Studies on the role of catecholamines in the regulation of the developmental pattern of hypothalamic aromatase

STUDIES ON THE ROLE OF CATECHOLAMINS OF THE Dads PA’JXERN OF ~C

IN THE REGULATION AROMATASE

Jacob A. &nick,” Stuart A. Tobet,b Michael J. Baum,f Dennis E. Vaccaro, Kenneth J. Ryan: Susan E. L.eeman,C and Thomas 0. Foxb Dqxutment of Pathology, Brown University, Women & Infants Hospital, Providence, RI 0290.5’; The E.K. Shriver Center, 200 Trapelo Road, Waltham, MA 02~254~;Depaiment of Biology, Boston University, Boston, MA 02219; Department of Obstetrics and Gynecology, Harvard Medical School, Boston, MA 0X15? Lkpartment of Physiology, University of Massachusetts School of Medicine, Worcester. MA 01605, USAe

ABSTRACT

Experiments were conducted to study the regulation of the developmental pattern of aromatase in the forebrain of the perinatal rat. Two experimental designs were used: aromatase measured in primary cultures of fetal hypothalamic cells and in cell-free preparations of forebrain tissue excised at varying ages. In cultured cells, aromatase decreased logarithmically at a slow rate (t1/2 = 7.8 days). Norepinephrine caused a pronounced dose (4 x IWM) and time-dependent (2-6 days) drop in aromatase without affecting the levels of Sa-reductase or substance P. In isolated tissue, ~omatase activity was compared with the concentrations of norepinephrine and dopamine in the forebrain of males vs females at different perinatal ages and in discrete forebrain areas at postnatal day 4. In no case was a sex difference in catecholamines seen. An overall developmental decline in aromatase was associated with developmental increases in catecholamine levels. Acute treatment with the beta-agonist, isoproterenol, had no effect on brain aromatase activity. INTRODUCTION Androgen metabolism in specific brain regions most likely plays an integral role in the process of sexual differentiation

of the male rat brain (1,2). Aromatization,

in particular,

may be essential for the defeminizadon of behavior and gonado~opin control and important for behavioral

masculinization

(3,4). Very little is known about the regulation

developmental

pattern of the aromatase enzyme in the brain during the period of sexual

differentiation,

although it is clear that aromatase activity in the preoptic area and temporal

lobe tissues is highest during the perinatal

period in mammals

and declines

of the

during

development (4-6). Throughout development, aromatase is somewhat higher in males than

STEROIDS 50 / 4-6 1987

510

Canick et al.

in females, synthesis

presumably

of aromatase

The possible effecters Since neural pathways concentrations serotonin

are increasing

beta-adrenergic

are forming

in the brain throughout

measurable

conducted

regulatory

factors,

decreased

marker, substance

hypothalami

decreased

administering

of neonatally

subsequently

effecters

may influence

and compared

examining

we attempted

into the nuclear an indirect

which

we have

have easily

of various

potential

androgen

metabolic

and another important

concentration

of a neuronal

was measured

with levels of norepinephrine

a potent beta-agonist,

hypothalamic

that apparent

aromatization,

in the presence

to manipulate

In fact,

treated female rats

of fetal rat hypothalamus,

with the cellular

and

(13).

P. In the other model, the activity of aromatase

(-)isoproterenol,

suggested

the uptake of [3Hjestradiol

(14,1.5), were grown

were compared

In addition,

dopamine,

of neonatal rats given [3H]-testosterone,

how possible

levels

of the rat forebrain

the same regions.

the

and since the

would include these compounds.

androgenization

and the levels of aromatase

Sex-reductase,

induce

are not yet known.

development

by Raum and his colleagues

In one, cell cultures

aromatase

which

such as norepinephrine,

of a possible effect on in vivo brain aromatization

used two models.

regions

androgens,

decline of brain aromatase

(9-l l), likely candidates

To study more precisely

enzyme,

circulating

neurotransmitters

such stimulation

of homogenized

indication

of the developmental

stimulation

(12). Additionally,

of higher

in males (7,8).

of putative

a series of experiments

fraction

because

beta-adrenergic

in discrete

and dopamine stimulation

in by

to neonatal male and female rats and

aromatase activity.

EXPERIMENTAL Animals. For cell-culture experiments, pregnant Sprague-Dawley-derived rats (CD strain) were obtained on the thirteenth day of gestation and were sacrificed on the eighteenth day. For tissue experiments, adult male and virgin female Long-Evans rats were used. Both strains were purchased from Charles River Breeding Laboratories (Wilmington, MA). Cell culture. Detailed methodology for the preparation of these cultures has been described previously (16). Hypothalami were removed from pooled male and female fetuses taken at

STEROIDS

50 14-6

1987

CATECHOLAMINES

AND BRAIN AROMATASE

511

day 18 of gestation. Cultures were treated with 2 x 10-e M I-beta-arabinofuranosylcytosine at the time of plating to enhance neuronal morphology. Aromatase was determined using 19-hydroxy-[6,7-3Hlandrostenedione as substrate. Sa-Reductase was determined using [lp,2@Hftestosterone as substrate. The incubation with labeled steroids and analysis of extracts for labeled metabolites and for substance P were as described previously (15). Tissue experiments. Rats were decapitated at various ages, ranging from the eighteenth day of gestation to the eighteenth postnatal day. For all ages and for all types of tissue dissected, the basic procedure was the same as previously described (6). Tissues for aromatase assay were stored in liquid nitrogen in an enzyme-stabilizing medium (20% glycerol, 0.1 mM EDTA, 50 mM potassium phosphate, pH 7.4, with 1 mM dithiothreitol added just prior to use). Tissue samples stored frozen for several weeks exhibited no loss of aromatase activity. Tissues for catecholamine (CA) analysis were homogenized and centrifuged before storage in liquid nitrogen. CA samples were never stored for longer tham 24 h. Aromatase w. The assay for aromatase activity was exactly as described previously (6). The substrate was 19-hydroxy-[6,7-XIJandrostenedione, SA=5 Ci/mmol and the isolated product was [3H]estrone. Data were expressed as total activity (pm01 estronef animal/30 min) or specific activity (pm01 estronelmg protein/30 min). Catecholamine m. Concentrations of norepinephrine (NE) and dopamine (DA} were determined in various brain regions dissected from individual animals using high performance liquid chromatography (HPLC) with electrochemical detection. Dissected tissue was homogenized by ultrasonic disruption in 0.1 M perchloric acid to which l-3 ngf100 FL dihydroxybenzylamine had been added as an internal standard. Details of the HPLC assay are given by Menniti and Baum (17). CA and aromatase data were analyzed by 2-factor ANOVA, sex x region (repeated measure). A w comparisons between sex in each region were done by Newman-Keuls test (18). Intraventricular isoproterenol. Four-day-old rats were administered one of four treatments 3-7 h before lights on: 1) 50 pg (-)isoproterenol (ISO) in 4 PL 0.9% saline injected intraventricularly (iv) according to a method modified by Noble a d (19); 2) 4 &Lof saline injected iv; 3) 60 pg of an~osta-l,4,6,-~ene-3,17-dione (ATD) suspended in 50 PL saline injected SC; or 4) 50 FL of saiine injected SC. Three hours after injection, pups were sacrificed, the HPOAs dissected, placed in stabilizing medium, and frozen until aromatase assay. Animals injected with IS0 were more active and vocalized more than either salineor ATD-injected animals, suggesting that this dose of IS0 was active behaviorally. The data were analyzed using a 2-factor analysis of variance, sex x treatment; aposteriori comparisons were made with Newman-Keuls tests (18). RESULTS Culture experiments.

The potential of the cultures to aromatize and fja-reduce added

substrate was determined as a function of the age of the culture (Fig. 1). Both enzymes are plotted as the total qu~tity

of product formed, either estrone or DHT, over a 24-h period.

Over the first month in culture, aromatase declined logarithmically days while Scl-reductase remained essentially constant.

STEROIDS

50 /4-6

1987

with a half life of 7.8

512

Canick et al.

0.2 -

I

I

I

I

I

I

5

IO

I5

20

25

30

Age of Culture

.

(days)

Figure 1. Aromatase (PHJestrone production) and So-reductase ([XIjdihydrotestosterone production) as a function of age of the culture. For aromatase, n=2; for Sa-reductase, n=4. Arrow indicates the half-life for aromatase. Several putative neurotransmitters and 5ol-reduction.

Neurotransmitters

were examined for their ability to affect aromatization were added twice daily for a period of 4 days

beginning 1 day after plating. Control wells received the app~p~ate vehicle. The cells were then allowed to incubate for consecutive 24-h periods with each of the tritiated substrates. Dopamine (DA) and epinephrine treatment did not alter aromatase or 5a-reductase 2B-C) except at the highest concentrations;

at these concentrations

(Fig.

both enzyme activities

decreased and necrosis of the cultures was apparent. The addition of y-aminobutyric

acid

(Fig. 2D) in the same manner did not affect either enzyme nor was it toxic in the range of concentrations

used. Incubation of the cultures with norepinephrine

effect on enzyme activity at low concentrations high concentrations

(NE) (Fig. 2A) had no

(<10-e M) and was toxic to the cultures at

+10-s M). However, at an intermediate concentration

(4 x IO-6 M),

which did not appear to be toxic to the cells, a selective decrease in aromatization occurred, while So-reduction was essentially unaffected. When the results of seven experiments were

STEROIDS

50 / 4-6 1987

CATECHOLAMINES

I t I 5 \

AND BRAIN AROMATASE

3.0

3.0

2.0

2.0

I.o

I .o

513

3.0 3.0 2.0

2.0

1.0

% 01

1.0 1 Y 0

w” + 1.5 W

t: 1.5 5

z I.0 E 0 0.5

1.0 -g a 0.5

I.5 1.0 0.5

lO-6 Neurotransmitter

lci5 ( M1

Figure 2. Production of [3H]estrogen (aromatization) and [3H]DHT (k-reduction) in culture in response to different doses of: A) norepinephrine, B) dopamine, C) epinephrine, D) y-aminobutyric acid. The dotted lines in B,C, and D describe the decline in aromatase in response to norepinephrine as shown in A. For all points, n=3. averaged using

NE at 4 x IO-6 M, a decrease of less than 2% in 5a-reductase

was

recorded, while aromatization decreased by 73% rt: 11%. To characterize

the selective decline in aromatization

caused by NE, the following

experiments

were performed. Figure 3 describes a time-course

of incubation

with NE

a~in~stered

twice daily. A decline in aromatase was apparent after 2 days of incubation

STEROIDS

50 / 4-6 1987

514

Canick

et al.

0

2

length’ of

4 6 8 NE treatment (days)

Figure 3. [3H]Estrone production (aromatization) in cultures as a function of the length of treatment with norepinephrine (4 x 10-b M). Each data bar represents the average of two cultures. and was virtually complete by 4 days. No decrease was observed after a 24-h incubation, nor did NE directly inhibit aromatization when included with tritiated substrate (data not shown). Washing away the NE did not reverse a decline once it had occurred. Figure 4 describes the effect of varying NE doses and treatment regimens on aromatase, Sex-reductase, and substance P, a peptide present in hypothalamic neurons. Treatment twice daily for 4 days with 10-c M NE had no effect (Fig. 4B) while 4 x 10-h M NE caused a selective decrease in aromatase only (Fig. 4C). Treatment with 10-S M NE twice daily for 4 days caused a sharp decrease in aromatase, but also caused a decrease in Sa-reductase and to a lesser extent in substance P (Fig. 4D), indicating a generalized toxic effect in this case. Multiple doses of NE appear to be necessary since the higher dose (lo-5 M), given only once, had no effect when tested 4 days later (Fig. 4E). Experiments

were performed to determine whether NE receptors were involved in

mediating this effect. NE was added twice daily to cultures, as described previously, with or without up to equimolar concentrations phentolamine,

of the alpha-blockers,

phenoxybenzamine

or

or the beta-blocker, propranolol. In all cases, no inhibition of the NE effect

was observed. Higher concentrations of these blockers were toxic. When NE was replaced

STEROIDS

50 / 4-6 1987

CATECHOLAMINES

by the alpha- or beta-agonists,

AND BRAIN AROMATASE

515

clonidine or isoprotereno1, no decrease in aromatization or

So-reduction

was observed in the range of concentrations

Isoproterenol

appeared to lower ~omatization

at which NE was effective.

isoproterenol were toxic. 0

.; 1

at a dose of 10-5 M; higher doses of

1

Aromatase 51x -Reductase

1 L

P

I

a

b

c

d

e

Figure 4. A comparison of [Wlestrogen production (aromatization), [VIJDHT production (5a-reduction), and substance P content of cultures after treatment for 4 days with norepinephrine (NE). a) vehicle alone, twice daily; b) 10-6 M NE, twice daily; c) 4 x10-6 M NE, twice daily; d) 10-s M NE, twice daily; e) 10-s M NE, administered only once and then left untreated for the 4-day period. Tissue em.

The developmental

patterns of NE, DA, and aromatase were

compared in the hypothalamus plus preoptic area (HPOA) of male and female rats from 2 days before birth to 18 days after birth (Fig. 5). Aromatase activity was consistently higher in males than females in the HPOA at all ages examined. The developmental

pattern of

aromatase activity appeared different when the data were expressed as total activity as compared to specific activity. Specific activity decreased with age, whereas total activity, especially in males, increased into the early postnatal period (days 3 and 4), fell on day 6, and rose to another apparent peak on days 9 and 10 before falling

STEROIDS

50 /4-6

1987

to the lowest

516

Canick et al.

Figure 5. Norepinephrine (NE) and dopamine (DA) concentrations and aromatase activity in the hypothalamus + preoptic area as a function of age in male and female rats. Aromatase data are displayed as the mean of the 4 determinations and are presented as pmol estrone/mg protein/30 min (specific activity) and pmol estrone/animal/30 min (total activity). NE and DA concentrations were determined in tissues from individual animals and the data are displayed as mean f SEM. levels on day 17. In contrast, from gestational catecholamine distribution

day 20 to postnatal content

of aromatase,

male and female females(p<.Ol), (p<.Ol),

there was an age-dependent

in NE and DA content

day 17. There were no significant

nor was there

a trend

for a sex difference

sex differences across

ages.

NE, and DA within the HPOA was studied in postnatal

rats (Fig. 6). evident

increase

Males had significantly

in the preoptic

but not in the anterior

area (~~01)

or posterior

more aromatase

and in the anterior

hypothalamus.

Furthermore,

STEROIDS

in The

day 4

activity

than

temporal

lobe

the greatest

50 / 4-6 1987

CATECHOLAMINES

AROA.fA JASE

AND BRAIN AROMATASE

517

5

Figure 6. Norepinephrine (NE), dopamine (DA), and aromatase activity in brain regions of male and female rats sacrificed on postnatal day 4. The number of animals in each group is given above each bar and the data are displayed as mean z!zSEM. Some of these data were reported in Tobet a d (6). Males significantly higher than females (*p<.O5,**p<.Ol). amount of activity was seen in the anterior temporal lobe followed by the preoptic area, anterior, and then posterior hypothalamus. In contrast, there were no reliable sex differences in catecholamine content in any brain region examined. There were, however, significant regional differences in catecholamine content

(~~01).

NE content

was highest

in the preoptic

area and the anterior

hypothalamus, lower in the posterior hypothalamus, and lower still in the anterior temporal lobe. DA content was greatest in the preoptic area. The anterior hypothalamus contained STEROIDS

50 / 4-6 1987

518

Canick et al.

less DA than the posterior comparable

administration

activity

subcutaneous m

The DA content of the anterior temporal lobe was

to that found in the anterior hypothalamus.

Intraventricular aromatase

hypothalamus.

in males

administration

by approximately

males had significantly

of the

beta-agonist,

or females

isoproterenol,

on postnatal

of the aromatase

day 4 (Fig.

inhibitor ATD reduced

70% in males and females.

had no effect

Regardless

on

7). In contrast,

aromatase

of pharmacologic

activity _ treatment,

more aromatase activity than females in the HPOA.

1

ATD

ControlIS0

ATD

Figure 7. Aromatase activity in the hypothalamus + preoptic area of male and female rats sacrificed on postnatal day 4, 3 h after drug or vehicle administration. The number of animals in each group is given above each bar and the data are expressed as mean + SEM. For control and IS0 groups, aromatase activity was significantly higher in males than in females (pc.01) and significantly lower in ATD-treated rats (pc.01) than in either IS0 or control animals.

DISCUSSION The experiments brain aromatase

reported

here were designed

and to study factors

to examine

which might regulate

the developmental that process.

STEROIDS

decline

in

Brain aromatase

50 / 4-6

1987

CATECHOLAMINES

519

AND BRAIN AROMATASE

was studied in two systems: primary culture of fetal rat hypothalamic cells and cell-free preparations of h~~~~c

tissue f?om rats at different times in the perinatal period.

The culture experiments demonstrate that norepinephrine causes a dramatic decline in the aromatization

of androgen to estrogen by hypothalamic cells. Treatment of the cells with

other putative neuro~ansmitters

causes a decline in aromatization only at concen~ations

which cause necrosis of the cells. When necrosis is observed, Sa-reductase and substance P also decrease. significantly

Only NE treatment

selectively

aromatization

altering either Sa-reductase or the intracellular concentration

and without visibly altering the mo~hology

high concentrations

without

of substance P

of the cultures. Alpha- and eta-agonists

antagonists were either ineffective in the concentration

Alternatively,

decreases

(>lO-5 M), thus masking

range tested or were cytotoxic at

a possible

it is possible that chronic administration

and

receptor-mediated

effect.

of NE is toxic to cells that contain

aromatase but not to others, perhaps in the same way that kainic acid can be toxic to some neurons

but not to most non-neuronal

cells (15,20).

Chronic

in vitro

effects of

catecholamines are not well-studied phenomena. Hypothalamic aromatase activity declines during development in vivQ (4-6) (Fig. 5) and declines during the lifetime of the cultured hypothalamic aromatization

cells (Fig. 1). The decline in

observed with age of culture appears not to be due to a general decline in

culture viability, since the levels of Sa-reductase are relatively constant throughout the first month of culture (Fig. 1) and somatostatin

and glutamine

synthetase levels have been

shown to increase during this period (21,22). Whether NE is responsible physiologically for the decreases in aromatase measured in vivo and in vitro is still to be determined. The developmental

decline in aromatase appears to correlate with a gross increase in NE

and DA within the HPOA (9-l 1) (Fig. 5). However, there are no correlations between sex or regional

increases

in aromatase

and lower NE or DA levels. With regard to sex

differences in aromatase, this well-documented

effect is not dramatic, generally ranging

from 50% to 100% higher in males than in females (4-6). The higher levels in males are

STEROIDS

50 / 4-6 1987

Canick et al.

520

most

likely

attributable

catecholamines. correlate

to induction

Regional differences

with catecholamine

lobe, the region

levels (Fig. 6). Sex or regional

not alter aromatase

in aromatase

aromatase

differences

aromatase

measured,

possible that estrogen translocation In summary, the possibility

brain and correlates perinatal

stress reduces

were not examined.

injection of isoproterenol

did

that acute beta-adrenergic

in aromatase

(13). Since they also

uptake into hypothalamic

nuclei, it is

administration.

of the catecholaminergic and dopamine.

activity (23) is particularly stimulation

temporal

catecholamine

may regulate the developmental

in norepinephrine

adrenergic

the lowest

further. The decline in aromatase

with maturation

do not appear to

the anterior

turnover

by

in vitro 3 h later (Fig. 7). This negative

is affected by beta-agonist

brain aromatase

directly related to increased

estrogen

that catecholamines

with increases

exhibits

in catecholamine

may cause a decline

should be investigated

period to adulthood coincides

although

of Raum and co-workers

as part of their assay product,

brain aromatase

activity,

activity in the HPOA measured

of the hypothalamus

than modulation

within the rat forebrain

activity, intraventricular

result does not support the conclusions stimulation

(7,8) rather

levels within those regions,

with the highest

In an attempt to manipulate

by androgens

pattern of

from the perinatal pathways

within the

The recent finding that intriguing

and may be

of the hypothalamus.

ACKNOWLEDGMENTS This work was supported by U.S. Public Health Service Grants HD 21094, HD 12337, and HD 20327 (M.J.B., J.A.C., T.O.F.), by Research Career Development Award MH 00392 (M.J.B.), and by a Mental Retardation Core Grant HD 04147 (E. K. Shriver Center).

REFERENCES 1. Baum MJ (1979). Differentiation of coital behavior in mammals: A comparative analysis. NEUROSCI BIOBEHAV REV 3:265-284. 2. MacLusky NJ and Naftolin F (1981). Sexual differentiation of the central nervous system. SCIENCE 211: 1294-1303. 3. McEwen BS, Lieberburg I, Chaptal C, and Krey LC (1977). Aromatization: Important for sexual differentiation of the neonatal rat brain. HORM BEHAV 9:249-263.

STEROIDS

50 / 4-6

1987

CATECHOLAMINES

AND BRAIN AROMAJASE

521

4. Naftolin F, Ryan KJ, Davies IJ, Reddy VV, Flores F, Petro Z, Kuhn M, White RJ,

Takaoka Y, and Wolin L (1975). The formation of estrogens by central neuroendocrine tissues. RECENT PROG HORM RES 31:295-3 19. 5. George FW and Ojeda SR (1982). Changes in aromatase activity in the rat brain during embryonic, neonatal, and infantile development. ENDOCRINOLOGY 111522-529. 6. Tobet SA, Baum MJ, Tang HB, Shim JH, and Canick JA (1985). Aromatase activity in the perinatal rat forebrain: Effects of age, sex and intrauterine position. DEV BRAIN RES 23:171-178. 7. Roselli CE, Ellinwood WE, and Resko JA (1984). Regulation of brain aromatase activity in rats. E~OCRINOL~Y 114:192-200. 8. Roselli CE and Resko JA (1984). Androgens regulate brain aromatase activity in adult male rats through a receptor mechanism. ENDOCRINOLOGY 114:2183-2189. 9. Hyyppa M (1969) A histochemical study of the primary catecholamines in the hypothalamic neurons of the rat in relation to the ontogenetic and sexual differentiation. ZEITSCHR ZELLFORSCH 98550-560. 10. Hyyppa M (1971). Hypo~~~ic mono~ines and pineal dopamine during the sexual ~fferentiation of the rat brain. EXPER~E~A 27:336-337. 11. Smith GC and Simpson RW (1970). Monoamine fluorescence in the median eminence of foetal, neonatal and adult rats. ZEITSCHR ZELLFORSCH 104541-556. 12. Raum WJ and Swerdloff RS (198 1). The role of hypothalamic adrenergic receptors in preventing testosterone-induced androgenization in the female rat brain. ENDOCRINOLOGY 109:273-278. 13. Raum WJ, Marcano M, and Swerdloff RS (1984). Nuclear accumulation of estradiol derived from the aromatization of testosterone is inhibited by hypothal~ic betareceptor stimulation in the neonatal female rat. BIOL REPROD 30:388-396. 14. Canick JA, Vaccaro DE, Ryan KJ, and Leeman SE (1977). The aromatization of androgens by primary monolayer cultures of fetal rat hypothalamus. ENDOCRINOLOGY 100:250-253. 15. Canick JA, Vaccaro DE, Livingston EM, Leeman SE, Ryan KJ, and Fox TO (1986). Localization of aromatase and Set-reductase to neuronal and non-neuronal cells in the fetal rat hypo~alamus. BRAIN RES 372:277-282. 16. Vaccaro DE and Messer A (1977). Preparation of fetal rat h~oth~~ic cells in primary monolayer culture. TISSUE CULTURE ASSN MANUAL 3561-563. 17. Menniti FS and Baum MJ (1981). Differential effects of estrogen and androgen on locomotor activity induced in castrated male rats by amphetamine, a novel environment, or apomorphine. BRAIN RES 216:89-107. 18. Brunning JL and Kintz BL. Comnutational Handbook pf Statistics, Scott, Foresman, Glenview, Illinois (1977), pp. 119-121. 19. Noble EP, Wurtman RJ, and Axelrod J (1967). A simple and rapid method for injecting 3H-norepinephrine into the lateral ventricle of the rat brain. LIFE SC1 6:281291. 20. Herndon RM and Coyle JT (1977). Selective destruction of neurons by a transmitter agonist. SCIENCE 198:7 l-72. 21. Vaccaro DE, Leeman SE, and Reif-Lehrer L (1979). Glutamine synthetase activity & m and in primary cell cultures of rat hypothalamus. J NEUROCHEM 33:953-957. 22. Gamse R, Vaccaro DE, Gamse G, DiPace M, Fox TO, and Leeman SE (1980). Release of i~unoreactive somatos~tin from hypo~al~i~ cells in culture: Inhibition by raminobutyric acid. PROC NATL ACAD SC1 USA 77:5552-5556. 23. Weisz J, Brown BL, and Ward IL (1982). Maternal stress decreases steroid aromatase activity in brains of male and female rat fetuses. NEUROENDOCRINOLOGY 35:374-379.

STEROIDS

50 / 4-6 1987