Neural sex steroids and hippocampal synaptic plasticity.

It was a widely held belief that sex steroids, namely testosterone and 17ß-estradiol (E2) of gonadal origin, control synaptic plasticity in the hippocampus. A new paradigm emerged when it was shown that these sex steroids are synthesized in the hippocampus. The inhibition of sex steroids in the hippocampus impairs synaptic plasticity sex-dependently in this region of the brain. In gonadectomized animals and in hippocampal cultures, inhibition of estradiol synthesis in female animals and in cultures from female animals, and inhibition of dihydrotestosterone synthesis in male animals and in cultures of male animals, cause synapse loss and impair LTP in the hippocampus, but not vice versa. Since the hippocampal cultures originated from perinatal animals, and due to the similarity of in vivo and in vitro findings, it appears that hippocampal neurons are differentiated in a sex-specific manner during the perinatal period when sexual imprinting takes place.

Structure-function-behavior relationship in estrogen-induced synaptic plasticity.

This article is part of a Special Issue "Estradiol and Cognition". In estrogen-induced synaptic plasticity, a correlation of structure, function and behavior in the hippocampus has been widely established. 17ß-estradiol has been shown to increase dendritic spine density on hippocampal neurons and is accompanied by enhanced long-term potentiation and improved performance of animals in hippocampus-dependent memory tests. After inhibition of aromatase, the final enzyme of estradiol synthesis, with letrozole we consistently found a strong and significant impairment of long-term potentiation (LTP) in female mice as early as after six hours of treatment. LTP impairment was followed by loss of hippocampal spine synapses in the hippocampal CA1 area. Interestingly, these effects were not found in male animals. In the Morris water maze test, chronic administration of letrozole did not alter spatial learning and memory in either female or male mice. In humans, analogous effects of estradiol on hippocampal morphology and physiology were observed using neuroimaging techniques. However, similar to our findings in mice, an effect of estradiol on memory performance has not been consistently observed.

[Gender and the effects of steroid hormones in the central nervous system]. / Geschlecht und Effekte von Steroidhormonen im Zentralnervensystem (ZNS).

Degenerative diseases of the central nervous system, the incidence and prevalence of which vary between men and women, often manifest in the hippocampus. Neurosteroids are hormones that are synthesized in the CNS, and it is here that they exert their influence. Estrogen and testosterone are examples of neurosteroid hormones. In the hippocampus, an area of the brain closely associated with learning and memory, the local synthesis of estrogen in females, but not in males, is essential for the plasticity and stability of the synapses. The inhibition of estrogen synthesis in the female hippocampus causes a reduction in long-term potentiation (LTP), an electrophysiological parameter of learning and memory, thus resulting in a significant loss of synapses. In light of this, the fact that estrogen has been attributed with many neuroprotective functions in degenerative diseases of the CNS suggests that therapeutic concepts involving the use of estrogen are possibly only effective in women, but not in men. These findings similarly provide a basis for explaining the gender dimorphism that has been found in certain degenerative illnesses of the CNS.

Hippocampal estradiol synthesis and its significance for hippocampal synaptic stability in male and female animals.

Increasing evidence points to an essential role played by neuron-derived neurosteroids, such as estrogen, on synaptic connectivity in the hippocampus. Inhibition of local estradiol synthesis results in synapse loss specifically in females, but not in males. Synapse loss in females, after inhibition of estradiol synthesis in hippocampal neurons, appears to result from impairment of long-term potentiation (LTP) and dephosphorylation of cofilin, and thereby the destabilization of postsynaptic dendritic spines. Such clear-cut effects were not seen in males. Cognitive deficits after inhibition of aromatase, the final enzyme of estrogen synthesis, have been seen in women, but not in men. Altogether, the data demonstrate distinct differences between genders in neurosteroid-induced synaptic stability.

Oestradiol-induced synapse formation in the female hippocampus: roles of oestrogen receptor subtypes.

During the oestrus cycle, varying spine synapse density correlates positively with varying local synthesis of oestradiol in the hippocampus. In this context, the roles of the oestrogen receptor (ER) subtypes ERα and ß are not fully understood. In the present study, we used neonatal hippocampal slice cultures from female rats because these cultures synthesise oestradiol and express both receptor subtypes, and inhibition of oestradiol synthesis in these cultures results in spine synapse loss. Using electron microscopy, we tested the effects on spine synapse density in response to agonists of both ERα and ERß. Application of agonists to the cultures had no effect. After inhibition of oestradiol synthesis, however, agonists of ERα induced spine synapse formation, whereas ERß agonists led to a reduction in spine synapse density in the CA1 region of these cultures. Consistently, up-regulation of ERß in the hippocampus of adult female aromatase-deficient mice is paralleled by hippocampus-specific spine synapse loss in this mutant. Finally, we found an increase in spine synapses in the adult female ERß knockout mouse, but no effect in the adult female ERα knockout mouse. Our data suggest antagonistic roles of ERß and ERα in spine synapse formation in the female hippocampus, which may contribute to oestrus cyclicity of spine synapse density in the hippocampus.

Oestrogen synthesis in the hippocampus: role in axon outgrowth.

Ovarian oestrogens have been postulated to be neuroprotective. It has also been shown that considerable amounts of oestrogens are synthesised in hippocampal neurones. In the present study, we focused on a potential role of hippocampus-derived oestradiol compared to gonad-derived oestradiol on axon outgrowth of hippocampal neurones. To address the role of hippocampus-derived oestradiol, we inhibited oestrogen synthesis by treatment of neonatal hippocampal cell cultures with letrozole, a specific aromatase inhibitor. As an alternative, we used siRNA against steroidogenic acute regulatory protein (StAR). Axon outgrowth and GAP-43 expression were significantly down-regulated in response to letrozole and in siRNA-StAR transfected cells. The effects after inhibition of oestrogen synthesis in response to letrozole and in siRNA-StAR transfected cells were reversed by oestrogen supplementation. No difference was found between ovariectomised animals, cycling animals at pro-oestrus and ovariectomised and subsequently oestradiol-treated animals. However, high pharmacological doses of oestradiol promoted axon outgrowth, which was possible to abolish by the oestrogen receptor antagonist ICI 182,780. Our results show that oestradiol-induced neurite outgrowth is very likely mediated by genomic oestrogen receptors and requires higher doses of oestradiol than physiological serum concentrations derived from the gonads.

Synaptic plasticity in the hippocampus: effects of estrogen from the gonads or hippocampus?

Different effects of estrogen on synaptic plasticity have [corrected] been reported. Here, we summarise effects of low, gonad-derived serum estrogen concentrations, of intermediate concentrations, provided by hippocampal cells, and of pharmacological doses of estrogen on synapses and spines and on the expression of synaptic proteins. No effects of low concentrations were found. To study the effects of hippocampus-derived estradiol, we inhibited hippocampal estrogen synthesis by treatment of hippocampal cell cultures with letrozole, an aromatase inhibitor. Alternatively, we used siRNA against Steroidogenic acute regulatory protein (StAR). Spines, synapses, and synaptic proteins were significantly down regulated in response to letrozole and in siRNA-StAR transfected cells. Application of high pharmacological doses of estradiol promoted only synaptophysin expression, a presynaptic protein, but did not increase the number of boutons. Our results point to an essential role of endogenous hippocampal estrogen in hippocampal synaptic plasticity rather than to a direct influence of estrogens derived from peripheral sources, such as the gonads.

Proliferation and apoptosis of hippocampal granule cells require local oestrogen synthesis.

Ovarian oestrogens have been demonstrated to influence neurogenesis in the dentate gyrus. As considerable amounts of oestrogens are synthesized in hippocampal neurones, we focused on the role of hippocampus-derived estradiol on proliferation and apoptosis of granule cells in vitro. We used hippocampal dispersion cultures, which allowed for cultivation of the cells under steroid- and serum-free conditions and monitoring of oestrogen synthesis. To address the influence of hippocampus-derived estradiol on neurogenesis, we inhibited oestrogen synthesis by treatment of hippocampal cell cultures with letrozole, a specific aromatase inhibitor. Alternatively, we used siRNA against steroidogenic acute regulatory protein (StAR). The number of proliferative cells decreased whereas the number of apoptotic cells increased dose-dependently, in response to reduced estradiol release into the medium after treatment with letrozole. This also held true for siRNA against StAR transfected cell cultures. Application of estradiol to the medium had no effect on proliferation and apoptosis whereas the anti-proliferative and pro-apoptotic effects of StAR knock-down and letrozole treatment were restored by treatment of the cultures with estradiol. Our findings suggest that neurogenesis and apoptosis in the hippocampus require a defined range of estradiol concentrations that is physiologically provided by hippocampal cells but not by gonads.

Direct and indirect effects of estrogen on rat hippocampus.

Estrogen-induced synaptic plasticity was frequently shown by an increase of spines at apical dendrites of CA1 pyramidal neurons after systemic application of estradiol to ovariectomized rats. Recent findings question this direct endocrine regulation of synaptogenesis by estradiol. We have shown, for the first time, that estrogens are synthesized de novo in rat hippocampal neurons. By using letrozole, an inhibitor of aromatase, estradiol levels in hippocampal dispersion cultures as well as in hippocampal slice cultures were significantly suppressed. Letrozole treatment resulted in a significant decrease in the density of spines and spine synapses and in the number of presynaptic boutons. Quantitative immunohistochemistry revealed a dose-dependent downregulation of spinophilin, a spine marker, and of synaptophysin, a presynaptic marker, in the hippocampus. Surprisingly, exogenous application of estradiol to the cultures had no effect. Indirect effects of estrogens, mediated via subcortical nuclei, may help to explain this phenomenon. Implantation of estrogen-filled cannulae into the median raphe, which projects to the hippocampus, resulted in a significant increase in spine density in the hippocampus after seven days of treatment. This increase was paralleled by a decrease in the density of serotonergic innervation of the strata lacunosum moleculare and radiatum of the CA1 region. Apart from direct endocrine mechanisms our findings suggest that estradiol-induced spinogenesis in the hippocampus is also mediated by indirect mechanisms and is furthermore regulated endogenously, in a paracrine manner.

Neurosteroid synthesis in the hippocampus: role in synaptic plasticity.

Neurosteroids are still found in the brain after steroidogenic glands were removed, indicating that they are synthesized either de novo or from endogenous precursors by enzymes present in the CNS. In fact, steroidogenic acute regulatory protein, and aromatase, two molecules essential for estrogen synthesis, are expressed in the hippocampus. We recently showed, for the first time, that estrogens are synthesized de novo in hippocampal neurons and that these hippocampus-derived estrogens are essential for synaptic plasticity. Both estrogen receptor isoforms, estrogen receptor alpha and estrogen receptor beta, are expressed in the hippocampus, and estradiol treatment of the cultures leads to an upregulation of estrogen receptor alpha. This finding confirmed the presence of functional estrogen receptors in hippocampal neurons and showed the responsiveness of the cultured hippocampal neurons to estradiol. By using letrozole, an inhibitor of aromatase, estradiol levels in hippocampal dispersion cultures as well as in hippocampal slice cultures were significantly suppressed which in turn led to a downregulation of estrogen receptor alpha. Letrozole treatment was followed by a significant decrease in the density of spines and spine synapses and in the number of presynaptic boutons. Quantitative immunohistochemistry revealed a dose-dependent downregulation of spinophilin, a spine marker, and of synaptophysin, a presynaptic marker, and of growth-associated protein 43 after letrozole treatment. Our data provide strong evidence for estrogens being potent modulators of structural synaptic plasticity and point to a paracrine rather than endocrine mechanism of estrogen action in the hippocampus.

Estrogen up-regulates estrogen receptor alpha and synaptophysin in slice cultures of rat hippocampus.

Previous studies have shown that estrogen application increases the density of synaptic input and the number of spines on CA1 pyramidal neurons. Here, we have investigated whether Schaffer collaterals to CA1 pyramidal cells are involved in this estrogen-induced synaptogenesis on CA1 pyramidal neurons. To this end, we studied estrogen-induced expression of both estrogen receptor (ER) subtypes (ERalpha and ERbeta) together with the presynaptic marker synaptophysin in the rat hippocampus. In tissue sections as well as in slice cultures mRNA expression of ERalpha, ERbeta and synaptophysin was higher in CA3 than in CA1, and mRNA expression and immunoreactivity for both ER subtypes were found in both principal cells and interneurons. By using quantitative image analysis we found stronger nuclear immunoreactivity for ERalpha in CA3 than in CA1. In slice cultures, supplementation of the medium with 10(-8) M estradiol led to an increase of nuclear immunoreactivity for ERalpha, but not for ERbeta, which was accompanied by a dramatic up-regulation of synaptophysin immunoreactivity in stratum radiatum of CA1. Together these findings indicate that estrogen effects on hippocampal neurons are more pronounced in CA3 than in CA1 and that ER activation in CA3 neurons leads to an up-regulation of a presynaptic marker protein in the axons of these cells, the Schaffer collaterals. We conclude that estradiol-induced spine formation on CA1 pyramidal cells may be mediated presynaptically, very likely by activation of ERalpha in CA3 pyramidal cells, followed by an increase in Schaffer collateral synapses.

Role of tumor necrosis factor in preovulatory follicles of swine.

The effects of tumor necrosis factor (TNF) on cultured porcine granulosa cells that were obtained from preovulatory follicles were studied with regard to following parameters: 1) TNF receptor type I expression, 2) progesterone receptor and transforming growth factor beta receptor type II (TbetaR II) as markers of luteinization, 3) proliferation, and 4) apoptosis. For comparative purposes the effects of TNF were also studied on insulin/forskolin-treated cells, as this treatment is well established to induce luteinization. Cytochemical methods followed by semiquantitative analysis were used. Our data show that TNF treatment upregulates TNF receptor type I expression in granulosa cells. TNF downregulates the expression of TbetaR II of insulin/forskolin-stimulated and of unstimulated cells. The progesterone receptor is also downregulated by the cytokine after insulin/forskolin-induced luteinization. Supplementation of the medium with TNF leads to increased proliferation and at the same time it induces apoptosis. Our results indicate that TNF exerts an inhibitory influence on luteinization and that TNF influences the balance between follicular growth (proliferation) and atresia (apoptosis).

Regulation of estrogen receptor alpha and progesterone receptor (isoform A and B) expression in cultured human endometrial cells.

The effects of RU 486 together with estradiol and progesterone on estrogen receptor alpha and progesterone receptor (isoforms A and B) expression were studied in human endometrial long term cultures at the mRNA and protein level. We asked whether ligand induced receptor regulation, found in mammals in vivo, is also found in human cultured endometrial cells with special regard to the progesterone isoforms A and B. Endometrial cultures were maintained for 27 days. Media were supplemented with progesterone and/or estradiol alone or in combination with RU 486. Receptor expression (estrogen receptor alpha and progesterone receptor isoform A and B) was examined at the mRNA level by RT-PCR and at the protein level by western blot analysis. All receptor types examined were expressed in our culture model. Estradiol led to a general increase of receptor expression whereas treatment with estradiol in combination with progesterone down regulated receptor expression. The receptor down regulation was not found when RU 486 was additionally supplemented into the medium. Activation or inhibition of expression due to these treatments was similar for both PR isoforms. Our results (1) show that in our culture system estradiol induced up regulation of estrogen receptor and progesterone receptor A and B and suggest that the estrogen induced up regulation is prevented by progesterone (2) a clear cut antigestagenic effect of RU 486 and (3) suggest that both progesterone isoforms are analogously regulated in our culture model. We conclude that human endometrial cell cultures are suitable for the study of the dynamics of steroid receptor expression.

Steroidogenic factor-1 expression in marmoset and rat hippocampus: co-localization with StAR and aromatase.

Steroidogenic factor-1 (SF-1), an orphan nuclear receptor, was studied with respect to the expression of steroidogenic enzymes in the hippocampus of rat and marmoset, since SF-1 is a regulator of steroid biosynthesis in the gonads. We used the steroidogenic acute regulatory protein (StAR) as a marker of the first step in the cascade of oestrogen synthesis and aromatase as a marker of the last. StAR transports cholesterol to the inner mitochondrial membrane where it is converted by the cytochrome P-450 enzyme complex. This is the rate-limiting step in steroid biosynthesis. Aromatase metabolizes testosterone to oestrogen. Using an anti-SF-1 antibody we show that SF-1 is highly expressed in neuronal cells of the pyramidal layer (CA1--CA3) and in the dentate gyrus of rat and marmoset hippocampi. Binding of the antibody was seen in more than 60% of all cells in the pyramidal layer and in the fascia dentata. In situ hybridization studies revealed the same expression pattern for StAR and aromatase. StAR and aromatase-positive cells were strictly correlated with SF-1 as shown by computer-assisted confocal microscopy in double labelling experiments (immunohistochemistry and in situ hybridization). This coexpression may imply SF-1 as a possible regulator of steroidogenesis in the hippocampus. However, a few interneurones express solely SF-1 and aromatase but are negative for StAR. Since the expression of StAR represents the first step in steroidogenesis its expression is suggestive for a de novo synthesis of steroids. A small population of interneurones must import precursors for oestrogen synthesis from other sources. Responsive cells, as evidenced by the presence of oestrogen receptor transcripts, were also found in the pyramidal layer and dentate gyrus. In conclusion, (1) SF-1 could play a regulatory role in steroidogenesis in the hippocampus of marmoset and rat and (2) with respect to the capacity of steroidogenesis two populations of hippocampal neurones coexist.

Mechanisms of estradiol inactivation in primate endometrium.

In uterine endometrium, the level of estradiol is controlled by oxidative 17beta-hydroxysteroid dehydrogenase (17HSD) activity which converts the bioactive hormone to the less active compound estrone. At least three different types of 17HSD (types 2, 4 and 8) use estradiol as their preferred substrate and may contribute to the overall rate of estradiol-inactivation in the uterus. In this study the marmoset monkey (Callithrix jacchus) was used for the investigation of the particular contribution of each type of 17HSD. Northern Blots revealed essentially the same tissue distribution as in the human. Likewise, uterine 17HSD enzyme activity increases in the secretory phase of the reproductive cycle, in parallel to the rise in circulating progesterone levels. Northern analysis of uteri from defined time points of the reproductive cycle showed that only the level of 17HSD2 expression is strongly upregulated in the secretory phase, whereas 17HSD4 and 17HSD8 seem to be expressed constitutively.

Inhibition of proliferation and differentiation by RU 486 in human endometrial stromal and epithelial cells in vitro.

The effects of progesterone and RU 486 on cellular proliferation and differentiation in long term cultures of mixed human endometrial cells were studied. The endometrial tissue was obtained from women with normal menstrual cycles who were undergoing hysterectomy for benign growths. Estradiol supplemented cultures were treated with progesterone and/or RU 486 for 27 days. Cell number was measured by crystal violet assay, and prolactin secretion was used as a marker of differentiation. Progesterone doubled the rate of proliferation, but the addition of RU 486 reduced it to baseline again. The gestagen increased prolactin secretion up to 30 times, while the addition of RU 486 suppressed it to baseline levels. When administered to cells that were pretreated with progesterone for 15 days RU 486 abolished the progesterone effects. RU 486 alone was without any effect. Our results indicate that (1) in vitro progesterone is essential for the initiation and maintenance of proliferation and differentiation of endometrial cells and (2) RU 486 acts as a pure progesterone antagonist in our culture model.

Spontaneous luteinization of antral marmoset follicles in vitro.

Large non-luteinized follicles of the marmoset monkey were cultured for up to 96 h in the presence of substances that are known to induce luteinization, i.e. LH, transforming growth factor (TGF)-beta and cyclic AMP. The state of the basal lamina, and the expression of connexin-43, alpha(2) integrin subunit and TGF-beta receptor type II (TbetaR-II) were chosen as parameters to judge the progress of luteinization. Antral follicles, cultured for 1 h, were not luteinized, as shown by an intact basal lamina, strong immunoreactivity of connexin-43 in granulosa cells, and no expression of TbetaR-II in the theca layer. After 12 h, most follicles showed a dissolution of the basal lamina, a faint reactivity of connexin-43, high expression of TbetaR-II in theca- and outer granulosa cells and high expression of alpha(2) integrin subunit in granulosa cells bordering at the basement membrane; all of which indicate luteinization. After 96 h of culture, luteal structures (e.g. corpora lutea accessoria) had developed. This was true for both non-stimulated and stimulated follicles. Our results strongly suggest that antral follicles luteinize spontaneously. The decisive determinant appears to be the follicular stage.

Differential expression of 17beta-hydroxysteroid dehydrogenases types 2 and 4 in human endometrial epithelial cell lines.

In the endometrium two enzymes are known to convert estradiol to its inactive metabolite estrone: microsomal 17beta-hydroxysteroid dehydrogenase type 2 (17beta-HSD2) and peroxisomal 17beta-HSD4. In order to elucidate the particular function of each of these two different enzymes, the human endometrial epithelial cell lines HEC-1-A and RL95-2 were examined with respect to the expression of 17betaHSD isozymes. They were compared with human endometrium in vivo. Non-radioactive in situ hybridization revealed both enzymes in glandular epithelial cells of human endometrium. The two cell lines were screened for mRNA expression of 17beta-HSD 1-4 by RT-PCR and Northern blot. 17beta-HSD2 and 4 could be detected by either method, 17beta-HSD1 only by RT-PCR, 17beta-HSD3 not at all. Both cell lines were proven to have no receptor for progesterone which is known as a physiological inducer of several 17beta-HSD isozymes. To study the regulation of 17beta-HSD2 and 17betaHSD4, the concentration of fetal calf serum in the cell culture media was reduced stepwise to 0.3% by dilution with a defined serum replacement. This treatment led to an inhibition of 17beta-HSD2 mRNA expression and an increase in the mRNA expression of 17beta-HSD4. Concomitantly, distinct morphological changes were observed, such as a decrease in the number and length of microvilli and a decrease in the formation of domes on top of the monolayers. The endometrial epithelial cell lines HEC-1-A and RL95-2 represent a suitable in vitro model for further studies of the differential expression of the major endometrial HSD isozymes, independent of the effect of progesterone.

TGF-beta could be involved in paracrine actions in the epididymis of the marmoset monkey (Callithrix jacchus).

The transforming growth factor-beta1 (TGF-beta1) and the transforming growth factor-beta receptor type II (TGF-betaRII) were studied in the epididymis of sexually mature marmoset monkeys (Callithrix jacchus) by immunohistochemical localization of the protein and by polymerase chain reaction (PCR) analysis of the mRNA level. In order to specify reactive cell types, the morphology of all three segments (caput, corpus, and cauda epididymidis) was evaluated by light microscopy. Six different cell types could be distinguished: principal, basal, apical, and clear cells, as well as intraepithelial lymphocytes and macrophages. Using immunohistochemistry, specific staining for TGF-beta1 in the caput was found in 47% of the apical cells, whereas the TGF-betaRII was located in the apical portion of 91% of all principal cells. In the corpus epididymidis, 20% of the apical cells were immunopositive for TGF-beta, and binding of the receptor antibody occurred in 17% of the principal cells (all numbers based on counts of counterstained nuclei). All differences between percentages in the caput and corpus were significant as determined by chi-square test. PCR analysis revealed detectable levels of TGF-beta1 mRNA in the marmoset epididymis. Our results indicate for the first time that TGF-beta1 is synthesized in the marmoset epididymis, possibly in a different subpopulation of epididymal cells than the TGF-beta receptor type II. Thus, TGF-beta might be of functional relevance in the primate epididymis.

Effects of single polychlorinated biphenyls on the morphology of cultured rat tubuli seminiferi.

The effects of single polychlorinated biphenyls (PCB) on isolated tubuli seminiferi of the rat were studied. Freshly isolated rat tubuli seminiferi were prepared according to their transillumination pattern, i.e. dark or pale. Tubuli seminiferi with the dark pattern included stages II to VIII and tubuli with the pale pattern represented stages IX to XIV and stage I of the seminiferous cycle. Afterwards, tubuli seminiferi were exposed to single polychlorinated biphenyls for 5 or 24 h in vitro. PCB 126, PCB 77, and PCB 118 were used in final parts per billion (p.p.b.) concentrations as determined by quantitative PCB analysis. Eventually, the specimens were plastic embedded, cut into semithin sections, stained, and morphology was evaluated by light microscopy. Single PCB congeners induced morphological alterations in cultivated rat tubuli seminiferi in a time- and stage-dependent manner. Effects comprised loosened intercellular contacts between germ cells and Sertoli cells as well as cellular fragmentation in the layer of round spermatids. Early spermiogenesis seems particularly susceptible to single PCB congeners in concentration of background magnitude. The target cell has still to be discovered.