GABAergic and catecholaminergic innervation of mediobasal hypothalamic β-endorphin cells projecting to the medial preoptic area

0306-4522/92$5.00+ 0.00 PergamonPress Ltd © 1992IBRO

Neuroscience Vol. 51, No. 2, pp. 391-399, 1992

Printed in Great Britain

GABAERGIC A N D CATECHOLAMINERGIC INNERVATION OF MEDIOBASAL HYPOTHALAMIC fl-ENDORPHIN CELLS PROJECTING TO THE MEDIAL PREOPTIC AREA T. L. HORVATH,*F. NAFTOLIN*and C. LERANTH*~ *Department of Obstetrics and Gynecology and "~Sectionof Neurobiology, Yale University School of Medicine, New Haven, CT 06510, U.S.A. Abstract--In the absence of cellular estrogen receptors or proven direct estrogen action in the rat, it is assumed that estrogen indirectly regulates the secretory activity of the preoptic area luteinizing hormone-releasing hormone-producing cells. We have previously shown that pro-opiomelanocortin neurons in the arcuate nucleus of the rat send axons rostrally to connect with luteinizing hormonereleasing hormone neurons of the preoptic area. An experiment combining retrograde tracing and doubleimmunostaining was used to test the hypothesis that rat GABAergic and/or catecholaminergic neurons can influence luteinizing hormone-releasing hormone-producing cells via mediobasal hypothalamic fl-endorphin neurons. The retrograde tracer horseradish peroxidase was injected into the medial preoptic area; two days later, arcuate nucleus Vibratome sections were double-immunostained for fl-endorphin and glutamate decarboxylase or tyrosine hydroxylase. Light and electron microscopic analysis of these triple-labeled sections demonstrated that a population of fl-endorphin-immunoreactive neurons concentrated in the ventromedial arcuate nucleus contain retrogradely transported horseradish peroxidase granules and form synaptic contacts with glutamate decarboxylase- and tyrosine hydroxylase-immunoreactive axon terminals. The present data suggest that arcuate nucleus GABA and catecholamine fibers may influenceluteinizing hormone-releasing hormone-containing neurons via projective pro-opiomelanocortin cells.

It has been suggested that estrogen regulates the secretory activity of luteinizing hormone-releasing hormone (LHRH)-producing cells indirectly, since these cells have not been demonstrated to contain estrogen receptors or to be targets of direct estrogen action. Therefore, the importance of estrogen receptor-containing hypothalamic neurons, such as opiatecontaining cells, 3° in the control of gonadotropin secretion becomes evident. Previous studies have shown that endogenous opiates are involved in the regulation of anterior pituitary luteinizing hormone (LH) secretion. 14 Administration of opiate agonists suppresses LH release in intact and orchidectomized rats, and intraventricular infusion of the endogenous opiate fl-endorphin completely eliminates LH secretion in ovariectomized rats.15 It has been suggested that the site of opiate action lies in the vicinity of LHRH neurons) 2 In support of this fact, we found that ventromedial arcuate nucleus (AN) opiate-containing neurons of the rat innervate LHRH cells. 24 :~To whom correspondence should be addressed. ABC, avidin-biotin-peroxidase; ACTH, adrenocorticotropin hormone; AN, arcuate nucleus; DA, dopamine; DAB, diaminobenzidine; GAD, glutamate decarboxylase; HRP, horseradish peroxidase; LH, luteinizing hormone; LHRH, luteinizing hormonereleasing hormone; MPOA, medial preoptic area; PB, phosphate buffer; POMC, pro-opiomelanocortin; TH, tyrosine hydroxylase.

Abbreviations:

The neurotransmitter and neuropeptide content of the AN is extensive, including GABA, proopiomelanocortin (POMC)-derived peptides, and catecholamines, all of which are involved in the control of gonadotropin secretion. ~4'28'43 In general, pharmacological effects of GABA administered into the area of the AN are stimulatory, while the effects of catecholamines are both stimulatory and inhibitory on LHRH release. This indicates that there may be intervening inhibitory neural connections between AN, GABA and catecholamine neurons and the LHRH neurons in the preoptic area. There are other indications that at least part of the GABA and catecholamine effect on LHRH neurons is indirect. For example, Clough4 demonstrated that the stimulation of LHRH release by GABA from hypothalamic explants is blocked by fl-endorphin, and another study implicates that opiate activity in the AN is influenced by catecholamines?6 Therefore, one can speculate that the functional interaction between the arcuate nucleus GABA and catecholamine system and the inhibitory POMCergic neurons may play an important role in LH release. Light and electron microscopic immunocytochemistry was used in combination with a retrograde tracer (horseradish peroxidase, HRP) technique to elucidate possible AN, GABA and catecholamine afferent connections of projective fl-endorphin neurons. After HRP injection into the medial preoptic

391

392

T. L. HORVATH et al.

area ( M P O A ) , d o u b l e - i m m u n o s t a i n i n g for fl-endorp h i n a n d either g l u t a m a t e decarboxylase ( G A D ) or tyrosine hydroxylase (TH) was p e r f o r m e d o n A N V i b r a t o m e sections.

EXPERIMENTAL PROCEDURES

Animals Six, normal cycling adult female Sprague Dawley rats ((Charles River, Kingston, NY); 200-250 g body wt) were used in this experiment. Animals were kept under standard laboratory conditions, with tap water and regular rat chow ad libitum, 12 h light/dark cycle.

Horseradish peroxidase injection Rats were deeply anesthetized with Brevital (50 mg/kg, i.p.), and placed in a stereotaxic apparatus (Kopf Instruments). HRP (Sigma grade VI; 0.1 ml 30% in saline) was injected unilaterally into the MPOA (L: 0.3 mm; V: 8.2 mm; AP: 0.3 mm according to Paxinos and Watson 34) via a micropipette (tip diameter 15 pm)connected to a Hamilton microsyringe. The HRP was injected over a period of 20 min, and the pipette was kept in place for an additional 10 min to avoid spread along the needle track. Twenty-four hours later the animals were re-anesthetized and received a colchicine injection into the contralateral lateral ventricle (80 #g in 20/~1 saline) to enhance the perikaryal labeling of GAD and fl-endorphin neurons. 23'24

Fixation Twenty-four hours after the colchicine injection, rats were killed under ether anesthesia by transaortic perfusion with 50 ml saline followed by 250 ml of fixative. The fixative consisted of 4% paraformaldehyde and 0.2% glutaraldehyde in 0.1 M phosphate buffer (PB), pH 7.35. The brains were removed and 3-mm-thick coronal blocks were cut containing the MPOA and AN. The blocks were postfixed for an additional 2 h at 4°C in glutaraldehyde-free fixative.

Tissue preparation and immunostaining Tissue blocks were rinsed in several changes of PB and 40-pm Vibratome (Lancer) sections were prepared and rinsed 4 x 15 min in PB. A modified version of the diaminobenzidine (DAB) glucose oxidase method was used to visualize retrogradely labeled neurons and the site of HRP injection: Vibratome sections were reacted for 90 min at room temperature with a freshly prepared solution containing 15mg DAB, 60mg fl-D-glucose, 12mg ammonium chloride, and 0.12 mg glucose oxidase (Sigma Type VII, St Louis, MO) in 100 ml PB. Following the glucose oxidase reaction, sections containing the site of injection were mounted on gelatin-coated glass slides, dehydrated and coverslipped with Permount. The sections of AN were rinsed 3 × 10 min in PB, transferred into vials containing 0.5ml of 10% sucrose (in PB), and rapidly frozen by immersing the vial into liquid nitrogen. They were then thawed to room temperature and repeatedly washed in PB, followed by treatment in 1% sodium borohydride in PB for 10 min to eliminate unbound aldehydes from the tissue/7

Single- and double-immunostaining Some of the sections were immunostained for flendorphin only, while the majority of the sections were double immunostained for fl-endorphin and GAD or TH. The light, microscopic single-immunostaining for flendorphin was performed to determine the location of the retrogradely labeled fl-endorphin neurons. In this experiment, sections were pretreated with 0.1% Triton X-100 in PB for 20min. After several washes in PB, Vibratome sections were incubated for 48 h at 4°C in rabbit anti-fl-

endorphin antibody 29 diluted 1:5000 in PB containing 1% normal goat serum. Following an extensive wash in PB, sections were incubated in second antibody (biotinylated anti-rabbit IgG, 1:250 in PB, Vector Labs, Burlingame, CA) for 2 h at room temperature, then washed again in PB, and incubated for 2 h at room temperature with avidin biotin-peroxidase (1:250 in PB; ABC Kit, Vector Labs, Burlingame, CA), followed by a DAB reaction (15 mg DAB, 165pl 0.3% H202 in 25ml PB for 5-10min) to visualize the tissue bound peroxidase. Sections were then thoroughly rinsed, mounted on slides, dehydrated and coverslipped by Permount. Sections for double-immunostainingn,23 also received freeze-thaw and borohydride treatment, and were divided into two groups. Sections of Group 1 were immunostained first for TH using rabbit anti-TH 35 1 : 2000 in PB containing 1% normal goat serum for 48 h at 4°C, then with biotinylated sheep anti-rabbit IgG (1:250, for 2 h at room temperature, Vector Labs, Burlingame, CA). Sections of Group II were immunostained for GAD using sheep-anti-GAD3~ as the primary antibody (1:2000 in PB containing 1% normal rabbit serum for 48 h at 4°C) followed by an incubation with biotinylated rabbit anti-sheep IgG (1:250, for 2 h at room temperature, Vector Labs, Burlingame, CA). In both experimental groups, the ABC technique9 combined with a DAB reaction was used to visualize the TH- and GAD-immunoreactive profiles, respectively. Following several rinses in PB, sections from both groups were further immunostained for fl-endorphin as follows: (1) incubation with rabbit-anti fl-endorphin antiserum, l: 5000 in PB containing I% normal goat serum, for 48 h at 4°C; (2) reaction with gold (5 nm)-conjugated goat anti-rabbit IgG (Polysciences, Warrington, PA), 1: 10 for 48 h at 4°C. Controls were performed according to our standard protocol described previously44 to exclude the possibility of non-specific cross-reactivity between primary antisera. Only single immunolabeling was found under these experimental conditions. After immunostaining of the second tissue antigen, sections were postosmicated (0.5% OsO 4 in PB) for 30 min, dehydrated through increasing ethanol concentrations (using 1% uranyl acetate in 70% ethanol, 30 min) and flat embedded) 9 Ribbons of ultrathin sections (Reichert-Jung Ultramicrotome) of the AN were collected on Formvarcoated single slot grids, contrasted with lead citrate, and examined using a Philips CM-10 electron microscope. RESULTS

Light microscopy A light m i c r o g r a p h taken from the M P O A of a n injected rat d e m o n s t r a t e s the site o f the H R P injection (Fig. la). U p o n studying sections of the A N of HRP-injected rats single-immunostained for fl-endorphin (Fig. lb, c), three groups of labeled n e u r o n s could be seen in the ipsilateral A N : (1) n e u r o n s in the ventromedial and ventrolateral A N containing i m m u n o p e r o x i d a s e reaction product; (2) n e u r o n s in the ventromedial part of the A N containing b o t h retrogradely transp o r t e d H R P granules a n d i m m u n o p e r o x i d a s e labeling (Fig. lb, c); a n d (3) cells in the ventromedial A N filled with H R P granules w i t h o u t immunoreactivity for fl-endorphin.

Electron microscopy In the d o u b l e - i m m u n o s t a i n i n g experiments, fle n d o r p h i n - i m m u n o r e a c t i v e profiles were labeled with

GABA and catecholamine fibers terminate on fl-endorphin neurons

393

Fig. 1. Light micrographs demonstrating the site of HRP injection (a), and projective fl-endorphinimmunoreactive ceils (b and c). (a) Light micrograph taken from the MPOA shows the site of HRP injection. V, third ventricle; Oc, optic chiasm. (b),(C) Light micrographs depicted from the arcuate nucleus after fl-endorphin immunostaining. Arrowheads point to fl-endorphin-immunoreactive perikarya which contain retrogradely transported HRP granules. Scale bars = 100 #m (a); 20/lm (b,c). immunogold while immunoreactivities for TH and G A D were visualized with an immunoperoxidase reaction. The small gold particles were loosely distributed throughout the cytoplasm of the flendorphin-immunoreactive neurons (Figs 2-4). In spite of the colchicine pretreatment, we were unable to find immunogold particles in the peripheral dendritic profiles, thus restricting this study to the region of the immunostained somata and the proximal dendrites. It is important to note that the gold was not observed over the peroxidase-labeled TH profiles, in spite of the fact that the primary antisera were raised in the same species. H R P injection into the MPOA combined with double-immunostaining for fl-endorphin and either G A D or TH resulted in four types of neurons in the arcuate nucleus: (1) cells filled with retrogradely transported HRP granules and immunogold particles located in the ventromedial part of the AN; (2) immunogold-labeled fl-endorphin neurons without H R P granules which were distributed throughout the ventromedial and ventrolateral parts of the AN, apparently nonprojective to the MPOA; (3) cells containing HRP granules without immunolabeling located in the ventromedial AN; and (4) heavily immunoperoxidase-labeled neurons representing TH (Group I) or G A D (Group II) immunoreactivity without HRP granules. These neurons were observed throughout the AN. NSC 5 I / 2 ~ G

Tyrosine hydroxylase ~fl-endorphin connections In the ventromedial AN, TH-immunoreactive cell bodies, dendrites and axons were found in close proximity to fl-endorphin cells. Most of the THimmunopositive axons established synaptic connections with immunonegative neurons. However, a population of TH-immunopositive boutons was observed in synaptic contact with projective flendorphin somata (Fig. 2). Fifty retrogradely labeled fl-endorphin somata were examined, and approximately 40% of these cells received TH-immunoreactive axon terminals. However, the vast majority of boutons terminating on the opiate cells were unlabeled. The synaptic connections between the TH-immunoreactive boutons and fl-endorphin somata appeared to be symmetric contacts. In addition, consistent with our previous observation, 2° axosomatic and axodendritic synaptic connections were detected between TH-immunoreactive profiles.

Glutamate decarboxylase ~ fl-endorphin connections In ultrathin sections, immunoperoxidase-labeled G A D neurons were observed in the ventromedial AN. The vast majority of G A D cell bodies contained infolded nuclei and were synaptic targets of G A D immunopositive boutons, corresponding to earlier morphological descriptions of G A D cells in this area. 22 None of these neurons contained retrogradely

Fig. 2. Electron micrographs taken from AN of (Group I) Animals after HRP injection into the MPOA and double-immunostaining for fl-endorphin and TH. (a) TH-immunopositive bouton (A) establishes a synaptic contact on the somatic spine of a retrogradely labeled fl-endorphin-immunoreactive neuron. (b) High power magnification of the TH bouton demonstrated in a. Arrowheads point to immunogold particles representing fl-endorphin immunoreactivity. (c) High power magnification of the area indicated in a. Long arrows point to HRP granules, arrowheads label the immunogold particles. Scale bars = 1 ~m. 394

GABA and catecholamine fibers terminate on fl-endorphin neurons

Fig. 3. Electron micrographs o f AN of (Group II) rats double-immunostained ff)r fl-endorphin and GAD. Panels a and b show retrogradely labeled (long arrows point to the H R P granules), fl-endorphin-immunoreactivity (arrowheads point to the immunogold particles) cell bodies. Both neurons are synaptic targets of immunoperoxidase-labeled GAD-containing boutons (A). Scale bars = 1 #m.

395

396

T.L. HORVATH et al.

Fig. 4. Electron micrographs taken from arcuate nucleus after HRP injection into the MPOA and double-immunostaining for/~-endorphin and GAD (Group II). A p-endorphin-immunopositive (arrowheads point to the immunogold particles) neuron containing HRP granules (long arrows) contacting a GAD-immunoreactive axon terminal (A). By applying a 10° angle of tilt by using a goniometer, b more clearly shows the symmetric synaptic connection between this GAD bouton (A) and the projective //-endorphin cell. Scale bars = 1/~m.

GABA and catecholamine fibers terminate on fl-endorphin neurons

397

transported HRP granules. GAD-immunopositive Both norepinephrine and DA have been implicated axon terminals were frequently found in synaptic in the control of LHRH release. 14 Norepinephrine contact with projective fl-endorphin-immunoreactive exerts both stimulatory and inhibitory effects on LH neurons. These GAD terminals contacted both the release/4 For example, Gallo and Drouva 7 have somata (Figs 3 and 4) and dendrites of fl-endorphin demonstrated that the frequency of LH discharge is cells, forming symmetric synaptic connections. In decreased during intraventricular infusion of norepicontrast to the TH-fl-endorphin interaction, all nephrine in ovariectomized rats. Furthermore, elecof the examined projective fl-endorphin cells trical stimulation of the dorsal mesencephalic (N = 50) were found to be synaptic targets of GADtegmentum, a region through which norepinephrine immunoreactive boutons. In addition, the same fl- fibers projecting to the diencephalon traverse, inhibits endorphin cells were frequently observed to contact LH releaseY In addition, activation of norepiseveral GAD boutons. The number of these axon nephrine receptors with clonidine reduces LH seterminals often exceeded the number of those with no cretion in ovariectomized rats. 25 Since LHRH labeling. neurons are not direct synaptic targets of norepinephrine axons, 23while the AN contains an abundant network of norepinephrine fibers,38 one can speculate DISCUSSION that norepinephrine can influence LHRH indirectly, This study presents the first direct evidence that via projective fl-endorphin cells of the AN. AN fl-endorphin-containingcells which project to the DA is also involved in gonadotropin secretion, MPOA are postsynaptic targets of GABAergic and although its effect remains controversial. While DA catecholaminergic axon terminals. It is likely that in the MPOA has been demonstrated to stimulate these AN POMC cells are the same group of neurons LHRH release, 2 the tuberoinfundibular dopamine that we have previously reported and shown to make system is considered to play an inhibitory role in contacts with LHRH cells in the MPOAfl4 This LHRH secretion. 2 For example, Gnodde and Schuilexperiment has also confirmed previous obser- ing8 observed that a significant rise in plasma LH vations24 that at least two populations of arcuate occurs if DA release is blocked by a centrally acting nucleus neurons project to the MPOA: (1) POMC DA release-blocking agent. In contrast, plasma LH [adrenocorticotropin hormone (ACTH)/fl-endorphin] levels rapidly decline if DA receptors are stimulated cells located exclusively in the ventromedial AN; and by apomorphine. In addition, electrical stimulation of (2) an uncharacterized population of neurons located the AN suppresses LH release in ovariectomized rats, in the same area. Based on our previous finding that presumably due to increased DA release. 6 In the ACTH 24- and fl-endorphin (unpublished obser- present study, fl-endorphin-immunoreactive cells invation)-containing boutons terminate on LHRH nervated by TH boutons were retrogradely labeled, cells, one can speculate that at least a number of the whereas none of the TH-immunopositive perikarya retrogradely labeled fl-endorphin neurons which are of the AN contained retrogradely transported HRP synaptic targets of TH- and/or GAD-immuno- granules. These observations implicate an indirect reactive boutons also make synaptic connections with effect of AN DA neurons on LHRH release via AN LHRH cells. POMC cells. Consistent with this idea, Rasmussen et al. 36 have observed parallel changes in mediobasal Tyrosine hydroxylase--*fl-endorphin connections hypothalamic DA activity and mediobasal hypoAn opioid-peptidergic mechanism has been impli- thalamic POMC mRNA content after treatment with cated in the control of gonadotropin secretion. L12,13,~6 estradiol, and suggested that DA regulates hypoIt was suggested that opiate-containing neurons in- thalamic POMC-containing neurons. directly communicate with LHRH neurons by regulating the influx of catecholamines. 16 However, our Glutamate decarboxylase ~ fl-endorphin connections recent results suggest that an alternative pathway Pharmacological studies have detected stimu(catecholamine~opiate~LHRH) may also be in- latory 32'33'41'42as well as inhibitory11 GABA effects on volved in gonadotropin hormone release regulation. LH secretion. Intraventricular administration of TH-immunopositive cell bodies in the AN rep- GABA elevates plasma LH, and this effect can be resent exclusively dopaminergic neurons, since im- blocked by administration of the GABA antagonist munostaining for the enzymes responsible for bicucullin.42 On the other hand, an inhibitory GABA norepinephrine or epinephrine synthesis have failed effect on gonadotropin release has also been observed. to reveal immunoreactive perikarya in the hypothala- For example, Rettori et al. 37 have demonstrated that mus. 3'39'4°However, TH-immunoreactive axons in this application of bicucullin into the MPOA stimulates area can represent dopamine (DA), norepinephrine, pituitary LH release. Turnover and release rate studies and epinephrine-containing fibers. The possible in- of GABA in the MPOA indicate that GABAergic volvement of epinephrine in gonadotropin hormone tonus is high in diestrus and early proestrus animals)'26 release will not be discussed here, since it has been Furthermore, implantation of muscimol, a specific established that epinephrine does not influence GABA receptor stimulating drug, into the MPOA gonadotropin secretion via the arcuate nucleus. had a strong inhibitory effect on LH release. 18

T. L. HORVATH et al.

398

Fig. 5. Schematic representation of the suggested connections between the MPOA LHRH, GAD and DA neurons, and the mediobasal hypothalamic projective flendorphin (fl-EN), GAD, DA, and ascending norepinephrine (NE) systems. While GABA and DA in the medial preoptic area have a direct effect on LHRH cells, they can exercise an opposite, indirect effect in the mediobasal hypothalamus, by inhibiting ( - ) or stimulating the projective fl-endorphin neurons, which in turn decreases ( - ) the activity of the LHRH neurons. The norepinephrine innervation of fl-endorphin neurons cannot be excluded. 3V, third ventricle; f, fornix; ac, anterior commissure; oc, optic chiasm.

CONCLUSIONS O u r present finding o f G A B A i n n e r v a t i o n of projective f l - e n d o r p h i n cells m a y help explain the contradiction between the inhibitory a n d stimulatory effects of G A B A o n g o n a d o t r o p i n secretion. In the case of a stimulatory G A B A effect, G A B A m a y suppress the

opiate inhibition o f L H R H release ~5 via its contacts on the projective f l - e n d o r p h i n neurons. On the other hand, G A B A can inhibit L H R H neurons directly, since G A B A b o u t o n s have been shown to form synaptic contacts o n L H R H cells. 2~ W e propose t h a t the two forms o f G A B A control originate in two different p o p u l a t i o n s of G A B A neurons. G A B A cells which t e r m i n a t e o n f l - e n d o r p h i n neurons are p r o b ably located in or a r o u n d the A N , while G A B A neurons with direct connections o n L H R H cells occupy areas in the M P O A . This idea is supported by the d e m o n s t r a t i o n t h a t G A B A administered into the A N facilitates L H R H release, 4'1°'27'28 while administration of G A B A into the M P O A inhibits gonadotropin secretion. 11 In s u m m a r y (Fig. 5), we have identified G A B A ergic and catecholaminergic connections o n M P O A projective P O M C neurons, at least a portion of which m a y c o n t r i b u t e to the P O M C - L H R H tract. O u r findings indicate t h a t the stimulation or inhibition of L H R H by catecholamines a n d the stimulation o f L H R H by G A B A could be due to indirect effects, via the P O M C - L H R H tract. O n the other hand, the inhibition o f L H R H by G A B A in the M P O A , a n d stimulation by D A in the same area, could be due to direct inhibitory and excitatory effects of these neurotransmitters, respectively.

Acknowledgements--We

are indebted to Marya Shanabrough for her excellent technical assistance, and to Drs W. H. Oertel, L Zaborszky, and T. Joh for kindly providing the GAD, fl-endorphin, and TH antisera, respectively. This work was supported by NIH grants HD23830 (C.L.) and HD13587 (F.N.).

REFERENCES

I. Adler B. A. and Crowley W. R. (1984) Modulation of luteinizing hormone release and catecholamine activity by opiates in the female rat. Neuroendocrinology 38, 248-253. 2. Barraclough C. and Wise P. (1982) The role of catecholamines in the regulation of pituitary luteinizing hormone and follicle stimulating hormone secretion. Endocr. Rev. 3, 91 119. 3. Bjorklund L., Lindvall O. and Nobin A. (1975) Evidence of an incerto-hypothalamic dopamine neuron system in the rat. Brain Res. 89, 29-42. 4. Clough R. W. (1991) Stimulation of luteinizing hormone releasing hormone (LHRH) release from hypothalamic explants by gamma-aminobutyric acid (GABA) is blocked by beta-endorphin. Abstract, 21st Annual Meeting of the Society for Neuroscience, New Orleans, LA, p. 1364, 544.9. 5. Demling J., Fuchs E., Baumert M. and Wuttke W. (1985) Preoptic catecholamine, GABA and glutamate release in ovariectomized and ovariectomized estrogen-primed rats utilizing a push-pull canulla technique. Neuroendocrinology 41, 212 218. 6. Gallo R. V. (1978) The effects of blockade of dopamine receptors in the inhibition of episodic luteinizing hormone release during electrical stimulation of the arcuate nucleus in ovariectomized rats. Endocrinology 102, 1026-1035. 7. Gallo R. V. and Drouva S. V. (1979) Effects of intraventricular infusion of catecholamines on luteinizing hormone release in ovariectomized and ovariectomized steroid-primed rats. Neuroendocrinology 29, 149-162. 8. Gnodde H. P. and Schuiling G. A. (1976) Involvement of catecholaminergic and cholinergic mechanisms in the pulsatile release of LH in the long-term ovariectomized rat. Neuroendocrinology 20, 212-223. 9. Hsu S., Raine L. and Fanger H. (1981) The use of avidin-biotin-peroxidase complex (ABC) in immunoperoxidase techniques. A comparison between ABC and unlabeled antibody (PAP) procedures. J. Histochem. Cytochem. 29, 577-590. 10. Jarry H., Perschl A. and Wuttke W. (1988) Further evidence that preoptic anterior hypothalamic GABAergic neurons are part of the GNRH pulse and surge generator. Acta endocr. Copenh. 118, 573-579. 11. Jarry H., Sprenger J. and Wuttke W. (1986) Rates of release of GABA and catecholamines in the mediobasal hypothalamus of ovariectomized and ovariectomized estrogen-treated rats: correlation with blood prolactin levels. Neuroendocrinology 44, 422-428. 12. Kalra S. P. (1981) Neural loci involved in naloxone-induced luteinizing hormone release: effects of norepinephrine synthesis inhibitor. Endocrinology 109, 1805-1810.

GABA and catecholamine fibers terminate on ]~-endorphin neurons

399

13. Kalra S. P. and Crowley W. (1982) Epinephrine synthesis inhibitors block naloxone-induced LH release. Endocrinology 111, 1403-1405. 14. Kalra S. P. and Kalra P. S. (1983) Neural regulation ofluteinizing hormone secretion in the rat. Endocr. Rev. 4, 311 351. 15. Kalra S. P. and Kalra P. S. (1984) Opioid adrenergi~steroid connection in regulation of luteinizing hormone secretion in the rat. Neuroendocrinology 38, 418-426. 16. Kalra S. P. and Simpkins J. (1981) Evidence for noreadrenergic mediation of opioid effects on luteinizing hormone secretion. Endocrinology 109, 776~82. 17. Kosaka T , Nagatsu I., Wu J.-Y. and Hama K. (1986) Use of high concentration of glutaraldehyde for immunocytochemistry of transmitter synthesizing enzymes in the central nervous system. Neuroscience 18, 975-990. 18. Lamberts R., Vijayan E., Graf M., Mansky T. and Wuttke W. (1983) Involvement of preoptic anterior hypothalamic GABA neurons in the regulation of pituitary LH and prolactin release. Expl Brain Res. 52, 356 362. 19. Leranth C. and Feher E. (1983) Synaptology and sources of vasoactive intestinal polypeptide (VIP) and substance P (SP) containing axons of the cat celiac ganglion. An experimental electron microscopic immunohistochemical study. Neuroscience 10, 947-958. 20. Leranth C., Sakamoto H., MacLusky N. J., Shanabrough M. and Naftolin F. (1985) Intrinsic tyrosine hydroxylase (TH) immunoreactive axons synapse with TH immunopositive neurons in the rat arcuate nucleus. Brain Res. 331, 371 375. 21. Leranth C., MacLusky N. J., Sakamoto M., Shanabrough M. and Naftolin F. (1985) Glutamic acid decarboxylasecontaining axons synapse in LHRH neurons in the rat medial preoptic area. Neuroendocrinology 40, 536-539. 22. Leranth C., Sakamoto H., MacLusky N. J., Shanabrough M. and Naftolin F. (1985) Application of avidin-ferritin and peroxidase as contrasting electron-dense markers for simultaneous electron microscopic immunocytochemical labelling of glutamic acid decarboxylase and tyrosine hydroxylase in the rat arcuate nucleus. Histochemistry 82, 165 168. 23. Leranth C., MacLusky N. J., Shanabrough M. and Naftolin F. (1988) Catecholaminergic innervation of luteinizing hormone-releasing hormone and glutamic acid decarboxylase immunopositive neurons in the rat medial preoptic area. Neuroendocrinology 48, 591-602. 24. Leranth C., MacLusky N. J., Shanabrough M. and Naftolin F. (1988) Immunohistochemical evidence for synaptic connection between pro-opiomelanocortin-immunoreactive axons and LH-RH neurons in the preoptic area of the rat. Brain Res. 449, 167 176. 25. Leung P. C. K., Arendash G., Whitmoyer D. I., Gorski R. A. and Sawyer C. H. (1981) Electrical stimulation of mesencephalic noradrenergic pathway: effect on luteinizing hormone levels in blood of ovariectomized and ovariectomized, steroid-primed rats. Endocrinology 109, 72~728. 26. Mansky T., Mestres-Ventura P. and Wuttke W. (1982) Involvement of GABA in the feedback action of estadiol on gonadotropin and prolactin release: hypothalamic GABA and catecholamine turnover rates. Brain Res. 231,353-364. 27. Masotto C. and Negro-Villar A. (1986) GABA and gonadotropin secretion: evidence from in vitro studies on regulation of LHRH secretion. In GABA and Endocrine Function (eds Racagni G. and Donoso A. O.), pp. 243-250. Raven Press, New York. 28. McCann S. M. and Rettori V. (1986) Gamma-aminobutyric acid (GABA) controls anterior pituitary secretion. In GABA and Endocrine Function (eds Racagni D. and Donoso A. O.), pp. 173 190. Raven Press, New York. 29. Mezey E., Kiss J. Z., Mueller G. P., Eskay R., O'Donohue T. L. and Palkovits M. (1985) Distribution of the pro-opiomelanocortin derived peptides, adrenocorticotrope hormone, a-melanocyte-stimulating hormone and fl-endorphin (ACTH, a-MSH, fl-END) in the rat hypothalamus. Brain Res. 328, 341 347. 30. Morell J. I., McGinty J. F. and Pfaff D. W. (1985) A subset of beta-endorphin- or dynorphin-containing neurons in the medial basal hypothalamus accumulates estradiol. Neuroendocrinology 41, 417-426. 31. Oertel W. H., Schmechel D. E., Tappaz M. L. and Kopin I. J. (1981) Production of a specific antiserum to rat brain glutamic acid decarboxylase by injection of an antige~antibody complex. Neuroscience 6, 2689-2700. 32. Ondo J. G. (1974) Gamma-aminobutyric acid effects on pituitary gonadotropin secretion. Science 185, 738-739. 33. Pass K. A. and Ondo J. G. (1977) The effects of gamma-aminobutyric acid on prolactin and gonadotropin secretion in the unanesthetized rat. Endocrinology 100, 1437 1442. 34. Paxinos G. and Watson C. (1986) The Rat Brain in Stereotaxic Coordinates. Academic Press, Orlando. 35. Pickel V., Joh T. and Reis D. (1975) Immunohistochemical localization of tyrosine hydroxylase in brain by light and electron microscopy. Brain Res. 85, 295 300. 36. Rasmussen D. D., Jakubowski M., Allen D. L. and Roberts J. L. (1990) Hypothalamic POMC mRNA levels during recovery from chronic estradiol: dopaminergic mechanism? Soc. Neurosci. Abstr. 16, 1202. 37. Rettori V., Karara A., Caceres Marizzano O., Ponzio R. and McCann M. (1989) The effect of bicuculline injected in the medial preoptic area on the secretion of prolactin and LH in the rat. Gynecol. Endocr. Suppl. 1, 83. 38. Sawyer C. H. and Clifton D. K. (1980) Aminergic innervation of the hypothalamus. Fedn Proc. 39, 2889-2895. 39. Swanson L., Sawchenko P., Berod A., Hartman B., Helle K. and Vanorden D. (1981) An immunocytochemical study of the organization of catecholaminergic cells and terminal fields in the paraventricular and supraoptic nuclei of the hypothalamus. J. comp. Neurol. 196, 271-285. 40. van den Pol A., Herbst R. and Powel J. (1984) Tyrosine hydroxylase-immunoreactive neurons of the hypothalamus. A light and electron microscopic study. Neuroscience 13, 1117-1156. 41. Vijayan E. and McCann M. (1978) The effects of intraventricular injection of gamma-aminobutyric acid (GABA) on prolactin and gonadotropin release in conscious female rats. Brain Res. 155, 35-43. 42. Vijayan E. and McCann M. (1979) Blockade of dopamine (DA) receptors with pimozide and pituitary hormone release in response to intraventricular injection of gamma-aminobutyric acid (GABA) in conscious ovariectomized rats. Brain Res. 162, 69-76. 43. Zaborszky L. (1982) Afferent connections of the medial basal hypothalamus. In Advances in Anatomy, Embryology and Cell Biology (eds Hild W., Limborgh J. van, Ortmann R., Pauly J. E. and Scheibler T. H.), Vol. 69, pp. 1-I06. Springer, Heidelberg. 44. Zaborszky L., Heimer L., Eckenstein F. and Leranth C. (1986) GABAergic input to cholinergic forebrain neurons: an ultrastructural study using retrograde tracing of HRP and double immunolabeling. J. comp. Neurol. 250, 282-295. (Accepted 1 July 1992)