Long non‐coding RNA deleted in lymphocytic leukaemia 1 promotes hepatocellular carcinoma progression by sponging miR‐133a to regulate IGF‐1R expression

Abstract Long non‐coding RNA (lncRNA) deleted in lymphocytic leukaemia 1 (DLEU1) was reported to be involved in the occurrence and development of multiple cancers. However, the exact expression, biological function and underlying mechanism of DLEU1 in hepatocellular carcinoma (HCC) remain unclear. In this study, real‐time quantitative polymerase chain reaction (qRT‐PCR) in HCC tissues and cell lines revealed that DLEU1 expression was up‐regulated, and the increased DLEU1 was closely associated with advanced tumour‐node‐metastasis stage, vascular metastasis and poor overall survival. Function experiments showed that knockdown of DLEU1 significantly inhibited HCC cell proliferation, colony formation, migration and invasion, and suppressed epithelial to mesenchymal transition (EMT) process via increasing the expression of E‐cadherin and decreasing the expression of N‐cadherin and Vimentin. Luciferase reporter gene assay and RNA immunoprecipitation (RIP) assay demonstrated that DLEU1 could sponge miR‐133a. Moreover, miR‐133a inhibition significantly reversed the suppression effects of DLEU1 knockdown on HCC cells. Besides, we found that silenced DLEU1 significantly decreased insulin‐like growth factor 1 receptor (IGF‐1R) expression (a target of miR‐133a) and its downstream signal PI3K/AKT pathway in HCC cells, while miR‐133a inhibitor partially reversed this trend. Furthermore, DLEU1 knockdown impaired tumour growth in vivo by regulating miR‐133a/IGF‐1R axis. Collectively, these findings indicate that DLEU1 promoted HCC progression by sponging miR‐133a to regulate IGF‐1R expression. Deleted in lymphocytic leukaemia 1/miR‐133a/IGF‐1R axis may be a novel target for treatment of HCC.

to elucidate the key molecular mechanisms of the pathogenesis and development of HCC for finding new therapeutic strategies for this disease.
Long non-coding RNAs (lncRNAs), a prominent class of transcripts longer than 200 nucleotides in length and limited protein-coding potential, 4 have recently gained wide attention due to their functional roles in a variety of biological processes. 5,6 Long non-coding RNAs have been highlighted to be involved in the occurrence and development of cancer, 7,8 offering the possibility of lncRNAs as novel diagnosis markers and therapy agent for cancer. Number of lncRNAs was identified to function as oncogene or tumour suppressors in HCC by modulation of cell proliferation, autophagy, apoptosis, cycle, invasion and metastasis via different pathway. 9,10 Long non-coding RNA deleted in lymphocytic leukaemia 1 (DLEU1), located on chromosome 13q14. 3,11 has been reported to be up-regulated and function as oncogene in several types of cancer, including oral squamous cell carcinoma, 12 colorectal cancer, 13 gastric cancer, 14 ovarian cancer 15 and endometrial carcinoma. 16 However, the role and potential molecular mechanism of DLEU1 in HCC remain unclear. Therefore, the aims of this study were to investigate the role of DLEU1 in HCC progression and explore the mechanism behind it in HCC action.

| Tissue specimens
A total of 56 HCC tissues and paired adjacent tissues were collected from patients with primary HCC with no previous others treatment who underwent curative resection between January 2013 and August 2014 at the Department of Hepatopancreatobiliary Surgery, the First Hospital of Jilin University. All tissues were rapidly frozen in liquid nitrogen following surgery and stored at the temperature of −80°C until RNA extraction. Tumour pathology and staging were determined by two independent pathologists form our hospital. Prior to operation, none of the patients received chemo-or radiotherapy and other therapy. The Clinical Research Ethics Committee of Jilin University approved this study. Informed consent was signed by each patient enroled in this study. Table 1 summarizes the relevant clinicopathological characteristics of all patients with HCC.

| Cell culture and transfection
Institute of Cell Biology of Chinese Academy of Science (Shanghai, China) was the source of human HCC cell lines (SMMC-7721, Hep3B,

| Wound healing assay
After seeding (5 × 10 5 cells/well, six-well plate), wounding was achieved by scratching with a sterile 200 μL pipette tip. Serum-free medium was then added, and cells were allowed to move towards the denuded area for 24 hours. The spread of the wound was observed under light microscopy and photographed at 0 and 24 hours.
The wound areas were analysed using Imagej software 3.2 (National Institutes of Health, Bethesda, MD).
Olympus fluorescence microscope (Tokyo, Japan), counted and calculated the mean in five randomly selected fields.

| Western blot
Total proteins were extracted from cultured cells and tumour tis-

| Tumour formation in BALB/c nude mice
Male athymic 5-week-old BALB/c nude mice (18-20 g) were bought from the Experiments Animal Center of Changchun Biological Institute (Changchun, China) and were kept in specific pathogen-free conditions.
All experiments for the in vivo nude mouse study were performed in accordance with the Guide for the Care and Use of Laboratory Animals of the National Institute of Health and were approved by the Animal Ethics Committee of Jilin University (Changchun, China).
Ten mice were randomly divided into two groups (each group five mice), and received 2 × 10 6 SMMC-7721 cells that had undergone stable transfection using sh-SHNG3 or sh-NC via subcutaneously injected respectively. Tumour volume was measured every 7 days by the formula: 0.5 × length × width 2 . Thirty-five days after injection, mice were killed, and the tumours were removed, photographed and partially the xenograft tumour processed for qRT-PCR and western blotted, as well as partially fixed for immunohistochemical staining.

| Immunohistochemistry
Immunostaining was performed on the paraffin-embedded tumour tissues from nude mice using streptavidin-peroxidase conjugated method as described previously. 17

| Statistical analysis
All results are showed as mean ± SD from at least three replicates measurements, and were analysed using spss v17.0 (IBM, Chicago, IL) and Graphpad Prism6.0 (San Diego, CA). Data from two groups were compared with Student's t tests, while one-way ANOVAs with Bonferroni's correction were used for comparisons between three or more groups. Kaplan-Meier method and log-rank test were applied to analysis overall survival ratio. Pearson's correlation coefficient was used to analyse correlation between two groups. The threshold of significance was *P < 0.05, **P < 0.01.

| DLEU1 was up-regulated in HCC tissues and was associated with the poor survival of patients
We examined the DLEU1 expression in HCC cell lines to analyse DLEU1 expression in hepatic carcinogenesis and progression. Results of qRT-PCR revealed that DLEU1 expressions were remarkably increased in HCC cell lines relative to LO2 cells lines ( Figure 1A). Moreover, the expression levels of DLEU1 were remarkably increased in HCC tissues compared with adjacent normal tissues ( Figure 1B). To further explore the clinical significance of The association between DLEU1 expression and clinical features is analysed and summarized in Table 1. There was no significant association between DLEU1 expression and patient's age, gender, differentiated, serum α-fetoprotein (AFP) and hepatitis C virus (HCV) antigen (All P > 0.05, Table1). However, increased DLEU1 was positively associated with vascular invasion and tumour-nodemetastasis (TNM) stage in HCC patients. In addition, Kaplan-Meier analysis revealed that patients with high DLEU1 expression had a significantly shorter overall survival compared to patients with low DLEU1 expression ( Figure 1C). These results indicated that DLEU1 might be involved in HCC progression.

| Knockdown of DLEU1 inhibits HCC cell proliferation, colony formation and cell cycle distribution
To evaluate the biological roles of DLEU1 on HCC progression, plasmid sh-DLEU1 or sh-NC were introduced into SMMC-7721 or HepG2 cells, subsequently, cell proliferation, colony formation and cell arrest were detected. As shown in Figure 2A, sh-DLEU1 produced a significant reduction in endogenous DLEU1 expression in

SMMC-7721 and HepG2 cells. After knockdown, both SMMC-7721
and HepG2 cell proliferation were significantly suppressed in a CCK-8 assay ( Figure 2B). We similarly found that SNHG3 knockdown clearly decreased SMMC-7721 and HepG2 cell colony formation ( Figure 2C).Flow cytometry assay revealed that knockdown of DLEU1 lead to a significant reduction in the percentage of S phase and promotion in the percentage of G1/G0 phase in SMMC-7721 and HepG2 cells compared to sh-NC group ( Figure 2D).

| Knockdown of DLEU1 impairs HCC cell migration and invasion
We next used wound healing and invasion assays to explore how  ( Figure 4I) cells. Taken together, these data indicated that miR-133a could directly bind to DLEU1 in HCC cells.

| miR-133a inhibitor rescued the inhibitory effect of HCC cells induced by DLEU1 depletion
To explore whether DLEU1 exerts biological functions through regulating miR-133a, we performed rescue experiment by inhibiting miR-133a expression in DLEU1 depletion-SMMC-7721 or HepG2 cells ( Figure 5A).
Moreover, we found that miR-133a inhibition partially reversed the inhibitory effect on cell proliferation, colony formation, cycle arrest, migration and invasion caused by DLEU1 depletion in both SMMC-7721 and HepG2 cells (Figur5B-F). These findings indicated that DLEU1 promoted HCC development partially by regulating miR-133a.

| DLEU1 regulated IGF-1R expression and PI3K/ AKT signal pathway via inhibition of miR-133a
Insulin-like growth factor 1 receptor, a known oncogene, was identified to serve as a downstream of miR-133a in HCC in our previous study. 18 Therefore, we wonder whether DLEU1 regulated IGF-1R via regulating miR-133a in HCC cells. sh-NC, sh-DLEU1 and sh-DLEU1 + miR-133a inhibitor were separately transfected into SMMC-7721 or HepG2 cells, then IGF-1R expression on mRNA and protein levels was determined by qRT-PCR and western blot analyses respectively. Our results demonstrated that knockdown of DLEU1 led to a significant decrease of IGF-1R mRNA ( Figure 6A) and protein expression ( Figure 6B) in both SMMC-7 and HepG2 cells, while miR-133a inhibitor partially reversed this trend. It was well known that IGF-1R could activate PI3K/AKT signalling pathways in HCC cells. 18,19 Here, we investigated whether DLEU1 affect activation of PI3K/ AKT pathway. We found that knockdown of DLEU1 significantly inhibited activation of PI3K/ AKT pathway in SMMC-7721 and HepG2 cells ( Figure 6B), while miR-133a inhibitor reversed the trends. We further investigated the correlation between the DLEU1, miR-133a and IGF-1R in HCC clinical samples. We found that DLEU1 expression was negatively correlated with miR-133a expression ( Figure 6C), while its expression was positively correlated with IGF-1R expression in HCC tissues ( Figure 6D). Collectively, these data suggest that DLEU1 modulated IGF-1R and PI3K/AKT pathway via regulation of miR-133a in HCC. We also examined the expression of DLEU1 and miR-133a in xenograft tumours. Our results showed that DLEU1 expression was obviously down-regulated ( Figure 7E), while miR-133a expression was up-regulated in sh-DLEU1 group compared with sh-NC group ( Figure 7F). In addition, we also investigated the effect of DLEU1 on IGF-1R expression and its downstream PI3K/AKT pathway in xenograft tumour. We found that silencing of DLEU1 significantly decreased IGF-1R expression and its downstream PI3K/AKT pathway ( Figure 7G). These results support the growth-promoting effect of DLEU1 in HCC in vivo. Therefore, we applied the Starbase2.0 software to identify the miRNAs that could bind to complementary sequences in DLUE1.

| D ISCUSS I ON
We found that miR-133a shares the complementary binding sites at DLEU1 3′-UTR, which was confirmed by the luciferase assay, RIP assay and qRT-PCR assay. Our published study showed that miR-133a expression was down-regulated in HCC, and functioned as tumour suppressor in HCC progression. 18 Moreover, the present study demonstrated that miR-133a inhibitor partially reversed the inhibitory effect caused by DLEU1 depletion. These results suggested that DLEU1 exerts tumour-promoting function in HCC via partially sponging miR-133a.
It has been shown that LncRNAs can affect the expression and biological functions of miRNA targets. 24,25 miR-133a was reported to exert tumour suppressor role in HCC by regulating IGF-1R 18 Growing evidence has supported the role of IGF-1R in promoting carcinogenesis and act as oncogene in multiple cancers including HCC. 26 In addition, abnormal expression of IGF-1R could regulate multiple downstream signal pathways including PI3K/AKT pathway. 27,28 Here, we found that knockdown of DLEU1 significantly decreased IGF-1R expression and inhibited the activation of PI3K/AKT pathway in SMMC-7721 and HepG2 cells, while miR-133a inhibitor reversed this trends.
Moreover, we found that DLEU1 expression was positively correlated with IGF-1R expression in HCC tissues. Interestingly, we found that knockdown of DLEU1 significantly inhibited DLEU1 expression, increased miR-133a expression, and inhibited IGF-1R expression and its downstream PI3K/AKT pathway in xenograft tumours. These data suggest that DLEU1 modulated IGF-1R and PI3K/AKT pathway via regulating miR-133a in HCC.
In conclusion, the present study demonstrated that DLEU1 was highly expressed in HCC tissues and its up-regulation was closely associated with TNM stage, vascular metastasis and poor overall survival ratio. DLEU1 could serve as an oncogenic lncRNA that promoted HCC tumourigenesis via acting as a ceRNA to regulate the expression of IGF-1R and its downstream PI3K/AKT signal pathway through directly sponging for miR-133a. These findings implied that DLEU1 might be a potential therapeutic target for HCC.