Mechanisms of Apoptosis-Related Long Non-coding RNAs in Ovarian Cancer

Ovarian cancer is a health-threatening malignancy of ovary in female reproductive systems and one of the most common gynecological malignancies worldwide. Due to rare early symptoms, ovarian cancers are often diagnosed at advanced stages and exhibit poor prognosis. Thus, efforts have been paid to develop alternative diagnostic and therapeutic strategies for the disease. Recent studies have presented that some long non-coding RNAs (lncRNAs) play roles in apoptosis of ovarian cancer cells through various mechanisms involved in the regulation of transcription factors, histone modification complexes, miRNAs, and protein stability. Because evasion of apoptosis in cancer cells facilitates to promote tumor progression and therapy resistance, apoptosis regulatory mechanisms of lncRNAs may be promising new targets in ovarian cancer. In this review, we introduce the recent findings in regard to the molecular mechanisms of apoptosis-related lncRNAs in ovarian cancer cells.


OVARY AND OVARIAN CANCER
The ovary is a female reproductive organ where oocyte development occurs (Motta et al., 1997;Virant-Klun, 2015;Yadav et al., 2018) and functions as an endocrine organ involved in the synthesis of the female sex steroid hormones and the regulation of reproduction such as the menstrual cycle, pregnancy, and lactation (Hiller-Sturmhöfel and Bartke, 1998). Thus, the health of ovaries is essential for reproduction and women's lives, rendering finding cures to ovarian diseases crucial. Ovarian cancer is one of the most common gynecological cancers (Momenimovahed et al., 2019). The GLOBOCAN 2018 data estimates ∼300,000 new cases of ovarian cancer and over 180,000 ovarian cancer-related deaths per year worldwide (Bray et al., 2018). Ovarian cancer is a heterogeneous disease and classified by type of originated cell. Epithelial ovarian cancer (EOC) is the most common ovarian cancer (∼90%) (Rojas et al., 2016;Momenimovahed et al., 2019). The disease is often advanced at diagnosis due to lack of early symptoms and the 5-year causespecific survival rate is <50% Torre et al., 2018;Trinidad et al., 2020). Based on the current limitations, alternative diagnostic and therapeutic approaches for ovarian cancer remain to be explored.

APOPTOSIS IN OVARY AND OVARIAN CANCER
Apoptosis is a process of programmed cell death triggered by intrinsic or extrinsic signals (Wong, 2011). Intrinsic signals are initiated by cellular stresses. These signals increase the mitochondrial permeability and release of the pro-apoptotic factors such as cytochrome-c, resulting in activation of cysteineaspartic acid proteases (caspases), which are essential enzymes for apoptosis execution. Meanwhile, extrinsic signals are mediated by death receptor signaling pathways. Death receptors, their ligands, and adaptor proteins form the death-inducing signaling complex (DISC), which triggers caspase activation (Wong, 2011).
Apoptosis plays physiological roles in normal ovary functions such as follicular atresia and corpus luteum regression (Vaskivuo and Tapanainen, 2003;Yadav et al., 2018). In malignant tumors, evasion of apoptosis facilitates cancer cell survival and tumor progression (Wong, 2011;Binju et al., 2019), thus efforts have been paid for cancer strategies to discover the molecules to exert apoptosis in cancer cells whereas not in normal cells. For ovarian cancer treatment, small chemicals that modulate apoptosis-related proteins such as inhibitors of apoptotic proteins (IAPs) have entered clinical trials (Binju et al., 2019).
Considering the importance of apoptosis in cancer pathophysiology, strategies targeting these apoptosis regulatory mechanisms may contribute to the development of novel ovarian cancer therapies.

LONG NON-CODING RNA (lncRNA)
Long non-coding RNAs (lncRNAs) are defined as >200nt transcripts that do not encode proteins and tens of thousands of lncRNA transcripts are identified throughout the human genome, the majority with unknown function. However, functional studies of some lncRNAs have revealed that they have a wide range of functions. For example, lncRNAs regulate transcription and chromatin remodeling by modulating the recruitment of transcription factors and PRC to specific genomic loci. Furthermore, lncRNAs are involved in gene regulation at post-transcriptional levels through interacting with mRNAs, miRNAs, and proteins (Marchese et al., 2017). Intriguingly, lncRNAs play important roles in pathophysiology of various cancers (Takayama and Inoue, 2016;Misawa et al., 2017;Arun et al., 2018;Mitobe et al., 2018;Kamada et al., 2020;Takeiwa et al., 2020). Particularly, several lncRNAs have been suggested to regulate the apoptosis of ovarian cancer cells (Figure 1 and Table 1). In the following sections, we will describe some apoptosis-related lncRNAs in ovarian cancer cells according to their mechanisms.

GAS5
Growth arrest-specific 5 is downregulated in ovarian cancer, with this low expression associated with shorter disease-free period and lower overall survival rate of ovarian cancer patients (Gao et al., 2015;Li et al., 2016;Zhao et al., 2018;Long et al., 2019). GAS5 overexpression promotes apoptosis of ovarian cancer cells such as A2780, HEY, OVCAR3, and SKOV3, and increases the sensitivity of HEY and SKOV3 cells to the anticancer agent cisplatin (Gao et al., 2015;Li et al., 2016;Zhao et al., 2018;Long et al., 2019). A functional study has shown that GAS5 recruits the E2F4 transcription factor to the poly(ADP-ribose) polymerase 1 (PARP1) promoter, repressing PARP1 transcription in HEY and SKOV3 cells (Long et al., 2019; Figure 1A). GAS5-mediated PARP1 repression might contribute to apoptosis by downregulating the mitogen-activated protein kinase (MAPK) pathway, but direct evidence will be required in the future study.
These results suggest a possibility that NRCP may modulate apoptosis by regulating cancer metabolism. NRCP is not annotated in National Center for Biotechnology Information (NCBI) Reference Sequence (RefSeq) database (on Feb 3rd, 2021) and requires further characterization of sequences and expression profiles.

ABHD11-AS1
Abhydrolase domain containing 11 antisense RNA 1 is upregulated in ovarian cancer Zeng et al., 2019). A functional study has shown that ABHD11-AS1 modulates the expression of ras homolog family member C (RhoC) by an unknown mechanism, suppressing apoptosis in A2780 and OVCAR3 cells . Another functional study has shown that ABHD11-AS1 binds to enhancer of zeste homolog 2 (EZH2), a component of PRC2. ABHD11-AS1 facilitates tri-methylation at the 27th lysine residue of the histone H3 protein (H3K27me3) on the tissue inhibitor of metalloproteinase 2 (TIMP2) promoter, as mediated by PRC2, and likewise suppresses TIMP2 expression in HO8910 cells and OVCA429 ovarian cancer cells ( Figure 1B). TIMP2 suppression mediated by ABHD11-AS1 promotes the proliferation of OVCA429 cells, suggesting that ABHD11-AS1 may also modulate apoptosis by this mechanism (Zeng et al., 2019).

FALEC/FAL1
Focally amplified lncRNA in epithelial cancer/focally amplified lncRNA on chromosome 1 was initially identified as an lncRNA whose gene copy number increased in multiple types of cancers, including ovarian cancer (Hu et al., 2014). Its high expression level and gain in genomic copy number correlate with a shorter overall survival rate of late-stage ovarian cancer patients (Hu et al., 2014). A functional study using A2780 cells has suggested that FALEC binds to a component of PRC1, B lymphoma Mo-MLV insertion region 1 homolog (BMI1) protein, and recruits PRC1 to the promoters of genes such as cyclin dependent kinase inhibitor 1A (CDKN1A), Bcell translocation gene 2 (BTG2), and FAS. Subsequently, PRC1 mediates the ubiquitination at the 119th lysine residue of the histone H2A (H2AK119ub) on these promoter regions and the suppression of these genes ( Figure 1B). The FALEC/PRC1 complex target genes such as CDKN1A, BTG2, and FAS are suggested to be involved in apoptosis regulation (El-Deiry, 2001;Mao et al., 2015). Thus, FALEC can be a regulator of ovarian cancer apoptosis.

TP73-AS1
Tumor protein p73 antisense RNA 1 is upregulated in EOC and associated with poor prognosis in EOC patients (Li Y. et al., 2019). A recent study has shown that TP73-AS1 knockdown induces apoptosis of SKOV3 cells, suppressing the proliferation in in vitro culture and the xenograft tumor formation in athymic mice. In contrast, TP73-AS1 overexpression suppresses apoptosis in CAOV3 ovarian cancer cells. Functional analyses have suggested that TP73-AS1 epigenetically suppresses CDKN1A expression by recruiting PRC2 to its promoter ( Figure 1B) and modulates apoptosis of SKOV3 cells through this mechanism (Li Y. et al., 2019).

UNC5B-AS1
Uncoordinated 5 netrin receptor B antisense RNA 1 is highly expressed in ovarian cancer, and a recent study has shown that its knockdown activates caspase 3 in ES2 and SKOV3 cells, suggesting the apoptosis-suppressive role of UNC5B-AS1 . Moreover, the same study has suggested that UNC5B-AS1 promotes PRC2 to repress the n-myc downstream-regulated gene 2 (NDRG2) expression epigenetically (Figure 1B), which may suppress ovarian cancer cell apoptosis . This study is limited in the elucidation of the mechanism by which UNC5B-AS1 regulates PRC2 and its in vivo function, and further functional analyses are required.

FEZF1-AS1
High levels of FEZF1-AS1 are detected in tumor tissues and the serum of EOC patients, with its high expression associated with shorter overall survival of EOC patients . Moreover, its knockdown promotes apoptosis in COC1 and SKOV3 ovarian cancer cells, suggesting the apoptosissuppressive role of FEZF1-AS1. In vitro analyses of FEZF1-AS1 have shown that it functions as a competing endogenous RNA (ceRNA) for miR-130a-5p, or sponges miR-130a-5p ( Figure 1C). Consequently, FEZF1-AS1 upregulates the expression of a miR-130a-5p target gene, sex-determining region Y (SRY)-box transcription factor 4 (SOX4), that promotes proliferation of COC1 and SKOV3 cells and may contribute to apoptosis suppression . Further analysis of FEZF1-AS1 function, especially in vivo, will clarify its role and significance in apoptosis of ovarian cancer cells.

GAS5
A recent functional study has suggested that GAS5 functions as a ceRNA for miR-196a-5p to upregulate homeobox A5 (HOXA5), promoting apoptosis of primary tumor cells from high-grade serous ovarian cancer tissues as well as A2780 and OVCAR3 cells ; Figure 1C).

HOTAIR
HOX transcript antisense RNA is upregulated in ovarian cancer, and the elevated expression level correlates with the shorter overall survival of ovarian cancer patients (Qiu et al., 2015;Wang Y. et al., 2015;Zhang et al., 2016;Luo et al., 2017;Yu et al., 2018). HOTAIR knockdown induces apoptosis in ovarian cancer cells such as A2780, HeyC2, and OVCA429, and decreases the cisplatin sensitivity of A2780 and SKOV3 cells (Qiu et al., 2015;Wang Y. et al., 2015;Zhang et al., 2016Zhang et al., , 2020Yu et al., 2018). A recent functional study using A2780 and SKOV3 cells has suggested that HOTAIR acts as a ceRNA for miR-138-5p, leading to cisplatin resistance of these cells ; Figure 1C). This study has shown that HOTAIR/miR-138-5p axis modulates EZH2 and sirtuin 1 (SIRT1) expression, but its biological significance has not been elucidated.

NCK1-DT/NCK1-AS1
Non-catalytic region of tyrosine kinase adaptor protein 1 divergent transcript is highly expressed in ovarian cancer. Mechanistically, it acts as a ceRNA for miR-137 to upregulate NCK1, which suppresses apoptosis of CAOV3 and SKOV3 cells and enhances their cisplatin resistance (Chang et al., 2020; Figure 1C).

NEAT1
Nuclear enriched abundant transcript 1 is upregulated in ovarian cancer and is associated with shorter overall survival of ovarian cancer patients (Chen et al., 2016). NEAT1 acts as a ceRNA for miR-34a-5p to upregulate BCL2 and suppresses apoptosis of OVCAR3 and SKOV3 cells (Ding et al., 2017).

UCA1
The lncRNA UCA1 is upregulated in ovarian cancer and is detected in exosomes derived from the serum of ovarian cancer patients (Li Z. et al., 2019;Li et al., 2020). Functional studies have shown that UCA1 acts as a ceRNA for miR-129 and miR-654-5p to upregulate ATP binding cassette subfamily B member 1 (ABCB1) and SALT INDUCIBLE KINASE 2 (SIK2), respectively, which contribute to the suppression of apoptosis and the enhancement of PTX resistance in HeyA8 and SKOV3 cells (Wang et al., 2018;Li et al., 2020). In addition, UCA1 functions as a ceRNA for miR-143 to increase Fos-related antigen 2 (FOSL2), and enhances cisplatin resistance in A2780 and SKOV3 cells (Li Z. et al., 2019; Figure 1C). However, the importance of the function of UCA1 as a ceRNA in vivo has not been fully analyzed. Recent studies have found that many other lncRNAs modulate ovarian cancer apoptosis through regulating miRNAs. For example, CYTOR/LINC00152 acts as a ceRNA of miR-125b to upregulate an antiapoptotic protein MCL1 in A2780 and SKOV3 cells . PVT1 suppresses apoptosis in OVACAR3 and TOV112D cells by inhibiting miR-543 and increasing a miR-543 target SERPIN1 (Qu et al., 2020). In contrast, MEG3 promotes apoptosis in OVCAR8 and SKOV3 cells by sponging miR-205-5p (Tao et al., 2020). The detail of lncRNAs regulating miRNAs is also reviewed in other articles (Braga et al., 2020;Salamini-Montemurri et al., 2020).

GHET1
The lncRNA GHET1 is upregulated in ovarian cancer and higher expression correlates with increased tumor size and distant metastasis (Liu and Li, 2019). Conversely, its knockdown induces apoptosis and downregulates glycolysis in A2780 and SKOV3 cells, where GHET1 binds to an E3 ubiquitin ligase, von Hippel-Lindau tumor suppressor (VHL), and prevents VHLmediated degradation of hypoxia-inducible factor 1α (HIF1α) (Figure 1D). Since the GHET1 function in ovarian cancer cells has been only examined by in vitro assays, in vivo analyses of GHET1 are needed. Although the role of the GHET1/VHL/HIF1α axis in apoptosis has not yet been elucidated, HIF1α and cancer metabolism have been shown to play important roles in apoptosis regulation, suggesting the possibility that this axis may also be involved in the phenomenon (Zhou et al., 2006;Matsuura et al., 2016).

NCK1-DT/NCK1-AS1
In addition to the function as a ceRNA, NCK1-AS1 increases the stability of NCK1: NCK1-AS1 binds to an E3 ubiquitin ligase, casitas B-lineage lymphoma (CBL), and prevents CBLmediated degradation of NCK1 (Chang et al., 2020; Figure 1D). The functions of NCK-AS1 in ovarian cancer have been suggested based on in vitro experiments, and thus needs to be evaluated using ovarian tumor specimens or in vivo ovarian cancer models.

CONCLUSION
In this review, we introduced the mechanisms of apoptosisrelated lncRNAs in ovarian cancer cells. Considering that dysregulation of apoptosis is involved in the resistance to ovarian cancer therapies, small molecule inhibitors/siRNAs targeting apoptosis-suppressing lncRNAs, or apoptosispromoting lncRNAs themselves may be applicable to ovarian cancer therapies. For nucleic acid-based therapeutics, it is important to develop the drug delivery systems (DDSs) with high target specificity and less non-specific toxicity in vivo. Particularly, for ovarian cancer, DDSs will be useful to treat metastatic cancer cells in peritoneal cavity (Amreddy et al., 2018;van den Brand et al., 2018). Moreover, apoptosis-related lncRNAs may be potential diagnostic and prognostic biomarkers. Especially, FEZF1-AS1 and UCA1 are detected in serum and exosomes recovered from serum of ovarian cancer patients, respectively, which suggested their potential as liquid biopsy markers for ovarian cancer.
Apoptosis-related lncRNAs have basically been studied using conventional ovarian cancer cell lines, and the functions of some lncRNAs have been examined by in vitro assays alone. For clinical application, it is required to elucidate the lncRNA functions in vivo. Moreover, previous studies have indicated some discrepancies between ovarian cancer cell lines and the original tumor clinical tissues in terms of genomic and histological features and gene expression profiles (Domcke et al., 2013;Beaufort et al., 2014). Thus, lncRNA studies using ovarian tumor specimens or other ovarian cancer models are strongly demanded. Three-dimensional cultures of patient-derived cancer cells (PDCs) and cancer models established by transplanting tumor specimens into host mice (patient-derived xenograft [PDX] models) retain the properties of original tumors and have attracted attention as promising models for cancer research and drug screening (Ishiguro et al., 2016;Maru and Hippo, 2019;Namekawa et al., 2019;Shiba et al., 2019). Further studies using PDC and PDX models would advance the application of apoptosis-related lncRNAs to ovarian cancer diagnosis, prognosis, and therapies.