Carole Lewis *, Bruce S. McEwen, Maya Frankfurt. Laboratory of Neuroendocrinology, Rockefeller University, 1230 York Avenue, New York, NY 10021, USA.
DEVELOPMENTAL BRAIN RESEARCH
ELSEVIER
DevelopmentalBrain Research 87 (1995) 91-95
Short communication
Estrogen-induction of dendritic spines in ventromedial hypothalamus and hippocampus: effects of neonatal aromatase blockade and adult GDX Carole Lewis *, Bruce S. McEwen, Maya Frankfurt Laboratory of Neuroendocrinology, Rockefeller University, 1230 York Avenue, New York, NY 10021, USA
Accepted 14 March 1995
Abstract
Treatment of male rats at birth with an aromatase inhibitor (Letrazole), followed by adult gonadectomy GDX, led to an increase in dendritic spine density on ventromedial hypothalamic neurons (VMN) when treated with estrogen as compared to a decrease when vehicle animals were given estrogen. In contrast, estrogen-treatment increased dendritic spine density on CA1 pyramidal neurons regardless of neonatal treatment. In addition, in CA1 there was a significant difference between the two estrogen groups. These results suggest that estrogen induction of dendritic spines in the VMN and CA1 is dependent on organizational effects of gonadal steroids. Keywords: Dendritic spine; Estrogen; Gonadectomy; Aromatase; VMN; Hippocampus
During pre-and postnatal development, gonadal steroids play a critical role in sexual differentiation of the brain [10,12,15,18]. Perinatal exposure to gonadal steroids organizes the sex-specific patterns of dendritic differentiation and synaptic organization in a number of brain regions [2-4,24]. Immediately after birth, males have higher levels of circulating testosterone (T) [12]. Unlike estradiol in newborn rats which binds to alpha fetoprotein, circulating T gains access to the brain [13,18]. Once inside the brain, T is aromatized by certain cells into estradiol, and binds, in those cells or adjacent cells, with specific estrogen receptors resulting in a defeminized neuronal circuitry [15]. In adulthood these hormones then reversibly activate the neural pathways involved in producing reproductive behaviors and influencing other sexually-differentiation neural processes [15,18]. The adult mammalian brain shows remarkable plasticity in response to g o n a d a l steroid m a n i p u l a t i o n s [1,2,19,23,26-29]. Fluctuations in dendritic spine density occur during the estrous cycle on neurons in the ventromedial nucleus (VMN) of the hypothalamus as well as in the CA1 region of the hippocampus [7,26-29]. In the VMN there is a sex difference in the synaptic pattern of innervation [17] and in dendritic spine density in response to
* Corresponding author.Fax: (1) (212) 327-8634. 0165-3806/95/$09.50 © 1995 Elsevier Science B.V. All rights reserved SSDI 0165-3806(95)00052-6
gonadectomy (GDX): i.e., in female rats, GDX caused a decrease in VMN dendritic spine and synaptic density [7,8] whereas in the male rats GDX induced an increase in dendritic spine density [9]. This was reversed through estrogen replacement of GDX adult female rats which increased dendritic spine and synapse density to levels seen in intact females in the VMN [5,7,8]. In the male GDX rat, however, estrogen replacement decreased dendritic spine density levels back to intact levels [9]. In the hippocampus of female rats, GDX decreased density of spines on dendrites of CA1 neurons and estrogen treatment reversed this decline [11]. The present study extends the above findings by exploring the structural adaptation of hormone sensitive neurons in the hippocampus of the adult male. In addition, the organizational a n d / o r activational effect of hormones on dendritic spine density were investigated by examining the estrogen-induced changes in dendritic spine density in adult males that were neonatally treated with the AI, CGSS2026T (Letrazole), in an attempt to alter the hormone-sensitive structure of both the VMN and the CA1 region of the hippocampus during adulthood. Fifteen day timed pregnant Sprague-Dawley female rats were housed under controlled light (14 h:10 h; light; dark) conditions with food and water available ad libitum. Within 6 h of birth animals were sexed and 20 of the male offspring were injected with the AI Letrozole (1 mg, CGSS2026T; CIBA-Geigy, Basel) in 2.5% carboxymethyl-
C. Lewis et al. / Developmental Brain Research 87 (1995) 91-95
92
cellulose vehicle (V) and twenty male pups injected with the V. Following treatment, male pups were cross-fostered, within treatment groups, and then weaned on day 50. On day 70, all animals were GDX under Metofane anesthesia. The animals were assigned to one of the following treatment groups during the interval between surgery and death: (1) beginning 5 days after surgery 10 AI-treated animals received two subcutaneous injections of estradiol benzoate (E, 1 0 / z g / 0 . 1 ml sesame oil) 24 h apart and sacrificed 48 h after the last estradiol injection; (2) 10 V-treated rats received the same E injection regimen; (3) 10 AI-treated animals received two subcutaneous injections of sesame oil alone at each of the time period described above; (4) 10 V-treated rats also received sesame oil injections at the above time periods; (5) 10 untreated adult intact males (I) were also processed at this same time. All rats were deeply anesthetized with Metofane and transcardially perfused with 100-150 ml of 4% paraformaldehyde in 0.1 m phosphate buffer with 1.5% picric acid. Brains were post-fixed for 24 h in a solution of the same composition as the perfusate and processed using a modified version of the single-section Golgi-impregnation method. Coronal sections, 100 /xm, were cut using a Vibratome into a bath of 3% potassium dichromate in distilled water and kept in this solution for 24 h. Following this the sections were rinsed in distilled water and mounted onto ungelatinized slides. Coverslips were glued over the tissue sections at the 4 comers and the slide assemblies were placed in 1.5% silver nitrate for 24 h in the dark. The slide assemblies were then dismantled and the sections rinsed in
distilled water, dehydrated, cleared, and coverslipped. The slides containing Golgi-impregnated brain sections were coded prior to quantitative analysis. The code was not broken until after the analysis was completed. In order to be selected for analysis, Golgi-impregnated hippocampal pyramidal cells had to possess the following characteristics: (1) location within the dorsal portion of the CA1 hippocampal field; (2) dark and consistent impregnation throughout the extent of the neuron; (3) location in the stratum radiatum between 250-400 /zm from the pyramidal cell body; (4) remain approximately in one plane of focus, and be greater than 15 /zm in length; and (5) relative isolation from neighboring impregnated cells in order to allow identification of dendrites which emanated from specific cells. For each dendritic segment selected, spine density was measured as follows: (1) the selected segment was traced (1250 × ) with a camera lucida drawing tube, (2) all the dendritic spines visible along that segment were counted, (3) the length of each segment was measured from its camera lucida drawing with the Zeiss Interactive Digitizing Analysis System (ZIDAS), (4) the data were then expressed as a number of spines per 10 /xm dendrite. Three to 5 dendritic segments per cell and 6 cells per animal were analyzed in this way. Means for each variable were calculated for each animal and the data were subjected to one way analysis of variance (ANOVA) with Tukey HSD post-hoc comparisons. Data are expressed as mean ___S.E.M.. All neurons readily distinguishable in the entire VMN were drawn using the above described drawing methods
Aromatase Inhibition and Estrogen in VMN
200
•
E
Spine Density
100 C
AlE
AIG
k,E
~3
I
Treatment Fig. 1. Spine d e n s i t y / m m primary dendrites of V M N neurons in male rats treated at birth with the AI, Letrazole, or with vehicle. Adult treatment consisted of GDX and treatment with estrogen or with vehicle as described in Methods. One way A N O V A showed significant treatment effects ( F = 6.002, 4,42 dr, P < 0.0007). Comparisons between individual means by T u k e y ' s HSD test showed significant differences between individual means at P < 0.05: * significant difference from VE and I; * * significant difference from VE and I. I, intact; V, vehicle at birth; E, estradiol; G, gonadectomized; AI, aromatase inhibitor treatment at birth.
C. Lewis et al. / Developmental Brain Research 87 (1995) 91-95
using the camera lucida. From these drawings, the number of primary dendrites and the number of dendritic branch points were counted (6 neurons per brain; 5-15 brains per group). For spine density measurements, the primary dendrite with the greatest number of spine was traced and the spines counted at 1250 X. Spine density was then expressed per millimeter primary dendrite. Data was analyzed by calculating means for each variable for each brain and then by one way ANOVA followed by post-hoc comparisons. Data are expressed as mean + S.E.M.. In the VMN, GDX of adult male rats caused spine density on primary dendrites to increase, whereas estrogen treatment caused a significant decrease in spine density, in agreement with our previous findings [9] (Fig. 1). This effect of GDX and estrogen treatment of males is opposite to that found for female rats, namely, a decrease after ovariectomy and an induction of spines after estrogen replacement that mimics an increased spine density during the estrous cycle [7]. Estrogen treatment of male rats treated neonatally with the aromatase blocker and GDX as adults resulted in a spine density in VMN that was not different from V-treated male rats that were GDX as adults and was significantly higher from V-treated adult GDX given estrogen treatment (Fig. 1). In the CA1 region of hippocampus, GDX caused dendritic spine density to decrease (Fig. 2). Intact male rats had higher apical dendritic spine densities than male rats that were GDX as adults as well as male rats treated with the AI at birth and GDX as adults (Fig. 2). There was no difference between adult GDX control males and GDX AI
93
males. Estrogen treatment of adult-GDX males induced a 15% increase of dendritic spines, whereas estrogen treatment of neonatally AI-treated, adult-GDX males increased spine density by 27% (Fig. 2). The differences between the two estrogen-treated groups was also significant, indicating that aromatase blockade at birth increases the response to estrogen treatment in adulthood. The results of the present study show that estrogen induction of morphological changes in hypothalamus and hippocampus in adult rats is qualitatively different. In the VMN, dendritic spine density increases after GDX in males [9] and decreases after GDX in females [7] and these spine changes are reversed by estrogen treatment. In the CA1 region of hippocampus there was a marked decrease in spine density after GDX in females which was reversed by estrogen treatment [11]. Similarly, in the present study we have found that GDX decreases dendritic spine density in neurons in the CA1 region of the hippoeampus of male rats. Regional and sex differences in response GDX and estrogen treatment indicate possible developmental programming of responses to estrogens, an effect supported by the results of the postnatal treatment with the AI, Letrazole, which increased in male rats both the CA1 and VMN responses to estradiol in a female direction. In the VMN of males, spine density on primary dendrites increased after adult GDX, and estrogen treatment reversed this post-GDX increase, in agreement with our previous report [9], and opposite to what happens in adult females [7]. However, the story for adult males and females is different from that found during prepubertal development: i.e., compared to same-age females, male VMN
Aromatase Inhibition and Estrogen in CA3
20
== •
SpineDensity
10
o AIE
AIG
VE
VG
I
Treatment
Fig. 2. Spine density/10 ~ m on primary apical dendrites of hippocampal CA1 pyramidal neurons in male rats treated at birth with the AI, Letrazole, or with vehicle. One-way A N O V A was significant ( F = 339.38, df = 4, P < 0.0001). Comparisions between individual means by Tukey's HSD test at P < 0.05: (1) Intact male (I) spine density was higher than each other treatment group; (2) Estrogen treatment induced increased spines in both control, GDX male (VE vs. VG) and aromatase-inhibtor, GDX males (AlE vs. AIG); (3) Estrogen treatment resulted in higher spine density in aromatase-treated, GDX males (A/E) vs. control, GDX males (VE). I, intact; V, vehicle at birth; E, estradiol; G, gonadectomized: Al, aromatase inhibitor treatment at birth.
94
C. Lewis et aL / Developmental Brain Research 87 (1995) 91-95
neurons showed a progressively smaller (but not opposite) response to exogenous estrogen priming after GDX at postnatal d16 or d36 [23]. Therefore, postpubertal development must result in further develomental changes in the VMN of males leading to the dramatic difference in their responses to estrogen from younger males and from females. While we do not know yet what these factors may be, we did show in the present study that neonatal treatment with the AI, Letrazole, produced an elevated spine density in estrogen-primed GDX animals This effect was in the direction of that seen in the VMN of GDX females after estrogen priming. Previous studies with an AI, 1,4,6-androstatriene-3,17dione, given to newborn male rats have shown a reversal of sex differences in estrogen-inducible progestin receptors in VMN, as well as blockade of defeminization of lordosis behavior, a response dependent on estrogen and progesterone priming of the VMN [21]. Although estradiol is present in the blood of newborn females and males, it is largely prevented from entering brain cells by estrogen binding proteins [13,18]. Yet, during perinatal development, males have higher levels of circulating testosterone that can pass freely past the estrogen binding protein [13]. Once inside cells containing aromatase, testosterone is aromatized into estradiol, which binds to estrogen receptors in the same or nearby cells and promoting the development of masculinized neuronal circuitry [13,15,18,25]. In the present study it appears that neonatal administration of the AI, Letrazole, at least partially prevented the defeminizing effects of testosterone with regard to VMN and CA1 response to estradiol given in adulthood. It is noteworthy that whereas the VMN expresses estrogen receptors and aromatase from late fetal life through adulthood [13,22,25], the hippocampus at birth expresses both aromatizing enzymes [16] and estrogen receptors [20] only transiently. In contrast to male VMN, spine density on apical dendrites of male CA1 pyramidal neurons decreased after GDX, and estrogen administration had only a small effect in increasing spine density on CA1 pyramidal neurons. Treatment of newborn males rats with Letrazole enhanced the ability of estrogen treatment to induce spines on apical dendrites of CA1 pyramidal neurons. Aromatase-treated males showed an almost two-fold increase in spine density in response to estradiol than V males, and the magnitude this increase was 64% of the increment in spine density of GDX females after estrogen treatment [11,25]. This data is the first report of its kind and we do not yet know whether androgen treatment of adult GDX male rats would fully reverse the GDX-induced decrease of dendritic spine density in hippocampus. Changes induced by estrogen in the brain are not uniform. For example, previous studies have shown that estrogen administration of GDX female rats did not alter dendritic spine density in either hippocampal pyramidal cells in the CA3 region or granule cells in the dentate
gyrus [11]. It is not surprising, therefore, that estrogen actions on spine density in the VMN and the CA1 region of the hippocampus are not the same. The VMN is known to contain a large population of estrogen concentrating cells and it is thought that estrogen directly alters VMN neurons [22]. In contrast, adult hippocampal pyramidal cells do not contain estrogen receptors [14] and therefore changes in synapse density may reflect changes in the number a n d / o r activity of presynaptic afferents. For example, local inhibitory interneurons project onto other inhibitory neurons as well as onto pyramidal cells, and some of these interneurons have estrogen receptors [6,14]. Another possibility is that CA1 neurons are influenced by estrogen concentrating cells elsewhere, such as the VMN. Therefore estrogen-induced spine changes in the CA1 area may be indirectly mediated and the hormonal basis of sex differences may reside in the afferent neurons rather than in the CA1 neurons themselves.
References [1] Arai, Y. and Matasumoto, A., Synapse formation of the hypothalamic arcuate nucleus during post-natal development in the female rat and its modification by neonatal estrogen treatment, Psychoneuroendocrinology, 3 (1986) 31-45. [2] Arnold, A. and Breedlove, S., Organizational and activational effects of sex steroids on brain and behavior: a reanalysis, Horm. Behav., 19 (1985) 469-498. [3] Arnold, A. and Gorski, R., Gonadal steroid induction of structural sex differences in the central nervous system, Annu. Rev. Neurosci., 7 (1984) 413-442. [4] Breedlove, S.M., Sexual dimorphism in the vertebrate nervous system, J. Neurosci., 12 (1992) 4133-4142. [5] Carrer, H. and Aoki, A., Ultrastructural changes in the hypothalamic ventromedial nucleus of ovariectomized rats after estrogen treatment, Brain Res., 240 (1982) 221-233. [6] DonCarlos, L.L., Monroy, E. and Morrell, J.I., Distribution of oestrogen receptor immunoreactive cells in the forebrain of the female guinea pig, J. Comp. Neurol., 305 (1991) 591-612 [7] Frankfurt, M., Gould, E., Woolley, C. and McEwen, B.S., Gonadal steroids modify dendritic spine density in ventromedial hypothalamic neurons: a Golgi study in the adult rat, Neuroendocrinology, 51 (1990) 530-535. [8] Frankfurt, M. and McEwen, B.S., Estrogen increases axodendritic synapses in the VMN of rats after ovariectomy, NeuroReport, 2 (1991) 380-382. [9] Frankfurt, M. and McEwen, B.S., 5,7-Dihydroxytryptamine and gonadal steroid manipulation alter spine density in ventromedial hypothalamic neurons, Neuroendocrinology, 54 (1991) 653-657. [10] Gorski, R.A., Steroid induced sexual characteristics in the brain, Neuroendocr. Persp., 2 (1983) 1-35. [11] Gould, E., Woolley, C., Frankfurt, M, and McEwen, B.S., Gonadal steroids regulate dendritic spine density in hippocampal pyramidal cells in adulthood, J. Neurosci., 10 (1990) 1286-1291. [12] Goy, R. and McEwen, B., Sexual Differentiation of the Brain, The MIT Press, Cambridge, MA, 1977. [13] Johnson, M. and Everitt, B., EssentialReproduction, VoL III, Blackwell Scientific Publications, London, 1988, pp. 35-49. [14] Loy R., Gerlach, J. and McEwen, B., Autoradiographic localization of estradiol-binding neurons in rat hippocampal formation and entorhinal cortex, Dev. Brain Res., 39 (1988) 245-251.
C. Lewis et al. / Developmental Brain Research 87 (1995) 91-95 [15] Maclusky, N.J. and Naftolin, F., Sexual differentiation of the central nervous system, Science. 211 (1981) 1294-1303. [16] MacLusky, N., Clark, A.S., Naftolin, F. and Goldman-Rakie, P.S., Oestrogen formation in the mammalian brain: possible role of aromatase in sexual differentiation of tht hippocampus and neocortex, Steroids, 50 (1987) 459-474. [17] Matsumoto, A. and Arai, Y., Male-female difference in synaptic organization of the ventromedial nucleus of the hypothalamus in the rat, Neuroendocrinology, 42 (1986) 232-236. [18] McEwen B., Gonadal steroid influences on brain development and sexual differentiation. In R.O. Greep (Ed.), Reproductive Physiology,/V 27 (1983) 99-145. [19] Naftolin, F., Garcia-Segura, L.M., Keefe, D., Leranth, C., Maclusky, N.J. and Brawer, J.R., Estrogen effects on the synaptology and neural membranes of the rat hypothalamic arcuate nucleus. The biology of reproduction, Biol. Reprod., 42 (1990) 21-28. [20] O'Keefe, J.A. and Handa, R.J., Transient elevation of oestrogen receptors in thc neonatal rat hippocampus, Dev. Brain Res., 57 (1990) 119-127 [21] Parsons, B., Rainbow, T.C. and McEwen, B.S., Organizational effects of testosterone via aromatization on feminine reproductive behavior and neural progestin receptors in rat brain, Endocrinology, 115 (1984) 1412-1417. [22] Pfaff, D.W. and Keiner, M., Atlas of estradiol-concentrating cells in the central nervous system of the female rat, J. Comp Neurol., 151 (1973) 121-1158.
95
[23] Segarra, A. and McEwen, B.S., Estrogen increases spine density in ventromedial hypothalamic neurons of peripubertal rats, Neuroendocrinology, 54 (1991) 365-372. [24] Toran-Allerand, C.D., Sex steroids and the development of the newborn hypothalamus and preoptic area in vitro. II. Morphological correlated and hormonal specificity, Brain Res., 189 (1980) 413427. [25] Tsuruo, Y., Ishimura, K., Fujita, H. and Osawa, Y., Immunocytochemical localization of aromatase-containing neurons in the rat brain during pre-and postnatal development, Cell Tiss. Res., 278 (1994) 29-39. [26] Woolley, C., Gould, E., Frankfurt, M. and McEwen, B.S., Naturally occurring fluctuation in dendritic spine density on adult hippocampal pyramidal neurons, .L Neurosci., 10 (1990) 4035-4039. [27] Woolley, C. and McEwen, B.S. Estradiol mediates fluctuation in hippocampal synapse density during the estrous cycle in the adult rat, J. Neurosci., 12 (1992) 2549-2554. [28] Woolley, C. and McEwen, B.S., Roles of estradiol and progesterone in regulation of hippocampal dendritic spine density during the estrous cycle in the rat, J. Comp. Neurol., 336 (1993) 293-306. [29] Woolley, C. and McEwen, B.S., Estradiol regulates hippocampal dendritic spine density via an N-methyl-D-aspartate receptor dependent mechanism, J. Neurosci.,14 (1994) 7680-7687.
