Pergamon

Neuroscienceand BiohehavioralReviews,Vol. 18, No. 3, pp. 373-383, 1994 Copyright1994ElsevierScienceLtd Printed in the USA. All riots reserved 0149-7634/94 $6.00 + .00 0149-7634(93)E0009-B

Physical Dependence Produced by Central Morphine Infusions: An Anatomical Mapping Study

MICHAEL A. BOZARTH

Department o f Psychology, Behavioral Neuroscience Program, State University o f New York at Buffalo, Buffalo, New York 14260-4110
Received 21 A u g u s t 1989 BOZARTH, M. A. Physical dependence produced by central morphine infusions: An anatomical mapping study. NEUROSCI BIOBEHAV REV 18(3) 373-383, 1994.-Morphine sulfate (1.5 nmoles/h) was chronically infused into various brain regions in rats. After 72 h of continuous infusions, an intraperitoneal injection of naloxone hydrochloride (5 mg/kg) was given to determine if physical dependence had developed. Various withdrawal signs were present following the naloxone challenge in rats chronically infused with morphine into the periventricular gray region. These withdrawal signs included escape behavior, wet-dog shakes, and teeth chattering. Several other brain regions (e.g., amygdala, caudate nucleus, lateral hypothalamus, nucleus accumbens, thalamus) were also tested, but morphine infusions into these areas produced only slight physical dependence. The magnitude of physical dependence produced by morphine infusions into the rostra] aspect of the periventricular gray was comparable to that seen following repeated systemic morphine injections, while the physical dependence produced by caudal infusions into the vicinity of the locus coeruleus was considerably less. These data confirm the importance of the periventricular gray region in the development of physical dependence on morphine and reveal that a drug action in the more rostral aspect of this brain region produces the strongest signs of physical dependence. Amygdala Caudate nucleus Chronic infusions Escape behavior Lateral hypothalamus Morphine, naloxone Nucleus accumbens Osmotic minipumps Periventricular gray region Physical dependence Teeth chattering Thalamus Ventral tegmental area Wet-dog shakes Withdrawal behavior

REPEATED exposure to moderate doses of opiates produces physical dependence. When the opiate injections are discontinued or when a narcotic antagonist such as naloxone is injected, withdrawal signs emerge that are generally proportional to the degree of opiate exposure (i.e., dosage administered, number of injections (see 3,48). The narcotic antagonist challenge has the advantage of precipitating withdrawal behavior with a predetermined time-course and usually of somewhat greater intensity than simple discontinuance withdrawal. Thus, studies involving the brain mechanisms mediating physical dependence generally use precipitated withdrawal behavior rather than simple discontinuance withdrawal. In early attempts to determine brain sites responsible for opiate dependence, rats were usually made physically dependent by systemic administration o f morphine and withdrawal was precipitated by central narcotic antagonist administration. In an experiment by Wei et al. (51), rats were made dependent by morphine pellet implantations and crystalline naloxone was applied to various brain sites to assess their 373

contributions to opiate dependence. Sites in the thalamic and periventricular gray regions were shown to be involved in physical dependence, while the application of naloxone to the caudate nucleus or other forebrain sites seldom precipitated withdrawal behaviors. Aware of the problem of extensive drug spread with lipophilic substances like naloxone, Laschka et al. (23) used intraventricular injections of a narcotic antagonist following the insertion o f "plugs" into various parts o f the ventricular system. This procedure restricted the entry o f the narcotic antagonist to certain parts of the cerebral ventricular system. These studies identified the caudal ventricular system (i.e., fourth ventricle) as being proximal to the opiate-receptor field responsible for physical dependence. The failure to find precipitated withdrawal from narcotic antagonist injections restricted to the rostral ventricular system would appear to be in conflict with Wei ct al.'s (51) report o f withdrawal from naloxone application in thalamic sites. This may be due to either (a) the spread of naloxone to more distal sites in the periventricular gray region in Wei et al.'s (51) study; (b) or the

374 failure o f ventricular narcotic antagonist infusions to produce effective concentrations in thalamic sites in Laschka et al.'s (23) study. Whatever the case, both lines of evidence eliminate a forebrain site of action for opiate-induced physical dependence and suggest the importance of caudal midbrain and periventricular gray opiate-receptor fields. A more direct approach to assessing which brain sites are responsible for the development of physical dependence on opiates involves the direct intracranial application of morphine into various brain regions. Wei (49) and Wei and Loh (50) have reported that rats chronically administered morphine into the periaqueductal gray-fourth ventricle region show pronounced withdrawal signs when challenged with systemic naloxone injections. In these studies, morphine was infused chronically using osmotic minipumps (Alzet Corp.; see 45) and various withdrawal signs were observed following a systemic challenge dose o f naloxone. Although a variety of withdrawal signs were demonstrable after central morphine infusions, the most clear appear to have been escape attempts from a cylindrical enclosure, teeth chattering, and wet-dog shakes. In the present study, the procedure developed by Wei (49,50) was used to anatomically map various brain regions for their abilities to produce physical dependence on chronic morphine infusions. Twelve brain sites were selected for study based on previous reports of their roles in physical dependence or on their proposed involvement in other opiate effects. These sites were systematically examined using a standard protocol so that the relative importance of these brain regions in the development of morphine dependence could be determined. The parameters selected for this study were the optimal infusion dose, duration o f infusion, and naloxone challenge dose reported in Wei's (49) study. METHODS
Subjects

BOZARTH TABLE 1 STEREOTAXIC COORDINATES* USED FOR MORPHINE INFUSIONS Placement FCX NAS CPU AMYG LHA THAL D3V VTA VTA-S VTA-30 RF SC PVG-R PVG-C AnteriorPosteriort + 4.4 + 3.5 + 2.0 +0.4 - 0.6 - 1.0 - 3.8 - 3.8 - 3.8 - 3.8 - 3.8 - 3.8 - 3.8 -6.8 Lateral~ + 2.5 + 1.6 + 3.0 +4.8 1.5 2.0 + 0.0 + 0.6 + 0.6 + 0.6 + 1.7 + 2.8 0.6 +0.0 to 0.6 Ventral 2.0 6.2 6.0 7.8 9.2 6.0 4.9 8.2 8.2 8.8 6.8 5.8 5.8 4.5 Subjects 14 15 12 16 10 17 13 12 14 19 14 10 14 20

Male, Long-Evans rats (weighing 300-375 g) were stereotaxically implanted with unilateral 21-gauge stainless steel cannulae aimed at the various brain regions listed in Table 1. While the rats were anesthetized with sodium pentobarbital (60 mg/kg, IP), cannulae that had been filled with sterile morphine sulfate solution were connected to osmotic minipumps (Alzet Corp., Model 2001), and the cannula tips were lowered to the brain regions under study. The minipumps delivered 1 #l/h of 0.5 #g/#l morphine sulfate (1.5 nmole/h morphine base) dissolved in Ringer's solution. Some o f the ventral tegmental cannulae were angled at 20-30 from the midline to avoid penetration o f the periventricular gray region which has previously been shown to produce physical dependence. The minipumps were implanted subcutaneously between the scapulae of the animals, and polyethylene tubing (PE 60) was used to connect the minipumps to the implanted cannulae (For details, see 49,54). PenidUin G (30,000 units, i.m.) was administered prophylactically following surgery. After recovery from the anesthetic, animals were housed in a 12-h light/12-h dark cycle of illumination. All behavioral testing occurred during the fight phase o f this cycle. Food and water were freely available in the home cages. Some rats used in this study to determine the effects o f other experimental manipulations did not receive implanted minipumps nor undergo surgery. Because of a latency o f about four h in reaching their nominal flow rate o f I / z l / h (55,56), the minipumps were kept overnight (range 16-24 h) in isotonic saline at 37C prior to

Two groups of animals implanted with cannulae in the ventral tegmental area had cannulae angled 20-30 from the sagittal plane to avoid the pcriventricular gray region. Superior colliculus and reticular formation placements were also angled 20-30 from the sagittal plane so that they followed the same trajectory as the angled ventral tegmental cannulae. The dorsal third-ventricle placement was angled at 15o from the sagittal plane to avoid the midsagittal sinus. All other cannulae were parallel to the sagittal plane. Abbreviations: FCX, frontal cortex; NAS, nucleus accumbens; CPU, caudate nucleus; AMYG, amygdala; LHA, lateral hypothalamic area; THAL, thalamus; D3V, medial aspect of the dorsal third-ventricle; VTA, ventral tegmental area (naloxone challenged); VTA-S, ventral tegmental area (physiological saline challenged); VTA-30 , ventral tcgmental area (eannulae angled 20-30); RF, reticular formation; SC, superior colliculus; PVG-R, periventricular gray substance (rostral aspect); PVG-C, pedventricular gray substance (caudal aspect including the fourth ventricle and locus caeruleus). Adapted from (33). *Upper incisor bar 5 nan above the interaural line; tram from bregma; ~mm from the midline; ram from dura; number of subjects tested.

implantation in the rats. Also, because the minipumps deliver 1/~l/h for a minimum of 200 h, each was reused in a second subject. The minipumps were bathed in isotonic saline at 37C to maintain their equilibrated flow rate between the first and second group of rats to receive the same minipumps. Fluorometric determinations of flow rate performed by Alzet Corporation showed that the mean infusion rate for this lot of minipumps was 0.99 #l/h + 0.1/zl (see 59). After completion of behavioral testing, rats were deeply anesthetized with sodium pentobarbital (approximately 90 mg/kg, IP) and intracardially perfused with isotonic saline followed by a 10% formalin solution. Their brains were then removed and f'LXedin 10% formalin for at least two days, and frozen sections were taken at 40 micron intervals. Following formal-thionin staining, the cannula placements were examined at approximately 10-times magnification, and the subjects were grouped according to the histological verification of cannula placements.
Procedure

After 72 ( + 2) h o f continuous morphine infusions, the rats were placed in a Plexiglas cylinder (23 x 25 cm). Follow-

BRAIN SITES OF MORPHINE DEPENDENCE ing a 5-min adaptation period, they were challenged with an intraperitoneal injection of naloxone hydrochioride (5 mg/kg) and placed back in the Plexiglas cylinder. The presence of teeth chattering and of wet-dog shakes and the number of escapes from the cylinder were scored in 5-min segments for the next 20 rain (see Table 2). Other withdrawal signs were also observed, but no attempt was made to quantify them because they have been shown to be less reliable than the measures selected in this study (see 51). RESULTS

375

~"

5O

E
O 4o w @ e~

0 M UJ

80

'3

2O

Mesencephalic Placements
Unilateral morphine infusions into either the rostral or candal regions of the periventricular gray substance produced marked physical dependence as quantified by escape behavior following naloxone challenge (see Fig. 1). The animals with cannulae in the ventral tegmental area also demonstrated significant escape behavior, although at about one third the rate of subjects that received morphine into the rostral periventricular gray region. Morphine infused into the medial aspect of the dorsal third-ventricle failed to produce appreciable escape behavior, and this suggests that the responding seen following morphine infusions into the other brain regions was not the result of ventricular diffusion to a distal site of action. An analysis of variance (53) revealed a significant effect for the factor of cannula placements (F (3,61) = 12.406, p < .001). A Newman-Keuls' test (53) further showed that the escape behavior associated with the rostral periventricular gray placement was significantly greater than that found with the other placements shown in Fig. 1 (p~ < .01). Ten unoperated rats were also tested for withdrawal signs following an injection of naloxone (5 mg/kg, IP). None of these rats demonstrated any signs of physical dependence including escape behavior, although several of the subjects did attempt to leave the cylin-

E = Z ca ~

10

o
NT
VTA PVG-R D3V

PVG-C

FIG. 1. Escape behavior following chronic morphine infusions into mesencephalicplacements. The figure shows the mean ( + SEM) number of escapes in 20 rain. NT = unoperated control animals challenged with naloxone. (See Table 1 for other abbreviations.)

TABLE 2 VARIOUS WITHDRAWALSIGNS USED TO SCORE PRECIPITATEDWITHDRAWAL Withdrawalsign Escape behavior Teeth-chattering Wet-dog shakes Criterion Three or more paws on the top rim of the cylindrical enclosure. An audible sound made by rapid mastication-like movements of the jaws. Rotational movements of the body which resemble the motion of a dog shaking himself when wet. A soft, formless stool without definite shape. Squealing upon touch or spontaneously. Attack behavior directed toward the experimenter when touching the subject. More than 2 g loss of body weight within 20 minutes of the naloxone challenge.

der during the 5-rain adaptation period prior to naloxone injections. Other withdrawal signs were also seen in the rats infused with morphine into the periventricular gray regions. As shown in Table 3, teeth chattering was seen with similar frequency in animals infused in either the rostral or caudal periventricular gray regions, but wet-dog shakes were primarily observed following infusions into the caudal periventricular gray region. In fact, this observation is even more apparent if a measure that not only represents frequency but is also sensitive to the intensity of the withdrawal sign is used. Figure 2 was derived by computing the mean number of periods each sign was present during testing. A subject that displayed a given withdrawal sign during only one of the four 5-min periods was assigned a score of one while a subject displaying the withdrawal sign for the entire 20 min of observation (i.e., all four 5-rain periods) received a score of four. These values were then converted to

TABLE 3
P E R C E N T A G E OF A N I M A L S S H O W I N G VARIOUS W I T H D R A W A L SIGNS FOR MORE T H A N ONE 5-MIN PERIOD FOLLOWING NALOXONE CHALLENGE

Placement

Escape Behavior

TeethChattering

Wet-Dog Shakes

Diarrhea Vocalization Aggression Weight-loss

Aggression, diarrhea, and weight-loss were observed in only a few cases of the 200 animals tested in the chronic infusion study. Vocalization scores, initially quantified, were deleted from the study when it became apparent that what was rated as spontaneous vocalization was frequently produced by the animal being approached by the experimenter.

FCX NAS CPU AMYG LHA THAI, D3V VTA VTA-S VTA-30 RF SC PVG-R PVG-C

0 33 0 12 40 41 21 76 25 32 100 0 100 70

0 27 0 0 0 0 14 0 0 10 14 0 43 25

0 7 0 6 0 0 0 0 0 0 0 0 14 45

See Table 1 for abbreviations.

376
40

BOZARTH saline instead of naloxone following chronic morphine infusions. This latter test was used to determine if any of the escape behavior observed in animals implanted in the ventral tegmentum was the result of mechanical irritation from the microinfusions or was related to chronic morphine stimulation in the ventral tegmental area. Figure 3 shows the results of these various treatment conditions. Rats with morphine infused through ventral tegmental cannulae angled to avoid the periventricular gray region showed a small number of escapes when challenged with systemic naloxone, while rats infused in the reticular formation showed escape responding that was intermediate between the rostral periventricular gray (cf. Fig. 1) and unangled ventral tegmental placements. Rats challenged with physiological saline displayed about the same level of escape behavior as those with angled ventral tegmental cannulae and challenged with naloxone. This suggests that some of the responding that remains with cannulae that avoid penetration of the pcriventricular gray region may be the result of mechanical irritation or of general activation produced by the ventral tegmentum morphine infusions. None of the rats with cannulae implanted in the superior colliculus showed any behavioral signs associated with withdrawal. An analysis of variance (53) demonstrated significant differences among animals infused in the various brain regions shown in Fig. 3 (F (4,67) = 9.080, p < .001). A Newman-Keuls' test (53) revealed that morphine infusions into the reticular formation produced significantly more dependence than infusions into the other brain sites (p~ < .05) and that angling the ventral tegmental cannulae produced a significant reduction in escape behavior (p < .05). Forebrain and Midbrain Placements Other brain sites were also tested to determine if chronic morphine infusions would produce physical dependence. Figure 4 shows the results of naloxone challenge after morphine infusions into several forebrain and midbrain regions. The only region that was associated with any appreciable escape behavior was the nucleus accumbens. In this group, three subjects showed a pronounced level of escape responding (mean = 20), but a post hoc histological examination failed to discern any differences between these rats and the other 12 subjects with cannula placements in this region. Likewise, several

C I1. cOD

30

20

O
P

E [': lo

PVG-R

PVG-C

PVG-R

PVQ-C

Teeth-Chattarlng

Wet-Dog Shakes

FIG. 2. Comparison of teeth chattering and wet-dog shakes from rostral (PVG-R) and caudal (PVG-C) periventricular gray morphine infusions. The figure shows the mean ( + SEM) percentage of the 20 min observation period that teeth chattering or wet-dog shakes were present.

the percentage of time the withdrawal sign was present during the observation period (four periods = 10070). A t-test showed that the incidence of wet-dog shakes was significantly higher for animals with cannulae in the caudal periventricular gray region (t (32) = 2.09, p < .05). This supports the notion that different brain regions may be responsible for the production of different signs of physical dependence (e.g., 51) and cautions against considering the withdrawal syndrome as a unitary phenomenon generated by a single brain mechanism. A comparison of Table 3 with Fig. 1 demonstrates the importance of using a measure that takes into account the influence of intensity or duration of the withdrawal sign. Tables only illustrating the percentage of animals that show a given response can be misleading, and rather arbitrary changes in response criterion can radically alter the conclusions derived from such tables. For example, changing the values in Table 3 to represent the percentage of animals that display more than 10 escapes in 20 min would drop the percentage of rostral periventricular gray subjects from 100-79070, but more dramatically, the percentage of angled ventral tegmental subjects (VTA-30 ) decreases from 32070 to 5070. Note also that both the rostral periventricular gray region and the reticular formation infusions of morphine produced physical dependence in 100070 of the subjects tested, but the intensity of the withdrawal escape behavior is markedly different (cf. Table 3 with Figs. 1 and 3). Angled Mesencephalic Placements The physical dependence produced by morphine infusions into the mesencephalic region was further investigated by implantation of cannulae in the ventral tegmental area at a 2030 angle to avoid penetration of the periventricular gray region. This is an important control procedure because appreciable quantities of drug can diffuse up the external wall of the cannula and affect tissue dorsal to the injection site (see 26,38). Additional animals were implanted with cannulae using the same angled penetration but approximately 2.5 and 3.5 mm dorsal to the angled ventral tegmental placement terminating in the reticular formation and superior colliculus, respectively. Another group of rats with unangled ventral tegmental cannula placements were challenged with physiological

A 2

25

E
0

20
@ 0 W 0 0

15

10

E
Z c @

o WrA-S VTA$O VTA RF $C

FIG. 3. Escape behavior following chronic mesencephalic morphine infusions: Comparison of control procedures. The figure shows the mean (+ SEM) number of escapes in 20 rain. (See Table I for abbreviations.)

BRAIN SITES OF M O R P H I N E DEPENDENCE

A r25

377 other investigators. These include abnormal posturing, diarrhea, hyperactivity, mastication, and vocalization (e.g., 3, 27,29,51). All of these signs except diarrhea were seen in the present study. Because there is difficulty in the quantification of these signs and because they tend to be less reliable than the withdrawal signs measured in this study, these other withdrawal signs were not used to assess the intensity of morphine withdrawal.

E
O ~1 @ n 0 m UJ 15 2O

"6
.0

10

E
Z C @

Histological Analysis
Representative histologies are illustrated in Fig. 6. Cannulae aimed at the frontal cortex, nucleus accumbens, and caudate nucleus were dispersed within about a 1 mm radius of the placements shown in the figure. No appreciable signs of withdrawal were seen in any of these rats in this range of placements except in the case of several nucleus accumbens infused animals previously discussed. Cannula placements in the thalamus were generally in the caudal portion of the ventral nucleus, and some lateral hypothalamic cannulae extended to the zona incerta. Sites tested in the amygdala included the central, cortical, and pars lateralis amygdaloid nuclei. Histological analyses were unable to discern any relationships between withdrawal behavior and within site variations in cannula placements for these brain regions. Even if regional variations existed, it would be unlikely that such fine anatomical distinctions could be made because morphine infused at this hourly rate is likely to diffuse more than a millimeter from the infusion site (see (26)). A number of subjects whose data were included in the caudal periventricular gray group had cannulae aimed specifically at the locus coeruleus. Because the size of this structure is about the same as the 21-gauge eannulae used to infuse the morphine, lesions were frequently created by the insertion of the cannulae. Furthermore, since the locus coeruleus is just ventral to the fourth ventricle (see Fig. 6H), infusions delivered to this area are accompanied by appreciable diffusion into the ventricular system. Likewise, morphine infusions into the caudal periventricular gray region readily reach the locus coeruleus by ventricular diffusion. In addition, several subjects tested early in the study had cannulae aimed at the lateral aspect of the caudal periventricular gray region. Even these

o
FOX NAS CPU AMYG THAL LHA

FIG. 4. Escape behavior following chronic morphine infusions into forebrain and midbrain regions. The figure shows the mean ( + SEM) number of escapes in 20 rain. (See Table I for abbreviations.)

of the animals that received morphine infusions into the amygdala, thalamus, and lateral hypothalamus displayed a low level of escape behavior, but no histological differences were apparent among responders and nonresponders in these groups. In general the level of responding seen after chronic infusions into forebrain and midbrain regions was very low and tests for ventricular diffusion were not undertaken because the primary site of action would appear to be associated with the mesencephalic regions which support a much higher level of escape behavior. An analysis of variance (53) failed to show any significant differences following morphine infusions into the brain regions illustrated in Fig. 4 (F (95,78) = 1.657, p > .05).

Characteristics of Withdrawal Responses

Bl~ig et al. (3) have suggested that the withdrawal syndrome precipitated after a narcotic antagonist challenge can be divided into dominant and recessive signs. They reported an inverse relationship between the appearance of dominant signs such as escape behavior and teeth chattering and recessive signs such as wet-dog shakes. They suggested this relationship was a function of both the level of physical dependence and the time since the antagonist challenge. Figure 5 illustrates the distribution of escape behavior, teeth chattering, and wetdog shakes over the 20 min observation period. There was no evidence that the incidence of wet-dog shakes increased toward the end of the observation period (cf. 3; Fig. 6). In fact, the percentage of animals showing these three signs was highest during the initial 5-min period. The discrepant findings of the present study and Blasig et al. (3) could be due to the fact that the latter study quantified the intensity of all three measures while the present study quantified escape behavior but simply tabulated the percentage of animals showing teeth chattering and wet-dog shakes. These signs might have increased in frequency within subjects while the number of animals showing a given sign decreased. Another and probably more important difference is that Bl~ig et al. (3) produced physical dependence with systemic drug injections, and the withdrawal syndrome noted in their study probably reflected the concurrent involvement of several opiate-receptor fields. Nonetheless, the present study suggests that the relationship between dominant and recessive withdrawal signs may be more complicated than previously suggested. A number of other withdrawal signs have been reported by

100 0 0 rn 90 8o 70 \ \ \

E
~

._E
~ @ ~ m

"~

~--A--

el~pQI

so
5O

wet-dog teeth

3O

~
a.

2o
10 0
S 10 15 20

FIG. 5. Distribution of the various withdrawal signs over the 20 min observation period. Only subjects that displayed the corresponding withdrawal sign during at least one of the observation periods were included in the computations. Means for escape behavior represent the number of escape responses in a given time interval while both wet-dog shakes and teeth chattering means represent the percentage of subjects showing these signs in each time interval.

378

BOZARTH

CC

/(

CS

~ 3.8 ~ --,O
B

,,

FIG. 6. Representative histologicalreconstructions of the various brain regions tested in this study. Each placement is indicated by a circle and the angle of cannula penctration by an arrow. It should be noted that the placements illustrated in sections A-F are approximate modal values and appreciable dispersion of cannulae occurred around these sites. Abbreviations: AL, lateral amygdaloid nucleus; BC, brachium conjunctivum (superior cerebellar peduncle); ca, anterior commissure; cc, corpus callosum; cci, commissure of the inferior colliculns; CI, internal capsule; CIF, inferior colliculus; CO, optic chiasm; CPU, caudate nuclcus-putamen; CS, superior colliculus; D3V, dorsal third-ventricle; DMH, dorsomedial nucleus of the hypothalamus; FX, fornix, GP, globns pallidus; IP, interpeduncular nucleus; LC, locus coeruleus; LM, medial lemniscus; LS, lateral septal nucleus; MD, dorsomedial nucleus of the thalamus; MS, medial septal nucleus; NAS, nucleus accumbens; OT, optic tract; POA, lateral preoptic area; PVF, periventricular gray substance; RF, reticular formation; SN, substantia nigra; TOL, lateral olfactory tract; V, ventricle; VD, central nucleus of the thalamus, dorsal part; VE, ventral nucleus of the thalamus; VMH,ventromedial nucleus of the hypothalamus (adapted from (33); ventricular nomenclature from (31).

placements, however, usually resulted in rupturing the fourth ventricle and subsequent cannulae were lowered through the midiine. Histological examination did not reveal any anatomical factor that could account for the variability o f behavioral responses associated with this brain region. Hence, these subjects were treated as a single group. Because the ventral tegmental area has been implicated in the rewarding effects of opiates [for a review, see (5,6,9)], the possible involvement o f this region in the development of physical dependence on morphine is particularly interesting. To fully characterize the anatomical specificity of the physical dependence produced by ventral tegmental morphine infusions, the dependence-producing effect o f morphine infusions into the area surrounding the ventral tegmentum was exam-

ined in detail (see 5). Some cannulae were angled 20-30 from the midllne to avoid penetration of the lateral aspect of the rostral periventricular gray substance; this is an extremely important control because unangled ventral tegmental cannulae penetrate the lateral aspect of the most sensitive brain site for producing physical dependence (see Fig. 6(3). Other cannulae were placed 2.5 and 3.5 mm dorsal to the ventral tegmentum along the same line of trajectory followed by the angled cannulae. This resulted in placements that included the reticular formation and the superior colliculus. Cannulae in the reticular formation were frequently more medial than the placement illustrated in Fig. 6(3. Cannulae in the superior colllculus ranged dorsal to the dorsal third-ventricle and as far anterior as the lateral nucleus o f the posterior thalamus and the lateral

BRAIN SITES OF M O R P H I N E DEPENDENCE geniculate body. The ventral tegmental region tested for its ability to produce physical dependence from morphine infusions included the zone shown to contain the reward-relevant opiate receptors (7,8) and ranged from about 3.2 to 4.2 mm posterior to bregma. The rostral periventricular gray region included an anterior-posterior dispersion similar to the ventral tegmental placements. Figure 7 illustrates the range of tissue damage produced by infusion of morphine sulfate at the rate of llzl/h for 72 h. Many of the histologies showed necrosis in the range illustrated in brain Section A but most were intermediate between the low (Section A) and moderate (Section B) levels of damage. Several animals had extensive damage produced by this chronic infusion regimen, and these subjects were deleted from the data analysis (i.e., see Fig. 7-C). The variability in cell loss was surprising because all subjects received identical osmotic minipumps and because care was exercised to insure that the nominal flow rate of l#l/h had been reached prior to implantation. Nonetheless, there were large variations in the extent of tissue damage following chronic morphine infusions. The modal level of tissue damage was higher than usually seen in studies using a similar infusion volume delivered over a shorter time interval (M. Bozarth, unpublished observations).
10 9 A O~ ~J 0 ...J O~ 0 8 7 8 5 4 3 2 1 0 0 .4 2 10

379

Naloxone Methiodide

(mg/kg)

FIG. 8. Withdrawal weight-loss following systemic morphine injections. The figure shows the mean ( + SEM) weight-loss (g) in 20 min. The 95% confidence interval for animals injected with naloxone hydrochloride is also shown.

Peripheral Component in Weight-Loss

The failure to observe withdrawal diarrhea or significant weight-loss was surprising because (a) opiates have both centrally and peripherally mediated effects on intestinal motility (43,44); (b) intraventricular clonidine inhibits withdrawal diarrhea (41); and (c) the systemic administration of a narcotic antagonist that does not cross the blood-brain barrier falls to precipitate withdrawal diarrhea (A. Herz, unpublished obser-

FIG. 7. Range of tissue damage produced by the chronic infusions. The f'flled circles illustrate the extent of necrotic tissue visible under 10-times magnification in formal-thionin stained brain section. Abbreviations: IP, interpeduncular nucleus; PVG, periventricular gray substance; SN, substantia nigra.

vation, cited in 22). The fact that morphine chronically infused into the caudal periventricular gray region (and the surrounding fourth ventricle) did not result in diarrhea following naloxone challenge does not support the notion that withdrawal diarrhea is mediated by an opiate action at the locus coernleus (cf. 14). Consideration of these data prompted a direct evaluation of the involvement of peripheral opiatereceptors in withdrawal diarrhea and weight-loss. To test the possible contribution of peripheral opiatereceptors to withdrawal diarrhea and weight-loss, 37 rats were made physically dependent from systemic morphine injections and withdrawal behavior was quantified following the administration of a peripherally acting narcotic antagonist. Naloxone methiodide is a quaternary derivative of naloxone which does not cross the blood-brain barrier. Therefore, withdrawal signs precipitated by this compound are presumably mediated by peripheral opiate-receptors. Four groups of rats (n = 812/group) were injected twice daily with morphine sulfate (80 mg/kg, IP every 12 + 2 h) for 3 days. Rats then received an intraperitoneal injection of naloxone methiodide (0.0, 0.4, 2, or 10 mg/kg) dissolved in physiological saline. Additional rats received the same morphine injection regimen and were tested following an intraperitoneal injection of naloxone hydrochloride (5 mg/kg, n = 14). This last group served as a positive control to assess physical dependence following this regimen of systemic morphine injections. Body weights were measured and the presence of diarrhea noted 20 rain following the naloxone injections. Figure 8 shows weight-loss following various doses of naloxone methiodide. This peripherally acting narcotic antagonist precipitated significant weight-loss (F(3,33)= 8.767, p < .001), although diarrhea was seen in only a few subjects. It is possible that diarrhea emerged some time later, but the failure to observe diarrhea shortly after administration of the peripherally acting narcotic antagonist is consistent with the earlier observation by Herz (see 22). Surprisingly, only 21% of the rats challenged with naloxone hydrochloride displayed diarrhea within 20 rain, and many rats showed spontaneous withdrawal as quantified by the weight-loss in subjects not injected with naloxone. Nonetheless, naloxone methiodide appeared to rapidly disinhibit fecal elimination and produced significant weight-loss. These data suggest a peripheral contri-

380 bution to withdrawal weight-loss, although unidentified central opiate-receptors may also be involved as suggested by other investigators. DISCUSSION Figure 9 is a sagittal reconstruction of the brain sites tested in this study. Escape behavior was chosen for this comparison because it only occurs after the development of moderate to severe physical dependence and because it is the most reliable measure for quantification (see 3,23,51). The degree of physical dependence produced by the same regimen of infusions is indicated by the relative density of the various brain sections shown in the figure. The highest level of escape behavior was associated with infusions into the rostral periventricular gray region, while the infusion of morphine into the reticular formation and caudal periventricular gray region also produced appreciable escape behavior. The withdrawal signs observed after infusions into the ventral tegmentum with unangled cannulae appear to be the result of diffusion up the cannula shaft because angling the cannulae to avoid the lateral aspect of the periventricular gray region virtually eliminated this behavior. Furthermore, the escape behavior seen after reticular formation infusions probably also resulted from drug diffusion to the rostral periventricular gray region which is just 1.5 mm medial to this infusion site (see Fig. 6G). Ventricular diffusion of morphine was eliminated as a possible explanation of these effects because infusions into the medial aspect of the dorsal

BOZARTH third-ventricle or the superior colliculus (just ventral to the lateral aspect of the dorsal third-ventricle) failed to produce any significant escape behavior. Teeth chattering was also seen after infusions into the rostral periventricular gray region, although wet-dog shakes were observed more frequently following caudal periventricular gray infusions. These observations provide empirical support for the suggestion of Calvino et al. (10) and Wei et al. (51) that different brain sites are responsible for the production of specific withdrawal signs. Furthermore, none of the groups chronically infused with central morphine showed significant withdrawal weight-loss, while subjects made physically dependent on systemic morphine and challenged with a peripherally acting narcotic antagonist showed significant weight-loss. This latter finding suggests a peripheral component is also involved in the opiate withdrawal syndrome. The relevance of the various withdrawal signs studied in animals to those observed in humans has not been systematically investigated (see 49). It is interesting to note, however, that in an early study by Wikler et al. (52), opiate addicts described nalorphine (an early narcotic antagonist) as a drug that "cleans you out of dope and makes you climb the wall (p. 12)." Unlike the mouse (2,48), morphine dependent rats do not show spontaneous withdrawal jumping when challenged with a narcotic antagonist (M. Bozarth, unpublished observations). The behavior termed "withdrawal jumping" in the rat is not observed when subjects are tested in an unenclosed area. It would appear that the escape behavior observed in rats is in fact escape behavior and motivated by withdrawal distress. Thus, the term "escape behavior" is probably more descriptive of the phenomenon in the rat than the term "jumping" which is frequently used. There are several lines of evidence that suggest that the locus coeruleus is involved in the production of physical dependence on opiates (see 14). First, morphine inhibits the spontaneous activity of locus coeruleus neurons (19) and tolerance develops to this effect following chronic morphine treatment (1). The microiontophoretic application of naloxone onto morphine-tolerant locus coeruleus neurons produces a "withdrawal response" characterized by a dramatic increase in the firing rate of these units (1). Second, electrical stimulation of the locus coeruleus produces behaviors that are similar, in some respects, to opiate withdrawal symptoms (35,36). Third, the locus coeruleus contains the largest concentration of noradrenergic cell bodies in the brain (25,32) and neurochemical evidence has suggested that norepinephrine mediates at least some of the symptoms of physical dependence on opiates: the systemic injection of clonidine, an alpha-2-noradrenergic agonist, attenuates many of the withdrawal signs in both animals (11,13,17,30) and humans (14,15,16,47). The proposed role of the locus coeruleus in withdrawal distress is particularly enticing because there is independent evidence that this brain system may be involved in anxiety and distress (28, 35,37). The caudal periventricular gray region tested in the present study is in close proximity to the locus coeruleus (see Fig. 6H). In fact, a number of animals received infusions directly into this structure, although preliminary data analysis showed no difference between these rats and others infused with morphine into the surrounding periventricular gray-fourth ventricle region. Therefore, the data from these subjects and that from subjects infused into the fourth-ventricle and the surrounding periventricular gray substance were treated as a single group. It should be noted that many of these cannula tracks were contiguous with the fourth ventricle, and physical

D3V

PVG-R

,,,..

,,:,,,,

\
LC

"PvG-c

30 All 1 0 3
::1

MEAN
ESCAPES

FIG. 9. Sagittal reconstruction of the brain regions tested for their ability to produce physical dependence on morphine. The relative density of the various regions indicates the degree of physical dependence produced by chronic morphine infusions as assessed by escape behavior during precipitated withdrawal. The reticular formation placement is lateral to the rostral periventricular gray region and is not shown on the figure. The data for the ventral tegmental placement is taken from the animals with cannulae angled to avoid the periventricular gray substance. LC = locus coerulens. (See Table 1 for other abbreviations.)

BRAIN SITES OF M O R P H I N E DEPENDENCE TABLE 4

A COMPARISON OF THE SEVERITY OF WITHDRAWAL JUMPING REPORTED FOR VARIOUS PROCEDURES IN THE RAT Investigator Agonist Antagonist Jumps* Minutes Tested

381

Bl~isig et al. (3) Calvino et al. (10) Calvino et al. (10) Kruszewska (20) Laschka et al. (22) Laschka et al. (23) Pinsky et al. (34) Schulz et al. (42) Wei et al. (51)
Wei (49)

systemic systemic systenuc systenuc systenuc systenuc systermc systermc systermc PVG-4th ventricle

systemic systemic amygdala systemic systemic 4th ventricle systemic 4th ventricle systemic systemic

30 8 25 13-23t 17:[ 12 7 12 11 25

30 30 30 20 30 30 10 15 to 20 10 15

*Mean number of jumps/escapes; tNote variability in the same study (cf. Figs. 3, 4, p. 660); :[3.3 mg dose; Periventricular gray-fourth ventricle region; 108 nmole dose.

dependence produced by morphine infusions into this placement is very likely to involve an opiate action distal to the infusion site. The present data regarding teeth chattering and wet-dog shakes are consistent with the proposed role of the locus coeruleus in opiate dependence. Escape behavior, however, does not seem to result primarily from a drug action in this region because a much more pronounced level of escape behavior was produced by infusions into the rostral periventricular gray region. This latter finding is consistent with previous reports (11,40) that clonidine falls to reduce escape behavior in animals during opiate withdrawal. Early work investigating the anatomical localization of brain sites involved in opiate dependence relied on the production of physical dependence from systemic drug injections or from pellet implantation and precipitated withdrawal by central narcotic antagonist application. The amygdaloid complex (10,21,46), medial thalamus (46,51), and parts of the striatum (46,51) have all been reported to be sensitive to direct naloxone application. In contrast to these reports, the present study failed to find any appreciable physical dependence after chronic morphine infusions into the amygdala, thalamus, or striatum (see Fig. 4), although if the criterion were changed to the percentage of animals showing a given withdrawal sign, the present study would conf'mn the reported role of the thalamus in opiate dependence (i.e., 46,51). Specifically, 4 1 0 of the subjects in the present study showed escape behavior (see Table 3), while Tremblay and Charton (46) reported responding in 42o70 of their animals; Wei et al. (51) reported that 58o70 of their animals challenged with naloxone in the medial thalamus showed withdrawal behavior (i.e., at least two escape attempts or three wet-dog shakes). Rather than confirming the role of the thalamus in the development of physical dependence, however, these comparisons would appear to emphasize the importance of using a measure that reflects not only the proportion of subjects responding but also the intensity of the response. Considering the intensity of the response, the physical dependence produced by morphine infusions into the thalamus is slight compared with the effect of the same infusions into the periventricular gray region (cf. Figs. 1,4). Because of the problem of excessive drug diffusion following the liquid injection of drugs, both Calvino et al. (10) and Wei et al. (51) used crystalline naloxone application to precipitate withdrawal. The study by Calvino et al. (10) showed anatomical resolution that would suggest crystalline naloxone

spreads only a fraction of a millimeter from the application site (see 10, Figs. 3,4). This is surprising because Routtenberg et al. (39) have shown that crystalline drug application can result in appreciable diffusion after only 11-12 min of application. In the case of dopamine, less diffusion is seen after liquid (4) than crystalline (39) application. Routtenberg (38) has suggested this may be due to the extremely high concentration of drug applied with the crystalline method. The Routtenberg et al. (39) study reported three diffusion patterns: (a) a 1-2 mm spherical distribution around the cannula tip; (b) appreciable "axonal streaming" orthograde to the cannula tip; and (c) significant diffusion to the cerebral ventricular system. Although the sites effective in precipitating withdrawal in the Calvino et al. (10) and Wei et al. (51) studies had cannulae penetrating various aspects of the ventricular system, effective and noneffective sites are not clearly discernible on the basis of proximity to the ventricular system. Nonetheless, diffusion to a distal site of action whether via the cerebral ventricular system or "axonal streaming" seems to be a viable explanation of the discrepant findings of these experiments and the present study. Another possibility would be that the high concentration reached after crystalline naloxone application caused nonspecific effects that resulted in the withdrawal reaction. Without controlling for actions produced by the physicochemical properties of naloxone (e.g., comparison of the potency of active and inactive stereoisomers), these studies are difficult to interpret (see 5). It is likely, however, that the amygdaloid complex is involved in some aspect of opiate withdrawal because lesions of this area (10) or amygdaloid kindling (24) attenuates the severity of the opiate withdrawal syndrome. The present study, however, suggests that a direct action of morphine in this brain region is probably not involved in the development of physical dependence. The importance of the periventricular gray region in the development of physical dependence on morphine was suggested by Herz et al. (17,23). Using a preparation in which a ventricular "plug" was inserted in the cerebral aqueduct, a narcotic antagonist was microinjected into the lateral or fourth ventricle after the animals were made physically dependent on systemic morphine. When the antagonist challenge was restricted to the region of the lateral and third ventricles, few signs of abstinence were precipitated. Microinjections into the fourth ventricle, on the other hand, produced a marked withdrawal syndrome which included many of the signs seen

382 after systemic narcotic antagonist challenge. The observation (in the present study) that infusions o f morphine into the periventricular gray region produced the greatest degree of physicai dependence is concordant with the Herz et al. (18) and Laschka et al. (23) studies. There is, however, one important difference among these studies: their procedure may have prevented drug diffusion to the most sensitive part of the periventricular gray substance. The exact anatomical location of the periventricular "plug" used by Herz et al. (18) and Laschka et al. (23) is difficult to determine based on their published reports. In another experiment using the same procedure, the approximate location of the "plug" was illustrated (see 42, Fig. 1). It would appear that the ventricular "plug" used to restrict the ventricular diffusion of microinjected narcotic antagonist would effectively block drug diffusion to the rostral portion of the periventricular gray region. Thus, the fourth ventricle injections of Laschka et al. (23) would reach the caudal periventricular gray substance tested in the present experiment, hut the "plug" would probably prohibit appreciable drug diffusion to the rostral periventricular gray region found in the present study to be the most sensitive area to the chronic infusion of morphine. Similarly, injections into the lateral ventricle would fail to reach the rostral periventricular gray region. Therefore, the studies using the ventricular "plug" probably missed what may have been the most effective site in producing the opiate withdrawal symptom of escape behavior. The level of responding seen after naioxone challenge in animals chronically infused with morphine into the rostral periventricular gray region compares favorably with that seen in other physical dependence studies. Table 4 summarizes various studies using several methods of demons~cating physical dependence. The intensity of the escape behavior seen in the present study (see Figure 1, PVG-R) is higher than that reported from systemic drug administration suggesting that the

BOZARTH degree of physical dependence in these subjects is as great as that obtainable with systemic morphine administration. This would be expected if the rostral periventricular gray region were a primary site of action for physical dependence produced by systemic opiate injections. Whether the periventricular gray region is heterogeneous with regard to its ability to produce physical dependence on morphine has not been clearly established. Because ventricular flow in this brain region is in a rostral to caudal direction (12), it is possible that this structure is homogeneous and that the greater level of escape behavior after rostral periventricular gray infusions is the result of drug diffusion (via the cerebral aqueduct) to a larger proportion of periventricniar gray opiate receptors. Regardless of the explanation for the strong physical dependence produce by rostra] periventricular gray infusions, this brain region appears to have an important role in the development of opiate physical dependence. Although other brain regions also contribute to the net withdrawal syndrome following abstinence from systemically delivered opiates, the strongest signs of physical dependence are associated with a drug action in a relatively well defined brain region. ACKNOWLEDGEMENTS Martha Asselin is thanked for animal surgery and behavioral testing. Cindy M. Pudiak is thanked for conducting the study with naioxone methiodide. Roy A. Wise is thanked for comments on an earlier version of the paper. And Endo Laboratories (Garden City, N J) is thanked for their generous donation of naioxone hydrochloride. This research was supported by grants from the National Institute on Drug Abuse (DA02285) and from the Natural Sciences and Engineering Research Council of Canada (NSERC). Some data for this report were collected while the author was an NSERC University Research Fellow at the Center for Studies in Behavioral Neurobiology, Concordia University, Montr~ai.

REFERENCES 1. Aghajanian, G. K., Tolerance of locus coeruleus neurones to morphine and suppression of withdrawal response by clonidine, Nature 276:186-188; 1978. 2. Bhargava, H. N.; Way, E. L. Acetylchofinesterase inhibition and morphine effects in morphine tolerant and dependent mice. J. Pharmacol. Exp. Ther. 183:31-40; 1972. 3. Blasig, J.; Herz, A.; Reinhold, K.; Zieglgansberger, S. Development of physical dependence on morphine in respect to time and dosage and quantification of the precipitated withdrawal syndrome in rats. Psychopharmacologia 33: 19-38; 1973. 4. Bondareff, W.; Routtenberg, A.; Narotzky, R.; Mclone, D. G. Intrastriatai spreading of biogenic amines. Exp. Neurol. 28:213229; 1970. 5. Bozarth, M. A. Opiate reward mechanisms mapped by intracraniai self-administration. In: Smith, J. E.; Lane, J. D., eds. Neurobiology of opiate reward process. Amsterdam: Elsevier/North Holland Biomedical Press; 1983:331-359. 6. Bozarth, M. A. The ventral tegmental reward system. In: Oreland, L.; Engel, J., eds. Brain reward systems and abuse. New York: Raven Press; 1987:1-17. 7. Bozarth, M. A. Neuroanatomical boundaries of the rewardrelevant opiate-receptor field in the ventral tegmental area as mapped by the conditioned place preference method in rats. Brain Res. 414:77-84; 1987. 8. Bozarth, M. A.; Wise, R. A. Localization of the reward-relevant opiate receptor. In: Harris, L. S., ed. Problems of drug detxmdance. National Institute on Drug Abuse Research Monograph 41, U.S. Government Printing Office, Washington D.C.: 1982; 158-164. 9. Bozarth, M. A.; Wise, R. A. Neural substrates of opiate reinforcement. Prog. Neuro-Psychopharmacol. Biol. Psychiat. 7: 569-575; 1983. Calvino, B.; Lagowska, J.; Ben-Ari, Y. Morphine withdrawal syndrome: Differential participation of structures located within the arnygdalnid complex and striatum of the rat. Brain Res. 177: 19-34; 1979. Cervo, L.; Rochat, C.; Romandiai, S.; Samanin, R. Evidence for a preferential role of brain serotonin in the mechanisms lending to naloxone-precipitated compulsive jumping in morphinedependent rats. Psychopharmacology 74:271-274; 1981. Clark, R. G. Manter and Gatz's essentials of clinical neuroanatomy and neurophysiology. Philadelphia: F. A. Davis; 1975. Fielding, S.; Wilder, J.; Hynes, M.; Szewczak, M.; Novick, W. J.; Lal, H. A comparison of clonidine with morphine for antinociceptive and antiwithdrawai actions. J. Pharmacol. Exp. Ther. 207:899-905; 1978. Gold, M. S.; Pottash, A. C.; The neurobiologicai impfications of clonidine HCL. Ann. N.Y. Acad. Sci. 362:191-202; 1981. Gold, M. S.; Pottash, A. L. C.; Sweeney, D. R.; Kleber, H. D. Efficacy of clonidine in opiate withdrawal: A study of thirty patients. Drug Alcohol Depend. 6:201-208; 1980. Gold, M. S.; Redmond, D. E.; Kleber, H. D.; Clonidine in opiate withdrawal. Lancet 1:929-930; 1978. Herz, A.; Bl~sig, J.; Papeschi, R. Role of catecholaminergic mechanisms in the expression of the morphine abstinence syndrome in rats. Psychopharmacologia 39:121-143; 1974. Herz, A.; Teschemacher, H.; Albns, K.; Ziegl~,~nsberger,S. Morphine abstinence syndrome in rabbits precipitated by injection of morphine antagonists into the ventricular system and restricted parts of it. Psychopharmacologia 26:219-235; 1972.

10.

11.

12. 13.

14. 15. 16. 17. 18.

B R A I N SITES O F M O R P H I N E

DEPENDENCE

383 Behavioral effects of stimulation of the locus coeruleus in the stumptail monkey (Macaca arctoides). Brain Res. 116:502-510; 19"76. Routtenberg, A. Intracranial chemical injection and behavior: A critical review. Behav. Biol. 7:601-642; 1972. Routtenberg, A.; Sladek, J.; Bondareff, W. Histoehemical fluorescence after application of neurochemicals to caudate nucleus and septal area in vivo. Science 161:272-274; 1968. Samanin, R.; Cervo, L.; Rochat, C.; Poggesi, E.; Mennini, T. Reduction in the number of serotonin receptors in the bralnstem of morphine dependent rats: Relation to blockade of naloxone precipitated jumping by serotonin agonists. Life Sci. 27:11411146; 1980. Schreier, W. A.; Burks, T. F. Suppression of morphine withdrawai diarrhea by clonidine. Pro. West. Pharmacol. Soc. 24: 341-345; 1980. Schulz, R.; Blfisig, J.; Laschka, E.; Herz, A. Site of naloxoneprecipitated opiate withdrawal dissociates from that at which apomorphine reinitiates this phenomenon. Naunyn-Schmiedeberg's Arch. Pharmacol. 305:1-4; 1978. Schulz, R.; Wuster, M.; Herz, A. Centrally and peripherally mediated inhibition of intestinal motility by opioids. NaunynSchmiedeberg's Arch. Pharmacol. 308:255-260; 1979. Stewart, J. J.; Wcisbrodt, N. W.; Burks, T. F. Central and peripheral actions of morphine on intestinal transit. J. Pharmacol. Exp. Ther. 205:547-555; 1978. Theeuwes, F.; Yum, S. I. Principles of the design and operation of generic osmotic pumps for the delivery of semisofid or liquid drug formulations. Ann. Biomed. Eng. 4:343-353; 1976. Tremblay, E. C.; Charton, G. Anatomical correlates of morphine-withdrawal syndrome: Differential participation of structures located within the limbic system and striatum. Neurosci. Lett. 23:137-142; 1981. Washton, A. M.; Resnick, R. B. Clonidine for opiate detoxification: Outpatient clinical trials. Am. J. Psychiatry 137:1121-1122; 1980. Way, E. L.; Loh, H. H.; Shen, F. H. Simultaneous quantitative assessment of morphine tolerance and physical dependence, J. Pharmacol. Exp. Ther. 167:1-8; 1969. Wei, E. T. Enkephalin analogs and physical dependence, J. Pharmacol. Exp. Ther. 216:12-18; 1981. Wei, E. T.; Loh, H. Physical dependence on opiate-like peptides, Science 193:1262-1263; 1976. Wei, E. T.; Loh, H. H.; Way, E. L. Brain sites of precipitated abstinence in morphine-dependent rats. J. Pharmacol, Exp. Ther. 185:108-115; 1973. Wikler, A.; Fraser, H. F.; Isbell, H. N-ailylnormorphine: Effects of single doses and precipitation of acute "abstinence syndromes" during addiction to morphine, methadone or heroin in man (postaddicts), J. Pharmacol. Exp. Ther. 107:8-20; 1953. Winer, B. Statistical principles in experimental design. New York: McGraw-Hill; 1971. Yates, F. E. Use of the minipump for prolonged, continuous microinfusions into the brains of small animals. ALZET Tech. Bull. 1:1-3; 1977. Yates, F. E. A new ALZET osmotic minipump: Model 2001. ALZET Tech. Bull. 3:1-2; 1979. Yates,~F. E. Advance information on a new ALZET product-The model 2001 osmotic pump. ALZET Tech. Bull. 5:1-2; 1981.

19. Korf, J.; Bunney, B. S.; Aghajanian, G. K. Noradrenerglc neurons: Morphine inhibition of spontaneous activity. Eur. J. Pharmacol. 25:165-169; 1974. 20. Kruszewska, A. Participation of serotonin in development of morphine dependence in mice and rats. Pol. J. Pharmacol. Pharm. 32:655-663; 1980. 21. Lagowska, J.; Calvino, B.; Ben-Ari, Y. Intra-amygdaloid applications of naloxone elicits severe withdrawal signs in morphine dependent rats. Neurosci Lett. 8:241-245; 1978. 22. Laschka, E.; Herz, A.; Bl~ig, J. Sites of action of morphine involved in the development of physical dependence in rats: I. Comparison of precipitated morphine withdrawal after intraperitoneal and intraventricular injection of morphine antagonists. Psychopharmacologia 46:133-139; 1976. 23. Laschka, E.; Teschemacher, H.; Mehraein, P.; Herz, A. Sites of action of morphine involved in the development of physical dependence in rats: II. Morphine withdrawal precipitated by application of morphine antagonists into restricted parts of the ventricular system and by microinfusion into various brain areas. Psychopharmacolngia 46:141-147; 1976. 24. Le Gal la Salle, G.; Lagowska, J. Amygdaloid kindling procedure reduces severity of morphine withdrawal syndrome in rats. Brain Res. 184:239-242; 1980. 25. Lindvall, O.; Bjtrklund, A. The organization of the ascending catecholamine neuron systems in the rat brain as revealed by the glyoxylic acid fluorescence method. Acta Physiol. Scand. 412:148; 1974. 26. Lum, J. T.; Nguyen, T.; Felpel, L. P. Drug distribution in solid tissue of the brain following chronic local perfusion utilizing implanted osmotic minipumps. J. Pharmacol. Meth. 12:141-147; 1984. 27. Martin, W. R.; Wikler, A.; Eades, C. G.; Pescor, F. T. Tolerance to and physical dependence on morphine in rats. Psychopharmacologia 4:247-260; 1963. 28. Mason, S. T.; Fibiger, H. C. Anxiety: The locus coeruleus disconnection. Life Sci. 25:2141-2147; 1979. 29. Menon, M. K.; Tseng, L. F.; Loh, H. H.; Clark, W. G. An electromyographic method for the assessment of naloxoneinduced abstinence in morphine-dependent rats. Naunyn-Schmiedeberg's Arch. Pharmacol. 313:43-49; 1980. 30. Meyer, D. R.; Sparber, S. B. Clonidine antagonizes body weight loss and other symptoms used to measure withdrawal in morphine pelleted rats given naioxone. Pharmacologist 18:236; 1976. 31. Mitro, A.; Palkovits, M.. Morphology of the rat brain ventricles, epindyma, and periventricular structures. Basel: Karger; 1981. 32. Moore, R. Y.; Bloom, F. E. Central catecholamine neuron systems: Anatomy and physiology of the norepinephrine and epinephrine systems. Annu. Rev. Neurosci. 2:113-168; 1979. 33. Pelligrino, L. J.; Pelligrino, A. S.; Cushman, A. J. A stereotaxic atlas of the rat brain. New York: Plenum Press; 1979. 34. Pinsky, C.; Dua, A. K.; Labella, F. S. Peptidase inhibitors reduce opiate narcotic withdrawal signs, including seizure activity in the rat. Brain Res. 243:301-307; 1982. 35. Redmond, D. E.; Huang, Y. H. Current concepts: II. New evidence for a locus coeruleus-norepinephrine connection with anxiety. Life Sci. 25:2149-2162; 1979. 36. Redmond, D. E.; Huang, Y. H.; Gold, M. S. Anxiety: The locus coeruleus connection. Soc. Neurosci. Abstr. 3:258; 1977. 37. Redmond, D. E.; Huang, Y. H.; Snyder, D. R.; Maas, J. W.

38. 39. 40.

41. 42.

43. 44. 45. 46.

47. 48. 49. 50. 51. 52.

53. 54. 55. 56.