LncRNA HOTAIR/miR‐613/c‐met axis modulated epithelial‐mesenchymal transition of retinoblastoma cells

Abstract Since lncRNAs could modulate neoplastic development by modulating downstream miRNAs and genes, this study was carried out to figure out the synthetic contribution of HOTAIR, miR‐613 and c‐met to viability, apoptosis and proliferation of retinoblastoma cells. Totally 276 retinoblastoma tissues and tumour‐adjacent tissues were collected, and human retinoblastoma cell lines (ie, Y79, HXO‐Rb44, SO‐Rb50 and WERI‐RB1) were also gathered. Moreover, transfections of pcDNA3.1‐HOTAIR, si‐HOTAIR, miR‐613 mimic, miR‐613 inhibitor, pcDNA3.1/c‐met were performed to evaluate the influence of HOTAIR, miR‐613 and c‐met on viability, apoptosis and epithelial‐mesenchymal transition (EMT) of retinoblastoma cells. Dual‐luciferase reporter gene assay was also arranged to confirm the targeted relationship between HOTAIR and miR‐613, as well as between miR‐613 and c‐met. Consequently, up‐regulated HOTAIR and down‐regulated miR‐613 expressions displayed associations with poor survival status of retinoblastoma patients (P < 0.05). Besides, inhibited HOTAIR and promoted miR‐613 elevated E‐cadherin expression, yet decreased Snail and Vimentin expressions (P < 0.05). Simultaneously, cell proliferation and cell viability were also less‐motivated (P < 0.05). Nonetheless, c‐met prohibited the functioning of miR‐613, resulting in promoted cell proliferation and viability, along with inhibited cell apoptosis (P < 0.05). Finally, HOTAIR was verified to directly target miR‐613, and c‐met was the direct target gene of miR‐613 (P < 0.05). In conclusion, the role of lncRNA HOTAIR/miR‐613/c‐met signalling axis in modulating retinoblastoma cells’ viability, apoptosis and expressions of EMT‐specific proteins might provide evidences for developing appropriate diagnostic and treatment strategies for retinoblastoma.

late-stage retinoblastoma that represented high chance of ocular extraction, early diagnosis and treatment for retinoblastoma appeared to be quite critical. 2,3 To save children's visual function, 4 diverse treatment strategies have been developed for varying degrees of retinoblastoma. 5 Among them, chemotherapy combined with adjuvant therapy has gradually become the preferred treatment, although radiotherapy could cause children's facial deformity and relapse of retinoblastoma. 6 More than that, the confined efficacies of the therapies also necessitated exploration of novel treatment biomarkers.
Long non-coding RNAs (lncRNAs) have been demonstrated to matter much in the aetiology of glioma, breast cancer, hepatocellular carcinoma, prostate cancer and retinoblastoma. [7][8][9][10] Mounting evidences also indicated that lncRNAs could serve as proto-oncogenes or anti-oncogenes by regulating tumour-related genes or pathways. 11,12 Of note, a microarray analysis confirmed that lncRNAs HOX antisense intergenic RNA (HOAIR), CCAT1 and HIF1A-AS1 were over-expressed within retinoblastoma tissues, yet lncRNAs MIR31HG, BCAR4 and H193 were evidently under-expressed within retinoblastoma tissues. 13 Among them, HOTAIR was identified to be a cis-acting parameter within the HOXC loca, [14][15][16] and its carcinogenicity has been authenticated within breast cancer, liver cancer, colorectal cancer, oesophageal squamous cancer, pancreatic cancer, non-small cell lung cancer and cervical cancer. [17][18][19][20][21][22][23][24] In addition, HOAIR was indicated as a potential therapeutic biomarker for retinoblastoma, and silencing of HOTAIR could attenuated proliferation, migration and invasion of retinoblastoma cells were determined after. 25 One vital mechanism underlying the involvement of HOTAIR in regulating tumour cell proliferation was to modify expressions of relevant miRNAs by acting as competitively endogenous RNAs. For example, HOTAIR could bind to inter-cellular miR-331-3p and miR-124 based on its characteristic RNA sponging, thereby up-regulating cancer cells' capacity to reproduce. 26 The sponging relationship between HOTAIR and miR-613 was also manifested within carcinomas, such as non-small cell lung cancer and pancreatic cancer. 27,28 The miR-613 herein was suggested as a suppressor for retinoblastoma formation by targeting E2F5. 29 However, whether HOTAIR could sponge miR-613 to modulate osteosarcoma progression remained inconclusive.
Besides, miR-613 was estimated to prohibit osteosarcoma development by directly targeting c-met, 30    Moreover, the cells were cultured in 5% CO 2 and 90% humidity at 37°C. The medium would be replaced for every 2 or 3 days.

| Micro-array analysis of lncRNA profile
We randomly gathered 5 pairs of retinoblastoma tissues and corre-

| Cell transfection
The

| Plate colony assay
The cells were inoculated within 6-well plates at the density of 1000/well. After 8-d culture, cell medium was discarded, and the cells were fixed with methanol. The colonies with >50 cells were calculated after staining cells with crystal violet.
The culture medium would be placed in 5% CO 2 at 37°C for 2 hours.
Microplate reader (Molecular Devices, Sunnyvale, CA, USA) was applied to measure the value of OD 450 .

| Apoptosis assay
The

| Western blotting
The cells were lysed within RIPA buffer that contained 1% PMSF, and the proteins were loaded onto SDS-PAGE microgels and transferred to PVDF membranes. Then 50 g/L skim milk powder  were assessed with Spearman correlation analysis method. Furthermore, the enumeration data were compared and analysed with chisquare test. The Bilateral P values <0.05 were considered as statistically significant.
Our RT-PCR results also verified that MEG3, HOTAIR, CCAT1 expressions within retinoblastoma tissues were far beyond those within para-carcinoma tissues (P < 0.01) ( Figure 1B). We chose HOTAIR for the following experiments, due to its relative stable and marked expression within retinoblastoma tissues in comparison to normal tissues.
In addition, miR-613 expression followed the trend opposite to HOTAIR, regarding its expression within retinoblastoma tissues (P < 0.05) (Figure 2A). It was also displayed that HOTAIR and miR-613 expressions within Y79, HXO-Rb44, SO-Rb50 and WERI-RB1 cell lines were profoundly different from those within ACBRI-181 cell line (P < 0.05) ( Figure 2B). Meanwhile, there was a negative correlation between HOTAIR and miR-613 expressions within retinoblastoma tissues (P < 0.05) ( Figure 2C). Furthermore, the expression of miR-613 was evidently lower in retinoblastoma patients with tumours that were located bilaterally, poorly or non-differentiated, larger than 10 mm, classified as T3 or T4 and were with lymph nodal metastasis (P < 0.05), whereas the expressions of HOTAIR was enormously higher within the above tumours than ones that were located unilaterally, well or moderately differentiated, <10 mm, classified as T1 or T2 and were without lymph nodal metastasis (P < 0.05) ( Table 2). As the multi-factor regression analysis showed, higher HOTAIR expression, lower miR-613 expression, larger tumour size and grade T3 + T4 had significantly statistical relevance with the undesirable OS rate (P < 0.05) ( Table 3) (Figure 2D-E).

| HOTAIR and miR-613 regulated viability, apoptosis and EMT process of human retinoblastoma cells
HOTAIR expression was up-regulated after transfection of pcDNA3.1-HOTAIR (P < 0.05), and was down-regulated after treatment with si-HOTAIR (P < 0.05) ( Figure 3A). Following a similar trend, miR-613 was highly expressed after treatment with miR-613 mimic (P < 0.05), and was lowly expressed after treatment with miR-613 inhibitor (P < 0.05). Moreover, when compared with NC group, the proliferative ability of Y79 and HXO-Rb44 cells in si-HOTAIR and miR-613 mimic groups were conspicuously inhibited, but the capacity was well as down-regulated E-cadherin expression (P < 0.05) ( Figure 4B). HOTAIR&autoId=1819&orgTable=mirLncRNAInteractionsAll). The results of dual luciferase reporter assay also displayed that the luciferase activity of cells transfected with pmirGLO-HOTAIR-WT and miR-613 mimic was conspicuously lower than that of NC group (P < 0.05) ( Figure 5A-B). Hardly, any significant distinctions were detected between cells transfected with pmirGLO-HOTAIR-MUT and miR-613 mimic and cells of NC group (P > 0.05).

| C-MET expression was modified by HOTAIR
and miR-613 | 5091 B). More than that, the c-met expression was down-regulated under the influence of lowly expressed HOTAIR or over-expressed miR-613, while c-met expression was up-regulated when miR-613 was knocked out or when pcDNA3.1-HOTAIR plasmid was added (P < 0.05) ( Figure 6C) It was implied that HOTAIR promoted c-met expression, and miR-613 played an opposite role. It was also mirrored that the luciferase activity of miR-613 binding to c-met-Wt was significantly lower than that of its binding to c-met-Mut (P < 0.05) ( Figure 6D-E), suggesting the targeted relationship between miR-613 and c-met.

| HOTAIR and miR-613 modified proliferation, apoptosis and EMT process of retinoblastoma cells via regulation of c-met
According to Figure 7A, c-met expression was significantly promoted in the pcDNA3.1/c-met group, when compared with NC group (P < 0.05). The results of CCK8 and cell account displayed that the OD value of miR-NC+pcDNA3.1/c-met group was obviously higher than that of miR-NC group (P < 0.05) ( Figure 7B-C), suggesting that the weakened effects induced by miR-613 on retinoblastoma cells' proliferation were inhibited by c-met. Through examination of EMT-related proteins, down-regulated E-cadherin and up-regulated N-cadherin, vimentin or α-SMA were found in miR-NC + pcDNA3.1/c-met group when compared with miR-NC group (P < 0.05) ( Figure 8A). Flow cytometry showed that the apoptotic rate of miR-NC + pcDNA3.1/c-met group was markedly lower than that of miR-NC group (P < 0.05) ( Figure 8B), reflecting that c-met inhibits the increased apoptosis of retinoblastoma cells induced by miR-613. All in all, it was show that c-met appeared as an important regulatory factor regulating the tumorigenicity of retinoblastoma cells.

| DISCUSSION
Retinoblastoma, an intraocular malignancy commonly found among children, possessed a worldwide prevalence of up to 1/200 000, [32][33][34] and the novel cases per year within China accounted for as high as 20% of that within the global range. 35 Despite the emergence of  REST co-inhibitory protein (CoREST). 43 It was also involved with epigenetic modulation of genes, and its interaction with polycomb repressive complex 2 (PRC2) turned as a crucial process within various cellular pathways. 44 Our study also demonstrated that HOTAIR acted as an oncogene for promoting retinoblastoma development (Figures 3-4).
Of note, lncRNAs could act as a competitive endogenous RNA (ceRNA) to affect miRNA expression and to influence actions of miRNA target genes, thereby modifying tumour processes, such as cell multiplication and cell apoptosis. 45,46 For example, HOTAIR could also serve as the competitive RNA for PIK3R3 to sponge miR-214 and miR-217, so then elevating PIK3R3 expression within ovarian cancer cells. 47 Furthermore, HOTAIR could sponge miR-1, miR-214-3p and miR-330-5p to raise MAPK1 expression, through which the proliferation, migration and invasion of ovarian carcinoma cells (ie, SKOV3) were largely facilitated. 48 The present study manifested that HOTAIR sponged miR-613 to encourage the EMT process within retinoblastoma cells ( Figure 4). Interestingly, another study insinuated that miR-613 was subjected to control of lncRNA DANCR in simultaneously acting on exacerbation of retinoblastoma. 49 It could be explained that several lncRNAs co-modified miR-613 to accelerate or decelerate the aggravation of retinoblastoma.
As for miR-613, its diminished expression seemed as a predictor for improved progression-free survival and overall survival of ovarian cancer patients, 50 and it also enhanced gastric cancer cells to progress by repression of brain-derived neurotrophic factor. 51 Our study displayed that c-met was the downstream molecule of miR-613, and they collaborated to modulate proliferation, viability, apoptosis and EMT process of retinoblastoma cells ( Conclusively, our study results initially manifested that highly expressed HOTAIR could promote development and aggravation of retinoblastoma by miR-613/c-met axis, providing strong evidences for developing potential diagnostic biomarkers and treatment targets for retinoblastoma. However, further investigations were in demand to verify the consequences of this study.

CONFLI CT OF INTEREST
None.