Evidence for an interrelationship between ventral noradrenergic bundle and CNS endorphins in the control of nociception in the rat

21

Pain, 14 (1982) 21-32 Elsevier Biomedical press

Evidence for an Interrelationship between Ventral Noradrenergic Bundle and CNS Endorphins in the Control of Nociception in the Rat M.J. Millan ‘, M.H. Millan and A. Herz Department of Neuropharmncology, Max-Planck-lnstitutfiir Psychiatric, Kraepelinstrasse 2, Q-8000 punier 40 (F. R. G.) (Received 10 July 1981, accepted 10 February 1982)

Summary

The present paper examines the role of the ventral noradrenergic bundle (VB) in relation to endorphins in the control of nociception in the rat. Selective, bilateral destruction of the VB produced a substantial fall in hypothalamic levels of noradrenaline. On day 4 post-surgery, VB-lesioned rats displayed a pronounced elevation in basal nociceptive threshold. This proved to be reversible by the specific opioid antagonist, naloxone, evidential of its mediation by endorphins. It was, however, unaffected by dexamethasone, a suppressor of corticotrophic secretion of j?-endorphin, indicative that this pituitary pool of p-endorphin was not responsible. On day 12, at which time the elevation in nociceptive threshold had disappeared, neither the time course nor the intensity of the antinociception elicited by acute stress or various doses of morphine was attenuated in VB-lesioned as compared to sham rats. These data are evidential that the VB may influence nociceptive thresholds via an interaction with a CNS endorphinergic network. They demonstrate, further, that the VB does not mediate a significant component of the antinociception generated by either morphine or stress.

Introduction

The influence of modulation of the activity of CNS monoaminergic systems upon nociceptive threshold (NT) and the antinociceptive efficacy of opiates has been extensively examined [2,5,9,15,25-28,331. Although evidence for an antinociceptive ’ To whom correspondence and requests for reprints should be addressed. 0304-3959/82/0000-0000/$02.75

0 1982 Elsevier Biomedical Press

action of noradrenaline (NA) in the spinal cord has been acquired [25,27.28,33]. the significance of NA in the brain remains comparatively unclear. This largely reflects the fact that non-selective pharmacological manoeuvres have failed to differentiate between the activity of particular brain noradrenergic networks. Further, although more recent studies have provided evidence that NA does not act antinociceptively in the brain [ 1,6- 10,231. the question as to the role of particular noradrenergic pathways has not, as yet. been resolved. The organization of these is complex and it would be naive to anticipate that independent pathways operate similarly [ 17.3 11. Indeed, in studies of the two major ascending noradrenergic networks, the dorsal bundle and the ventral bundle (VB), the need for selective manipulations of their activity appears paramount since there is evidence that they may play contrasting roles in the control of behaviour [14]. They also appear to interact differently with other monoaminergic pathways [14] and the question of their relationship to endorphinergic systems is of especial pertinence to nociception. Both opioid peptides and opiates have been found to reduce the release of NA from dorsal bundle terminals in the cortex [30] and to suppress the activity of noradrenergic neurones within the locus coeruleus, its primary origin [24]. Concerning the VB. its projection targets, the hypothalamus, limbic system and the periaqueductal and central grey, possess substantial quantities of opioid peptides and in the arcuate nucleus of the hypothalamus, noradrenergic fibres have been found in association with P-endorphinergic perikarya [22,23]. Further, stress mobilizes both hypothalamic pools of NA [ 11,211 and those of P-endorphin throughout the brain [ 19,21,22], whilst biochemical evidence that the VB inhibits the activity of particular brain and pituitary pools of opioid peptides, both tonically and under stress, has recently been obtained [ 18,20,21]. The above observations led us to examine the role of the VB in relation to endorphinergic systems in the control of nociception in the rat.

Methods (1) Production

of lesions

Male Sprague-Dawley rats, 180- 190 g in weight, were housed in groups of 4. with free access to rat chow and water and allowed 5 days adaptation to laboratory conditions and handling prior to surgery. Independent batches of sham and VB-lesioned rats were used in each particular study. Rats were ‘mounted’ under 40 mg/kg pentobarbital anaesthesia in a David Kopf small animal apparatus. The VB was bilaterally destroyed by radio-frequency with a model RFG-4 generator and TM electrode (Radionics, Burlington, Mass.). The electrode tip was positioned at coordinates derived from the studies of Ungerstedt [31] and based on the atlas of Kbnig and Klippel [ 131: frontal + 1.1 mm, vertical + 2.8 mm and lateral * 1.4 mm with respect to the intra-auricular line. A tip temperature of 50°C was held for 8 set for lesions, whereas, in sham operations, the electrode was lowered to a point 1 mm above lesion coordinates, for 8 sec.

23

(2) Verification

of lesion efficacy and selectivity

The NA content of the hypothalamus and cortex was determined on days 4 and 12 post-surgery (p.s.). Rats sacrificed on day 4 were untested and those killed on day 12 comprised animals in which the influence of naloxone upon basal NT on day 4 was tested. Rats were decapitated, their brains removed, the hypothalamus and cortex dissected out and levels of NA determined by liquid chromatography [12]. The remainder of the brain was frozen on dry ice for histological examination. All other sham and VB-lesioned rats were decapitated and their brains frozen for histology. Frontal sections were taken in a cryostat at 40 pm intervals and stained with cresyl violet. (3) Evaluation

of nociceptive

thresholds

(i) Analgesiometric tests Vocalization test. Rats were introduced into loosely restraining, horizontal, plexiglass cylinders and a moistened bipolar electrode was attached to the base of the tail. This was stimulated (rectangular pulses, frequency 50 Hz, duration 10 msec, delivered for 2 set) in a standardized sequence of incremental voltages, and the vocalization threshold, i.e., the current intensity required to elicit a scream, determined. Tail-flick test. The method involves the focussing of light onto the tip of the tail of the rat and the recording of latency to withdrawal. The mean of 7 values, with successive measurements separated by a 10 set interval, was evaluated. Beam intensity was adjusted as appropriate and a cut-off time imposed to preclude tissue damage (see below). Hot-plate test. Rats were placed onto a circular copper plate of diameter 18 cm, maintained at 51 “C by a thermostat-regulated pool of circulating water and surrounded by a vertical plexiglass cylinder 25 cm in height. The latency to licking of either hind paw was determined. (ii) Time course of changes in nociceptive threshold VB-lesioned and sham rats were, in each case, subdivided into two groups. The basal NT of those in the first were evaluated by use of the vocalization and tail-flick tests on days 4, 8, 12 and 20 p.s., whereas rats in the second group were tested only once, with the vocalization test, on day 12. Tail-flick latencies were read prior to vocalization threshold and beam intensity adjusted to give sham latencies of approx. 5-6 set, and a cut-off of 12 set imposed. (iii) Pharmacological manipulation of nociceptive threshold Naloxone. On day 4 p.s., in a blind design, 2 or 10 mg/kg naloxone hydrochloride (2.0 ml/kg, i.p.) or saline was injected and vocalization thresholds measured 10 min thereafter. On day 12 p.s., in an open design, 10 mg/kg naloxone (2.0 ml/kg, i.p.) or saline was administered and thresholds determined 10 min later. Dexamethasone. 500 pg/kg was injected (2.0 ml/kg, i.p.), approx. 24 h and 200

24

pg/kg 2-4 h prior to evaluation the control.

of vocalization

thresholds

on day 4 p.s.; saline was

(iv) Morphine-induced antinociception On day 12 p.s., basal tail-flick latencies and vocalization thresholds were read. rats injected with morphine (4.0 ml/kg, s.c.) or saline and placed in observation boxes. Vocalization thresholds, tail-flick and paw-lick latencies were determined after various doses of morphine at times indicated in Fig. 4. Where more than one test was used, the order - tail-flick, vocalization threshold, hot-plate - was taken. Beam intensity was adjusted to give basal tail-flick latencies of 5 set and a cut-off of 20 set imposed. (v) Stress-evoked antinociception On day 12 p.s., tail-flick beam intensity was adjusted to give basal latencies of 3-4 set with a cut-off of 8 sec. Basal latencies were read, rats exposed to 5 min intermittent, inescapable foot-shock (pulses of 3.5 MA, 300 msec in duration, 30/min) and latencies recorded as shown in Table II.

1610

1270

1020

Fig. 1. Frontal sections of rat brain (based on the atlas of Konig and Klippel): adjacent figures indicate section position according to anterior-posterior coordinates given in this atlas. The locations of the ventral noradrenergic bundle (VB), the dorsal noradrenergic bundle (DB) and the dorsal periventricular bundle (DPB) are shown on the right (stippled). The locations of the red nucleus (RN), medial lemniscus (LM) and substantia nigra (SN) are also indicated. The lesion position, shape and size are shown on the left (shaded area).

25

Results (1) Verification of lesion efficacy and selectivity Fig. 1 represents two series of schematic frontal sections through the lesioned region of the brain of the rat. It may be seen that bilaterally symmetrical zones of destruction were produced which corresponded well with the location of the VB as visualized in the histochemical studies of Ungerstedt and others [17,31]. The lesions incorporated the anterior extremity of the red nucleus and laterally, at their centre, slightly invaded, on a few occasions, the adjacent margin of the substantia nigra pars compacta. In no cases, however, were any motor defects and muscular rigidity or hypoactivity, aphagia and weight loss seen, effects which characterize disruption of, respectively, the red nucleus and substantia nigra. The lesions invaded neither the dorsal bundle nor the medial lemniscus. The primary target of the VB is the hypothalamus and, as evaluated on days 4 and 12 p.s., VB lesions precipitated a substantial, respectively 40 and 55%, fall in hypothalamic levels of NA (Table I). These decreases approximate to those reported by other authors [15], the remaining NA comprising that of the dorsal bundle and ventral periventricular system [ 17,311. Further evidence that these changes are associated with a functional inactivation of the VB is provided by our findings that on days 4 and 12 p.s., in each case, the decrease in hypothalamic levels of NA evoked by stress in other groups of sham animals was not manifested in VB-lesioned rats (not shown) [21]. The complete stability of cortical levels of NA on days 4 and 12 p.s. after VB destruction is corroborative of the histological data demonstrating that the dorsal bundle which, in contrast to the VB, innervates the cortex, is not interrupted by the present lesions (Table I). Any VB-lesioned rat, in which an appropriate pattern of lesioning was not observed upon histological examination and/or in which a fall in hypothalamic

TABLE

I

THE INFLUENCE OF LESIONS OF THE VENTRAL NORADRENERGIC LEVELS OF NORADRENALINE IN THE HYPOTHALAMUS AND CORTEX Day 4: sham (n= 12), VB-lesioned (n: 11); day 12: sham (n=21), indicated. Significance of sham vs. lesioned differences shown. Noradrenaline

VB-lesioned

(n=22).

(pg/g)

Day 4

Day 12

Sham

hypothalamus cortex

2.91 -to.21 0.22*0.02

3.15t0.12 0.19~0.02

VB lesion

hypothalamus cortex

1.81 kO.14 *** 0.23 it 0.02

1.42kO.07 0.20-c0.01

l

** P ~0.001,

Student’s

two-tailed

r-test.

BUNDLE

***

UPON

MeankS.E.M.

SHM

VBL

day L

,

, SHM VBL day 8

,

, SHM VBL , day 12

I SHM VBL day 20

,

Fig. 2. Time course of alterations in basal (a) tail-flick latency and (b) vocalization threshold produced by lesions of the ventral noradrenergic bundle (n= 12) or sham operations (n= 11). Mean 2S.E.M. presented. Asterisks indicate significance of sham vs. lesioned differences. * P GO.05. *** P
levels of NA was not detected. was eliminated from analyses of behavioural data. Those rats in which no fall in the hypothalamic content of NA was seen did not display the characteristic effects of discrete VB destruction. In the remaining rats, no further correlation was found between hypothalamic NA levels on day 12 p.s. and the magnitude of the rise in NT seen on day 4. This may reflect the fact that rats were killed 8 days post-testing or the low degree of variation in levels seen in lesioned rats (see TableI). It is also possible that the great majority of lesions decreased levels to below a ‘threshold’ for strong manifestation of effects [29]. (21 Time course

of lesion-induced changes in basaf noc~ceFt~ve rh~esho~d

As quantified by means of the vocalization and tail-flick tests a, respectively, 45% and 22% elevation in basal NT was seen in VB-lesioned rats on day 4 p.s. (Fig. 2). As measured in the same groups of rats, this antinoeiception had disappeared by day 12. A comparison of the vocalization thresholds of VB-lesioned rats on day 4 as compared to 8, and 8 as compared to 12, p.s., revealed these, in each case, to be significantly different from each other (P ,5 0.05, Wilcoxon matched-pairs test). The fall in sham thresholds seen between days 4 and 8 was also significant (P g 0.05) and probably reflects familiarization to testing. No difference in thresholds on day 12 p.s. emerged between groups of sham and VB-lesioned rats not previously tested,

27

c

rn

7 -

‘I

‘-1

,,

1

10

2

NAL

SAL

NAL

SAL

1 VBL

SHM

Fig. 3. The influence of naloxone upon basal vocalization thresholds of rats subjected to lesions of the ventral noradrenergic bundle or sham operations. Mean* S.E.M. presented. n= 7-8 per column. Significance of lesioned, saline vs. sham, saline and of lesioned, naloxone vs. lesioned, saline indicated. ** P ~0.01, l ** P =~0.001, Student’s two-tailed f-test.

evidental that the loss of antinociception did not represent or ‘learning’: sham (n = 12), 5002 25 PA; VB-lesioned (mean I+ S.E.M.).

a

;15ot

‘300.

an adaptation to the test (n = 12), 488 *25 pA

b

2 v P z L 5

lOO(

200.

500

100.

0 :

(

SAL

,2

5

mg /kg morphme

10

0'

i SAL

t

2

mg/kg

ml1

morphine

-0.

min post morphine

Fig. 4. The influence of lesions of the ventral noradrenergic bundle or sham operations upon the analgesic efficacy or morphine in the (a) vocalization, (b) hot-plate and (c) tail-flick tests. a: sham, A (30 mm), 0 (60 min) post-injection; VB-lesioned, A (30 mm) and 0 (60 min) post-injection. Data from 90 min are omitted for clarity. b: sham (0) and lesioned (0) 90 min post-injection. c: sham (@) and lesioned (0) at various times post-injection of 2 mg/kg morphine. Mean* S.E.M. presented. For all points, na5. No significant sham vs. lesioned differences in any tests. No significant saline vs. basal differences in sham or lesioned rats in any test emerged.

2x

TABLE

II

THE INFLUENCE ANTINOCICEPTION

OF LESIONS OF THE VENTRAL NORADRENERGIC EVOKED BY 5 MIN FOOT-SHOCK STRESS

BUNDLE

UPON

THE

Basal tail-flick latencies and the rise in these elicited by stresi is shown. Sham (n= 8), lesioned (n- 9). meaniS.E.M. indicated. No sham vs. lesioned significant differences. 0, 5. 10 and 20 min time-points significantly (P
latency

Basal

(set) Minutes

post-stress

0

5

IO

20

55

Sham

4.00-‘0.14

5.82’0.14

6.68 2 0.09

5.76 -c 0.07

6.17i0.18

4.08 ) 0.06

VB lesion

4.03 LO. 15

6.37 eO.23

6.51eO.24

5.83~~0.1

I

6.4Oi0.17

4.19*0.09

(3) Pharmacological

manipulation

of basal nociceptive

threshold

Neither 2 nor 10 mg/kg naloxone, on day 4 p.s.. altered the vocalization threshold of sham animals (Fig. 3). Both doses, in contrast, strongly attenuated the antinociception revealed by VB-lesioned rats (Fig. 3). evidential that a substantial component of this antinociception is endorphinergic in nature. Pretreatment with dexamethasone, which suppresses adenohypophyseal release of ACTH, P-lipotropin and /?-endorphin, was, however, ineffective in attenuating this antinociception on day 4 p.s., suggestive that an augmented corticotrophic secretion of /I-endorphin cannot underlie the antinociception produced by lesioning the VB (sham, saline 432 I+ 29 PA; sham, dexamethasone 386 2 17 PA; VB-lesioned, saline 605 -t 42 PA; VB-lesioned, dexamethasone 549 * 46 PA; mean * S.E.M., n 2 7). On day 12 p.s.. after subsidence of the antinociception, naloxone did not produce a significant hyperalgesia in VB-lesioned (or sham) animals: sham, saline 441 * 34 PA; sham. naloxone 461 * 41 PA; VB-lesioned, saline 469 * 5 1 PA; VB-lesioned, naloxone 467 i- 28 PA; mean * S.E.M., n 2 7). This observation is demonstrative of the loss of the endorphinergic contribution to basal thresholds in VB-lesioned rats. (4) Morphine-

and stress-evoked

antinociception

On day 12 p.s., at no time post-application, was any shift in the dose-response curve for morphine-evoked increases in vocalization thresholds seen between sham and VB-lesioned rats (Fig. 4a). At 60 mm, ED,,s (i.e., doses of morphine required for half-maximal analgesia) for sham and VB-lesioned rats were, respectively, 7.2 and 6.2 mg/kg in this test. Further, in the hot-plate test, no alteration in the

29

dose-response relationship for paw-lick latencies could be detected (Fig. 4b). At 90 min, ED,,s for sham and VB-lesioned rats were, respectively, 5.3 and 4.5 mg/kg in this test. Finally, neither the intensity nor the time course of the rise in tail-flick latencies evoked by a dose of 2.0 mg/kg morphine differed between sham and VB-lesioned animals (Fig. 4~). The pattern of antinociception generated by 5 min foot-shock stress was, also, unmodified in VB-lesioned as compared to sham rats on day 12 p.s. (Table II).

Discussion Selective interruption of the VB produced a temporary but pronounced rise in basal NT indicative of an involvement of this pathway in nociceptive processes. Consistent with the majority of studies naloxone failed to modify the NT of sham rats, indicative that endorphins are not major determinants of basal NT in intact animals [5,19,22,25]. It did, however, greatly attenuate the antinociception evoked by destruction of the VB, evidential of an endorphinergic mediation of this rise in NT. It is, further, suggested that the VB may modulate nociception via a, probably inhibitory, interaction with certain pools of endorphins. Such an interpretation is supported by our parallel studies in which naloxone-reversible alterations in other behaviours were observed in lesioned rats (Millan, in preparation) and in which biochemical evidence for an enhancement of the activity of particular pools of endorphins was acquired [l&20,21]. A rise in circulating levels of fi-endorphin immunoreactivity was, thus, detected in lesioned rats on day 4 p.s. [l&20,21]. This persisted, however, on day 12 at which time the antinociception had subsided. Further, the suppression of adenohypophyseal release of /3-endorphin by dexamethasone did not moderate this elevation in NT (see Results). An enhancement of pituitary secretion of /3-endorphin is, thus, unlikely to be responsible for this antinociception. Of especial interest were the indications of an augmented activity of met-enkephalin pools in the peri-aqueductal and central grey which paralleled the time course of the antinociception in being pronounced on day 4 but absent on day 12 p.s. [21]. This region, innervated by the VB [17,31], is a major site for the generation of naloxone-sensitive antinociception by microinjection of endorphins or electrical stimulation [1,33]. This stimulationevoked rise in NT is further potentiated by disruption of NA synthesis [l] whilst the cr-adrenergic antagonist phentolamine-naloxone reversibly suppresses sciatic nerve stimulation-evoked unit activity in the central grey [23]. Other studies employing a diversity of techniques have also recently provided evidence for a possible ‘hyperalgesic’ action of NA in the brain [6-8,101. The present data identify the VB as a particular noradrenergic pathway fulfilling such a role which contrasts with that of the antinociception-mediating descending spinal noradrenergic projection [25,27,28,33]. Other noradrenergic networks in the brain may, interestingly, act differently to the VB and selective lesions of neither the dorsal bundle nor locus coeruleus produce a rise in NT [9,14,25,26] (Millan, in preparation). The dissipation of the antinociception produced by VB destruction (Fig. 2)

30

suggests that this pathway is not essential for maintenance of a ‘normal’ NT. Since other effects of VB destruction, for example, hyperactivity, hyperphagia (Millan, in preparation) and elevated plasma levels of &endorphin immunoreactivity, are still prominent on day 12 [ 181, a supersensitivity to NA released by residual fibres is unlikely to account for this loss of antinociception [ 18,2 11. The lack of a diminution in the antinociceptive efficacy of morphine (Fig. 4) is, further, inconsistent with the possibility that this loss of lesion-evoked antinociception reflects the development of tolerance to the release of endorphins. The loss may, it is suggested, represent a decrease in the release of the antinociception-producing pool of endorphins, an interpretation in agreement with the fact that alterations in brain levels of endorphins, seen in lesioned rats on day 4, for example, of met-enkephalin in the midbrain, are not present on day 12 p.s. [ 18,20,21]. In pharmacological studies of morphine analgesia, the problem of non-selective manipulations of catecholaminergic systems has frequently been compounded by the failure to use more than a single analgesiometric test or dose of morphine [5,25]. Non-selective blockade of the activity of noradrenergic pathways has, thus, been found to either enhance, not effect, or to reduce morphine-evoked antinociception [5,25]. Studies of the VB have also led to conflicting results. Thus, 6-hydroxydopamine lesions failed to modify, whereas electrolytic lesions potentiated the rise in NT induced by morphine [ 15,261. Both studies, however, employed only a single dose of morphine and a single test. In the present study, neither the time course nor the intensity of the antinociception evoked by various doses of morphine in 3 distinctive tests differed between sham and lesioned rats (Fig. 4). These data strongly suggest that the integrity of the VB is not a requisite for the development of a normal pattern of morphine antinociception. The rise in NT evoked by stress is of related interest in view of its partial mediation by endorphins [ 19,221. Although a spinal pool of NA was postulated to be a mediator of the rise in NT associated with conditioned fear, with other stressors, no influence of pharmacological manipulation of noradrenergic activity was seen [3,4,16]. In line with these data, as shown in Table II, the VB is not essential for an antinociceptive response to acute foot-shock stress. The present data, in conclusion, provide evidence for a role of the VB in nociceptive processes. The VB apparently operates differently to the bulbo-spinal descending noradrenergic system with respect to its role in the tonic control of NT [27] and its lack of involvement in opiate-like antinociception. The data constitute. further, the first indication for an influence of a particular noradrenergic pathway upon CNS pools of endorphins. They also, finally, provide novel evidence for an antinociceptive action of CNS networks of endorphins.

Acknowledgements We thank E. Hofschuster for NA determination and provision of facilities for this. M.J. Millan was supported schungsgemeinschaft.

Prof. N. Matussek for by the Deutsche For-

31

References 1 Akil, H. and Liebeskind, J.C., Monoamine@ mechanisms of stimulation-produced analgesia, Brain Res., 94 (1975) 279-296. 2 Bllsig, J. and Herz, A., Interactions of opiates and endorphins with cerebra1 catecholamines. In: S. Szekeres (Ed.), Adrenergic Activators and Inhibitors, Springer, Berlin, 1980, pp. 463-497. 3 Buckett, W.R., Pharmacological studies on stimulation-produced analgesia in mice, Europ. J. Pharmacol., 69 (1981) 279-290. 4 Chance, W.T. and Schechter, M.D., Autoanalgesia: 5 6 7 8 9 10 11 12 13 14 15

16 17 18

19 20

21

22

blockade

by yohimbine,

Europ.

J. Pharmacol.,

58

(1979) 89-90. Dennis, S.G., Melzack, R., Gutman, S. and Boucher, F., Pain modulation by adrenergic agents and morphine as measured by three pain tests, Life Sci., 16 (1980) 1247-1259. Gage, F.H. and Springer, J.E., Opposite and antagonistic effects of intrahippocampal injection of serotonin and norepinephrine on behaviour, Neurosci. Abstr., 6 (1980) 454. Haigler, H.J. and Spring, D.D., Putative nociceptive neurotransmitters in the mesencephalic reticular formation, Life Sci., 29 (1981) 33-43. H.K., Hypoalgesia following microinjection of Hammond, D.L., Levy, R.A. and Proudfit, noradrenergic antagonists in the nucleus raphe magnus, Pain, 9 (1980) 85- 101. Hammond, D.L. and Proudfit, H.K., Effects of locus coeruleus lesions on morphine-induced antinociception, Brain Res., 188 (1980) 79-91. Han, J.-S., Tang, J., Ren, M.F., Zhou, Z.F., Fan, S.G. and Qui, X.C., Central neurotransmitters and acupuncture analgesia, Amer. J. Chin. Med., VIII (1980) 33 l-348. Keim, K.L. and Sigg, E.B., Physiological and biochemical concomitants of restraint stress in rats, Pharmacol. Biochem. Behav., 4 (1976) 289-297. Keller, R., Oke, A., Mefford, F. and Adams, R.N., Liquid chromatographic analysis of catecholamines: routine assay for regional brain mapping, Life Sci., 19 (1976) 995-1004. Konig, J.F.R. and Klippel, R.A., The Rat Brain (a Stereotaxic Atlas), Williams and Wilkins, Baltimore, Md., 1963. Kostowski, W., Noradrenergic interactions among central neurotransmitters. In: W. Essman (Ed.), Neurotransmitters, Receptors and Drug Action, Spectrum Publications, New York, 1980, pp. 47-65. Kostowski, W., Jerlicz, H., Bidzinski, A. and Hauptmann, M., Studies on the effect of lesions of the ventral noradrenergic bundle on the antinociceptive action of morphine, Psychopharmacology, 57 (1978) 189-192. Kulkarni, SK., Heat and other physiological stress-induced analgesia: catecholamine-mediated and naloxone-reversible analgesia, Life Sci., 27 (1980) 185- 188. Levitt, P. and Moore, R.Y., Origin and organization of brainstem catecholamine innervation in the rat, J. camp. Neurol., 186 (1979) 505-528. Millan, M.H., Millan, M.J. and Herz, A., Selective destruction of the ventral noradrenergic bundle but not the locus coeruleus elevates plasma levels of P-endorphin immunoreactivity in rats, Neurosci. Lett., 29 (1982) 269-273. Millan, M.J., Stress and endogenous opioid peptides, a review. In: H.E. Emrich (Ed.), The Role of Endorphins in Psychopharmacology, Karger, Basle, 1981, pp. 49-67. Millan, M.J., Jerlicz, M. and Herz, A., Lesions of the ventral noradrenergic bundle produce a naloxone-reversible antinociception and affect brain and pituitary pools of /3-endorphin, Pain, Suppl. 8 (1981) SlO4. Millan, M.J., Millan, M.H. and Herz, A., Evidence for an influence of the ventral noradrenergic bundle upon brain, pituitary and plasma pools of endorphins and the response of these to stress in rats, Neuroendocrinology, in press. Millan, M.J., Przewlocki, R., Jerlicz, M., Gramsch, C., Hollt, V. and Herz, A., Stress-induced release of brain and pituitary /3-endorphin: major role of endorphins in generation of hyperthermia, not analgesia, Brain Res., 208 (1981) 325-338.

23 Palmer, M.R. and Klemm, W.R., Adrenergic alteration of morphine-induced suppression of sciaticevoked unit responses in the central grey, Prog. Neuropsychopharmacol., 2 (1978) 337-347.

32

24 Pepper, C.M. and Henderson, G., Opiates and opioid peptides hyperpolarize locus coeruleus neurones in vitro, Science, 209 (1980) 394-396. 25 Pert, A., Psychopharmacology of analgesia and pain. In: L. Ng and J.J. Bonica (Eds.), Pain. Discomfort and Humanitarian Care, Elsevier, Amsterdam, 1980, pp. 139- 190. 26 Price, M.T.C. and Fibiger, H., Ascending catecholamine systems and morphine analgesia. Brain Res.. 99 (1975) 189-193. 27 Proudfit, H.K. and Hammond, D.L., Alterations in nociceptive threshold and morphine-induced analgesia produced by intrathecally administered amine antagonists, Brain Res., 218 (1981) 393-399. 28 Reddy, S.V. and Yaksh, T.L., Spinal noradrenergic terminal system mediates antinociception. Brain Res., 189 (1980) 391-401. 29 Scapagnini, V., Van Loon, G.R., Moberg, G.P., Preziosi, P. and Ganong, W.F., Evidence for central norepinephrine-mediated inhibition of ACTH-secretion in the rat, Neuroendocrinology, IO (1972) 155-160. 30 Taube, M.D., Borowski, E., Endo, T. and Starke, K., Enkephalin: a potential modulator of noradrenaline release in rat brain, Europ. J. Pharmacol., 38 (1976) 377-380. 31 Ungerstedt, V., Stereotaxic mapping of monoamine neurones in the rat brain, Acta physiol. stand., 90. Suppl. 367 (1971) l-48. 32 Watson, S.J., Richard, C.W., Cieranello, R.D. and Barchas, J.D., Interaction of opiate peptide and noradrenaline systems: light microscope studies, Peptides, 1 (1980) 23-30. 33 Yaksh, T.L. and Rudy, T.A., Narcotic analgesics: CNS sites and mechanisms of action as revealed by intracerebral injection techniques, Pain, 4 (1978) 2999359.