Introduction
Ovarian cancer is the most lethal gynecological malignancy. Approximately 80–90% of ovarian cancers originate from the ovarian surface epithelium. The etiology of ovarian epithelial cancer (OEC) is not yet fully clarified; currently, the gonadotropin theory of ovarian cancer proposes that elevated serum gonadotropins, follicle-stimulating hormone (FSH) and luteinizing hormone (LH) contribute significantly to the development of ovarian cancer. Wang et al reported that vascular endothelial growth factor (VEGF) plays a role in the development of ovarian cancer, and that elevated gonadotropin levels, as found in menopausal women and most ovarian cancer patients after surgery, could accelerate tumor growth and tumor recurrence by inducing expression of VEGF in OECs (1). However, the detailed molecular mechanisms by which FSH leads to expression of VEGF in OET remain unclear.
Increasing evidence supports the hypothesis that reactive oxygen species (ROS) are involved in the expression and regulation of VEGF and angiogenesis (2–6), and conversely, multiple ROS-mediated cellular functions can be induced by growth factors and hormones (7–10). It has been reported that estrogen-induced ROS-mediated signaling is involved in the development of breast cancer (11). In addition, LH-induced ROS generation contributes to ovulation (12). However, it is not clear whether FSH can induce ROS generation and in turn contribute to FSH-induced VEGF expression.
Cells have developed a variety of protective mechanisms in response to oxidative stress to escape ROS-mediated damage. Activated Nrf2 binds to the antioxidant-response element (ARE) leading to the upregulation of a large number of antioxidant genes (13–18). In a previous study, we observed that Nrf2 was overexpressed in ovarian epithelial carcinoma, and confirmed that FSH could induce the expression of Nrf2 in OEC cells, suggesting that activation of Nrf2 signaling contributes to the development of OEC (34). Kim et al reported that Nrf2 blockade suppresses angiogenesis in colon cancer by inhibiting hypoxia-induced activation of hypoxia-inducible factor 1α (HIF1α) (19), implying that HIF1α signaling is regulated by Nrf2, and suggesting that both HIF1α and Nrf2 signaling may regulate the expression of VEGF and cancer angiogenesis.
Aberrant activation of HIF1α signaling induces expression of VEGF in cancer (20–22). High levels of HIF1α expression are observed in several types of cancer and correlate with a poor prognosis. We previously confirmed that FSH induced the expression of HIF1α in ovarian cancer cells (23) and Lee et al revealed that lysophosphatidic acid (LPA) induced the binding of HIF1α to the VEGF promoter in cancer cells (24). However, it remains to be clarified how FSH-induced HIF1α activation and VEGF expression occurs in OEC cells. Therefore, in this study, we investigated whether ROS regulate FSH-induced expression of VEGF via Nrf2 and HIF1α signaling in OEC cells.
Materials and methods
Reagents and antibodies
Human FSH, dichlorofluorescein (DCF) and N-acetyl cysteine (NAC) were obtained from Sigma-Aldrich (St. Louis, MO, USA). Lipofectamine 2000, DMEM/F12 medium and fetal bovine serum were purchased from Invitrogen (Carlsbad, CA, USA). Anti-Nrf2, GAPDH and HIF1α primary antibodies were purchased from Abcam (Cambridge, UK).
Cell lines and cell culture
Human ES2 (clear cell adenocarcinoma) and Hey (papillary cystadenocarcinoma) cell lines were obtained from the American Culture Collection (Manassas, VA, USA) and were cultured in 1:1 DMEM/F12 supplemented with 100 U/ml penicillin, 100 μg/ml streptomycin and 10% fetal bovine serum at 37°C in a humidified incubator containing 95% room air and 5% CO2.
ROS detection
ES2 and Hey cells were seeded in 6-well plates and incubated for 24 h. The culture medium was replaced with OPTI-MEM and the cells were incubated for 24 h. The cells were then treated with 40 mIU/ml FSH for 20 min or 6 h, incubated with 10 μg/ml DCF for 30 min, and washed three times with PBS, fixed and imaged using a fluorescence microscope.
Western blotting
Western blotting was performed in a routine manner. Briefly, 60 μg protein samples was loaded on 10% SDS-PAGE gels, transferred to polyvinylidene fluoride (PVDF) membranes, incubated with specific primary antibodies at 4°C overnight, and incubated with the appropriate secondary antibody for 1 h at room temperature. The bands were visualized using the ECL Plus system (Amersham, GE Healthcare; Chalfont St. Giles, UK).
ELISA assay
To investigate the effect of NAC on FSH-induced VEGF expression, Hey cells were pretreated with different concentrations of NAC (10, 50, 100 μM) for 30 min, and then treated with 40 mIU/ml FSH for 48 h and the cell media were collected. To determine the effect of blocking ROS signaling and knockdown of Nrf2 or HIF1α on FSH-induced VEGF expression, Hey cells were treated with NAC in the presence or absence of siNrf2 or siHIF1α for the indicated times, and then the cell media were harvested. VEGF protein concentration was measured using an ELISA (R&D Systems, Minneapolis, MN, USA) according to the manufacturer’s instructions.
Chromatin immunoprecipitation assay
A human HIF-1α chromatin immunoprecipitation (ChIP) assay was performed using a kit purchased from R&D Systems, according to the protocol recommended by the manufacturer. Hey cells were treated as indicated in the figure legends, and then the human VEGF promoter was amplified using the primers: forward, 5′-CCTCAGTTCCCTGGCAACATCTG-3′ and reverse, 5′-GAAGAATTTGGCACCAAGTTTGT-3′. The amplification products were examined on 2% agarose gels using ethidium bromide staining.
RNA interference
Small interfering RNAs (siRNAs) against HIF-1α and Nrf2 were designed and synthesized by Dharmacon Thermo Scientific (Waltham, MA, USA). Hey cells were seeded in 6-well plates, cultured to 50% confluence and then serum starved for 24 h. The cells were transiently transfected with siRNA using DharmaFECT transfection reagents according to the manufacturer’s instructions, treated with 40 mIU/ml FSH or NAC for 48 h, and the protein expression levels of downstream target genes were determined by western blotting or using an ELISA.
Statistical analysis
Data are presented as the mean ± standard deviation (SD). Statistical significance was assessed using the Student’s t-test or one-way ANOVA with SPSS 11.5 software (SPSS, Chicago, IL, USA); P-values <0.05 were considered significant.
Results
FSH stimulates ROS generation in ovarian epithelial cancer cells
Increasing evidence supports the importance of ROS as secondary messengers in a variety of cellular functions (25–30). For example, ROS, particularly H2O2, can mimic LH-induced ovulation (12). To investigate the potential involvement of Nrf2 in FSH-induced VEGF expression, Hey and ES2 ovarian epithelial cancer cells were treated with 40 mIU/ml FSH for 20 min or 6 h. As shown in Fig. 1, FSH potently induced ROS production in both Hey and ES2 cell lines. These observations suggest that ROS play a role in FSH-induced cellular function.
ROS generation is required for FSH-induced Nrf2 signaling
Nrf2 is one of the most important cellular defense mechanisms against oxidative stress. Our observation of increased ROS production in FSH-treated cells prompted us to investigate whether Nrf2 is also involved in FSH-induced ovarian cellular function. As expected, FSH treatment induced Nrf2 expression in a dose-dependent manner (Fig. 2A). The peak Nrf2 expression level was observed in Hey and ES2 cells exposed to 40 mIU/ml FSH for 48 h (Fig. 2B).
To confirm that ROS production is required for FSH-induced Nrf2 signaling, we pretreated ovarian cancer cells with the broad-range ROS scavenger N-acetyl cysteine (NAC) at various concentrations for 30 min, and then treated the cells with 40 mIU/ml FSH for 48 h. As shown in Fig. 2C, NAC significantly and dose-dependently attenuated the ability of FSH to induce Nrf2 protein expression. As expected, H2O2 treatment also effectively induced Nrf2 expression in a dose-dependent manner (Fig. 2D), mimicking the effect of FSH; this effect was also inhibited by pretreatment with NAC (Fig. 2E). Collectively, these results suggest that ROS, especially H2O2, are required for FSH-induced activation of Nrf2 signaling.
FSH-induced HIF1α expression requires ROS generation and Nrf2 signaling activation
HIF1α accumulation is necessary for the induction of VEGF expression, and our previous research demonstrated that HIF1α is involved in FSH-induced VEGF expression. Therefore, we hypothesized that ROS production and activation of Nrf2 are necessary for FSH-induced HIF1α expression. To address this, Nrf2 was knocked down using Nrf2-specific siRNA. Knockdown of Nrf2 reduced the expression of HIF1α and abolished FSH-induced HIF1α expression, compared to control siRNA-transfected cells (Fig. 3A). To further confirm the role of ROS as a mediator in FSH-induced HIF1α expression, the cells were pretreated with various doses of NAC prior to treatment with 40 mIU/ml FSH for 48 h. FSH-induced HIF1α expression was considerably attenuated by NAC, suggesting that ROS play a role in FSH-induced HIF1α expression (Fig. 3B).
FSH induces VEGF expression through ROS production and activation of Nrf2-HIF1α signaling
As demonstrated in Fig. 4A, treatment of Hey cells with FSH dramatically enhanced the expression of VEGF; this effect was attenuated in a dose-dependent manner by pretreatment of the cells with NAC. Furthermore, siRNA-mediated knockdown of Nrf2 decreased FSH-induced VEGF expression in a dose-dependent manner (Fig. 4B). ELISA assays were performed to investigate the effect of combined NAC pretreatment and knockdown of HIF1α and/or Nrf2 on FSH-induced VEGF expression. NAC pretreatment, knockdown of HIF1α or knockdown of Nrf2 alone potently attenuated FSH-induced VEGF expression. Combined NAC pretreatment and knockdown of HIF1α did not lead to a significant reduction in FSH-induced VEGF expression, compared to pretreatment with NAC or knockdown of HIF1α alone (Fig. 4C). However, double knockdown of HIF1α and Nrf2 significantly inhibited FSH-induced VEGF expression, compared to pretreatment with FSH, knockdown of HIF1α or knockdown of Nrf2 alone (Fig. 4D).
FSH enhances binding of HIF1α to the VEGF promoter
To further investigate the role of HIF1α in FSH-induced VEGF expression, we explored the ability of FSH to affect the interaction of HIF1α with the native hypoxia response element of the VEGF promoter using the chromatin immunoprecipitation assay (ChIP). As shown in Fig. 5, HIF1α bound to the VEGF promoter could be precipitated using an anti-HIF1α antibody. Treatment of Hey cells with 40 mIU/ml FSH significantly enhanced the ability of HIF1α to bind the VEGF promoter; however, treatment with NAC, siNrf2 or siHIF1α alone or in combination blocked FSH-induced binding of HIF1α to the VEGF promoter.
Discussion
In the present study, our in vitro studies and molecular analyses provided evidence that ROS are necessary for FSH-induced VEGF expression in ovarian cancer, as ablation of ROS or knockdown of Nrf2 attenuated FSH-induced VEGF expression. Our data also demonstrated that the Nrf2 signaling pathway is involved in FSH-induced cellular function, and that FSH enhances the ability of HIF1α to bind the VEGF promoter. Increased FSH levels are a significant risk factor for the development of ovarian cancer. Our previous study demonstrated that activation of the PI3K/AKT pathway mediated FSH-stimulated VEGF expression in ovarian serous cystadenocarcinoma (23). That study also indicates that HIF1α is involved in FSH-induced VEGF expression.
ROS production was observed in OEC cells treated with FSH. ROS, such as hydrogen peroxide, the hydroxyl radical and superoxide anion radical are produced in measurable quantities by every aerobic system, and are considered to be toxic to living cells in high concentrations. Oxyradicals can act as important secondary messengers to regulate a variety of cellular functions. For example, estrogen-induced ROS production contributes to the development of breast cancer (11) and ROS can mimic LH-induced ovulation (12). In agreement with these reports which indicate that ROS mediated-signaling contributes to hormone-induced cellular function, our results clearly demonstrate that ROS are involved in FSH-induced VEGF expression. FSH-induced binding of HIF1α to the VEGF promoter and FSH-induced VEGF expression were attenuated by the antioxidant supplement NAC (Figs. 4A and 5), confirming that ROS are necessary for FSH-induced VEGF expression.
The transcription factor Nrf2 regulates the cellular antioxidant response which protects cells from various insults (17) and facilitates cell survival by inducing intracellular antioxidants, phase II detoxifying enzymes and other molecules that detoxify xenobiotics and neutralize ROS (31–33). In a previous study, we confirmed that Nrf2 was overexpressed in ovarian cancer tissues (34). In agreement with our previous study (34), the present study confirmed that FSH upregulated the expression of Nrf2 in ovarian cancer cells in a dose- and time-dependent manner (Fig. 2A and B). Moreover, H2O2 treatment mimicked FSH-induced Nrf2 expression; this effect was abolished by NAC (Fig. 2C–E). Moreover, knockdown of Nrf2 impaired FSH-induced VEGF expression and reduced the ability of HIF1α to bind the VEGF promoter (Figs. 4B and 5). These data imply that Nrf2 plays a critical role in ROS-mediated FSH-induced VEGF expression. Nrf2 normally exerts a protective role when oxyradicals are present, which raises the question of how the Nrf2-mediated antioxidant response fails to protect ovarian cells in patients with ovarian cancer. We suggest that Nrf2 signaling can easily eliminate FSH-induced ROS generation and prevent ROS-induced damage in the early stages of ovarian epithelial cancer.
According to the gonadotropin theory, persistent stimulation with high concentrations of FSH contributes to the progression of ovarian epithelial cancer. This stimulation may induce persistent ROS generation; therefore, the Nrf2-mediated protective mechanism may become saturated by excessive ROS, resulting in the development of ovarian epithelial cancer. Our data indicate that aberrant activation of Nrf2 in a highly oxidizing environment may facilitate angiogenesis and tumor cell survival in ovarian epithelial cancer.
In a previous study, we demonstrated that FSH regulated the expression of HIF1α in a dose-dependent manner (23). In the present study, we showed that knockdown of Nrf2 impaired FSH-induced HIF1α expression (Fig. 3A). In addition, elimination of ROS using the antioxidant NAC also abolished FSH-induced HIF1α expression (Fig. 3B). These data suggest that ROS and Nrf2 signaling are involved in the regulation of HIF1α expression. Zhou et al reported that HIF1α is indispensable during insulin-induced VEGF transcriptional activation (35). Another study demonstrated that Nrf2-deficient colon cancer cells failed to accumulate HIF1α protein, which limited the expression of VEGF and other HIF1α target genes (19). Our research highlights Nrf2 as a potential candidate molecular target for the control of tumor angiogenesis, as inhibition of Nrf2 may block HIF1α signaling (19). Blockage of ROS using NAC, knockdown of HIF1α or knockdown of Nrf2 attenuated FSH-induced binding of HIF1α to the VEGF promoter. However, knockdown of HIF1α combined with NAC and/or knockdown of Nrf2 to a more significant reduction in VEGF expression (Fig. 4C and D), indicating that Nrf2/HIF1α signaling is involved in FSH-induced VEGF expression. Most importantly, FSH induced the binding of HIF1α to the VEGF promoter, which explains why depletion of HIF1α abolished the expression of VEGF in our previous study (23).
In summary, this study suggests that FSH induces ROS generation, which activates Nrf2 signaling, which in turn triggers HIF1α signaling and promotes the binding of HIF1α to the VEGF promoter, which facilitates ovarian epithelial cancer progression. Prevention of ROS accumulation and targeting of the Nrf2/HIF1α signaling pathway may represent potential strategies to prevent the development of ovarian epithelial cancer.