Psychopharmacology(1992) 108:177-184
Psychopharmacology © Springer-Verlag 1992
Morphine withdrawal aggression: modification with D1 and D2 receptor agonists J . W . T i d e y and K.A. M i c z e k
Department of Psychology,Tufts University,490 Boston Avenue, Medford, MA 02155, USA ReceivedJuly 15, 1991 / Final version January 28, 1992 Abstract. Morphine withdrawal increases aggressive behaviors, induces explosive motor behaviors, and disrupts homeostatic functions in mice and rats. While many of these effects appear to result from altered dopaminergic activity during morphine withdrawal, the relative contributions of the D1 and Da receptor subtypes remain unclear. In the present experiments, the D1 agonist SKF 38393 and the D 2 agonist quinpirole were administered to male "resident" Swiss-Webster mice 5 h after the removal of a subcutaneously-implanted morphine or placebo pellet. These mice were then observed alone to determine changes in various motor activities and in confrontation with a group-housed male "intruder" to assess changes in aggressive behaviors. SKF 38393 decreased the display of aggressive behaviors by placebo and morphine-withdrawn mice without consistently altering walking or rearing. Quinpirole greatly decreased the display of aggressive behaviors by placebo mice and decreased aggressive behaviors in morphine-withdrawn mice to a lesser degree. The inhibitory effects of quinpirole were not specific to aggressive behaviors; low quinpirole doses also decreased the display of walking and rearing. In mice which received a low dose of SKF 38393 preceding quinpirole injection, pretreatment with the D1 agonist did not alter the effects of the Dz agonist quinpirole on motor activities but maintained high levels of aggression in morphine-withdrawn mice. The differential modification of aggressive and motor behaviors by selective dopaminergic agonists during morphine withdrawal further supports the suggestion that aggressive and motor behaviors are controlled independently; furthermore, D1 receptor stimulation appears to have particular relevance for the display of aggressive behaviors during morphine withdrawal. Key words: A g g r e s s i o n - Opiates - W i t h d r a w a l -
mine - Locomotion - Motor activity Offprint requests to:
K.A. Miczek
Dopa-
Heightened aggressive activity and explosive jumping are two striking behavioral events seen in laboratory mice and rats undergoing opiate withdrawal, a syndrome which also includes thermoregulatory and gastrointestinal disturbances, weight loss, writhing and "wet dog" shaking (Blfisig et al. 1973; Lal 1975a, b; Kantak and Miczek 1988; Vivian and Miczek 1991). Recently, we found that explosive and disrupted motor activities could be differentiated from heightened aggressive behavior on the basis of their emergence and disappearance during morphine withdrawal (Tidey and Miczek 1992). Within 5 h of subcutaneous morphine pellet removal, walking and rearing were markedly reduced, and self-grooming was increased concurrent with the emergence of explosive stereotyped jumping; these alterations in motor activity disappeared within 24-48 h. Heightened levels of threat and attack toward a group-housed male "intruder" mouse persisted for at least 4 days after morphine pellet removal. The temporal dissociation of the motoric and affective-aggressive components of opiate withdrawal suggests that their neural basis differs and may involve short- and long-term regulatory changes at distinct receptors. Pharmacological manipulations of dopamine neurotransmission alter both the aggressive and motor components of morphine withdrawal, d-Amphetamine, apomorphine, cocaine, methylphenidate and L-dopa potentiate morphine withdrawal-induced defensive aggression in rats (Lal et al. 1971; Puri and Lal 1973; Gianutsos et al. 1976) and offensive aggression in mice (Kantak and Miczek 1988) at doses which do not alter aggressive behavior in opiate-naive animals. In contrast, the alpha-2 noradrenergic agonist clonidine decreases attack and threat behaviors of morphine-withdrawn mice below control levels (Kantak and Miczek 1988). Walking and rearing, which are greatly reduced during the early stages of morphine withdrawal, are dose-dependently increased by d-amphetamine; similarly, the high levels of grooming observed in 5-h morphine-withdrawn mice are inhibited by d-amphetamine (Tidey and Miczek 1992).
178 Although the mechanisms by which direct and indirect dopamine agonists modify the behavioral consequences of opiate withdrawal remain unclear, several observations suggest that mesolimbic dopaminergic neurotransmission mediates several behavioral effects of opiates, including locomotor activity. Both 6 - O H D A lesions o f the nucleus accumbens and administration of dopamine antagonists block the motor-activating and reinforcing effects of opiates (Kelley et al. 1980; Stinus et al. 1980; Shippenberg and Herz 1988). Studies employing in vivo microdialysis techniques have demonstrated that m u and delta agonists increase D A release and metabolism in the nucleus accumbens; this release is augmented in freely moving rats and associated with increased m o t o r activity (Di Chiara and I m p e r a t o 1988; Spanagel et al. 1990, 1991). Microinjection o f e n k e p h a l i n analogues to the ventral tegmentum of rats also induces naloxone- and D A antagonist-reversible increases in m o t o r activity which is associated with increased D A metabolism in the nucleus accumbens (Kalivas et al. 1983). During morphine withdrawal, mesolimbic D A release appears to be greatly reduced for at least one week and these animals appear to be more sensitive to the DA-releasing effects of morphine (Acquas et al. 1991). Stimulation o f these sensitized dopamine receptors by dopaminergic agonists m a y result in the heightened aggression during morphine withdrawal (Lal 1975a, b). The development of drugs which target specific subtypes of the dopamine receptor m a y be useful in understanding the role of dopamine in the m o t o r and aggressive consequences of morphine withdrawal. Most behavioral effects of dopamine have been linked to activation of the O 2 receptor subtype, including m o t o r hyperactivity and the production of stereotypy (Seeman 1980). D2 receptor agonists are self-administered and substitute for d-amphetamine in drug discrimination tests (Woolverton et al. 1984; Arnt t988). In contrast, selective D1 agonist drugs such as S K F 38393 do not engender m o t o r activation or stereotypy (Arnt 1985a, b), are not self-administered in monkeys (Woolverton et al. 1984), and only partly substitute for d-amphetamine in drug discrimination tests (Arnt 1988). The only consistently reported behavioral effect of S K F 38393 is enhanced grooming activity (Molloy and Waddington 1984; Arnt 1985a, b; Braun and Chase 1986). A prominent hypothesis of D1 receptor function has been that of a "gating" mechanism, enabling or opposing the functional effects of the D 2 receptor (reviewed by Clark and White 1987). Interestingly, the D1 antagonist SCH 23390 potently decreases schedule-controlled responding and aggressive behavior (Haney et al. 1990), and antagonizes stereotypy induced by D2 receptor stimulation (Arnt 1985a, b). Furthermore, S K F 38393 induces m o t o r activation in 6 - O H D A lesioned or reserpinized rats (Arnt 1985a, b; Breese et al. 1985), suggesting that D1 agonist stimulation does generate behavioral effects in animals with sensitized dopamine receptors. Previous research suggests that interactions of opiate and dopaminergic systems are i m p o r t a n t in several behavioral facets of the opiate withdrawal syndrome, however the differential role of D t and D 2 receptor activation
during opiate withdrawal has not been explored with regard to heightened aggressive behavior. The present experiments use relatively selective D1 and D2 receptor ligands in order to examine the role of these receptor subtypes in different types of m o t o r behaviors which are suppressed during opiate withdrawal as well as heightened aggression which emerges during opiate withdrawal. By administering selective D~ and D2 agonists separately and in combination, we provide evidence that aggressive and motoric consequences of morphine withdrawal are not mutually dependent but involve distinct receptor activation.
Materials and methods Subjects. Subjects were male Swiss Webster (CFW) mice, weighing 23-25 g upon arrival (Charles River Breeding Labs, Wilmington, MA). Male "resident" mice were housed in male-female pairs in clear polycarbonate cages (48 x 27 x 20 cm high) with stainless steel lids, through which water and Purina rodent chow were freely available. Additional group-housed male mice, housed in 48 x 27 x 20 cm high clear polycarbonate cages with stainless steel lids, served as untreated "intruders". The animal housing room was maintained at 214- 1° C, with 3540% humidity and a 12/12 light/ dark photocycle (lights on at 8 a.m.). Animal care and use procedures were reviewed and approved by the Institutional Animal Care and Use Committee (IACUC) at Tufts University. Apparatus. Behavioral observation tests were videotaped using a camera and VHS recorder equipped with a time-date generator. Later, the video recordings were replayed and the start and end time of each behavioral event was quantified using a personal computer and customized software (Princeton Economics, Princeton, MA). Core temperature was measured using a Yellow Springs thermometer (Yellow Springs, OH) with a rectal probe inserted 2 cm into the rectum. Latency to remove the tail from a radiant heat stimulus (tail flick latency) was measured using an EMDE brand apparatus (Richmond, VA). The tail flick apparatus was set to obtain a 1.5-2 s latency under control conditions; a cutoff of 8 s was used in analgesic animals to avoid tissue damage. Procedure. Following our previous procedure (Kantak and Miczek 1988; Tidey and Miczek 1992), male resident mice were housed with females for 3 4 weeks. The residents were then screened for attack behavior toward an intruder male: the female and pups were removed from the home cage and a group-housed male was introduced into the cage. Following the first attack by the resident, attack latency was recorded and the intruder was immediately removed. Approximately 5-10% of the residents did not attack an intruder within three screening opportunities and were not used further. Prior to pellet implantation, core temperatures, tail flick latencies and body weights of the resident mice were obtained. A single 75 mg pellet, wrapped in nylon mesh (Cochin et al. 1979) to ensure later removal in entirety, was implanted subcutaneously on the dorsal surface of each mouse under light ether anesthesia. Four hours after implantation, temperature and tail flick latencies were again measured. Pellets were left undisturbed for 72 h, after which they were removed under light ether anesthesia. Five hours after pellet removal, the residents were administered saline, SKF 38393, quinpirole, or 3 mg/kg SKF 38393 plus quinpirole. Observations were conducted 30 rain after SKF 38393 administration or 15 rain after quinpirole administration. Preceding behavioral tests, temperature and tail flick latency measurements of the resident were obtained and the female and pups were removed from the home cage. The male residents were videotaped alone in
179 their home cage for 5 min for the assessment of walking, rearing and grooming. Immediately thereafter, an intruder was introduced into the home cage and the resident-intruder interaction was videotaped for 5 min after the first attack by the resident. If no attack occurred within 5 min, the intruder was removed; these procedures have been described elsewhere (Miczek and O'Donnell 1978). A trained observer replayed the videotapes and recorded the start and stop times of each element of motor, social and aggressive behavior performed by the resident. Two observers were used, one of which was blind to pellet and drug treatment. There was a high inter-rater reliability between the observers (> 95 %). In this experiment, motor activities consisted of walking, rearing and grooming, social behavior consisted of investigation by the resident of the intruder's anal or genital area, and aggressive behaviors consisted of attack bites, sideways threats, tail rattling and pursuits. Total frequencies and durations of these behaviors were generated by the data acquisition program; the data acquisition procedure has been described in detail elsewhere (Miczek 1983).
Drugs. Morphine base 75 mg and placebo pellets were provided by the National Institute on Drug Abuse. SKF 38393 (Research Biochemicals Inc., Wayland, MA) and quinpirole (Eli Lilly Inc., Indianapolis, IN) were dissolved in 0.9% saline and administered interperitoneally in a volume of 1 ml/100 g. SKF 38393 was administered in doses of 3, 10, 56 and 100 mg/kg. Quinpirole doses were 0.1, 0.3, 0.6 and 1.0 mg/kg. In tests where the D 1 and D2 agonist were given concurrently, 3 mg/kg SKF 38393 was administered 15 min before 0.1-3.0 mg/kg quinpirole. A separate group of 8-10 subjects was used to test each drug dose.
Table 1. Effects of morphine withdrawal on body weight, stereotyped jumping, motor and aggressive behaviors Measurement
Body weight (g) Explosive jumping frequency Walking duration (s) Rearing duration (s) Grooming duration (s) Sideways threat frequency Attack bite frequency Tail rattle frequency
Pellet type Placebo
Morphine
29.52i0.21 2.69 + 0.85 41.17±2.95 65.14±5.44 75.99 ± 8.57 36.02 ± 2.34 25.76 ± 1.98 16.094- 1 . 5 8
27.53:t:0.17"* 17.38 ± 1.99"* 27.53±3.81"* 34.53±5.14"* 101.94± 8.93* 43.24 ± 3.23 35.20 ± 2.75** 22.32± 1.99"
Values represent the mean ± SEM of measurements during a 5-min test session except for jumping (15 min test session) * P< 0.05 or ** P< 0.01 compared to placebo pellet removed control group
or placebo pellets had been in place for 72 h, revealed no differences in temperature or tail flick latencies between morphine-tolerant and placebo mice. Furthermore, m o r phine-tolerant and placebo mice did not significantly differ in frequency or duration o f any aggressive or m o t o r behavior.
Statistical analysis. The effect of morphine pellet implantation on temperature and tail flick latency, and the effect of morphine-withdrawal on body weight were analyzed using a one-way betweensubjects analysis of variance test (ANOVA). The effect of time since pellet removal on morphine-withdrawal jumping was analyzed in a separate group of animals using a two-way between-subjectsANOVA with PELLET as one factor and TIME as the other factor. The effect of morphine withdrawal on motor and aggressive behaviors was analyzed using a one-way between-subjects ANOVA in morphine-withdrawn and placebo mice which received saline injections. To determine the effects of SKF 38393 and quinpirole in morphinewithdrawn and placebo mice, a two-way ANOVA with PELLET as one factor and DOSE as another factor was performed on each behavioral measurement for each drug. Post-hoc analyses of differences between means were determined using the Duncan Multiple Range test.
Results
Morphine pellet implantation F o u r hours after pellet implantation, significant levels of analgesia and h y p o t h e r m i a were observed in the m o r phine-pelleted mice. In mice which were implanted with morphine pellets, tail flick latency increased to 4.53 4-0.16 s, while tail flick latency of placebo mice was 1.83:t:0.02 s; this difference was highly significant [F(1,375)=297.66, P
Psychopharmacology © Springer-Verlag 1992
Morphine withdrawal aggression: modification with D1 and D2 receptor agonists J . W . T i d e y and K.A. M i c z e k
Department of Psychology,Tufts University,490 Boston Avenue, Medford, MA 02155, USA ReceivedJuly 15, 1991 / Final version January 28, 1992 Abstract. Morphine withdrawal increases aggressive behaviors, induces explosive motor behaviors, and disrupts homeostatic functions in mice and rats. While many of these effects appear to result from altered dopaminergic activity during morphine withdrawal, the relative contributions of the D1 and Da receptor subtypes remain unclear. In the present experiments, the D1 agonist SKF 38393 and the D 2 agonist quinpirole were administered to male "resident" Swiss-Webster mice 5 h after the removal of a subcutaneously-implanted morphine or placebo pellet. These mice were then observed alone to determine changes in various motor activities and in confrontation with a group-housed male "intruder" to assess changes in aggressive behaviors. SKF 38393 decreased the display of aggressive behaviors by placebo and morphine-withdrawn mice without consistently altering walking or rearing. Quinpirole greatly decreased the display of aggressive behaviors by placebo mice and decreased aggressive behaviors in morphine-withdrawn mice to a lesser degree. The inhibitory effects of quinpirole were not specific to aggressive behaviors; low quinpirole doses also decreased the display of walking and rearing. In mice which received a low dose of SKF 38393 preceding quinpirole injection, pretreatment with the D1 agonist did not alter the effects of the Dz agonist quinpirole on motor activities but maintained high levels of aggression in morphine-withdrawn mice. The differential modification of aggressive and motor behaviors by selective dopaminergic agonists during morphine withdrawal further supports the suggestion that aggressive and motor behaviors are controlled independently; furthermore, D1 receptor stimulation appears to have particular relevance for the display of aggressive behaviors during morphine withdrawal. Key words: A g g r e s s i o n - Opiates - W i t h d r a w a l -
mine - Locomotion - Motor activity Offprint requests to:
K.A. Miczek
Dopa-
Heightened aggressive activity and explosive jumping are two striking behavioral events seen in laboratory mice and rats undergoing opiate withdrawal, a syndrome which also includes thermoregulatory and gastrointestinal disturbances, weight loss, writhing and "wet dog" shaking (Blfisig et al. 1973; Lal 1975a, b; Kantak and Miczek 1988; Vivian and Miczek 1991). Recently, we found that explosive and disrupted motor activities could be differentiated from heightened aggressive behavior on the basis of their emergence and disappearance during morphine withdrawal (Tidey and Miczek 1992). Within 5 h of subcutaneous morphine pellet removal, walking and rearing were markedly reduced, and self-grooming was increased concurrent with the emergence of explosive stereotyped jumping; these alterations in motor activity disappeared within 24-48 h. Heightened levels of threat and attack toward a group-housed male "intruder" mouse persisted for at least 4 days after morphine pellet removal. The temporal dissociation of the motoric and affective-aggressive components of opiate withdrawal suggests that their neural basis differs and may involve short- and long-term regulatory changes at distinct receptors. Pharmacological manipulations of dopamine neurotransmission alter both the aggressive and motor components of morphine withdrawal, d-Amphetamine, apomorphine, cocaine, methylphenidate and L-dopa potentiate morphine withdrawal-induced defensive aggression in rats (Lal et al. 1971; Puri and Lal 1973; Gianutsos et al. 1976) and offensive aggression in mice (Kantak and Miczek 1988) at doses which do not alter aggressive behavior in opiate-naive animals. In contrast, the alpha-2 noradrenergic agonist clonidine decreases attack and threat behaviors of morphine-withdrawn mice below control levels (Kantak and Miczek 1988). Walking and rearing, which are greatly reduced during the early stages of morphine withdrawal, are dose-dependently increased by d-amphetamine; similarly, the high levels of grooming observed in 5-h morphine-withdrawn mice are inhibited by d-amphetamine (Tidey and Miczek 1992).
178 Although the mechanisms by which direct and indirect dopamine agonists modify the behavioral consequences of opiate withdrawal remain unclear, several observations suggest that mesolimbic dopaminergic neurotransmission mediates several behavioral effects of opiates, including locomotor activity. Both 6 - O H D A lesions o f the nucleus accumbens and administration of dopamine antagonists block the motor-activating and reinforcing effects of opiates (Kelley et al. 1980; Stinus et al. 1980; Shippenberg and Herz 1988). Studies employing in vivo microdialysis techniques have demonstrated that m u and delta agonists increase D A release and metabolism in the nucleus accumbens; this release is augmented in freely moving rats and associated with increased m o t o r activity (Di Chiara and I m p e r a t o 1988; Spanagel et al. 1990, 1991). Microinjection o f e n k e p h a l i n analogues to the ventral tegmentum of rats also induces naloxone- and D A antagonist-reversible increases in m o t o r activity which is associated with increased D A metabolism in the nucleus accumbens (Kalivas et al. 1983). During morphine withdrawal, mesolimbic D A release appears to be greatly reduced for at least one week and these animals appear to be more sensitive to the DA-releasing effects of morphine (Acquas et al. 1991). Stimulation o f these sensitized dopamine receptors by dopaminergic agonists m a y result in the heightened aggression during morphine withdrawal (Lal 1975a, b). The development of drugs which target specific subtypes of the dopamine receptor m a y be useful in understanding the role of dopamine in the m o t o r and aggressive consequences of morphine withdrawal. Most behavioral effects of dopamine have been linked to activation of the O 2 receptor subtype, including m o t o r hyperactivity and the production of stereotypy (Seeman 1980). D2 receptor agonists are self-administered and substitute for d-amphetamine in drug discrimination tests (Woolverton et al. 1984; Arnt t988). In contrast, selective D1 agonist drugs such as S K F 38393 do not engender m o t o r activation or stereotypy (Arnt 1985a, b), are not self-administered in monkeys (Woolverton et al. 1984), and only partly substitute for d-amphetamine in drug discrimination tests (Arnt 1988). The only consistently reported behavioral effect of S K F 38393 is enhanced grooming activity (Molloy and Waddington 1984; Arnt 1985a, b; Braun and Chase 1986). A prominent hypothesis of D1 receptor function has been that of a "gating" mechanism, enabling or opposing the functional effects of the D 2 receptor (reviewed by Clark and White 1987). Interestingly, the D1 antagonist SCH 23390 potently decreases schedule-controlled responding and aggressive behavior (Haney et al. 1990), and antagonizes stereotypy induced by D2 receptor stimulation (Arnt 1985a, b). Furthermore, S K F 38393 induces m o t o r activation in 6 - O H D A lesioned or reserpinized rats (Arnt 1985a, b; Breese et al. 1985), suggesting that D1 agonist stimulation does generate behavioral effects in animals with sensitized dopamine receptors. Previous research suggests that interactions of opiate and dopaminergic systems are i m p o r t a n t in several behavioral facets of the opiate withdrawal syndrome, however the differential role of D t and D 2 receptor activation
during opiate withdrawal has not been explored with regard to heightened aggressive behavior. The present experiments use relatively selective D1 and D2 receptor ligands in order to examine the role of these receptor subtypes in different types of m o t o r behaviors which are suppressed during opiate withdrawal as well as heightened aggression which emerges during opiate withdrawal. By administering selective D~ and D2 agonists separately and in combination, we provide evidence that aggressive and motoric consequences of morphine withdrawal are not mutually dependent but involve distinct receptor activation.
Materials and methods Subjects. Subjects were male Swiss Webster (CFW) mice, weighing 23-25 g upon arrival (Charles River Breeding Labs, Wilmington, MA). Male "resident" mice were housed in male-female pairs in clear polycarbonate cages (48 x 27 x 20 cm high) with stainless steel lids, through which water and Purina rodent chow were freely available. Additional group-housed male mice, housed in 48 x 27 x 20 cm high clear polycarbonate cages with stainless steel lids, served as untreated "intruders". The animal housing room was maintained at 214- 1° C, with 3540% humidity and a 12/12 light/ dark photocycle (lights on at 8 a.m.). Animal care and use procedures were reviewed and approved by the Institutional Animal Care and Use Committee (IACUC) at Tufts University. Apparatus. Behavioral observation tests were videotaped using a camera and VHS recorder equipped with a time-date generator. Later, the video recordings were replayed and the start and end time of each behavioral event was quantified using a personal computer and customized software (Princeton Economics, Princeton, MA). Core temperature was measured using a Yellow Springs thermometer (Yellow Springs, OH) with a rectal probe inserted 2 cm into the rectum. Latency to remove the tail from a radiant heat stimulus (tail flick latency) was measured using an EMDE brand apparatus (Richmond, VA). The tail flick apparatus was set to obtain a 1.5-2 s latency under control conditions; a cutoff of 8 s was used in analgesic animals to avoid tissue damage. Procedure. Following our previous procedure (Kantak and Miczek 1988; Tidey and Miczek 1992), male resident mice were housed with females for 3 4 weeks. The residents were then screened for attack behavior toward an intruder male: the female and pups were removed from the home cage and a group-housed male was introduced into the cage. Following the first attack by the resident, attack latency was recorded and the intruder was immediately removed. Approximately 5-10% of the residents did not attack an intruder within three screening opportunities and were not used further. Prior to pellet implantation, core temperatures, tail flick latencies and body weights of the resident mice were obtained. A single 75 mg pellet, wrapped in nylon mesh (Cochin et al. 1979) to ensure later removal in entirety, was implanted subcutaneously on the dorsal surface of each mouse under light ether anesthesia. Four hours after implantation, temperature and tail flick latencies were again measured. Pellets were left undisturbed for 72 h, after which they were removed under light ether anesthesia. Five hours after pellet removal, the residents were administered saline, SKF 38393, quinpirole, or 3 mg/kg SKF 38393 plus quinpirole. Observations were conducted 30 rain after SKF 38393 administration or 15 rain after quinpirole administration. Preceding behavioral tests, temperature and tail flick latency measurements of the resident were obtained and the female and pups were removed from the home cage. The male residents were videotaped alone in
179 their home cage for 5 min for the assessment of walking, rearing and grooming. Immediately thereafter, an intruder was introduced into the home cage and the resident-intruder interaction was videotaped for 5 min after the first attack by the resident. If no attack occurred within 5 min, the intruder was removed; these procedures have been described elsewhere (Miczek and O'Donnell 1978). A trained observer replayed the videotapes and recorded the start and stop times of each element of motor, social and aggressive behavior performed by the resident. Two observers were used, one of which was blind to pellet and drug treatment. There was a high inter-rater reliability between the observers (> 95 %). In this experiment, motor activities consisted of walking, rearing and grooming, social behavior consisted of investigation by the resident of the intruder's anal or genital area, and aggressive behaviors consisted of attack bites, sideways threats, tail rattling and pursuits. Total frequencies and durations of these behaviors were generated by the data acquisition program; the data acquisition procedure has been described in detail elsewhere (Miczek 1983).
Drugs. Morphine base 75 mg and placebo pellets were provided by the National Institute on Drug Abuse. SKF 38393 (Research Biochemicals Inc., Wayland, MA) and quinpirole (Eli Lilly Inc., Indianapolis, IN) were dissolved in 0.9% saline and administered interperitoneally in a volume of 1 ml/100 g. SKF 38393 was administered in doses of 3, 10, 56 and 100 mg/kg. Quinpirole doses were 0.1, 0.3, 0.6 and 1.0 mg/kg. In tests where the D 1 and D2 agonist were given concurrently, 3 mg/kg SKF 38393 was administered 15 min before 0.1-3.0 mg/kg quinpirole. A separate group of 8-10 subjects was used to test each drug dose.
Table 1. Effects of morphine withdrawal on body weight, stereotyped jumping, motor and aggressive behaviors Measurement
Body weight (g) Explosive jumping frequency Walking duration (s) Rearing duration (s) Grooming duration (s) Sideways threat frequency Attack bite frequency Tail rattle frequency
Pellet type Placebo
Morphine
29.52i0.21 2.69 + 0.85 41.17±2.95 65.14±5.44 75.99 ± 8.57 36.02 ± 2.34 25.76 ± 1.98 16.094- 1 . 5 8
27.53:t:0.17"* 17.38 ± 1.99"* 27.53±3.81"* 34.53±5.14"* 101.94± 8.93* 43.24 ± 3.23 35.20 ± 2.75** 22.32± 1.99"
Values represent the mean ± SEM of measurements during a 5-min test session except for jumping (15 min test session) * P< 0.05 or ** P< 0.01 compared to placebo pellet removed control group
or placebo pellets had been in place for 72 h, revealed no differences in temperature or tail flick latencies between morphine-tolerant and placebo mice. Furthermore, m o r phine-tolerant and placebo mice did not significantly differ in frequency or duration o f any aggressive or m o t o r behavior.
Statistical analysis. The effect of morphine pellet implantation on temperature and tail flick latency, and the effect of morphine-withdrawal on body weight were analyzed using a one-way betweensubjects analysis of variance test (ANOVA). The effect of time since pellet removal on morphine-withdrawal jumping was analyzed in a separate group of animals using a two-way between-subjectsANOVA with PELLET as one factor and TIME as the other factor. The effect of morphine withdrawal on motor and aggressive behaviors was analyzed using a one-way between-subjects ANOVA in morphine-withdrawn and placebo mice which received saline injections. To determine the effects of SKF 38393 and quinpirole in morphinewithdrawn and placebo mice, a two-way ANOVA with PELLET as one factor and DOSE as another factor was performed on each behavioral measurement for each drug. Post-hoc analyses of differences between means were determined using the Duncan Multiple Range test.
Results
Morphine pellet implantation F o u r hours after pellet implantation, significant levels of analgesia and h y p o t h e r m i a were observed in the m o r phine-pelleted mice. In mice which were implanted with morphine pellets, tail flick latency increased to 4.53 4-0.16 s, while tail flick latency of placebo mice was 1.83:t:0.02 s; this difference was highly significant [F(1,375)=297.66, P
