Neurobiological mechanisms controlling aggression: Preclinical developments for pharmacotherapeutic interventions

Neuroscienceand BiobehavioralReviews,Vol. 18, No. 1, pp. 97-110, 1994 Copyright©1994ElsevierScienceLtd Printedin the USA.All rightsreserved 0149-7634/94$6.00 + .00

Pergamon

Neurobiological Mechanisms Controlling Aggression: Preclinical Developments for Pharmacotherapeutic Interventions K L A U S A. M I C Z E K , 1 E L I S E W E E R T S , M A R G A R E T H A N E Y A N D J E N N I F E R T I D E Y

Tufts University, Medford, M A 02155, USA MICZEK, K. A., E. WEERTS, M. HANEY AND J. TIDEY. Neurobiologicalmechanisms controlling aggression:Preclinicaldevelopmentsforpharmacotherapeutic interventions. NEUROSCI BIOBEHAV REV 18(1)97-110, 1994.-Current pharmacotherapentic approaches to the management of violent and aggressive behavior rely mostly on agents that act as receptor agonists or antagonists at subtypes of brain dopaminergic, GABAergic, and scrotonergic receptors. Ethological experimental studies in animals have shown that drugs may modulate aggression by inhibiting motor activity, by distorting aggression-provoking or -inhibiting signals, by fragmenting behavioral sequencesor temporal patterning, or by increasing the rate and intensity of aggressive acts. Evidence from animal studies points to large changes in sdected brain dopamine, serotonin, and GABA systems during and following aggressiveand defensive behavior. However, the specificityof drugs that are currently used to control aggressive behavior through their action as agonists or antagonists at subtypes of dopamine, serotonin or GABA receptors continues to be of concern. Similar to the effects of widely used traditional neuroleptics that nonselectively antagonize dopamine receptors, the range of behaviors which is suppressed by either DI or D2 receptor antagonists is pervasive. At present, systemic administration of dopamine receptor antagonists in animal preparations does not target aggression-specificmechanisms. The GABA^/Benzodiazepine/Chloride ionophore receptor complex is implicated in the aggression-heighteningeffects of alcohol and benzodiazepines. Although early reports focused on the "taming" effects of benzodiazepine anxiolytics, low doses may enhance aggression in both animals and humans. Benzodiazcpine antagonists block heightened aggression after low doses of alcohol or benzodiazepines. Agonists at certain 5-HT] receptor subtypes such as eltoprazine are potently effective in reducing aggressive behavior of males and females of various animal species under conditions that promote charging offensive-type aggression, without adversely affecting nonaggressive components of the behavioral repertoire. However, recent reports indicate that eltoprazine and related compounds may potentiate anxiety reactions in rodents, and question the behavioral specificity of these substances. Opioid receptor antagonists modulate primarily physiological and behavioral responses of defense and submission. Defeated animals show tolerance to opiate analgesia and withdrawal responses upon challenge with opioid receptor antagonists. Defensive and submissivevocalizations are potently blocked by opioid peptldes. Substances that target specific receptor subtypes at serotonergic, GABAergic and opioidergic synapses are most promising for the selective modification of aggressive, defensive and submissive behavior patterns. Aggression Alcohol Antipsychotic Benzodiazepinereceptor Defeat Defense Dopamine receptor antagonist 5-HT 5-HT receptor agonist 5-HT receptor antagonist Motor activity Opiates Opioid Schedules of reinforcement Tolerance

INTRODUCTION

Intruder

sion studies with a focus on receptor subtypes and their subunits (52). Current research strategies attempt to delineate the relative role of receptor subpopulations in different types of aggressive behavior; therapeutic agents need to be developed that target as selectively as possible the most pertinent receptor populations and their subunits. This objective is preferable to simplistic "receptorology", finking complex changes in aggressive behavior patterns to the functional state of a single type of receptor. The most important shift in the behavioral analysis of aggressive behavior in animal models during the last dozen years has been toward an ethological approach (Fig. 1; 18,50,61). In the sociological and psychological tradition, aggression is

MOST currently used pharmacotherapeutic approaches for the management of pathologically aggressive and violent behaviors rely on drugs that act as receptor agonists or antagonists at subtypes of brain dopaminergic, serotonergic,GABAergic, and opioid peptidergic receptors. During the last decade, the focus of preclinical as well as clinical research on aggressive behavior has shifted from synthetic and metabolic processes to pre and postsynaptic receptors for various amines, peptides and steroids as potential sites for intervention (Fig. 1). The most promising new pharmacological evidence originates from basic preclinical as well as applied clinical aggres-

Requests for reprints should be addressed to K. A. Miczek, Research Building, Tufts University, 490 Boston Ave., Medford, MA 02155. 97

98

MICZEK, WEERTS, HANEY AND TIDEY

Neuropharmacology

Aggression Research

Psychotropic Drugs Antispychotics Antidepressants Anxiolytics

Monoamines Norepinephrine Dopamine Serotonin

Receptor Research ! Acetylcholine Norepinepherine J Dopamine, Opiate Serotonin, GABA Benzodiazepine i

Neuropeptides Opioid Vasopressin CCK

ExperimentalI Psychology , : IsolationInduced tAggreseio~,,~ Pain-Elicited

Ethology Predatory Aggression Social

I Brain

Stimulation/Evoked

=Aggression ExtinctionInduced Aggression

r

ResidentIntruder

Defense and Predatory Aggression

Female Aggression

I I

FIG. 1. Major developments during the past four decades that influenced preclinical psychopharmacology research on aggression. The historical roots for neuropharmacology, experimental psychology, and ethology can be traced back considerably further.

viewed as antisocial behavior resulting primarily from exposure to aversive environmental events. Traditional laboratory models of animal aggression have employed pervasive manipulations such as prolonged isolated housing or more discrete stimulation such as the application of painful electric shocks. Otherwise, placid laboratory rats react to these types of events mostly with defensive acts and postures, often as behavioral fragments removed from the appropriate social context. However, it is not the reactive, but the active, explosive, charging form of aggressive behavior that is of social and public health concern. The production of androgenic steroids directly determines aggressive behavior in reptilian and avian phyla, while social and experiential factors modulate androgenic influence on aggressive behavior in lower mammals. In primates, the relationship between androgens and aggressive behavior is very complex and indirect. It is not surprising that synthetic steroids such as cyproterone acetate or methoxyprogesterone are unreliable and inefficient means of decreasing excessive levels of aggressive behavior in animals and humans (8). Ethological studies of aggressive behavior highlight the ontogenetic and phylogenetic adaptiveness of aggressive behavior (2,33). Establishment and maintenance of a territory, group formation and enforcement of social hierarchies, and defense of the young in the lactating female are some of the critical functions of aggressive acts, postures, and displays. This framework delineates the range of species-typical aggressive, defensive, submissive, and flight behavior, and identifies behavior in excess of the species-typical repertoire.

ETHOLOGICAL APPROACH: BEHAVIORAL METHODOLOGY A formidable challenge in the study of aggressive behavior is the adequate quantification of the behavior. During the past decade, the traditional rating scales or coarse overall incidence

measures (e.g., aggressive vs. nonaggressive) have been superseded by microprocessor-aided video- and audio-analysis of aggressive interactions (Fig. 2; e.g., 15,29,30,59). These methodological developments are significant because they (a) afford a higher resolution of the very rapid behavioral events in a social confrontation; Co) provide information on the moment-to-moment temporal pattern as well as the sequential structure of the agonistic interactions; and (c) quantitatively assess the respective agonistic behavior of each opponent (i.e., the interactive quality of this behavior). These features of the behavioral methodology begin to match the complexity of agonistic interactions that include rapid bursts of intensive behavior composed of characteristic acts, postures and displays (e.g., 49,52,59,67). In a confrontation between a resident rat with an intruder, agonistic acts and postures occur in a highly predictable sequence. The data in Fig. 3 are based on more than 100,000 transitions from one behavioral act to the next during about 500 resident-intruder confrontations (67). A lag sequential analysis (99) reveals that the probability of the sequence pursuit -* threat -~ attack ~ aggressive posture is about 3-4 times higher than chance. This highly likely sequence of behaviorai acts and postures by a resident rat is synchronized with a high-probabifity pattern of defensive, submissive and flight reactions on part of the intruder. Similar to bursts of neuronal firing, these agonistic acts and postures are emitted in a burst-fike temporal organiTation (59,67). In ca. 500 confrontations between resident and intruder rats, more than 90% of all aggressive acts followed each other within 7 s, whereas the remaining acts were separated by longer gaps (Fig. 4). A log-survivor analysis (21) identifies a criterion for the intervals separating consecutive aggressive acts that are part of a burst from the longer intervals that represent gaps between bursts (Fig. 4b). This pattern of longer gaps separating bursts with many events (most fre-

DRUGS AND NEURAL MECHANISMS OF AGGRESSION

99

VCR

~

~

~

Ti.me Code Generator

.....

Microphone

/U ~

Camera

SAS BMDP TemporalAnalysis SequenceAnalysis

_

ObserverKeypads I Telemetry

Graphics

I

I

Event Record

Histogram

Log Survivor Plot

IO ^ u t o g r = ~

IOOO C2H~gT.2~6 - Oh. 5

9 Uprtgh¢ 17 ^ L I o g r o o m r~ I1_1 It

I--'-I

--

7 Tall

IO

6 ACtack Bites 5 Sideways Threac

fl

it

Ratcte

hn

3 Rearing 2 Croomlng

[

2

3

~*

5

b

.

.

7

g

. 9

. I0

.

.

II 12

.

.

.

13 14 15

.

I Pursui~

16

20

Time (seconds)

Hi 40

tO

I[ !

~

60

• eO

H I00

5

20

40

60

5-sec Intervals

FIG. 2. Schematic representation of a system for detailed recording and analysis of behavioral events, as used in ethological studies of aggressive behavior.

Lag Sequential Analysis

PURSUIT - THREAT

~.

sol===_

m

__

THREAT - BITE

~

n=94 BITE - AGGRESSIVEPOSTURE

2~

~6 o

-5-4--3-2-1

1 2 3 4 5

o

-5-4-3-2-1

1 2 3 4 5

o

-5-4-3-2-I

1 2 3 4 5

Event Lag

FIG. 3. Lag sequential analysis of consecutive behavioral acts and postures during resident-intruder confrontations. The transition probability between pursuit, sideways threat, attack bite and aggressive posture as the first, second, third, fourth or fifth event following (+ lag) or prvceding ( - lag) a criterion event is calculated on the basis of all possible transitions. The random transition probability varies with the baserate of each behavioral act and is represented by a vertical line surrounded by 95% confidence intervals.

100

MICZEK, WEERTS, HANEY AND TIDEY A. Time Line

I

Ill

I

I lll n l i

rain

4

B. Criterion for Burst Length

C. Burst Composition

Log Survivor

20,000

E

40

o

g

"E

tJ ,@ f13

1,000

E

@ t~

100

@

I0

E @ []

@ ~n

"

I I 6.6 eec

2

3

4

Intervals (min)

~ 5

@ n

30

20 10 0

single 2 - )

12-) 22-) 32-)

42+

Elements per Buret

FIG. 4. Temporal analysis of bursts of aggressive behavior. (A) An event record depicting each aggressive behavior as a deflection from the time line during a selected encounter between a resident rat and an intruder. (B) The number of aggressive responses in 391 5-min encounters were plotted according to the size of the interval between each consecutive aggressive response. A "break point" was estimated at the intersection between two regression lines, one fitting the longer intervals and the other fitting the shorter intervals best. The burst criterion was estimated at 6.6. s. Fully 88% of aggressive responses followed each other in short intervals and the remaining aggressive acts were separated by "gaps" between bursts. (C) The percentage of aggressive behaviors meeting the burst criterion according to number of elements within a burst.

quently, 2-12; Fig. 4c) in short succession characterizes many biological activities, including aggressive behavior in rodents. Prerequisite to a productive and adequate neurobiological inquiry into the mechanisms of aggressive behavior is a comprehensive, unambiguous description of the salient behavioral acts, postures, movements, gestures, and displays, encompassing the sensory and motor apparatus. In addition to the ethogram of the behavioral repertoire, the formal characteristics of the initiation, execution, and termination of each behavioral act in space and time need to be appropriately assessed. Identification of the proximal and distal antecedents and consequences of aggressive acts represents the initial step in a causative analysis which may then continue toward a neurobiological level. This ethologically based behavior-analytic framework provides a more cogent basis of determining the level at which pharmacological interventions may achieve alterations in aggressive behavior: by inhibiting motor activity, by distorting potentially aggression-provoking or -inhibiting signals, by fragmenting behavioral sequences, by shortening latencies to initiate aggressive acts, by lengthening aggressive bursts and causing failures to terminate, or by increasing the rate and intensity of aggressive acts. BRAIN

DOPAMINE

AND

NEUROLEPTICS

The most frequently used therapeutic agents in the management of violent patients are neuroleptics (55). The primary mechanism of action for typical neuroleptic or antipsychotic drugs is antagonism of dopamine transmission at dopaminergic receptors of the D2 subtype (11,85). Dopamine receptors may influence aggressive and violent behavior at several levels: (a) by modulating motor capacities via neostriatal mecha-

nisms; (b) by altering aggression as a biologically reinforcing event, possibly via the mesolimbic or mesocortical systems; (c) by disrupting patterning or organizing mechanisms for behavioral repertoires; and (d) by regulating rhythms of behavioral activation and quiescence. Before considering the actions of drugs acting on dopaminergic receptors, it is important to recall that the initiation and performance of aggressive and defensive behavior in themselves alter brain dopamine (e.g., 70,77,97). Recent studies in resident mice that attacked and threatened an intruder for the first time illustrate how the execution of aggressive and defensive behavior results in increases in dopamine turnover in n. a c c u m b e n s (Fig. 5; 28). By contrast, dopamine turnover in striatum and n. a c c u m b e n s in aggressive as well as defensive animals with repeated fighting experience does not differ from that of nonaggressive animals (28). These findings suggest adaptive processes at the behavioral level as well as at the level of the dopaminergic mesolimbic system. The suppressive effects of the typical DA-receptor blocking neuroleptics on aggressive behavior in rodents and primates are part of a more pervasive suppression of the individual's active behavior (Fig. 6; 48). Classic substances such as chlorpromazine and haloperidol lack significant behavioral specificity with regard to the decrease in aggressive behavior as is apparent by a comparison of the dose-dependent decline in attack and threat behavior vs. nonaggressive walking and rearing (Fig. 6; e.g., 19,110). Traditionally, the antiaggressive effects of neuroleptics such as phenothiazines, butyrophenones as well as atypical substances such as clozapine are assessed in isolated mice tested for aggressive behavior, and then compared to the results of separate tests for locomotor activity (e.g., 14,46,83). The development of novel antipsychotics that

DRUGS AND NEURAL MECHANISMS OF AGGRESSION

101

do not exert their aggression-decreasing effects at the cost of behavioral sedation has obvious clinical relevance; this requires preclinical methods that simultaneously measure multiple aspects of drug action, preferably within the same subject. Recently, we have developed a protocol in mice that permits the concurrent assessment of schedule-controlled behavior, unconditioned motor activities, and aggressive behavior patterns during a single 60-min session (Fig. 7; 58). This protocol was designed to compare selective drug action in single individuals on behavioral processes with similar motor demands but differing behavioral context. A single experimental session consists of phases in which a multiple schedule of reinforcement maintains temporally patterned, low-rate f'Lxedinterval responding in alternation with high-rate f'Lxed-ratio IJ_ responding. In the middle of the session a 12-13 min phase is inserted in which first the motor activity (i.e. walking, rearing, grooming) and then the aggressive behavior of an animal toward an intruder are assessed during a 5-min social confrontation. Using this protocol, dopaminergic receptor antagonists that selectively target D~ and D2 subtypes were found to potently reduce low- as well as high-rate schedule-controlled be- (o havior, leave nonagonistic unconditioned motor activity unaf- (D fected, and decrease aggressive behavior such as attack bites co and threat responses only moderately in this dose range (Table t1). Again, behavioral specificity for a selective suppression of O aggressive behavior is lacking with the D~ antagonist Sch 23390 and the D2 antagonist raclopride. Notably, these agents suppress schedule-controlled behavior as or more potently than aggressive behavior. This profile of behavioral effects r'~ suggests that these agents may inhibit aggressive behavior by

HALOPERIDOL 30

Aggressive Behavior

* CL_oo ~ 1o T ~ * * ,

**

Motor Behavior 60

40

4

1'

ooming

3 c-

2

° ~

20

O

Wa,.,n0 I

=

1 =

V 0.125

0.25

=

0.5

=

0

1.0

Dose (mg/kg) F 1.2 :1.0

O

FIG. 6. Effects of haloperidol doses, given IP, to male resident mice confronting male intruders for 5 rain in their homecage, on salient measures of aggressive behavior (attack bites, sideways threats, tail rattles) and nonagonistic motor behavior (walking, grooming). Each individual resident mouse was administered all haloperidol doses in a Latin Square sequence, each dose once a week. Significant differences between vehicle control and haloperidol treatments are identified by asterisks. Vertical bars in each data point delineate I + SEM.

1.0 ~

0.4

1

0.3

~

0.2

~

OA

Resident Intruder

>

0.0 0

1

10

Fighting Experience (days) FIG. 5. DOPAC and DA were assayed immediately after an agonistic confrontation in resident and intruder mice with 0, 1, or 10 prior attack or defeat experiences. Each bar represents the mean and standard error of 8-11 mice (Comparison to saline: Dunnett's t-test

*p < 0.05).

"1"

disrupting the temporal or sequential patterning of behavior rather than by inhibiting motor activity. Surprisingly, the selective dopamine agonists tested using this protocol were found to have a similar behavioral profile: the D~ agonist SKF 38393 decreased schedule-controlled responding at doses which were insufficient to suppress aggressive or motor behavior, while the D 2 agonist quinpirole equipotently decreased unconditioned motor and aggressive behavior and schedulecontrolled responding (96). The finding that both agonists and antagonists of the D~ and D2 receptor subtypes nonspecifically reduce aggressive behaviors suggests that dopamine antagonists may be less useful anti-aggressive agents than previously thought; alternatively, a definitive characterization of the role of brain dopamine systems in the maintenance and control of aggressive behavior may require the development of dopamine receptor probes which are both more selective and more efficacious than those which are currently available. In spite of their widespread clinical use in the management of aggressive and violent patients with varying diagnoses, pre-

102

MICZEK, WEERTS, HANEY AND TIDEY EXPERIMENTAL PROTOCOL

Phase 1

/

Cycle 1

¢ Saline

/

F1600 see

&

3

"

4

5

Phase 2

// FIG. 7. Scheme and cumulative records depicting the phases of the experimental protocol: A cumulative record of performance under a multiple FI 10 min FR 30 from a 60-min experimental session is conducted in Phase 1 of the experiment (top). During the second experimental phase (bottom), the mouse is removed from the experimental chamber after completing two FI/FR cycles of the session, is alone in its homecage for 2 rain, confronts an intruder for ca. 5 rain and then returns to the experimental chamber for two more FI/FR cycles.

clinical studies have not established a cogent rationale for a specific role of brain dopamine systems and dopamine receptor subtypes in specific types of animal aggression. The more recent studies with more selectively acting D~ and De receptor antagonists demonstrate these substances to be potent in suppressing a range of active behaviors that include but are not specific to attack and threat responses. BRAIN SEROTONIN AND SERENICS

Probably no other neurotransmitter substance has been linked to aggressive behavior more often than brain serotonin (e.g., 9,43,55,72,101,102,103). The evidence on the role of the serotonin-containing and -releasing neurons as well as the various serotonin receptor subtypes along the neuroaxis in different types of aggressive behavior is equivocal: serotonindepleted rats may start to kill mice, 5-HT injections into lobsters' ventral cord may heighten their aggressive displays, lowimpulse control aggressive alcoholics tend to show low CSF 5-HIAA levels, but CSF 5-HIAA levels o f talapoin monkeys reveal no association with their preceding level o f aggressive

behavior (56). One of the major reasons for the complex pattern of effects of 5-HT manipulations on aggressive behavior are the alterations in serotonin turnover that result after the display of aggressive behavior in itself in the absence of any drug. As an estimate of the turnover of 5-HT, the ratio of 5H I A A / 5 - H T in amygdala is greatly elevated in mice that engage in aggressive behavior toward an intruder for the first time (Fig. 8; 28). When an animal has fighting experience, either as attacking resident or as defeated intruder, no significant alterations in 5-HT turnover are detectable in comparison to nonaggressive controls. The values for 5-HIAA and 5-HT after 10 confrontations between resident and intruder mice did not differ from each other and from those mice without fighting experience (Fig. 8). A major shift in research on serotonin and aggressive behavior has occurred with the advent of more selectively acting receptor agonists and antagonists (62,72). Early evidence with receptor antagonists such as methysergide were quite disappointing because of their relative lack of specificity with regard to 5-HT receptor subtypes and because of their poor behavioral specificity (Table 1; 42,45,54). A recent comparison of methysergide and ketanserin, a more selective 5-HT2 receptor antagonist, shows a systematic dose-dependent behavioral suppression that extends from low- to high-rate schedule-controlled behavior to motor activity and aggressive behavior (27). These results confirm the poor behavioral selectivity of 5-HT2 receptor antagonism with regard to attack and threat behaviors (Table 1). A more promising group of substances are the recently developed piperazine derivatives fluprazine and eltoprazine with 5-HT1A/B receptor blocking properties that appear to exert relatively selective anti-aggressive effects. A series of experiments with mice, rats, pigs, and primates has begun to identify the prototypic aryl-piperazine eltoprazine as remarkably potent in its anti-aggressive effects without inducing behavioral sedation (Table 2; 71, 72). In mice, ca. 0.5 mg/kg eltoprazine, p.o., was effective in reducing aggressive behavior. This effect was seen in mice fighting at relatively low basal levels in a neutral observation cage and in mice selected for a high incidence of aggressive behavior ("isolation-induced aggression"; Table 2; 73). Social or defensive behaviors remained unaffected by eltoprazine, even at doses 50 times higher than the EDs0 for suppressing aggressive behavior. The anti-aggressive effects of eltoprazine persisted after 1 wk of eltoprazine administration in the drinking water (3-10 mg/kg/day), suggesting no development of tolerance to the anti-aggressive effects. Male resident rats attacked an intruder less frequently and for a shorter time when given acute low doses of eltoprazine (1.25 mg/kg and higher); social interactions, exploratory behaviors, and avoidance remained intact. This may be considered a unique behavioral profile of a new drug class that has been referred to as "serenics" (Table 2). Under identical experimental conditions, eltoprazine shares the antiaggressive effectiveness of antipsychotic drugs such as haloperidol, whereas oxazepam and other benzodiazepines actually increase aggressive behaviors at lower doses (47). Neither eltoprazine nor oxazepam appear to induce the strong sedative effects of haloperidol. Eltoprazine (2.5 mg/kg and higher) dose-dependently and specifically decreased aggressive behavior by dominant and subordinate male rat colony members toward an intruder. While dominant rats attack an intruder much more frequently

DRUGS AND NEURAL MECHANISMS OF AGGRESSION

103

TABLE 1 BEHAVIORAL EFFECTS OF DA AND 5-HT ANTAGONISTS1N MICE EDge(mg/kg)* SCH 23390 I. Aggressive Behaviort Attack Bite Sideways Threat II. Motor Activity Rearing Walking III. Schedule Controlled Behavior Fixed Ratio 30 Fixed Interval l0 min

Raclopride Methysergide Ketanserin

0.1 0. l

3.0 3.0

3.0 10.0

10.0 10.0

> 0.3 0.3

10.0 > 10.0

10.0 > 10.0

10.0 > 10.0

1.0 1.0

10.0 > 10.0

10.0 10.0

0.03 0.03

Singly-housed, male CFW mice (n = 4-8) maintained at 80% free-feeding weight. *EDso represents the dose at which behavior was suppressed to 50% of saline control levels. tConditioned behavior was measured for 60 rain postinjection. After ca. 22 rain in the operant chamber, mice were returned to the home cage and motor activity was recorded for 2 rain. A group-housed male was placed into the home cage and aggressive behavior was measured for 5 min. Thereafter, mice completed the remainder of the conditioning session.

and for longer periods of time than subordinate group members, eltoprazine was effective in animals of both high and low social status. By contrast, low chlordiazepoxide doses actually increased aggressive behavior in dominant and subordinate colony members. The pro-aggressive effects of lower doses of benzodiazepine in rats and other animals may be of particular relevance to the clinical problems with these substances in the management of aggressive and violent patients (see 52). A striking form of intense attack leaps and bites can be evoked by electrical stimulation of lateral hypothalamic sites in male or female rats (37,39). The threshold current for evoking attacks, as opposed to locomotion, from rat hypothalamus is greatly increased by eltoprazine and fluprazine (98; Table 2). The more selective 5-HT~A agonist 8-OH-DPAT, the nonspecific agonist qulpazine and chlordiazepoxide were without selective effects in this preparation. In recent experiments, Olivier et al. (71) also report a specific threshold-increasing effect for evoking aggressive behavior and teeth-chattering after administration of DL-propranolol, a/3-adrenergic blocker (5-20 mg/kg IP). Eltoprazine (2 mg/kg and larger) decreased aggressive behavior in female rats, 3 to I 1 days after having given birth, in confrontations with an intruder male. However, this effect was not as aggression-specific as in males, because eltoprazine also affected social, exploratory, and nonaggressive motor activities indicating major shifts in the behavioral repertoire of the lactating female (Table 2). A significant animal husbandry problem in agricultural settings is the injurious aggressive behavior of young pigs during the establishment o f a social hierarchy. Eltoprazine shows potential in this veterinary application by maintaining the percent of fighting in pig pens at significantly reduced levels (Table 2). In contrast to the efficacious action in reducing offensive charging, threatening and attacking behavior patterns, it has been proposed that eltoprazine is largely ineffective in altering defensive responses (Table 2). This is consistent with converging evidence which suggests that attack and defense are medi-

ated by separate neural mechanisms (e.g. 3,87). Recently, reports have emerged that indicate that fluprazine and eltoprazine may potentiate fear and anxiety reactions in rodents (81,36,26). Using the plus-maze paradigm, it was found that both fluprazine and eltoprazine increased entries into closed arms while decreasing or not altering entries into open arms (81). Both fluprazine and eltoprazine have also been reported to increase latencies to enter novel environments and to reduce exploration of novel and brightly lit areas (36,26). These studies suggest that the antiaggressive effects of serenic substances are ancillary to an anxiogenic effect; if so, this may limit their clinical usefulness. Eltoprazine stands out as a potent and behaviorally selective antiaggressive substance. These promising preclinical results need to be expanded into issues of long-term maintenance effects and interactions with anxiolytics and antidepressants. The antiaggressive effect of eltoprazine correlates with the drug's capability to bind to 5-HT~A/B receptors. This may represent an important development in agents with a remarkably specific behavioral profile targeted at aggressive, but not defensive behavior. BRAIN GABA^ RECEPTORS AND ANXIOLYTICS Many substances that act on the GABAA/Benzodiazepine/ Chloride receptor complex have profound effects on aggressive, defensive, social, and submissive behaviors (38,47,74, 76). The "taming" effects of the benzodiazepines were highlighted at the time of their clinical introduction and also found application in veterinary medicine (32,84). However, the "paradoxical" rage phenomenon in certain individuals after benzodiazepine treatment has remained a matter of concern, both in animals and in humans 07,79,82). Benzodiazepine receptor agonists such as diazepam and chlordiazepoxide typically reduce aggression at moderate to high doses, but low doses have been shown to enhance intruder-provoked male resident aggression and maternal aggression in rats (47,51,69). Similarly, compounds with partial or full

104

MICZEK, WEERTS, HANEY AND TIDEY

1.0

[

[

~--l-tr3

,

~

Resident

iii/o "ruwi1 0.4

-

0.2

'>

0.5[

I

0.0

0.4

0.3

*.

2

> 0.1

0.0

0

1

10

Fighting Experience (days) FIG. 8. Bilateral punches (1.2 mm) of the amygdala: DOPAC, DA, 5-HT and 5-HIAA were assayed immediately after an agonistic confrontation in resident and intruder mice with 0, 1 or 10 prior attack or defeat experiences. Each bar represents the mean and standard error of 8-11 mice (comparison to saline: Dunnett's t-test; * p < 0.05). (From Haney et al. 1990.)

inverse agonist actions, such as/~-CCE, FG 7142 and Ro 154513 decrease these measures of aggression (4,69,105). Reductions in aggression due to the action of partial inverse agonists are accompanied by increased escape behaviors; the shift in the repertoire from aggressive behaviors to avoidance behaviors may be related to the anxiogenic properties of the drugs. Increases and decreases in aggressive behavior after administration of full agonists or inverse agonists, respectively, are antagonized by doses of flumazenil that are relatively inactive (3,51). Blockade of the pro- and anti-aggressive effects by the antagonist flumazenil points to a significant role of the benzodiazepine receptor in the modulation of aggressive behavior. A particularly intriguing and controversial possibility concerns the actions of alcohol on the GABAA/Benzodiazepine/ Chloride receptor complex (13,93). Alcohol is by far the most important substance that is associated with many types of violent and aggressive behavior (12,20,100). Whether or not alcohol's aggression-heightening effects are mediated, at least in part, by its potential action on the GABA^/Benzodiazepine/Chloride receptor complex has not been established.

It would be of considerable importance to determine whether or not it is possible to antagonize alcohol's aggressionheightening effects by modulating the GABAA/Benzodiazepine/Chloride receptor complex. The long-standing difficulties of demonstrating robust and reliable aggression-heightening effects of alcohol in animals may be overcome by choosing appropriate experimental conditions in socially housed rodents and in monkeys (6,67). For example, dominant squirrel monkeys exhibit a rich repertoire of agonistic and social behavior that varys with season and is very sensitive to neuropharmacological challenges (57,107, 108). In another experimental preparation, low doses of alcohol increased threats and attacks in a substantial proportion of resident male rats confronted with an intruder into their home cages. High alcohol doses systematically reduced aggression in all animals tested. Reductions in aggression are not a result of disruption of the typical aggressive sequence. A lag sequential analysis reveals that the aggressive sequence of p u r s u i t -* threat -* a t t a c k -* aggressive p o s t u r e remains intact regardless of alcohol dose (67). Once the aggressive sequence is initiated, it generally is completed even under otherwise sedative doses of alcohol. Alcohol's primary actions on aggressive behavior appear to be alteration of (a) the length of intervals separating bursts of aggressive behavior (i.e., gaps); and (b) the number of aggressive elements within an aggressive burst. Low doses of ethanol shorten gaps between bursts and increase the number of elements composing a burst, whereas high doses of alcohol lengthen the gaps between bursts, and reduced the number of aggressive elements within a burst. Very low, acute doses of alcohol reliably increase aggressive displays and threats in dominant, but not subordinate squirrel monkeys as demonstrated in several series of experiments (Fig.9; 106,107,108). Three experimental protocols were implemented to study these aggression-heightening effects of alcohol in both rats and squirrel monkeys: (a) the resident-intruder paradigm described above; (b) dyadic confrontations between two monkeys that are dominant in their respective social groups; and (c) observation of the established social groups where the dominant monkey engages in markedly lower levels of aggressive behavior toward other group members. The relatively neutral receptor antagonist ZK 93426, a /~-carboline derivative, (3 mg/kg) successfully blocked the aggression-heightening effects of acute alcohol doses (0.1, 0.3 g/kg) in A H A rats (not shown) and in dominant male squirrel monkeys during dyadic confrontations (Fig. 10, 106). Behaviorally suppressive effects of alcohol were not blocked by this antagonist. However, at specific doses, ZK 93426 has been reported to possess slight intrinsic anxiogenic actions in the Vogel conflict test and the social interaction test (25,34) and anxiolytic actions in the holeboard test (25,44). At doses that produce anxiogenic-like reductions in social interactions, ZK 93426 reduced aggression in resident male rats and dominant male squirrel monkeys within the social group (105). It appears that in a low dose range ZK 93426 has few detectable intrinsic effects, but is clearly capable in antagonizing the aggression-heightening effects of alcohol. In further experiments, pretreatment with the imidazobenzodiazepine flumazenil potentiated the sedative and aggression-reducing effects of alcohol. Subordinate males were threatened and attacked more frequently when administered flumazenil and alcohol. Pretreatment with flumazenil also increased alcohol-induced motor incoordination such as staggering, even at very low doses (0.1, 0.3 g/kg). These effects may be due to the partial agonist properties of flumazenil. When

DRUGS AND NEURAL MECHANISMS OF AGGRESSION

105

TABLE 2 SUMMARY OF DRUG EFFECTS IN ACJONISTICPARADIGMS Paradigms

Eltoprazine

Fluprazine

Chlordiazepoxide Haloperidol Chlorpromazine TFMPP

Fluvoxamine 80H-DPAT Buspirone

OFFENSE Isolation-induced aggression in mice

`[0.39 po `[0.1 ip

`[1.2 po `[0.7 ip

,[73 po

`[70 po

`[< lpo

`[4.7 po

`[0.2 po

> 20 po `[0.3 ip

(EDs0) Social interaction

< 0.5 po

< 1.25 po

> 15 po

< I po

3 po

< 1.25 po

< 25 po

< Iip

1.0 ip

in mice (LED)

`[sp

`[sp

`[nsp

`[nsp

`[nsp

`[sp

`[nsp

`[nsp

`[nsp

Resident-intruder aggression in rats (LED)

< 1.25 po `[sp

5 po `[sp

1'5-10 po `[> 20 po

0.1 ip `[nsp

-

Iip `[sp

5 ip `[nsp

0.1 sc `[nsp

2 ip `[nsp

Colony aggression in male rats (LED)

< 2.5 po `[sp

< 16 po `[sp

1.5-10 po `[20 po

.

Hypothalamicinduced aggression in rats (LED)

< 2 po `[sp

4po `[sp

ne > 20 po nsp

< 0.5 ip `[nsp

-

0.5 ip `[sp

10 ip `[nsp

ne(> 1.0 ip) --

Maternal aggression in female rats (LED)

2 po `[sp

5 ip `[sp

1"5-10 po `[> 20 po

0.1 ip `[nsp

-

0.5 ip `[sp

20 ip `[nsp

0.1 ip `[nsp

2.0 ip `[

Play fighting in juvenile rats (LED)

< 2 po J,sp

--

1.5-10 ip `[> 20 ip

.

Aggression in pigs

`[2.5 im `[10 im sp

> 30 im

.

>46.4 po

`[2.4 po sp

`[15.4 po nsp

--

ne unchanged strategy

ne unchanged strategy

`[ changed strategy

`[ changed strategy

`[5ip `[20po sp

`[Sip `[32po sp

ne > 20ip ne > 32po

`[0.lip `[8po nsp

ne > 5ip -

ne20ip --

DEFENSE Shock-induced defense in mice (EDso) Defense and flight of intruder PREDATION Mouse killingin rats(LED)

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

`[8.2 po nsp

`[5ip J.>20po nsp

`[1.0ip `[3po sp

`[20ip ne > 50po

t = increase; `[ = decrease; ne = no effect; sp = specific effect; nsp = nonspecific effect; - = not tested. All doses are given in mg/kg. EDs0 = dose that gives 50% suppression of aggression; LED = lowest effective dose. Table abridged from Olivier et al. 1990.

administered alone, flumazenil increased the duration o f foraging and feeding. Again, these results are consistent with intrinsic effects o f flumazenil that suggest partial agonist action. Flumazenil increased exploratory behavior in mice (24, 44) and generalizes to benzodiazepine agonists in rats trained to discriminate flumazenil (10 m g / k g ) from vehicle (16). However, flumazenil seems to possess some anxiogenic actions as well. It decreases social interactions (23) and increases the suppressive effects of aversive stimuli in a conflict situation (90). The partial agonist or partial inverse agonist actions o f flumazenil may be related to dose and sensitivity o f the testing situation. However, in our experiments, flumazenil appears to shift the alcohol dose-response curve to the left as evidenced by increased m o t o r coordination and sedation at low alcohol dosds (106). The "antagonism" o f the pro-aggressive effects o f

alcohol by flumazenil are more likely a potentiation o f the aggression-reducing properties of alcohol. High levels of aggressive behavior such as after alcohol administration exceed the species-typical range for this behavior and may be o f particular clinical relevance. The G A B A A / Benzodiazepine/Chloride receptor complex may be a critical site for novel substances that effectively reduce these "pathological" levels o f aggression. However, the currently available antagonists at the benzodiazepine receptor site possess intrinsic partial agonist and antagonist activities that alter aggressive and social behavior at specific doses. At present there are at least 15 different receptor subunits o f the GABAABenzodiazepine/Chloride receptor with diverse pharmacological sensitivities (104). The development o f newer compounds with more selective actions at these different receptor subunits

106

MICZEK, WEERTS, HANEY AND TIDEY AGGRESSIVE BEHAVIOR A. D o s e - E f f e c t

15

B. Time Course (0.6 q / k g ) 0 Dom i n o n l D Subordinate

,o

~

i

0

O.I

t

0 O.3

i

i

0.6

1.0

Log Ethenol Dose ( g / k g )

20

40

Time Since

60

80

Injection

I00

120

(min)

FIG. 9. (A). The effects of alcohol (0, 0.1, 0.3, 0.6, 1.0 g/kg, PO) on frequency of aggressive displays and threats in dominant (n = 5) and subordinate (n = 6) members of established groups of squirrel monkeys. Vertical lines represent + 1 SEM. B. Time course of effects of 0.6 g/kg alcohol on aggressive threats and displays in dominant male squirrel monkeys (n = 5), as measured in 6 consecutive 20 minute observations. The shaded area represents the control mean + 1 SEM.

the effects of opiates on aggression may be due in part to interactions with dopaminergic systems; this hypothesis is supported by reports that opiates activate mesolimbic dopamine (89) and that morphine withdrawal decreases striatal dopamine release (1) which may lead to receptor upregulation. In preclinical research, marked changes in release of opioid peptides and in the regulation and activation of opioid receptors are seen in animals that undergo important maternal, reproductive, and social events. In a recent series of experiments, opioid peptides were found to potently modulate affective defense in cats (88). Defensive responses evoked by electrical stimulation of the cat hypothalamus or periaqueductal grey area were facilitated by opioid receptor blockade and suppressed by the metenkephalin analog D-Ala2-Met5-Enkephalinamide (DAME; 10,86). In mice and rats, naloxone and naltrexone may increase defensive responses (e.g. 22,78, 80,91), but decrease attack behavior in mice, rats and squirrel monkeys (e.g. 4,74,109). One of the largest changes in opioid peptides and opioid receptor sensitivity to agonist challenges is seen in mice or

Aggression Sent A. In Dyad Test 30

may reveal more selective GABAA-Benzodiazepine/Chloride receptor mediated anti-aggressive actions. OPIATES AND OPIOID PEPTIDES While opioid peptides and their receptors have been demonstrated to be of critical significance in several social circumstances from infancy to adulthood (e.g. 5, 75), major therapeutic applications in pathologically violent individuals have not yet been offered. The recent successful treatment of self-injurious behavior by some mentally retarded or autistic youngsters with opiate antagonists naloxone and naltrexone represents a promising development (e.g. 6,31), but may only be indirectly relevant to the neurobiology of aggressive behavior. Furthermore, self-inflicted mutilation, injury or harm constitute responses that differ in causation, function and underlying neurobiology profoundly from threat and attack as well as defense and flight. Opiate use and opiate withdrawal produces severe physiological and behavioral consequences in humans and other animals. In mice, morphine withdrawal can be induced by the removal of subcutaneously implanted morphine pellets which have been in place for 3 days; the withdrawal syndrome includes disruptions of thermoregulatory and other homeostatic functions, explosive motor behaviors, and a strikingly intensive form of offensive aggression (41,94). Notably, while the effects of morphine withdrawal on explosive jumping, thermoregulation, and motor behaviors are most prominent within the first 24 hours o f pellet removal and return to control levels within 48 hours, aggression levels remain elevated for at least 4 days after pellet removal (94). Morphinewithdrawal aggression can be further potentiated with the administration of amphetamine, cocaine, apomorphine and Ldopa (35); co-administration o f the D1 agonist SKF 38393 and the De agonist qninpirole maintains high levels o f aggression in morphine-withdrawn mice at doses which decrease aggression in opiate-naive animals (95). These reports suggest that

,

,

__[

_[

• EtOH n +ZK93426,

\,

20 ¸

D 10.

D ~

D

(3 mg/kg)

i" i ~

*

,

~rl-------~ C~ coD ¢D

I

1

0

I

I

0.1

I

I

0.3

I

I

0.6

I

I

1.0

I

1.5

B. In Colony

k.

b_

80.

• EtOH A + Flumazenil, (10 rng/kg)

*

i

60. 4-0~

20I

0

I

I

0.1

• ~ A l

I

1

0.3

I

1

0.6

I

1

1.0

I

1.5

Dose (g/kg) FIG. 10. The effects of alcohol alone (solid symbols) and after pretreatment with ZK 93426 or flumazenil (open symbols) on aggressive behavior in dominant and rival male squirrel monkeys. Two test situations were used: the dyad test (top) and in the social group (bottom). Vertical lines represent + I SEM. Asterisks represent significance of p < 0.05 (from Weerts et al., 1993, b.).

DRUGS AND NEURAL MECHANISMS OF AGGRESSION rats that are exposed to the stress of a social confrontation (40,53,63,64,65). For example, a mouse that is defeated for the first time by an aggressive opponent shows large increases in diencephalic and mesencephalic /~-endorphin. However, after 3 or 5 more dally defeat experiences, reduced amounts of /~-endorphin are detectable in these brain regions (92). Fewer 3H-diprenorphine-binding sites are found in mice who exhibit defeat response for the first time in a social confrontation. After 5 or 7 defeat experiences the number of diprenorphine binding actually increases over control values pointing to important regulatory changes in receptor synthesis (92). The reduced amount of/3-endorphin and increased number of diprenorphin-binding sites coincide in time with the development of tolerance to the analgesic effects of opiates due to previous defeat experiences (53,63,65,68). When a mouse or a rat is challenged with a m receptor agonists 5-7 days after being defeated in a social confrontation, a 4-6 fold rightward shift in dose-effect curve indicates tolerance. The long-lasting nature of the tolerance to morphine analgesia after brief defeat experiences is demonstrated in mice and rats that require more than 4 times higher doses of morphine to achieve analgesia 2-3 months after the social confrontations (53,68). After repeated defeat experiences naloxone precipitates withdrawal jumps in mice comparable to those that are seen after the cessation of repeated morphine injections (Fig. 11). Over the course of 6 dally tests, the magnitude of analgesia diminishes in defeat-exposed as well as in morphine-injected mice, and when challenged with naloxone withdrawal jumps are evident after both types of manipulations. Alterations in endogenous opioid peptides and their receptors may be implicated in the naloxone-precipitated withdrawal jumps in defeat-experienced, opiate-naive mice. In a recently developed experimental protocol, the threat of being attacked was sufficient to cause large changes in sensitivity to challenges with opioid receptor agonists. Contrary to earlier observations with mice, short encounters with

107 a resident attacker did not cause "stress'-induced analgesia, but left the base response to pain unaltered in submissive intruders. However, while protected behind a wire screen, the intruders demonstrated an increased sensitivity to the analgesic effects of morphine. Delaying the morphine challenge by 1-7 days produced tolerance to the analgesic effects of morphine which started to diminish 30 days after the single social confrontation (66). Time-dependent regulatory changes at the opioid receptor sites may account for the large and longlasting tolerance to opiate challenges. These findings may be helpful in pointing to potential mechanisms for perplexing clinical observations of large individual differences in response to analgesics. Most intriguing is the functional selectivity of the potentiation and tolerance to the effects of morphine after exposure to social stress; only the analgesic, but not the hypothermic and other behavioral effects of morphine underwent the transition from potentiation to tolerance. The threat of attack in a social confrontation represents a potent challenge to an intruder rat which initiates a cascade of physiological and behavioral events in the short- and longterm. Shifts in dose-effect curves for drugs that interact with opioid receptors point to modulation of endogenous opioid activity by nonpharmacological, behavioral means. CONCLUSIONS

Aggression and submission cause large increases in dopamine, 5-HT, GABA turnover as well as regulatory changes in receptors for opioids, anxiolytics and catecholaminergic drugs. Among the most promising novel therapeutic agents for altering excessive violent and aggressive behavior are those with high selectivity for 5-HT receptor subtypes. The most significant contributions of precfinical aggression research derive from the accurate and sensitive assessment of the specificity with which drugs modulate aggressive behavior. The classic as well as the more recently developed antipsychotic dopamine

TOLERANCE

WITH DRAWAL

I00

I00

80

•

Doily Defeot

•

5 m g / k g / d a y MSO4

80

cn

:3" o =E .5"

1.6 EL

60

60

:3" t3_ "1

40

~

20

40

o

20

r-

c_

3

0

~

0

1

2

3

4

5

6

7

-13 uJ

Defeat MSO4

Days FIG. 11. Development of tolerance in mice subjected to repeated defeat or morphine (left). Intruder mice were subjected to attacks by resident stimulus mice or were given 5 mg/kg morphine IP every day for 7 days. Values are means and + SEM Nalxone (10 mg/kg), administered immediately after the seventh defeat, precipitated jumping in intruder mice during a 20-min observation period (right) (adapted from Miczek and Thompson 1984).

108

MICZEK, WEERTS, HANEY AND TIDEY

receptor antagonists continue to exhibit a disappointing lack o f specificity in decreasing aggressive behavior. The shift to an ethological experimental approach with its emphasis on biologically significant situations and on a detailed temporal and sequential analysis o f the salient acts and postures o f aggression, defense, submission and flight has enhanced the validity of preclinical contributions. This ethological approach should be applied to clinical aggression research, par-

ticularly with the advent o f automated behavior pattern analysis. ACKNOWLEDGEMENTS Preparation of this review and the original research were supported by U.S.P.H.S. research grants AA05122 and DA02632. The excellent assistance and advice of J.T. Sopko, W. Tornatzky, and J. Vivian are gratefully acknowledged.

REFERENCES 1. Acquas, E.; Carboni, E.; Di Chiara, G. Depression of basal dopamine release and sensitization to morphine-induced stimulation in the ventral striatum during abstinence. Society for Neuroscience Abstracts 17:328. 2. Archer, J. The behavioural biology of aggression. Cambridge: Cambridge University Press; 1988. 3. Bandier, R. Brain mechanisms of aggression as revealed by electrical and chemical stimulation: Suggestion of a central role for the midbraln periaqueductal grey region. In: Epstein, A. N.; Morrison, A. R., eds. Progress in psychobiology and physiological psychology. New York: Academic Press, Inc.; 1988:67-154. 4. Beck, C. H. M.; Cooper, S. J. /3-Carboline FG 7142-reduced aggression in male rats: Reversed by the benzodiazepine receptor antagonist, Ro 15-1788. Pharm. Biochem. Beh. 24:1645-1649; 1986. 5. Benton, D.; Brain, P. F. The role of opioid mechanisms in social interaction and attachment. In: Rodgers, R. J.; Cooper, S. J., eds. Endorphins, opiates and behavioural processes. Chichester: John Wiley & Sons; 1988:217-235. 6. Bernstein, G. A.; Hughes, J. R.; Mitchell, J. E.; Thompson, T. Effects of narcotic antagonists on seif-injurious behavior: A single case study. J. Am. Acad. Child Adolesc. Psych. 26:886-889; 1987. 7. Berry, M. S.; Smoothy, R. A critical evaluation of claimed relationships between alcohol intake and aggression in infra-human animals. In: Brain, P. F., ed. Alcohol and aggression. London: Croom Helm; 1986:84-137. 8. Brain, P. F. Hormonal aspects of aggression and violence. In: Violence Report. Washington, DC: National Academy Press; 1991 (in press). 9. Brown, G. L.; Goodwin, F. K.; Ballenger, J. C.; Goyer, P. F.; Major, L. F. Aggression in humans correlates with cerebrospinal fluid amine metabolites. Psych. Res. 1:131-139; 1979. 10. Brutus, M.; Zuabi, S.; Siegel, A. Microinjections of d-Ala2Met5-enkephalinamide placed into nucleus accumbens suppress hypothalamically elicited hissing in the cat. Exp. Nenrol. 104: 55-61; 1989. 11. Carlsson, A. Perspectives on the discovery of central monominergic neurotransmission. Annu. Rev. Neurosci. 10:19-40; 1987. 12. Coid, J. Alcoholism and violence. Drug and alcohol dependence 9:1-13; 1982. 13. Deitrich, R. A.; Dunwiddie, T. V.; Harris, R. A.; Erwin, V. G. Mechanism of action of ethanol: Initial central nervous system actions. Pharmacol. Rev. 41:489-537; 1989. 14. Deiini-Stula, A.; Vassout, A. Differential effects of psychoactive drugs on aggressive responses in mice and rats. In: Sandier, M., ed. Psychopharmacology of aggression. New York: Raven Press; 1979:41-60. 15. DePaulis, A. A microcomputer method for behavioural data acquisition and subsequent analysis. Pharmacol. Biochem. Behay. 19:729-732; 1983. 16. DeVry, J.; Slangen, J. L. Effects of chordiazepoxide training dose on the mixed agonist-antagonist properties of benzodiazepine receptor antagonist Ro 15-1788 in a drug discrimination procedure. Psychopharmacology 88:177-183; 1986. 17. DiMascio, A. The effects of benzodiazepines on aggression: Reduced or increased? Psychopharmacologia 30:95-102; 1973.

18. Dixon, K. Ethopharmacology: A new way to analyse drug effects on behavior. Triangle 21:95-105; 1982. 19. Dubinsky, B.; Robichaud, R. C.; Goldberg, M. E. Effects of (-)D9*trans-tetrahydrocannabinol and its selectivity in several models of aggressive behavior. Pharmacology 9:204-216; 1973. 20. Evans, C. M. Alcohol and violence: Problems relating to methodology, statistics and causation. In: Brain, P. F., ed. Alcohol and aggression. London: Croom Helm; 1986:138-160. 21. Fagen, R. M.; Young, D. Y. Temporal patterns of behaviors: Durations, intervals, latencies, and sequences. In: Colgan, P. W., ed. Quantitative ethology. New York: John Wiley; 1978: 79-114. 22. Fanselow, M. S.; Sigmundi, R. A.; Bolles, R. C. Naloxone pretreatment enhances shock-elicited aggression. Physiol. Psych. 8: 369-371; 1980. 23. File, S. E.; Lister, R. G. Interactions of ethyl-/5-carboline-3carboxylate and Ro 15-1788 with CGS 8216 in an animal model of anxiety. Neurosci. Lett. 39:91-94; 1983. 24. File; S. E.; Peilow, S. Intrinsic actions of the benzodiazepine receptor antagonist Ro 15-1788. Psychopharmacology 88:1-11; 1986. 25. File, S. E.; Pellow, S.; Jensen, L. H. Actions of the/~-carboline ZK 93426 in an animal test of anxiety and the holeboard: Interactions with Ro 15-1788. J. Neural Transm. 65:103-114; 1986. 26. Griebel, G.; Saffroy-Spittler, M.; Misslin, R.; Vogel, E.; Martin, J. R. Serenics fluprazine (DU 27716) and eitoprazine (DU 28853) enhance neophobic and emotional behaviour in mice. Psychopharmacology 102:498-502; 1990. 27. Haney, M.; Miczek, K. A. d-Amphetamine, MDMA and PCP effects on aggressive and conditioned behavior: 5-HT and dopamine antagonists. Society for Neuroscience Abstracts 15:635; 1989. 28. Haney, M.; Noda, K.; Kream, R.; Miczek, K. A. Regional 5-HT and dopamine activity: Sensitivity to amphetamine and aggressive behavior in mice. Aggress. Behav. 16:259-270; 1990. 29. Hausammann, K. F.; Dixon, K. An off-line data acquisitiondevice for measuring ethological elements of behaviour. Physiol. Behav. 28:743-745; 1982. 30. Hendrie, C. A.; Bennett, S. A microcomputer technique for the detailed analysis of animal behavior. Physiol. Behav. 30:233235; 1983. 31. Herman, B. H.; Hammock, M. K.; Arthur-Smith, A.; Egan, J.; Chatoor, I.; Werner, A.; Zelnik, N. Naltrexone decreases self-injurious behavior. Ann. Neurol. 22:550-552; 1987. 32. Heuscheie, W. P. Chlordiazepoxide for calming zoo animals. J. Am. Vet. Med. Assoc. 139:996-998; 1961. 33. Huntingford, F. A.; Turner, A. K. Animal Conflict. London, New York: Chapman and Hall; 1987. 34. Jensen, L. H.; Petersen, E. N.; Braestrup, C.; Honore, T.; Kehr, W.; Stephens, D. N.; Schneider, H.; Seideimann, D.; Schmiechen, R. Evaluation of the/~-carboline ZK 93426 as a benzodiazepine receptor antagonist. Psychopharmacology 83: 249-256; 1984. 35. Kantak, K.; Miczek, K. A.; Social, motor, and autonomic signs of morphine withdrawal: Differential sensitivities to catecholaminergic drugs in mice. Psychopharmacology 96:468-476; 1988.

DRUGS AND NEURAL MECHANISMS OF AGGRESSION 36. Kemble, E. D. Some further ethoexperimental studies of the anti-aggressive drug fluprazine hydrochloride. In: Blanchard, R. J.; Brain, P. F.; Blanchard, D. C.; Parmigiani, S., eds. Ethoexperimental approaches to the study of behavior. Kluwer: Dordrecht; 1989:484--493. 37. Koolhaas, J. M. Hypothalamically induced intraspecific aggressive behaviour in the rat. Exp. Brain Res. 32:365-375; 1978. 38. Krsiak, M. Behavioral changes and aggressivity evoked by drugs in mice. Res. Commun. Chem. Pathol. Pharmacol. 7:237-257; 1974. 39. Kruk, M. R.; Kuiper, P.; Meelis, W. An air pressure operated commutator system for electrical stimulation in a fighting rat. Physiol. Behav. 21:125-127; 1978. 40. Kfilling, P.; Frischknecht, H.-R.; Pasi, A.; Waser, P. G.; Siegfried, B. Social conflict-induced changes in nociception and 13endorphin-like immunoreactivity in pituitary and discrete brain areas of C57BL/6 and DBA/2 mice. Brain Res. 450:237-246; 1988. 41. Lal, H.; O'Brien, J.; Puri, S. Morphine-withdrawal aggression: Sensitization by amphetamines. Psychopharmacologia 22:217223; 1971. 42. Lindgren, T.; Kantak, K. M. Effects of serotonin receptor agonists and antagonists on offensive aggression in mice. Aggress. Behav. 13:87-96; 1987. 43. Linnoila, M.; Virkkunen, M.; Scheinin, M.; Nuutila, A.; Rimon, R.; Goodwin, F. K. Low cerebrospinal fluid 5-hydroxyindoleacetic acid concentration differentiates impulsive from nonimpulsive violent behavior. Life Sci. 33:2609-2614; 1983. 44. Lister, R. G. Partial reversal of ethanol-induced reductions in exploration by two benzodiazepine antagonists (flumazenil and ZK 93426). Brain Res. Bull. 21:765-770; 1988. 45. Maiick, J. B.; Barnett, A. The role of serotonergic pathways in isolation-induced aggression in mice. Pharmacol. Biochem. Behav. 5:55-61; 1976. 46. McMillen, B. A.; DaVanzo, E. A.; Song, A. H.; Scott, S. M.; Rodriguez, M. E. Effects of classical and atypical antipsychotic drugs on isolation-induced aggression in male mice. Eur. J. Pharmacol. 160:149-153; 1989. 47. Miczek, K. A. Intraspecies aggression in rats: Effects of d-amphetamine and chlordiazepoxide. Psychopharmacologia 39:275301; 1974. 48. Miczek, K. A. Differential antagonism of d-amphetamine effects on motor activity and agonistic behavior in mice. Neurosci. Abstr. 7:343; 1981. 49. Miczek, K. A. Ethopharmacology of aggression, defense, and defeat. In: Simmel, E. C.; Hahn, M. E.; Waiters, J. K., eds. Aggressive behavior: Genetic and neural approaches. Hiilsdale: Lawrence Erlbaum Associates; 1983:147-166. 50. Miczek, K. A. Ethological analysis of drug action on aggression and defense. Prog. Neuropsychopharmacol. Biol. Psychiatry 7: 519-524; 1983. 51. Miczek, K. A. Alcohol and aggressive behavior in rats: Interaction with benzodiazepines. Soc. Neurosci. Abstr. 11:1290; 1985. 52. Miczek, K. A. The psychopharmacology of aggression. In: Iversen, L. L.; Iversen, S. D.; Snyder, S. H., eds. Handbook of psychopharmacology, vol. 19: New Dir. Behav. Pharrnacol. New York: Plenum; 1987:183-328. 53. Miczek, K. A. Tolerance to the analgesic, but not discriminative stimulus effects of morphine after brief social defeat in rats. Psychopharmacology (in press). 54. Miczek, K. A.; DeBold, J. F. Hormone-drug interactions and their influence on aggressive behavior. In: Svare, B. B., ed. Hormones and aggressive behavior. New York: Plenum Press; 1983:313-347. 55. Miczek, K. A.; DeBold, J. F.; Haney, M.; Tidey, J.; Vivian, J.; Weerts, E. Neurochemistry and pharmacotherapeutic management of violence and aggression. In: Reis, A.; Roth, J., eds. Violence report. Washington, DC: National Academy Press (in press). 56. Miczek, K. A.; Donat, P. Brain 5-HT systems and inhibition of aggressive behavior. In: Archer, T.; Bevan, P.; Cools, A., eds.

109 Behavioral pharmacology of 5-HT. Hillsdale, New Jersey: Lawfence Eribaum Associates; 1989. 57. Miczek, K. A.; Gold, L. H. d-Amphetamine in squirrel monkeys of different social status: Effects on social and agonistic behavior, locomotion, and stereotypies. Psychopharmacology 81:183190; 1983. 58. Miczek, K. A.; Haney, M. Psychomotor stimulant effects of d-amphetamine, MDMA and PCP: Aggressive and schedulecontrolled behavior. Psychopharmacology (in press). 59. Miczek, K. A.; Haney, M.; Tidey, J.; Vatne, T.; Weerts, E.; DeBold, J. F. Temporal and sequential patterns of agonistic behavior: Effects of alcohol, anxiolytics and psychomotor stimulants. Psychopharmacology 97:149-151; 1989. 60. Miczek, K. A.; Krsiak, M. Drug effects on agonistic behavior. In: Thompson, T.; Dews, P. B., eds. Advances in behavioral pharmacology. New York: Academic Press, Inc; 1979:87-162. 61. Miczek, K. A.; Kruk, M. R.; Ofivier, B., eds. Ethopharmacological aggression research. New York: Alan R. Liss, Inc.; 1984. 62. Miczek, K. A.; Mos, J.; Ollvier, B. Brain 5-HT and inhibition of aggressive behavior in animals: 5-HIAA and receptor subtypes. Psychopharmacol. Bull. 25:399-403; 1989. 63. Miczek, K. A.; Thompson, M. L.; Shuster, L. Opioid-like analgesia in defeated mice. Science 215:1520-1522; 1982. 64. Miczek, K. A.; Thompson, M. L.; Shuster, L. Naloxone injections into periaqueductai grey area and arcuate nucleus block analgesia in defeated mice. Psychopharmacology 87:39--42; 1985. 65. Miczek, K. A.; Thompson, M. L.; Shuster, L. Analgesia following defeat in an aggressive encounter: Development of tolerance and changes in opioid receptors. In: Kelly, D. D., ed. Stressinduced analgesia. New York: Ann. NY Acad. Sci. 1986:14-29. 66. Miczek, K. A.; Thompson, M. L.; Tornatzky, W. Subordinate animals: Behavioral and physiological adaptations and opioid tolerance. In: Brown, M. R.; Koob, G. F.; Rivier, C., eds. Stress: neurobiology and neuroendocrinology. New York: Marcel Dekker; 1991:323-357. 67. Miczek, K. A.; Weerts, E. M.; Tornatzky, W.; DeBold, J. F.; Vatne, T. M. Alcohol and "bursts" of aggressive behavior: Ethological analysis of individual differences in rats. Psychopharmacology 107:551-563; 1992. 68. Miczek, K. A.; Winslow, J. T. Analgesia and decrement in operant performance in socially defeated mice: Selective crosstolerance to morphine and antagonism by naltrexone. Psychopharmacology 92:444--451; 1987. 69. Mos, J.; Olivier, B.; Van der Poel, A. M. Modulatory actions of benzodiazepine receptor ligands on agonistic behaviour. Physiol. Behav. 41:265-278; 1987. 70.Mos, J.; Van Valkenburg, C. F. M. Specific effect on social stress and aggression on regional dopamine metabolism in rat brain. Neurosci. Lett. 15:325-327; 1979. 71. Olivier, B.; Mos, J.; Rasmussen, D. Behavioural pharmacology of the serenic, eltoprazine. Rev. Drug Metab. Drug Inter. 8:3183; 1990. 72. Olivier, B.; Mos, J.; Tulp, M.; Schipper, J. Modulatory action of serotonin in aggressive behaviour. In: Archer, T.; Bevan, P.; Cools, A., eds. Behav. Pharmacol. of 5-HT. Hillsdale, N.J.: Lawrence Erlbanm Associates, 1989. 73. Olivier, B.; Mos, J.; Van tier Heyden, J.; Hartog, J. Serotonergic modulation of social interactions in isolated male mice. Psychopharmacology 97:154-156; 1989. 74. Olivier, B.; van Dalen, D. Social behaviour in rats and mice: An ethologieally based model for differentiating psychoactive drugs. Aggress. Behav. 8:163-168; 1982. 75. Panksepp, J. Brain opioids: A neurochemieal substrate for narcotic and social dependence. In: Cooper, S. J., ed. Theory in psychopharmacology. New York: Academic Press; 1981:149175. 76. Pellow, S.; File, S. E. Anxiolytic and anxiogenic drug effects on exploratory activity in an elevated plus-maze: A novel test of anxiety in the rat. Pharmacol. Biochem. Behav. 24:525-529; 1986.

110 77. Puglisi-Allegra, S.; Cabib, S. Effects of defeat experiences on dopamine metabolism in different brain areas of the mouse. Aggress. Behav. 16:271-284; 1990. 78. Puglisi-Allegra, S.; Oliverio, A. Naloxone potentiates shockinduced aggressive behavior in mice. Pharmacol. Biochem. Behay. 15:513-514; 1981. 79. Rickels, K.; Downing, R. W. Chlordiazepoxide and hostility in anxious outpatients. Am. J. Psychiatry 131:442-444; 1974. 80. Rodgers, R. J. Differential effects of naloxone and diprenorphine on defensive behavior in rats. Neuropharmacology 21: 1291-1294; 1982. 81. Rodgers, R. J.; Cole, J. C.; Cobain, M. R.; Daly, P.; Doran, P. J.; Ells, J. R.; Wallis, P. Anxiogenic-like effects of fluprazine and eltoprazine in the mouse elevated plus-maze: Profile comparisons with 8-OH-DPAT, CGS 12066B, TFMPP and mCPP. Behav. Pharmacol. 3:621-634; 1992. 82. Rodgers, R. J.; Waters, A. J. Benzodiazepines and their antagonists: A pharmacoethological analysis with particular reference to effects on "aggression". Neurosci. Biobehav. Rev. 9:21-35; 1985. 83. Rolinski, Z. Interspecies aggressiveness of rats towards mice after the application of p-chlorophenylalanine. Pol. J. Pharmacol. Pharmacy 27:223-229; 1975. 84. Scheckel, C. L.; Boff, E. Effects of drugs in aggressive behavior in monkeys. Excerpta Med. Internat. 129:789-795; 1966. 85. Seeman, P.; Lee, T.; Chan-Wong, M.; Wong, K. Antipsychotic drug doses and neuroleptic/dopamine receptors. Nature 261: 717-719; 1976. 86. Shaikh, M. B.; Shaikh, A. B.; Siegel, A. Opioid peptides within the midbrain periaqueductal gray supress affective defense behavior in the cat. Peptides 9:999-1004; 1988. 87. Siegel, A.; Mirsky, A. F. The neurobiology of violence and aggression. In: Violence Report. Washington, DC: Nat. Acad. Press (in press). 88. Siegel, A.; Pott, C. B. Neural substrates of aggression and flight inthe cat. Prog. Neurohiol. 31:261-283; 1988. 89. Spanagel, R.; Herz, A.; Shipperberg, T. The effects of opioid peptides on dopamine release in the nucleus accumbens: An in vivo microdialysis study. J. Neurochem. 55:1734-1740; 1990. 90. Stutzmann, J.; Bohme, A.; Cochan, M.; Roux, M.; Blanchard, J. Proconflict and electrocorticographic effects of drugs modulating GABAergic neurotransmission. GABAergic Neurotransm. 91:74-91; 1987. 91. Tazi, A.; Dantzer, R.; Mormede, P.; Le Moal, M. Effects of posttrial administration of naloxone and/~-endorphin on shockinduced fighting in rats. Behav. Neural Biol. 39:192-202; 1983. 92. Thompson, M. L.; Kream, R. M. Social conflict alters brain opioid mechanisms in mice. Eur. J. Pharm. (in press). 93. Ticku, M. K.; Kulkarni, S. K. Molecular interactions of ethanol with GABAergic system and potential of Ro 15-4513 as an ethanol antagonist. Pharmacol. Biocbem. Behav. 30:501-510; 1988. 94. Tidey, J. W.; Miczek, K. A. Heightened aggressive behavior during morphine withdrawal: Effects of d-amphetamine. Psychopharmacology 107:297-302; 1992. 95. Tidey, J. W.; Miczek, K. A. Morphine withdrawal aggression: Modification with D l and D 2 receptor agonists. Psychopharmacology 108:177-184; 1992

MICZEK, WEERTS, HANEY AND TIDEY 96. Tidey, J. W.; Miczek, K. A. Effects of SKF 38393 and quinpirole on aggressive, motor and schedule-controlled behaviors in mice. Behav. Pharmacol. 3:553-565; 1992. 97. Tizabi, Y.; O'Donohue, T. L.; Jacobowitz, D. M. Variations in plasma and adrenal catecholamines and related enzymes in isolated-aggressive mice. Commun. Psychopharmacol. 4:433439; 1980. 98. Van der Poel, A. M.; Olivier, B.; Mos, J.; Kruk, M. R.; Meelis, W.; Van Aken, J. H. M. Anti-aggressive effect of a new phenylpiperazine compound (DU27716) on hypothalamically induced behavioural activities. Pharmacol. Biochem. Behav. 17:147-153; 1982. 99. Van Hoof f, J. A. R. A. M. Categories and sequences of behavior: Methods of description and analysis. In: Scherer, K. R.; Ekman, P., eds. Handbook of methods in nonverbal behavior research. Cambridge: Cambridge University Press; 1982:362439. 100. Virkkunen, M. Alcohol as a factor precipitating aggression and conflict hehaviour leading to homicide. Br. J. Addict. 69:149154; 1974. 101. Virkkunen, M.; De Jong, J.; Barko, J.; Goodwin, F. K.; Linnoila, M. Relationship of psychobiological variables to recidivism in violent offenders and impulsive fire setters. Arch. Gen. Psych. 46:600-603; 1989. 102. Virkkunen, M.; De Jong, J.; Bartko, J.; Linnoila, M. Psychobiological concomitants of history of suicide attempts among violent offenders and impulsive fire setters. Arch. Gen. Psych. 46: 604-606; 1989. 103. Virkkunen, M. E.; Horrobin, D. F.; Jenkins, D. K.; Manku, M. S. Plasma phosphofipid essential fatty acids and prostaglandins in alcoholic, habitually violent, and impulsive offenders. Biol. Psychiatry 22:1987-1096; 1987. 104. yon Blankenfeld, G.; Sontheimer, H.; Pritchet, D.B.; Ewert, M.; Seeberg, P.; Kettenmann, H. Differential benzodiazepine pharmacology of recombinant GABAA receptors. Neurosci. Lett. 115:269-273; 1990. 105. Weerts, E. M.; Tornatzky, W.; Miczek, K. A. "Anxiolytic" and "anxiogenic" benzodiazepines and/5-carbolines: Effects on aggressive and social behaivor in rats and squirrel monkeys. Psychopharmacology 110:451-459; 1993. 106. Weerts, E. M.; Tornatzky, W.; Miczek, K. A. Prevention of the proaggressive effects of alcohol by benzodiazepine receptor antagonists in rats and in squirrel monkeys. Psychopharmacology 111:144-152; 1993. 107. Winslow, J. T.; Miczek, K. A. Social status as determinant of alcohol effects on aggressive behavior in squirrel monkeys (Saimiri sciureus). Psychopharmacology 85:167-172; 1985. 108. Winslow, J. T.; Miczek, K. A. Androgen dependency of alcohol effects on aggressive behavior: A seasonal rhythm in highranking squirrel monkeys. Psychopharmacology 95:92-98; 1988. 109. Winslow, J. T.; Miczek, K. A. Naltrexone blocks amphetamineinduced hyperactivity, but not disruption of social and agonistic behavior in mice and squirrel monkeys. Psychopharmacology 96:493-499; 1988. 110. Yen, C. Y.; Stanger, R. L.; Millman, N. Ataractic suppression of isolation-induced aggressive behavior. Archives internationales de Pharmacodynamie et de Therapie 123:179-185; 1959.