- Email: [email protected]
Discriminative stimulus effects of morphine: central versus peripheral training
Brain Research 847 Ž1999. 26–31 www.elsevier.comrlocaterbres
Research report
Discriminative stimulus effects of morphine: central versus peripheral training James P. Cleary b , Eugene O’Hare a , James D. Pomonis a , Patricia L. Dittel a , Jacki J. Hofmeister a , Melinda M. Fritz a , Charles J. Billington c,e , Allen S. Levine d,e,) a
Veterans Affairs Medical Center, Minneapolis MN, USA Department of Psychology, UniÕersity of Minnesota, Minneapolis, USA c Department of Medicine, UniÕersity of Minnesota, Minneapolis, USA d Department of Psychiatry, UniÕersity of Minnesota, Minneapolis, USA Minnesota Obesity Center, VA Medical Center, 1 Veterans DriÕe, Minneapolis, MN 55417, USA b
e
Accepted 17 August 1999
Abstract While it is well known that rats can discriminate a peripheral injection of morphine from a saline injection, to our knowledge no one has trained rats to discriminate a direct brain-site injection of morphine from saline. In the present series of studies, one group of rats was trained to discriminate morphine Ž0.3 mg. from saline injected into the perifornical area of the hypothalamus ŽPFA., a process that took rats about 37 sessions to learn. A dose response generalization curve for PFA-injected morphine Ž0.01, 0.03, 0.1, and 0.17 mg. was generated in which the two highest doses of morphine generalized to the morphine-appropriate training stimulus. Intraperitoneal Ži.p.. injection of 3 mgrkg, but not 1 mgrkg morphine, resulted in morphine-appropriate responding in the PFA morphine-trained rats. A second group of rats was trained to discriminate i.p. injections of 3 mgrkg morphine from injections of saline. A dose–response generalization test for i.p.-injected morphine Ž0.3, 0.56, 1.0, and 1.7 mgrkg. was conducted in which the 0.17 mgrkg dose of morphine generalized to the morphine-appropriate training stimulus. Generalization tests using PFA-injected morphine doses Ž0.17, 0.56, 1.0, and 3.0 mg. failed to result in morphine-appropriate responding in the i.p. morphine-trained rats. Naloxone administered into the PFA Ž50 mg. or the periphery Ž3 mgrkg, i.p.. blocked morphine discrimination in the PFA-trained rats. However, when naloxone was injected into the PFA Ž50 mg. together with i.p. morphine Ž3 mgrkg. in animals trained using i.p. injections, the antagonist failed to block morphine-appropriate responding. Thus, while peripheral injection of morphine generalized to the discriminative stimulus effects of morphine produced under PFA-injection training, the opposite effects were not noted. q 1999 Elsevier Science B.V. All rights reserved. Keywords: Discrimination; Morphine; Naloxone; Paraventricular nucleus; Perifornical area
1. Introduction The discriminative stimulus properties of morphine are well established w3,4,7,9x. Using these properties, it has been demonstrated that structurally dissimilar opioid agonists produce internal stimulus effects similar to that of morphine w3x. It has also been shown that opioid receptor antagonists can block morphine’s discriminative effects and can interfere with other opioid agonists’ ability to generalize to the morphine stimulus in morphine–saline
) Corresponding [email protected]
author.
Fax:
q 1-612-725-2093;
e-mail:
discrimination w9,14x. A few studies have attempted to identify the specific areas of the brain responsible for morphine’s discriminative effects. After successful training under the morphine-saline discrimination procedure, direct injections into the periaqueductal gray w17x, parabrachial nucleus w10x and ventral tegmental area w18x have produced generalization to the morphine stimulus, although these results are somewhat inconsistent. To our knowledge, however, no one has extended morphine discrimination or generalization to subjects that were trained to discriminate morphine from saline by repeated direct central injection. The current study tests the hypothesis that discrimination of morphine from saline can be trained and maintained by direct injection of these compounds into the
0006-8993r99r$ - see front matter q 1999 Elsevier Science B.V. All rights reserved. PII: S 0 0 0 6 - 8 9 9 3 Ž 9 9 . 0 2 0 0 1 - 6
J.P. Cleary et al.r Brain Research 847 (1999) 26–31
perifornical area of the hypothalamus ŽPFA. in rats. The PFA is an area directly adjacent to the paraventricular nucleus of the hypothalamus ŽPVN. which is a brain site where morphine alters rat behavior, namely food intake w1,13,16x. We chose the PFA to avoid damaging cells in the PVN resulting from the repeated injections necessary during discrimination training. There is considerable evidence that opioids play an important role in PVN function. Opioid agonists are known to increase feeding and opioid antagonists decrease feeding after injection into the PVN w2,5,6x. The messenger RNA for both mu and kappa opioid receptors are found in the PVN w15x. In addition, the PVN is known to be involved in a number of other systems which are related to opioid system function, including sexual behavioral and thermogenesis w11,12x. In addition to assessing discrimination training by direct brain injection, cross generalization studies were conducted using peripheral and central morphine injections and training. Finally, we tested the opioid antagonist naloxone’s ability to block discrimination under both central and peripheral injection procedures, after both central and peripheral discrimination training. Discrimination of drug stimuli following injection into selected brain nuclei should help identify central sites responsible for a drug’s stimulus profile.
2. Materials and methods 2.1. Discrimination training by intracranial injection Five male Sprague–Dawley rats, approximately 90 days old and weighing 250–275 g at the beginning of the experiment, were housed individually with free access to water. Initially, rats were maintained at 95% of their free-feeding weights and trained to press both levers in two-lever rat test chambers Žmodel E10-10, Coulbourn Instruments, Lehigh Valley, PA.. After animals learned to reliably press both levers for food reinforcement Ž45 mg pellets., they were anesthetized with sodium pentobarbital Ž40 mgrkg i.p.. and 26-gauge guide cannulae were aimed at the right PFA of the brain. Stereotaxic coordinates with the incisor bar set at 3.5 mm above the interaural line, were 0.75 mm lateral and 1.8 mm posterior to bregma, and 8.0 mm below the surface of the skull. The animals were allowed to recover for 7 days following surgery. During discrimination training, animals were injected via the indwelling cannula with either 0.3 mgr0.5 ml morphine or 0.5 ml saline, 25 min prior to the start of a session. Morphine sulfate was solubilized with 0.9% sodium chloride solution. Injections were given using a 5.0 ml microsyringe attached to a 33-gauge internal injection cannula which fit snugly into the guide cannula. The injection cannula extended 0.5 mm past the end of the guide cannula tip. Injections were given slowly, over a 3-min period, with the animal allowed to roam freely on a
27
table top. Following injection, internal injection cannulae were withdrawn and stainless steel stylets were placed in the guide cannulae. Immediately after injection animals were returned to their home cages until time for the training session Ž25 min.. Sessions preceded by morphine administration alternated randomly with those preceded by saline, with the restriction that no more than two consecutive sessions of either saline or morphine occurred. Training sessions started with 5 min of darkness during which time responses had no consequences. After this 5 min acclimation period, the house light was illuminated and training began. Responses on the right lever were reinforced following morphine administration and responses on the left lever were reinforced following saline administration. Completing the response requirement on the drug-inappropriate lever produced an 8 sec period of darkness Žtime-out., during which time lever presses had no consequences. Completing the response requirement on the drug-appropriate lever produced a food pellet. During the early training sessions, the number of responses required for delivery of each reinforcer or time-out was gradually increased from 1 to 20 responses Žfixed-ratio 20 or FR-20. for four animals. One animal did not perform well under the 20 response per reinforcer requirement and this animal was maintained under a fixed ratio 10 ŽFR 10. reinforcement schedule. Individual sessions ended after 25 reinforcers or time-outs were delivered. Animals were trained until at least 85% of responses prior to the first consequence Žreinforcement or time-out. occurred at the appropriate lever in 8 out of 10 consecutive sessions. This criterion was slightly more strict than typically employed for discrimination studies in our laboratory. After each animal met the training success criterion, test sessions utilizing PFA-injected morphine doses of 0.01, 0.03, 0.1, and 0.17 mgr0.5 ml were conducted instead of training sessions. During test sessions, one of the test doses was substituted for the training dose, no reinforcer was delivered, and the session ended after the first completed ratio of 20 responses. Each test session was separated by at least four training sessions. Following completion of the intracranial morphine dose-effect test regimen, naloxone’s ability to block morphine discrimination of the training dose was tested by administering either PFA test injections of 50 mgr0.5 ml of naloxone or i.p. injections of 3.0 mgrkg naloxone, twenty five minutes prior to a test session that also employed the training dose of PFA-injected morphine Ž0.3 mgr0.5 ml.. Naloxone was solubilized with 0.9% sodium chloride solution. Finally, generalization tests of i.p.-injected morphine at doses of 0.0, 1.0, and 3.0 mgrkg were conducted. These final i.p. morphine tests employed only four subjects because one rat died just prior to this testing phase. Whenever i.p. injections were used for testing, 0.9% saline was injected into the PFA cannula as a control for handling-related stress.
28
J.P. Cleary et al.r Brain Research 847 (1999) 26–31
2.2. Discrimination training by peripheral i.p. injection Seven male Sprague–Dawley rats were individually housed and maintained similarly to those described above, except that they were initially deprived to 80% of the their free feeding weights to facilitate lever press training. Experimental chambers and conditions, discrimination training, testing and injections were the same as described above except for those changes detailed below. On training days, subjects were injected i.p. with either morphine Ž3.0 mgrkg. or saline Ž0.9%., 25 min before the start of the session. Injection volumes were equal and the injection schedule was random except that neither morphine nor saline was ever given on more than two consecutive days. The injection-session interval was 25 min. As above, the first 5 min of each session consisted of darkness, with lever presses having no consequences, followed by house light illumination and lever activation. Responding on the drug-appropriate lever resulted in food reinforcement while responding on the drug-inappropriate lever resulted in time-out Ž8 s.. The final response requirement necessary for reinforcer delivery was 20 per reinforcer Žfixed-ratio 20 or FR 20.. Sessions lasted until 25 reinforces had been earned or until 25 min had elapsed, whichever came first. Subjects that met the criterion of successful discrimination entered the testing phase when at least 80% of responses prior to the first consequence Žreinforcement or time-out. occurred at the appropriate lever in 8 out of 10 consecutive sessions. As on training days, test days began with i.p. injection of either saline Ž0.9%. or morphine Ž0.3, 0.56, 1.0, 1.7, or 3.0 mgrkg.. Test sessions were always separated by at least two successful Ž) 80% correct responding before the first reinforcer or time out. training days. On test days, 20 responses to either lever ended the session without reinforcement and the animal was immediately returned to its home cage. After generalization to each i.p. morphine test dose was determined twice, intracranial cannulae were implanted exactly as described above, with cannulae terminating in the PFA. After recovery from surgery, additional training sessions insured i.p. morphine discrimination accuracy and accustomed the subjects to pre-session PFA injections Ž0.9% saline.. Following this, additional generalization test doses of morphine Ž0.17, 0.56, 1.0, and 3.0 mg in 1 ml. were administered directly into the PFA. Naloxone Ž50 mg in 1 ml. plus the training dose of morphine was also tested by injection into the PFA. Injections were given twenty five minutes prior to the session. IP naloxone Ž3.0 mgrkg. was also tested concurrently with the training dose of morphine. Finally, all subjects were given cannulae patency tests with neuropeptide Y ŽNPY.. NPY injected into the PFA produces voracious eating due to direct action on PFA neurons or by infusion into the paraventricular nucleus ŽPVN. which is adjacent to the PFA. For this test, subjects were allowed free access to food for two days in their
home cages. Injections of saline Ž1.0 ml, 0.9%. and NPY Ž1.0 mg in 1 ml. were administered on separate days and food consumption in the home cage was measured 2 h after injection. 2.3. Statistics and data analysis All drug-appropriate responding data were collected prior to the first completed ratio, before a reinforcer or time-out was given. Data for each dose-effect relationship under i.p.- or PFA-trained conditions were first subjected to Repeated Measures Analysis of Variance ŽRMANOVA. tests of significance. Only if the RMANOVA was significant were multiple comparison tests Ž t LSD . applied to individual dose-effect relationships. The protected multiple comparison p values are given in the text.
3. Results Rats learned to discriminate PFA injections of 0.3 mg morphine from injections of saline Ž0.9%. in an average of 37.4 sessions Žrange s 26–43.. The criterion for successful discrimination was selected as 85% of responses, prior to the first consequence, occurring at the appropriate lever in
Fig. 1. Dose–effect relationship for PFA-injected ŽB. and i.p.-injected morphine Žv . after discrimination training by direct injection of morphine into the PFA. Training doses were 0.3 mgr0.5 ml morphine or 0.5 ml of 0.9% saline. Data are from all responses emitted during completion of the first ratio Ži.e., before the first reinforcer or time out.. Asterisks ŽU . indicate responding that was significantly different than that produced by saline generalization tests. Brackets enclose "1 S.E.M.
J.P. Cleary et al.r Brain Research 847 (1999) 26–31
8 of 10 consecutive sessions. After criterion was met, the mean percentage of correct responses to the morphine lever under the training dose of PFA-injected morphine was 85% ŽS.E.M.s 6%.. Fig. 1 depicts the effect of doses of morphine under test conditions in which the session was terminated after completion of the first ratio Ži.e., 20 responses to either lever.. Mean responses to the morphine lever were 0% ŽS.E.M.s 0%., 20% Ž20%., 42% Ž19%., and 60% Ž24%., under respective PFA morphine doses of 0.01, 0.03, 0.1, and 0.17 mg. There was a significant main effect for dose Ž F s 5.5, df s 5,20 p0 - .01. and the two higher doses of morphine were significantly different from saline Ž p - 0.05, t LSD .. As can be seen from Fig. 1, responding on the morphine-appropriate lever after PFAinjected morphine, was directly related to morphine dose. Fig. 1 also depicts the effect of i.p.-injected morphine in animals trained by injections into the PFA. Mean responses to the morphine-appropriate lever after i.p. injections were 0% ŽS.E.M.s 0%., 25% Ž6%., and 100% Ž0%. at respective doses of 0.0, 1.0, and 3.0 mgrkg morphine. Peripherally injected morphine showed significant generalization to the morphine-appropriate lever Ž F s 13.0, df s 2.6, p - .01., but only at the highest dose of 3.0 mgrkg Ž p - .05, t LSD .. When rats were trained under i.p. injections of either 3.0 mgrkg morphine or an equivalent volume of saline
Fig. 2. Dose–effect relationship for PFA-injected ŽB. and i.p.-injected morphine Žv . after discrimination training by peripheral injection ŽIP. of morphine. Training doses were 3.0 mgrkg morphine or 1.0 mlrkg 0.9% saline. Data are from all responses emitted during completion of the first ratio Ži.e., before the first reinforcer or time out.. Asterisks ŽU . indicate responding that was significantly different than that produced by saline generalization tests. Brackets enclose "1 S.E.M.
29
Fig. 3. Effects of the opioid antagonist naloxone or saline on PFA- and i.p.-trained morphinersaline discrimination. Data are from all responses emitted during completion of the first ratio Ži.e., before the first reinforcer or time out..
Ž0.9%., they exhibited typical generalization dose–effect relationships. This effect is depicted in Fig. 2. Mean percent morphine-appropriate responses under i.p.-injected morphine were 0.0% ŽS.E.M.s 0%., 14% Ž13.6%., 20% Ž14.0%., 14% Ž14.3%., 66% Ž17.1%., and 97% Ž1.9%. at respective doses of 0.0, 0.3, 0.56, 1.0, 1.7, and 3.0 mgrkg Ž F s 11.2, df s 5,30, p - 0.01.. The highest two doses Ž1.7 and 3.0 mgrkg. were significantly different than saline in their generalization toward the morphine-appropriate lever in these peripherally trained animals Ž p .05, t LSD .. When these subjects were generalization tested under PFA injections, no dose of morphine tested produced significantly different discrimination from saline. Mean morphine-appropriate responses were 0.0% ŽS.E.M.s 0%., 10% Ž10.6%., 4% Ž5.1%., 22% Ž16.0%., and 26% Ž16.6%. at respective doses of 0.0, 0.17, 0.56, 1.0, and 3.0 mg. The PFA-injected doses tested included doses shown to be discriminable under PFA-training Ž0.17 mg. and doses several times higher Ž0.56, 1.0, and 3.0 mg.. Fig. 3 depicts the effect of the opioid antagonist naloxone on PFA-trained and i.p.-trained morphine discrimination. As might be expected, when the morphine discrimination was trained using PFA morphine Ž0.3 g. and saline injections, naloxone Ž50 mgr0.5 ml PFA. blocked morphine-appropriate responding under 0.3 mg morphine injected into the PFA. Peripherally injected Ži.p.. naloxone Ž3.0 mgrkg. also blocked morphine-appropriate responding under PFA trained and tested morphine. However, when naloxone was injected centrally Ž50 mg PFA. together with i.p. morphine Ž3.0 mgrkg. in animals trained using i.p. injections, the antagonist failed to block morphine-appropriate responding. Under PFA-trained and ad-
30
J.P. Cleary et al.r Brain Research 847 (1999) 26–31
Fig. 4. Drawings of coronal sections of rat brain showing cannulae placement. The upper section Ž1 cannula. is 1.8 mm posterior to Bregma and the lower section Ž4 cannulae. is 2.12 mm posterior to Bregma.
ministered morphine, saline produced 86% ŽS.E.M.s 6.7%. and naloxone produced 1% Ž0%. and 1% Ž0%. morphine-appropriate responding at respective naloxone doses of 50 mg ŽPFA. and 3.0 mgrkg Ži.p... Under i.p.-trained and administered morphine, naloxone Ž50 mg, PFA. produced 80% Ž10.4%. morphine-appropriate responding, while saline Ž0.9%. plus i.p. morphine Ž3.0 mgrkg. produced 100% morphine-appropriate responding. Fig. 4 illustrates the location of cannulae termination Žfilled dots. in drawings of coronal sections of rat brain for subjects trained under i.p. morphine injections. The cannulae terminated in the PFA, very near the paraventricular nucleus of the hypothalamus ŽPVN.. Cannulae patency in the PFA-trained group was assessed daily by successful discrimination of morphine and saline that was maintained throughout the study. For i.p.-trained subjects, the patency of the cannulae was assessed at the end of the study by injection of 1 mg of neuropeptide Y ŽNPY. under free feeding conditions. NPY injected into the PFA produces robust eating. All subjects showed dramatic increases in food consumed under NPY. The mean food consumption under PFA-injected saline Ž1.0 ml. was 4.1 g Žrange 2.1– 7.3. and under NPY Ž1.0 mg. was 11.4 g Žrange 8.1–15.6., with each animal at least doubling the amount eaten when injected with NPY. 4. Discussion We found that morphine Ž0.3 mg. administered into the PFA produced a discriminative stimulus that was distinctive from saline. The morphine doses of 0.3 and 0.17 mg which supported discrimination and generalization in our study are at least 1000 times lower than those used under peripheral training procedures. However, relatively high doses of morphine Ž3 mgrkg i.p.. given peripherally were necessary to generalize to the 0.3 mg dose given into the PFA. We also found, as have many others, that rats can discriminate a peripheral dose of morphine Ž3 mgrkg. from saline and that a peripheral injection of 1.7 mgrkg morphine generalizes to the training dose w7,8x. In this case, doses as high as 3 mg of morphine injected into the PFA failed to result in a generalization in rats trained to
discriminate 3 mgrkg of morphine peripherally. Others have noted that morphine injected into the ventral tegmental area w18x, the parabrachial nucleus w10x, or the periaqueductal gray w17x generalized in rats trained to discriminate subcutaneously injected morphine. In the ventral tegmental area, doses of 1–3 mg of morphine generalized in the peripherally injected morphine rats. However, in the parabrachial nucleus higher doses were needed Ž5–20 mg.. Clearly, a dose of 3 mg of morphine into the PFA is not sufficient to mimic the interoceptive stimulus effects of a 3 mgrkg dose of morphine. However, a rat can readily be trained to discriminate a much smaller dose Ž0.3 mg. of morphine in the PFA if that training is by PFA injection. Therefore, even though morphine in the PFA produces discriminative effects, it does not appear to be involved in the discriminative stimulus effects of peripherally injected morphine. We also noted that 50 mg of naloxone injected into the PFA and 3.0 mgrkg of naloxone injected peripherally, blocked morphine-appropriate responding under a relatively high morphine dose Ž3.0 mg. injected into the PFA. However, 50 mg of naloxone failed to affect morphine-appropriate responding in animals trained by i.p. injection of 3.0 mgrkg morphine. Once again, such data suggest that areas capable of supporting discriminated morphine effects are not necessarily involved in the discriminative stimulus effects of peripherally injected morphine. We chose to inject morphine into the PFA, an area directly adjacent to the PVN. This brain site was chosen since the PVN is a site where morphine alters rat behavior, namely food intake w1,13,16x. We found that PFA injection of morphine resulted in morphine becoming a discriminative stimulus. While peripheral injection of morphine generalized to the discriminative stimulus effects of morphine produced under direct injection training, the opposite effects were not observed. That is, relatively high doses of morphine injected into the PFA did not mimic the effects of a peripheral injection of morphine. Discrimination of morphine following direct site injection may help understand the neural pathways involved in interoceptive effects of morphine and perhaps may identify sites relevant to opiate abuse.
J.P. Cleary et al.r Brain Research 847 (1999) 26–31
Acknowledgements This work was supported by a Department of Veterans Affairs Merit Award ŽJPC and ASL.. P. Dittel was supported by a Research Experience for Undergraduate Grant ŽNSF SBR-96-00021. and E. O’Hare was supported under DA-07097. Additional support was provided by DA-03999 ŽASL.. We thank Dr. David Jewett for his assistance in the preparation of this manuscript. References w1x C.J. Billington, J.E. Briggs, S. Harker, M. Grace, A.S. Levine, Neuropeptide Y in hypothalamic paraventricular nucleus: a center coordinating energy metabolism, Am. J. Physiol. 266 Ž1994. R1765–R1770. w2x R.K., Bodnar, Opioid receptor subtype antagonist and ingestion, in: S.J. Cooper, P.G. Clifton ŽEds.., Drug Receptor Subtypes and Ingestive Behaviour, Academic Press, London, 1996, pp. 127–146. w3x F.C. Colpaert, Discriminative stimulus properties of narcotic analgesic drugs, Pharmacol. Biochem. Behav. 9 Ž1978. 863–887. w4x F.C. Colpaert, J.J. Kuyps, C.J. Niemegeers, P.A. Janssen, Discriminative stimulus properties of fentanyl and morphine: tolerance and dependence, Pharmacol. Biochem. Behav. 5 Ž1976. 401–408. w5x B.A. Gosnell, A.S. Levine, Stimulation of ingestive behavior by preferential and selective opioid agonists, in: S.J. Cooper, P.G. Clifton ŽEds.., Drug Receptor Subtypes and Ingestive Behaviour, Academic Press, London, 1996, pp. 147–166. w6x B.A. Gosnell, J.E. Morley, A.S. Levine, Opioid-induced feeding: localization of sensitive brain sites, Brain Res. 369 Ž1986. 177–184. w7x H.E. Hill, B.E. Jones, E.C. Bell, State dependent control of discrimination by morphine and pentobarbital, Psychopharmacologia 22 Ž1971. 305–313.
31
w8x S.G. Holtzman, Drug discrimination studies, Drug Alcohol Depend. 14 Ž1985. 263–282. w9x S.G. Holtzman, K.W. Locke, Neural mechanisms of drug stimuli: experimental approaches, in: F.C. Colpaert, R.L. Balster ŽEds.., Transduction Mechanisms of Drug Stimuli, Springer, Berlin, 1988, pp. 139–153. w10x T.V. Jaeger, D. van der Kooy, Morphine acts in the parabrachial nucleus, a pontine viscerosensory relay, to produce discriminative stimulus effects, Psychopharmacology 110 Ž1993. 76–84. w11x S.P. Kalra, J.T. Clark, A. Sahu, M.G. Dube, P.S. Kalra, Control of feeding and sexual behaviors by neuropeptide Y: physiological implications, Synapse 2 Ž1988. 254–257. w12x C.M. Kotz, J. Briggs, M. Grace, A.S. Levine, C.J. Billington, Divergence of the feeding and thermogenic pathways influenced by neuropeptide Y in the hypothalamic paraventricular nucleus of the rat, Am. J. Physiol. 275 Ž1998. R471–R477. w13x S.F. Leibowitz, C. Sladek, L. Spencer, D. Tempel, Neuropeptide Y, epinephrine and norepinephrine in the paraventricular nucleus: stimulation of feeding and the release of corticosterone, vasopressin and glucose, Brain Res. Bull. 21 Ž1988. 905–912. w14x K.W. Locke, S.G. Holtzman, Behavioral effects of opioid peptides selective for mu or delta receptors: I. Morphine-like discriminative stimulus effects, J. Pharmacol. Exp. Ther. 238 Ž1986. 990–996. w15x A. Mansour, C.A. Fox, H. Akil, S.J. Watson, Opioid-receptor mRNA expression in the rat CNS: anatomical and functional implications, TINS 18 Ž1995. 22–29. w16x J.D. Pomonis, A.S. Levine, C.J. Billington, Interaction of the hypothalamic paraventricular nucleus and central nucleus of the amygdala in naloxone blockade of neuropeptide Y-induced feeding revealed by c-fos expression, J. Neurosci. 17 Ž1997. 5175–5182. w17x J. Roscrans, G. Krynock, A possible role of the PAG in the mediation of subjective effects of morphine, Pharmacologist 19 Ž1977. 171–176. w18x M. Shoaib, R. Spanagel, Mesolimbic sites mediate the discriminative stimulus effects of morphine, Eur. J. Pharmacol. 252 Ž1994. 69–75.
Research report
Discriminative stimulus effects of morphine: central versus peripheral training James P. Cleary b , Eugene O’Hare a , James D. Pomonis a , Patricia L. Dittel a , Jacki J. Hofmeister a , Melinda M. Fritz a , Charles J. Billington c,e , Allen S. Levine d,e,) a
Veterans Affairs Medical Center, Minneapolis MN, USA Department of Psychology, UniÕersity of Minnesota, Minneapolis, USA c Department of Medicine, UniÕersity of Minnesota, Minneapolis, USA d Department of Psychiatry, UniÕersity of Minnesota, Minneapolis, USA Minnesota Obesity Center, VA Medical Center, 1 Veterans DriÕe, Minneapolis, MN 55417, USA b
e
Accepted 17 August 1999
Abstract While it is well known that rats can discriminate a peripheral injection of morphine from a saline injection, to our knowledge no one has trained rats to discriminate a direct brain-site injection of morphine from saline. In the present series of studies, one group of rats was trained to discriminate morphine Ž0.3 mg. from saline injected into the perifornical area of the hypothalamus ŽPFA., a process that took rats about 37 sessions to learn. A dose response generalization curve for PFA-injected morphine Ž0.01, 0.03, 0.1, and 0.17 mg. was generated in which the two highest doses of morphine generalized to the morphine-appropriate training stimulus. Intraperitoneal Ži.p.. injection of 3 mgrkg, but not 1 mgrkg morphine, resulted in morphine-appropriate responding in the PFA morphine-trained rats. A second group of rats was trained to discriminate i.p. injections of 3 mgrkg morphine from injections of saline. A dose–response generalization test for i.p.-injected morphine Ž0.3, 0.56, 1.0, and 1.7 mgrkg. was conducted in which the 0.17 mgrkg dose of morphine generalized to the morphine-appropriate training stimulus. Generalization tests using PFA-injected morphine doses Ž0.17, 0.56, 1.0, and 3.0 mg. failed to result in morphine-appropriate responding in the i.p. morphine-trained rats. Naloxone administered into the PFA Ž50 mg. or the periphery Ž3 mgrkg, i.p.. blocked morphine discrimination in the PFA-trained rats. However, when naloxone was injected into the PFA Ž50 mg. together with i.p. morphine Ž3 mgrkg. in animals trained using i.p. injections, the antagonist failed to block morphine-appropriate responding. Thus, while peripheral injection of morphine generalized to the discriminative stimulus effects of morphine produced under PFA-injection training, the opposite effects were not noted. q 1999 Elsevier Science B.V. All rights reserved. Keywords: Discrimination; Morphine; Naloxone; Paraventricular nucleus; Perifornical area
1. Introduction The discriminative stimulus properties of morphine are well established w3,4,7,9x. Using these properties, it has been demonstrated that structurally dissimilar opioid agonists produce internal stimulus effects similar to that of morphine w3x. It has also been shown that opioid receptor antagonists can block morphine’s discriminative effects and can interfere with other opioid agonists’ ability to generalize to the morphine stimulus in morphine–saline
) Corresponding [email protected]
author.
Fax:
q 1-612-725-2093;
e-mail:
discrimination w9,14x. A few studies have attempted to identify the specific areas of the brain responsible for morphine’s discriminative effects. After successful training under the morphine-saline discrimination procedure, direct injections into the periaqueductal gray w17x, parabrachial nucleus w10x and ventral tegmental area w18x have produced generalization to the morphine stimulus, although these results are somewhat inconsistent. To our knowledge, however, no one has extended morphine discrimination or generalization to subjects that were trained to discriminate morphine from saline by repeated direct central injection. The current study tests the hypothesis that discrimination of morphine from saline can be trained and maintained by direct injection of these compounds into the
0006-8993r99r$ - see front matter q 1999 Elsevier Science B.V. All rights reserved. PII: S 0 0 0 6 - 8 9 9 3 Ž 9 9 . 0 2 0 0 1 - 6
J.P. Cleary et al.r Brain Research 847 (1999) 26–31
perifornical area of the hypothalamus ŽPFA. in rats. The PFA is an area directly adjacent to the paraventricular nucleus of the hypothalamus ŽPVN. which is a brain site where morphine alters rat behavior, namely food intake w1,13,16x. We chose the PFA to avoid damaging cells in the PVN resulting from the repeated injections necessary during discrimination training. There is considerable evidence that opioids play an important role in PVN function. Opioid agonists are known to increase feeding and opioid antagonists decrease feeding after injection into the PVN w2,5,6x. The messenger RNA for both mu and kappa opioid receptors are found in the PVN w15x. In addition, the PVN is known to be involved in a number of other systems which are related to opioid system function, including sexual behavioral and thermogenesis w11,12x. In addition to assessing discrimination training by direct brain injection, cross generalization studies were conducted using peripheral and central morphine injections and training. Finally, we tested the opioid antagonist naloxone’s ability to block discrimination under both central and peripheral injection procedures, after both central and peripheral discrimination training. Discrimination of drug stimuli following injection into selected brain nuclei should help identify central sites responsible for a drug’s stimulus profile.
2. Materials and methods 2.1. Discrimination training by intracranial injection Five male Sprague–Dawley rats, approximately 90 days old and weighing 250–275 g at the beginning of the experiment, were housed individually with free access to water. Initially, rats were maintained at 95% of their free-feeding weights and trained to press both levers in two-lever rat test chambers Žmodel E10-10, Coulbourn Instruments, Lehigh Valley, PA.. After animals learned to reliably press both levers for food reinforcement Ž45 mg pellets., they were anesthetized with sodium pentobarbital Ž40 mgrkg i.p.. and 26-gauge guide cannulae were aimed at the right PFA of the brain. Stereotaxic coordinates with the incisor bar set at 3.5 mm above the interaural line, were 0.75 mm lateral and 1.8 mm posterior to bregma, and 8.0 mm below the surface of the skull. The animals were allowed to recover for 7 days following surgery. During discrimination training, animals were injected via the indwelling cannula with either 0.3 mgr0.5 ml morphine or 0.5 ml saline, 25 min prior to the start of a session. Morphine sulfate was solubilized with 0.9% sodium chloride solution. Injections were given using a 5.0 ml microsyringe attached to a 33-gauge internal injection cannula which fit snugly into the guide cannula. The injection cannula extended 0.5 mm past the end of the guide cannula tip. Injections were given slowly, over a 3-min period, with the animal allowed to roam freely on a
27
table top. Following injection, internal injection cannulae were withdrawn and stainless steel stylets were placed in the guide cannulae. Immediately after injection animals were returned to their home cages until time for the training session Ž25 min.. Sessions preceded by morphine administration alternated randomly with those preceded by saline, with the restriction that no more than two consecutive sessions of either saline or morphine occurred. Training sessions started with 5 min of darkness during which time responses had no consequences. After this 5 min acclimation period, the house light was illuminated and training began. Responses on the right lever were reinforced following morphine administration and responses on the left lever were reinforced following saline administration. Completing the response requirement on the drug-inappropriate lever produced an 8 sec period of darkness Žtime-out., during which time lever presses had no consequences. Completing the response requirement on the drug-appropriate lever produced a food pellet. During the early training sessions, the number of responses required for delivery of each reinforcer or time-out was gradually increased from 1 to 20 responses Žfixed-ratio 20 or FR-20. for four animals. One animal did not perform well under the 20 response per reinforcer requirement and this animal was maintained under a fixed ratio 10 ŽFR 10. reinforcement schedule. Individual sessions ended after 25 reinforcers or time-outs were delivered. Animals were trained until at least 85% of responses prior to the first consequence Žreinforcement or time-out. occurred at the appropriate lever in 8 out of 10 consecutive sessions. This criterion was slightly more strict than typically employed for discrimination studies in our laboratory. After each animal met the training success criterion, test sessions utilizing PFA-injected morphine doses of 0.01, 0.03, 0.1, and 0.17 mgr0.5 ml were conducted instead of training sessions. During test sessions, one of the test doses was substituted for the training dose, no reinforcer was delivered, and the session ended after the first completed ratio of 20 responses. Each test session was separated by at least four training sessions. Following completion of the intracranial morphine dose-effect test regimen, naloxone’s ability to block morphine discrimination of the training dose was tested by administering either PFA test injections of 50 mgr0.5 ml of naloxone or i.p. injections of 3.0 mgrkg naloxone, twenty five minutes prior to a test session that also employed the training dose of PFA-injected morphine Ž0.3 mgr0.5 ml.. Naloxone was solubilized with 0.9% sodium chloride solution. Finally, generalization tests of i.p.-injected morphine at doses of 0.0, 1.0, and 3.0 mgrkg were conducted. These final i.p. morphine tests employed only four subjects because one rat died just prior to this testing phase. Whenever i.p. injections were used for testing, 0.9% saline was injected into the PFA cannula as a control for handling-related stress.
28
J.P. Cleary et al.r Brain Research 847 (1999) 26–31
2.2. Discrimination training by peripheral i.p. injection Seven male Sprague–Dawley rats were individually housed and maintained similarly to those described above, except that they were initially deprived to 80% of the their free feeding weights to facilitate lever press training. Experimental chambers and conditions, discrimination training, testing and injections were the same as described above except for those changes detailed below. On training days, subjects were injected i.p. with either morphine Ž3.0 mgrkg. or saline Ž0.9%., 25 min before the start of the session. Injection volumes were equal and the injection schedule was random except that neither morphine nor saline was ever given on more than two consecutive days. The injection-session interval was 25 min. As above, the first 5 min of each session consisted of darkness, with lever presses having no consequences, followed by house light illumination and lever activation. Responding on the drug-appropriate lever resulted in food reinforcement while responding on the drug-inappropriate lever resulted in time-out Ž8 s.. The final response requirement necessary for reinforcer delivery was 20 per reinforcer Žfixed-ratio 20 or FR 20.. Sessions lasted until 25 reinforces had been earned or until 25 min had elapsed, whichever came first. Subjects that met the criterion of successful discrimination entered the testing phase when at least 80% of responses prior to the first consequence Žreinforcement or time-out. occurred at the appropriate lever in 8 out of 10 consecutive sessions. As on training days, test days began with i.p. injection of either saline Ž0.9%. or morphine Ž0.3, 0.56, 1.0, 1.7, or 3.0 mgrkg.. Test sessions were always separated by at least two successful Ž) 80% correct responding before the first reinforcer or time out. training days. On test days, 20 responses to either lever ended the session without reinforcement and the animal was immediately returned to its home cage. After generalization to each i.p. morphine test dose was determined twice, intracranial cannulae were implanted exactly as described above, with cannulae terminating in the PFA. After recovery from surgery, additional training sessions insured i.p. morphine discrimination accuracy and accustomed the subjects to pre-session PFA injections Ž0.9% saline.. Following this, additional generalization test doses of morphine Ž0.17, 0.56, 1.0, and 3.0 mg in 1 ml. were administered directly into the PFA. Naloxone Ž50 mg in 1 ml. plus the training dose of morphine was also tested by injection into the PFA. Injections were given twenty five minutes prior to the session. IP naloxone Ž3.0 mgrkg. was also tested concurrently with the training dose of morphine. Finally, all subjects were given cannulae patency tests with neuropeptide Y ŽNPY.. NPY injected into the PFA produces voracious eating due to direct action on PFA neurons or by infusion into the paraventricular nucleus ŽPVN. which is adjacent to the PFA. For this test, subjects were allowed free access to food for two days in their
home cages. Injections of saline Ž1.0 ml, 0.9%. and NPY Ž1.0 mg in 1 ml. were administered on separate days and food consumption in the home cage was measured 2 h after injection. 2.3. Statistics and data analysis All drug-appropriate responding data were collected prior to the first completed ratio, before a reinforcer or time-out was given. Data for each dose-effect relationship under i.p.- or PFA-trained conditions were first subjected to Repeated Measures Analysis of Variance ŽRMANOVA. tests of significance. Only if the RMANOVA was significant were multiple comparison tests Ž t LSD . applied to individual dose-effect relationships. The protected multiple comparison p values are given in the text.
3. Results Rats learned to discriminate PFA injections of 0.3 mg morphine from injections of saline Ž0.9%. in an average of 37.4 sessions Žrange s 26–43.. The criterion for successful discrimination was selected as 85% of responses, prior to the first consequence, occurring at the appropriate lever in
Fig. 1. Dose–effect relationship for PFA-injected ŽB. and i.p.-injected morphine Žv . after discrimination training by direct injection of morphine into the PFA. Training doses were 0.3 mgr0.5 ml morphine or 0.5 ml of 0.9% saline. Data are from all responses emitted during completion of the first ratio Ži.e., before the first reinforcer or time out.. Asterisks ŽU . indicate responding that was significantly different than that produced by saline generalization tests. Brackets enclose "1 S.E.M.
J.P. Cleary et al.r Brain Research 847 (1999) 26–31
8 of 10 consecutive sessions. After criterion was met, the mean percentage of correct responses to the morphine lever under the training dose of PFA-injected morphine was 85% ŽS.E.M.s 6%.. Fig. 1 depicts the effect of doses of morphine under test conditions in which the session was terminated after completion of the first ratio Ži.e., 20 responses to either lever.. Mean responses to the morphine lever were 0% ŽS.E.M.s 0%., 20% Ž20%., 42% Ž19%., and 60% Ž24%., under respective PFA morphine doses of 0.01, 0.03, 0.1, and 0.17 mg. There was a significant main effect for dose Ž F s 5.5, df s 5,20 p0 - .01. and the two higher doses of morphine were significantly different from saline Ž p - 0.05, t LSD .. As can be seen from Fig. 1, responding on the morphine-appropriate lever after PFAinjected morphine, was directly related to morphine dose. Fig. 1 also depicts the effect of i.p.-injected morphine in animals trained by injections into the PFA. Mean responses to the morphine-appropriate lever after i.p. injections were 0% ŽS.E.M.s 0%., 25% Ž6%., and 100% Ž0%. at respective doses of 0.0, 1.0, and 3.0 mgrkg morphine. Peripherally injected morphine showed significant generalization to the morphine-appropriate lever Ž F s 13.0, df s 2.6, p - .01., but only at the highest dose of 3.0 mgrkg Ž p - .05, t LSD .. When rats were trained under i.p. injections of either 3.0 mgrkg morphine or an equivalent volume of saline
Fig. 2. Dose–effect relationship for PFA-injected ŽB. and i.p.-injected morphine Žv . after discrimination training by peripheral injection ŽIP. of morphine. Training doses were 3.0 mgrkg morphine or 1.0 mlrkg 0.9% saline. Data are from all responses emitted during completion of the first ratio Ži.e., before the first reinforcer or time out.. Asterisks ŽU . indicate responding that was significantly different than that produced by saline generalization tests. Brackets enclose "1 S.E.M.
29
Fig. 3. Effects of the opioid antagonist naloxone or saline on PFA- and i.p.-trained morphinersaline discrimination. Data are from all responses emitted during completion of the first ratio Ži.e., before the first reinforcer or time out..
Ž0.9%., they exhibited typical generalization dose–effect relationships. This effect is depicted in Fig. 2. Mean percent morphine-appropriate responses under i.p.-injected morphine were 0.0% ŽS.E.M.s 0%., 14% Ž13.6%., 20% Ž14.0%., 14% Ž14.3%., 66% Ž17.1%., and 97% Ž1.9%. at respective doses of 0.0, 0.3, 0.56, 1.0, 1.7, and 3.0 mgrkg Ž F s 11.2, df s 5,30, p - 0.01.. The highest two doses Ž1.7 and 3.0 mgrkg. were significantly different than saline in their generalization toward the morphine-appropriate lever in these peripherally trained animals Ž p .05, t LSD .. When these subjects were generalization tested under PFA injections, no dose of morphine tested produced significantly different discrimination from saline. Mean morphine-appropriate responses were 0.0% ŽS.E.M.s 0%., 10% Ž10.6%., 4% Ž5.1%., 22% Ž16.0%., and 26% Ž16.6%. at respective doses of 0.0, 0.17, 0.56, 1.0, and 3.0 mg. The PFA-injected doses tested included doses shown to be discriminable under PFA-training Ž0.17 mg. and doses several times higher Ž0.56, 1.0, and 3.0 mg.. Fig. 3 depicts the effect of the opioid antagonist naloxone on PFA-trained and i.p.-trained morphine discrimination. As might be expected, when the morphine discrimination was trained using PFA morphine Ž0.3 g. and saline injections, naloxone Ž50 mgr0.5 ml PFA. blocked morphine-appropriate responding under 0.3 mg morphine injected into the PFA. Peripherally injected Ži.p.. naloxone Ž3.0 mgrkg. also blocked morphine-appropriate responding under PFA trained and tested morphine. However, when naloxone was injected centrally Ž50 mg PFA. together with i.p. morphine Ž3.0 mgrkg. in animals trained using i.p. injections, the antagonist failed to block morphine-appropriate responding. Under PFA-trained and ad-
30
J.P. Cleary et al.r Brain Research 847 (1999) 26–31
Fig. 4. Drawings of coronal sections of rat brain showing cannulae placement. The upper section Ž1 cannula. is 1.8 mm posterior to Bregma and the lower section Ž4 cannulae. is 2.12 mm posterior to Bregma.
ministered morphine, saline produced 86% ŽS.E.M.s 6.7%. and naloxone produced 1% Ž0%. and 1% Ž0%. morphine-appropriate responding at respective naloxone doses of 50 mg ŽPFA. and 3.0 mgrkg Ži.p... Under i.p.-trained and administered morphine, naloxone Ž50 mg, PFA. produced 80% Ž10.4%. morphine-appropriate responding, while saline Ž0.9%. plus i.p. morphine Ž3.0 mgrkg. produced 100% morphine-appropriate responding. Fig. 4 illustrates the location of cannulae termination Žfilled dots. in drawings of coronal sections of rat brain for subjects trained under i.p. morphine injections. The cannulae terminated in the PFA, very near the paraventricular nucleus of the hypothalamus ŽPVN.. Cannulae patency in the PFA-trained group was assessed daily by successful discrimination of morphine and saline that was maintained throughout the study. For i.p.-trained subjects, the patency of the cannulae was assessed at the end of the study by injection of 1 mg of neuropeptide Y ŽNPY. under free feeding conditions. NPY injected into the PFA produces robust eating. All subjects showed dramatic increases in food consumed under NPY. The mean food consumption under PFA-injected saline Ž1.0 ml. was 4.1 g Žrange 2.1– 7.3. and under NPY Ž1.0 mg. was 11.4 g Žrange 8.1–15.6., with each animal at least doubling the amount eaten when injected with NPY. 4. Discussion We found that morphine Ž0.3 mg. administered into the PFA produced a discriminative stimulus that was distinctive from saline. The morphine doses of 0.3 and 0.17 mg which supported discrimination and generalization in our study are at least 1000 times lower than those used under peripheral training procedures. However, relatively high doses of morphine Ž3 mgrkg i.p.. given peripherally were necessary to generalize to the 0.3 mg dose given into the PFA. We also found, as have many others, that rats can discriminate a peripheral dose of morphine Ž3 mgrkg. from saline and that a peripheral injection of 1.7 mgrkg morphine generalizes to the training dose w7,8x. In this case, doses as high as 3 mg of morphine injected into the PFA failed to result in a generalization in rats trained to
discriminate 3 mgrkg of morphine peripherally. Others have noted that morphine injected into the ventral tegmental area w18x, the parabrachial nucleus w10x, or the periaqueductal gray w17x generalized in rats trained to discriminate subcutaneously injected morphine. In the ventral tegmental area, doses of 1–3 mg of morphine generalized in the peripherally injected morphine rats. However, in the parabrachial nucleus higher doses were needed Ž5–20 mg.. Clearly, a dose of 3 mg of morphine into the PFA is not sufficient to mimic the interoceptive stimulus effects of a 3 mgrkg dose of morphine. However, a rat can readily be trained to discriminate a much smaller dose Ž0.3 mg. of morphine in the PFA if that training is by PFA injection. Therefore, even though morphine in the PFA produces discriminative effects, it does not appear to be involved in the discriminative stimulus effects of peripherally injected morphine. We also noted that 50 mg of naloxone injected into the PFA and 3.0 mgrkg of naloxone injected peripherally, blocked morphine-appropriate responding under a relatively high morphine dose Ž3.0 mg. injected into the PFA. However, 50 mg of naloxone failed to affect morphine-appropriate responding in animals trained by i.p. injection of 3.0 mgrkg morphine. Once again, such data suggest that areas capable of supporting discriminated morphine effects are not necessarily involved in the discriminative stimulus effects of peripherally injected morphine. We chose to inject morphine into the PFA, an area directly adjacent to the PVN. This brain site was chosen since the PVN is a site where morphine alters rat behavior, namely food intake w1,13,16x. We found that PFA injection of morphine resulted in morphine becoming a discriminative stimulus. While peripheral injection of morphine generalized to the discriminative stimulus effects of morphine produced under direct injection training, the opposite effects were not observed. That is, relatively high doses of morphine injected into the PFA did not mimic the effects of a peripheral injection of morphine. Discrimination of morphine following direct site injection may help understand the neural pathways involved in interoceptive effects of morphine and perhaps may identify sites relevant to opiate abuse.
J.P. Cleary et al.r Brain Research 847 (1999) 26–31
Acknowledgements This work was supported by a Department of Veterans Affairs Merit Award ŽJPC and ASL.. P. Dittel was supported by a Research Experience for Undergraduate Grant ŽNSF SBR-96-00021. and E. O’Hare was supported under DA-07097. Additional support was provided by DA-03999 ŽASL.. We thank Dr. David Jewett for his assistance in the preparation of this manuscript. References w1x C.J. Billington, J.E. Briggs, S. Harker, M. Grace, A.S. Levine, Neuropeptide Y in hypothalamic paraventricular nucleus: a center coordinating energy metabolism, Am. J. Physiol. 266 Ž1994. R1765–R1770. w2x R.K., Bodnar, Opioid receptor subtype antagonist and ingestion, in: S.J. Cooper, P.G. Clifton ŽEds.., Drug Receptor Subtypes and Ingestive Behaviour, Academic Press, London, 1996, pp. 127–146. w3x F.C. Colpaert, Discriminative stimulus properties of narcotic analgesic drugs, Pharmacol. Biochem. Behav. 9 Ž1978. 863–887. w4x F.C. Colpaert, J.J. Kuyps, C.J. Niemegeers, P.A. Janssen, Discriminative stimulus properties of fentanyl and morphine: tolerance and dependence, Pharmacol. Biochem. Behav. 5 Ž1976. 401–408. w5x B.A. Gosnell, A.S. Levine, Stimulation of ingestive behavior by preferential and selective opioid agonists, in: S.J. Cooper, P.G. Clifton ŽEds.., Drug Receptor Subtypes and Ingestive Behaviour, Academic Press, London, 1996, pp. 147–166. w6x B.A. Gosnell, J.E. Morley, A.S. Levine, Opioid-induced feeding: localization of sensitive brain sites, Brain Res. 369 Ž1986. 177–184. w7x H.E. Hill, B.E. Jones, E.C. Bell, State dependent control of discrimination by morphine and pentobarbital, Psychopharmacologia 22 Ž1971. 305–313.
31
w8x S.G. Holtzman, Drug discrimination studies, Drug Alcohol Depend. 14 Ž1985. 263–282. w9x S.G. Holtzman, K.W. Locke, Neural mechanisms of drug stimuli: experimental approaches, in: F.C. Colpaert, R.L. Balster ŽEds.., Transduction Mechanisms of Drug Stimuli, Springer, Berlin, 1988, pp. 139–153. w10x T.V. Jaeger, D. van der Kooy, Morphine acts in the parabrachial nucleus, a pontine viscerosensory relay, to produce discriminative stimulus effects, Psychopharmacology 110 Ž1993. 76–84. w11x S.P. Kalra, J.T. Clark, A. Sahu, M.G. Dube, P.S. Kalra, Control of feeding and sexual behaviors by neuropeptide Y: physiological implications, Synapse 2 Ž1988. 254–257. w12x C.M. Kotz, J. Briggs, M. Grace, A.S. Levine, C.J. Billington, Divergence of the feeding and thermogenic pathways influenced by neuropeptide Y in the hypothalamic paraventricular nucleus of the rat, Am. J. Physiol. 275 Ž1998. R471–R477. w13x S.F. Leibowitz, C. Sladek, L. Spencer, D. Tempel, Neuropeptide Y, epinephrine and norepinephrine in the paraventricular nucleus: stimulation of feeding and the release of corticosterone, vasopressin and glucose, Brain Res. Bull. 21 Ž1988. 905–912. w14x K.W. Locke, S.G. Holtzman, Behavioral effects of opioid peptides selective for mu or delta receptors: I. Morphine-like discriminative stimulus effects, J. Pharmacol. Exp. Ther. 238 Ž1986. 990–996. w15x A. Mansour, C.A. Fox, H. Akil, S.J. Watson, Opioid-receptor mRNA expression in the rat CNS: anatomical and functional implications, TINS 18 Ž1995. 22–29. w16x J.D. Pomonis, A.S. Levine, C.J. Billington, Interaction of the hypothalamic paraventricular nucleus and central nucleus of the amygdala in naloxone blockade of neuropeptide Y-induced feeding revealed by c-fos expression, J. Neurosci. 17 Ž1997. 5175–5182. w17x J. Roscrans, G. Krynock, A possible role of the PAG in the mediation of subjective effects of morphine, Pharmacologist 19 Ž1977. 171–176. w18x M. Shoaib, R. Spanagel, Mesolimbic sites mediate the discriminative stimulus effects of morphine, Eur. J. Pharmacol. 252 Ž1994. 69–75.