LncRNA FOXD2‐AS1 as a competitive endogenous RNA against miR‐150‐5p reverses resistance to sorafenib in hepatocellular carcinoma

Abstract The current study elucidated the role of a long non‐coding RNA (lncRNA), FOXD2‐AS1, in the pathogenesis of hepatocellular carcinoma (HCC) and the regulatory mechanism underlying FOXD2‐AS1/miR‐150‐5p/transmembrane protein 9 (TMEM9) signalling in HCC. Microarray analysis was used for preliminary screening of candidate lncRNAs in HCC tissues. qRT‐PCR and Western blot analyses were used to detect the expression of FOXD2‐AS1. Cell proliferation assays, luciferase assay and RNA immunoprecipitation were performed to examine the mechanism by which FOXD2‐AS1 mediates sorafenib resistance in HCC cells. FOXD2‐AS1 and TMEM9 were significantly decreased and miR‐150‐5p was increased in SR‐HepG2 and SR‐HUH7 cells compared with control parental cells. Overexpression of FOXD2‐AS1 increased TMEM9 expression and overcame the resistance of SR‐HepG2 and SR‐HUH7 cells. Conversely, knockdown of FOXD2‐AS1 decreased TMEM9 expression and increased the sensitivity of HepG2 and Huh7 cells to sorafenib. Our data also demonstrated that FOXD2‐AS1 functioned as a sponge for miR‐150‐5p to modulate TMEM9 expression. Taken together, our findings revealed that FOXD2‐AS1 is an important regulator of TMEM9 and contributed to sorafenib resistance. Thus, FOXD2‐AS1 may serve as a therapeutic target against sorafenib resistance in HCC.

didate lncRNAs in HCC tissues. qRT-PCR and Western blot analyses were used to detect the expression of FOXD2-AS1. Cell proliferation assays, luciferase assay and RNA immunoprecipitation were performed to examine the mechanism by which FOXD2-AS1 mediates sorafenib resistance in HCC cells. FOXD2-AS1 and TMEM9 were significantly decreased and miR-150-5p was increased in SR-HepG2 and SR-HUH7 cells compared with control parental cells. Overexpression of FOXD2-AS1 increased TMEM9 expression and overcame the resistance of SR-HepG2 and SR-HUH7 cells. Conversely, knockdown of FOXD2-AS1 decreased TMEM9 expression and increased the sensitivity of HepG2 and Huh7 cells to sorafenib. Our data also demonstrated that FOXD2-AS1 functioned as a sponge for miR-150-5p to modulate TMEM9 expression. Taken together, our findings revealed that FOXD2-AS1 is an important regulator of TMEM9 and contributed to sorafenib resistance. Thus, FOXD2-AS1 may serve as a therapeutic target against sorafenib resistance in HCC.

K E Y W O R D S
hepatocellular carcinoma (HCC), long non-coding RNA (lncRNA), pathogenesis, proliferation, resistance, sorafenib high risk of vascular invasion, metastasis, drug resistance and recurrence after surgical resection. 6 Therefore, unravelling the potential molecular mechanism underlying chemotherapeutic resistance in HCC (especially the changes of genetics and epigenetics) is a major focus of research activity. 7 LncRNAs are RNA transcripts >200 nucleotides in length, but lack an obvious open reading frame. 8,9 Although lncRNA is not translated into protein, lncRNA participates in multiple physiological activities, including chromosome modification, transcription activation and interference, as well as cell growth, differentiation, and apoptosis. 10 Recent studies have demonstrated that several abnormally expressed lncRNAs can mediate drug resistance. For instance, AFAP1-AS1 has been reported to mediate cisplatin resistance in laryngeal cancer cells through the miR-320a/RBP signalling pathway. 11 Overexpression of lncRNA MALAT1 enhances autophagy and chemotherapeutic resistance of gastric cancer (GC) cells through the miR-23B-3P/ATG12 signalling pathway. 12 The lncRNA, H19, up-regulates expression of the multi-drug resistance gene (MDR1), thereby promoting the accumulation of doxorubicin in HCC cells and increasing the acceptable level of toxicity 13 ; however, the role of the lncRNA, FOXD2-AS1, in sorafenib-resistant HCC cells remains elusive. In this study we determined the role of the lncRNA, FOXD2-AS1, which is involved in resistance of HCC to sorafenib and elucidated the underlying mechanism.

| Patient specimens
In the current study, human HCC specimens were obtained from 60 patients who underwent surgery (34 males and 26 females) be-

| Microarray analysis
Microarray analysis of gene expression was performed according to the manufacturer's instructions (Agilent Technologies Co., Ltd., Santa Clara, CA). Briefly, 50 ng of purified mRNA was amplified and transcribed into double-stranded complementary DNA (cDNA). As previously described, 8 the cDNA was labelled and hybridized to human lncRNA Array v3.0 (Arraystar, Inc, Rockville, MD), according to the manufacturer's instructions. The original data were standardized and corrected using GenePix Pro 4.0 software. The comparison between HepG2 and SR-HepG2 samples was analysed by a t test. LncRNAs with a P < 0.05 were selected and cluster analysis was carried out using the hierarchical method, average linkage and Euclidean distance metrics.

| RNA isolation and qRT-PCR
According to the manufacturer's instructions, total RNA was extracted from the cancer cells using TRIzol reagent (Invitrogen; Thermo Fisher Scientific, Inc, Waltham, MA). The first-strand cDNA was synthesized using a PrimeScript 1st Strand cDNA synthesis kit (Takara Bio Inc, Kusatsu, Japan). The synthesized cDNA template was supplemented with SYBR Select Master Mix (Thermo Fisher Scientific). The following cycling conditions were used: pre-denaturation at 95°C for 30 seconds; 35 denaturation cycles at 95°C for 5 seconds; annealing at 55°C for 40 seconds; extension at 72°C for 1 minute; and a final extension at 72°C for 10 minutes. qRT-PCR was performed using the 7500 Real-

| Plasmid construction
The scramble shRNA sequence or shRNA targeting FOXD2-

| Dual-luciferase reporter gene assay
Dual-luciferase reporter gene assay was carried out, as described below. Cells (3 × 10 5 ) were cultured in 24-well plates and cotransfected with 2 ng pRL-TK (Promega) 10 ng of luciferase plasmids, and 100 ng of miR-150-5p mimic or negative control. The luciferase activity in the cells was detected 48 hours after transfection using a luciferase assay kit (Promega) and standardized with Ranilla luciferase activity. The experiments were repeated three times.

| Statistical analysis
All statistical analyses were performed using GraphPad Prism 6.0 software (GraphPad Software, Inc, La Jolla, CA). The thricerepeated data are expressed as the mean ± standard deviation (SD). Inter-group comparisons were performed using t tests or one-way ANOVA. The correlation between the FOXD2-AS1 level and TMEM9 or miR-150-5p level was analysed with the Pearson correlation coefficient. A P < 0.05 was considered statistically significant.

| Down-regulated expression of FOXD2-AS1 in sorafenib-resistant HCC cells
To clarify the relationship between lncRNAs and sorafenib resistance in HCC cells, sorafenib-resistant cell lines (SR-HepG2 and SR-HUH7) were constructed according to an established protocol. As illustrated in Figure 1A, the half maximal inhibitory concentration (IC 50 ) value ranged from 9.  Figure 1E). Moreover, sorafenib down-regulated the level of FOXD2-AS1 expression in HepG2 and HUH7 cells in a dose-dependent manner ( Figure 1C). Taken together, these results suggest that FOXD2-AS1 plays a key role in sorafenib resistance in HCC.

| FOXD2-AS1 enhances TMEM9 expression
TMEM9 is an important regulator in the progression of HCC. 14 Interestingly, we found that the expression of TMEM9 was significantly lower in SR-HepG2 and SR-HUH7 cells than HepG2 and HUH7 cells ( Figure 3A). Moreover, sorafenib reduced the expression of TMEM9 in a dose-dependent manner ( Figure 3B). Stable overexpression of FOXD2-AS1 in SR-HepG22 and SR-HUH7 cells significantly up-regulates the expression of TMEM9 at the mRNA and protein levels ( Figure 3C-E). Conversely, silencing of FOXD2-AS1 in HepG2 and  Figure 3F-H). Cell fractionation testing revealed that FOXD2-AS1 was mainly located in the cytoplasm of HCC cells, implying that FOXD2-AS1 might play a role in post-transcriptional modification ( Figure 3I).
Moreover, the expression of FOXD2-AS1 in HCC tissue samples was positively correlated with TMEM9 expression (R 2 = 0.4207, P < 0.05, Figure 3J). Collectively, these results suggest that TMEM9 is a target of FOXD2-AS1 in HCC.

| D ISCUSS I ON
Sorafenib is regarded as a standard chemotherapy for advanced HCC in clinical trials; however, the low clinical efficacy limits the use of sorafenib. [15][16][17][18][19] Although biotechnological progress has been achieved in the past few decades, the precise molecular mechanism underlying sorafenib resistance has not been fully unravelled. In this study we demonstrated downregulation of FOXD2-AS and increased TMEM9 expression in HepG2 and HUH7 cells with sorafenib resistance.
Recent studies have confirmed that lncRNAs play an important functional role in multidrug resistance of cancer cells. Specifically, lncRNA AK126698 is involved in cisplatin resistance in non-small cell lung cancer cells and overexpression of lncRNA snaR enhances sorafenib-induced cell death in colon cancer. 20 It has been reported that the lncRNA, LEIGC, mediates sorafenib resistance and epithelial-mesenchymal transition in gastric cancer. 21 Linc-TUG1 provokes impaired sensitivity in oesophageal squamous cell carcinoma. 22 FOXD2-AS1 knockdown inhibits the tumour growth of gemcitabine-resistant bladder cancer cells via the miR-143/ABCC3 axis. 23 In the present study, a group of lncRNAs differentially expressed in sorafenib-resistant HCC cells were validated.
In this study, we confirmed that down-regulation of FOXD2-AS1 and TMEM9 expression was positively correlated with the increase in sorafenib resistance. Further investigation demonstrated that FOXD2-AS1 regulated TMEM9 expression by completely sponging miR-150-5p, which inhibited miR-150-5p-mediated degradation of TMEM9 mRNA.
Indeed, this is the first study to confirm that FOXD2-AS1 regulates TMEM9 expression by acting as a ceRNA ofmiR-150-5p. Ectopic expression of FOXD2-AS1 reversed sorafenib resistance in HCC cells, whereas silencing of TMEM9 or overexpressing mir-150-5p partially restored this effect, indicating that FOXD2-AS1 regulates sorafenib resistance via miR-150-5p/TMEM9 axis. In addition, the dual-luciferase assay confirmed that FOXD2-AS1 increased TMEM9 expression and suppressed the Nrf2 signalling pathway in SR-HepG2 and SR-HUH 7 cells, and these effects were partially blocked by miR-150-5p mimics.
In contrast, inHepG2 and HUH7 cells with silencing of FOXD2-AS1, ARE-driven luciferase activity was increased in a dose-dependent manner, which was partially blocked by miR-150-5p inhibitor. These results indicate that FOXD2-AS1 regulates the Nrf2 signalling pathway via the miR-150-5p/TMEM9 pathway.

| CON CLUS ION
Taken together, FOXD2-AS1 is a novel key regulator of TMEM9 and mediates sorafenib resistance in HCC cells. FOXD2-AS1 competes with the 3'UTR of TMEM9 for binding with miR-150-5p, which promotes the expression of TMEM9, inhibits the Nrf2-ARE signalling pathway, and reverses sorafenib resistance in HCC cells. The finding that the FOXD2-AS1/miR-150-5p/TMEM9 signalling pathway is involved in sorafenib resistance may provide novel strategies to overcome sorafenib resistance in HCC.

CO N FLI C T O F I NTE R E S T S
The authors declare that they have no competing interests.

DATA AVA I L A B I L I T Y S TAT E M E N T
Data are available in this manuscript.