Regulatory Peptides, 42 (1992) 135-144

135

© 1992 Elsevier Science Publishers B.V. All rights reserved 0167-0115/92/$05.00

REGPEP 01242

Regulation of somatostatin and growth hormone-releasing factor by gonadal steroids in fetal rat hypothalamic cells in culture G u m e r s i n d o Fern~mdez b, F r a n c o S~.nchez-Franco b, M a r i a Teresa de los Frailes a, R o s a M a r i a Tol6n a, M a r i a Jesfis L o r e n z o a, Judith L6pez a and L u c i n d a Cacicedo a a Servicio de Endocrinologia, Hospital Ramdn y Cajal and b CNIC-Instituto Carlos IH, Madrid (Spain)

(Received 20 February 1992; revised version received 30 July 1992; accepted 3 September 1992)

Key words: Gonadal steroid; Somatostatin; Growth hormone-releasing hormone; Fetal hypothalamic neuron

Summary The mechanism underlying the sexually dimorphic pattern of growth hormone (GH) secretion in the rat has not been clearly elucidated. In the present study, we assayed the possible direct effect of gonadal steroids on both somatostatin (SS) and growth hormone-releasing factor (GRF) in fetal rat hypothalamic cells in culture. Hypothalamic cells, obtained by mechanical dispersion, were maintained as monolayer cultures in serum-supplemented medium. After 20 days in culture, cells were incubated with serum free medium containing testosterone (T, 10, 20, 40 ng/dl) or estradiol (E, 0.1, 1, 10 ng/dl) for 48 h. At the end of the experiments, immunoreactive SS (IR-SS) and immunoreactive G R F (IR-GRF) were measured by specific radioimmunoassays (RIAs) in media and cell extracts. After 48 h of incubation with testosterone, somatostatin in both media and cells was significantly reduced. On the contrary, this treatment lead to a dose-dependent increase in media and cell G R F content. When cells were incubated with estradiol for 48 h, a significant inhibition in medium S S release was observed, whereas intracellular S S slightly increased at the highest concentration of 10 ng/dl. Estradiol treatment resulted in an inconsistent decrease in media and cells IR-GRF. Our results indicate that both SS and G R F are under the influence of testosterone and estradiol acting at the hypothalamic level, and furthermore suggest that at this stage

Correspondence to." L. Cacicedo, Servicio de Endocrinologia, Hospital Ram6n y Cajal, 28034 Madrid, Spain.

136

of brian development, gonadal steroids may regulate GH secretion through their ability to modulate hypothalamic SS and GRF.

Introduction

A sexual dimorphic pattern of growth hormone (GH) secretion has been demonstrated in the mature rat [ 1,2]. Males display high amplitude pulses at regular 3- to 3.5-h intervals, separated by low trough values of GH [3]. In contrast, female rats secrete GH more continuously, with lower irregular peaks and higher basal levels between pulses [4]. The pulsatile pattern of GH secretion is the net result of an interplay between hypothalamic GH-releasing factor (GRF) and GH release-inhibiting hormone or somatostatin (SS) [5-8]. In addition, other neural and endocrine signals modulate the pulsatile release of GH [ 1,9,10]. Among these, gonadal steroids appear to play an important role in inducing the sex differences in GH secretion. The mechanism by which this sex specificity occurs remains unclear. Gonadal steroids modulate endocrine systems at both the hypothalamic and pituitary levels [ 1,11]. Testosterone and estradiol have been shown to influence the synthesis and secretion of GH, as well as the pituitary cell response to SS and GRF, by exerting their effects directly at the somatotroph level [12]. However, increasing evidence suggests that they can also modulate GH by regulating hypothalamic SS and GRF. Recent studies in rats show increased levels ofhypothalamic SS mRNA in males as compared to those in females [ 13]. In addition, testosterone has been suggested to stimulate both GRF [14] and SS [13,15] gene expression through activation of an androgen receptor-dependent pathway. Sexual and developmental differences in both GRF and SS concentrations in the median eminence and hypothalamic tissues have also been reported [ 16,17] suggesting that GRF and SS hypothalamic neurons are a target for gonadal steroids. However, the possible direct action of sex steroids on hypothalamic GRF and SS secretion has not been well established. In the present study, we assayed the effect of testosterone and 17-13estradiol on both GRF and S S by using fetal rat hypothalamic cells in culture.

Materials and Methods

Reagents Crystalline bovine serum albumin (cB SA), bacitracin, pepstatin A, testosterone and 17-13estradiol were obtained from Sigma (St. Louis, MO). Cyclic synthetic somatostatin (SSx4), Tyr1-SS14, somatostatin 28 (SS28), synthetic rGRF-(1-43)OH, hpGRF(1-44)NH2, vasoactive intestinal peptide (VIP), ovine CRF, ACTH, GnRH, 13-endorphin, substance P, neurotensin, calcitonin, pancreatic polypeptide, secretin, gastrin inhibitory peptide, insulin, glucagon, arginine vasopressin (AVP), colecystokinin (CCK) and bradykinin, were supplied by Bachem (La Jolla, CA). ~25I was purchased

137

from New England Nuclear (Boston, MA). All other reagents were obtained from commercial sources and were of the highest purity available.

Buffers and media Hank's balanced salt solution (HBSS), phosphate-buffered saline (PBS), minimum essential medium (MEM), fetal calf serum (FCS), horse serum (HS), glutamine and streptomycin-penicillin were purchased from M.A. Bioproducts (Walkersville, MD). MEM supplemented (MEMs) used for the experiments consisted of MEM with glucose (6 g/l), insulin (80mU/ml), streptomycin (0.25 lag/ml), penicillin (100U/ml), glutamine (200 mg/1), 0.1 ~o cBSA, bacitracin (30 ~tg/ml)and pepstatin A (5 lag/ml). The medium used for cultures consisted of MEMs plus 10~o FCS and 10~o HS (MEMsFCS-HS). Cell culture Timed pregnant Wistar rats were raised in our laboratory. On day 17 of fetal life, the embryos were removed from the mother by aseptic surgical procedures, and the hypothalamic regions were dissected in PBS. After rinsing three times with HBSS, the tissues were placed in MEMs-FCS-HS and dissociated mechanically by gently passing them through a Pasteur pipette and then through 20-,21- and 22-gauge needles. Dispersed cells were resuspended in a final appropriate volume of MEMs-FCS-HS and 1.5 of this cell solution was plated on 30 mm diameter plastic petri dishes (Coming, NY). Viability, evaluated by trypan blue exclusion, was 80-90~o. There were two hypothalami per dish, at a concentration of 3.106 cells/dish. Cultures were kept in a humidified atmosphere of 5 ~ CO2-95~o air at 37°C. Medium was changed every 5 days. The morphological development of the cultures was monitored by phase contrast microscopy. Experiments were performed on day 20 of culture, when cells had formed a monolayer on the plate. Experiments with testosterone (T) and estradiol (E) T and E stock solutions were prepared in 96~o ethanol and diluted in MEMs just before used. On day 20 of culture, media were removed and cells washed with PBS (5 ml/plate). Then, each culture dish was incubated with 2 ml of freshly made MEMs containing the different concentrations of T (10, 20, 40 ng/dl) or E (0.1, 1, 10 ng/dl) during 48 h. Controls were incubated with 2 ml/plate of MEMs alone. IR-SS and IR-GRF were measured by specific RIAs in both cell extracts and incubation media. Sample preparation At the end of the experiments, media and cells were harvested separately. Media were removed and acidified with 1 M HCI (150 lal/2 ml). Cells were extracted by sonication in 0.1 M HC1 (1 ml/plate). Both media and cell extracts were boiled for 5 min and centrifuged at 14000 rpm for 30 min. The supernatants were stored at -20 ° C. Just before being used, cell extracts were centrifuged again at 14,000 rpm for 30 min (to remove completely the insoluble material). Standard GRF recoveries which were processed in the same way as that of samples were: cell extracts, 79 + 2.5~o (mean + S.E.

138

of 5 samples); media extracts 78 + 2% (mean + S.E. of 5 samples). HC1 acidified samples were neutralized in the assay tube to a final pH 7.1, using a 0.05 M phosphate buffer containing 0.25 ~o cB SA, 0.025 M EDTA, 0.1 ~J0 alkali-treated casein and 0.05 o bacitracin, for the case o f G R F RIA. The buffer for the SS RIA was 0.05 M phosphate containing 0.25~o cBSA, 0.025 M EDTA. Both SS and G R F concentrations were assayed in two different assays, one for cell extracts and the other for media extracts. Each standard curve, including blanks, was run on the same amounts of HCl as the samples: 100 pl of 0.1 M HC1 for cell extracts and 200 pl of 0.075 M HC1 for media extracts (final volumes were: 700 pl for G R F RIA and 900 pl for SS RIA, in the assay tube). Radioimmunoassay of SS IR-SS was quantified by RIA [ 18] using an antiserum against SS14 , raised by immunizing rabbits with S S 14 bound to bovine thyroglobulin with glutaraldehyde. Crossreactivity with SSe8 was 30~o. No cross-reactivity was demonstrated with VIP, rGRF, insulin, pancreatic polypeptide, ACTH, calcitonin, bradykinin, AVP, oxytocin, neurotensin, substance P, GnRH, CCK, CRF or glucagon. SS was iodinated by the Chloramine T method, resulting in a specific activity of 0.3 nCi/fmol. Sensitivity was 2 pg/tube. The intra-assay and inter-assay variations were 10~'o and 15~o, respectively. Radioimmunoassay of GRF IR-GRF was measured by RIA as previously described [19], with some modifications using a more sensible antiserum. Specific r G R F antiserum was raised in rabbits immunized with synthetic rGRF-(I-43)OH coupled to cBSA with carbodiimide. Cross-reactivity was 0.04~o with hpGRF-(1-44)NH2 and 0.037~o with hpGRF-(129). No crossreactivity was demonstrated with VIP, PHI, ovine CRF, ACTH, SS14, TRH, GnRH, 13-endorphin, substance P, neurotensin, calcitonin, pancreatic polypeptide, secretin, gastrin inhibitory peptide, insulin, glucagon, AVP and CCK. r G R F was iodinated by the Chloramine T method, resulting in a specific activity of 0.5 nCi/fmol. Sensitivity was 2 pg/tube. Intra-assay variation was 4.3 ~o and interassay variation was 8.7~o Statistics All experiments were performed at least twice. In every experiment, each group was the average of five culture dishes. Data were analyzed by the Kruskal-Wallis test or the Student's t-test, when appropriate.

Results

Effect of testosterone on IR-SS and IR-GRF Fetal rat hypothalamic cells kept in culture for 20 days were exposed to MEMs containing testosterone (T, 10, 20, 40 ng/dl) for 48 h. At the end of the experiments, IR-SS and IR-GRF were measured in both cell extracts and incubation media.

139

MEDIA 2000-

1600"

i zoo~ 800400O OI O.

i

i

i

1

CONTROL

I0

20

40

600-

t,o I

CELLS

500-

eY

H

400:500200-

I00" i

i

i

CONTROL

I0

20

40

TESTOSTERONE CONCENTRATION (ng/dl)

Fig. 1. Effect of 48 h incubation with testosterone (T, 10, 20, 40 ng/dl) on media and intracellular IR-SS content in fetal rat hypothalamic cells in culture. After 20 days 'in vitro', old media were removed and cells incubated with M E M s (2 ml/plate) containing testosterone for 48 h. The control group was incubated with M E M s alone. At the end of the experiments, IR-SS was measured by RIA in aliquots of media and cell extracts. The values represent the mean + S.E. (n = 5). n.s., not significant; ** P < 0 . 0 1 , vs. control.

As shown in Fig. 1, 48 h treatment with T had an inhibitory effect on IR-SS. This action occurred up to the highest testosterone concentration tested and was more evident in the media than in the cell content. A similar pattern of inhibition was observed in two other experiments. On the contrary, 48 h exposure to testosterone was associated with an increase in I R - G R F accumulation in the hypothalamic cultures (Fig. 2). Total (medium + cell) IR-GRF: control, 368 + 16; T (10 ng/dl), 422 + 14 (P