Developmental changes in the gonadotropin ...

Developmental changes in the gonadotropin releasing hormone neuron of the female rabbit: effects of t oxifen citrate and pregnant mare serum gonadotropin WARREN G. FQSTER~ Department of Obstetrics and Gymecology, McMaster University Medical Centre, Hamilton, ON L8N 32.5, Canada

Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by Guangzhou Jinan University on 06/05/13 For personal use only.

JOHNE JARRELL. Depcap-brnent of Obstetrics and Gynecology, Foo~hillsHospital, Calgary, AB T2N 2T9? Canada

EDWARD '$I. YOUNGLAI Department of Obstetrics c a d Gynecology, McMaster University Medical Centre, Hamilton, ON U N 323, Canada Received November 26, 1992 FOSTER,WG., JARRELL, J.E , and YOUNGLAH, E.Ve 1993. Developmental changes in the gonadotropin releasing hormone neuron of the female rabbit: effects of tamoxifen citrate and pregnant mare serum gonadotropin. Can. J. Physiol. Phamacol. 71: 761-767. Developmental changes in imunostained gonadotropin releasing hormone neurons were demonstrated in female rabbits assigned to the following treatment groups: (i) tamoxifen citrate, 10 mg . kgsL days1, in sesame seed oil (vehicle) (n = 24) or (ii) vehicle alone (control, n = 24) for 108 days; and (iii) 58 IU of pregnant mare serum gonadotropin on postnatal days 22 and 25 (n = 24) or vehicle on nontreatment days. Treatments had no effect on the total number of immunostained cells, but there was a significant ( p = 0.0160) developmental shift from cells with smooth processes to rough. Group comparisons revealed that there was a significant ( p < 8.001) age-related increase in the number of rough cells in pregnant mare serum treated rabbits between days 25 and 75, indicating an advancement in the shift from smooth to rough cells. Plasma gonadotropin levels, ovarian follicular development, and the developmental shift from smooth to rough cells were markedly suppressed by tamoxifen treatment compared with rabbits of the control group, while no difference in estradiol levels were found. Our results suggest that a developmental shift in gonadotropin releasing hormone cell morphology from smooth to rough precedes sexual maturity in the female rabbit. Key words : development, sexual maturation, gonadotropin releasing hormone, puberty, immunohistochemistry . FOSTER,W.G., JARWELL, J.E, et Y~UNGLAI, E.V. 1993. Developmental changes in the gonadotropin releasing hormone neuron of the female rabbit: effects of tamoxifen citrate and pregnant mare serum gonadotropin. Can. 3. Physiol. Phrmacol. 71 : 761-767. On a dtmontrt des variations dans le dtveloppement des neurones lits B l'hormone de liberation des gonadotrophines chez des lapines rtparties dans Bes groupes suivants : (i) citrate de tamoxifhe, 18 mg - kg-' - jour-l, dans de l'huile de stsame (vthicule) (n = 24) ou (ii) vthicule seul (ttmoin, n = 24) pendant 108 jours; et (iii) 50 UI de gonadotrophine extraite du strum de jument gravide aux jours 22 et 25 postnataux (n = 24) ou vthicule les jours sans traitement. Ees traitements n'ont pas eu cl'effet sur le nombre total de cellules immunoco1ortes, mais il y a eu une variation significative ( p = 0,0160) dans le dtveloppement des cellules avec des transformations texturales de lisses rugueuses. Ees comparaisons intergroupes ont rtvtlt une augmentation relite h l'lge significative ( p < 8,081) du nombre de cellules rugueuses chez les lapines traittes au strum de jument gravide entre Bes jours 25 et 75, indiquant une progression dans la variation des eellules de lisses 2 rugueuses. Les taux de gonadotrophines plasmatiques, He dtveloppement folliculaire ovarien et Ia variation dans le dtveloppement des eellules de lisses B mgueuses ont Ctt netaement supprimts par le traitement au tamoxifhe cebmparativement h ee qui a tt6 observt chez les lapins du grsupe ttmoin, alors qu'il n'y a eu aucune difference dans les taux d'estradiol. Nos rtsultats sugg2rent qu'une variation dans le dtveloppernent de la morphologie cellulaire de lisse 2 rugueuse de 19homonede liberation des gonadotrophines prtcMe la maturitt sexuelle chez la lapine. Moa cb&s : dtveloppement, maturation sexuelle, gonadolibkrine, pubehtt, immunohistochimie. [Traduit par la rtdaction]

Imtrsductiom In a prior study (Foster and YoungLai 1991), two distinct ) cell types in fie hypogonadotropin releasing hornone (e thalamus of the adult female rabbit were demonstrated by ~~wcells pssessing rough imunohistochemicd or spinous processes were the dominant type (64%), while cells with relatively aspinous processes (34%) were also observed. Morphologic~ly distinct G n w cell types have been reported previously in the rabbit (Weindl and Sofroniew 1980) and the rat (Kirsch 1980; Jennes et al. 1985; ray and Hoffman 1986a, 19866, 1 9 8 6 ~ ;Wray and Gainer 1987; 'Author for correspondence at the following address: Room 338, Environmental Health Centre, Tunney's Pasmre, Ottawa, ON KIA OE2, Canada. Prtnted in Canada I Hmpr~mdau Canada

Merchenthaler et A. 1989). Indeed, Wray and Hoffman (19 8 h , 198669 1986~)have demonstrated that there is a developmental increase in GnRH cells that possess spinous processes. Previously, elevated levels of luteinking hormone (LH) and follicle stimulating hormone (FSH) have been demonstrated between postnatal days (PNDs) 30 and 60 in the female rabbit (Younghi 1986; YoungLai et al. 1990). These levels coincided with a decrease in opiate binding in the hypothalamus (Wilkinson and YoungLai 1986) and the onset of ovarian follicular development (de TurcWleirn et al. 1983). 'The rise in circulating gonadotropins during this period may play an essential role in ovarian development and the control of sexual maturation the rabGit (YoungLai et al. 1990). In contrast to the rabbit, circulating LH and FSH concentrations achieve peak Bevels at about 28 weeks of intrauterine life in the

CAN. J . PHYSIOL. PHARMACOL. VOL. 71, 1993


762

human fetus (Ojeda et al. 1988) and are equivalent to those measured in pstmenopausd wcsmen (Beck-Pewz et d. 1991). In addition, oogenesis commences around 15 weeks of gestation in the human fetus, and the number of oogonia is maximal by the 20th week. The rabbit is a useful model for studying the role of LH and FSH in ovarian development and subsequent sexual maturation since, in the rabbit, oogenesis occurs entirely after birth (YoungEai and Byskov 1983) and is complete by the end of the 2nd postnatal week (Peters et al. 1965), and ovarian follicular development begins around PND 58 (de Turcfieim et al. 11983). It is well established that sexual maturation is affected by estrogens and pituitary gonadotropins. However, the consequence s f blocking estrogen, in the otherwise intact animal, on GnRH neuron morphology and sexual maturation requires investigation. The present study was designed to investigate the effect of estrogen receptor blockade or premature ovarian activation on developmental changes in GnRH neuron morphology. Immature female rabbits were injected with pregnant mare semm gonadotropin (PMSG), to prematurely activate ovarian development, or with the potent antiestrogen tamoxifen citrate (TAM), to competitively inhibit estradiol (E2) binding to E2 receptors. Circulating levels of LH, FSH, and estradiol and ovarian follicle maturation, body weight (BW), and sexual maturation were also studied.

Materials and methods Animks Female Mew Zealand White rabbits (n = 72) were obtained at 22 days of age from a local breeder and housed in environmental conditions of 22 " C , 40% relative humidity , and a 12 h light : 12 h dark cycle. Access to rabbit chow (hrina, St. Louis, Mo.) and water was on an ad libitum basis.

Experimental design Rabbits were randomly assigned t s one of three treatment groups. In the first group (n = 24) rabbits received TAM (Stuart Pharmaceuticals, Wilmington, Del.), 10 mg kg-, . day-', emulsified in sesame seed oil (vehicle), by S.C. injection, for a total of 108 days. Control rabbits (a = 24) were administered the vehicle daily by S.C. injection throughout the study, while the remainder of the animals (n = 24) were injected s.c. with 50 IU PMSG (Sigma Chemical Cs., St. Louis, Ms.) on PNDs 22 and 25, and vehicle s n the remaining days. TAM is an antiestrogen of the triphenyl ethylene type and is used in the treatment of some forms of breast cancer (Legha et al. 1978). It has mixed estrogen agonist and antagonist effects, depending on the dose and species (Jordan et a%.1980), and has previously been shown to suppress circulating gonadotropin levels (Labhsetwar 1970). In the present study a dose of BO mg . kg-' . day-' was selected, since this dose has previously been used in the immature female rabbit (Baski et al. 1985) without adverse effects. Blood was collected weekly for deteminaticen of plasma LH, FSH, and E2 concentrations by radioimmunoassay. Rabbits that were receptive to proven bucks on at least one of two trials were considered sexually mature. Six rabbits from each treatment group were killed on days 25, 40, 75, and 208 with 3 cm3 sodium pentobarbital (Somnotol, [email protected], Missisauga, Bnt.). Tissues were fixed by whole-body perftksion (4 % paraformaldehyde, 0.1 % glutaraldehyde, 0.1% acrolein, in 0.1 M Sorensen's phosphate buffer containing I % sucrose, pH 7.3). One ovary was fixed for routine histology and follicular morphometry . The hypothalamus was dissected out, sectioned serially in the frontal plane (40 pm), and immuncestaind for GnRH as described previously (Foster and YoarngLai 1991). Briefly, the polyclonal antiserum LR, 7-5-79 (donated by Dr. Robert Benoit, MontrCd General Hospital) was used at a dilution of I : 15 000, and the sections were shined by the percexidase antiperoxidase method. Every 10th section was used as a control for inmunostaining I m u n o -

FIG. I. Developmental changes in body weight of TAM (@) and PMSG ( A ) treated rabbits versus controls (M). Each data point represents the mean f SEM of 4 -6 rabbits. *Significantly different at p < 0.05: **significantly different at p < 0.01.

reactive GnRH cells were quantified according to the methods of Abercrombie (1946) and described in detail by Wray and Hoffman (1984b). Briefly, immunoreactive GnRH cells with rough or spinous processes and cells with smooth or aspinous processes were quantified for each treatment on PNDs 25,40, and 75. Since all rabbits in the PMSG group were pregnant by 108 days of age, immunoreactive GnRM cells were quantified only in the TAM and control groups at this age. Radisimmunsasscays Plasma LH and FSH were measured by established radioimmnoassay techniques (Moor and YoungLai 1975; Armstrong et al. 1978). Briefly, the LH standard was WP 360A (Dr. A.F. Parlow), 1 ng of which was equivalent to 30 pg pure rabbit pituitary LLH (EX 130 GB, Dr. L.E. Reichert, Jr.) and guinea pig anti-rabbit LH 7F GPaLH (Dr. R. Scaramumi). Bvine LH (LER-1056-C2, Dr. E.E. Reichert, Jr.) was used for iodination. The sensitivity of the assay was 42 pg of the pure pituitary LH standard. Intra- and inter-assay coefficients of variation were 4.8 and 18% , respectively. Reagents for PSM assays were provided by Dr. A.H. Parlow. Rabbit FSH (AFP-9488-C) was used for iodination and standards. The antiserum (AFP-4-7-2 B -76) was raised in guinea pigs. The sensitivity s f the assay was 80 pg of the pure FSH standard. Intra- and inter-assay coefficients of variation were 13.2 and 21 % , respectively. Estradiol antiserum was prepared in the rabbit to estradiol-6-oxime conjugated to bovine semm albumin (Wielgosz et al. 1980). Estradiol standard was obtained from Steraloids, Wilton, N.H. [3H]E2 (New England Nuclear, Boston, Mass.) was used for competitive binding. Bound steroid was separated from free by charcoal extraction. The sensitivity of the assay was 10 pg of the pure E, standard used. Hntra- and inter-assay coefficients of variation were 6.33 and 6.91 %, respectively. Folkicular mha~phomefy Ovarian follicle counts were perfom& on 10 prn thick hemtoxylinand eosin-stained sections, using a Biquant image analysis system (R&M Biometries, Toronto, Ont.). Every fifth section was examined? and only follicles in which the plane of section passed through the oocyte nucleus were recorded. The mean number of growing and antral follicles were counted for each treatment group. In addition, the maximal follicle diameter was measured using digitizing morphornetry features of the Bioquant image analysis system (Jarrell et d. 1987). Data analysis Data were log transformed and analyzed by two-way analysis of variance (ANOVA, Stats Plus program; Human Systems Dynamics,


FOSTER ET AE.

FIG.2. (A) Immunoreactive spinous GnRH cell with rough contours (arrows). (B) An aspinsus or smooth immunoreactiv GnRH cell. Bar in Fig. 2B represents 30 ,urn for Fig. 2A and 15pm for Fig. 2B.

-

TABLE 1. The effect of treatment of immature female rabbits with TAM (10 mg * kg-' day-') or PMSG on the mean (kSEM) ovarian weight and the diameter of growing follicles (follicles with more than six granulosa cell layers but no antmm) and antral follicles

Age (PNDs) 25 40

75 108

Treatment

N

Ovarian weight (g)

Diameter of growing follicles (W)

Diameter of antral follicles (~ffla>

TAM Control BMSG TAM Csntral PMSG TAM Control BMSG TAM Contra1 PMSG

"Significantlydifferent at g bSignificawtly different at g

< 0.05. < 0.01.

Northridge, Calif .) . Group comparisons s f treatment means were performed by the Mann -Whitney unpaired t-test. A p-value of 0.85 was considered significant.

Results Ge~eeraleflects Two-way ANOVA revealed that treatments had a significant ( p < 0.001) effect on body weight (Fig. 1). Group csrnparisons showed that PMSG-treated rabbits were significantly heavier than rabbits of the eontrd group on PND 4 4 ( p < 0.05) and PNDs 75- 108 (g < 8.81). TAM-treated rabbits weighed significantly less at 558 ( p < 8.05) and 75 ( p < 8.01) days of age than rabbits of the control group. Two-way ANOVA demonstrated that treatment with TAM

and PMSG had a significant effect ( p < 0.001) on ovarian weights (Table I ) . Group comparissns revealed no differences between treatment groups until PND 75, when ovaries from rabbits s f the control group were significantly ( p 9: 0.05) heavier thaw those of the TAM group. At PND 108 ovarian weights s f the control group were significantly ( p < 0.01) greater than svaries from TAM-treated rabbits. %mmunop.eactivecell counls Two distinct populations of immuworeactive GnRH cells were observed. As described previously (Foster and YouwgEai 1991) one population possessed rough or spinous cell processes (Fig. 2 4 , while the second group was characterized by relatively smooth or aspinous processes (Fig. 2B).

CAN. J . PHYSIOL. PHARMACBL. VOL. 71, 1993

764

Control A PMSG @ T a m o x i f en


PMSG

FIG. 3. The ratio of immnoreactive smooth to rough GnRH cells counted in the rabbit brain of control and TAM- and BMSG-treated rabbits on BNDs 25, 40, 75, and 188. Each bar represents the mean f SEM of 4-6 rabbits. *Significantly different at the 0.01 level. -

Two-way ANOVA demonstrated that treatments significantly ( p = 0.0160) decreased the number of aspinous or smooth immunoreactive cells. Group comparisons revealed that it was not until day 75 that the number of smooth cells in the PMSG group (374.11 35.473 was significantly less ( p < 0.01) than in the control group (613.70 f 24.40, n = 4 ) . At 108 days of age there were significantly ( p = 0.0085) fewer smooth cells in the control group (340.95 + 12.19; n = 3) than in the TAM group (721.76 50.11; n = 5 ) . In the PMSG group the number of smooth G n W cells decreased significantly ( p < 0.001) from 736.28 f 6.84 on PND 25 to 374.11 +_ 35.47 on PNB '75. A similar change was observed only in the control group at 108 days of age. The number of smooth cells was almost identical in day 75 controls, the day 44 PMSG-treated group, and the day 108 TAMtreated group. The developmental change in the ratio of smooth to rough immunoreactive GnRH cells is illustrated in Fig. 3. Results of two-way ANOVA revealed that the number of immunoreactive cells with rough processes was not significantly different among the treatment groups. However; significant differences (p < 0.001) were found within treatment groups. The number of rough cells in the PMSG group increased from 305.87 f 25.05 on PND 25 to 577.81 39.81 on PND 75. A similar pattern was observed for the control group but not the TAM-treated rabbits. The total number of immunoreactive cells counted was unaffected by treatments, but there were differences in the timing of the switch from smooth to rough.

PIG. 4. Developmental changes in plasma FSH concentration of TAM ($) and PMSG ( A ) treated rabbits versus controls (m) on sampling days 22, 25, 28, 34, 40, 44, 58, 60, 75, 85, 92, 101, and 108. Each data point represents the mean k SEM of 4 -6 rabbits. **Significantly different at p < 0.01.

+

+

Developrnentakal changes in phsnau gonadotropins Circulating plasma FSH levels were significantly different ( p < 0.001) for the three treatment groups studied (Fig. 4 ) . However, the pattern of circulating FSH was qualitatively similar for both the PMSG and control groups. In both cases FSH levels were monophasic, with a peak at 44 days of age for rabbits of the control group (17.98 f 11.87 ng/mL) compared with 50 days for the PMSG group (24.56 6.88 ng/mL). Although FSH levels in the PMSG group did not achieve a

*

I

I

I

I

I

I

I

i

I

20 30 40 50 60 70 80 90 100 110

Age (days) FIG. 5. Developmental changes in plasma LH concentration of TAM ($9 and PMSG (A) treated rabbits versus controls (HI) on the same bleeding internal as in Fig. 3. Each data point represents the mean f SEM of 4 -6 rabbits. *Significantly different at p < 0.05; **significantly different at p < 0.01.

peak until 18 days after that of the control group, group cornparisons revealed that the levels were not significantly different at 40 days of age. In addition, the mean plasma FSH levels in the $MSG-treated group were significantly different ( p < 0.01) compared with the control group om PNDs 4.4 and 50 (20.25 f 4.02 to 24.56 f 6.88 vs. 13.94 f 3.82 to 13.96 k 2.50 ng/mL). In TAM-treated rabbits FSH levels fell from 4.25 0.97 ng/mE at 22 days of age to levels under 2.0 ng/mL for the remainder of the study. Indeed, the mean plasma FSK concentrations were significantly greater ( p < 0.01) for rabbits of the control group compared with TAM-treated animals on PNDs 34, 463, 44, 50, and 60. Circulating plasma LH levels were significantly affected by treatments ( p < 8.Wl). Mean plasma LH levels showed a precocious rise in the PMSG-treated rabbits (Fig. 51, which however, were not significantly different from the control EH values. Increased plasma LH in this group may be due to cross-reaction of PMSG within the radioirnmunoassay used. In

+

765

FOSTER BT A&.

-

a

Control

A PMSG

*

0

Barnoxifen

-4

,

TABLE2. The effect of treatment of immature female rabbits with TAM (10 rng kg-' - day-" or $MSG on the number of growing ovarian follicles (follicles with more than six granulosa cell layers but no antmm) and antral follicles


Age (PNDs) 25

48

I

I

I

I

I

I

I

I

I

20 30 40 50 68 70 80 90 188 110 Age (days) FIG. 6 . Developmental changes in plasma estradiol concentration of TAM ( 0 ) and PMSG (A) treated rabbits versus controls (m). Plasma samples from PNDs 25, 34, 40, 50, Q8, $5, and 108 were assayed. Each data point represents the mean _+ SEM of 4 -6 rabbits. *Significantly different at p < 0.05.

the control group a biphasic pattern was observed, with peaks on PNDs 40 (1.10 k 0.30 ng/mL) and 75 (1.98 $ 0.76 ng/mL). The mean plasma LH levels were significantly greater ( p < 0.81) in the rabbits of the control group at PND 75 compared with PMSG- and TAM-treated rabbits. Both PMSG and control groups had adult levels (< 1 ng/mL) of circulating LH by the 85th day of age. Unlike PMSG or control rabbits, circulating levels of LH in the TAM group were suppressed throughout the study. Indeed, circulating LH levels in TAM-treated rabbits were significantly ( p < 0.01) lower than the values for rabbits of the control group on PNDs 34, 40, 44, 50, and 60. In the PMSG group, the mean plasma LH concentration peaked at PND 50 (1.81 & 0.66 ng/mL). Group comparisons revealed that the mean plasma LH level for the PMSG-treated group was greater than the control group only on PND 58 ( p c 0.05). Developmental changes in mean p l a s m estradisl Plasma & levels were significantly ( p = 0.003) affected by treatments, as shown by two-way ANOVA. Group comparisons revealed that in the PMSG group plasma E2 levels were 2.47 and 3.11 times control levels ( p < 0.05) on PNDs 25 and 34, respectively (Fig. 6). At 60 days of age plasma E2 concentrations in the PMSG group were 3.04 times those of control values. No difference in circulating levels of E2 could be found between the control and TAM-treated rabbits at any age studied despite the profound suppression of pituitary gonadotropin levels observed in the TAM group. Follicular morpksmetry Two-way ANBVA revealed significant treatment effects for growing and antral follicles ( p < 0.001). Group comparisons demonstrated that rabbits in the TAM group had significantly fewer growing follicles at PNDs 40 ( p = 0.8286), 75, and 108 ( p = 0.0143) than did rabbits of the control group (Table 2). Comparison of treatment means for antral follicles also revealed a significant difference ( p < 0.05) between the 'TAM group and rabbits of the control group at PND 75 (Table 2). The maximal diameter of growing follicles in the TAM group at PNDs 48 (p = 0.Q468) and 75 ( p = 0.0 137) were significantly less than the values for rabbits in the control group.

75 100

Treatment

M

Growing follicles

Antral follicles

TAM Control PMSG TAM Control PMSG TAM Control PMSG TAM Control BMSG

"Significantly different at p = 0.0286. bSignificantly different at p = 0.0143.

Treatment with PMSG had no demonstrable effect on ovarian follicle development compared with rabbits of the control group. The number s f antral follicles in the PMSG-treated group, although greater than in the control group, did not achieve significance (p = 0.057 1). Sexual receptivity Rabbits of the PMSG group weighed 3.0 kg by PND 85, and all were sexually receptive by PND 92 (BW 3.1 $_ 0.1 kg). In contrast, no successful matings were observed in the control or TAM-treated rabbits ( p = 0.02), neither of which had achieved a sexually mature body weight at this age. However, rabbits in the control group were sexually receptive at PND 108 and had a mean BW of 3.3 f 0.1 kg. TAM-treated rabbits failed to demonstrate sexual receptivity (p = 0.02) within the time frame of this study regardless of attaining a mean body weight of 3.1 kg.

Discussion Developmental changes in hypothalamic GnRH neuron morphology were demonstrated in this study. There was an agerelated decrease in the number of smooth or aspinous cells, which was advanced by BMSG treatment relative to the control group. In contrast, daily TAM treatment in the sexudly immature rabbit was shown to prevent the apparent shift in the number of smooth to rough cells. In addition, PMSG-treated rabbits attained sexual maturity earlier than did the control group, while 'TAM treatment profoundly impaired sexual development. In TAM-treated rabbits developmentd changes in GnRH neurons were found, but these changes were delayed relative to those found in the control and PMSG groups. The number of smooth or aspinous cells decreased, while rough or spinous cells increased as rabbits of the control group approached sexual maturity. These developmental changes were advanced in the PMSG-treated rabbits, and the change was significant at PND 75 compared with the control group. The control group was significantly different from TAMtreated rabbits on PND 108. The present findings support similar developmental changes observed in GnRH neurons in ; the rat brain (Wray and Hoffman 1986a. 1986b, 1 9 8 6 ~Wray and Gainer 1987). At present it is not clear what the develop-


766

CAN. 4. PHYSIOL. PHARMACOL. V8L. 71. I993

mental changes in the morphology of GnRH cells seen in the current study represent. Moreover, the functional significance of these morphological changes is unknown. Circulating gonadotropin levels in rabbits of the control and PMSG groups achieved peak levels between 30 and 60 days of age. These results are harmonious with prior reports in the rabbit (YoungLai 1986; Younghi et al. 1998). However, during this developmentd stage, plasma LH and FSH were suppressed by TAM treatment. In a prior report (Wilkinson and YoungLai 1986) it was proposed that increased circulating levels of gonadotropins occurring between 30 and 60 days of age in the female rabbit may be due to development of disinhibition of opioidergic inhibition of gonadotropin secretion, It is generally agreed that estrogens do not act directly on GnRH neurons, but E2 has been shown to colocalize with both catecholaminergic (Grant and Stumpf 1973; Sar and Stumpf 1981; Sar 1984) and opioidergic (Morrell et al. 1984, 1985; Jirikowski et al. 1986) neurons. It has been shown that TAM reverses E2 facilitation of Kg-evoked GnRH release in the rat (Brouva et al. 1984). Moreover, chronic estrogenization in the female rat induced by a single injection of estradiol valerate was shown to be selectively toxic to hypothalamic @-endorphin neurons (Des~ardinset al. 1993). Thus, we propose that TAM blocked the process of sexual maturation in the female rabbit by inhibiting E2-induced developmental changes in the morphology of hypothalamic GnRH neurons. The potential effects of 'PAM treatment on developmental changes in opiatergic inhibition of gonadotropin secretion in the female rabbit needs c which excite to be evaluated. In addition, h y p t l - ~ d hneurons G n W release or GnRH neurons themselves in addition to pituitary gonadstropes are also potential sites of TAM action. Circulating levels of E2 in the TAM group were comparable with control values regardless of the profound suppression of LH and PSH induced by 'FAM treatment. Consequently, it is suggested that TAM acts directly upon the ovary to induce E2 synthesis. This view is supported by the observation that in premenopausal women TAM induced a 2- to 8-fold increase in serum I!$ levels in the follicular phase without causing significant changes in the secretion of FSH and LH (Groom and Griffiths 1976). Alternatively, Groom and Griffiths (1976) suggest that TAM treatment may reduce prolactin (PWL) secretion and thus permit emharaced ovarian stimulation by normal concentrations of gonadotropins. In the current study, potential ovarian stimulation due to impaired PRL secretion in TAM-treated rabbits cannot be excluded. Absence of differences in circulating levels of E2 found in the control and TAM-treated rabbits may be the consequence of increased numbers of gonadotropin receptors, modulation of growth factor effects csn granulosa cells, or enhanced argzrnatase activity induced by the weak estrogenic effects of TAM. The effect of TAM on ovarian steroidogenesis requires study. A body weight of 3.0 kg and sexual receptivity were used as a means of assessing sexud maturity* The relationship between b ~ d yweight and sexual maturity in the female rabbit has been well established in prior studies (Hulat et al. 1982; YcsungLsai 1986). In the current study, BMSG-treated rabbits achieved sexual maturity (PND 92) earlier than control rabbits (BND 108), which in turn were advanced relative to TAMtreated rabbits. Results of the current study suggest that PMSG induces a precocious puberty by stimulating ovarian E2 secretion and subsequent weight gain. The initial increage in circulating E2 observed in this study is taken to be the result of PMSG stimulation of ovarian aromatase. In the rat PMSG has

been shown to cause superovulation by rescuing follicles from atresia (Brau and Tsafriri 1980). PMSG was also found to decrease atresia as well as to recruit "reserve" follicles in the hamster ovary (Chiras and Greenwald 1978). It is possible that PMSG activates the immature ovary prematurely and induces both maturation of the ovary and the hypothalamus. Thus, we propose that PMSG in the present study prematurely activated the ovary as demonstrated by elevated E2 levels on PNDs 25 and 34. Furthermore, elevated E2 levels induce maturational changes in the hypothalamic mechanisms that regulate gonadotropin secretion. In support of this interpretation, physiological doses of E2 have been demonstrated to advance puberty in the rat (Ramirez and Sawyer 1965). Moreover, E2 has been shown to induce synaptic remodeling of the rat brain (Nishizuka and Arai 1981; Garcia-Segura et al. 1986), and these changes have been correlated with puberty onset (Clough and Rodriguiz-Sierra 1983). Consequently, it is suggested that PMSG affects the rate of sexual maturation through premature activation of ovarian development and the subsequent maturation of hypothalamic mechanisms that regulate gonadotropin secretion. The results of the present study extend the literature with respect to the effect of estrogen on developmental changes in hypothalamic GnRH neurons of the rabbit hypothalamus. Moreover, it was shown that developmental changes in GnRH neurons are impaired by 'FAM treatment and sexud maturation is retarded. At present the morphological characteristics and functional significance of these alterations is unknown. Regardless, our findings suggest that a developmentally related transition in GnRH neuron cell populations occurs in the rabbit hypothalamus, and this change is in part estrogen dependent. Abercrombie, M. 1946. Estimation of nuclear population from microtome sections. Anat. Rec. 94: 239 -247. Armstrong, R.W., Gauldie, B., and YoungLai, E.V. 1978. Effects of active in~~llunizationof female rabbits against testosterone. J. Endocrinol . 79: 339 - 347. Baski, S.N., Hughes, M.B., and Light, K,E. 1985. Alteration s f dopamine metabolism in different brain regions of the rabbit by estradiol and tamoxifen. Neuroscience (Oxford), 14: 1053- 18359. Beck-Pe~csz,P., Badmanabhan, Y , Baggiani, A.M., Cortelaazi, D., Buscaglia, M.M.,M d r i , G., Marcsni, A.M., Pardi, G., and Beitins, 1.Z. 1991. Maturation of hypothalamic -pituitary gonadal function in normal h u m n fetuses: circulating levels of gonadotropins, their common a-subunit and free testosterone, and discrepancy between immueaslogical and biological activities of circulating follicle-stimulating hormone. $. Glin. Endocrinol. Metab. 73: 525-532. Brau, R.H., and Tsafriri, A. 1984). Effect of PMSG on follicular atresia in the immature rat ovary. B. Weprod. Fertil. 59: 267 -272. Chiras, D.B., and Greenwald, G ,S. 1978. Ovarian follicular development in cyclic hamster treated with a superovulatory dose of pregnant mare's serum. Biol, Reprod. 19: 895 -801 . Clough, R.W, and Rodripiz-Sierra, J.F. 1983. Synaptic changes in the hypothalamus of the prepubertal female rat administered estrogen. Am. I. Anat. 167: 205-214. Besjardins, G.C., Brawer, J.R., and Beaudet, A. 1993. EstradioH is selectively weuratsxie to hypothalamic 6-endorphin neurons. Endacrinalogy, 132: 86 -93. de Turcheirn, M.,Berger, M . , Jean-Faticher, C.L., Veyssiere, G . , and Jean, C,J. 1983. Changes in ovarian oestrogens and in plasma gonadotropins in female rabbits from birth t s adulthood. Acta Endocrinol. 103: 125- 1383. Drouva, S.V., Laplante, E., Gautron, B.-P., and Kordon, C. 1984. Effects of 17 B-estradiol on LH-RH release from rat mediobasal hypothalamic slices. Neuroendocrinology , 38: H 52 - 157.


FOSTER ET AL.

Foster, W.G., and YoungLai, E.V. f 99%.An immunohistochemical study of the GnRH neuron morphology and topography in the adult female rabbit hypothalamus. Am. J. Anat. 191: 293-300. Garcia-Segura, L.M ., Baetens, D., and Naftolin, E 1986. Synaptic remodelling in arcuate nucleus after injection of estradiol valerate in adult female rats. Brain Res. 366: 13 1 - 136. Grant, L.D., and Stumpf, W.E. 1973. Lmalization s f 3H-estradiol and catecholamines in identical neurons in the hypothalamus. J. Histochern. Cytochem. 21: 404. Groom, G.V., and Griffiths, K. 1976. Effect of the antiestrogen tamoxifen on plasma levels of luteinizing hormone, prolactin, oestradiol and progesterone in normal pre-menopausal women. J. Endocrinol. 70: 42 1 -428. Harlot, L.R., Mariana, J.C., and Lebas, F. 1982. L'Ctablissement de la pubertC chez la lapine (folliculogenkse et ovulation) : effet du rationnement alimentaire. Weprod. Nutr. Bev. 22: 439-453. Jarrell, J., YoungLai, E.V., Ban, W., McMahon, A., and Belkck, L. 1987. Ovarian toxicity of cyclophospharnide alone and in combination with ovarian irradiation in the rat. Cancer Res. 47: 23402343. Jennes, L., Staampf, W.E., and Sheedy, M.E. 1985. Ultrastructural characterization of gonadotropin-releasing hormone (GnRW)producing neurons. J. Comp. Neurol. 23%: 534 -547. Jirikowski, G,R,, Merchenthaler, I., Wieger, G.E., and Stumpf, W.E. 1986. Estradiol target sites immunoreactive for 0-endorphin in the arcuate nucleus of rat and mouse hypothalamus. Neurosci. Lett. 65: 121 -126. Jordan, V.C., Allen, K.E., and Bix, C.J. 1980. Pharmacology of tamoxifen in laboratory animals. Cancer Treat. Rep. 64:745 -759. Kirsch, B. 1980. Two types of luliberin-immunsgeactive perikarya in the preoptic area of the rat. Cell Tissue Res. 212: 44.3-455. Labhsetwar, A.P. 1970. Role of estrogens in ovulation: a study using the estrogen antagonist ICH 46474. Endocrinology, 87: 542 -5%1. Legha, S.S., Davis, H.L., and Muggia, E M . 1978. WormePnal therapy of breast cancer: new approaches and concepts. Ann. Intern. Med, 88: 69-77. Merchenthaler, I., Setals, G., Csontos, C., Petwsz, P., Flerko, B., and Negro-Vilar, A. 1989. Combined retrograde tracing and imunocybchemica1 identification sf luteinizing hsmone-releasing hormone and somatostatin containing neurons projecting to the median eminence of the rat. Endocrinology , 225: 28 12 -282 1. Moor, B.C., and YoungEai, E.V. 1975. Variations in peripheral levels sf EH and testosterone in adult male rabbits. J. Weprod. Fertil. 42: 259-266. Morrell, J J . , Schwamel-Fukuda, M., Fahrbach, S.E., and Pfaff, D.W. 1984. Axonal projections and peptide content s f steroid hormone concentrating neurons. Peptides (Elmsford) , 24(Suppl. 1): 227 -239. Morrell, J.I., McGinty, J.E, and Pfaff, D.We 1985. A subset of

767

P-endorphin- or dynorphin-containing neurons in the medial basal hypothalamus accumulates estradiol. Neuroendocrinology , 41: 417 -426. Nishizuka, M.. and Arai, Y. 198 1 . Organizational action of estrogen on synaptic pattern in the amygdala: implications for sexual differentiation of the brain. Brain Res. 283: 422-426. Ojeda, S.R.. Andrews, WW., Advis, J.P., and Smith-White, S. 1980. Recent advances in the endocrinology of puberty. Endocrinol . Rev. 1: 228-257. Peters, H.. Levy, E., and Crone, M. 1965. Oogenesis in rabbits. J. Exp. Zool. 158: 169-180. Wamirez, V.D., and Sawyer, C .H. 1965. Advancement of puberty in the female rat by estrogen. Endocrinology, 76: 1 158 - 1 168. Sar, M. 1984. Estradiol is concentrated in tyrosine hydroxy lasecontaining neurons of the hypothalamus. Science (Washington, D.C.), 223: 938-940. Sar, M., and Staampf, W.E. 1981. Central noradrenergic neurons concentrate %I-oestradiol . Nature (London), 289: 303 - 3 13. Wielgosz, G.J., Low, M.J., and YoungLai, E.V. 1980. Effects s f immunizing rabbits against oestradiol-6-oxime -BSA on ovarian follicular development. Acta Endocrinol. 94: 235 -24 1. Weindl, A.. and Sofroniew , M.V. 1980. The morphology of LRW and sxytocin neurons. Bxp. Brain Wes. Suppl. 3: H - 17. Wilkinson, M., and YoungLai, E.V. 1986. Development of opiate ([3H]-naloxone) binding sites in female rabbit brain: correlation with prepubertal gonadotropin secretion. Biol. Reprod. 35: 572578. Wray , S., and Gainer, H. 1987. Effect sf nesnatal gonadectomy on the postnatal development of LHWW cell subtypes in male and female rats. Neuroendwrinology , 45: 4 13 -4 19. Wray , S., and Hoffman, G. 1986a. Postnatal morphological changes in rat LHWH neurons correlated with sexual maturdtion. Neuroendocrinology, 43: 93 -97. Wray, S., and Hoffman, G. H986b. Cateeholamine innervation of LH-RH neurons: a developmental study. Brain Res. 399: 327 - 33 1. Wray, S., and Hoffman, G. 1986~.A developmental study of the quantitative distribution of LHRH neurons within the central nerGous system of postnatal male and female rats. J. Comp. Neurol. 25%:522-531. YoungEai, E.V. 1986. Age related changes in the concentrations of semm gonadotrophins and cholesterol in the female rabbit. J. Endscrinol. 109: 287 -290. YoungLai, E.V., and Byskov, A.G. 1983. Relationship of meiotic prophase and ornithine decarbsxylase in the neonatal rabbit ovary Cell Tissue Res. 231: 56%-570. YoungEai, E.V., Thompson, N., and Foster, W. 1990. Effects of steroid implants s n the pubertal increase in circulating gonadotropins and sexual receptivity in the female rabbit. J. Steroid Bischem. 35: 415 -419.