- Email: [email protected]
Ovarian effects of SK&F 86002-A2 in the rat: Site of action
TOXICOLOGYANDAPPLIEDPHARMACOLOGY
94,276-286(1988)
Ovarian Effects of SK&F 86002-A* RICHARD
in the Rat: Site of Action
F. WALKER,* LESTER W. SCHWARTZ,? THEODORE JOHN F. NEWTON,~
J. TORPHY,$
AND JEANNE M. MANSON*
Departments of *Reproductive and Developmental Toxicology, TExperimental Pathology, SPharmacology, and QDrug Metabolism, Smith Kline & French Laboratories. Research and Development, King of Prussia, Pennsylvania 19406-0939
Received October 6, 1987; accepted February 25, I988 Ovarian Effects of SK&F 86002-A2 in the Rat: Site of Action. WALKER, R. F., SCHWARTZ, L. W.,TORPHY,T.J.,NEWTON,J.F.,ANDMANSON,J.M. (1988). Toxicol. Appl. Pharmacol. 94,276-286. In a preliminary 30-day study, oral administration of SK&F 86002-A2, an inhibitor of prostaglandin and leukotriene synthesis, blocked ovulation and altered ovarian structure and hormone production in rats. The purpose of the present study was to determine if the locus of action of SK&F 86002-A* for these effects was the ovary or some other site in the female reproductive system, using a number of experimental approaches. A single SCor intraovarian injection of SK&F 86002-A* did not block spontaneous or gonadotropin-induced ovulation in proestrous rats, whereas indomethacin, a positive control, acutely disrupted the ovulatory process. Since neither route of administration blocked ovulation, integrated pituitary and ovarian events were not negatively affected by a single injection of SK&F 86002-A* at doses which caused ovarian dysfunction when administered repeatedly for 30 days. In contrast to a single dose, oral administration of SK&F 86002-A* to hypophysectomized rats for 2 weeks suppressed follicular growth and estradiol production in response to sc administration of pregnant mare serum gonadotropin. Although ovarian function was suppressed in hypophysectomized rats, LH surges induced by estradiol in ovariectomized rats were not at&ted by administration of SK&F 86002AZ for 2 weeks. Thus, hypothalamic/pituitary dysfunction did not contribute to the ovarian effects of SK&F 86002 that occurred after repeated dosing. In conclusion, these results indicate that disruption of ovarian cycles by SK&F 86002-A* is related to a direct effect on the ovary, and not to altered hypothalamic/pituitary function and LH release. Specifically, SK&F 86002AZ may suppress the ovarian response to gonadotrophin, retarding follicular growth and estrogen production. The ovarian effects are consistent with a pharmacological expression of the inhibitory action of SK&F 86002-A* on prostaglandin and leukotriene synthesis. o 1988 Academic Press, Inc.
SK&F 86002-AZ, [5-(4-pyridyl)-6(4-fluorophenyl) - 2,3 - dihydroimidazo - (2,1- 6) - thia zole] dihydrochloride, inhibits cyclooxygenase and 5-lipoxygenase pathways of the arachidonic acid metabolic cascade (DiMartino et al., 1987). The dual inhibitory action of SK&F 86002-A* was confirmed by the following observations: ( 1) impaired production of the 5-lipoxygenase products leukotriene B4 (diHETE) and 5-HETE by RBL-1 cells, (2) impaired production of cyclooxygenase products (PGE2) by inflammatory macro0041-008x/88
$3.00
Copyright 8 1988 by Academic Press, Inc. All rights of reproduction in any form reserved.
phages in vitro, and by (3) reduced production of LTB., in vitro by peritoneal exudate cells harvested from SK&F 86002-AZ-treated mice. The results of a previous study from our laboratory showed that oral administration of SK&F 86002-A* to rats for 30 days caused estrous cycle irregularity in a dose-related fashion (Walker et al., 1988). Disruption of regular estrous cycles correlated with structural and biochemical changes in the ovaries, which were often hypertrophied to the extent 216
277
SK&F 86002-A* SITE OF ACTION
that they could be differentiated from controls by visual inspection. Histological examination revealed increased numbers of luteinized follicles with retained ova and occasional cystic follicles. The histologic changes were accompanied by altered levels of plasma estradiol and progesterone, while the levels of plasma pituitary hormones (LH, FSH and prolactin) were comparable to controls. These preliminary findings suggested that SK&F 86002-A* disrupted cyclic ovarian function by a local, cumulative action that inhibited ovulation and altered steroid secretion. However, basal but not surge levels of pituitary hormones were estimated, and the data were insufficient to conclude that SK&F 86002-A* did not alter preovulatory secretion of gonadotropins. The ovulation-blocking action of prostaglandin inhibitors such as indomethacin can occur at hypothalamic/pituitary (Orczyk and Behrman, 1972; Ojeda et al., 1975; Ojeda and Campbell, 1982) or ovarian (Tsafiiri et al., 1973; Armstrong and Grinwich, 1972; Behrman et al., 1972; Wallach et al., 1975) sites depending upon the dose administered. Similarly, inhibitors of leukotriene synthesis have the potential to alter ovarian function by blocking follicular rupture (Reich et al., 1983), steroidogenesis (Dix et al., 1984), or the release of luteinizing hormone (Snyder et al., 1983; Conte et al., 1986). Since SK&F 86002-A2 is a dual inhibitor of cyclooxygenase and lipoxygenase (DiMartino et al., 1987) it may cause ovarian dysfunction by affecting one or more sites within the female reproductive/neuroendocrine axis. For example, ovulation may be prevented indirectly as the result of hypothalamic/pituitary dysfunction in which the preovulatory LH surge is lost, or directly as the result of ovarian dysfunction in which follicular rupture is blocked. Alternatively, events which precede and promote ovulation such as folliculogenesis and steroidogenesis may be altered causing ovarian dysfunction. Thus, the purpose of this study was to determine the site at which SK&F 86002-A2 blocks ovula-
tion, and to better understand the mechanism by which it causes ovarian dysfunction in the rat. METHODS
AND
MATERIALS
SK&F 86002-A?, which is the dihydrochloride salt of SK&F 86002 [6-(4-fluorophenyl)2,3-dihydro-5-(4-pyridinyl)imidazo(2,1 -b)-thiazole], was synthesized locally (SK&F Laboratories, King of Prussia, PA) (DiMartino et al., 1987). It is readily soluble in water (>80 mg/ml) and was prepared for use in these studies as a sterile aqueous solution [ 100 mg (base)/ml, pH 1.41. Animals andfacilities. Sexually mature (75-85 days of age) female rats (Sprague-Dawley/CD-VAF) that were intact, ovariectomized, or hypophysectomized were purchased from Charles River Breeding Laboratories (Cambridge, MA). Prepubertal rats (22-28 days old) that were used to test the effect of SK&F 86002-A* on gonadotropin-induced superovulation were from the same source. All animals were housed in a barrier facility (SK&F R&D Labs, King of Prussia, PA) in clear polycarbonate boxes containing Alpha-dri bedding (Sheperd Specialty Papers, Inc., Kalamazoo, MI) and maintained under standard conditions of light ( 12 hr; 0600- 1800 hr), temperature (72 f 4°F) humidity (50 f 10% relative), and nutrition (Certified Rodent Chow, Ralston Purina Co., St. Louis, MO, and tap water, ad libitum). Quantitation of SK&F 86002-A2 and metabolites in plasma. In this study, SK&F 86002-A* was administered by gavage as well as subcutaneous injection. Plasma concentrations of SK&F 86002-A? and its metabolites (SK&F 86096 and SK&F 104343) were measured for up to 6 hr following sc or po administration. Blood samples (100 ~1) were collected from a tail vein into heparinized Eppendorf microtubes at various times after dosing. Plasma was separated by centrifugation at 4’C and a known volume was transferred to a polypropylene test tube which contained internal standard, SK&F 85838 (6-(4’-lluorophenyl)-5-(4’-methoxyphenyl)-2,3-dihydroimidazo-(2,1 -b)-thiazole- 1-oxide, hydrate: 100 ~1 of a 2 &ml solution), and the mixture was stored at -20°C until analysis. Plasma concentrations of SK&F 86002 and its sulfoxide and sulfone metabolites, SK&F 86096 and 104343, were quantitated by HPLC following solid phase extraction using Baker 10 SPE (J.T. Baker Chemical Co., Phillipsburg, NJ) I ml Cls columns preconditioned with methanol and water. Samples containing internal standard were added to the columns which were then washed with ammonium acetate (100 mM). SK&F 86002 and its metabolites were eluted from the column with methanol (2 X 250 ~1). Samples were dried under nitrogen and resuspended in HPLC mobile phase (100 pl), and aliquots (50 ~1) of the reconstituted samples were quantitated on
278
WALKER
an HPLC system consisting of a Model 590 solvent delivery system, a Model 7 1OB autoinjector, a Model 44 1 uv detector set at 254 nm, and a Model RCM 100 radial compression module which contained a 5-pm NOVAPAK C,s (8 X IO-mm) cartridge (Waters Assoc., Inc., Milford, MA). SK&F 86002, metabolites, and the internal standard were separated isocratically using a solvent system consisting of 29.5% acetonitrile in 100 mM ammonium acetate at a flow rate of 2.0 ml/min. Plasma concentration of SK&F 86002 and metabolites were determined from peak area ratios (compared to internal standard) that were compared to extracted authentic standard curves. Ovarian function: Single dose studies. The first objective of this study was to determine if SK&F 86002-A* directly affects the ovary by altering its response to spontaneous or gonadotropin-induced ovulation. This was accomplished in one set of experiments by direct injection of SK&F 86002-A2 into the ovaries of proestrous rats followed by measurement of spontaneous ovulation in response to the endogenous LH surge. Sexually mature female rats with unambiguous proestrus vaginal smears were anesthetized at 1430 hr with CO2 and ether. Each ovary was surgically exposed and injected with 2 pl of vehicle or vehicle containing 200 pg SK&F 86002-Al. Unanesthetized, uninjected rats were also included in the test to control for the possible effects of inhalation anesthesia on the preovulatory LH surge. At 0700 hr on the following morning, the rats were killed and tubal ova were counted and compared among groups. Next, SK&F 86002-A2 was tested for potential effects on gonadotropin-induced superovulation. Prepubertal (24-27 days of age) female rats were given SCinjections of pregnant mare serum (PMS, 30 IU, 0.2 ml) at 1200 hr. Forty-eight hours after receiving PMS to initiate follicular growth, the rats were anesthetized with ether and their ovaries were exteriorized through bilateral incisions in the dorsolumbar flank. The ovaries were then injected with SK&F 86002-A2 (2 ~1 containing 10 or 200 rg) or an equivalent volume of vehicle. Shortly thereafter, the rats received an SCinjection of human chorionic gonadotropin (HCG; 20 IU) to induce ovulation (Zarrow et al., 1958). Twenty hours after receiving HCG the rats were killed by COZ asphyxiation. Their fallopian tubes were removed and examined microscopically for the presence of ova which were then counted and the numbers compared among treated and control groups. Subcutaneous injections of SK&F 86002-A* were also administered to adult rats to test for a direct ovarian effect on ovulation. Sexually mature female rats with unambiguous proestrus vaginal smears were injected with pentobarbital(30 mg/kg, ip) at 1430 hr to block the neurogenic stimulus for the preovulatory LH surge (Everett and Sawyer, 1950). A single injection of ovine LH (30 rg) was administered 15 min later. Barbiturate blockade of the LH surge followed by LH administration removes central elements of the reproductive axis from influenc-
ET AL. ing ovarian function. Thus, any effect of systemic drug administration on ovulation in this model must occur at the level of the ovary. SK&F 86002-A2 (60 mg/kg) was administered by a single sc injection. Indomethacin (10 mg/kg) was given to other barbiturate-blocked, gonadotropin-treated rats as a positive control. The rats were killed at 0700 hr on the following morning, and tubal ova were counted and the numbers compared among drugtreated and control groups. Ovarian function: Multiple dose studies. Hypophysectomized rats were allowed a postsurgery recovery period of 30 days during which time ovarian atrophy occurred in the absence of endogenous gonadotropins. Then, SK&F 86002-A2 (60 mg/kg/day) or vehicle was administered by gavage for 18 consecutive days. In addition, each rat received daily injections of PMS ( 100 IU) for the last 4 consecutive days of dosing. Vaginal smears were monitored during PMS administration for signs of cellular comification as an indication of gonadal recrudescence. At the end of gonadotropin treatment the rats were killed, their ovaries were fixed for histological examination, and their blood was collected for estimation of estradiol levels by radioimmunoassay. Pituitary function. Pituitary function was assessedby examining estrogen-stimulated release of LH in ovariectomized rats that were repeatedly dosed with SK&F 86002-A*. In intact rats, the preovulatory LH surge occurs in response to increasing levels of serum estradiol, produced by growing ovarian follicles (Legan et al., 1975). After ovulation, progesterone produced by the corpora lutea prevents an LH surge from recurring. Cohorts of growing follicles that produce estradiol in subsequent cycles trigger the intermittent LH surges that are associated with each ovulation. This ovarian influence on preovulatory LH release can be eliminated by ovariectomy (OVX). When serum estradiol levels are pharmacologically elevated in OVX rats, an LH surge will occur each day because the inhibitory effects of progesterone are eliminated. OVX rats are useful when investigating the effect of drugs on central components of the female reproductive axis since LH surges are artificially induced by estrogen in the absence of the ovaries. Thus, the estradiol-supplemented OVX rat was used to test the effects of SK&F 86002-A* on the pituitary mechanism controlling the LH surge. Sexually mature, ovariectomized rats were gavaged with SK&F 86002-A2 (10 or 60 mg/kg) or vehicle for 14 consecutive days. On Day 9 of dosing, the rats received SCimplants of Silastic tubing (5 mm) containing estradiol 170, which initiated daily LH surges (Legan et al., 1975). Each rat was bled from a tail vein at 1000, 1300, 1500, and 1700 hr, five days after the estradiol capsules were implanted to estimate serum levels of LH by radioimmunoassay. These times were selected because the LH surge begins at approximately 1400 hr and peaks at approximately 1700 hr under the lighting conditions of this study. Serum LH values were compared among groups
279
SK&F 86002-A* SITE OF ACTION to determine if repeated daily dosing with SK&F 86002Az disrupts the hypothalamic/pituitary mechanism for preovulatory secretion of LH. Radioimmunoassay. Serum estradiol concentrations were estimated using commercially available RIA kits purchased from Diagnostic Products Corp. (Los Angeles, CA 90045). The assay was sensitive to 10.0 pg/ml estradiol with interassay and intraassay coefficients of variation of 10.3 and 7.2%, respectively. Serum LH levels were estimated using reagents provided by the National Pituitary Agency (NIH). Assay sensitivities as well as interassay and intraassay coefficients of variation for the LH RIA were 460 pg, 10.1, and 8.2%, respectively. Serum levels of LH were expressed in terms ofthe reference preparation, LHRP-2. Phosphodiesterase assay. Cyclic AMP phosphodies terase (PDE) was prepared from monocytes (75 X IO6 cells) that were recovered from two units of human peripheral blood (Colotta et al., 1984). The monocytes were suspended in 2.5-3.0 ml of 50 mM Tris-HCl buffer (pH 7.5) containing 0.1% Triton X- 100 and a protease inhibitor mixture of 5 mM EDTA, 10 &ml soybean trypsin inhibitor, 17 pg/ml benzamidine, 100 fig/ml bacitracin, and 10 pM phenylmethyl sulfonyl fluoride. The cells were disrupted by sonification for 30 set using a Branson Sonifier Cell Disruptor 200. Sonilication of human monocyte suspensions in the presence of Triton X- 100 releases essentially all particulate bound PDE activity. The cell-free suspension was then centrifuged at 20,OOOgfor 30 min at 4’C and the supernatant collected and applied to a DEAE-Sepharose column to separate PDE isozymes according to the method of Reeves et al. (1987). Enzyme activity was measured in column fractions containing the CAMP-selective PDE by a modification of the method of Davis and Daly (1979). All assayswere conducted in the linear range of the reaction where less than 20% of the initial substrate was hydrolyzed. Final concentrations in the reaction mixture were 50 mM Tris-HCl buffer (pH 7.5), 5 mM MgCl*, 50 WM S’AMP (carrier), 1 pM 3’5’-[3H]cAMP (approximately 2000 dpm/pmol) and 50-~1 fraction aliquots. The reaction was run at 3o”C, initiated by the addition of substrate, and terminated after 30 min by placing the reaction vesselsin a 1WC heating block for 1 min. Various concentrations (l-300 pM) of SK&F 86002-A2 were tested for PDE inhibition using [3H]cAMP concentrations of 0.1,0.3, and 1.OpM. Aliquots ofthe enzyme were preincubated with SK&F 86002-A2 at room temperature for 10 min prior to addition of [3H]cAMP. The 3’5’-cyclic nucleotide was separated from the 5’ nucleotide product by first adding 0.5 ml ofO.1 M Hepes buffer (pH 8.5) containing 0.1 M NaCl to each sample. The samples were then applied to individual polyacrylamide-boronate gel column (0.5 g of Biorad Affi-gel 601 on a 0.7 X IO-cm Biorad econo-column) which had been equilibrated with the 0.1 M Hepes/O. 1 NaCl buffer (pH 8.5) and the unreacted cyclic nucleotides were eluted from the column
unmiected
vehicle
?,K.w
86002-A*
TREATMENT FIG. 1. Effect of intraovarian injections of SK&F 86002-A* on spontaneous ovulation. *p < 0.05 compared to uninjected controls. N = 9 (uninjected), 6 (vehicle), 5 (SK&F 86002-A&
with 10 ml of Hepes buffer. The 5’ nucleotide product was then eluted into a scintillation vial with 10 ml 0.25 M acetic acid. Ten milliliters of Beckman Ready SolvMP scintillation cocktail was added to the eluate and shaken to form a gel, and the radioactivity was measured via liquid scintillation spectrometry. The CAMP PDE inhibition constant (&) for SK&F 86002-A* was determined according to the method of Dixon (1952). Statistical analyses of hormone levels, shed ova, and ovarian follicle counts were conducted using ANOVA and Duncan’s multiple range test.
RESULTS Data presented in Fig. 1 compare the numbers of ova shed from ovaries of sexually mature rats that received intraovarian injections of either SK&F 86002-A2 or vehicle on proestrus, to the numbers of ova shed from ovaries of unanesthetized, uninjected control rats. The mean number of ova shed from SK&F 86002-AZ-treated rats was lower than that from unanesthetized, uninjected controls. However, the differences were not statistically significant. In contrast, significantly fewer ova (p c 0.01) were shed by vehicleinjected controls compared with SK&F 86002-AZ-treated rats or unanesthetized, un-
280
WALKER
subcapsular
intraovarian
INJECTION SITE FIG. 2. Superovulation in immature gonadotropintreated rats receiving SK&F 86002-A2 by ovarian injection N = 9 for each treatment and route.
injected controls. Blockade of ovulation in the vehicle-injected control group was most likely due to anesthesia which depressed the preovulatory LH surge. Intraovarian injection of vehicle or SK&F 86002-A* in prepubertal female rats did not suppress ovulation when exogenous gonadotropin was administered, as indicated from data presented in Fig. 2. Thus, when the endogenous LH surge was blocked by anesthesia, SK&F 86002-A2 may have actually stimulated ovulation in the presence of the anesthetic block. The data in Fig. 3 compare the effects of subcutaneously administered SK&F 86002A2 on gonadotrophin-induced ovulation in the “pentobarbital-blocked” proestrus rat. As seen in Fig. 4, oral doses of SK&F 86002-A2 above 30 mg/kg did not increase maximal plasma concentrations of the drug, so subcutaneous doses were administered. The subcutaneous route produced higher plasma concentrations than equivalent doses given orally. Pentobarbital was administered SCto
ET AL.
block the neurogenic stimulus for the LH surge and consequently, exogenous LH was injected to eliminate central factors governing ovulation from consideration in this experiment. Under these conditions, ovulation was blocked by indomethacin (p -C0.0 1) but not by SK&F 86002-A2 or vehicle (Fig. 3). Although SK&F 86002-A2 did not block gonadotropin-induced ovulation, it was of interest to determine whether a systemic injection of the drug would suppress ovulation in response to the preovulatory LH surge. The experiment was repeated with proestrus rats that were not pentobarbital-blocked. As seen in Fig. 5, SK&F 86002-A* did not block spontaneous ovulation in these rats and in fact, the mean number of tubal ova was slightly higher in drug-treated than in control rats. Thus, the preceding experiments showed that single injections of SK&F 86002-A* did not block ovulation at the ovarian level, and may in fact have stimulated it slightly. The effect of 14 consecutive, daily doses of SK&F 86002-A* on PMS-induced follicular maturation and estradiol synthesis in hy-
FIG. 3. Comparison of the effects of SK&F and indomethacin on gonadotropin-induced in “pentobarbital-blocked” rats. *p < 0.05 with values from the same group at 1000 hr. each treatment.
86002-A* ovulation compared N = 6 for
SK&F 86002-A2 SITE OF ACTION 20
1
SK&F 86002
11
Sulfoxide
a~
281
I]
Sulfone
15
Time
After
Administration
(hrs)
FIG. 4. Plasma concentrations of SK&F 86002 and its metabolites following oral or istration of SK&F 86002 [30 mg (base)/kg].
pophysectomized rats is shown in Table 1. Vaginal cornification did not occur, and the number of developing follicles and serum estradiol levels in hypophysectomized rats treated with PMS and SK&F 86002-A* were significantly less (p < 0.05 for both parameters) than in PMS-treated control rats. Histological evaluation of ovaries from these
subcutaneous
admin-
hypophysectomized rats corroborated the vaginal cytology and serum estradiol concentrations in that mature or preovulatory follicles were only infrequently observed in SK&F 86002-AZ-treated rats (Fig. 6). The effect of 14 days of dosing with SK&F 86002-A* on estrogen-induced LH release in ovariectomized rats is presented in Fig. 7. Serum levels of LH rose comparably from 1000 to 1700 hr after 5 days of estradiol exposure in ovariectomized rats dosed with 0, 10, or 60 mg/kg SK&F 86002-A* for 14 consecutive days, indicating a lack of drug effect on central sites controlling the LH surge. The data in Fig. 8 show that SK&F 86002A* is a competitive inhibitor of phosphodiesterase with a Ki of approximately 10 j.&M. DISCUSSION
“EHlCLE
SKW
aeow-A2
TREATMENT FIG. 5. Effect of systemic injection of SK&F 86002-A* on spontaneous ovulation. N = 5 for each treatment.
The major finding of this study was that repeated doses of SK&F 86002-A2 retarded follicular maturation and inhibited estradiol production in the rat ovary. These ovarian effects could not be attributed to defects in gonadotropin secretion since estradiol-induced LH surges occurred in rats that received multiple doses of SK&F 86002-AZ under conditions which disrupted estrous cycles and blocked ovulation in an earlier study (Walker et al., 1988).
282
WALKER TABLE
ET
AL.
1
EFFECT OF REPEATED DOSING WITH SK&F 86002-A2 ON OVARIAN IN GONADOTROPIN-TREATED HYPOPHYSECTOMIZED
*
p
Treatment
N
Vehicle SK&F 86002-A,
6 6
< 0.05 compared
Vaginal comification + -
Follicles (2 + SE) 12.3 + 2.7 6.1 f 1.5
RECRUDESCENCE RATS Serum (p&ml;
estradiol 1 f SE)
40.5 + 4.1 28.4 + 3.1*
with control.
Unlike other inhibitors of prostaglandin or leukotriene synthesis that block ovulation after acute administration (Orczyk and Behrman, 1972; Armstrong and Grinwich, 1972; Reich et al., 1983), single doses of SK&F 86002-A2 had no effect on spontaneous or gonadotropin-induced ovulation. This unexpected observation may have resulted from the activity of SK&F 86002-A* to inhibit phosphodiesterase in addition to its inhibitory effects on arachidonic acid metabolism. The biological actions of prostaglandins are directly linked to changes in CAMP. When prostaglandins increase intracellular CAMP, they enhance target cell functions (Lee and Katayama, 1981). Following the LH surge, CAMP rises in preovulatory follicles, initiating the process of rupture, ovulation, and luteinization (Nilsson et al., 1977). Since inhibition of phosphodiesterase causes intracellular CAMP to accumulate (Nathanson, 1977), this additional effect of SK&F 86002-AZ may have circumvented its potential antiovulatory action resulting from blockade of prostaglandin synthesis. This hypothesis is supported by the fact that SK&F 86002-A2 stimulated ovulation in rats whose LH surges were partially blocked by COz/ether anesthesia on the day of proestrus. When proestrus rats were anesthetized for surgery and given intraovarian injections of SK&F 86002-A2 or vehicle, significantly more eggs were shed by drug-treated than control rats. The low number of eggs shed by control rats is explained by the fact that anesthesia administered on the day of proestrus can block the neurogenic
stimulus for the preovulatory LH surge (Everett and Sawyer, 1950). COz/ether anesthesia given during proestrus in this study reduced ovulation in control rats, while ovulation was not as profoundly affected in rats injected with SK&F 86002-A*. If SK&F 86002-A2 inhibited ovarian phosphodiesterase in addition to cyclooxygenase, then it mimicked in part the ovulatory effects of LH by raising ovarian CAMP. Since a full complement of ova was not shed in these rats, the ovulatory blockade expected from inhibition of prostaglandin synthesis by SK&F 86002A2 may have been only partially reversed by phosphodiesterase inhibition. Stimulatory effects of SK&F 86002-A2 on ovulation were also seen in gonadotropin-treated rats in which the mean number of eggs shed was higher than in controls. Thus, as a consequence of its phosphodiesterase inhibiting action, the potential antiovulatory action of SK&F 86002-A2 resulting from inhibition of prostaglandin synthesis may have been masked, and late stage ovulatory events such as follicular rupture were not blocked by a single dose. In contrast to late stage events, earlier events in the ovarian cycle were disrupted following repeated dosing with SK&F 86002AZ. In cycling, adult female rats, SK&F 86002-A2 caused estrous cycle irregularity and ovarian changes that included luteinization of unruptured follicles (Walker et al., 1988). In an attempt to determine the cause of these effects, the estrogen-stimulated LH surge was examined in rats treated for 14 days
FIG. 6. Photomicrographs of ovaries obtained from hypophysectomized rats administered gonadotropins. (A) Ovary containing two preovulatory follicles from a control rat administered gonadotropins and vehicle. (B) Ovary obtained from a gonadotropin-treated rat administered SK&F 86002-A2 for 14 consecutive days. Maturing or preovulatory follicles were not present. Magnification 164X. 283
284
WALKER
6 ‘55 E $4 3 5” 5 m2
1
0 lOOOh
I 1300h
J
1500h
1700h
TIME
FIG. 7. Estrogen-induced LH surges in ovariectomized rats dosed repeatedly with SK&F 86002-A2 *p < 0.05 compared with values at 1000 hr. N = 6 for each treatment.
with SK&F 86002-AZ. Repeated administration of SK&F 86002-A2 had no effect on LH surges produced by exogenous estradiol in OVX rats, suggesting that the hypothalamic/ pituitary mechanism for preovulatory release of LH remained intact, despite repeated exposure to SK&F 86002-AI. A direct ovarian effect on steroidogenesis was discovered in hypophysectomized rats. The ovaries of these animals were refractory to exogenous gonadotropin following repeated doses of SK&F 86002-A2 since follicular growth and estrogen production were significantly retarded. Knowing that estrogen priming increases the sensitivity of ovaries in hypophysectomized rats to gonadotropins (Zeleznik et al., 1974), that follicular growth is inhibited when the biologic activity of estrogen is neutralized with antibody (Harman et al., 1975), that estradiol exerts a direct mitogenic effect on granulosa cells to increase their number and thereby enhance the effects of FSH on follicular proliferation (Hsuek et
ET AL.
al., 1983), and that leukotriene inhibitors block steroid synthesis in response to gonadotropin (Dix et al., 1984), the data suggest that the ovarian effects of SK&F 86002-A2 resulted from inhibition of estradiol production. Since the synergy between estradiol and FSH represents an intraovarian positive feedback mechanism to promote follicular growth, inhibition of steroid synthesis resulting from blockade of leukotriene synthesis by SK&F 86002-A2 would explain why ovaries from hypophysectomized, SK&F 86002-Altreated rats were refractory to gonadotropins. This antisteroidogenic activity coupled with the phosphodiesterase inhibitory activity of SK&F 86002-A2 would also explain why a 30-day treatment of cycling females with the drug produced unruptured, luteinized follicles. Comparable ovarian changes have been produced pharmacologically (Killick and Elstein, 1987) and also occur spontaneously in women with luteinized unruptured follicle syndrome (LUF) (Marik and Hukla, 1978; Koninckx et al., 1980) in which gonadal steroid production may be disturbed. Since CAMP causes luteinization of granulosa cells, and blockade of leukotriene synthesis inhibits steroidogenesis, the combined effects of SK&F 86002-A* could retard follicular proliferation and block ovulation while promoting spontaneous luteinization.
SK&F
66002-4
x 10’
M
FIG. 8. Dixon Plot of phosphodiesterase inhibition SK&F 86002-A*.
by
SK&F 86002-A2 SITE OF ACTION
In conclusion, the results of this study suggest that the effects of SK&F 86002-AZ on reproductive function in the female are complex, resulting from multiple interactions involving the inhibition of enzymes for prostaglandin and leukotriene synthesis, as well as that for catabolism of CAMP. Apparently the antiovulatory effects of prostaglandin inhibition seen with drugs like indomethacin are masked in SK&F 86002-A* by elevation of CAMP following phosphodiesterase inhibition. However, the effects on leukotriene inhibition may be more insidious, retarding steroid synthesis and thus, gradually desynchronizing the events that are required for normal cyclic ovarian function. These changes ultimately lead to ovulatory blockade and follicular luteinization. Since similar events occur spontaneously in some women, SK&F 86002-A2 may be useful for producing an experimental model for study and treatment of LUF.
ACKNOWLEDGMENTS The authors thank Dr. William Johnson for purifying the human monocytes, Ms. Lenora Cieslinski for technical assistance, and Ms. Janet Koster for her expert secretarial help in preparation of this manuscript.
REFERENCES ARMSTRONG, D. T., AND GRINWICH, D. L. (1972). Blockade of spontaneous and LH-induced ovulation in rats by indomethacin, an inhibitor of prostaglandin biosynthesis. Prostaglandin 1,2 l-28. BEHRMAN, H. R., ORCZYK, G. P., AND GREEP, R. 0. ( 1972). Effect of synthetic gonadotrophin-releasing hormone (Gn-RF) on ovulation blockade by aspirin and indomethacin. Prostaglandins 1,245-256. COLOrrA, F., PER!, G., VILLA, A., AND MANTOVANI, A. (1984). Rapid killing of actinomycin D-treated tumor cells by human mononuclear cells. J. Immunol. 132, 936-944. CONTE, D., FALASCHI, P., PROIETX, A., D’ARso, R., CL TARELLA, F., NORDIO, M., ROMANELLI, F., MAGGI, R., MOTTA, M., AND ISIDORI, A. (1986). Role of arachidonate metabolism on the in vitro release of luteinizing hormone and prolactin from the anterior pitu-
285
itary gland: Possible involvement of lipoxygenase pathway. Neuroendocrinology 43,428-434. DAVIS, C. W., AND DALY, J. W. (1979). A simple direct assay of 3’,5’-cyclic nucleotide phosphodiesterase activity based on the use of polyacrylamide-boronate affinity gel chromatography. J. Cyclic Nucieotide Res. 5,65-74. DIMARTINO, M. J., GRISWOLD, D. E., BERKOWITZ, B. A., POSTE, G., AND HANNA, N. (1987). Pharmacologic characterization of the antiinflammatory properties of a new dual inhibitor of lipoxygenase and cyclooxygenase. Agents Actions 20, 1I3- 123. DIX, C. J., HABBERFIELD, A. D., SULLIVAN, H. F., AND COOHE, B. A. ( 1984). Inhibition of steroid production of Leydig cells by non-steroidal anti-inflammatory and related compounds: Evidence for the involvement of lipoxygenase products in steroidogenesis. Biochem. J. 219,529-537.
DIXON, M. ( 1952). The determination of enzyme inhibitor constants. Biochem. J. 55,170-l 7 1. EVERETT, J. W., AND SAWYER, C. H. (1950). A 24-hour periodicity in the “LH-release apparatus” of female rats disclosed by barbiturate sedation. Endocrinology 47,198-218. HARMAN, S. M., LOUVET, J. P., AND Ross, G. T. ( 1975). Interaction of estrogen and gonadotropins on follicular atresia. Endocrinology 96, 1145- 1155. HSUEK, A. J. W., JONES,P. B. C., ADASHI, E. Y., WANG, C., ZHUANG, L. Z., AND WELSH, T. H. (1983). Intraovarian mechanisms in the hormonal control of granulosa cell differentiation in rats. J. Reprod. Fertil. 69,325-337. KILLICK, S., AND ELSTEIN, M. (1987). Pharmacologic production of luteinized unruptured follicles by prostaglandin synthetase inhibitors. Fertil. Steril. 47,773777. KONINCKX, P. R., DEMOOR, P., AND BROSEN, I. A. (1980). Diagnosis of the luteinized unruptured follicle syndrome by steroid hormone assays on peritoneal &id. Brit. J. Obstet. Gynaeco/.87,929-936. LEE, J. B., AND KATAYAMA (1981). Prostaglandins, thromboxanes and leukotrienes. In Williams Textbook of Endocrinology, (J. D. Wilson and D. W. Foster, Eds.), 7th ed., pp. 1345-1378. Saunders, Philadelphia. LEGAN, S. J., COON, G. A., AND KARSCH, F. J. (1975). Role of estrogen as initiator of daily LH surges in the ovariectomized rat. Endocrinology 96.5 56-56 1. MARIK, J., AND HUKLA, J. F. (1978). Luteinized unrup tured follicle syndrome: A subtle cause of infertility. Fertil. Steril. 29,270-279. NATHANSON, J. A. (1977). Cyclic nucleotides and nervous system function. Physiol. Rev. 57, 157-256. NILSSON, L., HILLENSJO, T., AND EKHOLM, C. (1977). Preovulatory changes in rat follicular cyclic AMP and sensitivity to gonadotropins. Acta Endocrinol. (Copenhueen) u , 86.384-396. ,
286
WALKER
OJEDA, S. R., AND CAMPBELL, W. B. ( 1982). An increase in hypothalamic capacity to synthesize prostaglandin E2 precedes the first preovulatory surge of gonadotropins. Endocrinologylll, 1031-1037. OJEDA, S. R., HARMS, P. G., AND MCCANN, S. M. (1975). Effects of inhibitors of prostaglandin synthesis on gonadotropin release in the rat. Endocrinology 97, 843-854. ORCZYK, G. P., ANDBEHRMAN, H. R. (1972). Ovulation blockage by aspirin or indomethacin-In vivo evidence for a role of prostaglandin in gonadotropin secretion. Prostuglandins 1,3-20. REEVES, M. L., LEIGH, B. K., AND ENGLAND, P. J. (1987). The identification of a new cyclic nucleotide phosphodiesterase activity in human and guinea-pig cardiac ventricle. Biochem. J. 241,535-54 1. REICH, R., KOHEN, F., NAOR, Z., AND TSAFRIRI, A. ( 1983). Possible involvement of lipoxygenase products of arachidonic acid pathway in ovulation. Prostaglandins 26,101 l-1020. SYNDER, G. D., CAPDEVILA, J., CHACOS, N., MANNA, S.. AND FALCK, J. R. (1983). Action ofluteinizing hor-
ET AL. mone-releasing hormone: Involvement of novel arachidonic acid metabolites. Proc. N&l. Acad. Sci. USA 80,3504-3507. TSAFRIRI, A., KOCH, Y., AND LINDER, H. R. (1973). Ovulation rate and serum LH levels in rats treated with indomethacin or PGE2. Prostaglandins 3,46 l-467. WALKER, R. F., SCHWARTZ, L. W., AND MANSON, J. M. (1988). Ovarian effects ofan anti-intIammatoryimmunomodulatory drug in the rat. Toxicol. Appl. Pharmacol. 94,266-275. WALLACH, E. E., DELACRUZ, A., HUNT, J., WRIGHT, K. H., AND STEVENS, V. C. (1975). The effect of indomethacin on HMG-HCG induced ovulation in the rhesus monkey. Prostaglandins 9,645-658. ZARROW, M. X., CALDWELL, A., HAFEZ, E. S. F., AND PINCUS, G. (1958). Superovulation in the immature rat as a possible assayfor LH and HCG. Endocrinology 63,748-75. ZELEZNIK, A. J., MIDCLEY, A. R., AND REICHERT, L. E. (1974). Granulosa cell maturation in the rat: Increased binding of human chorionic gonadotropin following treatment with follicle-stimulating hormone in vivo. Endocrinology 95,8 18-829.
94,276-286(1988)
Ovarian Effects of SK&F 86002-A* RICHARD
in the Rat: Site of Action
F. WALKER,* LESTER W. SCHWARTZ,? THEODORE JOHN F. NEWTON,~
J. TORPHY,$
AND JEANNE M. MANSON*
Departments of *Reproductive and Developmental Toxicology, TExperimental Pathology, SPharmacology, and QDrug Metabolism, Smith Kline & French Laboratories. Research and Development, King of Prussia, Pennsylvania 19406-0939
Received October 6, 1987; accepted February 25, I988 Ovarian Effects of SK&F 86002-A2 in the Rat: Site of Action. WALKER, R. F., SCHWARTZ, L. W.,TORPHY,T.J.,NEWTON,J.F.,ANDMANSON,J.M. (1988). Toxicol. Appl. Pharmacol. 94,276-286. In a preliminary 30-day study, oral administration of SK&F 86002-A2, an inhibitor of prostaglandin and leukotriene synthesis, blocked ovulation and altered ovarian structure and hormone production in rats. The purpose of the present study was to determine if the locus of action of SK&F 86002-A* for these effects was the ovary or some other site in the female reproductive system, using a number of experimental approaches. A single SCor intraovarian injection of SK&F 86002-A* did not block spontaneous or gonadotropin-induced ovulation in proestrous rats, whereas indomethacin, a positive control, acutely disrupted the ovulatory process. Since neither route of administration blocked ovulation, integrated pituitary and ovarian events were not negatively affected by a single injection of SK&F 86002-A* at doses which caused ovarian dysfunction when administered repeatedly for 30 days. In contrast to a single dose, oral administration of SK&F 86002-A* to hypophysectomized rats for 2 weeks suppressed follicular growth and estradiol production in response to sc administration of pregnant mare serum gonadotropin. Although ovarian function was suppressed in hypophysectomized rats, LH surges induced by estradiol in ovariectomized rats were not at&ted by administration of SK&F 86002AZ for 2 weeks. Thus, hypothalamic/pituitary dysfunction did not contribute to the ovarian effects of SK&F 86002 that occurred after repeated dosing. In conclusion, these results indicate that disruption of ovarian cycles by SK&F 86002-A* is related to a direct effect on the ovary, and not to altered hypothalamic/pituitary function and LH release. Specifically, SK&F 86002AZ may suppress the ovarian response to gonadotrophin, retarding follicular growth and estrogen production. The ovarian effects are consistent with a pharmacological expression of the inhibitory action of SK&F 86002-A* on prostaglandin and leukotriene synthesis. o 1988 Academic Press, Inc.
SK&F 86002-AZ, [5-(4-pyridyl)-6(4-fluorophenyl) - 2,3 - dihydroimidazo - (2,1- 6) - thia zole] dihydrochloride, inhibits cyclooxygenase and 5-lipoxygenase pathways of the arachidonic acid metabolic cascade (DiMartino et al., 1987). The dual inhibitory action of SK&F 86002-A* was confirmed by the following observations: ( 1) impaired production of the 5-lipoxygenase products leukotriene B4 (diHETE) and 5-HETE by RBL-1 cells, (2) impaired production of cyclooxygenase products (PGE2) by inflammatory macro0041-008x/88
$3.00
Copyright 8 1988 by Academic Press, Inc. All rights of reproduction in any form reserved.
phages in vitro, and by (3) reduced production of LTB., in vitro by peritoneal exudate cells harvested from SK&F 86002-AZ-treated mice. The results of a previous study from our laboratory showed that oral administration of SK&F 86002-A* to rats for 30 days caused estrous cycle irregularity in a dose-related fashion (Walker et al., 1988). Disruption of regular estrous cycles correlated with structural and biochemical changes in the ovaries, which were often hypertrophied to the extent 216
277
SK&F 86002-A* SITE OF ACTION
that they could be differentiated from controls by visual inspection. Histological examination revealed increased numbers of luteinized follicles with retained ova and occasional cystic follicles. The histologic changes were accompanied by altered levels of plasma estradiol and progesterone, while the levels of plasma pituitary hormones (LH, FSH and prolactin) were comparable to controls. These preliminary findings suggested that SK&F 86002-A* disrupted cyclic ovarian function by a local, cumulative action that inhibited ovulation and altered steroid secretion. However, basal but not surge levels of pituitary hormones were estimated, and the data were insufficient to conclude that SK&F 86002-A* did not alter preovulatory secretion of gonadotropins. The ovulation-blocking action of prostaglandin inhibitors such as indomethacin can occur at hypothalamic/pituitary (Orczyk and Behrman, 1972; Ojeda et al., 1975; Ojeda and Campbell, 1982) or ovarian (Tsafiiri et al., 1973; Armstrong and Grinwich, 1972; Behrman et al., 1972; Wallach et al., 1975) sites depending upon the dose administered. Similarly, inhibitors of leukotriene synthesis have the potential to alter ovarian function by blocking follicular rupture (Reich et al., 1983), steroidogenesis (Dix et al., 1984), or the release of luteinizing hormone (Snyder et al., 1983; Conte et al., 1986). Since SK&F 86002-A2 is a dual inhibitor of cyclooxygenase and lipoxygenase (DiMartino et al., 1987) it may cause ovarian dysfunction by affecting one or more sites within the female reproductive/neuroendocrine axis. For example, ovulation may be prevented indirectly as the result of hypothalamic/pituitary dysfunction in which the preovulatory LH surge is lost, or directly as the result of ovarian dysfunction in which follicular rupture is blocked. Alternatively, events which precede and promote ovulation such as folliculogenesis and steroidogenesis may be altered causing ovarian dysfunction. Thus, the purpose of this study was to determine the site at which SK&F 86002-A2 blocks ovula-
tion, and to better understand the mechanism by which it causes ovarian dysfunction in the rat. METHODS
AND
MATERIALS
SK&F 86002-A?, which is the dihydrochloride salt of SK&F 86002 [6-(4-fluorophenyl)2,3-dihydro-5-(4-pyridinyl)imidazo(2,1 -b)-thiazole], was synthesized locally (SK&F Laboratories, King of Prussia, PA) (DiMartino et al., 1987). It is readily soluble in water (>80 mg/ml) and was prepared for use in these studies as a sterile aqueous solution [ 100 mg (base)/ml, pH 1.41. Animals andfacilities. Sexually mature (75-85 days of age) female rats (Sprague-Dawley/CD-VAF) that were intact, ovariectomized, or hypophysectomized were purchased from Charles River Breeding Laboratories (Cambridge, MA). Prepubertal rats (22-28 days old) that were used to test the effect of SK&F 86002-A* on gonadotropin-induced superovulation were from the same source. All animals were housed in a barrier facility (SK&F R&D Labs, King of Prussia, PA) in clear polycarbonate boxes containing Alpha-dri bedding (Sheperd Specialty Papers, Inc., Kalamazoo, MI) and maintained under standard conditions of light ( 12 hr; 0600- 1800 hr), temperature (72 f 4°F) humidity (50 f 10% relative), and nutrition (Certified Rodent Chow, Ralston Purina Co., St. Louis, MO, and tap water, ad libitum). Quantitation of SK&F 86002-A2 and metabolites in plasma. In this study, SK&F 86002-A* was administered by gavage as well as subcutaneous injection. Plasma concentrations of SK&F 86002-A? and its metabolites (SK&F 86096 and SK&F 104343) were measured for up to 6 hr following sc or po administration. Blood samples (100 ~1) were collected from a tail vein into heparinized Eppendorf microtubes at various times after dosing. Plasma was separated by centrifugation at 4’C and a known volume was transferred to a polypropylene test tube which contained internal standard, SK&F 85838 (6-(4’-lluorophenyl)-5-(4’-methoxyphenyl)-2,3-dihydroimidazo-(2,1 -b)-thiazole- 1-oxide, hydrate: 100 ~1 of a 2 &ml solution), and the mixture was stored at -20°C until analysis. Plasma concentrations of SK&F 86002 and its sulfoxide and sulfone metabolites, SK&F 86096 and 104343, were quantitated by HPLC following solid phase extraction using Baker 10 SPE (J.T. Baker Chemical Co., Phillipsburg, NJ) I ml Cls columns preconditioned with methanol and water. Samples containing internal standard were added to the columns which were then washed with ammonium acetate (100 mM). SK&F 86002 and its metabolites were eluted from the column with methanol (2 X 250 ~1). Samples were dried under nitrogen and resuspended in HPLC mobile phase (100 pl), and aliquots (50 ~1) of the reconstituted samples were quantitated on
278
WALKER
an HPLC system consisting of a Model 590 solvent delivery system, a Model 7 1OB autoinjector, a Model 44 1 uv detector set at 254 nm, and a Model RCM 100 radial compression module which contained a 5-pm NOVAPAK C,s (8 X IO-mm) cartridge (Waters Assoc., Inc., Milford, MA). SK&F 86002, metabolites, and the internal standard were separated isocratically using a solvent system consisting of 29.5% acetonitrile in 100 mM ammonium acetate at a flow rate of 2.0 ml/min. Plasma concentration of SK&F 86002 and metabolites were determined from peak area ratios (compared to internal standard) that were compared to extracted authentic standard curves. Ovarian function: Single dose studies. The first objective of this study was to determine if SK&F 86002-A* directly affects the ovary by altering its response to spontaneous or gonadotropin-induced ovulation. This was accomplished in one set of experiments by direct injection of SK&F 86002-A2 into the ovaries of proestrous rats followed by measurement of spontaneous ovulation in response to the endogenous LH surge. Sexually mature female rats with unambiguous proestrus vaginal smears were anesthetized at 1430 hr with CO2 and ether. Each ovary was surgically exposed and injected with 2 pl of vehicle or vehicle containing 200 pg SK&F 86002-Al. Unanesthetized, uninjected rats were also included in the test to control for the possible effects of inhalation anesthesia on the preovulatory LH surge. At 0700 hr on the following morning, the rats were killed and tubal ova were counted and compared among groups. Next, SK&F 86002-A2 was tested for potential effects on gonadotropin-induced superovulation. Prepubertal (24-27 days of age) female rats were given SCinjections of pregnant mare serum (PMS, 30 IU, 0.2 ml) at 1200 hr. Forty-eight hours after receiving PMS to initiate follicular growth, the rats were anesthetized with ether and their ovaries were exteriorized through bilateral incisions in the dorsolumbar flank. The ovaries were then injected with SK&F 86002-A2 (2 ~1 containing 10 or 200 rg) or an equivalent volume of vehicle. Shortly thereafter, the rats received an SCinjection of human chorionic gonadotropin (HCG; 20 IU) to induce ovulation (Zarrow et al., 1958). Twenty hours after receiving HCG the rats were killed by COZ asphyxiation. Their fallopian tubes were removed and examined microscopically for the presence of ova which were then counted and the numbers compared among treated and control groups. Subcutaneous injections of SK&F 86002-A* were also administered to adult rats to test for a direct ovarian effect on ovulation. Sexually mature female rats with unambiguous proestrus vaginal smears were injected with pentobarbital(30 mg/kg, ip) at 1430 hr to block the neurogenic stimulus for the preovulatory LH surge (Everett and Sawyer, 1950). A single injection of ovine LH (30 rg) was administered 15 min later. Barbiturate blockade of the LH surge followed by LH administration removes central elements of the reproductive axis from influenc-
ET AL. ing ovarian function. Thus, any effect of systemic drug administration on ovulation in this model must occur at the level of the ovary. SK&F 86002-A2 (60 mg/kg) was administered by a single sc injection. Indomethacin (10 mg/kg) was given to other barbiturate-blocked, gonadotropin-treated rats as a positive control. The rats were killed at 0700 hr on the following morning, and tubal ova were counted and the numbers compared among drugtreated and control groups. Ovarian function: Multiple dose studies. Hypophysectomized rats were allowed a postsurgery recovery period of 30 days during which time ovarian atrophy occurred in the absence of endogenous gonadotropins. Then, SK&F 86002-A2 (60 mg/kg/day) or vehicle was administered by gavage for 18 consecutive days. In addition, each rat received daily injections of PMS ( 100 IU) for the last 4 consecutive days of dosing. Vaginal smears were monitored during PMS administration for signs of cellular comification as an indication of gonadal recrudescence. At the end of gonadotropin treatment the rats were killed, their ovaries were fixed for histological examination, and their blood was collected for estimation of estradiol levels by radioimmunoassay. Pituitary function. Pituitary function was assessedby examining estrogen-stimulated release of LH in ovariectomized rats that were repeatedly dosed with SK&F 86002-A*. In intact rats, the preovulatory LH surge occurs in response to increasing levels of serum estradiol, produced by growing ovarian follicles (Legan et al., 1975). After ovulation, progesterone produced by the corpora lutea prevents an LH surge from recurring. Cohorts of growing follicles that produce estradiol in subsequent cycles trigger the intermittent LH surges that are associated with each ovulation. This ovarian influence on preovulatory LH release can be eliminated by ovariectomy (OVX). When serum estradiol levels are pharmacologically elevated in OVX rats, an LH surge will occur each day because the inhibitory effects of progesterone are eliminated. OVX rats are useful when investigating the effect of drugs on central components of the female reproductive axis since LH surges are artificially induced by estrogen in the absence of the ovaries. Thus, the estradiol-supplemented OVX rat was used to test the effects of SK&F 86002-A* on the pituitary mechanism controlling the LH surge. Sexually mature, ovariectomized rats were gavaged with SK&F 86002-A2 (10 or 60 mg/kg) or vehicle for 14 consecutive days. On Day 9 of dosing, the rats received SCimplants of Silastic tubing (5 mm) containing estradiol 170, which initiated daily LH surges (Legan et al., 1975). Each rat was bled from a tail vein at 1000, 1300, 1500, and 1700 hr, five days after the estradiol capsules were implanted to estimate serum levels of LH by radioimmunoassay. These times were selected because the LH surge begins at approximately 1400 hr and peaks at approximately 1700 hr under the lighting conditions of this study. Serum LH values were compared among groups
279
SK&F 86002-A* SITE OF ACTION to determine if repeated daily dosing with SK&F 86002Az disrupts the hypothalamic/pituitary mechanism for preovulatory secretion of LH. Radioimmunoassay. Serum estradiol concentrations were estimated using commercially available RIA kits purchased from Diagnostic Products Corp. (Los Angeles, CA 90045). The assay was sensitive to 10.0 pg/ml estradiol with interassay and intraassay coefficients of variation of 10.3 and 7.2%, respectively. Serum LH levels were estimated using reagents provided by the National Pituitary Agency (NIH). Assay sensitivities as well as interassay and intraassay coefficients of variation for the LH RIA were 460 pg, 10.1, and 8.2%, respectively. Serum levels of LH were expressed in terms ofthe reference preparation, LHRP-2. Phosphodiesterase assay. Cyclic AMP phosphodies terase (PDE) was prepared from monocytes (75 X IO6 cells) that were recovered from two units of human peripheral blood (Colotta et al., 1984). The monocytes were suspended in 2.5-3.0 ml of 50 mM Tris-HCl buffer (pH 7.5) containing 0.1% Triton X- 100 and a protease inhibitor mixture of 5 mM EDTA, 10 &ml soybean trypsin inhibitor, 17 pg/ml benzamidine, 100 fig/ml bacitracin, and 10 pM phenylmethyl sulfonyl fluoride. The cells were disrupted by sonification for 30 set using a Branson Sonifier Cell Disruptor 200. Sonilication of human monocyte suspensions in the presence of Triton X- 100 releases essentially all particulate bound PDE activity. The cell-free suspension was then centrifuged at 20,OOOgfor 30 min at 4’C and the supernatant collected and applied to a DEAE-Sepharose column to separate PDE isozymes according to the method of Reeves et al. (1987). Enzyme activity was measured in column fractions containing the CAMP-selective PDE by a modification of the method of Davis and Daly (1979). All assayswere conducted in the linear range of the reaction where less than 20% of the initial substrate was hydrolyzed. Final concentrations in the reaction mixture were 50 mM Tris-HCl buffer (pH 7.5), 5 mM MgCl*, 50 WM S’AMP (carrier), 1 pM 3’5’-[3H]cAMP (approximately 2000 dpm/pmol) and 50-~1 fraction aliquots. The reaction was run at 3o”C, initiated by the addition of substrate, and terminated after 30 min by placing the reaction vesselsin a 1WC heating block for 1 min. Various concentrations (l-300 pM) of SK&F 86002-A2 were tested for PDE inhibition using [3H]cAMP concentrations of 0.1,0.3, and 1.OpM. Aliquots ofthe enzyme were preincubated with SK&F 86002-A2 at room temperature for 10 min prior to addition of [3H]cAMP. The 3’5’-cyclic nucleotide was separated from the 5’ nucleotide product by first adding 0.5 ml ofO.1 M Hepes buffer (pH 8.5) containing 0.1 M NaCl to each sample. The samples were then applied to individual polyacrylamide-boronate gel column (0.5 g of Biorad Affi-gel 601 on a 0.7 X IO-cm Biorad econo-column) which had been equilibrated with the 0.1 M Hepes/O. 1 NaCl buffer (pH 8.5) and the unreacted cyclic nucleotides were eluted from the column
unmiected
vehicle
?,K.w
86002-A*
TREATMENT FIG. 1. Effect of intraovarian injections of SK&F 86002-A* on spontaneous ovulation. *p < 0.05 compared to uninjected controls. N = 9 (uninjected), 6 (vehicle), 5 (SK&F 86002-A&
with 10 ml of Hepes buffer. The 5’ nucleotide product was then eluted into a scintillation vial with 10 ml 0.25 M acetic acid. Ten milliliters of Beckman Ready SolvMP scintillation cocktail was added to the eluate and shaken to form a gel, and the radioactivity was measured via liquid scintillation spectrometry. The CAMP PDE inhibition constant (&) for SK&F 86002-A* was determined according to the method of Dixon (1952). Statistical analyses of hormone levels, shed ova, and ovarian follicle counts were conducted using ANOVA and Duncan’s multiple range test.
RESULTS Data presented in Fig. 1 compare the numbers of ova shed from ovaries of sexually mature rats that received intraovarian injections of either SK&F 86002-A2 or vehicle on proestrus, to the numbers of ova shed from ovaries of unanesthetized, uninjected control rats. The mean number of ova shed from SK&F 86002-AZ-treated rats was lower than that from unanesthetized, uninjected controls. However, the differences were not statistically significant. In contrast, significantly fewer ova (p c 0.01) were shed by vehicleinjected controls compared with SK&F 86002-AZ-treated rats or unanesthetized, un-
280
WALKER
subcapsular
intraovarian
INJECTION SITE FIG. 2. Superovulation in immature gonadotropintreated rats receiving SK&F 86002-A2 by ovarian injection N = 9 for each treatment and route.
injected controls. Blockade of ovulation in the vehicle-injected control group was most likely due to anesthesia which depressed the preovulatory LH surge. Intraovarian injection of vehicle or SK&F 86002-A* in prepubertal female rats did not suppress ovulation when exogenous gonadotropin was administered, as indicated from data presented in Fig. 2. Thus, when the endogenous LH surge was blocked by anesthesia, SK&F 86002-A2 may have actually stimulated ovulation in the presence of the anesthetic block. The data in Fig. 3 compare the effects of subcutaneously administered SK&F 86002A2 on gonadotrophin-induced ovulation in the “pentobarbital-blocked” proestrus rat. As seen in Fig. 4, oral doses of SK&F 86002-A2 above 30 mg/kg did not increase maximal plasma concentrations of the drug, so subcutaneous doses were administered. The subcutaneous route produced higher plasma concentrations than equivalent doses given orally. Pentobarbital was administered SCto
ET AL.
block the neurogenic stimulus for the LH surge and consequently, exogenous LH was injected to eliminate central factors governing ovulation from consideration in this experiment. Under these conditions, ovulation was blocked by indomethacin (p -C0.0 1) but not by SK&F 86002-A2 or vehicle (Fig. 3). Although SK&F 86002-A2 did not block gonadotropin-induced ovulation, it was of interest to determine whether a systemic injection of the drug would suppress ovulation in response to the preovulatory LH surge. The experiment was repeated with proestrus rats that were not pentobarbital-blocked. As seen in Fig. 5, SK&F 86002-A* did not block spontaneous ovulation in these rats and in fact, the mean number of tubal ova was slightly higher in drug-treated than in control rats. Thus, the preceding experiments showed that single injections of SK&F 86002-A* did not block ovulation at the ovarian level, and may in fact have stimulated it slightly. The effect of 14 consecutive, daily doses of SK&F 86002-A* on PMS-induced follicular maturation and estradiol synthesis in hy-
FIG. 3. Comparison of the effects of SK&F and indomethacin on gonadotropin-induced in “pentobarbital-blocked” rats. *p < 0.05 with values from the same group at 1000 hr. each treatment.
86002-A* ovulation compared N = 6 for
SK&F 86002-A2 SITE OF ACTION 20
1
SK&F 86002
11
Sulfoxide
a~
281
I]
Sulfone
15
Time
After
Administration
(hrs)
FIG. 4. Plasma concentrations of SK&F 86002 and its metabolites following oral or istration of SK&F 86002 [30 mg (base)/kg].
pophysectomized rats is shown in Table 1. Vaginal cornification did not occur, and the number of developing follicles and serum estradiol levels in hypophysectomized rats treated with PMS and SK&F 86002-A* were significantly less (p < 0.05 for both parameters) than in PMS-treated control rats. Histological evaluation of ovaries from these
subcutaneous
admin-
hypophysectomized rats corroborated the vaginal cytology and serum estradiol concentrations in that mature or preovulatory follicles were only infrequently observed in SK&F 86002-AZ-treated rats (Fig. 6). The effect of 14 days of dosing with SK&F 86002-A* on estrogen-induced LH release in ovariectomized rats is presented in Fig. 7. Serum levels of LH rose comparably from 1000 to 1700 hr after 5 days of estradiol exposure in ovariectomized rats dosed with 0, 10, or 60 mg/kg SK&F 86002-A* for 14 consecutive days, indicating a lack of drug effect on central sites controlling the LH surge. The data in Fig. 8 show that SK&F 86002A* is a competitive inhibitor of phosphodiesterase with a Ki of approximately 10 j.&M. DISCUSSION
“EHlCLE
SKW
aeow-A2
TREATMENT FIG. 5. Effect of systemic injection of SK&F 86002-A* on spontaneous ovulation. N = 5 for each treatment.
The major finding of this study was that repeated doses of SK&F 86002-A2 retarded follicular maturation and inhibited estradiol production in the rat ovary. These ovarian effects could not be attributed to defects in gonadotropin secretion since estradiol-induced LH surges occurred in rats that received multiple doses of SK&F 86002-AZ under conditions which disrupted estrous cycles and blocked ovulation in an earlier study (Walker et al., 1988).
282
WALKER TABLE
ET
AL.
1
EFFECT OF REPEATED DOSING WITH SK&F 86002-A2 ON OVARIAN IN GONADOTROPIN-TREATED HYPOPHYSECTOMIZED
*
p
Treatment
N
Vehicle SK&F 86002-A,
6 6
< 0.05 compared
Vaginal comification + -
Follicles (2 + SE) 12.3 + 2.7 6.1 f 1.5
RECRUDESCENCE RATS Serum (p&ml;
estradiol 1 f SE)
40.5 + 4.1 28.4 + 3.1*
with control.
Unlike other inhibitors of prostaglandin or leukotriene synthesis that block ovulation after acute administration (Orczyk and Behrman, 1972; Armstrong and Grinwich, 1972; Reich et al., 1983), single doses of SK&F 86002-A2 had no effect on spontaneous or gonadotropin-induced ovulation. This unexpected observation may have resulted from the activity of SK&F 86002-A* to inhibit phosphodiesterase in addition to its inhibitory effects on arachidonic acid metabolism. The biological actions of prostaglandins are directly linked to changes in CAMP. When prostaglandins increase intracellular CAMP, they enhance target cell functions (Lee and Katayama, 1981). Following the LH surge, CAMP rises in preovulatory follicles, initiating the process of rupture, ovulation, and luteinization (Nilsson et al., 1977). Since inhibition of phosphodiesterase causes intracellular CAMP to accumulate (Nathanson, 1977), this additional effect of SK&F 86002-AZ may have circumvented its potential antiovulatory action resulting from blockade of prostaglandin synthesis. This hypothesis is supported by the fact that SK&F 86002-A2 stimulated ovulation in rats whose LH surges were partially blocked by COz/ether anesthesia on the day of proestrus. When proestrus rats were anesthetized for surgery and given intraovarian injections of SK&F 86002-A2 or vehicle, significantly more eggs were shed by drug-treated than control rats. The low number of eggs shed by control rats is explained by the fact that anesthesia administered on the day of proestrus can block the neurogenic
stimulus for the preovulatory LH surge (Everett and Sawyer, 1950). COz/ether anesthesia given during proestrus in this study reduced ovulation in control rats, while ovulation was not as profoundly affected in rats injected with SK&F 86002-A*. If SK&F 86002-A2 inhibited ovarian phosphodiesterase in addition to cyclooxygenase, then it mimicked in part the ovulatory effects of LH by raising ovarian CAMP. Since a full complement of ova was not shed in these rats, the ovulatory blockade expected from inhibition of prostaglandin synthesis by SK&F 86002A2 may have been only partially reversed by phosphodiesterase inhibition. Stimulatory effects of SK&F 86002-A2 on ovulation were also seen in gonadotropin-treated rats in which the mean number of eggs shed was higher than in controls. Thus, as a consequence of its phosphodiesterase inhibiting action, the potential antiovulatory action of SK&F 86002-A2 resulting from inhibition of prostaglandin synthesis may have been masked, and late stage ovulatory events such as follicular rupture were not blocked by a single dose. In contrast to late stage events, earlier events in the ovarian cycle were disrupted following repeated dosing with SK&F 86002AZ. In cycling, adult female rats, SK&F 86002-A2 caused estrous cycle irregularity and ovarian changes that included luteinization of unruptured follicles (Walker et al., 1988). In an attempt to determine the cause of these effects, the estrogen-stimulated LH surge was examined in rats treated for 14 days
FIG. 6. Photomicrographs of ovaries obtained from hypophysectomized rats administered gonadotropins. (A) Ovary containing two preovulatory follicles from a control rat administered gonadotropins and vehicle. (B) Ovary obtained from a gonadotropin-treated rat administered SK&F 86002-A2 for 14 consecutive days. Maturing or preovulatory follicles were not present. Magnification 164X. 283
284
WALKER
6 ‘55 E $4 3 5” 5 m2
1
0 lOOOh
I 1300h
J
1500h
1700h
TIME
FIG. 7. Estrogen-induced LH surges in ovariectomized rats dosed repeatedly with SK&F 86002-A2 *p < 0.05 compared with values at 1000 hr. N = 6 for each treatment.
with SK&F 86002-AZ. Repeated administration of SK&F 86002-A2 had no effect on LH surges produced by exogenous estradiol in OVX rats, suggesting that the hypothalamic/ pituitary mechanism for preovulatory release of LH remained intact, despite repeated exposure to SK&F 86002-AI. A direct ovarian effect on steroidogenesis was discovered in hypophysectomized rats. The ovaries of these animals were refractory to exogenous gonadotropin following repeated doses of SK&F 86002-A2 since follicular growth and estrogen production were significantly retarded. Knowing that estrogen priming increases the sensitivity of ovaries in hypophysectomized rats to gonadotropins (Zeleznik et al., 1974), that follicular growth is inhibited when the biologic activity of estrogen is neutralized with antibody (Harman et al., 1975), that estradiol exerts a direct mitogenic effect on granulosa cells to increase their number and thereby enhance the effects of FSH on follicular proliferation (Hsuek et
ET AL.
al., 1983), and that leukotriene inhibitors block steroid synthesis in response to gonadotropin (Dix et al., 1984), the data suggest that the ovarian effects of SK&F 86002-A2 resulted from inhibition of estradiol production. Since the synergy between estradiol and FSH represents an intraovarian positive feedback mechanism to promote follicular growth, inhibition of steroid synthesis resulting from blockade of leukotriene synthesis by SK&F 86002-A2 would explain why ovaries from hypophysectomized, SK&F 86002-Altreated rats were refractory to gonadotropins. This antisteroidogenic activity coupled with the phosphodiesterase inhibitory activity of SK&F 86002-A2 would also explain why a 30-day treatment of cycling females with the drug produced unruptured, luteinized follicles. Comparable ovarian changes have been produced pharmacologically (Killick and Elstein, 1987) and also occur spontaneously in women with luteinized unruptured follicle syndrome (LUF) (Marik and Hukla, 1978; Koninckx et al., 1980) in which gonadal steroid production may be disturbed. Since CAMP causes luteinization of granulosa cells, and blockade of leukotriene synthesis inhibits steroidogenesis, the combined effects of SK&F 86002-A* could retard follicular proliferation and block ovulation while promoting spontaneous luteinization.
SK&F
66002-4
x 10’
M
FIG. 8. Dixon Plot of phosphodiesterase inhibition SK&F 86002-A*.
by
SK&F 86002-A2 SITE OF ACTION
In conclusion, the results of this study suggest that the effects of SK&F 86002-AZ on reproductive function in the female are complex, resulting from multiple interactions involving the inhibition of enzymes for prostaglandin and leukotriene synthesis, as well as that for catabolism of CAMP. Apparently the antiovulatory effects of prostaglandin inhibition seen with drugs like indomethacin are masked in SK&F 86002-A* by elevation of CAMP following phosphodiesterase inhibition. However, the effects on leukotriene inhibition may be more insidious, retarding steroid synthesis and thus, gradually desynchronizing the events that are required for normal cyclic ovarian function. These changes ultimately lead to ovulatory blockade and follicular luteinization. Since similar events occur spontaneously in some women, SK&F 86002-A2 may be useful for producing an experimental model for study and treatment of LUF.
ACKNOWLEDGMENTS The authors thank Dr. William Johnson for purifying the human monocytes, Ms. Lenora Cieslinski for technical assistance, and Ms. Janet Koster for her expert secretarial help in preparation of this manuscript.
REFERENCES ARMSTRONG, D. T., AND GRINWICH, D. L. (1972). Blockade of spontaneous and LH-induced ovulation in rats by indomethacin, an inhibitor of prostaglandin biosynthesis. Prostaglandin 1,2 l-28. BEHRMAN, H. R., ORCZYK, G. P., AND GREEP, R. 0. ( 1972). Effect of synthetic gonadotrophin-releasing hormone (Gn-RF) on ovulation blockade by aspirin and indomethacin. Prostaglandins 1,245-256. COLOrrA, F., PER!, G., VILLA, A., AND MANTOVANI, A. (1984). Rapid killing of actinomycin D-treated tumor cells by human mononuclear cells. J. Immunol. 132, 936-944. CONTE, D., FALASCHI, P., PROIETX, A., D’ARso, R., CL TARELLA, F., NORDIO, M., ROMANELLI, F., MAGGI, R., MOTTA, M., AND ISIDORI, A. (1986). Role of arachidonate metabolism on the in vitro release of luteinizing hormone and prolactin from the anterior pitu-
285
itary gland: Possible involvement of lipoxygenase pathway. Neuroendocrinology 43,428-434. DAVIS, C. W., AND DALY, J. W. (1979). A simple direct assay of 3’,5’-cyclic nucleotide phosphodiesterase activity based on the use of polyacrylamide-boronate affinity gel chromatography. J. Cyclic Nucieotide Res. 5,65-74. DIMARTINO, M. J., GRISWOLD, D. E., BERKOWITZ, B. A., POSTE, G., AND HANNA, N. (1987). Pharmacologic characterization of the antiinflammatory properties of a new dual inhibitor of lipoxygenase and cyclooxygenase. Agents Actions 20, 1I3- 123. DIX, C. J., HABBERFIELD, A. D., SULLIVAN, H. F., AND COOHE, B. A. ( 1984). Inhibition of steroid production of Leydig cells by non-steroidal anti-inflammatory and related compounds: Evidence for the involvement of lipoxygenase products in steroidogenesis. Biochem. J. 219,529-537.
DIXON, M. ( 1952). The determination of enzyme inhibitor constants. Biochem. J. 55,170-l 7 1. EVERETT, J. W., AND SAWYER, C. H. (1950). A 24-hour periodicity in the “LH-release apparatus” of female rats disclosed by barbiturate sedation. Endocrinology 47,198-218. HARMAN, S. M., LOUVET, J. P., AND Ross, G. T. ( 1975). Interaction of estrogen and gonadotropins on follicular atresia. Endocrinology 96, 1145- 1155. HSUEK, A. J. W., JONES,P. B. C., ADASHI, E. Y., WANG, C., ZHUANG, L. Z., AND WELSH, T. H. (1983). Intraovarian mechanisms in the hormonal control of granulosa cell differentiation in rats. J. Reprod. Fertil. 69,325-337. KILLICK, S., AND ELSTEIN, M. (1987). Pharmacologic production of luteinized unruptured follicles by prostaglandin synthetase inhibitors. Fertil. Steril. 47,773777. KONINCKX, P. R., DEMOOR, P., AND BROSEN, I. A. (1980). Diagnosis of the luteinized unruptured follicle syndrome by steroid hormone assays on peritoneal &id. Brit. J. Obstet. Gynaeco/.87,929-936. LEE, J. B., AND KATAYAMA (1981). Prostaglandins, thromboxanes and leukotrienes. In Williams Textbook of Endocrinology, (J. D. Wilson and D. W. Foster, Eds.), 7th ed., pp. 1345-1378. Saunders, Philadelphia. LEGAN, S. J., COON, G. A., AND KARSCH, F. J. (1975). Role of estrogen as initiator of daily LH surges in the ovariectomized rat. Endocrinology 96.5 56-56 1. MARIK, J., AND HUKLA, J. F. (1978). Luteinized unrup tured follicle syndrome: A subtle cause of infertility. Fertil. Steril. 29,270-279. NATHANSON, J. A. (1977). Cyclic nucleotides and nervous system function. Physiol. Rev. 57, 157-256. NILSSON, L., HILLENSJO, T., AND EKHOLM, C. (1977). Preovulatory changes in rat follicular cyclic AMP and sensitivity to gonadotropins. Acta Endocrinol. (Copenhueen) u , 86.384-396. ,
286
WALKER
OJEDA, S. R., AND CAMPBELL, W. B. ( 1982). An increase in hypothalamic capacity to synthesize prostaglandin E2 precedes the first preovulatory surge of gonadotropins. Endocrinologylll, 1031-1037. OJEDA, S. R., HARMS, P. G., AND MCCANN, S. M. (1975). Effects of inhibitors of prostaglandin synthesis on gonadotropin release in the rat. Endocrinology 97, 843-854. ORCZYK, G. P., ANDBEHRMAN, H. R. (1972). Ovulation blockage by aspirin or indomethacin-In vivo evidence for a role of prostaglandin in gonadotropin secretion. Prostuglandins 1,3-20. REEVES, M. L., LEIGH, B. K., AND ENGLAND, P. J. (1987). The identification of a new cyclic nucleotide phosphodiesterase activity in human and guinea-pig cardiac ventricle. Biochem. J. 241,535-54 1. REICH, R., KOHEN, F., NAOR, Z., AND TSAFRIRI, A. ( 1983). Possible involvement of lipoxygenase products of arachidonic acid pathway in ovulation. Prostaglandins 26,101 l-1020. SYNDER, G. D., CAPDEVILA, J., CHACOS, N., MANNA, S.. AND FALCK, J. R. (1983). Action ofluteinizing hor-
ET AL. mone-releasing hormone: Involvement of novel arachidonic acid metabolites. Proc. N&l. Acad. Sci. USA 80,3504-3507. TSAFRIRI, A., KOCH, Y., AND LINDER, H. R. (1973). Ovulation rate and serum LH levels in rats treated with indomethacin or PGE2. Prostaglandins 3,46 l-467. WALKER, R. F., SCHWARTZ, L. W., AND MANSON, J. M. (1988). Ovarian effects ofan anti-intIammatoryimmunomodulatory drug in the rat. Toxicol. Appl. Pharmacol. 94,266-275. WALLACH, E. E., DELACRUZ, A., HUNT, J., WRIGHT, K. H., AND STEVENS, V. C. (1975). The effect of indomethacin on HMG-HCG induced ovulation in the rhesus monkey. Prostaglandins 9,645-658. ZARROW, M. X., CALDWELL, A., HAFEZ, E. S. F., AND PINCUS, G. (1958). Superovulation in the immature rat as a possible assayfor LH and HCG. Endocrinology 63,748-75. ZELEZNIK, A. J., MIDCLEY, A. R., AND REICHERT, L. E. (1974). Granulosa cell maturation in the rat: Increased binding of human chorionic gonadotropin following treatment with follicle-stimulating hormone in vivo. Endocrinology 95,8 18-829.