Testosterone modulates the dendritic architecture of

Brain Research 890 (2001) 78–85 www.elsevier.com / locate / bres

Research report

Testosterone modulates the dendritic architecture of arcuate neuroendocrine neurons in adult male rats Steve C. Danzer b

a,b ,1

b,c

, Nathaniel T. McMullen , Naomi E. Rance

a,b,c ,

*

a Department of Pathology, University of Arizona College of Medicine, 1501 N. Campbell Avenue, Tucson, AZ 85724, USA Department of Cell Biology and Anatomy, University of Arizona College of Medicine, 1501 N. Campbell Avenue, Tucson, AZ 85724, USA c Department of Neurology, University of Arizona College of Medicine, 1501 N. Campbell Avenue, Tucson, AZ 85724, USA

Accepted 3 October 2000

Abstract Recent studies have demonstrated that gonadectomy of adult male rats induces dendritic growth of neuroendocrine neurons in the arcuate nucleus. We have hypothesized that these changes are secondary to the loss of testosterone negative feedback. In the present study, we examined the effects of testosterone replacement on the dendritic morphology of arcuate neuroendocrine neurons in castrated rats. Rats were orchidectomized and implanted with silastic capsules designed to produce physiological levels of plasma testosterone (n59) or empty silastic capsules (n59) for 2 months. Retrograde labeling with systemically injected Fluoro-Gold, followed by intracellular injection of labeled neurons in a fixed slice preparation, were used to visualize arcuate neuroendocrine neurons. Quantitative analysis of dendritic morphology was performed using three-dimensional computer reconstruction. Serum levels of LH (luteinizing hormone) and testosterone were measured by radioimmunoassay. Treatment of castrated rats with physiological levels of testosterone significantly reduced dendritic length, volume and terminal branch number relative to the castrated rats receiving empty silastic capsules. Dendritic spine density was also greater in the testosterone-treated animals, although the total numbers of spines per dendrite was not significantly different between the two groups. In addition, testosterone replacement was effective in reducing serum LH to levels found in intact rats. These studies demonstrate that testosterone replacement suppresses the dendritic outgrowth of arcuate neuroendocrine neurons that occurs in response to castration. The parallel changes in dendritic arbor and serum LH after castration and hormone replacement suggests that the suppressive effects of testosterone are related to steroid negative feedback.  2001 Elsevier Science B.V. All rights reserved. Theme: Endocrine and autonomic regulation Topic: Hypothalamic–pituitary–gonadal regulation Keywords: Hypothalamus; Fluoro-Gold; Median eminence; Castration; Steroid feedback

1. Introduction Despite nearly a century of research, the mechanisms of steroid negative feedback on the reproductive axis are still poorly understood. The first description of the suppressive effects of gonadal hormones on the endocrine axis was made by Fichera, who observed enlargement of the cells of the anterior pituitary gland in response to castration [12]. In subsequent studies, Addison determined that the hy*Corresponding author. Tel.: 11-520-626-6099; fax: 11-520-6261027. E-mail address: [email protected] (N.E. Rance). 1 Present address: Department of Medicine, Duke University, Durham, NC 27710, USA.

pertrophy in response to gonadectomy is specific for basophils in the anterior lobe [1]. Steroid negative feedback was originally hypothesized as a reciprocal relationship between the gonads and pituitary [30], however, it was soon discovered that development of ‘castration cells’ was dependent upon the close anatomical relationship between the anterior pituitary gland and the hypothalamus [17]. The subsequent discovery of directional flow in the hypothalamic portal system [15,46] and the characterization of gonadotropin-releasing hormone (GnRH, [2,24]) firmly established the hypothalamus as the central neuroendocrine control center. The hypothalamic arcuate nucleus plays a pivotal role in the control of reproductive function and steroid negative feedback [4,8,16]. Although the rat arcuate nucleus does

0006-8993 / 01 / $ – see front matter  2001 Elsevier Science B.V. All rights reserved. PII: S0006-8993( 00 )03083-3

S.C. Danzer et al. / Brain Research 890 (2001) 78 – 85

not contain GnRH neurons, a subpopulation of neurons (arcuate neuroendocrine neurons) project to the primary capillary plexus of the hypothalamic portal system [6,19,36,43,45]. These neurons express a rich variety of classical neurotransmitters and peptides [7,11,25,26,28,32,40] that have been shown to modulate gonadotropin secretion from the anterior pituitary gland [3,10,20,21,44]. In addition, arcuate neuroendocrine neurons regulate the anterior pituitary gland secretion of prolactin [5] and growth hormone [27,38]. Our recent studies, using a combination of retrograde labeling with systemically injected Fluoro-Gold and intracellular injection of labeled neurons in a fixed-slice preparation, demonstrated that gonadectomy dramatically alters the morphology of arcuate neuroendocrine neurons in the adult male rat [9]. Orchidectomy resulted in the growth of arcuate neuroendocrine neurons as evidenced by increased somatic size, dendritic length and dendritic volume. Dendritic branch and spine numbers were also increased by castration [9]. We hypothesized that these changes in dendritic architecture were secondary to the loss of testosterone. This interpretation is problematic, however, because there are also a wide variety of protein hormones secreted by the testes that regulate gonadotropin secretion at the level of the anterior pituitary gland [22]. These protein hormones also have the potential to regulate arcuate neuroendocrine neurons because these neurons project to fenestrated capillaries beyond the blood–brain barrier. To determine if the castration-induced growth of arcuate neuroendocrine neurons was secondary to the loss of testosterone, we examined the effects of testosterone replacement on the dendritic architecture of arcuate neuroendocrine neurons in castrated animals.

2. Materials and methods

2.1. Animal treatments and tissue preparation Eighteen adult male Sprague–Dawley rats (Harlan Sprague–Dawley, Inc., Houston, TX) weighing between 200 and 300 g were maintained on a 12 h on / 12 h off light cycle. The rats were provided with food and water ad libitum. Animal protocols were approved by the University of Arizona Institutional Animal Care and Use Committee and conformed to NIH guidelines. Under metofane anesthesia, animals were castrated and subcutaneously implanted with either an empty 30 mm silastic capsule (GDX; capsule inner diameter, 1.57 mm; outer diameter, 3.18 mm; Dow Corning, Midland, Mi.; n59), or with a 30 mm silastic capsule filled with crystalline testosterone (GDX1T; Sigma, St. Louis, MO; n59). The capsules were incubated in 0.1 M phosphate buffered saline (PBS) at 378C for 2 days prior to surgical implantation. The capsules were removed and replaced with fresh capsules after thirty days. Four days before sacrifice, the animals

79

were given a single intraperitoneal injection of FluoroGold dissolved in 0.9% saline (20 mg / kg, Fluorochrome, Inc., Englewood, Co.). Two months after castration, the animals were administered an overdose of sodium pentobarbital (115 mg / kg i.p.) and perfused through the ascending aorta with 100 ml of heparinized 0.1 M PBS (1 unit / ml, 378C) followed by 800 ml of 4% paraformaldehyde in 0.1 M PBS (378C). The brains were removed and post-fixed for an additional 2–4 h in the same fixative at 48C. The brains were then coded so that all subsequent procedures could be conducted with the experimenter blind to treatment group. Serial 250 mm thick sagittal sections through the hypothalamus were cut using a tissue slicer (Stoelting Instruments, Wood Dale, Il.) and collected in 0.1 M PBS.

2.2. Intracellular injection of Fluoro-Gold labeled neurons The injection of neurons was limited to a region approximately 150–500 mm lateral to the third ventricle within the ventrolateral subdivision of the arcuate nucleus. The dye injection protocol used for these experiments has been previously described in more detail [9]. Briefly, Fluoro-Gold labeled neurons were viewed using a Zeiss ACM Fixed Stage microscope equipped with Lucifer yellow (Zeiss) and Fluoro-Gold (Omega Optical, Brattleboro, VT) filter sets and a Nikon 40X (NA50.55) longworking distance, water-immersion objective. Labeled neurons were impaled with glass microelectrodes filled with a cocktail composed of 0.5% Lucifer yellow, 1% biotinylated Lucifer yellow, and 2.5% biocytin (all obtained from Molecular Probes, Eugene, OR) in deionized H 2 O. Cells were filled with direct negative current for 3–10 min (to inject the Lucifer Yellow) followed by direct positive current for 3–10 min (to inject biocytin). After filling, the slices were post-fixed at room temperature for 45 min in 0.1% glutaraldehyde, followed by 4% paraformaldehyde overnight at 48C. Tissue slices were processed to convert the biocytin and biotinylated Lucifer yellow to a reaction product visible with brightfield microscopy. To enhance reagent penetration, slices were dehydrated and rehydrated in ETOH (50, 70, 95, 100, 95, 70, 50%), rinsed in 0.1 M PBS and then placed for 1 h in 0.1% collagenase (Boehringer Mannheim, Indianapolis, IN) in 0.1 M PBS. The tissue slices were then treated with 1% H 2 O 2 for 30–45 min to inhibit endogenous peroxidase activity. This was followed by an overnight incubation in Vector ABC Elite reagent (Vector Laboratories, Burlingame, CA) in 0.1 M PBS with 1% Triton X-100 (Sigma) at 4 8C. Finally, slices were reacted with a filtered solution of 0.05% 3,39-diaminobenzedine tetrahydrochloride (Sigma) and 0.003% H 2 O 2 in cold PBS. To avoid shrinkage artifacts induced by dehydration [14], slices were cleared and mounted in dimethyl sulfoxide (Sigma) under coverslips sealed with silicone caulk.

80


2.3. Computer morphometry Three-dimensional reconstructions of the filled neurons were made using an image-combining computer microscope (Neurolucida Software, Microbrightfield Inc., Colchester, VT) equipped with a Zeiss 633 (NA51.25) oilimmersion objective. A minimum of four arcuate neuroendocrine neurons were reconstructed per animal. To be selected for reconstruction, a neuron had to be well-labeled with reaction product and possess at least one dendrite that terminated within the tissue section. Filled neurons that met these criteria were randomly selected for reconstruction. Sections from two animals did not contain sufficient numbers of filled neurons that met these selection criteria, and therefore only somatic area was digitized in these animals. Data from incomplete dendrites was also excluded from the study. Although only measurements from complete dendrites were analyzed, some of these dendrites belonged to neurons that also possessed truncated dendrites. Neurons with longer dendrites are more likely to have dendrites that are truncated at the surface of the slice, and therefore a difference in the number of truncated dendrites per neuron could introduce a bias into the study. To access the impact of truncation, the number of truncated dendrites per neuron was determined for each group. This analysis showed no significant differences in the number of truncations per neuron between the two groups (GDX, 0.9760.03; GDX1 T, 1.0160.02). To insure that a sampling bias did not occur in the location of neurons selected for reconstruction, the anterior / posterior and dorsal / ventral location of each neuron was measured. The location of reconstructed arcuate neuroendocrine neurons from the castrated (GDX) and testosterone treated (GDX1T) groups is shown in Fig. 1. A statistical analysis (Nested ANOVA) found no significant difference in neuronal location between the two groups. Dendritic spines were defined as all protrusions that extended 5 mm or less from the dendritic shaft [34]. All dendritic processes greater than 5 mm in length were traced as branches. The computer microscope reconstructions were used to calculate soma area, number of primary dendrites, total dendritic length, dendritic volume, number and length of terminal branches per dendrite and spine number per dendritic length. Spine density was calculated by dividing the number of spines per dendrite by dendritic length. Results were analyzed using a hierarchical (nested) analysis of variance, which takes into account the variance within animals when testing for significant differences between groups [42].

2.4. Radioimmunoassay of testosterone and LH Blood samples were taken at the time of sacrifice from each rat by cardiac puncture and allowed to clot at 48C.

Fig. 1. Scatterplot of the location of the arcuate neuroendocrine neurons selected for intracellular dye filling and morphometric analysis in the gonadectomized (GDX) and gonadectomized plus-testosterone (GDX1T) rats. The location of reconstructed neurons in the sagittal plane are shown relative to the optic chiasm (x-axis) and the ventral surface of the brain ( y-axis). The location of the reconstructed neurons was not different between the two groups.

The samples were centrifuged and the serum was removed and stored at 2208C. For comparative purposes, hormone levels were also assayed in a group of five intact, agematched, adult male rats that had been injected 4 days earlier with Fluoro-Gold. Serum testosterone levels were determined using a Coat-A-Count Total Testosterone kit with tri-level human serum-based immunoassay controls (Diagnostic Products Corporation, Los Angeles, CA). The lower limit of detection of this assay was 0.04 ng / ml. Luteinizing hormone (LH) levels were determined by standard double-antibody radioimmunoassay kits generously provided by Dr. A.F. Parlow and the National Hormone and Pituitary Program of the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK). The lower limit of detection for LH was 0.02 ng / ml, and the intraassay coefficient of variation was 5.4%. The rat luteinizing hormone reference preparation (NIDDK-rLHRP-3) supplied by the NIDDK was used as the standard. For each hormone, all of the samples were measured in duplicate in a single assay.

3. Results Computer microscope reconstructions of arcuate neuroendocrine neurons from GDX and GDX1T rats are shown in Figs. 2 and 3, respectively. Similar to the results of our previous study [9], the arcuate neuroendocrine neurons from long-term castrated rats had dendrites with complicated branching patterns and growth cone-like processes. The morphological appearance of dendrites from testosterone-treated rats appeared intermediate in complexity between those of GDX and intact rats [9] with occasional


81

Fig. 2. Computer microscope reconstructions of neuroendocrine neurons from gonadectomized (GDX) rats. Neurons are all oriented so that the ventral surface of the brain is toward the bottom of the page. Spines are represented by dots. Scale bar550 mm.

complicated terminal branches and growth cone-like processes. In addition, the dendrites of testosterone treated animals were notably thinner than dendrites from castrated animals, particularly in the proximal regions. Results of the quantitative analyses of dendritic architecture are shown in Fig. 4 and Table 1. The overall dendritic length of arcuate neuroendocrine neurons from the GDX1 T animals was reduced 23% relative to the GDX animals [F(1,14)59.1; P,0.01]. The decrease in overall dendritic length was due to a reduction in branching: the number of dendritic branches was reduced 25% in the GDX1T animals [F(1,14)54.4; P,0.05] while the mean branch

length was significantly increased. The overall reduction in dendritic length and dendritic thickness in the GDX1T animals resulted in a significant loss of dendritic volume (ca. 34%) relative to the GDX rats [F (1,14)59.6; P, 0.01; Fig. 4]. Because the terminal dendritic branches were predominantly involved in the growth of arcuate neurons in response to orchidectomy [9], this parameter was also analyzed. GDX1T animals exhibited a significant reduction in the number of terminal dendritic branches relative to the GDX rats [F (1,14)54.4; P,0.05; Fig. 4]. Finally, dendritic spine density was greater in GDX1T rats

82


Fig. 3. Computer microscope reconstructions of neuroendocrine neurons from gonadectomized-plus-testosterone (GDX1T) treated rats. Neurons are all oriented so that the ventral surface of the brain is toward the bottom of the page. Spines are represented by dots. Scale bar550 mm.

[F(1,14)56.8; P,0.05] although there was no difference in the total number of spines / dendrite between the two groups (Table 1). In contrast, testosterone replacement had no effect on soma area or the number of primary dendrites.

3.1. Radioimmunoassay The mean testosterone level in intact male rats was 2.0460.19 ng / ml (range: 1.45–2.61 ng / ml). In GDX animals, serum testosterone was undetectable (less than 0.04 ng / ml). Implantation of subcutaneous capsules containing crystalline testosterone into castrated rats resulted in serum testosterone levels of 0.9960.06 ng / ml (range: 0.69–1.25 ng / ml). An analysis of variance revealed that

testosterone levels in the replaced animals were significantly different from both castrated and intact rats (Tukey’s HSD, P,0.05). The serum LH was significantly elevated in the GDX animals (23.462.26 ng / ml) compared to both the intact animals (0.4963.0 ng / ml) and the testosterone-treated group (0.6912.4 ng / ml). There was no significant difference in serum LH levels between intact and GDX1T rats.

4. Discussion In the present study, a combination of retrograde labeling with systemically injected Fluoro-Gold and in-


Fig. 4. Mean (6S.E.M.) dendritic length (top), dendritic volume (middle) and terminal branch number (bottom) of gonadectomized (GDX) and gonadectomized-plus-testosterone (GDX1T) treated rats. *, P,0.05; **, P,0.01.

tracellular injection of labeled neurons in a fixed slice preparation was used to determine the effects of testosterone replacement on the dendritic architecture of arcuate neuroendocrine neurons. Our previous studies showed that

83

orchidectomy resulted in dendritic growth of arcuate neuroendocrine neurons in adult male rats [9]. In comparison to intact controls, the neuroendocrine neurons in the orchidectomized group had significantly larger somatic profile areas and exhibited significant increases in dendrite length, dendrite volume, terminal branch number, and spines per unit length of dendrite. The increase in terminal branch number in orchidectomized animals was due to the appearance of short branches giving a striking, claw-like appearance to many of the distal dendrites [9]. In the present study, testosterone replacement in castrated animals was found to prevent many of these changes. Testosterone-treated rats displayed dendrites with significantly reduced length, volume and terminal branch number relative to castrate-untreated animals. Qualitatively, the neurons from testosterone-treated animals appeared intermediate in morphology between those of castrated and intact animals. Although the dendritic complexity of neurons from testosterone-treated animals occasionally resembled that of castrated animals, dendritic thickness, particularly in the proximal portions of the tree, was more comparable to intact animals. Quantitative analysis also revealed that testosterone only partially reversed the morphological effects of castration. For example, in our previous study, castration produced an elevation in the total number of dendritic spines [9] and testosterone replacement did not affect this parameter. The somatic hypertrophy produced by castration [9] was also unaffected by testosterone replacement. These findings may reflect the very low levels of testosterone achieved by the replacement regimen in this study. Alternatively, replacement regimens of subcutaneous capsules do not fully mimic the intact state because the circadian or ultradian rhythms of hormone secretion are not duplicated. It is noteworthy, however, that significant effects of testosterone were observed with levels below that of intact animals, indicating that the effect of the hormone is well within the physiological, rather than pharmacological, range. Although the dendritic changes were not completely reversed by the testosterone, this treatment was effective in decreasing the level of serum LH to that of intact animals. The spine density was increased in testosterone-treated animals relative to the castrates, but the total spine number was not significantly different between the two groups. Rather, the increase in spine density was due to the shorter dendritic length of neurons from the animals treated with testosterone. Therefore, in this instance, significant changes in spine density were not indicative of absolute changes in the total number of spines. This distinction is important because an increase in the total number of spines is correlated with changes in afferent input [13,39,41], whereas an increase in spine density per se may exert very different effects, such as altering synaptic integration [35]. These findings illustrate the importance of dendritic reconstruction in assessing the effects of experimental manipulations on spine number.


84

Table 1 Quantitative analysis of the effects of testosterone on arcuate neuroendocrine neurons in gonadectomized rats a

GDX (n59) GDX1T (n59)

Number of neurons

Soma area (mm 2 )

Number of primary dendrites

Number of dendritic branches

Dendritic branch length (mm)

Dendritic spines / 10 mm

Number of spines per dendrite

45 48

27467 276610

2.1460.09 2.1160.04

24.662.4 18.562.0*

24.763.5 35.265.3**

1.4660.12 1.7260.08*

71.065.5 65.264.8

a

All measures represent group means6standard error of the mean. *P,0.05, **P,0.01.

The parallel changes in dendritic arbor and serum LH after castration and hormone replacement suggests that the suppressive effects of testosterone are related to steroid negative feedback. This hypothesis is supported by the location of these neurons within the hypothalamic arcuate nucleus (a control center for reproduction) in a subpopulation of neurons projecting to the primary capillary plexus of the median eminence. More importantly, recent studies have revealed an important physiological role for hypothalamic peptides in the modulation of gonadotroph function at the level of the anterior pituitary gland. These peptides are released from hypothalamic neurons into the portal system and facilitate the effects of GnRH through direct actions, by interaction with GnRH receptors, or potentiation of GnRH self-priming [10]. Several of the peptides and neurotransmitters that have been shown to modulate anterior pituitary function in this manner are present within arcuate neuroendocrine neurons in the ventrolateral subdivision [3,10,20,21,25,37,44]. Confirmation of this hypothesis awaits the identification of the precise peptide or neurotransmitter content of the arcuate neuroendocrine neurons that undergo structural remodeling in response to gonadectomy. The ultrastructural pattern of synaptic connectivity in the rat arcuate nucleus is sexually dimorphic and can be modified by testosterone administration during the critical period for brain sexual differentiation [23]. Testosterone administration also alters the number of dendritic spines on Golgi-impregnated neurons in the rat arcuate nucleus of both males and females during the critical period [29]. In the female rat, considerable plasticity has been demonstrated in the adult as well. There are alterations in synaptic density during distinct phases of the estrous cycle [31] and a dramatic decrease in axosomatic synapses during the LH surge [33]. The decrease in putative inhibitory synapses has been postulated to be the basis of the facilitatory feedback effect (positive feedback) of estrogen on the proestrus surge of gonadotropins [31]. Consistent with this hypothesis, hormone treatments designed to induce positive feedback and synaptic remodeling in the female fail to produce both synaptic remodeling and facilitation of an LH surge in the male [18]. Our studies, however, utilizing dendritic reconstruction of neurons with identified target sites, shows that there is considerable plasticity in the arcuate nucleus of adult male rats in response to steroid withdrawal [9] or replacement

(present study). Remodeling of dendritic structures thus provides a powerful mechanism for gonadal steroids to alter the function of the arcuate nucleus in both males and females.

Acknowledgements The authors wish to thank Keri Kaeding and Sally Krajewski for careful reading of an earlier version of this manuscript. Grant Sponsor: NIH NIA; AG-09214. Steve Danzer is a recipient of a Predoctoral Fellowship from the Robert S. Flinn Biomedical Research Initiative.

References [1] W.H.F. Addison, The cell-changes in the hypophysis of the albino rat, after castration, J. Comp. Neurol. 28 (1917) 441–463. [2] M. Amoss, R. Burgus, R. Blackwell, W. Vale, R. Fellows, R. Guillemin, Purification, amino acid composition and N-terminus of the hypothalamic luteinizing hormone releasing factor (LRF) of ovine origin, Biochem. Biophys. Res. Commun. 44 (1971) 205–210. [3] R.A. Anderson, R. Mitchell, Effects of g-aminobutyric acid receptor agonists on the secretion of growth hormone, luteinizing hormone, adrenocorticotrophic hormone and thyroid-stimulating hormone from the rat pituitary gland in vitro, J. Endocrinol. 108 (1986) 1–8. [4] C.A. Barraclough, Sex steroid regulation of reproductive neuroendocrine processes, in: R.O. Greep, E.B. Astwood (Eds.), Handbook of Physiology, Endocrinology II, Part 1, Waverly Press, Baltimore, Maryland, 1973, pp. 29–56. [5] N. Ben-Jonathan, L.A. Arbogast, J.F. Hyde, Neuroendocrine regulation of prolactin release, Prog. Neurobiol. 33 (1989) 399–447. ´ [6] M. Bodoky, M. Rethelyi, Dendritic arborization and axon trajectory of neurons in the hypothalamic arcuate nucleus of the rat, Exp. Brain Res. 28 (1977) 543–555. [7] P. Ciofi, W.R. Crowley, A. Pillez, L.L. Schmued, G. Tramu, M. Mazzuca, Plasticity in expression of immunoreactivity for neuropeptide Y, enkephalins and neurotensin in the hypothalamic tuberoinfundibular dopaminergic system during lactation in mice, J. Neuroendocrinol. 5 (1993) 599–602. [8] J.A. Clemens, M.E. Roush, R.W. Fuller, C.J. Shaar, Changes in luteinizing hormone and prolactin control mechanisms produced by glutamate lesions of the arcuate nucleus, Endocrinology 103 (1978) 1304–1312. [9] S.C. Danzer, N.T. McMullen, N.E. Rance, Dendritic growth of arcuate neuroendocrine neurons following orchidectomy in adult rats, J. Comp. Neurol. 390 (1998) 234–246. [10] J.J. Evans, Modulation of gonadotropin levels by peptides acting at the anterior pituitary gland, Endocr. Rev. 20 (1999) 46–67. ˚ ¨ [11] B.J. Everitt, B. Meister, T. Hokfelt, T. Melander, L. Terenius, A.


[12] [13] [14]

[15] [16]

[17]

[18]

[19]

[20]

[21]

[22]

[23]

[24]

[25]

[26] [27]

[28]

[29]

¨ Rokaeus, E. Theodorsson-Norheim, G. Dockray, J. Edwardson, C. Cuello, R. Elde, M. Goldstein, H. Hemmings, C. Ouimet, I. Walaas, P. Greengard, W. Vale, E. Weber, J.-Y. Wu, K.-J. Chang, The hypothalamic arcuate nucleus-medium eminence complex: immunohistochemistry of transmitters, peptides and DARPP-32 with special reference to coexistence in dopamine neurons, Brain Res. Rev. 11 (1986) 97–155. G. Fichera, Sur l’hypertrophie de la gland pituitire consecutive a la` castration, Arch. Ital. Biol. 43 (1905) 405–426. M. Frankfurt, Gonadal steroids and neuronal plasticity: Studies in the adult rat hypothalamus, Ann. NY Acad. Sci. 743 (1994) 45–60. ´ Morphological artifacts induced in intracelluA.A. Grace, R. Llinas, larly stained neurons by dehydration: circumvention using rapid dimethyl sulfoxide clearing, Neuroscience 16 (1985) 461–475. J.D. Green, G.W. Harris, Observation of the hypophysio-portal vessels of the living rat, J. Physiol. (Lond) 108 (1949) 359–361. ´ B. Halasz, L. Pupp, Hormone secretion of the anterior pituitary gland after physical interruption of all nervous pathways to the hypophysiotropic area, Endocrinology 77 (1965) 553–562. ¨ regulierung der W. Hohlweg, K. Junkmann, Die hormonal-nervose funktion des hypophysenvorderlappens, Klin. Wochenschr. 11 (1932) 321–323. T.L. Horvath, L.M. Garcia-Segura, F. Naftolin, Lack of gonadotropin-positive feedback in the male rat is associated with lack of estrogen-induced synaptic plasticity in the arcuate nucleus, Neuroendocrinology 65 (1997) 136–140. R.M. Lechan, J.L. Nestler, S. Jacobson, The tuberoinfundibular system of the rat as demonstrated by immunohistochemical localization of retrogradely transported wheat germ agglutinin (WGA) from the median eminence, Brain Res. 245 (1982) 1–15. ´ F.J. Lopez, I. Merchenthaler, M. Ching, M.G. Wisniewski, A. Negro-Vilar, Galanin: a hypothalamic–hypophysiotropic hormone modulating reproductive functions, Proc. Natl. Acad. Sci. USA 88 (1991) 4508–4512. V. Lux-Lantos, E. Rey, C. Libertun, Activation of GABA B receptors in the anterior pituitary inhibits prolactin and luteinizing hormone secretion, Neuroendocrinology 56 (1992) 687–693. J.P. Mather, A. Moore, R.H. Li, Activins, inhibins, and follistatins: further thoughts on a growing family of regulators, Proc. Soc. Exp. Biol. Med. 215 (1997) 209–222. A. Matsumoto, Y. Arai, Sexual dimorphism in ‘wiring pattern’ in the hypothalamic arcuate nucleus and its modification by neonatal hormonal environment, Brain Res. 190 (1980) 238–242. H. Matsuo, Y. Baba, R.M.G. Nair, A. Arimura, A.V. Schally, Structure of the porcine LH-and FSH-releasing hormone. I. The proposed amino acid sequence, Biochem. Biophys. Res. Commun. 43 (1971) 1334–1339. ¨ ´ M. Anden, ´ E. B. Meister, S. Ceccatelli, T. Hokfelt, N.-E. Anden, Theodorsson, Neurotransmitters, neuropeptides and binding sites in the rat mediobasal hypothalamus: effects of monosodium glutamate (MSG) lesions, Exp. Brain Res. 76 (1989) 343–368. I. Merchenthaler, The hypophysiotropic galanin system of the rat brain, Neuroscience 44 (1991) 643–654. I. Merchenthaler, C.R. Thomas, A. Arimura, Immunocytochemical localization of growth hormone releasing factor (GHRF)-containing structures in the rat brain using anti-rat GHRF serum, Peptides 5 (1984) 1071–1075. R. Mitchell, G. Grieve, R. Dow, G. Fink, Endogenous GABA receptor ligands in hypophysial portal blood, Neuroendocrinology 37 (1983) 169–176. J.A. Mong, E. Glaser, M.M. McCarthy, Gonadal steroids promote glial differentiation and alter neuronal morphology in the developing

[30]

[31]

[32]

[33]

[34]

[35]

[36]

[37]

[38]

[39]

[40]

[41]

[42]

[43]

[44]

[45] [46]

85

hypothalamus in a regionally specific manner, J. Neurosci. 19 (1999) 1464–1472. C.R. Moore, D. Price, Gonad hormone functions and the reciprocal influence between gonads and hypophysis with its bearing on the problem of sex hormone antagonisms, Am. J. Anat. 50 (1932) 13–71. F. Naftolin, C. Leranth, T.L. Horvath, L.M. Garcia-Segura, Potential neuronal mechanisms of estrogen actions in synaptogenesis and synaptic plasticity, Cell Mol. Neurobiol. 16 (1996) 213–223. M. Niimi, J. Takahara, M. Sato, K. Kawanishi, Immunohistochemical identification of galanin and growth hormone-releasing factorcontaining neurons projecting to the median eminence of the rat, Neuroendocrinology 51 (1990) 572–575. G. Olmos, F. Naftolin, J. Perez, P.A. Tranque, L.M. Garcia-Segura, Synaptic remodeling in the rat arcuate nucleus during the estrous cycle, Neuroscience 32 (1989) 663–667. A. Peters, I.R. Kaiserman-Abramof, The small pyramidal neuron of the rat cerebral cortex. The perikaryon, dendrites and spines, Am. J. Anat. 127 (1970) 321–355. W. Rall, Core conductor theory and cable properties of neurons, in: J.M. Brookhart, V.B. Mountcastle, E.R. Kandel, S.R. Geiger (Eds.), Handbook of Physiology: the Nervous System, American Physiological Society, Bethesda, Maryland, 1977, pp. 39–97. ´ M. Rethelyi, Dendritic arborization and axon trajectory of neurons in the hypothalamic arcuate nucleus of the rat-updated, Neuroscience 16 (1985) 323–331. M. Sar, A. Sahu, W.R. Crowley, S.P. Kalra, Localization of neuropeptide-Y immunoreactivity in estradiol-concentrating cells in the hypothalamus, Endocrinology 127 (1990) 2752–2756. P.E. Sawchenko, L.W. Swanson, J. Rivier, W.W. Vale, 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 hypothalamic GRF, J. Comp. Neurol. 237 (1985) 100–115. M.G. Stewart, M. Lowdnes, A. Hunter, T. Doubell, Memory storage in chicks involves an increase in dendritic spine number and synaptic density, Brain Dysfunct. 5 (1992) 50–64. M.L. Tappaz, M. Wassef, W.H. Oertel, L. Paut, J.F. Pujol, Light- and electron-microscopic immunocytochemistry of glutamic acid decarboxylase (GAD) in the basal hypothalamus: Morphological evidence for neuroendocrine g-aminobutyrate (GABA), Neuroscience 9 (1983) 271–287. A.M. Turner, W.T. Greenough, Differential rearing effects on rat visual cortex synapses. I. Synaptic and neuronal density and synapses per neuron, Brain Res. 329 (1985) 195–203. H.B.M. Uylings, J. van Pelt, R.W.H. Verwer, P. McConnell, Statistical analysis of neuronal populations, in: J.J. Capowski (Ed.), Computer Techniques in Neuroanatomy, Plenum, New York, 1989, pp. 241–264. A.N. Van den Pol, J.R. Cassidy, The hypothalamic arcuate nucleus of rat — a quantitative Golgi analysis, J. Comp. Neurol. 204 (1982) 65–98. M.A. Virmani, S.S. Stojilkovic, K.J. Catt, Stimulation of luteinizing hormone release by g-aminobutyric acid (GABA) agonists: Mediation by GABAA -type receptors and activation of chloride and voltage-sensitive calcium channels, Endocrinology 126 (1990) 2499–2505. S.J. Wiegand, J.L. Price, Cells of origin of the afferent fibers to the median eminence in the rat, J. Comp. Neurol. 192 (1980) 1–19. G.B. Wislocki, L.S. King, The permeability of the hypophysis and hypothalamus to vital dyes, with a study of the hypophyseal vascular supply, Am. J. Anat. 58 (1936) 421–472.