Platelet-activating factor-stimulated production of reactive oxygen species in ovarian granulosa cells from periovulatory follicles



The platelet-activating factor (PAF) is a proinflammatory lipid present in the fluid of ovarian Graafian follicle. Ovarian blockage of the PAF receptor (PAFr) reduces ovulations in the rat whereas underlying mechanism is poorly understood. Mural granulosa cells (MGC) were mechanically isolated from the theca interna of bovine periovulatory follicle. The mRNA abundance for PAFr, progesterone receptor and cyclooxygenase-2 were measured by real-time PCR. Cytosolic calcium (Ca2+) concentration was assayed by microscopy using Fura-2 AM as indicator, 8-isoprostaglandin F (8-isoPGF) by an ELISA kit. Fluorescent products arising from intracellular oxidation of hydroethidine (HE) and dihydrorhodamine (DHR) were quantified by flow cytometry. The cells expressed PAFr mRNA and PAFr protein and responded to cPAF (nonhydrolyzable form of PAF) with a pulsating increase in Ca2+, demonstrating functional PAFr. Elevation of Ca2+ was reversed by WEB-2086, an inverse PAFr agonist. cPAF elevated the level of 8-isoPGF in the medium of MGC cultured with luteinizing hormone (LH). cPAF alone had no significant influence on the oxidation of HE and DHR, or 8-isoPGF level. In MGC from vital periovulatory follicle, PAF and LH signaling plays an important role in regulating the production of excessive oxidants. Blockage of PAFr seems to interfere with these regulatory processes essential for ovulation. © 2009 International Society for Advancement of Cytometry


The platelet-activating factor (PAF) is known to participate in both male and female reproductive processes (1–6). The bioactivity of PAF is mediated by the PAF receptor which exists in many tissues. Evidence for its role in the regulation of the female reproduction has been obtained by studies on the rat (5, 6). Subsequently to ovarian stimulation with gonadotropin, the local administration of a PAF receptor antagonist reduced follicle ruptures. This inhibition was in part reversed by simultaneous administration of PAF (5).

PAF is present in the fluid of human periovulatory Graafian follicles (7). The cultured ovine follicular wall has been shown to secrete PAF (8). However, inconsistent results have been reported regarding the regulation of the PAF level by the periovulatory surge of luteotropin (also called luteinizing hormone, LH). The PAF production of ovine follicle increased after the LH surge (8), whereas the PAF concentration in rat ovarian tissue decreased within several hours after the human chorionic gonadotropin - triggered LH surge (9).

Many of the cell types, including leukocytes of the myeloid lineage, lymphocytes, endothelial cells, which produce PAF in response to stimulants, can also become the target of PAF bioactions. Apart from the function of PAF as chemoattractant, PAF stimulates the production of the interleukins 6 and 8 (10, 11), the monocyte chemoattractant protein-1 (10, 12), leukotriene (12) and is involved in the cell cycle regulation (11).

PAF activates cell surface Gi and Gq protein-coupled receptors to induce these diverse biological functions (12–16). The functional Gq protein-coupled receptor signals Ca2+ mobilization, induces cross-talk between Ca2+-sensitive and Ca2+-independent phospholipases A2 (PLA2) and regulates the formation of leukotrienes. The Gi and Gq protein-coupled PAF receptor is required for activation of phospholipase Cβ, mobilization of cytosolic calcium, activation of protein kinase C (PKC) and chemoattractant generation. Thus, stimulation of both functional receptor types results in cytosolic calcium increase. One of the possible consequences is the activation of calcium-sensitive PLA2 which preferentially releases arachidonic acid from phospholipids (17). This fatty acid can serve as substrate for peroxidations by cyclooxygenases and lipoxygenases. The resulting lipohydroperoxides are converted in enzymatically controlled manner to constituents of inflammatory and noninflammatory pathways.

However, arachidonic acid is susceptible to random oxidation by reactive oxygen species (ROS), and periovulatory follicle has been reported to produce ROS (18). nonenzymatic oxidation of phospholipids by ROS gives rise to the generation of iso-prostanes (19, 20). At least, 8-isoPGF has been shown to have potent vasoconstrictor activity (21, 22), this effect may be important for the ovulatory regulation. However, role of PAF in the generation of 8-isoPGF by periovulatory follicle is unknown. The aim of this study was to get information about the expression of functional PAF receptor, progesterone receptor and cyclooxygenase-2 which are also known to be essential for final follicular development, in bovine periovulatory follicles, and to elucidate the regulation of ROS and 8-isoPGF levels by LH and PAF.


Isolation and Culture of Granulosa Cells

Periovulatory follicles were induced following a protocol described elsewhere (23). Briefly, mid-cyclic Holstein heifers were injected with a prostaglandin Fanalog (cloprostenol) to regress the corpus luteum, then with an analog of the gonadotropin releasing hormone (GnRH, depherelin) to trigger the LH surge. Luteal regression and follicular development were monitored in vivo by daily transrectal ultrasonography. Twenty hours post-GnRH the ovary bearing the periovulatory follicle was removed by ovariectomy. The procedures with animals were approved by a governmental Animal Care and Use Committee.

The follicle was dissected from the ovary, the follicle fluid was aspirated, and the cells in the fluid were spun off (100 × g, 10 min, 4°C). Aliquots of the supernatant were stored (−80°C) for analyses. The thecal layers were separated using surgical tools. The mural granulosa cells were prepared by scraping the luminal theca interna with fine forceps in serum-free Hepes (20 mM, pH 7.4)-buffered DPBS containing BSA (0.5%, w/v). After filtration (cell strainer, 100 μm, Falcon), the cells were collected by centrifugation (60 × g, 10 min, 4°C), the sediment was resuspended in ice-cold water (10 sec) to lyse erythrocytes. Lysis was termed by NaCl (to 150 mM), the cells were spun off (60 × g, 10 min, 4°C), resuspended in Mega cell medium (Sigma, Deisenhofen, Germany), and counted (Multisizer, Coulter). The cell concentration was adjusted to 0.5 × 106 per ml, and aliquots (150 μl) were plated on 96-well culture plates. Viability was tested by trypan blue exclusion (>98% after preparation) as recommended by the manufacturer (Sigma). Cultures were maintained at 37°C in humidified chambers gassed with 5% CO2 air.

Assay of mRNA

Total RNA was extracted from freshly isolated and cultured mural granulosa cells. The RNA from cells was extracted using Invisorb Spin Cell RNA Mini kits (Invitek, Germany) as recommended by the manufacture. Concentration and quality of the extracted RNA were measured using a NanoDrop ND-1000 Spectrophotometer (Peqlab Biotechnologie GmbH, Erlangen, Germany). Ratios of absorbance at 260 and 280 nm of all preparations were about 2.0. Integrity of RNA was checked by denaturing agarose gel electrophoresis and ethidium bromide staining. The iScriptTM cDNA Synthesis Kit (Bio-Rad Laboratories GmbH, München, Germany) was used to synthesize cDNA from 100 ng of total RNA from each sample according to the manufacturer's instruction. A negative control, without reverse transcriptase, was processed for each sample to detect possible contaminations of genomic DNA or environmental DNA.

The abundance of mRNA for ribosomal protein S18 (RPS18), cyclooxygenase-2 (COX-2) and progesterone receptor were quantified by real-time PCR (iCycler, Bio-Rad Laboratories GmbH, München, Germany) as described previously (23). Briefly, 1 μl aliquots of each RT reaction (1/20 of total) were primed, in each 10 μl PCR, using an iQ-SYBR green supermix (Bio-Rad Laboratories GmbH, München, Germany) and gene-specific oligonucleotides (final concentration of 0.2 μM). The sequences of specific bovine primers used are shown in Table 1. PCR was performed in 40 cycles with 180 s at 94°C, 10 s at 94°C followed by 30 s at 60°C and 225 s at 70°C. The specificity of amplification was determined by melting curve analysis and agarose gel electrophoresis in comparison with an oligonucleotide molecular mass ladder to confirm that the calculated molecular mass of the cDNA corresponded to the produced cDNA. The cDNA structure was checked by sequencing. The reported sequences matched exactly to those published in GenBank. Each cDNA was quantified in duplicate; the average value of each sample value minus the corresponding negative control value was used to calculate the cDNA product corresponding to the abundance of mRNA. Relative amount of mRNA was calculated by dividing the specific real-time PCR product by the corresponding RPS18 mRNA.

Table 1. Primer sites and corresponding mRNA fragments
  1. RPS18, ribosomal protein S18; COX-2, cyclooxygenase-2; P4R, progesterone receptor.

COX-21033-1052 ATG GGG CGA TGA GCA GTT GT1248-1229 ACG TCA GGC AGA AGG GGA TG216gi 2921285

Measurement of Cytosolic Calcium

The procedure followed a described technique (24, 25) with fura-2 as indicator. Briefly, cells in Ca2+ free DPBS were loaded with the calcium-sensitive dye fura-2 AM (to 2 μM) in the dark (30 min with gentle shaking). After washing, the calcium signal (fura-2 ratio, excitation at 340 nm and 390 nm, emission at 510 nm) was recorded using a high-resolution residual-light fluorescence imaging system (Till photonics, Martinsried, Germany). Cells were viewed continuously under the microscope (IX70, Olympus) in an open dish incubation chamber (ΔTC3, Bioptech, Butler, PA) at 37°C.

Specific stimulation of the PAF receptor by 10 nM carbamyl PAF (cPAF, 15 min), a metabolic - stable PAF analog was checked by preincubation (15 min) with WEB2086 (9.5 μM), a selective PAF receptor antagonist (26) and an inverse PAF receptor agonist which stabilizes inactive conformation of the receptor in absence of an agonist (27).

A positive control was obtained by exposure of the cells to 1 μM thapsigargin which mobilizes calcium ions from the endoplasmic reticulum (28) independent of the PAF receptor.

Detection of Intracellular ROS

Acute intracellular production of ROS was assessed by incubation (30 min) of cell aliquots with hydroethidine (HE, 0.7 μM) and dihydrorhodamine (DHR, 0.7 μM) in DPBS. These virtually nonfluorescent probes are oxidized to fluorescent products. Fluorescence of single cell was measured by a flow cytometer (Epics XL, Beckman-Coulter) using an excitation wavelength at 488 nm and emission at 515 ± 10 nm (DHR) and 600 ± 10 nm (HE). Granulosa cells were gated in the flowcytometric histogram (forward scatter versus side scatter signal) on the analogy to the technique described previously (29).

Assay of PAF

The concentration of PAF was assessed by a Platelet Activating Factor (PAF) [3H] Scintillation Proximity Assay (SPA) system (Amersham Bioscience). The assay is based on the competition between unlabelled PAF and a fixed quantity of tritium labeled PAF for a limited number of binding sites on a PAF specific antibody. Radioactive counts were countered in a liquid β-scintillation counter (Berthold, Germany). The concentration of unlabelled PAF in a sample was determined by interpolation from a standard curve. The sensitivity of the assay has been reported by the manufacturer to be 20 pg/ml. The inter- and intra-assay coefficients of variation were 15.3 and 11.9%.

Determination of 8-Isoprostaglandin F (8-isoPGF)

The concentration of 8-isoPGF was assayed by an EIA kit according to the direction of the manufacturer (Cayman). Culture media samples were diluted with an equal volume of EIA buffer, then used for competition between 8-isoPGF and a constant concentration of 8-isoPGF tracer for a limited number of 8-isoPGF -specific rabbit antiserum binding sites (18 hours at 4°C). The lowest dose of 8-isoPGF which could measured reproducibly was 5 pg/ml. Intra- and inter- assay coefficients of variation were 19 and 14%. Cross-reactivity of the antibody has been reported by the manufacturer to be limited to 8-iPGF (20.6%) and 2,3-dinor-8 iPG F (4.0%).

Statistical Analysis

Effects were assessed by one way analysis of variance (ANOVA). Test (pair-wise comparisons or comparisons vs control) were performed by the post hoc Newman–Keuls procedure and Dunnet's test as appropriate. Values of P < 0.05 were considered statistically significant. Results are presented as means ± SD when not otherwise stated. Test and analyses of linear regression were performed by SAS statistical package (SAS Institute Inc. Cary, NC) and Sigma Stat of the Jandel Scientific Software (Erkrath, Germany). Graphic presentations were accomplished by Sigma Plot of the Jandel Scientific program package.


Mural granulosa cells from the periovulatory follicle expressed the PAF receptor at the level of the transcript, 0.7 ± 0.3 pg mRNA/100 ng total RNA, 0.003 ± 0.001 mRNA/RPS18 mRNA respectively, and at the level of protein, roughly 1/3 of the concentration observed in monocytes (data not shown) as reported previously (29). The PAF receptor was found to be functional demonstrated by the cytosolic calcium mobilization in response to cPAF (Fig. 1A). The PAF receptor antagonist, WEB2086, suspended the cPAF-provoked phasic cytosolic calcium increase (Fig. 1B) and the thapsigargin-triggered Ca2+ spike was not reversed by WEB2086 (Fig. 1C).

Figure 1.

Mobilization of cytosolic calcium by cPAF. Elevation of cytosolic calcium (Ca2+) in mural granulosa cells from periovulatory follicle was detected using fura-2 AM and computer-aided microscopic imaging as described in methods. (A) Demonstrating pseudo-colored image (green, basal level) of the rapid cell-dependent response to cPAF (10 nM). The punctual color (insert) denotes the cells (n = 3) used in (B) and (C) to quantify the phasic Ca2+ mobilization in a single cell. (B) shows cPAF-provoked biphasic Ca2+-spikes followed by a more sustained increase. The negative control (C) was obtained by preincubating the cells with the PAF receptor antagonist, WEB2086 (10 μM, 10 min), followed by the addition of cPAF (10 nM). Note that during the 400 s interval used to demonstrate the cPAF effect shown in (B), the WEB2086-treated cells failed to mobilize cytosolic calcium. As the positive control, thapsigargin (1 μM) was added into the WEB2086-containing incubation medium (C).

The mural granulosa cells from periovulatory follicles produced PAF (289 ± 83 pg/h/106 cells/ml culture medium). This production may largely contribute to the PAF concentration of the follicle fluid (2.1 ± 0.5 ng/ml). The rate of PAF formation by pieces of the theca interna was 9.4 ± 2.3 pg/mg wet weight/h. An inhibitor of cytosolic phospholipase A2, arachidonyl trifluormethyl ketone (AACOCF3, 6 μM), and the PAF receptor antagonist, WEB2086 (1 μM), reduced the production of PAF (0.3- and 0.6-times the basal value), suggesting the remodeling pathway as the main source of PAF formation because the remodeling pathway require phospholipase A2, but de novo pathway require phosphocholine-transferase. Luteinizing hormone (LH) inhibited the PAF production in a dose-dependent manner (5 and 25 ng/ml of LH, 0.8 and 0.45 times the basal value).

In a dose similar to the preovulatory LH surge, LH moderately decreased the intracellular oxidation of DHR (Fig. 2D) and HE (Fig. 2E). Indomethacin, an inhibitor of cyclooxygenase-2, caused a strong reduction of the fluorescence production. The drug particularly reduced the fluorescence arising from the oxidation of DHR (0.3-times the control value, Fig. 2D).

Figure 2.

Intracellular oxidation of fluorogenic probes. Single cell fluorescence (AC) was measured by flow cytometry. MFI denotes mean fluorescence intensity, bars stand for mean ± SD (n = 3–4 independent experiments), *P < 0.05 versus control. Autofluorescence (MFI, arbitary units) was 1.43 ± 0.27 (n = 6). Incubation only with dihydrorhodamine yielded MFI = 6.5 ± 1.5, with hydroethidine, MFI = 4.1 ± 1.4. Mural granulosa cells from periovulatory follicles were exposed to luteinizing hormone (LH, 4 h), then cPAF (15 min). Cells were incubated (4 h) with indomethacin (Indo) and the PAF receptor antagonist, WEB2086, alone and WEB2086 (4 h), then cPAF (15 min). Controls were obtained by incubation without the effectors, in some experiments with the carrier DMSO < 0.01%. Next, the cells were loaded (30 min) with (D) dihydrorhodamine or (E) hydroethidine, 0.7 μM each. The portion of propidium iodide–positive cells was <2% after culture. [Color figure can be viewed in the online issue, which is available at]

Exposure of the cells to allopurinol (100 μM) and apocynin (500 μM), which is the inhibitors of xathine oxidase and NADPH oxidase, respectively, did not significantly change the fluorescence production (data not shown).

In contrast indomethacin also substantially decreased the level of 8-isoPGF (Fig. 3A). However, strong elevation (roughly twofold the control value) of 8-isoPGF concentration was only observed by preincubating the cells with LH (4 h) followed by an exposure (15 min) to a physiological dose of cPAF (Fig. 3A). The oxidation of DHR and HE remained virtually unchanged in these conditions (Fig. 2).

Figure 3.

Response of 8-isoPGF and periovulatory-related mRNA to LH, cPAF, WEB2086, and indomethacin. Periovulatory mural granulosa cells were exposed to LH (4 h), then to cPAF (15 min) or only to cPAF (15 min), WEB2086, LH and indomethacin (4 h each) in doses indicated on the abscissa. (A), showing the extracellular level of 8-isoPGF (B), demonstrating the relative abundance of the progesterone receptor mRNA and (C) that of the cyclooxygenase-2 mRNA. Differential cellular response of relative mRNA abundance to an exposure to LH, then to cPAF or to the indicated effectors alone is illustrated by empty (insensitive cells) and filled (responsive cells) bars. They stand for mean ± SD (three independent experiments each), *P < 0.05 vs control.

Because of the substantial inhibitory effect of indomethacin, we looked for the expression of COX-2 which is under control of the transcription factor, progesterone receptor. In cells which did not respond to LH and cPAF the relative abundance of progesterone receptor (Fig. 3B) and COX-2 mRNA (Fig. 3C) were elevated, the 8-isoPGF level increased 2.2-times the control value (Fig. 3A) after a preincubation with LH followed by cPAF exposure. Responsive cells (Figs. 3B and 3C) elevated the 8-isoPGF level only 1.5-times the control value (Fig. 3A). Thus, the data demonstrate a correlated reduction of fluorescence production and 8-isoPGF concentration by indomethacin but divergent response of elevating the production of fluorescence and excessive oxidant which generates 8-isoPGF. The divergence associates with the differential response of the expression of the progesterone receptor and cyclooxygenase-2 to LH and cPAF.


Rodent ovarian periovulatory follicles have been shown to respond to PAF receptor antagonist with reduced number of follicular rupture whereas PAF restored in part normal ovulation (6). Therefore, PAF and PAF receptor signaling has been implicated in the ovulatory regulation. However, direct evidence for the expression of functional PAF receptor is lacking to the best of our knowledge. Here, we demonstrated that bovine periovulatory follicle express PAF receptor transcript and protein.

Evidence for functional PAF receptor was provided by the PAF-induced mobilization of cytosolic calcium. Mural granulosa cells responded to cPAF with pulsatile spikes of cytosolic calcium. The phasic response is consistent with studies on other cells which showing rapid desensitization of the PAF receptor on agonist exposure followed by functional recruitment by phosphorylation and dephosphorylation (12, 30). Preincubation with the PAF receptor antagonist, WEB2086 (27, 31, 32), abolished the calcium spikes, indicative of receptor-mediated response.

The cells not only responded to PAF but also produced PAF, suggesting an autocrine regulation of the PAF receptor. Mural granulosa cells produce PAF, agreed with earlier reports (8, 9). Our results also indicate that the production of PAF is likely to be mediated by the remodeling pathway, because the inhibitor of phospholipase A2, AACOCF3 used in a dose selectively to cPLA2 (33) suppressed the production of PAF (34, 35). In agreement with a previous study (9), we found that LH inhibited dose-dependently the PAF production.

Luteinizing hormone drives diverse intracellular calcium second messenger signals in ovarian cells (36) and activates a number of intracellular signals cascades. Among the well-studied signaling modes are the adenylyle cyclase-cAMP protein kinase A (PKA) and the Ca2+-phosholipase C- inositol triphosphate-PKC transduction systems (37). The latter is also activated by the stimulated PAF receptor (30). Thus, our results suggest that crosstalk or interference may occur between PAF receptor and LH signaling pathways.

In our experiment, LH elevates the level of mRNA for the progesterone receptor and COX-2. These results are consistent with reports that the PKA pathway regulates the expression of COX-2 (38). The progesterone receptor acts as a transcription factor that stimulates the generation of COX-2 mRNA in bovine periovulatory follicle (39). Our data additionally indicate that PAF enhanced the LH-mediated increase in the mRNA abundance of the progesterone receptor and COX-2. This result supports the notation of interactive PAF and LH signaling pathways.

The mRNA level of COX-2 largely corresponds to the activity of the enzyme because of the suicidal action of the enzyme activity (40). However, the absolute requirement of a hydroperoxide to peroxide arachidonate, the initial step of prostanoid formation, prompted us to study the production of ROS in periovulatory mural granulosa cells in response to LH and cPAF. cPAF itself exerted small effect on both intracellular ROS and extracellular 8-isoPGF concentration. However, in granulosa cells which failed to respond to LH plus cPAF with an increase in the mRNA abundance of the progesterone receptor and COX-2, cPAF elevated strongly the 8-isoPGF concentration. Differently, in cells which responded to an exposure to LH, then cPAF with a rise in these mRNA species, cPAF elevated moderately the 8-isoPGF level. These different reactions provided additional evidence for an interaction between the LH and PAF signaling pathways. This regulatory mode may play a role in the ovulatory physiology, because the 8-isoPGF has been reported to exert vasoconstrictive effect on microvasculature (27, 28).

The strong reduction of both the 8-isoPGF concentration and the oxidation of HE and DHR to fluorescent products by indomethacin, the cyclooxygenase inhibitor, suggests, that the cyclooxygenase activity is involved in these response. The drug binds to the lipoxygenase site of the enzyme (40), thereby inhibiting the formation of instable PGG2 followed by the nonenzymatic conversion of PGG2 to PGH2. Therefore, these reactions may associate with spontaneous oxidant production capable of generation 8-isoPGF. The oxidants are likely to include a peroxynitrite component because granulosa cells are known to produce nitric oxide (41). The stronger reduction of the fluorescence from DHR than that from HE supports this view. Because the oxidation of DHR is largely independent of the peroxidase species and peroxynitrite-mediated oxidation of DHR yields fluorescence with one of the highest efficiencies yet determined (42), but HE depend on superoxide anions. To approximate a linear response of fluorescence with concentration of intracellular oxidant, a low dose of the fluorogenic probes was used in our experiments.

In conclusion, periovulatory mural granulosa cells express functional PAF receptor. The cells produce and secrete 8-isoPGF, and these processes are enhanced by cPAF in presence of LH, indicate PAF and LH has a cross-talk or interference to regulate ovulation. Because indomethacin substantially suppressed both the 8-isoPGF production and the cellular fluorescence arising from the oxidation of DHR and HE, the COX-2 activity appears to contribute to an excessive production of oxidant in periovulatory granulosa cells.


The authors thank Renate Hantel, Uta Naumann, and Simone Rackow for excellent technical assistance. The generous donation of WEB2086 by Boehringer–Ingelheim is appreciated.