Long intergenic non-protein coding RNA 662 accelerates the progression of gastric cancer through up-regulating centrosomal protein 55 by sponging microRNA-195-5p

ABSTRACT Long non-coding RNAs (lncRNAs) are important players in regulating diverse human diseases, including cancers. Nonetheless, the function of long intergenic non-protein coding RNA 662 (LINC00662) in gastric cancer (GC) carcinogenesis and progression remains to be delineated. In the present study, LINC00662, microRNA-195-5p (miR-195-5p) and centrosomal protein 55 (CEP55) mRNA expression levels were quantified by qRT-PCR. GC cell proliferation, migration and invasion were analyzed by CCK-8, BrdU and Transwell assays. Besides, dual-luciferase reporter and RNA pull-down assays were conducted for verifying the targeting relationships of LINC00662, miR-195-5p and CEP55. The regulatory functions of LINC00662 and miR-195-5p on CEP55 were examined utilizing Western blot. In this study, it was revealed that LINC00662 expression level was elevated in GC tissues and cells. LINC00662 overexpression facilitated the malignant biological behaviors of GC cells whereas knockdown of LINC00662 worked oppositely. In terms of mechanism, LINC00662 targeted miR-195-5p to modulate CEP55 expression. In conclusion, LINC00662 facilitates the malignant biological behaviors of GC cells via miR-195-5p/CEP55 axis, and therefore, it may be a promising target for GC treatment.


Introduction
Gastric cancer (GC) is a life-threatening cancer, which is with high incidence in East Asia [1,2]. A majority of patients unfortunately were in the advanced stage by the time of diagnosis, frequently accompanied by lymph node or distant metastasis [3]. Delineating the molecular mechanism of GC progression is vital.
Long non-coding RNAs (lncRNAs), incapable of encoding proteins, is able to interact with DNA, RNA or proteins, and they can regulate gene expression via multiple mechanisms [4,5]. In recent years, it has been elucidated that lncRNAs are associated with tumorigenesis [6][7][8]. Long intergenic non-protein coding RNA 662 (LINC00662), an emerging lncRNA identified in recent years, reportedly, is associated with the progression of oral squamous cell carcinoma, hepatocellular carcinoma and lung cancer [9][10][11]. Intriguingly, it is reported that silencing LINC00662 impedes GC cell proliferation and increases chemosensitivity [12]. Nonetheless, the detailed biological functions and underlying mechanism of LINC00662 in GC progression require further delineation.
Centrosomal protein 55 (CEP55), as a regulator in spindle assembly, exhibits regulatory functions in cytokinesis [18]. CEP55, overexpressed in glioma tissues and cells, facilitates cancer cell proliferation and suppresses apoptosis through PI3K/ Akt/p21 signaling pathway [19]. CEP55 also facilitates epithelial-mesenchymal transition of renal cell carcinoma cells via PI3K/AKT/mTOR pathway [20]. The abnormal high expression of CEP55 also contributes to GC progression [21]. Nevertheless, the upstream regulatory mechanism of CEP55 in GC warrants further investigation. This work was aimed to decipher how LINC00662 functioned in GC progression. It was hypothesized that, LINC00662 could promote GC progression as a competitive endogenous RNA (ceRNA). It was revealed that LINC00662 was highly expressed in GC, and it, via repressing miR-195-5p and up-regulating CEP55, increased the malignancy of GC cells.

Samples collection
From March 2018 to March 2019, GC tissues and adjacent non-tumor tissues obtained during surgery in Qinghai Provincial People's Hospital were rapidly frozen in liquid nitrogen. All of the subjects signed the informed consent. The inclusion criteria: patients with gastric adenocarcinoma receiving no neoadjuvant therapy. Exclusion criteria: adenocarcinoma of esophagogastric junction; patients receiving neoadjuvant therapy; patients with severe cardiovascular disease, cerebrovascular disease, metabolic diseases and other cancers; patients with incomplete clinicopathologic information. This work was permitted by the Ethics Committee of Qinghai Provincial People's Hospital (Approval No. 20,170,334). This collection and usage of human samples followed Declaration of Helsinki.

Subcellular fractionation
A Cytoplasmic and nuclear RNA purification kit (Amyjet, Wuhan, China) was adopted to extract cytoplasmic and nuclear RNA from SGC-7901 and BGC823 cells.

Bromodeoxyuridine (BrdU) staining
The transfected GC cells in each group were transferred onto coverslips (Beyotime, Shanghai, China) and then cultured for 12 h. Subsequently, the cells were incubated BrdU solution (Beyotime, Shanghai, China) for 6 h. The GC cells, after the medium was discarded, were fixed with 4% paraformaldehyde for 30 min, and incubated with anti-BrdU antibody (Beyotime, Shanghai, China) at ambient temperature for 1 h, and then rinsed with phosphate buffer saline (PBS). Subsequently, the nuclei were stained employing Hoechest staining solution (Beyotime, Shanghai, China). After the cells were washed by PBS again, the numbers of BrdU and Hoechst positive cells were counted.

Transwell assays
After transfection, BGC823 and SGC-7901 cell lines (1 × 10 5 cells/ml) were harvested for the preparation of single-cell suspensions with serumfree medium. Then, the upper and lower compartments of each Transwell chamber (Corning, Shanghai, China) were added with 200 μL of single-cell suspension and medium supplemented with 20% FBS, respectively. After 36 h, the cells upper surface of the membrane were removed, and the cells on the below surface of the membrane were fixed with 4% paraformaldehyde and then stained with 0.5% crystal violet solution. Finally, the stained cells were observed and counted under a microscope. Notably, the membranes, prior to usage in the invasion assay, were coated with a layer of Matrigel (5.00 μg/well; BD, San Jose, CA, USA).

Animal experiment was authorized by the Animal Protection and Use Committee of Qinghai Provincial
People's Hospital (Approval No. 20,200,085). Female BALB/c nude mice available from SIPPR-BK Laboratory Animal Co. Ltd. (Shanghai, China) were divided into NC group and LINC00662 overexpression group, 15 mice in each group. BGC823/NC and BGC823/LINC00662 cells were injected into the nude mice via caudal vein (1 × 10 7 per mouse). The mice, after 2 weeks, were euthanized, and the lung metastases were checked with hematoxylin-eosin (HE) staining by the pathologists.

Dual-luciferase reporter assay
PmiR-RB-REPORT™ (Ribobio, Guangzhou, China) reporter plasmids containing the wildtype LINC00662 and CEP55 3ʹ-UTR sequences with predicted binding sites for miR-195-5p were constructed. HEK293 cells were co-transfected with the reporter plasmids and miR-195-5p mimics or miR-con, and after 48 h, the cells were collected to examine the luciferase activities.

Western blot
RIPA lysis buffer (Beyotime, Shanghai, China) was adopted to conduct protein extraction, with the concentration of protein samples determined with a Bicinchoninic acid (BCA) kit (Beyotime, Shanghai, China). After SDS-PAGE, the proteins on the gel were transferred onto PVDF membranes (Beyotime, Shanghai, China). After that, the membranes were incubated with anti-CEP55 rabbit polyclonal antibody (ab170414, 1:1000, Abcam, Shanghai, China) and anti-GAPDH rabbit polyclonal antibody (ab181602, 1:2000, Abcam, Shanghai, China) overnight at 4°C, followed by the rinse with tris buffered saline tween (TBST) 3 times, with 15 min each time. Following that, the membranes were incubated, at ambient temperature, with horseradish peroxidase (HRP)-labeled secondary antibodies (1: 2000, Beyotime, Shanghai, China) for 1 h and subsequently, washed with TBST with the same procedure. After adding the enhanced chemiluminescence kit (Beyotime, Shanghai, China) onto the membranes, the protein bands were detected employing ChemiDoc™ Touch Imaging System (Bio-Rad, Shanghai, China).

Bioinformatics analysis
The targeting relationships among LINC00662, miR-195-5p and CEP55 were predicted by StarBase database (http://starbase.sysu.edu.cn/) [22]. SPSS 22.0 statistical software (SPSS Inc., Chicago, IL, USA) was employed to analyze the data, which were shown as 'mean ± standard deviation'. The difference between two groups was compared employing student's t-test. Pearson's correlation analysis was used to evaluate the correlation. P < 0.05 implicated statistical significanc.

Result
The present work explored the expression levels of LINC00662 and miR-195-5p in GC, and investigate their biological function in modulating the malignant phenotypes of GC cells with in vitro and in vivo experiments. Additionally, the regulatory functions of LINC00662 and miR-195-5p on CEP55 were studied. Our study showed that LINC00662, miR-195-5p and CEP55 formed a ceRNA network to regulate the progression of GC.

LINC00662 and miR-195-5p in GC tissues and cells
First of all, the expression levels of LINC00662 and miR-195-5p in GC tissues and cell lines were detected by qRT-PCR. It was revealed that, LINC00662 was markedly elevated while miR-195-5p expression was observably decreased in GC tissue samples (Figure 1(a,b)). Additionally, LINC00662 was uncovered to be highly expressed whereas miR-195-5p was under-expressed in GC cells (Figure 1(c,d)).

LINC00662 targeted miR-195-5p to facilitate the malignant phenotypes of GC cells
For deciphering how LINC00662/miR-195-5p axis worked in GC progression, BGC823 cells were transfected with LINC00662 overexpression plasmid or co-transfected with miR-195-5p mimics while SGC-7901 cells were transfected with si-LINC00662#3 or co-transfected with miR-195-5p inhibitors (Figure 3(a)). As shown, LINC00662 overexpression promoted the malignant phenotypes of BGC823 cells, whereas knockdown of LINC00662 worked oppositely on SGC-7901 cells (Figure 3(b-f)). Besides, transfecting miR-195-5p mimics weakened the promoting impacts of LINC00662 overexpression on the malignant biological behaviors of GC cells, whereas the inhibitory impacts of si-LINC00662#3 were abolished by miR-195-5p inhibitors (Figure 3(b-f)). Besides, the number of lung metastases in mice injected with BGC823 cells with LINC00662 overexpression was significantly increased; 10 of 15 mice in the BGC823/LINC00662 group showed severe lung metastasis, and the incidence was observably higher than that of BGC823/NC group (2 of 15 mice) (Supplementary Figure S1). Collectively, it was concluded that LINC00662 facilitated the malignant phenotypes of GC cells, probably via sponging miR-195-5p.

CEP55 was the target of miR-195-5p
For clarifying how miR-195-5p worked in GC, StarBase database (http://starbase.sysu.edu.cn/) was searched again, and it predicted miR-195-5p targeted CEP55 (Figure 4(a)). Subsequently, in dual-luciferase reporter assay, it was revealed that the transfection of miR-195-5p mimics suppressed the luciferase activities of WT, MUT1 and MUT2 CEP55 reporters, whereas no effects were observed on that of MUT1&2 CEP55 reporter (Figure 4(b)). Besides, miR-195-5p in BCG823 cells was elevated after the transfection of miR-195-5p mimics, whereas miR-195-5p in SGC-7901 cells was up-regulated after the transfection of miR-195-5p inhibitors (Figure 4(c)). Moreover, miR-195-5p overexpression repressed the expression of CEP55 mRNA and protein in Figure 2. MiR-195-5p was the target of LINC00662. a. qRT-PCR was used to detect LINC00662 expression in nuclear and cytoplasmic fractions of BGC823 and SGC-7901 cells. B. The binding sequences between miR-195-5p and LINC00662 were predicted. c. Luciferase activities of the cells were detected to validate the predicted binding sites. d-e. RNA pull-down assay was adopted to validate the interaction between miR-195-5p and LINC00662. F. qRT-PCR data were utilized to examine the correlation between miR-195-5p expression and LINC00662 expression in GC samples (N = 45). g. LINC00662 overexpression plasmids were transfected into BGC823 cells, and siRNAs were transfected into SGC-7901 cells, and then the expression of LINC00662 in GC cells was detected by qRT-PCR. h. qRT-PCR was utilized to quantify miR-195-5p expression in GC cell lines after transfection. All of the experiments were performed in triplicate. **P < 0.01, ***P < 0.001 and ns: P > 0.05.  assay (c, d). e-f. Transwell experiment was adopted for assessing the migration and invasion of transfected GC cells. All of the experiments were performed in triplicate. *P < 0.05, **P < 0.01 and ***P < 0.001. BGC823 cells but inhibiting miR-195-5p induced the expression of CEP55 mRNA and protein in SGC-7901 cells (Figure 4(d,e)). Additionally, as opposed to normal tissues adjacent to cancer, CEP55 mRNA and protein exhibited enhanced expression in GC tissues (Figure 4(f,g)). Besides, miR-195-5p expression was in negative association with CEP55 mRNA expression in GC tissue samples (Figure 4(h)). These data suggested that CEP55 was a target gene of miR-195-5p and could be negatively modulated by it.

LINC00662 accelerated GC progression by regulating CEP55
Next, the correlation between CEP55 mRNA expression and LINC00662 expression in GC tissues was analyzed, and the results showed that there was a positive correlation ( Figure 5(a)). Subsequently, BGC823 cells were transfected with LINC00662 overexpression plasmid or cotransfected with si-CEP55; SGC-7901 cells were transfected with si-LINC00662#3 or co- a. Bioinformatics analysis was employed for predicting the binding sequences between miR-195-5p and 3ʹUTR of CEP55. B. Luciferase activity was performed to validate the predicted binding sites after reporter vectors and miR-195-5p mimics (or miR-con) were cotransfected into the cells. c. qRT-PCR was adopted to quantify miR-195-5p in GC cells after the cells were transfected with miR-195-5p mimics or inhibitors. d-e. CEP55 mRNA and protein in GC cells with miR-195-5p overexpressed or inhibited were quantified by qRT-PCR (d) and Western blot (e). F. qRT-PCR was utilized to detect the expression level of CEP55 mRNA in GC tissues (N = 45) and non-cancerous tissues (N = 45). g. Western blot was adopted to quantify CEP55 protein in GC tissues (N = 3) and non-cancerous tissues (N = 3). h. The association between miR-195-5p and CEP55 mRNA expressions in GC samples (N = 45) was evaluated. All of the experiments were performed in triplicate. ***P < 0.01 and ns: P > 0.05. a. The correlation between LINC00662 and CEP55 mRNA expressions in GC samples (N = 45) was evaluated. b-d. BGC823 cells were transfected with LINC00662 overexpression plasmid or co-transfected with si-CEP55, respectively, and SGC-7901 cells were transfected with si-LINC00662#3 or co-transfected with CEP55 overexpression plasmid. CEP55 mRNA and protein were quantified by qRT-PCR (c) and Western blot (d). e-f. The proliferation of transfected cells was measured employing CCK-8 assay (E) and BrdU assay (F).G. Transwell experiment was conducted to measure the migration and invasion of transfected cells. All of the experiments were performed in triplicate. *P < 0.05, **P < 0.01 and ***P < 0.001. transfected with CEP55 overexpression plasmid ( Figure 5(b-d)). It was demonstrated that CEP55 knockdown partly abrogated the promoting effects of LINC00662 overexpression on malignant biological behaviors of GC cells whereas overexpression of CEP55 partly counteracted the inhibitory effects of LINC00662 knockdown on these biological behaviors ( Figure 5(e-g)). These data suggested that LINC00662 was a participant in regulating GC progression via modulating CEP55.

Discussion
LncRNA is widely involved in multiple biological processes such as cell proliferation, differentiation, autophagy, apoptosis and so on [23]. Many lncRNAs are aberrantly expressed in human cancers including GC, and they are pivotal regulators in the growth, migration, invasion, apoptosis and drug resistance of GC cells. For instance, lncRNA CERS6-AS1 promotes the progression of GC [24]. What's more, LINC00240 promotes GC progression via modulating miR-338-5p/METTL3 axis [25]. Additionally, knockdown of LINC00483 represses GC cell viability and induces apoptosis [26]. LINC00662 exhibits a carcinogenic role in multiple types of human cancers. For instance, LINC00662 expedites the proliferation and invasion of liver cancer cells by activating Wnt/β-catenin signaling [10]. Besides, LINC00662, facilitating the metastasis of lung cancer cells through interaction with LIN28, is in association with the unfavorable prognosis of lung cancer patients [11]. We, herein, observed that LINC00662 was remarkably up-regulated in GC tissue samples and cell lines. Furthermore, LINC00662 overexpression promoted the malignant phenotypes of GC cells, but its knockdown exhibited opposite effects. These data suggest that LINC00662 plays a carcinogenic role in GC and was a potential biomarker and therapeutic target for GC.
CEP55 (also known as FLJ10540 or C10orf3), as a component of the centrosome of cells, is recognized as a regulator of cell cycle and mitosis [18]. CEP55 participates in spindle formation by anchoring microtubule polymerization associated proteins, thereupon regulating cell proliferation [18]. Except in thymus and testis, CEP55 is weakly expressed in other mature tissues [35]. Overexpression of CEP55 can trigger the dysfunction of mitosis, giving rise to the presence of multinucleated cells [36]. CEP55 is in close relationship with the occurrence and development of human cancers [19][20][21]. For instance, CEP55, highly expressed in osteosarcoma, facilitates the proliferation and invasion of tumor cells by activating AKT signaling [37]. In pancreatic cancer, it is in association with the adverse prognosis of the patients; its overexpression promotes the aggressiveness of cancer cells via activating NF-κB signaling [38]. Importantly, it is reported that knockdown of CEP55 suppresses GC cell proliferation by inducing G2/M cell cycle arrest [21]. Herein, it was revealed that, CEP55 was negatively regulated by miR-195-5p. What's more, knockdown of CEP55 counteracted the effects of LINC00662 overexpression on GC biological behaviors, indicating that LINC00662 participated in GC progression by regulating miR-195-5p/CEP55 axis.

Conclusion
Our work authenticates that LINC00662 is capable of regulating CEP55 by targeting miR-195-5p, thereupon facilitating the malignant biological behaviors of GC cells. Our work presents novel insights into the diagnosis and therapy of GC.

Research highlights
(1) LINC00662 is up-regulated in gastric cancer tissues, and it promotes the malignant phenotypes of gastric cancer cells.