size, we next disclosed the biological roles of MCM3AP-AS1 in HCC cell growth. MCM3AP-AS1 were stably depleted in HepG2 and Hep3B cells with different specific shRNAs (P < 0.05, Fig. 3a). CCK-8 assays indicated that MCM3AP-AS1 knockdown significantly inhibited the proliferation of HepG2 and Hep3B cells (P < 0.05, Fig. 3b). Colony formation and EdU incorporation assays also indicated that MCM3AP-AS1 silencing prominently suppressed the growth of HepG2 and Hep3B cells (P < 0.05. Figure 3c and d). Moreover, flow cytometry assays revealed that the percentage of apoptotic HCC cells were obviously increased by MCM3AP-AS1 knockdown (P < 0.05. Figure 3e). Depletion of MCM3AP-AS1 led to cell cycle arrest at G1 phase in HepG2 and Hep3B cells (P < 0.05, Fig. 3f ). Furthermore, MCM3AP-AS1 knockdown led to increased levels of cleaved PARP1, cleaved caspase-3, cleaved caspase-7 and p21, and decreased Cyclin D1 expression in HCC cells (P < 0.05, Additional file 2: Figure S2). Notably, MCM3AP-AS1 knockdown did not significantly affected the growth of LO2 cells, which had low MCM3AP-AS1 expression (Additional file 3: Figure S3). Thus, these results showed that knockdown of MCM3AP-AS1 repressed the proliferation, cell cycle progression and induced apoptosis of HCC cells in vitro.

MCM3AP-AS1 knockdown restrains tumorigenesis of HCC in vivo
To further elucidate the biological roles of MCM3AP-AS1 in HCC tumorigenesis in vivo, Hep3B cells with MCM3AP-AS1 knockdown were implanted into nude mice via subcutaneous injection. The results of tumor growth curves and tumor weight indicated that MCM3AP-AS1 knockdown obviously reduced tumor growth in mice (P < 0.05, Fig. 4a and b). Tumor tissues were harvested for qRT-PCR analysis of MCM3AP-AS1. We confirmed that lower expression of MCM3AP-AS1 was detected in tumor tissues arising from MCM3AP-AS1 knockdown group compared to control group (P < 0.05, Fig. 4c). Ki-67 immunostaining indicated that the subcutaneous tumors formed by MCM3AP-AS1 knockdown Hep3B cells showed less Ki-67 positive cells compared to those formed by control Hep3B cells (P < 0.05, Fig. 4d). Altogether, these results suggested that MCM3AP-AS1 knockdown suppressed HCC tumorigenesis in vivo.

FOXA1 restoration attenuates the effect of MCM3AP-AS1 knockdown on HCC cells
To explore whether the FOXA1 was critical for cell proliferation restriction, G1 arrest and apoptosis upon MCM3AP-AS1 knockdown, HepG2 cells with MCM 3AP-AS1 knockdown were transfected with FOXA1 expression vectors. CCK8 assays indicated that restoration of FOXA1 attenuated the proliferation suppressive role of MCM3AP-AS1 knockdown in HepG2 cells (P < 0.05, Fig. 7a). Flow cytometry assays revealed that FOXA1 restoration reversed apoptosis and G1 arrest in HepG2 cells with MCM3AP-AS1 knockdown (P < 0.05, Fig. 7b and c). Moreover, colony formation and EdU incorporation assays also indicated that overexpression of FOXA1 attenuated the growth arrest of HepG2 cells induced by MCM3AP-AS1 knockdown (P < 0.05, Fig. 7d and e). Thus, these results strongly suggested that MCM3PA-AS1 knockdown induced cell proliferation restriction, cell cycle arrest and apoptosis was at least partially mediated by FOXA1 inhibition in HCC.

Discussion
Tens of thousands lncRNAs are identified by human transcriptome sequencing. Increasing studies have revealed more and more cancer-related lnRNAs, some of them play essential roles in tumorigenesis and progression of HCC [39]. Beside the well characterized lncRNAs, it is still worth to investigate potential essential lncRNAs in controlling HCC initiation and progression. Thus, we re-analyzed the microarray data from public available database about the differentially exp ressed lncRNAs in HCC. Interestingly, we found a novel lncRNA MCM3AP-AS1, which was markedly overexpressed in HCC tissues and cell lines compared with  endothelial cells (GECs) [40]. TCGA data from OncoLnc also reveals that high expression of MCM3AP-AS1 indicates poor survival of patients with colon adenocarcinoma. Thus, the aberrant expression and clinical significance of MCM3AP-AS1 in other human cancers are worth to be investigated. Next, loss-of-function assays indicated that MCM3AP-AS1 knockdown inhibited cell proliferation, colony formation and cell cycle progression, and induced apoptosis of HCC cells in vitro. In vivo experiments found that MCM3AP-AS1 silencing suppressed HCC tumor growth in mice. Taken together, these results suggested an oncogenic role of MCM3AP-AS1 in HCC. MCM3AP-AS1 promotes angiogenesis of glioblastoma in vitro [40]. This implies that MCM3AP-AS1 may regulate other malignant behaviors of HCC cells including metastasis and angiogenesis, which needs further study.
One of the most popular functional model for lncRNAs is that lncRNAs function as ceRNAs to sponge miRNAs via sequence complementarity and subsequently influence functional roles of miRNAs [39]. Yang et al. find that MCM3AP-AS1 acts as a ceRNA to promote KLF5/AGGF1 axis, and activate PI3K/AKT and ERK1/2 signaling pathways by sponging miR-211 in glioblastoma [40]. Here, we found that MCM3AP-AS1 mainly located in the cytoplasm and the abundance of MCM3AP-AS1 was comparable to that of miR-194-5p in HCC cells. Moreover, MCM3AP-AS1 silencing prominently up-regulated miR-194-5p expression in HCC cells. An inverse correlation between MCM3AP-AS1 and miR-194-5p expression was confirmed in HCC tissues from our cohort and TCGA database. The following luciferase reporter assay and RNA pull down assay demonstrated that MCM3AP-AS1 acted as molecular sponge for miR-194-5p by directly binding to complementary sequence in HCC cells. Since miR-194-5p suppresses cell proliferation and blocks G1-S transition in HCC cells [41]. And a recent study also reports the lncRNA XIST regulation of miR-194-5p in HCC [42]. Thus, we considered that MCM3AP-AS1 played an oncogenic role in HCC via down-regulating miR-194-5p expression. FOXA1, a transcription factor, promotes tumor growth of HCC [27,35,36]. Our previous study shows that FOXA1 mediates the tumor suppressive role of miR-212 in HCC tumor growth [27]. MSL2 contributes to the growth of HCC cells in vitro and in vivo and is up-regulated HBx-mediated activation of YAP/FOXA1 signaling [35]. Moreover, tumor suppressive lncRNA MT1DP inhibits cell proliferation and colony formation, but induces apoptosis of HCC cell in a FOXA1 depen dent manner [36]. In this study, FOXA1 was identified as a direct target of miR-194-5p in HCC cells. miR-194-5p regulation of FOXA1 is also observed in lung cancer [43]. Notably, MCM3AP-AS1 positively regulated FOXA1 abundance in HCC cells, while miR-194-5p showed an opposite regulatory effect. A positive correlation between MCM3AP-AS1 and FOXA1 and a negative correlation between miR-194-5p and FOXA1 were observed in HCC tissues. Importantly, FOXA1 restoration reversed MCM3AP-AS1 knockdown induced HCC cell proliferation restriction, cell cycle arrest and apoptosis. To conclude, our study provided a novel insight that MCM3AP-AS1/miR-194-5p/FOXA1 axis contributed to the growth of HCC. MCM3AP-AS1/ miR-194-5p/FOXA1 axis might be potential therapeutic targets for HCC.

Conclusions
In summary, our findings identified a novel lncRNA MCM3AP-AS1, which was up-regulated in HCC and associated with poor prognosis of HCC patients. MCM3AP-AS1 knockdown inhibited the proliferation, cell cycle progression and induced apoptosis of HCC cells, and suppressed tumor growth of HCC in vivo. Mechanistically, MCM3AP-AS1 functioned as an oncogenic lncRNA by acting as a ceRNA to sponge miR-194-5p and subsequently promoted FOXA1 expression. Our data suggested that MCM3AP-AS1 might be a potential prognostic biomarker and therapeutic target for HCC.