Regulation of hypoxia-induced release of corticotropin-releasing factor in the rat hypothalamus by norepinephrine

Corticotropin-releasing factor (CRF) peptide release was activated by hypoxia in the rat hypothalamus. The mechanisms, however, of the hypoxia-induced CRF release remains unclear. In this study, we demonstrated that the norepinephrine (NE) and its receptors in the paraventricular nucleus (PVN) mediated the CRF release in a simulated altitude hypoxia. When rats were exposed to 5 or 7 km altitude of hypoxia for a short or long term: (1) NE levels in the PVN and the CeA, using the HPLC analysis, were intensity and time course dependently increased, but the increase in the PVN were potential than in the CeA. Restraint-induced NE increase was much higher in both the PVN and the CeA, compared with hypoxia-induced response. (2) Hypoxia and restraint significantly enhanced CRF release in the ME and the PVN but not in the CeA, through RIA assay, which result in stimulating corticosterone secretion. (3) Hypoxia-induced CRF release was reversed by an injection of prazosin (i.c.v.), an alpha-1 adrenoceptor antagonist, while administration of yohimbine (i.c.v.), an alpha-2 receptor antagonist, facilitated further CRF release. These data suggested that hypoxia induced NE activation centrally, via alpha-1 and -2 receptors, leading to improving hypothalamic CRF release, which in turn stimulated pituitary and adrenal cortex. Restraint presented much potential action on NE activation than hypoxia.     

Type 1 corticotropin-releasing factor receptors in the ventromedial hypothalamus promote hypoglycemia-induced hormonal counterregulation

Type 2 corticotropin-releasing factor (CRF) receptors (CRFR2) within the ventromedial hypothalamus (VMH), a key glucose-sensing region, play a major role in regulating the hormonal counterregulatory responses (CRRs) to acute hypoglycemia. The VMH expresses both subtypes of CRF receptors, CRFR1 and CRFR2. The objective of this study was to examine the role of the CRFR1 receptor in the VMH in the regulation of the CRR to acute hypoglycemia. To compare the hormonal CRR to hypoglycemia, awake and unrestrained Sprague-Dawley rats were bilaterally microinjected to the VMH with either 1) aECF, 2) CRF (1 pmol/side), 3) CRFR1 antagonist Antalarmin (500 pmol/side), or 4) CRF + Antalarmin prior to undergoing a hyperinsulinemic hypoglycemic (2.8 mM) clamp. A second series of studies also incorporated an infusion of [(3)H]glucose to allow the calculation of glucose dynamics. In addition the effect of CRFR1 antagonism in the paraventricular nucleus (PVN) was studied. Activation of VMH CRFR1 increased, whereas inhibition of CRFR1 suppressed hypoglycemia-induced CRRs. Inhibition of VMH CRFR1 also increased peripheral glucose utilization and reduced endogenous glucose production during hypoglycemia, whereas VMH CRF reduced peripheral glucose utilization. In contrast CRFR1 inhibition in the PVN blunted corticosterone but not epinephrine or glucagon CRR to hypoglycemia. In contrast to CRFR2 activation, CRFR1 activation within the VMH amplifies CRRs to acute hypoglycemia. The balance between these two opposing CRFRs in this key glucose-sensing region may play an important role in determining the magnitude of CRRs to acute hypoglycemia.     

Disruption of the CRF/CRF1 receptor stress system exacerbates the somatic signs of opiate withdrawal

Escape from the extremely stressful opiate withdrawal syndrome may motivate opiate seeking and taking. The corticotropin-releasing factor receptor-1 (CRF1) pathway mediates behavioral and endocrine responses to stress. Here, we report that genetic inactivation (CRF1-/-) as well as pharmacological antagonism of the CRF/CRF1 receptor pathway increased and prolonged the somatic expression of opiate withdrawal. Opiate-withdrawn CRF1-/- mice also showed aberrant CRF and dynorphin expression in the paraventricular nucleus of the hypothalamus (PVN) and the striatum, indicating profound impairments in stress-responsive brain circuitry. Intake of nonstressful amounts of corticosterone effectively reduced the exaggerated somatic reactions of CRF1-/- mice to opiate withdrawal. Exogenous corticosterone also restored "wild-type-like" patterns of CRF and dynorphin gene expression in the PVN and the striatum of opiate-withdrawn CRF1-/- mice, respectively. The present findings unravel a key role for the hypothalamus-pituitary-adrenal (HPA) system and brain extra-hypothalamic CRF/CRF1 receptor circuitry in somatic, molecular, and endocrine alterations induced by opiate withdrawal.     

The role of corticotropin-releasing factor receptors in stress and anxiety

Corticotropin releasing factor (CRF) is a critical integratorof the hypothalamic-pituitary-adrenal (HPA) axis in responseto stress. CRF and its related molecule urocortin (UCN) bindCRF receptor 1 (CRFR1) and CRFR2 with distinct affinities. Micedeficient for CRFR1 or CRFR2 were generated in order to determinethe physiological role of these receptors. While CRFR1-mutantmice show a depleted stress response and display anxiolytic-likebehavior, CRFR2-mutant mice are hypersensitive to stress anddisplay anxiogenic-like behavior. Both CRFR1- and CRFR2-mutantmice show normal basal feeding and weight gain, but CRFR2-mutantmice exhibit decreased food intake following a stress of fooddeprivation. While CRFR2-mutant mice display increased levelsof CRF mRNA in the central nucleus of the amygdala (cAmyg) butnot in the paraventricular nucleus of the hypothalamus (PVN),the CRFR1-mutant mice express high levels of CRF in the PVNbut normal levels in the cAmyg. CRFR2-mutant mice also displayincreased levels of Ucn mRNA and protein in the edinger westphalnucleus (EW) as well as an increased number of cells expressingUcn. The levels of these CRF-receptor ligands reflect the stateof the receptor-deficient mice. These results demonstrate apossible modulatory function of CRFR2 in response to CRFR1 stimulationof the HPA axis or anxiety.     

Age-specific threats induce CRF expression in the paraventricular nucleus of the hypothalamus and hippocampus of young rats

Young animals respond to threatening stimuli in an age-specific way. Their endocrine and behavioral responses reflect the potential threat of the situation at a given age. The aim of the present study was to determine whether corticotropin-releasing factor (CRF) is involved in the endocrine and behavioral responses to threat and their developmental changes in young rats. Preweaning 14-day-old and postweaning 26-day-old rats were exposed to two age-specific threats, cat odor and an adult male rat. The acute behavioral response was determined during exposure. After exposure, the time courses of the corticosterone response and of CRF expression in the paraventricular nucleus of the hypothalamus (PVN) and in extrahypothalamic areas were assessed. Preweaning rats became immobile when exposed to cat odor or the male rat, whereas postweaning rats became immobile to cat odor only. Male exposure increased serum corticosterone levels in 14-day-old rats, but cat odor failed to increase levels at either age. Exposure induced elevation of CRF mRNA levels in the PVN that paralleled changes in corticosterone levels. CRF may thus play a role in endocrine regulation and its developmental changes during early life. Neither cat odor nor the adult male altered CRF mRNA levels in the bed nucleus of the stria terminalis (BNST) or the amygdala, but both stimuli increased levels in the hippocampus. Hippocampal CRF mRNA expression levels did not parallel cat odor or male-induced immobility, indicating that CRF is not involved in this response in young rats but may be involved in aspects of learning and memory.     

Residues of corticotropin releasing factor-binding protein (CRF-BP) that selectively abrogate binding to CRF but not to urocortin 1

Corticotropin releasing factor-binding protein (CRF-BP) binds CRF and urocortin 1 (Ucn 1) with high affinity, thus preventing CRF receptor (CRFR) activation. Despite recent progress on the molecular details that govern interactions between CRF family neuropeptides and their cognate receptors, little is known concerning the mechanisms that allow CRF-BP to bind CRF and Ucn 1 with picomolar affinity. We conducted a comprehensive alanine scan of 76 evolutionarily conserved residues of CRF-BP and identified several residues that differentially affected the affinity for CRF over Ucn 1. We determined that both neuropeptides derive their similarly high affinity from distinct binding surfaces on CRF-BP. Alanine substitutions of arginine 56 (R56A) and aspartic acid 62 (D62A) reduce the affinity for CRF by approximately 100-fold, while only marginally affecting the affinity for Ucn 1. The selective reduction in affinity for CRF depends on glutamic acid 25 in the CRF peptide, as substitution of Glu(25) reduces the affinity for CRF-BP by approximately 2 orders of magnitude, but only in the presence of both Arg(56) and Asp(62) in human CRF-BP. We show that CRF-BP(R56A) and CRF-BP(D62A) have lost the ability to inhibit CRFR1-mediated responses to CRF that activate luciferase induction in HEK293T cells and ACTH release from cultured rat anterior pituitary cells. In contrast, both CRF-BP mutants retain the ability to inhibit Ucn 1-induced CRFR1 activation. Collectively our findings demonstrate that CRF-BP has distinct and separable binding surfaces for CRF and Ucn 1, opening new avenues for the design of ligand-specific antagonists based on CRF-BP.     

Novel high-affinity photoactivatable antagonists of corticotropin-releasing factor (CRF) photoaffinity labeling studies on CRF receptor, type 1 (CRFR1).

Novel photoactivatable antagonists of human/rat corticotropin-releasing factor (h/rCRF) have been synthesized and characterized. The N-terminal amino acid D-phenylalanine in astressin ?cyclo(30-33) [D-Phe12, Nle21,38, Glu30, Lys33]h/rCRF-(12-41)?, a potent CRF peptide antagonist, was replaced by a phenyldiazirine, the 4-(1-azi-2,2,2-trifluoroethyl)benzoyl (ATB) residue. Additionally, His32 of astressin was substituted by either alanine or tyrosine for specific radioactive labeling with 125I at either His13 or Tyr32, respectively. The photoactivatable CRF antagonists were tested for their ability to displace 125I-labeled Tyr0 ovine CRF ([125I-labeled Tyr0]oCRF) in binding experiments and to inhibit oCRF-stimulated adenylate cyclase activity in human embryonic kidney (HEK) 293 cells, permanently transfected with cDNA coding for rat CRF receptor, type 1 (rCRFR1) or human Y-79 retinoblastoma cells known to carry endogenous functional human CRFR1 (hCRFR1). ATB-cyclo(30-33)[Nle21,38, Glu30, Ala32, Lys33]h/rCRF-(13-41) (compound 1) was found to bind with higher affinity to rat or human CRFR1 when compared with ATB-cyclo(30-33)[Nle21,38, Glu30, Tyr32, Lys33]h/rCRF-(13-41) (compound 2) and exhibited higher inhibition of oCRF-stimulated cAMP accumulation in HEK 293 cells stably transfected with cDNA coding for rCRFR1 (HEK-rCRFR1 cells) or Y-79 cells. A highly glycosylated, 66-kDa protein was identified with SDS/PAGE, when the radioactively iodinated compounds 1 or 2 were covalently linked to rCRFR1. The specificity of the photoactivatable 125I-labeled CRF antagonists was demonstrated with SDS/PAGE by the finding that these analogs could be displaced from the receptor by their corresponding nonlabeled form, but not other unrelated peptides such as vasoactive intestinal peptide. The observed molecular size of the receptor was in agreement with the size of CRFR1 found in rat pituitary (66 kDa), but was significantly larger than the size of CRFR1 found in rat cerebellum and olfactory bulb (53 kDa).     

低氧暴露条件下高原鼠兔和大鼠HPA轴活动的比较

采用人工模拟低气压低氧的方法比较研究了不同程度（模拟海拔5 km和7 km)和不同时间（24d和5d）低氧暴露,对大鼠和高原鼠兔（Ochotona curzoniae）下丘脑-垂体-肾上腺皮质 (hypothalamo-pituitary-adrenalcortex,HPA)轴活动的影响。结果如下：7 km低氧暴露24 h,大鼠下丘脑的促肾上腺皮质激素释放激素（corticotropin-releasing actor,CRF）和肾上腺皮质激素皮质酮分泌显著增加,大鼠HPA低氧暴露对大鼠HPA 轴活动无显著差异。低氧暴露5天后,大鼠7 km、5 km组的HPA轴活动与对照相比无明显差异。低氧暴露对高原鼠兔的HPA轴无明显影响。上述结果表明：低氧暴露的时间和程度与大鼠HPA的活动密切相关;从HPA的活动来看,高原鼠兔表现出较强的低氧耐受性。     

In vivo and in vitro release of ACTH by synthetic CRF

The 41-residue corticotropin releasing factor (CRF) was synthesized by the solid phase method. The synthetic CRF and arginine vasopressin (AVP) were examined for ACTH releasing activity and effects on the release of 5 other pituitary hormones in vivo and in vitro. Injection of the CRF into pharmacologically blocked rats increased plasma corticosterone levels in a dose-related manner. The minimum effective dose was 1.6 x 10(-12) mol/100 g body weight. CRF also significantly stimulated release of ACTH-like immunoreactivity in a dose-related manner from rat pituitary quarters beginning at a concentration of 10(-9) M. AVP, a peptide known to have CRF activity, exhibited slightly lower corticotropin releasing activity than the CRF at equimolar dose levels. Secretion of other pituitary hormones was not appreciably altered by either the CRF or AVP.     

Regulation of corticotropin-releasing factor receptor type 2beta mRNA by mitogen-activated protein kinases in aortic smooth muscle cells

The actions of the corticotropin-releasing factor (CRF) family of peptides are mediated by the seven transmembrane-domain G-protein-coupled receptors, the CRF receptors. CRF receptor type 2beta (CRFR2beta) messenger RNA (mRNA) is expressed primarily in the cardiovascular system, where its levels are decreased by urocortin 1 (Ucn1), a novel peptide in the CRF family. In a previous study, we reported that CRFR2beta mRNA levels were partially down-regulated via the cAMP-protein kinase A pathway. This study focused on the involvement of the intracellular mitogen-activated protein (MAP) kinase pathway in the modulation of CRFR2beta mRNA levels. Ribonuclease protection assays showed that decreases in CRFR2beta mRNA levels induced by Ucn1 and cAMP were attenuated by the p38 MAP kinase inhibitor SB202190 or SB203580. This finding suggested that the p38 MAP kinase pathway was involved in this regulation. Anisomycin, a classic p38 kinase activator, increased CRFR2beta mRNA levels in A7r5 cells. This effect of anisomycin was completely reversed by H7, a serine/threonine kinase inhibitor, while both p38 kinase and MAP kinase kinase inhibitors failed to block the increase in CRFR2beta mRNA levels caused by anisomycin. As anisomycin can activate Jun amino terminal kinases, as well as p38 MAP kinase, it is possible that other MAP kinases, such as Jun amino terminal kinases, also contribute to the increase in gene levels. Alternatively, anisomycin may increase CRFR2beta mRNA levels indirectly as a consequence of blocking protein synthesis.     

Interactions between RAMP2 and CRF receptors: The effect of receptor subtypes,splice variants and cell context

Corticotrophin releasing factor (CRF) acts via two family B G-protein-coupled receptors, CRFR1 and CRFR2. Additional subtypes exist due to alternative splicing. CRFR1α is the most widely expressed subtype and lacks a 29-residue insert in the first intracellular loop that is present in CRFR1β. It has been shown previously that co-expression of CRFR1β with receptor activity modifying protein 2 (RAMP2) in HEK 293S cells increased the cell-surface expression of both proteins suggesting a physical interaction as seen with RAMPs and calcitonin receptor-like receptor (CLR). This study investigated the ability of CRFR1α, CRFR1β and CRFR2β to promote cell-surface expression of FLAG-tagged RAMP2. Four different cell-lines were utilised to investigate the effect of varying cellular context; COS-7, HEK 293T, HEK 293S and [ΔCTR]HEK 293 (which lacks endogenous calcitonin receptor). In all cell-lines, CRFR1α and CRFR1β enhanced RAMP2 cell-surface expression. The magnitude of the effect on RAMP2 was dependent on the cell-line ([ΔCTR]HEK 293 > COS-7 > HEK 293T > HEK 293S). RT-PCR indicated this variation may relate to differences in endogenous RAMP expression between cell types. Furthermore, pre-treatment with CRF resulted in a loss of cell-surface FLAG-RAMP2 when it was co-expressed with CRFR1 subtypes. CRFR2β co-expression had no effect on RAMP2 in any cell-line. Molecular modelling suggests that the potential contact interface between the extracellular domains of RAMP2 and CRF receptor subtypes is smaller than that of RAMP2 and CRL, the canonical receptor:RAMP pairing, assuming a physical interaction. Furthermore, a specific residue difference between CRFR1 subtypes (glutamate) and CRFR2β (histidine) in this interface region may impair CRFR2β:RAMP2 interaction by electrostatic repulsion.     

Coexistence of CRF peptide and oxytocin mRNA in the paraventricular nucleus

Several studies have reported coexistences of peptides in parvocellular neurons of the paraventricular nucleus (PVN). However, the coexistence of peptides in the magnocellular PVN is less clear. Controversy exists in particular about the coexistence of corticotropin-releasing factor (CRF) and oxytocin (OX). Although these peptides are present in distinct areas of the PVN, some overlap may exist. This study investigated a potential coexistence of OX and CRF in magno- and parvocellular PVN. The data demonstrate with clarity that neurons containing both the mRNA for OX and the peptide CRF are present in subpopulations of magnocellular and parvocellular neurons of the PVN.     

Effects of experimental hypothyroidism on the development of the hypothalamo-pituitary-adrenal axis in the rat

Hypothyroid pups were obtained by adding methimazole in the mother's drinking water from day 15 of gestation and sacrificed at 4, 8 or 15 days. Circulating corticosterone decreased at all ages, while CBG concentrations diminished at day 4, increased at day 8 and did not change at day 15 in hypothyroid rats. As opposed to controls, plasma ACTH concentrations decreased steadily with age while there was an accumulation of ACTH in the anterior pituitary of hypothyroid 15-day-old rats. Anterior pituitary POMC contents were unaffected by the treatment. In the hypothalamic PVN, CRF mRNA levels in the total population of CRF-synthesizing cells and in the CRF+/AVP+ subpopulation were below those of controls whatever the age considered while AVP mRNA in the CRF+/AVP+ subpopulation did not change at day 4 and decreased at day 8 and 15 in hypothyroid animals. Both the number of cell bodies expressing detectable levels of CRF mRNA and the percentage of CRF and AVP colocalization decreased at day 4 and were unchanged thereafter. CRF and AVP immunoreactivity in the zona externa of the median eminence increased with age but was not affected by methimazole treatment. The concentration of AVP mRNA in the magnocellular cell bodies of the PVN and the SON as well as AVP immunoreactivity in the zona interna of the median eminence were not changed by the treatment at days 4 and 8. In hypothyroid 15-day-old rats, SON AVP mRNA increased, AVP immunoreactivity decreased while plasma osmolality was enhanced. In conclusion, our data demonstrate that experimental hypothyroidism impairs specifically the maturation of hypothalamic parvocellular CRF and AVP gene expression during the stress hyporesponsive period. These observations suggest that the physiological peak in plasma thyroxine concentrations that occur between day 8-12 may participate in the maturation of hypothalamic CRF- and AVP-synthesizing cells.     

The stress response neuropeptide CRF increases amyloid‐β production by regulating γ‐secretase activity

The biological underpinnings linking stress to Alzheimer's disease (AD) risk are poorly understood. We investigated how corticotrophin releasing factor (CRF), a critical stress response mediator, influences amyloid‐β (Aβ) production. In cells, CRF treatment increases Aβ production and triggers CRF receptor 1 (CRFR1) and γ‐secretase internalization. Co‐immunoprecipitation studies establish that γ‐secretase associates with CRFR1; this is mediated by β‐arrestin binding motifs. Additionally, CRFR1 and γ‐secretase co‐localize in lipid raft fractions, with increased γ‐secretase accumulation upon CRF treatment. CRF treatment also increases γ‐secretase activity in vitro, revealing a second, receptor‐independent mechanism of action. CRF is the first endogenous neuropeptide that can be shown to directly modulate γ‐secretase activity. Unexpectedly, CRFR1 antagonists also increased Aβ. These data collectively link CRF to increased Aβ through γ‐secretase and provide mechanistic insight into how stress may increase AD risk. They also suggest that direct targeting of CRF might be necessary to effectively modulate this pathway for therapeutic benefit in AD, as CRFR1 antagonists increase Aβ and in some cases preferentially increase Aβ42 via complex effects on γ‐secretase.     

Glucocorticoids regulate peptidyl-glycine alpha-amidating monooxygenase gene expression in the rat hypothalamic paraventricular nucleus

Peptidyl-glycine alpha-amidating monooxygenase (PAM) is a posttranslational processing enzyme which catalyzes the formation of biologically active alpha-amidated peptides. The two major neuropeptides involved in the regulation of ACTH secretion [CRF and arginine vasopressin (AVP)], synthesized in the parvocellular part of the hypothalamic paraventricular nucleus (PVN), are amidated, and their synthesis and/or release is negatively regulated by glucocorticoids. In this study, using in situ hybridization, we have shown that PAM mRNA is abundantly expressed in the hypothalamic paraventricular and supraoptic nucleus. Surgical adrenalectomy (ADX) induced increases in PAM, CRF, and AVP mRNA in the parvocellular part of the PVN, while corticosterone treatment normalized these values. PAM and AVP gene expression were not changed in the magnocellular part of the PVN or in the supraoptic nucleus. These observations suggest that in addition to stimulation of CRF and AVP synthesis, ADX induces an increase in PAM synthesis in the PVN and, thus, support the hypothesis of increased secretion of both CRF and AVP after ADX.     

High-affinity binding of urocortin and astressin but not CRF to G protein-uncoupled CRFR1

The structure-activity relationship (SAR) between the recently identified neuropeptide urocortin (Ucn) and corticotropin-releasing factor (CRF) receptor, type 1 (CRFR1), has been investigated. To this end, rat Ucn (rUcn), ovine CRF (oCRF) and chimeric peptides of rUcn and oCRF were synthesized and tested for their binding affinity and potency to stimulate cAMP production in human embryonic kidney (HEK) 293 cells stably transfected with cDNA encoding rat CRFR1 (rCRFR1). In binding studies with [125I-TyrO]oCRF or [3H-Leu9]rUcn as radioligand, it was observed that rUcn but not oCRF bound in a similar fashion as the CRF antagonist astressin with high affinity to rCRFR1 coupled to G protein or uncoupled from G protein by guanosine 5'-O-(3-thiotriphosphate) (GTPgammaS). Consequently, rUcn was found to exert a significantly lower potency than oCRF to stimulate cAMP accumulation in transfected cells. CD spectroscopic investigations and reverse-phase HPLC (RPHPLC) retention behavior of the peptides suggested a more pronounced amphipatic alpha-helical character of rUcn when compared to oCRF and the chimeric peptides.     

Urocortins and corticotropin releasing factor type 2 receptors in the hypothalamus and the cardiovascular system

In addition to urocortin (Ucn I), Ucn II and Ucn III were identified as endogenous ligands for corticotropin-releasing factor type 2 receptor (CRF2 receptor). CRF2 receptor is abundantly located in central hypothalamic ventromedial nucleus (VMH) and in peripheral cardiovascular system. In this mini-review, we focused on the roles of these urocortins and CRF2 receptor in the hypothalamus and the cardiovascular system. Ucn II mRNA was increased in the parvocellular part or the magnocellular part of the hypothalamic paraventricular nucleus (PVN) following immobilization stress or 3 days of water deprivation, respectively. Therefore, it is thought that Ucn II may modulate CRF and vasopressin synthesis in the PVN in a paracrine or autocrine fashion through PVN CRF2 receptor. The early and later phases of Ucn I-mediated feeding suppression may be CRF1 and CRF2 receptor-mediated events, respectively. Ucn II decreases food intake at a later phase, beyond 4 h post injection. A large dose of corticosterone increased plasma leptin and insulin levels as well as the levels of CRF2 receptor mRNA. Adrenalectomy, starvation, and immobilization each lowered plasma leptin and insulin levels and were associated with decrements in CRF2 receptor mRNA levels in the VMH. Peripheral injection of leptin increased VMH CRF2 receptor mRNA, as can induce reductions of food intake and body weight, indicating that circulating leptin is involved in the regulation of VMH CRF2 receptor mRNA expression. Therefore, it is also plausible that VMH CRF2 receptor transduces the anorexogenic effects of leptin as well as those of urocortins. The systemic administration of Ucn II decreases mean arterial pressure (arterial vascular tone) and causes tachycardia via vascular CRF2 receptor in rats, similar to the effects of Ucn I. Thus, CRF2 receptor seems to mediate cardioprotective effects of urocortins.     

Starvation-induced changes in rat brain corticotropin-releasing factor (CRF) and pituitary-adrenocortical response

Starvation-induced changes in CRF concentration in major brain regions and abnormalities in the pituitary-adrenal axis were examined in rats using rat CRF radioimmunoassay. The CRF concentrations in the hypothalamus and cerebellum were significantly reduced in the completely starved rats, while those in the midbrain, thalamus and neurointermediate lobe of the pituitary were significantly increased in the semi-starved or completely starved rats. No significant changes in the CRF concentrations were found in the pons, medulla oblongata and cerebral cortex. In the completely starved rats, the serum ACTH level was significantly reduced, whereas the serum corticosterone level was markedly elevated. These observations suggest that starvation may stimulate the CRF-ACTH-corticosterone system and that not only hypothalamic CRF but also extrahypothalamic CRF may be discretely related to feeding behavior or starvation. The reduced serum ACTH level in starved rats may be ascribed to the negative feedback effect of the elevated serum corticosterone.