NRF2-mediated signaling is a master regulator of transcription factors in bovine granulosa cells under oxidative stress condition

Transcription factors (TFs) are known to be involved in regulating the expression of several classes of genes during folliculogenesis. However, the regulatory role of TFs during oxidative stress (OS) is not fully understood. The current study was aimed to investigate the regulation of the TFs in bovine granulosa cells (bGCs) during exposure to OS induced by H2O2 in vitro. For this, bGCs derived from ovarian follicles were cultured in vitro till their confluency and then treated with H2O2 for 40 min. Twenty-four hours later, cells were subjected to various phenotypic and gene expression analyses for genes related to TFs, endoplasmic reticulum stress, apoptosis, cell proliferation, and differentiation markers. The bGCs exhibited higher reactive oxygen species accumulation, DNA fragmentation, and endoplasmic reticulum stress accompanied by reduction of mitochondrial activity after exposure to OS. In addition, higher lipid accumulation and lower cell proliferation were noticed in H2O2-challenged cells. The mRNA level of TFs including NRF2, E2F1, KLF6, KLF9, FOS, SREBF1, SREBF2, and NOTCH1 was increased in H2O2-treated cells compared with non-treated controls. However, the expression level of KLF4 and its downstream gene, CCNB1, were downregulated in the H2O2-challenged group. Moreover, targeted inhibition of NRF2 using small interference RNA resulted in reduced expression of KLF9, FOS, SREBF2, and NOTCH1 genes, while the expression of KLF4 was upregulated. Taken together, bovine granulosa cells exposed to OS exhibited differential expression of various transcription factors, which are mediated by the NRF2 signaling pathway. Supplementary Information The online version contains supplementary material available at 10.1007/s00441-021-03445-4.


Introduction
In the mammalian follicle, the oocyte is surrounded by multiple layers of somatic cells including granulosa cells, which play a significant role during oocyte maturation by producing paracrine factors including estradiol and progesterone (Wen et al. 2010). Thus, disruption of granulosa cell functions may result in poor oocyte quality and subsequently reduction in pregnancy rate (Lai et al. 2018). Previous studies have evidenced that exposure of granulosa cells to oxidative stress has led to impaired function and subsequently poor quantity and quality of oocytes (Jancar et al. 2007;Lai et al. 2018;Rajani et al. 2012). Despite the fact that accumulation of reactive oxygen species (ROS)-induced oxidative stress harms granulosa cell function, elevated levels of ROS at the time of ovulation play an indispensable role in facilitating the ovulation process (Tanabe et al. 2015).
ROS are endogenously produced during metabolic pathways as a byproduct from different cellular compartments including the plasma membrane, mitochondria, endoplasmic reticulum, and peroxisomes (Di Meo et al. 2016). Cells are equipped with a well-balanced system including antioxidant machinery to eliminate the excessive levels of ROS and maintain cellular redox homeostasis (He et al. 2017).
However, disruption of this balanced system due to overwhelmed cellular ROS production leads to oxidative stress (Agarwal et al. 2006). Excessive ROS accumulation and subsequent oxidative stress could also arise from external and/or internal stressors such as heat stress (Alemu et al. 2018), inflammation (Mittal et al. 2014), and endoplasmic reticulum stress (Liu et al. 2018). Exposure of cells to oxidative stress has a negative impact on cell membrane permeability and mitochondrial activity and causes DNA damage (Kadenbach et al. 2004;Zhang et al. 2016) resulting in cell cycle arrest (Saeed-Zidane et al. 2017) and apoptosis (Sohel et al. 2019).
Mammalian cells respond to intra and extracellular signals by altering several transcription factors (TFs) that activate and/or inhibit genes involved in various biological functions (Leite et al. 2017). Some of those TFs are believed to be involved in cellular defense mechanisms against oxidative stress (Mikaeili et al. 2016;Weng et al. 2016). In bovine embryos, the capability of in vitro produced embryos to survive under oxidative stress conditions relies on the activation of nuclear factor (erythroid-derived 2)-like 2 (NRF2 or NFE2L2) transcription factor at mRNA and protein levels (Amin et al. 2014). Similarly, NRF2 is responsible for the activation of cellular antioxidant machinery in bovine granulosa cells under oxidative stress conditions (Alemu et al. 2018;Khadrawy et al. 2019;Sohel et al. 2019Sohel et al. , 2018. Moreover, NRF2 could be extracellularly released via exosomes and horizontally transferred to neighboring cells as a means of inducing adaptive response against oxidative stress in recipient cells (Saeed-Zidane et al. 2017).
NRF2 might particularly be involved in the regulation of other TF networks such as NOTCH (Zhao et al. 2016), Krüppel-like factor (KLFs) family members (Jang et al. 2014;Zucker et al. 2014), and sterol regulatory elementbinding transcription factor (SREBFs) family (Amin et al. 2014;Kamisako et al. 2014), which are involved in cell proliferation, differentiation, apoptosis (Andersson et al. 2011;Miao et al. 2017), fatty acids, and cholesterol biosynthesis (Daemen et al. 2013;Shimano 2001). Some of those TFs significantly contribute to the regulation of ovarian functions (Murta et al. 2015;Natesampillai et al. 2008) such as regulation of genes involved in steroidogenesis (Liu et al. 2009). Furthermore, the overexpression of the KLF family showed a negative effect on the expression of CYP11A1 in granulosa cells (Natesampillai et al. 2008). However, increasing intracellular ROS accumulation accompanied by overexpression of NRF2 resulted in promoting cell death via regulation of KLF9 (Zucker et al. 2014). Moreover, apoptosis of granulosa cells could be induced via regulation of FOS after exposure to mycotoxin (Guerrero-Netro et al. 2015), which is one of the activator protein-1 (AP-1) members (Dias et al. 2013). However, little is known regarding the regulation of granulosa cell functions via various TFs under oxidative stress conditions and their relevance with the NRF2 signaling pathway. Therefore, the current study aimed to investigate the expression patterns of stress-induced TFs and their potential regulation by the NRF2 signaling pathway in bovine granulosa cells exposed to oxidative stress.

Experimental setup
To investigate the effect of H 2 O 2 on the expression of the stress-related TFs and their association with the NRF2 signaling pathway in bovine granulosa cells, two independent experiments were conducted. In the first experiment, subconfluent bovine granulosa cells were treated with 5 µM H 2 O 2 for 40 min as optimized previously (Saeed-Zidane et al. 2017) to induce moderate oxidative stress. Moreover, based on previous pieces of evidence and in silico analysis, the second experiment was conducted to figure out the cross-talk between NRF2 and other candidate TFs. For that, sub-confluent bovine granulosa cells were transfected with 200 nM small interference RNA (siRNA) targeting NRF2.
In both experiments, the cells were incubated at 37 °C in 5% CO 2 for another 24 h after the treatment and then subjected to various phenotypic and genotypic analyses.

In vitro culture of bovine granulosa cells
Bovine ovaries were obtained from a local abattoir and transported in warm (37 °C) physiological saline (0.9% NaCl) solution. Upon arrival, ovaries were washed three times with warm (37 °C) 0.9% NaCl, followed by prewarmed (37 °C) 70% ethanol washing. Granulosa cells were aspirated from small follicles (3-5 mm) in 15-mL tubes containing prewarmed (37 °C) phosphate-buffered saline, calcium-and magnesium-free (PBS-CMF) and then kept standing for 15 min at 37 °C to let the cellular debris settled down at the bottom. Subsequently, the supernatants containing the granulosa cells were collected and centrifuged at 750g for 7 min to get granulosa cell pellets. Afterwards, the cell viability and concentration were checked under the microscope using trypan blue staining. Ultimately, a total of 2.5 × 10 5 cells per well was cultured in CytoOne® 24-well plate (Starlab GmbH, Hamburg, Germany) containing 600-µL DMEM/F-12 medium (Sigma-Aldrich Chemie GmbH, Taufkirchen, Germany) supplemented with 10% fetal bovine serum (FBS) (Gibco®, Karlsruhe, Germany) and 1% penicillin/streptomycin and incubated at 37 °C in 5% CO 2 .

Small interference RNA transfection
Briefly, sub-confluent granulosa cells were transfected with 200 nM siRNA targeting NRF2 (siRNA-NRF2) or scrambled siRNA as a negative control (siRNA-ve) (Exiqon, Vedbaek, Denmark) in the presence of Lipofectamine® 2000 (Invitrogen, Carlsbad, California, USA) as transfection reagent diluted in Opti-MEM I reduced-serum medium (Invitrogen, Carlsbad, California, USA). Afterwards, cells were incubated for 24 h in a humidified 5% CO 2 incubator at 37 °C. Thereafter, cells were harvested and used for further analyses.

Cell proliferation assay
A total of 1.5 × 10 4 cells were cultured in 96-well plates using the same procedure described above. Cells from both experiments were used for proliferation assay using CCK-8 kits (Dojindo EU GmbH, München, Germany) after 24 h post-treatment either with H 2 O 2 or siRNA according to the manufacturer's protocol. The proliferation rate was indicated by the absorbance at 450 nm wavelength using Synergy™ H1 Multi-Mode Reader (BioTek Germany, Bad Friedrichshall, Germany).

Intracellular ROS detection
Granulosa cells were cultured in a 96-well plate and then the sub-confluent cells were treated with H 2 O 2 . After 24 h, cells were co-incubated with FBS-free medium containing 75 µM 2′,7′-dichlorodihydrofluorescein diacetate (H2DCFDA) (Invitrogen, Carlsbad, California, USA) for 20 min at 37 °C in 5% CO 2 . Subsequently, the cells were washed twice with PBS-CMF, and then, the images were captured using a green-fluorescence filter of the inverted fluorescence microscope (Leica DM IRB, Leica, Wetzlar, Germany). The fluorescence intensity was quantified using ImageJ 1.48v software (National Institutes of Health, Maryland, USA, http:// imagej. nih. gov).

Mitochondrial activity assay
To assess the cellular mitochondrial activity, a total of 4 × 10 4 live granulosa cells were seeded into 8-well slide chamber according to the manufacturer's protocol. Briefly, the granulosa cells were co-incubated with 200 nM MitoTracker red dye (MitoTracker® Red CMXRos, M7512; Invitrogen) for 30 min in a CO 2 incubator. Then, the cells were rinsed twice with PBS-CMF and subsequently fixed with 4% paraformaldehyde overnight at 4 °C. Fixed cells were mounted with a Vectashield mounting medium containing DAPI. Finally, the images were acquired using confocal microscopy CLSM LSM-780 and the fluorescent intensity was analyzed with ImageJ 1.48v software (National Institutes of Health, Maryland, USA, http:// imagej. nih. gov).

Lipid accumulation assay
Cells were cultured in a 96-well plate and treated with either H 2 O 2 or siRNA. Twenty-four hours post-treatment, plates containing granulosa cells were gently rinsed with 50 µL of PBS-CMF. Subsequently, cells were fixed using 4% paraformaldehyde overnight at 4 °C and then subjected to lipid accumulation assay. Briefly, a working solution was prepared by mixing three parts (30 mL) of oil red O (Sigma-Aldrich Chemie GmbH, Munich, Germany) from stock solution with 2 parts (20 mL) deionized water and then stored at room temperature for 10 min followed by filtration via filter paper. After removing the 4% paraformaldehyde, each well was gently rinsed by sterile water followed by co-incubation with 50 µL of 60% isopropanol at room temperature for 5 min. Afterwards, the isopropanol was poured off and 100-µL working solution of oil red O was pipetted to each well and left for 5 min. Following this, cells were rinsed with tap water and the images were captured using an inverted microscope with 40× magnification.

DNA fragmentation assay
The fragmented DNA was isolated according to the Abcam protocol (Cambridge, UK). Briefly, the floated and adhered cells were collected and then lysed in 500 µL of 10 mM Tris (pH 7.4), 5 mM EDTA, 0.2% Triton-X100 followed by vortexing for several times. The lysate was incubated on ice for 30 min followed by centrifugation at 27,000g for 30 min. Subsequently, the supernatants were divided into two parts and 50-µL ice-cold 5 M NaCl was added to each part coupled with several times vortexing. Subsequently, a mixture containing 600 µL ethanol and 150-µL 3 M sodium acetate (pH 5.2) was added to each aliquot with gentle mixing by pipetting and then incubated at − 80 °C for 1 h followed by centrifugation for 20 min at 20,000g. After discarding the supernatant, the pellets were pooled together and redissolved in 10 mM Tris and 5 mM EDTA accompanied by adding DNase-free RNase 10 mg/mL and then incubated in a water bath at 37 °C for 5 h. Afterwards, a combination of 25 µL proteinase K at 20 mg/mL (Qiagen GmbH, Hilden, Germany) and 40 µL buffer (100 mM Tris pH 8.0, 100 mM EDTA, and 250 mM NaCl) was added and subsequently incubated overnight at 65 °C. Thereafter, the DNA was extracted using phenol/chloroform/isoamyl alcohol mixture (25:24:1) (Biomol GmbH, Hamburg, Germany) and then precipitated using ethanol. The supernatant was carefully removed followed by air drying. The pellets were resuspended in 15-µL 1× TE buffer (Carl Roth GmbH, Karlsruhe, Germany). Finally, the samples were electrophoretically separated using a 2% agarose gel containing ethidium bromide and then visualized using ChemiDoc™ XRS+ system (Bio-Rad Laboratories GmbH, Germany).

RNA isolation and cDNA synthesis
Total RNA from granulosa cells was extracted using the miRNeasy® mini kit (Qiagen GmbH, Hilden, Germany) following the manufacturer's protocol. Before starting the total RNA extraction, granulosa cells were washed using 1× PBS to remove the remaining culture medium. At the end of the extraction protocol, trapped RNAs in the membrane of spin column were eluted using 30 μL RNase-free water. RNA concentration was assessed using NanoDrop 8000 UV-Vis Spectrophotometers (Thermo Scientific, Wilmington, USA). Equal amounts of total RNA were reverse transcribed using the RevertAid first-strand cDNA synthesis kit (Thermo Fisher Scientific, Schwerte, Germany). Briefly, a combination from oligo-dT and random primer was added to each sample followed by incubation at 65 °C for 5 min. After incubation time, a mixture containing 1 µL of RiboLock RNase inhibitor, 2 µL of 10 mM dNTP mix, 4 µL of 5× reaction buffer, and 2 µL of MMuLV reverse transcripts was added to each sample. Afterwards, samples were incubated at 25 °C for 5 min, then 37 °C for 60 min followed by 70 °C for 5 min to terminate the reaction. Ultimately, the cDNA was kept at − 20 °C for further analysis.

Immunoblotting
Total protein from granulosa cells was isolated using 1× passive lysis buffer (Promega GmbH, Mannheim, Germany) and the concentration was determined using Coomassie Protein Assay Reagent (Life Technologies GmbH, Darmstadt, Germany). Thirty-five micrograms of protein was loaded onto 12% SDS-polyacrylamide gel. After electrophoresis, proteins were transferred onto a nitrocellulose membrane (Bio-Rad Laboratories GmbH, Germany) using Trans-Blot® SD Semi-Dry Transfer Cell (Bio-Rad Laboratories GmbH, Germany). Thereafter, the membrane was blocked in Roti®-Block (Carl Roth GmbH, Germany) for 1 h at room temperature. The membrane was incubated overnight at 4 °C with polyclonal primary antibodies specific for KLF4 (1:100), Grp78 (1:300), PRKAA 1/2 (1:350), StAR (1:250), and ACTB (1:500) mouse monoclonal antibody (Santa Cruz Biotechnology Inc, Germany). After the incubation with the primary antibody, the membrane was incubated with goat anti-rabbit or goat antimouse (1:5000; Santa Cruz Biotechnology Inc, Germany) secondary antibody with the respective candidate primary antibody for 1 h at room temperature. The proteins were visualized using Clarity™ Western ECL Substrate (Bio-Rad Laboratories Inc, USA). The images were captured using the ChemiDoc™ XRS+ system (Bio-Rad Laboratories GmbH, Germany). Subsequently, the membrane was stripped using a middle stripping buffer in order to remove primary and secondary antibodies followed by blocking and incubation with another antibody using the same protocol. As shown in supplementary Fig. 3, in both control and H 2 O 2 -treated cells, the StAR protein was detected at 30 kDa.

Immunocytochemistry
Localization and subsequent quantification of NRF2 and KLF4 proteins were conducted using the immunocytochemistry technique. For this, granulosa cells were seeded into 8-well slide chamber and subsequently treated with either H 2 O 2 or siRNA. Twenty-four hours post-treatment, granulosa cells were fixed using 4% paraformaldehyde and kept at 4 °C overnight. Subsequently, the cells were rinsed three times with PBS-CMF and then incubated with 0.3% Triton X-100 (Roche Diagnostics , mitochondrial activity c-f, the relative mRNA expression profile of endoplasmic reticulum stress marker genes g, and protein level of Grp78 f of bovine granulosa cells under oxidative stress conditions. Green color a, a' indicates the fluorescence of the 2′,7′-dichlorodihydrofluorescein diacetate (H2DCFDA). The blue color c, c' indicates the nuclear staining using 40,6-diamidino-2-phenylindole (DAPI) and the red color d, d' indicates mitochondrial activity. The fluorescence and band intensity were measured by ImageJ software. Data were statistically analyzed with Student's two-tailed t test and are shown as means ± SEM of three independent replicates with different pools of cells (*p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001). Scale bars: 100 µm a, a', 20 µm c-e' GmbH, Mannheim, Germany) for 10 min followed by blocking with 3% donkey serum (Sigma-Aldrich Chemie GmbH, Munich, Germany) for 1 h at room temperature. Then, the cells were co-incubated with the primary antibody for NRF2 (1:100) or KLF4 (1:100) at 4 °C overnight. Thereafter, the cells were washed three times with PBS-CMF followed by co-incubation with fluorescent secondary antibody (Lifespan Biosciences, Alexa Fluor goat anti-rabbit 1:350) for 3 h at 37 °C in dark. Thereafter, the cells were washed three times with PBS-CMF and subsequently mounted with Vectashield mounting medium containing DAPI. Finally, the cells were visualized under confocal microscopy CLSM LSM-780 and the fluorescence intensity of proteins was analyzed using ImageJ 1.48v software (National Institutes of Health, Maryland, USA, http:// imagej. nih. gov).

Statistical analysis
Data were represented as means ± SEM from at least three independent replicates with different pools of cells. Data were statistically analyzed with Student's two-tailed t test using GraphPad Prism software version 7 (GraphPad Software, La Jolla, California, USA). The difference between means was considered significant when P value ≤ 0.05.

Oxidative stress impaired granulosa cell functions
Bovine granulosa cells treated with H 2 O 2 exhibited shrinkage in shape and decreased confluency compared with untreated counterparts (Suppl. Figure 1). Moreover, Fig. 2 Relative abundance of the transcription factors in bovine granulosa cells exposed to oxidative stress. The expression pattern of NRF2 mRNA a and its protein b-e in response to oxidative stress, whereby the blue color b, b' indicates the nuclear staining using 40,6-diamidino-2-phenylindole (DAPI) and the red color c, c' indicates the NRF2 protein; the effect of oxidative stress on candidate transcription factors, which showed up f and downregulation g. The transcript expression data were normalized to internal ACTB and GAPDH as well as relatively compared with the control group. Data were statistically analyzed using Student's two-tailed t test and are represented as means ± SEM of triplicates (*p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001). The ImageJ software was utilized to calculate the fluorescence intensity of NRF2 protein. Scale bars: 20 µm b-d' exposure of cells to H 2 O 2 elevated (P ≤ 0.01) the intracellular ROS amount and decreased (P ≤ 0.001) mitochondrial activity compared with the untreated controls ( Fig. 1a-f). Moreover, the higher intracellular ROS accumulation in H 2 O 2 -challenged cells was accompanied by the induction of DNA fragmentation (Suppl. Figure 2). Overall, these results indicated that bovine granulosa cells treated with H 2 O 2 resulted in the induction of oxidative stress phenotypes.

H 2 O 2 -induced oxidative stress coupled with endoplasmic reticulum stress
The mRNA expression level of glucose-regulated protein (Grp)78 and Grp94 was significantly (P ≤ 0.01 and P ≤ 0.05, respectively) higher in H 2 O 2 -treated cells compared with the controls (Fig. 1g). Furthermore, the protein level of Grp78 was in agreement with the corresponding mRNA expression level (Fig. 1h). However, the expression level of Calpain1 did not show any significant differences.

H 2 O 2 -induced oxidative stress altered the expression pattern of candidate TFs
The temporal expression level of the candidate TFs which have pivotal roles in stress response (NRF2, FOS, and FOSB), apoptosis, differentiation, and proliferation (KLF4, KLF6, KLF9, and NOTCH1), cell cycle (E2F1 and E2F4), fatty acids and cholesterol biosynthesis (SREBF1 and SREBF2), and immune response (IRF7) was investigated. Upon oxidative stress induction, the expression pattern of NRF2 was significantly increased (P ≤ 0.01, Fig. 2a). Similarly, the mRNA transcript level of NOTCH1, FOS, SREBF1, SREBF2, KLF6, KLF9, and E2F1 was significantly higher (P ≤ 0.05) in the H 2 O 2 -treated group compared with the untreated controls (Fig. 2f). Contrary to this, the expression pattern of KLF4 and IRF7 was significantly lower (P ≤ 0.05) in H 2 O 2 -treated granulosa cells compared with the untreated control group (Fig. 2g). However, no significant differences were observed in the expression of E2F4 and FOSB (Fig. 2g).

Oxidative stress influenced the protein level of NRF2 and KLF4
The protein level of NRF2 and KLF4 was assessed using immunocytochemistry assay after exposure of granulosa cells to oxidative stress. Similar to mRNA expression, the protein abundance of NRF2 was significantly upregulated (P ≤ 0.001) in H 2 O 2 -challenged cells compared with controls (Fig. 2c, c', and e). Furthermore, the NRF2 protein was found to be localized in the nucleus of treated cells. However, the protein abundance of KLF4 was significantly lower (P ≤ 0.001) in the H 2 O 2 -challenged cells (Fig. 3a-e).

Alterations of downstream genes of KLF4 under oxidative stress condition
The downregulation of KLF4 in the H 2 O 2 -treated group was accompanied by a significant reduction (P ≤ 0.01) of CCNB1 (Fig. 3f). However, the mRNA expression of CCND2 and BAX was not significantly different between H 2 O 2 -treated and untreated groups (Fig. 3f).

Oxidative stress induced lipid accumulation
In addition to KLF4, TFs involved in fatty acids and cholesterol biosynthesis (SREBF1 and SREBF2) and cellular lipid accumulation were investigated under oxidative stress condition. Results showed a significant increment (P ≤ 0.05) in mRNA expression of PRKAA1 (Fig. 4a) as an upstream gene of SREBF1 and SREBF2. Additionally, the PRKAA1/2 protein level (Fig. 4b) and lipid accumulation (Fig. 4c and c') were higher in the H 2 O 2 -challenged cells compared with control ones.

Oxidative stress-induced granulosa cell differentiation marker genes
To clarify whether the exposure of bovine granulosa cells to oxidative stress affected their ability to differentiate, the expression pattern of differentiation marker genes was investigated. The mRNA expression of CYP11A1, CYP19A1, and STAR was significantly upregulated (P ≤ 0.001, P ≤ 0.01, and P ≤ 0.01, respectively) upon exposure of granulosa cells to H 2 O 2 (Fig. 5a). Additionally, the protein abundance of StAR showed upregulation in challenged cells (Fig. 5b). Moreover, the expression pattern of EGFR was significantly upregulated (P ≤ 0.01) in the H 2 O 2 -treated group compared with untreated controls. However, no significant differences were noticed in the mRNA expression level of Inhibin Beta A (INHBa) and Forkhead box O1 (FOXO1) between the treatment groups (Fig. 5a).

Oxidative stress induced the expression of the pro-apoptotic gene and subsequently impaired granulosa cell proliferation
Unlike BAX, the mRNA level of CASP3, a pro-apoptotic gene, was significantly increased (P ≤ 0.05) in H 2 O 2 -challenged cells (Fig. 5c). However, the expression of the anti-apoptotic marker gene (BCL2L1) was significantly decreased (P ≤ 0.05) under the H 2 O 2 challenge. On the other hand, the mRNA and protein expression level of the proliferation marker gene (PCNA) was significantly decreased (P ≤ 0.01) under oxidative stress condition (Fig. 5 d and f). Accordingly, a reduction (P ≤ 0.001) in cell proliferation rate was observed upon exposure of granulosa cells to H 2 O 2 challenge (Fig. 5e).

Fig. 4
The effect of oxidative stress on mRNA expression of PRKAA1 and PRKAA2 a, the protein level of PRKAA1/2 b, and lipid accumulation c, c' in bovine granulosa cells. The red color c, c' indicates the oil red O stain. Data were statistically analyzed using Student's two-tailed t test and represented as means ± SEM of triplicates (*p ≤ 0.05). Scale bars: 25 µm c, c'

Selective knockdown of NRF2 altered the expression pattern of stress-related candidate TFs
The granulosa cells transfected with siRNA targeting NRF2 exhibited reduced expression of NRF2 (P ≤ 0.001; Fig. 6). Upon NRF2 knockdown, the mRNA expression of KLF9, FOS, NOTCH1, and SREBF2 was significantly downregulated (P ≤ 0.05; Fig. 6f). Interestingly, the relative mRNA expression of KLF4 was significantly increased (P ≤ 0.001) by approximately fivefold. However, the mRNA expression levels of E2F1 and SREBF1 were not significantly altered.

The consequence of NRF2 knockdown on the expression of KLF4 and its downstream genes and cell proliferation
Targeted suppression of NRF2 resulted in a significant increase in the protein level of KLF4 (Fig. 7a-d). Next, to demonstrate the effect of NRF2 knockdown and induced expression of KLF4 on granulosa cell functions, the expression pattern of KLF4 downstream genes and cell proliferation markers was investigated. Consequently, the mRNA expression of CCND2 was upregulated (P ≤ 0.05; Fig. 7e). Intriguingly, the expression pattern of BAX was upregulated (P ≤ 0.01) accompanied by a significant reduction (P ≤ 0.01) in cell proliferation rate (Fig. 7f).

Fig. 5
The mRNA expression pattern of differentiation-related genes (a), StAR protein (b), expression pattern of anti-apoptotic, pro-apoptotic c, proliferation marker d genes, cell viability e, and PCNA protein level f in bovine granulosa cells exposed to H 2 O 2 -induced oxidative stress. The geometric mean of ACTB and GAPDH was used for mRNA normalization and the data were relatively compared with the control group. The band intensity of immunoblotting f was measured by ImageJ software. Data were statistically analyzed with Student's two-tailed t test and are represented as means ± SEM of triplicates (*p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001). The level of ACTB protein was used as an internal control in protein analysis

NRF2 knockdown altered the expression pattern of genes associated with differentiation and lipid accumulation
The mRNA expression level of the CYP11A1 transcript was significantly reduced (P ≤ 0.05) in siRNA-NRF2transfected cells (Fig. 8a) and the mRNA level of INHBa was significantly increased (P ≤ 0.01). On the other hand, the NRF2 knockdown resulted in dysregulation of fatty acids and cholesterol biosynthesis TFs, which was confirmed by lower lipid accumulation (Fig. 8 b and b').

Discussion
In the current study, granulosa cells were exposed to moderate oxidative stress via 5 µM H 2 O 2 treatment as we have used previously (Saeed-Zidane et al. 2017). The H 2 O 2 , one of the non-radical groups of ROS, is diffused through the cell membrane (Tamma et al. 2018) and subsequently increased the accumulation of intracellular ROS level (Sohel et al. 2019;Zhang et al. 2016). This strategy is orchestrated during the ovulation process by leukocytes, which are deemed to be the source of ROS (Sugino 2005).
Upon increasing intracellular ROS accumulation, the free radicals attack the cellular bioactive molecules such as DNA, lipids, and proteins (Guo et al. 2013;Kadenbach et al. 2004;Zhang et al. 2016), as has been found in the current investigation (Supp. Figure 2). Moreover, the compromising of mitochondrial DNA could cause mutations and subsequently mitochondrial dysfunction (Hollensworth et al. 2000). The permeability of the mitochondrial membrane is prone to lipid peroxidation resulted from the higher levels of ROS, which ultimately impairs the functions of mitochondria and decreases its activity (Guo et al. 2013;Kirkinezos et al. 2005), as shown in the present study (Fig. 1d, d', and f). This could lead to disturbing the cellular metabolism (Kadenbach et al. 2004), increase apoptosis (Yang et al. 1997), and decrease cell proliferation as well as alter the cell cycle (Antico Arciuch et al. 2012;Saeed-Zidane et al. 2017). Besides that, H 2 O 2 -induced oxidative stress contributes to altering the protein folding mechanism and induction of endoplasmic reticulum stress (Kim et al. 2016;Malhotra et al. 2008). The endoplasmic reticulum stress has a dual role in ovarian functions either during atresia (Lin et al. 2012) or maturation of the follicle and corpus luteum formation (Huang et al. 2017), which was confirmed in our study by the upregulation of endoplasmic reticulum marker genes (Grp78 and Grp94) mRNA and the protein level of Grp78 ( Fig. 1g and h). Both Grp78 and Grp94 contribute to the maintenance of endoplasmic reticulum homeostasis and induction of apoptosis via activation of caspase under endoplasmic reticulum stress condition (Liu et al. 2011).
Besides their roles in defense mechanisms in a celldependent manner, several TFs have been proven to play a pivotal role in controlling ovarian functions (Sirotkin 2010). For instance, highly abundant NRF2 protein was found to be localized in the porcine follicular cells of the preovulatory follicle and newly formed corpus luteum (Likszo et al. 2019) to trigger the release of antioxidant machinery and subsequently protect the cells from apoptosis (Khadrawy et al. 2019;Likszo et al. 2019;Saeed-Zidane et al. 2017;Sohel et al. 2019). Similarly, we found upregulation of NRF2 expression in granulosa cells exposed to oxidative stress (Fig. 2), which is the master of TFs involved in oxidative stress response. Furthermore, KLF9-induced apoptosis has been found to be regulated by oxidative stress in cancer cells via NRF2 (Zucker et al. 2014), as has been found in the current study (Fig. 2). Therefore, we speculated the potential role of NRF2 in the regulation of other TFs under oxidative stress conditions and subsequently alteration of ovarian functions.
In ovarian cells, the expression of KLF4 and KLF13 mRNA is regulated by luteinizing hormone (LH) or/and insulinlike growth factor 1 (IGF1) in follicular granulosa cells (Natesampillai et al. 2008), while the dysregulation of KLF2 and KLF4 expression was noticed in polycystic ovary syndrome (PCOS) compared with the normal ovaries (Jansen et al. 2004). In the present study, the mRNA expression of KLF4 was significantly decreased under oxidative stress conditions (Fig. 2g), which was subsequently confirmed by the reduction of its protein level (Fig. 3b, b', and d) and its CCND2 target gene (Fig. 3f). CCND2, which is known to be involved in the regulation of apoptosis (Li et al. 2015) and cell cycle , is found to be associated with the reduction in cell proliferation rate (Fig. 5e). Furthermore, our previous study showed a reduction in the expression of CCND2 and subsequently cell cycle arrest at the G 2 /M phase in H 2 O 2 -treated granulosa cells (Saeed-Zidane et al. 2017). The upregulation of KLF4 was found to interfere with the expression of CYP19A1 , which is responsible for estrogen biosynthesis. However, the reciprocal expression pattern of KLF4 and NRF2 was found during NRF2-knockdown (Fig. 6) resulting in the induction of BAX (Fig. 7e), which is responsible for stimulating the apoptosis in granulosa cells (Yang et al. 2017). This indicates the novel protective role of the NRF2 signaling pathway by inhibiting KLF4-induced apoptosis (Li et al. 2015) under oxidative stress conditions. The reduction of KLF4 expression could be indirectly regulated by the higher expression of SREBF2, as has been shown in endothelial cells exposed to oxidative stress . SREBF1 and SREBF2 are TFs involved in the regulation of fatty acids and cholesterol biosynthesis (Shimano 2001). In the preovulatory follicle, the reduction of cholesterol as a consequence of LH-induced steroidogenesis leads to the stimulation of genes coding for cholesterol biosynthesis (Likszo et al. 2019). In agreement with that, our results showed that SREBF1 and SREBF2 were found to be upregulated in granulosa cells exposed to oxidative stress conditions (Fig. 2f), which was coupled with a reduction of KLF4 expression and induction of genes responsible for steroidogenic enzymes (Fig. 5). On the other hand, the elevated ROS-induced endoplasmic reticulum stress (Fig. 1) was found to be responsible for increasing the SREBF2-cleavage (Colgan et al. 2007), which is in agreement with our results. Accordingly, the induction of those TFs could lead to increasing lipid accumulation and subsequently promoting of granulosa cell apoptosis and differentiation (Regan et al. 2018), as evidenced by the higher expression of differentiation marker genes (Fig. 5a) and apoptosis (Fig. 5c), as has been previously reported in granulosa cells treated with FSH (Liu et al. 2009). These facts indicate the significant contribution of ROS in granulosa cell transdifferentiation (Shkolnik et al. 2011). Additionally, the dysregulation of SREBFs' levels resulted in reducing lipid accumulation upon NRF2 knockdown, which indicates the potential role of NRF2 signaling in cellular metabolism through regulation of SREBFs' family (Kamisako et al. 2014). This led to a reduction of CYP11A1 as a differentiation marker gene in granulosa cells (Fig. 8a), which demonstrated the vital role of NRF2 in harmonic with other TFs in the regulation of granulosa cell differentiation.
Moreover, the E2F1, one of E2Fs family involved in cell cycle regulation and apoptosis (Crosby and Almasan 2004), was found to be increased in H 2 O 2 -treated granulosa cells (Fig. 2f). Consistently, the upregulation of E2F1 was accompanied by increasing the apoptosis in various cell types (Ginsberg 2002), which is evidenced by increasing DNA fragmentation (Suppl. Figure 2) and decreasing of cell proliferation rate (Fig. 5e) in the present study. Interestingly, as has been evidenced in the current study, there are E2F-binding sites in the NOTCH promotor resulting in the upregulation of NOTCHs' family expression (Viatour et al. 2011).
There is clear evidence that the NOTCH signaling pathway is implicated in ovarian follicular development, through the spatial and temporal regulation of its associated genes, during follicular growth and selection (Murta et al. 2015;Vanorny et al. 2014;Zhang et al. 2016) as well as corpus luteum steroidogenesis and regression (Murta et al. 2015). Additionally, the stimulation of NOTCH signaling acts as a suppressor of apoptosis signal-regulating kinase 1 (ASK1) 1 and consequently p38 The expression level of ACTB and GAPDH was used to normalize mRNA transcript and the data were relatively compared with the negative control group (siRNA-ve). Data were analyzed with Student's two-tailed t test and represented as means ± SEM of triplicates (*p ≤ 0.05, **p ≤ 0.01). Scale bars: 25 µm b, b' MAPK signaling pathway (Mo et al. 2013). Accordingly, the upregulation of the NOTCH1 transcript in the H 2 O 2 -treated group (Fig. 2f) could be a defense mechanism to suppress the elevation of apoptosis under oxidative stress conditions. This is in agreement with the fact that the NOTCH signaling pathway is involved in the regulation of granulosa cell functions during follicle development (Jing et al. 2017). This novel results in bovine granulosa cells open a new window for further investigations related to TF cross-talk during oxidative stress and their association with apoptosis and differentiation.
Collectively, granulosa cells exposed to oxidative stress showed higher intracellular ROS accumulation accompanied by increased DNA fragmentation and reduced mitochondrial activity. The challenged cells activated the NRF2 signaling pathway in parallel with other TFs involved in apoptosis and differentiation (Suppl. Figure 4), which was remarkably evidenced by the reduction of cell proliferation and induction of steroidogenic genes and lipid accumulation. Those impacts were reversed by NRF2 knockdown. Therefore, the NRF2 signaling pathway in cooperation with other stress-related TFs plays a vital role in the oxidative stress response through regulation of bovine granulosa cells apoptosis and differentiation, which showed its implication for follicular development. Funding Open Access funding enabled and organized by Projekt DEAL. Mohamed Omar Taqi was financially supported by the German Academic Exchange Service (DAAD).

Declarations
Ethical approval This article does not contain any studies with animals performed by any of the authors.

Conflict of interest
The authors declare that they have no conflict of interest.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.