Long non‐coding RNA LUCAT1 promotes tumourigenesis by inhibiting ANXA2 phosphorylation in hepatocellular carcinoma

Abstract Long non‐coding RNAs (lncRNAs) play essential roles in diverse biological processes; however, current understanding of the mechanism underlying the regulation of tumour proliferation and metastasis is limited. Lung cancer‐associated transcript 1 (LUCAT1) has been reported in a variety of human cancers, while its role in hepatocellular carcinoma (HCC) remains unclear. This study aimed to determine the biological role and underlying mechanism of LUCAT1 on progression and metastasis in HCC cells and clinical specimens. Our results demonstrated that LUCAT1 was up‐regulated in HCC tissues and cells. Loss‐ and gain‐of‐function studies revealed that LUCAT1 promotes the proliferation and metastasis of HCC cells in vitro and in vivo. Furthermore, RNA pulldown and Western blot assays indicated that LUCAT1 inhibited the phosphorylation of Annexin A2 (ANXA2) to reduce the degradation of ANXA2‐S100A10 heterotetramer (AIIt), which in turn accelerated the secretion of plasminogen into plasmin, thereby resulting in the activation of metalloprotease proteins. In conclusion, we propose that LUCAT1 serves as a novel diagnostic and therapeutic target for HCC.


| INTRODUCTION
Hepatocellular carcinoma (HCC) is one of the most common cancers in adults, accounting for one-fifth of the incidence of malignant tumours in China. 1,2 Its main pathogenic factors include chronic hepatitis B infection, chronic alcohol abuse and non-alcoholic fatty liver disease. 3 Despite advances in the understanding of the molecular mechanisms underlying progression and therapeutic treatments for HCC, the median survival rate remains at approximately 50% (17%-69%) after 5 years. 4 This poor prognosis is mainly led by the frequent intrahepatic or extrahepatic metastasis of HCC. Cancer classification using biomarkers may define the risk of recurrence, which provides for the use of appropriate treatments for a better prognosis. But to date, few measurable biomarkers for predicting HCC metastasis or therapeutic targets for HCC have been identified.
Long non-coding RNAs (LncRNAs) are a class of poorly conserved non-coding RNA transcripts that are longer than 200 nucleotides. 5 It is now clear that various LncRNAs can function as signals, decoys, guides or scaffolds for other regulatory proteins. [6][7][8][9] For example, the LncRNA HULC promotes the phosphorylation of YB-1 through the extracellular signal-regulated kinase pathway, which then regulates the interaction of YB-1 with certain oncogenic mRNAs, thereby accelerating their translation to promote tumourigenesis. 10 Additionally, the LncRNA TSLNC8 exerts its tumour suppressive activity through inactivation of the IL-6/STAT3 signaling pathway by physically interacting with TKT and STAT3, which then inhibits STAT3 phosphorylation and its transcriptional activity in HCC. 11 These studies indicate that despite various ways through which LncRNAs affect the development of HCC, a part of LncRNAs can interact with proteins to regulate post-translational modifications such as phosphorylation or ubiquitination, ultimately influencing their activities and functions.
Annexin A2 (ANXA2) is a pleiotropic calcium-and anionic phospholipid-binding protein that exists as a monomer and as heterotetrameric complex with the plasminogen receptor protein, S100A10. 12 The complex is also named as AIIt. Annexin A2 has been proposed as an important part in many processes, including plasmin generation and matrix invasion. [13][14][15] It is proven that ANXA2 stabilizes S100A10, which in turn directly participates in binding to tPA and plasminogen and enhancing plasminogen activation. Plasminogen is then converted into its active proteolytic form plasmin. Plasmin mediates the hydrolysis and remoulding of the extracellular matrix (ECM), and activates metalloprotease (MMPs) such as MMP-9 12 .
Lung cancer-associated transcript 1 was first reported to be involved in smoking-related lung cancer and was previously called smoke and cancer-associated LncRNA1 (SCAL1). 6 It is also associated with non-small lung cancer, oesophageal squamous cell carcinoma and colorectal cancer and glioma. [7][8][9] It is proved that LUCAT1 regulates cell proliferation via epigenetically repressing p21 and p57 expression in non-small lung cancer. 9 Nevertheless, the expression pattern, underlying molecular mechanisms and biological roles of LUCAT1 in HCC tumourigenesis remain unclearly defined. In the present study, we found that LUCAT1 was up-regulated in tumour tissues compared to adjacent non-cancerous tissues and correlated with tumour size, metastasis and Edmondson grade of HCC. Our results indicated that LUCAT1 promotes tumour progression and metastasis in vivo and in vitro. Furthermore, it is demonstrated in our study that LUCAT1 inhibits the phosphorylation of ANXA2 at Ser-25 site to reduce the degradation of AIIt, which in turn accelerated the secretion of plasminogen into plasmin, thereby resulting in the activation of MMP proteins. Our results provide insights into the importance of LUCAT1 in the tumourigenesis and progression of HCC and as a predictor for survival.

| RNA isolation and quantitative reverse transcription-PCR
Total RNA was isolated from tissues using TRIzol and purified with the RNeasy MinElute Clean up kit (Qiagen, Hilden, Germany) according to the manufacturer's instruction. cDNAs were synthesized from the total RNA using the random priming method. Transcript levels were measured in duplicate by quantitative reverse transcription-PCR (qRT-PCR) (ABI 7900; Life Technologies). Expression levels were calculated relative to that of GAPDH.

| Transfection of cell lines
Full-length LUCAT1 was subcloned into the lentivirus vector GV367 to overexpress LUCAT1.1 (Gene, Shanghai, China). To knock down the expression of LUCAT1, a short hairpin RNA (shRNA) sequence aiming for LUCAT1 was subcloned into the lentivirus vector pLL3.7 (Gene), and the negative control hairpin shRNA with no sequence homology to any human gene was subcloned into the control groups. For overexpressing ANXA2, sequence of full-length ANXA2 was subcloned into the lentiviral vector GV367 (Gene). All vectors were labelled with luciferase. Transfection was conducted according to the manufacturer's instructions, and qRT-PCR was performed to determine transfection efficiency. The cells were then subjected to RNA extraction or functional assays.
In the invasion assay, over the membrane, a thin layer of ECM was dried. The ECM layer occluded the membrane pores and blocked the non-invasive cells from migrating. To the top chamber, 5 × 10 3 cells in serum-free medium were added. In the lower chamber, DMEM with 10% FBS was added. After 48 hours, cells that had invaded through the membrane were fixed with methanol, stained with crystal violet and counted. In the migration assay, similar protocol was performed that was used for the invasion assay described above, except that the ECM layer was not added to the chamber, and 2 × 10 3 cells were added per chamber.

| RNA binding protein immunoprecipitation
According to the manufacturer's instructions, we used the special antibody for RNA binding protein immunoprecipitation (RIP) assays that specifically targeted ANXA2. The co-precipitated RNAs were detected by RT-PCR. To ensure that the detected signals were from the RNAs specifically binding to ANXA2, negative controls and input controls were performed as well.

| RNA pulldown assays
For RNA pulldown assays, the biotin-labelled LUCAT1 transcript and the LUCAT1 intron sequence and its antisense were obtained using in vitro transcription with T7 RNA polymerase and Biotin RNA Labeling Mix (Roche), followed by 4 hours of incubation at 37°C with oncosphere cell lysates. Then, streptavidin-conjugated agarose beads were used for centrifugal enrichment. The precipitated components were separated using SDS-PAGE, followed by silver staining. Differential bands were isolated for mass spectrometry (MS) (LTQ Orbitrap XL).

| RNA fluorescence in situ hybridization
RNA fluorescence in situ hybridization (RNA-FISH) assays were performed as described below. Digoxygenin (DIG) labelled LUCAT1 probes were used for RNA FISH. MHCC97H cells were fixed in 4% formaldehyde and permeabilized with 0.5% Triton X-100 for 5 minutes, washed with PBS three times and once in 2X SSC buffer.
Hybridization was carried out using DNA probe sets at 37°C for 12 hours. Images were obtained with a FV1000 confocal laser microscope (Olympus). | 1875 charged nylon membrane (NC). The RNA was then fixed to the NC membrane using UV crosslinking. The cross-linked membrane was pre-hybridized with Ultrahyb-oligo hybridization buffer and hybridized with the LUCAT1-specific and GAPDH-specific oligonucleotide probes labelled with DIG-ddUTP in roller bottles.

| Western blot assay
Tissue samples and cultured cells were incubated with RIPA reagent plus phenylmethanesulfonylfluoride (Beyotime, Nantong, China) to isolate the proteins. Approximately 30 mg of total proteins were loaded into each lane and then transferred onto PVDF membranes.

| Statistical analysis
All experimental assays were performed in triplicate. The data were expressed as the mean ± SEM. Differences between two independent groups were tested with the student's t test using SPSS

| LUCAT1 is overexpressed in HCC tissues
To confirm the elevated expression of LUCAT1 in HCC tissues, we first examined its expression levels in 90 pairs of liver cancer and adjacent non-cancerous tissues by qRT-PCR. An increase in LUCAT1 F I G U R E 1 Lung cancer associated transcript 1 (LUCAT1) is aberrantly up-regulated in HCC tissues. (A) An increase in the expression of LUCAT1 occurs in hepatocellular carcinoma (HCC) tumour tissues compared to matched adjacent non-tumour tissues (n = 90). Lung cancer associated transcript 1 expression was tested by quantitative real-time PCR and measured using the 2 ÀΔΔCT method, analysed with paired student's t test, and presented as means ± SEM. (B) Based on the median value of the LUCAT1 expression in HCC tissues, patients were divided into two groups (LUCAT1-high expression group and LUCAT1-low expression group), the Kaplan-Meier survival analysis was used to calculate the overall survival. (C,D) The differential expression level of LUCAT1 in HCC cells, detected by quantitative reverse transcription-PCR and northern blot assays. (E) RNA fluorescence in situ hybridization was conducted to detect the sub-location of LUCAT1 (red) in MHCC97H cells, which revealed that it was mainly located in cytoplasm. A scale is presented at the lower right of the first panel. Magnification: 400× expression was found in the HCC samples (P = 0.004; Figure 1A). To further explore the clinical significance of the aberrant expression of LUCAT1 in HCC, according to the bimodal expression pattern of LUCAT1 in HCC patients ( Figure S1A), we separated all the HCC patients into two groups: LUCAT1-high expression group and LUCAT1-low expression group based on the median value of LUCAT1.
Statistical analysis showed that LUCAT1 expression is highly associated with tumour size, metastasis and tumour node metastasis (TNM) stage but not sex, age, HBsAg, cirrhosis and tumour number (Table 1).
Additionally, survival analysis showed that HCC patients with high LUCAT1 expression levels had poorer prognosis than those with low LUCAT1 expression and a shorter overall survival (OS) time ( Figure 1B).

| LUCAT1 promotes proliferation and metastasis of HCC cells in vitro
To investigate the biological functions of LUCAT1 in vitro, we detected the expression levels of LUCAT1 in the HCC cell lines, human normal liver cell LO2 as a control. The results of qRT-PCR and northern blot assays showed that the expression of LUCAT1 is significantly higher in the HCC cell lines compared to the LO2 cells, especially MHCC97H cells ( Figure 1C,D) Figure 2C). However, the results of the cell apoptosis assays showed no evidence of an association between LUCAT1 and cell apoptosis ( Figure S1D). Knockdown of LUCAT1 in MHCC97H and Hep3B cell lines sharply decreases its migration and invasion abilities ( Figure 3B).

| LUCAT1 enhances tumour growth and metastasis in vivo
To further determine the functional effects of LUCAT1 on tumourigenesis in vivo, a xenotransplantation model was built, in which nude mice were subcutaneously injected with MHCC97H cells stably  (Figures 4F,H). All the lung colonization of HCC cells was validated by histological examination ( Figure 4G). Taken together, these results reveal that LUCAT1 promotes growth and metastasis of HCC cells in vivo.

| LUCAT1 interacts with ANXA2
Many studies reveal that LncRNAs can regulate the function of proteins through RNA-protein interactions. 16 We used RNA pulldown assays followed by silver staining and MS to identify LUCAT1-binding proteins. Annexin A2 was found to specifically bind to LUCAT1 ( Figure 5A). The binding site of LUCAT1 and ANXA2 was predicted using the RNA sequence of LUCAT1 and the amino acid sequence of ANXA2 as input in Cat-Rapid database. We found an enrichment region encompassing 315 to 366 nt of the RNA sequence of LUCAT1 ( Figure 5B). RIP assay using the ANXA2 antibody was also conducted. We designed two independent primers of LUCAT1 for PCR assays after hybridization (primer 1 targeting the 315 to 366 nt region, whereas primer 2 was used as control). The binding status of LUCAT1 and ANXA2 was confirmed ( Figure 5C). Moreover, RNA- FISH and immunofluorescence assays indicated that LUCAT1 colocalized with ANXA2 in the cytoplasm of HCC cells ( Figure 5D).
Because previous studies have proven that ANXA2 plays an essential role in tumourigenesis and progression of HCC, [17][18][19][20] we propose that LUCAT1 exert its function through the LUCAT1/ANXA2 axis in HCC cell lines.

| LUCAT1 inhibits the phosphorylation of ANXA2 and accelerates the secretion of plasminogen into plasmin
Previous studies have proven that ANXA2 exist as monomers or as heterotetrameric complexes with S100A10 (P11), and the ANXA2/ S100A10 complex is referred to as AIIt, which was known to function in adherens junction formation in epithelial 21 and endothelial cells based on its association with epithelial E-cadherin and endothelial VE-cadherin. 22 Before any further investigation, we conducted ANXA2-overexpressing HepG2 cell lines ( Figure S2A,B). Figure 6A shows the binding of ANXA2 and S100A10 in HCC cell lines using co-immunoprecipitation (Co-IP) assays. As LUCAT1 could bind to ANXA2, we wonder whether the expression or phosphorylation status of ANXA2 could be affected by LUCAT1 in HCC cell lines. Figure 6B shows that there was no differential change in the total protein expression of ANXA2 upon LUCAT1 alteration, but the Ser-25 phosphorylation status of ANXA2 was apparently affected by LUCAT1. The expression of pSer25-ANXA2 decreased in LUCAT1overexpressing cells and increased in LUCAT1 knockdown cells. In the AIIt complex, ANXA2 played an obligatory role in the regulation of S100A10 by protecting S100A10 from rapid ubiquitin-mediated degradation. [23][24][25][26] Furthermore, it has been proven that the phosphorylation of Ser-25 of ANXA2 triggers the dissociation of the AIIt complex, which results in the release of S100A10 as it undergoes ubiquitin-mediated proteasomal degradation. 12,27 Next, to investigate whether LUCAT1 influences the AIIt complex, we conducted Co-IP assays, which showed that ANXA2 precipitated less S100A10 in LUCAT1 knockdown cells ( Figure 6C) and more S100A10 in LUCAT1 overexpressing cells ( Figure 6D).
The colocalization of the plasminogen activators and plasminogen driven by AIIt could increase the cleavage of plasminogen into plasmin, which further activates pro-MMPs (matrix MMPs) into active MMPs. 12,[28][29][30] Previous studies have revealed that plasmin and MMPs could function as carcinogenic factors by enhancing the proliferation and metastasis of various tumours, including HCC. [31][32][33][34] Next, to explore alterations in the expression of plasmin and MMP proteins caused by LUCAT1, ELISAs were conducted. Figures 6E,F show that the expression of plasmin and MMP-9 increased when LUCAT1 was overexpressed and decreased when LUCAT1 was knocked down. Also, the classical epithelial-mesenchymal transition  showed decreased expression of pSer25ANXA2 when LUCAT1 was up-regulated ( Figure 6B), suggesting that LUCAT1 inhibited the phosphorylation of ANXA2 at Ser-25, while other phosphorylation sites showed no changes likewise (data not shown). It has been proven that the phosphorylation of Ser-25 of ANXA2 triggers the dissociation of the AIIt complex, which results in the release of S100A10 which then undergoes ubiquitin-mediated proteasomal degradation. 12,27 Co-immunoprecipitation assays confirmed the combination between ANXA2 and S100A10, and it is obvious that ANXA2 precipitated less S100A10 in LUCAT1 knockdown cells and more S100A10 in LUCAT1 overexpressing cells, which indicated that In conclusion, we revealed that LUCAT1 plays a key role in HCC tumour progression and metastasis. Our results therefore suggest that the LncRNA LUCAT1 might be a predictive marker for HCC metastasis and target for drug development.

ACKNOWLEDG EMENTS
The study is supported by the National Natural Science Foundation

CONFLI CT OF INTEREST
The authors declare that they have no financial conflict of interest.