Granulosa Cell Apoptosis Induced by a Novel FSH Binding Inhibitory Peptide From Human Ovarian Follicular Fluid

Abstract

Pituitary gonadotropins, follicle-stimulating hormone and luteinizing hormone, are the key regulators of ovarian folliculogenesis; these are known to be directly or indirectly modulated by many intraovarian factors. Our group has identified and studied one such novel peptide from human ovarian follicular fluid. Its partial N-terminal eight amino acid sequence has been deduced, referred to as octapeptide (OP). OP induces follicular atresia in mice and interferes with normal ovarian function in non-human primates, this action being similar to the native peptide. Thus, in this study, an attempt has been made to elucidate the mechanism of action of the synthetic OP by studying the pathway of follicular atresia in mouse ovary. Changes in granulosa cells were studied using various apoptotic markers by flow cytometry and immunohistochemistry. An increase in apoptotic cell population in atretic- and peptide-treated groups was observed compared with normal controls. Interestingly, both these groups exhibited differences in the apoptotic pathway. Results showed that the mitochondrial pathway was predominant in the atretic group, whereas the Fas-FasL pathway was predominant in the peptide-treated groups. The ultrastructural study also showed apoptotic changes in the OP-treated and atretic groups; the pattern of apoptosis differed at the subcellular level. (J Histochem Cytochem 56:961–968, 2008)

Ovarian follicular development is a continuous process wherein the resting follicles develop to form a mature Graffian follicle under the influence of pituitary gonadotropins follicle-stimulating hormone (FSH) and luteinizing hormone (LH) (Yang et al. 2003). It is well documented that, in addition to gonadotropins, intraovarian factors are also involved in the regulation of folliculogenesis (Armstrong and Webb 1997). Inhibin and its related family of glycoproteins, cytokines, growth factors, and neuropeptides (Pezzani et al. 1994; Findlay et al. 2000) are examples of such nonsteroidal regulatory molecules.

During folliculogenesis, the follicles, which fail to develop, undergo a degenerative process called atresia (Tsafriri and Braw 1984; Hirshfield 1991). It is now well established that granulosa cells in these atretic follicles show apoptotic changes (Hughes and Gorospe 1991; Tilly et al. 1991; Nandedkar et al. 1996,1998).

The FSH binding inhibitor (FSHBI) is one of the intraovarian nonsteroidal regulators that has been shown to modulate FSH actions by inhibiting the binding of FSH to granulosa cells. Reichert and colleagues purified FSHBI from bovine, porcine, and human follicular fluid (Darga and Reichert 1978; Sluss and Reichert 1984; Lee et al. 1990).

Our group has purified FSHBI from human ovarian follicular fluid collected from in vitro fertilization (IVF) clinics. The follicular fluid was ultrafiltered and passed on a gel column. The active fraction (inhibit binding of FSH to granulosa cells in vitro by radioreceptor assay) is referred as human gel filtration fraction-2 (hGF₂). This was further purified on a preparative followed by analytical reverse phase–high performance liquid chromatography (RP-HPLC). The single peak obtained was sequenced, and a partial N-terminal eight amino acid sequence (AESNEDGY) was obtained, which was referred to as octapeptide (OP) (Wadia et al. 2003).

Our group previously reported that hGF₂ was immunolocalized in granulosa cells using anti-hGF₂ polyclonal antibodies (Nandedkar et al. 1992,1994). Furthermore, hGF₂ inhibited steroid production by granulosa cells (Wadia et al. 2007), indicating an autocrine role of the peptide on granulosa cells of the follicles. Its ability to modulate ovarian function strengthens the need to understand its mechanism of action. Both the partially purified (hGF₂) peptide and its fragment (OP) inhibited binding of FSH to granulosa cells in vitro. Thus, the biological activity of OP (commercially synthesized) was found to be similar to hGF₂ (Wadia et al. 2003; Kulkarni-Chitnis and Nandedkar 2007). Furthermore, hGF₂ also caused apoptosis in granulosa cells of atretic follicles (Nandedkar et al. 1998,2001). Thus, in this study, specific apoptotic markers have been used to understand the mechanism of action of the peptides hGF₂ and OP with reference to the intrinsic mitochondrial and extrinsic Fas–Fas L–associated pathways of apoptosis by the techniques of flow cytometry and IHC. The apoptotic changes in the ovarian follicles of OP-treated mice were also studied at the ultrastructural level and were compared with the normal and atretic groups.

Materials and Methods

Purification of FSHBI

The FSHBI was purified from human ovarian follicular fluid as previously reported (Wadia et al. 2003). In brief, the follicular fluid obtained from patients undergoing IVF embryo transfer in the Fertility Clinic, Mumbai, India, was pooled and ultrafiltered through Amicon PM-10 (Millipore; Carrigtwohill, Ireland) membranes (cut-off < 10 kDa). The filtrate was further purified by passing it through a G-25 Sephadex (Amersham; Buckinghamshire, UK) gel column. The active fraction (hGF₂) was identified by radioreceptor assay (RRA), inhibiting FSH binding to granulosa cells in a dose-dependent manner (Wadia et al. 2003). hGF₂ was further purified and analyzed on RP-HPLC. A single peak was obtained that was sequenced. A partial sequence was obtained, and the N-terminal eight amino acid peptide, OP, was commercially procured (Imgenex; San Diego, CA). Its purity was checked on RP-HPLC. The activity of this synthetic batch was also assessed by RRA.

Animals

Swiss mice used in this study were bred in an animal house of the National Institute for Research in Reproductive Health (NIRRH), maintained at constant light (14 hours light: 10 hours dark), temperature (24C), and humidity (60%) and supplied with food and water ad libitum. Animal use was approved by the Institutional Animal Ethics Committee (IAEC), and all the experiments were conducted as per the norms of the Committee for the Purpose of Supervision of Experiments on Animals, India.

Treatment of Animals

Immature female Swiss mice (21–23 days old) were used for the study. Animals (n=3/group) from Groups I (normal) and II (atretic) were injected with 10 IU equine chorionic gonadotropin (eCG) and autopsied after 48 and 72 hr to procure ovaries predominantly containing normal and atretic follicles, respectively. Animals in Groups III and IV were primed with 10 IU eCG and administered 20 μg hGF₂ or 100 μg OP 24 hr after eCG injection, respectively. Animals of both groups were autopsied 48 hr after eCG injection (i.e., 24 hr after peptide injections).

The ovaries collected from all four groups were used for flow cytometric analysis and IHC. For the electron microscopic studies, only Groups I, II, and IV were used because the native peptide hGF₂ and OP have shown similar mechanisms of action.

Flow Cytometry

Ovaries were collected on ice in serum-free MEM 199 supplemented with l-glutamine (Hi-Media Labs; Mumbai, India), and follicles were needle punctured under the stereomicroscope (Olympus; Hamburg, Germany) to release granulosa cells into the medium. These cells were centrifuged at 1000 rpm, and the pellet was resuspended in 1 ml MEM 199. Viable cells were counted on a hemocytometer by trypan blue exclusion test, and aliquots at a cell density of 1 × 10⁶ cells/ml were used. Granulosa cells from ovaries of three animals were pooled in each group, and each experiment was repeated three times. A total of 10,000 cells were recorded for each experiment on FACS Vantage SE with Argon laser (Becton Dickinson; San Diego, CA). Data were analyzed using CellQuest Pro 3.1 software (BD Biosciences; San Jose, CA).

Identification of Cell Population Expressing Death Receptor Fas

One million cells/100 μl were incubated with 5 μl monoclonal anti-Fas antibody (0.5 mg/ml; Becton Dickinson) at 37C for 15 min. Cells were pelleted by centrifugation, washed with fresh medium, and resuspended in 100 μl medium. Normal goat serum (NGS; 5 μl) was added for blocking and incubated for 30 min at room temperature. Secondary rabbit monoclonal antibody (2 μl) labeled with FITC (Becton Dickinson) was added after washing the cells with fresh medium. The cells were incubated in the dark for 30 min at room temperature. Fluorescence was detected using a single filter to identify the cell population expressing the apoptotic marker Fas. Cells were acquired at 530 nm and analyzed.

Detection of Mitochondrial Membrane Potential (ΔΨ_m)–Rhodamine 123 (R123)

R123 is a mitochondrial dye that is taken up against the concentration gradient by functionally active cells. Thus, the normal cells accumulate a higher concentration of R123 and are identified as cells that fluoresce at higher intensity (R123_high). Apoptotic cells, on the other hand, have a lowered ΔΨ_m (Alonso-Pozos et al. 2003). Thus, these cells accumulate a lower concentration of R123 (which is that taken up by osmotic pressure) and are identified as cells that fluoresce at low intensity (R123_low). Thus, by measuring the percentage of R123_high and R123_low, one can identify non-apoptotic and apoptotic cell populations. For this study, 1 × 10⁶ cells/ml were stained with 2 μl R123 (10 μg/ml methanol) and incubated at 37C for 15 min. Cells were acquired at 530 nm and analyzed.

Identification of Apoptotic Population by Anti-caspase 3 Antibodies

Granulosa cells (1 × 10⁶ cells /ml) were washed in cold PBS and incubated in Cytoperm–Cytofix solution (BD Pharmingen; San Diego, CA) (provided in kit) on ice. These permeabilized and fixed cells were centrifuged at 1000 rpm and resuspended in 500 μl Perm/Wash buffer at room temperature. Cells were incubated with 20 μl anti-caspase 3 monoclonal antibodies (ready to use) labeled to FITC from the kit (Becton Dickinson). Cells were acquired on FACS Vantage SE at 530 nm and analyzed.

IHC

Ovaries from all the four groups were dissected, freed of fat, and fixed in Bouin's fixative for 24 hr. The tissues were processed for histology. Five-μm sections were cut on a microtome (RM 2125RT; Leica, Wetzlar, Germany) and mounted on slides coated with 0.1% poly-l-lysine (Sigma; St. Louis, MO). The IHC localization was carried out for different antibodies for apoptotic markers Fas, cytochrome c, caspases 3, 8, and 9, and BID. The antigen retrieval of the deparaffinized sections was performed using citrate buffer (pH 6.0) in a microwave at high for 3 min. Sections were quenched in 0.3% (v/v) H₂O₂ in methanol at −20C for 20 min to block endogenous peroxidase. The details of blocking and primary antibody treatment are given in Table 1. Sections with antibody to Fas were incubated at 37C for 1 hr, and the rest were incubated with respective primary antibodies (Table 1) at 4C overnight. After a brief wash in 0.01 M PBS, incubation was done with horseradish peroxidase (HRP)-labeled appropriate secondary antibodies as mentioned in Table 1. The HRP was visualized by reaction with DAB (Sigma) and 0.3% (v/v) H₂O₂ in PBS as substrate mix for 2–3 min, and the sections were counterstained with hematoxylin, dehydrated in xylene, and mounted in Canada balsam for microscopic examination. Primary antibody was omitted in negative controls and substituted with normal goat serum/SP₂O medium depending on the use of polyclonal or monoclonal primary antibody, respectively. Quantitative comparison of staining intensities in granulosa cells of the four groups for immunolocalization of Fas, caspase 8, cytochrome c, caspase 9, BID, and caspase 3 was made using Biovis Image Analysis Software (Expert Vision Laboratories; Mumbai, India), and results are expressed in terms of mean absorbance.

Table 1.

Details of blocking, primary antibodies, and secondary antibodies used for immunolocalization of Fas, caspase 8, cytochrome c, caspase 9, BID, and caspase 3

Marker	Blocking	Primary antibody (dilution used)	Secondary antibody (dilution used)
Fas	Superblock (Pierce; Rockford, IL)	Mouse monoclonal (1:50) (Novocastra; Newcastle Upon Tyne, UK)	HRP-anti-mouse rabbit polyclonal (1:100) (Dako; Glostrup, Denmark)
Caspase 8	Superblock (Pierce)	Mouse monoclonal (1:50) (Calbiochem; Germany)	HRP-anti-mouse rabbit polyclonal (1:100) (Dako)
Cytochrome c	5% NFDM (Aarey; Mumbai, India)	Rabbit polyclonal (1:80) (Santa Cruz Biotechnology; Santa Cruz, CA)	Swine anti-rabbit biotinylated antibody (1:200) (Dako)
Caspase 9 and BID	Normal donkey serum (Immunocruz kit; Santa Cruz Biotechnology)	Goat polyclonal (1:80) (Santa Cruz Biotechnology)	Donkey anti-goat biotinylated antibody (prediluted) (Immunocruz kit; Santa Cruz Biotechnology)
Caspase 3	Superblock (Pierce)	Mouse monoclonal (1:25) (Calbiochem)	Swine anti-rabbit biotinylated antibody (1:200) (Dako)

HRP, horseradish peroxidase; NFDM, non-fat dry milk.

Electron Microscopy

Ovaries were fixed in modified Karnovsky's fluid (Karnovsky 1965) for 4–6 hr. The tissues were rinsed twice in 0.1 M sodium cacodylate buffer at 4C. After postfixation with 1% osmium tetroxide, the tissues were dehydrated in ascending grades of acetone and embedded in Araldite resin (Pelco International; Clovis, CA). Ultrathin sections (60–70 nm) were cut on a UCT-R ultramicrotome (Leica) and picked up on uncoated copper grids (200 mesh). The sections were stained with uranyl acetate and subsequently contrasted with lead citrate. The grids were viewed in the transmission electron microscope Philips Tecnai G²12 (Eindhoven, The Netherlands) at an accelerating voltage of 80 KV. The images captured were analyzed with version 3.1 of the SIS software (Eindhoven, The Netherlands) supplied by the manufacturer.

Three grids per group were scanned, and at least six granulosa cells of each cell type, based on their location, antral, cumulus, and basal, of the normal and atretic- and OP-treated groups were compared by statistical analysis. Changes in oocyte and oocyte–granulosa cell interaction were also observed and analyzed.

Statistical Analysis

The data for all the experiments are represented as mean ± SEM from three independent experiments for the normal and atretic- and hGF₂/OP-treated groups. The statistical analysis was performed by Student's t-test, and p<0.05 was taken as significant.

Results

Flow Cytometric Analysis

Presence of the Death Receptor Fas on Granulosa Cells

Fas, a cell surface receptor protein belonging to the tumor necrosis factor (TNF)-α receptor family mediates apoptosis-inducing signals (Sakamaki et al. 1997). The anti-Fas antibody labeled to FITC identifies the cell population positive for Fas. It was observed that there was a significant increase (p<0.05) in the apoptotic cell population in the atretic- (7.28 ± 0.7) and hGF₂/OP-treated (10.0 ± 1.1/9.67 ± 2.1, respectively) groups (Figure 1) compared with normal (0.79 ± 0.01) cells expressing the receptor.

Figure 1.

Comparative percentage of apoptotic cell population as identified by expression of death receptor Fas (A), change in mitochondrial membrane potential (B), and expression of active caspase 3 (C) in granulosa cells from normal and atretic- and human gel filtration fraction-2(hGF₂)/octapeptide (OP)-treated groups by flow cytometry.

Detection of ΔΨ_m–R123

It was observed that there was a significant (p<0.05) increase in the R123_low population in atretic- (47.08 ± 7.2; p<0.05) and hGF₂/OP-treated cells (40.75± 10.5/42.71 ± 0.5, respectively) compared with that in cells from the normal group (21.03 ± 5.2). It can be noted that the percentage increase in the apoptotic cell proportion in the atretic group was higher than that in the peptide-treated groups.

Apoptotic Population by Anti-caspase 3 Antibodies

Both the Fas-FasL and mitochondrial pathways converged at the activation of caspase 3, the ultimate effector caspase. The anti-caspase 3 antibody FITC binds to the active form in the cells that are in late apoptotic phase. Thus, the percentage of cells that showed green fluorescence was in late apoptotic phase. There was a significant (p<0.05) increase in the caspase 3–activated population in atretic- (5.64 ± 0.3) and hGF₂/OP-treated (12.59 ± 0.8/9.44 ± 0.1, respectively) cells (Figure 1) compared with normal cells (4.09 ± 0.2).

IHC Localization

Immunolocalization of the apoptotic markers studied, Fas, caspase 8, cytochrome c, caspase 9, BID, and caspase 3, was observed in granulosa cells of the atretic-, hGF₂-, and OP-treated groups. Staining intensities were quantified and are depicted in Table 2. It was observed that expression of all the apoptotic markers was not significantly expressed in normal group compared with the negative control run parallel in each experimental set. The comparison is thus done between atretic/hGF₂/OP-treated groups and the normal group.

Table 2.

Quantitative image analysis for IHC staining (OD) of apoptotic markers in normal and atretic- and peptide-treated groups

Marker	Normal	Atretic	GF2 treated	OP treated
Fas	6.36 ± 0.75	21.36 ± 3.36a	21.05 ± 1.15b	22.01 ± 2.16b
Caspase 8	7.78 ± 0.51	13.83 ± 1.82c	11.04 ± 0.84a	9.81 ± 0.91a
Cytochrome c	11.61 ± 1.06	21.67 ± 1.61d	17.25 ± 1.38c	22.09 ± 2.00e
Caspase 9	10.21 ± 1.22	19.01 ± 0.52e	15.38 ± 1.76c	15.53 ± 1.11a
BID	14.91 ± 1.17	26.41 ± 1.07d	22.14 ± 2.15d	23.01 ± 1.22d
Caspase 3	7.29 ± 0.56	25.16 ± 2.94b	17.56 ± 1.15d	21.30 ± 1.83d

^ap<0.01.

^bp<0.005.

^cp<0.05.

^dp<0.0001.

^ep<0.001.

OD, optical density; GF₂, gel filtration fraction-2; OP, octapeptide.

A significant change in localization of the death receptor Fas (Figure 2) was observed in the atretic- and peptide-treated groups compared with the normal group. Furthermore, quantification of the staining intensities showed a significant visible enrichment of Fas in the peptide-treated groups (p<0.005) compared with the atretic group (p<0.01). A significant expression of caspase 8 (Figure 2) was observed in the peptide-treated groups (p<0.01) and in the atretic group (p<0.05).

Figure 2.

IHC localization of apoptotic markers using specific antibodies to Fas (A–D), caspase 8 (E–H), cytochrome c (I–L), caspase 9 (M–P), BID (Q–T), and caspase 3 (U–X) in ovaries of normal and atretic-, hGF₂-, and OP-treated ovarian follicles (N, negative control). A significantly high expression of apoptotic markers was observed in the atretic- and peptide-treated groups compared with the normal group. Bar = 50 μm.

The immunolocalization of cytochrome c and caspase 9 (Figure 2), which represent the mitochondrial pathway, was studied. It was observed that the expression of these markers was more significant (p<0.0001 and p<0.001, respectively) in the atretic group compared with the peptide-treated groups (Table 2). Interestingly, a conspicuous visible enrichment of caspase 9 was observed in the atretic group compared with the peptide-treated group. Thus, in the atretic group, the mitochondrial pathway of apoptosis is predominant.

The localization of BID and caspase 3 (Figure 2) was observed in all three groups and was highly significant compared with the normal group (Table 2).

Electron Microscopic Observations

The transmission electron microscopic examination of the ovarian tissue from the atretic and OP-treated groups showed a change in gross morphology exhibiting characteristic features of apoptosis-like membrane blebbing, chromatin condensation, and vacuolization (Figure 3). The observations are focused on three cell types wherein the membrana granulosa cells lining the antrum are designated as antral granulosa cells, those closer to the basal lamina as basal granulosa cells, and the cells in close proximity to the oocyte as cumulus granulosa cells. A significant decrease in size was observed in the atretic antral granulosa cells (p<0.005) compared with those in the normal group (Table 3). In the OP-treated group, cumulus granulosa cells showed a significant decrease (p<0.05) in cell size (Table 3). Although a decrease in size was noted in the apoptotic cells, there was no change in the nucleus to cytoplasm ratio, suggesting that the shrinkage occurs in both the nucleus and the cytoplasm because of apoptotic changes.

Figure 3.

Representative ultramicrographs of basal (A–C), antral (D–F), and cumulus (G–I) granulosa cells and oocyte–granulosa cell interaction (J–L) in follicles from normal and atretic- and OP-treated ovaries of mice (O, oocyte; GC, granulosa cells). Characteristic apoptotic features like the intense vacuolation and condensation of the nucleus can be observed in basal granulosa cells of atretic- and OP-treated follicles. Condensation of nucleus can also be observed in antral granulosa cells of atretic follicles. Bar = 2000 nm.

Table 3.

Comparison of zona pellucida thickness, oocyte diameter, nucleus:cytoplasm ratio, and cell size of granulosa cells from normal and atretic- and OP-treated ovarian follicles as observed by transmission electron microscopy

	Normal	Atretic	OP treated
Zona pellucida thickness(um)	2.87 ± 0.06	3.26 ± 0.17	4.96 ± 0.47a
Oocyte diameter (um)	46.37 ± 1.45	45.36 ± 0.65	46.56 ± 2.00
Nucleus /cytoplasm ratio
Antral GC	0.45 ± 0.02	0.39 ± 0.06	0.30 ± 0.04
Cumulus GC	0.44 ± 0.03	0.48 ± 0.02	0.42 ± 0.05
Basal GC	0.50 ± 0.04	0.42 ± 0.05	0.46 ± 0.05
Cell size
Antral GC	47.92 ± 2.77	26.97 ± 4.04b	50.66 ± 6.20
Cumulus GC	51.64 ± 3.71	50.69 ± 3.37	40.74 ± 2.69a
Basal GC	38.10 ± 5.47	29.54 ± 5.02	37.64 ± 3.59

^ap<0.05.

^bp<0.002.

OP, octapeptide.

A significant decrease in the size of mitochondria was observed in the cumulus granulosa cells of the atretic group (p<0.005), and in the antral granulosa cells of the OP-treated group when compared with the respective cell types in the normal group (p<0.005). A significant decrease (p<0.05) in the number of mitochondria was observed in the cumulus granulosa cells of the atretic group and the basal granulosa cells of both the atretic- and OP-treated groups.

Interestingly, a significant (p<0.05) increase in the width of the zona pellucida of the ovarian follicles of the OP-treated group was observed. A remarkable loss in the transzonal oocyte–granulosa cell communication [transzonal projections (TZPs)] was also noted. There was no change in the oocyte diameter in all three groups.

Discussion

Apoptosis in granulosa cells occurs by two main pathways: the Fas–Fas L receptor pathway and the mitochondrial pathway. The membrane receptor pathway of apoptosis is initiated by binding of Fas L to the membrane-associated Fas, leading to stimulation of the Fas-associated death domain, inducing a cascade of events causing activation of caspase 3, which finally ends in apoptosis (Ashkenazi and Dixit 1998). On the other hand, in the mitochondrial pathway, imbalance in pro- and anti-apoptotic bcl2 family members causes a drastic change in mitochondrial membrane potential, which in turn leads to a change in membrane permeability. This induces the release of membrane-bound apoptogenic proteins such as cytochrome c, which leads to activation of caspases (Green and Reed 1998).

FSH is one of the key survival factors for growth and development of ovarian follicles. Action of FSH is directly or indirectly modulated by various intraovarian factors. Our group has identified one such novel partially purified intraovarian peptide that inhibits the binding of FSH to granulosa cells. OP, the partial N-terminal eight amino acid synthetic peptide, has similar biological action as the native peptide (Wadia et al. 2003). hGF₂ has been shown to induce apoptosis in ovarian follicles (Nandedkar et al. 1996,1998). The aim of this study was to elucidate the mechanism of action of these peptides. Specific markers of the two pathways were used to study the apoptotic changes in the ovary by two techniques: flow cytometry and IHC.

The results showed that an overall change in expression of apoptotic markers Fas and caspase 8 (Fas–Fas L pathway) and ΔΨ_m, cytochrome c, and caspase 9 (mitochondrial pathway) was observed in granulosa cells of the atretic- and hGF₂/OP-treated groups. Here, BID, which links the two pathways, and caspase 3, the effector caspase, were also studied. A significantly higher expression (p<0.0001) of BID in the atretic- and peptide-treated groups suggested interlinkage of the two pathways of the death mechanism in these groups. Also a significant increase in caspase 3 activation in atretic- (p<0.005) and peptide-treated (p<0.0001) groups compared with the normal group confirmed induction of apoptosis in growing follicles. Interestingly, a specific pattern of these markers was observed. It can be noted from the results that expression of Fas and caspase 8 was higher in peptide-treated groups compared with that in the atretic group, whereas the expression of mitochondrial pathway markers was vice versa in the atretic group. These results may underline the fact that apoptosis in the atretic group is initiated by intrinsic signals, whereas that in the peptide-treated groups is by extrinsic stimuli.

The initiation of apoptosis in atretic- and peptide-treated groups is distinct; this is possibly because of the difference in stimulatory mechanisms. eCG is a combination of gonadotropins, FSH and LH, and therefore acts as a natural stimulant during follicular growth. In the absence of human chorionic gonadotropin (hCG), no ovulation occurs, and the follicles undergo atresia, mimicking the normal phenomenon. On the other hand, hGF₂/OP is an exogenous inducer of apoptosis following the membrane receptor pathway. Interestingly, injections of hGF₂/OP do not cause necrosis of granulosa cells as assessed by propidium iodide staining in viable cells (data not shown). This can be explained because, although hGF₂/OP were injected exogenously, hGF₂/OP are purified from ovarian follicular fluid and are secreted by granulosa cells (Nandedkar et al. 1992).

Because a difference in initiation of apoptosis was observed in the atretic- and peptide-treated groups, an attempt has been made to study the morphological changes at the ultrastructural level. OP has a similar biological action as hGF₂ and also exhibits a similar action in granulosa cells; therefore, the ultrastructural studies were conducted in only the OP-treated group.

It has been reported that antral atresia fits the classic description of atresia wherein apoptosis is initiated near to the antrum and progresses toward the basal lamina (Irving-Rodgers et al. 2001). Our ultrastructural studies showed a similar prototype wherein apoptosis was observed in antral granulosa cells of the atretic group. However, in the OP-treated group, maximum apoptosis was observed in cumulus granulosa cells. This observation may corroborate our earlier findings that there is a drastic decrease in progesterone (P₄) secretion by granulosa cells when cultured in the presence of OP (Wadia et al. 2007). Khamsi and Roberge (2001) reported that cumulus granulosa cells have the potential to secrete six times more progesterone compared with antral or basal granulosa cells.

Disappearance of oocyte microvilli and granulosa cell projections from the zona pellucida of atretic follicles reported here has been supported by earlier studies (Apkarian and Curtis 1981). This leads to decreased communication between oocyte and granulosa cells. A thickened zona pellucida was also observed in the OP-treated group, which may possibly be caused by lipid accumulation as suggested by Guraya (1966).

In conclusion, these studies showed that the difference in the pattern of apoptosis in the atretic- and peptide-treated groups may be caused by the difference in initiation of the apoptotic pathway, spontaneously by deprivation of hormone in the atretic group, and by induction of the membrane receptor pathway in the peptide-treated groups. Moreover, the ultrastructural studies also supported the difference in the initiation of the apoptotic pathways being antral granulosa cell atresia in the atretic group and cumulus granulosa cell atresia in the peptide-treated groups.