MECHANISM AND CHARACTERISTICS OF DRUG-INDUCED AGGRESSION

Gerald Gianutsos, Ph.D.: Mechanism and Characteristics of Drug-Induced Aggression. This research investigated the pharmacological alteration of drug-induced aggression in naive, morphine-dependent and chronically haloperidol treated rats. Naive rats were treated with apomorphine (1.25 20 mg/Kg) and aggregated in groups of four for aggression. It was determined that 20 mg/Kg of apomorphine produced maximal aggression, as measured by number of attacks, duration of aggressive posturing and number of vocalizations. Doses less than 5 mg/Kg were without any effect. The neuroleptics, haloperidol and oxyperomide, the narcotic, morphine, and the cholinergic agonist, pilocarpine all reduced apomorphine-induced aggression in a dosedependent manner. Centrally acting anti-cholinergic drugs partially reversed the blockade of aggression produced by all the drugs except morphine which was preferentially antagonized by naloxone, a narcotic antagonist. Aggression was elicited by sub-threshold doses of apomorphine when combined with clonidine; large doses of dexetimide, an anticholinergic; or when administered three days after an injection of p-chloroamphetamine (12 mg/Kg). None of these drugs were capable of eliciting aggression by themselves. Amphetam.ine potentiated the apomorphine-induced aggression, but failed to elicit aggression with sub-threshold doses of apomorphine unless combined with cyproheptadine. I n morphine-dependent rats, aggression was observed during withdrawal produced by withholding of injections, but not in withdrawal induced by naloxone . The aggression which was observed at 72 hours and at 30 days of withdrawal was reduced by morphine, haloperidol or by lesions of the dopaminergic nigro-striatal bundle, and enhanced by small doses of apomorphine. Apomorphine was also capable of eliciting aggression when administered at four hours of withdrawal, with or without naloxone. The 72 hour withdrawal aggression was similar to apomorphine-induced aggression in naive rats since it was dose-dependently increased by dexetimide or clonidine and decreased by pilocarpine. However, in contrast to the effect in naive rats, anticholinergic drugs failed to reverse the blockade of aggression produced by haloperidol in morphine dependent rats. Following chronic treatment with haloperidol, rats demonstrated signs of dopaminergic supersensitivity as shown by enhanced apomorphineor amphetamine-induced stereotypy, increased spontaneous hyperactivity and a shift to the left of the dose-response curve for amphetamine stimulation and for apomorphine-induced reduction in striatal dopamine turnover. In addition, the rats exhibited an increased sensitivity to the stimulation of activity produced by dexetimide and a decrease in the depression of activity produced by pilocarpine. Although there was no spontaneous aggression after discontinuation of chronic halperidol, the threshold dose of apomorphine required to elicit aggression was dramatically reduced. Amphetamine failed to produce aggression after chronic haloperidol, in contrast to the effect normally seen after chronic morphine. The results of this study demonstrated the requirement of central dopaminergic stimulation for drug-induced aggression, and suggested that the aggression was antagonized by the activity of acetylcholine and serotonin and possibly facilitated by norepinephrine. In addition, it suggested that morphine and haloperidol produce an anti-aggression action by different mechanisms, possibly involving a cholinergic component in the case of haloperidol. Finally, the research provided evidence that the dopaminergic supersensitivity following chronic treatment with morphine may be qualitatively or quantitatively different from the supersensitivity following chronic treatment with haloperidol, since spontaneous and amphetaminestimulated aggression are noted only in the former case. It was proposed that morphine interferes with cholinergic and/or serotonergic compensatory mechanisms and that these contribute to t~e aggression.

failed to elicit aggression with sub-threshold doses of apomorphine unless combined with cyproheptadine. I n morphine -dependent rats, aggression was observed during withdrawal produced by withholding of injections, but not in withdrawal induced by naloxone . The aggression -which was observed at 72 hours and at 30 days of withdrawal -was reduced by morphine, haloperidol or by lesions of the dopaminergic nigro-striatal bundle, and enhanced by small doses of apomorphine. Apomorphine was also capable of eliciting aggression when administered at four hours of withdrawal, with or without naloxone. The 72 hour withdrawal aggression was similar to apomorphine-induced aggression in naive rats since it was dose-dependently increased by dexetimide or clonidine and decreased by pilocarpine. However, in contrast to the effect in naive rats, anticholinergic drugs failed to reverse the blockade of aggression produced by haloperidol in morphine dependent rats.
Following chronic treatment with haloperidol, rats demonstrated signs of dopaminergic supersensitivity as shown by enhanced apomorphine-or amphetamine-induced stereotypy, increased spontaneous hyperactivity and a shift to the left of the dose-response curve for amphetamine stimulation and for apomorphine-induced reduction in striatal dopamine turnover.
In addition, the rats exhibited an increased sensitivity to the stimulation of activity produced by dexetimide and a decrease in the depression of activity produced by pilocarpine. Although there was no spontaneous aggression after discontinuatio n of chronic halperidol, the threshold dose of apomorphine required to elicit aggression was dramatically reduced. Amphetamine failed to produce aggression after chronic haloperidol, in contrast to the effect normally seen after chronic morphine.
The results of this study demonstrated the requirement of central dopaminergic stimulation for drug-induced aggression, and suggested that the aggression was antagonized by the activity of acetylcholine and serotonin and possibly facilitated by norepinephrine.
In addition, it suggested that morphine and haloperidol produce an anti-aggression action by different mechanisms, possibly involving a cholinergic component in the case of haloperidol. Finally, the research provided evidence that the dopaminergic supersensitivity following chronic treatment with morphine may be qualitatively or quantitatively different from the supersensitivity following chronic treatment with haloperidol, since spontaneous and amphetaminestimulated aggression are noted only in the former case. It was proposed that morphine interferes with cholinergic and/or serotonergic compensatory mechanisms and that these contribute to t~e aggression.

ACKNOWLEDGEMENTS
The author wishes to express his gratitude to his wife, Adeline, and to his parents, Mr. and Mrs. Costas apomorphine (Senault, 1970) and amphetamine (Chance, 1948).
Therefore, a working hypothesis may be formulated which states 1 2 that a hyperstimulation of dopamine neuronal systems may be responsible for aggression.
A study of the types of drugs which both increase and decrease the degree of aggression may provide some insight into the mechanisms involved in this type of behavior and a method for comparing different types of drugs.
the two drugs.
Yet, there are obvious differences between Dopamine's behavioral effects are modulated by activity of other neurotransmitters. For instance, amphetamine induced stereotypy is decreased by cholinergic (Arnfred and Randrup, 1968) or serotonergic (Weiner ~ al., 1973) activity. These types of interactions are rarely studied in drug-induced aggression.
The objectives of this research, therefore, are threefold. First, the research will examine the role of brain dopamine indrug-induced aggression. Secondly, it will investigate the interaction between dopamine and acetylcholine and serotonin in this behavior, and thirdly, it will compare and contrast haloperidol and morphine after both acute and chronic treatment. 3 The significance of this research is manifold. The primary significance will be the determination of the role of brain dopamine in the expression of drug-induced aggression.
Secondly, this research will provide information relevant to the interaction between dopamine and acetylcholine and serotonin in the CNS, especially in the control of drug-induced aggression.
Thirdly, this research will examine similarities and differences between morphine and haloperidol after both acute and chronic treatment.
Since only the former produces physical dependence (though both share ~any behavioral actions, see above), this research may provide information relevant to drug addiction.
Lastly, this research may have practical significance by serving as a potential model for schizophrenia research.
It has been suggested that schizophrenia may be due to a hyperactivity of dopamine neuronal systems in the CNS (Klawans ~al., 1972a;Matthysse, 1973). It is proposed here, that a similar hyperactivity may be the cause underlying the drug-induced aggression in rodents. If so, the drug-induced aggression may serve as a useful tool for studying brain changes which may be related ~o schizophrenia and serve as a test for screening new potential therapeutic agents.

II. LITERATURE SURVEY
A. AGGRESSION

1) General Aspects
Violence is a well recognized component of human society. However, fighting is not restricted to humans, but is found in a wide variety of animals. Because of the ubiquity of this behavior, it has long . attracted the attention of psychologists, psychiatrists, ethologists, sociologists and other investigators in the behavioral sciences.
In recent years, it has also been studied by pharmacologists who have used aggression to provide information on the complex workings of the brain, to gain insight into the actions of centrally acting drugs and to attempt to rationally develop drugs for the treatment of psychopathological disorders.
Aggression has been defined by Lorenz (1966) as "the fighting instinct in beast and man which is directed against members of the same species." This behavior, in the view of Valzelli (1967), is directed towards removing or overcoming what is menacing the physical or psychologi-4 5 cal integrity of the organism. Aggression is such a widespread phenomenon, that Scott (1958) concluded that it is not an accident of evolution but rather a necessary and useful part of an animal's life. Aggression is necessary, Lorenz (1966) concluded, to establish territoriality, for the selection of the strongest member of a species, and for the defense of the young. Collias (1944) investigated the role of aggression in the ecological pattern of living systems and observed that a great deal of fighting in vertebrates is associated with territoriality.
In addition, he observed that fighting in animal groups tends to become organized into social dominance hierarchies which has the net effect of greatly reducing the amount of fighting and that one of the major physiological factors connected with fighting in vertebrates is the male sex hormone.
Considerable attention has been focused on further delineating the variables underlying aggressive behavior.
The physiological mechanisms responsible for aggression are very complex and involve interactions between neural, hormonal and sensory factors, as well as equally complex psychological factors. For example, it is believed that aggression is strongly stimulus bound.
As an illustration, it is known that rats will kill mice, but not rat pups; male mice will attack other males, but will not attack females; attacks precipitated by stimulation of the 6 hypothalamus of the cat will be directed at a live rat, but not at a stuffed rat (see Moyer, 1968). Furthermore, it is well recognized that animals respond to postures of their opponents which results in inhibition of aggression. In addition, other psychological factors such as success at fighting influence aggression. Successful, experienced animals are much more likely to fight than inexperienced ones or animals which have lost a battle (Scott, 1962).
Thus it is clear that aggression is an extremely complex behavior and any hypothesis which purports to explain aggressive behavior would have to be equally complex. Moyer (1968) has classified aggression i~to seven categories. These are predatory, inter-male, fear-induced, irritable, territorial defense, maternal and instrumental.
These types of aggression were differentiated on the basis of stimulus situations which elicit the behavior. Moyer ~ointed out that different neuroanatomical loci, neural mechanisms and topographical patterns may accompany the different types of aggression, so that a global explanation encompassing all types of aggression may be an insurmountable task. Therefore, it is not surprising that a drug may affect different types of aggression in different ways. Some of these methods used to study tpe effects of drugs on aggression are presented in the next section.

2) Models of Aggression
There is a startling variety of methods by which aggression may be studied in the laboratory. The most widely used procedures are aggression produced by a painful stimulation (Tedeschi~ al., 1959;Ulrich and Azrin, 1962), prolonged social isolation of mice (Yen ~ al., 1959) or the administration of drugs. Some other frequently utilized models for the study of aggression include mouse killing (muricide) by rats (Karli, 1956); frog killing (ranacide) by rats (Bernstein and Moyer, 1970); hyperemotionality induced by lesions of the septum (Brady and Nauta, 1953); rage responses following intracerebral injections of chemicals (see Myers, 1975); aggression during extinction of an operant task (Azrin ~al., 1966); and the introduction of a strange rat into the residence of -another (Miczek, 1974). It is probable that the models each have a different physiological basis. The characteristics of the most popular models will be briefly described.
In the isolation paradigm, male mice are housed singly for a period of at least three weeks (Yen et al., 1959).
The aggression occurs spontaneously when the isolated mouse is exposed to another mouse. The mice become progressively more aggressive with increasing duration of isolation and progressively less aggressive as the size of the group in which they live increases (Welch and Welch, 1970). The isolated mice are more sensitive to the effects of CNS 8 stimulants and less sensitive to .the effects of CNS depressants (Lal et al., 1972a). This type of aggression has been extensively utilized for the testing of drugs and is blocked by a wide variety of pharmacological substances (Janssen ~ al., 1960;Valzelli, 1967) apparently with little selectivity.
The isolation report~dly produces a decrease in the turnover of brain serotonin and norepinephrine and an increase in the turnover of brain dopamine (Valzelli, 1970).
In pain-induced aggression, two animals are enclosed in an area from which they cannot escape and a painful stimulus, usually an electric shock, is delivered to the animals. The aggression occurs in females a~ well as males, but is not directed towards inanimate objects in the case of rats (Ulrich and Azrin, 1962). This procedure has also been used for screening drugs (Tedeschi et al., 1959).
Pain-induced aggression appears to be primarily related to the activity of brain catecholamines, especially norepinephrine (Eichelman, 1973).
Pharmacological investigation of aggression may also be carried out in fighting that occurs as a result of the prior administration of a drug. Several types of drugs are capable of producing aggression in laboratory animals, especially apomorphine (Senault, 1970) and amphetamines (Lal ~al., 1970;Hasselager ~al., 1972). Aggression also occurs during withdrawal following chronic treatment with narcotics (Boshka ~al., 1966;Lal~ al., 1971). The apomorphine-induced and morphine withdrawal aggression 9 will be treated more fully below. The drug-induced aggression paradigm appears to be related to dopaminergic activity in the CNS a) Apomorphine-Induced Aggression-Large doses of apomorphine, a dopamine receptor agonist (Anden ~al., 1967), produce aggression in rats (Senault, 1970;McKenzie, 1971).
The aggression is directed only towards a similarly treated rat and is not observed when a non-treated rat or a mouse is put into contact with an apomorphine treated rat (McKenzie, 1971). The aggression occurs with increased frequency in older (84 day old) rats and is absent in rats less than 49 days of age (McKenzie, 1971).
This effect is seen only in male rats (McKenzie, 1971).
Castration or hypophysectomy decreases, while large doses of testosterone increase the degree of aggression (Senault, 1972).
When lesions are made in the amygdala or the substantia nigra of rats, apomorphine-induced aggression is reduced. In contrast, lesions which destroy the septum or olfactory bulb enhance the degree of aggression but fail to produce aggression in normally non-aggressive rats (Senault, 1973).
b) Morphine Withdrawal Aggression -Rats can be made dependent on narcotics and will exhibit a number of behavioral changes during withdrawal from these drugs . Dependent rats undergoing withdrawal from morphine are hyperirritable and, when placed together, will show vigorous fighting (Boshka et al., 1966). The aggression is blocked by continued treatment with morphine, but emerges when drug administration is discontinued. The aggression reaches its maximum approximately three days after withdrawal has been initiated (Thor et al., 1970;Puri and Lal, 1973a). The fighting consists of vocaliza-11 tions, aggressive postures, biting and other forms of attack and often leads to severe injury or death of some of the rats (Lal, 1975). Abstinence from other narcotics, such as etonitazene ( F lorea and Thor, 1968), methadone (Singh, 1973) or fentanyl (Lal~ al., 1975b) also results in withdrawal aggression, but spontaneous aggression is not noted after chronic treatment with amphetamine, alcohol or barbiturates (Pu~i and Lal, 1974b)~ Morphine withdrawal aggression has also been reported in mice (Weissman, 1973) and guinea pigs (Goldstein and Schulz, 1973). Rage and diffi-.
culty in handling has also been reported during narcotic withdrawal in monkeys (Seevers and Deneau, 1963).
The aggression differs from natural aggression in that posturing by the subordinate animal which usually leads to cessation of aggression, does not inhibit fighting during withdrawal, and a withdrawn rat will attack a dead rat (Crabtree and Moyer, 1973). Some aggression, although to a lesser degree, is shown by females in morphine withdrawal (Crabtree and Moyer, 1973). Older rats are more responsive than younger ones (Davis and Khalsa, 1971).
The aggression occurs spontaneously during withdrawal but can be markedly increased by directly (Puri and Lal, 1973a) or indirectly (Thor, 1969;Lal~ al., 1971) acting dopamine receptor stimulants. These observations first suggested that dopamine supersensitivity develops during narcotic addiction (Puri and Lal, 1973a). Small doses of d-or 1-amphetamine potentiate the aggression and are near-ly equipotent in this regard (Lal !U_ al., 1971). Hydroxyamphetamine, which has only a peripheral action, does not en~ance the aggression (Lal et al., 1971). Similarly, the aggression is markedly enhanced by small doses of apomorphine or L-DOPA and is decreased by dopamine receptor blockade by haloperidol or by inhibition of dopamine synthesis by alpha-methyl-p-tyrosine (Puri and Lal, 1973a). Apomorphine also induces aggression in morphine dependent mice (Iorio~ al., 1975). These results strongly suggest the involvement of brain dopamine in morphine addiction and withdrawal.
In addition, morphine withdrawal aggression is also enhanced by extracts of Cannabis sativa (Carlini and Gonzalez, 1972 Based upon these studies, the following dopaminergic path-13 ways have been described: a) Nigro-neostriatal Dopamine neurons b) Mesolimbic Dopamine neurons c) Tubero-infundibular Dopamine neurons a) Nigro-Neostriatal Dopamine Neurons -The nigro-neostriatal system seems to originate mainly in the pars compacta of the substantia nigra (Anden et al., 1964). Support for this comes from fluoTescence microscopy and the high dopamine content observed in this area (Hornykiewicz, 1966).
Unilateral pars compacta lesions result in a 60 per cent lowering of the dopamine level in the corpus striatum of the operated side when compared with the unoperated side (Faull and Laverty, 1969). The caudate nucleus and putamen show a fairly strong green to yellow fluorescence due to the high dopamine content. This fluorescence was reduced in animals with lesions of the-substantia nigra.
A clear correlation was found between the fluorescence reduction and the extent of destruction of the pars compacta (Anden il_ al·, 1964). Some cell bodies are also found in the zona reticulata and the pars lateralis of the substantia nigra, which also belong to the nigroneostriatal dopamine system (Fuxe et al., 1970). Recently, studies after the removal of the nucleus caudatus putamen suggest that the cell bodies of the ventrolateral part of the midbrain tegmentum belong to this uncrossed neuron system, in as much as they show marked reduction in fluorescence intensity and signs of atrophy after such 14 operations (Fuxe et al., 1970). There is now a virtually  York, 1967;York, 1970 (Connor, 1968; , 1974).
These cyclase systems are considered to resemble the "dopamine receptor" (Kebabian et al., 1972 (Fuxe, 1963;Fuxe and Hokfelt, 1966;. In the external layer of the median eminence, the axons give rise to a densely packed plexus 6f dopaminergic nerve terminals, which exerts an axo-axonic influence in the layer (Hokfelt, 1967).
This very short dopaminergic intrahypothalamic system regulates the discharge of the releasing and inhibiting factors from the median eminence (Fuxe ~al., 1970). The release of doapmine in the median eminence acts locally on terminals storing luteinizing hormone releasing factor (LHRF) to inhibit the release of LHRF from the median eminence. This system also participates in mediating the negative feedback action of estrogen and testosterone on gonadotropic secretion, because estrogen and testosterone markedly increase the turnover of tubero-infundibular dopamine in neurons of castrated rats, resulting in increased release of dopamine in the area (Fuxe ~ ~., 1967;. The blockade of ovulation by synthetic estrogens and their derivatives may at least be partly mediated via activation of this neuron system . This system is also highly sensitive to prolactin, which markedly increases the turnover of dopamine in tubero-infundibular neurons .
d} Cortical Dopamine Neurons -In addition to the well defined tracts described above, large quantities of dopamine have also been found in the cortex of various species (Bertler and Rosengren, 1959 (Thierry~&., 1973a). In addition, a specific dopamine reuptake process has been demonstrated in the cerebral cortex of normal rats and in rats whose ascending noradrenergic pathways have been selectively destroyed (Tassin~ al., 1974). The dopamine cells in the cortex have recently been visualized by histofluorescence techniques (Hokfelt ~al., 1974). The dopamine was shown to be distributed throughout the limbic cortex, but was not present in the neocortex (Hokfelt et al . ,19 7 4) • Part of the dopamine terminals in the cortex originate in the A9 and AlO areas of the substantia nigra (Lindvall and Bjorklund, 1974). A dopamine-sensitive adenylate cyclase has also been found in the cortex (Von Hungen and Roberts, 1273}. 2) Dopamine Release and Turnover,in the CNS Using a push-pull cannula, McLennan (1964) demonstrated the resting release of dopamine from the caudate nucleus, which was estimated to occur at the rate of lng/min.
Electrical stimulation of the substantia nigra produced an increase in dopamine outpu~ in the putamen ('McLennan, 1965) and an increase in dopamine and its metabolite, homovanillic acid (HVA) in the lateral ventricle of the cat (Portig and Vogt, 1969). A frequency and intensity related release of labelled dopamine from the caudate nucleus was also demonstrated after stimulation of the nigro-striatal bundle or the caudate (Von Voightlander and Moore,197la;197lb).
Similar results were shown when caudate dopamine stores were specifically labelled by radioactive tyrosine {Chiueh and Moore, 1973). Electrode placements more medial or lateral in the hypothalamus failed to induce dopamine release (Chiueh and Moore, 1973). These results clearly implicate dopamine as the neurotransmitter in the nigrostriatal pathway.
The release of dopamine can be further enhanced by potassium or by drugs such as amphetamine (Besson~ al., 1971). The amphetamine induced release is prevented by lesions of the nigro-striatal bundle (Von Voightlander and Moore, 1973a}. Lesioning of the nigro-striatal bundle produces a gradual degeneration of dopamine terminals and a marked ·reduction of dopamine levels in the striatum (Poirer and 20 Sourkes, 1965;}loore··l l ·a1., 1971;Oltsman and Harvey, 1972). However, inhibition of the firing of the nigrostriatal bundle by lesions or administration of gammahydroxybutyrate causes an immediate increase in the levels of caudate dopamine (Walters et al., 1973).

--
The effect of drugs on the release and utilization of brain dopamine is usually indirectly determined after assessing the drug's effect on dopamine turnover. The turnover of dopamine refers to the overall rate at which the dopamine stores are replaced in a given tissue. The turnover of rat brain dopamine has been estimated to occur at the rate of 0.21 ug/g/hr, with a turnover time of 3.6 hours (Neff and Costa, 1966). The turnover rate can be influenced by a variety of physiological situations and can also be affected by drugs (see section B.4). The overall rate of catecholamine formation is governed by the activity of tyrosine hydroxylase (Levitt~ al., 1965), the first enzyme in the synthetic pathway. The synthesis of new dopamine is regulated by a complex process involving end-product inhibition (Udenfriend g a l . , 1959); activity of post-synaptic receptors (Carlsson and Lindqvist, 1963); activity of pre-synaptic autoreceptors (Roth g ·al., 1975); and brain tyrosine concentrations (Wurtman g a l . , 1974).
Drugs may interact at any potnt, but drug effects are generally considered to occur oy increasing or decreasing dopamine receptor stimulation. In general, a drug which !µcreases dopamine receptor activity would decrease dopa-mine turnover, while a d~ug that decreases dopamine receptor ac ti:v:t ty· would increase dopamine turnover.
will be briefly described.
These effects a) Motor Functions -The degeneration of the nigro-striatal pathway causes the motor dysfunctions associated with Parkinson's Disease (Hornykiewicz, 1966). This degeneration leads to akinesia, rigidity and tremors in human patients.
It is possible to restore normal motor function in Parkinsonian patients by treatment with L-DOPA (Cotzias et .2J:.., 1967) which markedly elevates the level of brain dopamine (Everett and Borcherding, 1971). In animals, the deficiency of striatal dopamine leads to an immobility termed catalepsy. This can be accomplished by large lesions of the mesencephalic-diencephalic areas of the brain which disrupt dopamine activity in the neostriatum (Anden ~al., 1966e) . or by the administration of neuroleptic drugs either systemically (Janssen·~ al., 1965) or directly into the caudate-putamen or globus pallidus (Costa11· et al., 1972).

22
Unilateral lesions o~ the nigro-striatal pathway or the corpus striatum are known to produce asymmetries in movement and posture (Poirer and Sourkes, 1965;Anden et al., 19660}. The difference oetween the two sides of the brai.n may Be fur ther aggravat ed by treatment with drugs that release dopamine from the non-lesioned side. Such animals show a pronounced rotational behavior (Anden ~ al. , 19' 6 6b) • The rotational behavior is further linked to the differences in dopamine levels on the tw~ sides of the _ brain by the finding that unilateral striatal injections of dopamine causes the rats to turn or slowly rotate away from the side where dopamine was injected . Spontaneous rotations towards the intact side are seen 24 to 34 hours after a lesion of the nigro-striatal dopamine system. The direction of the rotation as well as the time point of its occurence is indicative of a degeneration release of dopamine from the lesioned side (Ungerstedt,197la). In a chronically lesioned animal, there is a striking difference between the effects of dopamine-releasing drugs and dopamine receptorstimulating drugs. Amphetamine causes the animal to rotate towards the lesioned side, apomorphine causes it to rotate towards the intact side (Ungerstedt,197la).
rntrastriatal injections of dopamine produce motor hyperactivity in rats (Benkert and Kohler, 1972). Similarly, large doses of L~DO~A combined with a peripheral aromatic acid decarooxylase inhibitor produce an increase 23 in spontaneous locomotor activity which is associated with a seven-fold increase in orain dopamine levels, while norepinephrine remains unchanged (Butcher et al., 1972).
rn addition, the locomotor stimulation produced by amphetamine is believed to be due to release of dopamine (Van Rossum and Hurkmans, 1964;Thornburg and Moore, 1973a;Hollister~ al., 1974). Depletion of brain norepinephrine by inhibition of dopamine-beta-hydroxylase (Thornburg and Moore, 1973a) or treatment with 6-hydroxydopamine (Hollister ~ !:!.1·, 1974) failed to alter the amphetamine-induced stimulation, while selective dopamine depletion did reduce the amphetamine effect.
b) Reward Mechanisms -Reward systems are usually thought of as being noradrenergic (Stein, 1968), but recent evidence has suggested a major role for dopamine as well.
In additiort to the traditional sites of electrode placement, self-stimulation behavior may also be maintained from electrodes placed in the substantia nigra (Arbuthnott et .i!!. •, 1970) or the nucleus accumbens (Phillips et al., ]975), both of which are important dopaminergic sites. Self-stimulation from the lateral hypothalamus is blocked by the dopamine receptor blockers, haloperidol and pimozide (Wauquier and Niemegeers, 1972;Lippa~ al., 1973), but not by phentolamine, a noradrenergic receptor blocker, or oy the inhibition of norepinephrine synthesis with FLA-63 (Lippa· ·et al. , 19 7 3) . Furthermore, dep le ti on of dopamine by 6-hrdroxydopamine depressed brain self-stimulation, wh:tle s:tm:tlar treatment:;;: which deplete norepinephrine were ineffect :tve . The action of amp he tamine in enhancing self-stimulation was similarly affected oy 6-hydroxydopa1lline treatment (Cooper i l al., 1974}. Lastly, it was shown that d-and 1-amphetamine were equipotent in enhancing self-stimulation from dopaminergic sites while d-amphetamine was more potent in noradrenergic sites (Phillips and Pibiger, 1973;Phillips et ·al., 1975).

--
These differences in potencies closely parallelled the in vitro effect of amphetamine in releasing norepinephrine or dopamine (Azzaro and Rutledge, 1973).
In addition, the self-administration of amphetamine is apparently dependent on dopaminergic mechanisms since low doses of dopamine receptor blockers selectively increase lever pressi~g for amphetamine (analagous to the effect seen after decreasing the dose of amphetamine in the self-administered solution), while higher doses produce extinction (Yokel and Wise, 1975). Lastly, apomorphine, a direct . dopamine agonist, is also self-administered by rats (Bax- , 1974). c) Consummatory Behavior -Aphagia ~nd adipsia result from the destruction of dopamine fibers coursing through the lateral hypothalamus by electrolytic lesions or 6-hydroxydopamine treatment (Ungerstedt,197lb). This behavior is correlated with biochemical and histochemical evidence of 25 decreased dopaminergic input to the striatum (Ungerstedt, 19710;Zigmond and Stricker, lg72;Glick et' ·al., 1974; Hynes~ .2.1·, lg75). Similar destruction of noradrenergic fiBers failed to produce these deficits in eating and drinking (Ungerstedt, 19710). d) Bizarre Behaviors -Injections of drugs which increase brain dopaminergic · activity, such as apomorphine, DOPA or amphetamine, result in a compulsive gnawing behavior (Harnack, 1874;Randrup and Udsen, 1963;Ernst, 1965) termed stereotypy. This behavior consists of repetitive sniffing, licking and chewing of the cage. This behavior is also noted after direct injection of apomorphine into the striatum (Smelik and Ernst, 1966) or the olfactory tubercles (McKenzie, 1972).
Various dopaminergic structures h~ve been reported to be the site responsible for stereotypy on the basis of studies involving lesions and local injections. These include the striatum (Smelik and Ernst, 1966;Fog ~al., 1970;; the olfactory tubercles (McKenzie, 1972); and various other portions of the mesolimbic dopamine system (Costall and Naylor, 1973).
It is interesting to note that recent theories concerning the etiology of schizophrenia have suggested that it is t~e result of dopaminergic hyperactivity in the striatum (Klawans i l ·al., 19'72a); the limbic system (Matthysse, 1973); or the cortex (Hokfelt ~ · .2.1., 1974).

26
The blockade of stereotypy has be~n used as a screening procedure ~or detecting anti-p$ychotic actiy t t~ of drugs (Janssen ·et al., 1~651.

--
The hypothesis put forward in this thesis is that druginduced aggression is also tlie result of dopaminergic hyperactivity. Apomorphine and amphetamine generally produce aggression in rats and ~ice, respectively (see Section II. A.2). Apomorphine is believed to directly stimulate dopamine receptors (Anden ~al., 1967), while amphetamine re-lea~es dopamine (Besson~~., 1971). ET-495, another dopamine receptor stimulant (Corrodi ~al., 1972a), also produces aggression in rats (Senault, 1974). Similarly, DOPA administration results in a substantial increase in brain dopamine and elicits fighting, while similar treatments which raise norepinephrine are ineffective in producing fighting (Benkert~ al., 1973a). A more detailed discussion of the relevant data is found in Section V.

4) Effects of Drugs on Dopamine Levels and Turnover
This research is primarily concerned with the actions of apomorphine, haloperidol and morphine. All of these drugs have profound effects on brain dopamine, which are important in the determination of their mechanisms of action. These effects will be examined in this section a} Apomorphine -Apomorphine is believed to d~rectly stimulate dopamine receptors. This conclusion is based on Be-havioral e.xperlments in wh_ich apomorphine-induced stereotypy was unaffected oy· depletion of dopamine oy alph.amethyl-p-tyros ine ('MPT) CErns-t, 1967) or By 6-hydroxydopamine lesions of the suBstantia nigra (Price and Pibiger, 1974). These treatments aoolished the stereotypy produced oy amphetamine (Ernst, 1967;~rice and Pibiger, 1974), which is Believed to release dopamine from pre-synaptic terminals (Besson ·et al., 1971;Chiueh arid Moore, 1973).
In experiments where histofluorescence was measured, apomorphine retarded the reduction in dopamine fluorescence which normally occurs after MPT (Anden ~· al., 1967}. Small doses of apomorphine also depress the firing rate of dopamine cells (Bunney 'et ·al., 1973a), presumably oy a . feedback mechanism. These effects are likewise blocked 28 by haloperidol while the effects on firing rate are unaffected DY' X:eT.
These results ha~e Been interpreted as d~e to a decrease in activity of dopamine neurons as a compensatory mechanismsfor receptor stimulation by the apomorphine (Anden i l al., 1967;Lahti ·et al., 1972;Puri· i l al., 1973). Although apo~orphine can inhibit tyrosine hydroxylase (Goldstein et al., 1970), the doses required ·in vitro are substantially higher than necessary to produce functional changes.
b) Haloperidol -Haloperidol, and related anti-psychotic drugs, are believed to block dopamine receptors. The Behavioral consequences of this effect are the induction of catalepsy, ptosis and inhibition of operant avoidance and brain self-stimulation (Fielding and Lal, 1974).
Neurochemically, haloperidol increases the turnover of dopamine in the striatum (Anden et al., 1970a), the meso-limbic areas (Anden i l al., 1970a) and the cortex (Scatton i l al., 1975). This is accomplished without altering the levels of dopamine in the brain (Anden et al., 1966a). There is also no change in brain dopamine levels after chronic haloperidol treatment (Gyorgy et al., 1969;Puri and Lal, 1974a;Hyttel, 1974). Tolerance to the dopamine turnover enhancing effect of haloperidol has been reported after chronic treat1!lent (Asper · i l al., 1973), but Puri and Lal (1974a) failed to observe this phenomenon, although they did find that morphine dependent rats at two and four hours after drug administration has been reported in the rat (Clouet and Ratner, 1970). Morphine causes a decrease in dopamine levels in the mouse (Takagi and Nakama, 1966;Rethy ~ al., 1971) and an increase in levels in the monkey (Segal i l al., 1972).
The exact relationship between dopamine and narcotic drugs may not be defined by the measurements of dopamine . content of nervous tissue since these reflect only gross changes and may not reveal local changes or alterations i~ the turnover of dopamine.
Morphine also causes an increase in the depletion of dopamine in the rat orain after inhibition of catecholamine synthesis by MPT CGunne · il al., 1969; Puri et ·al., 1973}.
This depletion was determined By histofluorescence to occur in the dopamine terminals of the striatum, the nucleus accumbens and the olfactory tubercles (Gunne · il . §!.1.., 1969).

--
The mechanism by which morphine enhances dopamine turnover is not well established. Morphine does not alter either in vitro (Clouet il . §!.1.., 1973) or in vivo (Cicero il al., 1973) tyr9sine hydroxylase, nor does it alter the affinity of tyrosine hydroxylase for its cofactor, substrate or end product (Cicero et al., 1973). Puri and coworkers (1973} proposed that morphine blocks dopamine receptors and the increased turnover reflects a compensatory increase in neuronal activity analagous to the effect of neuroleptics. Consistent with this hypothesis is the observation that the increased synthesis of new dopamine produced by· ~orphine is associated with an increased release of dopamine into an incubation medium from striatal slices 19"73). Methadone lias also been proposed to act via post synaptic dopamine receptor blockade (Sasame ~al., 1972J. Alternatively, it has been suggested that morphine produces a functional deficiency of dopamine by diverting newly synthesized dopamine from storage sites to sites of cataoolism (Kuschinsky and Hornykiewicz,1972).
· After chronic treatment with morphine, there is a slight increase in the dopamine content of the brain in rats (Sloan~ al., 1963;Johnson et !!.l·• 1974), but dopamine levels remain unchanged after prolonged treatment in dogs (Gunne, 1963) or monkeys (Segal~ al., 1972).
Although most researchers report that a tolerance develops to the effect of morphine on brain dopamine turnover (Gunne et al., 1969;Smith et al., 1972; ---------1973; Puri and Lal, 1974a), it has also been reported that morphine continues to increase the synthesis of dopamine aft er chronic treatment (Johnson~ al., 1974}. This effect may be associated with the increase in tyrosine hydroxylase activity which reportedly occurs during chronic morphine treatment (~eis ~al., 1970). However, decreased tyrosine hydroxylase activity has also been reported in rats implanted w· ith morphine pellets (Branchey ~G· 1974).
During morphine withdrawal, brain dopamine was deereased in dogs, 72 hours after w~thdrawal was initiated (Gunne, 1963). In rats, Brain dopamine levels appear to remain unchanged during wit h drawal (Maynert and Klingman, 1962;Puri and Lal, 1974a), although a decrease in levels, relative to inflated levels during addiction, has been reported (Sloan et al., 1963).
--In contrast to these results, which were obtained in animals which were withdrawn by withholding of morphine injections, abstinence precipitated by naloxone administration reportedly increases the level of brain dopamine in rats and mice (Iwamoto i l tl•, 1973).
The turnover of brain dopamine also appears to remain unchanged during morphine withdrawal (Puti and Lal, 1974a) although decreased dopamine turnover has also been reported in rats (Gunne i l tl·, 1969) and mice (Rosenman and Smith, 1972). In addition, caudate tyrosine hydroxylase activity is reportedly elevated at 48 hours of withdrawal in rats (Reis i l al., 1970) and decreased at 36 hours of withdrawal in mice (Marshall and Smith, 1974).

C. DOPAMINE rNTERACTIONS IN TRE CNS
The The most widely studied of dopamine's interactions are with acetylcholine and serotonin.

1) Acetylcholine
The concept of an interaction between dopamine (DA) and acetylcholine (ACh) developed from the clinical observations that Parkinsonism could be relieved either by increasing dopamine activity or by decreasing cholinergic activity (Klawans, 1968). Subsequently, considerable behavioral and neurochemical evidence has been gathered to support the idea of a reciprocal balance between ACh and DA.
Likewise, the amphetamine-induced locomotor stimulation is enhanced by scopolamine (Fibiger ~ al., 1970) and inhibited by tremorine (Mennear, 1965)., while the locomotor stimulation produced by anticholinergics is blocked by reducing the synthesis of dopamine (Thornburg and Moore, 1973b 1974). Similarly, apomorphine, amphetamine or L-DOPA increases the level of ACh in the striatum (Sethy and Van Woert, 1974). The increase in ACh levels produced by apomorphine (Guyenet ~al., 1975) or ET-495 (Ladinsky ~ ..21..·, is blocked by neuroleptics but is not blocked by alphamethyl-p-tyrosine (Ladinsky ~al., 1974) or lesions of the nigro-striatal bundle (Guyenet ~al., 1975), suggesting that the effect is mediated via dopaminergic receptor stimulation. Similarly, neuroleptic drugs increase the release of ACh with a resultant decrease in levels in the striatum Guyenet ~-al., 1975), and increase ACh turnover . Promethazine, a phenothiazine related to chlorpromazine which has little or no DA receptor blocking activity, does not alter ACh levels . This effect is noted in the striatum, but not in other dopaminergic areas , nor is it noted in the hippocampus, which receives a rich cholinergic input but does not contain dopamine terminals (Rommelspacher and Kuhar, 1975). These results are poorly correlated with the direct anticholinergic activity of the neuroleptics . (Rommelspacher and Kuhar, 1975) again suggesting that the effect is mediated by dopamine.
These results have been interpreted to mean that striatal ACh is under an inhibitory influence of DA. However, lesions of the nigro-striatal pathway did not affect striatal ACh levels (Butcher and Butcher, 1974;Guyenet ~ ~., 1975).
Of considerable interest is the additional interaction between ACh and DA in which anticholinergics have been reported to inhibit the increase in DA turnover produced by the neuroleptics (O'Keefe ~al., 1970; Anden, 1972;Bowers and Roth, 1972). It appears that this effect is limited to the neuroleptic induced increase in striatal dopamine turnover, since the increase in turnover in limbic areas is un-affected by anticholinergics (Anden and Bedard, 1971;Anden, 1972).

2) Serotonin
Behavioral evidence supporting an interaction between serotonin and DA has grown in recent years, although the evidence is not as strong as in the case of ACh (previous section).
In general, increasing serotonergic activity opposes the effect of dopaminergic stimulation. Lycke and coworkers (1969) found that depressing the level of serotonin by parachlorophenylalanine (PCPA), an inhibitor of serotonin synthesis (Koe and Weissman, 1966), increased the excitation produced by DOPA and resulted in aggression in their mice. Similarly, PCPA enhances the stimulation produced by amphetamine (Mabry and Campbell, 1973), while blocking serotonin receptors enhanced the stimulation produced by apomorphine (Grabowska and Michaluk, 1974). Breese and coworkers (1974) further investigated this relationship and observed thatpargyline reduced the stimulant effect of amphetamine and that this was correlated with an increase in serotonin levels.
PCPA reversed the inhibition.
Similarly, destruction of serotonin neurons with 5,6-dihydroxytryptamine likewise potentiated the amphetamine response. PCPA was unable to alter the inhibition of amphetamine-induced stimulation produced by 6-hydroxydopamine , suggesting that the serotonin effect is indirect.
Further evidence for an interaction between DA and serotonin has been provided by studies which measured stereotypy produced after DA agonists, but these results have been contradictory.
Decreasing serotonergic activity has been reported to increase the stereotypy produced by either amphetamine or apomorphine (Weiner et~·, 1973;. However, it has also been reported that decreasing serotonergic stimulation has no effect  and even decreases the stereotypy (Costall and Naylor, 1975) induced by these DA agonists.
It has also been reported that the catalepsy induced by neuroleptics is antagonized by PCPA or by . the destruction of raphe neurons (Kostowski ~al., 1972), lending further support to an interaction.
The evidence thus suggests that activation of dopaminergic neurons leads to a compensatory increase in the activity of serotonergic neurons which results in a reduction of the effect of the initial dopamine stimulation (Grabowska and Michaluk, 1974;Cools, 1974).
• 39 D. DOPAMINERGIC SUPERSENSITIVITY Supersensitivity refers to the phenomenon in which the amount of a substance required to ptoduce a given biological response is less than normal. This implies a shift in the dose response curve for an agonist to lower agonist concentrations. Such a shift in the dose response curve is considered the major criterion fo~ supersensitivity (Trendelenburg, 1966). Evidence for the concept of supersensitivity was provided by experiments in the peripheral nervous system.
In early experiments, it was shown that spinal motor neurons became more reactive to a variety of stimuli following transection of the spinal cord or destruction of sensory · nerve fibers (Cannon and Rosenblueth, 1949). Supersensitivity in the periphery is considered to be of two types.
In the first type, termed denervation supersensitivity, physical removal of the pre-junctions! nerve terminals leads to an enhanced effect of the neurotransmitter due to loss of the re-uptake function of the nerve terminal (Langer and Trendelenburg, 1966 Ln the ~entral nervous system, the term supersensitivity is l.aas:e:Ly a _ p-p-1.i_ ed to any procedure which results in an e.xag_ge rateci response to an agonis t. This has been studied, in dapa:nrine:rgLc brain areas, after eithe-r the destruction of neurons whLch supposedly synapse with dopamine receptors or after the ~dministration of drugs which decrease the availabil.ity of doµamine for its receptors. In both cases, an enhanced r~s-p-orrse to stimulation by dopamine and related ~ganists i~ p-resumed to occur as a compensatory adaptation to the l.~c:k of ~timulation as a result of the effect of trhe 1.es:i.an or t -h e drug • It should be noted that supersensitivity has been used to explain the development of toler-ance to and physical dependence on drugs (e. g. Jaffe and Sharpless, 1968;Collier, 1966;Sharpless and Jaffe, 1969;Puri and Lal, 1973a), since withdrawal phenomena seem to represent rebound effects which are opposite in character to those produced by the drug itself. The characteristics of the two methods of producing supersensitivity in the CNS are described below.

1) Lesion Induced Supersensitivity
Intracerebral injections of 6-hydroxydopamine produce a degeneration of catecholamine containing nerve terminals in the CNS (Breese and Traylor, 1970;Uretsky and Iversen, 1970). Following this treatment, rats exhibit a marked increase in locomotor activity in response to L-DOPA (Uretsky and Schoenfeld, 1971) or apomorphine (Schoenfeld and Uretsky, 1972) as well as enhanced apomorphine-induced stereotypy (Iversen and Creese, 1975) and ET-495 induced hypothermia (Reid, 1975). Similarly, destruction of the substantia nigra results in enhancement of apomorphineinduced stereotypy, while the stereotypy produced by amphetamine is blocked (Iversen and Creese, 1975).
Unilateral destruction of dopaminergic nigro-striatal fibers or terminals provides an interesting model for the study of dopaminergic supersensitivity (Ungerstedt, 197la;Von Voightlander and Moore, 1973c). After such lesions, animals circle away from the side of the lesion in response to apomorphine and towards the side of the lesion in response to amphetamine, presumably because apomorphine has more of an effect on the lesioned (supersensitive) side. An injection of 6-hydroxydopamine into the striatum, which induces this behavior, markedly decreases the concentration of dopamine but has little effect on norepinephrine or serotonin (Von Voightland~r and Moore, 1973c).
In addition, the dose response curve for eliciting the turning by apomorphine or L-DOPA gradually shifts to the left over the course of 30 days . Similarly, the rate of apomorphine-induced circling is well correlated with the reduction of forebrain dopamine .
Following lesions of the substantia nigra, an enhancement of striatal dopamine-stimulated adenylate cyclase has been reported (Mishra et .21.·, 1974) which would provide neurochemical substantiation for the supersensitivity.
Furthermore, caffeine and theophylline enhanced the turning response to apomorphine in unilaterally lesioned rats, but failed to alter turning by themselves or that produced by amphetamine (Fuxe and Ungerstedt, 1974) which was interpreted as due to a preferential effect of these compounds on cyclic AMP formed in the lesioned (supersensitive) side.
However, it has also been reported that there is no difference in caudate dopamine-stimulated adenylate cyclase after destruction of pre-synaptic nerve terminals by 6-hydroxydopamine (Von Voightlander g al., 1973). Further evidence for denervation supersensitivity is provided by Fibiger and Grewaal (1974) who found that in rats with unilateral substantia nigra lesions, the effect of apomorphine in increasing striatal acetylcholine levels was greater on the lesioned side compared with the non-lesioned side, although the response to hal0peridol was unaffected by the lesion.
In addition, it is interesting to note that the dopamine cells which do not degenerate after 6-hydroxydopamine treatment have a higher rate of dopamine turnover than cells on the non-lesioned side , which may represent another type of compensation for reduced dopamine stimulation on the lesioned side.

2) Drug Induced Supersensitivity
Changes in behavior during and after chronic administration of drugs has long been recognized. Boyd (1960) noted a tolerance to the depressant effect of chlorpromazine after 30 weeks of treatment without a cross tolerance to pentobarbital. Furthermore, he noted an increase in locomotor activity during "withdrawal" of the drug. Similarly, Moore (1968) reported a tolerance to the depressant effect of alpha-methyl-p-tyrosine (MPT) during chronic feeding of mice, despite the fact that catecholamine levels remained depressed.
Similarly, daily treatment with reserpine led to enhanced locomotor activity when the drug was discontinued which was correlated with an increase in norepinephrine stimulated adenyl cyclase (Williams and Pirch, 1974).
Subsequently, it was noted that exaggerated responses to dopamine agonist drugs could be elicited following chronic treatment with drugs which would normally block their effect in naive rats. The stereotypy produced by amphetamine, apomorphine or methylphenidate was augmented following the cessation of repeated treatment with neuroleptics in rats (Schelkunov, 1967;Tarsy and Baldessarini, 1973), mice (Fjalland and Moller-Nielsen, 1974b) and guinea-pigs (Klawans and Rubovits, 1972). This effect was also noted following chronic MPT, but was not seen following chronic treatment with the non-neuroleptics, pentobarbital or diazepam (Tarsy and Baldessarini, 1973), which do not affect CNS dopamine.
Interestingly, it has been reported that this effect also fails to develop following chronic treatment with the novel neuroleptic, clozapine (Sayers~ al., 1975), a drug which does not block apomorphine stereotypy in naive rats (Stille~~., 1971).
Similarly, amphetamine stimulation of locomotor activity was enhanced following chronic MPT, but the stimulation produced by methylphenidate or pipradol, which is not blocked by MPT in naive animals, was unaffected following chronic MPT (Dominic and Moore, 1969). Concomitant administration of L-DOPA along with MPT, which would bypass the inhibition of catecholamine synthesis and prevents the de-· pletion of dopamine produced by the MPT, blocked the development of supersensitivity (Gudelsky ~al., 1975).
Recently, it was shown that supersensitivity following chronic haloperidol could also be demonstrated neurophysiologically (Yarbrough, 1975), such that a larger proportion of caudate cells were inhibited by dopamine following chronic haloperidol treatment. However, there is apparently no change in striatal dopamine-sensitive adenylate cyclase activity in mice chronically treated with neuroleptics, although behavioral evidence of supersensitivity was present (Von Voightlander ~al., 1975). Puri and Lal (1973a) proposed that a dopaminergic supersensitivity also develops during chronic treatment with narcotics. This suggestion was based on experiments involving aggression during morphine withdrawal in which normally ineffective doses of apomorphine or amphetamine stimulated the aggression (Lal et~., 1971;Puri and Lal, 1973a). Subsequently, it was also shown that the reduction in striatal dopamine turnover produced by apomorphine could be elicited with ordinarily ineffective doses in morphine withdrawn rats (Puri and Lal, 1973b    The rats were observed for one hour during which time, three parameters of aggression were measured as described in Lal (1975). The relay was modified to detect sounds in the range of 2400-4800 Hz.
All three parameters were recorded for six consecutive 10 minute intervals and were summed to arrive at the figures for responses per hour. Each group of rats was treated statistically as one subject.

G. Locomotor Activity
At various time intervals after drug treatment (see During the observation period, the rats were rated for the intensity of stereotypy according to the five point scale devised by Ernst (1967). The ratings were as follows: 0 -No stereotyped movements.
1 -Rats walking around cage, sniffing over grid, occasionally licking the wires and putting t ·heir nose in the grids.
2 -Rats moving around, occasionally biting and gnawing at the wires.
3 -Rats restricting their locomotion to a small area and gnawing intensely at the bottom.
4 -Rats remaining on the same spot for at least 5 minutes while jerkingly gnawing and clinging their teeth around the wires convulsively for longer periods, sometimes interrupted by short intermissions. A two and a half ml sample was transferred to a test tube and assayed spectrofluorometrically (Carlsson and Waldeck, 1958 were made in these cases by Dunn's Procedure (Hollander and Wolfe, 1973). The 20 mg/kg dose reliably produced a substantial aggression and was utilized in further studies. Doses of 5 mg/kg or less were used in studies where it was predicted that the treatment would increase the aggression.

1) Effect of Anti-Dopaminergic Drugs
If the aggression is due to dopaminergic hyperactivity, it would expected that drugs which reduce dopaminergic activity would also reduce the aggression. Drugs from two The neuroleptics, haloperidol and oxyperomide, produced a dose-dependent reduction of the aggression induced by the 20 mg/kg dose of apomorphi~e (Table 3). This effect was noted on all three parameters of aggression measured.
Haloperidol was maximally ~ffective at a dose of 2.5 mg/kg, while the 5 mg/kg dose of oxyperomide abolished the aggression.
Similarly, morphine produced a dose-dependent blockade of the apomorphine-induced aggression, with a dose of 10 mg/kg being maximally effective (Table 4).
The morphine-induced blockade of aggression was prevented by the narcotic antagonist, naloxone, which had no effect of its own. The specificity of this effect is evident from the fact that naloxone had no effect on the reduction in aggression produced by haloperidol (Table 4).
These results confirmed the suspected anti-aggression action of anti-dopaminergic drugs.
In contrast to the effect of naloxone, an anticholinergic drug, dexetimide, partially reversed the anti-aggression action of haloperidol or oxyperomide as illustrated in Table 5.
This effect was produced by doses of dexetimide which apparently had little effect on apomorphine-induced aggression of their own (c.f. Table 9). However, dexetimide was unable to alter the blockade of aggression produced by morphine (   1 Apomorphine (20 mg/Kg) was injected into all groups 2 Morphine was injected 40 minutes before testing, haloperidol was injected 2 hours before testing and naloxone was injected 10 minutes before testing.
3 Number of groups (4 rats/group) tested.  can be seen in Table 6, the combination of an approximately ED50 dose of oxyperomide w~th an ED 50 dose of the cholinergic drug, pilocarpine, resulted in a reduction of aggression which was greater than either drug alone. These results suggest a reciprocal interaction between dopamine and acetylcholine.
If so, cholinergic stimulation should decrease apomorphine-induced aggression.

2) Effect of Drugs Affecting Acetylcholine
Data summarized in Table 7 shows that pilocarpine reduced the aggression induced by the 20 mg/kg dose of apomorphine in a dose-dependent manner, with the dose of 5 mg/kg of pilocarpine being maximally effective. The experiments summarized in Table 8 were designed to test the effect of various anticholinergics on the pilocarpine-induced blockade of aggression. The data is divided into two consecutive 30 minute segments, rather than a one hour observation period as usually presented. This is because it was noted that the centrally active anticholinergics, dexetimide and atropine, when combined with the high dose of apomorphine used in this experiment, resulted in convulsive-like toxic effects towards the middle of the observation period. These convulsions became apparent after the rats had begun to fight. Therefore, the measurements recorded in the first period . were analyzed separately from  (Winer, 1971.) 1 20 mg/Kg of apomorphine injected into all groups.
2 Number of (4 rats per group) tested. groups 3 2.5 mg/Kg 10 minutes before testing.    (Janssen and Niemegeers, 1967). Isopropamide effectively blocked the dlarrhea produced by the pilocarpine at the dose used (2.5 mg/kg) but had no effect on the aggressionblocking property of pilocarpine, indicating that the effect of cholinergic stimulation in inhibiting aggression is centrally mediated.
Since the cholinergic drug, pilocarpine, blocked the aggression produced by apomorphine,it would be expected that an anticholinerg i c ma y increase the aggression. Large doses of dexetimide were combined with a threshold dose (5 mg/kg) and a sub-threshold dose (2.5 mg/kg) of apomorphine (see Figure 2). Dexetimide alone, in any dose including one as high as 20 mg/kg, was unable to elicit aggression in the rats (Table 9)  ~Number of groups (4 rats/group) tested.
phine at a dose of 5 mg/kg (Table 9).
3) Effect of Amphetamine~ rt was predicted that amphetamine, which releases dopamine, would intensify the apomorphine-induced aggression.
Amphetamine, at a dose of 8 mg/kg, did exacerbate the aggression produced by threshold doses of apomorphine ·cs mg/kg), but failed to elicit aggression when combined with sub-threshold doses of apomorphine (Table 10) Clonidine, which stimulates alpha noradrenergic receptors (Anden ~ ~., 1970b) was used to explore the role of norepinephrine. As is illustrated in Table 11, combinations of clonidine and sub-threshold doses of apomorphine resulted in a marked and dose-dependent increase in the aggression.
Doses of clonidine as small as 0.13 mg/kg produced a violent behavior characterized by an excessively large proportion of biting when combined with normally ineffective doses of apomorphine.
The serotonergic influence was measured in two ways.
Some preliminary data is presented in Table 12 in wh~ch

1) 72 Hour Withdrawal Aggression
Morphine withdrawal aggression was studied because it also may have a dopaminergic basis (Puri and Lal, 1973a).
Any similarities between morphine withdrawal aggression and apomorphine-induced aggression would provide further support for the mechanisms proposed.
a} Effect of Dopamine Rats were made dependent on morphine by the three methods described in Section III.c.l. As expected, the Fig. 3--Dose related elicitation of aggression by apomorphine in p-chloroamphetamine (PCA) treated rats. Apomorphine was injected into groups of four rats, three days after PCA (12 mg/kg in a single injection). Points are mean+ S. E. for four groups, except for the 2.5 mg/kg dose of apomorphine which represents data for one group.
Apomorphine or saline was injected five minutes before observations. 80 rats addicted to a terminal dose of 405 mg/kg/day demonstrated considerable aggression 72 hours after their last morphine injection ( It is interesting to note that the aggression peaks at 72 hours of withdrawal, while other signs of withdrawal (e. g. "wet shakes", ptosis, hypothermia), peak at 24-48 hours of withdrawal (Martin i l al., 1963;Gianutsos i l tl·, 1975). Spontaneous locomotor activity was also measured during withdrawal to determine if any pattern was evident.
It can be seen from the data summarized in Table 14 that locomotor activity was significantly lower than in normal rats at 48 hours of withdrawal, when most signs -of withdrawal are at their most intense. The activity was significantly higher than control at 72 hours of withdrawal, when aggression was at its maximum.
24 hours of withdrawal.
The activity was normal at Final daily dose of morphine at which rats were maintained (see methods) .
Total number of days during which morphine injections were given.
3 Number of groups (4 rats per group). by a gradual depletion of striatal dopamine, ~ith the larger lesion producing a more complete depletion (Hynes .!:l. al.", 1975). These results provide further evidence for the dopaminergic basis of morphine withdrawal aggression.
In contrast to the effects in normal rats (c.f. Table   5), dexetimide did not affect the blockade of aggression by ·haloperidol in morphine withdrawn rats (Table 17)  Number of groups (4 rats/group) tested.
2 Apomorphine or saline was injected 5 minutes before testing.
3 Haloperidol was injected 2 hours before testing.  1 Haloperidol was injected two hours before testing while dexetimide and apomorphine were injected 40 minutes and 10 minutes, respectively, before testing.  (Table   19) .
Clonidine produced a similar enhancement of apomorphine-induced aggression, suggesting that noradrenergic activity may be important in the aggression.

2) ELICITED WITHDRAWAL AGGRESSION
Naloxone, a narcotic-antagonist, is capable of precipitating a withdrawal syndrome if administered to dependent rats within a few hours of a narcotic (Wei et al., 1973), -but certain characteristics of the withdrawal differ from spontaneous withdrawal (see . It    (Table 20). When rats treated with naloxone were grouped, they failed to consistently show aggression even when a higher dose of naloxone (8 mg/kg) was used as illustrated in Table 21.
In contrast, when these rats were injected with a small dose of apomorphine (1.25 mg/kg) either alone or in combination with naloxone four hours after their last morp~ine injection, they exhibited considerable aggression (Table 21). Amphetamine likewise produced some aggression, either alone or when combined with naloxone (Table 21). In order to control for a possible direct anti-aggression action of naloxone, the drug was administered to 72 hour morphine-withdrawn rats.
As can be seen from the data in Table 22,   Drugs injected 4 hours after last morphine injection (135mg/ Kg). Apomorphine was injected 5 minutes, while naloxone was injected 10 minutes before observation.
Amphetamine was injected 60 minutes prior to testing.  1 Number of group tested (4 rats/group)~ 2 Apomorphine was injected 5 minutes before testirrg whi:Ia naloxone was injected 10 minutes before testing. Injections were made 72 hours after last morphine injection. 94 3) frotracted Aggres~ion rt is known that some of the signs of narcotic abstin-.
ence can oe detected even weeks or months after witlidrawal.
in rats (Martin and Sloan, 1971) and humans (Himmelsoach, 1941), a phenomenon known as protracted abstinence. lesion of the nigro-stri&tal bundle produced 24 hours before testing completely blocked the aggression. When the effect of the lesion was bypassed by stimulating the postsynaptic receptors with apomorphine, the aggression was reinstated (Table 23). Lesions of the adjacent medial forebrain bundle, which carries noradrenergic and serotonergic fibers, did not significantly affect the aggression. These results suggest that the dopaminergic supersensitivlty which develops during morphine dependence is a long lasting phenomenon.

C. Chronic Haloperidol
If it is assumed that morphine-withdrawal aggression reflects a dopaminergic supersensitivity produced as a result of dopamine receptor Blockade by morphine, then certain similarities should Be noted between chronic morphine and chronic haloperidol, since haloperidol is generally regarded to block dopamine receptors (Carlsson and Lindqvist, 1963;Anden ~al., 1270a). ~or this reason, haloperidol was administered to rats chronically and observations were made after the drug was discontinued.
Following the cessation of chronic treatment with haloperidol, rats exhibited an apparent supersensitivity to the stereotypy-producing actions of apomorphine (Table 24) or amphetamine (Table 25). It is widely believed that stereotypy produced by these drugs is dependent upon dopaminergic stimulation (Ernst, 1967;Snyder~ al., 1970). Doses of apomorphine (1.25 mg/kg) and amphetamine (7.5 mg/kg) were selected which produced a low to moderate degree of stereotypy in normal rats on the basis of pilot experiments.
When these drugs were administered to rats three days after stopping chronic treatment with haloperidol, the stereotypy was significantly increased. When the rats treated chronically with haloperidol were grouped after discontinuation of the drug, no spontaneous aggression was observed (Table   26). However, when these rats received small doses of apomorphine {0.63-2.5 mg/kg), considerable aggression was ~--------~~---~~     3 Housed one per cage for three weeks.
4 Animals injected chronically with haloperidol as described in Methods.
..... 0 observed (Table 26). This indicated that rats who exhibit dopaminergic supersensitivity, exhibit a supersensitivity to the aggression eliciting effects of apomorphine as well, which would be consistent with the explanation proposed in morphine withdrawal. However, in contrast to morphine-withdrawal aggression, there was no spontaneous aggression after chronic haloperidol. The aggression produced by apomorphine was particularly intense seven days after "withdrawal" but had subsided by 18 days (Table 26).
If amphetamine was injected instead of apomorphine, aggression was observ~d infrequently and inconsistently (Table 27), in contrast to the effect seen during morphine withdrawal. There was virtually no aggression even with a dose of amphetamine as high as 16 mg/kg.
Despite the fact that aggression was not elicited by amphetamine in rats after chronic haloperidol, the locomotor stimulation produced by amphetamine was increased. It can be seen in Figure 5 that a normally ineffective dose of amphetamine (0.2 mg/kg) produced a significant stimulation of locomotor activity after chronic treatment with haloperidol. This effect of amphetamine was further analyzed in  2 Number of groups tested (3 rats per group).
3 Amphetamines were injected 30 minutes prior to observations. Apomorphine was injected 5 minutes before testing.
...... Amphetamine was injected, at the indicated doses, to normal rats and to rats who had received their last of a series of haloperidol injections three to six days prior to testing. Activity was measured for one hour, 60 minutes after the amphetamine injection.
Each point is the mean + S. E . for 10 rats.
Asterisks (*) indicate that the dose significantl y increased locomotor activity (p(0.05) compared with the corresponding saline (O mg/kg) control (Dunnet's Test, Winer, 1971 Fig. 6--Effect of amphetamine, apomorphine or alpha-methyl-p-tyrosine (MPT) on locomotor activity in normal and chronically haloperidol treated rats. The rats were chronically treated with haloperidol, then withdrawn for three to five days before locomotor activity testing.
MPT was injected either 4 hours or 4 and 16 hours before testing.
Locomotor activity in haloperidol treated rats was significantly higher than control rats in each comparison (p<,0.01).
MPT significantly reduced locomotor activity in all groups (MPT vs. no MPT, p<0.01). After MPT treatment, amphetamine induced locomotor stimulation was not higher than the control activity of normal rats, but was significantly higher (p(0.01) in haloperidol treated rats, except when these rats were given two doses of MPT (stripped bars). Apomorphine failed to increase locomotor activity in normal rats (p>0.05), but did produce stimulation in the chronic haloperidol group (p< 0.05).  Dexetimide was injected at the indicated doses into normal rats and to rats who had received their last of a series of haloperidol in~ections three to six days before testing.
Activity was measured for one hour, 30 minutes after injection. Asterisks (*) indicate that the dose significantly increased (p"-0.05) locomotor activity compared with the corresponding saline (O mg/kg) cont~ol group (Dunnet's Test, Winer, 1971).
. Pilocarpine was injected at the indicated doses to normal rats and to rats who had received their last of a series of haloperidol injections three to six days before testing.
Activity was measured for one hour, ten minutes after the injection.
Lastly, the dynamics of striatal dopamine were directly measured following termination of chronic treatment with haloperidol, in order to determine if neurochemical substantiation for the dapaminergic supersensitivity could be obtained.
It can be seen from Table 29 that neither the steady state level nor the turnover of striatal dopamine differed in the chronic haloperidol group when compared with naive rats.
However, doses of apomorphine which produced no change in dopamine turnover in normal rats, significantly reduced striatal dopamine turnover when given seven days after the terminal dose of haloperidol (Figure 9), supporting a concept of dopaminergic supersensitivity.  Values are mean and standard error based upon four animals used in each group. There was no difference (p 0.05) between controls and chronic haloperidol groups on either measure (Student's t test).
2 Rats treated chronically with haloperidol and then left drug free for seven days before testing. Fig. 9--Effect of apomorphine on striatal dopamine turnover in normal rats and after chronic haloperidol.
Rats were sacrificed 30 minutes after apomorphine or saline.
The drugs were injected into either normal rats or rats who had received haloperidol for 17 days.
Tests were made seven days after the last haloperidol injection.
The dose of 0.63 mg/kg significantly reduced the striatal dopamine turnover in the chronic haloperidol group (p~0.05) while the dose of 2.5 mg/kg was the minimum effective dose in normal rats.
~oints are mean + S. K. for four rats.
Dunnet's Test was used for statistical evaluations. In addition, there will be a section which will specifically compare haloperidol with morphine.
In brief, the postulated mechanisms behind drug-induced aggression may be summarized as shown in Table 30. Dopamine plays a major role in the initiation and maintenance of the behavior, with acetylcholine and serotonin having an antagonistic effect.
Stimulation by norepinephrine appears to facillitate the aggression, but further experimentation is necessary to clearly establish this effect. In general, apomorphine-induced aggression and morphine-withdrawal aggression appear to operate under the same mechanisms.
During morphine withdrawal, there is spontaneous and amphetamine stimulated aggression in contrast to the behaviors noted after chronic haloperidol, despite the fact that after either treatment there is evidence of dopaminergic supersensitivity. Several possibilities for this apparent contradiction will be discussed. In brief, these ? Evidence is insufficient to make a decision or contradictory.
115 116 explanations are that chronic haloperidol may lead to an inhibition of neurotransmitter release in selected brain areas; the supersensitivity after each drug may be qualitatively or quantitatively different; morphine may directly interfere with compensatory mechanisms to alter the normal balance in brain functio~; or any combination of these effects.

1) Role of Dopamine
Apomorphine induced aggression in a dose-dependent manner. This was reliably established using three different parameters of aggression. It is generally accepted that apomorphine stimulates dopamine (DA) receptors on the basis of biochemical (Anden ~al., 1967), histochemical (Anden ~ ~·, 1967), behavioral (Ernst, 1967) and neuroanatomical (Price and Fibiger, 1974) studies. In addition, large doses of apomorphine have pre-synaptic actions on dopamine neurons, causing both an inhibition of DA uptake and a dose-dependent release (Ferris~ ,!l., 1974).
These mechanisms may be considered to be responsible for the aggression produced by apomorphine, since it was found that che aggression was blocked by neuroleptics, drugs which are believed to block DA receptors with little if any direct effect on norepinephrine (Nyback ~al., 1968;Anden ~al., 1970a). In addition the aggression was shown to be intensified by amphetamine. Previous reports have also shown that the aggression produced by apomorphine plus reserpine is increased by intraventricular dopamine (Patni and Dandiya, 1974).
Evidence for a dopaminergic basis for drug-induced aggression is also supported by several reports in the literature.
For example, the aggression in mice elicited by amphetamine is blocked by neuroleptics while drugs which inhibit the synthesis of norepinephrine (NE), such as FLA-63, or block NE receptors, such as phenoxybenzamine do not have a clear-cut blocking action (Hasselager £.!_al., 1972). Similarly, fighting was observed in rats after treatment with DOPA plus reserpine and a peripheral decarboxylase inhibitor, which produced a substantial elevation of brain DA (Benkert ~al., 1973a). Elevation of brain NE levels with pargyline plus dihydroxyphenylserine failed to produce fighting (Ben-kert£.!_ al., 1973a). In addition, the aggression produced by ET-495 was increased by L-DOPA (Butterworth£.!_ al., 1975).
A detailed search of the literature revealed that drug-induced aggression not associated with elevated DA receptor activity is rarely, if ever, observed. Therefore, it may be concluded that DA plays a primary role in drug-induced aggression.

2) Cholinergic -Dopaminergic Interaction
Cholinergic activity in the CNS produces a variety of effects on aggression, depending on the model studied. For example, local injection of cholinergic compounds into the 118 hypothalamus (Myers, 1964) or amydgala (Allikmets ~ .21.·, 1969) of cats produces a rage response which includes attacks on the experimenter. Similarly, anticholinergic drugs are potent inhibitors of isolation-induced aggression in mice (Janssen~ .21.·, 1960). On the ooher hand, cholinergic drugs suppress the hyperirritability resulting from septa! lesions (Stark and Henderson, 1972), while the anticholinergic, scopolamine, when combined with an inhibitor of mono amine oxidase, res~lts in aggression-like responses in rats (Randrup and Munkvard, 1969).
In the present experiment, isopropamide (Janssen and Niemegeers, 1967) at a dose which .
blocked the pilocarpine-induced diarrhea.
The extensive evidence for a balance between acetylcho-line and dopamine in the CNS (see ·Section II.C.l) suggests an explanation for the apparent paradoxical effects of cholinergic stimulation on aggression. Acetylcholine (ACh) has a widespread distribution in the CNS and it may be possible that activity in certain brain areas enhances, while activity in other brain areas reduces aggression. It is interesting in this regard that . cholinergic drugs injected into the central nucleus of the amygdala suppress the rage response produced by cholinergic stimulation of the hypothalamus (Decsi ~al., 1969). It is possible that the proposed cholinergic inhibitory system directly interacts with a dopaminergic system which initiates the aggressive responding~ Therefore, aggression elicited by DA hyperactivity -as a result of apoorphine injection or morphine withdrawal -is modulated by the activation of a cholinergic inhibitory system in a manner analogous to the control of motor movements proposed by Klawans (1968). Pilocarpine would synergize with this cholinergic system, while dexetimide would antagonize it. It should be remembered from Section II.C.l that many other behaviors produced as a result of DA stimulation are antagonized by ACh, such as stereotypy and locomotor stimulation (Mennear, 1965;Arnfred and Randrup, 1968;Klawans ~ ~., 1970) while the locomotor stimulation produ~ed by anticholinergics is dependent on DA (Thornburg and Moore, 1973b).
However, neurochemical support for this hypothesis is lacking. Dopaminergic drugs reduce the release of ACh in the striatum (Sethy and Van Woert, 1974) while anticholinergics 120 reduce the utilization of DA (O'Keefe ~al., 1970;Haubrich and Goldberg, 1975). It would be expected from the above hypothesis that DA stimulation should facilitate the release of ACh.
It should be pointed out that the neurochemical ananyses were made only in the striatum, which may not be the area responsible for aggression. Interestingly, amphetamine causes an increase in ACh output from the cortex, which is prevented by alpha-methyl-p-tyrosine (Nistri ~al., 1972).
Alternatively, it may be hypothesized that the reduced ACh as a result of DA stimulation is the mechanism which triggers the aggression, or that still other interactions are necessary.
In addition to the effects enumerated above, low doses of anticholinergic drugs were also able to partially reverse the blockade of aggression produced by the neuroleptics but not the blockade produced by morphine. Anticholinergics also reverse the increase in dopamine turnover produced by the neuroleptics (Anden, 1972) as well as some of the anti-psychotic activity of these drugs (Singh and Smith, 1973).

3) Dopaminergic-Serotonergic Interaction
Evidence in favor of a serotonergic system which antagonizes aggression was also provided by this study. It was predicted that amphetamine, which releases dopamine (Besson et al., 1971;Chiueh and Moore, 1973), would intensify apomorphine induced aggression. Although amphetamine did in-crease the aggression occuring after threshold doses of apomorphine, no aggression was elicited by the combination of amphetamine and sub-threshold doses of apomorphine. It should be remembered that release of serotonin by amphetamine reduced the locomotor stimulation produced by the drug .
It was demonstrated that the combination of amphetamine . plus cyproheptadine, a serotonin recep~ tor blocker, and a sub-threshold dose of apomorphine did result in significant aggression, while cyproheptadinc at the dose used had no effect of its own. This suggests that serotonin release antagonizes aggr~ssion. However, cyproheptadine is non-specific in its effect, since it is anticholinergic and antihistaminic in addition to being anti-serotonergic and these other effects would have to be ruled out before making a definite conclusion.
An additional test of this hypothesis utilized p-chloroamphetamine (PCA) which produces a long-lasting depletion of brain serotonin (Sanders-Bush ~ al., 1972). This drug produced a marked reduction in the threshold dose of apomorphine necessary for eliciting aggression. Recent evidence (Yunger ~al., 1974) suggests that large doses of PCA cause a degeneration of substantia nigra cells, so that it may be interpreted that the apomorphine effect is due to a dopaminergic denervation supersensitivity after PCA. Alternatively, it may be hypothesized that PCA causes reduced activity in a serotonin-containing inhibitory system and therefore removes an inhibition of dopaminergic activity. Consistent ·122 with the latter interpretation are studies which show that parachlorophenylalanine (PCPA), an inhibitor of serotonin biosynthesis (Koe and Weissman, 1966), elicits aggression after treatment with DOPA in mice (Lycke ~ ~·, 1969) and rats (Benkert~~·, 1973b). Furthermore, three days after an injection of 20 mg/kg of PCA, there is spontaneous aggression in mice .
Apomorphine causes a transient elevation in serotonin turnover which is antagonized by neuroleptics . Furthermore, activity in the dopaminergic caput nuclei caudati rostromedialis of the caudate nucleus is abolished by activation of serotonergic nuclei in the caudate (Cools, 1974). These results suggest that dopaminergic activity activates a serotonergic system which attenuates the effects of the dopaminergic stimulation that initiated the cycle.
In addition, the disruption of behavior produced by scopolamine is antagonized by elevating brain serotonin with 5-hydroxytryptamine (Swonger and Rech, 1972), while pilocarpine increases brain serotonin turnover (Haubrich and Reid, 1972). These findings suggest a co-operation between ACh and serotonin in the CNS. Since both neurotransmitters appear to reduce drug-induced aggression, such a co-operative relationship would be consistent with the data presented.

4) Effect of Clonidine
Clonidine dose-dependently and reliably elicited aggression when combined with a sub-threshold dose of apomorphine. It was previously reported that a very high dose of clonidine (1 mg/kg) could induce some aggression in rats after apomorphine or ET-495 (Senault, 1974), while a smaller dose (0.2 mg/kg) increased the "ferocity" and motor activity of rats produced by ET-495 while inhibiting the stereotypy (Butterworth et al., 1975). The present results confirmed and extended the subjective impressions gained by these fixed dose experiments.
elf, did not produce aggression in rats.
Clonidine, by its-Clonidine is considered to be a centrally acting noradrenergic stimulant since it decreases the rate of depletion of norepinephrine in the brain after alpha-methyl-ptyrosine (Anden et .2..1·, 1970b). This effect was measured both biochemically and histochemically. However, clonidine in vitro is also reported to inhibit the stimulation induced release of norepinephrine from brain slices (Starke and Mentel, 1973) and block the increase in cyclic AMP formation by norepinephrine (Vetulani ~al., 1975), which suggests that clonidine may have an anti-adrenergic effect.
Similarly, clonidine in low doses decreased the self-stimulation of the medial forebrain bundle, which is usually enhanced by norepinephrine (Vetulani et al., 1975). Clonidine does not affect the turnover of dopamine in the striatum (Anden et al., 1970b;Rochette and Bradet, 1975).
If it is assumed that clonidine is a noradrenergic agonist, it may be argued that dopaminergic activity initiates noradrenergic activity (or vice versa) and that this combination leads to the expression of aggression. Some evidence which would support this notion is as follows: Apomorphine-induced aggression is blocked by large doses of the alpha-adrenergic receptor blocker, phenoxybenzamine (Senault, 1974) or the beta noradrenergic receptor blocker, L-propranolol (Butterworth~ !!l·, 1975). In mice, a very large dose of clonidine induces aggression by itself, but this aggression is blocked by haloperidol and not by p~en-

oxybenzamine.
Neurochemically, apomorphine increases the turnover of norepinephrine, apparently by an indirect action through stimulation of dopamine receptors (Persson and Waldeck, 1970). On the other hand, Geyer and Segal (1974) reported that intracerebral injections of dopamine enhance shock induced aggression, while intracerebral norepinephrine reduces theaggression. Similarly, it was shown in REM-sleep deprived rats that aggression could be elicited by apomorphine, L-DOPA or phenoxybenzamine (Carlini and Lindsey, 1974). These latter results suggest that norepinephrine reduces aggression.
An alternative explanation may be that the clonidine effect is mediated by serotonin. Clonidine, at a dose which does not affect serotonin metabolism on its own (1 mg/Kg), antagonized the increase in the accumulation of the serotonin metabolite, 5-hydroxyindoleacetic acid produced by apomorphine (Maj ~ tl·, 1973).
Thus, it may be argued that, as above, stimulation of dopamine receptors produces a compensatory antagonistic activation of serotonin neurons.
Clonidine, therefore, would exacerbate the aggression induced by apomorphine by preventing the release of serotonin which would normally antagonize the aggression in a manner analogous to the arguement for PCA (above).

B. Morphine Withdrawal Aggression
Aggression during morphine withdrawal has been considered to reflect a dopaminergic supersensitivity in the CNS since it is enhanced by apomorphine (Puri and Lal, 1973a), amphetamine (Thor, 1969;Lal~ 2-1_., 1971) and L-DOPA (Puri and Lal, 1973a) and reduced by haloperidol (Puri and Lal, 1973a It was found to be blocked by morphine, haloperidol or pilocarpine and increased by dexetimide, apomorpnine, amphetamine and clonidine, implying that the same mechanisms responsible for apomorpnine-induced aggression also govern morphine-withdrawal aggression. However, in contrast to these effects in naive rats, anticholinergics failed to antagonize the blockade of aggression produced by neuroleptics in addicted rats. This will be discussed further below.

1) Elicited Aggression
Naloxone precipitated withdrawal in addicted rats when administered within a few hours of the last morphine injection, but the symptamatology had some differences from the withdrawal produced by witholding of morphine injections.
In the naloxone precipitated withdrawal, there was evidence of teeth chattering and lacrimation which are not seen during spontaneous withdrawal and the naloxone treated rats showed more diarrhea and fewer "wet shakes" than spontaneously withdrawn rats (see Gianutsos ~ .21.·, 1975 Naloxone's action could not be attributed to a direct antiaggression action since it failed to affect either the 72 hour withdrawal aggression or the weak elicitation of aggression produced by amphetamine at four hours of withdrawal. Rowever, it is interesting that naloxone causes an increase in brain dopamine levels in morphine dependent rats (Iwamoto ~al., 1973), which is suggestive of decreased transmitter release. Clearly, such an effect would prevent the expression of withdrawal aggression.
In contrast to the effect of naloxone, apomorphine administered as little as four hours after the last morphine injection was capable of eliciting intense aggression in addicted rats without producing other signs of withdrawal.
The aggression was produced either in the presence or absence of naloxone. These results strongly suggest that the agonists reduce the severity of withdrawal, except for aggression which is intensified (Gianutsos, Hynes and Lal, in preperation). Exactly how this may take plac~ is highly speculative, but Marshall and Smith (1974) showed that, in mice, caudate tyrosine hydroxylase activity is depressed at 36 hours of withdrawal, but returns to normal by 72 hours.
Thus, one could predict that aggression emerges as the biphasic . changes in tyrosine hydroxylase activity return towards normal.

2) Protracted Aggression
Following the chronic administration of morphine, spontaneous aggression as well as the supersensitivity to apomor-

C. Chronic Haloperidol
Following prolonged treatment with haloperidol, rats exhibited an enhanced sensitivity to the actions of apomorphine and amphetamine in stereotypy and locomotor activity.
Since these effects are believed to be dependent on DA stimulation (Ernst, 1967;Snyder et al., 1970;Costa et al., 1972;Thornburg and Moore, 1973a), it may be proposed that a dopaminergic supersensitivity develops during chronic haloperidol treatment.
Since haloperidol is believed to block dopamine receptors (Nyback ~ ~., 1968;Anden et al., 1970a;Koe, 1974), this supersensitivity may reflect a compensatory adaptation to the prolonged receptor blockade. This supersensitivity was also demonstrated neurochemically by a shift to . the left in the apomorphine dose-response curve for re-ducing dopamine turnover in the striatum.
Normally inef fective doses of apomorphine were capaole of reducing striatal dopamine turnover after chronic haloperidol treatment.
In addition, there was a marked shift to the left of the apomorphine-induced aggression dose-response curve after repeated injections with haloperidol. However, in constast to the effect of chronic morphine, there was no spontaneous or amphetamine-induced aggression after chronic haloperidol, despite the presence of supersensitive dopamine receptors. It is interesting that amphetamine induces aggression after chronic morphine (Thor, 1969;Lal et al., 1971) or chronic ethanol (Mann and Lal, 1975) both of which are abused drugs, but failed to induce aggression after prolonged treatment with haloperidol, a non-abused drug. The reason for these effects is not clear, but several explanations may be offered.
For example, it may be proposed that haloperidoi stabilizes the pre-synaptic dopamine neuronal membrane (Janssen, 1967 Long-term neuroleptic therapy reportedly results in tardive dyskinesias in humans which is associated with dopaminergic supersensitivity and a cholinergic subsensitivity (Klawans, k973;Gerlach~~., 1974). A similar relationship was demonstrated with drug effects on locomotor activity in rats following chronic haloperidol in the present experiment. The stimulation produced by dopaminergic agonists (amphetamine, apomorphine) was enhanced in these rats. A similar effect was reported in mice (Thornburg and Moore, 1974;Von Voightlander ~al~, 1975 ' is not dependent on catecholamines (Rech and Moore, 1968).
Several anticholinergics are known to block the reuptake of dopamine (Coyle and Snyder, 1969) and, although this has not been studied with dexetimide, it could provide an explanation for the enhanced stimulatory effects since the dopamine which would be prevented from returning to the pre~synaptic neuron for inactivation would react with supersensitive receptors.
However, it has been demonstrated that anticholinergics produce their effects on locomotor stimulation by an anticholinergic rather than an uptake-blocking mechanism (Thornburg ·133 and Moore, 1973c). Furth~rmore, the decreased effectiveness of pilocarpine could not be accounted for by a mechanism which is based on blockade of catecholamine uptake. Since cholinergic drugs antagonize the stimulation produced by dopaminergic drugs (Mennear, 1965)

D. Comparison of Haloperidol and Morphine
Haloperidol and morphine share a number of similar behavioral and biochemical actions related to dopamine.
However, equally striking are the differences between the two drugs. Most notable, of course, is the well recognized ability of morphine to produce physical dependence and reinforcing properties. Recently, Glick and Cox (1975) demonstrated that very small doses of haloperidol would be self administered by the rat, but it is generally recognized that haloperidol is not abused and does not produce physical dependence.
In addition, the behavioral effects of morphine and haloperidol are differentially antagonized by different compounds. Anticholinergics reverse the blockade of aggres-• sion produced by neuroleptics, but not that produced by morphine.
On the other hand, morphine's actions were antagon-· ized by naloxone, but haloperidol's were not. Similar results have been demonstrated for the catalepsy produced by narcotics and neuroleptics (Kuschinsky and Hornykiewicz, 1972;Costall and Naylor, 1974) and for the blockade of mouse jumping induced by DOPA + amphetamine (Colpaert et ~·, 1975) and the blockade of brain self-stimulation (Wauquier et~., 1974; produced by either drug.
This points up an important difference in the dopaminergic/cholinergic balance between these two compounds and suggests that the reduction of dopaminergic activity by morphine may be indirect.
Furthermore, the ability of anticholinergics to reverse the neuroleptic-induced blockade of aggression was lost after chronic treatment.
It was pointed out previously that chronic treatment with either haloperidol or morphine produces measurable dopaminergic supersensitivity, but only treatment with the latter results in spontaneous aggression. One possible explanation has already been offered (see previous section).
Alternatively, there may be differences in the supersensitivity produced by the two drugs. Kuschinsky (1975) observed that morphine dependent rats failed to show an enhanced sensitivity to apomorphine-induced stereotypy, while a dopaminergic supersensitivity was demonstrated neurochemically.
Some unpublished, preliminary observations in our laboratory would be consistent with these result5. However, following chronic haloperidol, enhanced apomorphine-induced stereotypy is readily observed (this thesis and Schelkunov, 1967;Tarsy and Baldessarini, 1973).
Furthermore, the supersensitivity produced by morphine appears to be longer lasting than the comparable supersensitivity produced by haloperidol (N.B. apomorphine-induced aggression appeared to fade 18 days after stopping haloperidol, while the morphine effect lasted for at least 30 days of withdrawal). This suggests that morphine and haloperidol may produce a qualitatively different supersensitivity, ·perhaps by interacting with different types or groups of receptors.
It should be remembered that after chronic haloperidol, a subsensitivity to cholinergic stimulation was noted. increase the synthesis and release of DA in order to override the effect of the neuroleptic or other drug. There is ample evidence that this occurs after neuroleptics Carlsson and Lindqvist, 1963). Since this method is inefficient over the long-term, the post-synaptic neuron could also compensate by becoming supersensitive to stimulation by DA.
In the meantime, the cholinergic sensitive receptors could also compensate for the unusally large number of messages reaching them by becoming less sensitive to cholinergic stimulation. By these adaptations, a new homeostatic balance is developed. · (Incidentally, if one was to speculate that an enhanced release of ACh from DA sensitive sites was responsible for the anti-aggression action of the neuroleptics, it would explain the antagonism by anticholinergics of the neuroleptic effect.) Although this hypothesis was presented in ·terms of DA and ACh, it would work equally well for DA and serotonin (i.e., DA receptor blockade would reduce the need for serotonin to antagonize DA and so the serotonin neuron would decrease its activity. In the long run, one would predict that the serotonin receptors ~~would become supersensitive due to a lack of stimulation and this would contribute to the lack. of aggression during "withdrawal" of haloperidol).
Morphine, on the other hand, may be thought of as interfering with and modifying the normal compensatory proceses, so that a more drastic or entirely different mechanism for adaptation would be called into play. There are some differences in the effects of morphine and haloperidol on ACh.
It is believed, for example, that morphine blocks the release of ACh (Beleslin and Polak, 1965), althpugh this effeet has been questioned (Mullin and Phillis, 1974). A resulting cholinergic supersensitivity has been proposed as being responsible for narcotic withdrawal (~aton, 1969). Although it is known that ACh utilization is inr.reased durin g withdra~al of narcotics (Domino and Wilson. 1973), this effect does not appear to follow the temporal pattern of the abstinence syndrome (Mullin and Phillis, 1974). It has been noted that neuroleptics accelerate the turnover of ACh in the striatum, but do not affect cortical ACh turnover , while morphine decreases the turnover of ACh in the cortex, with no effect on the stria tum . This has been attributed to a combined direct effect on morphine on ACh neurons and an indirect effect mediated by DA . Furthermore, Yarbrough (1974) was unable to demonstrate an increased sensitivity of cortical neurons to ACh when measured six hours after a terminal injection of morphine, but the efficacy of atropine in blocking the effects of ACh was reduced (another apparent distinction between morphine and haloperidol, since the behavioral effects of anticholinergics were increased after chronic haloperidol).
In spite of the seeming cholinergic supersensitivity, there are some inconsistencies. The ratio of the effective anti-aggression dose of pilocarpine to haloperidol in blocking apomorphine-induced aggression in naive rats was 2:1 (5 mg/kg:2.5 mg/kg). The ratio of these drugs in blocking morphine withdrawal aggression was 16:1 (10 mg/kg:0.64 mg/kg).
Thus, despite the fact that there is supposed to be a dopa- In addition, morphine is known to increase the turnover of serotonin in the brain (Way ~ al., 1968;Yarbrough ~ ~·, 1971;Haubrich and Blake, 1973). There is apparently no tolerance to this effect when rats are made dependent by morphine injections (Bowers and Kleber, 1971). Administration of PCPA prevents the development of morphine tolerance and physical dependence (Way~ al., 1968). If one proposes that the increased turnover of serotonin is a consequence of in-  However, a small dose of apomorphine was able to elicit aggression when given as little as four hours after the last morphine injection. These results further provide evidence that chronic morphine administration results in dopaminergic supersensitivity and that this is reflected in the withdrawal aggression, but ~uggests that mere removal of morphine from its "receptors" by naloxone is insufficient to elicit the aggression.

5.
Morphine withdrawal aggression was noted even 30 days after withdrawal was initiated and was enhanced by small doses apomorphine and blocked by nigro-striatal lesions, suggesting that the supersensitivity produced during narcotic dependence is a long-lasting phenomenon.

6.
Morphine withdrawal aggression pharmacologically resembled apomorphine-induced aggression since both were enhanced by dexetimide, amphetamine and clonidine and both were reduced by haloperidol, morphine and pilocarpine. However, the blockade of aggression produced by haloperidol was reversed by anticholinergics in naive rats, but not in morphine-withdrawnrats, suggesting that an alteration in a cholinergic inhibitory system or in the balance between dopaminergic and cholinergic activity develops during the course of narcotic addiction.

7.
Following discontinuation of long-term treatment with haloperidol, enhanced apomorphine or amphetamine induced stereotypy or locomotor stimulation was observed, leading to the conclusion that this treatment induces supersensitivity of dopamine receptors. This was further supported by neurochemical evidence which showed that the dose response curve for apomorphine-induced reduction in striatal dopamine turnover was shifted to the left after chronic treatment with haloperidol was discontinued.

8.
The dose of apomorphine required to elicit aggression was also markedly reduced after cessation of chronic haloperidol injections, supporting the hypothesis of dopaminergic supersensitivi~y as the mechanism for morphine withdrawal aggression.
In contrast to the effect of chronic morphine, there was no spontaneous o~ amphetamine-induced aggression after chronic haloperidol.
These results suggest either that other mechanisms besides dopaminergic supersensitivity are necessary for morphine withdrawal aggression; or that the supersensitivity produced after these two treatments are dissimilar; or that after chronic haloperidol treatment there is interference with the process of transmitter release; or that there are differences in the balance between dopamine and other transmitters after treatment with the two drugs; or any combination of these factors.

9.
Following chronic haloperidol, the stimulation of locomotor activity produced by dexetimide was increased and the locomotor depression produced by pilocarpine was decreased, while pentobarbital wasequi-effective in naive and chronically haloperidol treated rats. These results suggested that a cholinergic subsensitivity develops during chronic haloperidol treatment, perhaps as a compensatory mechanism.