Effects of a nitric oxide donor and nitric oxide synthase inhibitors on ...

Human Reproduction vol.14 no.10 pp.2537–2543, 1999

Effects of a nitric oxide donor and nitric oxide synthase inhibitors on luteinizing hormone-induced ovulation in the ex-vivo perfused rat ovary

Kenrokuro Mitsube1, Masato Mikuni, Markus Matousek and Mats Bra¨nnstro¨m Department of Obstetrics and Gynecology, Go¨teborg University, Sahlgrenska University Hospital, S-413 45 Go¨teborg, Sweden 1To

whom correspondence should be addressed

The aim of this study was to investigate the role of nitric oxide (NO) in ovulation and ovarian steroidogenesis by the use of NO synthase (NOS) inhibitors and an NO donor administrated to the luteinizing hormone (LH)-stimulated ex-vivo perfused pre-ovulatory rat ovary. The ovaries were stimulated with LH (0.2 µg/ml) alone or in combination with the phosphodiesterase inhibitor IBMX (200 µmol/l). The presence of both endothelial NOS (eNOS) and inducible NOS (iNOS) in the perfused rat ovary were detected by immunoblotting and a clear increase in amount of iNOS protein was seen after LHFIBMX stimulation. The addition of a non-selective NOS inhibitor, NG-monomethyl-Larginine (L-NMMA; 300 µmol/l), to the perfusate significantly decreased ovulation numbers (median J 4.0, range J 1–14) as compared with LH F IBMX stimulated control (12.0, 6–17). In contrast, an inhibitor with relative selectivity towards iNOS, aminoguanidine bicarbonate (AG, 300 µmol/l and 1 mmol/l), did not change the ovulation rate (11.5, 6–18 and 11.0, 7–15 respectively). In perfusions with only LH, a lower ovulation rate was seen but with similar effects (0.0, 0–8 for L-NMMA; 7.5, 3–12 for control and 7.0, 1–15 for AG 300 µmol/l). The administration of an NO donor, spermine NONOate, resulted in similar ovulation numbers as in LH-stimulated controls. The NO inhibitors did not affect steroid concentrations in the perfusion media, while 100 µmol/l NONOate increased progesterone production. Key words: nitric oxide/NOS inhibitor/ovarian perfusion/ ovulation/rat

Introduction Nitric oxide (NO) is an important paracrine mediator that has a variety of physiological functions, including regulation of vascular dilatation/permeability (Powers et al., 1995; Stones et al., 1995), neurotransmission (Snyder et al., 1992) and cytotoxic effects (Moncada et al., 1991; Palacois et al., 1993). Nitric oxide is produced from the oxidation of L-arginine by nitric oxide synthase (NOS), which exists in three isoforms: two constitutive forms, endothelial NOS (eNOS) and neuronal NOS (nNOS), and one inducible form (iNOS). Several lines of evidence suggest that NO is involved in © European Society of Human Reproduction and Embryology

various cycle-dependent ovarian events, such as ovulation and regulation of luteal function (Van Voorhis et al., 1994). The expression of eNOS and iNOS, but not nNOS, in the periovulatory rat ovary has been documented (Zackrisson et al., 1996a; Jablonka-Shariff and Olson, 1997). In the latter study it was demonstrated that expression of both eNOS and iNOS protein was mostly localized to the theca layer and increased after human chorionic gonadotrophin (HCG). The effects of inhibition of NOS have been studied both in vivo (Shukovski et al., 1994) and ex vivo (Bonello et al., 1996; Yamauchi et al., 1997) with observed reductions in the numbers of oocytes released. Nitric oxide also affects ovarian steroidogenesis. In studies using cell culture models, NO consistently inhibits oestradiol and progesterone production in a dose dependent manner while NOS inhibitors enhance steroid synthesis by granulosa–lutein cells (Van Voorhis et al., 1994; Olson et al., 1996). However, in the rat ovary perfusion model, in which the three-dimensional structure and cell-to-cell contacts are intact, contradictory results from the cell culture studies were observed (Bonello et al., 1996). The aims of the present study were to investigate further the role of the two ovarian expressed isoforms of NOS in ovulation and steroid production and to evaluate the effects of exogenously donated NO on ovulation and steroidogenesis stimulated by luteinizing hormone (LH). To exclude secondary systemic effects of NOS inhibitors and a NO donor, the method of ex-vivo ovarian perfusion was used.

Materials and methods Animals Immature female Sprague–Dawley rats (B&K Universal, Sollentuna, Sweden) were kept under controlled light (14 h light, 10 h dark) and had free access to pelleted food and water. All experiments were carried out according to the principles and procedures outlined in the National Institute of Health Guide for the Care and Use of Laboratory Animals, and were approved by the Animal Ethics Committee of Go¨teborg University. Hormones and chemicals Ovine luteinizing hormone (NIADDK-oLH-26) was kindly provided by the NIADDK and National Hormone and Pituitary Program (Rockville, MD, USA). Pregnant mares’ serum gonadotrophin (PMSG), 3-isobutyl-1-methylxanthine (IBMX), NG-monomethyl-Larginine (L-NMMA) and aminoguanidine bicarbonate (AG) were purchased from Sigma Chemical Company (St Louis, MO, USA). Spermine NONOate was supplied by Alexis Corporation (Leufelfingen, Switzerland); ketamine from Parke Davis (Barcelona, Spain); xylazine from Bayer (Leverkusen, Germany); medium 199 from GIBCO BRL (Rockville, MD, USA); gentamycin sulphate from Biological Industries (Kibbutz Beit Haemek, Israel); bovine serum

2537

K.Mitsube et al.

albumin (fraction V) from Boehringer Mannheim (Mannheim, Germany); insulin from Novo (Bagsvaerd, Denmark) and heparin was purchased from Lo¨vens (Ballerup, Denmark). Ovarian perfusion At 28 days of age, the rats were given 20 IU of PMSG s.c. to promote growth and maturation of a first generation of large antral follicles. On the morning of day 30, the rats were anaesthetized with ketamine and xylazine (40 and 6.5 mg/kg respectively), and 300 IU of heparin sulphate was injected i.v. through a femoral vein. Laparotomy was performed and the right ovary was surgically removed with its feeding and draining vessels as previously described in detail (Bra¨nnstro¨m et al., 1987). The bursa was gently opened and the ovary was placed in a perfusion chamber. The perfusion was performed in a recirculating system with 30 ml of medium (Medium 199 with Earle’s salts supplemented with 0.026 mol/l sodium bicarbonate, 0.2 IU/ml insulin, 50 µg/ml gentamycin sulphate and 4% bovine serum albumin). The perfusion pressure was maintained at 80 mm Hg and gassed with 5% CO2 and 95% O2. The perfusion was continued for 20 h after the administration of LH into the perfusion medium. Samples of medium were taken at 0, 1, 3, 5, 8 and 20 h time points and stored at –70°C for later analysis. At the end of the perfusion, the ovulation numbers were determined by counting the ovulated oocytes present in the perfusion chamber under a stereomicroscope. Experimental design Detection of NOS proteins in the perfused rat ovary In order to validate this ex-vivo perfusion system further with regard to NOS expression, immunoblotting was conducted with perfused ovaries stimulated by gonadotrophin. The ovaries were perfused for 10 h in the presence (n 5 6) or absence (n 5 4) of LH 1 IBMX (0.2 µg/ml and 200 µmol/l respectively). At the end of the perfusion period, ovaries were snap frozen in liquid nitrogen and stored at –70°C until later analysis. NOS inhibitor experiment The first sets of experiments were undertaken to determine the contributions of the two different isoforms of NOS, iNOS and eNOS to ovulation rate and steroidogenesis. The non-selective NOS inhibitor, L-NMMA (300 µmol/l) and the relatively selective iNOS inhibitor (Corbett et al., 1992), AG (300 µmol/l and 1 mmol/l), were administrated 30 min prior to LH (0.2 µg/ml) or LH 1 IBMX (0.2 µg/ml and 200 µmol/l respectively). IBMX is a nonselective phosphodiesterase inhibitor, which was added to potentiate LH effects and increase the number of ovulations (Peterson et al., 1993). Samples of perfusion medium were analysed for oestradiol and progesterone concentrations at 0, 1, 3, 5, 8 and 20 h time points. The total amount of the two stable products of NO, nitrite (NO2–) and nitrate (NO3–), in the perfusion medium was analysed at 0, 8 and 20 h time points. At the end of the 20 h perfusion, ovulation numbers were assessed as described earlier (Bra¨nnstro¨m et al., 1987). NO donor experiment The second set of experiments was conducted to determine whether the administration of an NO donor has any further effects on ovulation stimulated by LH (0.2 µg/ml). Spermine NONOate is an NO complex, which releases NO spontaneously without enzymatic involvement (Maragos et al., 1991). Spermine NONOate was added to the perfusion media to obtain two different initial concentrations (10 µmol/l and 100 µmol/l). As the half life of this NO donor is short (39 min at 37°C and pH 7.4), it was administrated at three time points (30min before LH, 4 h and 8.5 h after LH). The concentrations of oestradiol, progesterone and total amount of NO products in perfusion media were analysed in the same way as in the NO inhibitor experiment. Ovulation numbers were assessed at the end of the perfusion.

2538

Immunoblotting Soluble tissue extracts were prepared as previously described (Piontkewitz et al., 1993) with minor modifications. Briefly, the tissues were homogenized by Pellet Pestle® homogenizer (Kontes, Vineland, NJ, USA) in PE buffer (10 mmol/l potassium phosphate buffer, pH 6.8 and 1 mmol/l EDTA) containing 10 mmol/l 3-[(3-cholamidoprpyl) dimethyl-ammonio] 1-propan sulphonate (CHAPS), aprotinin (1 mg/ml), leupeptin (1 mg/ml), pepstatin (1 mg/ml) and Pefablock® (1 mg/ml). All protein inhibitors were purchased from Boehringer Mannheim (Mannheim, Germany). The homogenate was sonicated (2310 s) and centrifuged for 10 min at 10 800 g, 4°C. Supernatants were stored at –70°C until analysis. The samples were diluted in sodium dodecyl sulphate (SDS) sample buffer before loading on a one-dimensional SDS-polyacrylamide gel (4.5% stacking gel, 6–8% separating gel; Novex system, San Diego, CA, USA). The amount of total protein loaded into each lane was 25–35 µg. The proteins were transferred to a polyvinyldifluoride membrane (PVDF; Amersham, Bucks, UK) by electroblotting (Novex system, San Diego, CA, USA). The membranes were then incubated with blocking buffer containing antibodies against eNOS (N30020, Transduction Laboratories, Lexington, KY, USA; dilution 1:1000) or iNOS (N32024, Transduction Laboratories, dilution 1:1000). Immunoreactive proteins were visualized by chemiluminescense using an alkaline phosphate-conjugated second antibody and CDP-star® (Tropix, Bedford, MA, USA) as substrate. The membranes were exposed to enhanced chemiluminescence (ECL) film (Amersham, Bucks, UK) for 10 s to 2 min and subsequently developed. Assays Oestradiol and progesterone concentrations in perfusion media were analysed by DELFIA® assay kits (Wallac, Oy, Finland). Nitric oxide produced by tissues is rapidly scavenged and its final products are nitrite (NO2–) and nitrate (NO3–). As the relative proportion of nitrite and nitrate is variable and cannot be predicted, the total amount of these two final products of NO was colourimetrically analysed (Green et al., 1982). Ten µl each of nitrate reductase and co-factors (Alexis Corp.) were added to 100 µl of samples to convert nitrate to nitrite. Then Greiss reagent (110 µl; Alexis Corp.) was added to each well with converted sample. After 60 min incubation at room temperature, the absorbance was measured at a wavelength of 550 nm. Statistical analysis Non-parametric tests were used in data analysis, since in some of the experiments the data were not normally distributed. The results of ovulation number, concentrations of steroids and NO products at each sample point were evaluated by Kruskal–Wallis rank test followed by Mann–Whitney U test. P , 0.05 was considered to be statistically significant.

Results Expression of eNOS and iNOS proteins in the ex-vivo perfused rat ovary The presence of both iNOS and eNOS proteins in the perfused rat ovary was detected by immunoblotting. Perfusions with an ovulatory dose of LH 1 IBMX enhanced iNOS expression as compared to those without gonadotrophin stimulation (Figure 1). Endothelial NOS was detected at an approximate size of 140 kDa, with no major difference in eNOS expression between the treated groups observed (data not shown).

Nitric oxide and ovulation

Figure 1. Representative immunoblot demonstrating the induction of inducible nitric oxide synthase ( iNOS) in rat ovaries perfused for 10 h in the presence or absence of luteinizing hormone (LH) 1 3-isobutyl-l-methylxanthine (IBMX). Treatment with LH 1 IBMX induced higher amounts of iNOS protein (130 kDa) as compared to control without gonadotrophin stimulation. Extracts of mouse macrophages activated by lipopolysaccharides were used as a positive control.

Effects of NOS inhibitors There was no significant difference in nitrite/nitrate concentrations in the perfusion media when LH 1 IBMX and LH controls were compared (data not shown). Nitric oxide concentrations in the perfusate increased with time in all treatment groups and the administration of NOS inhibitors resulted in suppressed NO production. At the 20 h time point, nitrite/nitrate concentrations were significantly lower in the L-NMMA and AG 1 mmol/l groups compared with the LH 1 IBMX control (Figure 2a). Ovulation numbers in the NOS inhibitor groups are presented in Figure 2b. The LH 1 IBMX control group (n 5 10) exhibited 12.0 (median, range 5 6–17) ovulations per ovary. Significant reduction of ovulation rate was observed in the L-NMMA group (median 5 4.0, range 5 1–14; n 5 5), while ovaries perfused with AG at both 300 µmol/l and 1 mmol/l concentrations exhibited similar ovulation numbers as control (11.5, 6–18; n 5 6 and 11.0, 7–15; n 5 5 respectively). In all groups, there were marked increases in secreted oestradiol and progesterone after the administration of LH 1 IBMX. None of the NOS inhibitors significantly altered the LH 1 IBMX-induced steroid production (data not shown). To exclude the possibility that IBMX may influence the ovarian action of NO in this model, an additional set of experiments was performed, in which IBMX was excluded. Nitrite/nitrate concentrations were significantly lower at 8 and 20 h in the L-NMMA (n 5 5) and AG 300 µmol/l (n 5 5) groups compared to the LH control (n 5 4; Figure 3a), while ovulation numbers were reduced only in the L-NMMA group (median 5 0.0, range 5 0–8 for L-NMMA versus 7.5, 3–12 for control and 7.0, 1–15 for AG; Figure 3b).

Figure 2. (a) Concentrations of nitrite/nitrate in the perfusion media. Administration of NOS inhibitors (aminoguanidine bicarbonate, AG 1 mmol/l and NG-monomethyl-L-arginine, L-NMMA) resulted in significantly lower nitrite/nitrate concentrations at 20 h (*P , 0.05). Bars indicate 10–90% range; boxes indicate 25–75% range and medians. (b) Ovulation numbers in perfusions with NOS inhibitors. The L-NMMA (300 µmol/l; n 5 5) treated group exhibited significantly (*P , 0.05) lower ovulation numbers than the LH 1 IBMX control group (n 5 10). The relatively iNOS selective inhibitor, AG, did not significantly alter the ovulation number at either 300 µmol/l (n 5 6) or 1 mmol/l (n 5 5) concentrations. Individual values are shown and medians are indicated by horizontal bars.

Effects of NO donor The concentrations of NO products in perfusion media are presented in Figure 4a. Significantly higher concentrations of nitrite/nitrate were observed in the NONOate 100 µmol/l group (n 5 6), while there were no differences between LH control (n 5 6) and the group with the lower NONOate concentration (10 µmol/l; n 5 7). No significant differences in ovulation numbers were observed (Figure 4b) between the three experimental groups (median 5 6.0, range 5 2–9 for LH control; 3.0, 0–8 and 2.5, 0–6 for NONOate 10 µmol/l and 100 µmol/l respectively). The progesterone production by the rat ovary was increased in the NONOate 100 µmol/l group with higher concentrations than in control at 1 h and 3 h time points. The treatment with the NO donor resulted in slightly lower oestradiol concentrations, but the difference was not significant (Figure 5). Discussion Nitric oxide has been postulated to be one of several paracrine mediators in a number of cycle dependent ovarian events, such 2539

K.Mitsube et al.

Figure 3. (a) Concentrations of nitrite/nitrate in the media of perfusions with NOS inhibitors in the absence of IBMX. Both of L-NMMA and AG significantly decreased nitrite/nitrate concentrations at the 8 and 20 h time points (*P , 0.05). Bars indicate 10–90% range; boxes indicate 25–75% range and medians. (b) Ovulation numbers in perfusions with only LH as the ovulatory stimulator. The administration of L-NMMA caused significant reductions in ovulation number (*P , 0.05), while AG treatment did not significantly alter the ovulation number from control. Individual values are shown and medians are indicated by horizontal bars.

as ovulation (Shukovski and Tsafiri, 1994; Van Voorhis et al., 1994; Jablonka-Shariff and Olson, 1998). Several mechanisms for NO involvement in the ovulatory process have been suggested. Ovarian blood flow is increased by the pre-ovulatory LH surge (Janson, 1975) and NO may be one of the mediators in this event by dilating ovarian vasculature (Stones et al., 1995). Nitric oxide also affects the blood–follicle barrier during the ovulatory period (Powers et al., 1995), and NO induction by gonadotrophins may contribute to the increased vascular permeability (Hess et al., 1998), hyperaemia and leukocyte migration around ovulating follicles (Bonello et al., 1996). Another ovulation-associated pathway which may be NOmediated is eicosanoid synthesis, since enhanced prostaglandin (PG)E2 and PGF2α production after exogenous NO donor administration was reported in both ovarian cell culture (Salvemini et al., 1993) and perfusion study models (Yamauchi et al., 1997). Furthermore, an overproduction of NO may cause direct cytotoxic/tissue remodelling effects in the ovary (Ellman et al., 1993). The expressions of two different isoforms of NOS, iNOS 2540

Figure 4. (a) Concentrations of nitrite/nitrate in the perfusion media. The nitric oxide (NO) donating substance, spermine NONOate (10 µmol/l; n 5 7, 100 µmol/l; n 5 6), was added to the perfusion media at –0.5, 4.5, and 9 h time points. The NONOate 100 µmol/l group exhibited significantly higher nitrite/nitrate concentrations (*P , 0.01), while no significant difference from control was observed in the NONOate 10 µmol/l group. Bars indicate 10–90% range; boxes indicate 25–75% range and medians. (b) The administration of the NO donor, at either concentrations, did not significantly alter the ovulation numbers. Individual values are shown and medians are indicated by horizontal bars.

and eNOS, and their regulation by gonadotrophins have been extensively examined in the rat ovary (Van Voorhis et al., 1995; Zackrisson et al., 1996a; Jablonka-Shariff and Olson, 1997). Nevertheless, the contribution of each isoform to ovulation is still not clear. In this report, it is demonstrated that both iNOS and eNOS are expressed in the ex-vivo perfused rat ovary and that the presence of a non-selective NOS inhibitor, L-NMMA, results in reduced ovulation numbers after gonadotrophin stimulation. This result is in accordance with a previous study (Bonello et al., 1996), where another non-selective inhibitor, L-NAME (N-omega-nitro-L-arginine methyl ester), was used. However in the present study, an inhibitor with relative selectivity towards iNOS, AG, did not influence LH-induced ovulation rate in contrast to previous reports on effects in vivo in the rat (Shukovski et al., 1994) and in the ex-vivo perfused rabbit ovary (Yamauchi et al., 1997). In both of the NOS inhibitor groups of the present study, the total quantities of NO produced by the ovary were most likely reduced to similar concentrations as judged by

Nitric oxide and ovulation

Figure 5. The effects of the NO donating substance, spermine NONOate, on steroid secretion by the ovary stimulated by LH. (a) The treatment with the NO donor resulted in slightly lower oestradiol concentrations, but the difference was not significant. (b) Progesterone concentrations in the 100 µmol/l NONOate treated group were significantly higher than LH control at 1 h and 3 h time points (*P ,0.05). Bars indicate 10–90% range; boxes indicate 25– 75% range and medians.

comparable nitrite/nitrate concentrations, but yet there were clear differences in effects on ovulation rate. Thus, it seems as if relatively small quantities of NO could support the full ovulatory function in the AG-treated ovaries, in which eNOS would be 10 to 100 times less potently inhibited compared to iNOS (Corbett et al., 1992). In contrast, in the L-NMMAtreated ovaries, in which eNOS and iNOS would be equally inhibited, ovulation numbers were significantly lower than in the AG-treated group, but with similar concentrations of nitrite/ nitrate in the perfusion media. These findings in perfusions with L-NMMA were consistently observed in both experimental groups with and without IBMX. The non-specific phosphodiesterase inhibitor IBMX was used to increase further the LH-induced cyclic AMP concentrations and thereby optimally stimulate the ovary to ovulate (Peterson et al., 1993). The presence of IBMX would also result in higher concentrations of cyclic GMP (cGMP), which is one of the second messengers of NO (Moncada et al., 1991), and could possibly counteract the influence of NOS inhibitors. However, the similar results in the two experimental models of the present study suggest that the presence of IBMX did not affect the actions of NOS inhibitors on ovulation in this ex-vivo ovarian perfusion model. A possible explanation for the difference in ovulation numbers in the presence of the two NOS inhibitors may be the differences in tissue localization of NO production, since the total NO produced by the ovaries was similar in the two

inhibitor groups. In the rat ovary, eNOS is localized mainly in endothelial cells of the vascular walls of the theca and ovarian stroma components during the pre-ovulatory and ovulatory stages, and both eNOS mRNA and protein are upregulated by gonadotrophin stimulation (Van Voorhis et al., 1995; Zackrisson et al., 1996a; Jablonka-Shariff and Olson, 1997). There exist some discrepancies in the descriptions of localization and regulation of iNOS in the peri-ovulatory ovary. Jablonka-Shariff and Olson (1997) reported that iNOS protein is localized in the theca cell layer, ovarian stroma and nonparenchymal cells of corpus luteum, but not in the granulosa cell layer. In their study, administration of LH resulted in a marked enhancement of iNOS protein expression in the theca cell layer and stroma. A somewhat different pattern was observed in an earlier study (Zackrisson et al., 1996a) with a weaker expression of iNOS in the granulosa cell and theca cell layers compared with eNOS and a slight increase of iNOS protein expression in the granulosa cell layer 6 h after HCG. In contrast to the protein, iNOS mRNA, which is localized exclusively to the granulosa cell layer, is maximally expressed in unstimulated ovaries and HCG administration results in decreased concentrations of iNOS mRNA (Van Voorhis et al., 1995). Based on results of the present study, it can be speculated that eNOS-derived NO from endothelial cells may be more accessible to modulate the ovarian vascular system in a way that would facilitate ovulation. Another explanation for the differences in the effects of the two NOS inhibitors is the presumed time difference in NO production by eNOS and iNOS. Endothelial NOS is calciumcalmodulin dependent and produces small amounts of NO in response to certain stimulus for a relatively short duration. In contrast, iNOS is regulated at the transcriptional level and once expressed it produces large amounts of NO for long periods. As it takes several hours for iNOS to be induced, the NO-related ovarian events in the early phase after LH-surge may be more dependent on constitutively expressed eNOS than iNOS. Consequently, it is conceivable that the increase and maintenance of ovarian blood in this early period after LH stimulation may be more severely disrupted in the L-NMMA group, in which eNOS was potently inhibited, than in the AG groups. A rapid effect on blood flow by nonselective NOS inhibitors was indicated in vitro (Bonello et al., 1996), and it has recently been shown that the ovulatory process of the rat is more sensitive to reductions of blood flow during the first third of the ovulatory process than during later phases (Zackrisson et al., 1996b). A study by Powers et al. concerning the rat ovarian blood–follicle barrier also supports this hypothesis (Powers et al., 1995). They reported that the permeability of the blood–follicle barrier, which was measured as an influx of inter-α-trypsin inhibitor protein into the follicle, increased within seconds after HCG stimulation and that this influx was completely blocked by non-selective NOS inhibitors with reduced ovulation numbers observed. It was recently reported in studies of NOS gene-targeted mice that there exists a reduction in ovulation number in both iNOS (40% reduction) and eNOS (60% reduction) knock-out mice (Olson et al., 1997; Jablonka-Shariff and Olson, 1998). These findings are 2541

K.Mitsube et al.

in line with the results of the present study, in that they also indicate a more critical role in ovulation for eNOS than iNOS. A major difference of the present ex-vivo perfusion model from the in-vivo situation is the complete exclusion of leukocytes, which are the major sources of cytokines and the production sites of NO by iNOS, from circulation. In the ex-vivo ovarian perfusion model, the ovulation numbers are lower than in in-vivo models, and this reduced ovulation rate can be partly restored by the administration of leukocytes (Hellberg et al., 1991) or cytokines such as interleukin-1β (IL-1β) (Bra¨nnstro¨m et al., 1993) into the perfusion system. Increased production of NO by the LH and IL-1β stimulated ovary compared to LH controls was reported in a study (Bonello et al., 1996), where the augmentation of NO synthesis by IL-1β was suggested to be one of the causes for the difference in ovulation numbers. Tao et al. reported that human follicular nitrite/nitrate concentrations positively correlated with follicular IL-1β concentrations and the culture of follicular cells with IL-1β resulted in elevated nitrite production (Tao et al., 1997). These types of interactions between cytokines and the NOS system have been discussed in several reproductive organs (Rosselli et al., 1998). In the present study, even in the absence of exogenously added IL-1β, an up-regulation of iNOS could be demonstrated in the perfused ovary stimulated by LH 1 IBMX and a significant increase in nitrite/nitrate concentrations in the perfusion media was observed. These findings suggest that the pathways which regulate NOS expression remain intact in the ex-vivo ovarian perfusion model. The perfused ovary, containing residual leukocytes, may produce cytokines including IL-1β by itself or activation of other LHinduced pathways may be sufficient for the induction of NOS. Recently, we have reported that the transcription factor CCAAT/enhancer binding protein (C/EBP) β, which is strongly up-regulated by LH 1 IBMX, plays a major role in the ovulatory process (Pall et al., 1997) and this factor promotes iNOS gene expression (Hecker et al., 1997). To examine further if NO concentrations of higher magnitude than those obtained in the LH-stimulated perfusions would facilitate ovulation, experiments were conducted with an NO donating compound. The addition of an NO donor, spermine NONOate, to the LH-stimulated ovary did not change the ovulation rate as compared to control, even when nitrite/nitrate concentrations in the perfusion media were increased. These findings suggest that LH-induced endogenous NO production by the ovary is sufficient for a full ovulatory function to be operative and NO from exogenous sources, such as leukocytes, may not have additional effects on ovulation. In general, even low concentrations of NO mediate effects through the cGMP pathway to induce physiological responses such as vasorelaxation and neural transmission, whereas relatively higher concentrations of NO are associated with cytotoxic effects (Xie et al., 1992; Jun et al., 1997). The results of the present study suggest that a cytotoxic effect of NO may not be indispensable for ovulation to occur, which is in accordance with the findings that IL-1β potentiates the LH-induced ovulation (Peterson et al., 1993; Bra¨nnstro¨m et al., 1993) but that IL-1β induced ovarian cellular cytotoxicity is not mediated by NO (Ben-Shlomo et al., 1994). As previously described, it 2542

may not only be the total NO production by the ovary but also the tissue/time specific production of this mediator, which is of importance. Recently, the involvement of the neuronal isoform of NOS in the events around the ovulatory process was suggested (Klein et al., 1998). Neuronal NOS was localized in nerves innervating the reproductive tract and substantial reproductive defects and fewer oocyte numbers in the oviducts after ovulation were observed in nNOS knock-out mice. By comparing the number of the oocytes in the oviducts and the number of ovarian rupture sites, it was suggested that this reproductive deficit was not due to the inhibition of the rupture of the follicular wall with the extrusion of oocytes from the ovary, but due to an impaired transfer of ovulated oocytes from the ovarian bursa to the oviducts by the lack of neuronally derived NO. Nitric oxide has been postulated as one of several paracrine/ autocrine regulators of ovarian steroidogenesis. In the present study, exogenous addition of an NO donor enhanced progesterone production by the rat ovary. The concentrations of oestradiol in the NO donor-treated groups were nearly half of those in the control group at certain time points, but these were not statistically different. In accordance with our observation, some previous studies reported that NO donors suppressed oestradiol production and NOS inhibitors enhanced steroid synthesis (Olson et al., 1996; Yamauchi et al., 1997). The direct inhibitory effect of NO on cytochrome P450 aromatase was demonstrated in granulosa–luteal cells (Snyder et al., 1996). It is suggested that nitric oxide binds to the iron-containing haem of the enzymes, altering their activity. It is also reported that NO can affect the gene expression of steroidogenic enzymes in extra-ovarian tissues (Sun et al., 1997). On the other hand, there are discrepancies in results between different experimental models regarding the effect of NO on ovarian progesterone production. In studies with cultured ovarian cells, NO inhibited progesterone production (Olson et al., 1996; Ahsan et al., 1997), while in a whole ovarian culture study, the administration of an NO donor resulted in higher concentrations of progesterone (Dong et al., 1999), which is consistent with the findings of the present study. As the preservation of the three-dimensional tissue structure and cell-to-cell contacts seems to be the major difference between the models, these factors may be of importance in the regulation of steroid production. In the present study, neither of the two NOS inhibitors changed the steroid concentrations compared to control. It may be that in this NO-suppressed milieu, possible modulations in oestradiol and progesterone production by the ovarian cells would be counteracted by the reduced flow through the vasculature. This result suggests that the intrinsic ovarian NO synthesis may be sufficient to support full ovarian steroidogenic function and that only supraphysiological concentrations of NO may exert effects on ovarian steroidogenesis. Acknowledgements This study was supported by grants from the Swedish Medical Research Council (11607 to M.B.), Hjalmar Svensson Research Foundation and Medical Faculty of Go¨teborg University.

Nitric oxide and ovulation

References Ahsan, S., Lacey, M. and Whitehead, S.A. (1997) Interactions between interleukin-1β, nitric oxide and prostaglandin oestradiol in the rat ovary: effects on steroidogenesis. Eur. J. Endocrinol., 137, 293–300. Ben-Shlomo, I., Adashi, E.Y. and Payne, D.W. (1994) The morphologic/ cytotoxic and prostaglandin-stimulating activities of interleukin-1β in the rat ovary are nitric oxide independent. J. Clin. Invest., 94, 1463–1469. Bonello, N., McKie, K., Jasper, M. et al. (1996) Inhibition of nitric oxide; effects on interleukin-1β-enhanced ovulation rate, steroid hormones, and ovarian leukocyte distribution at ovulation in the rat. Biol. Reprod., 54, 436–445. Bra¨nnstro¨m, M., Wang, L. and Norman, R.J. (1993) Ovulatory effect of interleukin-1β on the perfused rat ovary. Endocrinology, 213, 399–404. Bra¨nnstro¨m, M., Johanson, B.M., Sogn, J. et al. (1987) Characterization of an in vitro perfused rat ovary model: ovulation rate, oocyte maturation, steroidogenesis and influence of PMSG priming. Acta Physiol. Scand., 130, 107–114. Corbett, J.A., Tilton, R.G., Chang, K. et al. (1992) Aminoguanidine, inhibitor of nitric oxide formation, prevents diabetic vascular dysfunction. Diabetes, 41, 552–556. Dong, Y.L., Gangula, P.R.R., Fang, L. et al. (1999) Nitric oxide reverses prostaglandin-induced inhibition of ovarian progesterone secretion in rats. Hum. Reprod., 14, 27–32. Ellman, C., Corbett, J.A., Misko, T.P. et al. (1993) Nitric oxide mediates interleukin-1-induced cellular cytotoxicity in the rat ovary. J. Clin. Invest., 92, 3053–3056. Green, L.C., Wagner, D.A., Glogowski, J. et al. (1982) Analysis of nitrate, nitrite and [15N] nitrite in biological fluids. Anal. Biochem., 126, 131–138. Hecker, M., Preiss, C. and Schini-Kerth, V.B. (1997) Induction by staurosporine of nitric oxide synthase expression in vascular smooth muscle cells: role of NF-kB, CREB and C/EBP β. Br. J. Pharmacol., 120, 1067–1074. Hellberg, P., Thomsen, P., Janson, P.O. et al. (1991) Leukocyte supplementation increases the luteinizing hormone-induced ovulation rate in the in vitroperfused rat ovary. Biol. Reprod., 44, 791–797. Hess, K.A., Chen, L. and Larsen, W.J. (1998) The ovarian blood follicle barrier is both charge- and size-selective in mice. Biol. Reprod., 58, 705–711. Jablonka-Shariff, A. and Olson, L.M. (1997) Hormonal regulation of nitric oxide synthase and their cell-specific expression during follicular development in the rat ovary. Endocrinology, 138, 460–468. Jablonka-Shariff, A. and Olson, L.M. (1998) The role of nitric oxide in oocyte meiotic maturation and ovulation: meiotic abnormalities of endothelial nitric oxide synthase knock-out mouse oocytes. Endocrinology, 139, 2944–2954. Janson, P.O. (1975) Effects of luteinizing hormone on blood flow in the follicular rabbit ovary, as measured by radioactive microspheres. Acta Endocrinol., 79, 122–133. Jun, C.D., Park, S.J., Choi, B.M. et al. (1997) Potentiation of the activity of nitric oxide by the protein kinase C activator phorbol ester in human myeloid leukemic HL-60 cells; association with enhanced fragmentation of mature genomic DNA. Cell. Immunol., 176, 41–49. Klein, S.L., Carnovale, D., Burnett, A.L. et al. (1998) Impaired ovulation in mice with targeted deletion of the neuronal isoform of nitric oxide synthase. Mol. Med. 4, 658–664. Maragos, C.M., Morley, D. and Wink, D.A. (1991) Complexes of NO as agents for controlled biological release of nitric oxide. J. Med. Chem., 34, 3242–3247. Moncada, S., Palmer, R.M.J. and Higgs, E.A. (1991) Nitric oxide: physiology, pathophysiology and pharmacology. Pharmacol. Rev., 43, 109–142. Olson, L.M., Jones-Burton, C.M. and Jablonka-Shariff, A. (1996) Nitric oxide decreases estradiol synthesis of rat luteinized ovarian cells: possible role for nitric oxide in functional luteal regression. Endocrinology, 137, 3531–3539. Olson, L.M. (1997) Nitric oxide in reproductive function. Biol. Reprod., 56 (Suppl. 1), 32–33 (abstract). Palacois, C.M., Konowles, R.G., Palmer, R.M.J. et al. (1993) Nitric oxide from L-arginine stimulates the soluble guanylate cyclase in adrenal glands. Biochem. Biophys. Res. Commun., 165, 802–809. Pall, M., Hellberg, P., Bra¨nnstro¨m, M et al. (1997) The transcription factor C/EBP-β and its role in ovarian function; evidence for direct involvement in the ovulatory process. EMBO J., 16, 5273–5279. Peterson, C.M., Hales, H.A., Hatasaka, H.H. et al. (1993) Interleukin-1β modulates prostaglandin production and the natural IL-1 receptor antagonist inhibits ovulation in the optimally stimulated rat ovarian perfusion model. Endocrinology, 133, 2301–2306.

Piontkewitz,Y., Enerba¨ck, S. and Hedin, L. (1993) Expression and hormonal regulation of the CCAAT Enhancer Binding Protein-alpha during differentiation of rat ovarian follicles. Endocrinology, 133, 2327–2333. Powers, R.W., Chen, L., Russell, P.T. et al. (1995) Gonadotropin-stimulated regulation of blood–follicle barrier is mediated by nitric oxide. Am. J. Physiol., 269, 290–298. Rosselli, M., Keller, P.J. and Dubey, R.K. (1998) Role of nitric oxide in the biology, physiology and pathophysiology of reproduction. Hum. Reprod. Update, 4, 3–24. Salvemini, D., Misko, T.P., Masferrer, J.L. et al. (1993) Nitric oxide activates cyclooxygenase enzymes. Proc. Natl Acad. Sci. USA, 90, 7240–7244. Shukovski, L. and Tsafriri, A. (1994) The involvement of nitric oxide in the ovulatory process in the rat. Endocrinology, 135, 2287–2290. Stones, R.W., Loesch, A., Beard, R.W. et al. (1995) Substance P: endothelial localization and pharmacology in the human ovarian vein. Obstet. Gynecol., 85, 273–278. Snyder, G.D., Holmes, R.W., Bates, J.N. et al. (1996) Nitric oxide inhibits aromatase activity: mechanisms of action. J. Steroid. Biochem. Molec. Biol., 58, 63–69. Snyder, S.H. (1992) Nitric oxide: first in a new class of neurotransmitters. Science, 257, 494–496. Sun, K., Yang, K. and Challis, J.R.G. (1997) Differential regulation of 11βhydroxysteroid dehydrogenase type 1 and 2 by nitric oxide in cultured human placental trophoblast and chorionic cell preparation. Endocrinology, 138, 4912–4920. Tao, M., Kodama, H., Kagabu, S. et al. (1997) Possible contribution of follicular interleukin-1β to nitric oxide generation in human pre-ovulatory follicles. Hum. Reprod., 12, 2220–2225. Van Voorhis, B.J., Dunn, M.S., Snyder, G.D. et al. (1994) Nitric oxide: an autocrine regulator of human granulosa–luteal cell steroidogenesis. Endocrinology, 135, 1799–1806. Van Voorhis, B.J., Moore, K., Strijbos, P.J.M.L. et al. (1995) Expression and localization of inducible and endothelial nitric oxide synthase in the rat ovary. J. Clin. Invest., 96, 2719–2726. Xie, Q.W., Cho, H.J., Calaycay, J. et al. (1992) Cloning and characterization of inducible nitric oxide synthase from mouse macrophages. Science, 256, 225–228. Yamauchi, J., Miyazaki, T., Iwasaki, S. et al. (1997) Effects of nitric oxide on ovulation and ovarian steroidogenesis and prostaglandin production in the rabbit. Endocrinology, 138, 3630–3637. Zackrisson, U., Mikuni, M., Wallin, A. et al. (1996a) Cell-specific localization of nitric oxide synthases in the rat ovary during follicular development, ovulation and luteal formation. Hum. Reprod., 11, 2667–2673. Zackrisson, U., Mikuni, M., Makinoda, S. et al. (1996b) Ovulation rate in the rat is decreased by acute ligation of the ovarian artery or the ovarian branch of the uterine artery. Biol. Reprod., 54 (suppl. 1), 68 (abstract 46). Received on March 25, 1999; accepted on July 15, 1999

2543