Long Noncoding RNA LINC00202 Promotes Tumor Progression by Sponging miR-3619-5p in Retinoblastoma

. Retinoblastoma (RB) is the most common intraocular malignancy in childhood, and the prognosis in the advanced RB is poor. It is urgent to find novel therapeutic targets. Long noncoding RNAs (lncRNAs) have critical functions in cancer progression, and lncRNA LINC00202 is found associated with poor prognosis in RB. However, the functions of LINC00202 in RB remain unclear. We employed qRT-PCR and immunoblot to detect the expression levels of mRNAs and proteins, respectively. Cell proliferation was determined by CCK-8 assay and colony formation assay. Transwell assays were applied to evaluate the cell abilities of migration and invasion. Luciferase reporter assay was applied to examine RNA stability, and RNA pulldown assays were used to detect interaction between lncRNA and microRNA (miRNA). LINC00202 expression in RB tissues is higher than that in the paired adjacent normal tissues, which has correlation with poor prognosis in RB. RB cell proliferation, migration and invasion were weakened by LINC00202 depletion, but enhanced by LINC00202 overexpression. MiR-3619-5p was identified to directly bind and mediate LINC00202-promoted RB progression, meanwhile, miR-3619-5p directly regulated expression of an oncongene, RIN1. Moreover, RIN1 knockdown completely blocked miR-3619-5p-enhanced RB progression. In summary, high LINC00202 levels are correlated with poor prognosis in RB, and it promotes RB progression by sponging miR-3619-5p and therefore up-regulating RIN1 expression.


Introduction
Retinoblastoma (RB) is the most common malignant neural retinal intraocular tumor of childhood, affecting one out of 16,000-18,000 neonates (Mansoor et al., 2016;Villegas et al., 2013). RB is curable at the early stage, and the survival rates of the patients with early stage RB are nearly 100% † These authors contributed equally to this work. *To whom correspondence should be addressed: Lianzhi Yu (Broaddus et al., 2009). However, the survival rates are dramatically declined in undeveloped countries or regions, only 20-46% in Africa (Bowman et al., 2008) due to delayed diagnosis and treatment (Dimaras et al., 2010). The advanced RB can grow into the periocular tissues and finally into the brain, and even metastases to distant organs leading to death (Villegas et al., 2013). Although some treatments have been developed to benefit RB patients, the prognosis in the advanced RB is still poor (Sengupta et al., 2016). Thus, it is urgent to discover new effective therapeutic targets.
Long noncoding RNAs (lncRNAs) and microRNAs (miRNAs) are two classes of important RNA regulators engaged in cancer development (Ayers and Vandesompele, 2017). LncRNAs are a kind of RNAs more than 200 nucleotides in length that do not encode proteins (Quinn and Chang, 2016). LncRNAs regulate multiple biological processes, including cell proliferation, tumorigenesis and metastasis (Wang and Chang, 2011). For instance, lncRNA BANCR is involved in RB cell growth and metastasis, and its higher levels are related to poor prognosis in RB (Su et al., 2015). RB development is regulated by lncRNA BDNF-AS (Shang et al., 2018). MiRNAs are small RNAs with a length of 21-25 nucleotides and inhibit gene expression by either degrading the targeted transcripts or inhibiting translation of the targeted mRNAs. One approach of lncRNA exerting its functions is to sponge miRNAs and down-regulate their levels, that in turn up-regulates the expression levels of miRNA-targeted genes (Quinn and Chang, 2016). For example, lncRNA MALAT1 elevates Slug expression via miR-124 to modulate RB cell proliferation, migration and invasion (Liu et al., 2017). H19 directly binds miR-17-92 to de-suppress p21 expression, and consequently prevents RB progression (Zhang et al., 2018a). LINC00202 is a lncRNA related to poor prognosis in RB. It has been reported that it is also associated with overall survival (OS) time in patients with renal cancer . However, the biological functions of LINC00202 have not been clarified.
The expression of Ras and Rab interactor 1 (RIN1) expression increases and is related to poor prognosis in many tumors, including non-small cell lung cancer , melanoma (Fang et al., 2012), gastric adenocarcinoma  and renal cancer (Wei et al., 2015). It has been reported that RIN1 overexpression aggravated tumor cells through activating EGFR signaling (Feng et al., 2017). RIN1 is predicted to be one of the potential targets of miR-3619-5p, while miR-3619-5p was predicted to be one of LINC00202 targets. In this study, we investigated the functions of LINC00202 in RB and the underlying mechanism.

Tissue samples and Patients
Fifty cases of RB tissues and the paired adjacent normal tissues were obtained from the Affiliated Yantai Yuhuangding Hospital of Qingdao University. Once dissected out of patients, the tissues were immediately frozen in liquid nitrogen until use. This study had been approved by The Human Research Ethics Committee of the hospital and patients had written informed consents.

Cell culture and establishment of stable cell lines
The cells were cultured in RPMI-1640 (Gibco, ThermoFisher, Grand Island, NY, USA) containing 10% fetal bovine serum (FBS) (Hyclone, USA), and maintained in a humidified incubator at 37°C in the presence 5% CO 2 .
To establish the cell line stable expressing LINC00202, LINC00202 DNA fragment was amplified by PCR with primers 5'-CTAGAATTCGTCAGCCATATGCCT-3' and 5'-CTAGGAT-CCTGAACGTTTAACTTTTA-3'. The DNA fragment was cloned into pBabeMNires vector. Weri-Rb1 cells were transfected with LINC00202 overexpression plasmid and control vector, followed by puromycin selection for 1 week to obtain stable cell lines.

Cell proliferation assay
CCK-8 assay and colony formation assay were conducted to measure cell proliferation. To execute CCK-8 assay using a CCK-8 assay kit (CCK-8 assay kit, Dojindo, Japan), cells were plated into 96-well plate and cultured for 24 h, followed by starvation by removal of FBS for 12 h prior to adding 10 μl CCK-8 reagent. The absorbance at 450 nm was examined. Each group had 6 repeats, and the average was calculated. To do colony formation assay, cells were plated into 6-well culture plate at a density of 1000 per well and cultured for 2 weeks in a cell incubator. The cells were stained with 0.1% crystal violet (Beyotime, Shanghai, China) for 0.5 h after fixed with 4% paraformaldehyde for 0.5 h. The number of colonies was counted under a microscope.

Transwell migration and invasion assays
For transwell migration assay, 1×10 5 cells were plated into the upper chamber of the transwell chamber (Millipore, Billerica, MA, USA). To do transwell invasion assay, 1×10 5 cells were plated to the matrigel-coated upper chamber (BD Biosciences, San Jose, CA, USA). For both assays, the upper chamber was added with FBS-free medium, and the lower chamber was supplied with 10% FBS-contained medium. After cultured for 24 h, the cells staying on the upper surface of membrane were swiped off, and the membrane was dyed with 0.5% crystal violet in 20% methanol. The cell abilities to migration and invasion were assessed by counting the number of cells through the membrane.

Western blot
Cells were lyzed in RIPA buffer (50 mM Tris-HCl, pH 7.5, 0.1% SDS, 1% Nonidet P-40, 0.1% sodium deoxycholate, 150 mM NaCl). The protein samples were subject to electrophoresis in SDS-PAGE gels, and then the separated proteins were transferred onto PVDF membranes (Millipore, Bedford, USA). The membranes were blocked with 5% non-fat milk, before incubated with primary antibodies to RIN1 (Abcam, Cambridge, MA, USA) and β-actin (Sigma, St. Louis, MO, USA) at 4°C overnight. After washed 3 times with washing buffer, the membranes were blotted with HRP-coupled secondary antibodies (Promega). The enhanced chemiluminescence Detection System (ThermoFisher, USA) was used to detect Protein bands.

Statistical analysis
In this study, the data were expressed as the mean value±SD (standard deviation). P values were calculated by Student's t-test or ANOVA analysis and P<0.05 was defined as statistical significance difference.

LINC00202 is up-regulated in RB and high LINC00202 expression predicts poor prognosis
LINC00202 is one of the lncRNAs associated with overall survival (OS) time in patients with renal cancer . To investigate if LINC00202 functions in RB, we first examined this lncRNA expression in RB tissues by qRT-PCR. LNC00202 expression in RB tissues was significantly richer comparing to the adjacent normal tissues (Fig.  1a). Similarly, LINC00202 mRNA levels expressed in RB cell lines, including Weri-Rb1, SORB50, HXO-RB44 and Y79, were 3 to 10 folds higher than those in normal retinal pigment epithelial cell lines, ARPE-19 and hTERT-RPE1 (Fig. 1b). These data demonstrate that LINC00202 is upregulated in RB. Next, we explored correlation between LINC00202 expression levels and OS rate or disease-free survival (DFS) rate in 50 RB cases. We divided RB patients into 2 groups according to LINC00202 expression levels, i.e. low LINC00202 expression and higher LINC00202 expression, and each group contained 25 cases. Kaplan-Meier survival analysis revealed that the group of high LINC00202 expression had a significantly lower OS rate (Fig. 1c) and a lower DFS rate (Fig. 1d), suggesting that high LINC00202 expression level is related to poor prognosis and LINC00202 is an oncogene in RB.

LINC00202 regulates RB cell proliferation, migration and invasion
To explore the function of LINC00202 in RB, LINC00202 LINC00202 Promotes Retinoblastoma Progression expression in RB cells was disturbed by knockdown or overexpression. Y79 cells express the highest LINC00202 level in all examined RB cell lines, so we knocked down LINC00202 in Y79 cells. Two LINC00202 shRNAs both efficiently reduced LINC00202 expression, and the knockdown efficiency was over 60% (Fig. 2a). LINC00202 knockdown significantly retarded cell proliferation (Fig.  2b). Y79 ability of colony formation was also severely impaired by LINC00202 shRNAs (Fig. 2c). As shown in the right panel of Fig. 2c, the numbers of colonies in LINC00202 knockdown cells were reduced by over 60%. These data indicate that LINC00202 regulates Y79 cell proliferation. Meanwhile, transwell assays showed that Y79 abilities of both migration and invasion were declined by over 50% in response to LINC00202 knockdown (Fig. 2d). In contrast, LINC00202 overexpression in Weri-Rb1 cells (Fig. S1a), which expressed the lowest LINC00202 level among all examined RB cell lines (Fig. S1b), significantly promoted cell viability (Fig. S1b) and increased the number of colonies over 3 folds in the colony formation assay (Fig.  S1c). Meantime, transwell assays unveiled that Weri-Rb1 abilities of both migration and invasion were elevated over 3 folds in response to LINC00202 overexpression (Fig.  S1d).
In summary, LINC00202 is engaged in manipulating RB cell proliferation, migration and invasion.

Discussion
This study showed that lncRNA LNC00202 was a biomarker for prognosis in RB, and higher LNC00202 expression could predict poor prognosis. Furthermore, we elucidated LNC00202 functions in RB cells and the related mechanism. LNC00202 promoted RB cell progression by inhibiting miR-3619-5p and thereby leading to upregulation of miR-3619-5p targeted oncogene, RIN1. Thus, LNC00202, miR-3619-5p and RIN1 form an axis to modulate RB cell progression. This is the first study reporting LNC00202 pro-malignant functions as well as the LNC00202/miR-3619-5p/RIN1 regulation axis in RB.
Recently, lncRNAs have gained attention as their involvement in many physiological and pathological processes (Quinn and Chang, 2016). Accumulating literatures have highlighted that the expression change of lncRNAs has critical and clinically predictive roles in tumorigenesis (Elchuri et al., 2018). More and more lncRNAs were repor- Fig. 4. MiR-3619-5p inhibits RB cell proliferation, migration and invasion. Y79 and Wei-Rb1 cells were treated with negative control mimic (miR-NC) or miR-3619-5p mimic (miR-3619-5p). The cells were then used for qRT-PCR analysis of miR-3619-5p expression (a), CCK-8 assay (b, c) and colony formation assay (d) to analyze cell viability, and transwell assays to examine cell abilities to migration and invasion (e, f). The data display the mean±SD based on three independent experiments. *P<0.05; **P<0.01; ***P<0.001.
Several studies have reported that miR-3619-5p acts as a tumor suppressor in various cancers. MiR-3619-5p decreases β-catenin expression to suppress progression of nonsmall cell lung cancer (Niu et al., 2015) and bladder cancer (Zhang et al., 2018b). MiR-3619-5p binds to the promoter of CDKN1A and induces its expression to inhibit prostate cell proliferation (Li et al., 2017). This study shows that miR-3619-5p also had a tumor suppressive function in RB, and RIN1 was identified as a new miR-3619-5p target.
RIN1 possesses an oncogenic role via Ras signaling pathway (Han et al., 1997). High RIN1 expression is related to tumor progression and poor prognosis in bladder urothelial carcinoma (Shan et al., 2012), gastric adenocarcinoma  , melanoma (Fang et al., 2012) and non-small cell lung cancer .
In conclusion, LINC00202 could be a prognosis biomarker for RB, and has an important role in RB cell prolif- Fig. 5. LINC00202 modulates RB cell progression via regulating miR-3619-5p. Y79 cells stably expressing LINC00202 shRNAs were transfected with negative control miRNA mimic (miR-NC) or miR-3619-5p mimic (miR-3619-5p). Then the cells were applied for qRT-PCR analysis of miR-3619-5p expression (a), CCK-8 assay (b) and colony formation assay (c) to analyze cell viability, and transwell assays to examine cell abilities to migration and invasion (d). The data display the mean±SD based on three independent experiments. *P<0.05; **P<0.01. eration, migration and invasion via inhibiting miR-3619-5p and consequently inducing RIN1 expression. Our results imply that LINC00202 could be a potential therapeutic target for RB, and miR-3619-5p is a promising drug candidate for RB therapy.

Conclusion
High LNC00202 level is related to poor prognosis in RB, and LNC00202 promotes RB proliferation, migration and invasion via miR-3619-5p and its target RIN1. Fig. 6. RIN1 is a direct target of miR-3619-5p in RB cells. MiR-3619-5p binding site within RIN1 3'UTR was predicted (a), and the DNA fragments of wild-type (WT) and mutant (MT) sequences (a) were cloned into luciferase reporter vector. Luciferase reporter assay was conducted in Weri-Rb1 (b) and Y79 cells (c) after co-transfection with these luciferase reporter plasmids and miR-3619-5p or miR-NC. Weri-Rb1 and Y79 cells were treated with miRNA mimic, miR-3619-5p or its inhibitor, and subsequently qRT-PCR and Western blot were conducted to detectbRIN1 mRNA levels (d) and protein levels (e), respectively. Weri-Rb1 cells stably overexpressing LINC00202 were treated with miR-3619-5p, or Y79 cells stably expressing LINC00202 shRNA were treated with miR-3619-5p inhibitor, and then qRT-PCR and Western blot were conducted to detect RIN1 mRNA levels (f, g) and protein levels (h, i), respectively. The data display the mean±SD based on three independent experiments. *P<0.05; **P<0.01.