Upregulated LINC00565 Accelerates Ovarian Cancer Progression By Targeting GAS6

Background

Ovarian cancer (OC) is one of the familiar malignant cancer of the female reproductive system.¹ The overall incidence of OC ranks second among gynecological malignant tumors.² OC, with the high mortality rate, causes serious damage to female physical and mental health, especially in postmenopausal women aged 55–65.³ Epithelial ovarian cancer (EOC) is the most common histological type of OC, accounting for 85% to 90%. Because the early symptoms of OC are insidious, malignant degree of OC is high, the majority of OC patients are already at an advanced stage when diagnosed.^4,5 Therefore, early diagnosis and effective treatment are the key factors to improve the survival and prognosis of OC.

Long non-coding RNAs (lncRNAs) are known as a class of RNA molecules longer than 200 nucleotides in length, located in the nucleus or cytoplasm and were once considered as the “noise” of the genome.^6,7 As further studies have performed, growing evidences have proved that lncRNAs can participate in gene regulation at transcriptional, post-transcriptional and epigenetic levels.^8,9 In addition, accumulating studies have suggested that lncRNAs serve as tumor promoters or/and tumor suppressors in the process of tumorigenesis and tumor development, depending on the specific circumstances. Multiple lncRNAs, for instance, ROR1,^10,11 HOTAIR,^12,13 PVT1,^14,15 HOST2^16,17 and H19^18,19 are well-known oncogenes, whereas MEG3,²⁰ ADAMTS9-AS2,^21,22 CASC2²³ are distinguished as tumor-suppressor lncRNAs. Thus, identification of OC-related lncRNAs may provide a new direction for the diagnosis and targeted treatment of OC.

In this study, we analyzed four microarrays that were downloaded from Gene Expression Omnibus (GEO, http://www.ncbi.nlm.nih.gov/geo) datasets and OC patient information downloaded from the Cancer Genome Atlas (TCGA, https://cancergenome.nih.gov/). LINC00565 was screened out as the investigative target for its significantly high-expression in OC.

Materials And Methods

Data Acquisition And Differentially Expressed Genes (DEGs) Screening

Raw data files were downloaded from TCGA and GEO dataset. After filtering all the data of OC gene expression downloaded, the missing data were supplemented. Limma package, insilicoMerging package and Rankprod package were used to analyze the DEGs in OC.

Sample Collections And Ethics Statement

All the tissue samples involved in our study were harvested from OC patients undergoing surgery in the Department of Gynecology and Obstetrics, the First Affiliated Hospital of Nanjing Medical University, dating from January 2010 to January 2018. All patients had signed the informed consent form. Resected tissue samples were immediately frozen and stored in liquid nitrogen for further research. The criteria of tumor stage and grade were in accordance with that proposed by the International Federation of Gynecologists and Obstetricians (FIGO). The Research Ethics Committee of Nanjing Medical University, China approved our study. Basic clinical pathological characteristics of the OC patients are listed in Table 1, including age, tumor size, FIGO stage, lymph node metastasis and histological grade.

Table 1 Association Between LINC00565 Expression And Clinicopathological Characteristics Of Patients With OC (n=74)

Cell Culture And Cell Transfection

Human-derived OC cell lines (OVCAR3, SKOV3, HO8910, A2780 and HEY) and human-derived normal ovarian cell line (IOSE) were provided by the Institute of Biochemistry and Cell Biology of the Chinese Academy of Sciences (Shanghai, China). Cells were cultured in RPMI1640 (Gibco, Grand Island, NY, USA) containing 10% fetal bovine serum, 100 U/mL penicillin and 100 μg/mL streptomycin((Invitrogen, Carlsbad, CA), incubating at 37°C and 5% CO₂ in a humidified atmosphere. LINC00565 siRNA, over-expression plasmid of GAS6 and negative controls were designed and synthesized by GenePharma (Shanghai, China). Transfection was performed using Lipofectamine 3000 (Invitrogen, USA) until cells seeded in the 6-well plate reaching 50–70% of confluence. Transfection vector sequences are illustrated in Table 2.

Table 2 Primer Sequences

RNA Extraction And Quantitative Real-Time Polymerase Chain Reaction (qRT-PCR)

On the basis of the manufacturer’s instructions, total RNA was extracted from related tissues and treated cells using TRIzol reagent (Invitrogen, USA). Then, cDNA was synthesized by reverse transcription according to Takara PrimeScript RT Master Mix instructions. qRT-PCR was then performed in the light of SYBRGreen (TaKaRa, USA) instructions and detected by LC480 II. Relative gene expression was normalized to GAPDH and calculated using the formula of 2^−△Ct. The primers involved in the study are shown in Table 2.

5-Ethynyl-2ʹ-Deoxyuridine (EdU) Proliferation Assay

For detecting the cell proliferation ability, EdU assay was performed. A total of 2×10⁴ transfected cells per well were seeded into 96-well plates. After cell adherence, 100 μL complete medium with 0.1% EdU solution (RiboBio, Guangzhou, China) was applied into each well and cultured for 1 hr. Living cells were fixed in 4% phosphate-buffered paraformaldehyde, neutralized by 2 mg/mL glycine and subjected to permeabilization in 0.5% Triton X-100. After repeated washing with PBS, cells were dyed with Apollo staining solution and Hoechst reaction mixture in dark for 30 mins. Finally, the analysis of EdU incorporation rate was measured by fluorescence microscope (magnification 20×).

Transwell Migration And Invasion Assay

Transwell chambers with 8 µm pore size polycarbonate membrane (Corning, NY, USA) were inserted into a 24-well plate. For the migration assay, 2×10⁴ cells were resuspended in 100 μL of serum-free medium and filled into the upper chambers. Meanwhile, 600 μl of RPMI1640 complete medium was supplied into the bottom chambers. Then, cells were incubated at 37°C with 5% CO₂ for 24 hrs. For the invasion assay, 100 μL of Matrigel was pre-coated on the upper chamber. Following steps of the invasion, assay were the same as the abovementioned. Cells were fixed in 4% phosphate-buffered paraformaldehyde and stained by 0.5% crystal violet solution (Beyotime, Nantong, China) after the incubation. At last, the numbers of migrated and invasive cells were pictured by calculating from five random fields per sample.

Flow Cytometry

Cell cycle was analyzed by flow cytometry. Briefly, treated cells were digested by trypsin (Gibco, Grand Island, NY, USA), washed by cold PBS, and fixed with 70% cold ethanol at 4°C overnight. Cell samples were stored at −20°C. During analysis, fixed cells were washed and incubated in 1 mL of protease inhibitor (PI) solution in dark for 15 mins at room temperature (RT). The percentages of cell cycle distribution were evaluated by PI staining, and all samples were analyzed using FACSCalibur (BD Biosciences).

Protein Extraction And Western Blot Analysis

Total proteins of tissues and cultured cells were extracted by using RIPA buffer containing of 1% protease inhibitors (Sigma-Aldrich). The concentrations of soluble protein were measured by bicinchoninic acid (BCA) method (Beyotime, Nantong, China). Equal quality of total proteins of every group were separated with 10% SDS-PAGE (Beyotime, Nantong, China), transferred to PVDF, and blocked in 5% milk for 1 hr at RT. Membranes were incubated overnight at 4°C with the primary antibody against p16 (Abcam, ab51243), p21 (Abcam, ab109199), Cyclin E1 (Abcam, ab71535), Cyclin D1 (Abcam, ab16663), CDK4 (Abcam, ab137675), and GAPDH (Beyotime, Nantong, China). The secondary antibody against rabbit (Beyotime, Nantong, China) and mouse (Beyotime, Nantong, China) was incubated at RT for 1 hr. At last, ECL detection kit (Millipore, USA) was used to detect the protein bands.

Tumor Xenograft In Animals

All procedures involving animals were approved by the Experimental Animal Ethical Committee of Nanjing Medical University and all procedures referring to animal experiments were performed in accordance with the principles and procedures of the National Institutes of Health Guide for the Care and Use of Laboratory Animals. Ten female BALB/c mice (aged 4–6 weeks) were randomly divided into 2 groups. HO8910 cells were stably transfected with sh-LINC00565 or sh-NC, and resuspended in PBS at 1×10⁸ cells/mL. 100 μL of cell suspension was subcutaneously injected into the left axilla of nude mice. Tumor size was measured every 3 days, and the xenografts were harvested from the injection site for histological analysis at 6 weeks after injection. Tumor volumes were estimated by using the formula: Volume = 1/2 × length × width². Tumor tissues were used to examine the expression of LINC00565 by qRT-PCR. Slides of xenograft tumors were paraffin-embedded, formalin-fixed, and finally subjected to H&E staining and immunostaining.

Immunohistochemistry (IHC) Analysis

Firstly, slides of the xenograft tumors were pretreated for staining. Subsequently, goat serum was used to block the slides and diluted primary antibodies were used to incubate overnight at 4°C. On the next day, the slides were incubated with secondary antibodies for 2 hrs in a 37°C incubator. Then, the slides were stained with DAB for 1–5 mins and washed in distilled water. After the slides were re-stained with the hematoxylin solution for 3–5 mins, they were dehydrated by gradient ethanol from low to high concentration. Lastly, the slides were soaked in the xylene solution twice for 10 mins and sealed. Images were observed and photographed with a digital slice scanner.

Statistical Analysis

SPSS 20.0 statistical software was applied to all data analysis and GraphPad Prism 5.0 software (GraphPad Software, San Diego, CA) was used to generate the data results. We applied Chi-squared test, Student’s t-test, and the Kaplan–Meier method appropriately to compare intergroup differences. The statistically significant differences were defined as p value <0.05 (*p<0.05, **p<0.01, ***p<0.001).

Results

LINC00565 Expression Is Upregulated In OC And Negatively Correlated With The Prognosis

Initially, GSE52037,²⁴ GSE38666,^25,26 GSE40595²⁷ and GSE26193²⁸ were filtered from GEO datasets by searching the key words of “ovarian cancer” and “GPL570”. By analyzing raw microarray data of three datasets (GSE52037, GSE38666, GSE40595) using bioinformatics, we found that LINC00565 expression was dramatically upregulated in OC tissues compared to that of normal ovarian tissues (Figure 1A–C). Further analysis of TCGA database and GSE26193 demonstrated that LINC00565 expression was negatively related to prognosis of OC patients (Figure 1D). In addition, 22 normal ovarian tissues and 74 OC samples involved in our hospital were used to detect the expression of LINC00565. Identically, LINC00565 was markedly upregulated in OC tissues compared to that of normal ovarian tissues (Figure 1E). According to the medium expression of LINC00565 in OC tissues, 74 OC patients were divided into low expression group (n=37) and high-expression group (n=37) (Figure 1F). Correlation analysis between the LINC00565 expression and pathological characteristics of OC revealed that higher expression level of LINC00565 was significantly related to higher FIGO stage (p= 0.0325) and higher tumor size (p= 0.0023), rather than other clinical characteristics such as age, histological subtype, histological grade, lymph node metastasis, ascites, or CA125 level (Table 1). Moreover, Kaplan–Meier analysis indicated that high expression of LINC00565 was associated with poor outcome in OC patients (p=0.0078, log-rank test; Figure 1G).

Figure 1 LINC000565 is highly expressed in ovarian cancer and is negatively correlated with the prognosis. (A–C) LINC00565 was up-regulated in GSE52037, GSE38666, GSE40595; (D) analysis of TCGA database and GSE26193 showed that patients with high levels of LINC00565 expression showed worse prognosis (log-rank test); (E) qRT-PCR of clinical samples showed that the levels of LINC00565 in ovarian tumor tissues were significantly higher than those in normal ovarian tissues. (F) Ovarian tumor patients were divided into high-expression group and low-expression group according to the median of the expression of LINC00565; (G) LINC00565 expression was negatively correlated with the prognosis of ovarian cancer patients (log-rank test). ***P < 0.001.

Knockdown Of LINC00565 Inhibits The Progression Of OC Cells

As illustrated in Figure 2A, we found that LINC00565 expression was higher in OC cell lines (OVCAR3, SKOV3, HEY, A2780, HO8910) compared to normal ovarian cell line (IOSE), especially in A2780 and HO8910 cell lines. Hence, A2780 and HO8910 cell lines were chosen to be representative of OC cells in subsequent experiments. To investigate the potential functional role of LINC00565 in OC cells, we initially silenced LINC00565 by transfection of siRNAs (si-LINC00565 1#, si-LINC00565 2#, si-LINC00565 3#) in HO8910 and A2780 cell lines. Transfection efficiencies of siRNAs were measured by qRT-PCR. HO8910 and A2780 cells transfected with si-LINC00565 presented a marked reduction in LINC00565 expression relative to those transfected with si-NC (Figure 2B and C). Subsequently, we picked out si-LINC00565 3#, the most pronounced one to reduce the expression of LINC00565, for the following experiments. EdU assay revealed that knockdown of LINC00565 significantly suppressed cell proliferation both in HO8910 and A2780 cells (Figure 2D). Next, Transwell assays showed that downregulated LINC00565 significantly decreased migration and invasion of OC cells (Figure 2E and F). Above data suggested that LINC00565 may serve as an oncogene in promoting the progression of OC.

Figure 2 LINC00565 mainly facilitates the progress of OC cells. (A) Compared to normal ovarian cell line (IOSE), LINC00565 expression was higher in ovarian cancer cell lines (OVCAR3, SKOV3, HEY, A2780, HO8910), especially in A2780 and HO8910 cell lines. (B–C) The efficiency of small interference RNAs was detected by qRT-PCR. (D) EDU analysis indicated that knockdown LINC00565 can inhibit HO8910 and A2780 cells proliferation. (E–F) HO8910 and A2780 cell migration and invasion were evaluated by transwell assays after transfected with si-LINC00565 3#, and the results showed that cell migration and invasion were significantly decreased compared with that of control. **P < 0.01; ***P < 0.001.

Knockdown Of LINC00565 Induces Cell Cycle Arrest

As LINC00565 dramatically accelerated OC cells proliferation, we next explored the role of LINC00565 in the regulation of cell cycle. Flow cytometry was applied to detect the changes of cell cycle in OC cells with LINC00565 knockdown. Dramatically, silence of LINC00565 led to a sharply block of cell cycle in OC cells (Figure 3A). To further explore the underlying mechanism of LINC00565 in facilitating the progression of OC, we paid attention to its effects on cell cycle. Considering that knockdown of LINC00565 caused a reduction in cell proliferation through arresting cell cycle in G0/G1 phase, Cyclin D1, a protein involved in the transition from G0/G1 phase to S phase, was markedly downregulated in si-LINC00565 group compared to si-NC group. Further studies found that after knockdown of LINC00565, as expected, cell cycle promoting factors Cyclin E1 and CDK4 were downregulated. Nevertheless, cell cycle inhibiting factors P16 and P21 had the opposite trends (Figure 3B). These data implied that LINC00565 could accelerate the cell cycle of OC cells to stimulate tumorigenesis.

Figure 3 Interference with LINC00565 blocks the cell cycle in G0/G1 phase. (A) The cell cycle of HO8910 and A2780 cells was blocked in G0/G1 phase after LINC00565 knockdown. (B) Western blot analysis of protein expressions of Cyclin D1, Cyclin E1, CDK4 as well as P16 and P21 after cells transfected with si-LINC00565 3# and si-NC. GAPDH was used as control. **P < 0.01; ***P < 0.001.

Synergistic Roles Of LINC00565 And GAS6 In Promoting OC

Raw data downloaded from TCGA database including 379 OC patients were analyzed using bioinformatics. Results showed that LINC00565 was positively correlated with GAS6 (Figure 4A). Previous studies revealed that GAS6 was highly expressed in plentiful tumors, such as lung adenocarcinoma,²⁹ ovarian cancer^30,31 and breast cancer.³² To further validation, clinical OC samples were used to detect the expression of GAS6. qRT-PCR verified that GAS6 was highly expressed in 74 OC samples compared to 22 normal ovarian tissues (Figure 4B). Meanwhile, silence of LINC00565 in OC cells correspondingly downregulated mRNA expression of GAS6 (Figure 4C and D), so was the protein expression of GAS6 (Figure 4E). Transfection of overexpression plasmid of GAS6 sufficiently upregulated its level in OC cells (Figure 4F). Subsequently, rescue experiments were conducted to notarize whether LINC00565 can influence the pro-cancer effect of GAS6 on OC cells. Overexpression of GAS6 in HO8910 and A2780 cells markedly facilitated proliferation, migration and invasion, but were further reversed by LINC00565 knockdown (Figure 4G–I). These results collectively indicated that LINC00565 and GAS6 played a synergistic role in the progression of OC.

Figure 4 LINC00565 and GAS6 play a synergistic role in promoting cancer. (A) LINC00565 was positively correlated with GAS6 analyzed by TCGA. (B) qRT-PCR verified that GAS6 was highly expressed in 74 OC samples compared to 22 normal ovarian tissues. (C–D) The mRNA expression of GAS6 was reduced after HO8910 and A2780 cells transfected with si-LINC00565. (E) The protein expression of GAS6 was reduced after HO8910 and A2780 cells transfected with si-LINC00565. (F) After transfecting with plasmid, GAS6 protein expression was increased distinctly. (G–I) Overexpression of GAS6 in HO8910 and A2780 cells markedly facilitated proliferation, migration and invasion, but was further reversed by LINC00565 knockdown. *P<0.05; **P < 0.01; ***P < 0.001. #P<0.05; ##P < 0.01; ###P < 0.001.

LINC00565 Regulates The Tumor Growth Of OC In Vivo

To further verify the role of LINC00565 in in vivo tumorigenesis of OC, HO8910 cells stably transfected with lentivirus were injected to nude mice subcutaneously. The subcutaneous xenograft tumor size was observed every 3 days. As shown in Figure 5A, tumor growth in sh-LINC00565 group was markedly slower relative to sh-NC group. Tumor weight in sh-LINC00565 group was lighter than that of sh-NC group (Figure 5B). In addition, qRT-PCR was applied to appraise LINC00565 expression in xenograft tumors. The sh-LINC00565 group showed a low-expression level of LINC00565 compared to that of sh-NC group (Figure 5C). Xenograft tumor tissues were stained with cell cycle-related proteins conducted by immunohistochemistry. As expected, PCNA, Ki-67, Cyclin D1, Cyclin E1 and CDK4 were markedly reduced in sh-LINC00565 group compared with those of sh-NC group; nevertheless, P16 and P21 showed the opposite trends (Figure 5D). The above results further indicated that LINC00565 promoted the in vivo proliferation of OC.

Figure 5 LINC00565 regulates the tumor growth in vivo. (A) The growth of tumors from sh-LINC00565 group were markedly slower than these from sh-NC group. (B) The tumors dissected from sh-LINC00565 group node mice were lighter than the sh-NC group. (C) qRT-PCR verified that LINC00565 expression was lower in sh-LINC00565 group than that in sh-NC group. (D) Immunohistochemistry showed that PCNA, Ki-67, Cyclin D1, Cyclin E1 and CDK4 were markedly reduced in LINC00565 knockdown group compared with NC group, nevertheless, P16 and P21 showed the opposite effect. ***P < 0.001.

Discussion

With the rapid improvement of medical technology, the diagnosis and treatment of OC has been obviously advanced. However, due to the high degree of biological malignancy of OC itself, the 5-year survival of OC is still not optimistic. Extensive evidences have demonstrated that lncRNAs play vital roles in the pathogenesis of OC by functioning as oncogenes or tumor-suppressor genes.

Since Okazaki et al³³ first discovered lncRNAs from the DNA transcription products, a great number of studies have identified the roles of lncRNAs in the occurrence and development of diseases, especially cancers.^34,35 At present, multiple tumor-related lncRNAs have been discovered. For instance, the MALAT1,^36,37 H19³⁸ and HOTAIR³⁹ were closely associated with the occurrence and development of many kinds of cancers. Gao et al revealed that lncRNA membrane-associated guanylate kinase inverted 1 intronic transcript (MAGI1-IT1) promoted EOC cell invasion and metastasis through miR-200a/ZEB axial.⁴⁰ Yong et al demonstrated that NEAT1 promotes the progression of high-grade serous ovarian cancer via miR-506.⁴¹ Therefore, finding OC-related lncRNAs is of great significance to the effective diagnosis and treatment of OC. In this study, we first found that LINC00565 was significantly up-regulated in OC tissues, higher than that in normal ovarian tissues. In vitro assays demonstrated that LINC00565 could promote cell proliferation, migration and invasion, and facilitate cell cycle of OC cells. In vivo assays further confirmed the promotive role of LINC00565 in tumor growth of OC. These results suggested that LINC00565 exerted a promotion role in the occurrence and development of OC.

As is well known, the balance between cell proliferation and apoptosis is an important factor to maintain the homeostasis of the body, and this process is regulated by many genes. Once the balance is broken, it will lead to the occurrence and development of diseases and even tumors. Cell proliferation is mainly regulated by the cell cycle, which is a complex network regulation process involving multiple signaling pathways. To explore the role of LINC00565 in OC, we focused on the changes of cell cycle networks. EdU assay showed that knockdown of LINC00565 can restrain the proliferation of HO8910 and A2780 cells. Flow cytometry assay proved that the cell cycle was blocked in G0/G1 phase due to LINC00565 knockdown. Western blot showed downregulation of Cyclin D1, Cyclin E1 and CDK4, but upregulation of p16 and p21 by LINC00565 knockdown. Immunohistochemistry obtained the same results.

Cyclin-dependent kinase (CDK) family and cyclins, key proteins that regulate cell cycle, can directly participate in the regulation of cell cycle.⁴² Cyclin D1 and Cyclin E1 are the starting factors of the cell cycle for the transformation from G1 phase to S phase. Inhibited activities of Cyclin D1 and Cyclin E1 arrest cells in G0/G1 phase.^43,44 Cyclin D1 controls the progression from G1 phase to S phase by binding to CDK4/CDK6.⁴⁵ P21 is an important member of Cyclin-dependent kinase inhibitors (CDK inhibitor, CKI) family, which can inhibit a variety of cyclin-CDK complexes, block CDK kinase activities, eventually leading to the blockage of cell cycle.^46–48 P16 competes with Cyclin D1 to bind to and inactivate CDK4/CDK6, thereby preventing the cell cycle progression from G1 phase to S phase.⁴⁹ Our study showed that silence of LINC00565 down-regulated Cyclin D1, Cyclin E1 and CDK4, but upregulated P16 and P21. These results implied that LINC00565 can inhibit the expressions of P21 and P16, resulting in the overexpression of cell cycle-related proteins to shorten the cell cycle, eventually leading to the development of OC.

Moreover, information analysis of 379 OC patients downloaded from TCGA database showed that LINC00565 expression was positively correlated with growth arrest-specific gene 6 (GAS6). GAS6, a number of vitamin K-dependent protein family, is expressed by several tissues.⁵⁰ By binding with TAM receptor, GAS6 plays an important role in regulating cell proliferation, differentiation and other processes.⁵¹ Studies have pointed out⁵² the crucial function of GAS6 in the occurrence of diabetic vascular complications. Moreover, GAS6 has been proved to be engaged in the occurrence of various tumors,^53,54 implying the promoting role of GAS6 in the OC. Our study showed that GAS6 promoted OC cells to proliferate, migrate and invade, while the oncogenic effect was reversed due to LINC00565 knockdown. In summary, these results declared that LINC00565 and GAS6 can jointly promote the progression of OC.

However, there are still some deficiencies in this study. In vitro and in vivo experiments conducted in this study mainly elucidated the function of LINC00565 in the proliferation of OC. The diagnostic and prognostic potentials of LINC00565 and the specific functional relationships between LINC00565 and GAS6 have not been well explored, which are the future focuses for further investigations.

Conclusions

To sum up, these findings showed that LINC00565 functioned as a tumor supporter in OC and was associated to the prognosis of OC patients. LINC00565 and GAS6 jointly regulated the occurrence and development of OC. LINC00565 knockdown in OC cells induced cell cycle arrest in G0/G1 phase. These results indicated LINC00565 may be utilized as a molecular marker to predict the progression, prognosis and therapeutic effect of OC. Therefore, clear identification of abnormally expressed lncRNAs will enhance the understanding of their functions in malignant tumors. These lncRNAs may contribute to develop diagnostic and prognostic hallmarks of OC.