Functional consequences of reduction in NMDA receptor glycine affinity in mice carrying targeted point mutations in the glycine binding site.

We have used site-directed mutagenesis in conjunction with homologous recombination to generate two mouse lines carrying point mutations in the glycine binding site of the NMDAR1 subunit (Grin1). Glycine concentration-response curves from acutely dissociated hippocampal neurons revealed a 5- and 86-fold reduction in receptor glycine affinity in mice carrying Grin1(D481N) and Grin1(K483Q) mutations, respectively, whereas receptor glutamate affinity remained unaffected. Homozygous mutant Grin1(D481N) animals are viable and fertile and appear to develop normally. However, homozygous mutant Grin1(K483Q) animals are significantly lighter at birth, do not feed, and die within a few days. No gross abnormalities in CNS anatomy were detected in either Grin1(D481N) or Grin1(K483Q) mice. Interestingly, in situ hybridization and Western blot analysis revealed changes in the expression levels of NMDA receptor subunits in Grin1(D481N) mice relative to wild type that may represent a compensatory response to the reduction in receptor glycine affinity. Grin1(D481N) mice exhibited deficits in hippocampal theta burst-induced long-term potentiation (LTP) and spatial learning and also a reduction in sensitivity to NMDA-induced seizures relative to wild-type controls, consistent with a reduced activation of NMDA receptors. Mutant mice exhibited normal prepulse inhibition but showed increased startle reactivity. Preliminary analysis indicated that the mice exhibit a decreased natural aversion to an exposed environment. The lethal phenotype of Grin1(K483Q) animals confirms the critical role of NMDA receptor activation in neonatal survival. A milder reduction in receptor glycine affinity results in an impairment of LTP and spatial learning and alterations in anxiety-related behavior, providing further evidence for the role of NMDA receptor activation in these processes.

Influence of the selective ORL1 receptor agonist, Ro64-6198, on rodent neurological function.

Identification of synthetic agonists and antagonists at orphan receptors represents an important step for understanding their physiological function and therapeutic potential. Accordingly, we have recently described a non-peptide agonist at the opioid receptor like (ORL1) receptor (1S,3aS)-8-(2,3,3a,4,5,6-hexahydro-1H-phenalen-1-yl)-1-phenyl-1,3,8-triazaspiro[4.5]decan-4-one (Ro64-6198; Jenck et al., PNAS 94 (2000) 4938; Wichmann et al., Eur. J. Med. Chem. 35 (2000) 839). We have investigated the effects of this compound in various tests of rodent neurological function, utilising ORL1 knockout mice to examine the pharmacological specificity of Ro64-6198. In male C57BL/6J mice, effects on balance and motor co-ordination were detected following low doses (0.3-1mg/kg IP) of Ro64-6198. At higher doses (1-3mg/kg IP), effects on swim behaviour and hypothermia was observed. At 10mg/kg, each effect became more profound and a severe neurological disturbance appeared, including loss of righting reflex. These effects of Ro64-6198 (10mg/kg IP) were absent in ORL1 receptor knockout mice. In male, hooded Lister rats, Ro64-6198 (6-10mg/kg IP), produced some disturbance of neurological function, including hypoactivity, rotarod performance, grip strength and mild hypothermia. An impairment of food responding under a variable interval (VI) 20s schedule of reinforcement was noted at 3mg/kg. These results confirm Ro64-6198 to be a highly selective pharmacological tool to investigate ORL1 receptor function in vivo and, furthermore, that activation of this receptor is accompanied by a variety of effects on neurological function.

Novel agonists of 5HT2C receptors. Synthesis and biological evaluation of substituted 2-(indol-1-yl)-1-methylethylamines and 2-(indeno[1,2-b]pyrrol-1-yl)-1-methylethylamines. Improved therapeutics for obsessive compulsive disorder.

The syntheses of a series of substituted 2-(indol-1-yl)-1-methylethylamines and 2-(indeno[1,2- b]pyrrol-1-yl)-1-methylethylamines are reported. The binding affinities of the compounds at 5HT2C and 5HT2A receptors (79% homology in the transmembrane domain) were determined. The ligands displayed selectivity for 5HT2C receptors relative to 5HT2A receptors. Compounds were functionally characterized both in vitro and in vivo as 5HT2C receptor agonists. 5f, 5l, 5n, 5o, 5q, 14c, 14f, 14k, and 14m exhibited anticompulsive activity in an animal model of obsessive compulsive disorder.

High-affinity, non-peptide agonists for the ORL1 (orphanin FQ/nociceptin) receptor.

The discovery of 8-(5,8-dichloro-1,2,3,4-tetrahydro-naphthalen-2-yl)-1-phenyl-1,3,8-triazaspiro[4.5]decan-4-one, 1a, as a high-affinity ligand for the human ORL1 (orphanin FQ/nociceptin) receptor led to the synthesis of a series of optimized ligands. These compounds exhibit high affinity for the human ORL1 receptor, exhibit moderate to good selectivity versus opioid receptors, and behave as full agonists in biochemical assays. In this paper we present the synthesis, structure-activity relationship (SAR), and biochemical characterization of substituted 1-phenyl-1,3,8-triazaspiro[4.5]decan-4-ones culminating in the discovery of 8-(5-methyl-1,2,3,4-tetrahydro-naphthalen-1-yl)-1-phenyl-1,3,8-triazaspiro[4.5]decan-4-one, 1p, and 8-acenaphten-1-yl-1-phenyl-1,3,8-triazaspiro[4.5]decan-4-one 1q, two high-affinity, potent ORL1 receptor agonists with good to moderate selectivity versus the other opioid receptors.

Involvement of 5-HT1C-receptors in drug-induced penile erections in rats.

Drug-induced penile erections (PE) were initially suggested to be 5-HT1B receptor mediated. However, since the discovery of the 5-HT1C receptor a number of compounds, considered to be 5-HT1B selective, appear to bind more strongly to the 5-HT1C receptor and this prompted a re-evaluation of the receptor subtype involved in PE induction. PE could be induced by the 5-HT agonists mCPP (0.22-2.2 mg/kg), TFMPP (0.46-1.0 mg/kg) and MK 212 (0.1-1.0 mg/kg). The 5-HT agonist DOI (0.022-0.22 mg/kg) did not induce PE in placebo-pretreated rats but in rats pretreated with various 5-HT2 antagonists it did. These compounds have in common a strong affinity for the 5-HT1C receptor, mCPP (0.46 mg/kg)-induced PE could be antagonized by the 5-HT antagonists metergoline, cyproheptadine, mesulergine, mianserin, ritanserin and ketanserin. Their ED50S were 0.04, 0.4, 0.03, 0.06, 0.4 and 2 mg/kg, respectively. The potency of both the agonists to induce, and the antagonists to inhibit, PE was found to be dependent on their selectivity for the 5-HT1C receptor versus the 5-HT2 receptor. Spiperone (0.1-1.0 mg/kg) and GR 38032F (1-10 mg/kg) did not antagonise mCPP-induced PE. 8-OH-DPAT and 5MeODMT counteracted mCPP (0.46 mg/kg)-induced PE. Their ED50S were 0.03 and 0.4 mg/kg, respectively. DOI counteracted mCPP induced PE only at doses above 1 mg/kg, whereas CGS 12066B (1.0-10 mg/kg) was inactive. The results suggest that PE are induced by activation of the 5-HT1C receptor and are functionally inhibited by activation of 5-HT1A or 5-HT2 receptors.

Precipitated withdrawal in squirrel monkeys after repeated daily oral administration of alprazolam, diazepam, flunitrazepam or oxazepam.

The lowest dose of alprazolam, diazepam, flunitrazepam and oxazepam consistently to induce loss of righting reflex in squirrel monkeys or vehicle was orally administered to monkeys on 18 consecutive days: 2 mg/kg alprazolam (n = 4), 30 mg/kg diazepam (n = 4), 1 mg/kg flunitrazepam (n = 4), 280 mg/kg oxazepam (n = 5), or vehicle (n = 4). Tolerance developed rapidly for loss of righting reflex, more slowly for sleep and only minimally for muscle relaxation observed during the period immediately following daily oral administration. Injection of the specific benzodiazepine receptor antagonist flumazenil (10 mg/kg i.v.) 5 h after the ninth daily oral treatment produced signs of precipitated withdrawal (tremor, vomiting and/or convulsions) in one alprazolam-, four diazepam-, one flunitrazepam- and four oxazepam-treated monkeys, but not in the vehicle-treated monkeys. Physiological saline injected intravenously several days later under these same experimental conditions failed to provoke a precipitated withdrawal reaction. When flumazenil-induced precipitated withdrawal was again evaluated after the 18th daily oral treatment, withdrawal signs were observed in all alprazolam- and all diazepam-treated monkeys, as well as in three flunitrazepam- and three oxazepam-treated monkeys, but not in the vehicle-treated monkeys (convulsions were observed in one alprazolam-, two diazepam-, one flunitrazepam- and two oxazepam-treated monkeys). No signs of spontaneous withdrawal were observed in any of the monkeys during a subsequent 3-week drug-free period. Thus, repeated administration of approximately equieffective doses of these four benzodiazepines resulted in a similar development of tolerance and physical dependence (indicated by the occurrence of a precipitated withdrawal reaction).

Drug-induced potentiation of prepulse inhibition of acoustic startle reflex in mice: a model for detecting antipsychotic activity?

RATIONALE: Schizophrenic patients typically have impaired startle habituation (SH) and prepulse inhibition of the startle reflex (PPI). PPI can be disrupted in rats by psychomimetics, and drug-induced reversal of this deficit is considered to predict potential antipsychotic properties. Certain strains of mice, such as C57BL/6J, naturally display poor PPI. OBJECTIVE: To test whether mice spontaneously showing low levels of PPI might prove a useful tool for detecting novel antipsychotics. METHODS: PPI and SH were evaluated in four strains of mice: BALB/cByJ, MORO, 129/SvEv and C57BL/6J. The effects of antipsychotic [haloperidol (1, 3 and 6 mg/kg), clozapine (0.3, 1, 3 and 30 mg/kg) and risperidone (0.1, 0.3 and 1 mg/kg)] and non-antipsychotic [diazepam (3, 10 and 30 mg/kg), buspirone (1, 3 and 10 mg/kg), desipramine (3, 10 and 30 mg/kg), morphine (3, 10 and 30 mg/kg) and scopolamine (0.3, 1 and 3 mg/kg)] drug treatments were studied on PPI. RESULTS: Haloperidol (6 mg/kg), clozapine (3 and 30 mg/kg), and risperidone (1 mg/kg) all significantly enhanced PPI in C57BL/6J. All non-antipsychotics failed to improve PPI in this strain, except diazepam. Facilitation of PPI was also obtained in the other strains; however, clear interstrain differences were observed depending on the class of antipsychotic used and on the level of prepulse intensity. CONCLUSION: Antipsychotic-induced facilitation of PPI is clearly detected in mice naturally exhibiting poor levels of sensorimotor gating (e.g., C57BL/6J), but is also observed in other strains of mice. The use of this procedure as a potential screening test for detecting novel antipsychotic medications is discussed.

Effects of serotonin receptor antagonists on PAG stimulation induced aversion: different contributions of 5HT1, 5HT2 and 5HT3 receptors.

The effects of serotonin receptor antagonists with differential selectivity for the various classes of 5HT receptors (5HT1, 5HT2 and 5HT3) were tested for their effects on the response to aversive brain stimulation. Electrical stimulation was administered to the dorsal part of the periaqueductal gray matter (PAG), one of the main cerebral structures subserving negative reinforcement. Stimulation frequency thresholds for escape responses were recorded before and following administration of the compounds. Ketanserin (0.32-32 mg/kg IP), trazodone (1.0-22 mg/kg), pirenperone (0.032-1.0 mg/kg) and spiperone (0.1-0.2 mg/kg) dose-dependently increased stimulation frequency thresholds necessary to induce escape responses. Opposite effects were observed with mianserin (0.01-32 mg/kg) and metergoline (0.032-10 mg/kg) which decreased threshold for escape. ICS 205-930 (0.01-10 mg/kg), did not affect the stimulation frequency threshold for escape. Prazosin (0.1-22 mg/kg) did not specifically affect aversive brain stimulation. Haloperidol (0.02-1.0 mg/kg) increased the frequency threshold for escape responses but with some motoric side effects. These data show that the various types of 5HT receptors differentially contribute to the control of central aversive systems in rats. It is suggested that blockade of 5HT2 receptors suppresses the central aversive system, whereas blockade of some 5HT1 receptors enhances aversion and overcomes the 5HT2-mediated suppression. Blockade of 5HT3 receptors has no effects. Dopamine receptor blockade further contributes to the suppression of the central aversive system. The relevance of these findings to some pathophysiological mechanisms of anxiety and depressive disorders is discussed.

5-HT1C receptors in the serotonergic control of periaqueductal gray induced aversion in rats.

The functional role of brain 5-HT and 5-HT receptor subtypes in periaqueductal gray (PAG) induced aversion has been investigated in rats. Antiaversive effects were found with the serotonin agonists TFMPP, mCPP and DOI but not with RU 24969 which was found to facilitate PAG aversion. The first three serotonin agonists share potent 5-HT1C activity while RU 24969 differs with a high 5-HT1A activity. Proaversive effects were found with the mixed 5-HT1C/5-HT2 antagonists cyproheptadine and ritanserin; this effect was already reported for the mixed 5-HT1C/5-HT2 antagonists metergoline and mianserin and is opposite to the effects of the selective 5-HT2 antagonists ketanserin, pirenperone, trazodone and spiperone. The antiaversive effects of mCPP (1 mg/kg) could be prevented by pretreatment of the animals with mianserin (1 and 10 mg/kg). These results suggest that 5-HT1C receptors play an important role in the serotonergic control of PAG aversion. 5-HT1C receptor activation seems to mediate antiaversive effects whereas acute 5-HT1C receptor blockade appears to facilitate PAG aversion. Functional interactions take place between several receptor types in the in vivo control of PAG aversion, where 5-HT1C receptors appear to play a predominant function.

Effects of repeated mild stress and two antidepressant treatments on the behavioral response to 5HT1C receptor activation in rats.

This study investigated the possible involvement of 5HT1C receptors in the development of depressive states and in the mode of action of antidepressants. The effects of repeated unpredictable mild stress (a regimen known to induce an anhedonic state in the rat) and of chronic administration of either of two recognized antidepressant treatments (sleep deprivation or inhibition of monoamine oxidase type A) in rats were studied on a 5HT1C receptor initiated response, i.e. mCPP-induced penile erection. A 3-week period of repeated, but unpredictable exposure to mild stressors induced a shift to the left of the dose-response curve for mCPP-induced penile erection. In contrast, 72-h REM sleep deprivation resulted in a shift to the right of the mCPP dose-response curve and 10-day administration of the monoamine oxidase type A inhibitor moclobemide (20 mg/kg IP bid) also resulted in a decreased number of mCPP-induced penile erections. These findings support the hypothesis that neuronal activities initiated via 5HT1C receptor stimulation may play a role in the pathophysiology and treatment of depression.

Pharmacological characterization of benzodiazepine receptor ligands with intrinsic efficacies ranging from high to zero.

Several benzodiazepine receptor ligands were pharmacologically characterized in a battery of functional tests after oral administration in mice, rats, and monkeys. Previous experiments have consistently demonstrated that diazepam exhibits high intrinsic efficacy, bretazenil exhibits intermediate intrinsic efficacy, Ro 42-8773 and Ro 41-7812 both show low intrinsic efficacy, and flumazenil exhibits virtually zero intrinsic efficacy. In the test battery used here it appears that nearly full intrinsic efficacy is required for clear anterograde amnesia or rotarod impairment. In contrast, full protection in the pentetetrazol test was achieved with intermediate-to-high intrinsic efficacy and nearly full protection with lower intrinsic efficacy. In the audiogenic seizure test full anticonvulsant effects were produced with intrinsic efficacy ranging from low to high. Clear inhibition of punished operant responding was observed for all test compounds except for Ro 41-7812 and flumazenil, which exhibit the lowest intrinsic efficacies. All of the test compounds enhanced palatable food consumption, with even those having low intrinsic efficacy producing maximum effects approximating that of diazepam. By additionally taking into consideration the degree of receptor occupancy required to obtain pharmacological activity in each of the tests in this battery it is possible to order the compounds with respect to intrinsic efficacy: diazepam > bretazenil > Ro 42-8773 > Ro 41-7812 > flumazenil. The latter four compounds all exhibited a maximum antagonistic activity in tests involving reversal of meclonazepam- or flunitrazepam-induced central nervous system depression. Thus, using these tests appears to permit the accurate ordinal classification of benzodiazepine receptor ligands for intrinsic efficacy.(ABSTRACT TRUNCATED AT 250 WORDS)

Acute and chronic treatment with 5-HT reuptake inhibitors differentially modulate emotional responses in anxiety models in rodents.

This study investigated behavioural effects of very potent 5-HT reuptake inhibitors after acute treatment (cianopramine and citalopram), as well as after chronic treatment (cianopramine), in two behavioural models of anxiety: 1) the light/dark choice procedure in mice and 2) the elevated plus-maze test in rats. In addition, the responses of mice to novelty in a free exploration paradigm were assessed after acute administration of both drugs. A single injection of cianopramine or citalopram increased neophobic reactions in the free exploration test. Furthermore, these drugs increased the avoidance reaction to a brightly illuminated chamber in the light/dark choice procedure as well as to open arms in the elevated plus-maze test. In contrast, after chronic treatment (10 mg/kg IP, once daily for 21 days) of cianopramine, anxiogenic-like effects were no longer produced in the light/dark choice paradigm whereas in the elevated plus-maze test, anxiolytic-like effects appeared. These results shed more light on the 5-HT hypothesis of anxiety, insofar as the increased availability of 5-HT resulting here from reuptake inhibition seems to initially result in an increased emotional reactivity which, however, subsequently disappears during chronic treatment.

Urocortin, a novel neuropeptide with anxiogenic-like properties.

Potential anxiogenic-like properties of urocortin, a corticotropin-releasing factor (CRF)-related neuropeptide, were investigated in models of anxiety in rodents. In the elevated plus-maze, CRF- and urocortin-treated rats (0.1 nmol, i.c.v.) spent less time and made fewer entries into open arms. In the light-dark test in mice, urocortin (0.006-0.06 nmol, i.c.v.) dose-dependently reduced time and number of transitions into the lit area. Urocortin also dose-dependently (0.02-0.2 nmol, i.c.v.) reduced mice exploratory behaviour in an open field. This effect was reversed by diazepam (0.1-1 mg/kg, i.p) and by the CRF receptor antagonist alpha-helical CRF (0.8-8 nmole, i.c.v.). These data show that urocortin produces anxiety-like effects in several behavioural paradigms in rodents.

Aniracetam improves radial maze performance in rats.

The memory enhancing effect of the pyrrolidinone derivative aniracetam was investigated in rats trained in a delayed-response task in an 8-arm radial maze. Oral administration of aniracetam (100, 200, 400, or 800 mg kg-1) 16 h and again 1 h prior to a first trial of exposure to a given configuration of 4 baited arms resulted in a significant improvement in performance during a second trial in the maze given 3 h later in which there was access to all 8 arms but only the other 4 arms were baited. The pattern of baited arms was varied daily. The performance enhancement was greatest for the highest doses. These results extend the demonstration of the cognition enhancing effects of aniracetam to a spatial memory task in rats.

Impact of environmental housing conditions on the emotional responses of mice deficient for nociceptin/orphanin FQ peptide precursor gene.

Nociceptin/orphanin FQ (N/OFQ) is a newly discovered neuropeptide that has been implicated in the neurobiological regulation of the behavioral responses to stress and fear. To investigate the role of this peptide in the expression of stress/anxiety-related behaviors in mice, a gene targeting approach to disrupt N/OFQ in the pre-proN/OFQ gene was used. The impact of environmental housing conditions (single and social housing) was assessed on N/OFQ-knockout male and female mice in different experimental paradigms known to trigger distinctive types of stress and anxiety states. Neurological examination of homozygous mutant adult animals indicated that basic neurological functions (vision, audition, olfaction, tactile and pain sensitivity, motor performances) were normal. When housed individually, N/OFQ-knockout animals displayed responses similar to control animals in behavioral tests of emotional reactivity (behavioral despair, locomotor activity, light-dark preference, and acoustic startle tests). In contrast, increased emotional responses were detected when individually housed mice were crowded together (five per cage) under conditions of competitive access to food, water, space, and social contacts. Under those conditions, male mice deficient for N/OFQ developed greater home-cage aggression and increased fear/anxiety-like behaviors in the light-dark and acoustic startle tests, when compared to their wild-type littermates. Group-housed female mutants also showed higher level of anxiety in the acoustic startle test, but needed additional restrain stress to express detectable levels of anxiety in the light-dark test. These data indicate a clear environment-induced rise in fear reactions of N/OFQ-knockout mice. They further suggest that N/OFQ system is essential for development of adequate coping strategies to acute and chronic stress.

A neuropharmacological study of the periventricular neural substrate involved in flight.

This paper reviews results obtained in experiments concerning the neurochemical characteristics of the substrate involved in the control of flight reactions and the induction of aversive effects in the rat. These experiments investigated the behavioural effects produced by microinjecting into the periaqueductal grey matter (PAG) or the medial hypothalamus (MH) compounds known to interfere with the functioning of some neurotransmitter systems known to exist in these structures. The data obtained show that: the activity of the substrate involved in the production of flight reactions is tonically inhibited by the release of GABA (gamma-aminobutyric acid); the behavioural reactions produced by microinjecting GABA antagonists can be clearly distinguished, depending on whether such drugs were injected into the PAG or the MH, despite the fact that jumps were produced from either level; behavioural effects, comparable to some extent to those produced by microinjections of GABA antagonists, can be obtained by injecting drugs which act on non-GABAergic neurochemical substrates, namely opioidergic or cholinergic systems; and behavioural effects, comparable to those produced by injecting GABA antagonists into the PAG, can be obtained by injecting such drugs into various sites located in other parts of the tectum such as the inferior colliculus or adjacent structures.

Morphine injected into the periaqueductal gray attenuates brain stimulation-induced effects: an intensity discrimination study.

A discrimination situation was used to study the effects of morphine microinjected into the periaqueductal gray (PAG) on the aversive effects induced by PAG or medial hypothalamic (MH) electrical stimulation. Rats were trained in a T-maze to interrupt a high intensity (HI) stimulation inducing a short escape latency by pressing the lever (HI lever) located in one arm of the maze and a low intensity (LI) stimulation inducing a longer escape latency by pressing the lever (LI lever) located in the other arm. Microinjections of 15 nmol (5 micrograms) or 40 nmol (13 micrograms) of morphine both lengthened the escape latencies and shifted towards the LI lever the animal's choice in order to interrupt HI stimulations. This effect of morphine showed a similar time course as regards both escape latency and lever choice; it was more marked on PAG than on MH stimulation-induced aversive effects. The data are discussed in terms of morphine microinjections into PAG lengthening the escape latency by decreasing the aversiveness of PAG or MH stimulation rather than by affecting the animals' ability to respond to such stimulations.

Contraversive circling induced by ventral tegmental microinjections of moderate doses of morphine and [D-Pen2, D-Pen5]enkephalin.

Unilateral ventral tegmental area (VTA) injections of morphine and [D-Pen2,D-Pen5]enkephalin (DPDPE), caused contraversive circling at doses of 1.2, 12, and 24 nmol. Similar doses of the selective kappa-agonist U-50,488H were ineffective. These data suggest a common mechanism for the circling, locomotion and facilitation of brain stimulation reward caused by VTA morphine, and distinguish this mechanism for that of feeding which is caused by both morphine and kappa-actions in this region.

Opioid receptor subtypes associated with ventral tegmental facilitation of lateral hypothalamic brain stimulation reward.

Rats were trained to lever-press for lateral hypothalamic electrical stimulation, and tested following ventral tegmental microinjections of morphine, delta ([D-Pen2,D-Pen5]enkephalin: DPDPE), or kappa (U-50, 488H) receptor agonists or saline. Across the range of effective stimulation frequencies, morphine (8 nmol) and DPDPE (8 nmol) reduced the current necessary to sustain responding, while U-50,488H (8 nmol) had no effect. Naloxone (1 mg/kg) blocked the effects of morphine and DPDPE. Since each of these opioids facilitates eating induced by stimulation of the lateral hypothalamus, but only the mu and delta agonists facilitate brain stimulation reward, it would appear that ventral tegmental kappa receptors are linked to circuit elements which play a role in eating but not lateral hypothalamic brain stimulation reward. Ventral tegmental mu and delta receptors, on the other hand, appear to be linked to circuit elements involved in both behaviors.

Opioid receptor subtypes associated with ventral tegmental facilitation and periaqueductal gray inhibition of feeding.

Eating was induced in sated animals by lateral hypothalamic electrical stimulation following central microinjections of mu- (morphine), delta-([D-Pen2,D-Pen5]enkephalin) or kappa-(U-50,488H) receptor agonists, or saline. With stimulation intensity fixed at a moderate level, time to eat 3 45-mg food pellets decreased with increases in stimulation frequency, approaching an asymptote near 7 s at ca. 70 Hz. Ventral tegmental injections (8 but not 0.8 nmol) of each of the 3 drugs reduced the minimum frequency required to produce eating of 3 pellets within 20 s and reduced the frequency at which asymptotic performance was produced; the drugs were equally effective at these doses. Naloxone (2 mg/kg) reversed the effects of each drug; naloxone was slightly more effective against morphine than against DPDPE or U-50,488H. These data suggest that all 3 receptor classes may contribute to the ventral tegmental facilitation of feeding. Periaqueductal gray injections (16 but not 1.6 nmol) of morphine had the opposite effect; they increased the stimulation frequency required to cause eating of 3 pellets in 20 s, and decreased the speed of eating across all stimulation frequencies. Periaqueductal gray injections of the delta- and kappa-agonists were each without effect. These data indicate that the periaqueductal gray inhibition of feeding is mediated solely by mu-receptors and their associated periaqueductal gray circuitry.