Brain Research, 595 (1992) 32-38 © 1992 Elsevier Science Publishers B.V. All rights reserved 0006-8993/92/$05.00

32

BRES 18211

Antinociception produced by receptor selective opioids: modulation of spinal antinociceptive effects by supraspinal opioids C h r i s t i n e Miaskowski

a,

J o n D. Levine b,c

Schools of a Nursing, b Medicine and " Dentistry c Unit,ersity of California at San Francisco, San Francisco, CA 94143 (USA)

(Accepted 2 June 1992)

Key words: Intracerebrovenlricular; Intrathecal, ~ Opioid; 80pioid; x Opioid; DAMGO, DPDPE; US0,488H; Morphiceptin; Motor coordination; Synergy; Antagonism

The effect of intracerebroventricular administration of Iow-antinociceptive doses of selective p.- (DAMGO) or 6- (DPDPE) opioid agonists on the dose-dependent antinociceptive effects produced by intrathecal administration of sequentially increasing doses ot selective /~-, 8-, or K(U50,488H) opioid agonists was evaluated, in the rat, using the RandalI-Selitto paw-withdrawal test. When DPDPE or US0,488H was administered intrathecally, the low doses of both intracerebroventricular DAMGO and intracerebroventricular DPDPE markedly enhanced the antinociceptive effects of both intrathecal opioids, in contrast, when DAMGO was administered intrathecally, both intracerebroventricular DAMGO and intracerebroventricular DPDPE, administered in low doses, markedly antagonized the antinociceptive effects of the intrathecal opioid. In addition, the intracerebroventricular administration of a low-antinociceptive dose of a second /~-opioid agonist, morphiceptin, antagonized the antinociceptive effects of intrathecal morphiceptin. The antagonism of the antinociceptive effects observed with spinal administration of DAMGO is dose-dependent, with the effect observed only at low doses. Furthermore, the antagonism cannot be explained by a reduction in motor deficits produced hy intrathecal administration of DAMGO, because there were no differences in motor deficits, measured with an accelerating Rotarod treadmill, between intrathecal DAMGO administered as a single agent or as part of a combination regimen. The differences in antinociceptive effects obtained with the various supraspinal and spinal combinations are discussed in terms of the interactions that may occur between brainstem and spinal opioid receptor sites.

INTRODUCTION Extensive evidence exists for endogenous antinociceptive control mechanisms. These antinociceptive controls contain opioidergic components with rostralcaudal projections from supraspinal sites to the spinal cord Lz. For example, stimulation of the midbrain periaqueductal gray or hindbrain nuclei in the region of the raphe produces antinociception in animals 13,25. Evidence supporting a spinal opioidergic contribution to descending antinociceptive controls comes from studies demonstrating that the antinociceptive effects produced by electrical stimulation of the ventromediai medulla 42 or intracerebroventricular (i.c.v.) administration of opiates 16 are reversed by intrathecal (i.t.) administration of naloxone. In fact, recent studies suggest

that there are multiple descending antinociceptive controis ~. To further evaluate the suggestion of multiple descending controls, we studied the interactions between i.c.v, and i.t. opioid antinociceptive effects, by administering combinations of /~-, 8-, and x-opioids at supraspinal and spinal sites 2°. In these experiments, we demonstrated that the antinociceptive effects produced by i.c.v, administration of selective /z- and 8-opioid agonists are modulated differentially by low-antinociceptive doses of selective/z-, 8-, or K-opioid agonists administered i.t. 2°. To further evaluate the interactions between the supraspinal and spinal opioidergic components of the endogenous antinociceptive circuits, we examined in this study the reciprocal relationship; that is, the modulation of the antinociceptive effects pro-

Correspondence: J.D. Levine, Division of Rheumatology, S1334/Box 0452A, University of California, San Francisco, California 94143-0452A, USA. Fax: (1) (415) 476 4845.

33 duced by i.t. administration of DAMGO, DPDPE, or U50,488H with a low-antinociceptive i.c.v, dose of D A M G O or DPDPE. MATERIALS AND METHODS The experiments were performed using 240-260 g, male Sprague-Dawley rats (Bantin and Kingman, Fremont, CA). Animals were housed individually in a temperature controlled room and fed standard rat chow with both food and water available ad libitum. A 12-h light-dark cycle (lights on at 7.00 h) was used in the animal care facility. One week prior to the experiments, under pentobarbital anesthesia (65 mg/kg), i..~. catheters were inserted through the atlanto-occipital membrane and passed caudally 8.5 cm to a site just rostrai to the lumbosacral enlargement 4° and 22-gauge stainless-steel cannulae were stereotactically implanted into the third ventricle. Both devices were anchored to the skull with bone screws and acrylic dental cement. Stainless steel stylettes (30-gauge) were placed in the i.c.v, cannulae to keep them patent. Rats that exhibited neurological deficits, at any time, following the surgical procedure were eliminated from the study.

Antinociceptit,e testing The RandalI-Selitto paw-withdrawal test 3° was used to measure mechanical nociceptive threshold. The mechanical stimulus was applied with a Basile analgesymeter (Stoelting Co., Chicago, IL) which generates a linearly increaslng mechanical force, applied by a domeshaped plastic tip to the corsal surface of the rat's hindpaw. The nociceptive threshold was defined as the force in grams at which the animal withdraws its paw. The rats were trained in the test procedure, for three hours daily, for five days prior to data collection3f'. On the day of the experiments, rats were first tested, at 5 min intervals, for 2 h. Baseline thresholds were defined as the mean of the last six measurements prior to drug administration. The average baseline nociceptive threshold, for all the rats used in these experiments was 5.17+0.49 g (mean +S.E.M., n = 198). The effect of each of the opioid agonists on nociceptive threshold was tested 15, 20, and 25 rain after drug administration, at the time of peak effect of these

agents12,'4,29. Drug administration protocol Dose-response curves for t.t. administration of the selective /L-opioid agonists, [D-Ala2,N-Me-Phe4,GlyS-ol]-enkephalin (DAMGO) m and morphiceptinS; the selective ~-opioid agonist, [DPenZS].enkephalin (DPDPE) 2~' (all from Peninsula Laboratories; Belmont, CA); and the selective K-opioid agonist, US0,488H, 3s (a generous gift from Dr. Robert Lahti, Upjohn, Kalamazoo, MI), as single agents, were determined by administering sequentially increasing doses of each drug, in a volume of 10 ~1, at 30 rain intervals. The effect of a Iow-antinociceptive i.c.v, dose of an opioid agonist on the i.t. dose-dependence relationship, was tested using the following combinations, in different groups of rats: (i) a low-antinociceptive dose (i.e., a dose which produces approximately a 20% increase in nociceptive threshold) of DAMGO (5 fg, 50 fg, and 0.5 pg) or DPDPE (50 fg) administered i.c.v, with each sequentially increasing dose of i.t. DPDPE; (ii) a low-antinociceptive dose of morphiceptin (5 fg) administered i.c.v, wit~a each sequentially increasing dose of i.t. morphiceptin; (iii) a Iow-antinociceptive dose of DAMGO (50 fg) or DPDPE (50 fg) administered i.c.v, with each sequentially increasing dose of i.t. U50,488H; (iv) a low-antinociceptive dose of DAMGO (50 fg) or DPDPE (50 fg) administered i.c.v, with each sequentially increasing dose of i.t. DAMGO. Each i.c.v, injection was administered first, in a volume of 1 /~l, followed immediately by the i.t. injection, in a volume of 10/~!. All drugs were dissolved in saline. The number of rats, in each group, is indicated in the figure legends. The i.c.v, dose used was chosen to produce a small (20%) increase in nociceptive thresholds. We believe that the increases in nociceptive thresholds following i.c.v, administration of Iow-antinociceptive doses of selective opioid agonists results from activation

of supraspinal, rather than, spinal opioid receptors, because doses four to five orders of magnitude higher were required to produce approximately a 20% increase in nociceptive thresholds when administered at the spinal site. We further believe that the increases in nociceptive thresholds following i.t. administration of the selective opioid agonists resulted from activation of spinal rather than supraspinal opioid receptors because, as previously reported a3, the administration of these opioid agonists at the highest doses used in this study, to rats with i.t. catheters placed at the C 5-C 6 level of the spinal cord, demonstrated significant attenuation in antinociceptive effects compared of these opioids compared with the effects of spinal administration. Since the cumulative dosing regimen might have an effect on nociceptive thresholds, we determined whether i.t. administration of a single dose (i.e., 5/~g or 50/~g) of DAMGO, DPDPE, or US0,488H produced an increase in nociceptive threshold of similar magnitude as when the same dose was administered as the highest dose in an ascending series, as performed in these experiments. We found no differences in the antinociceptive effects produced by administration of the highest dose of opioid agonist compared to the same dose administered as the last dose in an ascending series (data not shown). Therefore, neither the lower doses, preceding the highest dose; nor the time in the apparatus; nor the time of exposure to elevated paw pressures appear to contribute to the antinociceptive effects produced, in these experiments, by the highest dose of opioid agonist.

Motor coordination testing in the group of rats that received 5 fg of i.c.v. DAMGO with sequentially increasing doses of i.t. DAMGO, motor coordination was tested using an accelerating Rota-Rod Treadmill (Ugo Basile, Stoelting Co., Chicago, IL)7'2L39. The rotarod was set in motion at a constant speed and the rats were placed into individual sections of the apparatus. Once all the rats were in position, the timers were set to zero and the rotarod was switched to the accelerating mode. The rotarod accelerated from a rate of 3.7 to 37.5 rpm in a period of 5 rain. The animals performance time in seconds was recorded when the rat, unable to stay on the rotarod, tripped a plate and stopped the timer. A minimum of two training sessions, of three hours duration, were performed to condition the animals to the treadmill and establish consistent baseline performance scores. On the day of the experiment, the rats were tested, on the rotarod, at 5 rain intervals for 2 h. Baseline rotarod performance score was defined as the mean of the last six measurements prior to drug administration. The average rotarod performance score for these rats was 103.83 :t: 13.35 s (mean + S.E.M., n = 17). The effect of the opioid agonists alone and in combination on rotarod performance scores (i.e., length of time on the rotarod) was tested using the same drug administration paradigm used in the antinociceptive testing. The effect of the drugs on rotarod performance score was also calculated as a percentage change from baseline performance score.

Statistical analyses Analysis for antinociceptit'e synergy. Statistical analysis of the dosedependent effects of i.t. administration of DPDPE, U50,488H, DAMGO, and morphiceptin as single agents, on nociceptive thresholds was performed using a one-way analysis of variance (ANOVA). A two-factor, repeated measures ANOVA was used to analyze for interactions between the antinociceptive effects of the receptor selective opioid agonists6'1s:9'35. ANOVA, which has been used to identify interactions between biologically active agents in cardiovascular physiology4'14'1s'37, immunologyeT'3a, oncology34, and toxicology TM studies, was chosen because this method of analysis allows one to evaluate for a statistically significant interaction between two groups (i.e., the effects of a single drug compared to the effects of a combination regimen) across a series of measures (i.e., cumulative dosing across several orders of magnitude). The dose of opioid agonist in the cumulative dose response curve that produced approximately a 20% increase in nociceptive threshold was chosen as the Iow-antinociceptive dose in the combination

34 regimen. A iow-antinociceptive dose was chosen, in this experimental design, in order to enhance the ability to detect a synergistic interaction between combinations of receptor selective opioids across the full range of doses tested. However, a statistical analysis for interactions between the dosedependent antinociceptive effects of a single agent and a combination regimen is not possible over the range of doses where the dose-dependent effects for either the single agent or the combination regimen have reached a plateau. Therefore, we determined whether the opioid-induced increases in nociceptive thresholds for either the single agents or the combination regimens reached a plateau at the highest doses tested, using the Scheffe post-hoc test 6'1s'19'35. The results of these post-hoe contrasts are presented, in the results section, using successively larger integers to represent each successively higher dose in the i.t. cumulative dose-response curve (from 0 = either the vehicle or Iow-antinociceptive dose of the i.c.v, drug up to 5 -- the highest dose of the opioid agonist administered at the i.t. site). Following the statistical procedure for determining the presence of a plateau for either the single agent or the combination regimen, a two-factor, repeated measures ANOVA was done using the portion of the dose-response curves below the plateau effect. If the dose-response curves are divergent, then a statistically significant interaction term, in a two-factor, repeated measures ANOVA, demonstrates that the differences between the two groups, over the repeated measure (i.e., increasing doses of opioid agonist), deviate from parallelism ;rod therefore are more than additive, that is synergistic 6'1s'i9'35 (tbr the mathematical derivation of the interaction term in the ANOVA see references 8, 17, and 33). Analysis of motor coordination data. Statistical analysis of the doseresponse effects of i.t. administration of DAMGO compared to the dose-response effects of 5 fg of i.c.v. DAMGO co-administered with sequentially increasing i.t. doses of DAMGO on rotarod performance was done using a two-factor, repeated measures ANOVA. For all the statistical analyses, differences were considered statistically significant at the P < 0.05 level.

RESULTS

l,t. &agonist (DPDPE) l.t. administration of sequentially increasing doses of DPDPE produced statistically si~,if;:-ant, dose-de. pendent increases in mechanical nociceptive threshold (Fs,t15-92.59, P 0.05) or group × dose interaction (F5.75 = 2.14, p > 0.05). DISCUSSION In this study, we found that co-administration of a low-antinociceptive i.c.v, dose of either a selective #or 8-opioid agonist with sequentially increasing i.t. doses of selective 8- or K-opioid agonists produced synergistic (i.e., more than additive) antinociceptive effects while the i.c.v, administration of a low-antinociceptive dose of the same/~- and 6-opioids with i.t. administration of #-opioids produced markedly antagonistic effects. The findings of antinociceptive synergy

produced with combinations of i.c.v, and i.t. opioid agonists is compatible with the suggestion from previous studies 2°m'4~ that supraspinal opioid receptor circuitry can interact with spinal opioidergic mechanisms. The antagonism of the antinociceptive effects of i.t. D A M G O by both i.c.v. DAMGO and i.c.v. DPDPE was unexpected, since previous work demonstrated that each agonist administered alone produced dose-dependent increases in nociceptive thresholds 2~ and that co-administration of a low-dose of i.t. D A M G O with sequentially increasing doses of i.c.v. DAMGO, enhanced the dose response effects of i.c.v. DAMGO 22. The antagonism does not appear to be a non-specific effect of DAMGO, since antagonism was also observed with a second/~-opioid agonist, morphiceptin and since the antagonistic effects were observed only with the lower doses of D A M G O (Fig. 4). Furthermore, the antagonism cannot be explained by a reduction in motor side-effects produced by i.t. administration of D A M G O because there were no differences in motor deficits between i.t. D A M G O administered as a single agent or as part of the combination regimen (Fig. 6). The basis tbr both the synergy and the antagonism of the antinociceptive effects produced by combinations of supraspinal and spinal opioids is presently unknown. The asymmetrical opioidergic interactions (i.e., synergy versus antagonism based on where the low antinociceptive dose and varying dose of opioid agonist were administered), might be explained, at least in part, by differential activation of sub-types of /z-opioid receptors 2s such that low i.c.v, doses of /~opioids, when given in combination with spinal /~opioids elicit the observed antagonism, while higher doses acting at a second /~-receptor subtype are responsible for the supraspinal antinociceptive effects. This hypothesis could be tested using selective agonists and antagonists for the/~-1 and/~-2 opioid receptors. In summary, in this study of the modulation of spinal antinociceptive effects by supraspinal opioids, we observed that the antinociceptive effects produced by spinal administration of 8- and K-opioids was markedly enhanced by supraspinal administration of /~- and ~5-opioids. In contrast, the i.t. administration /~-opioids were antagonized by i.c.v, administration of /.~- and 6-opioids. When compared to findings from a previous study, 2° which reversed the site at which the low-antinociceptive and varying doses of receptor selective opioids were administered, marked asymmetries in antinociceptive effects were observed. Further studies to elucidate the mechanisms underlying the asymmetries in the interactions between supraspinal and spinal antinociceptive effects of receptor selective opioids are currently being performed.

38 Acknowledgements. We would like to thank Dr. Steven Paul for statistical consultation and Dr. Philip Heller for his thoughtful review of the manuscript. This work was supported by NIH Grant DE08973 and a Biomedical Research Support Grant RR05604 REFERENCES 1 Basbaum, A.I. and Fields, H.L., Endogenous pain control systems: Brainstem spinal pathways and endorphin circuitry, Annu. Rev. Neurosci., 7 (1984) 309-338. 2 Basbaum, A.I. and Fields, H.L., Endogenous pain control mechanisms: Review and hypothesis, Ann. Neurol., 4 (1978) 451-462, 1978. 3 Brennan, J.F. and Jastreboff, P.J., Interaction of salicylate and noise results in mortality of rais, E.tperientia, 45 (1989) 731-734. 4 Caplan, R.A. and Su, J.Y., Interaction of halothane and verapamil in isolated papillary muscle, Anesth. Analg., 65 (1986) 463-468. 5 Chang, K.J., Wei, E.T., Killian, A. and Chang, J.K., Potent morphiceptin analogs: Structure activity relationship and morphine-like activities, J. PharmacoL Exp. Ther., 227 (1983) 403-408. 6 Cohen, J. and Cohen, P., Applied Multiple Regression/ Correlational Analysis for the Behavioral Sciences, Lawrence Erlbaum Associates, Hillsdale, NJ, 1983, pp. 301-350.. 7 Dunham, N.W. and Miya, T.S., A note on a simple apparatus for detecting neurological deficits in rats and mice, J. Am. Pharma. ceut. Assoc., 46 (1957) 208-209. 8 Dunn, O.J. and Clark, V.A., Applied Statistics: Analysis of Variance and Regression, Wiley, NY, 1987, pp. 123-151. 9 Fields, H.L., Pain, McGraw-Hill. NY, 1987, pp. 99-131. 10 Handa, B.K., Lane, A.C., Lord, J.A.H., Morgan, B.A., Rance, M.J. and Smith, C.F.C., Analogues of beta-LPH61-64 possessing selective agonist activity at mu-opiate receptors, Fur. J. Pharmacol., 70 (1981) 531-540. 11 Harvey, M.J. and Klaassen, C.D., Interaction of metals and carbon tetrachloride on lipid peroxidation and hepatotoxicity, Toxicol. Appl. Pharmacol., 71 (1983) 316-322. 12 Heyman, J.S., Koslo, R.J., Mosberg, H.I., Tallarida, R.J. and Porreca, F., Estimation of the affinity of naloxone at supraspinal and spinal opioid receptors in vivo: studies with receptor selective agonists, Life Sci., 39 (1986) 1759-1803. 13 Hodge, C.J., Jr., Apkarian, A.V., Stevens, R.T., Vogelsang, G.D., Brown, O. and Frank J.I., Dorsolateral pontine inhibition of dorsal horn cell responses to cutaneous stimulation: Lack of dependence on catecholaminergic system in the cat, J. Neurophys. iol., 50 (1983) 1220-1235. 14 Katahira, K., Mikami, H., Ogihara, T., Kohara, K., Otsuka, A., Kumahara, Y., and Khosla, M.C., Synergism of intraventricular NaCi infusion and subpressor angiotensin in rats, Am. J. PhysioL, 256 (1989) HI-H8. 15 Kopia, G.A., Kopaciewicz, L.J., Ohlstein, E.H., Horohonich, S., Storer, B.L., and Shebuski, R.J., Combination of the thromboxane rect.'ptor antagonist, sulotroban (BM 13.177; SK&F 95587) with streptokinase: demonstration of thrombolytic synergy, J. Pharmetcol. Exp. Ther., 250 (1989)887-895. 16 Levine, J.D., Lane, S.R., Gordon, N.C. and Fields, H.L., A spinal opioid synapse mediates the interaction of spinal and brainstem sites in morphine analgesia, Brain Res., 236 (1982) 85-91. 17 Lindman, H.R., AnalysL, of Variance in Complex Experimental Designs, Freeman, San Francisco, CA, 1974, pp. 92-97. 18 Marascuilo, L.A. and Serlin, R.C., Statistical Methods for the Social and Behavioral Sciences, Freeman, NY, 1988, pp. 448-454. 19 Melton, J.S. and Tsokos, J.O., Statistical Methods in Biological and Health Sciences, McGraw-Hill, NY, 1983, pp. 271-331. 20 Miaskowski, C., Taiwo, Y.O. and Levine, J.D., Antinociception produced by receptor selective opioids: Modulation of supraspinal antinociceptive effects by spinal opioids, submitted. 21 Miaskowski, C., Sutters, K.A., Taiwo, Y.O. and Levine, J.D., Comparison of the antinociceptive and motor effects of intrathecal opioid agonists in the rat, Brain Res., 553 (199i) 105-109. 22 Miaskowski, C., Taiwo, Y.O. and Levine, J.D., Contribution of

supraspinal mu-and delta-opioid receptors to antinociception in the rat, Fur. I. Pharmacol., 205 (1991) 247-252. 23 Miaskowski, C., Taiwo, T.O. and Levine, J.D., Kappa- and deltaopioid agonists synergize to produce potent analgesia, Brain Res., 509 (1990) 165-168. 24 Millan, M.J., Kappa-opioid receptor-mediated antinociception in the rat. I. Comparative actions of mu- and kappa-opioids against noxious thermal, pressure, and electrical stimuli, I. Pharmacol. Exp. Ther., 251 (1989)334-341. 25 Mokha, S.S., McMillan, J.A. and Iggo, A., Descending control of spinal nociceptive transmission. Actions produced on spinal multireceptive neurons from the nuclei locus coeruleus (LC) and raphe magnus (NRM), Exp. Brain Res., 58 (1985) 213-226. 26 Mosberg, H.I., Hurst, R., Hruby, V.J., Gee, K., Yamamura, H.I., Ga!!jgan, JJ. and Burks, T.F., Bis-penicillamine enkephalins possess highly improved specificity toward delta opioid receptors, Proc. Natl. Acad. Sci. USA, 80 (1983) 5871-5874. 27 Oksenberg, D., Dieckmann, B.S., and Greenberg, P.L., Functional interactions between colony-stimulating factors and the insulin family hormones for myeloid leukemic cells, Cancer Res., 50 (1990) 6471-6477. 28 Pasternak, G.W. and Wood, PJ., Minireview: Multiple mu opiate receptors, Life Sci., 38 (1986) 1889-1898. 29 Porreca, F., Heyman, J.S., Mosberg, H.I., Omnaas, J.R. and Vaught, J.L., Role of mu and delta receptors in the supraspinal and spinal analgesic effects of [D-Pen2,D-Pen5] enkephalin in the mouse, J. Pharmacol. Exp. Ther., 241 (1987)393-400. 30 Randall, L.O. and Selitto, J.J., A method for measurement of analgesic activity on inflamed tissue, Arch. Int. Pharmacodyn., 111 (1957) 409-419. 31 Roerig, S.C. a~d Fujimoto, J.M., Muitiplicative interaction between intrathecally and intracerebroventricularly administered mu opioid agonists but limited interactions with delta and kappa agonists for antinociception in mice, J. Pharmacol. Exp. Ther., 249 (1989) 762-768. 32 Sato, K., Fujii, Y., Kasono, K., Ozawa, M., Imamura, H., Kanaji, Y., Kurosawa, H., Tsushima, T., and Shizume, K., Parathyroid hormone-related protein and interleukin-la synergistically stimulate bone resorption in vitro and increase the serum calcium concentration in mice in vivo, Endocrinology, 124 (1989) 21722178. 33 Searle S.R., Linear Models For Unbalanced Data, Wiley, NY, 1987, pp. 101-125. 34 Shaikh, N,A., Owen, A.M., Ghilehik, M.W., and Braunsberg, H., Actions of medroxyprogesterone acetate on the efficacy of cytotoxic drugs: studies with human breast cancer cells in culture, Int. J. Cancer, 43, (1989) 458-463. 35 Sokal, R.R. and Rohlf, F.J., Introduction to Biostatistics, Freeman, NY, 1987, pp. 185-210. 36 Taiwo, Y.O., Coderre, T.J. and Levine, J.D., The contribution of training to sensitivity in the nociceptive paw-withdrawal test, Brain Res., 487 (1989) 148-151. 37 Thompson, C.I. and Epstein, A.N., Salt appetite in rat pups: ontogeny of angiotensin !I - aldosterone synergy, Am. J. Physiol., 260 (1991) R421-R429. 38 VonVoigtlander, P.F., Lahti, R.A. and Ludens, J.H., U-50,488H: a selective and structurally novel non-mu (kappa) opioid agonist, J. Pharmacol. Exp. Ther., 224 (1983) 7-12. 39 Watzman, N., Barry Ill, H., Kinnard Jr., W.J. and Buckley, J.P., Influence of certain parameters on the performance of mice on the rotarod, Arch. Int. Pharmacodyn., 169 (1967) 362-374. 40 Yaksh, T.L. and Rudy, T.A., Chronic catheterization of the spinal subarachnoid space, Physiol. Behav., 17 (1976) 1031-1036. 41 Yeung, J.C. and Rudy, T.A., Multiplicative interaction between narcotic agonisms expressed at spinal and supraspinal sites of antinociceptive action as revealed by concurrent intrathecal and intracerebroventricular injection of morphine, J. Pharmacol. Exp. Ther., 215 (1980) 633-642. 42 Zorman, G., Belcher, G., Adams, J.E. and Fields, H.L., Lumbar intrathecal naloxone blocks analgesia produced by microstimulation of the ventromedial medulla in the rat, Brain Res., 236 (1982) 77-84. •