Adrenal steroids and plasticity of hippocampal neurons: toward an understanding of underlying cellular and molecular mechanisms

On the role of glucocorticoid receptors in brain plasticity

1. The mapping of glucocorticoid receptors (GR) in the rat central nervous system (CNS) has demonstrated their widespread presence in large numbers of nerve and glial cell populations also outside the classical stress regions. 2. The present paper summarizes the evidence that glucocorticoids via GR in the CNS can act as lifelong organizing signals from development to aging. The following examples are given. (a) In the prepubertal and adult offspring, prenatal corticosterone treatment can produce long-lasting changes in striatal dopaminergic communication. (b) In adulthood, the evidence suggests complex regulation by adrenocortical hormones of neurotrophic factors and their receptors in the hippocampal formation. (c) In aging, the strongly GR-immunoreactive pyramidal cell layer of the CA1 hippocampal area appears to be preferentially vulnerable to neurotoxic actions of glucocorticoids, especially in some rat strains. 3. Strong evidence suggests that each nerve cell in the CNS is supported by a trophic unit, consisting of other nerve cells and glial cells, blood vessels, and extracellular matrix molecules. Due to multiple actions on nerve and glial cell populations of the different trophic units, the glucocorticoids may exert either an overall trophic or a neurotoxic action. It seems likely that with increasing age, the endangering actions of glucocorticoids on nerve cells prevail over the neurotrophic ones, leading to reduced nerve cell survival in some trophic units.

Chronic exposure to stress levels of corticosterone alters GABAA receptor subunit mRNA levels in rat hippocampus

Chronic exposure to stress levels of corticosteroids alters many aspects of hippocampal function and may lead to neurodegeneration. Male rats were treated for 10 days with corticosterone (CORT) or vehicle pellets, and mRNA levels for six gamma-aminobutyric acid (GABAA) receptor subunits were measured. Effects of castration on subunit mRNA levels in CORT- and vehicle-treated animals were also examined. In situ hybridization studies demonstrated that mRNA levels for hippocampal GABAA receptor alpha 1, alpha 2, beta 1, beta 2, beta 3, and gamma 2 subunits were differentially altered by CORT treatment. Levels of alpha 1 and alpha 2 mRNA decreased in the dentate gyrus, and beta 1 mRNA levels decreased in CA1 and dentate gyrus of CORT-, compared to vehicle-treated, animals. In contrast, beta 2 subunit levels increased in all hippocampal regions examined, beta 3 levels increased in the dentate gyrus, and gamma 2 levels increased in CA1-CA3. The alpha 1, beta 1, and beta 2 mRNA levels all increased in the cingulate cortex of CORT-treated animals. There was no significant effect of gonadal state on any of the subunits examined, but there was a significant negative correlation between testosterone levels and mRNA levels of alpha 1, alpha 2 and beta 3 in specific regions. These data demonstrate that chronic exposure to stress levels of CORT produces complex changes in the mRNA levels of multiple GABAA receptor subunits, independently of the CORT-induced suppression of circulating testosterone.

Chronic psychosocial stress induces morphological alterations in hippocampal pyramidal neurons of the tree shrew

The effect of sustained psychosocial stress on the morphology of hippocampal pyramidal neurons was analysed in male tree shrews after 14, 20, and 28 days of social confrontation. A variety of physiological changes such as constantly elevated levels of urinary cortisol and norepinephrine and reduced body weight, which are indicative of chronic stress were observed in the subordinate, but not in the dominant males. Light microscopic analysis of Nissl-stained hippocampal sections showed that the staining intensity of the nucleoplasm in the CA1 and CA3 pyramidal neurons was increased after prolonged psychosocial stress, indicating a change in the nuclear chromatin structure. These alterations were observed only in subordinate animals and increased in a time dependent manner in accordance with the length of the stress period. There was, however, neither a reduction in density nor a degeneration of pyramidal neurons in chronically stressed animals. Mechanisms which may possibly account for the observed alterations are discussed.

Increased glucocorticoid receptor gene expression in the rat hippocampus following combined serotonergic and medial septal cholinergic lesions

Glucocorticoid excess is associated with hippocampal neuronal dysfunction and loss, mainly affecting CA1. Degeneration of both cholinergic and serotonergic (5-HT) hippocampal afferents is prominent in aged rats and Alzheimer's disease. Lesions of these individual pathways alter hippocampal expression of mineralocorticoid (MR) and glucocorticoid (GR) receptor mRNAs; both transcripts are increased by cholinergic lesions, but markedly decreased by serotonergic denervation. In the present study we found that combined medial septal cholinergic and central 5-HT lesions increase hippocampal GR mRNA expression, specifically in CA1 and CA2 subfields, whereas MR mRNA expression was similar to controls. Thus the effects of the cholinergic lesion, at least upon GR gene expression, appear to predominate while the effects of the lesions upon MR gene expression were additive. Increased hippocampal GR gene expression per neuron may increase hippocampal neuronal vulnerability with age or disease.

Coincident plasma ACTH and corticosterone time series: comparisons between young and old rats

Senescence is accompanied by a reduced ability to respond to a variety of physical and behavioral stressors. A sizable literature has been devoted to the interplay between hypothalamic-pituitary-adrenocortical axis dysfunction and senescence; yet, the precise interactions remain an enigma. Adrenocorticotropic hormone (ACTH) is secreted in pulsatile bursts generating complex signals in the plasma compartment that must be "read" by adrenocortical cells in order to initiate appropriate secretory responses. We have previously demonstrated subtle differences between young and old rats in the pattern of fluctuations in plasma ACTH concentrations over time, despite no difference in mean levels. The present work addressed the physiological significance of these differences in the plasma ACTH signal by analyzing the corresponding plasma corticosterone concentration time series and the relationship between these two hormones over time. Time series of integrated 10-min ACTH and corticosterone concentrations were collected over 4 h at the time of diurnal activation and analyzed in the time and frequency domains. The time of onset of the diurnal surge occurred 20 min later in old rats, and the ratio of corticosterone to ACTH was less at the time of onset and peak of the diurnal surge. Corticosterone levels were lower in old rats and mean ACTH and corticosterone levels were correlated in young but not old rats, as were maximum levels of the two hormones. Cross-correlation of ACTH and corticosterone time series and comparison of spectra were consistent with smoother fluctuations in plasma corticosterone in old animals with less variability at time scales less than 55 min. We conclude that age may be associated with a delay in diurnal activation of the HPA axis, a loss of sensitivity of adrenal corticosterone secretion to plasma ACTH levels, and a relative loss of high frequency variability in the corticosterone signal, as seen in many physiological systems with age.

Glucocorticoid regulation of mRNA encoding (Na+K) ATPase alpha 3 and beta 1 subunits in rat brain measured by in situ hybridization

The effect of glucocorticoids on (Na+K)ATPase mRNA synthesis was studied in 19 brain areas of adrenalectomized (ADX) rats untreated or receiving dexamethasone (DEX). For in situ hybridization, we employed a [35S]oligonucleotide probe for the alpha 3-subunit isoform, and a [3H]cDNA coding for the beta 1-subunit of the enzyme. Mean levels of grain density for the alpha 3 subunit mRNA of DEX-treated rats were significantly higher by a 't' test in medial septum, amygdala lateralis (AL) and medialis (AME), gyrus dentatus, CA4 hippocampal area, substantia nigra and periventricular gray, compared to untreated rats. For the beta 1-subunit, mean levels after DEX were significantly higher in AL and lateral preoptic area. In addition, the Kolmogorov-Smirnov test applied to frequency histograms of neuronal densities indicated a coordinate increase in alpha 3 and beta 1-subunit mRNA expression for the CA2 subfield and preoptic area medialis (POA MED). We conclude that (1) glucocorticoids are positive modulators of (Na+K)ATPase mRNA; (2) analysis of frequency histograms suggests that glucocorticoids promote in a few regions (AL, POA MED, CA2 subfield) a coordinate increase in the biosynthesis of the alpha 3 and beta 1-subunit mRNA. In 11 other areas stimulation occurs for one subunit mRNA only, whereas 5 areas were insensitive to glucocorticoid effects on this enzyme.

Hippocampal glucocorticoid receptor expression in the tree shrew: regulation by psychosocial conflict

1. This study was conducted to determine whether chronic psychosocial conflict alters the expression of glucocorticoid receptor (GR) mRNA in the hippocampus of male tree shrews (Tupaia belangeri). 2. To generate probes for the in situ hybridization, the tree shrew GR gene was partly cloned. There was a 90% homology between the deduced amino acid sequence of the cloned tree shrew GR and that of the corresponding human GR sequence. 35S-Labeled riboprobes which had been transcribed from the tree shrew GR clone hybridized to pyramidal neurons in all subregions of the tree shrew hippocampal formation and to granule neurons in the dentate gyrus. 3. After in situ hybridization, the expression of GR mRNA was semiquantitatively determined by counting silver grains over single neurons of the hippocampal formation of psychosocially stressed tree shrews and control animals. After 12 days of social conflict, the number of silver grains in the CA1 and CA3 pyramidal neurons of stressed animals was significantly lower than in controls. No statistically significant differences in mRNA expression were observed in the pyramidal neurons of the subiculum and in the granule neurons of the dentate gyrus. 4. The present results suggest that psychosocial stress leads to a site-specific down-regulation of hippocampal GR via modification of mRNA expression.

Effects of deoxycorticosterone acetate (DOCA) and aldosterone on Sar1-angiotensin II binding and angiotensin-converting enzyme binding sites in brain

1. It is known that regulation of salt appetite is a complex behavior controlled in the brain by interaction of mineralocorticoids (MC) and angiotensin II (ANGII). To investigate the effects of MC on ANGII receptors and ANGII synthesis, we have studied two models of salt appetite control. 2. In the first model, doses of DOCA sufficient to induce salt appetite of intact rats were given. In the second one, we studied the effects of aldosterone (ALDO) in doses sufficient to suppress salt appetite developed by prior adrenalectomy (ADX). 3. Binding to ANGII receptors was determined in brain sections incubated with 3 nM [125I]Sar1 ANGII, exposed to [3H]Hyperfilm with an optical density of autoradiograms measured by computerized densitometry. Sar1-ANGII binding was increased by DOCA treatment in the median preoptic nucleus (MnPO) and subfornical organ (SFO) but not in the paraventricular nucleus (PVN) in comparison to vehicle-treated rats. ALDO treatment was without effect on the MnPO but increased ANGII binding in the SFO and PVN. Neither hormone affected binding in the median eminence or anterior pituitary (AP). 4. In contrast to effects on Sar1-ANGII binding in selected areas, [125I]351A binding to angiotensin-converting enzyme (ACE) was unchanged by DOCA or ALDO administration in the SFO, caudate putamen, AP, or posterior pituitary. 5. These findings suggest that MC modulation of the renin-angiotensin system is exerted at the central, and not at the pituitary level. ANGII receptors were modulated in a dose- and region-specific manner: while DOCA may promote their actions upon the MnPO and SFO, ALDO actions may occur at the PVN and SFO. This mechanism may not require increased generation of ANGII in the brain or pituitary.