Anorexigenic C75 alters c-Fos in mouse hypothalamic and hindbrain subnuclei

The fatty acid synthase inhibitor C75 reduces feeding rapidly and for several days. We investigated brain sites potentially involved in actions of i.p. C75 in mice by examining c-Fos. At 3 h C75 increased numbers of c-Fos-immunoreactive cells in hindbrain feeding-related nuclei, and in the paraventricular nucleus (PVN), lateral aspects of the arcuate nucleus (ARC), and in the central amygdala. At 24 h C75 prevented fasting-induced c-Fos expression in the medial ARC and three of its targets: lateral magnocellular PVN, lateral hypothalamus, and dorsomedial hypothalamus. C75, but not fasting, increased c-Fos in parvocellular PVN. This pattern of results suggests a shift from hindbrain-initiated short-term actions to activation of hypothalamic mechanisms that could mediate the long-term anorectic responses to C75.     

Effect of calcitonin gene-related peptide (CGRP) on avian appetite-related processes

Calcitonin gene-related peptide (CGRP) is released from the gastrointestinal tract following ingestion and causes satiety in mammals. Its effects on appetite in non-mammalian vertebrates are unreported. In Experiment 1, fasted chicks reduced food and water intake after central injection of CGRP. These effects were not associated with increased plasma corticosterone concentration. In Experiment 2, we showed that the effect on water intake was independent of food intake. In Experiment 3, central CGRP caused increased c-Fos immunoreactivity in the arcuate (ARC) nucleus, paraventricular nucleus (PVN), periventricular (PHN) and ventromedial (VMH) hypothalamic nuclei. The results of Experiment 4 demonstrate that intraperitoneal injection of CGRP also causes reduced food and water intake. c-Fos immunoreactivity was increased in the ARC, PHN, PVN and VMH in Experiment 5 after intraperitoneal injection of CGRP. Lastly in Experiment 6, we showed that central CGRP changes the type of pecks from feeding to exploratory, and reduces the number of escape attempts. The effect of CGRP appears to be primary on appetite in chicks. In conclusion, the mechanisms of CGRP induced satiety have some similarities and differences between avian and rodent models. The results presented here provide new insight into the evolution of vertebrate satiety regulatory mechanisms.     

Food intake elicited by central administration of orexins/hypocretins: identification of hypothalamic sites of action

Orexin A and B, a recently identified pair of neuropeptides, are produced in perikarya located in the lateral and perifornical hypothalamus (LH and PFH). Immunoreactive fibers from these neurons innervate several nuclei in the hypothalamus. Orexin A and orexin B stimulate feeding when administered intracerebroventricularly to rats. To identify the specific sites of orexin action, orexin A and B were microinjected into a number of hypothalamic and extrahypothalamic sites in rats. Orexin A was found to enhance food intake when injected into four hypothalamic sites, the paraventricular nucleus (PVN), the dorsomedial nucleus (DMN), LH and the perifornical area, but was ineffective in the arcuate nucleus (ARC), the ventromedial nucleus (VMN), and the preoptic area (POA) as well as the central nucleus of the amygdala (CeA) and nucleus of the tractus solitarius (NTS). Orexin B was not effective at any site tested. These findings demonstrate that orexin A receptive sites for stimulation of food intake exist primarily in a narrow band of neural tissue within the hypothalamus that is known to be involved in control of energy homeostasis.     

The central anorexigenic mechanism of adrenocorticotropic hormone involves the caudal hypothalamus in chicks

Adrenocorticotropic hormone (ACTH), consisting of 39 amino acids, is most well-known for its involvement in an organism's response to stress. It also participates in satiety, as exogenous ACTH causes decreased food intake in rats. However, its anorexigenic mechanism is not well understood in any species and its effect on appetite is not reported in the avian class. Thus, the present study was designed to evaluate central ACTH's effect on food intake and to elucidate the mechanism mediating this response using broiler chicks. Chicks that received intracerebroventricular (ICV) injection of 1, 2, or 4 nmol of ACTH reduced food intake, under both ad libitum and 180 min fasted conditions. Water intake was also reduced in ACTH-injected chicks under both feeding conditions, but when measured without access to feed it was not affected. Blood glucose was not affected in either feeding condition. Following ACTH injection, c-Fos immunoreactivity was quantified in key appetite-associated hypothalamic nuclei including the ventromedial hypothalamus (VMH), dorsomedial hypothalamus, lateral hypothalamus (LH), arcuate nucleus (ARC) and the parvo- and magno-cellular portions of the paraventricular nucleus. ACTH-injected chicks had increased c-Fos immunoreactivity in the VMH, LH, and ARC. Hypothalamus was collected at 1 h post-injection, and real-time PCR performed to measure mRNA abundance of some appetite-associated factors. Neuropeptide Y, pro-opiomelanocortin, glutamate decarboxylase 1, melanocortin receptors 2–5, and urocortin 3 mRNA abundance was not affected by ACTH treatment. However, expression of corticotropin releasing factor (CRF), urotensin 2 (UT), agouti-related peptide (AgRP), and orexin (ORX), and melanocortin receptor 1 (MC1R) mRNA decreased in the hypothalamus of ACTH-injected chicks. In conclusion, ICV ACTH causes decreased food intake in chicks, and is associated with VMH, LH, and ARC activation, and a decrease in hypothalamic mRNA abundance of CRF, UT, AgRP, ORX and MC1R.     

Feeding and drinking elicited by central injection of neuropeptide Y: Evidence for a hypothalamic site(s) of action

Neuropeptide Y (NPY), which exists in very high concentrations in the brain, has been shown to elicit a powerful feeding response and a small drinking response in satiated rats. In order to delineate the brain sites sensitive to these effects, NPY was injected through chronic guide cannulas into seven different brain regions, and the food and water intake of satiated rats was measured one hr postinjection. Injection of NPY (78 pmoles) into hypothalamic areas, namely the paraventricular nucleus (PVN), ventromedial hypothalamus (VMH), and lateral hypothalamus (LH), elicited a strong feeding response; in contrast, injections into extra-hypothalamic areas, namely the amygdala, thalamus, and periaqueductal gray, were completely ineffective. Administration of NPY into the PVN and VMH also elicited a small drinking response; however, all other areas, including the LH, were insensitive to this effect. The findings that NPY was effective in the hypothalamus, as opposed to sites anterior, posterior, lateral or dorsal to this structure, suggest a hypothalamic site(s) of action for this neuropeptide.     

Hypothalamic mechanisms associated with corticotropin-releasing factor-induced anorexia in chicks

Central administration of corticotropin-releasing factor (CRF), a 41-amino acid peptide, is associated with potent anorexigenic effects in rodents and chickens. However, the mechanism underlying this effect remains unclear. Hence, the objective of the current study was to elucidate the hypothalamic mechanisms that mediate CRF-induced anorexia in 4 day-old Cobb-500 chicks. After intracerebroventricular (ICV) injection of 0.02 nmol of CRF, CRF-injected chicks ate less than vehicle chicks while no effect on water intake was observed at 30 min post-injection. In subsequent experiments, the hypothalamus samples were processed at 60 min post-injection. The CRF-injected chicks had more c-Fos immunoreactive cells in the arcuate nucleus (ARC), dorsomedial nucleus (DMN), ventromedial hypothalamus (VMH), and paraventricular nucleus (PVN) of the hypothalamus than vehicle-treated chicks. CRF injection was associated with decreased whole hypothalamic mRNA abundance of neuropeptide Y receptor sub-type 1 (NPYR1). In the ARC, CRF-injected chicks expressed more CRF and CRF receptor sub-type 2 (CRFR2) mRNA but less agouti-related peptide (AgRP), NPY, and NPYR1 mRNA than vehicle-injected chicks. CRF-treated chicks expressed greater amounts of CRFR2 and mesotocin mRNA than vehicle chicks in the PVN and VMH, respectively. In the DMN, CRF injection was associated with reduced NPYR1 mRNA. In conclusion, the results provide insights into understanding CRF-induced hypothalamic actions and suggest that the anorexigenic effect of CRF involves increased CRFR2-mediated signaling in the ARC and PVN that overrides the effects of NPY and other orexigenic factors.     

Cervical stimulation activates A1 and locus coeruleus neurons that project to the paraventricular nucleus of the hypothalamus

In female rats, stimulation of the uterine cervix during mating induces two daily surges of prolactin. Inhibition of hypothalamic dopamine release and stimulation of oxytocin neurons in the paraventricular nucleus (PVN) are required for prolactin secretion. We aim to better understand how stimulation of the uterine cervix is translated into two daily prolactin surges. We hypothesize that noradrenergic neurons in the A1, A2, and locus coeruleus (LC) are responsible for conveying the peripheral stimulus to the PVN. In order to determine whether projections from these neurons to the PVN are activated by cervical stimulation (CS), we injected a retrograde tracer, Fluoro-Gold (FG), into the PVN of ovariectomized rats. Fourteen days after injection, animals were submitted to artificial CS or handling and perfused with a fixative solution. Brains were removed and sectioned from the A1, A2, and LC for c-Fos, tyrosine hydroxylase (TH), and FG triple-labeling using immunohistochemistry. CS increased the percentage of TH/FG+ double-labeled neurons expressing c-Fos in the A1 and LC. CS also increased the percentage of TH+ neurons expressing c-Fos within the A1 and A2, independent of their projections to the PVN. Our data reinforce the significant contributions of the A1 and A2 to carry sensory information during mating, and provide evidence of a functional pathway in which CS activates A1 and LC neurons projecting to the PVN, which is potentially involved in the translation of CS into two daily prolactin surges.     

Cholecystokinin activates catecholaminergic neurons in the caudal medulla that innervate the paraventricular nucleus of the hypothalamus in rats

Stimulation of gastric vagal afferents by systemic administration of cholecystokinin octapeptide (CCK) inhibits gastric motility, reduces food intake, and stimulates pituitary secretion of oxytocin and adrenocorticotropic hormone in rats. To characterize further the central neurol circuits responsible for these effects, the present study used triple-labeling immunocytochemical methods to determine whether or not exogenous CCK activates cFos expression in catecholaminergic neurons in the caudal medulla that project to the paraventricular nucleus of the hypothalamus (PVN). To identify these neurons, the retrograde tracer fluorogold (FG) was iontophoresed into the PVN of anesthetized rats under stereotaxic guidance. After 2 weeks, rats were injected with CCK (100 μg/kg, i. p.) and then anesthetized and killed 1 hour later by perfusion fixation. Medullary sections were processed for triple immunocytochemical localization of cFos, retrogradely transported FG, and tyrosine hydroxylase (TH). In rats with FG injections centered in the PVN (n = 10), approximately 70% of the FG-labeled neurons in the caudal nucleus of the solitary tract (NST) and ventrolateral medulla (VLM) expressed cFos. Of these activated PVN-projecting neurons, approximately 78% in the NST and 89% in the VLM were catecholaminergic (TH positive). These results indicate that PVN-projecting catecholaminergic neurons within the caudal medulla are activated by periph eral administration of CCK, further implicating these ascending catecholaminergic path ways in the neuroendocrine, physiological, and behavioral effects produced by gastric vagal stimulation. © 1995 Wiley-Liss, Inc.     

Effects of Cholecystokinin in the Supraoptic Nucleus and Paraventricular Nucleus are Negatively Modulated by Leptin in 24‐h Fasted Lean Male Rats

Cholecystokinin (CCK) and leptin are two important satiety factors that are considered to act in synergy to reduce meal size. Peripheral injection of CCK activates neurones in several hypothalamic nuclei, including the supraoptic (SON) and paraventricular (PVN) nuclei and neurones in the brainstem of fed rats. We investigated whether peripheral leptin would modulate the effects of CCK on neuronal activity in the hypothalamus and brainstem of fasted rats by investigating Fos expression in the PVN, SON, arcuate nucleus, ventromedial hypothalamus (VMH), dorsomedial hypothalamus (DMH), area postrema (AP) and the nucleus tractus solitarii (NTS). Male rats, fasted for 24 h, received either one i.p. injection of vehicle, leptin or CCK‐8 alone, or received one injection of vehicle or leptin before an i.p. injection of CCK‐8. We found that CCK increased Fos expression in the PVN and SON as well as in the NTS and AP, but had no effect on Fos expression in the arcuate nucleus, VMH or DMH compared to vehicle. Leptin injected alone significantly increased Fos expression in the arcuate nucleus but had no effect on Fos expression in the VMH, DMH, SON, PVN, AP or NTS compared to vehicle. Fos expression was significantly increased in the AP in rats injected with both leptin and CCK compared to rats injected with vehicle and CCK. Unexpectedly, there was significantly less Fos expression in the PVN and SON of fasted rats injected with leptin and CCK than in rats injected with vehicle and CCK, suggesting that leptin attenuated CCK‐induced Fos expression in the SON and PVN. However, Fos expression in the NTS was similar in fasted rats injected with vehicle and CCK or with leptin and CCK. Taken together, these results suggest that leptin dampens the effects of CCK on Fos expression in the SON and PVN, independently from NTS pathways, and this may reflect a direct action on magnocellular neurones.     

5-HT1B receptors modulate the feeding inhibitory effects of enterostatin

Serotonin (5-HT) is considered to play an important role in control of appetite. Enterostatin has been shown to alter 5-HT release in the brain, and non-specific 5-HT antagonists blocked the anorectic response to icv enterostatin. The aim of this study was to further identify which 5-HT receptor subtype mediates the enterostatin feeding behavior and whether this effect occurs due to action in the PVN. Wild-type and 5-HT2C receptor-/- (KO) mice and normal Sprague-Dawley rats were used in these experiments. All animals were fed a high fat diet. Enterostatin (120 nmol, i.p.) reduced the intake of high fat diet in 5-HT2C receptor mutant mice (saline 4.54 +/- 0.47 kcal vs. Ent 2.53 +/- 0.76 kcal) 1 h after injection. A selective 5-HT1B antagonist (GR55526, 40 mg/kg body weight, i.p.) blocked the enterostatin hypophagic effects in these KO mice. Rats were implanted with cannulas into the amygdala and the ipsilateral PVN. The 5-HT receptor antagonists metergoline (non-specific receptor subtypes 1 and 2), or ritanserin (selective 2C), or GR55562 (selective l B) was injected into the PVN prior to enterostatin (0.01 nmol) injection into the amygdala. Enterostatin reduced food intake (saline: 5.80 +/- 0.59 g vs. enterostatin 3.47 +/- 0.56 g, P < 0.05 at l h). Pretreatment with either metergoline (10 nmol) or GR55526 (10 nmol) but not ritanserin (10 nmol) into the PVN attenuated the anorectic response to amygdala enterostatin. The data imply that the enterostatin anorectic response may be modulated by 5-HT1B receptors and that a neuronal pathway from the amygdala to the PVN regulates the enterostatin response through activation of 5-HTlB receptors in PVN.     

Effects of Gold Thioglucose Treatment on Central Corticotrophin‐Releasing Hormone Systems in Mice

Systemic administration of gold thioglucose (GTG) causes a hypothalamic lesion that extends from the ventral part of the ventromedial hypothalamus (VMH) to the dorsal part of the arcuate nucleus (ARC), resulting in hyperphagia and obesity in mice. In the present study, we used in situ hybridisation histochemistry to explore the effects of GTG on the central corticotrophin‐releasing hormone (CRH) system, which regulates feeding and energy homeostasis. Type 2 CRH receptor (CRHR‐2) mRNA expression decreased by 40% at 8 weeks in the VMH and by 40–60% at 2 and 8 weeks in the ARC after GTG injection. By contrast, CRHR‐2 mRNA expression in the hypothalamic paraventricular nucleus (PVN) and lateral septum was unchanged. Urocortin (Ucn) 3 mRNA expression in the perifornical area and medial amygdala decreased, whereas CRH mRNA expression in the PVN increased at 2 and 8 weeks after GTG injection. Ucn 1 mRNA expression in the Edingher–Westphal nucleus and Ucn 2 mRNA expression in the PVN were unchanged. Because Ucn 3 is an anorexigenic and a possible endogenous ligand for VMH CRHR‐2, our results suggest that decreased Ucn 3 expression and decreased VMH CRHR‐2 expression contribute, in part, to GTG‐induced hyperphagia and obesity. To determine whether VMH CRHR‐2 mediates the anorexigenic effects of Ucn 3, Ucn 3 was administered i.c.v. and food intake was measured 8 weeks after GTG treatment. Ucn 3 decreased cumulative food intake on days 4–7 after surgery compared to i.c.v. administration of vehicle in control mice. By contrast, the anorexigenic effects of i.c.v. Ucn 3 were abolished in GTG‐treated mice. Taken together, our results indicate that the Ucn 3 pathway, which innervates the VMH, is involved in appetite regulation via CRHR‐2. It remains to be determined whether CRHR‐2 in the ARC has additional roles in appetite regulation by Ucn 3.     

Central neuropeptide FF reduces feed consumption and affects hypothalamic chemistry in chicks

Information on the physiological functions of neuropeptide FF; NPFF, a morphine modulating octapeptide in avians is lacking. Thus, we designed a study to investigate the effects of central NPFF with particular emphasis on appetite-related processes. Cobb-500 chicks were intracerebroventricularly (ICV) injected with 0, 4.16, 8.32 or 16.6nmol NPFF, and feed and water intake were quantified. Feed intake was linearly decreased as NPFF dose increased, and this effect decayed over time and was not significant by 120min post-injection. Water intake was not affected by ICV NPFF. In a second exp, we observed that naloxone completely reversed the NPFF-induced decrease in feed intake. The amount of time a visible marker took to travel through the total length of the alimentary canal linearly increased as NPFF dose increased. We measured neuronal activation in the lateral hypothalamus (LH), paraventricular nucleus (PVN) dorsomedial nucleus (DMN) and ventromedial hypothalamus (VMN) of the hypothalamus, and nucleus dorsomedialis posterior thalami (DMP) of the thalamus. The DMN, DMP, PVN and VMH were all activated by ICV NPFF while the LH was not affected. Finally, we determined that the anorexigenic effect of ICV NPFF is primarily behavior specific, since behaviors unrelated to ingestion were not increased the same duration of time as was consumatory pecking. We conclude that NPFF causes anorexigenic effects in chicks that are primarily behavior specific.     

Development of obesity and neurochemical backing in aurothioglucose-treated mice

To clarify the neurochemical backing of aurothioglucose (ATG)-induced obesity in mice, we investigated lesion sites, hypothalamic neurotransmitters and c-Fos-like immunoreactivity (Fos-IR). At day 2 after ATG, tissue loss or cells death was observed in several parts of the ventral area of the ventromedial hypothalamic nucleus (VMH), and the dorsal area of arcuate nucleus and in the nucleus of the solitary tract (NTS). However, the greater part of the VMH was retained. Body weight began to increase in week 1. Hypothalamic serotonin (5-HT) and the metabolites were increased at day 2. The contents of acetylcholine, norepinephrine and dopamine in the hypothalamus showed no significant change. In week 1, the area shown tissue loss was compacted and plugged up. In the control group, most obvious c-Fos-like immunoreactive region was paraventricular nucleus (PVN). At day 2, Fos-IR was observed around destroyed regions in the hypothalamus and NTS, but few Fos-IR was found in the other regions including PVN. The Fos-IR around destroyed regions diminished after week 1. In week 3, Fos-IR in the PVN increased. These results suggest that the development of ATG-induced obesity cannot be attributed to solely VMH destruction. The restoration processes of the neuronal dysfunction involving PVN seem to play an important role in the development of obesity. NTS lesion and 5-HT system might contribute to decrease in food intake for several days after ATG.     

Efferents from medial basal forebrain and hypothalamus in the rat. II. An autoradiographic study of the anterior hypothalamus.

Using tritiated amino acid autoradiography, the efferent projections of the anterior hypothalamic area (AHA) were studied in albino rats. Axons from AHA neurons were not confined to local projections in the hypothalamus. Ascending AHA axons ran through the preoptic region, joined the diagonal band and distributed in the lateral septum. Descending AHA efferents within the hypothalamus coursed in a bundle ventromedial to the fornix. Projections were observed to the dorsomedial, ventromedial, arcuate and dorsal premammillary nuclei, and to the median eminence. Sweeping dorsomedially in the posterior hypothalamus, some AHA axons distributed in the central grey. AHA axons staying ventral projected to the supramammillary region, ventral tegmental area, raphe nuclei and midbrain reticular formation. Other AHA efferents distributed to the periventricular thalamus, to the medial amygdala via the stria terminalis or supraoptic commissure, and to the lateral habenula through the stria medullaris. For comparison with the AHA, efferent projections from the paraventricular nucleus (PVN) and from the ventromedial nucleus and adjacent basal hypothalamus (VMR) were studied. Projections from PVN neurons were not restricted to the median eminence and neurohypophysis. PVN efferents also distributed to many of the same regions as did those of the AHA but had somewhat different fiber trajectories and longer descending projections. VMR efferents were more widespread than those of the AHA, with projections extending into the lateral zona incerta and pontine reticular formation. Projections from the AHA were distinct from those of the medial preoptic area (mPOA). For example, while AHA axons descended in a bundle ventromedial to the fornix, mPOA axons ran in the medial forebrain bundle. Such anatomical differences may underlie experimentally demonstrated functional differences between the mPOA and AHA, for instance, in mediation of male and female sex behaviors.     

Xenin reduces feed intake by activating the ventromedial hypothalamus and influences gastrointestinal transit rate in chicks

This study was conducted to determine the effects of xenin on appetite related processes in chicks. Chicks were centrally and peripherally administered xenin, and feed and water intake were quantified. Chicks responded with a linear dose-dependent decrease in feed intake to central xenin and had a quadratic type response to peripheral administration. Water intake was not affected by treatment. To determine if the lateral hypothalamus (LH) or ventromedial hypothalamus (VMH) was involved in this effect, chicks were both centrally and peripherally injected with xenin and an immunocytochemistry assay for c-Fos was conducted. Central and peripheral xenin caused increased activation of the VMH but had no effect on the LH. Finally, to determine if gastrointestinal transit rate was affected, chicks received central xenin and were gavaged with chicken feed slurry containing a visible marker. Chicks exhibited a quadratic dose-dependent response to transit rate after central xenin. These results suggest that xenin affects feeding and gastrointestinal motility through hypothalamic interactions in chicks.     

Opposite dopaminergic activity in lateral and median hypothalamic nuclei in relation to the feeding effect ofd-Ser-Leu-Enk-Thr (DSLET)

The Leu-enkephalin analogue D-Ser2-Leu-Enk-Thr6 (DSLET) had been shown to enhance feeding in rats, increase dopaminergic activity in the striatum like other opiate agonists, and particularly to decrease dopaminergic activity in the hypothalamus. In this study, the latter effect was found to be localized in the hypothalamic nuclei involved in the regulation of feeding such as the paraventricular (PVN), ventromedian (VMH), dorsomedian (DMH) nuclei and the lateral hypothalamus (LH). DSLET produced the same decrease in dopaminergic activity in the LH as in the whole hypothalamus. In the median nuclei (PVN and VMH and to a lesser extent in the DMH), an opposite effect was observed, resembling that in the striatum. The relevance of these opposite variations with regard to the feeding effect of DSLET is discussed. The decreased dopaminergic activity in the LH would appear to be the most specifically related to the behavioural effect given the known role of dopamine in this region. These data reconcile apparently contradictory aspects of the role of dopamine and the functional opposition between the lateral and median hypothalamus in food intake control.     

Distribution and efferent projections of corticotropin-releasing factor-like immunoreactivity in the rat amygdaloid complex

Using cobalt-enhanced immunohistochemistry, the tracing of retrograde transport of horseradish peroxidase (HRP) and experimental manipulations, a widespread localization of corticotropin-releasing factor-like immunoreactive (CRFI) structures in the rat amygdaloid complex, and CRFI-containing pathways from the amygdala to the lower brainstem, bed nucleus of the stria terminalis (bst) and ventromedial nucleus of the hypothalamus (VMH) have been demonstrated. By means of cobalt-enhanced immunohistochemistry, CRFI cells were detected in almost all the regions of the amygdala, including the central amygdaloid nucleus (Ce), basolateral amygdaloid nucleus (B1), intra-amygdaloid bed nucleus of the stria terminalis (Abst), medial amygdaloid nucleus (Me), amygdalohippocampal area (Ahi), posterior cortical amygdaloid nucleus (Aco), lateral amygdaloid nucleus (La), anterior amygdaloid area (AAA) and basomedial amygdaloid nucleus (Bm). Neural processes with CRFI were found in all of the above areas. The greatest density of CRFI fibres was observed in the Ce, the Me and Ahi. Unilateral lesions located in the Ce and adjacent areas caused an ipsilateral decrease in CRFI fibre number in the lateral hypothalamic area (LH), mesencephalic reticular formation (RF), dorsal (Dpb) and ventral (Vpb) parabrachial nuclei, mesencephalic nucleus of the trigeminal nerve (MeV) and in the lateral division of the bst (bstl). In addition, ipsilateral CRFI fibres decreased in number in the core and shell of the VMH after unilateral lesions of the corticomedial amygdala (CoM) and ventral subiculum (S). These findings suggest that the CRFI cells in the Ce and adjacent areas innervate the Dpb, Vpb and MeV through the LH and RF; the CRFI fibres in the bstl are supplied by the Ce and adjacent areas; and the CoM and S give rise to the CRFI fibres to the VMH. The distribution of retrogradely transported HRP has confirmed these projections. Furthermore, combined HRP and immunohistochemical staining has demonstrated double labeled cells in the Ce following HRP injection into the Dpb, Vpb, MeV and bstl. This provides direct evidence for the amygdalofugal CRF-containing projections to the lower brainstem and bstl. Double-labeled cells were not seen in the CoM and S after HRP injection into the VMH.