Growth hormone-releasing hormone, somatostatin, galanin and β-endorphin afferents to the hypothalamic periventricular nucleus

llourmllof.

Chem~al

Neuroanatom~y

E L S E V 1E R

Journal of Chemical Neuroanatomy 8 (1994) 61-73

Growth hormone-releasing hormone, somatostatin, galanin and /3-endorphin afferents to the hypothalamic periventricular nucleus Mariann Fodor a'b, Zsolt Csaba b, Claude Kordon b, Jacques Epelbaum *b alNSERM U 159, 2 let Rue d'AlOsia, 75014 Paris, France bLaboratory of Neuromorphology, Semmelweis University Medical School, Budapest, Hungary

Accepted 14 September 1994

Abstract

A combined retrograde tracing (wheat germ agglutinin-horseradish peroxidase-gold complex)-immunohistochemical technique was used to identify the origin of growth hormone-releasing hormone (GHRH)-immunoreactive (ir), /3-endorphin-ir, galanin (GAL)-ir and somatostatin (SRIH)-ir terminals in the hypothalamic periventricular nucleus, which contains all the hypophysiotrophic SRIH-ir neurons. Retrogradely labeled cells were mostly observed ipsilaterally in the arcuate, dorsomedial (DMH), suprachiasmatic nuclei and the parvocellular part of the paraventricular nucleus. They were less abundant in the ventromedial and periventricular nuclei and in the lateral hypothalamus. The proportion of retrogradely labeled GHRH cells was greater at the outer rim of the ventromedial nucleus (10%) than in the arcuate nucleus proper (3%). In the arcuate nucleus, 14% of the SRIH-ir cells projected to the periventricular nucleus. Of the GAL-ir cells in the arcuate and the DMH 10% were double-labeled. Scattered retrogradely labeled GAL-ir cells were observed in paraventricular and perifornical nuclei and in the lateral hypothalamus, Of the /3-Endorphin-ir cells in the ventral part of the arcuate nucleus 15% were retrogradely labeled. It is concluded that: (1) There is no major direct connection between the hypophysiotropic GHRH and SRIH neurons, respectively, located in the arcuate and periventricular nucleus. (2) GHRH projections to the periventricular nucleus arise mainly from cells located at the outer rim of the ventromedial nucleus. (3) Intrahypothalamic SRIH projections to the periventricular nucleus arise from arcuate SRIH neurons located along the wall of the third ventricle. (4) GAL neurons from the DMH and the arcuate nucleus innervate to the same extent the periventricular nucleus. (5) 13-Endorphin arcuate neurons strongly innervate the periventricular nucleus.

Keywords: lmmunohistochemistry; Wheat germ agglutin-horseradish peroxidase-gold (WGA-HRP-gold); Growth hormone regulation

I. Introduction

The ultradian pulsatility of plasma growth hormone (GH) levels is based on the rhythmic secretion patterns of two antagonistic neurohormonal neuronal systems releasing somatostatin (SRIH) and GH-releasing hormone (GHRH). In rats, G H pulsatility is intrinsic to the hypothalamus since animals with a completely deafferented hypothalamus still exhibit the same pattern of secretion (Willougby et al., 1976). S R I H neurons projecting to the median eminence are located within the rostral periventricular nucleus and the parvocellular paraventricular nucleus (Ishikawa et al., 1987; Kawano * Corresponding author. 0891-0618/94/$07.00 © 1994 Elsevier Science B.V. All rights reserved SSDI 0891-0618(94)00029-S

and Daikoku, 1988) while hypophysiotropic G H R H neurons are confined to the ventrolateral part of the arcuate nucleus (Sawchenko et al., 1985; Niimi et al., 1989). Other S R I H cell bodies and fibers are scattered throughout the hypothalamic structures such as the suprachiasmatic, arcuate, ventromedial and dorsomedial nuclei (DMH) (Johansson et al., 1984), while G H R H - i r neurons are only located in the arcuate nucleus and around the rim of the ventromedial nucleus (Sawchenko et al., 1990). Much ultrastructural evidence exists for synaptic connections between G H R H - and SRIH-containing hypophysiotrophic neurons (Daikoku et al., 1988; Liposits et al., 1988; Horv/tth et al., 1989). S R I H receptors on the majority of the G H R H m R N A containing neurons in the ventrolateral arcuate nucleus

62

M, Fodor et al./J. Chem. Neuroanat. 8 (1994) 61-73

Table 1 Data on antisera used in studies Primary antiserum a

Dilution

Source

Reference

/~-Endorphin GAL GHRH SRIH

1:5000 1:2000 i:5000 1:2000

R.L. Eskay A. R6kaeus T. G6rcs J. Benoit

Mezey et al., 1985 Melander et al., 1986 Horv~ith et al., 1989 Morrison et al., 1982

aAII antiscra were rabbit polyclonal.

have been characterized (Bertherat et al., 1992). SRIH receptors were also found on the same proportion of GHRH-ir cells in that region (McCarthy et al., 1992). However, the orion of the reciprocal innervation of GHRH- and SRIH-ir fibers has not been elucidated to date. Combined retrograde tracing and immunohistochemistry revealed very few SRIH-ir neurons in the rostral periventricular nucleus after injecting with fluorogold in the mediobasal hypothalamus (Willoughby et al., 1989a). Nevertheless, a large number of physiological experiments indicate unequivocally that reciprocal interaction occurs between the two neurohormortal systems (for review, see Epelbaum, 1992). In the present study, the connections between the SRIH and GHRH neuron systems are specified, using a combination of immunohistochemistry for both peptides and retrograde tracing of wheat germ agglutininapohorseradish peroxidase conjugate coupled to 10 nm colloidal gold particles (WGA-HRP-gold complex) into the periventricular nucleus of the hypothalamus (Men6trey, 1985). In addition, the connections of two hypothalamic neuropeptider0c systems, /3-endorphin and galanin, which stimulate central GH secretion (Shaar et al., 1977; Ottlecz et al., 1988), possibly through inhibition of SRIH secretion (Drouva et al., 1981; Tanoh et al., 1993), were also investigated. 2. Materials aMI metheds

2.1. Animals Male Spragne-Dawley (CD, Charles River, France) rats (n = 30), weighing 300-350 g, were used in the experiments. They were housed in a temperature-

controlled room maintained on a 12:12 h light/dark cycle. For all procedures, animals were anesthetized with pentobarbital (6 mg/100 g B.W.; Sanofi, Libourne, France) intraperitoneally.

2.2. Retrograde tracing studies To determine the orion of the hypothalamic periventricular afferents, the retrograde tracer (wheat germ agglutinin-apohorseradish peroxidase conjugate coupled to I0 nm colloidal gold particles WGA-HRPgold complex) was used. Animal heads were ftr,ed in a stereotaxic apparatus and 0.I ml of WGA-HRP-gold complex injected unilaterally into the periventricular nucleus of the hypothalamus at different rostro~audal levels from bregma. Pressure injections were performed through a glass micropipette (tip diameter 20-40 mm) connected to a l ml Hamilton syringe. The tracer was prepared as previously described (Men~rey, 1985; Men, trey and de Pommery 1989). The extent and location of the injection sites were easily controlled by SRIH immunohistuchemistry in the periventricular nucleus. Two days later the animals were reanesthetized and 10 ml of I% cokhicine in saline injected into the lateral ventricle. Within 24-48 h after colchicine treatment, the animals were perfused transcardially trader ether anesthesia with 0.9% physiological saline followed by 300 ml fixative (4% paraformaldehyde in 0.I M phosphate buffer (PB) pH 7.4). After peffusion, the brains were removed from the skull, diencephalons were dissected out and posffixed overnight at 4°C. Tissue blocks were cryoprotected in a phosphate buffered 30% sucrose solution for 24 h at 40C. Frozen serial sections (50 mm thick) were cut on a cryostat from the entire rostro-caudal area of the hypo-

Fig. 1. Extent of WGA-HRP-gold injection site in the periventricular nucleus and location of perikarya containing retrograde tracer and SRIH-ir in the arcuate nucleus. (A) The extent of the tracer diffusion is restricted to the periventricular nucleus u delineated by the ~ y of SRIH-ir cell bodies seen on the contralateral side. Drawinp illustrate the rostro-caudal distribution of the tracer. (B) Dark field illm~mtfioo was used for the quantifw.ation of projecting neurons. ((2) Innatmoptnitive cell bodies were located by bright field illumination. Arrows indicate the correspondence between the cells labeled retrogradely with WGA-HRP-gold (white arrows) and immunoreactive for somatostatin (black arrows). Scale bars: 200/an (A), and 100 0ma (B,C). Abbreviations used in figures: AHA, anterior hypothalamic area; ARH, arcuate nucleus; DMH, dorsomedial nucleus; fx, fomix; ic, internal capsula; mt, mamillothalamic tract; or, optic tract: PVH. paraventricular nucleus: SO, supraoptic nucleus~ VMH, ventromedial nucleus: V3, third ventricle.

63

M. Fodor et al./J. Chem. Neuroanat. 8 (1994) 61-73

thalamus and divided into 4 series. Sections were collected in a 0.1 M PB to be silver-intensified before the immunohistochemistry, as described previously (Men6trey, 1985).

2.3. l m m u n o h i s t o c h e m i c a l technique

Following observation of the tracer, free-floating sections were permeabilized with 0.5% Triton X-100 overnight, and treated with 3% H202 in PB for 15 min in

A

010

0

%

9328

B

A

64

M. Fodor et al./J. Chem. Neuroanat. 8 (1994) 61-73

order to abolish the non-specific peroxidase activity. Background staining was reduced by an initial incubation in 10% normal goat serum. Sections were incubated with primary antibodies for 48 h at 4°C, then rinsed in PB and incubated in biotinylated anti-rabbit IgG (1:500 dilution) for 1 h at room temperature, followed by incubation with avidin-biotin-peroxidase complex (1:250, both of Vectastain ABC Kits, Vector Laboratories) also for t h at room temperature. Between each period of incubation, sections were washed 3 times in 0.1 M PB. The dilutions of the primary antisera, sources from which they were obtained and references indicating their specificity are given in Table I. All antibody incubations were conducted in a 0.1 M PB containing 1% normal goat serum, 0.3% Triton X100, 0.1% sodium azide. After the final wash, tissues were further rinsed in 50 mM Tris-HC1 buffer (pH 7.6),

and reacted in Tris-HCl buffer containing 0.05% 3,3'diaminobenzidine tetrahydrochloride (DAB) and 0.002% H202 in order to view the antigens. After the DAB reaction, the sections were mounted, coverslipped and examined with a light microscope (BHS, Olympus) using both bright- and dark-field optics. 2.4. Analysis o f the results

The number of immunoreactive perikarya was evaluated in 10 rats. Immunoreactive cell bodies were counted in 1 section every 150 mm along the rostrocaudal extent of the hypothalamus (mean -,- S.E.M. for GHRH-ir arcuate perikarya, 125 ± 21; for GHRH-ir outer VMH perikarya, 76 ± 6; for SRIH-ir arcuate perikarya, 148 4-16; for SRIH-ir DMH perikarya, 115 4- 17; for GAL-ir arcuate perikarya, 233 4- 20; for GAL-ir DMH perikarya, 176 4- 44; and for/~-endor-

Fig. 2. Projecting neurons-containing GHRH-ir at the outer rim of the ventromedial nucleus (A-C) and in the arcuate nucleus (D). (A) The black arrow indicates a strongly labeled GHRH neuron-co~ning WGA-HRP-lgokl deposits, while the empty arrow points to a less intensively immunosmined cell body. (B and C) Double labeled neurons displayed a great v ~ in size (note the same magnifioatioa), dendritic orientation and Ihe intensity of inununostaining, Both strong (B) and weak ((2) immunolabeling ~ found among the projecting neurons. (D) Single retrogradely label. ed (small black arrows) and GHRH-immunostained neurons (empty arrows) were numerous in the arcuate nucleus but only a small portion of the immunolaheled cells (D) seemed to project to the periventricular nucleus. Scale bars: 30/tm.

M. Fodor et al./J. Chem. Neuroanat. 8 (1994) 61-73 phin arcuate perikarya, 528 ± 59). The proportion of immunoreactive retrogradely labeled cells per total number of immunostained cell bodies was calculated. Double-labeled perikarya, i.e. immunoreactive projecting neurons, exhibited a reddish-brown cytoplasmic stain (DAB chromogen), as well as distinct round black granules (silver precipitatum). For the quantification of retrogradely labeled cell bodies and immunolabeled neurons, dark field (Fig. 1A) and bright field (Fig. I B) illuminations were applied, respectively. Thus, the cooccurence of the DAB chromogen and silver particulum was identified on the same section.

3. Results The extent and location of WGA-HRP-gold injections in the periventricular nucleus are illustrated on Fig. I A. The average number of retrogradely labeled cells, neuropeptide immunoreactive and double labeled cells at six frontal levels of the hypothalamus is summarized in Figs. 6-10. 3.1. Retrograde transport study The distribution of the retrogradely labeled cells is illustrated for the arcuate nucleus region in Fig. 1B and summarized for the whole hypothalamus in Fig. 6. Retrogradely labeled cells were mainly observed, ipsilaterally, in several nuclei of the hypothalamus. Tracer-containing perikarya were most numerous in the arcuate, DMH and suprachiasmatic nuclei, in the parvocellular part of the paraventricular nucleus and less abundant in the ventromedial and periventricular nuclei and in the lateral hypothalamus.

65

3.2. Combined immunohistochemistry and retrograde tracing study 3.2.1. Growth hormone-releasing hormone. Growth hormone-releasing hormone (GHRH)-containing neurons were found throughout the rostrocaudal extent of the arcuate nucleus where they were concentrated ventrally as well as around the ventromedial nucleus (Fig. 7). Only a small proportion of the immunolabeled cells (2.7 q- 0.6%) in the arcuate nucleus appeared retrogradely labeled (Fig. 2D). In contrast, in the outer rim of the ventromedial nucleus 9.5 ± 1.6% of the G H R H ir neurons contained WGA-HRP-gold deposits (Fig. 2A). Double labeled neurons displayed multiple shapes with irregular dendritic arborization (Fig. 2). Both strongly and weakly stained neurons were retrogradely labeled (Fig. 2B-C). 3.2.2. Somatostatin. Numerous hypothalamic nuclei contained SRIH-ir perikarya. The largest number of cells were located in the periventricular, suprachiasmatic, arcuate, D M H as well as in the parvocellular portion of the paraventricular nucleus. Fewer labeled neurons were visualized in the anterior and lateral hypothalamus, the anterior part of the ventromedial nucleus and in the perifornical region. Double labeled perikarya were most numerous (14.0 ± 2.3%) in the arcuate nucleus, exclusively in the periventricular stripe, with dendritic arborization parallel to the wall of the third ventricle (Fig. 3B). A less abundant proportion of double labeled neurons (5.5 ± 1.8%) was noted in the middle portion of the D M H (Fig. 3A). Retrogradely labeled SRIH-ir cells were occasionally ( < 4 % ) found in the anterior hypothalamus, the posterior part of the periventricular nucleus, the parvocellular portion of the

B

V3

Fig. 3. Retrogradelylabeled cells after unilateral injection of WGA-HRP-gold complex into the periventricularnucleus followed by SRIH-ir immunostaining in the DMH (A) and arcuate (B) nuclei. (A) Single immunostained neurons are indicated by the empty arrows, and a projecting neuron containingsomatostatin is indicated by a black arrow. (B) Double labeled cell in the dorsomedialpart of the arcuate nucleus with dendritic arborization oriented parallel to the wall of the third ventricle(V3). Scale bars: 30 tzm.

66

M. Fodor et al./J. Chem. Neuroanat. 8 (1994) 61-73

PVH and the ventromedial nucleus. In the suprachiasmatic nucleus retrogradely labeled cells were evenly localized in the nucleus but the medial part contained a slightly higher number of projecting neurons. In spite of the high number of retrogradely labeled immunoreactive perikarya, projecting neuronscontaining somatostatin were never observed in the suprachiasmatic nucleus (not shown). 3.2.3. Galanin. Galanin-immunopositive neurons were visualized along the whole rostrocaudal extent of the hypothalamus but with higher densities in the supraoptic, the magnocellular part of the paraventricular, the arcuate and DMH. Less immunostained perikarya were

detected in the parvocellular portion of the PVH, in the periventricular (Per) and perifornical nuclei and the lateral hypothalamus (Fig. 9). The source of the GAL-ir fibers in the Per was primarily found in the middle and posterior third of the hypothalamus (Fig. 4A), namely in the DMH (Fig. 4B) and arcuate (Fig. 4 C - D ) nuclei, since 9.7 4- 2.2% of the D M H and 10.3 ± 2.4% of the arcuate nuclei GAL-ir neurons were double labeled ipsilaterally to the injection site. In the DMH, small retrogradely labeled/GAL-ir neurons (Fig. 4B) were mainly localized in the medial half of the nucleus; both dorsally and ventrally. A large group of GAL-ir cell bodies was also seen in the ventrolateral aspects of the

Fig. 4. ProjectingneuronscontainingGAL-ir in the DMH (A,B)and arcuate (C,D) nuclei. (A) NumerousGAL and retrogradelylabeledcell bodies weref o ~ in the entire rostrocaudalextent of the DMH. (B) The enclosedarea of the sectionof (A) is shown at a highermagttificationto illustrate the occurenceof small doublelabeledneurons(arrows). (C,D) Non retrogradelylabeledGAL-ir neurons(emptyarrow)and double labeledneurons (black arrows) in the arcuate nucleus. Scale bars: 50/tin (A) and 30/~m (B-D).

67

M. Fodor et al./J. Chem. Neuroanat. 8 (1994) 61-73

F

m

")~

D

Fig. 5./~-Endorphin-ir cells projecting from the arcuate to the periventricular nucleus. (A) General view. (B-D) Higher magnificationsof inserts in (A), showing that double labeled neurons are evenly located among immunostained cell bodies in the arcuate nucleus. Black arrows point to double labeled cells. The empty arrow points to a non-retrogradely labeled jS-endorphin-containingperikarya. Scale bars: 200/~m (A) and 30 #m (B-D).

arcuate nucleus. Projecting neurons were located in the medial zone of this population (Fig. 9). Occasionally, a few double labeled cells were also visible in the contralateral arcuate nucleus. In addition, a small number of double labeled cells was found in the perifornical, periventricular, parvocellular paraventricular nuclei and in the anterior hypothalamus. 3.2.4. f3-Endorphin. In the hypothalamus B-endorphin containing neurons were restricted to the arcuate nucleus (Fig. 10). These relatively large neurons were located mostly in the ventral portion of the nucleus (Fig. 5A-D). with the highest density in its caudal portion (Fig. 10). A significant proportion 05.1 ± 2.1%) of the population was double labeled and these neurons were evenly distributed among single labeled immunoreactive cell bodies.

4. Discussion The topographical organization of the periventricular nucleus makes it difficult to investigate the origin of its afferences, since a few layers of cells are not an easy target for tract-tracing studies using conventional techniques. For this reason W G A - H R P - g o l d complex was chosen as a retrograde tracer. This complex provides the best available tool for local injections into a restricted area, since it does not easily diffuse and remains concentrated at the injection site. Previous studies concluded that direct injections of W G A - H R P - g o l d into peripheral or central nerve tracts produced little or no retrograde labeling (Men6trey and de Pommery, 1989). Another great advantage of this modification of the horseradishwheat germ agglutinin tract-tracing method is that it

68

M. Fodor et al./J. Chem. Neuroanat. 8 (1994) 61-73

A

B

C

D I ..."

°

-...--.. • •

'1 •

VMH ARH ~

%*

I

®.

• .......~.

."

:

• • °VMH

." . . •.i

.~

Fig. 6. Distributional mapping of retrogradely labeled neurons after unilateral injection of WGA-HRP-gold complex into the h y p o t h a l ~ ~ v e n tricular nucleus. (A-F) Line drawings of six coronal sections through representative levels of the hypothalamus, Each symbol f e i n t s one ~rikaryon.

provides very clear pictures after gold intensification and thus permits the identification of neuropeptideprojecting cells by immunohistochemistry at the light microscope level. The distribution of neuropeptide-containing cells in the h ~ t ~ u s as observed herein is perfectly identical with previously reported studies (Johansson et al., 1984; Khaehaturian et al., 1985; Sawchenko et al., 1985; Melander et al., 1986). Fibers originating in (or crossing through) the arcuate and ventromedial nuclei have been qualitatively shown to terminate in the hypothalamic periventricular nucleus by anterograde tracing with tritlated aminoacids autoradiography (Conrad and Pfaff, 1976; Saper et at., 1976) or degeneration techni-

ques (Z~borszky and Makara 1979). The present study identified both the origin and neuropeptide content of these periventricular afferents by coupling retrograde tract-tracing technique with irmnunohistoehemistry. The most significant finding in terms o f intrahypothalamic control of GH secretion is the very li~ted projections from the hypophysiotropic GHRH neurons, located in the ventrolateral part of the arcuate nucleus, to the periventricular nucleus. Taken together with the lack of projections of periventricular SRIHergie neurons to the arcuate nucleus (Willoughby et al., 1989a, Magoul et al., 1993) and the lack of SRIH receptors in the median eminence (Epelbaum, 1992), these findings seem to rule out a model implying a simple reciprocal

69

M. Fodor et a l . / 1 Chem. Neuroanat. 8 (1994) 61-73

growth hormone-releasing hormone

A

B

PVH

G

SO ARH

:i

I D

C

DMH

~

~

DMH :.~: •

E

I

Fig. 7. Distributional mapping of neurons with GHRH alone (closed circle) and with GHR H plus tracer (asterisk) after unilateral injection of WGAHRP-gold complex into the hypothalamic periventricular nucleus. (A-F) Line drawings of six coronal sections through representative levels of the hypothalamus. Each symbol represents one perikaryon.

loop at the perikaryal level, between the two neurohormonal systems. In addition, synaptic associations of SRIH and G H R H fibers have not been seen in the external layer of the median eminence (Daikoku et al., 1988). Thus, the numerous physiological demonstrations of reciprocal interactions between SRIH and G H R H have to be rethought of in terms of indirect neuroanatomical connections. Accordingly, this study demonstrates that the G H R H innervation of the neurohormonal SRIH neurons arises mostly from the limited population of G H R H neurons located along the rim of the ventromedial nucleus. This is in agreement with previous

findings indicating that these neurons are likely to project anteriorly and do not innervate the median eminence (Sawchenko et al., 1985). Furthermore, these non-arcuate G H R H neurons do not bear SRIH receptors in contrast to those which are located in the ventrolateral part of the arcuate nucleus (Bertherat et al., 1992). The origin of the SRIH fibers innervating the arcuate G H R H neurons seems to be intrinsic (Daikoku et al., 1988) and we have demonstrated that the intrinsic arcuate SRIH neurons located along the wall of the third ventricle do innervate the anterior periventricular nucleus. Thus a four component network has now to be

70

M. Fodor et al./J. Chem. Neuroanat. 8 (1994) 61-73

somatostatin

A

B +,

so

At4A

•

~+•

.

."

" °

'"2

C

t

~?~+~

D

"

'

OUH

I

xt

". ;" V;H* *.i, ++ IJ\ ~..

m

L _ ............

r

.~..

" ":'". . . . +"H:+'. :.~ •

.

-

Fig. 8. Distributionalmappingof neurons with SRIH alone (closedcircle)and with SRIH plus tracer (asterisk)after unilaterali ~ of WGAHRP-goldcomplex into the hypothalamicperiventricularnucleus. (A-F) Line drawingsof six coronal sections throngh representativelevelsof the hypothalamus.Each symbolrepresents one perikaryon.

taken into aca~unt to integrate the reciprocal interactions between SRIH and G H R H in the hypothalamus. Namely, median eminenc~ projecting SRtH and G H R H neurons from the periventricular and arcuate nucleus, respectively, would be connected indirectly by intrahypothalamic arcuate SRIH and outer rim ventromedial nucleus G H R H neurons. Another peptide which appears to stimulate GH secretion through intrahypothalamic mechanisms is GAL (Ottlecz et al., 1986) and direct innervation of periventricular SRIH neurons has recently been demonstrated (Liposits et al., 1993). Of the numerous GAL

containing cell groups in the hypothalamus, neurons in the arcxtate and DMH project more abuadantly to the periventricular nucleus. This was already reported for the intrahypothalamic GAL.ir inputs to the paraventricular nucleus (Levin et al., 1987). In this latter study, an additional contribution from extrahypotlmlamic GAL-containing neurons from the norudrenersic A1 and A6 or serotoninergic dorsal raphe nuclei neurons was also demonstrated. Such brain stem projections cannot be excluded at the present time in the case of the periventricular GALergic inputs. Taken together with the results of this study on

71

M. Fodor et al./J. Chem. Neuroanat. 8 (1994) 61-73

galanin

A

B

".~

..

;..

:~. " ° :~,'~ PVH

";-

so

AHA VMH"

C

D •

.4,;i • " • *

.''*

?'~

.

..:

DMH

• .....;..;' VMH

"

.'"

;"

ARH

""*

Fig. 9. Distributional mapping of neurons with GAL alone (closed circle) and with GAL plus tracer (asterisk) after unilateral injection of WGAHRP-gold complex into the hypothalamic periventricular nucleus. (A-F) Line drawings of six coronal sections through representative levels of the hypothalamus. Each symbol represents one perikaryon.

GHRH innervation, it is however rather unlikely that the GAL fibers in the periventricular nucleus, arising from the arcuate nucleus, originate from GHRH/GAL neurons. They might instead correspond to the GAL/tyrosine hydroxylase neurons, abundant in both the ventrolateral and dorsomedial part of the arcuate nucleus (Everitt et al., 1986). Finally, an important contribution to the innervation of the periventricular nucleus arises from the arcuate/3endorphin-ir cell group. Furthermore, a very dense network of/3-endorphin containing nerve endings is located in the periventricular nucleus (Khachaturian et al.,

1985) and they have been recently observed in close apposition with SRIH-containing cell bodies (Fodor et al., unpublished data). More studies at the electron microscopic level will be needed in order to determine whether they correspond to actual axosomatic synapses. Whether different POMC-related peptides, such as adrenocorticotropic hormone, ct-melanocyte stimulating hormone and /3-endorphin, can directly influence hypothalamic SRIH activity is not definitely proven. At any rate, it has recently been shown, that t~ opiate agonists injected in the vicinity of the SRIH periventricular cell group results in a sustained rise in GH secretion

72

M. Fodor et al./J. Chem. Neuroanat. 8 (1994) 61-73

~B-endorphin

A

B

0 Pv. : C

6

D

Fig. 10. Di|tributional ~ of neurons with ~8-endorphin alone (closed circle) and with 0-eudorphin plus tracer (asterisk) after ~ injection of WGA-HRP,gold complex into the hypothalamic periventricular nucleus. (A-F) Line drawings of six coronal sections throul|h ~ t i v e levels of the hypothalamus. Each symbol represents one perikaryon.

(Willoughby et al., 1989b; Chapman et al., 1993) and this is consistent with a direct inhibition of SRIH release by opiates (Drouva et al., 1981). A

~

This study was supported by INSERM. M.F. was a re~pient of an INSERM fellowship and Z.Cs. of a MAE fellowship. The authors gratefuUy acknowledge the valuable contributions of J. de Pommery, Drs. D. Me.trey and P. Krzywkowsky (INSERM U 161, Paris,

France).

itefereaees Bertherat, J., Dournaud, P., B6rod, A., Normand, E., Bloch, B., Rost6ne, W., Kordon, C. and ElYeibaum, J. (1992) Growth hormone-releasing h ~ s y n ~ g neurons are a subpopulation of somatostatin receptor-labelled cells in the rat areuate nucleus: A combined in titu hybridization and receptor light.microscopic radioautographi¢ study. Neuroem~crtnotofy 56, 25-31. Chapman, I.M., Kapoor, R. and W i l k ~ b y , J.O. (1993) Interactions between the effects of opioid, serotonin and a-2-adreneq~ agonists on growth hormone release in the male rat. tCeuroewdocrir,ology 57, 921-927.

M. Fodor et al./J. Chem. Neuroanat. 8 (1994) 61-73 Conrad, L.C.A. and Pfaff, D.W. (1976) Efferents from the medial basal forebrain and hypothalamus in the rat. II. An autoradiographic study of the anterior hypothalamus. J. Comp. Neurol. 169, 221-262. Daikoku, S.S., Hisano, H. Kawano, Chikamori-Aoyama, M., Kagotani, Y., Zhang, R. and Chihara, K. (1988) Ultrastructural evidence for neuronal regulation of growth hormone secretion. Neuroendocrinology 47, 405-415. Drouva, S.D., Epelbaum, J., Tapia-Arancibia, L., Laplante, E. and Kordon, C. (1981) Opiate receptors modulate LHRH and SRIF release from mediobasal hypothalamic neurons. Neuroendocrinology 32, 163-167. Epelbaum, J. (1992) lntrahypothalamic neurohormonal interactions in the control of growth hormone secretion. In Functional Anatomy of the Neuroendocrine Hypothalamus (eds Chadwick, D.J. and Marsh, J.), pp. 54-63. John Wiley & Sons, Chichester, England. Epelbaum, J., Moyse, E., Tannenbaum, G.S., Kordon, C. and Beaudet A. (1989) Combined autoradiographic and immunohistochemical evidence for an association of somatostatin binding sites with growth hormone-releasing factor containing nerve cell bodies in the arcuate nucleus. J. Neuroendocrinol. I, 109-115. Everitt, B.J., Meister, B., Hfkfelt, T., Melander, T., Terenius, L., Rfkaeus, A., Theodorsson-Norheim, E., Dockray, G., Edwardson, J., Cuello, C., Elde, R., Goldstein, M., Hemmings, H., Ouimet, C., Walaas, I., Greengard, P., Vale, W., Weber, E., Wu, J.-Y. and Chang K.J. (1986) The hypothalamic arcuate nucleusmedian eminence complex: lmmunohistochemistry of transmitters, peptides and DARPP-32 with special reference to coexistence in dopamine neurons. Brain. Res. Rev. 11, 97-155. Horvhth, S., Palkovits, M., G6rcs, T. and Arimura A. (1989) Electron microscopic immunocytochemical evidence for the existence of bidirectional synaptic connections between growth hormonereleasing hormone- and somatostatin-containing neurons in the hypothalamus of the rat. Brain Res. 481, 8-15. Ishikawa, K., Taniguchi, Y., Kurosumi, K., Suzuki, M. and Shinoda M. (1987) lmmunohistochemical identification of somatostatincontaining neurons projecting to the median eminence of the rat. Endocrinology 121, 94-97. Johansson, O., H6kfelt, T. and Elde, R. (1984) Immunohistochemical distribution of somatostatin-like immunoreactivity in the central system of the adult rat. Neuroscience 13, 265-339. Kawano, H., Daikoku, S. and Saito S. (1982) lmmunohistochemical studies of intrahypothalamic somatostatin containing neurons in rat. Brain Res. 242, 227-232. Kawano, H. and Daikoku S. (1988) Somatostatin-containing neuronal systems in the rat hypothalamus: Retrograde tracing and immunohistochemical studies. J. Comp. Neurol. 271,293-299. Khachaturian, H., Lewis, M.E., Tsou, K. and Watson S.J. (1985)/3endorphin, ct-MSH, ACTH, and related peptides. In Handbook of Chemical Neuroanatomy vol. 4, GABA and Neuropeptides in the CNS, part 1., (eds Bjfrklund, A. and H6kfelt, T.), pp. 216-272. Elsevier, Amsterdam. Levin, M.C., Sawchenko, P.E., Howe, P.R.C., Bloom, S.R. and Polak J.M. (1987) Organization of galanin-immunoreactive inputs to the paraventricular nucleus with special reference to their relationship to catecholaminergic afferents. J. Comp. Neurol. 261, 562-582. Liposits, Zs., Merchenthaler, I., Paull, W. and Flerko, B. (1988) Synaptic communication between somatostatinergic axons and growth hormone releasing factor (GRF) synthetizing neurons in the arcuate nucleus of the rat. Histochemistry 89, 247-252. Liposits, Zs., Merchenthaler, I., Reid, J.J. and Negro-Vilar A. (1993) Galanin-immunoreactive axons innervate somatostatinsynthesizing neurons in the anterior periventricular nucleus of the rat. Endocrinology 132, 917-923.

73

Magoul, R., Onteniente, B., Benjelloun W. and Tramu, G. (1993) Tachykinergic afferents to the rat arcuate nucleus. A combined immunohistochemical and retrograde tracing study. Peptides 14, 275-286. McCarthy, G.F., Beaudet, A. and Tannenbaum, G.S. (1992) Colocalization of somatostatin receptors and growth hormone releasing factor immunoreactivity in neurons of the rat arcuate nucleus. Neuroendocrinology 56, 18-24. Melander, T., H6kfelt, T. and R6kaeus, A. (1986) Distribution of galanin-like immunoreactivity in the rat central nervous system. ,L Comp. Neurol. 248, 475-517. Men6trey, D. (1985) Retrograde tracing of neuronal pathways with a protein gold complex. Light microscopic detection after silver intensification. Histochemistry 83, 391-395. Men6trey, D. and de Pommery J. (1989) Protein-gold complexes as neuronal markers for long-term tracing studies. Histochemistry 93, 155-159. Mezey, 1~., Kiss, J.Z., Mueller, G.P., Eskay, R.L., O'Donohue, T.L. and Palkovits, M. (1985) Distribution of the pro-opiomelanocortin derived peptides, adrenocorticotrop hormone, c~-melanocyte stimulating hormone and/~-endorphin (ACTH, ct-MSH,/3-END) in the rat hypothalamus. Brain Res. 328, 341-347. Morrison, J.H., Benoit, R., Magistretti, P.J., Ling, N. and Bloom, F.E. (1982) Immunohistochemical distribution of prosomatostatin related peptides in hippocampus. Neurosci. Lett. 34, 137-142. Niimi, M., Takahara, J., Satoh, M. and Kawanishi, K. (1989) Sites of origin of growth hormone-releasing factor-containing neurons projecting to the stalk-median eminence of the rat. Peptides 10, 605-608. Ottlecz, A., Samson, W.K. and McCann, S.M. (1986) Galanin: evidence for a hypothalamic site of action to release growth hormone. Peptides 17, 51-53. Saper, C.B., Swanson, L.W. and Cowan, W.M. (1976) The efferent connections of the ventromedial nucleus of the hypothalamus of the rat. J. Comp. Neurol. 169, 409-442. Sawchenko, P.E. and Swanson, L.W. (1990) Growth hormone releasing hormone. In Handbook of Chemical Neuroanatomy, vol. 9, Neuropeptides in the CNS, part 11. (eds. Bj6rklund, A., H6kfelt, T. and Kuhar, M.J.), pp. 131-163. Elsevier, Amsterdam. Sawchenko, P.E., Swanson, L.W., Rivier, J. and Vale, W.W. (1985) The distribution of growth hormone releasing factor (GRF) immunoreactivity in the central nervous system of the rat: An immunohistochemical study using antisera directed against rat hypothlamic GRF. J. Comp. Neurol. 237, 110-115. Shaar, C.J., Frederickson, R.C.A., Dininger, N.B. and Jackson, L, (1977) Enkephalin analogue and naloxone modulate the release of growth hormone and prolactin - - evidence for regulation by an endogenous opiate in the brain. Life Sci. 21, 853-860. Tanoh, T., Shimatsu, A., lshikawa, Y., Ihara, C., Yanaihara, N. and Imura, H. (1993) Galanin-induced growth hormone secretion in conscious rats: evidence for a possible role involvement of somatostatin. J. Neuroendocrinol. 5, 183-187. Willoughby, J.O., Terry, L.C., Brazeau, P. and Martin, J.B. (1976) Pulsatile growth hormone, prolactin and thyrotropin secretion in rats with hypothalamic deafferentation. Brain Res. 127, 137-143. Willoughby, J.O., Brogan, M. and Kapoor, R. (1989a) Hypothalamic interconnections of somatostatin and growth hormone releasing factor neurons. Neuroendocrinology 50, 584-591. Willoughby, J.O., Brogan, M. and Kapoor R. (1989b) lntrahypothalamic actions of somatostatin and growth hormone releasing factor on growth hormone secretion. Neuroendocrinology 50, 592-596. Zhborszky, L. and Makara, G. (1979) Intrahypothalamic connections: An electron microscopic study in the rat. Exp. Brain Res. 34, 201-215.