Inhibition of Nitric Oxide: Effects on Interleukin-lβ-Enhanced Ovulation ...

BIOLOGY OF REPRODUCTION 54, 436-445 (1996)

Inhibition of Nitric Oxide: Effects on Interleukin-lp-Enhanced Ovulation Rate, Steroid Hormones, and Ovarian Leukocyte Distribution at Ovulation in the Rat' Nigel Bonello,2 '3 Kylie McKie, 3 Melinda Jasper, 3 Lucy Andrew,3 Nicki Ross, 3 Emily Braybon,3 Mats Brinnstr6m, 3 ' 4 and Robert J. Norman3

Reproductive Medicine Unit,3 Department of Obstetrics and Gynaecology, The University of Adelaide The Queen Elizabeth Hospital, Woodville, SA 5011, Australia Departmentof Obstetrics and Gynaecology, 4 The University of GCteborg Sahlgrenska Sjukhuset, Goteborg S-41345, Sweden ABSTRACT The ovulatory process resembles an inflammatory reaction with an infiltration of leukocytes, production of inflammatory mediators such as cytokines, and a general edema and hyperemia. Nitric oxide (NO), a potent vasodilator and the main mediator of macrophage tumoricidal and bacteriocidal activities, is known to participate in inflammatory reactions and has been shown to mediate the interleukin1[ (IL-lp)-directed tissue-remodeling events within the ovary. The regulation by NO of ovulation rate, leukocyte distribution, and steroid release in the rat ovary was investigated through use of a combination of in vivo and in vitro models of ovulation and a competitive -4 inhibitor, N-omega-nitro-L-arginine methyl ester (L-NAME), of the NO synthase (NOS) enzyme. Subcutaneous L-NAME (1.5 x 10 mol/kg) administration significantly reduced the in vivo ovulation rate of eCG/hCG-primed rats (L-NAME-treated: 10.6 + 1.8 [mean ± SEMI oocytes per ovary [O/O, 11.0 ± 1.2 rupture sites per ovary [RS/O]; saline-treated: 18.0 + 1.8 0/O, 19.4 ± 1.1 RS/O; p < 0.01) at 20 h post-hCG. These results were reflected in vitro, where addition of L-NAME (3.5 x 10-5 mol/L) to LH (0.1 pg/ml)-perfused ovaries decreased ovulation rate from 8.2 ± 1.6 to 2.7 1ovulations per ovary (p < 0.05) and simultaneously decreased nitrite accumulation at the completion of perfusions from 16.5 ± 1.9 to 4.1 ± 0.5 nmol/ml (p < 0.001). The addition of L-NAME to LH + IL-1P (4 ng/ml)-perfused ovaries decreased ovulation rate from 15.2 ± 2.4 to 0.8 ± 0.8 ovulations per ovary (p < 0.001) and simultaneously decreased nitrite accumulation at 22 h from 22.8 ± 2.2 to 1.9 ±0.6 nmol/ml (p < 0.001). Studies analyzing and manipulating perfusion flow rate indicated that the L-NAME effects on ovulation rate are primarily due to a reduction in flow rate resulting from inhibition of NO, which may be a consequence of the known vasoconstrictor effects of NOS inhibitors. The observed reduction of in vivo ovulation rate by NO inhibition at 20 h post-hCG was associated with a significant reduction in thecal MCA149+ neutrophils at 12 h post-hCG, the expected time of ovulation (L-NAME-treated: 11.5 cells per thecal area; p < 0.001), while ED1 + monocytes/macrophages 98.4 ± 9.2 cells per thecal area; saline-treated: 211.5 underwent similar but nonsignificant changes. Plasma (20 h post-hCG) and perfusate progesterone were not different with L-NAME treatment, while perfusate estradiol levels were markedly reduced upon addition of L-NAME, suggesting a role for NO in ovulation but not in the process of luteinization. In summary, deprivation of NO by use of the competitive inhibitor, L-NAME, led to fewer ovulations, reduced accumulation of nitrite, a decreased neutrophil count in the theca of preovulatory follicles, and reduced estradiol secretion, while progesterone release remained unaffected. The NO pathway may therefore play an important role inthe regulation of ovulation and the mediation of IL-1l's pro-ovulatory effects. There are likely to be primarily vascular effects, but also a nonvascular component, to the NO regulation of ovulation, with both components indirectly affecting ovulatory leukocyte distribution and steroid secretion. INTRODUCTION

diverse cell types in both physiological and pathophysiological conditions, have all been shown to play a role in the process of follicle rupture [8]. Of particular interest is the cytokine interleukin-lp (IL-1p), which has been shown to be present in the ovary at both the mRNA and protein levels [9, 10] and which undergoes an hCG-dependent rise in mRNA just before ovulation [11]. IL-1P enhances ovulation rate in perfused rat ovaries when administered alone, but especially so in the presence of LH [12], and it modulates the activity of ovulatory mediators such as those within the plasminogen activator system [13, 14]. Furthermore, local administration of the natural inhibitor of the IL-1 system, the IL-1 receptor antagonist, is known to markedly reduce the hCG-induced ovulation rate of the rat in vivo [15]. Nitric oxide (NO), a potent vasodilator, is formed from the oxidation of L-arginine by NO synthase (NOS), which has constitutive calcium/calmodulin-dependent and cytokineinducible calcium/calmodulin-independent forms. Other

Ovulation is a complex process involving not only gonadotropins and steroid hormones but also many mediators common to inflammatory reactions [1]. At the cellular level, white blood cells-especially neutrophils and monocytes/ macrophages-have been shown to infiltrate the ovary just before ovulation [2-4] and to enhance the LH-induced ovulation rate of in vitro-perfused rat ovaries [5], while their depletion from the peripheral blood is associated with a reduction in ovulation rate [6]. At the molecular level, products of inflammatory leukocytes such as cytokines [7], prostaglandins, leukotrienes, plasminogen activators, histamine, and bradykinin, which are now known to be secreted by Accepted September 27, 1995. Received June 30, 1995. 'This research was funded by the NH &MRC (Australia). N.B. is supported by a University of Adelaide Medical Faculty Postgraduate Scholarship and a Queen Elizabeth Hospital Scholarship. M.B. was supported by the Swedish Medical Research Council (B96-17X). 2Correspondence. FAX: 61 8 2687978.

436

OVULATORY EFFECTS OF NITRIC OXIDE INHIBITION

physiological functions of NO include smooth muscle relaxation, neurotransmission, and the mediation of macrophage tumoricidal and bacteriocidal actions [16, 17]. An important regulatory role for NO in follicle rupture is indicated in a recent study revealing a reduction in the in vivo ovulation rate of rats administered inhibitors of the NOS enzyme intrabursally [18]. Detection of NOS in ovarian cells [19-21] and in hypothalamic neurons that regulate LHRH secretion [22] and the observation of cyclicity in the concentration of serum nitrite (a stable metabolic breakdown product of NO) [23] during a woman's menstrual cycle, have further implied a role for NO in ovarian function. Furthermore, NO mediates the cytotoxic actions of IL-I 3 on ovarian cells [24], suggesting that through remodeling of ovarian tissue NO may be involved in the pro-ovulatory pathway of IL-1. The present study addresses several questions concerning the relative importance of NO in the ovulatory process and the mechanisms through which it may act. By using a wellcharacterized NOS competitive inhibitor [25, 26], N-omeganitro-L-arginine methyl ester (L-NAME), we wished to confirm that NO is indeed important for ovulation in vivo and to investigate whether the same effect can be reproduced by using the in vitro rat ovary perfusion model-a system that is devoid of systemic influences. The level of nitrite accumulation by perfused ovaries and the changes in steroid hormone production attributable to L-NAME treatment were also examined. In addition, we studied the role of NO in mediating the known effects of IL-1i on ovulation rate and steroid release, and we investigated the ovulatory effects of NO ascribed to its known vasodilatory actions. Finally, given the increasing direct and indirect evidence supporting a role for leukocytes in ovulation, we used immunohistochemistry to examine the effect of NO deprivation on the distribution of neutrophils and monocytes/macrophages in the ovary.

MATERIALS AND METHODS Chemicals and Monoclonal Antibodies L-NAME, sulfanilamide, and N-(1-naphthyl) ethylamide dihydrochloride were from Sigma Chemical Company (St. Louis, MO); isopentane was from BDH (Poole, UK); and Tissue-Tek (OCT) compound was from Miles Inc. (Elkhart, IN). Ovine LH (NIH LH s-25; specific activity [SA] 2.3 U/mg) was provided by the NIDDK and the National Hormone and Pituitary Program, University of Maryland School of Medicine (Baltimore, MD). Equine CG was from Intervet (Boxmeer, The Netherlands), and hCG was from Organon (Oss, The Netherlands). Monoclonal antibodies MCA149 and ED1 were from Serotec (Oxford, UK), while human recombinant IL-13 (spec. act. 5 X 108 U/mg) was from Genzyme (Boston, MA).

437

Animals Immature female Sprague-Dawley rats (from the University of Adelaide colony) weighing 50-75 g were maintained under controlled conditions (14L:10D) with free access to

pelleted food and water. At 27 days of age, all rats received an s.c. injection of eCG (in vivo study, 16 IU; in vitro study, 20 IU) at 1200 h on experimental Day -2 to promote the growth and maturation of a first generation of antral follicles. In Vivo Study The animals were divided into two groups; the experimental group received L-NAME s.c. (1.5 X 10 - 4 mol/kg), and the control group received a matched volume of saline s.c. Animals were first administered L-NAME or saline at 1100 h on Day 0. Human CG was administered i.p. at 1200 h on Day 0, followed by further L-NAME or saline administrations at 1400 h, 1700 h, 2000 h, and 2300 h on Day 0. Animals required for ovarian immunohistochemistry were killed at 0000 h on Day 1, and their ovaries were frozen. Animals required for ovulation assessment were killed at 0800 h on Day 1. A cardiac blood sample was collected within 30 sec for subsequent plasma progesterone assay, and the ovaries and oviducts were immediately examined. In Vitro (Rat Ovary Perfusion) Study Ovary perfusions in a 35-ml recirculating system were performed on the morning of Day 0 through use of surgery and perfusion procedures that have been described in detail previously [27]. The perfusion pressure was maintained at 80 mm Hg (except in the reduced-pressure group), resulting in an average flow of 0.9 ml/min in untreated and LH-perfused ovaries. The ovaries were initially perfused for 1 h to allow metabolic stabilization of the tissue before treatment was commenced. Treatments of L-NAME (3.5 X 10- 5 mol/ L), LH (0.1 lg/ml), and IL-13 (4 ng/ml), alone or in combination, were added to perfusion medium; in the case of controls, the ovaries were perfused with Medium 199 (Gibco, Grand Island, NY) alone. Samples of the circulating medium were obtained at 0, 1, 2, 3, 4, 8, and 22 h following treatment administration. Samples were stored at - 20°C for subsequent steroid hormone and NO assays. Assessment of Ovulation Rate The oocytes in the ampulla region of each oviduct were counted at 20 h post-hCG (i.e., 0800 h on Day 1). These data were supported by surface examination of the corresponding ovary for rupture sites. In vitro-perfused ovaries were removed from the perfusion system and washed gently with saline to dislodge any adherent oocytes before the number of oocytes found in the perfusion chamber was counted. All counting was coded and conducted by investigators blind to the treatment protocol.

438

BONELLO ET AL.

Perfusion Flow Rate Analysis To assess the effects of L-NAME on ovarian vessel constriction and flow, the perfusion flow rate was measured in LH- and LH + L-NAME-treated ovaries. The flow rate of the LH + L-NAME-treated ovaries was then reproduced in LH + IL-1S3 (reduced pressure; RP)-perfused ovaries by reducing the perfusion pressure to achieve the required flow rate. Progesteroneand EstradiolAssays Progesterone levels in perfusion media and rat plasma were analyzed through use of RIA kits from Amersham (Buckinghamshire, UK) with inter- and intraassay coefficients of variation (CVs) of < 7% and < 5%, respectively. Estradiol levels in perfusion media were analyzed with an RIA kit from Diagnostic Systems Laboratories (Webster, TX) with inter- and intraassay CVs of < 10% and < 9%, respectively. NO Assay Accumulation of NO in perfusion samples was analyzed by measuring nitrite as described previously [28]. Briefly, Greiss reagent was prepared; this consisted of one part 0.1% naphthylethylenediamine dihydrochloride in distilled water plus one part 1% sulfanilamide in 5% concentrated H3 PO 4. Fifty microliters of Greiss reagent was added to wells in a 96-well plate (Disposable Products, Adelaide, Australia), followed by 50 ll sample or standard (2.5, 5, 10, 20, 40, 80 gIM NaNO 2 ). After a 10-min incubation at room temperature, the plate was read for absorbance in an ELISA plate reader set at a wavelength of 540 nm. Immunohistochemistry The ovaries of animals killed at 0000 h on Day 1 were embedded in OCT and frozen in isopentane in liquid nitrogen before being stored at - 85°C. Air-dried 5-jm-thick cryostat sections then underwent immunohistochemical staining using the murine antibodies MCA149 (1:800 dilution), which detects rat neutrophilic granulocytes [29], and ED1 (1:400 dilution), which detects rat monocytes/macrophages [30], exactly as described previously [14] except that normal sheep serum (nonimmune serum of the species from which the linking antibody was derived) was used to block nonspecific binding. For negative controls, the mAb was replaced with either the antibody diluent or mouse IgG1. Only stained cells in the theca of follicles > 0.6 mm (average of 4.4 follicles per ovary) were counted, as it has previously been shown that follicles of this size at proestrus can be considered to have escaped atresia and are destined to ovulate [31]. Two independent observers (interobserver difference < 16%) counted the number of positively stained cells within an entire thecal region and in 1 1-mm areas of medullary stroma (average of 1.9 fields per ovary).

StatisticalAnalysis Unpaired Student's t-tests were utilized to compare the means of the saline- and L-NAME-treated groups in the in vivo component of the study, while one-way ANOVAs were used to examine differences between group variances (at each sample point in the time courses) in the in vitro-perfused ovary studies. A Tukey-Kramer Multiple Comparisons post-hoc test then detected the specific groups with different variances. In all cases p < 0.05 was considered significant. RESULTS In Vivo Study Ovulation rate (Fig. 1). Examination of oocyte numbers in the oviducts of rats at 20 h post-hCG revealed that the LNAME-treated rats ovulated significantly (p < 0.01) fewer oocytes per ovary than rats in the control group (mean + SEM, 10.6 ± 1.8 vs. 18.0 + 1.8, respectively). When follicle rupture points on the surface of ovaries were counted and used as the parameter for ovulation having occurred, almost identical results were obtained (L-NAME 11.0 ± 1.2; control group 19.4 1.1 rupture sites per ovary; p < 0.001). Numbers of ovulated oocytes correlated well with numbers of observed rupture sites for each ovary, with correlation coefficients of r = 0.86 and r = 0.84 in the control and LNAME-treated groups, respectively. Mean levels of progesterone in the plasma of rats at 20 h post-hCG were not significantly different (p = 0.65) between the L-NAMEtreated group (24.3 + 3 pmol/ml) and the control group (22.3 3.3 pmol/ml). Estradiol was not measured because of lack of plasma. Ovarian leukocyte distribution (Figs. 2 and 3). Untreated animals had more MCA149 + neutrophil cells than ED1 + monocyte/macrophage cells in the thecal region of follicles greater than 0.6 mm in diameter and had greater numbers of ED1 + compared to MCA149+ in 1 X -mm areas of medullary stroma in ovaries at the expected time of ovulation. The effect of L-NAME was most noticeable with regard to changes in neutrophil distribution. There was a significant 2-fold reduction in thecal neutrophils and a corresponding 2-fold rise in medullary neutrophils. Negative controls revealed negligible nonspecific cell staining: 0.5 + 0.2 cells per thecal area and 0.2 0.2 cells per medullary stroma when the antibody diluent was used and 1.6 + 0.8 cells per thecal area and 0.8 ± 0.3 cells per medullary stroma when the matching isotype, mouse IgG1, was used as the negative control. In Vitro Study Perfusionflow rate (Fig. 4). L-NAME had an immediate effect on the flow rate of LH-perfused ovaries, resulting in a 46% decrease in the initial flow rate within the first 15 min after addition of the compound. This decreased level was

OWVULATORY EFFECTS OF NITRIC OXIDE INHIBITION 250

439

_

]1 Control IS L-NAME 200

_

. 150 E

0

THE,CA

MEDULLA

z

CI

100 0 50

I_

n ay

Neutrophils

.i:

Neutrophils

Monocyte/ Macrophages

I

Monocyte/ Macrophages

FIG. 1. In vivo ovulation rate (mean + SEM ovulations per ovary, n = 12 rats per group) at 20 h post-hCG. Effects of L-NAME treatment were analyzed by counting the number of oocytes recovered from the ampulla region of the oviduct and assessing the number of separate follicle rupture sites found on the corresponding ovary. **/***Significantly (p < 0.01/p < 0.001, respectively) different from untreated controls.

FIG. 2. Ovarian leukocyte distribution at 12 h post-hCG. Effects of L-NAME treatment on ovarian leukocyte distribution (mean SEM cells per unit area, n = 6 rats per group) in the thecal area of unruptured follicles with a diameter greater than 0.6 mm (designated preovulatory), and in 1 x 1-mm areas of the medullary stroma were immunohistochemically analyzed through use of mAbs MCA149 (specific for rat neutrophils) and ED1 (specific for rat monocytes/macrophages). */***Significantly (p < 0.05/p < 0.001, respectively) different from untreated controls.

maintained for the first 2 h of perfusion, normalizing by 3 h into the perfusions and until their completion. The flow rate profile of LH + IL-13 (RP)-perfused ovaries was subsequently made to exactly mimic that of the LH + L-NAME group by adjusting perfusion pressure between 82.8 and 53.3 mm Hg. IL-i1I was included with LH in this comparison group because of its known stimulation of ovulations [121, therefore facilitating detection of any reduction in ovulation rate attributable to flow rate depression caused by L-NAME. Ovulation rate (Fig. 5). The reduction of ovulation rate in rats administered L-NAME in vivo was similarly reflected in vitro. Ovaries co-perfused with LH and L-NAME ovulated significantly less (2.7 + 1 ovulations per ovary, p < 0.05) than their LH-perfused counterparts (8.2 + 1.6), while coperfusion of LH + IL-1, resulted in an ovulation rate (15.2 + 2.4) that was significantly greater than that of the LHperfused group. Adding L-NAME to the ovaries perfused

with LH + IL-1 led to a dramatic attenuation (p < 0.001) in the ovulatory response, resulting in 0.8 0.8 ovulations per ovary. A major portion of this effect could be attributed to the decrease in ovarian flow as demonstrated by the LH + IL-1 (RP)-perfused group, which ovulated 4.2 + 1.5 oocytes per ovary. Control ovaries perfused with medium alone showed no ovulations. NO accumulation (Table 1). Nitrite levels as a measure of NO accumulation during the perfusion period were measured in samples of the perfused ovaries obtained at 0, 1, 8, and 22 h after addition of the treatments. No significant difference in nitrite concentration existed at 0 h between the different treatment groups. Comparison of the rise in nitrite levels, standardized to each ovary's initial (0 h) nitrite concentration, showed only an approximately 2-fold rise in nitrite accumulation during the entire perfusion period of ovaries treated with LH + L-NAME and LH + IL-1P + L-

TABLE 1. Nitrite accumulation (mean t SEM nmol/ml) in perfusion medium at 0, 1, 8, and 22 h of in vitro rat ovary perfusions. Hours of perfusion Treatment' Control LH LH + L-NAME LH + IL-1p LH + IL-15 + L-NAME LH + IL-11 (RP)

0 3.7 4.6 2.4 1.4 1.1 3.2

+ 1.1 + 1.4 ± 0.2 ± 0.1 + 0.2 + 0.5

1 4.6 3.1 2.6 1.9 1.8 2.3

+ 1.6 ± 0.5 ± 0.2 ± 0.4 ±+0.4 ± 0.3

8 6.4 4.7 2.6 6.2 2.4 4.5

± 2.2

+ 0.8 ± 0.3 ± 1.0 ± 0.9 + 1.0

22 11.9 16.5 4.1 22.8 1.9 17.6

+ 0.7 ± 1.9 + 0.5*** + 2.2 + 0.6*** a + 1.9* **b

Fold rise' 4.7 5.4 1.8 17.2 2.1 6.1

+ 1.7 + 1.0 + 0.3 + 2.1** ± 1.0* **a ± 1.3***a

tDoses and experimental numbers as shown in Figure 5 legend. 'Fold rise at 22 h of perfusion. Standardized by dividing 22-h nitrite concentration by initial (0 h) concentration for each ovary within each treatment. ***/***a/***bSignificantly different (p < 0.001) from LH, LH + IL-1p, or LH + IL-1p+ L-NAME, respectively.

440

BONELLO ET AL.

441

OVULATORY EFFECTS OF NITRIC OXIDE INHIBITION

0

5

10

15

20

25

Control

LH

LH +L-NAME

LH +IL-18

Hours of Perfusion

FIG. 4. Flow rate (mean ± SEM ml/min, n = 4 ovaries per group) of perfusate passing through in vitro-perfused ovaries treated with LH (0.1 Vlg/ml) ± L-NAME (3.5 x 10-5 mol/L). Flow rate measurements were obtained every 15 min for the first 2 h of perfusion and then at 3, 4, 8, and 22 h after commencement of the perfusions. ***All samples of LH + L-NAME-treated ovaries within 2 h of commencing perfusions were significantly different (p < 0.001) from ovaries treated with LH alone.

NAME, demonstrating the effectiveness of L-NAME at inhibiting the production of NO in the system. An approximately 4.7-fold increase in nitrite accumulation occurred at the 22h time point in ovaries perfused with medium alone in comparison to L-NAME-perfused ovaries, while there were 5.4and 6.1-fold increases in nitrite levels in LH- and LH + IL1P (RP)-perfused ovaries over the same time period and a highly significant 17-fold increase in the LH + IL-13-perfused group; this indicated that IL-1P is a potent stimulator of NO production in perfused rat ovaries. Steroidsecretion (Fig. 6). In all groups of perfused ovaries except the group perfused with medium alone, there was the expected dramatic increase in secreted progesterone within the first 1-2 h after the addition of LH + treatments (Fig. 6, a and b). The peak progesterone production in ovaries perfused with LH + IL-113 + L-NAME occurred a little later, however, at 4 h after the beginning of treatment; but in this group, as in all groups exhibiting an initial peak, detection of progesterone in the medium gradually declined over the remaining period of perfusion because of the known binding of progesterone with the glass walls of the perfusion apparatus [27]. Inhibition of NO via the addition of L-NAME had no real effect on either LH- or LH + IL-13-

FIG. 3. Effects of NO inhibition on immunohistochemical localization of neutrophils (MCA149+) and monocytes/macrophages (ED1+) in the rat ovary at the expected time of ovulation (12 h post-hCG). MCA149+ cells (darkly stained) were localized to the thecal region of preovulatory follicles with significantly larger cell numbers inthe thecal area of saline-treated controls (a)than of L-NAME-treated rats (b). ED1 + cells were present in large numbers in the medullary stroma region of the ovary, while their density did not vary significantly between saline-treated controls (c)and L-NAME-treated rats (d). Magnification: a-b, x 100; c-d, x200.

LH+ IL-I1

+ L-NAME

LH

IL-IL (RP)

FIG. 5. In vitro ovulation rate. Oocytes (mean + SEM) ovulated from untreated perfused ovaries or from ovaries treated with LH (0.1 ig/ml) + IL-1I(4 ng/ml) LNAME (3.5 x 10- 5 mol/L). Experimental numbers: LH, LH + L-NAME, n = 10; LH + IL-1p (RP), n = 6; LH + IL-1p, n = 5; controls, LH + IL-1i + L-NAME, n = 4.

*/***Significantly different (p < 0.05/p < 0.001) from the LH-perfused group and the LH + IL-lp-perfused group, respectively.

perfused ovaries, resulting in a noticeably (p < 0.05) reduced progesterone level only at the 22-h time point in LH + IL-1 3-perfused ovaries. Estradiol secretion did seem to be at least partly regulated by NO (Fig. 6, c and d). L-NAME treatment of LH-perfused ovaries resulted in a reduction in estradiol secretion (p < 0.05), while treating LH + IL-13-perfused ovaries with the NO inhibitor invoked an even more substantial attenuation of estradiol release (p < 0.01) over most time points. Peak secretion of estradiol in each group generally occurred between 4 and 8 h after the commencement of perfusions and then plateaued for the remainder of the perfusion. DISCUSSION Administration of the inducible NOS inhibitors aminoguanidine and NG-methyl-L-arginine, either i.p. or into the ovarian bursal cavity, has previously been shown to significantly reduce ovulation rate in rats, an effect that could be reversed with the NO generator, sodium nitroprusside [18]. In the current study we reduced ovulation rate to similar levels in vivo, and to an even greater extent in the in vitroperfused rat ovary model, by treating rats, or their excised and perfused ovaries, with the NOS competitive inhibitor LNAME. The reduction in ovulation rate was associated with altered steroid secretion profiles in vitro and altered leukocyte distribution in vivo at the time of ovulation. The finding of a good correlation between follicle rupture sites and oocytes collected from the oviduct suggested that while L-NAME was inhibiting ovulation rate in vivo, the treatment was having no effects on transportation of the oocyte from the site of rupture to the ampullary region of

442

BONELLO ET AL.

Hours of Perfusion

Hours of Perfusion

CA -2

E S 0 4 -0 to

a.

5

5 Hours of Perfusion

Hours of Perfusion

FIG. 6. Time course of steroid secretion (mean + SEM pmol/ml) throughout the period of perfusion. a-b) Progesterone secretion by perfused ovaries treated as shown. cd) Estradiol secretion by perfused ovaries treated as shown. Treatment doses and experimental numbers as described in Figure 5 legend. b) */**/***Significantly different (p < 0.05/p < 0.01 /p < 0.001, respectively) from LH + IL-1 (RP) group; *"Significantly different (p < 0.05) from LH + IL-1p group. c) */**Significantly different (p < 0.05/ p < 0.01) from LH group. d) */**/***Significantly different (p < 0.05/p < 0.01/p < 0.001) from LH + IL-1I group.

the rat oviduct. A similar reduction of the ovulation rate in vitro by L-NAME indicated that the in vivo results were not attributable solely to indirect systemic effects of L-NAME but resulted from direct effects of the NOS inhibitor on the ovary. Interestingly, our in vitro results indicate that L-NAME was particularly effective in inhibiting the IL-13-mediated enhancement of the LH-induced ovulation rate reported by us previously [12]; this suggested that a major component of the increase in ovulation rate caused by IL-11 may be mediated by NO. Indeed it is noteworthy that the maximal production of NO occurred in the group co-perfused with LH and IL-13. Previous studies using ovarian dispersates from immature rats have shown the ability of IL-I1 to induce production of NO in the ovary [24, 32]. The present study indicates that this effect persists in the rat, after exposure of whole perfused ovaries to LH and up until and including the time of rupture, as shown in a similar study on whole ovarian dispersates of the gonadotropin-primed rat [331. IL-1p activates the inducible NOS (iNOS) in the ovary [24, 32], presumably requiring protein synthesis; this may account for the time delay in ni-

trite accumulation we observed in IL-l1-perfused ovaries. Furthermore, the induction by IL-13 of ovarian NO is receptor-mediated, and IL-i1 requires the participation of both granulosa and theca-interstitial cells to exert its cytotoxic and other effects [24, 32]. Hence a complex intrafollicular cellular communication system comprising IL-13B, its receptors, NOS, and NO is likely to exist at ovulation. While mRNA for IL- 1p and the IL-1 receptor type 1 (IL-IRtl) have been found in aspirated granulosa-luteal cells in the human [9, 13], rodent mRNA and protein for IL-10 and IL-1Rtl are known to be concentrated in the thecal-interstitial layer before ovulation [10, 11]. The mRNA and protein for a constitutive NOS (eNOS) have been localized to preovulatory human granulosa-luteal cells [19], and NOS activity has been shown in luteal phase thecal, granulosa, and CL cells of cynomolgus monkeys [20]; but to our knowledge no localization of iNOS in the ovary has yet been published. The nitrite data also confirm that i.-NAME effectively inhibits NO production throughout the course of perfusions and that NO production is not significantly increased in LHperfused ovaries as compared to ovaries perfused with me-

OVULATORY EFFECTS OF NITRIC OXIDE INHIBITION

dium only; this indicates that as shown previously [32], gonadotropin alone is not sufficient to induce NO production in vitro. However, the ovarian infiltration of leukocytes (a potent source and site of action for NOS-activating cytokines such as IL- 13) that follows hCG stimulation of rats in vivo [4] (but obviously cannot occur in the in vitro-perfused rat ovary) may well increasingly stimulate NO production as ovulation approaches. Perfusions with IL-13 increased LH-induced ovulations to a frequency comparable to that for hCG-induced ovulations in vivo in this study, perhaps indicating that exogenous IL-113 compensates for the lack of infiltrating leukocytes in vitro. The fact that the LH-perfused ovaries ovulated while untreated ovaries did not, when both groups demonstrated similar nitrite accumulation, is probably attributable to factors other than NO (such as the eicosanoids, serine proteases, kinins, and histamine) that are activated by LH and required in addition to basal NO levels in order for ovulation to occur. LH-perfused ovaries deprived of this apparently necessary background level of NO by L-NAME had very few ovulations, indicating that NO is obligatory for the ovulatory process. Human ovarian vein rings have been shown to constrict in vitro in response to stimulation with L-NAME [341. Hence, inhibitory effects on ovulation induced by the use of NOS inhibitors may be due to inhibition of NO-derived relaxation of ovarian vascular smooth muscle and unmasking of the vasopressor response to endothelin [35], leading to decreases in ovarian blood flow or, in the case of the in vitro-perfused ovaries, decreases in perfusate flow. Ovarian blood flow normally increases as ovulation approaches [36, 371 and contributes to the hyperemia and macroscopically observable "follicular blushing" that accompany the general increase in vascular permeability and edema observed in ovulating ovaries. By comparing the flow rates of LH-perfused and LH + L-NAME-perfused ovaries we have shown that there is an approximately 46% decrease in flow rate upon addition of the NOS inhibitor at a dose of 3.5 X 10 -5 mol/L; as mimicked in optimally stimulated ovaries, this resulted in a decreased ovulation rate that accounted for approximately 76% of the effect observed in this group when perfused with L-NAME. This result implies that there are vascular and nonvascular components to the reduced ovulation rate observed with NO inhibition and that the vascular effects are probably dominant. NO also promotes vascular permeability and edema formation [38]. Hence, it seems likely that NO is important for maintaining the state of ovarian vasorelaxation required to accommodate the ovulatory increase in blood flow, blood volume, and plasma exudation, and that at least part of this effect is regulated by IL-113. The importance of the role of leukocytes in ovulation has been established [5, 6]. Immunohistochemical localization of ovarian leukocytes in the current study confirmed the previously reported abundance of neutrophils over monocytes/macrophages in the theca and its reversal in the med-

443

ullary stroma of ovulating ovaries in the rat [4]. The significant depression in thecal and the concomitant increase in medullary neutrophils upon L-NAME treatment, which may be reflected by changes in the distribution of monocytes/ macrophages, may indicate a direct role for NO in modifying leukocyte trafficking at ovulation. NO has previously been implicated in promoting monocyte chemotaxis in humans [39]; this may explain the decreased numbers of thecal leukocytes found in animals treated with the NOS inhibitor. Similarly, ovarian NO may enhance leukocyte adhesion in thecal vessels during the periovulatory period. Evidence exists, however, that NO is an inhibitor of platelet and leukocyte adhesion rather than a promoter of the process [17]. An alternative explanation for the considerable trafficking of neutrophils from the follicular theca to the ovarian medullary stroma is preferential vasoconstriction of capillaries in the thecal area of follicles as a result of inhibition of the vasodilatory response to NO. Analysis of the effect of NO inhibition on steroidogenesis revealed a significant reduction in the secretion of estradiol (but not progesterone) in those perfused ovaries that exhibited a marked reduction in nitrite accumulation, suggesting that NO may have a role in modulation of ovarian estradiol secretion. This result complements findings of another study showing that serum nitrite levels increase during the follicular phase of the ovarian cycle in women and decrease after ovulation, correlating with serum estradiol but not with progesterone, FSH, or LH [231. It has previously been shown that inhibitors of the NOS enzyme significantly increase secretion of estradiol from human granulosa-luteal cells [19]; however, these experiments on isolated and cultured cells, lacking an intact vasculature and from another species, may not serve as an appropriate comparison to this study. The opposing view that estradiol regulates NO production has also been presented [23, 40, 41]. A similar study investigating the role of NO in regulating flow rate and steroidogenesis in the in vitro-perfused rat adrenal gland revealed that the marked decrease in flow rate resulting from L-NAME treatment was similarly associated with a drop in hormone secretion [42]. Hence, it is possible that a component of the decrease in estradiol secretion observed with L-NAME treatment may be explained by the resultant reduced flow rate. This possibility is supported by the partial reduction in secretion of estradiol observed in the LH + IL13 (RP) group. The specific importance of estradiol to the rupture process is not currently understood [43], although a possible role for estrogen in augmenting blood pressure control by NO in the rat has been suggested and may apply during the dramatic vascular changes that occur in the ovulating ovary [44]. Progesterone output during the periovulatory period was not altered by NO inhibition of perfused ovaries. Plasma progesterone was similarly unchanged during the early luteal period (20 h post-hCG) of in vivo-treated rats, sug-

444

BONELLO ET AL.

gesting that while follicle rupture was affected by L-NAME treatment, luteinization was proceeding normally, possibly because of the phenomenon of luteinized unruptured follicles [45]. We are unable to account for the marked increase in progesterone levels observed in LH + IL-p3 (RP) ovaries perfused with the artificially decreased flow rate, and further investigations are required to establish its physiological significance. There are alternative mechanisms by which NO may control ovulation. The physical processes of rupture and cellular reorganization accompanying the transformation of ovulatory follicles into young CL require extensive tissue remodeling and cell death. For example, NO has already been found to mediate the cytotoxic actions of IL-1 3 within the ovary [24] and so may be involved with the degradative events of ovulation. As with other free radicals, NO can damage DNA by base deamination, causing activation of ATP-consuming DNA repair enzymes, which can lead to cell death by energy depletion [17]. NO can also destroy cells by binding the heme- or non-heme iron of numerous enzymes to catalyze inhibition of vital DNA synthesis or can combine with free oxygen to produce a variety of cytotoxic factors [46]. Furthermore, a neural pathway for the NO effects on ovulation may exist as evidenced by reports identifying hypothalamic NOS-containing neurons that regulate LHRH secretion via guanylate cyclase [22] and NOS-containing nerve fibers present in the rat [21] but not the mouse ovary [47]. This study has shown that inhibition of NO significantly reduced the rate of ovulation in rats both in vivo and in vitro, and suggests that NO is likely to mediate the proovulatory actions of IL-1P3 that enhance LH-induced ovulations in vitro. Control of ovarian vessel relaxation to accommodate the necessary changes in blood flow, volume, and plasma exudation that accompany rupture is likely to be the most important role of NO in ovulation; but effects on steroidogenesis, leukocyte trafficking, and cytotoxicity may also exist and warrant further investigation. ACKNOWLEDGMENTS We wish to thank Dr. Hideo Tozawa and Vivian Pascoe for helpful advice and Jane Lomas, Maria Bellis, and Suzie Dosljak for expert technical assistance. We are also grateful to The National Hormone and Pituitary Program, University of Maryland, and NIDDK for the generous gift of LH.

REFERENCES 1. Espey L.Ovulation as an inflammatory reaction-a hypothesis. Biol Reprod 1980; 22:73106. 2. Cavender JL, Murdoch WJ. Morphological studies of the microcirculatory system of periovulatory ovine follicles. Biol Reprod 1988: 39:989-997. 3. Gerdes U, GAfvels M, Bergh A, Cajander S. Localized increases in ovarian vascular permeability and leucocyte accumulation after induced ovulation in rabbits. J Reprod Fertil 1992; 95:539-550. 4. Brinnstrom M, Mayrhofer G, Robertson SA. Localization of leukocyte subsets in the rat ovary during the periovulatory period. Biol Reprod 1993; 48:277-286.

5. Hellberg P,Thomsen P Janson PO, Brinnstrom M.Leukocyte supplementation increases the LH-induced ovulation rate in the perfused rat ovary. Biol Reprod 1991; 44:791-797 6. Brinnstrom M, Bonello N, Norman RJ,Robertson SA. Reduction of ovulation rate in the rat by administration of a neutrophil-depleting monoclonal antibody. J Reprod Immunol 1995; (in press). 7. Adashi EY. The potential relevance of cytokines to ovarian physiology: the emerging role of resident ovarian cells of the white blood series. Endocr Rev 1990; 11:454-464. 8. Brinnstrom M,Janson PO. Biochemistry of ovulation. In: Hillier S ed.), Ovarian Endocrinology. Oxford: Blackwell; 1991: 132-166. 9. Hurwitz A, Loukides J, Ricciarelli E, Botero L, Katz E, McAllister JM, Garcia JE, Rohan RILM, Adashi EY, Hernandez ER. Human intraovarian interleukin-1 (IL-) system: highly compartmentalized and hormonally-dependent regulation of the genes encoding IL-1, its receptor, and its receptor antagonist. J Clin Invest 1992; 89:1746-1754. 10. Simon C, Frances A, Piquette G, Polan ML. Immunohistochemical localization of the interleukin-1 system in the mouse ovary during follicular growth. ovulation and luteinization. Biol Reprod 1994; 50:449-457. 11. Hurwitz A, Ricciarelli E, Botero L,Rohan RM, Hernandez ER, Adashi EY. Endocrine- and autocrine-mediated regulation of rat ovarian (theca-interstitial) interleukin-l1 gene expression: gonadotropin-dependent preovulatory acquisition. Endocrinology 1991; 129:3427-3429. 12. BrnnstrOm M, Wang L,Norman RJ. Ovulatory effect of interleukin-lp on the perfused rat ovary. Endocrinology 1993; 132:399-404. 13. Piquette GN, Simon C, Danasouru IE, Frances A,Polan ML. Gene regulation of interleukin-1l, interleukin- receptor type 1, and plasminogen activator inhibitor-1 and -2 in human granulosa-luteal cells. Fertil Steril 1994; 62:760-770. 14. Bonello NP, Norman RJ. Brinnstr6m M. Interleukin-lp inhibits luteinizing hormoneinduced plasminogen activator activity in rat preovulatory follicles in vitro. Endocrine

1995; 3:49-54. 15. Simon C, Tsafriri A, Chun S-Y, Piquette GN, Dang W. Polan ML. Interleukin receptor antagonist suppresses human chorionic gonadotropin-induced ovulation in the rat. Biol Reprod 1994; 51:662-667. 16. Moncadla S, Palmer RM, Higgs EA. Nitric oxide: physiology, pathophysiology and pharmacology. Pharmacol Rev 1991; 43:109-142. 17. Bredt DS, Snyder SH. Nitric oxide: a physiologic messenger molecule. Annu Rev Biochem 1994; 63:175-195. 18. Shukovski L, Tsafriri A. The involvement of nitric oxide in the ovulatory process in the rat. Endocrinology 1994: 135:2287-2290. 19. Van Voorhis BJ, Dunn MS, Snyder GD, Weiner CP. Nitric oxide: an autocrine regulator of human granulosa-luteal cell steroidogenesis. Endocrinology 1994; 135:1799-1806. 20. Bukowski R. Hsiu JG, Queenan JT, Wu W, Williams RE, Goodman AL. Nitric oxide synthase in monkey and human granulosa and theca cells. In: Program of the 50th annual meeting of the American Fertility Society; 1994; San Antonio, TX. Abstract 0-031. 21. Jarrett WA, Price GT, Lynn VJ, Burden HW. NADPH-diaphorase positive neurons innervating the rat ovary. Neurosci Lett 1994; 177:47-49. 22. Moretto M,LOpez FJ, Negro-Vilar A.Nitric oxide regulates luteinizing hormone-releasing hormone secretion. Endocrinology 1993; 133:2399-2402. 23. Rosselli M, Imthurm B, Macas E, Keller PJ, Dubey RK. Circulating nitrite/nitrate levels increase with follicular development: indirect evidence for estradiol mediated NO release. Biochem Biophys Res Commun 1994; 202:1543-1552. 24. Ellman C, Corbett JA, Misko TP, McDaniel M, Beckerman KPRNitric oxide mediates in terleukin-l-induced cellular cytoxicity in the rat ovary. A potential role for nitric oxide in the ovulatory process. J Clin Invest 1993; 92:3053-3056. 25. Rees DD, Palmer RMJ, Schulz R, Hodson HE, Moncada S. Characterization of three inhibitors of endothelial nitric oxide synthase in vitro and in vivo. Br J Pharmacol 1990; 101:746-752. G 26. Wang Y X, Poon Cl. Pang CCY. Vascular pharmacodynamics of N nitro-L-arginine methyl ester in vitro and in vivo. J Pharmacol Exp Ther 1993; 267:1091-1099. 27. Brdnnstrdm M,Johansson BM, SognJ, Janson PO. Characterization of an in vitro perused rat ovary model: ovulation rate, oocyte maturation, steroidogenesis and influence of PMSG priming. Acta Physiol Scand 1987; 130:107-114. 28. Green LC,Wagner DA, Glogowski J, Skipper PL, Wishnok JS, Tannenbaum SR. Analysis of nitrate, nitrite and [N]nitrite in biological fluids. Anal Biochem 1982: 126:131-138. 29. Billet EE, Gunn B, Mayer RJ. Characterization of two monoclonal antibodies obtained by immunization with human liver mitochondrial membrane preparations. Biochem 1984; 221:765-776 30. Dijkstra JM, Dopp EA, Joling P, Kraal G. The heterogeneity of mononuclear phagocytes in lymphoid organs: distinct macrophage subpopulations in the rat recognized by monoclonal antibodies EDI, ED2 and ED3. Immunology 1985: 54:589-599. 31. Osman P. Rate and course of atresia during follicular development in the adult cyclic rat. J Reprod Fertil 1985; 73:261-270. 32. Ben-Shlomo 1, Kokia E, Jackson MJ, Adashi EY, Payne DW. Interleukin-l1 stimulates nitrite production in the rat ovary: evidence for heterologous cell cell interaction and for

OVULATORY EFFECTS OF NITRIC OXIDE INHIBITION

33.

34. 35.

36.

37.

38. 39.

the insulin-mediated regulation of the inducible isoform of nitric oxide synthase. Biol Reprod 1994; 51:310-318. Olson LM,Ellman CF, Beckerman KP. Ovarian nitric oxide production at different stages of the cycle. In: Program of the 75th annual meeting of the Endocrine Society; 1993; Las Vegas, NV. Abstract 1633. Stones RW, Loesch A, Beard RW, Burnstock G. Substance P: endothelial localization and pharmacology in the human ovarian vein. Obstet Gynecol 1995; 85:273-278. Richard V, Hogie M,Clozel M, Loffler BM, Thuillez C. In vitro evidence of endothelininduced vasopressor tone after inhibition of nitric oxide synthesis in rats. Circulation 1995; 91:771-775. Brown BW, Cognie Y,Chemineau P, Poulin N, Salama OA. Ovarian capillary blood flow in seasonally anoestrus ewes induced to ovulate by treatment with GnRH. J Reprod Fertil 1988; 84:653-658. Abisogun AO, Daphna-lken D, Reich R, Kranzfelder D, Tsafriri A. Modulatory role of eicosanoids in vascular changes during the preovulatory period in the rat. Biol Reprod 1988; 38:756-762. Ialenti A, lanaro A, Moncada S, DiRosa M.Modulation of acute inflammation by endogenous nitric oxide. EurJ Pharmacol 1992; 211:177-182. Belenky SN, Robbins RA, Rubinstein I. Nitric oxide synthase inhibitors attenuate human monocyte chemotaxis in vitro. J Leukocyte Biol 1993; 53:498-503.

445

40. Stallone JN. Sex differences in nitric oxide-mediated attenuation of vascular reactivity to vasopressin are abolished by gonadectomy. EurJ Pharmacol 1994; 259:273-283. 41. HuangJ, Roby KF, Pace JL, Russel SW, HuntJS. Cellular localization and hormonal regulation of inducible nitric oxide synthase in cycling mouse uterus. J Leukocyte Biol 1995;

57:27-35. 42. Cameron LA, Hinson JP. The role of nitric oxide derived from L-arginine in the control of steroidogenesis, and perfusion medium flow rate in the isolated perfused rat adrenal gland. J Endocrinol 1993; 139:415-423. 43. Tsafriri A, Dekel N. Molecular mechanisms in ovulation. In: Findlay J (ed.), Molecular Biology of the Female Reproductive System. San Diego: Academic Press; 1994: 207-258. 44. Brosnihan KB, Moriguchi A, Nakamoto H, Dean RH, Ganten D, Ferrario CM. Estrogen augments the contribution of nitric oxide to blood pressure regulation in transgenic hypertensive rats expressing the mouse Ren-2 gene. AmJ Hypertens 1994; 7:576-582. 45. Check JH, Dietterich C, Nowroozi K, Wu CH. Comparison of various therapies for the luteinized unruptured follicle syndrome. Int J Fertil 1992; 37:33-40. 46. Stamler JS, Singel DJ, Loscalzo J. Biochemistry of nitric oxide and its redox-activated forms. Science 1992; 258:1898-1902. 47. Grozdanovic Z, Mayer B, Baumgarten HG, Bruning G. Nitric oxide synthase-containing nerve fibers and neurons in the genital tract of the female mouse. Cell Tissue Res 1994; 275:355-360.