Subcellular location of unoccupied and occupied glucocorticoid receptor by a new immunohistochemical technique

Recent immunohistochemical studies suggest that the unoccupied glucocorticoid receptor (GR) is cytoplasmic and that the ligand causes its translocation into the target cell nucleus. The subcellular location of GR is especially interesting in that other members of the steroid receptor superfamily appear to be nuclear. The intracellular distribution of GR was studied immunohistochemically using a new freeze-drying and vapor fixation method which eliminates the protein diffusion and redistribution possibly caused by liquid fixation techniques. We used two monoclonal antibodies against rat liver GR. Dried samples of the adrenalectomized rat brain and uterus were fixed in p-benzoquinone vapor for 3 h at 60°C and embedded in paraffin. Sections were stained with a biotinylated mouse monoclonal GR antibody using the avidin-biotin-peroxidase complex. Both unoccupied and occupied GR were found in the nucleus of the target cells, fibroblasts in the uterus and nerve cells in the cortex of the brain. The staining was saturated with the cytosol of cos cellls transfected with GR. No cytoplasmic staining was seen even 2 days after adrenalectomy. In conclusion we propose that GR is also located in the nucleus independently of occupation.     

Ultrastructural localization of glucocorticoid receptor (GR) in hypothalamic paraventricular neurons synthesizing corticotropin releasing factor (CRF)

Summary Corticotropin releasing factor (CRF) synthesizing neurons, located in the hypothalamic paraventricular nucleus (PVN), are the main central regulators of the pituitary-adrenal cortex endocrine axis. The hormone production and release of CRF-synthesizing neurons is regulated by neuronal messages and feedback action(s) of glucocorticoids secreted by the adrenal gland. In order to characterize the latter mechanism, glucocorticoid receptor (GR)-immunoreactive (IR) sites were studied in hypothalamic paraventricular neurons of intact, long-term adrenalectomized, and adrenalectomized plus glucocorticoid treated animals, by means of ultrastructural immunocytochemical labelling. In intact animals, glucocorticoid receptor immunoreactivity was found predominantly in the nuclei of parvocellular neurons. Following adrenalectomy GR-immunoreactivity was localized in the cytoplasm of the cells, and there was a concomitant disappearance of the label from the nuclei. After corticosterone administration to adrenalectomized animals, GR-IR sites were again concentrated within the cell nuclei. Immunocytochemical double labelling studies performed on adrenalectomized plus corticosterone-replaced animals demonstrated glucocorticoid receptor-IR sites in the cell nuclei of parvocellular paraventricular neurons that expressed CRF-immunoreactivity in their cytoplasm.These ultrastructural data indicate that the intracellular location of glucocorticoid receptor is dependent on the availability of glucocorticoids by the neurons. The simultaneous expression of GR- and CRF-immunoreactivity in parvocellular paraventricular neurons supports the concept of a direct feedback action of glucocorticoids upon CRF-synthesizing neurons.Supported by NIH Research Grants NS19266 (W.K.P. and Zs.L.), NS20832 (M.C.B.) and a joint grant (INT-8703030) awarded by the National Science Foundation and the Hungarian Academy of Sciences (Zs.L. and W.K.P.). R.M.U. is a recipient of NIMH Pre-doctoral Fellowship and M.C.B. an NIH Research Carcer Development Award     

Central peptidergic neurons as targets for glucocorticoid action. Evidence for the presence of glucocorticoid receptor immunoreactivity in various types of classes of peptidergic neurons.

By means of double immunolabeling procedures it has been possible to demonstrate glucocorticoid receptor (GR) immunoreactivity (IR) in large numbers of various peptidergic neurons of the brain including neurons containing gastrointestinal peptides, opioid peptides, and peptides with a hypothalamic hormone function. For each peptide system, however, marked heterogeneities exist among brain regions. Thus, in the neocortex and the hippocampal formation most of the brain peptide neurons lack GR IR, while the same types of peptide neurons in the arcuate and paraventricular nucleus [e.g. neuropeptide Y (NPY), somatostatin (SRIF) and the cholecystokinin (CCK) neurons] possess strong GR IR. Furthermore, in the arcuate, parvocellular part of the paraventricular nuclei and the central amygdaloid nucleus practically all the peptidergic neurons are strongly GR IR, while in the lateral hypothalamus, mainly the neurotensin (NT) and galanin (GAL) IR neurons are GR IR. These marked differences among areas probably reflect functional differences dependent upon their participation in stress regulated circuits. All the paraventricular NT, corticotropin-releasing factor (CRF), growth hormone-releasing factor (GRF), thyrotropin-releasing hormone (TRH) and SRIF IR neurons appear to contain GR IR, while the luteinizing hormone-releasing hormone (LHRH) IR neurons lack GR IR, underlying the importance of glucocorticoids (GC) in controlling endocrine function. Finally, the GC may influence pain and mood control mainly via effects on enkephalin (ENK) neurons especially in the basal ganglia (mood) and on all beta-endorphin (beta-END) neurons of the arcuate nucleus, while most of the dynorphin neurons are not directly controlled by GC.     

Ah receptor for 2,3,7,8-tetrachlorodibenzo-p-dioxin. Codistribution of unoccupied receptor with cytosolic marker enzymes during fractionation of mouse liver, rat liver and cultured Hepa-1c1 cells

In early experiments Ah receptor appeared to be localized in cytosol when in its unoccupied state and it was thought that the receptor translocated into nuclei only when occupied by its ligands. However, a recent report [Whitlock and Galeazzi (1984) J. Biol. Chem. 259, 980-985] concluded that unoccupied Ah receptor in the intact cell was primarily located within the nucleus and that apparent 'cytosolic' Ah receptor was a redistribution artifact caused by fractionation of cells in large volumes of buffer. We examined the effect of buffer volume and ionic strength on apparent 'cytosolic' versus 'nuclear' distribution of unoccupied Ah receptor in liver from C57BL/6J mice and Sprague-Dawley rats as well as Hepa-1c1c9 cells in culture. In all three systems the Ah receptor appears to shift out of the nuclear fraction and into the cytosolic fraction as the volume of buffer is increased or when the ionic strength of the buffer is increased. In each system, however, the distribution of the Ah receptor was identical to the distribution of each of three standard cytosolic marker enzymes: aldolase, glucose-6-phosphate dehydrogenase and lactate dehydrogenase. Co-distribution of unoccupied Ah receptor with these cytosolic marker enzymes during fractionation at varied buffer volumes and ionic strengths makes it seem unlikely that the unoccupied receptor is predominantly a 'nuclear' component in intact cells. Marker enzyme data favor an interpretation that unoccupied Ah receptor is primarily cytoplasmic or that this soluble protein is in equilibrium between cytoplasm and nucleus.     

Ultrastructural analysis of estrogen receptor immunoreactive neurons in the medial preoptic area of the female rat brain

Neurons of the medial preoptic area were studied in the brain of the female rat by means of ultrastructural immunocytochemistry using a monoclonal antibody generated against purified estrogen receptor (ER), in order to delineate the morphological correlates of estrogen feedback mechanisms. In addition to the preoptic area, the bed nucleus of the stria terminalis, the arcuate and ventromedial nuclei of the hypothalamus exhibited an intense labelling for estrogen receptor. At the light microscopic level, the cell nuclei were immunoreactive. No major alterations were detected in the ER expression of medial preoptic neurons sampled during the estrous cycle, but proestrous rats did exhibit a slightly increased intensity of staining. At the ultrastructural level, the ER immunoreactivity was primarily confined to the nuclei and associated with the chromatin. Long term steroid deprivation elicited by either ovariectomy or ovariectomy plus adrenalectomy resulted in a marked intensity of nuclear labelling. This pattern was not influenced by acute estradiol replacement. These morphological data indicate that neurons of the medial preoptic area have the capacity to detect estrogens via receptor mechanisms and that changes in the level of the circulating ligand are manifested in an alteration in the staining for the estrogen receptor. The study also supports the revised concept of estrogen receptor action by demonstrating the presence of receptors in the nuclei of the cells, whether or not they are occupied by their ligand.     

Structure and dynamics of the estrogen receptor

To evaluate the structure and function of estrogen receptor (ER) in various mammalian systems, the cytosolic forms of receptor from calf uterus and from MCF-7 human breast cancer cells have been purified to virtual homogeneity by sequential selective adsorption to estradiol-Sepharose and heparin-Sepharose. In both cases, the purified steroid-receptor complex appears to exist as an activated 5S homo- or heterodimer of mol. wt 65,000 (4S) steroid-binding subunits. Purified ER has high affinity for DNA and serves as a substrate for phosphorylation by a purified rat brain kinase. Several monoclonal antibodies prepared against affinity-purified MCF-7 cytosol ER have been used to localize receptor by an indirect immunoperoxidase technique in fixed, frozen sections of human breast tumors, human uterus, rabbit uterus and in other mammalian reproductive tissues and cancers, as well as in fixed MCF-7 cell cultures and in paraffin-embedded sections of breast tumors and human endometrium. In all cases, we have observed only nuclear localization of immunoreactive receptor in tissues and whole cells, even under conditions in which virtually all of the receptor is found in a low-salt extract (cytosol) of the target cells. Treatment of cells or tissues in vivo or in vitro with estradiol alters the intensity but not the distribution of specific staining for ER. By immunoelectron microscopy, receptor was localized in the euchromatin, but not in the marginated heterochromatin or nucleoli of MCF-7 nuclei and epithelial and stromal nuclei of postmenopausal human endometrium. These observations suggest that the majority of the unoccupied receptor may actually reside in the nucleus, rather than in the cytoplasm as previously thought. Thus, hormone action may involve binding of the steroid directly to receptor loosely associated with nuclear components, followed by conversion of the steroid-receptor complex to an activated form which becomes more tightly associated with chromatin.     

Studies on the molecular mechanism of the translocation of estrogen receptor from the cytoplasm into the nucleus

A detailed study of the molecular mechanism of the translocation of estrogen receptor (ER) from the cytoplasm into the nucleus was undertaken in an in vitro system of porcine uterus. The capabilities of vero-ER . E (basic ER molecular bound with estradiol) (sedimentation coefficient 4.5S; Stokes radius 44 A) and the complexes ["5S" ER . E, (vero-ER . E) . (component A); "6S" ER . E, (vero-ER . E) . (component B)6; "8S" ER . E, (vero-ER . E) . (component B)6 . (component A)] with ER-binding factors (ERBFs) to translocate into the isolated nuclei were estimated by subtracting the amounts of ER adsorbed by the nuclear envelopes from those of ER bound to the whole nuclei. The results strongly supported our previous assumption that vero-ER . E translocates into the nuclei, and the complexes with ERBFs do not. The results suggested also that the binding site of vero-ER to ERBFs is required to be unoccupied in the process of the translocation of ER from the cytoplasm into the nucleus. The presence of a cytoplasmic factor (component C) which binds specifically with "5S" ER . E under low salt conditions was indicated. The complex, ("5S" ER . E) . (component C), was shown to possess relatively high affinity towards nuclear envelopes, but not to translocate into the nuclei.(ABSTRACT TRUNCATED AT 250 WORDS)     

雌激素受体α和β免疫反应在雄性和雌性棕色田鼠脑区分布的比较

单配制和多配制动物社会行为有差异,这些差异可能与雌激素受体类型有关（ERs）。虽然多配制大鼠和小鼠中枢神经雌激素受体α（ERα）和β（ERβ）免疫反应在大脑的分布已有报道,单配制雄性草原田鼠中枢神经ERα的分布也有报道,但单配制田鼠ERα和（或）ERβ在雌性和雄性分布差异未见报道。本研究对雄性和雌性棕色田鼠前脑区域ERα和ERβ免疫反应（IR）细胞数量进行比较。研究结果表明：（1）免疫反应主要分布在细胞核中。 （2）ERα-IR和ERβ-IR细胞广泛分布于整个雌性和雄性前脑区域,在许多脑区表达有重叠。然而,不同受体在雌雄不同脑核中的分布数量是不同的。（3）ERα 和ERβ的分布存在性别差异。例如,雌性ERα在视前核中部（MPN）,终纹床和（BNST）和杏仁内侧核（MeA）比雄性多,相反雄性ERβ在MPN和BNST比雌性多。这些研究结果可能为我们理解如何通过ERα和ERβ调节动物的社会行为,及雌性和雄性社会行为的差异提供一个重要的神经解剖学基础。     

Estrogen receptor and progesterone receptor-immunoreactive cells are not co-localized with gonadotropin-releasing hormone in the brain of the female mink (Mustela vison)

The distribution of gonadal steroid (estrogen, progesterone) receptors in the brain of the adult female mink was mapped by immunocytochemistry. Using a monoclonal rat antibody raised against human estrogen receptor (ER), the most dense collections of ER-immunoreactive (IR) cells were found in the preoptic/anterior hypothalamic area, the mediobasal hypothalamus (arcuate and ventromedial nuclei), and the limbic nuclei (amygdala, bed nucleus of the stria terminalis, lateral septum). Immunoreactivity was mainly observed in the cell nucleus and a marked heterogeneity of staining appeared from one region to another. A monoclonal mouse antibody raised against rabbit uterine progesterone receptor (PR) was used to identify the PR-IR cells in the preoptic/anterior hypothalamic area and the mediobasal hypothalamus (arcuate and ventromedial nuclei). This study also focused on the relationship between cells containing sex-steroid receptors and gonadotropin-releasing hormone (GnRH) neurons on the same sections of the mink brain using a sequential double-staining immunocytochemistry procedure. Although preoptic and hypothalamic GnRH neurons were frequently in close proximity to perikarya containing ER or PR, they did not themselves possess receptor immunoreactivity. The present study provides neuroanatomical evidence that GnRH cells are not the major direct targets for gonadal steroids and confirms for the first time in mustelids the results previously obtained in other mammalian species.     

Ultrastructural analysis of estrogen receptor immunoreactive neurons in the medial preoptic area of the female rat brain

Summary Neurons of the medial preoptic area were studied in the brain of the female rat by means of ultrastructural immunocytochemistry using a monoclonal antibody generated against purified estrogen receptor (ER), in order to delineate the morphological correlates of estrogen feedback mechanisms. In addition to the preoptic area, the bed nucleus of the stria terminalis, the arcuate and ventromedial nuclei of the hypothalamus exhibited an intense labelling for estrogen receptor. At the light microscopic level, the cell nuclei were immunoreactive. No major alterations were detected in the ER expression of medial preoptic neurons sampled during the estrous cycle, but proestrous rats did exhibit a slightly increased intensity of staining. At the ultrastructural level, the ER immunoreactivity was primarily confined to the nuclei and associated with the chromatin. Long term steroid deprivation elicited by either ovariectomy or ovariectomy plus adrenalectomy resulted in a marked intensity of nuclear labelling. This pattern was not influenced by acute estradiol replacement.These morphological data indicate that neurons of the medial preoptic area have the capacity to detect estrogens via receptor mechanisms and that changes in the level of the circulating ligand are manifested in an alteration in the staining for the estrogen receptor. The study also supports the revised concept of estrogen receptor action by demonstrating the presence of receptors in the nuclei of the cells, whether or not they are occupied by their ligand.Supported by grants from the IBRO/MacArthur Foundation Network Grant, the National Science Foundation (NSF INT 8703030), the Hungarian Academy of Sciences (OTKA 104), the National Institutes of Health (NS 19266), the National Foundation of Technical Development (OKKFT Tt 286/1986) and the Well-come Trust (14685/1.5)     

Rapid important paper on the cellular localization and distribution of estrogen receptors in the rat tel- and diencephalon using monoclonal antibodies to human estrogen receptor

By means of indirect immunoperoxidase procedures using the biotin- avidin method in combination with monoclonal antibodies to the human estrogen receptor it has been possible to map out distinct populations of nerve cells possessing nuclear estrogen immunoreactivity in rat brain. High densities of strongly estrogen immunoreactive nerve cells were especially observed in the medial preoptic area and the bed nucleus of the stria terminalis but also in the magnocellular part of the arcuate nucleus, the ventral premammillary nuclei and in the area between the medial and lateral hypothalamus including the lateral component of the ventromedial hypothalamic nucleus. Similar results were obtained in the male and female adult brain. Following castration of the male and female adult rat, the nuclear estrogen immunoreactivity did not change its location but the degree of immunoreactivity was increased. Administration of 50 μg/kg of estrogen benzoate in the castrated animals induced a marked disappearence of the estrogen immunoreactivity in the nerve cells in all regions analyzed. The results give further evidence for the existence of a selective population of estrogen receptor containing neurons in the female and male brain of adult animals and that the estrogen free receptor is associated with the nucleus. Upon activation the nuclear estrogen receptors appear to loose this immunoreactivity probably due to a change in the conformation of the receptor protein.     

Growth factor regulation of cell growth and proliferation in the nervous system

This article discusses a novel intracrine mechanism of growth-factor action in the nervous system whereby fibroblast growth factor-2 (FGF-2) and its receptor accumulate in the cell nucleus and act as mediators in the control of cell growth and proliferation. In human and rat brain the levels and subcellular localization of FGF-2 differ between quiescent and reactive astrocytes. Quiescent cells express a low level of FGF-2, which is located predominantly within the cytoplasm. In reactive astrocytes, the expression of FGF-2 increases and the proteins are found in both the cytoplasm and nucleus. In glioma tumors, FGF-2 is overexpressed in the nuclei of neoplastic cells. Similar changes in FGF-2 expression and localization are found in vitro. The nuclear accumulation of FGF-2 reflects a transient activation of the FGF-2 gene by potentially novel transactivating factors interacting with an upstream regulatory promoter region. In parallel with FGF-2, the nuclei of astrocytes contain the high-affinity FGF-2 receptor, FGFR1. Nuclear FGFR1 is full length, retains kinase activity, and is localized within the nuclear interior in association with the nuclear matrix. Transfection of either FGF-2 or FGFR1 into cells that do not normally express these proteins results in their nuclear accumulation and concomitant increases in cell proliferation. A similar regulation of nuclear FGF-2 and FGFR1 is observed in neural crest-derived adrenal medullary cells and of FGF-2 in the nuclei of cerebellar neurons. Thus, the regulation of the nuclear content of FGF-2 and FGFR1 could serve as a novel mechanism controlling growth and proliferation of glial and neuronal cells.     

A modified immunohistochemical procedure for the detection of estrogen receptor in mouse tissues

A horseradish peroxidase (HRP) labeled antibody method was developed for use with a monoclonal antibody to detect estrogen receptor (ER) in mouse tissue. Combined use of HRP labeled F(ab')2 fragment absorbed with mouse liver protein to minimize background staining and imidazol-DAB reaction gave the most reliable and sensitive immunostaining. The method was applied to uterine, vaginal, pituitary and liver tissues in ovariectomized adult mice. In uterus and vagina, ER was recognized in nuclei of epithelial cells, stromal cells and smooth muscle cells of the muscle layer and blood vessels. Liver tissue showed positive nuclear immunostaining in parenchymal cells; however, no reaction was present in endothelial cells, Kupffer cells, bile ductal cells, and smooth muscle cells of blood vessels. ER was localized in the nuclei of anterior pituitary cells while weak reaction was also recognized in cells of the intermediate lobe. No staining was detected in the posterior pituitary. Results demonstrate that both occupied and unoccupied ER are localized in the cell nucleus from several target tissues. Weak immunostaining in samples could not be enhanced by multiple procedures. It is suggested that nuclear ER is partially hidden by nuclear components such as nucleic acid and chromatin proteins.     

Natural variation in paternal behavior is associated with central estrogen receptor alpha and oxytocin levels

In monogamous mammals paternal care plays an important role in the neural and behavioral development of offspring. However, the neuroendocrine mechanisms underlying paternal behavior remain poorly understood. Here, we investigate the association between natural variation in paternal responsiveness and central levels of oxytocin (OT) and estrogen receptor alpha (ERα). We used the frequency of licking and grooming behavior to distinguish low paternal responsiveness and high paternal responsiveness in virgin mandarin voles (Microtus mandarinus). Males that engaged in high paternal behavior had elevated levels of OT immunoreactive neurons in the paraventricular nuclei of the hypothalamus and supraoptic nuclei of the hypothalamus compared with males that displayed low paternal behavior. Likewise, males of high paternal responsiveness had more ERα immunoreactive neurons in the medial preoptic area, bed nucleus of the stria terminalis, arcuate nucleus of the hypothalamus and medial amygdaloid nucleus compared to low responsive males. The level of ERα immunoreactive neurons in the ventromedial hypothalamic nucleus was lower in highly paternal males compared to less paternal males. These results suggest that natural variation in paternal responsiveness may be directly related to variation in central OT and ERα.     

Opiocortin and catecholamine input to CRF-immunoreactive neurons in rat forebrain

Neural input to distinct and separate populations of CRF-immunoreactive (ir) neurons in rat forebrain was investigated. The relationship of opiocortin and/or catecholamine fibers to different groups of CRF-containing neurons was elucidated using single and dual labeling immunocytochemical procedures. Antibodies to CRF, ACTH(1–39) and the catecholamine synthesizing enzymes which are tyrosine hydroxylase (TH), dopamine β-hydroxylase (DBH) and phenylethanolamine-N-methyltransferase (PNMT) were utilized. CRF-ir neuronal populations are localized predominantly in the following regions of rat forebrain: bed nucleus of stria terminalis, medial preoptic area, suprachiasmatic and paraventricular (PVN) nuclei of hypothalamus and central nucleus of amygdala. The present study demonstrates that CRF-ir neuronal groups in rat forebrain are not homogenous in that each population received a characteristic neural input. CRF-ir neurons in the PVN received a dense input of ACTH-, TH-, DBH-, and PNMT-ir fibers. In contrast, CRF-ir neurons in the central nucleus of amygdala are colocalized predominantly with TH-ir fiber/terminals. In the ventral portion of the bed nucleus of stria terminalis, TH-, ACTH- and DBH-ir fibers are demonstrated in close anatomical proximity to CRF-containing perikarya; in the dorsal portion of this nucleus, TH-ir fiber/terminals are colocalized with CRF-ir neurons. In the suprachiasmatic nucleus, neither opiocortin- nor catecholamine-immunostained fibers are observed in association with CRF-ir neurons. Our data suggest that there is a transmitter specificity of neural input to each CRF-ir neuronal population in rat forebrain.