Somatostatin-containing neuron systems in the rat hypothalamus: retrograde tracing and immunohistochemical studies

By employing a combination of the immunohistochemistry for somatostatin (SRIF) and retrograde tracing with biotinylated wheat germ agglutinin (b-WGA) injected into the posterior pituitary (group 1) or into the median eminence (group 2), functional topography of hypothalamic SRIF neurons was determined in the rat hypothalamus. In group 1, large numbers of WGA-labeled neurons appeared in the rostral periventricular region and in the magnocellular division of the paraventricular and supraoptic nuclei; none of them were SRIF immunoreactive. In group 2, WGA-labeled neurons were numerous in the rostral periventricular region, the parvicellular division of the paraventricular nucleus, and the arcuate nucleus; most of the WGA-labeled neurons in the rostral periventricular region and some in the paraventricular nucleus were SRIF immunoreactive, but none in the arcuate nucleus showed immunoreactivity for SRIF. It is concluded that, in the rat hypothalamus, the locations of neurons containing hypophysiotrophic SRIF are confined within the rostral periventricular region and the parvicellular paraventricular nucleus. Our results do not support previous suggestions that SRIF immunoreactive axons innervate the posterior lobe of the pituitary.     

Effects of periventricular lesions on the release of somatostatin during perifusion

Hypothalamic periventricular (PV) nucleus lesions reduce median eminence (ME) SRIF content by 80% without affecting non-stress plasma growth hormone (GH) levels or the GH response to stress. Our aim was to study the effects of PV lesions on SRIF released during perifusion of preoptic-anterior hypothalamic (PO-AH) tissue. Female rats received anterior or posterior PV lesions; sham-lesioned and intact rats served as controls. Non-stress and stress plasma GH levels were similar in all groups at 2,4 and 16 weeks after surgery. At 18 weeks after surgery, the perifussed PO-AHs of the PV- and sham-lesioned rats released similar amounts of SRIF, and these were higher (P <0.001) than amounts released from PO-AHs of intact rats. The PO-AHs from all groups showed similar increases in SRIF release after 56 mM K⁺. Two rats were chosen randomly from each group to asses ME SRIF content; PV lesions caused almost 80% depletion of SRIF, sham lesions did not. These results confirm that most SRIF neurons in the PV nucleus and 80% of ME SRIF content are not essential for the contrl of GH levels under non-stress conditions of for the GH responses to stress and indicate that PV or sham lesions in the rostal forebrain enhance in vitro SRIF release, perhaps from neurons outside the PV nucleus.     

Regulation of somatostatin synthesis by GABAA receptor stimulation in mouse brain.

Neuroanatomical data have documented the existence of synaptic contacts between gamma-aminobutyric acid (GABA) terminals and periventricular hypothalamic somatostatin (SRIF) neurons. In other brain regions, like the cortex or hippocampus, GABA and SRIF are colocalized in short interneurons. These observations suggest that GABA modulates SRIF neuronal activity. In order to test this hypothesis, we studied the effects of the in vivo stimulation of the GABAA receptor (muscimol, 0.75 mg/kg + diazepam, 2.5 mg/kg) on SRIF content and preproSRIF mRNA levels, in mouse brain. Chronic (7 days), but not acute, treatment induced a 38% decrease in hypothalamic SRIF content (as estimated by RIA), a 20% decrease in cortex and no effect in the striatum. The decrease in hypothalamic and cortical SRIF levels lasted until 24 h after cessation of the treatment. In the hypothalamus, prosomatostatin mRNA levels were estimated by Northern blot analysis using a 32P-labeled 45-mere oligoprobe. ProSR1F mRNA hypothalamic levels were equally (48%) decreased by the acute and chronic treatments and remained lower than controls 48 h after the last injection. Quantitative in situ hybridization was used to examine the regional distribution of GABA-induced acute inhibition of proSR1F mRNA densities, using the same oligomere labeled with 35S. ProSR1F mRNA levels were decreased by 35% in the periventricular hypothalamic nucleus. In contrast, no significant modification was observed in cortex, striatum and hilus of the dentate gyrus of the dorsal hippocampus. The present data demonstrate a regionally selective inhibitory action of GABA, mediated by GABAA receptors stimulation, on the biosynthetic mechanisms of the long projecting neuroendocrine SRIF neurons of the anterior periventricular nucleus of the hypothalamus.     

Enkephalin-immunoreactive neurons in the parvicellular subdivisions of the paraventricular nucleus project to the external zone of the median eminence.

The enkephalin-immunoreactive neurons that project to the external zone of the median eminence were identified on thin paraffin and thick vibratome sections using a combination of retrograde labeling with peripherally administered Fluoro-Gold and immunocytochemistry. The vast majority of the enkephalin-immunoreactive neurons that project to the external zone of the median eminence (ME) reside in the paraventricular nucleus (PVN) of the hypothalamus. Within the PVN, the majority of these hypophysiotropic neurons are located in the medial parvicellular subdivision, while a smaller number can be detected in the anterior and the periventricular subdivisions. Although many enkephalin-immunoreactive neurons are present in other hypophysiotropic areas of the hypothalamus, such as the medial preoptic area, the anterior periventricular area, and the arcuate nucleus, only a few of these can be retrogradely labeled from the ME. These results provide morphological evidence for the key role of paraventricular enkephalin-immunoreactive neurons in the regulation of neuroendocrine functions.     

Somatostatin in catecholamine-rich nuclei of the brainstem

Somatostatin (SRIF) concentrations in catecholamine-rich nuclei of the rat brainstem were measured by radioimmunoassay. The study was performed both in control or sham operated animals and after transecting the major projections of hypothalamic SRIF-containing neurons. Concentrations of the peptide were found to be relatively high in the locus coeruleus, the parabrachial nucleus and the nucleus of the solitary tract; they were intermediate in the lateral reticular nucleus (A1 cell group) and low in the substantia nigra. Transection of hypothalamic periventricular efferents resulted in a 58% depletion of SRIF content in the locus coeruleus, while concentrations of the peptide in other areas were unaffected. Transection of the medial forebrain bundle at the level of the lateral hypothalamus decreased SRIF content by 55% in the substantia nigra, but not in the other nuclei tested. It is concluded that the hypothalamus contributes significantly to the somatostatinergic innervation of the locus coeruleus and the substantia nigra, whereas SRIF in the other nuclei is intrinsic or originates outside the hypothalamus.     

Somatostatin connections between the hypothalamus and the limbic system of the rat brain

Somatostatin (SRIF) content of several brain structures was evaluated by radioimmunoassay in rats bearing various types of hypothalamic transections, as well as lesions of the amygdala. Analysis of the regional changes in SRIF concentrations after surgery suggest the following conclusions: (1) hypothalamic somatostatinergic neurons project to the limbic system, with the exception of the amygdaloid nuclei; (2) the olfactory tubercle, the lateral septal nucleus, the habenula and probably the hippocampus receive somatostatin projections from periventricular SRIF-containing cells; (3) somatostatin-containing fibers take a lateral course after leaving periventricular cells and join the medial forebrain bundle; (4) somatostatin innervation of the amygdala seems to be intrinsic.     

Hypophysiotrophic TRH-producing neurons identified by combining immunohistochemistry for pro-TRH and retrograde tracing

To determine hypophysiotrophic thyrotropin-releasing hormone (TRH)-producing neurons in the rat hypothalamus, we employed a combination of the immunohistochemistry for TRH prohormone (pro-TRH) and the retrograde tracing of neurons that project to the median eminence (ME) by injecting biotinylated wheat germ agglutinin (WGA) into the ME. In intact rats, immunoreactive pro-TRH-positive neurons occurred in the parvicellular paraventricular nucleus (parvi-PVN), basal part of the anterior and lateral hypothalamus, perifornical area and dorsomedial nucleus, especially accumulating in the parvi-PVN. Twenty-four hours after injection of the WGA into the middle portion of the ME, we found neurons that incorporated the lectin in the anterior periventricular area, the PVN, and the arcuate nucleus. When we examined serial sections consecutively stained with anti-WGA, anti-pro-TRH, and anti-WGA, most of the pro-TRH-labeled neurons in the medial parvi-PVN and a part of the neurons in the anterior periventricular area and in the anterior, lateral, and dorsal parvi-PVN appeared to incorporate WGA. These neurons may correspond with the hypophysiotrophic TRH-synthesizing neurons in the rat hypothalamus.     

Alz-50 immunoreactivity in the hypothalamus of the normal and Alzheimer human and the rat

Alz-50 is a monoclonal antibody recognizing a 68 kilodalton protein that is abundant in Alzheimer's disease (AD) but not detectable by immunoblotting methods in normal brains. When used for immunohistochemistry in AD cortex, Alz-50 recognizes large numbers of neurofibrillary tangles (NFT), neuritic plaques, and some neurons that show no evidence of neurofibrillary degeneration by conventional histopathological staining methods. Alz-50 immunoreactivity is described at the light and electron microscopic levels in the hypothalamus of brains obtained at autopsy from normal and AD subjects. Alz-50 immunoreactivity in the rat hypothalamus is also described. A well-defined population of Alz-50 immunoreactive hypothalamic neurons was identified in both the normal human and rat. At the light microscopic level in the normal human, immunoreactive neurons were most concentrated in the periventricular region, but were also scattered throughout the arcuate nucleus (ARC), lateral hypothalamic area, and tuberal region. Immunoreactive fibers were seen in the periventricular region, dorsal division of the ventromedial nucleus (VMNd), ARC, and external layer of the median eminence (ME). In the rat, reactive neurons were seen only in the periventricular region, and reactive fibers were seen in the periventricular zone, medial preoptic nuclear complex, suprachiasmatic nucleus, VMNd, ARC, and external layer of the ME. Ultrastructurally, all immunoreactivity in the normal human and rat hypothalamus was associated with intraneuronal vesicles. In the AD hypothalamus, Alz-50 identified numerous senile plaques and NFT in addition to the cells and fibers that were stained in the normal brains. Immunoreactive plaques and NFT were most numerous in regions previously reported to undergo neurofibrillary degeneration. At the ultrastructural level, the immunoreactivity in the AD hypothalamus was associated with filaments as well as vesicles. The significance of the selective staining of a specific population of vesicles by Alz-50 is unknown; however, the present results suggest that it is independent of AD pathology.     

A Subset of Estrogen Receptor-Containing Neurons Project to the Median Eminence in the Ewe

The neural pathways responsible for conveying the steroid feedback signals that ultimately affect reproductive neuroendocrine function remain largely undefined. One possibility involves a direct projection from estrogen receptor (ER)-containing neurons to the median eminence (ME), a site of neuroendocrine peptide release. To examine this possibility, 8 ewes received stereotaxic injections of the retrograde neuronal tract-tracing compound cholera toxin-β subunit (CTβ) into the ME. Neurons sending projections to the ME and containing ER were identified using a dual-label immunoperoxidase method. Double-labeled cells were found in distinct regions: (1) the ER-rich arcuate nucleus (ARC) that contained the greatest number of double-labeled cells, and (2) the organum vasculosum of the lamina terminalis (OVLT) which contained a very consistent, but low, number of double-labeled cells. While a fairly large number of retrogradely-labeled ARC neurons containing ER were identified, the majority of ER-containing ARC neurons were unlabeled and thus send projections elsewhere. Other regions containing high concentrations of ER-positive cells such as the medial preoptic area (MPOA), anterior hypothalamic area, and ventrolateral portion of the ventromedial hypothalamic nucleus, were devoid of double-labeled cells. Similarly, regions rich in neuroendocrine neurons such as the periventricular hypothalamus and paraventricular and supraoptic hypothalamic nuclei contained no double-labeled cells. These results suggest that modulation of neuroendocrine secretory activity may occur directly at the level of the ME by ER-containing neurons located within restricted regions of the hypothalamus and forebrain. However, the relatively low proportion of ER-containing neurons projecting to the ME suggests that the influence of estradiol upon neuroendocrine function also may include target sites other than the ME.     

Comparative distribution of neuropeptide-immunoreactive systems in the brain of the green molly, Poecilia latipinna

The comparative distribution of peptidergic neural systems in the brain of the euryhaline, viviparous teleost Poecilia latipinna (green molly) was examined by immunohistochemistry. Topographically distinct, but often overlapping, systems of neurons and fibres displaying immunoreactivity (ir) related to a range of neuropeptides were found in most brain areas. Neurosecretory and hypophysiotrophic hormones were localized to specific groups of neurons mostly within the preoptic and tuberal hypothalamus, giving fibre projections to the neurohypophysis, ventral telencephalon, thalamus, and brain stem. Separate vasotocin (AVT)-ir and isotocin (IST)-ir cells were located in the nucleus preopticus (nPO), but many AVT-ir nPO neurons also displayed growth hormone-releasing factor (GRF)-like-ir, and in some animals corticotrophin-releasing factor (CRF)-like-ir. The main group of CRF-ir neurons was located in the nucleus recessus anterioris, where coexistence with galanin (GAL) was observed in some cells. Enkephalin (ENK)-like-ir was occasionally present in a few IST-ir cells of the nPO and was also found in small neurons in the posterior tuberal hypothalamus and in a cluster of large cells in the dorsal midbrain tegmentum. Thyrotrophin-releasing hormone (TRH)-ir cells were found near the rostromedial tip of the nucleus recessus lateralis. Gonadotrophin-releasing hormone (GnRH)-ir cells were present in the nucleus olfactoretinalis, ventral telencephalon, preoptic area, and dorsal midbrain tegmentum. Molluscan cardioexcitatory peptide (FMRF-amide)-ir was colocalized with GnRH-ir in the ganglion cells and central projections of the nervus terminalis. Melanin-concentrating hormone (MCH)-ir neurons were restricted to the tuberal hypothalamus, mostly within the nucleus lateralis tuberis pars lateralis, and somatostatin (SRIF)-ir neurons were numerous throughout the periventricular areas of the diencephalon. A further group of SRIF-ir neurons extending from the ventral telencephalon into the dorsal telencephalon pars centralis also contained neuropeptide Y (NPY)-, peptide YY (PYY)-, and NPY flanking peptide (PSW)-like-ir. These immunoreactivities were, however, also observed in non-SRIF-ir cells and fibres, particularly in the mesencephalon. Calcitonin gene-related peptide (CGRP)-like-ir had a characteristic distribution in cells grouped in the isthmal region and fibre tracts running forward into the hypothalamus, most strikingly into the inferior lobes. Antisera to cholecystokinin (CCK) and neurokinin A (NK) or substance P (SP) stained very extensive, separate systems throughout the brain, with cells most consistently seen in the ventral telencephalon and periventricular hypothalamus. Broadly similar, but much more restricted, distributions of cells and fibres were seen with antisera to neurotensin (NT) and vasoactive intestinal peptide (VIP).(ABSTRACT TRUNCATED AT 400 WORDS)     

Fiber degeneration associated with hyperphagia-inducing knife cuts in the hypothalamus

The pattern of axonal degeneration associated with hyperphagia-producing hypothalamic knife transections was investigated using the Fink-Heimer method for staining of degenerating axons and their terminal endings. Histological analysis of silver-stained material after parasagittal knife cuts which result in hyperphagia and obesity revealed fiber degeneration coursing longitudinally in the medial forebrain bundle including the perifornical component to reach the nucleus accumbens, the diagonal band, the preoptic-anterior hypothalamic junction, the lateral hypothalamus, the zona incerta, the periventricular thalamus, the parafascicular thalamic nucleus, the substantia nigra pars compacta, the central gray matter, the ventral tegmental area of T'sai and the superior colliculus. The data obtained in the present study lend support to the suggestion that projections coursing in the medial forebrain bundle interconnect the anteriomedial hypothalamus and the midbrain tegmentum and may underlie the hyperphagia and obesity produced by hypothalamic knife cuts.     

Distribution of somatostatin in the mouse brain: Effects of neonatal MSG treatment

Immunocytochemical analysis using antisera generated against the brain peptide somatostatin (SRIF) was examined in the brain of normal mice and in mice with chemical lesions of the arcuate nucleus produced neonatally by the administration of monosodium glutamate (MSG). In the normal mouse brain, SRIF immunoreactivity was seen in perikarya of the preoptic and hypothalamic periventricular nuclei. The normal distribution of SRIF fibers was apparent in several hypothalamic nuclei including the arcuate nucleus and in the internal and external zones of the median eminence. Extrahypothalamic sites of SRIF immunoreactive neurons and fibers were also observed throughout the telencephalon.At 60 days of age, certain neuroendocrine deficiencies, including growth parameters and obesity, were apparent in MSG-treated newborn mice. Analysis of SRIF projections in the brain of MSG-treated mice demonstrated a neurotoxic effect on arcuate neurons and a loss of SRIF projections to this region as well. Other components of the SRIF system in brain appeared unaffected. SRIF fibers of the arcuate region seem to originate from neuronal perikarya of the periventricular nucleus suggesting that MSG-induced endocrine deficiencies may be due to SRIF interactions at the level of the arcuate nucleus.     

Distribution of tyrosine-hydroxylase (TH)-immunoreactive neurons in the diencephalon of the pigeon (Columba livia domestica)

The distribution of tyrosine-hydroxylase (TH)-immunoreactive cell bodies and fibers in the diencephalon has been investigated with immunohistological techniques in the pigeon. The results suggest that TH is present in a number of morphologically distinct neuronal systems. Preoptic and hypothalamic TH neurons were subdivided into a medial periventricular and a lateral group. The medial group starts with a rostral collection of small cells in the preoptic region. A significantly larger collection of TH neurons occupies the paraventricular nucleus (PVN) (stratum cellulare internum) and mainly consists of large multipolar cells. Further caudally, the main concentration of cells is in the hypothalamic posteromedial and the periventricular regions of the tuberoinfundibular (arcuate) nucleus. No TH neuron was found in the ventral and lateral parts of the tuberoinfundibular region, suggesting that the prominent tuberoinfundibular dopaminergic system described in mammals is absent in the pigeon. This further substantiated by the relative scarcity of TH immunoreactive fibers and varicosities in the neurohemal zone of the median eminence (ME). The caudalmost components of the medial group appear to be continuous with the large population of TH neurons distributed in the midline of the mesencephalon. Tyrosine-hydroxylase-immunopositive cells have not been found in the paraventricular organ. The lateral group consists of TH neurons loosely arranged in the lateral hypothalamus, including regions of the supraoptic nucleus and hypothalamic posterolateral nucleus. Tyrosine-hydroxylase containing neurons vary widely in size, shape, and dendritic arborization in each diencephalic region. However, it is possible to distinguish two main cell types. Small bipolar neurons with two simple arborizing dendrites were concentrated in the medial periventricular system. The second type of cell is large, multipolar with four to five branching dendrites. This latter cell type occurs mainly in the lateral system and in the PVN. Major fiber bundles containing TH immunoreactivity were identified in the lateral and periventricular hypothalamus. The paraventricular organ and the organum vasculosum laminae terminalis contained the densest arborization of fibers and varicosities. In the ME, dense innervation was found in the subependymal layer. Dense arborizations of TH positive fibers and varicosities were located in the septal nuclei and the paleostriatum augmentatum.     

Effects of lesions in the periventricular nucleus of the preoptic-anterior hypothalamus on growth hormone and thyrotropin secretion and brain somatostatin

Two experiments were performed to study the role of somatostatin (SRIF) neurons of the preoptic-anterior hypothalamic area (PO-AHA) in regulating growth hormone (GH) and thyrotropin (TSH) secretion in rats. Small lesions were placed in the periventricular (PV) zone and blood was collected at 24 h and 15 days after surgery. Blood samples were obtained at 3 min and at 15 min after ether exposure for assessing non-stress levels, respectively, of plasma GH and TSH. Non-stress blood samples were also collected at decapitation at 4 weeks. The brains from the first experiment were dissected and processed for measuring SRIF content in several regions. At 24 h and 15 days, non-stress GH and TSH levels were significantly elevated in rats with PV lesions. Stress-induced decrements in GH levels persisted in all groups. Although non-stress plasma GH and TSH levels returned to normal in lesioned rats at 4 weeks, SRIF content was decreased 83% in the median eminence and 33% in the hypothalamus. These results show that discrete lesions in the PV zone of the PO-AHA cause transient elevations in non-stress secretion of GH and TSH and that normal levels of such secretion can be reinstated despite reductions of SRIF in the median eminence and hypothalamus.     

Regional distribution of gastrin-releasing peptide- and somatostatin-like immunoreactivity in the rabbit hypothalamus

Regional distribution of gastrin-releasing peptide- (GRP) and somatostatin (SRIF)-like immunoreactivity in the discrete nuclei of the hypothalamus was examined in the rabbit according to Palkovits' microdissection method. GRP-like immunoreactivity (LI) was detected abundantly in the hypothalamus as compared with the cerebral cortex when measured by radioimmunoassay using the antiserum recognizing the C-terminal portion of synthetic porcine GRP. On gel-filtration chromatography of the hypothalamic extracts, two major peaks of GRP-LI were eluted; the peak with larger molecular size corresponded to synthetic porcine GRP1-27 and the smaller size to porcine GRP14-27. A concentration of GRP-LI was highest in the infundibular nuclei (IFN) as well as the ventromedial nuclei (VMN), and next high in the paraventricular nuclei (PVN), suprachiasmatic nuclei (SCN) and periventricular nuclei (PEV). The content of GRP-LI in the median eminence was not so much when compared with them. On the other hand, SRIF was localized in the highest concentration in the ME, followed by the VMN and IFN, as well as the PEV. The findings indicate that porcine GRP-LI exists in the hypothalamus of rabbits with characteristic regional distribution. Concurrent localization of GRP-LI and SRIF in some parts of the hypothalamus may suggest the interaction of both peptides in these areas under various physiological and pathological status.     

Activation of neurons in the hypothalamic dorsomedial nucleus via hypothalamic projections of the nucleus of the solitary tract following refeeding of fasted rats

We report that satiation evokes neuronal activity in the ventral subdivision of the hypothalamic dorsomedial nucleus (DMH) as indicated by increased c‐fos expression in response to refeeding in fasted rats. The absence of significant Fos activation following food presentation without consumption suggests that satiation but not craving for food elicits the activation of ventral DMH neurons. The distribution pattern of the prolactin‐releasing peptide (PrRP)‐immunoreactive (ir) network showed remarkable correlations with the distribution of activated neurons within the DMH. The PrRP‐ir fibers and terminals were immunolabeled with tyrosine hydroxylase, suggesting their origin in lower brainstem instead of local, hypothalamic PrRP cells. PrRP‐ir fibers arising from neurons of the nucleus of the solitary tract could be followed to the hypothalamus. Unilateral transections of these fibers at pontine and caudal hypothalamic levels resulted in a disappearance of the dense PrRP‐ir network in the ventral DMH while PrRP immunoreactivity was increased in transected fibers caudal to the knife cuts as well as in perikarya of the nucleus of the solitary tract ipsilateral to the transections. In accord with these changes, the number of Fos‐expressing neurons following refeeding declined in the ipsilateral but remained high in the contralateral DMH. However, the Fos response in the ventral DMH was not attenuated following chemical lesion (neonatal monosodium glutamate treatment) of the hypothalamic arcuate nucleus, another possible source of DMH inputs. These findings suggest that PrRP projections from the nucleus of the solitary tract contribute to the activation of ventral DMH neurons during refeeding, possibly by transferring information on cholecystokinin‐mediated satiation.     

Distribution of immunoreactive substance P neurons in the hypothalamus and pituitary of the rhesus monkey

The distribution of immunoreactive substance P (IR-SP) neurons was examined in the hypothalamus and pituitary gland of the rhesus monkey by using the peroxidase-antiperoxidase immunocytochemical technique. Immunoreactive SP cell bodies were observed in the arcuate nucleus, in the region lateral to the arcuate nucleus, and in the median eminence (ME). Immunoreactive SP cells were also seen in the periventricular area of the dorsal tuberal region. A rich network of SP fibers was concentrated in the arcuate region, and the fiber stain was particularly dense in the external zone of the median eminence and in the external layer of the infundibular stalk. Also, substance P fibers were seen in the internal layer of the pituitary stalk and in the neural lobe of the pituitary gland. Outside the hypothalamus a dense network of IR-SP fibers was observed in the globus pallidus.     

谷氨酸单钠对大鼠下丘脑生长抑素神经元性差别的影响

本实验用原位杂室组化方法研究了谷氨酸单钠对大鼠下丘脑室周核生长抑素神经元性差别的影响。结果证明对照组下丘脑室周核生长抑素神经元在基因转录水平表现出明显的性别差异，谷氨酸单钠损伤下丘脑弓状核神经元后这种性别差异不再存在。这些结果提示下丘脑室周核生长抑素神经元在基因转录水平所表现出的性别差异可能是通过下丘脑弓状核神经元间接地受性激素的调节。     

Morphometric changes of specific located vasopressin-reacting parvicellular neurons in the paraventricular nucleus of the rat after adrenalectomy

The morphological-morphometric consequences of bilateral adrenalectomy on vasopressin-reacting neurons of the paraventricular nucleus of the rat hypothalamus were analyzed. Bilateral adrenalectomy led to a dramatic increase in the cellular area as well as the number of immunoreactive cells (when compared to those obtained in normal colchicine-treated animals) in the neurons located in the anterior, medial and periventricular parvicellular subdivisions of the paraventricular nucleus. By contrast, no changes were observed in either the dorsal or lateral parvicellular subdivisions or in any of the magnocellular subdivisions of the paraventricular nucleus.