Oleic acid induces speciﬁc alterations in the morphology, gene expression and steroid hormone production of cultured bovine granulosa cells General and Comparative Endocrinology

After parturition, one of the major problems related to nutritional management that is faced by the majority of dairy cows is negative energy balance (NEB). During NEB, excessive lipid mobilization takes place and hence the levels of free fatty acids, among them oleic acid, increase in the blood, but also in the follicular ﬂuid. This accumulation can be associated with serious metabolic and reproductive disorders. In the present study, we analyzed the effects of physiological concentrations of oleic acid on cell morphol- ogy, apoptosis, necrosis, proliferation and steroid production, and on the abundance of selected transcripts in cultured bovine granulosa cells. Increasing oleic acid concentrations induced intracellular lipid droplet accumulation, thus resulting in a foam cell-like morphology, but had no effects on apoptosis, necrosis or proliferation. Oleic acid also signiﬁcantly reduced the transcript abundance of the gonadotro- pin hormone receptors, FSHR and LHCGR , steroidogenic genes STAR , CYP11A1 , HSD3B1 and CYP19A1 , the cell cycle regulator CCND2 , but not of the proliferation marker PCNA . In addition, treatment increased the transcript levels of the fatty acid transporters CD36 and SLC27A1 , and decreased the production of 17-beta-estradiol and progesterone. From these data it can be concluded that oleic acid speciﬁcally affects morphological and physiological features and gene expression levels thus altering the functionality of granulosa cells. Suggestively, these effects might be partly due to the reduced expression of FSHR and thus the reduced responsiveness to FSH

levels, the levels of oleic acid were nearly the same (200-300 lM, Aardema et al., 2013). Oleic acid promotes apoptosis and necrosis of human lymphocytes through the activation of caspase-3 (Cury-Boaventura et al., 2006). At high concentration oleic acid inhibits the proliferation of scar and normal fibroblasts and causes an excessive and continued inflammatory reaction by promoting the secretion of pro-inflammatory mediators (Jiang et al., 2012). In lymph-node carcinoma cells of the prostate (LNCaP), oleic acid significantly inhibits proliferation (Liu et al., 2009).
In a previous report it was shown that oleic acids reduced proliferation, but stimulated 17-beta-estradiol secretion in bovine granulosa cells . However, there are no data available on the effects of oleic acid on gene expression in bovine granulosa cells. The aim of the present study was to analyzed effects of elevated oleic acid concentrations as observed in the follicular fluid during negative energy balance or starvation on granulosa cell function on the molecular level. Therefore, during this work, we comprehensively studied the impact of oleic acid on cell morphology, cell cycle, proliferation, steroid hormone secretion, and on the expression of key genes involved in the function of granulosa cells in a serum-free, estrogen-producing granulosa cell long term culture model.

Materials
Unless otherwise indicated, all chemicals used were purchased from Sigma Aldrich (Steinheim, Germany).

Granulosa cells cultures
Granulosa cells were isolated and cultured as previously described, (Baufeld and Vanselow, 2013;Gutierrez et al., 1997;Hamel et al., 2005). Briefly, granulosa cell preparations (n = 3) were cell pools derived from 30 to 40 ovaries and 10-15 small to medium sized follicles (2-6 mm diameter), each. Only follicles containing clear, turbid free antral fluid were aspirated with the help of a syringe and an 18-G needle in 1Â PBS that was supplemented with antibiotics. Viability of cells was measured in a hemocytometer as per the trypan blue exclusion method. The proportion of living cells was 60-80% in freshly isolated granulosa cell samples (data not shown). Only living cells were included in cell counts. Cells were seeded on 24-well collagen coated plates at 1.25x10 5 cells per 0.5 ml of basal a-MEM supplemented with 2 mM L-Glutamin, 0.1% BSA, 0.084% sodium bicarbonate, 20 mM HEPES, transferrin (5 lg/ml), sodium selenite (4 ng/ml), 1 mM non-essential amino acids, penicillin (100 IU), streptomycin (0.1 mg/ml), 10 ng/ml insulin (above chemicals from Biochrom, Berlin, Germany), 20 ng/ml FSH, 50 ng/ml IGF-I and 2 lM androstenedione. For initial 48 h, cells were grown in a-MEM at 37°C in the presences of 5% CO 2 .
After initial 48 h, the media were replaced with media (supplemented a-MEM) containing different concentrations of oleic acid (C 18 H 34 O 2 ). Oleic acid was dissolved in 100% ethanol as described in Vanholder et al. (2005) and diluted to stocks of 0.35 M and final concentrations of 100 lM, 200 lM and 400 lM in culture medium.
All controls were treated with 0.11% of the solvent ethanol as vehicle corresponding to the ethanol percentage of the highest oleic acid treatment group (400 lM). Similar oleic acid concentrations were found also in follicular fluid after starvation in vivo (Aardema et al., 2013). Media were replaced every other day according to (Baufeld and Vanselow, 2013). After 8 days in culture the cells were lysed for RNA isolation and conditioned media were stored at-20°C for steroid determination.

Staining of intracellular lipid droplets
Intracellular lipid droplets were stained with the Lipid Staining Kit Oil Red O according to the manufacturer's instructions. Briefly, cells were cultured in collagen-coated 96-well plates and treated with different concentrations of oleic acid as described. After treatment, the medium was removed and the cells were gently washed twice with 100 ll of PBS. After the addition of 100 ll of 10% formalin, the cells were incubated for 30 min at room temperature. Subsequently, the formalin was discarded and the cells were washed twice with 100 ll of water and incubated with 100 ll of 60% isopropanol for 5 min. Then, isopropanol was removed, Oil Red O working solution was added, and the cells were incubated at room temperature for 15 min. Oil Red O solution was removed and the cells were washed five times with water to remove excess Oil Red O solution. For nuclear staining, 100 ll of hematoxylin was added for 1 min, and the cells were washed five times with water to remove the excess hematoxylin. Then, 100 ll of water was added and the staining was evaluated in a Nikon TMS-F inverted microscope.

Flow cytometry analysis
Different phases of the cell cycle and the percentage of apoptotic and necrotic cells were determined by flow cytometry analysis. Briefly, after culturing and treatment with different oleic acid concentrations the medium was removed completely, 250 ll of accutase enzyme (Biochrom, Berlin, Germany) was added to each well of the 24-well plates and the cells were incubated at 37°C in 5% CO 2 for 20 min. After microscopic control, cells from three wells were pooled, and transferred to a 1.5 ml micro centrifuge tube. To collect all remaining cells, the wells were washed with 300 ll of PBS and transferred to the same tube. After centrifugation at 300Âg for 5 min at 4°C, the supernant was removed and the sedimented cells were re-suspended in 300 ll PBS. The cells were then added to 10 ml of ethanol (70% v/v, ice cold) dropwise and stored at À20°C for 1-2 h. Subsequently, the cells were subjected to centrifugation at 300Âg at 4°C for 5 min. The pellets were resuspended in 1 ml of RNase solution (1 mg/ml) and incubated at 37°C in 5% CO 2 for 30 min. For nuclear staining 100 ll of propidium iodide solution (70 lM final concentration) was added and the cells were incubated at 37°C for 30 min. Then, propidium iodide fluorescence was quantified from single cells (10,000 counts) at 488 nm (argon laser band) and an emission at 600 nm ± 10 nm using a flow cytometer (EPICS-XL, Beckman-Coulter, Krefeld, Germany). The data were recorded and analyzed using the Multicycle software (Phoenix, USA) as described in previous studies (Darzynkiewicz et al., 1992;Lohrke et al., 1998).

RNA isolation and cDNA synthesis
Total RNA was isolated using the Nucleo Spin RNA II Kit (Macherey-Nagel, Düren, Germany) according to the manufacturer's instructions and quantified using a NanoDrop1000 Spectrophotometer (Thermo Scientific, Bonn, Germany). cDNA was prepared with M-MLV reverse transcriptase and RNasin ribonuclease inhibitor (both Promega, Mannheim, Germany) using oligo-(dT) primers (2 ng/ll) mixed with random hexamer primers (4 ng/ll; both Roche, Mannheim, Germany) from 200 ng RNA as previously described (Baufeld and Vanselow, 2013).

Real-time RT-PCR (qPCR)
Quantification of FSHR, LHGCR, CCND2, CYP19A1, STAR, CYP11A1, HSD3B1, CD36, PCNA, SLC27A1, relative to TBP reference transcripts by real time PCR was performed using SensiFast SYBR No-ROX (Bioline, Luckenwalde, Germany) and gene-specific primers (Table 1) in a LightCycler 96 instrument (Roche) as described previously (Baufeld and Vanselow, 2013). Briefly, we amplified 0.25 and 0.5 ll of cDNA of each sample in a 12-ll reaction mixture as follows: pre-incubation at 95°C for 5 min, 40 cycles of denaturation for 20 s at 95°C, annealing for 15 s at 60°C, extension at 72°C for 15 s and single-point fluorescence acquisition for 10 s. To ensure amplification of correct products melting points of products were routinely analyzed. Initially, all amplicons were cloned and sequenced for authenticity. These cloned plasmids were used as external standards. For preparation of external standard curve we prepared five fresh dilutions with concentrations of 5 Â 10 À12 to 5 Â 10 À16 g DNA/reaction. These were co-amplified with each run of samples. After amplification, values of 0.25 and 0.5 ll of cDNA were averaged considering different dilutions. The size of product was routinely controlled by agarose gel electrophoresis. The abundance of transcripts was calculated relatively to TBP as an appropriate housekeeping control (Baddela et al., 2014).

Quantification of 17-beta-estradiol and progesterone concentrations in culture media
After, 8 days of culture, media conditioned for 2 days from individual samples were collected and the concentrations of 17-betaestradiol and progesterone were determined using radioimmunoassay (RIA) with rabbit-raised antibodies purified by chromatography as described previously (Schneider and Brüssow, 2006). The levels of 17-beta-estradiol were determined with a modified competitive 3H-RIA. The tracer, [2,4,6,7-3H]estradiol-17b, was purchased from Hartmann Analytic (Braunschweig, Germany). The lower limit for detection of 17-beta-estradiol was 3 pg/ml and intra-and interassay coefficients of variation were 6.9% and 9.9%, respectively. To prepare the standards, 17-betaestradiol was dissolved in 100% ethanol and diluted in RIA buffer.
The analysis of media was done with 10 ll of undiluted samples in duplicates. The levels of progesterone were quantified by a competitive 3H-radioimmunoassay. The tracer, [1,2,6,7-3H(N)] progesterone, was purchased from PerkinElmer (Boston, MA, USA). The minimum detection limit was 7 pg/ml and intra-and interassay coefficients of variation were 7.6% and 9.8%, respectively. Progesterone standard dilutions were prepared in RIA buffer after dissolving the steroid in 100% ethanol. For analysis, media were diluted 1:40 in RIA-buffer and 10 ll of the dilution was measured in duplicates. The levels of radioactivity were measured in a liquid scintillation counter (LSC) that also contains an integrated RIA-calculation programme (TriCarb 2900 TR; PerkinElmer).

Cell viability assay
The cell viability assay indicating the numbers of viable cells was performed as previously described (Baufeld and Vanselow, 2013). Briefly, cells were grown in 96-well plates and treated with oleic acid as described. After 8 d in culture, the medium was replaced with 100 ll of fresh medium and 20 ll of CellTiter 96 AQueous One Solution Reagent (Promega, Mannheim, Germany) in each well. Subsequently, the cells were incubated for 1 h at 37°C in 5% CO 2 and the OD was measured at 450 nm in a MR5000 plate reader (Dynatech, Denkendorf, Germany).

Statistical analyses
Data of treatment groups were analyzed by one-way ANOVA (ANOVA, all pairwise multiple comparison procedures, Holm-Sidak method) or unpaired t-test using the Sigma Plot 11.0 Statistical Analysis System (Jandel Scientific, San Rafael, CA, USA). All experimental data are presented as means ± SD (standard deviation) of three independent experiments. Differences were considered significant at P < 0.05.

Cell morphology, lipid droplet formation and expression of fatty acid transporter transcripts
In vitro treatment of granulosa cells with increasing concentrations of oleic acid induced dramatic morphological and structural changes. At 400 lM, nearly all cells showed a typical morphology ( Fig. 1) that has been described as ''foam cell-like" in other cell types [2,32,35]. Oil Red O staining of the cells revealed numerous lipid droplets inside the cells, in contrast to untreated controls. The numbers of lipid droplets however, varied between different concentration groups. In the samples that received 400 lM oleic acid Oil Red O staining was more intense as compared with the 100 lM and 200 lM treatment groups (Fig. 2). As a first mechanistic approach, the transcript abundance of possibly involved fatty acid transporters CD36 and SLC27A1 was determined in treated (400 lM oleic acid) and untreated granulosa cells. The data showed that oleic acid treatment induced a significant increase of CD36 and SLC27A1 transcripts (Fig. 2e, f).

Effects on apoptosis, necrosis and proliferation, and on the expression of associated transcripts
Flow cytometry analysis of untreated granulosa cells and cells treated with different concentrations of oleic acid showed no significant differences of the proportions of cells at different phases of the cell cycle ( Fig. 3a-h). The percentage of cells in the sub G0/G1 phase of the cell cycle indicated that only a small number of cells underwent necrosis with 4.67% ± 1.04%, 3.3% ± 0.12%, 4.83% ± 0.7% and 6.12% ± 0.67% in the controls and in the 100 lM, 200 lM and 400 lM treatment groups, respectively, however without significant differences. There were virtually no signs of apoptosis in controls and oleic acid treated cells. The proliferation activity of the cells was also low under all experimental conditions as indicated by the low percentage of cells in the S-phase or G2/M-phase, but high percentage in the G0/G1-phase. CellTiter 96 AQueous One Solution Cell Proliferation Assay results showed no change in the numbers of viable cells treated with oleic acid compared with untreated cells. Real-time PCR results showed that PCNA was not significantly altered, whereas CCND2 transcript levels were significantly reduced by oleic acid treatment at all concentrations (Fig. 4).

Expression of gonadotropin receptors and transcripts involved in steroidogenesis
The data showed that oleic acid reduced the levels of FSHR and LHCGR transcripts at all concentrations, but maximum down regulation was observed in case of 400 lM (Fig. 5). Quantitative PCR analysis of transcripts involved in steroidogenesis, such as STAR, CYP11A1, HSD3B1 and CYP19A1, also revealed significant downregulation with maximum inhibition at 400 lM (Fig. 6).
3.4. 17-beta-estradiol and progesterone production of oleic acid treated granulosa cells 17-beta-estradiol and progesterone levels in granulosa cell conditioned media were determined after treatment with different oleic acid concentrations. The results showed that all concentrations of oleic acid gradually reduced the production of 17-betaestradiol, whereas that of progesterone was significantly decreased only in case of 400 lM (Fig. 7).

Oleic acid induced intracellular lipid droplet accumulation and the transformation into foam cells without changing the proportion of apoptotic, necrotic and proliferative cells
In the present study, we describe for the first time that oleic acid induced specific and profound morphological, physiological and molecular transformations of bovine granulosa cells. The marked morphological changes closely resemble foam cells, which were previously observed in smooth muscle cells and macrophages (Aqel et al., 1984;Doran et al., 2008;Ma et al., 2011;Oh et al., 2012). Foam cells are characterized by the accumulation of lipid droplets, which can be identified by the Oil Red O staining (Ii et al., 2008;Lee et al., 2009;Xu et al., 2010). In smooth muscle cells, it was observed that oleic acid can induce foam cell formation  by increasing lipid droplet accumulation in a concentrationdependent manner (Ma et al., 2011). In addition to this, oleic acid also increased lipid droplet formation in Huh7 cells (Rohwedder et al., 2014). In a recent paper it has been described that lipid droplet accumulation in the cumulus cell layer may have a protective function (Lolicato et al., 2015). Here, we describe that oleic acid induced the expression of CD36 and SLC27A1, which are involved in the transport of fatty acids in different cell types   (Bonen et al., 2007). Our results are consistent with other studies in bovine mammary epithelial cells, where oleic acid induced CD36 mRNA expression (Yonezawa et al., 2004) and where it was shown that inhibition of CD36 lead to inhibition of lipid droplet formation (Ma et al., 2011). In the light of these studies, our findings suggest that oleic acid may also induce the accumulation of lipid droplets  via elevated CD36 and SLC27A1 synthesis in bovine granulosa cells. However, it is not yet clear, whether or how lipid droplet accumulation is causally involved in other effects induced by oleic acid treatment as altered gene expression and reduced steroid production.
In spite of these dramatic morphological changes, our results also showed that treatment with oleic acid did not induce cell death (apoptosis or necrosis). Also the number of viable granulosa cell was not affected even at highest oleic acid concentrations. This result is consistent with some previous studies, where oleic acid was reported to have no effects even at high concentrations in CHO cells and b-islet cells on cell viability, but can even prevent palmitate-induced apoptosis (Listenberger et al., 2001(Listenberger et al., , 2003Maedler et al., 2001). In contrast, however, it has also been reported that oleic acid promotes apoptosis and necrosis in human lymphocytes, where it mediates cell death by the activation of caspase-3 (Cury-Boaventura et al., 2006). From our results, we conclude that oleic acid at physiological concentrations does not significantly affect the viability of cultured granulosa cells or the proportion of apoptotic and necrotic cells. With regards to gene expression, however, cell cycle regulation might be slightly affected, because the abundance of CCND2 transcripts showed significant down regulation in cultured granulosa cells.

The decreased steroid hormone production might be due to the down-regulation of associated key transcripts by oleic acid
To determine effects of oleic acid on granulosa cell function, we analyzed key genes involved in gonadotropin signaling such as FSHR and LHCGR, and in steroidogenesis such as STAR, CYP11A1, HSD3B, and CYP19A1. Our data showed that oleic acid decreased the expression of these genes. FSHR has been frequently shown to be present on granulosa cells (Camp et al., 1991;Sprengel et al., 1990;Zeleznik et al., 1974). The binding of FSH leads to the activation of FSHR, which in turn can induce nearly 500 target genes (Escamilla-Hernandez et al., 2008;Hsueh and Rauch, 2012) that are involved in processes of cytodifferentiation and proliferation thus leading to the development of preovulatory follicles (Erickson, 1983;Hirshfield, 1991).
FSH also activates the transcription of genes involved in the cascade of steroidogenesis, as STAR, encoding steroidogenic acute regulatory protein, which is involved in the initial step of steroidogenesis, the transport of cholesterol into the mitochondria. Subsequent steps are catalyzed by cholesterol side chain cleavage enzyme encoded by CYP11A1, and by the 3 beta hydroxyl steroid dehydrogenase transcribed from HSD3B1, and CYP19A1, encoding aromatase, the key enzyme of estrogen synthesis (Clarke et al., 1993;Eimerl and Orly, 2002;Ke et al., 2004Ke et al., , 2005. According to our data, the treatment of granulosa cells with oleic acid significantly reduced the expression of FSHR and LHCGR along with STAR, CYP11A1, HSD3B1 and CYP19A1. Accordingly, the levels of progesterone and 17-beta-estradiol found in granulosa cell conditioned media were also lower compared with the control. Very likely, the reduced production of progesterone and particularly of 17-beta-estradiol might be due to the reduced transcription of STAR, CYP11A1, HSD3B1 and CYP19A1. These results appear contradictory to previously published data , from which it was concluded that oleic acid inhibited proliferation, but stimulated the production of 17-beta-estradiol in cultured bovine granulosa cells. However, our study and the study by Vanholder et al. (2005) are based on very different experimental approaches.
In the study of Vanholder et al. (2005) granulosa cells were harvested from large follicles and cultured for two days. In the present study, which was strongly focused on effects of oleic acid on gene expression granulosa cells were obtained from small to medium sized follicles and were cultured for 8 days, because it has been demonstrated that the abundance of CYP19A1 transcripts , and decreased expression of FSH regulated genes. Also the expression of FSHR was largely down-regulated, thus suggesting that the cells become less sensitive to FSH and as a consequence downstream processes like estradiol and progesterone production were also compromised.
or of 17-beta-estradiol production in granulosa cells isolated from large follicles are rapidly (i.e. within the first 48 h) down-regulated under culture conditions (Baufeld and Vanselow, 2013;Gutierrez et al., 1997). In contrast, CYP19A1 expression preferentially directed from the granulosa cell-specific promoter region 2 (Hamel et al. 2005), and as a consequence, 17-beta-estradiol production is considerably restored after 6 days in culture and further increases to day 8 of culture (data not shown), if granulosa cells descend from small-to medium-sized follicles (Gutierrez et al., 1997;Silva and Price, 2000). Accordingly, the observed effects of oleic acid on in vitro differentiated granulosa cells in the present work reflect long-term effects, thus including profound effects on cell type-specific gene expression profiles. In contrast, the effects observed by Vanholder et al. (2005) might be explained by rapid mechanisms including translational and/or enzyme activity. Possible changes of transcript abundance have not been determined in previous studies.

Oleic acid-induced alterations might be due to reduced sensitivity to gonadotropic hormones
An important clue to understand the multiple and specific effects of oleic acid on granulosa cells is provided by the observation that oleic acid clearly reduced the expression levels of FSHR. In rat granulosa cells, it was shown that FSH induces the expression of LHCGR via protein kinase A (PKA) and phosphoinositide3-kinase (PI3K) pathways (Law et al., 2013). LHCGR is highly expressed in granulosa cells during the preovulatory stage to enable responsiveness to the LH surge thus leading to ovulation, oocyte maturation and corpus luteum formation (Zhang et al., 2001). Thus, reduced FSHR expression might be responsible for the observed decrease in LHCGR transcripts abundance. FSH in cooperation with activin also mediates CCND2 expression, which is required for granulosa cell proliferation (Hsueh et al., 1984;Park et al., 2005;Sicinski et al., 1996). Accordingly, follicular maturation is seriously compromised in CCND2 deficient mice (Sicinski et al., 1996). The down regulation of FSHR may therefore affect the above signaling pathways thus leading to compromised granulosa cell function.
In the present study, we showed that oleic acid induced specific morphological and physiological changes and alterations in the gene expression levels of cultured bovine granulosa cells. These data support the hypothesis that during NEB increased oleic acid levels in the follicular fluid can directly affect the functionality of granulosa cells and thus the follicular development and differentiation, which might be responsible for the compromised fertility of dairy cows under these conditions. Suggestively, these alterations might be primarily attributed to the reduced expression of FSHR, which is absolutely essential for the development and maturation of ovarian follicles (Fig. 8).