Identification of aberrantly expressed long non‐coding RNAs in ovarian high‐grade serous carcinoma cells

Abstract Purpose To identify the aberrantly expressed long non‐coding RNAs (lncRNAs) in ovarian high‐grade serous carcinoma (HGSC). Methods Total RNA was isolated in HGSC cell lines, ovarian surface epithelial cells, and normal ovaries. Aberrantly expressed lncRNAs in HGSC were identified by PCR array, which analyzes 84 kinds of lncRNAs. To infer their functions, HGSC cell lines with different levels of expression of the identified lncRNAs were established, and then, activities of proliferation, migration, and apoptosis were examined. Expression levels of the identified lncRNAs were also examined in multiple ovarian HGSC tissues. Results Ten aberrantly expressed lncRNAs, six upregulated and four downregulated, were identified in the HGSC cell lines. The authors established four HGSC cell lines: in two of the cell lines, one of the upregulated lncRNAs was knocked down, and in two other cell lines, one of the downregulated lncRNAs (MEG3 and POU5F1P5) was overexpressed. Migration activities were inhibited in the HGSC cell lines overexpressing MEG3 or POU5F1P5 while there were no differences in proliferation and apoptosis between the established and control cell lines. The four lncRNAs downregulated in the HGSC cell lines were also observed to be downregulated in ovarian HGSC tissues. Conclusion The authors identified four downregulated lncRNAs in ovarian HGSC.

mechanisms involved in the progression and malignant behaviors of ovarian cancers.
In recent years, with the advances in large-scale sequencing technology, it has been widely accepted that more than 98% of the human genome is transcribed into non-coding RNAs that do not encode obvious proteins. [5][6][7] A type of non-coding RNA, long non-coding RNA (lncRNA), consists of transcripts longer than 200 nucleotides in length. lncRNAs have roles in a variety of biological processes such as X chromosome inactivation, genomic imprinting, chromatin remodeling, and cell fate determination. 8,9 In addition, ln-cRNAs are closely associated with tumorigenesis in various types of cancers. In fact, the number of publications related to the biological roles of lncRNAs in cancers has increased exponentially in the past few years. 7,10 In ovarian cancer, the lncRNAs such as H19, HOTAIR, HOXA11-AS, LSINCT5, MALAT1, and PVT1 have been reported to be involved in cell proliferation and malignancy. [10][11][12][13] In this study, we focused on HGSCs, which are the most frequently occurring and aggressive subtypes among ovarian cancers, and aimed to identify lncRNAs involved in the progression and malignant behaviors of HGSC. We compared lncRNA expression profiles between HGSC cell lines and normal ovarian cells and ovarian tissue samples, and identified the aberrantly expressed lncRNAs specific to the HGSC cell lines. In addition, some of the lncRNAs downregulated in the HGSC cell lines were observed to be downregulated in ovarian HGSC tissues. Our results also suggest that some of the identified lncRNAs have roles in malignant behaviors.

| Cell culture and tissues samples
Human ovarian cancer cell lines, KURAMOCHI and TYK-nu, were used as HGSC cell models. Both cell lines were purchased from Japanese Collection of Research Bioresources Cell Bank. KURAMOCHI and TYK-nu cells were cultured in RPMI-1600 (Wako) and eagle's minimal essential medium (Sigma-Aldrich) supplemented with 10% fetal bovine serum, respectively. Human ovarian surface epithelial cells (HOSE) were purchased from Cosmo-bio and cultured in ovarian surface epithelial cell medium (ScienCell Research Laboratories).
Twenty-two cases of HGSC specimens were obtained from the patients who underwent operation at Shimane University hospital.
Ten cases of normal ovarian tissues were obtained from the patients who underwent oophorectomy by a reason different from HGSC at Yamaguchi University hospital. Dissected specimens were immediately immersed in liquid nitrogen and stored at −80°C until RNA isolation.

| RNA isolation and real-time reverse transcription polymerase chain reaction (RT-PCR)
Total RNA was isolated from the cell lines using an RNeasy mini kit (Qiagen) according to the manufacturers' instructions. Total RNA was also isolated from the tissue specimens by ISOGEN reagent (Nippon Gene), followed by chloroform extraction and 2-propanol precipitation. 14 First-strand cDNA was synthesized from 1 μg of total RNA by random hexamers using a QuantiTect Reverse Transcription Kit (Qiagen) as previously reported. 14 Real-time RT-PCR was carried out using TB green premix Ex taq II (Takara) and primer sets listed in Table 1 under the cycling condition (40 cycles of 95°C for 5 seconds and 60°C for 20 seconds with an initial step of 95°C for 10 seconds).
The relative expression levels were calculated with the delta-delta Ct method using GAPDH as a reference gene.

| PCR Array
cDNA synthesis was performed using the RT 2 First Strand Kit (Qiagen) according to the manufacturers' instructions. 15,16 In this study, PCR array covering 84 kinds of lncRNAs was used, which are reported to be related with variable cancers (RT 2 lncRNA PCR Array Human Cancer Pathway Finder (Qiagen)). Real-time PCR was carried out with the RT 2

| Cell proliferation assay
TYK-nu lines, in which the lncRNA expression was altered, and the control lines were plated at approximately 6 × 10 4 cells in 6-well plates, respectively. 18,19 At every 24 hours, single cell suspension was prepared in each line by trypsinization and was counted with TC20 Automated Cell Counter (Bio-Rad Laboratories). Each experiment was carried out in triplicate.

| Wound healing assay
Cells of TYK-nu lines overexpressing MEG3 or POU5F1P5 and the control lines were plated at approximately 2 × 10 5 cells in 6-well plates and cultured until they reached confluence. Linear scratch wounds were created on the cell layer in the center of each well with a 1000-µL sterile pipette tip. After 72 hours, images were taken to evaluate the wounds at the same fields under the microscope, and the separation distances between wound sides were quantified as previously reported. 18 Each experiment was performed in triplicate, and the mean distances were obtained from three independent experiments.

| Migration assay
Cell migration assay was performed with a BioCoat Matrigel Invasion Chamber (Corning Life Science) according to the manufacturer's protocol as previously reported. 18 In brief, the cells of TYK-nu lines over-

| Statistical analysis
The significance of the difference between the two groups was analyzed by Student's t test and Wilcoxon test. A probability value of P < .05 was considered to be significant. All the statistical analyses were performed by using the SPSS 5.0 J for Windows software package (SAS Institute).

| Identification of aberrantly expressed lncRNAs in HGSC cell lines
To identify aberrantly expressed lncRNAs in HGSC, we used HGSC cell lines because using cell lines generally yields more uniform data than using tissue specimens that reflect individual differences. To ( Figure 1).

| Establishment of the HGSC cell lines in which expressions of upregulated or downregulated lncRNAs are altered
To  (Figure 2A and B).

| Cell proliferation of the established cell lines
Cell proliferation was analyzed in the HGSC cell lines in which LINC00152 or LINC01234 was knocked down ( Figure 3A) and in which MEG3 or POU5F1P5 was overexpressed ( Figure 3B). There was no significant difference in cell proliferation between the established cell lines and control cell lines (Figure 3).

| Cell cycles and apoptosis of the established cell lines
The cell cycle and apoptosis were analyzed in the HGSC cell lines in which MEG3 or POU5F1P5 was overexpressed (Figure 4), be-

| Cell migration activities of the established cell lines
Cell migration activity was assessed in the HGSC cell lines in which MEG3 or POU5F1P5 was overexpressed by two methods: a wound healing assay ( Figure 5A) and a cell migration assay ( Figure 5B). In the wound healing assay, in the MEG3-and POU5F1P5-overexpressing  Figure 5B). Therefore, the overexpression of these two lncRNAs inhibited the cell migratory activities in the HGSC cell lines.

| Expression levels of the aberrantly expressed lncRNAs in ovarian HGSC tissues
To  Values are mean ± SD Cancer Cell Line Encyclopedia (CCLE) databases, respectively. A recent comparison of the data in these databases revealed that the abovementioned cell lines are not suitable as HGSC models due to differences in mutation, chromosomal copy number, and mRNA expression profiles. 27 Therefore, we used KURAMOCHI and TYK-nu as HGSC cell lines because their mutations and mRNA expression profiles are more similar to those of HGSC specimens. 27 Downregulation of ADAMTS9-AS2 or XIST is associated with progression of cell proliferation and malignant behaviors in ovarian cancer. 13,22,28 Although the functions of ADAMTS9-AS2 and XIST

| D ISCUSS I ON
were not examined in this study, the remarkably low expression in the HGSC tissues suggests that ADAMTS9-AS2 or XIST has inhibitory effects on malignant behaviors of ovarian HGSC.
About the four upregulated lncRNAs (CDKN2B-AS1, DLEU2, LINC00152, and LINC01234), knockdown using siRNAs did not show any inhibitory effects and their expression levels were not different between ovarian HGSC tissues and normal ovarian tissues.
We cannot clearly explain the difference in actions and expressions F I G U R E 6 Expression levels of aberrantly expressed lncRNAs in HGSC cell lines in multiple specimens of HGSC and ovarian tissues. Box plots of relative expression levels of lncRNAs downregulated in HGSC cell lines (ADAMTS9-AS2, MEG3, POU5F1P5, and XIST), which were analyzed by real-time RT-PCR in triplicates, in the twenty-two HGSC tissues and ten ovaries. GAPDH was used as an internal control. A plotted circle indicates each sample. *P < .05 and **P < .01 compared to the control (Mann-Whitney U test) between upregulated and downregulated lncRNAs. We speculate as follows; the lncRNAs in this study were originally identified in HGSC cell lines. Although the HGSC cell lines used in this study are derived from ovarian HGSC cells, ovarian HGSC tissues reflect individual differences, indicating the heterogeneity. Therefore, the character of the HGSC cell lines is not always consistent with the character of the ovarian HGSC tissues.
In conclusion, we identified four new lncRNAs that are downregulated in ovarian HGSC. These four lncRNAs may have uses in the diagnosis and treatment of HGSC.

ACK N OWLED G EM ENTS
We would like to thank Professor KYO Satoru (Shimane University), for providing the HGSC tissues used in this study.

CO N FLI C T O F I NTE R E S T
The authors declare that there are no conflicts of interest.