Psycho pharmacology
Psychopharmacology53, 135-145 (1977)
9 by Springer-Verlag1977
Further Studies of the Aggressive Behavior Induced by A9-Tetrahydrocannabinol in REM Sleep-Deprived Rats E. A. C A R L I N I Depto. de Psicobiologia, Escola Paulista de Medicina, Rua Botucatu, 862, 04023 $5_oPaulo, Brazil
Abstract. The aggressive behavior induced by A9tetrahydrocannabinol in pairs of R E M sleep-deprived rats was studied in five experiments by measuring dominant and submissive behavioral patterns. When 2 REM-deprived rats received A9-THC, one of the animals displayed very aggressive postures, while its partner assumed incomplete defensive postures. The intensity of these behavioral postures was dosedependent. In pairs composed of one REM-deprived rat injected with A9-THC and one normal or one REM-deprived partner injected with control solution the deprived/drugged rat showed an aggressive posture and catatonia, or a strikingly bizarre behavior, while the control partner displayed typical defensive postures. The behavioral alterations induced in REMdeprived rats by amphetamine, LSD-25, and pentobarbital failed to provoke defensive postures in the normal rats paired with them; however, apomorphine partially mimicked the Ag-THC effects. It is concluded that in REM-deprived rats A9-THC not only provokes aggressive behavior but also impairs the defensive-submissive behavioral patterns.
Key words: R E M deprivation huana - Aggressive behavior
A9-THC -
Mari-
facing each other, including instances in which one animal forced the other to assume different patterns of submissive postures, were scored. Recently, several reports have examined in detail the response topography of rats engaged in fighting after receiving drugs (Crabtree and Moyer, 1973; McKenzie, 1971; Miczek, 1974; Miczek and Barry, 1974a, b). These studies prompted us to reanalyze the aggressive behavior of rats deprived of R E M sleep and treated with marihuana by using the methodology employed by those authors. On the other hand, Miczek (1974) has questioned how meaningful it is to subject simultaneously two opponents to a drug treatment, because the drug effects could either increase aggressive behavior of one of the contenders or indirectly change the defensive reactions of the opponent. Sbordone and Carder (1974) also stated that when both rats of a pair are treated with a drug (in their case mescaline) it is not possible to conclude whether the observed fighting is due to drug effects in the attacker, the victim, or both. Thus, the second aim of the present paper was to analyze the behavior of pairs of rats in which only one rat had received Ag-THC. MATERIALS AND METHODS
Subjects Cannabis extracts and (-)A9-trans-tetrahydrocannabinol (A9-THC) provoke aggressive behavior in paired rats previously deprived of rapid eye movement (REM) sleep (Alves et aI., 1973; Carlini and Lindsey, 1974). In these observations both rats of e a c h pair were REM-deprived and injected with the drug. Aggressive behavior was scored when both animals remained in an upright position, standing on the hind legs (Alves et al., 1973); that is, all postures the pair assumed in which they were in bodily contact and
Male Wistar albino rats 80-100 day old and weighing 220-250 g at the beginning of the experiments were used. They were raised in litters of 6 young to a mother and weaned when 25 days old. After weaning they were kept in groups of 6 in wooden cages measuring 49 • 20 x 29 cm until the beginning of the experiments. All animals were used only once unless otherwise stated.
Drugs An alcoholic solution of (-)A9-trans-tetrahydrocannabinol (A9THC) was kindly provided the National Institute on Drug Abuse (NIDA), U.S.A. The alcohol was evaporated and the residue thoroughlymixedwithTween-80for 5 min; the amount of Tween-80
136 was such that its final concentration in the suspensions did not exceed 0.87oo. Two drops of saline were then added, followed by 5 more min of thorough mixing. After that, 4 drops of saline were added and 5 more min of mixing took place. One milliliter portions of saline were added, followed by agitation to complete the desired volume. The control solution was composed of saline and Tween-80 in same proportions. Amphetamine sulfate (Sigma Chem. Co.), apomorphine hydrochloride (Sandoz A.G. Basle), LSD-25 (Delysid, Sandoz A.G. Basle; furnished by NIDA), and pentobarbital sodium (Abbott Laboratories) were dissolved in saline.
Psychopharmacology 53 (t 977) vigorously, leaving a mark on the pencil, one bite was scored. Thus, the maximum score for an animal was four bites during any one period of observation.
Vocalization. It was scored from 0 to 4 points as the animals did not vocalize at all (zero) or did it continuously throughout the 15 min of observation.
Aggressive Posture. The animal dominates the opponent by assuming an upright position, extending the forepaws, and exposing the teeth; the opponent assumes a defensive upright or submissive supine posture.
Statistics. Wide limits of error were obtained by authors when analyzing the behavioral effects of cannabis through several methods (Jones and Pertwee, 1972; Carlini et al., 1974). Feeney (/976) emphasized the dramatic effects of Ag-THC on measures of variability when compared to the effects on measures of central tendency. The present work also shows broad limits of error in the measures (aggressive posture, catatonia, submissive supine posture, defensive upright posture) employing an interval scale (s or min), as shown in the figures and tables. This is probably not attributable to uncontroled variables but is rather inherent to A9-THC effects. To better evidence this fact, the results are presented as means _+ SD or SEM and were analyzed through parametric methods. Furthermore, some of the measured behaviors-such as the submissive supine and the defensive upright postures-were mutually exclusive. For example, if half the submissive rats of all pairs in one experiment assumed submissive supine posture and the other half defensive upright posture, we would have large variabilities for each measure, although 100~ of these rats were dominated by their respective cagemates. Due to this fact, statistical analysis was not performed for all the experiments and sessions but only on those occasions we felt it would be pertinent. In these cases the comparison within the same group, during 2 successive experimental sessions, was carried out either through the t paired test (for aggressive, mutual upright, submissive supine, defensive upright, and catatonia postures) or through the Wilcoxon matched-pairs signed rank test (for bites and attacks). Comparisons between 2 different groups were made through the Student's t test and the Mann-Whitney U test.
Submissive Supine Posture. The submissive rat lies on its back with all 4 feet extended and the belly exposed; the dominant opponent maintains its aggressive posture.
EXPERIMENT 1
R E M Deprivation Procedure Rats were introduced into water containers in which a 6 cm platform was placed in the middle. The water level was 1 cm below the top of the platform and was changed every morning. Food was available ad libitum. The animals were REM-deprived for 4 days.
Behavior Measurements About 10 rain after the end of REM deprivation the rats were injected with the drugs under study and immediately paired in wire cages measuring 16 x 20 x 20 cm. One of the rats was marked with dye for easy identification. Several times after the injections (see Experiments below), the duration of eight categories of behavior was scored with the heIp of stopwatches for each rat during 15-rain periods. These categories are described betow.
Mutual Upright Posture. When both rats of the pair assume this position, it is impossible to know which animal dominates the other; thus, it is an item common to both ratsl They stand on their hind legs with the forepaws extended and often touching each other.
Defensive Upright Posture. The submissive rat stands on its hindlegs with its head tilted upward and forepaws abducted; the animal's back touches the Wail of the cage, frequently one of the corners. The teeth are not exposed and, when the dominant rat increases its aggressive posture, there is often teeth chattering. Immobile Crouch Posture. The submissive rat stays immobile with its 4 legs on the ground. Allogrooming. The rat grooms its partner. Autogrooming. The rat grooms itself, mostly in genital areas, paws, or as face washing. Catatonia. After assuming an upright position, the rat slowly bends forward, as melting, and remains immobile, arched, and with one or two forepaws in the air. The frequency with which the following two categories of behavior appeared was also scored during the 15-min intervals: Attack. Any form of direct approach by one of the rats toward the other, followed by an upright aggressive posture. Threat. The dominant ratarches its back and circles the submissive partner. Finally, bites and vocalization were quantified as follows: Bites. At 0, 5, 10, and 15 min of each 15-min period of observation, a black pencil was introduced once in front of the head of each of the two animals in the cage. If at each introduction the animal bit
I n t h e first e x p e r i m e n t a l s e s s i o n 24 r a t s (12 p a i r s ) w e r e d e p r i v e d o f R E M sleep. Six p a i r s w e r e t h e n i n j e c t e d w i t h 10 m g / k g o f A 9 - T H C ( g r o u p D A 9 • D A 9 ; Deprived + A9-THCxDeprived + A9-THC), and t h e o t h e r 6 p a i r s w i t h 1.0 m l / k g o f c o n t r o l s o l u t i o n ( g r o u p D C x D C ; D e p r i v e d + C o n t r o l sol. • D e p r i v e d + C o n t r o l sol.). B e g i n n i n g 30, 75, a n d 120 r a i n after injection, the behavioral measurements were
undertaken during the 15-rain periods described above. The scores obtained for the three periods were summed up so that the final score corresponded to 45 rain measurements. The entire procedure was repeated two more times (experimental sessions 2 and 3) with a resting period of 5 - 6 days before beginning the new periods of R E M deprivation. The scores of the 12 rats in group D A 9 x D A 9 in the 3 experimental sessions were analyzed and the rats of each pair were classified as dominant or submissive. A rat was considered dominant if in the 3 experimental sessions it scored at least 5 times more aggressive postures and 5 times fewer submissive and defensive u p r i g h t p o s t u r e s t h a n its o p o n e n t , w h i c h w a s t h e n
E. A. Carlini: REM Sleep, Ag-THC, and Aggression
137
Group DAgxDA g *
40
/
30
-
10
g u
/
\ i
\.,
/\
T x
o
?
',,;
20 .
~/"
* rl6 I
/ \
i
\
/~//f " 1
fo
/
0
-o
,
I
I
c 40
,
\
I
I
~
12
I A ~ I•
/:\
I
[] 9 subms.supine a 9 defens, upright
\\
E
7< H i % I
I
o 9 egress, posture x mutual upright
~
i i o/
/ ~ . , ~ . \'1
o
;,.12
S .~. / s L oo~
/
.x
r~
\,
•
m - - - - A ...... -A
I
Oe bites(median) [] 9 catatonia ( min )
-..
'.,
\
, I
x
I
I
I
experim. sessions
A attacks x vocelization
~r
B1
A1 3O /
20
!iI
/i
!
/
!
i
• i ..... g
~x.o I
~-
~
I
D1
12
/
I
!
,/"
10
C1
12
I
I
I
2
3
4
~It2 I
001IC]
~13I
!
/
i
i
-E -4
i f~
i5
.--eU
I
I
I
2
3
4
IOio I
1
/
/'3t
i
//i tICI~ I
2
ffCIo l
3
Group DC x DC
/
4 DO ]
X I
• X I
4
1
2
i i dx/ L 3
/
1
!
,,x experim. i sessions 4 /
Fig. 1. Categories of aggressive behavior induced by 10 mg/kg of A%THC in REM-deprived rats. Panels A, B, C, and D refer to pairs of rats composed of 2 REM-deprived + Ag-THC-injected rats (group DA 9 x DAn). Panels A 1, B1, C J, and D1 correspond to pairs composed of 2 REM-deprived + control solution-injected rats (group D C x DC). Heavy symbols indicate the dominant (panels A - D ) and the marked (panels A 1 - D 1 ) rats. Open symbols indicate the submissive (panels A - D ) and the unmarked (panels A 1 - D 1 ) rats. At the first three experimental sessions AO-THC and control solution were injected as stated above; at the fourth session Ag-THC was withheld from the dominant rats (panels A - D ) and injected into the marked rats (panels A 1 - D 1). The vertical bars represent the standard errors of the means. The asterisks indicate statistically significant differences from the previous experimental session within the same group (* P % 0.01 and P _< 0.025, respectively, for t paired test and the Wilcoxon matched-pairs test)
considered submissive9 The criterion was achieved in all 6 pairs. The rats in pairs DC • DC were not classified as dominant and submissive because they did not fight, and their behavior was scored, respectively, for the marked (with dye) and the unmarked .rats of each pair. Finally, in a fourth experimental session, the animals were again REM-deprived and injected, but now the dominant rats of group D A 9 • D A 9 received control solution while their submissive partners still received A 9 - T H C . [n group D C x D C the marked animals now received A g - T H C and their unmarked partners still received control solution. The items of behavior were assessed as before, after which the animals were discarded. Results. The rats of group DC x DC did not display aggressive behavior in the first three experimental sessions (Fig. 1, panels A1 to DI). The exception was one pair in the first session, which continually presented mutual upright postures (panel A1) that nevertheless disappeared completely during the second session. In marked contrast to this group, all 6 pairs of group D A 9 • D A 9 evidenced marked alterations of
behavior9 Since the first session, the rats were very excited and by frequently jumping they bumped each other and fighting started. These 'attacks' contrast markedly with the description of Miczek (1974), and it was impossible to identify the responsible rats. The 'attacks' decreased considerably in the second (Fig. 1, panel D; P < 0.025, Wilcoxon matched-pairs test) and following sessions. At the beginning of the first experimental session and after each 'attack', both rats assumed the mutual upright posture (Fig. 2) but in all 6 pairs one of the rats soon dominated the other and assumed a more aggressive posture (Fig. 1, panel A). Most of the time their opponents showed defensive upright or submissive supine postures (Fig. 1, panel B), but these defensive postures were somewhat incomplete. They extended their forepaws, fully exposed their teeth, and tilted their heads only partially upward (Fig. 3, parts A and B). They also vocalized constantly and reacted to the dominants provoking from these more aggressive postures. It seemed that the submissive rats were fighting back but, by 'losing the fight', were driven backward toward one corner of the cage in a defensive upright posture (Fig. 3, part A) or fell with their backs on the floor in submissive supine
138
Fig.2. Typical mutual upright posture of two REM-deprived + A9-THC-injectedr a t s ( D A 9 rats) during the first experimental session of Experiment
posture (Fig. 3, part B), due to the constant pressure of the dominant animal. This impression was strengthened by the number of times the submissive rats vigorously bit the pencil (Fig. 1, panel C). The dominant and submissive roles in all pairs became still more clear in the second and third sessions. Thus, there was a decrease of jumps, attacks (Fig. 1, panel D), and mutual uptight posture (Fig. 1, panel A), whereas the time of engagement was almost maximal (45 min of observation) with the dominant rats spending more time in aggressive postures (Fig. 1, panel A) and the submissive ones in the submissive and defensive upright postures (Fig. 1, panel B). The importance of A9-THC in producing this behavior became clear in the fourth session, in which A9-THC was withdrawn from the dominant rats of group DAgx DA 9 and given to the marked rats of group DC x DC. The once submissive rats of group DA9x DA 9 who were still receiving A9-THC in the fourth session, now assumed a clearly dominant role. They showed nearly 25 rain of aggressive posture (Fig. l, panel A; P < 0.001, t paired test) and no submissive supine or defensive upright postures (Fig. 1, panel B; P _< 0.001, t paired test). Conversely, the once dominant rats, now without A9-THC, showed nearly 0 rain of aggressive posture (Fig. 1, panel A; P < 0.001, t paired test) and more than 30 rain of defensive upright and submissive supine postures (Fig. l, panel B; P < 0.001, t paired test). These defensive postures were now typical as when in defensive upright the animals were motionless, chattering their teeth, tilting their heads upward, and abducting their forepaws (Fig. 3, part C); when in submissive supine, they were motionless (Fig.3, part D). Also,
Psychopharmacology53 (1977)
Fig.3. Panels A and B show further aspects of the aggressive behavior between two D A 9 rats. In panel A the submissive DA9 rat (right), in spiteof beingin defensiveuprightposture,is vocalizing, has its teeth exposed, and extends the forepaws against its DA9 dominant partner. In panel B the submissive D A 9 rat (left) is in submissive supine posture, but is also reacting against its DA9 dominant partner. Pictures were taken during the second experimental session. Panels C and D show the behavior of rats of the s a m e DA9 • D A 9 group at the fourth experimentalsession.The rats on the right sides, which were clearly dominant in the 3 previous sessions, are now without Ag-THC and assumetypicallydefensive upright (panel C) and submissivesupine (panel D) postures
they did not bite the pencil (Fig. 1, panel C; P < 0.025, Wilcoxon matched-pairs test). Due to the 'frozen' state of these animals, their A9-THC-treated partners began to present periods of catatonic reactions (Fig. 1, panel C), bending slowly (as 'melting') in front of their motionless partners and staying in a semireared position. In group DC x DC similar facts were revealed in the fourth session. The marked rats, now injected with A9-TI-IC, assumed a clearly aggressive posture (Fig. 1, panel A 1 ; P < 0.001, t paired test), frequently bit the pencil, (P < 0.025, Wilcoxon matched-pairs test) and stayed 15 min in a catatonic posture (Fig. 1, panel C1 (P _< 0.001, t paired test). Sometimes the Ag-THC-treated rat jumped, interrupting the aggressive posture against its partner, and after attacking the wires on the opposite wall of the cage assumed the catatonic state; its partner, on the other side of the cage, still remained frozen in the defensive upright or submissive supine posture. Consequently, as shown in panels A 1 and B 1 of Figure 1, the time spent by the rats without A9-THC in defensive upright and supine postures (38.5 min) was longer than that spent by their Ag-THC-treated partners in aggressive postures (24 min). These phenomena were even more evident in the rats of Experiment 2, described below.
E. A. Carlini: REM Sleep, Ag-THC, and Aggression 1A
30 ~"
Group DA 9 x DC
T
40
. / ~
u
-
",,
-
r ".
10
-
~ o o ogress.posture X mutual upright
D
12
8 ,:"
i i ~ [ ] 9 subm. supine A A defens.upright
~
/ ~
\.
\.8
I '.,4 / ; .
E o
C
=t2
/ B ' ~
20-
~39
\,4
ou
ou ou-'-'u~, ~ ~ ~ ~ ~ o o bites {median) A attocks 131 cototonia (min] X vocolizotion
x.~experim. ~ sessions
4.0 A1
~ ~T /---q
30
20 g
.
\
T
C1 12
IE'~8 ~
m,~-----~-------- -": i i
D1
12
9 ou i
I \E
ou t
ou q
Group DA s x N c
Fig.4. Categories of aggressive deprived + control solution) rats AI-D1). Asterisks in the first 3 submissive partners (DC and NC
%*
~
mm ~ ~
\
"" :'*
"l~':~Zexperim. 88sessions J
behavior exhibited by pairs composed of one DA ~ (REM-deprived + AKTHC) and one DC (REM-
(panels A - D), or from pairs composed of one DA9 and one NC (normal + control solution) rats (panels sessions indicate statistically significant differences between dominants (in this case D~t9 rats) and their rats). Asterisks in the fourth session are to be read as in Figure
Blood was visible in about 20 % of the fighting pairs, regardless of the group considered; in all cases the blood originated from the forepaws. In the four experimental sessions, none of the A9-THC-treated animals presented threats or lateral blocks, immobile crouches, allogrooming, or autogrooming. The control solution-treated rats presented allo- and autogrooming that varied from a few seconds to a maximum of 3 rain, depending on the experimental session.
EXPERIMENT 2 In Experiment 1 all pairs were composed of similarly treated rats. In Experiment 2, the REM-deprived and A9-THC-treated rats (DA 9 rats) were paired with either REM-deprived or normal rats, both of which were injected with control solution (DC or NC rats). Eighteen rats were REM-deprived. Twelve of them constituted group DA 9 x DC and the remaining 6 entered in the pairs of group DA9x NC. The dose of A9-THC was J0 mg/kg. Therefore, these pairs basically differed from group DA 9 x DA 9 (Experiment 1) in that only one animal of each pair was treated with Ag-THC. Four experimental sessions with three 15-min observation periods each were carried out. At the last session the Ag-THC injection was withheld and all rats received
control solution. The classification criterion for dominant and submissive rats was the same as in Experiment 1. Results. Beginning with the first experimental session, the DA 9 rats clearly assumed the dominant role with their DC and NC partners. In the first three experimental sessions they showed very aggressive postures against their oponents (Fig. 4, panels A and A1), very little or no submissive supine or defensive upright postures (Fig.4, panels B and B1), and bit the pencil (Fig. 4, panels C and C1). They also showed 'attacks' (Fig. 4, panels D and D 1) and periods of catatonic state (Fig.4, panels C and C1). Their behavior was quite bizarre. After displaying the aggressive posture against the partners for a period of time, they turned around very fast or jumped backward and started to 'fight' against the wires of the cages (Fig. 5, panel A), vocalizing a lot. Next, they either jumped again (falling or not on their opponents and assuming once again the aggressive posture) or remained in the same place, bending slowly and entering a catatonic state (Fig. 5, panel B). Their DC or NC partners showed no aggressive postures (Fig.4, panels A and A1), did not bite the pencil (Fig.4, panels C and C l), and stayed for long periods of time in a typically defensive upright posture (Fig.4, panels B and BI). Several actually
140
Psychopharmacology53 (1977)
Fig. 5 Bizarre behavior of DA9 rats when paired with control solution-injectedrats (DC or NC rats). Panel A shows one DA9 rat (left) 'attacking' the wires of the cage; panel B shows one DA9 rat (left) in catatonic position. Panels" C and D were taken in sequencefrom the same pair; the DA9 rat (right) seems to be fighting against an imaginary foe. In all panels the submissiveNC or DC partners are completely immobile in either typical submissive supine or defensiveupright postures
remained frozen in a corner of the cage during the entire 2 h 15 min duration of the session, disregarding the DA 9 rats that were in aggressive postures, 'fighting' the wires in the other corner of the cage, or even in catatonic states (Fig.5, panels B, C, and D). The immobility was so complete that they refused to change their posture even when pushed with a pencil. However, an interesting change in behavior took place in some of these rats, mostly during the third experimental session. In 2 pairs of group DA 9 x DC and in 2 pairs of group DA 9 x NC, the control solutioninjected rats seemed to learn how to deal with the DA 9 rats. Although assuming what seemed to the observer a defensive upright p o s t u r e - they presented chattering of the teeth, tilted their heads upward, abducted their forepaws, and did not expose their t e e t h - t h e y began to 'walk' toward their opponents on their hindlegs, though they maintained some distance from them (Fig. 6). At the same time they started a very distinct vocalization, emitting short, soft, and regularly spaced squeals that contrasted with the irregular, louder, and longer vocalizations o f the DA 9 animals. By assuming this posture, they sometimes managed to control their 'partners' behavior: the DA 9 rats retreated, fell backward, and remained in what seemed defensive upright or submissive supine postures, though they still vocalized and moved their forepaws (Fig.6). A few moments later, however, the DA 9 rats again assumed their aggressive posture by jumping or turning around. These observations led to higher scores of the mutual upright posture, attacks, and also aggressive posture for the DC and NC rats during the third experimental session (Fig. 4, panels A and A 1); conversely, the DA 9 partners showed submissive supine and defensive upright positions (Fig.4, panels B and B1). Figure 4 also shows again the importance of A9-THC in inducing the behavior being described. At the follrth
Fig.6. Picture taken from a DAgx NC pair, at the third experimental session (Experiment2). The NC rat (left) slowly 'walks' toward the partner but still maintains features of defensive upright posture, causing a bizarre reaction from the DA9 rat
experimental session, in which all rats received only control solution, the items of behavior were practically absent and were thus significantly smaller than those obtained in the third session. Allogrooming, autogrooming, threats, or lateral blocks and immobile crouches were not observed in the DA 9 animals.
EXPERIMENT 3 This experiment was designed to further assess the importance of the association of R E M deprivation and A9-THC in inducing aggressive behavior. Fifteen pairs of rats were divided into 3 groups of 5 pairs each. In group NA 9 x NA 9 each pair was composed of 2 normal rats injected with 10 mg/kg Aq-THC.
141
E. A. Carlini: REM Sleep, Ag-THC, and Aggression Table 1.
Categories of aggressive behavior in pairs of rats subjected to different experimental conditions
Groups
Categories of behavior (in s + SD or median' shown by Dominant or marked rat Rats
Aggressive posture
Bites (median)
Submissive or unmarked rat Catatonia (s)
Rats
(s) NA 9 x NA 9 DC x NC
NA 9 DC
D A 9 X NA 9
DA 9
0 12 _+ 18 724 _+ 49
0 0 4
0 0 46 • 39
Group DC x NC consisted of REM-deprived rats injected with control solution and their normal partners injected with control solution. In group D A 9 x N A 9 the treatment consisted of deprivation plus 10 mg/kg of A9-THC and only 10 mg/kg A9-THC, respectively, for each member of the pairs. The categories of behavior were assessed as in Experiment 1, but only one 15-rain measure, from ~/5-90 rain after injections, was used. Results. Neither Ag-THC alone (group N A 9 x N A 9) nor REM deprivation alone (group DC x NC) was able to induce aggressive behavior (Table i). On the other hand, all 5 D A 9 animals in group D A 9 x N A 9 bit the pencil and displayed aggressive posture toward t h e i r N A 9 partners, which scored defensive upright postures and catatonia (Table 1). However, these subm i s s i v e N A 9 rats presented differences when compared to the DC or NC submissive animals. Their defensive upright posture was incomplete because they vocalized and extended and moved their forepaws, which provoked more aggressive postures from their DA 9 partners.
EXPERIMENT 4 In the three previous experiments the dosage of A g - T H C w a s 10 mg/kg. Experiment 4 was designed to verify whether aggressive behavior appeared with smaller doses. Six groups of 5 pairs each were used. In 3 groups both rats of each pair were RElV~-deprived and received, respectively, 1.25, 2.5, and 5.0mg/kg of A g - T H C . In the remaining 3 groups one rat of each pair was REM-deprived and received the above doses of A 9 - T H C , while their normal partners were treated with control solutionl One session of 15 min duration, beginning 75 rain after the injection, was scored. Results. Figure 7 shows a dose-dependent increase in all categories of behavior. Thus, the larger the dose, the larger the amount of aggressive postures (panels A
NA 9 NC NA 9
Submissive supine
Defensive upright
(s)
(s)
0 0 77 -t- 85
12 _+ 18 637 • 78
0
Bites (median)
0
0 3.5
Catatonia (s)
0
0 144 _+ 48
and AI) and bites (panels C and C1) displayed by the dominant DA 9 rats; conversely, the submissive rats showed submissive supine and defensive upright postures (panels B and B 1). The submissive NC rats displayed typical submissive supine and defensive upright postures (880 s; Fig.7, panel BI; 5 mg/kg) that contrasted with the D A 9 submissive animals who showed less defensive postures (355 s; Fig. 7, panel B; 5 mg/kg) and engaged more in mutual upright postures (Fig. 7, panel A). Thus, the postures assumed by the submissive rats clearly affected the behavior of dominant rats. The frozen submissive postures assumed by the NC rats resulted in the absence of mutual upright postures (Fig. 7, panel A1) and in more states of catatonia in the D A 9 rats (Fig. 7, panel C1). To study further the behavioral patterns of the d o m i n a n t D A 9 rats, according to the different partners with which they were provided ( D A 9, D C , N C , and N A 9 rats), a joint analysis of the four experiments was performed. The second 15-rain period ( 7 5 - 9 0 rain after injection) of the first experimental sessions in which A9-THC was given to the animals was selected. Only data obtained with 5 mg/kg of A9-THC in Experiment 4 was included. In all, 12 D A 9 • 9 pairs, 20 D A 9 x N C plus DAgxDc pairs, and 5 DA 9 x N A 9 pairs were analyzed. The data are summarized in Figure 8. When paired with submissive DA 9 rats, the dominant D A 9 animals presented about 300 s of aggressive posture, little catatonia, and engaged in about 400 rain of mutual upright postures (Fig.8, left side); conversely, the submissive D A 9 animals did not present many defensive upright postures (Fig. 8, right side). Paired with NC and DC submissive partners, the dominant D A 9 rats displayed more catatonia and less mutual upright posture. On the other hand, the submissive NC and DC rats showed typical defensive upright postures. Finally, when the D A 9 dominant rats were paired with N A 9 rats, the longest time of aggressive posture is observed, with little catatonia and mutual upright posture; conversely, the submissive N A 9 rats remained most of the time in defensive upright postures, which as mentioned
142
Psychopharmacology53 (1977)
Group D.49xDA 9
f 8
I
A
Z,
B
c ~ l
6
3
D
12
0
~2
8
o •
1
2
1 i
o e agress.posture mutual upright
S g
I
z, I ?o
i
O 9 subm. supine z&9 defensive upright
.o
A1
B•
6
[] 9 catatonia ( rain )
3
./.. 1.25
2.5
5.0
1.25
\
2.5
A attacks voccdization
o 9 bites ( median )
4-
~8
16
C1/ E8~12
-
1
jL/" D I~ D O I
1.25 2.5 Group D/19xNC 5.0
mglkg Ag-THC
.o
DI
4 DOI
5.0
1.25
2.5
5.0 Ag-THC
J
Fig. 7. Effects of several doses of Ag-THC on several categories of aggressive behavior. To be read as in Figure J. Note that the Ag-THC effects are dose-dependent. Asterisks indicate statistically significant differences from the preceding dose (P < 0.05 or less; Student's t-test or Mann-Whitney U-test)
DOMINANT DAs RATS 800-
[ ] aggress, posture 700-
SUBMISSIVE RATS [ ] subm. supine [ ] defens, upright N A 9
[] catatonia
DC+NC
[ ] mutual upright
EXPERIMENT 5
600-
500-
u~ 4 0 0 o
300=
g
"o
200-
=-
too.
D~~
NC+DC
submissive
Fig.8.
before (Experiment 3) was also accompanied by vocalization and motion of forepaws.
N , ~g
partners
DZ~ dominant
Dz~
DA*
partners
Categories of aggressive behavior of dominant DA 9 rats
(right) and their submissive DA 9, NC, D e , or NA 9 partners (left). Right: aggressive posture (hatched columns), catatonia (open columns), and mutual upright posture (dotted columns) of the dominant DA9 rat when paired with different submissivepartners. Left: submissive supine (open columns) and defensive upright (columns with vertical lines) postures of different submissive rats when paired with dominant DA9 animals
In the previous experiments the submissive DC and NC rats displayed a defensive behavior whether the dominant DA 9 rats were in an aggressive posture, 'attacking' other parts of the cage, or in a state of catatonia. Experiment 5 was designed to verify whether or not other drugs that induce behavioral alterations in rats would also provoke defensive postures from control rats. The drugs chosen were apomorphine, amphetamine, pentobarbital sodium, and LSD-25. Groups of 5 pairs of rats were deprived of R E M sleep and then injected with the several drugs at the dosages stated in Table 2. Fifteen minutes after apomorphine, pentobarbital, and LSD-25, and 60 rain after amphetamine injections, each treated animal was paired with a normal rat and the categories of behavior were scored for the next 15 min. Results. Table 2 shows that with the exception of the apomorphine-treated rats the animals injected with the other three drugs were not able to elicit defensive
E. A. Carlini: REM Sleep, Ag-THC, and Aggression
143
Table 2. Behavioral responses of normal rats when paired with REM-deprived rats injected with several drugs (challenger rat). Five pairs of rats for each treatment Drug to challenger rat (mg/kg)
Posture (in s • SD) of normal rat Aggressive
Submissive supine
Defensive upright
Allogrooming
Autogrooming
-
0
0
0
21 •
18
53 • 42
Control solution
0
0
0
30 • 15
78 • 28
J-THC (10)
0
Apomorphine HC1 1.0 2.0 4.0
0 0 4 • 6
0 0 0
Pentobarbital sodium 10 20
0 0
0 0
6• 0
LSD-25 0.1
0
0
0
101 + 92
688 • 51
0
184 • 38 350 • 68 567 • 37
0 0 0
6
0
Mutual upright (s • SD)
13 • 21 0 107 • 68
14 • 10 12 • 6 19 • 13
0 0 0
0 21 • 13
0 42•
25 •
7
40_+ 15
0
22_+ 9 • 10+_ 6_+ 0 0
8 12 2 3
9• 3 17 • II 8• 4 16• 9 18 • 3 23 • 14
6• 0 0 0 0 0
21
29• 0
7
Amphetamine sulfate 0.05 0.1 0.5 1.0 2.0 4.0 8.0
4• 0 0 0 0 0 0
3• 8 0 0 6• 6 97 • 78 0 0
11 • 15 0 3 7 • 42 75 • 101 0 0 5 • 10
postures from the normal rats. REM-deprived rats treated with pentobarbital displayed gross ataxia and increased ambulation, touching their normal partners while awkwardly traveling in the cage. They often tried to rear up and fell backward onto their partners. REM-deprived rats treated with LSD-25 traveled in the cages with their bodies pressed against the floor and their noses very near the floor; consequently they often crawled under their control partners. The smaller doses of amphetamine induced mild excitation in the REM-deprived rats, and the larger ones caused intense stereotyped movements; and yet these rats were unable to elicit clear defensive postures from normal partners. These data were in marked contrast with the almost 13 rain of defensive postures observed in normal rats challenged by DA 9 rats.
DISCUSSION Experiment 1, with DA 9 paired rats, indicates that the fighting behavior observed does not conform to the typical patterns of fights between untreated rats described previously (Barnett, 1963; Davis, 1933;
Eibl-Eibesfeldt, 1961; Grant, 1963; Grant and Makintosh, 1963). The aggressive postures observed never began with such ritualized postures as lateral blocks or threats. On the contrary, they started explosively through a jump of one of the animals, preceded or not by a tactile contact between the partners. Allogrooming and autogrooming were essentially absent. On the other hand, the submissive DA 9 rats never presented immobile crouches, and even when in defensive upright and submissive supine postures they bit the pencil and reacted to their partners with vocalization and movement of forepaws. Consequently, their postures did not effectively deter their dominant partners. These results suggest that A9-THC induced aggressive behavior in both rats of each pair, and consequently also blocked the full expression of defensive reactions in the defeated animal. Therefore, the topography of this aggressive behavior clearly differed from that observed in normal rats. Miczek (I 976 a) proposed that this behavior should be distinguished from natural aggressive behavior and further suggested that druginduced aggressive reactions should be considered forms of behavioral pathologies (Miczek, 1976b). The capacity of A9-THC to induce aggressive behavior in REM-deprived rats was further confirmed
144
in Experiment 2. The D A 9 rats vigorously bit the pencil and assumed aggressive postures against their control solution-injected partners. Conversely, these DC and NC rats evidenced striking defensive postures that were frequently characterized by complete immobility in the defensive upright posture. That the D A 9 rats elicited these postures from the control animals may be the best evidence of the aggressioninducing properties of A 9 - T H C in REM-deprived rats. After all, it is likely that rats are quite capable of perceiving aggressive postures in other rats. Other results further confirm the aggressioninducing properties of A9-THC. In Experiment 1, withholding A9-THC transformed into submissives the rats which were consistently dominants in the three previous sessions. In Experiment 2, fourth session, Ag-THC was not given and the aggressive behavior observed in the three previous sessions disappeared. In Experiment 3, only the DA 9 rats fought. Finally, in Experiment 4 the A9-THC property of inducing aggressive behavior was clearly shown to be dosedependent. Other drug-induced alterations of behavior failed to elicit defensive postures from normal rats: the altered behavior induced by LSD-25, pentobarbital, and amphetamine was ineffective in this regard. The ability of low doses of apomorphine to induce aggressive behavior in REM-deprived rats has been described before (Carlini and Lindsey, 1974). The double t r e a t m e n t - R E M deprivation plus Ag-THC-also produced other behavioral abnormalities (Experiment 2). When the aggressiveness of the DA 9 rats was not stimulated by their immobile submissive partners, catatonia and bizarre behavior followed. Catatonia is a well-known effect of Ag-THC in normal rats, and its longer or shorter duration in our experimental conditions depended on the stimuli provided by the partners. The bizarre behavior was remarkable. The author has been working with A9THC and cannabis compounds for the last 14 years and has never observed anything similar to it in normal rats and mice. The attacks and biting against the wires of the cages, the fighting postures assumed against imaginary foes, the vocalization, and the jumps and rapid circular turning (while the normal partners were in complete immobility) all indicated that these rats were undergoing profound behavioral abnormalities (Fig. 4). At present there is no way to classify these manifestations, but it is rather tempting to suggest that they could be considered an experimental model of hallucinatory behavior in rats. The impairment by A9-THC of the defensive patterns of behavior, suggested in Experiment 1, became clear in group D A 9 • N A 9 of Experiment 3. The large number of defensive upright positions dis-
Psychopharmacology 53 (1977)
played by the N A 9 rats failed to effectively deter their D A 9 partners, which maintained their aggressive postures most of the time (Table 1). This was probably due to an incomplete defensive upright posture, as it was accompanied by vocalization and motions of the forepaws that provoked a more aggressive posture in the D A 9 rats. Miczek and Barry (1974a) have also shown that A9-THC impairs the defensive submissive behavior of rats, but in contrast with the present data, they attributed the effect to an increase in immobile crouching postures that decreased defensive upright and submissive supine postures. A comparison of Experiments I and 2 may be helpful in analyzing the meaningfulness of simultaneously treating both rats of a pair with drugs (Miczek, 1974). Both experiments reveal that, at least in our experimental conditions, treating both rats of each pair is also a sound procedure. In fact, aggression can only manifest itself through an interaction between at least 2 animals. If one of the contenders displays efficient defensive postures, the aggressor may decrease or even stop its aggression. In Experiment 1 both rats were under A9-THC and consequently the defensive postures of the submissive animal were impaired, resulting in more aggressive postures from the dominant partner. On the other hand, in Experiment 2 the efficient defensive postures assumed by the undrugged rats reduced the aggression of the D A 9 partners. Sbordone and Carder (1974) also reported that when both rats of a pair were treated with mescaline aggressive behavior occurred, and that when only one of the rats was treated fighting did not occur at all. Sheard (1973) studied the aggressive behavior of a rat habituated to its homecage toward intruder rats. The aggression of the homecage rat was greater when the intruder rats were treated with amphetamine and p-chlorophenylalanine. He concluded that the drugs impaired the defensive postures of the intruder rats and hence decreased the normal cues for turning off the aggression of the dominant rat. Actually, the situation may bear some resemblance to the effects of alcohol on human behavior. One would not expect a sober person to react to the provocation of an alcoholintoxicated individual, and yet fighting is a common behavioral manifestation between two such individuals (Tinklenberg, 1973). Acknowledgements. I gratefully acknowledge the generous gifts of Sandoz and Abbott Laboratories.
REFERENCES Alves, N. C., Goyos, A. C., Carlini, E. A. : Aggressiveness induced by marihuana and other psychotropic drugs in REM sleep deprived rats. Pharmacol. Biochem. Behav. 1, 183-189 (1973)
E. A. Carlini: REM Sleep, Ag-THC, and Aggression Barnett, S. A. : A study in behavior. London: Methuen 1963 Carlini, E. A., Karniol, I. G., Renault, P. F., Schuster, C. R.: Effects of marihuana in laboratory animals and in man. Brit. J. Pharmacol. 50, 299-309 (1974) Carlini, E. A., Lindsey, C. J. : Pharmacological manipulations of brain catecholamines and the aggressive behavior induced by marihuana in REM-sleep-deprived rats. Aggres. Behav. 1, 81-99 (1974) Crabtree, J. M., Moyer, K. E.: Sex differences in fighting and defense induced in rats by shock and d-amphetamine during morphine abstinence. Physiol. Behav. 1t, 337-343 (1973) Davis, F. C. : The measurement of aggressive behavior in laboratory rats. J. genet. Psychol. 43, 213-217 (1933) Eibl-Eibesfeldt, I. : The fighting behavior of animals. Sci. Amer. 205, 112-122 (1961) Feeney, D. M. : The marihuana window: a theory of cannabis use. Behav. Biol. 18, 455-471 (1976) Grant, E. C.: An analysis of the social behaviour of the male laboratory rat. Behaviour 21, 260-261 (1963) Grant, E.C., Mackintosh, J. H.: A comparison of the social postures ofsome common laboratory rodents. Behaviour 21, 246- 259 (1963) Jones, G., Pertwee, R. G. : A metabolic interaction in vivo between cannabidiol and Al-tetrahydrocannabinol. Brit. J. Pharmacol. 45, 375- 377 (1972) McKenzie, G. M. : Apomorphine-induced aggression in the rat. Brain Res. 34, 323-330 (1971)
145 Miczek, K. A. : !ntraspecies aggression in rats: effects of d-amphetamine and chlordiazepoxide. Psychopharmacologia (Berl.) 39, 275- 301 (1974) Miczek, K. A. : In the discussion of the papers on marihuana and aggressive behavior. In: The pharmacology of marihuana, M. C. Braude and S. Szara, eds., p. 533. New York: Raven Press 1976a Miczek, K. A. : Mouse-killing behavior and motor activity: effects of chronic A9-tetrahydrocannabinol and pitocarpine. Psychopharmacology 47, 5 9 - 64 (1976 b) Miczek, K. A., Barry III, H.: Effects of A~-tetrahydrocannabinol on aggressive behavior in laboratory rats. In: Drug addiction, vol. III, J. M. Singh and H. Lal, eds., pp. 19-38. New York: Stratton 1974a Miczek, K. A., Barry III, H. : Ag-Tetrahydrocannabinol and aggressive behavior in rats. Behav. Biol. 11, 261- 267 (1974b) Sbordone, R. J., Carder, B. : Mescaline and shock induced aggression in rats. Pharmacol. Biochem. Behav. 2, 777-782 (1974) Sheard, M. H. : Aggressive behaviour: modification by amphetamine, p-chlorophenylalanine and lithium in rats. Aggressologie 14, 323-326 (1973) Tinklenberg, J. R. : Alcohol and violence. In: Alcoholism progress in research and treatment, pp. :195- 210. New York: Academic Press 1973
Received September 29, 1975; Final Version January 18, 1977