Site selective activation of lateral hypothalamic mGluR1 and R5 receptors elicits feeding in rats.

Recent findings from our lab indicate that metabotropic glutamate receptor (mGluR) activation elicits eating, and the goal of the current study was to specify whether the lateral hypothalamus (LH) is the actual brain site mediating this effect. To examine this issue we injected the selective mGluR group I agonist (S)-3,5-dihydroxyphenylglycine (DHPG) unilaterally into the LH and surrounding regions (n=5-8 subjects/brain site) of satiated adult male Sprague-Dawley rats and measured elicited feeding. We determined that 1.0 nmol elicited food intake only within the LH. Increasing the dose to 10 or 25 nmol produced a more sustained effect in the LH, and also elicited eating in several other brain sites. These results, demonstrating that the LH mediates the eating elicited by low doses of DHPG, suggest that the LH may contain mGluR whose activation can produce eating behavior.

N-methyl-D-aspartate receptor subunit NR2B is widely expressed throughout the rat diencephalon: an immunohistochemical study.

Glutamate (Glu), a major excitatory neurotransmitter within the hypothalamus and thalamus, acts upon many receptors, including the N-methyl-D-aspartate (NMDA) subtype. Abundant evidence suggests that variations in the subunit composition of NMDA receptors (NMDA-Rs) contribute to differences in Glu's immediate electrophysiological effects as well as to the patterns of signal transduction cascades it triggers to mediate long-term changes in neuronal function. We have previously shown that hypothalamic NMDA-Rs containing the NR2B subunit may be involved in the control of eating as well as in the mediation of physiological responses to osmotic stimuli. To broaden our understanding of diencephalic NMDA-R participation in other functions, we localized the NR2B subunit in the diencephalon of the adult male rat using immunoperoxidase, immunogold, and immunofluorescence techniques and an affinity-purified polyclonal antibody specific for the NR2B subunit of the NMDA-R. In addition, we used a monoclonal NR2B antibody with immunoperoxidase detection to confirm the NR2B distribution seen with the polyclonal antibody. In the hypothalamus, the highest levels of NR2B immunoreactivity (-ir) were found in the magnocellular neurosecretory system, including the paraventricular and supraoptic nuclei. A new finding was that intense NR2B-ir was present within perivascular "accessory" magnocellular groups of this system, including the nucleus circularis, anterior fornical nucleus, and scattered clusters of lateral hypothalamic cells apposed to blood vessels. Robust NR2B-ir was also present within the arcuate nucleus, the median eminence, and the tuberal nucleus, and light immunostaining was found in all other hypothalamic nuclei examined. In the thalamus, the highest NR2B-ir was observed in the medial habenula and the anterodorsal, paraventricular, rhomboid, reticular, and dorsal lateral geniculate nuclei. As in the hypothalamus, all thalamic nuclei examined displayed at least light immunostaining for this subunit. Control sections, including those incubated with the polyclonal NR2B antibody preadsorbed with its fusion protein, were virtually devoid of immunostaining. This demonstration that the NR2B subunit of the NMDA-R is widely distributed in the diencephalon, implicates it in a wide variety of functions, and provides a useful anatomical framework for establishing a comprehensive map of Glu receptor populations within this major subdivision of the brain.

Self-injection of amphetamine directly into the brain.

Rats learned to self-administer d-amphetamine (10 micrograms/microliter) through a cannula implanted in the nucleus accumbens. They responded more frequently for 65 +/- 15 nl of amphetamine than for equal amounts of saline. When presented with two levers (one amphetamine, one blank) they responded more on the correct lever for amphetamine. They would also switch levers, when necessary, to maintain access to the drug. When half the usual drug intake was given automatically, animals reduced their response rate by half, thus self-regulating the total amount of amphetamine they received. In tests for leakage into the ventricles, eight rats that self-injected with an accumbens cannula showed response extinction when switched to a ventricular cannula. We conclude that amphetamine self-injected into the accumbens is a positive reinforcer. This localization of 'amphetamine reward' suggests that the nucleus accumbens contains a synaptic mechanism underlying amphetamine abuse and, perhaps, also natural reinforcement of behavior.

Feeding responses to perifornical hypothalamic injection of neuropeptide Y in relation to circadian rhythms of eating behavior.

Hypothalamic injection of neuropeptide Y (NPY) can elicit eating in satiated rats, and the perifornical hypothalamus (PFH) is the site where this effect is most pronounced (48). Additionally, there is a well-documented circadian rhythm of spontaneous eating behavior. Our objective was to determine whether there are daily rhythms of sensitivity to NPY in the PFH that might contribute to this behavioral rhythm. To accomplish this, the effectiveness in eliciting eating of PFH injection of NPY was examined at six different time points in the light-dark cycle. Neuropeptide Y (78 pmol/10 nl) or vehicle (10 nl) were injected through chronically implanted guide cannulas into the PFH of satiated adult male rats and food intake was measured 1, 2, and 4 h later. In animals on 12-12 h light-dark cycles, these injections were given 1 h before and after the onset of the light and dark phases, and in the middle of these phases. Additionally, dose-response effects of NPY were examined at two points: the first hour of both the dark and the light phases. The results show that NPY was effective at every time tested, and that the magnitude of the peptide-elicited intakes was primarily additive to the underlying patterns of spontaneous intake, with only a modest daily cycle of sensitivity to NPY. Consistent with this, NPY dose-dependently increased intake in the early light and in the early dark, and the magnitude of these effects across doses was similar at these times. This suggests that the sensitivity of the PFH system mediating NPY eating exhibits only a modest daily cycle.(ABSTRACT TRUNCATED AT 250 WORDS)

Neuropeptide Y chronically injected into the hypothalamus: a powerful neurochemical inducer of hyperphagia and obesity.

Neuropeptide Y (NPY), a putative neurotransmitter abundant in the brain, has recently been shown to act within the hypothalamus, inducing a powerful eating response and a specific appetite for carbohydrates. In the present study, NPY (235 pmol) injected bilaterally in the paraventricular nucleus three times a day for 10 days caused approximately a two-fold increase in daily food intake, a six-fold increase in the rate of body weight gain and a three-fold increase in the body fat of female rats. Subsequently, the food intake and body weight of these subjects decreased precipitously, reaching control levels 20 days postinjection. These findings, demonstrating that exogenous NPY is capable of overriding mechanisms of satiety and body weight control, suggest that disturbances in NPY function may play a role in some disorders of eating behavior and body weight regulation.

Bombesin-induced anorexia: sites of action in the rat brain.

To determine the brain sites at which centrally injected bombesin (BBS) may act to suppress feeding behavior, this peptide (1.0 micrograms/0.3 microliter) was microinjected into one of twelve brain regions in 6 hr food deprived rats, and food intake was measured 45 min postinjection. Bombesin produced its strongest suppression of feeding (47-65%) when injected into hypothalamic areas, namely, the paraventricular, dorsomedial, ventromedial nuclei and lateral hypothalamus, and also when administered into the amygdala and the periaqueductal gray. Insensitive areas included the septum, ventral tegmental area and reticular formation. In contrast to these somewhat site-specific effects on feeding behavior, observation of BBS' effects on other behaviors revealed that, in all brain areas tested, there was a significant increase in grooming behavior and decrease in time spent resting and sleeping. In conjunction with high levels of BBS-like immunoreactivity and BBS receptors in the brain areas where injected BBS suppresses feeding, these results suggest that the effects of centrally administered BBS on feeding behavior may be mediated by multiple hypothalamic and extra-hypothalamic brain regions.

Neuropeptide Y receptor agonists: multiple effects on spontaneous activity in the paraventricular hypothalamus.

In vitro rat hypothalamic slices were used to examine the ability of neuropeptide Y (NPY), and the putative Y1 and Y2 receptor agonists [Pro34]NPY and [C2]NPY, to modify spontaneous single-neuron discharge in the paraventricular nucleus (PVN). NPY and [Pro34]NPY, at high concentrations (1500 nM), decreased discharge rates. At intermediate concentrations (150 nM) these peptides produced multiple effects, including increases, decreases, and biphasic changes. At lower concentrations (0.15-15 nM), they typically increased discharge rates. In contrast, [C2]NPY, at all concentrations (1.5-1500 nM), predominantly increased discharge rates. Thus, these NPY subtype agonists have multiple effects on discharge rate, which may be due to actions on multiple NPY receptor subtypes.

Neurotensin: effects of hypothalamic and intravenous injections on eating and drinking in rats.

To investigate a role for the brain-gut peptide neurotensin (NT) in ingestive behavior, changes in food and water intake of food-deprived rats were examined following injection of NT into the paraventricular hypothalamic nucleus (PVN) or the mesenteric vein. Unilateral PVN NT (2.5, 5.0, 10.0 micrograms/0.3 microliter) produced substantial dose-dependent reductions in total food intake 0.5, 1, and 4 hr postinjection. In contrast, PVN NT had no effect on water intake and produced no change in grooming, rearing, sleeping, resting or locomotor activity. Bilateral PVN NT at a high dose (10.0 micrograms/side) suppressed consumption of solid or liquid diet in food-deprived rats, but did not affect water intake in water-deprived rats. This specificity is consistent with a role for CNS NT in feeding behavior. Intravenous NT (1-1000 pmole/kg/min for 30 min) did not specifically suppress food intake; however, low doses did increase water intake in food-deprived rats. These findings do not support a role for plasma NT in feeding, but do suggest that it may play a role in drinking behavior.

Evidence for neuropeptide Y mediation of eating produced by food deprivation and for a variant of the Y1 receptor mediating this peptide's effect.

Neuropeptide Y (NPY) elicits eating when injected directly into the paraventricular nucleus (PVN) or perifornical hypothalamus (PFH). To identify the essential regions of the NPY molecule and the relative contributions of Y1 and Y2 receptors, the eating stimulatory potency of NPY was compared to that of its fragments, analogues, and agonists when injected into the PVN or PFH of satiated rats. Additionally, antisera to NPY was injected into the cerebral ventricles (ICV) to determine whether passive immunization suppresses the eating produced by mild food deprivation. Tests with NPY fragments revealed that NPY(2-36) was surprisingly potent, nearly three times more so than intact NPY. In contrast, fragments with further N-terminal deletions were progressively less effective or ineffective, as was the free acid form of NPY. Collectively, this suggests that both N- and C-terminal regions of NPY participate in the stimulation of eating. Tests with agonists revealed that the putative Y1 agonist [Pro34]NPY elicited a strong dose-dependent feeding response, while the putative Y2 agonist, C2-NPY, had only a small effect at the highest doses. Although this suggests mediation by Y1 receptors, the uncharacteristically high potency of NPY(2-36) may additionally suggest that the receptor subtype underlying feeding is distinct from that mediating other responses. Additional results revealed that ICV injection of antisera to NPY, which should inactivate endogenous NPY, produced a concentration-dependent suppression of eating induced by mild food deprivation. This finding, along with published work demonstrating enhanced levels of hypothalamic NPY in food-deprived rats, suggests that endogenous NPY mediates the eating produced by deprivation.

Stimulation of feeding by galanin: anatomical localization and behavioral specificity of this peptide's effects in the brain.

The neuropeptide galanin (GAL) has been found to elicit eating after injection into the hypothalamic paraventricular nucleus (PVN). To determine whether GAL's effect in the brain is anatomically specific, this peptide (0.1 or 0.3 nmol) was microinjected into one of 14 different brain areas of rats, and its impact on subsequent food intake was measured. Among the hypothalamic sites tested, only the PVN and the adjacent periventricular region yielded a significant eating response to GAL. With injection into the PVN, a feeding response was observed without apparent changes in other food-associated behaviors, e.g., drinking, grooming, resting and sleeping, or low and high levels of activity. All other hypothalamic and extrahypothalamic sites tested were unresponsive to GAL, with the exception of the amygdala where a significant eating response was observed. These findings suggest that central GAL elicits feeding by acting in an anatomically localized and behaviorally specific manner. In light of other pharmacological and anatomical evidence, it is suggested the PVN GAL, in modulating feeding behavior, may work in association with the catecholamine norepinephrine (NE) which is known to coexist with GAL in PVN neurons.

Paraventricular nucleus injections of peptide YY and neuropeptide Y preferentially enhance carbohydrate ingestion.

Neuropeptide Y (NPY) injected into the paraventricular nucleus (PVN) is known to elicit a powerful feeding response in satiated, brain-cannulated rats. The present experiment investigates the effect of peptide YY (PYY), a structurally-related peptide, on feeding behavior and, in addition, the effects of both PYY and NPY on the pattern of macronutrient selection. Injection of PYY directly into the PVN, in doses ranging from 7.8 to 235 pmol/0.3 microliters, caused a strong, dose-dependent stimulation of feeding behavior, as well as a small stimulation of drinking behavior, in satiated rats. The mean latency to eat was 9.3 min, with substantial feeding occurring within 30 min of the injection. At low doses, the increase in feeding was seen predominantly during the first hr. At the highest dose, in contrast, food intake continued to increase progressively over the next few hr. such that by 4 hr postinjection food intake was more than 20 g over vehicle baseline. In 1 hr tests with 3 pure macronutrient (protein, fat and carbohydrate) diets simultaneously available, PYY and NPY (78 pmol/0.3 microliters) both elicited a strong and selective increase in carbohydrate consumption, with little or no effect on protein or fat consumption. These results suggest that hypothalamic receptors sensitive to PYY and NPY may participate in the control of carbohydrate consumption.

Microdialysis analysis of norepinephrine levels in the paraventricular nucleus in association with food intake at dark onset.

It has been proposed that norepinephrine (NE) in the paraventricular nucleus (PVN) acts to stimulate carbohydrate feeding specifically at the start of the active (dark) cycle in rats. This study used microdialysis to examine the relationship between endogenous levels of NE in the PVN at dark onset and the amount of food consumed at this time. The results indicated that: (1) in satiated rats on a lab chow diet, NE levels in the PVN, as opposed to sites anterior or lateral to this nucleus, were significantly higher during large meals around dark onset then they were during small meals or during intervals that preceded the large meals, and (2) in food-deprived animals given a pure carbohydrate diet at dark onset, PVN levels of NE just before the initiation of the meal were significantly higher in animals that consumed a relatively large carbohydrate meal (> 2.0 g) during the first dark hour, and they were significantly correlated across animals with the size of their carbohydrate meal. These findings are consistent with other evidence linking endogenous PVN NE to food intake at the beginning of the natural feeding cycle.

Evidence that neuropeptide Y and dopamine in the perifornical hypothalamus interact antagonistically in the control of food intake.

Mapping studies have revealed that the perifornical hypothalamus (PFH) is a primary locus for both the feeding-stimulatory effect of neuropeptide Y (NPY) and the anorectic effect of catecholamines (CAs), suggesting that NPY and CAs may interact antagonistically there. To investigate this, the CA-releasing agent amphetamine (AMPH) was injected through indwelling guide cannulas into the PFH of satiated adult male rats 5 min prior to injection of NPY (78 pmol/0.3 microliters) and food intake was measured 1, 2, and 4 h later. Amphetamine (50-200 nmol) dose-dependently reduced NPY feeding, usually eliminating it at the higher doses. The receptors mediating this effect were investigated by sequential injection of various CA antagonists, AMPH, and NPY into the PFH. Neither the alpha- nor beta-adrenergic receptor antagonists phentolamine (100 nmol) or propranolol (200 nmol) significantly affected AMPH suppression of NPY feeding. In contrast, the dopamine receptor antagonist haloperidol (5 nmol) abolished AMPH suppression of NPY feeding, suggesting that dopamine (DA) mediates the AMPH effect. To examine this, epinephrine (EPI, 50-200 nmol) and DA (25-200 nmol) were tested for suppression of NPY-induced feeding. While EPI had no significant effect, DA at the maximally effective dose (50 nmol) reduced the NPY feeding response by 36% or more. These findings provide convergent evidence for antagonistic interactions between endogenous DA and NPY in the control of eating behavior.

Brain regions where cholecystokinin suppresses feeding in rats.

The gut-brain peptide, cholecystokinin (CCK), inhibits food intake when injected either systemically or within the brain. To determine whether CCK's effect in the brain is anatomically specific, CCK-8 (0. 8, 4, 20, 100, 500 pmol) was microinjected into one of 14 different brain sites of rats, and its impact on subsequent food intake was measured. CCK-8 at 500 pmol significantly suppressed intake during the first hour post-injection following administration into six hypothalamic sites (anterior hypothalamus, dorsomedial hypothalamus, lateral hypothalamus, paraventricular nucleus, supraoptic nucleus, ventromedial hypothalamus) and two hindbrain sites (nucleus tractus solitarius, fourth ventricle). Although lower doses were sometimes effective (anterior hypothalamus, dorsomedial hypothalamus, nucleus tractus solitarius), there appeared to be no significant difference in potency among sites. Injections into the medial amygdala, nucleus accumbens, posterior hypothalamus, dorsal raphe, and ventral tegmental area were either ineffective or produced a delayed response. The higher doses required for most sites, as well as the widespread effectiveness of CCK-8 within the hypothalamus, suggest that spread of CCK-8 to adjacent brain sites, and (or) to the periphery, may have been required for anorexia to occur. Findings reported in an accompanying paper provide strong evidence that paraventricular nucleus injection of CCK-8 (500 pmol) did not increase plasma CCK-levels sufficiently to suppress feeding by a peripheral mechanism. Together, these results suggest that CCK may be acting as a neurotransmitter or neuromodulator within two different brain regions to produce satiety - one region which includes the nucleus tractus solitarius in the hindbrain, and another more distributed region within the medial-basal hypothalamus.

Disproportionate increases in locomotor activity in response to hormonal and photic stimuli following regional neurochemical depletions of serotonin.

The role of forebrain serotonin in behavior-related energy output was assessed in two locomotor activity tests conducted 3 and 6 months after bilateral, intrahypothalamic microinfusion of the serotonin neurotoxin, 5,7-dihydroxytryptamine (5,7-DHT). The serotonin-depleted animals exhibited a long-lasting and stable increase in energy expenditure as locomotor activity. This increased activity was investigated at the behavioral level by relating the hyperactivity to estrous cycle, photoperiod and body weight. Although the serotonin depletion-induced hyperactivity occurred in all photoperiod and estrous cycle stages, its magnitude was disproportionately increased during light and estrus. This hyperactivity could not be related to decreases in body weight because the serotonin-depleted animals weighed significantly more than the control animals. These animals responded to the weight loss that normally accompanies wheel running by increasing their activity to the same proportion as the other groups. The neuroanatomical and neurochemical substrate of the increased locomotor activity was investigated with a regional neurochemical assay for serotonin, dopamine and norepinephrine. This assay revealed that the toxin had no effect on dopamine or norepinephrine in any structure analyzed; however, serotonin was depleted in the hippocampus, septum and, to a lesser degree, in the hypothalamus. Serotonin levels were negatively correlated with overall activity. The magnitude of the disproportionate increase in activity during light and estrus was negatively correlated with hippocampal serotonin level. These results indicate that forebrain depletions of serotonin differentially affect the control of activity exerted by the phases of the photoperiod and estrous cycle. However, the modulation of activity levels by decreases in body weight remains intact.

Similar feeding patterns are induced by perifornical neuropeptide Y injection and by food deprivation.

Although hypothalamic injections of neuropeptide Y (NPY) induce robust feeding, there is little information about the patterns of feeding elicited by this peptide. To reveal these patterns, NPY (0, 8, 24, 78, 235 pmol/10 nl) was injected into the perifornical hypothalamus (PFH) of satiated adult male rats and their subsequent food intake was monitored every minute for 24 h. For comparison, feeding patterns were similarly observed following fasts of 0, 3, 6, 9, 12, and 24 h. The results demonstrated that NPY and food deprivation both produced dose- or deprivation-dependent increases in food intake that were most evident in the first 6 h. The increased intakes induced by NPY were characterized by combinations of increased meal size and frequency, with the predominant effects being increases in the size of and decreased latency to eat the first meal. Similarly, fasting progressively increased food intake by combinations of increased meal size and frequency, with the predominant effects being increases in the size of and decreased latency to eat the first meal. These similarities between NPY-induced and food deprivation-induced feeding are consistent with a stimulatory role for endogenous NPY in deprivation-induced feeding. These findings also suggest that NPY may increase eating by acting on mechanisms of both meal initiation and of meal termination.

Suppression of neuropeptide Y-elicited eating by adrenalectomy or hypophysectomy: reversal with corticosterone.

Neuropeptide Y (NPY) injected into the paraventricular hypothalamus (PVN) stimulates a robust eating response in the satiated rat. To examine whether the NPY-feeding system interacts with the pituitary-adrenal axis, the eating response to PVN injections of NPY (78 pmol) was tested in adult male rats before and after sham surgery, adrenalectomy (ADX), hypophysectomy (HYPX), and/or corticosterone (CORT) replacement therapy. In unoperated or sham groups, NPY elicited 5.7-8.8 g of food intake in 1 h as compared to 0.4-1.1 g for vehicle-injected animals. In ADX groups, the NPY-elicited response was reduced by 60-71%, to between 2.4 and 2.8 g. Likewise, the average response of the HYPX group was reduced by 69%, to 1.7 g. Corticosterone replacement, via subcutaneous implant of a 100 mg CORT pellet, normalized the NPY-induced feeding response in both the ADX and HYPX groups. These findings suggest that the hypothalamic NPY-feeding system is largely dependent upon circulating CORT and that no other adrenal or pituitary hormone is essential.

The perifornical area: the major focus of (a) patchily distributed hypothalamic neuropeptide Y-sensitive feeding system(s).

Neuropeptide Y (NPY), a neurochemical found in high concentrations within hypothalamic neurons, is believed to participate in the control of eating behavior and body energy balance and elicits a powerful eating response when injected into the hypothalamus. To delineate precisely the locus of this effect, NPY (78 pmol) or its artificial cerebrospinal fluid vehicle was injected in the extremely small volume of 10 nl through chronic guide cannulae into an array of 47 different hypothalamic areas in satiated rats and the elicited food intake was measured. To determine the anatomical resolution of this technique, the spread and recovery of [125I]NPY injected in 10 nl was also assessed. Results indicate that as much as 95% of the injected label was recovered within the brain tissue and guide cannulae and that 100% of the tissue label was localized to within 0.8 mm of the injection site. Behavioral results show that the perifornical hypothalamus (PFH), at the level of the caudal paraventricular nucleus, is the most sensitive hypothalamic site for NPY-induced eating. NPY there elicited mean increases in food intake of 12.5 g over baseline at 1 h and 20.0 g at 4 h postinjection. Injections bracketing the PFH in all directions were substantially less effective. Additionally, significant effects were also observed in at least seven other sites that were distributed throughout the hypothalamus. These findings suggest both that the PFH may be the primary hypothalamic site containing feeding-related NPY-sensitive receptors and that other sites distributed within the hypothalamus also can mediate NPY's effects.

Lateral hypothalamic injections of glutamate, kainic acid, D,L-alpha-amino-3-hydroxy-5-methyl-isoxazole propionic acid or N-methyl-D-aspartic acid rapidly elicit intense transient eating in rats.

A convergence of evidence suggests that stimulation of lateral hypothalamic (LH) neurons can elicit eating, but the neurotransmitters that mediate this effect are unknown. To determine whether glutamate might be involved, it was injected through chronic guide cannulas directly into the LH of satiated adult male rats and consequent food intake was measured. Glutamate produced a dose-dependent eating response (mean intakes of 3.7 g at 300 nmol and 5.2 g at 900 nmol) only within the first hour after injection. As a first step in determining the receptor types mediating this response, agonists for specific excitatory amino acid (EAA) receptors were similarly tested. Kainic acid (KA), D,L-alpha-amino-3-hydroxy-5-methyl-isoxazole propionic acid (AMPA) or N-methyl-D-aspartic acid (NMDA) injected into the LH each elicited eating in a dose-dependent fashion beginning at 0.33 to 1.0 nmol. At maximally effective doses (1.0-33 nmol), each agonist elicited food intakes of approximately nine grams within 1 h. Finally, analysis of meal and behavioral patterns produced by LH injection of glutamate (600 nmol) and KA (1.0 nmol) revealed that the elicited eating usually began 2-3 min postinjection and consisted of a single normal to large size meal. There were no other behavioral effects during this initial postinjection period and no effects on other oral behaviors, like drinking or gnawing, at any time. Collectively, these findings suggest that glutamate may act through several subtypes of its receptors on some LH neurons to elicit eating.

The lateral hypothalamus: a primary site mediating excitatory amino acid-elicited eating.

Lateral hypothalamic (LH) injections of the excitatory neurotransmitter glutamate, or its excitatory amino acid (EAA) agonists, kainic acid (KA), D,L-alpha-amino-3-hydroxy-5-methyl-isoxazole propionic acid (AMPA), or N-methyl-D-aspartic acid (NMDA), can rapidly elicit an intense feeding response in satiated rats. To determine whether the LH is the actual locus of this effect, we compared these compounds' ability to stimulate feeding when injected into the LH, versus when injected into sites bracketing this region. Food intake in groups of adult male rats was measured 1 h after injection of glutamate (30-900 nmol), KA (0.1-1.0 nmol), AMPA (0.33-3.3 nmol), NMDA (0.33-33.3 nmol) or vehicle, through chronically implanted guide cannulas, into one of seven brain sites. These sites were: the LH, the anterior and posterior tips of the LH, the thalamus immediately dorsal to the LH, the amygdala just lateral to the LH, or the paraventricular and perifornical areas medial to the LH. The results show that across doses and agonists the eating-stimulatory effects were largest with injections into the LH. In the LH, glutamate between 300 and 900 nmol elicited a dose-dependent eating response of up to 5 g within 1 h (P < 0.01). Each of the other agonists at doses of 3.3 nmol or less elicited eating responses of at least 10 g with injections into this site. Injections into the other brain sites produced either no eating, or occasionally smaller and less consistent eating responses.(ABSTRACT TRUNCATED AT 250 WORDS)