The immunostaining for the hypothalamic vasoactive intestinal peptide, but not for beta-endorphin, dynorphin-A or methionine-enkephalin, is affected by the glucocorticoid milieu in the rat: correlation with the prolactin secretion

Regulation of estrogen receptor-beta expression in the female rat hypothalamus: differential effects of dexamethasone and estradiol

Estrogen and glucocorticoids interact in multiple aspects of endocrine regulation by exerting opposing influences on the expression of selective genes. In rats, estrogen receptor (ER)-beta is the predominant form of ER present in the hypothalamic paraventricular (PVN) and supraoptic (SON) nuclei, suggesting its involvement in neuroendocrine regulation. To date, the hormonal regulatory profile of the ERbeta gene in the rat central nervous system has not been closely elucidated. In the present study, we first examined the effects of dexamethasone (DEX) and estradiol benzoate (EB) on the ERbeta protein expression in the PVN and SON of ovariectomized female rats. In the SON and parvocellular and magnocellular parts of the PVN, the number of ERbeta immunoreactive nuclei significantly increased after DEX treatment, compared with the control group, whereas EB treatment caused a significant decrease. The effect of EB was consistent across other brain nuclei such as the anteroventral periventricular nucleus and medial preoptic nucleus. To determine the molecular level at which DEX and EB control ERbeta expression, we examined the effects of these steroids on ERbeta mRNA levels using real-time RT-PCR. EB significantly decreased the expression of ERbeta mRNA in the PVN (P = 0.0006) and SON (P < 0.01). In contrast, DEX did not change ERbeta mRNA levels. These results indicate that glucocorticoids and estrogen exert opposing regulatory influences on the ERbeta gene expression. This may represent a mechanism by which these steroids can alter the cellular sensitivity of ERbeta-expressing neurons to subsequent steroidal activation.

Effect of adrenalectomy on mating-induced prolactin surges and pseudopregnancy in the female rat

In the estrous female rat, mating stimulation induces an acute surge of prolactin (PRL) within 20 min after mating followed by the onset of twice-daily PRL surges which persist for an 8- to 13-day period of acyclicity called pseudopregnancy. In Experiment 1, we examined whether the release of adrenal hormones after mating modulates mating-induced PRL secretion during the first 38 h after mating. Ovariectomized females were adrenalectomized (Adx) or sham-operated (Sham) and were implanted with jugular vein catheters 2 days later. They were given estrogen and progesterone and mated 6 days after the last surgery until they received 15 intromissions or 15 mounts-without-intromission from a male. Blood samples were collected beginning 20 min before mating at 23:00 h and continuing for 38 h. Plasma PRL concentrations were measured using radioimmunoassay. Mating that included intromissions induced an acute (20-min) PRL response which was higher in Adx than in Sham animals, and advanced in the Adx animals in the onset of the first daily PRL surge to 10:00 h, some 18 h before the surge was observed at 04:00 h in the Sham-mated animals. A small but measurable nocturnal surge was observed in Adx and Sham groups 18-24 h later at 04:00-10:00 h. In Experiment 2, Adx- and Sham-cycling animals received 5 (5I) or 7 (7I) intromissions from a male 12-16 days after surgery. Adx animals receiving 5I showed a significantly higher incidence of pregnancy or pseudopregnancy (%P/PSP) than did Sham 5I animals, while there was no difference in %P/PSP in the 7I groups. We conclude that adrenal gland secretions normally suppress plasma PRL concentrations immediately post-mating and before the onset of the nocturnal mating-induced PRL surge and also inhibit pseudopregnancy when females receive a subthreshold number of intromissions normally required for its induction.

Neuroendocrine pharmacology of stress

Exposure to hostile conditions initiates responses organized to enhance the probability of survival. These coordinated responses, known as stress responses, are composed of alterations in behavior, autonomic function and the secretion of multiple hormones. The activation of the renin-angiotensin system and the hypothalamic-pituitary-adrenocortical axis plays a pivotal role in the stress response. Neuroendocrine components activated by stressors include the increased secretion of epinephrine and norepinephrine from the sympathetic nervous system and adrenal medulla, the release of corticotropin-releasing factor (CRF) and vasopressin from parvicellular neurons into the portal circulation, and seconds later, the secretion of pituitary adrenocorticotropin (ACTH), leading to secretion of glucocorticoids by the adrenal gland. Corticotropin-releasing factor coordinates the endocrine, autonomic, behavioral and immune responses to stress and also acts as a neurotransmitter or neuromodulator in the amygdala, dorsal raphe nucleus, hippocampus and locus coeruleus, to integrate brain multi-system responses to stress. This review discussed the role of classical mediators of the stress response, such as corticotropin-releasing factor, vasopressin, serotonin (5-hydroxytryptamine or 5-HT) and catecholamines. Also discussed are the roles of other neuropeptides/neuromodulators involved in the stress response that have previously received little attention, such as substance P, vasoactive intestinal polypeptide, neuropeptide Y and cholecystokinin. Anxiolytic drugs of the benzodiazepine class and other drugs that affect catecholamine, GABA(A), histamine and serotonin receptors have been used to attenuate the neuroendocrine response to stressors. The neuroendocrine information for these drugs is still incomplete; however, they are a new class of potential antidepressant and anxiolytic drugs that offer new therapeutic approaches to treating anxiety disorders. The studies described in this review suggest that multiple brain mechanisms are responsible for the regulation of each hormone and that not all hormones are regulated by the same neural circuits. In particular, the renin-angiotensin system seems to be regulated by different brain mechanisms than the hypothalamic-pituitary-adrenal system. This could be an important survival mechanism to ensure that dysfunction of one neurotransmitter system will not endanger the appropriate secretion of hormones during exposure to adverse conditions. The measurement of several hormones to examine the mechanisms underlying the stress response and the effects of drugs and lesions on these responses can provide insight into the nature and location of brain circuits and neurotransmitter receptors involved in anxiety and stress.

Annexin 1 (lipocortin 1) mediates the glucocorticoid inhibition of cyclic adenosine 3',5'-monophosphate-stimulated prolactin secretion

Our previous studies have identified a role for annexin 1 (also called lipocortin 1) in the regulatory actions of glucocorticoids (GCs) on the release of PRL from the rat anterior pituitary gland. In the present study we used antisense and immunoneutralization strategies to extend this work. Exposure of rat anterior pituitary tissue to corticosterone (1 nM) or dexamethasone (100 nM) in vitro induced 1) de novo annexin 1 synthesis and 2) translocation of the protein from intracellular to pericellular sites. Both responses were prevented by the inclusion in the medium of an annexin 1 antisense oligodeoxynucleotide (ODN; 50 nM), but not by the corresponding sense and scrambled ODN sequences. Unlike the GCs, 17beta-estradiol, testosterone, and aldosterone (1 nM) had no effect on either the synthesis or the cellular disposition of annexin 1; moreover, none of the steroids or ODNs tested influenced the expression of annexin 5, a protein closely related to annexin 1. The increases in PRL release induced in vitro by drugs that signal via cAMP/protein kinase A [vasoactive intestinal polypeptide (10 nM), forskolin (100 microM), 8-bromo-cAMP (0.1 microM)] or phospholipase C (TRH, 10 nM) were attenuated by preincubation of the pituitary tissue with either corticosterone (1 nM) or dexamethasone (100 nM). The inhibitory actions of the steroids on the secretory responses to vasoactive intestinal polypeptide, forskolin, and 8-bromo-cAMP were specifically quenched by inclusion in the medium of the annexin 1 antisense ODN (50 nM) or a neutralizing antiannexin 1 monoclonal antibody (antiannexin 1 mAb, diluted 1:15,000). By contrast, the ability of the GCs to suppress the TRH-induced increase in PRL release was unaffected by both the annexin 1 antisense ODN and the antiannexin 1 mAb. In vivo, interleukin-1beta (10 ng, intracerebroventricularly) produced a significant increase in the serum PRL concentration (P < 0.01), which was prevented by pretreatment of the rats with corticosterone (100 microg/100 g BW, sc). The inhibitory actions of the steroid were specifically abrogated by peripheral administration of an antiannexin 1 antiserum (200 microl, sc); by contrast, when the antiserum was given centrally (3 microl, intracerebroventricularly), it was without effect. These results support our premise that annexin contributes to the regulatory actions of GCs on PRL secretion and suggest that it acts at point distal to the formation of cAMP.

Evaluation of the role for prolactin-releasing peptide in prolactin secretion induced by ether stress and suckling in the rat: comparison with vasoactive intestinal peptide

Prolactin (PRL)-releasing peptide (PrRP) is a recently discovered hypothalamic peptide possessing a specific stimulatory action on PRL secretion. In this study, we examined whether PrRP plays a role in mediating ether stress- and suckling-induced PRL secretion in rats through administering anti-PrRP antisera intracerebroventricularly. For comparison, we also tested the effect of anti-vasoactive intestinal peptide (VIP) antisera on the hormonal responses, since VIP is another candidate for a physiological PRL-releasing factor. The immunoneutralization of VIP, but not of PrRP, led to a significant suppression of PRL responses to both ether and suckling. These results suggest that PrRP may not play a significant role, or at least play a much weaker role than VIP, in mediating PRL release induced by ether stress and suckling in the rat.

Colocalization of VIP with AVP in neurons of the human paraventricular, supraoptic and suprachiasmatic nucleus

Aim of this study was to investigate, with the aid of a recently developed immunofluorescence technique, cellular colocalization of vasoactive intestinal peptide (VIP) with arginine-vasopressin (AVP) in the paraventricular nucleus (PVN), the supraoptic nucleus (SON) and the suprachiasmatic nucleus (SCN) of the human hypothalamus. To this end, six hypothalami resected from patients who had died suddenly served as material of research. After formaldehyde fixation and subsequent storage in 30% sucrose, 25-microm thick cryosections were cut of one half of each hypothalamus. These sections were double-immunolabeled with primary antibodies against AVP and VIP followed by fluorophore-conjugated secondary antibodies. Autofluorescence, mainly caused by lipofuscin granules in neurons and glial cells, was blocked by a specially developed procedure consisting of incubating the immunolabeled sections in a Sudan Black B solution. Quantitative analysis with a confocal laser scanning microscope showed that of all stained cellular profiles the percentages of profiles immunoreactive exclusively for AVP or VIP or for both neuropeptides (colocalization) were for the SCN approximately 76.5%, 19.6% and 3.9%, for the SON 97.7%, 0.2% and 2. 1% and for the PVN 93.2%, 1.6% and 5.2%, respectively. These data illustrate that colocalization between AVP and VIP is not only present in neurons of the PVN and SON, but also in neurons of the SCN. This unexpected finding illustrates that the human SCN may also use a highly differentiated language to transmit its circadian signal to the rest of the brain.

Regulation of gene expression by corticoid hormones in the brain and spinal cord

Glucocorticoids (GC) and mineralocorticoids (MC) have profound regulatory effects upon the central nervous system (CNS). Hormonal regulation affects several molecules essential to CNS function. First, evidences are presented that mRNA expression of the alpha3 and beta1-subunits of the Na,K-ATPase are increased by GC and physiological doses of MC in a region-dependent manner. Instead, high MC doses reduce the beta1 isoform and enzyme activity in amygdaloid and hypothalamic nuclei, an effect which may be related to MC control of salt appetite. The alpha3-subunit mRNA of the Na,K-ATPase is also stimulated by GC in motoneurons of the injured spinal cord, suggesting a role for the enzyme in GC neuroprotection. Second, we provide evidences for hormonal effects on the expression of mRNA for the neuropeptide arginine vasopressin (AVP). Our data show that GC inhibition of AVP mRNA levels in the paraventricular nucleus is sex-hormone dependent. This sexual dimorphism may explain sex differences in the hypothalamic-pituitary-adrenal axis function between female and male rats. Third, steroid effects on the astrocyte marker glial fibrillary acidic protein (GFAP) points to a complex regulatory mechanism. In an animal model of neurodegeneration (the Wobbler mouse) showing pronounced astrogliosis of the spinal cord, in vivo GC treatment down-regulated GFAP immunoreactivity, whereas the membrane-active steroid antioxidant U-74389F up-regulated this protein. It is likely that variations in GFAP protein expression affect spinal cord neurodegeneration in Wobbler mice. Fourth, an interaction between neurotrophins and GC is shown in the injured rat spinal cord. In this model, intensive GC treatment increases immunoreactive low affinity nerve growth factor (NGF) receptor in motoneuron processes. Because GC also increases immunoreactive NGF, this mechanism would support trophism and regeneration in damaged tissues. In conclusion, evidences show that some molecules regulated by adrenal steroids in neurons and glial cells are not only involved in physiological control, but additionally, may play important roles in neuropathology.

Lactation as a model for naturally reversible hypercorticalism plasticity in the mechanisms governing hypothalamo-pituitary- adrenocortical activity in rats

Steady state levels of hypothalamic expression of the genes encoding corticotropin-releasing hormone (CRH), proopiomelanocortin (POMC), arginine vasopressin (AVP), and oxytocin (OT) were studied in rats to investigate the mechanisms underlying the transitions between hypercorticalism during lactation and normocorticalism upon weaning. During lactation, CRH mRNA levels and blood titers of adrenocorticotropin (ACTH) were found to be significantly reduced, although POMC mRNA levels in the anterior pituitary were not significantly different from those found in cycling virgin (control) rats; during all phases of lactation, an inverse relationship was observed between the blood levels of ACTH and corticosterone (CORT). Plasma prolactin (PRL) concentrations were elevated approximately 30-fold during lactation. Whereas steady state levels of OT mRNA were markedly increased throughout lactation, those of AVP mRNA were only transiently (initially) elevated, and the blood levels of these hormones were not significantly altered in lactating as compared with cycling virgin and postlactating rats. CRH and POMC gene expression and blood levels of ACTH, CORT, and PRL were normalized within 1-3 d of removal of suckling pups. The temporal relationships between the biosynthetic profiles of the various peptide hormones and the patterns of ACTH and CORT secretion during the two physiological states suggest that lactation-associated hypercorticalism does not merely result from increased ACTH secretion; although still not well substantiated at this time, the evidence points to contributory roles of PRL, OT, and AVP in the hypercorticalismic state found during lactation.

Regulation of VIP mRNA expression by thyroid hormone in different brain areas of adult rat

A role of thyroid hormone in the regulation of neuropeptide synthesis has been demonstrated in different tissues. In this paper we investigated the vasoactive intestinal peptide (VIP) mRNA expression by means of in situ hybridization in several brain areas of hypo- and hyperthyroid adult rats. Neither hypo- nor hyperthyroidism modified the VIP mRNA levels in the thalamus and in the hypothalamic suprachiasmatic nucleus. In contrast, in the anterior cingulate and frontoparietal motor cortex of hypothyroid rats there was a marked increase in the signal for VIP mRNA per cell, but the number of VIP expressing neurons did not change. These data indicate that also central VIP synthesis can be influenced by the levels of circulating thyroid hormone, but that this effect is confined to specific areas and cell populations of the brain.

VIP binding sites in adult rat hypothalamus: nuclear distribution and daily variations

Vasoactive intestinal peptide (VIP) neurons are present in the suprachiasmatic nucleus (SCN) and project to various hypothalamic nuclei. Several results suggest that hypothalamic VIP is involved in the mediation of synchronizing information. [125I]Vasoactive intestinal peptide binding has been reported to be especially high in the hypothalamus. We show here by radioautography by dipping that [125I]VIP binding is high and homogeneously distributed in the SCN. The correspondence between VIP binding and terminals densities supports the hypothesis that VIP acts through synapses in the SCN. Moderate and fairly homogeneous VIP binding is also found in the paraventricular nucleus (PaVN), whereas VIP terminals are reported to be restricted to the sub-paraventricular area. This suggests that VIP could diffuse and thus function as a paracrine neuromodulator in the PaVN. [125I]Vasoactive intestinal peptide binding is stable along the nycthemer in the suprachiasmatic, peri-, and paraventricular nuclei as well as in all other areas investigated.

Median eminence-afferent vasoactive intestinal peptide (VIP) neurons in the hypothalamus: localization by simultaneous tract tracing and immunocytochemistry

Retrograde tract tracing and immunocytochemistry were used to investigate the CNS source of the VIP that is present in high concentrations in the hypophysial portal blood and has been shown to have a stimulatory effect on pituitary prolactin secretion. Fluoro-gold (FG), which enters the CNS through areas devoid of the blood-brain barrier, such as median eminence, was injected peripherally. Brain sections from FG-treated animals were immunostained for VIP. A small population of VIP-containing cell bodies in the parvocellular and periventricular parts of the paraventricular nucleus (PVN) was also labeled with FG. Vasoactive intestinal peptide-immunoreactive perikarya not labeled with FG were also observed in the PVN, as well as FG-labeled cells that did not contain VIP. The results suggest that some VIP-producing neurons in the PVN project to the median eminence and are, therefore, functionally related to pituitary regulation; the function of other VIP neurons in the PVN is unknown.

Daily variations in concentration of vasoactive intestinal peptide immunoreactivity in hypothalamic nuclei of rats rendered diurnal by restricted-schedule feeding

We previously described that in the suprachiasmatic (SCN), peri-(PeVN) and paraventricular (PaVN) nuclei of normal rats, vasoactive intestinal peptide (VIP)-like immunoreactivity (VIP-LI) accumulates during the night period and decreases during the day. In order to determine whether these variations are linked to the light-dark cycle or are a consequence of sleep-wake rhythm expression, we dissociated these two parameters by restricting feeding to diurnal hours. In these conditions which inverse the paradoxical sleep rhythm, the VIP-LI pattern is perturbed and its minimum advanced by 4 h in the SCN. In the PeVN, the daily pattern is maintained but the minimum is also advanced by 4 h. Finally, the rhythm is abolished in the PaVN. These circadian fluctuations indicate that the hypothalamic VIP-LI rhythm is not linked to the paradoxical sleep rhythm but could be sensitive to photic and non-photic Zeitgeber.

Disturbance of fluid homeostasis leads to temporally and anatomically distinct responses in neuropeptide and tyrosine hydroxylase mRNA levels in the paraventricular and supraoptic nuclei of the rat

The response of six mRNAs (for prepro-corticotropin-releasing hormone, prepro-enkephalin, prepro-vasoactive intestinal polypeptide/peptide histidine isoleucine, prepro-neurotensin/neuromedin N, prepro-cholecystokinin, and prepro-tyrosine hydroxylase) was measured in the hypothalamic paraventricular and supraoptic nuclei after increasing periods of osmotic stimulation caused by the replacement of regular drinking water with hypertonic saline (up to five days) or by forced dehydration (up to three days). In addition, hematocrits and concentrations of corticosterone were determined after the different periods of osmotic stimulation and correlated with the effects on the content of the various mRNAs. The temporal response of the mRNAs within the paraventricular and supraoptic nuclei to osmotic stimulation was different within the three compartments of these nuclei. First, in response to overnight osmotic stimulation, magnocellular neurosecretory neurons increased their mRNA content for two molecules (prepro-corticotropin-releasing hormone and tyrosine hydroxylase). As the stimulus was maintained over the next two to four days, these cells accumulated the mRNAs for at least three other peptides (cholecystokinin, vasoactive intestinal polypeptide/peptide histidine isoleucine and enkephalin). Second, the response of peptide-coding mRNAs in parvicellular neurosecretory neurons of the paraventricular nucleus appeared to be slower; no changes could be measured after overnight stimulation. However, after a further two- to four-days of continued osmotic stimulation, the content of the mRNA coding for corticotropin-releasing hormone markedly decreased while that for cholecystokinin increased. No change in the content of the mRNAs coding for prepro-vasoactive intestinal polypeptide/peptide histidine isoleucine, enkephalin, and prepro-neurotensin/neuromedin N could be seen at any time after osmotic stimulation in parvicellular neurosecretory neurons. Third, increases in the content of mRNA coding for corticotropin-releasing hormone in the parvicellular neurons that provide descending projections from the paraventricular nucleus could only be detected after longer periods of osmotic stimulation. The effect of osmotic stimulation on plasma corticosterone concentrations was quickly apparent; plasma corticosterone concentrations were significantly elevated on the first morning after the beginning of salt-loading, and demonstrated the rapid effects of osmotic stimulation on the mechanisms controlling corticosterone release. These results show that the synthetic capability of cells in all three compartments of the paraventricular and supraoptic nuclei are modified by osmotic stimulation over different time scales, thereby allowing differential modulation of the neuroendocrine, autonomic, and behavioral components of the animal's response to disturbances in fluid homeostasis.(ABSTRACT TRUNCATED AT 400 WORDS)

Neonatal androgenization increases vasoactive intestinal peptide levels in rat anterior pituitary: possible involvement of vasoactive intestinal peptide in the neonatal androgenization-induced hyperprolactinemia

Accumulating evidence suggests that vasoactive intestinal peptide (VIP) may be a physiological PRL-releasing factor. In the present study, we examined a possible involvement of VIP in the neonatal androgenization (NA)-induced hyperprolactinemia. Twenty-four hours after birth, newborn female rats were injected sc with 1,000 micrograms of testosterone (NA) or with oil vehicle only (control). Both groups were sacrificed at 8 weeks of age. Compared to controls, NA rats showed significantly higher plasma PRL levels (7.3 fold), anterior pituitary (AP) PRL content (2.1 fold) and plasma estradiol levels (2.1 fold). AP VIP content was extremely higher (61 fold) in NA rats than in controls. However, NA did not affect VIP content in the suprachiasmatic nucleus, paraventricular nucleus or median eminence. These results suggest that the NA-induced hyperprolactinemia may be mediated, at least in part, by paracrine and/or autocrine effects of the increased AP VIP on PRL secretion. However, since the potentiation by NA of the AP VIP content was extremely marked compared to those of the other parameters, the possibility was also raised that the increased AP VIP may be involved in other endocrine and/or nonendocrine events occurring in the AP.

Vasoactive intestinal peptide levels in the hypothalamus and anterior pituitary do not change in association with the pre-ovulatory prolactin surge in the rat

We examined whether vasoactive intestinal peptide (VIP) is involved in the preovulatory prolactin (PRL) surge in the rat. Ovariectomized adult rats implanted sc with an estradiol capsule 2 days previously were decapitated at hourly intervals between 1100-1800 h. The plasma PRL and luteinizing hormone showed an evident surge at 1700 h and 1800 h, respectively. The anterior pituitary PRL content was significantly lower at and after 1600 h than at 1100 h. However, there was no significant change in the VIP content in the suprachiasmatic nucleus, paraventricular nucleus, median eminence or anterior pituitary throughout the entire period of observation. Since all of these regions are known to be involved in the VIP synthesis and/or release, the present results appear to cast some doubt on previous studies reporting a significant role for VIP in the generation of the pre-ovulatory PRL surge.

Ether anesthesia differentially affects the content of prepro-corticotropin-releasing hormone, prepro-neurotensin/neuromedin N and prepro-enkephalin mRNAs in the hypothalamic paraventricular nucleus of the rat

The content of the mRNAs that code for prepro-corticotropin releasing hormone (ppCRH), prepro-neurotensin/neuromedin N (ppNT/NMN) and prepro-enkephalin (ppENK) was measured using in situ hybridization histochemistry in the various subdivisions of the hypothalamic paraventricular nucleus (PVH) 4 h after a bout of ether anesthesia. This treatment increased the content of ppENK mRNA only in a selected region of the dorsal medial parvicellular PVH (PVHmpd). PpNT/NMN mRNA was undetectable in the PVH of control animals, but hybridization was observed in the PVHmpd of anesthetized animals. The content of ppCRH mRNA was unchanged 4 h after ether aesthesia.

Vasoactive intestinal polypeptide/peptide histidine isoleucine immunoreactive neuron systems in the basal hypothalamus of the rat with special reference to the portal vasculature: an immunohistochemical and in situ hybridization study

Using immunohistochemistry and in situ hybridization, we have analysed the distribution of vasoactive intestinal polypeptide (VIP) and peptide histidine isoleucine (PHI) in hypothalamus of male and female Sprague-Dawley rats under normal and experimental conditions. In most cases there was a good overlap between the distribution of VIP- and PHI-immunoreactive structures. At the median eminence level precapillary arterioles along its lateral aspect were surrounded by dense networks of VIP/PHI-positive fibers, suggesting that these peptides, via their vasodilatory property, may be involved in control of blood flow through portal vessels. Furthermore, a thick VIP/PHI-containing nerve bundle of variable size was observed on the surface of the median eminence in coronal, horizontal and sagittal sections. Also this bundle could be of importance for portal circulation, but VIP/PHI released may reach the anterior pituitary level and play a role in, for example, control of prolactin release. Although different lesions were performed, the origin of the VIP/PHI nerves around lateral blood arterioles and of the bundle is still unclear, but is in all probability peripheral. Within the median eminence of untreated rats only few positive nerve endings were seen in the external layer, but after 48 h hypophysectomy a large number of PHI-immunoreactive fibers could be observed. With regard to cell bodies the suprachiasmatic nucleus contained VIP/PHI-immunoreactive neurons even in untreated rats. After colchicine administration fluorescent cells were in addition seen in several other hypothalamic nuclei, including the parvocellular paraventricular nucleus. After hypophysectomy, with in situ hybridization, VIP mRNA could be demonstrated in magno- and parvocellular neurons of the paraventricular nucleus, whereas in control rats VIP mRNA was undetectable. These results demonstrate that VIP/PHI are present in at least three systems of direct neuroendocrine importance: (1) in nerves related to the blood vessels in the median eminence and presumably involved in control of blood flow through the portal system; (2) in parvocellular paraventricular neurons, presumably related to stress-induced prolactin release; and (3) in magnocellular neurons after certain experimental manipulations.