Morphine and GABA: Effects on perception, escape response and long-term habituation to a danger stimulus in the crab Chasmagnathus

Brain Research Bulletin, Vol. 26, pp. 699-706. Q Pergamon Press plc, 1991. Printed in the U.S.A.

0361.9230/91 $3.00 + .oO

Morphine and GABA: Effects on Perception, Escape Response and Long-Term Habituation to a Danger Stimulus in the Crab Chasmagnathus DANIEL

TOMSIC,

HECTOR

MALDONADO’

AND ANA RAKITIN

Laboratorio de Fisiologia de1 Comportamiento Animal, Facultad de Ciencias Exactas y Naturales UBA, Buenos Aires, Argentina Received

5 September

1990

TOMSIC, D., H. MALDONADO AND A. RAKITIN. Morphine and GABA: Effects on perception, escape response and longterm habituafion to a danger stimulus in the crab Chasmagnathus. BRAIN RES BULL 26(5) 699-706, 1991.-Prior results (37) showed that morphine pretreatment reduces the escape response of the crab Chasmqnathus to a shadow passing overhead and prevents the acquisition of a long-term habituation. These results were explained by a reduction in the danger signalled by the stimulus, and to test this hypothesis methods other than morphine injection were used herein to abolish response during training. GABA pretreatment induced a dose-dependent reduction in responsiveness to the danger stimulus, and instances of autotomy were shown with doses larger than 12 kg/g. A response was rarely displayed with a 9 p,g GABAIg dose given 5 min before training, but long-term memory was acquired. In one experiment, both morphine and GABA pretreatment produced similar mild response inhibition during training, but morphine, not GABA impaired long-term habituation. Morphine administered immediately after training had no amnesic effect. These results support the hypothesis that morphine effects may be explained by transient disruption between the stimulus and its danger meaning, ruling out alternative explanations such as response inhibition or amnesia due to either storage or retrieval failure. Habituation

Response-independent

habituation

Morphine

that habituates after repeated stimulation (9). Morphine produced a dose-dependent naloxone-reversible reduction in this response (30). In contrast, doses of morphine with strong inhibitory effect on the escape response (75-100 pg/g) failed to affect the optokinetic response (42), suggesting that the response decrement to a danger stimulus is stimulus-specific, and that morphine acts centrally rather than on visual input and/or motor ability. Two alternative hypotheses may account for the putative central action of morphine on the crab’s escape response to a shadow passing overhead. As previously advanced (30), the morphine effect could result from an interference with the decoding of the danger signal, namely by a reduction in the magnitude of the danger signal associated with the stimulus. An alternative proposal is that morphine raises the response threshold specifically to a danger stimulus. Should the Fist hypothesis be correct, repeated shadow stimulation to preinjected crabs would be expected not to produce long-term (between-session) habituation, since a proper danger stimulation was impaired during training. In contrast, if the second hypothesis is correct, long-term habituation is expected since a meaningful stimulus to H. Maldonado,

Danger stimulus

Crustacea

was present despite a low response level during training. Tomsic and Maldonado (42) demonstrated that morphine pretreatment sharply reduces reactivity during training and impairs the acquisition of long-term habituation, a result that they interpreted as supporting the Fist hypothesis. However, the failure to acquire a long-term habituated response might be accounted for by the impairment of responsiveness itself and not by an impairment in perception. This argument the authors contested on theoretical grounds, recalling that habituation studies support the belief that the stimulus rather than the response is the critical factor in this learning process (4, 11, 13, 34). To decide the issue on experimental grounds, a promising approach would be to compare the morphine effect on long-term habituation to a danger stimulus, with that obtained by using a drug that, without impairing vision, inhibits the escape response regardless of the type of visual stimulus presented, i.e., a drug acting on the efferent limb. For this purpose, GABA appears to be a good candidate, since it is well established as the transmitter substance released by crustacean inhibitory motor axons (5). Accordingly, present experiments are aimed at making such a comparison by testing the effect of GABA pretreatment both on the escape response level to a repeated danger stimulation and on the long-term habituation to the iterated stimulus.

WHEN a passing shadow (a danger stimulus) is presented to the crab Chasmagnathus granulatus, an escape response is elicited

‘Requests for reprints should be addressed taria, (1428) Buenos Aires, Argentina.

GABA

Laboratorio

699

Fisiologia

Comportamiento,

Depto Biologia,

Pabellon 2, Ciudad Universi-

700

TOMSIC, MALDONADO

GENERAL METHOD Animals The animals were adult male Chasmagnathus crabs, 2.8-3.0 cm across the carapace, collected from water less than 1 m deep in the rias (narrow coastal inlets) of San Clemente de1 Tuyu, Argentina, and transported to the laboratory, where they were lodged in plastic tanks (35 x 48 X 27 cm) filled to 2 cm depth with water without aeration, at a density of 35 crabs per tank. Water used in tanks and other containers during experiments was collected from the same place animals were captured (salinity lo-14%0, pH 7.4-7.6). The holding room was maintained on a 12-h light-dark cycle (lights on 0700-1900 h). Animals were fed rabbit pellets (Nutrientes SA) every 3 days and after feeding the water was changed. Temperature of both holding and experimental rooms as well as the alley between them was maintained within a range of 19-24°C. Experiments were conducted during daylight between the 2nd and 6th day after animals’ arrival, except when the response level failed to reach the acceptance criterion (see below) since in such case animals were kept for a previous adaptation period (3 or 4 days) in their home tanks. Each crab was used in only one experiment. Mean crab weight was determined as described elsewhere (9) (17.3 g, SE 0.2, n = 60) and absolute drug doses calculated according to this mean. Chasmagnuthus can be captured during the whole year except for the few coldest winter days (water temperature below 10°C). The level of response to a passing shadow is usually lower in animals coming from capture efforts corresponding to winter or the beginning of spring. Experiments of the present study were performed between October and May (i.e., late spring, summer and fall). Apparatus The apparatus used is described in detail elsewhere (39). Briefly, the experimental unit was the uctometer: a bowl-shaped plastic container with steep concave walls and a circular central flat floor 10 cm in diameter, covered to a depth of 0.5 cm with water. The crab was lodged in the container which was suspended by three strings from an upper wooden framework (23 x 23 x 30 cm) and illuminated by a 10-W lamp placed 30 cm above the animal. An opaque rectangle screen (25 x 13 cm) could be moved horizontally across the upper border of the framework by a motor at an angular speed which allowed it to cover the entire opening in 2.3 s, thereby projecting a shadow on the crab. Screen displacements provoked a crab’s running response and consequently container oscillations. A stylus was centrally cemented to the bottom of the container and connected to a piezoelectric transducer. Container oscillations induced, through the transducer, electrical signals proportional to the velocity of the oscillations (10). Such signals were amplified, integrated during the recording time (9 s) and translated into numerical units ranging from zero to 1020, before being processed by computer. Thus the scores were correlated proportionally to the velocity and number of the container oscillations recorded during 9 s. The amplification of the voltage changes was kept at such a gain that scores remained below 1020. The same activity unit and amplification gain has been used in all recent experiments at our laboratory [e.g., (28, 29, 40, 42)]. The experimental room had 40 actometers, isolated from each other by lateral partitions and a frontal wall. In order to avoid unobserved malfunctioning the actometers were periodically calibrated against one another by throwing a small lead ball from the upper border of the framework to the center of the

AND RAKITIN

container and recording the score for 9 s. A noticeable mity of scores was obtained (e.g.. a mean of 513 i9.48 40 actometers).

uniforfor the

Experimental Procedure A stimulation session comprised 15 or 3 trials given with 180-s intertrial intervals and was preceded by 30 min of adaptation in the actometer. Each trial lasted 9 s and consisted of passing the screen 4 times over the actometer, recording the crab’s activity during the entire trial time. Short-term experiments included a single 15-trial session, and long-term experiments two sessions, a first one of 15 trials and a second one of 3 trials, separated by a rest interval. When 2 sessions were run, crabs were individually housed in plastic containers both for 24 h before the first session and for the rest interval. The water in the plastic containers was changed daily and that in the actometers just before each experiment. Throughout this paper, either the single session of a short-term experiment or the first session of a long-term experiment is called training session, whereas the second session is termed testing session. A computer was employed to program trial sequences, trial duration and intertrial intervals, as well as to monitor experimental events. Since the number of actometers was insufficient to run all groups of each experiment simultaneously, replications during the same day were necessary. An equal number of crabs per group was used in each replication, but animals of a same group were placed in different actometers each time. Thus any potential effect of time of day and/or between-actometer differences was offset. The crab’s baseline responsiveness to the passing shadow proved remarkably consistent up to 10 days after arrival, but on occasion animals coming from different capture efforts presented differences in response level. Therefore, only crabs belonging to a same capture were used in each experiment, and in addition, an acceptance test aimed at ensuring a minimum level of responsiveness was employed. During the 2nd day after arrival and prior to any experimental use, one group of 40 crabs was given a 3-trial session. If the mean response obtained by averaging accumulated scores for the 3-trial session proved less than 1000, a second acceptance test was done 3-4 days later and if scores again failed to meet the criterion, no experiments were performed with crabs coming from such capture effort. Before animals were placed in the actometers or in the rest containers to start an experiment, they underwent a selection test: each crab was turned on its back and only animals that immediately returned to their normal position were used. The rationale behind this selection is that crabs with a slow righting reaction show a low responsiveness to a large diversity of stimuli, and at a later time, they usually present unhealthy symptoms. No more than 10% of tested crabs were eliminated. Injections Animals were injected by means of a Hamilton syringe through the right side of the cephalothoracic-abdominal membrane, i.e., the right metabranchial region. A small rubber ring placed in the needle 4 mm from the tip acted as a stop, ensuring the injected solution was released roughly at the center of the pericardial sac. Unlike the method used in other work at our laboratory, where injections were given through a needle chronically implanted in the carapace (9,30), the present technique allows animals to be readily moved from the actometers to be housed in individual containers during the intersession interval.

MORPHINE AND GABA: DANGER STIMULUS

IN CRAB

701

Injections consisted of 100 pl of the vehicle (NaCl, 1.6 %) or a drug solution. Morphine-HCl was purchased from SaporitiArgentina and gamma-aminobutyric acid from Sigma Chemicals. USA. Statistics The data from each experiment were analyzed with a oneway analysis of variance (ANOVA) followed by Duncan multiple range tests. Description of the Escape Response in the Actometer The escape response in the actometer consists of the crab starting to run in an attempt to move away from the passing shadow. However, since the steep concavity of the walls prevents the animal from climbing up, each running effort is confined to the flat center of the container, in such a way that the escape response during a single trial looks like a series of flights from the center toward the base of the walls. The extension and speed of each flight as well as its frequency vanish over each 9-s trial, and the total values per trial of these parameters also decrease over training. It is apparent that the lower the strength of an escape response, the lower the velocity and number of container oscillations during the 9-s trial. Therefore, a roughly linear correlation between response and score is expected since, as above mentioned, the electrical signal produced by the piezoelectric transducer is in turn proportional to the velocity and number of oscillations (10). This prediction is borne out by results from a test often performed in our laboratory in which the intensity of the escape response is scored by an observer as belonging to one of three categories (low, medium and high) and then compared with the score actually recorded. Values higher than 700 corresponded to responses classified as high responses; lower than 300 to low responses; and between 200 and 800, though with a noticeable preponderance of 400-600, to medium responses. The test showed higher interobserver reliability. Container movements caused by crab’s response to stimulation last less than 9 s, namely, less than the trial time. A spontaneous exploratory behavior consisting of slow displacements occurs mainly during the first minutes of the adaptation time and sporadically during intertrial intervals. Indeed, during most of the intertrial intervals the crab remains motionless at the center of the actometer. Thus no escape response is displayed either during the adaptation time or during intertrial intervals, and vice versa, no exploratory activity is shown during the trial time. During both training and testing session, no changes in behavior other than those related with the strength of the escape response, are observed.

0.0

0.06

0.6

60.0

1. GABA administration produces a dose-dependent reduction in responsiveness to a danger stimulus. Ordinates: average of accumulated FIG.

scores pe.r animal for 15 trials. Asterisks stand for significant differences between the saline control group and each one of the other groups (Duncan test) *p
trials. GABA produced dose-dependent reduction in the crab’s reactivity to the danger stimulus. The ANOVA on these data resulted in a significant treatment mean square, F(4,155)=4.4, p
EXPERIMENT

6.0

GABA pWG

2

1

The purpose of this experiment was to study the effect of GABA on the escape response to a danger stimulus. METHOD

One hundred and sixty crabs were injected 30 min before the first trial of a 1Wrial training session. Five groups of 32 crabs each were randomly formed: each group received 5 different GABA doses: 0.0, 0.06, 0.6, 6.0, and 60.0 pglg, respectively. Groups were run simultaneously. RESULTS AND DISCUSSION

Figure 1 presents the mean response score for each group, obtained by averaging accumulated scores per animal for the 15

This experiment was aimed at obtaining a drastic reduction in responsiveness during training by means of GABA administration, in order to test the effect of such absence of reactivity to the danger stimulus, on the acquisition of a long-term habituation. METHOD

During pilots experiments a 9 kg/g dose of GABA administered 5 min before training was found to induce a clearcut decrement in reactivity to the passing shadow while producing few instances of autotomy. Additional pilot experiments on animals injected with 9 Fg GABA/g and tested several hours later with a single 15-trial session showed that the inhibitory effect of the

702

TOMSIC,

SALGABA

2 0e 1

3

5

7

9

11

13

15

TRIALS

GAB&O

GABA-TR

GROUPS

MALDONADO

group did not respond during the entire training session. Figure 2B presents the mean response score for each group during the testing session. Three points prove apparent from an inspection of this figure. First, after a 4-h 30-min rest interval, the inhibitory effect of GABA pretreatment disappears, i.e., the response level of GABA-O during testing is similar to that of SALINE-O. Second, the response level of SALINE-TR is, as usual [e.g., (28, 29, 40, 42)], lower than that of SALINE-O, suggesting that habituation formed during training persists after the rest interval. Third, the response level of GABA-TR during testing is lower than that of GABA-0, suggesting that a longterm habituation is acquired during training, in spite of the fact that responding is almost completely suppressed. Accordingly, the ANOVA on data from testing gave, F(3,116)=6.97. p
FIG. 2. GABA administration

inhibits response to the danger stimulus during training without impairing long-term habituation. (A) Results of training session. Ordinates: mean response score per trial. Open circles: group injected with saline 5 min before training. Closed circles: group injected with GABA (9 pg/g) 5 min before training. Inset: average of accumulative scores for the 15 training trials per group. (B) Results of testing session. Ordinates: average of accumulative scores for the 3 testing trials per group. SAL-O: no training after saline injection. SAL-TR: training after saline injection. GABA-0: no training after GABA injection. GABA-TR: training after GABA injection. **Significant difference (Duncan test, p
drug had disappeared after 4 h and 30 min. Therefore, in this experiment animals underwent a 15trial training session followed 4 h and 30 min later by a 3-trial testing session, as in prior experiments with morphine (37). Saline or GABA solutions were injected 5 min before the 1st training trial. One hundred and twenty crabs were randomly assigned in equal numbers to a each cell in a 2 x 2 factorial design; the factors being drug (SALINE: the vehicle, or GABA: 9 p&g), and training (0: no training session, or TR: training session), so that 4 groups were formed: SALINE-O, SALINE-TR, GABA-O and GABA-TR. Groups were run simultaneously. Zero-groups remained in the actometers during the time corresponding to the training session (75 min), but without being stimulated with the passing shadow. RESULTS AND DISCUSSION Figure 2A shows the curves of responding score vs. trials for the SALINE-TR and GABA-TR groups during the training session. GABA markedly depressed response scores, as depicted in the inset of Fig. 2A where the averages of accumulative scores for 1.5 trials are shown. Indeed, several crabs of the GABA-TR

AND RAKITIN

3

Prior results from our laboratory (28) suggest that a 15-trial session is the minimum training necessary to ensure a 24-h habituation. Consistently, pilot experiments show that when the training session is reduced by a few trials, instances of nonretention at testing are observed. Such an effect may be explained, according to the above hypothesis, by a reduction in the amount of stimulation rather than by a decrease in the amount of responding, so that a small depressing effect of morphine on the response level should be sufficient to diminish retention. Therefore, this experiment was conducted to compare the effect of GABA and morphine on long-term habituation, when drug doses, injection-training intervals and intertrial intervals were tailored to produce a similar small reduction in responding during training; in other words, to test the hypothesis that despite similar small effects on responding, morphine but not GABA impairs long-term habituation. METHOD Pilot experiments demonstrated that 6 pg GABA/g and 100 pg morphine/g injected 30 min before the 1st training trial, induced similar reduction in responding during a 15-trial training session with 180-s intertrial intervals. Two sessions separated by a 24-h rest interval were run: a 15-trial training session and a 3-trial testing one. Two hundred and four crabs were randomly divided into 6 groups of 34 each: SAL-TR injected with the vehicle; MP-TR, with 100 pg/g of morphine; GABA-TR, with 6 p,g/g of GABA; and three O-groups (SAL-O, MP-0 and GABA-O) injected with the respective solutions but without receiving shadow stimulation during the training session time. RESULTS AND DISCUSSION

Figure 3A illustrates the performance of trained crabs (SALTR, MP-TR and GABA-TR) during the training session with the

MORPHINE

AND GABA: DANGER STIMULUS

703

IN CRAB

@<0.05) between SAL-O and SAL-TR or GABA-TR, but not between SAL-O and GABA-O or MP-0 or MP-TR. Thus acquisition of a long-term habituation to a danger stimulus, appears to be impaired by a small depressing effect of morphine on the escape response during training, but not by a similar effect of GABA pretreatment. EXPERIMENT

* II

Although foregoing results support the hypothesis that morphine inhibits the formation of long-term habituation by interfering with stimulus processing, an alternative hypothesis is available. In this hypothesis, morphine would reduce the response level to a danger stimulus but leave stimulus processing intact, so that when the shadow is iteratively presented habituation would take place. However, long-term habituation is not observed because morphine also induces amnesia, either due to a storage failure or to a retrieval failure. Experiment 4 was aimed at testing this alternative explanation.

E9

SAL-O SAL-lR

“P-0

4

METHOD

MP-TR

GABAQ

GABkTR

GROUPS

PIG. 3. Morphine but not GABA impairs long-term habituation, despite similar effect of both drugs on training performance. (A) Results of training session. Ordinates: mean response score per trial. Open circles: group injected with saline 30 min before training. Closed circles: group injected with GABA (9 kg/g) 30 min before training. A: group injected with morphine (100 pg/g) 30 min before training. Inset: average of accumulative scores for the 15 training trials per group. (B) Results of testing session. Ordinates: average of accumulative scores for tbe 3 testing trials per group. White bars: SAL-O, no training after saline injection; SALT-TR, training after saline injection. Striped bars: MP-0, no training after morphine injection; MP-TR, training after morphine injection. Black bars: GABA-0, no training after GABA injection; GABA-TR, training after GABA injection. *Significant difference (Duncan test, p
mean accumulative scores for 15 trials shown in an inset. As expected, the response level of GABA-TR was far higher than that of the corresponding group in Experiment 2, since 6 instead of 9 (*g/g was injected at 30 min instead of 5 min before training. The responding of the MP-TR group was here noticeably higher than that obtained in prior work with the same dose (100 bg/g) (42); a difference attributable to the fact that in the present experiment 180-s instead of 27-s intertrial intervals were used, i.e., the 15 trials were spread out between 30 and 75 min after injection whereas in the prior experiment between 30 and 38 min, so that dissipation of the morphine effect over time was more apparent (2740). The ANOVA on accumulative scores for 15 trials (Fig. 3A, inset) revealed a significant treatment effect, F(2,117)= 3.1, pcO.5. The Duncan test disclosed a significant difference @<0.05) between SAL-TR and GABA-TR or MPTR, but no significant difference between drug-treated groups. The ANOVA performed on data of testing (Fig. 3B) revealed significant treatment mean square, F(5,198)=2.8, pcO.025. The Duncan test showed a significant difference

Two sessions were run 24 h apart: a 15-t&l training session and a 3-trial testing one. Injections were given immediately after the last training trial or at an equivalent time for crabs which were not presented the shadow stimulus (O-groups). One hundred and sixty crabs were randomly assigned in equal numbers to each cell of a 2 x 2 factorial design, the factors being drug (SALINE: the vehicle, or MP: 100 pg morphine/g), and training (0: no training session, or TR: training session), so that 4 groups were formed: O-SALINE, TR-SALINE, 0-MP and TR-MP. Groups were run simultaneously. Zero-groups remained in the actometers during all the time corresponding to the training session (75 min), but without being stimulated with the passing shadow. RESULTSAND DISCUSSION The ANOVA on data of testing (Fig. 4), showed significant treatment square, F(3,156)=6.01, p
704

TOMSIC,

1500 ki 0

1200

s:;;i t): 2 ZE 0,

**

900

%P g 1* 600 Z II: r

300

0

i

_ O-SAL

TR-SAL

O-HP

TR-MP

GROUPS FIG. 4. Posttraining injection of morphine fails to impair long-term habituation. Results of testing session. Ordinates: average of accumulative scores for the 3 testing trials per group. Injection given immediately after training. White bars: O-SAL, no training before saline injection; TRSAL.: training before saline injection. Striped bars: 0-MP, no training before morphine injection (100 p&g); TR-MP: training before morphine injection. **Significant difference (Duncan test, p
internal state. On the other hand, several instances drawn from studies on state dependency phenomenon (23,24), indicate that when modifications in the internal state are produced by changes in the level of certain endogenous compounds due to training, exogenous posttraining supply of these compounds exerts an amnesic effect because memory depends on the relationship between “interoceptive cues” (37) both after training and at testing time (23). Therefore, the finding that mo~hine posttraining injection has no effect on memos retention appears to rule out the state dependency hypothesis, too. GENERAL DISCUSSION Previous administration of GABA blocks the escape response to a danger without attenuating long-term habituation. According to our results, a detrimental effect of GABA on the shadow perception ability appears to be excluded. In addition, GABA induces autotomy, mainly to doses higher than 12 pgig, so that crabs spontaneously shed chelipeds and walking legs. Crabs acquire a long-term habituation in spite of the fact that the escape response to the danger stimulus is suppressed during training by GABA pretreatment, thus offering a new example of the so-called using-inde~ndent habituation (33). Several instances of this phenomenon are reported for animals other than crabs (4, 11, 13, 34), supporting the concept that stimulation is far more critical than responding to acquire long-term habituation. However, it is worthwhile to stress that the term responseindependent habituation does not mean acquisition of long-term habitation without any response whatsoever to the stimulus during training. Indeed, it means that the specific response under study, which correlates with the eliciting stimulus, wanes after iterative stimulation and recovers partially after an intersession interval, is not required during training to ensure its long-term habituation. In the present work such a specific response is the escape response, ~.e., a well-defined overt response. The res~nse-inde~ndent habituation is similar to the condihoning-without-res~nse paradigm that forced changes in asso-

MALDONADO

AND RAKITIN

ciative learning theory in the 1950s (33,35). Studies on the eye withdrawal reflex in the green crab from the laboratory of Abramson and Feinman, demons~ated that conditioning occurs despite the lack of the UR, either because the UR is habituated or the eye is restrained (l-3, 17. 18). Thus the equivalence between response-independent habituation and conditioning-without-response is illustrated by findings in the s_ame field of crustacean learning and memory research. The data from the present series of experiments along with those stemming from prior reports (9,42), provide reasonable support to the hypothesis that morphine induces in crabs a transient disruption between stimulus (a passing shadow) and meaning (an impending threat of a predator’s attack). According to a current view, the inhibitory effect of morphine, and the often reported enhancing effect of naloxone, seems mainly confined to behavior elicited by or related to a nociceptive stimulus [e.g., in rats, postshock activity burst (14): defensive boxing (15); tail withdrawal reflex (45); jump-avoidance (2O)j. Consistently. Warren and Ison (44), failed to show opiate modulation of reactions elicited by nonnociceptive stimuli. Indeed, these researchers matched tones and shock for the magnitude of startle provoked in rats and found that exogenous opioid analgesic morphine suppressed only the startle elicited by shock. Fanselow’s ~rceptual-defensive-recu~mtive model (IQ, includes a pain-fear interaction, but it contends that pain and not fear can be reduced by the action on an endogenous opioid mechanism. However, Davis (12) reported that morphine produced a reduction of the potentiated startle effect in rats, i.e., an effect where acoustic startle amplitude is normally increased in the presence of light previously paired with shock (8). Thus, morphine appears to inhibit a response elicited by Pavlovian fear cues. At first analysis, the latter result parallels those obtained with crabs, since in both cases morphine seems to modulate a response to a danger stimulus. Therefore, our results might be interpreted as in Davis’ report, i.e., that morphine attenuates or blocks the crab’s escape response to the danger stimulus by reducing fear or anxiety. We prefer, however, to explain them in the above terms of interference with stimulus decoding, based on the following reasons. First, the term fear is used to designate an emotional state that in rats is inferred from a set of behavioral changes concomitant with the unconditioned response under study. Thus the so-called conditioned fear stimuli have been shown to provoke a wide variety of responses, including hear? rate changes (6), suppression of ongoing appetitive behavior (7). enhancement of avoidance behavior (36) and freezing (41). In contrast, neither overt nor suggestive changes in behavior appear concomitantly with the Chasmagnathus’s escape response to a passing shadow overhead. Conditioning studies performed with the green crab (Car&us maenas) used an aversive stimulus (l-3, 17, 18), but no behavior changes concomitant to UR were described. Thus up to now it seems unjustified to assume that the crabs responses to danger or aversive stimuli are mediated by a generalized fear process. Second, there is a sharp difference between Davis’ preparation and the present one: in the former, light becomes a token of danger because of a pairing with shock, and it is due to this previous experience that light later on enhances the normal response to a startling stimulus; in the latter, instead, the passing shadow is assumed to be an innate omen of damage (25). In this regard, the crab’s danger-induced escape response resembles the rat’s normal acoustic startle reflex, which is unaffected by morphine pretreatment (12,44). Third, evidence from present and prior papers suggests that opioids alter the danger stimulus percept, since morphine both attenuates escape response during training and impairs ac-

MORPHINE AND GABA: DANGER STIMULUS IN CRAB

quisition of long-term habituation, while naloxone enhances the escape response (39). No such conclusion may be drawn from Davis’ experiments, since the possibility that morphine attenuates the rat’s potentiated startle by altering the light-danger link was not tested. Fourth, the hypothesis of the morphine action in terms of interference with stimulus decoding implies a model quite close to the wiring scheme suggested to explain how crabs can detect the approach of an object and initiate the escape response (32). Therefore, some predictions from our hypothesis may be made in electrophysiological studies based on the assumption that the critical stimulus parameter to elicit the escape response is a suprathreshold level of excitation in the lobula

705

looming cells (21,32). In fact, it would be expected that neither GABA nor morphine were capable to altering the normal behavior of lamina or medulla neurons, and on the other hand, that morphine but not GABA were able of decreasing the firing of the looming neurons. ACKNOWLEDGEMENTS

This work was supported by the Consejo National de Investigaciones Cientificas y Tecnicas de Argentina, grant PID 218-89, by the Buenos Aires University, grant Ex-134, and by Fundacion Antorchas. The authors are indebted to Angel Vidal for technical assistance.

REFERENCES 1. Abramson, F. I.; Feinman, R. D. Operant punishment in the green crab Curcinus maenas. Behav. Neural Biol. 48:259-277; 1987. 2. Abramson, F. I.; Feinman, R. D. Classical conditioning of the eye withdrawal reflex in the green crab. J. Neurosci. 8:2907-2912; 1988. 3. Abramson, C. I.; Armstrong, P. M.; Feinman, R. A.; Feinman, R. D. Signalled avoidance in the eye withdrawal reflex in the green crab. J. Exp. Anal. Behav. 50:483-492; 1988. 4. Applewhite, B. P.; Gardner, F. T.; Lapan, E. Physiology of habituation learning in a protozoan. Trans. NY Acad. Sci. 31:842-849; 1969. _ In: Bliss, D. E. The 5. Atwood, H. L. Synapses and neurotransmitters. bioloev of Crustacea. vol. 4. New York: Academic Press; 1982: lOS-lso. 6. Black, A. H.; de Toledo, L. The relationship among classically conditioned responses: Heart rate and skeletal behavior. In: Black, A. H.; Prokasy, W. F., eds. Classical conditioning II: Current theory and research. New York: Appleton, Century, Crofts; 1972:290311. 7. Bouton, M. E.; Bolles, R. C. Conditional fear assessed by freezing and by the suppression of the three different baselines. Anim. Learn. Behav. 8:429434; 1980. 8. Brown, J. S.; Kalish, H. I.; Farver, I. E. Conditional fear as revealed by magnitude of startle response to an auditory stimulus. J. Exp. Psychol. 41:317-327; 1951. 9. Brunner, D.; Maldonado, H. Habituation in the crab Chasmagnarhus granularus: Effect of morphine and naloxone. J. Comp. Physiol. [Al 162:687-694; 1988. New York: Dover; 1964. 10. Cady, W. G. Piezoelectricity. E. Funktionsweise und Stammesgeschichte des Flug 11. Curio, feinderknnens einiger Darwin&ken (Geospizinae). Z. Tierpsychol. 26:394-487; 1960. 12. Davis, M. Morphine and naloxone: effects on conditioned fear as measured with the potentiated startle paradigm. Eur. J. Pharmacol. 54:341-347; 1979. 13. Davis, M.; Wagner, A. R. Habituation of startle response under incremental sequence of stimulus intensities. J. Comp. Physiol. Psychol. 67:486-492; 1967. 14. Fanselow, M. S. Opiate modulation of the active and inactive components of the postshock reaction: parallels between naloxone pretreatment and shock intensity. Behav. Neurosci. 98:269-277; 1984. 15. Fanselow, M. S.; Sigmundi, R. A. The enhancement and reduction of defensive fighting by naloxone pretreatment. Physiol. Psychol. 10:313-316; 1982. 16. Fanselow, M. S. Conditioned fear-induced opiate analgesia: A competing motivational state theory of stress analgesia. In: Kelly, D. D., ed. Stress-induced analgesia. New York: New York Academy of Sciences; 198640-54. 17. Feimnan, R. D.; Abramson, C. I.; Forman, R. R. Classical conditioning in crustaceans. In: Wiese, K.; Kren, W. D.; Tautz, J.; Reichert, H.; Mulloney, B. eds. Frontiers in crustacean neurobiology. Berlin: Birkhauser Verlag AG; 1990:215-222. 18. Feinman, R. D.; Llinas, R. H.; Abramson, C. I.; Forman, R. R. Electromyographic record of classical conditioning of eye withdrawal in the crab. Biol. Bull. 178:187-194; 1990. 19. Fredericq, L. Sur l’autotomie ou mutilation par voie reflexe comme

20.

21.

22. 23.

24.

25.

26. 27.

28.

29.

30.

31.

32.

33.

34.

35. 36. 37.

moyen de defense chez les animaux. Arch. Zool. Exp. Gen. 1:413426; 1883. Galizio, M.; Gore, S.; Sanderson, P. Effect of naloxone on acquisition and extinction of jump-up avoidance behavior in rats. Learn. Motiv. 14:48-57; 1983. Glantz, R. M. Visual input and motor output of command intemeurons of the defense reflex pathway in the crayfish. In: Hoyle, G., ed. Identified neurons and behavior of arthropods. New York: Plenum; 1977:259-274. Gold, P. E. The use of avoidance training in studies of modulation of memory storage. Behav. Neural Biol. 46:87-98; 1986. Izquierdo, I.; Dias, R. D. Endogenous state-dependency: memory regulation by post-training and pre-testing administration of ACTH, B-endorphin, adrenaline and tyramine. Braz. J. Med. Res. 16:5564; 1983. Izquierdo, I.; Netto, A. Role of the brain B-endorphin system and hormonal mechanisms in the regulation of retrieval at 0, 3, 6 or 24 hs. from training. In: McGaugh, J., ed. Contemporary psychology: Biological processes and theoretical issues. North-Holland: Elsevier Science Publishers B. V.; 1985:95-108. Krasne, F. B. Learning in Crustacea. In: Coming, W.; Dyal. J.; Willows, A. 0. D., eds. Invertebrate learning. vol. 2. New York: Plenum; 1972:49-129. Lewis, D. J. Psychobiology of active and inactive memory. Psychol. Bull. 86:1054-1083; 1979. Lozada, M.; Romano, A.; Maldonado, H. Effect of morphine and naloxone on a defensive response of the crab Chasmagnathus grundams. Pharmacol. Biochem. Behav. 30:635-640; 1988. Lozada, M.; Romano, A.; Maldonado, H. Long-term habituation to a danger stimulus in the crab Chasmagnathus grumdatus. Physiol. Behav. 47:3541; 1990. Lozada, M.; Romano, A.; Maldonado, H. Effect of intertrial interval length on escape response habituation in the crab Chasmagnathus. Stimulus timing as a memory component. Submitted. Maldonado, H.; Romano, A.; Lozada, M. Gpioid action on response level to a danger stimulus in the crab (Chasmagnathus grandams). Behav. Neurosci. 103:1139-l 143; 1989. Martinez, J. L.; Jensen, R. A.; McGaugh, J. L. Facilitation on memory consolidation. In: Deutsch, J. A., ed. The physiological basis of memory. New York: Academic Press; 1983:49-70. Nalbach, H. 0. Visually elicited escape in crabs. In: Wiese, K.; Kren. W. D.; Tautz, J.; Reichert, H.; Mulloney, B., eds. Frontiers in crustacean neurobiology. Berlin: Birkhauser Verlag AG; 1990: 165-173. Peeke, H. V. S. Habituation and the maintenance of territorial boundaries. In: Peeke, H. V. S.; Petrinovich, L. Habituation, sensitization and behavior. New York: Academic Press; 1988:393-421. Peeke, H. V. S.; Veno, A. Stimulus specificity of habituated aggression in the three-seined sticklebacks (Gasrerosteus acualearus). Z. Tierpsichol. 40:53-58; 1976. Rescorla, R. A. Behavioral studies of Pavlovian conditioning. Annu. Rev. Neurosci. 11:329-352; 1988. Rescorla, R. A.; LoLordo, V. M. Inhibition of avoidance behavior. J. Comp. Physiol. Psychol. 59406-412; 1965. Riccio, D. C.; Concannon, J. T. ACTH and the reminder phenomenon. In: Martinez, J. L., Jr. ; Jensen, R. A.; Messing, R. B.; Rigter,

TOMSIC. MALDONADO AND RAKITIN

706

38. 39.

40.

41.

42.

H.; McCaugh, J. L., eds. Endogenous peptides and learning and memory processes. New York: Academic Press; 1981:117-142. Rodgers, R. J. Differential effects of naloxone and diprenorphine on defensive behavior. Neurophatmacology 21:1291-1294; 1982. Romano, A. G.; Lozada, M.; Maldonado, H. Effect of naloxone pretreatment on habituation in the crab Chasmagnarhus grandam. Behav. Neural Biol. 53:113-122; 1990. Romano, A.; Lozada, M.; Maldonado, H. Non-habituation processes affect stimulus specificity of the response-habituation in the crab Chasmagnarhus. Behav. Neurosci.; in press. Sigmundi, R. A.; Bouton, M. E.; Bolles, R. C. Conditioned freezing in a rat as function of shock intensity and CS modality. Bull. Psychonom. Sot. 15:254-256; 1980. Tomsic. D.; Maldonado, H. Central effect of morphine pretreatment

on short- and long-term habituation to a danger stimulus in the crab Chasmagnathus. Pharmacol. Biochem. Behav. 36:787-793; 1990. 43. Valeggia, C.; Femandez Duque, E.; Maldonado, H. Danger stimulus-induced analgesia in the crab Chasmagnathus granularus. Brain Res. 481:304-308; 1989. 44. Warren, P. H.; Ison, J. R. Selective action of morphine on reflex expression to nociceptive stimulation in the rat: A contribution to the assessment of analgesia. Pharmacol. Biochem. Behav. 16:869874; 1982. 45. Watkins, L. R.; Cobelli, D. A.; Mayer, D. J. Classical conditioning of front paw and hind paw footshock induced analgesia (FSIA): Naloxone reversibility and descending pathways. Brain Res. 243: 119-132: 1982.