A re-evaluation of the effects of gonadal steroids on neuronal activity in the male rat

0361-9230/93

$h.OO + i)i)

Copyright d 1993 PergamonPresst.td.

A Re-Evaluation of the Effects of Gonadal Steroids on Neuronal Activity in the Male Rat HEIKO T. JANSEN, CAROLE L. FOPIELA, GARY L. JACKSON AND GARY A. I~,~~UT~*

Received 15 July 199 1; Accepted 1 October 1992 JANSEN,

H. T.. C. L. POPIELA, G. L. JACKSON

AND G. A, IWAM~~~*

14 r~~-~~)~ll~u~j~~ t$fh@ &kfs

~?r~~}~~~~~~~ .sterrkdr

MI n~w~md acfivity in ihe male ror. BRAIN RES BULL 31( i/2) 2 17-223,

1993.-Single unit activity (SUA) was recorded from 77 cells located in the arcuate nucleus (ARC) and medial preoptic area (MPA) of anesthetized, intact male rats. Animals were adminjstered vehicle, testosterone (T; 5 or 50 19) or 17B-estradiol (E: 0.5 fig) intravenously and SUA was monitored for 8-i 2 min. T (SO pgj reduced WA in 50% of ARC units and 44% of MPA units within 2. i +-0.46 and 3.3 c 0.92 min. respectively. Inbibit~~l~ of ARC WA was more pronoun& than MPA SUA. A small percentage (9%) of ARC units were excited by T. E reduced SUA in 29% of ARC units and 27% of A/IPA units, Single doses of 5 rg T did not affect ARC activity. However, when followed within IO min by an additional dose of 5 or 50 pg T, 30% and 43% of ARC units were inhibited, respectively. Doses (10 pg) of T produced plasma T concentrations within physiological limits, although 50 pg doses produced supraphysiologicai T levels. Neither dose affected circulating LH concentrations, We conclude that physiological and supraphysiolo~cai concentrations ofT can rapidly affect SUA within the ARC. Testosterone

Estradiol

Unit activity

Arcuate nucleus

THE gonadal steroid testosterone (T) affects a variety of neural systems in the male including those involved in reproductive behavior ( 17, t 8) and luteinizing hormone-re~ea~ng hormone (LHRH) release f13,14). T is thought to influence neuroendocrine activity directly or fo~low~ng its conversion to dihydrotestosterone (DHT) or estradiol (E) (30). The active steroid(s) are taken up by cells distributed throughout the brain ( 12,32). The events following steroid receptor binding are not completely understood; however, numerous investigators have demonstrated effects of E and T on electrical activity (6,7,23,34,35). Because these responses (in both neuronal and nonneuronal cells) can occur within seconds to minutes (1,3,7) they appear to be mediated by nongenomi~ mechanisms and may, therefore, represent a separate mechanism in the cascade of steroid effects. Two regions important for male sexual behavior and LHRH pulse-generator activity, and which contain both androgen and estrogen receptors, include the medial preoptic area (MPA) and medial basal hypothalamus (MBH). Neurons that synthesize LHRH are located within the MPA (26,3 I), and approximately 70% of the LHRH terminals within the median eminence originate in the MPA (21). The arcuate nucleus (ARC} within the MBH does not contain LHRH neurons that project to the median eminence. However, substances released from cells within the ARC are thought to affect LHRH release via actions on LHRH terminal fields located in the median eminence (13,27,34). In the female rat some neurons within the ARC/MB~ (I 5,191, but not the MPA (1 S), exhibit bursting activity coin’ Requests for reprints should be addressed to

Dr. G.

Medial preoptic area

Rat

Male

cident with pulses of LH. A functional relationship between gonadal steroids, ARC activity, and LH secretion appears to exist based on observations that electrical stimulation of the ARC increases LH secretion in the presence of E (IO). Furthermore, because a percentage of units that respond rapidly to steroids, in both females and males, project to the median eminence (6.7), they may be involved in steroid feedback mechanisms (7). Less is known about electrical activity in the ARC as it relates to steroid feedback and LH retease in the male rat. Dopaminergic and peptidergic (NPY, /3-endorphin) neurons are localized within the ARC and have been implicated in the modulation of LHRH release (13,27). In addition, the demonstration that certain peptidergic and GABAergic neurons accumulate estradiol(9,22,29) makes these potential target sites for steroid feedback actions in the male. The objectives of the current studies were, therefore: af to determine if the gonadal steroids T and E produce short-term effects on unit activity in the ARC and MPA of male rats, b) to determine whether the effects of T are dose dependent, and if so, if these effects are produced by physiologica or supraphysiological blood T concentrations, and 3) if alterations in LH can be detected within minutes following T administration. METHOD

Adult gonadal-intact male Sprague-Dawley rats weighing 30~-6~ g were anesthetized with sodium thio~nt~l (35 mg/ kg, IP). The carotid artery and both externat jugular veins were

A. Iwamoto, College of Veterinary Medicine, 2001 S. Lincoln YMRSB, Urbana, fL 61801.

217

cannulated; animals were then maintained on a constant infusion of 0.5% sodium thiopental (approximately 4 mg/h). Body temperature was maintained at 37°C using a circulating water heating pad. After being placed into a stereotaxic apparatus the cranium was opened and stainless steel electrodes (Frederick Haer, 3-4 or 9- 11MQ) were inserted into the MPA and ARC according to the atlas of Paxinos and Watson (25). Single-unit activity (SUA) was recorded on magnetic tape. Blood pressure. heart rate, the output of a window discriminator. and rate/interval monitor (Frederick Haer) were recorded on a chart recorder. Test solutions were prepared in mammalian Locke-Ringer’s solution plus propylene glycol (LPG; 9: I vol/vol). Steroids (obtained from Sigma Chemical Co., St. Louis, MO) were prepared in IOOYoethanol and diluted with LPG to a final concentration of 5. 10. or 50 pg/OS ml (T), 0.5 pg/O.5 ml (E), and containing a final concentration of 0.05% ethanol. Vehicle solutions consisted of LPG containing 0.05% ethanol. All test solutions were administered intravenously as a rapid bolus. A stabilization period of 2-4 h following electrode placement preceded the initial steroid injections. Recordings were made either in the ARC (nine rats) or the MPA (six rats) following injections of vehicle and 50 pg T or 0.5 pg E. Two to four units per animal were analyzed for steroid responsiveness. Each animal received up to three steroid injections at an interval of at least 30 min. In additional rats, ARC or MPA activity was recorded for approximately 10 min following an initial 5 fig T injection, then 10 min later, a second T injection (either 5 or 50 fig) was given and SUA again recorded for IO-1 2 min. Those units exhibiting a 30% change in activity following steroid treatment were considered steroid sensitive. A 30% decrease in activity was chosen as a conservative criterion based on previous reports evaluating the effects of substances on neuronal activity (6,24). An electrolytic lesion (50 WAfor 10 s) was placed at the last recording site in an electrode track. Animals were then euthanized with an overdose of pentobarbital. The brains were removed. formalin fixed (4% with Prussian blue reagents) and then processed histologically to verify the electrode placements. To monitor blood T and LH concentrations blood samples (300 ~1) were collected from a separate group of four gonadalintact, anesthetized, male rats (500-600 g) via indwelling atrial catheters. Blood samples were taken at 10 min intervals for 60 min to determine basal LH and T concentrations and then T (10 or 50 pg/O.5 ml LPG) was injected IV. Following T injections, blood samples were taken at 1, 2, 5, 15, 30, and 60 min postinjection. Blood was centrifuged and the plasma separated and frozen until assayed for LH and T by radioimmunoassay.

tima Gold, Packard) was added to each munvial and the vials counted in a liquid scintillation counter. Testosterone binding curves for standards and various volumes of rat and ram plasma are shown in Fig. 1. Increasing volumes of ram or rat plasma displaced the binding curves in parallel to the standard curve’. Stutisticul Analysis

SUA was initially grouped into I-min bins. Units were classified as sensitive or insensitive based on a change in activity of 230% for at least five consecutive l-min periods (i.e., approximately one-half of the recording period). Mean activity (at Imin intervals) was then compared to the 95% confidence intervals (CI) for the control period (only if no significant time trends were observed during that period, see below) to confirm whether the initial 30% criterion correlated with significant changes in activity. SUA was also analyzed within and between sensitive and insensitive cells during vehicle and post-steroid periods for treatment, time, and interactive effects using two-factor analysis of variance (ANOVA) with repeated measures (2). Comparisons were considered significant if p < 0.05. Chi-square analysis was performed to determine whether the percentage of ARC and MPA units classified as sensitive or insensitive changed as a function of steroid injection number or electrode placement. RESULTS

The plasma testosterone profiles for four animals receiving 50 or IO pg T are shown in Fig. 2. T was rapidly eliminated from the plasma. Mean T levels for the 60 min sampling period were significantly higher (p < 0.01) following 50 gg injections than following 10 Kg injections (54.1 vs. 13.9 @ml, respectively). Although concentrations of testosterone returned to near preinjection levels by 60 min following the 10 pg injections (2.5 vs. 1.5 ng/ml, respectively), they were well above preinjection levels in the animals receiving 50 pg doses (12.6 vs. 1.6 ng/ml). Our criterion for steroid effects on SUA was deemed acceptable because a 30% change SUA correlated with activity which fell beyond the 95% CI for vehicle periods in all but MPA units following T. In the ARC, 11 of 22 units (50%) responded

60 70 60 -

Plasma concentrations of LH were determined using a previously validated assay (20). Ovine LH S 12 was used as the standard. The intraassay coefficient of variation was 2.3%. Plasma concentrations of testosterone were determined using a slight modification of the double antibody method described previously (28). The assay buffer used was 2.5% bovine serum albumin in phosphate buffered saline (pH 7.0). Samples (2-200 ~1)were dispensed into mini liquid scintillation vials (mini vials) and then incubated with antitestosterone antiserum (Immunocarp, Montreal, Quebec; 1:lOOO) and 3H-testosterone (Amersham; 75 Ci/mmol) for 24 h at 4°C. Charcoal-treated antirabbit gamma globulin serum (raised in rams; diluted 1: 10) was then added and the tubes incubated for an additional 24 h at 4°C. Minivials were then centrifuged at 2300 rpm for 30 min and the supernatant was discarded. Liquid scintillation cocktail (Ul-

504a30 20 10 -

01 1

.

‘.....’

“““’ 10 pg

100

“.“”

+ 1000

or pl added

FIG. 1. Testosterone standard curve and parallel displacement by increasing sample volumes of rat and ram plasma. See the Method section for assay details.

GONADAL

STEROIDS

AND

UNIT

219

ACTIVITY

60

mean

=

54.5

1

100

$

Y ._

D

C

0; -6

l/l

mean

10.6 ng/ml

=

=

17.3

-6

-4

-2

0

6O :lv

CL

t-lleCt”

1

T

v

0

insensitive

l

sensitive

ng/ml

\ 1

il

T

2

4

6

6

10

12 1

2

4

6

6

10

12

T

1

60 -

ng/ml

40

-

10 20 -

; 1

I ., 0

10

., 20

., 30

., 40 min

.,

.,

50

60

post

! ., 10

., 20

, ., 30

40

., 50

.,

*

I

-6

60

I

-6

-4

-2

0

injection

FIG. 2. Plasma testosterone profiles in four intact. anesthetized rats, receiving 50 ~g (A,B) or 10 r.qg(C.D) testosterone by rapid IV injection. Blood samples were taken via indwelling atrial catheter at 10 min intervals for 60 min prior to testosterone injection, then at I, 2, 5. 15, 30. and 60 min after testosterone injection. Mean testosterone values (ng/ml + SD) prior to injection were: (A) I .35 + 0.14. (B) 1.19 ? 0.24, (C) I .35 -t 0.04. and (D) I .X2 :t- 0.09.

to T with a decrease in SUA (ANOVA p < 0.05; Fig. 3. top panel). T-sensitive responses differed from T-insensitive responses overall (treatment, p < 0.01, treatment X time, p < 0.0001). The inhibition in T-sensitive units was observed within 2. I + 0.46 min of dosing and remained for the duration of the recording period in many cells. Two units (9%) increased their activity following T when compared to the control period (data not shown). A transient increase in SUA following vehicle administrations was observed in the T-insensitive (time effect p < 0.05) but not the T-sensitive units. T-insensitive units exhibited a significantly higher basal activity than the T-sensitive units @ < 0.02). The percentage of sensitive units increased (p < 0.05) with injection number or unit, although the percentage of insensitive units decreased (Table I). Within the MPA, 7 of I6 units (44%) exhibited a significant reduction (230%) in firing rate within 3.3 f 0.92 min following T (Fig. 3, bottom panel). The overall response to T was characterized by a moderate decrease of short duration (6.1 rt 0.7 min). T-sensitive responses differed from T-insensitive responses overall (treatment X time effect, p < 0.0001); however, no significant effect of T was revealed by ANOVA when comparing T and V periods in sensitive units. The percentage of units classified as sensitive or insensitive did not vary significantly with injection number or unit recorded (Table I).

Time

(Min)

FIG. 3. Effects of IV vehicle (V) and testosterone (T: 50 pg) on mean single-unit activity (spikes/min) within the arcuate nucleus (ARC: upper panel: n = 22) and medial preoptic area (MPA: lower panel; n = 16). *Indicates values that fell beyond the 95% Cl for vehicle period mean. tlndicates reduction by 230’7 of vehicle period mean. (?SEM)

In the ARC, two of seven (29’%) units responded to E with a decrease in SUA (Fig. 4. top panel). These two E-sensitive units were also inhibited by T. Of the five units that did not respond to E, two responded previously to T and the remainder did not respond to either steroid. In the MPA. 3 of I1 (37%) units re-

TABLE SC’MMARY OF

OF

ARC

RESPONSE

AND

MPA

TESTOSTERONE

1 CHARACTERISTI(‘S

UNITS

FOLLOWING

INJECTIONS

(50

Number Total lnjectwn

& ol’Un~ts/

Numhcr

Numhcr

or Unit

Region

Response

I

7

3

ARC

Sensitive* Insensitive Sensitive

319 619 316 316

619 319 ‘16 416

3/4t I /4t l/3 73

MPA

Insensitive

* Includes both increases and decreases in activity. i Percentage of units changed signiticantly with injection number or unit (p < 0.05).

220

!,4NSE!\ 1-:-i Al 0

insensitive

l

sensitive

Plasma concentrations of LH prior to I administration were similar among animals (mean = 3.7 f 0.8 ng/ml) and were not affected by either the low (10 pg) or high (50 fig) T doses (3.9 i 1.7 ng/ml and 3. I i 0.8 ng/ml, respectivelvl. DISCUSSION

I

-8

,

-4

-2

0

2

4

I

I

I

-6

-4

-2

0

2

4

6

6

, -6

I

-6

Time

6

6

10

12

10

12

,

(Min)

FlG. 4. Effectsof IV vehicle (V) and 17&estradiol (E; 0.5 rg) on mean @EM) single-unit activity (spikes/min) within the arcuate nucleus (ARC; upper panel; n = 9) and medial preoptic area (MPA; lower panel; n = 11). *Indicates values that fell beyond the 95% CI for vehicle period mean

sponded with a decrease in SUA following E (Fig. 4, bottom panel). Two of these three units were also inhibited by T. The magnitude of the decrease in unit activity was greater in the ARC than in the MPA (mean percent of inhibition = 67.3 + 4.5 and 45 + 5.5, respectively). None of the units recorded from the ARC (n = 17) altered their activity significantly following a single 5 pg dose of T (data not shown). However, a second T injection of either 5 or 50 pg, given IO min after the first, resulted in 30 and 43%, respectively, of cells exhibiting a decrease in SUA (ANOVA treatment effect, p -c 0.05 and p < 0.03, respectively). Within the MPA, 5 PLgT was inneffective (n = 2); however, both units were then briefly inhibited by a second 50 rg dose. The changes in unit activity following steroid administrations were not related to changes in blood pressure which often rose slightly following injections and lasted only a few seconds. Heart rate remained unaffected following vehicle or steroid injections. Nearly all T-and E-sensitive units in the MBH were located within the ARC (Fig. 5A). The majority of T- and E-insensitive units were located outside of the ARC and in or near the ventromedial hypothalamic nucleus (VMH) or dorsomedial hypothalamic nucleus (DMD). In contrast, T- and E-sensitive units in the MPA were found in the same general areas as the insensitive units (Fig. 5B).

The results of this study demonstrate that neurons within the MPA and ARC of the male rat are capable of rapidly altering their firing rates in the presence of gonadal steroids, The effects of T in the MPA appeared transient compared to ARC responses and occured only at the higher dose (50 gg). Basal activity in sensitive ARC units was significantly less than in nonresponsive units and probably reflects a different population of cells because sensitive units were located within the ARC, whereas insensitive units were located outside of the ARC. E responses were robust in both the ARC and MPA even though the percentage of responsive units was generally less than with T. Because the percentage of ARC units responding to T increased with injection number we cannot completely rule out the possibility of carryover effects of T. However, because T sensitivity correlated with proximity to the ARC, the observed responses are probably specific rather than merely a reflection of increases in plasma steroid concentrations. Furthermore, because the number of sensitive and insensitive units within the MPA did not change significantly with injection number, any carryover effects of T, even at supraphysiological plasma concentrations, appear minimal. The reason for the longer lasting effects of T in the ARC versus the MPA is not clear. Differences in the duration of steroid inhibition have been reported for MPA units in the rabbit (7) and appears to be an intrinsic property of these cells. We observed that the interval between T administration and inhibition of activity was similar in both MPA and ARC; only the duration of inhibition differed. This may reflect differences in: a) local steroid concentrations and clearance rates, b) different enzymatic activities (e.g., aromatization) (30) or c) an adaptive property of MPA units. In view of the rapid clearance of testosterone from the plasma and brain ( I l), the first possibility seems likely. Early studies of the effects of T on unit activity used very high T doses (23,35) that would be considered pharmacological. Our observation that testosterone-sensitive units were detected in the ARC following 50 pg doses or two consecutive 5 pg doses, but not following single 5 rg T injections, suggests that a minimum effective dose for T lies above 5 pg. Based on our plasma T profiles (Fig. 2) between 5 and IO pg T would produce physiological concentrations (~10 ng/ml) (4) of this steroid for at least 1 h. T is rapidly eliminated from the hypothalamus (1 I) and plasma [(I I), current study] following a single IV injection, and

is highly protein bound in blood. Therefore, microgram doses of T apparently are required to attain tissue concentrations high enough to alter electrical activity. It should be noted, however, that a single 10 wg T dose produced physiological plasma concentrations within 10 min (see Fig. 2). Our observation that two consecutive 5 pg T doses (which would produce similar, or lower blood T concentrations than a single 10 pg dose) effectively suppressed ARC activity indicates that physiological T concentrations are sufficient. Furthermore, based on our results, we would suggest that a possible explanation for the lack of effectiveness of T on septal neurons in a recent study (6) relates to the doses given. The doses used (100 ng-2 gg) would, based on our T profiles, produce T concentrations of insufficient amplitude and/ or duration to affect unit activity. Our study and that of Donevan and Ferguson (6) are similar, though, in that the effects of T

A a.20

I

! A 7.60

I

I

I

A' 8.08

PIG;. 5. Location of electrodetips tn the ARC (A) and MPA (B) from all units exammed for steroid sensitivit!. (0.m) inhibited by T. (A) stimulated bq T. (0) inhibited by E. (circled bullet) inhibited bk T but not E. (half-shaded circle) inhibited by both E and 7. (X.0) not affected by T. (circled X) unaffected by either steroid. Square symbols represent those units tested for T sensitivity (5 pg. tuice). Arc-arcuate nu. ME-median eminence. VMH-ventromedial hypothalamic nu. t)M-dorsomedial hypothalamic nu. Pe-periventricular hypothalamic nu. oxoptic chiasm.MPA-medial preoptic area. SCh-suprachiasmatic nu. MPO-medial preoptic nu. Numbers at the bottom of each block represent distance from interaural line. according to the atlas of Pavinos and Watson (25).

1

A a.70

6.20

5.86

A a.74

A 6.44

fl

222

differed markedly from those of E and may relate to pharmacokinetic differences between these two hormones. In support of this hypothesis are observations that E remains in the brain much longer than T (I 1). The inhibitory effects of E within the MPA are also consistent with earlier reports using female rats (35) and rabbits (7). However, we did not observe increases in activity following E administration. In the male rat, as in the ram for example (5), I‘ is secreted in a pulsatile fashion. In the rat, these pulses are often superimposed on baseline shifts and are associated with a decline in LH (8). Based on these observations, it has been hypothesized that the elevations in plasma T are responsible for moment-tomoment regulation of LH secretion in the rat (8). However. Clifton and Steiner (4) demonstrated that constant T infusions required 2-4 h to inhibit pituitary LH secretion and concluded that short-term feedback by T does not exist. Although our results of a lack of short-term T feedback on LH would favor Clifton and Steiner’s (4) hypothesis, that study and others have not been able to identify, nor rule out, any effects of gonadal steroids on electrical activity that may precede a change in LH secretion. In the female rat, unit activity in the ARC increases dramatically 2-9 min prior to LH pulses (15), and stimulation of this region alters LH secretion (IO). Taken together, these results, as welt as unpublished observations from our laboratory that ARC stimulation in the male rat increases LH secretion, indicates that a relationship between unit activity and LH secretion exists. We would predict, therefore, that longer infusions of T or E would result in rapid changes in ARC unit activity followed by

reductions in LH several hours later. This hypothesis remalnr to be tested. Although our experimental protocol did not allow us to dctermine whether the observed steroid effects occurred direct11 on responding units or if these units projected to the median eminence, our E-response data are consisent with observations of E-sensitivity in cells projecting to the median eminence as well as in cells that do not project there (7). Furthermore. the possibility of steroid actions on intemeurons or pulse-modulator cells that indirectly alter LHRH release is entirely consistent with the complex nature of LHRH pulse-generator circuitr). The rapid changes in unit activity following steroid injections, which we and others (1,2,6.7.25) have observed, are indicative of membrane-mediated events and would precede those occuring via nuclear receptor mechanisms. Although the genomic actions of gonadal steroids have been extensively studied, the mechanisms and physiological significance of the short-term modulation of unit activity remains unclear and awaits further study. In summary, T and E elicited rapid effects on ARC and MPA unit activity. Responses to T in the MPA were transient compared to ARC responses, although E produced robust effects in both regions. The effects of T. in particular. occured at doses producing physiological, as well as supraphysiological plasma concentrations of this steroid. ACKNOWLEDGEMENTS

We gratefully acknowledge D. Kuehl for performing the testosterone RIA. This work was supported by the National Science Foundation.

REFERENCES I. Blackmore, P. F.: Beebe, S. J.; Danforth, D. R.; Alexander. N. Pro-

2. 3.

4.

5.

6.

I.

8. 9.

10.

Il.

12.

gesterone and I7cu-hydroxyprogesterone novel stimulators of calcium influx in human sperm. J. Biol. Chem. 265: l376- 1380, 1990. Bruning, J. L.; Kintz, B. L. Computational handbook of statistics (2nd edition). Glenview: Scott Foresman and Company; 1977. Calzada, L.; Bernal, A.: Loustaunau, E. Effect of steroid hormones and capacitation on membrane potential of human spermatozoa. Arch. Androl. 21:121-128; 1988. Clifton, D. K.; Steiner, R. A. Effects of acutely increased plasma testosterone concentrations on pulsatile luteinizing hormone secretion in the male rat. Biol. Reurod. 34:624-629: 1986. D’Occhio, M. J.; Schanbacher, 6. D.; Kinder, J. E. Relationship between serum testosterone concentration and patterns of luteinizing hormone secretion in the male sheep. Endccrinology 1IO:I547- 1554; 1982. Donevan, S. D.: Ferguson, A. V. The characteristics of media1 septal neurons antidromically identified as projecting to the median eminence and their response to gonadal steroids. J. Neuroendocrinol. 2:575-581; 1990. Dufy, B.; Partouche, C.; Poulain, D.; Dufy-Barbe, L; Vincent, J. D. Effects of estrogen on the electrical activity of identified and unidentified hypothalamic units. Neuroendocrinology 22:38-47; 1976. Ellis, G. B.; Desjardins, C. Male rats secrete luteinizing hormone and testosterone episodically. Endocrinology 110:16 18- 1627: 1982. Flugge, G.; Oertel, W. H.; Wuttke, W. Evidence for estrogen-receptive GABAergic neurons in the preoptic/anterior hypothalamic area of the rat brain. Neuroendocrinology 43: 1-5; 1986. Gallo, R. V.; Osland, R. B. Electrical stimulation of the arcuate nucleus in ovariectomized rats inhibits episodic luteinizing hormone (LH) release but excites LH release after estrogen priming. Endocrinology 99:659-668; 1976. Henderson, S. B.; Ciaccio, L. A.; Kincl, F. A. Dynamics of estradiol and testosterone uptake in the brain of adult male rats. J. Steroid Biochem. 11:1601-1607; 1979. Heritage, A. S.; Stumpf, W. E.; Sar, M.; Grand, L. D. (3H)-dihydrotestosterone in catecholamine neurons of rat brain stem: Combined localization by autoradiography and formaldehyde-induced fluorescence. J. Comp. Neurol. 200~289-307; I98 I.

13. Kalra, S. P.; Kalra, P. S. Neural regulation of luteinizing hormone secretion in the rat. Endocr. Rev. 4:31 l-351; 1983. 14. Kawakami. M.: Kimura. F. Possible roles of CNS estrogen-neuron systems in ;he control of gonadotropin release. In: Stuipf, W. E.: Grant, L. D., eds. Anatomical neuroendocrinology. Basel: Karger: 1975:197-219. 15. Kawakami, M.; Uemura, T.; Hayashi, R. Electrophysiological correlates of pulsatile gonadotropin release in rats. Neuroendocrinology 35~63-67; 1982. 16. Kelly, M. J. Electrical effects of steroids in neurons. In: McKerns, K. W.; Pantic, V., eds. Hormonally active brain peptides: Structure and function. New York: Plenum Press; 1982:253-277. 17. Kendrick, K. M. Electrophysiological effects of testoterone on the medial preoptic-anterior hypothalamus of the rat. J. Endocrinol. 6: 35-42; 1983. 18. Kendrick, K. M. Effect of testosterone on medial preoptic/anterior hypothalamic neurone responses to stimulation of the lateral septum. Brain Res. 262:136-142; 1983. 19. Kimura, F.; Nishihara, M.; Hiruma, H.; Funabashi, T. Naloxone increases the frequency of the electrical activity of luteinizing hormone-releasing hormone pulse generator in long-term ovariectomized rats. Neuroendocrinology 53:97-102; 1991. 20. Matteri, R. L.; Roser, J. F.; Baldwin, D. M.; Lipovetsky, V.; PapkolT, H. Characterization of a monoclonal antibody which detects luteinizing hormone from diverse mammalian species. Dom. Anim. Endoctinol. 4:157-165; 1987. 21. Merchenthaler, I.; Setalo, G.; Csontos, C.; Petrusz, P.; Flerko, B.; Negro-Vilar, A. Combined retrograde tracing and immunocytochemical identification of luteinizing hormone-releasing hormoneand somatostatin-containing neurons projecting to the median eminence of the rat. Endocrinology 125:2812-2821; 1989. 22. Morrell, J. I.; McGinty, J. F.: Pfaff, D. W. A subset of pendorphinor dynorphin-containing neurons in the medial basal hypothalamus accumulate estradiol. Neuroendocrinology 4 1:417-426; 1985. 23. Orsini, J.-C. Hypothalamic neurons responsive to increased plasma level of testosterone in the male rat. Brain Res. 212:489-493: 1981.

GONADAL

STEROIDS

AND

UNlT

ACTIVITY

24. Orsini. J. C.: Wiser. A. K.; Himmi, T.; Boyer, A.; Pet-tin, J. Sensitivity of lateral hypothalamic neurons to glycemic level: Possible involvement of an indirect adrenergic mechanism. Brain Res. Bull. 26:473478; I99 I. 35. Paxinos G.: Watson, C. The rat brain in stereotaxic coordinates. New York: Academic Press: 1986. 16. Ronnekleiv, 0. K.: Naylor. B. R.: Bond C. T’.: Adelman, J. P. Combined immunohistochemistry for gonadotropin-releasing hormone (GnRH) and Pro-GnRH, and in situ hybridization for GnRH messenger ribonucleic acid in tat brain. Mol. Endocrinol. 3:363-371; 1989. 27. Sahu. A.: Kalra. S. P.: Crowley. W. R.: Kalra. P. S. Testosterone raises neuropeptide Y concentrations in selected hypothalamic sites and in vitro release from the medial basal hypothalamus ofcastrated male rats. Endocrinology I24:4 IO-4 14: 1989. 28. Sanford. L. M. Evidence that estrogen regulation of testosterone secretion in adult rams is mediated by both indirect (gonadotropin dependent) and direct (gonadotropin independent) means. J. Androl. 6:306-314; 1985. 79. Sar. M.; Sahu. A.: Crowley. W. R.: Kalra. S. P. Localization of neuropeptide-Y immunoreactivity in estradiol-concentrating cells in the hypothalamus. Endocrinology 127:2752-2756: 1990.

223

30. Selmanoff, M. K.; Brodkin. L. D.; Weiner, R. I.; Siiteri. P. K. Aromatization and Sn-reduction of androgens in discrete hypothalamic and limbic regions of the male and female rat. Endocrinology IO I: 841-848; 1977. 3 I. Silverman. A. J.: Jhamandas. J.; Renaud. L. P. Localization of luteinizing hormone releasing hormone (LHRH) neurons that project to the median eminence. J. Neurosci. 723 12-23 19: 1987. 32. Stumpf. W. E.: Sar. M. Anatomical distribution of estrogen. androgen. progestin. corticosteroid and thyroid hormone target sites rn the brain of mammals: Phylogeny and ontogeny. Am. Zool. IX: 435-445; 1978. 33. Weller. M. J.: Terasawa. E. Infusion of neuropeptide Y into the stalk-median eminence stimulates in vivo LHRH release in castrated rhesus monkeys. 72nd Annual Meeting of The Endocrine Society. Atlanta. GA: 1990:337. 34. Yagi. K. Changes in firing rates of single preoptic and hypothalamtc units following an intravenous administration of estrogen in the castrated female rat. Brain Res. 53:343-352; 1973. 35. Yamada. Y.. Effects of testosterone on unit activity in rat hypothalamus and septum. Brain Res. 172: 165-16X: 1979.