Analysis of apoptosis and expression of bcl-2 gene family members in the human and baboon ovary

Recent data support a role for apoptosis, under tight regulatory control by bcl-2, oxidative stress response, tumor suppressor, and CASP gene family members, in mediating granulosa cell demise during follicular atresia in the rodent and avian ovary. Herein we evaluated the occurrence of apoptosis in the human and baboon ovary relative to follicular health status, and analyzed expression of several cell death genes in these tissues. In situlocalization of DNA strand breaks in fixed human and baboon ovarian tissue sections indicated that apoptosis was essentially restricted to granulosa cells of atretic antral follicles. Biochemical analysis of DNA oligonucleosomes in individual follicles isolated from baboon ovaries during the ovulatory phase revealed the presence of apoptotic DNA fragments in subordinate but not dominant follicles, thus substantiating the in situ labeling studies. Messenger RNA transcripts encoded by the bax death susceptibility gene, the bcl-xlong survival gene, the bcl-xshort pro-apoptosis gene, the p53 tumor suppressor gene, and two members of the CASP gene family (CASP-2/Ich-1, CASP-3/CPP32), were detected by Northern blot analysis of total RNA prepared either from human ovaries or from Percoll-purified granulosa-lutein cells obtained from patients undergoing assisted reproductive technologies. Lastly, immunohistochemical localization of the BAX death-susceptibility protein in the human ovary revealed abundant expression in granulosa cells of early atretic follicles, whereas BAX protein was extremely low or non-detectable in healthy or grossly-atretic follicles. We conclude that apoptosis occurs during, and is probably responsible for, folicular atresia in the human and baboon ovary. Moreover, apoptosis in the human ovary is likely controlled by altered expression of the same cohort of cell death regulatory factors recently implicated as primary determinants of apoptosis induction or suppression in the rodent ovary.

Gene regulation of interleukin-1 beta, interleukin-1 receptor type I, and plasminogen activator inhibitor-1 and -2 in human granulosa-luteal cells

OBJECTIVE

To investigate the regulation of messenger ribonucleic acid (mRNA) levels of interleukin-1 beta (IL-1 beta), interleukin-1 (IL-1) receptor type 1, and plasminogen activator (PA) inhibitor-1 and -2 in cumulus cells, granulosa-luteal cells, and macrophage-depleted granulosa-luteal cells obtained from human preovulatory follicles.

DESIGN

Prospective longitudinal study.

SETTING, PATIENTS

Patients undergoing assisted reproductive technologies (ART) in the Department of Gynecology and Obstetrics, Stanford University, Stanford, California.

INTERVENTIONS

Cumulus cells and granulosa-luteal cells were collected by ultrasound-guided transvaginal aspiration at the time of ART.

MAIN OUTCOME MEASURES

Northern blot analysis of mRNA levels of IL-1 beta, IL-1 receptor type 1, PA inhibitor-1 and -2 in cumulus cells, granulosa-luteal cells and macrophage-depleted granulosa-luteal cells, and indirect immunocytochemical analysis of the IL-1 system and macrophages in granulosa-luteal cell preparations were performed.

RESULTS

Interleukin-1 beta mRNA levels in uncultured cumulus cells were less than those of uncultured granulosa-luteal cells with no differences in IL-1 receptor type 1 mRNA levels between these two cell types. Granulosa-luteal cell IL-1 receptor type 1 mRNA levels were expressed constitutively throughout 24 hours of culture with no effect by hCG, whereas IL-1 beta mRNA levels increased within 6 hours, and then remained elevated for 24 hours with no effect by hCG. Interleukin-1 beta significantly increased granulosa-luteal cell mRNA levels of IL-1 beta (over twofold), IL-1 receptor type 1 (over twofold), PA inhibitor-1 (approximately 1.4-fold), and PA inhibitor-2 (approximately 1.6-fold). In contrast, IL-1 beta had no effect on IL-1 beta and IL-1 receptor type 1 mRNA levels in macrophage-depleted granulosa-luteal cells. Granulosa-luteal cells, not macrophages, account for the majority of the immunocytochemical staining for IL-1 beta and IL-1 receptor type 1 in follicular aspirates.

CONCLUSIONS

These studies suggest that the IL-1 system is regulated in human granulosa-luteal cells during the periovulatory period. Furthermore, the augmentation of PA inhibitor-1 and -2 mRNA levels by IL-1 beta suggests a potential role for IL-1 beta in remodeling of the human ovary during the periovulatory period.

Interleukin-1 receptor antagonist suppresses human chorionic gonadotropin-induced ovulation in the rat

Indirect evidence has implicated the interleukin-1 (IL-1) system in ovulation. Thus, the ability of IL-1 beta to induce ovulation in rat and rabbit perfused ovaries has been demonstrated. In the present study, the involvement of the IL-1 system in ovulation was directly tested in vivo, in the rat model. For this purpose, the natural inhibitor of the IL-1 system, interleukin-1 receptor antagonist (IL-1ra), was administered locally by use of an intrabursal injection route. Twenty-six-day-old Sprague-Dawley rats received injections of eCG (10 IU), followed 56 h later by hCG (15 IU). IL-1ra (75 micrograms/bursa) was administered locally into the periovarian sac, 6 h (n = 5), 2 h (n = 11), and 0 h (n = 5) before hCG administration. Control animals (n = 10) received injections of the same volume (50 microliters) of vehicle (PBS). IL-1ra administered locally into the periovarian sac inhibited ovulation from the treated ovary, reaching 40% inhibition (p < 0.05) when injected 2 h prior to hCG, as compared to the untreated contralateral ovary (6 +/- 1.4 ova vs. 10 +/- 1.8 ova) and PBS-injected control ovaries (6 +/- 1.4 ova vs. 8.2 +/- 0.7). Injection of IL-1ra 6 h before or concomitantly with hCG did not affect the ovulation rate. Internucleosomal DNA fragmentation was evaluated by 3' end-labeling and autoradiography for detecting apoptotic changes. No difference in DNA fragmentation was found between treated and untreated ovaries.(ABSTRACT TRUNCATED AT 250 WORDS)

Detection of apoptosis in human and rat ovarian follicles

OBJECTIVE

The purpose of this study was to determine whether cells acquired from individual human preovulatory follicles undergo apoptosis (physiologic cell death) and, if so, to correlate the degree of apoptosis with characteristics of the follicles or the oocytes derived from the follicles.

METHODS

We devised a sensitive nonradioactive method for detecting apoptotic DNA fragmentation in small numbers of cells derived from rat atretic follicles and follicular aspirates of patients undergoing assisted reproductive technologies.

RESULTS

Using this method, apoptotic DNA was detected in rat atretic follicles, with optimal detection at 10-100 ng. Furthermore, apoptotic DNA was detected in some, but not all individual human follicular aspirates from several patients, and was found in follicles that produced oocytes that fertilized and developed into embryos.

CONCLUSION

Apoptosis occurs in cells from human ovarian preovulatory follicles and may be a normal physiologic process of the follicle during luteinization.

The effect of interleukin-1 beta (IL-1 beta) on the regulation of IL-1 receptor type I messenger ribonucleic acid and protein levels in cultured human endometrial stromal and glandular cells

Because we hypothesize that the interleukin-1 (IL-1) system may be important in the dialogue between mother and embryo during the implantation process, we have analyzed the effect of IL-1 beta, a secretory product of the human embryo and human endometrium, on the mRNA and protein levels of IL-1 receptor type I (IL-1R tI) in the human endometrium. For this purpose, endometrial epithelial cells (EEC) and stromal cells (ESC) were isolated and cultured with progesterone (3.18 micrograms/mL) and epidermal growth factor (20 ng/mL) for 8 days in the presence or absence of hrIL-1 beta (20 pg/mL). EEC from proliferative and secretory endometrium expressed high levels of IL-1R tI mRNA compared to ESC, and these levels were not modulated by IL-1 beta. However, prostaglandin E2 levels peaked on day 4 in EEC treated with progesterone, epidermal growth factor, and IL-1 beta (208.7 +/- 92 ng/10(7) cells), whereas no prostaglandin E2 was detectable in cells not treated with IL-1 beta, indicating that these cells responded to IL-1 beta. With regard to ESC from secretory endometrium, IL-1 beta increased its own receptor mRNA levels (4 +/- 0.5-fold increase) after 8 days in culture. However, when ESC were isolated from proliferative endometrium, an up-regulation of IL-1R tI (3.5 +/- 0.5-fold increase) was observed on days 6 and 8 of culture regardless of the presence or absence of IL-1 beta. Immunoreactive IL-1R tI was identified in cultured EEC and ESC, and patterns similar to those of mRNA were observed. The constitutive presence of IL-1R tI in EEC, which was not affected by IL-1 beta, and the up-regulation of IL-1R tI mRNA by its ligand IL-1 beta in ESC isolated during the luteal phase suggest a role for the IL-1 system in human implantation.

Embryonic implantation in mice is blocked by interleukin-1 receptor antagonist

We have investigated the relevance of interleukin-1 receptor type I (IL-1R tI) in the implantation process in vivo in a murine model. Indirect immunofluorescence experiments demonstrate that IL-1R tI is located in mouse endometrial lumenal epithelium with increased intensity in the periimplantation period, whereas IL-1 beta staining is located in the mouse placenta. PMSG/human CG (hCG)-stimulated and mated 12-week-old B6C3F-1 female mice were randomly allocated to three groups: A, control noninjected; B, buffer-injected animals; and C, animals injected ip with 20 micrograms recombinant human IL-1 receptor antagonist (rhIL-1ra) every 12 h beginning on pregnancy day 3. Injections were continued until day 9, and animals were killed 12 h after the last injection. Pregnancy rates in the three groups were: noninjected, 58.8% (10 of 17); buffer-injected, 73.7% (14 of 19); rhIL-1ra-injected, 6.7% (1 of 15), P = 0.0001155, Fisher exact test. To rule out the possibility that pregnancy failure was due to an embryotoxic effect of rhIL-1ra, 2-cell mouse embryos (n = 276) were flushed from the same group of animals used for in vivo experiments and cultured with increasing concentrations of rhIL-1ra: 0 microgram/ml (n = 91), 1 microgram/ml (n = 36), 50 micrograms/ml (n = 36), 100 micrograms/ml (n = 52), and 200 micrograms/ml (n = 61) rhIL-1ra. The percentages of 2-cell mouse embryos reaching the blastocyst stage after 72 h in culture were 85.7%, 91.6%, 94.4%, 96%, and 85.2%, respectively. We further cultured these blastocysts for 5 days on fibronectin-coated plates with or without 200 micrograms/ml rhIL-1ra. In both groups, hatching, attachment to fibronectin, outgrowth, and migration were documented to be similar. Furthermore, our longitudinal morphological study of embryonic implantation in control and rhIL-1ra-injected mice shows that the blockade of IL-1R tI interferes with the attachment of mouse blastocysts to maternal endometrium in vivo. In summary, we demonstrate that blockade of maternal endometrial IL-1R tI with IL-1ra prevents implantation in the mouse by interfering with embryonic attachment, without adverse effects on blastocyst formation, hatching, fibronectin attachment, outgrowth, and migration in vitro.

Localization of interleukin-1 type I receptor and interleukin-1 beta in human endometrium throughout the menstrual cycle

Previous studies in the human suggest that the interleukin-1 (IL-1) system, may be an important paracrine/autocrine mediator in local intercellular interaction in endometrial tissue. In this study we have determined that IL-1 receptor type I (IL-1R tI) is expressed at the messenger RNA (mRNA) and protein levels in glandular cells and its ligand, IL-1 beta has been localized by immunohistochemical methods in endothelial cells and isolated stromal cells in the human endometrium throughout the menstrual cycle. IL-1R tI mRNA was detected in glandular epithelium using both specific complementary DNA and complementary RNA 32P-labeled probes. Human glandular epithelium contains a 5.1-kilobase mRNA transcript throughout the complete menstrual cycle. Quantitative densitometric analysis of slot blot hybridization signals shows an increase of IL-1R tI mRNA in both early and mid-late secretory phases in comparison with the proliferative phase (P < 0.05). IL-1R tI protein was localized in endometrial glandular epithelial cells using both indirect immunofluorescence and avidin-biotin-peroxidase methods. However, more intense staining for IL-1R tI was observed in lumenal epithelial cells compared with the staining present deep in the endometrial glands. Using the same methods, IL-1 beta was detected in endothelial cells of spiral vessels and isolated stromal cells throughout the menstrual cycle, and an increased staining from proliferative to secretory phase was observed. The detection of IL-1R tI in the human endometrial epithelium and its ligand, IL-1 beta, in isolated stromal cells and endothelial cells, is another example of possible communication between the immune and reproductive systems with special relevance to human implantation.

Interleukin-1 type I receptor messenger ribonucleic acid expression in human endometrium throughout the menstrual cycle

OBJECTIVE

To investigate the messenger ribonucleic acid (mRNA) expression of interleukin-1 (IL-1) type I receptor in the endometrial tissue of normal patients during the menstrual cycle.

DESIGN

Prospective longitudinal study.

SETTING

Department of Obstetrics and Gynecology, Stanford University Medical Center, Stanford, California.

PATIENTS

Twenty fertile women between 19 and 41 years of age underwent hysterectomy for benign reasons (n = 9) and laparoscopy for tubal ligation (n = 11). In all cases, endometriosis was not visualized.

INTERVENTIONS

Endometrial biopsy using the Novak curette was obtained at the time of surgery.

MAIN OUTCOME MEASURE

Total RNA extracted from unfractioned endometrial tissue was analyzed on Northern blots by using specific complementary deoxyribonucleic acid probes.

RESULTS

We found IL-1 type I receptor mRNA expression in endometrial tissue throughout the entire menstrual cycle. However, IL-1 type I receptor mRNA levels were significantly higher during both early and late luteal phases than follicular and midluteal phases.

CONCLUSIONS

Our results demonstrate the presence of the IL-1 system in the human endometrium and that the receptor is regulated throughout the menstrual cycle with a 4.1-fold increased expression of the IL-1 receptor gene in the early luteal phase compared with preovulatory endometrium.

Regulation of plasminogen activator inhibitor-1 and -2 messenger ribonucleic acid levels in human cumulus and granulosa-luteal cells

Plasminogen activators and their inhibitors have been implicated in the process of fibrinolysis, tissue remodeling, and ovulation. Epidermal growth factor (EGF), a paracrine hormone found in the human ovary, increases plasminogen activator (PA) activity and the gene expression of PA and plasminogen activator inhibitor (PAI) in human endothelial cells and human cell lines. Gonadotropins also increase PA activity and gene expression in rat preovulatory granulosa cells. We have now analyzed the gene expression of PAI-1 and PAI-2 in uncultured human cumulus cells (CC), uncultured granulosa-luteal cells (GLC), and cultured GLC obtained from preovulatory follicles of patients undergoing assisted reproductive technologies. We also studied the effects of hCG and EGF on PAI-1 and PAI-2 mRNA levels in cultured GLC; GLC were cultured in serum-free medium for various times within 24 h with or without hCG and for 6 h with or without hCG, EGF, or EGF plus hCG. Total RNAs from CC and GLC were extracted, and blot hybridizations with 32P-labeled PAI-1, PAI-2, or 28S ribosomal RNA cDNA probes were performed. Both CC and GLC expressed PAI-1 and PAI-2 genes. In GLC, steady state levels of PAI-1 mRNA levels steadily increased within 24 h of culture, whereas PAI-2 levels peaked at 6 h of culture. PAI-1 mRNA levels were not affected by hCG or EGF at 6 h of culture, but PAI-2 mRNA levels were significantly increased by EGF at 6 h of culture. These studies demonstrate that human GLC PAI-1 and PAI-2 mRNA levels are differentially regulated and suggest that EGF may be involved in modulation of the human ovarian PA system during the periovulatory period.

Regulation of luteinizing hormone receptor messenger ribonucleic acid levels by gonadotropins, growth factors, and gonadotropin-releasing hormone in cultured rat granulosa cells

The induction of LH receptors in granulosa cells is prerequisite for ovarian follicles to ovulate and form corpora lutea. Earlier studies have demonstrated the modulatory role of gonadotropins, growth factors, and GnRH on ovarian LH receptor content. We have now analyzed the influences of gonadotropins (FSH, LH, and PRL), several growth factors, and GnRH on LH receptor mRNA levels in cultured granulosa cells. Cells were obtained from immature estrogen-treated rats and cultured in medium containing FSH with or without growth factors or GnRH for 48 h. Some cells were also treated with FSH for 48 h, followed by treatment with FSH, LH, or PRL for another 2 days. Cellular total RNA was extracted, and blot hybridization with 32P-labeled LH receptor cRNA or 28S ribosomal RNA cDNA probes was performed. Treatment of granulosa cells with FSH increased the levels of five species of LH receptor mRNAs in a dose- and time-dependent manner. In FSH-primed cells, LH receptor mRNA levels were maintained by FSH, LH, and PRL. In contrast, treatment of cells with basic fibroblast growth factor or epidermal growth factor suppressed FSH induction of LH receptor mRNA in a dose-dependent manner, whereas treatment with insulin-like growth factor-I had no effect. In addition, GnRH suppressed FSH-stimulated LH receptor mRNA levels in a dose-dependent manner; the effects of GnRH could be counteracted by coincubation with a GnRH antagonist, suggesting mediation by specific GnRH-binding sites. These studies demonstrated that the observed stimulatory effects of gonadotropins (FSH, LH, and PRL) and the inhibitory effects of growth factors (epidermal growth factor and basic fibroblast growth factor) and GnRH on LH receptor content are correlated to their regulation of LH receptor mRNA levels. The granulosa cell culture system should provide a useful model for studying LH receptor gene regulation.

Equine granulosa-theca cell tumors express inhibin alpha- and beta A-subunit messenger ribonucleic acids and proteins

The association of equine granulosa-theca cell tumors with atrophied contralateral ovaries and abnormal estrous cycles suggests that these tumors produce hormones that affect pituitary gonadotropin production. Because inhibin, a heterodimer protein secreted by granulosa cells, decreases FSH production, we examined the presence of inhibin alpha- and beta A-subunits and their mRNAs in ovarian tumors obtained from three mares. These tumors contained neoplastic cords and nodules, multiple fluid-filled cysts, and a predominance of neoplastic granulosa cells. Reduced proteins from tumor-conditioned media were analyzed by electrophoresis and immunoblotting using antibodies directed against peptide fragments of the alpha- and beta A-chains of porcine inhibin. Specific bands at 50-kDa and 36-kDa for the inhibin alpha-subunit and at 44 kDa and 13 kDa for the inhibin beta A-subunit were observed in these tumors. Northern blot hybridization of 32P-labeled rat inhibin alpha- and beta A-subunit complementary RNAs to total RNA from each tumor revealed predominant bands of activity in all three tumors at 1.5 and 7 kb for the alpha- and beta A-subunit mRNAs, respectively. These results demonstrate that equine granulosa-theca cell tumors express the mRNAs for inhibin alpha- and beta A-subunits and also secrete inhibin subunits that could potentially affect gonadotropin production in afflicted mares. Furthermore, cells derived from these tumors may provide a useful model for understanding inhibin gene regulation and ovarian tumorigenesis.

Regulation of inhibin subunit messenger ribonucleic acid levels by gonadotropins, growth factors, and gonadotropin-releasing hormone in cultured rat granulosa cells

Cultured rat granulosa cells have provided a useful model to examine the hormonal regulation of inhibin secretion. In the present study we have used the cloned rat inhibin alpha- and beta A-subunit cDNAs to characterize the influences of gonadotropins, growth factors, and GnRH on inhibin subunit mRNA levels in granulosa cells obtained from immature estrogen-treated rats. Cells were cultured in medium with or without added hormones. Total RNA from cultured cells was extracted and hybridized with 32P-labeled inhibin alpha- and beta A-subunit cRNA or beta-actin cDNA probes, and inhibin subunit mRNA levels were normalized with beta-actin mRNA levels. Treatment of granulosa cells with FSH increased inhibin alpha- and beta A-subunit mRNA levels in a dose-dependent manner. Similarly, LH, but not PRL, increased alpha- and beta A-subunit mRNA levels in granulosa cells pretreated with FSH to induce functional LH and PRL receptors. The effects of FSH and LH on inhibin subunit mRNA levels were mimicked by forskolin, which increased alpha- and beta A-subunit transcripts in a dose- and time-dependent manner, suggesting involvement of the cAMP-dependent protein kinase-A pathway. Since several growth factors have been shown to influence inhibin secretion, their effects on inhibin subunit mRNA levels were also studied. Treatment of cells with transforming growth factor-beta 1 increased both basal and FSH-stimulated inhibin alpha- and beta A-subunit mRNA content, whereas insulin-like growth factor-I had no significant effect. In contrast, both epidermal growth factor (EGF) and basic fibroblast growth factor (FGF) markedly suppressed both basal and FSH-stimulated inhibin subunit transcript levels. The inhibitory effects of EGF and basic FGF were dose dependent and persisted from 12-72 h of incubation. The regulatory peptide GnRH, which decreases inhibin secretion, was also found to suppress FSH-stimulated inhibin alpha- and beta A-subunit mRNA levels in a dose-dependent manner. Furthermore, the effects of GnRH could be counteracted by coincubation with a GnRH antagonist, suggesting the involvement of specific GnRH-binding sites in GnRH action. These studies indicate that, except for insulin-like growth factor-I, the effects of gonadotropins, growth factors (EGF, basic FGF, and transforming growth factor-beta 1), and GnRH on inhibin secretion are related to their regulation of inhibin alpha- and beta A-subunit mRNA levels.

Isolation and characterization of rabbit ovarian surface epithelium, granulosa cells, and peritoneal mesothelium in primary culture

Mammalian ovarian surface epithelial (OSE) cells and peritoneal mesothelial (PM) cells have a common embryologic origin, yet certain morphologic and histochemical characteristics are different in the adult. In this study, a two-step culture method was developed to examine the characteristics of these two cell types in vitro. OSE, PM, and ovarian granulosa (GC) cells were isolated from estrous rabbits and cultured for 6 d in 5% serum-supplemented D-valine medium (to inhibit fibroblast growth), then incubated for a further 2 d in serum-free McCoy's 5A medium. This study showed that rabbit OSE and PM cells in vitro maintained certain in vivo morphologic characteristics; OSE cells exhibited distinct cell borders and abundant microvilli of homogeneous size and shape, whereas PM cells were characterized by obscure cell borders and abundant microvilli of heterogeneous form. GC in vitro exhibited overlapping cell borders and sparse microvilli of homogeneous structure. This study showed for the first time that cultured rabbit OSE and PM cells, but not GC, contain distinct filaments of cytokeratin 18. In addition, rabbit OSE cells and GC, but not PM cells, contained 17 beta-hydroxysteroid dehydrogenase. However, only GC contained delta 5-3 beta hydroxysteroid dehydrogenase. OSE, PM, and GC maintained their ultrastructural and histochemical characteristics in serum-free medium. These results suggest that rabbit OSE cells in vitro could be distinguished from PM cells by histochemical and ultrastructural differences. Furthermore, because these characteristics were not altered in serum-free medium, the two-step culture method will be valuable in further hormonal studies of these cells in vitro.