C S I R O
P U B L I S H I N G
Reproduction, Fertility and Development Volume 11, 1999 © CSIRO Australia 1999
A journal for the publication of original work, review and comment in the field of reproductive biology, reproductive endocrinology and developmental biology, including puberty, lactation and fetal physiology when they fall within these fields
w w w. p u b l i s h . c s i r o . a u / j o u r n a l s / r f d All enquiries and manuscripts should be directed to Reproduction, Fertility and Development CSIRO PUBLISHING PO Box 1139 (150 Oxford St) Collingwood Telephone: 61 3 9662 7629 Vic. 3066 Facsimile: 61 3 9662 7611 Australia Email:
[email protected]
Published by CSIRO PUBLISHING for CSIRO Australia and the Australian Academy of Science
Reprod. Fertil. Dev., 1999, 11, 293–302
Role of peripheral and central aromatization in the control of gonadotrophin secretion in the male sheep T. P. SharmaA, D. Blache, M. A. Blackberry and G. B. MartinB A
Faculty of Agriculture (Animal Science), The University of Western Australia, Nedlands, WA 6907, Australia. Current address: Reproductive Sciences Program, University of Michigan, 300 N. Ingalls Building, 11th floor, Ann Arbor, MI 48109–0404, USA. B To whom correspondence should be addressed. email:
[email protected]
Abstract. Both testosterone and its aromatized metabolite, oestradiol-17β, are known to act centrally on the secretion of GnRH, but the major site of aromatization is not clear as aromatase activities are found in numerous tissues including brain and testis. Here, we tested the importance of central aromatization of testosterone using a non-steroidal aromatase inhibitor, fadrozole. To distinguish between testicular and nontesticular sites, five intact and five testosterone-infused castrated rams (600 µg kg–1 per 24 h for 3 days) were given four injections of fadrozole (i.m; 500 µg kg–1) at 48, 52, 64 and 68 h relative to the start of testosterone infusion. Control rams (n = 5) received vehicle only. Fadrozole treatment decreased plasma oestradiol-17β concentrations and increased the LH pulse frequency in both intact rams and testosterone-treated castrates, suggesting that non-testicular sites of aromatization are important in the control of pulsatile LH secretion. To test the importance of central aromatization, intact rams (n = 5) were infused into the third ventricle with vehicle (artificial cerebrospinal fluid) or with fadrozole (20 and 200 µg kg–1 per day). After two weeks, the same two doses of fadrozole were infused intravenously instead of intracerebrally. Central infusion of fadrozole did not affect plasma oestradiol concentrations but increased LH pulse frequency. Only the highest dose increased LH pulse frequency when infused intravenously. In conclusion, central aromatization is involved in the control of pulsatile LH secretion in male sheep. Keywords: oestradiol, testosterone, fadrozole, male, ram.
Introduction Oestradiol-17β and dihydrotestosterone (DHT) are metabolites of testosterone produced by the action of aromatase and 5α-reductase respectively. These hormones are produced in many tissues (Naftolin et al. 1975; Roselli et al. 1985; Hotzel et al. 1995; Negri-Cesi et al. 1996), the testis and brain being the most important with respect to the control of gonadotrophin secretion. All three steroids appear to exert negative feedback on the secretion of GnRH/LH in male sheep. The clearest evidence comes from studies showing (1) an immediate rise in gonadotrophin secretion following castration (Pelletier 1970; Riggs and Malven 1974; Parrott and Davies 1979) and restoration of gonadotrophin secretion to intact levels following treatment with testosterone, dihydrotestosterone, or oestradiol-17β (Pelletier 1970; Bolt 1971; Riggs and Malven 1974; Karsch and Foster 1975; Schanbacher and Ford 1977; Parrott and Davies 1979; Tilbrook et al. 1991); (2) an increase in the secretion of LH after immunization against oestradiol-17β in intact or testosterone-treated castrated rams (Schanbacher 1984; Sanford 1985; Monet-Kuntz et al. 1988); and (3) an increase in the secretion of LH following treatment with the aromatase inhibitor, aminoglutethimide, in testosterone-treated, castrated rams (Schanbacher 1984). Whereas such observations ©CSIRO Australia 1999
suggest that the products of aromatization and 5α-reduction play a role in negative feedback, they do not indicate whether this role is primarily due to the conversion of testosterone in the central nervous system or in other tissues. To distinguish between testicular and other sites of aromatization, we compared intact rams with testosterone-treated castrates. We used the non-steroidal aromatase inhibitor, fadrozole, which has been reported to be 200–400 times more potent and more specific than aminoglutethimide (Miller 1996), the inhibitor used in sheep by Schanbacher (1984). To test whether aromatization occurring within the brain is involved, we infused a low dose of fadrozole into the third ventricle. Materials and methods Mature Merino sheep were used in all experiments. They were kept indoors (latitude 31°56´S) in individual pens under natural lighting after arrival from the farm, and were only subjected to treatments after at least two weeks of acclimatization. All the animals were fed with 800 g wheaten chaff plus 100 g lupin grain per day supplemented with minerals. Water was provided ad libitum. All the experiments performed were approved by the Animal Ethics and Experimentation Committee of The University of Western Australia. Experiment 1 was done in April (breeding season) and the preliminary study and Experiment 2 were done in September–October 10.1071/RD99084
1031-3613/99/050293
T. P. Sharma et al.
294
(non-breeding season). Merino rams respond very poorly to photoperiod and show very little seasonal change (Martin et al.1994b). Preliminary study: effect of fadrozole on oestradiol concentrations during the follicular phase in female sheep Fadrozole is effective in inhibiting aromatization in humans, rats and monkeys, but its use in sheep has not been reported. Gonadotrophin-treated females were used to determine an effective dose of fadrozole because they have higher plasma concentrations of oestradiol-17β than males and thus any effects of fadrozole can be detected more readily. The oestrous cycles of 20 ewes (live weight 50.1 ± 2.7 kg) were synchronized using two injections (11 days apart) of Dinoprost trometamol, a synthetic analogue of prostaglandin F2α (Lutalyse; Upjohn Pty Ltd, Rydalmere, NSW, Australia). Twenty-four hours before the second injection, the ewes were injected (i.m.) with 500 IU of pregnant mare serum gonadotrophin (Pregnecol; Horizon Technology Pty Ltd, Sydney NSW, Australia). The ewes were allocated to five groups of four, with each group having the same average bodyweight. Twenty-four hours after the second prostaglandin injection, when the ewes were expected to be in an early preovulatory stage, four groups were injected (i.m.) with saline containing 125, 250, 500 or 1000 µg kg–1 bodyweight of fadrozole (CGS 16949A; Novartis Pharma Inc., Basel, Switzerland). The treatment was repeated 4 h later. The 4-h interval was chosen beccause fadrozole has a half-life of 10–12 h and two injections 4 h apart should maintain minimum inhibitory concentrations for 16 h, the end point for collection of follicular fluid. The fifth group (control) received injections of normal saline instead of fadrozole. Blood was sampled every 4 h starting from 4 h before the first injection (0 h) to 16 h after the last injection. Plasma was separated and stored at –20°C. Experiment 1: effect of fadrozole treatment on secretion of gonadotrophins and steroid hormones in male sheep We used ten intact (48.5 ± 1.35 kg; 2–3 y old) and ten castrated (55.3 ± 2.1 kg); 2–3 y old poll Merino rams in this experiment. Castration was done six months before the experiment began and the animals were implanted with a subcutaneous silastic implant containing oestradiol-17β (2 cm long, OD 5 mm, ID 3 mm) at the time of gonadectomy to maintain some steroid action at the hypothalamus and pituitary. These implants were removed 1 month before the experiment and were replaced with 14 pellets (Ropel®, Dover Laboratories Pty Ltd, Silverwater NSW, Australia), each containing 23.8 mg of testosterone propionate (equivalent to 19.7 mg of testosterone). This treatment was sufficient to maintain testosterone concentrations in plasma in the range seen in intact rams, but failed to reduce LH pulse frequency to normal values seen in intact rams. We therefore added extra testosterone to the castrated group by continuous jugular infusion (600 µg kg–1 bodyweight per 24 h in ethanol : normal saline, 1 : 1) at a flow rate of 400 µL h–1 for 72 h. The animals were fitted with indwelling jugular catheter for blood sampling one day before the start of experiment. Five intact and five castrated rams were injected (i.m.) with 500 µg kg–1 bodyweight of fadrozole at 48, 52, 64 and 68 h (0 h = start of testosterone infusion), the 4-h intervals being chosen because the preliminary study suggested that they provided sufficient fadrozole for 12 h (from 52 to 64 h). We then repeated the regime to maintain effective levels up to 72 h, when the experiment ended. The remaining five rams from each group served as controls and received only saline injections. Blood was sampled from all rams every 20 min from 36 to 72 h relative to the start of testosterone infusion. In the animals receiving i.v. infusion, blood was sampled from the opposite jugular vein through a second indwelling catheter. Plasma LH concentration was measured in all samples and all other hormones were measured in 12 h pools made by subsampling 100 µL from each serial sample during the periods 36–48 h, 48–60 h and 60–72 h.
Experiment 2: effect of central inhibition of aromatization on secretion of gonadotrophins and steroid hormones in male sheep Three groups of five testis-intact, mature Merino rams were fitted with intracerebroventricular cannulae using the technique described by (FabreNys et al. 1991). Two weeks after surgery, blood was collected every 20 min for 24 h to measure the pretreatment concentrations of hormones. After a 12 h break, 2 groups were infused constantly for 36 h into the third cerebral ventricle with fadrozole dissolved in artificial cerebrospinal fluid [aCSF; (Nilsson et al. 1991)]. The doses infused were 20 µg kg–1 per 24 h and 200 µg kg–1 per 24 h, and flow rate was 400 µL h–1. These two doses are 1/100 and 1/10th of the dose (2 mg kg–1 per 24 h) given i.m. in Experiment 1. Rams in the third group were infused with vehicle only. Blood was again sampled every 20 min during the last 24 h of infusion of fadrozole or aCSF. After two weeks, the rams were re-allocated randomly to the three treatments and the experiment was repeated, except that fadrozole was dissolved in normal saline and infused into the jugular vein instead of the third ventricle. Plasma LH concentration was measured in all samples and all other hormones were measured in pools made by subsampling from each serial sample over the 24-h periods. Hormone assays Luteinizing hormone was estimated in all plasma samples using a double antibody radioimmunoassay (Martin et al. 1980). The ovine LH (CNRS M3, biopotency 1.8 IU per mg NIH-LH-S1) used for standards was kindly supplied by Dr. M. Jutisz, College de France, Paris and the ovine LH (oLH-1-2) used for iodination was kindly provided by NIDDK, Torrance, CA, USA. The mean limit of detection of the four assays was 0.3 ng mL–1. The between-assay coefficients of variation (CV) were calculated using six replicates of three pooled samples containing 0.8 (7.4%), 2.4 (1.9%) and 4.2 ng mL–1 (4.7%). The within-assay CV (mean ± SEM) were 6.5 ± 0.5, 5.0 ± 1.38 and 6.4 ± 1.17% for the same pools. Plasma FSH concentrations were estimated by using a double antibody RIA as described by (Atkinson and Adams 1988). The ovine FSH (oFSH-11) for iodination and standards and the antisera were kindly donated by NIDDK. All samples were analysed in a single assay for which the limit of detection was 0.3 ng mL–1. The within-assay CV were 5.4% at 0.9 ng mL–1 and 10.8% at 1.9 ng mL–1. Plasma oestradiol-17β concentration was measured using affinity chromatography (antibody–sepharose) for extraction of samples and using an iodinated tracer as described by (Webb et al. 1985). The limit of detection was 0.4 pg per tube. All the samples were measured in one assay for which the within-assay CV was 12.1% at 6 pg mL–1. Plasma testosterone and DHT were measured by the method of (Puri et al. 1981) after re-validation in our laboratory for sheep plasma. Plasma is extracted twice using diethyl ether and reconstituted in gelatine phosphate buffer. One aliquot of the reconstituted extract is used for measuring total androgens. The rest is treated with 0.5% potassium permanganate for 30 min at room temperature to oxidize testosterone, and then extracted using diethyl ether. This second extract is assayed for DHT. Oxidized testosterone does not cross react with the antibody and the potassium permanganate does not affect the DHT assay. Testosterone values are calculated by subtracting the DHT values from total androgens. The bound and unbound hormones were separated using dextran-coated charcoal. Antiserum #457 used in the assay was kindly provided by Dr R. I. Cox (CSIRO, NSW, Australia) and it cross reacts 100% with testosterone, 98% with DHT, 4.7% with androstenedione, 17% with 4-androsten-3β,17β-diol, and less than 1% with oestradiol-17β, oestrone, oestriol and cortisol. Validation of the procedure: (1) we assayed samples over the range of the standard curve (0.72–200 pg per tube) with our assay and with a commercial kit (Testosterone/DHT Assay System; Amersham Australia) and the values were closely correlated (r = 0.96 for both DHT and testosterone); (2) we measured recovery in samples to which we had added DHT (76% at 1000 pg mL–1; 92% at 500 pg mL–1; 91% recovery at 250 pg mL–1, this latter concentration being similar to values seen in
Aromatization and gonadotrophin secretion in rams
295
most of our samples); (3) DHT was undetectable in samples to which we added testosterone up to a final concentration of 32 ng mL–1. Data analysis Pulses of LH were located using the Munro pulse analysis program (Zaristow Software, East Lothian, Scotland). In all experiments, data were analysed using repeated measures analysis of variance (ANOVA) and Fisher’s least significant difference test for comparisons between groups if ANOVA revealed significant effects. In Experiment 1, the data were grouped into periods before treatment (36–48 h) and after treatment (48–60 h and 60–72 h) for analysis.
Results Preliminary study Oestradiol concentrations were high or rising in all groups, as expected since the ewes were in mid-follicular phase. Saline injections did not significantly affect plasma oestradiol concentrations. The ewes treated with fadrozole at 250, 500 and 1000 µg kg–1 showed significant decreases in the plasma concentrations of oestradiol at 4 and 8 h after the first injection, relative to the pretreatment values at 0 h (Fig. 1). In the ewes treated with fadrozole at 500 and 1000 µg kg–1, the values at 4 and 8 h were also significantly (P
