Proc. Natl Acad. Sci. USA Vol. 79, pp. 3360-3364, May 1982 Neurobiology

Regional cerebral glucose utilization during morphine withdrawal in the rat (cerebral metabolism//drug dependence) G. F. WOOTEN, P. DISTEFANO, AND R. C. COLLINS Departments of Neurology and Pharmacology, Division of Clinical Neuropharmacology, Washington University School of Medicine, St. Louis, Missouri 63110 Communicated by Oliver H. Lowry, February 26, 1982 ABSTRACT Regional cerebral glucose utilization was studied precipitated morphine withdrawal in the rat. A preliminary re- by 2-deoxy['4C]glucose autoradiography in morphine-dependent port of this work has appeared as an abstract (17). rats and during naloxone-induced morphine withdrawal. In mor- phine-dependent rats, glucose utilization was increased compared MATERIALS AND METHODS with naive controls uniformly (23-54%) in hippocampus, dentate gyrus, and and reduced in frontal cortex, , an- Preparation of Animals. Male Sprague-Dawley rats weigh- terior ventral , and medial habenular nucleus. On pre- ing 275-325 g were used. On experimental day 1, a single pellet cipitation ofmorphine withdrawal by subcutaneous administration containing 75 mg of morphine as free base was implanted sub- of naloxone at 0.5 mg/kg to morphine-dependent rats, glucose cutaneously under light ether anesthesia. On day 4, two pellets, utilization was increased in the central nucleus ofamygdala (51%), each containing 75 mg of morphine as free base, were im- lateral mammillary nucleus (40%), lateral habenular nucleus planted. On day 7, after being deprived of food for 12 hr, the (39%), medial mammillary nucleus (35%), and medial septal nu- rats were lightly anesthetized with 2% halothane, the pellets cleus (35%) (all, P < 0.01). Significant increases also occurred in were removed, and a polyethylene cannula was implanted into several other limbic structures including interpeduncular nucleus, the right externaljugular vein for subsequent injections. At least anterior medial and ventral thalamic nuclei, and lateral septal 2 hr of recovery time was allowed before injection of naloxone nucleus. Knowledge of the functional cerebral anatomy of the at 0.5 mg/kg subcutaneously to induce morphine withdrawal. morphine-withdrawal syndrome should facilitate studies directed Controls consisted of naive rats (n = 5), rats made dependent toward understanding the molecular mechanisms of opiate on morphine but injected on day 7 with normal saline rather withdrawal. than naloxone (morphine dependent; n = 4), and naive rats re- ceiving naloxone at 0.5 mg/kg subcutaneously (n = 3). Five rats Abrupt termination of chronic morphine treatment or admin- were studied during naloxone-induced morphine withdrawal. istration of opiate antagonists to morphine-dependent animals 2-Deoxy['4C]glucose Autoradiography. 2-Deoxy['4C]glucose results in a complex withdrawal syndrome composed of both autoradiography was carried out according to the method of stereotyped behavioral and autonomic features (1, 2). The be- Sokoloff et aL (8). Ten minutes after subcutaneous injection of havioral concomitants of morphine withdrawal in rodents in- naloxone at 0.5 mg/kg or normal saline, 2 deoxy['4C]glucose clude wet dog shakes, jumping or escape attempts, teeth chat- (25 ,uCi per rat; 1 Ci = 3.7 X 10"' becquerels) was injected into tering, and abnormal posturing; autonomic features include the awake freely moving rats via the venous cannula. Forty-five diarrhea, weight loss, hypothermia, seminal emissions, and minutes later, the animals were killed with intravenous sodium ptosis. Although much information has accumulated in recent pentobarbital and perfused with 0.2 M sodium cacodylate (pH years regarding the identity and distribution of endogenous 7.3) and then with 3.3% paraformaldehyde in the same buffer. opiates (3, 4) and their receptor sites (5, 6) in the central nervous The brain was removed, rapidly frozen in liquid Freon, and system, there is little knowledge of those areas of the central mounted on brass chucks. Duplicate 20-lim sections were taken nervous system critical for expression of the morphine-with- every 200 Aum, mounted on cover glasses, and exposed to Kodak drawal syndrome (7). SB-5 x-ray film for exactly 5 days. Selected sections were then The 2-deoxyglucose autoradiographic method of Sokoloff et taken for thionin staining and densitometry was carried out on aL (8) has been ofgreat utility in characterizing functional cere- the autoradiographs. OD results are expressed as OD ofa given bral anatomy. The physiological activity of neurons requires region/OD of the corpus callosum ratios. Mean blood glucose energy derived from the metabolism of glucose. By mapping levels (mg/ml; ± SD) during naloxone-precipitated morphine regional cerebral glucose utilization, inferences can be made withdrawal were 1.87 ± 0.57; levels in naive rats treated acutely about regional physiological neuronal activity. The method has with naloxone were 1.15 ± 0.23 and those in morphine-depen- been used to study a variety ofexperimental paradigms includ- dent rats treated acutely with normal saline were 1.66 ± 0.10. ing the routes ofcentral processing ofperipheral sensory stimuli Therefore, changes in endogenous blood glucose levels were (9, 10), the spread offocal seizure activity (11, 12), central sites not likely to independently alter the autoradiographic images ofdrug action (13, 14), and the consequences ofrestricted brain in rats during naloxone-precipitated morphine withdrawal. lesions (15, 16). In an attempt to characterize the functional Student's t test was used to determine the significance ofthe cerebral anatomy of morphine withdrawal, we have used the differences in glucose utilization between various treatment 2-deoxyglucose autoradiographic method to describe relative groups (18). regional cerebral rates of glucose utilization during naloxone- RESULTS The publication costs ofthis article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertise- Morphine-dependent rats that received naloxone at 0.5 mg/ ment" in accordance with 18 U. S. C. §1734 solely to indicate this fact. kg subcutaneously on day 7 rapidly developed a morphine-with- 3360 Downloaded by guest on September 24, 2021 Neurobiology: Wooten et al. Proc. Natl. Acad. Sci. USA 79 (1982) 3361 drawal syndrome consisting of tachypnea, increased motor ac- had no overt behavioral changes. tivity, wet dog shakes, and diarrhea. The animals vocalized Regional brain glucose utilization in naive controls, in mor- when handled but did not exhibit the "jumping" and "flying" phine-dependent rats, and in rats during naloxone-induced phenomena, behaviors that are thought to be environmentally morphine withdrawal is summarized in Table 1. Increased glu- determined. In contrast, morphine-dependent rats that re- cose utilization relative to naive controls in morphine-depen- ceived normal saline subcutaneously on day 7 remained quiet dent rats not given naloxone was found in the medial, lateral, and alert, exhibiting no apparent abnormal behavior. Likewise, and magnocellular preoptic areas, bed nucleus of the stria ter- naive rats treated with naloxone at 0.5 mg/kg subcutaneously minalis, parasubiculum, subiculum, medial and lateral entorhi-

Table 1. 2-Deoxy[14C]glucose incorporation into brain regions in morphine-dependent and -withdrawing rats Morphine dependent Morphine withdrawal Brain region Naive control (n = 5) (n = 4) (n = 5) Cortex Frontal motor 3.41 ± 0.19 2.80 ± 0.15 3.00 ± 0.49 Medial frontal 3.72 ± 0.22 3.30 ± 0.16 3.21 ± 0.44 Cingulate gyrus 3.46 ± 0.31 3.31 ± 0.33 3.33 ± 0.13 Pyriform 2.30 ± 0.13 2.39 ± 0.16 2.58 ± 0.20 Orbital frontal 4.28 ± 0.22 3.74 ± 0.62 4.12 ± 0.39 Basal ganglia Striatum 3.56 ± 0.36 2.94 ± 0.26 3.09 ± 0.21 Globus pallidus 2.17 ± 0.13 1.96 ± 0.13 2.00 ± 0.17 Substantia nigra (P.C.) 2.49 ± 0.43 2.28 ± 0.16 2.61 ± 0.21 Substantia nigra (P.R.) 1.83 ± 0.36 1.80 ± 0.04 2.09 ± 0.34 Subthalamic nucleus 3.11 ± 0.13 2.99 ± 0.20 3.15 0.35 Thalamus Paratenial nucleus 2.77 ± 0.26 2.94 ± 0.17 3.67 ± 0.31 T T Anterior ventral nucleus 3.72 ± 0.23 3.13 0.25 4.08 ± 0.50 t Anterior medial nucleus 3.25 ± 0.28 2.87 ± 0.24 3.73 ± 0.23 T T Preoptic area Medial preoptic area 1.49 ± 0.25 2.11 ± 0.22 T T 1.98 ± 0.33 Lateral preoptic area 1.61 ± 0.12 2.24 ± 0.30 T T 2.02 ± 0.21 Bed nucleus 1.74 ± 0.21 2.28 0.24 t 2.70 ± 0.24 T Preoptic magnocellularis 1.99 ± 0.32 3.07 ± 0.44 T 3.06 ± 0.29 "Limbic system" -central nucleus 2.72 ± 0.25 2.55 ± 0.28 3.85 ± 0.34 T T T Amygdala-medial nucleus 1.76 ± 0.15 1.61 ± 0.07 2.18 ± 0.34 T T Amygdala-basolateral nucleus 2.09 ± 0.16 2.09 ± 0.15 2.28 ± 0.17 Medial septal nucleus 2.43 ± 0.21 2.30 ± 0.13 3.10 ± 0.26 1 1 Lateral septal nucleus 2.06 ± 0.23 2.02 ± 0.11 2.54 ± 0.24 T t Medial habenular nucleus 3.27 ± 0.47 2.44 ± 0.30 2.81 ± 0.21 Lateral habenular nucleus 4.08 ± 0.26 3.64 ± 0.53 5.25 ± 0.78 T Medial mammillary nucleus 3.89 ± 0.44 3.41 ± 0.58 4.70 ± 0.43 T Lateral mammillary nucleus 3.89 ± 0.54 3.41 ± 0.60 4.77 ± 0.47 t Interpeduncular nucleus 3.80 ± 0.38 3.67 ± 0.34 4.69 ± 0.50 T 2.86 ± 0.27 2.51 ± 0.23 2.93 ± 0.18 T Dorsal presubiculum 2.92 ± 0.32 3.28 ± 0.10 3.28 ± 0.73 Parasubiculum 2.23 ± 0.28 3.02 ± 0.10 T T T 2.60 ± 0.55 Subiculum 2.21 ± 0.29 3.03 ± 0.04 T T 2.36 ± 0.40 Medial 1.82 ± 0.16 2.85 ± 0.25 T T T 2.00 ± 0.34 Lateral entorhinal cortex 1.85 ± 0.35 2.69 ± 0.13 TT 1.94 ± 0.36 Dorsal hippocampus CA-1 pyramidal layer 1.56 ± 0.12 1.96 ± 0.26 T 1.77 ± 0.18 CA-1 stratum radiatum 1.58 ± 0.14 2.18 ± 0.34 T T 1.71 ± 0.16 Perforant path 2.23 ± 0.25 2.98 ± 0.23 1T 2.60 ± 0.39 Dentate gyrus 1.43 ± 0.11 2.21 ± 0.20 T 1 T 1.97 ± 0.18 CA-3 pyramidal layer 1.80 ± 0.14 2.27 ± 0.23 T t 1.94 0.10 "CA-2" 2.08 ± 0.34 2.57 ± 0.26 T 2.28 ± 0.24 Ventral hippocampus CA-1 pyramidal layer 1.80 ± 0.31 2.46 ± 0.42 1 2.17 ± 0.22 CA-1 stratum radiatum 1.75 ± 0.34 2.37 ± 0.38 T 1.90 ± 0.21 Perforant path 2.56 ± 0.45 3.31 ± 0.24 1 T 2.76 ± 0.58 Dentate gyrus 1.78 ± 0.45 2.53 ± 0.34 T 2.22 ± 0.42 Ventral subiculum 1.79 ± 0.42 2.50 ± 0.56 2.34 ± 0.38 Results represent mean ± SD of the OD of the region in question/OD of the corpus callosum ratio. Arrows denote those regions in which a statistically significant change occurred and the direction of the change: 1 arrow, P < 0.05; 2 arrows, P < 0.01; 3 arrows, P < 0.001. The morphine-dependent group was compared with the naive control kroup and the morphine- withdrawal group was compared with the morphine-dependent group. Downloaded by guest on September 24, 2021 3362 Neurobiology: Wooten et al. Proc. Nad Acad. Sci. USA 79 (1982) nal cortex, CA-1 pyramidal layer and stratum radiatum, per- Autoradiographs depicting regional glucose utilization dur- forant path, and dentate gyrus. Glucose utilization was decreased ing morphine withdrawal are shown in Fig. 1. in frontal motor cortex, medial frontal cortex, striatum, medial habenular nucleus, and anterior ventral thalamic nucleus in morphine-dependent rats. DISCUSSION Naloxone-induced morphine withdrawal was associated with The principal finding ofthis study was that morphine addiction increased glucose utilization in anterior thalamic nuclei (para- and withdrawal were accompanied by discrete, selective, and tenial, anterior ventral, and anterior medial), bed nucleus of highly reproducible changes in regional cerebral glucose utili- stria terminalis, central nucleus ofamygdala, medial and lateral zation. Basically, two major regional patterns of cerebral me- septal nuclei, lateral habenular nucleus, medial and lateral tabolism appeared during morphine addiction and withdrawal. mammillary nuclei, interpeduncular nucleus, and nucleus ac- One general pattern was that ofa small increase in metabolism cumbens. Glucose utilization was decreased during morphine during addiction and a return toward control duringwithdrawal. withdrawal in the subiculum, medial and lateral entorhinal cor- This pattern occurred primarily in the hippocampal formation, tex, CA-1 stratum radiatum, and CA-3 pyramidal layer. subiculum, and entorhinal cortex. These structures are stimu- Injection ofnaloxone at 0.5 mg/kg subcutaneously produced lated or disinhibited by enkephalins and endorphin (19-22) and no significant changes in regional glucose utilization compared are metabolically activated when opiate-like compounds are with naive controls (data not shown). administered to cause seizures (23). Of particular interest are

A D

B E

rl.

I -t T

1. -, I 4

C F

FIG. 1. 2-Deoxy['4C]glucose autoradiographs of rat brain during morphine withdrawal. Glucose utilization is prominent in the medial and lat- eral septal nuclei (A), the anterior ventral and parataenial thalamic nuclei (B), thecentral nucleus ofthe amygdala (C), the lateral habenular nucleus (D), the mammillary nuclei (E), and the interpeduncular nucleus (F). Downloaded by guest on September 24, 2021 Neurobiology: Wooten et al. Proc. Natl. Acad. Sci. USA 79 (1982) 3363 the recent observations of relatively high concentrations of lin content ofany ofthe amygdaloid nuclei (30, 31), and (iii) local opiate "receptors" in the hippocampus (5, 6). injection of naloxone into the central nucleus of the amygdala The other general pattern of regional changes in brain me- in rats made dependent on morphine induced the "jump" sign tabolism could be characterized as a small reduction or no while injection ofnaloxone into nearby structures failed to pro- change during morphine addiction with a moderate to large in- duce evidence of withdrawal (7). Further, electrolytic lesions crease during withdrawal. This pattern was seen in several lim- of the central nucleus ofamygdala markedly attenuated the ap- bic structures including the central nucleus of the amygdala, pearance of the jump sign in response to systemic naloxone several midline anterior thalamic nuclei, the bed nucleus ofthe administration to dependent animals (7). stria terminalis, the medial and lateral septal and mammillary As far as we know, except for a 1976 abstract in which it was nuclei, the lateral habenular nucleus, and the interpeduncular reported that glucose utilization is particularly affected in the nucleus. Each ofthese structures bears some synaptic proximity habenula during morphine withdrawal (32), this is the first study to the other. For example, the central nucleus of the amygdala of regional cerebral glucose utilization during opiate addiction projects via the stria terminalis to both the bed nucleus of the and precipitated withdrawal. Our results serve to focus atten- stria terminalis and the lateral hypothalamic area (24-26). Some tion on the role of several lImbic structures, particularly the neurons ofthe lateral hypothalamic area project to the septum, central nucleus of the amygdala, in production ofthe behavioral which sends efferents into the stria medullaris joining projec- and autonomic properties of opiate withdrawal. With the avail- tions from the bed nucleus of the stria terminalis en route to ability of maps of the functional anatomy of morphine with- the habenula (27). The habenula, in turn, projects back to the drawal, biochemical and pharmacological studies directed to- central nucleus of the amygdala (28) and via the fasciculus ret- ward understanding the molecular mechanisms underlying roflexus to the interpeduncular nucleus (27). Another major opiate withdrawal should be facilitated. projection ofthe central nucleus ofthe amygdala is to the preop- tic (27, 28), which sends fibers via the median We thank Dr. Eugene Johnson for helpful discussion (supported by forebrain bundle to mammillary nuclei, which in turn project National Institutes of Health Grant HL 23293). This work was sup- principally to anterior thalamic nuclei via the mammillo-tha- ported by a grant from the Institute for Medical Education and Research lamic tract (27). Each of the structures that are metabolically of the City of St. Louis and National Institute of Neurological and Com- active during morphine withdrawal are closely connected syn- municative Disorders and Stroke Grant 1-PO1-14834-1. G.F.W. is a aptically. Therefore, one or a few of those structures may be George C. Cotzias Fellow of the American Parkinson Disease Asso- directly or primarily activated with secondary increases in me- ciation. tabolism in other structures occurring as a consequence of syn- aptic connections. 1. Wei, E., Loh, H. H. & Way, E. L. (1973) J. Phar-macol. Exp. An alternative hypothesis is that each ofthe structures show- Ther. 184, 398-403. ing prominent increases in metabolism during morphine with- 2. Blasig, J., Herz, A., Reinhold, K. & Zieglgansberger, S. (1973) drawal may be activated independently and directly by manip- Psychopharmacology 33, 19-38. 3. Simantov, R., Kuhar, M. J., Uhl, G. R. & Snyder, S. H. (1977) ulation of opiate agonist and antagonist drugs. Extensive Proc. Nati. Acad. Sci. USA 74, 2167-2171. mapping studies of both endogenous opiate distribution (3, 4) 4. Sar, M., Stumpf, W. E., Miller, R. J., Chang, K-J. & Cuatraca- and opiate receptor binding sites (5, 6) are now available. Im- sas, P. (1978)J. Comp. Neurol. 182, 17-38. munohistochemical maps ofenkephalin distribution reveal con- 5. Atweh, S. F. & Kuhar, M. J. (1977) Brain Res. 134, 393-405. centrations in midline thalamic nuclei, the lateral edge of the 6. Herkenham, M. & Pert, C. B. (1980) Proc. Natl. Acad. Sci. USA medial habenula, the central nucleus of the amygdala, and the 77, 5532-5536. 7. Calvino, B., Lagowska, J. & Ben-Ari, Y. (1979) Brain Res. 177, lateral septum; each of these structures are metabolically active 19-34. during morphine withdrawal. Enkephalin immunoreactivity is 8. Sokoloff, L., Reivich, M., Kennedy, C., DesRosiers, M. H., Pat- quite low, however, in the mammillary nuclei (which showed lak, C. S., Pettigrew, K. D., Sahurada, 0. & Shenohara, M. increased glucose utilization during withdrawal) and quite high (1977) J. Neurochem. 28, 879-916. in the globus pallidus (a structure that showed little change in 9. Kennedy, C., DesRosiers, M. H., Sakurada, O., Shinohara, M., glucose utilization). Thus, there is no clear correlation between Reivich, M., Jehle, J. W. & Sokoloff, L. (1976) Proc. Natl. Acad. Sci. USA 73, 4230-4234. those structures that are particularly active during morphine 10. Sharp, F. R., Kauer, J. S. & Shepherd, G. M. (1975) Brain Res. withdrawal and the enkephalin content of the structures. Fur- 98, 596-600. thermore, autoradiographic studies of opiate receptor distri- 11. Collins, R. C., Kennedy, C., Sokoloff, L. & Plum, F. (1976) bution revealed a relatively high density in the interpeduncular Arch. Neurol. 33, 536-542. nucleus (an active structure during withdrawal) as well as the 12. Collins, R. C. (1978) Brain Res. 150, 487-501. hippocampus (a structure not activated during withdrawal). 13. Weinberger, J., Greenberg, J. H., Waldman, M. T. G., Sylves- tro, A. & Reivich, M. (1979) Brain Res. 177, 337-345. Thus, neither endogenous enkephalin content nor opiate re- 14. Wooten, G. F. & Collins, R. C. (1980) Brain Res. 201, 173-184. ceptor density correlate consistently with those regions of brain 15. Steward, 0. & Smith, L. K. (1980) Exp. Neurol. 69, 513-527. that are most metabolically active during morphine withdrawal. 16. Wooten, G. F. & Collins, R. C.'(1981)J. Neurosci. 1, 285-291. Whether anatomical relationships or individual biochemical 17. DiStefano, P., Wooten, G. F., Collins, R. C. & Johnson, E. M. properties are of primary importance in generating the func- (1981) Soc. Neurosci. Abstr. 7, 52 (abstr.). tional anatomical map of morphine withdrawal, the central nu- 18. Snedecor, G. W. & Cochran, W. G. (1978) Statistical Methods (Iowa State Univ. Press, Ames, IA). cleus of the amygdala would appear to be a particularly impor- 19. Nicoll, R. A., Siggins, G. R., Ling, N., Bloom, F. E. & Guiller- tant site. That the amygdala, particularly the central nucleus, min, R. (1977) Proc. Natl. Acad. Sci. USA 74, 2584-2588. is important in expression of morphine withdrawal has been 20. Zieglgansberger, W., French, E. D., Siggins, G. R. & Bloom, F. suggested by several lines of evidence: (i) the central nucleus E. (1979) Science 205, 415-417. of the amygdala projects widely to the hypothalamus and other 21. Lee, H. K., Dunwiddie, T. & Hoffer, B. (1980) Brain Res. 84, limbic structures and its stimulation has been found to elicit 331-342. 22. Dunwiddie, T., Mueller, A., Palmer, M., Stewart, J. & Hoffer, behavioral and autonomic manifestations similar to some as- B. (1980) Brain Res. 184, 311-330. pects of the morphine-withdrawal syndrome (26, 29), (ii) the 23. Frenk, H., Urea, G. & Liebeskind, J. C. (1978) Brain Res. 147, central nucleus of the amygdala possesses the highest enkepha- 327-337. Downloaded by guest on September 24, 2021 3364 Neurobiology: Wooten et aL Proc. Natd Acad. Sci. USA 79 (1982)

24. Lammers, H. J. (1972) in The Neurobiology oftheAmygdala, ed. 29. Kaada, B. (1972) in The Neurobiology of the Amygdala, ed. Eleftheriou, B. E. (Plenum, New York), pp. 123-144. Eleftheriou, B. E. (Plenum, New York), pp. 205-281. 25. Cowan, W. M., Raisman, G. & Powell, T. P. S. (1965)J. Neurod 30. Gros, C., Pradelles, P., Humbert, J., Dray, F., LaSalle, G. & Neurosurg. Psychiatr. 28, 137-151. BenAri, Y. (1978) Neurosci. Lett. 10, 193-196. 26. Krettek, J. E. & Price, J. L. (1978) J. Comp. NeuroL 178, 31. Uhl, G. R., Kuhar, M. J. & Snyder, S. H. (1978) Brain Res. 149, 225-253. 223-228. 27. Isaacson, R. L. (1974) The Limbic System (Plenum, New York). 32. Sakurada, O., Shinohara, M., Klee, W. A., Kennedy, C. & So- 28. Kretlek, J. E. & Price, J. L. (1978) J. Comp. NeuroL 178, koloff, L. (1976) Soc. Neurosci. Abstr. 2, 613 (abstr.). 255-279. Downloaded by guest on September 24, 2021