Abstract
Hypoxia-inducible factor 1 alpha (HIF1A) and endothelin 2 (EDN2) are transiently expressed during the same time window in the developing corpus luteum (CL). In this study, we sought to investigate the involvement of LH/cAMP, reactive oxygen species (ROS), and a hypoxia-mimetic compound (CoCl2) on HIF1A expression and how it affected EDN2 levels, using transformed human granulosa cells (thGCs) and primary bovine granulosa cells (GCs). CoCl2 elevated HIF1A protein levels in thGCs in a dose-dependent manner. Forskolin alone had no significant effect; however, forskolin and CoCl2 together further induced HIF1A protein and EDN2 mRNA expression in thGCs. Similarly, in primary GCs, LH with CoCl2 synergistically augmented HIF1A protein levels, which resulted in higher expression of EDN2 and another well-known hypoxia-inducible gene, VEGF (VEGFA). Importantly, LH alone elevated HIF1A mRNA but not its protein. The successful knockdown of HIF1A in thGCs using siRNA abolished hypoxia-induced EDN2 and also the additive effect of forskolin and CoCl2. We then examined the roles of ROS in thGCs: hydrogen peroxide (20 and 50 μM) elevated HIF1A protein as well as the expression of EDN2, implying that induction of HIF1A protein levels is sufficient to stimulate the expression of EDN2 (and VEGF) in normoxia. A broad-range ROS scavenger, butylated hydroxyanisole, inhibited CoCl2-induced HIF1A protein with a concomitant reduction in the mRNA expression of EDN2 and VEGF in thGCs. The results obtained in this study suggest that HIF1A, induced by various stimuli, is an essential mediator of EDN2 mRNA expression. The results may also explain the rise in the levels of HIF1A-dependent genes (EDN2 and VEGF) in the developing CL.
Introduction
Endothelin 2 (EDN2) is a new player in reproductive processes implicated in ovulation and corpus luteum (CL) formation (Palanisamy et al. 2006, Klipper et al. 2010, Cacioppo et al. 2014). Prevention of EDN2 action in rodents using EDN receptor antagonists resulted in unruptured follicles that failed to develop into corpora lutea (CL; Ko et al. 2006, Palanisamy et al. 2006, Cacioppo et al. 2014). Consistent with the contention that EDN2 is essential for CL formation, we reported that EDN2 mRNA levels were transiently expressed immediately after ovulation, in the early, developing bovine CL. EDN2 in follicles or CL was localized mainly to the luteal steroidogenic cells or luteinized human granulosa cells (GCs; Ko et al. 2006, Klipper et al. 2010, Imbar et al. 2012). GCs collected from patients with polycystic ovary syndrome (PCOS), characterized among other symptoms, by chronic oligo-ovulation or anovulation, had significantly lower EDN2 mRNA expression when compared with normally ovulating women (Imbar et al. 2012). Thus, this study helps in further establishing the roles of EDN2 in follicular rupture, ovulation, and CL formation.
In order to define the regulation of EDN2 gene expression, we had previously reported that hypoxia and luteinizing hormone (LH) each elevated EDN2 levels in GCs (Klipper et al. 2010). In fact, hypoxia (either the mimetic compound, CoCl2, or reduced oxygen) was found to strongly induce EDN2 and VEGF expression in GCs of several species examined thus far (Na et al. 2008, Kim et al. 2009, Klipper et al. 2010). Glucose transporter type 1 (SLC2A1) is yet another known hypoxia-induced gene (Shih & Claffey 1998). Induction of EDN2 expression by LH and hypoxia is physiologically relevant. Hypoxia is an important physiological cue in the developing CL (Nishimura & Okuda 2010) and, together with LH, it orchestrates the formation of CL and the consequent robust angiogenic process. Hypoxia in the CL and also in other fast-growing tissues, most notably tumors, induces the expression of a broad genetic program by stabilizing the transcription factors hypoxia-inducible factor 1 alpha (HIF1A) and hypoxia-inducible factor 2 alpha (HIF2A) (Semenza 2007, Brahimi-Horn & Pouyssegur 2009). Under normoxic conditions, HIF1A undergoes O2-dependent hydroxylation by members of the prolyl hydroxylase domain (PHD) family (for reviews, see Schofield & Ratcliffe (2004) and Kaelin (2005)). Hydroxylation marks HIF1A for ubiquitination and proteasomal degradation (Semenza 2007). When cellular oxygen levels fall, HIF1A hydroxylation is suppressed and HIF1A rapidly accumulates. Consequently, HIF1A is translocated into the nucleus and dimerizes with the constitutively expressed HIF2A subunit. The active HIF1 complex then initiates transcription of target genes by binding to hypoxia-responsive elements located in their promoter or enhancer regions (Hewitson et al. 2007), as described for genes such as VEGF and SLC2A1 (Shih & Claffey 1998).
HIF1A can also be hormonally regulated in an oxygen-independent manner. Insulin-like growth factor 1, for instance, augmented the accumulation of HIF1A in normoxia in several cell types (Alam et al. 2009, Sartori-Cintra et al. 2012, Yu et al. 2012). Reactive oxygen species (ROS) is another factor implicated in HIF1A accumulation. Evidence suggests that ROS increases HIF1A by directly inhibiting PHD catalytic activity (Cash et al. 2007, Kaelin & Ratcliffe 2008). Finally, reports have also suggested that hypoxia can act independently of the HIF1A transcription factor (Bindra et al. 2005, Arany et al. 2008, Fujisaka et al. 2013).
In this study, we wished to investigate whether: i) hypoxia-induced EDN2 is HIF1A dependent, ii) LH or other cAMP-elevating agents such as forskolin also utilize HIF1A as a tool to induce EDN2, and iii) other potential inducers of HIF1A, such as ROS, would also affect EDN2 levels in GCs. To resolve these questions, we first studied the effects of LH (or cAMP) and a hypoxia-mimetic compound separately or together on HIF1A expression (mRNA and protein) and how it affected EDN2 levels. To critically determine the involvement of HIF1A in EDN2 mRNA induction, HIF1A was knocked down with specific siRNA molecules. The involvement of ROS in these processes on HIF1A accumulation and gene expression in GCs was studied using H2O2 and a broad-range ROS scavenger.
Materials and methods
Cell cultures
Transformed human granulosa lutein cells (thGCs; also known as SVOG) were a generous gift from N Auersperg (University of British Columbia, BC, Canada) (Lie et al. 1996). Cells were cultured in M-199/MCDB 105 (1:1), containing 10% FCS, 2 mM l-glutamine, 400 ng/ml hydrocortisone, and 50 μg/ml gentamicin sulfate (Lie et al. 1996, Kisliouk et al. 2003, Klipper et al. 2010). Cells from passages 6–22 were used. Cells were grown to 80–90% confluence on Petri dishes, trypsinized with trypsin–EDTA solution, and plated (0.3×106 cells/well) on six-well plates. The following day, cells were incubated with a hypoxia-mimetic compound CoCl2 (25–150 μM) alone or in the presence of 10 μM forskolin in 1% FCS for 3–24 h as indicated. Forskolin, as an inducer of adenylyl cyclase, was used as these cells have lost most of their response to LH/hCG (Lie et al. 1996). For experiments examining the effects of ROS, thGCs were incubated with varying doses of H2O2 (20, 50, or 100 μM), given as repeated boluses every 15 min in a serum-free medium (Chandel & Schumacker 2000). After 2 h, cells were collected in a sample buffer for protein analysis of HIF1A by western blot or, after 3 h, for RNA analysis of HIF1A-inducible genes by real-time PCR, or thGCs were preincubated with a broad-range ROS scavenger, butylated hydroxyanisole (BHA; 250 μM), for 1 h and then incubated with CoCl2 for 4 or 16 h (for protein isolation and RNA extraction respectively) with or without forskolin in the presence or absence of BHA.
Primary bovine GC culture
Ovaries bearing large follicles (>10 mm in diameter) were collected at a local slaughterhouse as described previously (Meidan et al. 1990, Aflalo & Meidan 1993, Mamluk et al. 1998). Only follicles containing >4 million viable cells were included in these experiments. GCs were enzymatically dispersed using a combination of collagenase type IA (5000 units), hyaluronidase III (1440 units), and DNase I (390 units) (Meidan et al. 1990, Aflalo & Meidan 1993, Mamluk et al. 1998), and cultured overnight in DMEM–F12 containing 3% FCS, 2 mM l-glutamine, and penicillin (100 U/ml)/streptomycin (1 mg/ml) solution (Biological Industries, Kibbutz Beit Hemeek, Israel). One day after isolation, bovine GCs (bGCs) were exposed to CoCl2 and/or bLH (100 ng/ml) for various time points as indicated (for RNA extraction and protein isolation respectively) at 1% FCS. Unless otherwise stated, biochemicals were purchased from Sigma, and tissue culture material from Biological Industries.
Cell transfection
ThGCs were trypsinized and plated immediately on six-well plates and cultured for 24 h. Then cells were transfected using Lipofectamine 2000 reagent in 1% FCS, as described previously (Rayhman et al. 2008, Klipper et al. 2010). Cells were transfected with 50 nM siRNA constructs targeting HIF1A or with scrambled siRNA (the negative control). The HIF1A siRNA (si-HIF1A) sequence was sense, CUGAUGACCAGCAACUUGAdTdT and antisense, UCAAGUUGCUGGUCAUCAGdTdT. Scrambled siRNA sequence-negative controls were sense, UUCUCCGAACGUGUCACGUdTdT and antisense, ACGUGACACGUUCGGAGAAdTdT.
RNA isolation and real-time PCR
Total RNA was isolated from tissue and cells using Tri-Fast reagent (Peqlab Biotechnologie GmbH, Erlangen, Germany) according to the manufacturer's instructions. Real-time PCRs were performed using the Mx3000P quantitative PCR system (Stratagene, Garden Grove, CA, USA), using Platinum SYBR Green qPCR SuperMix–UDG (Invitrogen), as described previously (Kisliouk et al. 2005, Zalman et al. 2012). ACTB was used as the reference gene. Dissociation curves after each real-time run confirmed the presence of only one product and the absence of the formation of primer dimers. The threshold cycle number (Ct) for each tested gene X was used to quantify the relative abundance of the gene; arbitrary units were calculated as 2−ΔCt=2–(Ct target gene X–Ct reference gene). The primer sequences were listed in Table 1.
Lists of primers for bovine and human genes.
| Primer sequence (5′–3′) | Accession no. | |
|---|---|---|
| Bovine genes | ||
| β-actin | F: CGGGCAGGTCATCACCATC | AC_000182 |
| R: GCCATCTCGTACTCGAAGTCC | ||
| EDN2 | F: TTGCCACCTGGACATTATCT | NM_175714 |
| R: CCTGGCACTGTAGCACTCA | ||
| VEGF | F: CCATGAACTTTCTGCTCTCTTGG | NM_174216 |
| R: TCCATGAACTCCACCACTTCG | ||
| HIF1A | F: ACTCATCCATGTGACCACG | NM_174339 |
| R: TAGTTCTCCCCCGGCTAG | ||
| SLC2A1 | F: CGCTTCCTGCTCATTAACCG | NM_174602 |
| R: CCTTCTTCTCCCGCATCAT | ||
| Human genes | ||
| EDN2 | F: GCCAGCGTCCTCATCTAT | NM_001956 |
| R: GCCGTAAGGAGCTGTCTGTTC | ||
| VEGF | F: ATCGAGACCCTGGTGGACA | NM_001025366 |
| R: CCTCGGCTTGTCACATCTGC | ||
| HIF1A | F: ACTCATCCATGTGACCACG | NM_001530.3 |
| R: TAGTTCTCCCCCGGCTAG | ||
| SLC2A1 | F: CGCTTCCTGCTCATTAACCG | NM_006516.2 |
| R: CCTTCTTCTCCCGCATCAT | ||
| HPSE | F: CTGAAGGCTGGTGGAGAAGT | NM_006665.5 |
| R: CTAACCAGACCTTCTTGC | ||
| PTGS2 | F: CCCTTCCTCCTGTGCCTGATGA | NM_000963.3 |
| R: GTGAAGTGCTGGGCAAAGAATG | ||
F, forward; R, reverse.
Western blot analyses
Cells were washed with PBS and proteins were extracted by scraping cells in a sample buffer (100 mmol/l Tris–HCl (pH 6.8), 20% (v/v) glycerol, 4% (w/v) SDS, 0.0002% (w/v) bromophenol blue, and 1% (v/v) β-mercaptoethanol). All steps were carried out on ice, and samples were kept frozen until use. The protein samples were separated by 7.5–10% SDS–PAGE under reducing conditions. Proteins were electrically transferred onto a nitrocellulose membrane. After having been blocked for 2 h at 4 °C in TBST (20 mmol/l Tris, 150 mmol/l NaCl, and 0.1% Tween 20; pH 7.6) containing 5% low-fat milk, the membranes were incubated with the following respective primary antibodies overnight at 4 °C: rabbit anti-HIF1A (H-206; Santa Cruz Biotechnology, Inc., diluted 1:500) used to detect human HIF1A, monoclonal anti-HIF1A (Novus Biological (Cambridge, UK), diluted 1:500), used to detect bovine HIF1A, and rabbit anti-bVEGF (kindly provided by D Schams (Germany), diluted 1:1000) and rabbit anti p44/42 total MAPK (diluted 1:50 000; used as the loading control) for 2 h at room temperature. Then, the membranes were washed with TBST (4×5 min) and incubated with secondary antibodies diluted in 1% nonfat dry milk in TBST, peroxidase-conjugated goat anti-rabbit IgG (Jackson ImmunoResearch, West Grove, PA, USA), or goat anti-mouse IgG (H+L) (KPL, Baltimore, MD, USA) for 1 h at room temperature and washed as described previously. A chemiluminescent signal was generated with EZ-ECL Chemiluminescence Detection Kit for HRP (Biological Industries), and the membranes were exposed to X-ray films. The films were scanned and analyzed using the Gel-Pro 32 Software (Media Cybernetics, Silver Spring, MD, USA).
Statistical analysis
Statistical analysis was performed by one-way ANOVA followed by the Bonferroni's multiple comparison test (GraphPad Prism Software, version 6.00, San Diego, CA, USA). Results represent the mean±s.e.m. for at least three separate experiments as specified. Significance was defined at a value of P<0.05 or lower.
Results
Effects of CoCl2, forskolin, and LH on HIF1A and EDN2 expression
The studies outlined herein have utilized CoCl2 as a hypoxia-mimetic compound. CoCl2 as well as reduced oxygen tension both augmented EDN2, VEGF, and SLC2A1 mRNA levels by GCs (Supplementary Figure, see section on supplementary data given at the end of this article). HIF1A protein was determined in cells incubated with various concentrations (25–150 μM) of CoCl2 only or with forskolin (Fig. 1A). Forskolin (used to activate adenylyl cyclase) given alone did not elevate HIF1A protein levels in thGCs, but it additively augmented HIF1A levels that were stimulated by CoCl2 (1.7-fold on average compared with CoCl2 alone, P<0.04, P<0.05, and P<0.05 for 25, 100, and 150 μM CoCl2+forskolin when compared with CoCl2 alone respectively; Fig. 1B). Examination of EDN2 gene expression revealed a significant additive effect (P<0.02) after incubation with forskolin and CoCl2 (Fig. 1C). To examine whether LH produces a similar effect, we employed primary bGCs that express the functional LH/hCG receptors. A day after isolation, bGCs were incubated with varying concentrations of CoCl2 (25–100 μM) with or without LH (100 ng/ml) (Fig. 2). As with thGCs, there was a dose-dependent increase in CoCl2-induced HIF1A protein in bGCs (Fig. 2A and B). LH alone did not elevate the HIF1A protein above the control levels. However, there was a synergistic, dose-dependent effect of LH with CoCl2 on HIF1A protein levels (Fig. 2A and B). For instance, in the presence of LH, HIF1A protein levels were approximately threefold higher when compared with levels induced by 100 μM CoCl2 alone (Fig. 2). The data presented in Fig. 2 also indicate that LH and CoCl2 exert different effects on the protein (Fig. 2B) and mRNA levels (Fig. 2C) of HIF1A. LH only elevated the mRNA levels of HIF1A without affecting its protein concentration. LH and CoCl2 synergized in elevating HIF1A protein levels but not its mRNA levels (Fig. 2B and C). CoCl2 slightly, but significantly, elevated HIF1A gene expression (Fig. 2C). Importantly however, HIF1A protein levels were elevated only in the presence of CoCl2 (Fig. 2B).
Effect of CoCl2 and forskolin on HIF1A protein and EDN2 mRNA. ThGCs were cultured with varying concentrations of CoCl2 with or without forskolin (10 μM) for 4 h. (A) Representative western blots of HIF1A protein and total MAPK. (B) Quantification of HIF1A protein using GelPro 32 is depicted as the percentage change from the control (designated 100%); the signal of anti-total MAPK (p44/42) antibody was used to correct for protein loading. (C) Cells were harvested for RNA extraction and the relative EDN2 mRNA levels were measured by real-time PCR. *Significant differences from their respective controls. #Statistically significant differences between CoCl2 alone and CoCl2 in combination with forskolin (P<0.05 for the protein and P<0.02 for EDN2 mRNA). The results (mean±s.e.m.) were obtained from four independent experiments.
Citation: REPRODUCTION 149, 1; 10.1530/REP-14-0409
Effect of LH and CoCl2 on mRNA and protein levels of HIF1A in bGCs. Cells were incubated with varying concentrations of CoCl2 with or without LH (100 ng/ml) for 4 h. (A) Representative western blot (left panel) and quantification of HIF1A protein (right panel). (B) Quantification of HIF1A protein using GelPro 32 is depicted as the percentage change from the control (designated 100%); the membranes were also probed with an anti-total MAPK (p44/42) antibody to correct for protein loading. (C) Relative HIF1A mRNA levels were measured by real-time PCR. *Significant differences from their respective controls. #Statistically significant differences between CoCl2 alone and CoCl2 in combination with LH (P<0.01). The results (mean±s.e.m.) were obtained from three and five independent experiments for protein and for mRNA levels respectively.
Citation: REPRODUCTION 149, 1; 10.1530/REP-14-0409
To examine whether the synergistic effect of LH and CoCl2 on HIF1A induction was transcriptionally active, we exposed cells to these stimuli and examined gene expression. The data presented in Fig. 3 indicate that the combined effect of LH and CoCl2 on HIF1A induction was manifested in the profile of EDN2 and two known hypoxia-induced genes: VEGF and SLC2A1; approximately twofold increase).
Effect of CoCl2 and LH on the expression of hypoxia-induced genes in bGCs. Cells were cultured with CoCl2 (100 μM) with or without LH for 4–24 h. At the end of the incubation period, the cells were harvested for RNA extraction and EDN2, VEGF, and SLC2A1 mRNA levels were measured by real-time PCR. *Significant differences from their respective controls. The horizontal bar denotes the statistical difference between CoCl2 alone and CoCl2 in combination with LH. The results (mean±s.e.m.) were obtained from four independent experiments.
Citation: REPRODUCTION 149, 1; 10.1530/REP-14-0409
Effects of HIF1A inhibition by siRNA on EDN2 expression
To critically establish the role of HIF1A in EDN2 expression in GCs, we silenced its expression with specific siRNA. Transfection of HIF1A constructs effectively reduced HIF1A protein concentrations induced by CoCl2 approximately to 15% (P<0.001) of levels present in cells transfected with scrambled siRNA (Fig. 4A). Specificity of HIF1A silencing was demonstrated by the lack of inhibition in levels of non-hypoxia-dependent genes such as heparanase (HPSE) and prostaglandin-endoperoxide synthase 2 (PTGS2) (Fig. 4B). To examine the effect of HIF1A silencing on gene expression, cells were transfected with either scrambled siRNA or HIF1A-specific siRNA and, 48 h later, were exposed to CoCl2 for an overnight incubation. The results, presented in Fig. 5, demonstrate that HIF-silenced and CoCl2-treated cells exhibited significantly lower expression of EDN2 as well as of VEGF and SLC2A1. This was evident for both 25 and 100 μM CoCl2. HIF1A silencing also successfully reduced VEGF protein present in cells transfected with scrambled siRNA (Fig. 5). Importantly, as can be observed from Table 2, HIF1A knockdown not only reduced the effect of a hypoxia-mimetic compound but also abolished the additive effect of forskolin and CoCl2 on EDN2 mRNA levels. HIF1A silencing lowered EDN2 levels induced by forskolin alone, but this was not statistically significant (Table 2).
HIF1A silencing. ThGCs were transfected with 50 nM of scrambled siRNA (negative control) or a HIF1A siRNA-specific sequence. Forty-eight hours after transfection, the cells were treated with 100 or 150 μM CoCl2 for 4 h. (A, upper panel) Representative western blot of HIF1A protein in silenced vs non-silenced thGCs. (A, lower panel) Quantification of HIF1A protein levels (percentage change from the negative control designated 100%). (B) At the end of the incubation period, the cells were harvested for RNA extraction and the mRNA levels of HPSE and PTGS2 were measured by real-time PCR. *Significant differences from their respective controls. #Statistically significant differences (P<0.001) between cells treated with HIF1A and scrambled siRNA. The results (mean±s.e.m.) were obtained from three independent experiments.
Citation: REPRODUCTION 149, 1; 10.1530/REP-14-0409
Effect of HIF1A silencing on EDN2, VEGF, and SLC2A1 expression. Forty-eight hours after transfection, thGCs were incubated overnight with 25 or 100 μM CoCl2. At the end of the incubation period, the cells were harvested for RNA extraction and EDN2 (A), SLC2A1 (B), and VEGF (C) mRNA levels were measured by real-time PCR. Cells were also harvested for VEGF protein (D) determination using a specific antibody (representative blot – upper panel and quantification lower panel). *Significant differences from their respective controls. #Statistically significant differences (P<0.001 for SLC2A1 and EDN2; P<0.05 for VEGF mRNA and protein) between cells treated with HIF1A and scrambled siRNA. The results (mean±s.e.m.) were obtained from three independent experiments.
Citation: REPRODUCTION 149, 1; 10.1530/REP-14-0409
Effect of HIF1A silencing on EDN2 mRNA levels by CoCl2 and forskolin.
| Treatment | Scrambled siRNA | HIF1A siRNA |
|---|---|---|
| Control | 100 | 94±20 |
| Forskolin (10 μM) | 213±55 | 140±32, NS |
| CoCl2 (25 μM) | 268±46 | 73±21* |
| CoCl2+forskolin | 475±98 | 110±39* |
ThGCs were transfected with siRNA molecules as indicated in the legend to Fig. 5. EDN2 mRNA levels were determined by quantitative real-time PCR. The results (mean±s.e.m.) were obtained from three independent experiments. *Significant differences between scrambled and HIF1A siRNA. NS, non significant.
Effects of ROS on HIF1A and EDN2 expression
The involvement of ROS was studied using H2O2 and a broad-range ROS scavenger (BHA). The results presented in Fig. 6 A and B indicate that H2O2 augmented the levels of HIF1A protein by approximately fourfold when compared with the control. The magnitude of stimulation by 20 or 50 μM was similar, suggesting that 20 μM of H2O2 already attained the maximal stimulatory effect. H2O2 also significantly elevated the expression of EDN2 and VEGF mRNAs (Fig. 6); actually the effect of H2O2 on EDN2 mRNA was higher than that of forskolin (Fig. 6). To further investigate the role of ROS, we examined the effects of BHA, ROS scavenger. The presence of BHA significantly reduced HIF1A protein levels elevated by hypoxia (2.5 times, on an average, less compared with CoCl2 alone; P<0.05; Fig. 7). Notably, BHA not only reduced the levels of HIF1A and the genes induced by CoCl2 but also eliminated the combined stimulatory effect of CoCl2 and forskolin on HIF1A protein and EDN2 and VEGF mRNA levels (Figs 7 and 8).
Effect of H2O2 on HIF1A protein and gene expression. (A) ThGCs were incubated with H2O2 for 2 h and harvested for protein extraction. Quantification of HIF1A protein western blot was carried out as described in the legend to Fig. 1. (B) Cells were incubated with H2O2 or forskolin for 3 h, then were harvested for RNA extraction, and EDN2 and VEGF mRNA levels were measured by real-time PCR. Results are expressed as mean±s.e.m. for four independent experiments. *Significant differences from their respective controls.
Citation: REPRODUCTION 149, 1; 10.1530/REP-14-0409
Effect of ROS scavengers on HIF1A protein levels. ThGCs were incubated with CoCl2 (100 μM), forskolin (10 μM), or the two treatments together in the presence or absence of BHA (250 μM) for 4 h. The incubation period was preceded by preincubation for 1 h (with BHA or a control medium). Western blot and quantification of HIF1A protein were carried out as described in the legend to Fig. 1, values are relative to the control (100%). Results are expressed as mean±s.e.m. for four independent experiments. *Significant differences from respective controls (without CoCl2). @Statistically significant differences between cells incubated with the same treatments with and without BHA. Horizontal bar represents the statistical difference between CoCl2 alone and CoCl2 in combination with forskolin.
Citation: REPRODUCTION 149, 1; 10.1530/REP-14-0409
Effect of BHA on expression of EDN2 and VEGF. ThGCs were preincubated with BHA (250 μM) for 1 h and then incubated with CoCl2 (25 or 100 μM) with or without forskolin (10 μM) in the presence or absence of BHA (250 μM). After overnight incubation, RNA was extracted and gene expression was determined by real-time PCR. The results (mean±s.e.m.) were obtained from four independent experiments. *Significant differences from their respective controls (without CoCl2). #Statistically significant differences between CoCl2 alone and CoCl2 in combination with forskolin for VEGF (P<0.06 and P<0.05 for 25 and 100 μM respectively), and for EDN2 (P<0.01 and P<0.02 for 25 and 100 μM between CoCl2 alone and CoCl2 in combination with forskolin respectively).
Citation: REPRODUCTION 149, 1; 10.1530/REP-14-0409
Discussion
There was a close relationship between the profiles of HIF1A protein and EDN2 expression in the various experiments utilizing primary GCs and thGCs described in this study. Stimulation of HIF1A by either the hypoxia-mimetic compound alone or in combination with cAMP-elevating agents (forskolin and LH), or yet by H2O2, resulted in higher EDN2 mRNA expression (Fig. 9). Moreover, the synergistic effect of LH and CoCl2 on HIF1A induction was also manifested in the profile of EDN2. In accordance, inhibition of HIF1A by siRNA or by an ROS scavenger reduced EDN2 levels. Thus, the findings reported in this study imply that HIF1A is a critical mediator of EDN2 expression in GCs. The profile of EDN2 expression closely resembled that of VEGF and SLC2A1 – two hypoxia-induced genes that are known to be activated by HIF1A, further portraying EDN2 as a bona fide hypoxia-dependent gene.
Schema summarizing HIF1A protein accumulation induced by a hypoxia-mimetic compound (CoCl2), LH/cAMP, and ROS (H2O2) in GCs. CoCl2 and ROS inhibit PHD-dependent hydroxylation of HIF1A, thereby stabilizing its protein levels (Chandel et al. 1998, Cash et al. 2007, Semenza 2007, Kaelin & Ratcliffe 2008). LH and forskolin stimulate HIF1A mRNA further elevating HIF1A protein levels in the presence of a hypoxia-mimetic compound. In addition, ROS production is enhanced by CoCl2 (Chandel & Schumacker 2000, Grasselli et al. 2005) and LH (Yacobi et al. 2007). Stabilized HIF1A protein then translocates into the nucleus, where it binds hypoxia-responsive elements (HREs) in the 5′ ends of hypoxia-dependent genes such as EDN2, VEGF, and SLC2A1.
Citation: REPRODUCTION 149, 1; 10.1530/REP-14-0409
The studies outlined herein have utilized CoCl2 as a hypoxia-mimetic agent, though, CoCl2 may not mimic all the effects of hypoxia (reduced oxygen tension); it induces HIF1A protein and also mimics many of its physiological effects (Chandel & Schumacker 2000). Critically, Supplementary Figure demonstrates that CoCl2 and reduced oxygen tension both significantly augmented EDN2, VEGF, and SLC2A1 mRNA levels by GCs.
Hypoxic conditions exist in the newly formed CL, because the angiogenic process lags behind the intense luteal cell proliferation. Many studies utilizing various animal species have demonstrated the accumulation, nuclear localization, and activity of HIF1 proteins following ovulation and CL development (Boonyaprakob et al. 2005, Duncan et al. 2008). For example, HIF1A was detected in the early luteal stage in the cow, but no expression could be found in later stages of the cycle (Nishimura & Okuda 2010, Jiang et al. 2011). Similarly, in human CL, nuclear HIF1A protein in granulosa lutein cells was highest during luteal formation and was absent from fully functional CL (van den Driesche et al. 2008). Hypoxia is therefore an important physiological cue in the developing CL. It is well established that hypoxia plays an important role in the up-regulation of VEGF that occurs at this stage and is responsible for the angiogenic process in early CL (Tesone et al. 2005, van den Driesche et al. 2008, Nishimura & Okuda 2010, Meidan et al. 2013). The current study shows that the hypoxia-mimetic compound elevated HIF1A protein and EDN2 mRNA levels, whereas HIF1A silencing (employed herein for the first time in GCs) reduced EDN2 expression and ablated its response to hypoxia. This therefore suggests that HIF1A in the developing CL may be responsible for high EDN2 mRNA levels present at this stage (Klipper et al. 2010). In addition to hypoxia, EDN2 and VEGF were induced in cells treated with forskolin or LH/hCG on their own, as demonstrated in quite a few previous studies (Klipper et al. 2010, Zhang et al. 2012, this study). This increase appears to be independent of HIF1A, because HIF1A protein was not induced by the cAMP-elevating agents and HIF1A silencing did not inhibit forskolin-induced EDN2 expression. But more than each stimuli alone, we show herein that there was a synergism between LH and the hypoxia-mimetic compound in inducing the levels of HIF1A and EDN2 (Fig. 9). Indeed, the highest EDN2 mRNA levels were observed in early bovine CL (∼30 h after ovulation) when both LH and hypoxia occur (Klipper et al. 2010). We show that the combined effect is the result of HIF1A being induced transcriptionally and post-transcriptionally, first enhanced transcription by LH, which cannot be manifested in higher HIF1A protein levels unless the protein is stabilized under hypoxic conditions (Fig. 9). This issue was not clearly resolved in earlier studies that have not measured the separate and combined effects of hypoxia and cAMP in the mRNA and protein levels.
Exposure of GCs to the hypoxia-mimetic compound also enhances the levels of HIF1A and HIF2A proteins (Kim et al. 2009). Although HIF1A and HIF2A have many overlapping functions, recent studies have illustrated distinct roles for each α-isoform in both normal physiology and disease (Keith et al. 2012). Delineation of a potential role for HIF2A in ovarian function will require further research.
PCOS is one of the most common endocrine metabolic disorders in women of reproductive age (Asuncion et al. 2000). It is characterized by anovulation, oligomenorrhea or amenorrhea, hyperandrogenism, and obesity and is related to insulin (Legro et al. 2004). Recently, we have reported that significantly lower levels of EDN2 were present in GCs derived from women with PCOS, compared with normally ovulating women (Imbar et al. 2012). Their response in culture to luteotropic agents was also different from that in the control group (Imbar et al. 2012), further demonstrating the importance of LH/hCG in EDN2 induction. The involvement of hypoxic conditions in the abnormal responses of GCs derived from women with PCOS is yet to be investigated.
There is increasing evidence that ROS, particularly H2O2, also operate as signaling molecules to mediate various responses (Rhee 2006, Forman et al. 2010). It may seem paradoxical that cells respond to low O2 levels by elevating ROS formation, yet there is ample evidence that ischemia, chemical hypoxia, or low oxygen tension can induce ROS in some or most tissues (Guzy et al. 2005, Guzy & Schumacker 2006, Walshe & D'Amore 2008). ROS can modulate HIF1A activity by directly inhibiting the PHD catalytic activity (Chandel et al. 1998, Cash et al. 2007, Kaelin & Ratcliffe 2008). Substantial data link LH and ROS; for instance, LH elevated ROS levels in ovarian follicles (Yacobi et al. 2007) and hydrogen peroxide up-regulated the expression of ovulatory genes, mimicking the effect of LH (Shkolnik et al. 2011). Furthermore, LH-induced expansion of cumulus mass could be inhibited by the ROS scavenger, BHA (Shkolnik et al. 2011). However, these studies have not explored the involvement of HIF1A in ROS action. Our data are the first to provide evidence, suggesting that ROS actions in ovarian cells are mediated by HIF1A (Fig. 9). We found that H2O2 increased the levels of HIF1A protein in normoxic conditions. Notably, H2O2 significantly elevated EDN2 and VEGF mRNA levels confirming the ability of H2O2 to mimic LH actions as proposed for rat GCs (Shkolnik et al. 2011). In this study, the involvement of ROS in stabilizing HIF1A protein was also observed in cells treated with BHA. In thGCs, BHA significantly reduced the levels of HIF1A protein induced by CoCl2. The combined stimulatory effect of forskolin and CoCl2 on HIF1A protein was also significantly decreased by the presence of BHA. Notably, the expression levels of HIF1A-dependent genes (EDN2 and VEGF) were markedly reduced by BHA, emphasizing the importance of HIF1A protein for transactivation of these genes. Our findings are in agreement with those of Chandel et al. (1998). These authors reported that Ebselen (a synthetic glutathione peroxidase mimetic that degrades H2O2 to H2O) abolished mRNA expression of erythropoietin (a model protein expressed upon hypoxia), VEGF, and glycolytic enzymes during low oxygen tension and CoCl2 exposure (Chandel et al. 1998). The results of this study and those cited earlier indicate that ROS, most probably H2O2, contributes to the induction of gene expression by hypoxia (Fig. 9). These data also indicate that the mechanism underlying the synergism between LH and hypoxia may also result from the direct effects on HIF1A (mRNA and protein respectively), but may involve ROS as well. LH and hypoxia can each generate ROS production and their combined effect can result in a stronger inhibition of PHD activity, elevating consequently higher HIF1A protein levels. It should be noted that although both low oxygen levels and CoCl2 can generate ROS (most probably H2O2) (Chandel et al. 1998, Grasselli et al. 2005), CoCl2 acts via a mitochondria-independent mechanism (Chandel et al. 1998).
In conclusion, the data presented in this study indicate that HIF1A is an essential mediator of LH/cAMP, hypoxia, and ROS in their induction of EDN2 expression (Fig. 9). These findings are physiologically significant as they suggest a possible mechanism for optimal up-regulation of EDN2 and VEGF during ovulation and CL formation.
Supplementary data
This is linked to the online version of the paper at http://dx.doi.org/10.1530/REP-14-0409.
Declaration of interest
The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the research reported.
Funding
This work was supported by a grant to R Meidan from the Israel Science foundation (ISF) grant number 263/10, and a grant to T Imbar and R Meidan from the Joint Research Fund between the Hebrew University Faculty of Medicine and Hadassah Hebrew University Medical Center.
References
Aflalo L & Meidan R 1993 The hormonal regulation of cholesterol side-chain cleavage cytochrome P450, adrenodoxin, and their messenger ribonucleic acid expression in bovine small-like and large-like luteal cells: relationship with progesterone production. Endocrinology 132 410–416.
Alam H, Weck J, Maizels E, Park Y, Lee EJ, Ashcroft M & Hunzicker-Dunn M 2009 Role of the phosphatidylinositol-3-kinase and extracellular regulated kinase pathways in the induction of hypoxia-inducible factor (HIF)-1 activity and the HIF-1 target vascular endothelial growth factor in ovarian granulosa cells in response to follicle-stimulating hormone. Endocrinology 150 915–928. (doi:10.1210/en.2008-0850)
Arany Z, Foo SY, Ma Y, Ruas JL, Bommi-Reddy A, Girnun G, Cooper M, Laznik D, Chinsomboon J & Rangwala SM et al. 2008 HIF-independent regulation of VEGF and angiogenesis by the transcriptional coactivator PGC-1α. Nature 451 1008–1012. (doi:10.1038/nature06613)
Asuncion M, Calvo RM, San Millan JL, Sancho J, Avila S & Escobar-Morreale HF 2000 A prospective study of the prevalence of the polycystic ovary syndrome in unselected Caucasian women from Spain. Journal of Clinical Endocrinology and Metabolism 85 2434–2438.
Bindra RS, Schaffer PJ, Meng A, Woo J, Maseide K, Roth ME, Lizardi P, Hedley DW, Bristow RG & Glazer PM 2005 Alterations in DNA repair gene expression under hypoxia: elucidating the mechanisms of hypoxia-induced genetic instability. Annals of the New York Academy of Sciences 1059 184–195. (doi:10.1196/annals.1339.049)
Boonyaprakob U, Gadsby JE, Hedgpeth V, Routh PA & Almond GW 2005 Expression and localization of hypoxia inducible factor-1 alpha mRNA in the porcine ovary. Canadian Journal of Veterinary Research 69 215–222.
Brahimi-Horn MC & Pouyssegur J 2009 HIF at a glance. Journal of Cell Science 122 1055–1057. (doi:10.1242/jcs.035022)
Cacioppo JA, Oh SW, Kim HY, Cho J, Lin PC, Yanagisawa M & Ko C 2014 Loss of function of endothelin-2 leads to reduced ovulation and CL formation. PLoS ONE 9 e96115. (doi:10.1371/journal.pone.0096115)
Cash TP, Pan Y & Simon MC 2007 Reactive oxygen species and cellular oxygen sensing. Free Radical Biology & Medicine 43 1219–1225. (doi:10.1016/j.freeradbiomed.2007.07.001)
Chandel NS & Schumacker PT 2000 Cellular oxygen sensing by mitochondria: old questions, new insight. Journal of Applied Physiology 88 1880–1889. (doi:10.1063/1.1303764)
Chandel NS, Maltepe E, Goldwasser E, Mathieu CE, Simon MC & Schumacker PT 1998 Mitochondrial reactive oxygen species trigger hypoxia-induced transcription. PNAS 95 11715–11720. (doi:10.1073/pnas.95.20.11715)
van den Driesche S, Myers M, Gay E, Thong KJ & Duncan WC 2008 HCG up-regulates hypoxia inducible factor-1 alpha in luteinized granulosa cells: implications for the hormonal regulation of vascular endothelial growth factor A in the human corpus luteum. Molecular Human Reproduction 14 455–464. (doi:10.1093/molehr/gan040)
Duncan WC, van den Driesche S & Fraser HM 2008 Inhibition of vascular endothelial growth factor in the primate ovary up-regulates hypoxia-inducible factor-1 alpha in the follicle and corpus luteum. Endocrinology 149 3313–3320. (doi:10.1210/en.2007-1649)
Forman HJ, Maiorino M & Ursini F 2010 Signaling functions of reactive oxygen species. Biochemistry 49 835–842. (doi:10.1021/bi9020378)
Fujisaka S, Usui I, Ikutani M, Aminuddin A, Takikawa A, Tsuneyama K, Mahmood A, Goda N, Nagai Y & Takatsu K et al. 2013 Adipose tissue hypoxia induces inflammatory M1 polarity of macrophages in an HIF-1α-dependent and HIF-1α-independent manner in obese mice. Diabetologia 56 1403–1412. (doi:10.1007/s00125-013-2885-1)
Grasselli F, Basini G, Bussolati S & Bianco F 2005 Cobalt chloride, a hypoxia-mimicking agent, modulates redox status and functional parameters of cultured swine granulosa cells. Reproduction, Fertility, and Development 17 715–720. (doi:10.1071/RD05059)
Guzy RD & Schumacker PT 2006 Oxygen sensing by mitochondria at complex III: the paradox of increased reactive oxygen species during hypoxia. Experimental Physiology 91 807–819. (doi:10.1113/expphysiol.2006.033506)
Guzy RD, Hoyos B, Robin E, Chen H, Liu L, Mansfield KD, Simon MC, Hammerling U & Schumacker PT 2005 Mitochondrial complex III is required for hypoxia-induced ROS production and cellular oxygen sensing. Cell Metabolism 1 401–408. (doi:10.1016/j.cmet.2005.05.001)
Hewitson KS, Schofield CJ & Ratcliffe PJ 2007 Hypoxia-inducible factor prolyl-hydroxylase: purification and assays of PHD2. Methods in Enzymology 435 25–42.
Imbar T, Klipper E, Greenfield C, Hurwitz A, Haimov-Kochman R & Meidan R 2012 Altered endothelin expression in granulosa–lutein cells of women with polycystic ovary syndrome. Life Sciences 91 703–709. (doi:10.1016/j.lfs.2012.06.006)
Jiang YF, Tsui KH, Wang PH, Lin CW, Wang JY, Hsu MC, Chen YC & Chiu CH 2011 Hypoxia regulates cell proliferation and steroidogenesis through protein kinase A signaling in bovine corpus luteum. Animal Reproduction Science 129 152–161. (doi:10.1016/j.anireprosci.2011.12.004)
Kaelin WG 2005 Proline hydroxylation and gene expression. Annual Review of Biochemistry 74 115–128. (doi:10.1146/annurev.biochem.74.082803.133142)
Kaelin WG Jr & Ratcliffe PJ 2008 Oxygen sensing by metazoans: the central role of the HIF hydroxylase pathway. Molecular Cell 30 393–402. (doi:10.1016/j.molcel.2008.04.009)
Keith B, Johnson RS & Simon MC 2012 HIF1α and HIF2α: sibling rivalry in hypoxic tumour growth and progression. Nature Reviews. Cancer 12 9–22.
Kim J, Bagchi IC & Bagchi MK 2009 Signaling by hypoxia-inducible factors is critical for ovulation in mice. Endocrinology 150 3392–3400. (doi:10.1210/en.2008-0948)
Kisliouk T, Levy N, Hurwitz A & Meidan R 2003 Presence and regulation of endocrine gland vascular endothelial growth factor/prokineticin-1 and its receptors in ovarian cells. Journal of Clinical Endocrinology and Metabolism 88 3700–3707. (doi:10.1210/jc.2003-030492)
Kisliouk T, Podlovni H & Meidan R 2005 Unique expression and regulatory mechanisms of EG-VEGF/prokineticin-1 and its receptors in the corpus luteum. Annals of Anatomy 187 529–537. (doi:10.1016/j.aanat.2005.07.005)
Klipper E, Levit A, Mastich Y, Berisha B, Schams D & Meidan R 2010 Induction of endothelin-2 expression by luteinizing hormone and hypoxia: possible role in bovine corpus luteum formation. Endocrinology 151 1914–1922. (doi:10.1210/en.2009-0767)
Ko C, Gieske MC, Al-Alem L, Hahn Y, Su W, Gong MC, Iglarz M & Koo Y 2006 Endothelin-2 in ovarian follicle rupture. Endocrinology 147 1770–1779. (doi:10.1210/en.2005-1228)
Legro RS, Castracane VD & Kauffman RP 2004 Detecting insulin resistance in polycystic ovary syndrome: purposes and pitfalls. Obstetrical & Gynecological Survey 59 141–154. (doi:10.1097/01.OGX.0000109523.25076.E2)
Lie BL, Leung E, Leung PC & Auersperg N 1996 Long-term growth and steroidogenic potential of human granulosa–lutein cells immortalized with SV40 large T antigen. Molecular and Cellular Endocrinology 120 169–176. (doi:10.1016/0303-7207(96)03835-X)
Mamluk R, Wolfenson D & Meidan R 1998 LH receptor mRNA and cytochrome P450 side-chain cleavage expression in bovine theca and granulosa cells luteinized by LH or forskolin. Domestic Animal Endocrinology 15 103–114. (doi:10.1016/S0739-7240(97)00085-4)
Meidan R, Girsh E, Blum O & Aberdam E 1990 In vitro differentiation of bovine theca and granulosa cells into small and large luteal-like cells: morphological and functional characteristics. Biology of Reproduction 43 913–921. (doi:10.1095/biolreprod43.6.913)
Meidan R, Klipper E, Zalman Y & Yalu R 2013 The role of hypoxia-induced genes in ovarian angiogenesis. Reproduction, Fertility, and Development 25 343–350. (doi:10.1071/RD12139)
Na G, Bridges PJ, Koo Y & Ko C 2008 Role of hypoxia in the regulation of periovulatory EDN2 expression in the mouse. Canadian Journal of Physiology and Pharmacology 86 310–319. (doi:10.1139/Y08-025)
Nishimura R & Okuda K 2010 Hypoxia is important for establishing vascularization during corpus luteum formation in cattle. Journal of Reproduction and Development 56 110–116. (doi:10.1262/jrd.09-162E)
Palanisamy GS, Cheon YP, Kim J, Kannan A, Li Q, Sato M, Mantena SR, Sitruk-Ware RL, Bagchi MK & Bagchi IC 2006 A novel pathway involving progesterone receptor, endothelin-2, and endothelin receptor B controls ovulation in mice. Molecular Endocrinology 20 2784–2795. (doi:10.1210/me.2006-0093)
Rayhman O, Klipper E, Muller L, Davidson B, Reich R & Meidan R 2008 Small interfering RNA molecules targeting endothelin-converting enzyme-1 inhibit endothelin-1 synthesis and the invasive phenotype of ovarian carcinoma cells. Cancer Research 68 9265–9273. (doi:10.1158/0008-5472.CAN-08-2093)
Rhee SG 2006 Cell signaling. H2O2, a necessary evil for cell signaling. Science 312 1882–1883. (doi:10.1126/science.1130481)
Sartori-Cintra AR, Mara CS, Argolo DL & Coimbra IB 2012 Regulation of hypoxia-inducible factor-1 alpha (HIF-1α) expression by interleukin-1 beta (IL-1β), insulin-like growth factors I (IGF-I) and II (IGF-II) in human osteoarthritic chondrocytes. Clinics 67 35–40. (doi:10.6061/clinics/2012(01)06)
Schofield CJ & Ratcliffe PJ 2004 Oxygen sensing by HIF hydroxylases. Nature Reviews. Molecular Cell Biology 5 343–354. (doi:10.1038/nrm1366)
Semenza GL 2007 Hypoxia-inducible factor 1 (HIF-1) pathway. Science's STKE 2007 cm8. (doi:10.1126/stke.4072007cm8)
Shih SC & Claffey KP 1998 Hypoxia-mediated regulation of gene expression in mammalian cells. International Journal of Experimental Pathology 79 347–357. (doi:10.1046/j.1365-2613.1998.00088.x)
Shkolnik K, Tadmor A, Ben-Dor S, Nevo N, Galiani D & Dekel N 2011 Reactive oxygen species are indispensable in ovulation. PNAS 108 1462–1467. (doi:10.1073/pnas.1017213108)
Tesone M, Stouffer RL, Borman SM, Hennebold JD & Molskness TA 2005 Vascular endothelial growth factor (VEGF) production by the monkey corpus luteum during the menstrual cycle: isoform-selective messenger RNA expression in vivo and hypoxia-regulated protein secretion in vitro. Biology of Reproduction 73 927–934. (doi:10.1095/biolreprod.105.039875)
Walshe TE & D'Amore PA 2008 The role of hypoxia in vascular injury and repair. Annual Review of Pathology 3 615–643. (doi:10.1146/annurev.pathmechdis.3.121806.151501)
Yacobi K, Tsafriri A & Gross A 2007 Luteinizing hormone-induced caspase activation in rat preovulatory follicles is coupled to mitochondrial steroidogenesis. Endocrinology 148 1717–1726. (doi:10.1210/en.2006-1533)
Yu J, Li J, Zhang S, Xu X, Zheng M, Jiang G & Li F 2012 IGF-1 induces hypoxia-inducible factor 1 alpha-mediated GLUT3 expression through PI3K/Akt/mTOR dependent pathways in PC12 cells. Brain Research 1430 18–24. (doi:10.1016/j.brainres.2011.10.046)
Zalman Y, Klipper E, Farberov S, Mondal M, Wee G, Folger JK, Smith GW & Meidan R 2012 Regulation of angiogenesis-related prostaglandin f2α-induced genes in the bovine corpus luteum. Biology of Reproduction 86 92. (doi:10.1095/biolreprod.111.095067)
Zhang J, Zhang Z, Wu Y, Chen L, Luo Q, Chen J, Huang X, Cheng Y & Wang Z 2012 Regulatory effect of hypoxia-inducible factor-1 alpha on hCG-stimulated endothelin-2 expression in granulosa cells from the PMSG-treated rat ovary. Journal of Reproduction and Development 58 678–684. (doi:10.1262/jrd.2012-089)
