ERRα expression in ovarian cancer and promotes ovarian cancer cells migration in vitro

Ovarian cancer is the leading cause of death from a gynaecological malignancy in the developed world, and is characterized by invasion and metastasis and thus causes a high fatality rate. Estrogen-related receptor alpha (ERRα) has been demonstrated to play a widespread and pathophysiological relevant role in tumourigenesis and development. The aim of this study was to investigate the effect of ERRα expression on the progression of ovarian cancer. The correlation between ERRα expression level and clinical pathological parameters in ovarian cancer tissues were analysed via cancer public database CPTAC. The expression level of ERRα in ovarian cancer cells were confirmed by RT-qPCR and Western blot methods. The cellular ERRα expression was up-regulated by lentivirus transfection and down-regulated by specific antagonist. The invasion and metastasis capabilities of ovarian cancer cells were characterized by wound healing assay and trans-well chamber assay. The CPTAC database showed that the ERRα expression levels were higher in the late-stage and high-grade ovarian cancer tissues than in early-stage and low-grade tissues. Ovarian cancer cells with higher-expression ERRα exhibited stronger invasion and metastasis capabilities in vitro. After up-regulating the ERRα expression level, the invasion and metastasis capabilities of ovarian cancer cells were enhanced, while down-regulation weakened. Moreover, the wound sealing rate was positively correlated with the expression of ERRα mRNA expression level (r = 0.921, P < 0.01), and the cell invasiveness was also positively correlated with the cellular ERRα mRNA expression level (r = 0.926, P < 0.01). Our results suggest that ERRα may promote the progression of ovarian cancer, and may serve as a promising predictive biomarker.


Introduction
Ovarian cancer is the leading cause of death from gynecologic malignancies in developed countries, posing a significant threat to women's health. It has a high risk of invasion and metastasis, post-treatment recurrence and drug resistance. Approximately 125,000 people die from ovarian cancer every year in the world [1], posing a serious threat to women's lives and health. Ovarian cancer patients lack typical symptoms at the early stage, and about 70% of the cancer cases are diagnosed at an advanced stage, with a 5-year survival rate of only 20-30% [2]. Ovarian cancer poses a major challenge to its diagnosis and treatment as it is characterized by metastasis, recurrence and poor prognosis. To improve the early diagnosis and prognosis of ovarian cancer, there is an urgent need for biomarkers or therapeutic targets for it.
The estrogen-related receptor alpha (ERRα) is an orphan nuclear receptor (NR) that can perform biological functions without binding to a ligand. Nuclear receptor superfamily (NRS) is a family of transcription factors with a wide range of functions. Most of them bind directly to genomic DNA in response to specific ligands, and control the expression of target genes in the presence of co-activators or co-inhibitors, so as to regulate the metabolism and homeostasis. ERRα coding gene is located at the 11q13 site of human chromosome and mainly contains three functional domains, N terminal domain (NTD), DNA binding domain (DBD) and ligand binding domain (LBD) [3]. ERRα competes with estrogen receptor alpha (ERα) for binding to the same target genes, transcription factors and co-activator proteins, thereby interacting and bypassing the classic estrogen signaling pathway [4]. Recently, many studies [5] have identified the ability of ERRα to influence the metabolic reorganisation of tumour cells to meet the energy requirements during tumourigenesis and progression, bringing the ERR receptor ERRα to widespread attention in the field of cancer. ERRα have been reported to play a crucial role in tumorigenesis and progression of breast cancer [6], endometrial cancer [7], prostate cancer [8], colorectal cancer [9] and lung cancer [10]. However, only few studies have reported the ERRα expression level of ovarian cancer and its biological function in ovarian cancer. Therefore, our purpose is to study the role of ERRα expression level in the progression of ovarian cancer.

Data acquisition and processing
The bioinformatics portal UALCAN (http:// ualcan. path. uab. edu) was used to access ERRα protein expression levels in normal ovarian tissues and in ovarian cancer tissues. This resource for expression analysis uses data from the Clinical Proteomic Tumor Analysis Consortium (CPTAC) Confirmatory/Discovery datasets (proteomics by mass-spectrometry) [11]. Protein data are expressed as Z-values, and representative of standard deviations from the median across samples for the given cancer type. Log2 spectral count ratio values from CPTAC were first normalized within each sample profile, then normalized across samples.

Western blotting
Total protein was extracted from ovarian cancer cells using RIPA lysis buffer (Sig-ma-Aldrich, St. Louis, MO, USA) and low-temperature centrifugation at 10,000 × g for 10 min at 4 °C. Total protein was quantified using a bicinchoninic acid (BCA) protein assay reagent kit (Pierce, Rockford, IL, USA). 35 μg of whole cell protein was loaded per lane and separated via SDS-PAGE on a 12% gel at 110 V. Proteins were blotted onto nitrocellulose membranes. Blotted membranes were respectively incubated with the anti-human ERRα rabbit monoclonal antibody (1:500 dilution; Cell Signaling Technology, Beverly, MA, USA) and β-Actin rabbit monoclonal antibody (1:1000 dilution; Affinity Biosciences, no.AF7018) overnight at 4 °C. The membranes were washed with Tris-buffered saline containing 0.04% Tween-20 (TBST), followed by incubation with HRP-labeled goat anti-rabbit IgG antibody (1:1000; Affinity Biosciences, no. S0001) for 1 h at room temperature. An enhanced chemiluminescence (ECL) detection system (Thermo Fisher, Waltham, MA, USA) was used to visualize the bands.

Construction of stable ERRα-expressing ovarian cancer cells by lentivirus transfection
A customized ERRα-overexpression lentivirus vector and lentivirus-negative control (plasmid sequence: Ubi-MCS-3FLAG-SV40-EGFP-IRES-puromycin) was obtained from Genechem Co., LTD. (Shanghai, China). Ovarian cancer cells (HO8910) were transfected with the ERRα lentivirus and screened/selected using puromycin (1 µg/ml; Genechem; China). The surviving cells were cultured into multiple monoclonal cell lines and were assessed for the expression of ERRα using PCR analysis. Three groups of cells were set: cells infected with ERRα-overexpression lentivirus vector (group OV-ERRα), lentivirus-negative control (group NC), and cells treated with DMSO as the control (group CON) ( Fig. 1).

Treatment with ERRα-specific antagonist XCT790
The XCT790 was purchased from Sigma-Aldrich (St. Louis, MO, USA). Prior to treatment, the ovarian cancer cells (HO8910PM) were seeded in six-well plates at a density of 1 × 10 5 cells/well and cultured in 3-ml serum-free DMEM for 12 h to achieve adherence, and then in DMED supplemented with 10% fetal bovine serum, 1% penicillin (100 IU/mL), and 1% streptomycin (100 IU/mL). When the cell population reached about 80% (logarithmic phase), as we previously studied [13], ovarian cancer cells were treated with 10 µM XCT790 for 24 h, and then were collected for RNA and protein extraction. Similarly, prior to the cellular wound healing assay and trans-well chamber assay, cells were incubated with final concentrations of 10 µM XCT790 for 24 h.

In vitro wound healing assay
Ovarian cancer cells were seeded at density 1 × 10 5 and cultured in six-well plates to 80-90% confluence and were serum-starved for 24 h. Two scratches were then introduced to the cell layer in each well using a 100-1000 µl tip. Following washing twice with PBS, the cells (HO8910, HO8910PM, group OV-ERRα, group NC) were incubated in DMED supplemented with 10% fetal bovine serum, 1% penicillin (100 IU/mL), and 1% streptomycin (100 IU/mL). The cells treated with ERRα-specific antagonist XCT790 (group XCT790-treated) were incubated in DMED supplemented with fetal bovine serum, penicillin, streptomycin and 10 µM XCT790. Images of the same regions were captured at 0 and 24 h following stimulation with light microscope (TE2000-U; Nikon, Japan); the paired images were analysed (the ratio of the difference in scratches area between 0 and 24 h to the 0 h scratches area).

Trans-well chamber migration assay
After thawing overnight at 4 °C on ice, Matrigel™ Basement Membrane Matrix (50 μL; BD, USA) was added to a Millicell Hanging Cell Culture Insert (Millipore, USA) to coat the membrane and incubated at 37 °C for 30 min. 200-μL cell suspensions containing 0.5% FBS (5.0 × 105 cells/mL) were added to the insert, placed in 24-well plates containing 1300 μL of DMEM supplemented with 10% FBS. The plates were incubated for 24 h at 37 °C. Then, non-invading cells on the top of the filter were removed with a cotton swab, and the filters were fixed with methanol and stained with crystalline violet. The filters were removed from the inserts and mounted onto slides for imaging and quantification as described in a previous study [14].

Statistical analysis
Statistical analysis was performed using the average results of three experiments under identical conditions. Numerical data are presented as the mean ± standard deviation. Differences between two means were compared by Student's t-test. A one-way analysis of variance was performed for multiple comparisons of groups, which was followed by the Fisher's least significant difference post hoc test, and associated parameters were further analysed using the Spearman's correlation test. Data were analysed using 19.0 for Windows (SPSS Inc., Chicago, IL, USA). P < 0.05 was considered to indicate a statistically significant difference.

ERRα expression correlates with clinicopathological parameters in ovarian cancer patients
UALCAN bioinformatics analysis was used to analyse protein levels of ERRα in normal ovarian tissues and ovarian cancer tissues. The protein expression data were obtained from the mass-spectrometry proteomic profiles, generated by the Clinical Proteomic Tumor Analysis Consortium (CPTAC). ERRα expression level showed an increasing tendency (P < 1E-12) in late-stage ovarian cancer [stage 3 (n = 75); stage 4 (n = 16)], compared to early stage [stage 1 (n = 2)] ( Fig. 2A). Although the number of cases was lower for stage 1 and 2, this result still reflects some tendencies. In addition, the results showed that the expression level of ERRα was higher in Grade 3 ovarian cancer tissues [Grade 3 (n = 79)] compared to Grade 2 ovarian cancer tissues [Grade 2 (n = 6)]. These all suggests the promoting effect for ERRα in tumor progression, P < 1E-12 (Fig. 2B).

The expression level of ERRα in ovarian cancer cell lines
The in vitro screening of the highly metastatic ovarian cancer cell line HO8910PM, from which cell lines with different metastatic potential were further isolated, provides an ideal cell model for the in-depth study of the metastatic mechanism of ovarian cancer. Real-time PCR (RT-PCR) analysis revealed the relative mRNA expression level of ERRα in HO8910 cell lines was lower than its metastatic cell line HO8910PM (P = 0.002; Fig. 3A). There was no significant difference in ERRα mRNA expression level between ovarian cancer HO8910 cells and the lentivirusnegative control group (P = 0.077), while cells infected with ERRα-overexpression lentivirus vector (OV-ERRα group) had higher ERRα mRNA expression level compared with those untreated HO8910 cell lines (P = 0.001, Fig. 3B). After treatment with XCT790, a widely used ERRα-specific antagonist [15,16], the relative mRNA expression level of ERRα was lower in XCT790-treated group compared with those untreated HO8910PM cells (P < 0.001, Fig. 3C). A similar tendency was observed for ERRα protein expression in these ovarian cancer cells. The protein expression level of ERRα in HO8910 was lower than its metastatic equivalent HO8910PM (0.637 ± 0.023 vs. 0.876 ± 0.105, P = 0.018; Fig. 4A). The protein expression levels of ERRα in ovarian cells HO8910, lentivirus-negative control group (NC group) and OV-ERRα group were 0.637 ± 0.023, 0.669 ± 0.043 and 1.713 ± 0.035, respectively. (HO8910 vs. NC group, P = 0.318 and HO8910 vs. OV-ERRα group, P < 0.001, respectively; Fig. 4B). As an agonist of ERRα, XCT790 is found to decrease mitochondrial masses as well as affect the mitochondrial membrane potential, results in a subsequent dysregulation of mitochondrial functions; XCT790 also reduces the expression level of PGC-1α, which is a coactivator of ERRα to regulate mitochondrial biogenesis [17,18]. After treatment with ERRα specific antagonist XCT790, the protein expression levels of ERRα were lower in XCT790-treated group compared with untreated HO8910PM cells (0.378 ± 0.028 vs. 0.871 ± 0.110, P = 0.002, Fig. 4C).
The results of the invasion assay (Fig. 5B) were consistent with the results of the wound healing assay. HO8910 and its metastatic equivalent HO8910PM exhibited transmembrane cell counts of 112.333 ± 12.284 and 265.667 ± 30.663, respectively (P = 0.003). Ovarian cells of HO8910, NC group (lentivirus negative control group) and OV-ERRα group (cancer cells infected with ERRα high expressing lentivirus vector) exhibited transmembrane cell numbers of Fig. 4 Expression level of ERRα protein in ovarian cancer cells, a similar trend to the mRNA expression. A There was higher expression level of relative ERRα protein in the ovarian cancer HO8910PM cells than in HO8910 cell lines. B The ERRα protein in HO8910 cell lines transfected with ERRα-overexpression lentivirus vector (OV-ERRα) showed an up-regulation of ERRα, while those treated with lentivirus-negative control (NC Group) did not; C The ERRα protein in HO8910PM cell lines treated with 10 µM XCT790 for 24 h showed a down-regulation of ERRα induced by XCT790. As indicated, *P-value < 0.05 was defined as statistically significant 112.333 ± 12.284, 120.333 ± 9.978 and 482.333 ± 24.500, respectively (HO8910 vs. NC group, P = 0.514. HO8910 versus OV-ERRα group, P < 0.001). After treatment with the ERRα-specific antagonist XCT790, the XCT790-treated group exhibited a smaller number of transmembrane cells than the untreated HO8910PM cells (107.667 ± 10.339 vs. 265.667 ± 30.663, P = 0.002) (Fig. 5D).

Discussion
This study is the first to investigate the impact of ERRα expression level on the progression of ovarian cancer. Cancer public database analysis showed that the expression levels of ERRα were related to the clinic stage and pathological grade of ovarian cancer tissue. Specifically, the expression levels of ERRα were higher in the late-stage and high-grade ovarian cancer tissues compared with those in early-stage and low-grade tissues. The in vitro experiments results showed that ovarian cancer cells with strong invasion and metastasis capability have higher ERRα mRNA and protein expression levels. In addition, after up-regulating the ERRα expression level, the invasion and metastasis of ovarian cancer cells was enhanced, while down-regulation weakened. Ovarian cancer cells with different invasion and metastasis capabilities have different expression levels of ERRα, and ovarian cancer cells with different ERRα expression levels also have different invasion and metastasis capabilities. Using Spearman's correlation analysis, we found that both the percentages of wound closure and the cell counts of invasion were significantly positively correlated with the cellular ERRα mRNA expression level. In conclusion, that the expression level of ERRα in ovarian cancer cells is related to its invasion and metastasis capability. These results were consistent with the results of cancer public database analysis, suggesting that ERRα may promote the progression of ovarian cancer. Further research, however, needs to involve more samples to confirm this conclusion.
ERRα is involved in the tumorigenesis and is considered as a potential biomarker in various cancers. Compared with normal tissues, ERRα expression levels were increased in breast cancer [19], prostate cancer [20], endometrial cancer [21] and colorectal cancer [22]. High levels of ERRα expression not only serve as a diagnostic biomarker, but also correlate closely with tumour progression. Lu et al. [23] studies show that ERRα can activate the pS2 promoter in breast cancer cell lines, which significantly increases the synthesis of estrogen, and the abnormal increase of estrogen can lead to tumorigenesis. Chang et al. [24] reported that the activation of the Her2/IGF-1R signaling pathways and subsequent C-MYC stabilization upregulate the expression of PGC-1β that can increase the expression of ERRα, and thus promoting the proliferation of breast cancer cell lines. Cai et al. [25] found that ERRα also stimulate glycolysis under normoxia and activate promoters of many genes encoding glycolytic enzymes in breast cancer cells, suggesting that ERRα can also promote the proliferation of breast cancer cells without involving estrogen signaling pathway. ERRα is not only involved in the proliferation of cancer cells, but also in the invasion and metastasis. Huang et al. [21] reported that ERRα can trigger the migration and invasion of endometrial cancer cells via up regulation of TGFB1. Jia et al. [26] studies showed that bisphenol S can induce the expression of miR-10b in PC12 cells via ERRα. The upregulated miR-10b inhibited the expression of KLF4, which can suppress the migration and invasion of cancer cells. Zhang et al. [10] found that activation of NF-κB/IL-6 is involved in ERRα induced migration and invasion of non-small cell lung cancer (NSCLC) cells, suggesting that ERRα might be a potential target for NSCLC treatment.
Ovarian cancer is one of the three common malignant tumors in gynecology, and it is also the cancer with the highest mortality rate in gynecology. Lack of early diagnostic biomarkers, easy metastasis and recurrence are the main causes of high mortality and poor prognosis. Therefore, it is of great significance to search for tumor biomarkers related to cancer for studying the pathogenesis of ovarian cancer and exploring novel treatment targets. Although many studies have reported ERRα is involved in the tumorigenesis and progression of various cancers, only a few studies have explored the effect of ERRα on ovarian cancer. Sophia Sn Lam et al. [27] reported that ERRα functions in epithelialmesenchymal transition and in subsequent stem cell traits responsible for the highly aggressive and metastatic characteristic of ovarian cancer, which suggests ERRα activation is a mechanism of tumor aggressiveness and imply that targeting ERRα may be a promising approach in ovarian cancer treatment. Our study is the first to suggest that the expression of ERRα is associated with the invasion and metastasis of ovarian cancer cells, and that ERRα may promote the progression of ovarian cancer, providing a reference for further studies on mechanism of ovarian carcinogenesis and molecular targeting therapy. The invasiveness and metastatic capacities of ovarian cancer cells. A Wound healing assay (10 ×). B Trans-well chamber assay (10 ×). C The percentages of wound closure exhibited by ovarian cancer HO8910PM cells was higher than HO8910 cell lines. The HO8910 cells transfected with ERRα-overexpression lentivirus vector (OV-ERRα group) had a higher percentage of wound closure than its control group, while the HO8910PM cell lines treated with 10 µM XCT790 for 24 h showed a lower percentage of wound closure than its control group. D Transmembrane cell counts exhibited by HO8910PM cells were more than HO8910 cell lines. The OV-ERRα group had more transmembrane cell counts than its control group, while the HO8910PM cell lines treated with 10 µM XCT790 for 24 h had less, a similar trend to wound healing assay. As indicated, *P-value < 0.05 was defined as statistically significant ◂ Fig. 6 Spearman's correlation analysis. A Correlation between the percentages of wound closure and cellular ERRα mRNA expression level (P < 0.01). B Correlation between the cell counts of invasion and the cellular ERRα mRNA expression level (P < 0.01)