Monoamines as Mediators of Avoidance-Escape Behavior* HOWARD I. GLAZER, PHD, JAY M. WEISS, PHD, LARISSAA. POHORECKY,PHD, ANDNEALE. MILLER, PHD A series of three experiments were carried out to test the hypothesis ("motor activation deficit" hypothesis) that the avoidance-escape deficits observed following certain highly stressful conditions result from changes in activity of noradrenergic (or other monoaminergic) neural systems. These studies indicate that: (1) Depletion of monoamines by a single injection of tetrabenazine produces an active avoidance-escape deficit when the avoidance-escape response involves a relatively high degree of motor activity but not when a minimum of motor activity is required. This parallels results found when animals are acutely exposed to a stressor prior to avoidanceescape testing. (2) Daily injections of tetrabenazine for a period of two weeks, like daily exposure to inescapable shock for the same period of time, markedly attenuates the magnitude of the avoidance-escape deficit produced by either a single injection of tetrabenazine or single session of inescapable shock. (3) Decreasing the stress-induced depletion of monoamines by the use of an MAO inhibitor serves to protect the animals from the effects of inescapable shock, markedly reducing the avoidance-escape deficit produced by such shock. It is concluded that these results are consistent with the motor activation deficit hypothesis.
INTRODUCTION
presented indicating that these stressors produced this behavioral deficit by temThis paper is the third in a series study- porarily reducing neurotransmitter activing how exposure to severe stressors af- ity important for active motor behavior. fects behavioral responses. In two previ- The first paper described six studies which ous papers (1,2), it was shown that both showed, in essence, that exposure to cold brief swim in cold water and strong ines- swim and strong inescapable shock procapable shocks interfered with subsequent duced a transient avoidance-escape deficit avoidance-escape behavior. Evidence was that was highly sensitive to the amount of motor activity required in the avoidanceescape task, with the avoidance-escape This is the third paper in a series of three on the deficit being intensified, reduced, or even subject of avoidance-escape behavior. This series is eliminated completely depending on the published in its entirety in this issue. From The Rockefeller University, New York, NY amount of motor activity required of the 10021. animal in order to execute the correct reSupported by Public Health Service Grants AA sponse. The second paper, in testing the 00296 and MH 13189, and by grants from the Alfred P. Sloan, Grant and Mobil Foundations, the Merrill hypothesis that this stress-induced Trust, the Scottish Rite Schizophrenia Research Pro- avoidance-escape deficit was mediated by gram, and Hoffmann-La Roche. HIG held a fellowship noradrenergic depletion in the brain, from the Foundations' Fund for Research in Psychiatry during the course of the research and is showed that if one counteracted brain presently at Loyola Campus of Concordia University, noradrenergic depletion by repeatedly exMontreal, Quebec. posing animals to the inescapable shock or Received for publication 9 January, 1975; revision the cold swim, then the avoidance-escape received 9 June 1975. Psychosomatic Medicine Vol. 37, No. 6 (November-December 1975) Copyright ° 1975 by the American Psychosomatic Society, Inc. Published by American Elsevier Publishing Company, Inc.
535
HOWARD I. GLAZER, et al.
deficit also did not appear. The present paper now describes experiments which further test for neurochemical mediation of the avoidance-escape deficit by attempting to (a) produce the avoidance-escape deficit by using a drug treatment instead of a stressor, and (b) eliminate the stressinduced avoidance-escape deficit by appropriate pharmacological intervention. Before describing the experiments which were undertaken, it is important to comment on the hypothesis that strong stressors, such as severe inescapable shock and cold swim, produce an avoidanceescape deficit by reducing central neurotransmitter activity important for active motor behavior. This hypothesis, called the "motor activation deficit" hypothesis, has indeed focused on the role of norepinephrine, but it is in principle more broadly based, being linked to those neurotransmitters which are (a) affected by stressful conditions, and (b) mediate active motor behavior (1). For example, as we have pointed earlier (1), dopamine may prove to be involved as well as norepinephrine. The drugs used in the present experiments—tetrabenazine and pargyline—reflect this and similar possibilities in that they affect a whole class of putative neurotransmitters, the monoamines, which includes norepinephrine, dopamine, and serotonin. The first experiment in the present paper determined whether an avoidanceescape deficit similar to that produced by stressful conditions could be produced by drug action. The motor activation deficit hypothesis predicts that pharmacologically-induced depletion of monoamines should produce an avoidance-escape deficit comparable to that observed after exposure to a severe stressor such as cold swim or inescapable shock. Therefore, the effects of a stressful 536
condition (cold swim) and pharmacological depletion of monoamines by tetrabenazine were compared on avoidanceescape responding in the shuttle task. A further prediction of this hypothesis is that the pharmacologically-induced deficit, like the stress-induced deficit, will be observed only when the avoidance-escape response requires a high level of motor activity. Therefore the effects of cold swim and tetrabenazine were also compared on the low activity "nosing" avoidanceescape task. The second experiment studied the effects of repeated exposure to a pharmacological agent that depletes brain monoamines. Such a procedure should produce neurochemical "habituation" by inducing higher synthesis of norepinephrine, as was seen with repeated exposures to inescapable shock (2), as well as increased synthesis of other monoamines. Thus, according to the motor activation deficit hypothesis, animals treated in this manner should subsequently show little, if any, avoidance-escape deficit, whereas animals given only a single injection of the monoamine depleter should, as stated in the previous paragraph, show the avoidance-escape deficit. In short, acute and repeated treatment with the monoamine depleter tetrabenazine should produce effects on avoidance-escape responding parallel to those produced by acute and repeated exposure to a severe stressor. The third experiment attempted to prevent a stress-induced avoidance-escape deficit by pharmacological intervention. Monoamines released at the presynaptic membrane are taken back up into the cell and destroyed by monoamine oxidase (MAO). This constitutes the major mechanism by which all monoamines are broken down (3). Therefore, by blocking
Psychosomatic Medicine Vol. 37, No. 6 (November-December 1975)
AVOIDANCE-ESCAPE BEHAVIOR
the activity of MAO, it should be possible to attenuate the depletion of monoamines that results from exposure to a stressor. If, as suggested by the motor activation deficit hypothesis, reduction of noradrenergic activity stemming at least in part from norepinephrine depletion is indeed responsible for avoidance-escape deficits observed after exposure to stress, then it might be possible to protect animals from the adverse behavioral effects of a stressor by the use of an MAO inhibitor. Finally, in presenting the motor activation deficit hypothesis, it was argued that it provided a more satisfactory explanation for an avoidance-escape deficit that appeared in dogs after severe inescapable shock than did another hypothesis which proposed that the animals had learned to be "helpless" as a result of the inescapable shock (4,5). In contrasting these two hypotheses, it should be noted that learned helplessness does not necessarily make any predictions about avoidance-escape behavior after manipulations of central monoamines, so that this hypothesis cannot be evaluated on the basis of the results described in this paper. However, very recently one of the proponents of the learned helplessness hypothesis has attempted to deal with the fact that severe inescapable shock depletes brain norepinephrine (6, pp. 67-74), suggesting that this may occur when an animal experiences that it is "helpless," or that it might even be one of the neurochemical events that occur centrally as a part of the "helpless" cognition. Unfortunately, such suggestions have the serious disadvantage, from a scientific point of view, of being untestable given our present inability to measure cognitions in nonverbal animals. But this objection aside, the present experiments provide evidence indicating that depletion of central monoamines is not the neurochem-
ical basis of a "helplessness" cognition. EXPERIMENT 1 This experiment examined the effects of cold swim and pharmacological depletion of monamines on avoidance-escape responding in both the shuttle task and the nosing task. Method Subjects. The subjects were naive male albino rats derived from a Sprague-Dawley strain, obtained from Holtzman Farms. Animals were 200-250 g in weight when received. They were individually housed and maintained on ad libitum food and water. Apparatus. The apparatus included a swim tank, a two-way shuttle box, and nosing apparatus described earlier by Weiss and Glazer (1). Procedure. The animals were divided into 6 groups of 6 subjects each. Two of the groups received a swim in cold water (2° C) for a duration of 3.5 min. Two of the groups received an injection (2 mg / kg body weight) of tetrabenazine (RO-1-9569/7), a fast-acting, reserpine-like agent that depletes brain monoamines (7). The last two groups received, in equal volume, a placebo injection of physiological saline. All animals were subjected to an avoidanceescape test 30 min after drug or placebo treatment. For the avoidance-escape test, one cold swim group (swim-shuttle test), one drug group (tetrabenazine-shuttle test), and one placebo group (placebo-shuttle test) was tested in the barrier-jump shuttle apparatus, and one of each type of group was tested in the nosing apparatus (swim-nosing test, tetrabenazine-nosing test, placebo-nosing test). Under both test conditions, a compound conditioned stimulus (CS) consisting of a tone (80 db, 1000 cps) and a click (1 /sec) initiating a trial. Five seconds after CS initiation, the tone was terminated, the click continued, and an unconditioned stimulus (1.25 mA electric shock) was initiated. If a subject responded during the CS, the CS was terminated and a 1-min intertrial interval (ITI) was initiated (avoidance). A response during the shock terminated the click and the shock and initiated a 1-min ITI (escape). If no response occurred after 20 sec of shock, the shock and click were terminated and the 1-min ITI was initiated.
Psychosomatic Medicine Vol. 37, No. 6 (November-December 1975)
537
HOWARD I. GLAZER, et al. In the barrier-jump apparatus, the shock was delivered through the bars of the floor, and jumping the barrier was designated as the instrumental response. In the nosing apparatus, the shock was delivered through a tail electrode, and the instrumental response was designated as a forward movement of the head with the nose protruding through the hole in the Plexiglas covering. During the ITI, the hole remained covered. For more detailed description, see Weiss and Glazer [1, p. 503, 515, this issue].
performed on the latency data. Main effects of conditions (F = 81.07,d/ = 2,30,p < .01) and test response (F = 300.40, df = 1, 30, p < .01) were found. Most notable is the condition x test response interaction (F = 62.91, df = 2, 30, p < .02), which indicates that both cold swim and tetrabenazine conditions (swim-shuttle test, tetrabenazine-shuttle test) showed an avoidance-escape deficit when the test reResults sponse was shuttle, but no deficit when the Figure 1 shows the average latency of all test response was nosing. Significant groups to perform the correct response. An within-group factors yielded by the analysis of variance, using conditions (tet- analysis include: trials (F = 4.83, df = 24, rabenazine, placebo, cold swim), test re- p < .01), condition x trials (F = 4.39,df = sponse (shuttle, nosing) and trials, was 48, 720, p < .01), test response x trials (F
25
20 CO
o
15
10
tetrabenazine- \ nosing test
\placebonosing test
0
I
Blocks of 5 trials Fig. 1. Mean latency (in sec) to perform the correct response in either the nosing or the shuttle avoidanceescape test, showing the effects of cold swim and tetrabenazine (compared with placebo control) on performance of either response. 538
Psychosomatic Medicine Vol. 37, No. 6 (November-December 1975)
AVOIDANCE-ESCAPE BEHAVIOR
= 3.81, d/ = 40,72, p < .01). As can be seen from Fig. 1, the three-way interaction indicates that animals given cold swim or tetrabenazine and tested in the shuttle response did not show any learning of this response relative to all the remaining groups, which did show learning (i.e., declining latencies).
EXPERIMENT 2
This experiment examined the effects of repeated drug depletion of monoamines on shuttle avoidance-escape performance shortly after inescapable shock or drug treatment. Method Subjects. The same kind of rats were used, and they received the same kind of housing and maintenance, as those in Experiment 1. Apparatus. The apparatus consisted of the shuttle box described earlier (1) and a restraining box for delivery of tail shock described by Weiss (8). Procedure. The subjects were divided into five groups of 6 animals each. Two of the groups were given a pretreatment consisting of a single injection of tetrabenazine (2 mg /kg body weight.) on each of 14 successive days. The remaining three groups received an equal volume of physiological saline on the same regimen. On the 15th day, one of the groups pretreated with repeated tetrabenazine (2 mg/kg body weight) followed 30 min later by a shuttle avoidance-escape test (tetrabenazine pretreatmenttetrabenazine before test) while one of the placebo groups received tetrabenazine and the avoidanceescape test (placebo pretreatment-tetrabenazine before test) On this 15th day, the second group pretreated with repeated tetrabenazine received a single session of inescapable shock, followed by the avoidance-escape test (tetrabenazine pretreatmentshock before test), and a second placebo group received the same treatment (placebo pretreatmentshock before test). The inescapable shock was 50 min in duration, and consisted of a 4.0 mA shock, 2 sec in duration, delivered every 20 sec through electrodes
attached to the tail. The remaining placebo group, which was essentially the "no treatment" control, simply received another placebo injection before the avoidance-escape test (placebo pretreatmentplacebo before test). The shuttle avoidance-escape task was carried out as described in Experiment 1.
Results Figure 2 shows the average latency for all groups to perform the shuttle response. A groups x trials analysis of variance performed on the latency data yielded significant effects of groups (F = 70.04, df = 4, 25, p < .01). Pairwise comparisons showed the placebo pretreatment—tetrabenazine before test group was slower (longer escape latencies) than the placebo pretreatment—shock before test group, which, in turn, was slower than all remaining groups (Newman-Keuls, p < .01 in all cases). A significant groups x trials interaction was also found (F = 2.85, df = 96, 600, p < .01). As can be seen from Fig. 2, the groups x trials interaction indicated that the placebo pretreatment —tetrabenazine before test group and the placebo pretreatment—shock before test group did not show any learning (i.e., declining latency over trials) whereas the remaining groups did.
EXPERIMENT 3
This experiment determined whether the debilitating effects of inescapable shock on shuttle avoidance-escape responding could be attenuated by pharmacological "protection" of monoamines. Procedure The same kind of rats were used, and they received the same kind of housing and maintenance, as those in Experiments 1 and 2; the same apparatus was used as in Experiment 2. The subjects were divided into
Psychosomatic Medicine Vol. 37, No. 6 (November-December 1975)
539
HOWARD I. GLAZER, et al. 25r placebo preireotment tetrabenazine before test
placebo pretrealmenlshock before test
tetrabenazine pretreatmenttetrabenazine before test
tetrabenazine pretreatment shock before test 0
1
2
3
4
B l o c k s of 5 t r i a l s Fig. 2. Mean latency (in sec) to cross barrier (correct response) in shuttle avoidance-escape test, showing the effects of repeated exposure (pretreatraent) to tetrabenazine (compared with placebo) before administration of either tetrabenazine or inescapable shock before the avoidance-escape test.
four groups of 6 subjects each, on the basis of a 2 x 2 factorial design. Two groups of subjects were given injections (100 ing/kg body weight) of pargyline hydrochloride, a monoamine oxidase (MAO) inhibitor; the subjects in the remaining two groups were injected with an equal volume of physiological saline (placebo). One group injected with the MAO inhibitor received a single session of inescapable shock beginning 5 min after the injection. The shock session lasted for 50 min and consisted of a 4.0 mA electric shock, 2 sec in duration, delivered every 20 sec through tail electrodes. Thirty minutes after the end of the shock session, these subjects were tested in the shuttle avoidance-escape apparatus (MAO inhibitor-shock). One of the groups that received the placebo injection similarly received shock before the shuttle test (placebo-shock). The second group injected with the MAO inhibitor received no shock before the shuttle test (MAO inhibitor-no shock), each subject simply remaining in its home cage for 80 min after receiving its injection before the avoidance-escape test. The second group that was injected with placebo was treated in the same way, except that it received an initial placebo injection 540
(placebo-no shock); this group was essentially the "no treatment" control.
Results Figure 3 shows average latency for all groups to perform the shuttle avoidanceescape response. A 3-way analysis of variance—drug treatment (MAO vs placebo) x condition (shock vs no shock) x trials—was performed on the latency data. The results indicated a main effect of the drug treatment (MAO inhibitor vs placebo,F = 34.85, df = l , 2 0 , p < .01) and condition (shock vs no shock, F = 86.63, df = 1, 20, p < .01). Most important, a significant interaction of these two factors was found (F = 105.00, df = 1,20, p < . 0 1 ) . As can be seen from Fig. 3, this interaction reflects the fact that MAO inhibitor given
Psychosomatic Medicine Vol. 37, No. 6 (November-December 1975)
AVOIDANCE-ESCAPE BEHAVIOR 25r
-S 20 o
e
placebo-shock
15
10
placebo-no shock
0
1
2
3
4
5
Blocks of 5 t r i a l s Fig. 3 Mean latency (in sec) to cross barrier in shuttle avoidance-escape test, showing the effect of administering an MAO inhibitor (pargyline) before administration of inescapable shock and subsequent avoidance-escape test.
before the inescapable shock session greatly attenuated the avoidance-escape deficit whereas placebo injections had no effect on avoidance-escape responding. The analysis also showed a significant trials effect [F = 1.93, d/ = 24, 48, p < .01). DISCUSSION The first experiment yielded two important findings that are in accord with the motor activation deficit hypothesis explained in the introduction. First, by pharmacologically depleting norepinephrine and other monoamines, an avoidanceescape deficit in the shuttle task was pro-
duced that was comparable to that produced by cold swim as well as by inescapable shock (1). Second, when animals were tested in a task requiring minimal motor activity, pharmacological depletion of monoamines did not produce a deficit relative to control animals, as was also found to be the case with cold swim and inescapable shock. The fact that, after tetrabenazine, animals were able to learn and perform the nosing avoidance-escape response appears to rule out the suggestion that norepinephrine depletion mediates the cognition of helplessness. If an animal experiences itself to be "helpless", it would be expected to show a deficit of some sort
Psychosomatic Medicine Vol. 37, No. 6 (November-December 1975)
541
HOWARD I. GLAZER, et al.
in any task requiring voluntary activity; since none was seen in the nosing test after a treatment which depletes central monoamines, it is suggested that depletion of monoamines does not mediate "helplessness". On the other hand, that animals learned the nosing response after tetrabenazine is quite consistent with the motor activation deficit hypothesis. This hypothesis does not state that exposure to a severe stressor will produce a deficit in learning capacity; it states only that exposure should produce a deficit in the ability to mediate active motor behavior. Thus, any response that requires little motor activity to perform should be learned, as was the case with the nosing task. The second experiment showed that repeated injections of tetrabenazine had the same effect as was found with repeated exposure to cold swim or inescapable shock; that is, repeated tetrabenazine caused shuttle avoidance-escape responding not to be impaired after another injection of tetrabenazine. Most important, repeated tetrabenazine also attenuated the effects of inescapable shock before the shuttle test. This result is important since the motor activation deficit hypothesis states that attenuation of the avoidanceescape deficit derives from the attenuation of norepinephrine depletion, so that any treatment that produces this should attenuate the avoidance-escape deficit. In agreement with this prediction, Weiss et al. (2) showed that repeated exposure to cold swim had the effect of reducing norepinephrine reuptake in response to a session of inescapable shock. Experiment 2 in this paper tested this prediction in another manner. In this study, multiple exposure to a stressor was replaced by multiple injections of the monoamine depleter tetrabenazine. As can be seen in Fig. 2, a series of tetrabenazine injections mar542
kedly attenuated the avoidance-escape deficit after inescapable shock as well as after another injection of tetrabenazine. The results support the notion that, if norepinephrine is depleted on a succession of occasions by a variety of means, the animal will build up a type of "immunity" to the norepinephrine-depleting effects of a stressor, even if the stressor is different from the agent used as the successive depleter. Experiment 3 represented another approach to protecting animals against the monoamine-depleting effects of stress. Intraneuronal oxidation of monoamines can be greatly reduced either by preventing the transmitter that is released from being taken back up or by blocking the action of monoamine oxidase. In the present experiment, the MAO inhibitor, pargyline hydrochloride, was used to prevent the oxidative metabolism of monoamines and thus decrease the norepinephrine depletion resulting from stress. As seen in Fig. 3, the placebo-shock group, receiving a single session of shock, showed an avoidanceescape deficit identical to that shown by the similarly treated group in the previous study. The remaining groups all showed much shorter latencies overall as well as learning over trials (declining latencies). Although the administration of pargyline HC1 prior to the inescapable shock session greatly attenuated the avoidance-escape deficit and animals treated in this way do show learning, the "immunization" was not complete. However, the fact that the group pretreated with pargyline alone (no shock) also showed slightly poorer performance than the control group suggests that the incomplete immunization might have been due to a side effect of pargyline rather than to its incomplete effectiveness in preventing monoamine depletion,
Psychosomatic Medicine Vol. 37, No. 6 (November-December 1975)
AVOIDANCE-ESCAPE BEHAVIOR
REFERENCES 1. 2. 3. 4. 5. 6. 7. 8.
Weiss JM, Glazer HI: Effects of acute exposure to stressors on subsequent avoidance-escape behavior. Psychosom Med 37:499-521, 1975, this issue. Weiss JM, Glazer HI, Pohorecky LA, Brick J, Miller NE: Effects of chronic exposure to stressors on avoidance-escape behavior and on brain norepinephrine. Psychosom Med 37:522-534,1975, this issue. Weiner N: Regulation of norepinephrine biosynthesis. Annu Rev Pharmacol 10:273-290, 1970 Overmier JB, Seligman MEP: Effects of inescapable shock upon subsequent escape and avoidance learning. J Comp Physiol Psychol 63: 23-33, 1967 Seligman MEP, Maier SF: Failure to escape traumatic shock. J Exp Psychol 74: 1-9, 1967 Seligman MEP: Helplessness. Freeman, San Francisco, 1975 Pletscher A: Significance of monamine oxidase inhibition for the pharmacological and clinical effects of hydrazine derivatives. Ann NY Acad Sci 80: 1039-1046, 1959 Weiss fM: Effects of coping behavior in different warning signal conditions on stress pathology in rats. J Comp Physiol Psychol 77:1-13, 1971
Psychosomatic Medicine Vol. 37, No. 6 (November-December 1975)
543
INTRODUCTION
presented indicating that these stressors produced this behavioral deficit by temThis paper is the third in a series study- porarily reducing neurotransmitter activing how exposure to severe stressors af- ity important for active motor behavior. fects behavioral responses. In two previ- The first paper described six studies which ous papers (1,2), it was shown that both showed, in essence, that exposure to cold brief swim in cold water and strong ines- swim and strong inescapable shock procapable shocks interfered with subsequent duced a transient avoidance-escape deficit avoidance-escape behavior. Evidence was that was highly sensitive to the amount of motor activity required in the avoidanceescape task, with the avoidance-escape This is the third paper in a series of three on the deficit being intensified, reduced, or even subject of avoidance-escape behavior. This series is eliminated completely depending on the published in its entirety in this issue. From The Rockefeller University, New York, NY amount of motor activity required of the 10021. animal in order to execute the correct reSupported by Public Health Service Grants AA sponse. The second paper, in testing the 00296 and MH 13189, and by grants from the Alfred P. Sloan, Grant and Mobil Foundations, the Merrill hypothesis that this stress-induced Trust, the Scottish Rite Schizophrenia Research Pro- avoidance-escape deficit was mediated by gram, and Hoffmann-La Roche. HIG held a fellowship noradrenergic depletion in the brain, from the Foundations' Fund for Research in Psychiatry during the course of the research and is showed that if one counteracted brain presently at Loyola Campus of Concordia University, noradrenergic depletion by repeatedly exMontreal, Quebec. posing animals to the inescapable shock or Received for publication 9 January, 1975; revision the cold swim, then the avoidance-escape received 9 June 1975. Psychosomatic Medicine Vol. 37, No. 6 (November-December 1975) Copyright ° 1975 by the American Psychosomatic Society, Inc. Published by American Elsevier Publishing Company, Inc.
535
HOWARD I. GLAZER, et al.
deficit also did not appear. The present paper now describes experiments which further test for neurochemical mediation of the avoidance-escape deficit by attempting to (a) produce the avoidance-escape deficit by using a drug treatment instead of a stressor, and (b) eliminate the stressinduced avoidance-escape deficit by appropriate pharmacological intervention. Before describing the experiments which were undertaken, it is important to comment on the hypothesis that strong stressors, such as severe inescapable shock and cold swim, produce an avoidanceescape deficit by reducing central neurotransmitter activity important for active motor behavior. This hypothesis, called the "motor activation deficit" hypothesis, has indeed focused on the role of norepinephrine, but it is in principle more broadly based, being linked to those neurotransmitters which are (a) affected by stressful conditions, and (b) mediate active motor behavior (1). For example, as we have pointed earlier (1), dopamine may prove to be involved as well as norepinephrine. The drugs used in the present experiments—tetrabenazine and pargyline—reflect this and similar possibilities in that they affect a whole class of putative neurotransmitters, the monoamines, which includes norepinephrine, dopamine, and serotonin. The first experiment in the present paper determined whether an avoidanceescape deficit similar to that produced by stressful conditions could be produced by drug action. The motor activation deficit hypothesis predicts that pharmacologically-induced depletion of monoamines should produce an avoidance-escape deficit comparable to that observed after exposure to a severe stressor such as cold swim or inescapable shock. Therefore, the effects of a stressful 536
condition (cold swim) and pharmacological depletion of monoamines by tetrabenazine were compared on avoidanceescape responding in the shuttle task. A further prediction of this hypothesis is that the pharmacologically-induced deficit, like the stress-induced deficit, will be observed only when the avoidance-escape response requires a high level of motor activity. Therefore the effects of cold swim and tetrabenazine were also compared on the low activity "nosing" avoidanceescape task. The second experiment studied the effects of repeated exposure to a pharmacological agent that depletes brain monoamines. Such a procedure should produce neurochemical "habituation" by inducing higher synthesis of norepinephrine, as was seen with repeated exposures to inescapable shock (2), as well as increased synthesis of other monoamines. Thus, according to the motor activation deficit hypothesis, animals treated in this manner should subsequently show little, if any, avoidance-escape deficit, whereas animals given only a single injection of the monoamine depleter should, as stated in the previous paragraph, show the avoidance-escape deficit. In short, acute and repeated treatment with the monoamine depleter tetrabenazine should produce effects on avoidance-escape responding parallel to those produced by acute and repeated exposure to a severe stressor. The third experiment attempted to prevent a stress-induced avoidance-escape deficit by pharmacological intervention. Monoamines released at the presynaptic membrane are taken back up into the cell and destroyed by monoamine oxidase (MAO). This constitutes the major mechanism by which all monoamines are broken down (3). Therefore, by blocking
Psychosomatic Medicine Vol. 37, No. 6 (November-December 1975)
AVOIDANCE-ESCAPE BEHAVIOR
the activity of MAO, it should be possible to attenuate the depletion of monoamines that results from exposure to a stressor. If, as suggested by the motor activation deficit hypothesis, reduction of noradrenergic activity stemming at least in part from norepinephrine depletion is indeed responsible for avoidance-escape deficits observed after exposure to stress, then it might be possible to protect animals from the adverse behavioral effects of a stressor by the use of an MAO inhibitor. Finally, in presenting the motor activation deficit hypothesis, it was argued that it provided a more satisfactory explanation for an avoidance-escape deficit that appeared in dogs after severe inescapable shock than did another hypothesis which proposed that the animals had learned to be "helpless" as a result of the inescapable shock (4,5). In contrasting these two hypotheses, it should be noted that learned helplessness does not necessarily make any predictions about avoidance-escape behavior after manipulations of central monoamines, so that this hypothesis cannot be evaluated on the basis of the results described in this paper. However, very recently one of the proponents of the learned helplessness hypothesis has attempted to deal with the fact that severe inescapable shock depletes brain norepinephrine (6, pp. 67-74), suggesting that this may occur when an animal experiences that it is "helpless," or that it might even be one of the neurochemical events that occur centrally as a part of the "helpless" cognition. Unfortunately, such suggestions have the serious disadvantage, from a scientific point of view, of being untestable given our present inability to measure cognitions in nonverbal animals. But this objection aside, the present experiments provide evidence indicating that depletion of central monoamines is not the neurochem-
ical basis of a "helplessness" cognition. EXPERIMENT 1 This experiment examined the effects of cold swim and pharmacological depletion of monamines on avoidance-escape responding in both the shuttle task and the nosing task. Method Subjects. The subjects were naive male albino rats derived from a Sprague-Dawley strain, obtained from Holtzman Farms. Animals were 200-250 g in weight when received. They were individually housed and maintained on ad libitum food and water. Apparatus. The apparatus included a swim tank, a two-way shuttle box, and nosing apparatus described earlier by Weiss and Glazer (1). Procedure. The animals were divided into 6 groups of 6 subjects each. Two of the groups received a swim in cold water (2° C) for a duration of 3.5 min. Two of the groups received an injection (2 mg / kg body weight) of tetrabenazine (RO-1-9569/7), a fast-acting, reserpine-like agent that depletes brain monoamines (7). The last two groups received, in equal volume, a placebo injection of physiological saline. All animals were subjected to an avoidanceescape test 30 min after drug or placebo treatment. For the avoidance-escape test, one cold swim group (swim-shuttle test), one drug group (tetrabenazine-shuttle test), and one placebo group (placebo-shuttle test) was tested in the barrier-jump shuttle apparatus, and one of each type of group was tested in the nosing apparatus (swim-nosing test, tetrabenazine-nosing test, placebo-nosing test). Under both test conditions, a compound conditioned stimulus (CS) consisting of a tone (80 db, 1000 cps) and a click (1 /sec) initiating a trial. Five seconds after CS initiation, the tone was terminated, the click continued, and an unconditioned stimulus (1.25 mA electric shock) was initiated. If a subject responded during the CS, the CS was terminated and a 1-min intertrial interval (ITI) was initiated (avoidance). A response during the shock terminated the click and the shock and initiated a 1-min ITI (escape). If no response occurred after 20 sec of shock, the shock and click were terminated and the 1-min ITI was initiated.
Psychosomatic Medicine Vol. 37, No. 6 (November-December 1975)
537
HOWARD I. GLAZER, et al. In the barrier-jump apparatus, the shock was delivered through the bars of the floor, and jumping the barrier was designated as the instrumental response. In the nosing apparatus, the shock was delivered through a tail electrode, and the instrumental response was designated as a forward movement of the head with the nose protruding through the hole in the Plexiglas covering. During the ITI, the hole remained covered. For more detailed description, see Weiss and Glazer [1, p. 503, 515, this issue].
performed on the latency data. Main effects of conditions (F = 81.07,d/ = 2,30,p < .01) and test response (F = 300.40, df = 1, 30, p < .01) were found. Most notable is the condition x test response interaction (F = 62.91, df = 2, 30, p < .02), which indicates that both cold swim and tetrabenazine conditions (swim-shuttle test, tetrabenazine-shuttle test) showed an avoidance-escape deficit when the test reResults sponse was shuttle, but no deficit when the Figure 1 shows the average latency of all test response was nosing. Significant groups to perform the correct response. An within-group factors yielded by the analysis of variance, using conditions (tet- analysis include: trials (F = 4.83, df = 24, rabenazine, placebo, cold swim), test re- p < .01), condition x trials (F = 4.39,df = sponse (shuttle, nosing) and trials, was 48, 720, p < .01), test response x trials (F
25
20 CO
o
15
10
tetrabenazine- \ nosing test
\placebonosing test
0
I
Blocks of 5 trials Fig. 1. Mean latency (in sec) to perform the correct response in either the nosing or the shuttle avoidanceescape test, showing the effects of cold swim and tetrabenazine (compared with placebo control) on performance of either response. 538
Psychosomatic Medicine Vol. 37, No. 6 (November-December 1975)
AVOIDANCE-ESCAPE BEHAVIOR
= 3.81, d/ = 40,72, p < .01). As can be seen from Fig. 1, the three-way interaction indicates that animals given cold swim or tetrabenazine and tested in the shuttle response did not show any learning of this response relative to all the remaining groups, which did show learning (i.e., declining latencies).
EXPERIMENT 2
This experiment examined the effects of repeated drug depletion of monoamines on shuttle avoidance-escape performance shortly after inescapable shock or drug treatment. Method Subjects. The same kind of rats were used, and they received the same kind of housing and maintenance, as those in Experiment 1. Apparatus. The apparatus consisted of the shuttle box described earlier (1) and a restraining box for delivery of tail shock described by Weiss (8). Procedure. The subjects were divided into five groups of 6 animals each. Two of the groups were given a pretreatment consisting of a single injection of tetrabenazine (2 mg /kg body weight.) on each of 14 successive days. The remaining three groups received an equal volume of physiological saline on the same regimen. On the 15th day, one of the groups pretreated with repeated tetrabenazine (2 mg/kg body weight) followed 30 min later by a shuttle avoidance-escape test (tetrabenazine pretreatmenttetrabenazine before test) while one of the placebo groups received tetrabenazine and the avoidanceescape test (placebo pretreatment-tetrabenazine before test) On this 15th day, the second group pretreated with repeated tetrabenazine received a single session of inescapable shock, followed by the avoidance-escape test (tetrabenazine pretreatmentshock before test), and a second placebo group received the same treatment (placebo pretreatmentshock before test). The inescapable shock was 50 min in duration, and consisted of a 4.0 mA shock, 2 sec in duration, delivered every 20 sec through electrodes
attached to the tail. The remaining placebo group, which was essentially the "no treatment" control, simply received another placebo injection before the avoidance-escape test (placebo pretreatmentplacebo before test). The shuttle avoidance-escape task was carried out as described in Experiment 1.
Results Figure 2 shows the average latency for all groups to perform the shuttle response. A groups x trials analysis of variance performed on the latency data yielded significant effects of groups (F = 70.04, df = 4, 25, p < .01). Pairwise comparisons showed the placebo pretreatment—tetrabenazine before test group was slower (longer escape latencies) than the placebo pretreatment—shock before test group, which, in turn, was slower than all remaining groups (Newman-Keuls, p < .01 in all cases). A significant groups x trials interaction was also found (F = 2.85, df = 96, 600, p < .01). As can be seen from Fig. 2, the groups x trials interaction indicated that the placebo pretreatment —tetrabenazine before test group and the placebo pretreatment—shock before test group did not show any learning (i.e., declining latency over trials) whereas the remaining groups did.
EXPERIMENT 3
This experiment determined whether the debilitating effects of inescapable shock on shuttle avoidance-escape responding could be attenuated by pharmacological "protection" of monoamines. Procedure The same kind of rats were used, and they received the same kind of housing and maintenance, as those in Experiments 1 and 2; the same apparatus was used as in Experiment 2. The subjects were divided into
Psychosomatic Medicine Vol. 37, No. 6 (November-December 1975)
539
HOWARD I. GLAZER, et al. 25r placebo preireotment tetrabenazine before test
placebo pretrealmenlshock before test
tetrabenazine pretreatmenttetrabenazine before test
tetrabenazine pretreatment shock before test 0
1
2
3
4
B l o c k s of 5 t r i a l s Fig. 2. Mean latency (in sec) to cross barrier (correct response) in shuttle avoidance-escape test, showing the effects of repeated exposure (pretreatraent) to tetrabenazine (compared with placebo) before administration of either tetrabenazine or inescapable shock before the avoidance-escape test.
four groups of 6 subjects each, on the basis of a 2 x 2 factorial design. Two groups of subjects were given injections (100 ing/kg body weight) of pargyline hydrochloride, a monoamine oxidase (MAO) inhibitor; the subjects in the remaining two groups were injected with an equal volume of physiological saline (placebo). One group injected with the MAO inhibitor received a single session of inescapable shock beginning 5 min after the injection. The shock session lasted for 50 min and consisted of a 4.0 mA electric shock, 2 sec in duration, delivered every 20 sec through tail electrodes. Thirty minutes after the end of the shock session, these subjects were tested in the shuttle avoidance-escape apparatus (MAO inhibitor-shock). One of the groups that received the placebo injection similarly received shock before the shuttle test (placebo-shock). The second group injected with the MAO inhibitor received no shock before the shuttle test (MAO inhibitor-no shock), each subject simply remaining in its home cage for 80 min after receiving its injection before the avoidance-escape test. The second group that was injected with placebo was treated in the same way, except that it received an initial placebo injection 540
(placebo-no shock); this group was essentially the "no treatment" control.
Results Figure 3 shows average latency for all groups to perform the shuttle avoidanceescape response. A 3-way analysis of variance—drug treatment (MAO vs placebo) x condition (shock vs no shock) x trials—was performed on the latency data. The results indicated a main effect of the drug treatment (MAO inhibitor vs placebo,F = 34.85, df = l , 2 0 , p < .01) and condition (shock vs no shock, F = 86.63, df = 1, 20, p < .01). Most important, a significant interaction of these two factors was found (F = 105.00, df = 1,20, p < . 0 1 ) . As can be seen from Fig. 3, this interaction reflects the fact that MAO inhibitor given
Psychosomatic Medicine Vol. 37, No. 6 (November-December 1975)
AVOIDANCE-ESCAPE BEHAVIOR 25r
-S 20 o
e
placebo-shock
15
10
placebo-no shock
0
1
2
3
4
5
Blocks of 5 t r i a l s Fig. 3 Mean latency (in sec) to cross barrier in shuttle avoidance-escape test, showing the effect of administering an MAO inhibitor (pargyline) before administration of inescapable shock and subsequent avoidance-escape test.
before the inescapable shock session greatly attenuated the avoidance-escape deficit whereas placebo injections had no effect on avoidance-escape responding. The analysis also showed a significant trials effect [F = 1.93, d/ = 24, 48, p < .01). DISCUSSION The first experiment yielded two important findings that are in accord with the motor activation deficit hypothesis explained in the introduction. First, by pharmacologically depleting norepinephrine and other monoamines, an avoidanceescape deficit in the shuttle task was pro-
duced that was comparable to that produced by cold swim as well as by inescapable shock (1). Second, when animals were tested in a task requiring minimal motor activity, pharmacological depletion of monoamines did not produce a deficit relative to control animals, as was also found to be the case with cold swim and inescapable shock. The fact that, after tetrabenazine, animals were able to learn and perform the nosing avoidance-escape response appears to rule out the suggestion that norepinephrine depletion mediates the cognition of helplessness. If an animal experiences itself to be "helpless", it would be expected to show a deficit of some sort
Psychosomatic Medicine Vol. 37, No. 6 (November-December 1975)
541
HOWARD I. GLAZER, et al.
in any task requiring voluntary activity; since none was seen in the nosing test after a treatment which depletes central monoamines, it is suggested that depletion of monoamines does not mediate "helplessness". On the other hand, that animals learned the nosing response after tetrabenazine is quite consistent with the motor activation deficit hypothesis. This hypothesis does not state that exposure to a severe stressor will produce a deficit in learning capacity; it states only that exposure should produce a deficit in the ability to mediate active motor behavior. Thus, any response that requires little motor activity to perform should be learned, as was the case with the nosing task. The second experiment showed that repeated injections of tetrabenazine had the same effect as was found with repeated exposure to cold swim or inescapable shock; that is, repeated tetrabenazine caused shuttle avoidance-escape responding not to be impaired after another injection of tetrabenazine. Most important, repeated tetrabenazine also attenuated the effects of inescapable shock before the shuttle test. This result is important since the motor activation deficit hypothesis states that attenuation of the avoidanceescape deficit derives from the attenuation of norepinephrine depletion, so that any treatment that produces this should attenuate the avoidance-escape deficit. In agreement with this prediction, Weiss et al. (2) showed that repeated exposure to cold swim had the effect of reducing norepinephrine reuptake in response to a session of inescapable shock. Experiment 2 in this paper tested this prediction in another manner. In this study, multiple exposure to a stressor was replaced by multiple injections of the monoamine depleter tetrabenazine. As can be seen in Fig. 2, a series of tetrabenazine injections mar542
kedly attenuated the avoidance-escape deficit after inescapable shock as well as after another injection of tetrabenazine. The results support the notion that, if norepinephrine is depleted on a succession of occasions by a variety of means, the animal will build up a type of "immunity" to the norepinephrine-depleting effects of a stressor, even if the stressor is different from the agent used as the successive depleter. Experiment 3 represented another approach to protecting animals against the monoamine-depleting effects of stress. Intraneuronal oxidation of monoamines can be greatly reduced either by preventing the transmitter that is released from being taken back up or by blocking the action of monoamine oxidase. In the present experiment, the MAO inhibitor, pargyline hydrochloride, was used to prevent the oxidative metabolism of monoamines and thus decrease the norepinephrine depletion resulting from stress. As seen in Fig. 3, the placebo-shock group, receiving a single session of shock, showed an avoidanceescape deficit identical to that shown by the similarly treated group in the previous study. The remaining groups all showed much shorter latencies overall as well as learning over trials (declining latencies). Although the administration of pargyline HC1 prior to the inescapable shock session greatly attenuated the avoidance-escape deficit and animals treated in this way do show learning, the "immunization" was not complete. However, the fact that the group pretreated with pargyline alone (no shock) also showed slightly poorer performance than the control group suggests that the incomplete immunization might have been due to a side effect of pargyline rather than to its incomplete effectiveness in preventing monoamine depletion,
Psychosomatic Medicine Vol. 37, No. 6 (November-December 1975)
AVOIDANCE-ESCAPE BEHAVIOR
REFERENCES 1. 2. 3. 4. 5. 6. 7. 8.
Weiss JM, Glazer HI: Effects of acute exposure to stressors on subsequent avoidance-escape behavior. Psychosom Med 37:499-521, 1975, this issue. Weiss JM, Glazer HI, Pohorecky LA, Brick J, Miller NE: Effects of chronic exposure to stressors on avoidance-escape behavior and on brain norepinephrine. Psychosom Med 37:522-534,1975, this issue. Weiner N: Regulation of norepinephrine biosynthesis. Annu Rev Pharmacol 10:273-290, 1970 Overmier JB, Seligman MEP: Effects of inescapable shock upon subsequent escape and avoidance learning. J Comp Physiol Psychol 63: 23-33, 1967 Seligman MEP, Maier SF: Failure to escape traumatic shock. J Exp Psychol 74: 1-9, 1967 Seligman MEP: Helplessness. Freeman, San Francisco, 1975 Pletscher A: Significance of monamine oxidase inhibition for the pharmacological and clinical effects of hydrazine derivatives. Ann NY Acad Sci 80: 1039-1046, 1959 Weiss fM: Effects of coping behavior in different warning signal conditions on stress pathology in rats. J Comp Physiol Psychol 77:1-13, 1971
Psychosomatic Medicine Vol. 37, No. 6 (November-December 1975)
543
