Physiology&Behavior.Vol. 52, pp. 449-453, 1992
0031-9384/92 $5.00 + .00 Copyright © 1992 Pergamon Press Ltd.
Printed in the USA.
Halothane Anesthesia Causes State-Dependent Retrieval Failure in Mice EDWIN
ROSMAN,
DAVID
QUARTERMAIN,
~ RICHARD
PANG
AND
HERMAN
TURNDORF
Departments of Anesthesiology and Neurology, New York University School of Medicine, 550 First Avenue, New York, N Y 10016 R e c e i v e d 5 F e b r u a r y 1992 ROSMAN, E., D. QUARTERMAIN, R. PANG AND H. TURNDORF. Halothane anesthesia causes state-dependent retrieval failure in mice. PHYSIOL BEHAV. 52(3) 449-453, 1992.--Effects of exposure to halothane on memory processing was studied using single-trial inhibitory avoidance learning to measure retention. Mice were anesthetized with halothane either before training, immediately after training, or both before training and before testing. Results showed that memory was not impaired by posttraining halothane exposure, indicating that the anesthetic does not cause retrograde amnesia. Mice trained after recovery from halothane showed a robust memory loss 24 h later. This deficit could be alleviated by reexposure to the anesthetic before the retention test. Mice given multiple training trials following recovery from the anesthetic showed a normal rate of learning when compared with controls, but deficient retention. This indicates that the performance deficit was the result of impaired retention (anterograde amnesia) rather than disrupted acquisition. Anterograde amnesia occurred when training was delayed up to 2 h after recovery from anesthesia. These findings indicate that the memory impairment following halothane anesthesia is the result of a statedependent retrieval failure. Halothane anesthesia
Anterograde amnesia
State dependent learning
A L T H O U G H it is well established that halothane produces amnesia, there is ambiguity about the causes of the m e m o r y loss. Animal studies using an imprinting paradigm and chicks as subjects have shown that halothane introduced after a training trial decreases the strength of the following response (8,11,14). These data have been interpreted as evidence that halothane can disrupt m e m o r y consolidation (5,10), but demonstrations that the anesthetic can have proactive aversive effects on behavior (3,7,14,15) make this interpretation problematical. There is evidence that retrieval processes can be impaired by anesthetics. Adam (1) showed that material learned following 6 h of deep anesthesia was recalled poorly if the retention interval was 10 min or 2 h; however, performance was significantly improved if the subjects were tested 1 week after anesthesia. Additional evidence that anesthetics can cause retrieval failure is provided by a study which demonstrated that isoflurane can result in statedependent learning (2). Subjects tested after recovery from anesthesia were unable to recall verbal material learned during anesthesia, yet m e m o r y could be restored if the subjects were reexposed to isoflurane before the retention test. These studies indicate that the basis of the m e m o r y disturbance induced by general anesthetics is not well understood. Few experiments have attempted to explicitly distinguish between anterograde and retrograde effects of anesthetics on m e m ory, and only limited attention has been given to the effect of anesthetics on retrieval processes.
Memory retrieval
Inhibitory avoidance
The purpose of the present study was to investigate the retrograde and anterograde effects of halothane anesthesia on m e m o r y processing and attempt to determine the relative contribution of disturbances of storage and retrieval mechanisms to the etiology of the amnesia. METHOD Male Swiss Webster mice (Taconic, Germantown, NY), 10 weeks of age and between 30 and 40 g body weight, were the subjects for these experiments. Mice were housed five per cage with food and water available ad lib.
Behavioral Task and Apparatus Single-trial inhibitory avoidance learning was used to study the effects of halothane anesthesia on retention. Mice were trained in a two-compartment shuttle chamber (LVE #MSC002). The dark compartment was 23 cm long, 9 cm wide, and 11 cm high, and was constructed from black Plexiglas. The floor was made from stainless steel rods (0.3 cm diameter, 7 cm between rods) through which a scrambled shock could be delivered from a Coulbourn constant current shocker. The safe side of the chamber, which was white, was the same size as the shock side, had a solid floor, and was covered by a lid in the center of which was a 28 V lamp which was illuminated during training
Requests for reprints should be addressed to David Quartermain.
449
R O S M A N ET AL.
450 and testing. The two compartments were separated by a wall which contained a guillotine door 8 cm high and 4 cm wide.
300
Procedure
250
A training trial was begun by placing a mouse in the white side and opening the door. W h e n the animal entered the dark compartment the door was lowered and a 0.2 m A shock was administered for 1.2 s. A latency timer automatically recorded the time to cross into the dark side. Retention of this learning was tested by returning the mouse to the white compartment and recording the time to reenter the dark compartment. Mice failing to cross within 300 s were given the m a x i m u m latency as the test score.
200 150 10 0
50
0
Halothane Anesthesia Mice were anesthetized in groups of 10 by passing a gas mixture of halothane (2%) and oxygen into a 5 1 acrylic chamber using a calibrated vaporizer. Soda lime was sandwiched between the bottom of the chamber and the perforated floor plate. Excess gasses were scavenged. Induction time was measured from time of initiating gas flow to loss o f spontaneous m o v e m e n t after falling over. Wake-up time was determined by initiation of organized locomotion following termination of anesthesia. EXPERIMENT 1 The objective of this experiment was to determine ifhalothane Would induce retrograde amnesia (RA) by anesthetizing groups of mice for different durations immediately following termination of the training trial.
Procedure
TEST LATENCY (SEC)
AIR CONTROL
115MIN
30MJN HALOTHANE
60MIN I
FIG. 1. Absence of retrograde amnesia in halothane-treated mice. Immediately following the training, trial groups of mice were anesthetized for 15, 30, or 60 rain. Retention was tested 24 h after training.
to air. Some mice in both air and halothane groups were tested l0 min after learning and others were tested 24 h after learning. To evaluate possible state-dependent effects ofhalothane, a group of 10 mice treated with halothane before training, and a group of 10 mice exposed to air prior to the learning trial were exposed to halothane prior to the retention test. Mice were tested 15 min after recovery. The design of this experiment is shown in Table 1.
Results
Mice were trained as described above and immediately after placed in the halothane chamber. The time between completion of training and anesthesia was approximately 80 s. Different groups of mice (N = 10 per group) were exposed to halothane for 15, 30, or 60 min. A control group of 10 mice was left in r o o m air in the laboratory for 30 rain after the learning trial. Retention was tested 24 h after training.
Results Mean test latencies for the four treatment groups is shown in Fig. 1. It is evident from these data that halothane anesthesia failed to produce retrograde amnesia. This was confirmed by the results of a one-way A N O V A ( F = 0.89).
There was no significant difference in initial step-through latencies between halothane- and air-treated mice (AIR = 18.3 s; H A L = 20.1 s) indicating that anesthesia had not altered activity levels. There was also no behavioral evidence of residual analgesia in the halothane-treated mice. Responses to foot shock (running, squealing, and jumping) were virtually identical in the two groups. The results of air- a n d halothane-treated mice tested 10 rain or 24 h after training are shown in Fig. 2. A two-way A N O V A applied to these data indicated that there was a significant difference between treatment groups, F(I, 36) = 9.34, p = 0.004, a significant effect of time of testing, F(1, 36) = 6.34, p = 0.015, and a significant interaction between treatment group and test interval, F(1, 36) = 5.64, p = 0.021. Figure 2 shows that while the test scores of the two groups were similar 10 min
EXPERIMENT 2 The aim of this experiment was to examine possible anterograde amnestic effects of halothane by training animals following recovery and testing retention 24 h later. An additional objective was to evaluate the contribution of state-dependent retrieval failure to anterograde m e m o r y loss.
TABLE 1 DESIGN FOR EXPERIMENT 2 Group
Procedure Thirty mice were anesthetized in three batches of 10 as previously described. After 30 min they were removed to the home cage and recovery time was determined. Mice were considered to have recovered from the anesthetic when they were spontaneously ambulating and rearing. Mean recovery time was 8.5 min. Training was initiated 15 min after recovery, a time at which mice were behaviorally indistinguishable from control animals. Thirty control mice were trained after 45-min exposure
1 2 3 4 5 6
Treatment Before Training Air Air HAL HAL Air HAL
Treatment Before Testing
Test Time
None None None None HAL HAL
24 Hours 10 Minutes 24 Hours 10 Minutes 24 Hours 24 Hours
HAL: Exposure to Halothane for 30 min. Mice were trained or tested 15 min after recovery from anesthesia.
HALOTHANE-INDUCED AMNESIA IN MICE
LATENCY (SEC) 300[ 2501
~
HALOTHANE
NN
/
200[
oot=
150[-
50
0
10 min
24 hr
TRAIN
-
TEST INTERVAL
FIG. 2. Anterogradeamnesia.Micewere exposedto halothane and trained 15 min after recovery. Half of the group was tested 10 rain after training and half at 24 h. A control group left in room air was trained and tested similarly.
after training, substantial differences were evident by 24 h. At this time point, air-treated mice had developed a strong memory of the avoidance response, while halothane mice continued to show poor retention. The slow development of learned avoidance in untreated animals has been observed previously [e.g., (6,13)], and suggests that some avoidance behaviors require many hours to reach maximum strength. The results of exposure to halothane prior to the retention test are shown in Fig. 3. Groups 1 and 4 (Table 1) from Fig. 2 are included for comparison and analysis. A one-way ANOVA carried out on the data in Fig. 3 show that there was a significant difference in test scores among the four groups, F(3, 36) = 6.0, p = 0.002. Post hoc comparisons indicated that the HAL/HAL group had significantlybetter retention of passive avoidance than the HAL/AIR group, t(18) = 2.15, p = 0.04, and that the AIR/ HAL group had significantly poorer retention of passive avoidance than the AIR/AIR subjects, t(18) = 3.72, p = 0.001. While it appears that reexposure to halothane reverses the AA, an alternative, though unlikely, explanation should be ruled out. It is possible that exposure to halothane twice within 24 h is aversive, and that the long lateneies on the test day reflect conditioned aversion rather than restored memory. To evaluate this possibility, we tested two additional groups of 10 mice each. Procedure was the same as usual except that no shock was administered on the training trial. One group was anesthetized prior to training and the other was exposed to air. Before testing, 24-h later the halothane group was reexposed to the anesthetic and tested 15 min after recovery. Results showed no difference between the two groups (AIR = 17.9 + 1.2; HAL/HAL 19.9 + 1.7). These results show: 1. that animals trained following air exposure become amnestic if they are anesthetized with halothane before testing and 2. mice ordinarily amnestic following training after halothane exposure show a restoration of memory if they are reexposed to halothane before testing. These findings indicate that halothane anesthesia results in a symmetrical state-dependent retrieval failure. EXPERIMENT 3
When animals are treated with drugs prior to the training trial, it is frequently difficult to determine whether poor perfor-
451 mance on a subsequent test is the result of impaired acquisition or disrupted retention. A procedure which can sometimes help distinguish between these two possibilities is to test performance shortly after training (e.g., 30 min to 1 h) before information is transferred to long-term storage. For example, Kubanis, Gobbel, and Zornetzer (9) have shown that retention performance of old and young mice is similar 2 h after the training trial, but at 24 h the old animals exhibit a profound deficit. This finding indicates that the deficit in the aged rats is the result of abnormally rapid forgetting rather than impaired acquisition. The intention of the 10-min test in Experiment 2 was to provide information on the effects of halothane on acquisition processes, but because of the poor performance by both halothane- and air-treated mice on this test, the interpretation of these data is ambiguous. The objective of Experiment 3 was to examine the effects of halothane on the acquisition of step-through inhibitory avoidance. Procedure Twenty-five mice were the subjects for this experiment. Thirteen mice were exposed to halothane for 30 min as previously described and the 12 control animals were left in room air. As in Experiment 2, training was begun 15 min following recovery. The training session was begun by placing mice in the white compartment and raising the door. When the mouse entered the dark compartment the shock (0,2 mA) was initiated and remained on for the remainder of the training session. The mouse could escape the shock by running back into the white compartment. The door remained raised and mice received foot shock each time they stepped onto the bars of the dark compartment. Training was continued until mice remained in the white compartment for 100 s without a foot shock. The number of shocks taken until the learning criterion was attained was recorded for all animals. Retention was tested 24 h later. Results Mean number of shocks taken to attain criterion was 12.75 for the air-exposed mice and 12.41 for the halothane-treated subjects. Twenty-four h retention test scores were 207.6 + 27.2 s for the air controls and 112.7 + 26.7 s for the halothane-exposed subjects. This difference was statistically significant, t(23) = 2.488, p = 0.021. These data indicate that mice anesthetized
300
LATENCY(SEC) ~lB
250
TRAINING TESTING
200
150
T
100
50
0 BEFORE TRAINING BEFORE TESTING
AIR AIR
HALOTHANE AIR HALOTHANE AIR HALOTHANE HALOTHANE
FIG. 3. State dependent learning. Mice were trained after exposure to halothane or air. Half of each group received halothane or air before testing.
ROSMAN ET AL.
452 with halothane learned the avoidance response normally but forgot that learning significantly faster than the nonexposed control subjects.
TEST LATENCY (SEC) 140 120
EXPERIMENT 4
The purpose of this experiment was to determine the temporal gradient ofhalothane-induced anterograde amnesia. To accomplish this, we trained different groups of mice at varying times after recovery from anesthesia and tested for retention 24 h after training. Procedure
tO0 80 60 40
Forty mice were anesthetized in batches of 10 as described above and randomly assigned to one of four groups (N = 10 per group) which were trained either 15, 120, 240, or 1440 rain after recovery from halothane. A control group (N = 10) was not anesthetized. Retention was tested 24 h following training. Results The test latencies of the five treatment groups are shown in Fig. 4. A one-way ANOVA applied to these data revealed a significant difference among the test latencies of the five groups, F(4, 45) = 3.07, p = 0.025. Inspection of Fig. 4 shows that amnesia occurs if mice are trained up to 2 h after recovery from halothane, but retention is normal if training is delayed for 4 h, 120 vs. 240 min, t(18) = 2.14, p -- 0,044. DISCUSSION
The principal finding of this study is that halothane anesthesia causes state-dependent anterograde amnesia. If training is initiated within 2 h following anesthesia, learned information is not accessible to normal retrieval cues at the retention test. However, if mice are reexposed to the anesthesia prior to the retention test, restoration of the memory of the avoidance response occurs. This finding, in conjunction with the observation that unanesthetized mice become amnestic if exposed to halothane before the test, is evidence that halothane causes memory failure by disrupting retrieval processes. The symmetrical state dependence can be most readily explained by supposing that halothane anesthesia produces an altered central state accompanied by novel interoceptive stimuli which changes the way the animal perceives the training environment. These altered stimulus conditions form the context within which the avoidance response is learned and they become incorporated, along with other contextual features, as attributes of the memory. These contextual stimuli are known to function as retrieval cues, and a sufficient number have to be available at testing in order for the event to be remembered. Amnesia occurs 24 h after halothane because the central state has changed and the stimulus conditions no longer match those which existed at the time of training. Reexposing mice to halothane prior to testing restores remembering because the anesthesia reinstates a significant part of the context within which the original learning took place. The demonstration that retrieval is disrupted by halothane should not be taken to imply that encoding processes are unaffected by the anesthetic. The results of the present experiment make it clear that it is the interaction between encoding and retrieval which determines retention performance. Whether mice remember being shocked in the dark compartment depends on the conditions (halothane or air cues) which exist at the time of
20 0 tSmin
2 Hrs
4 hra
24 hra
Control
Recovery-Training Interval FIG. 4. Temporal gradient of anterograde amnesia, Mice were anesthetized with halothane and trained either 15 rain, 2, 4 or 24 h after recovery from anesthesia. Retention was tested 24 h after training.
testing. Furthermore, the effectiveness of cues in the test environment in retrieving the memory of shock depends on the conditions (halothane or air) under which original learning took place. A similar phenomenon has been reported in anestheticinduced amnesia in man (1). In this study subjects were read a list of words during recovery from a period of anesthesia. Retention was tested by a forced-choice recognition test in which errors could be categorized as acoustically similar to the item, semantically similar, or unrelated. Analysis of errors showed that subjects selected significantly more acoustically similar items than would be expected under undrugged conditions, suggesting that they had been encoding information in terms of the acoustical rather than the semantic features of the material. Amnesia occurred in this study because the cues used during encoding did not match the cues used at the retrieval test. Taken together, these two studies suggest that explanation of the disrupting effects of anesthetics on remembering should be sought in the interaction of encoding and retrieval processes rather in either process alone (13). The clear absence of retrograde amnesia in this study is noteworthy. This finding confirms a similar experiment using mice as subjects (4), but conflicts with a rat experiment which demonstrated a gradient of retrograde amnesia (12). The demonstration of a retrograde disruption of performance with halothane is not strong evidence for amnesia without the inclusion of control groups for the aversive effects of anesthetics which are known to occur in rats (3,7). Overall, it does not appear that halothaneinduced retrograde amnesia is a robust phenomenon. The alterations in neural functioning which underlie the halothane-induced retrieval deficit remain to be determined. A research strategy which may reveal which neurochemical systems are involved in the retrieval failure is to attempt to block or reverse the amnesia by administering selected pharmacological agents (e.g., cholinergic agonists, calcium channel blockers, GABA antagonists) before anesthesia, after learning or before testing. This line of investigation is currently being pursued in our laboratory.
HALOTHANE-INDUCED
AMNESIA IN MICE
453
REFERENCES 1. Adam, N. Effects of general anesthetics on memory functions in man. J. Comp. Physiol. Psychol. 83:294-305; 1973. 2. Adam, N.; Castro, A. D.; Clark, D. L. State-dependent learning with a general anesthetic (Isoflurane) in man. Life Sci. 4:125134; 1974. 3. Alexinsky, T.; Chapouthier, G. Halothane anesthesia and DMS performance in rats: Memory impairment or avoidance behavior? Physiol. Behav. 22:99-105; 1979. 4. Angel, C.; Bounds, H. M., Jr.; Perry, A. A comparison of the effects of halothane on blood-brain barrier and memory consolidation. Dis. Nerv. Syst. 33(2):87-93; 1972. 5. Cherkin, A.; Lee-Teng, E. Interruption by halothane of memory consolidation in chicks. Fed. Proc. 24:328; 1965. 6. Deutsch, J. A.; Rogers, J. B. Cholinergic excitability and memory: Animal studies and their clinical implications. In: Davis, K. L.; Berger, P. A., eds. Brain acetylcholine and neuropsychiatric disease. New York: Plenum Press; 1979:175-204, 7. Gisquest-Verrier, P. Accelerated extinction after post trial halothane anesthesia: An aversive effect. Physiol. Behav. 26:223-231 ; 1981.
8. Gutekunst, R.; Youniss, J. Interruption of imprinting following anesthesia. Percept. Mot. Skills 16:340; 1963. 9. Kubanis, P.; Gobbel, G.; Zornetzer, S. F. Age related memory deficits in swiss mice. Behav. Neural Biol. 32:241-247; 1981. 10. Lecanuet, J. P.; Alexinsky, T.; Cbapouthier, G. The following response in chicks: Conditions for the resistance of consolidation to a disruptive agent. Behav. Biol. 12:365-372; 1974. 11. Lecanuet, J. P.; Deweer, B.; Bloch, V. Effect of post-exposure anesthesia on the retention of imprinting. Behav. Biol. 12:365-372; 1974. 12. Penrod, W. C.; Boice, R. Effects of halothane anesthesia on the retention of passive avoidance task in rats. Psychon. Sci. 23:205207; 1971. 13. Sehacter, D. L.; Tulving, E. Amnesia and memory research. In: Cermak, L. S., ed. Human memory and amnesia. Hillsdale, N J: Lawrence Erlbaum Assoc. Inc.; 1982:1-32. 14. Schmaltz, G. Can halothane anesthesia have any aversive effects on the rat? Physiol. Behav. 22:25-29; 1979. 15. Zinkin, S.; Lecanuet, J. P.; Deweer, B. Retroactive and proactive effects of anesthesia on following in chicks. Physiol. Behav. 16:185189; 1976.
0031-9384/92 $5.00 + .00 Copyright © 1992 Pergamon Press Ltd.
Printed in the USA.
Halothane Anesthesia Causes State-Dependent Retrieval Failure in Mice EDWIN
ROSMAN,
DAVID
QUARTERMAIN,
~ RICHARD
PANG
AND
HERMAN
TURNDORF
Departments of Anesthesiology and Neurology, New York University School of Medicine, 550 First Avenue, New York, N Y 10016 R e c e i v e d 5 F e b r u a r y 1992 ROSMAN, E., D. QUARTERMAIN, R. PANG AND H. TURNDORF. Halothane anesthesia causes state-dependent retrieval failure in mice. PHYSIOL BEHAV. 52(3) 449-453, 1992.--Effects of exposure to halothane on memory processing was studied using single-trial inhibitory avoidance learning to measure retention. Mice were anesthetized with halothane either before training, immediately after training, or both before training and before testing. Results showed that memory was not impaired by posttraining halothane exposure, indicating that the anesthetic does not cause retrograde amnesia. Mice trained after recovery from halothane showed a robust memory loss 24 h later. This deficit could be alleviated by reexposure to the anesthetic before the retention test. Mice given multiple training trials following recovery from the anesthetic showed a normal rate of learning when compared with controls, but deficient retention. This indicates that the performance deficit was the result of impaired retention (anterograde amnesia) rather than disrupted acquisition. Anterograde amnesia occurred when training was delayed up to 2 h after recovery from anesthesia. These findings indicate that the memory impairment following halothane anesthesia is the result of a statedependent retrieval failure. Halothane anesthesia
Anterograde amnesia
State dependent learning
A L T H O U G H it is well established that halothane produces amnesia, there is ambiguity about the causes of the m e m o r y loss. Animal studies using an imprinting paradigm and chicks as subjects have shown that halothane introduced after a training trial decreases the strength of the following response (8,11,14). These data have been interpreted as evidence that halothane can disrupt m e m o r y consolidation (5,10), but demonstrations that the anesthetic can have proactive aversive effects on behavior (3,7,14,15) make this interpretation problematical. There is evidence that retrieval processes can be impaired by anesthetics. Adam (1) showed that material learned following 6 h of deep anesthesia was recalled poorly if the retention interval was 10 min or 2 h; however, performance was significantly improved if the subjects were tested 1 week after anesthesia. Additional evidence that anesthetics can cause retrieval failure is provided by a study which demonstrated that isoflurane can result in statedependent learning (2). Subjects tested after recovery from anesthesia were unable to recall verbal material learned during anesthesia, yet m e m o r y could be restored if the subjects were reexposed to isoflurane before the retention test. These studies indicate that the basis of the m e m o r y disturbance induced by general anesthetics is not well understood. Few experiments have attempted to explicitly distinguish between anterograde and retrograde effects of anesthetics on m e m ory, and only limited attention has been given to the effect of anesthetics on retrieval processes.
Memory retrieval
Inhibitory avoidance
The purpose of the present study was to investigate the retrograde and anterograde effects of halothane anesthesia on m e m o r y processing and attempt to determine the relative contribution of disturbances of storage and retrieval mechanisms to the etiology of the amnesia. METHOD Male Swiss Webster mice (Taconic, Germantown, NY), 10 weeks of age and between 30 and 40 g body weight, were the subjects for these experiments. Mice were housed five per cage with food and water available ad lib.
Behavioral Task and Apparatus Single-trial inhibitory avoidance learning was used to study the effects of halothane anesthesia on retention. Mice were trained in a two-compartment shuttle chamber (LVE #MSC002). The dark compartment was 23 cm long, 9 cm wide, and 11 cm high, and was constructed from black Plexiglas. The floor was made from stainless steel rods (0.3 cm diameter, 7 cm between rods) through which a scrambled shock could be delivered from a Coulbourn constant current shocker. The safe side of the chamber, which was white, was the same size as the shock side, had a solid floor, and was covered by a lid in the center of which was a 28 V lamp which was illuminated during training
Requests for reprints should be addressed to David Quartermain.
449
R O S M A N ET AL.
450 and testing. The two compartments were separated by a wall which contained a guillotine door 8 cm high and 4 cm wide.
300
Procedure
250
A training trial was begun by placing a mouse in the white side and opening the door. W h e n the animal entered the dark compartment the door was lowered and a 0.2 m A shock was administered for 1.2 s. A latency timer automatically recorded the time to cross into the dark side. Retention of this learning was tested by returning the mouse to the white compartment and recording the time to reenter the dark compartment. Mice failing to cross within 300 s were given the m a x i m u m latency as the test score.
200 150 10 0
50
0
Halothane Anesthesia Mice were anesthetized in groups of 10 by passing a gas mixture of halothane (2%) and oxygen into a 5 1 acrylic chamber using a calibrated vaporizer. Soda lime was sandwiched between the bottom of the chamber and the perforated floor plate. Excess gasses were scavenged. Induction time was measured from time of initiating gas flow to loss o f spontaneous m o v e m e n t after falling over. Wake-up time was determined by initiation of organized locomotion following termination of anesthesia. EXPERIMENT 1 The objective of this experiment was to determine ifhalothane Would induce retrograde amnesia (RA) by anesthetizing groups of mice for different durations immediately following termination of the training trial.
Procedure
TEST LATENCY (SEC)
AIR CONTROL
115MIN
30MJN HALOTHANE
60MIN I
FIG. 1. Absence of retrograde amnesia in halothane-treated mice. Immediately following the training, trial groups of mice were anesthetized for 15, 30, or 60 rain. Retention was tested 24 h after training.
to air. Some mice in both air and halothane groups were tested l0 min after learning and others were tested 24 h after learning. To evaluate possible state-dependent effects ofhalothane, a group of 10 mice treated with halothane before training, and a group of 10 mice exposed to air prior to the learning trial were exposed to halothane prior to the retention test. Mice were tested 15 min after recovery. The design of this experiment is shown in Table 1.
Results
Mice were trained as described above and immediately after placed in the halothane chamber. The time between completion of training and anesthesia was approximately 80 s. Different groups of mice (N = 10 per group) were exposed to halothane for 15, 30, or 60 min. A control group of 10 mice was left in r o o m air in the laboratory for 30 rain after the learning trial. Retention was tested 24 h after training.
Results Mean test latencies for the four treatment groups is shown in Fig. 1. It is evident from these data that halothane anesthesia failed to produce retrograde amnesia. This was confirmed by the results of a one-way A N O V A ( F = 0.89).
There was no significant difference in initial step-through latencies between halothane- and air-treated mice (AIR = 18.3 s; H A L = 20.1 s) indicating that anesthesia had not altered activity levels. There was also no behavioral evidence of residual analgesia in the halothane-treated mice. Responses to foot shock (running, squealing, and jumping) were virtually identical in the two groups. The results of air- a n d halothane-treated mice tested 10 rain or 24 h after training are shown in Fig. 2. A two-way A N O V A applied to these data indicated that there was a significant difference between treatment groups, F(I, 36) = 9.34, p = 0.004, a significant effect of time of testing, F(1, 36) = 6.34, p = 0.015, and a significant interaction between treatment group and test interval, F(1, 36) = 5.64, p = 0.021. Figure 2 shows that while the test scores of the two groups were similar 10 min
EXPERIMENT 2 The aim of this experiment was to examine possible anterograde amnestic effects of halothane by training animals following recovery and testing retention 24 h later. An additional objective was to evaluate the contribution of state-dependent retrieval failure to anterograde m e m o r y loss.
TABLE 1 DESIGN FOR EXPERIMENT 2 Group
Procedure Thirty mice were anesthetized in three batches of 10 as previously described. After 30 min they were removed to the home cage and recovery time was determined. Mice were considered to have recovered from the anesthetic when they were spontaneously ambulating and rearing. Mean recovery time was 8.5 min. Training was initiated 15 min after recovery, a time at which mice were behaviorally indistinguishable from control animals. Thirty control mice were trained after 45-min exposure
1 2 3 4 5 6
Treatment Before Training Air Air HAL HAL Air HAL
Treatment Before Testing
Test Time
None None None None HAL HAL
24 Hours 10 Minutes 24 Hours 10 Minutes 24 Hours 24 Hours
HAL: Exposure to Halothane for 30 min. Mice were trained or tested 15 min after recovery from anesthesia.
HALOTHANE-INDUCED AMNESIA IN MICE
LATENCY (SEC) 300[ 2501
~
HALOTHANE
NN
/
200[
oot=
150[-
50
0
10 min
24 hr
TRAIN
-
TEST INTERVAL
FIG. 2. Anterogradeamnesia.Micewere exposedto halothane and trained 15 min after recovery. Half of the group was tested 10 rain after training and half at 24 h. A control group left in room air was trained and tested similarly.
after training, substantial differences were evident by 24 h. At this time point, air-treated mice had developed a strong memory of the avoidance response, while halothane mice continued to show poor retention. The slow development of learned avoidance in untreated animals has been observed previously [e.g., (6,13)], and suggests that some avoidance behaviors require many hours to reach maximum strength. The results of exposure to halothane prior to the retention test are shown in Fig. 3. Groups 1 and 4 (Table 1) from Fig. 2 are included for comparison and analysis. A one-way ANOVA carried out on the data in Fig. 3 show that there was a significant difference in test scores among the four groups, F(3, 36) = 6.0, p = 0.002. Post hoc comparisons indicated that the HAL/HAL group had significantlybetter retention of passive avoidance than the HAL/AIR group, t(18) = 2.15, p = 0.04, and that the AIR/ HAL group had significantly poorer retention of passive avoidance than the AIR/AIR subjects, t(18) = 3.72, p = 0.001. While it appears that reexposure to halothane reverses the AA, an alternative, though unlikely, explanation should be ruled out. It is possible that exposure to halothane twice within 24 h is aversive, and that the long lateneies on the test day reflect conditioned aversion rather than restored memory. To evaluate this possibility, we tested two additional groups of 10 mice each. Procedure was the same as usual except that no shock was administered on the training trial. One group was anesthetized prior to training and the other was exposed to air. Before testing, 24-h later the halothane group was reexposed to the anesthetic and tested 15 min after recovery. Results showed no difference between the two groups (AIR = 17.9 + 1.2; HAL/HAL 19.9 + 1.7). These results show: 1. that animals trained following air exposure become amnestic if they are anesthetized with halothane before testing and 2. mice ordinarily amnestic following training after halothane exposure show a restoration of memory if they are reexposed to halothane before testing. These findings indicate that halothane anesthesia results in a symmetrical state-dependent retrieval failure. EXPERIMENT 3
When animals are treated with drugs prior to the training trial, it is frequently difficult to determine whether poor perfor-
451 mance on a subsequent test is the result of impaired acquisition or disrupted retention. A procedure which can sometimes help distinguish between these two possibilities is to test performance shortly after training (e.g., 30 min to 1 h) before information is transferred to long-term storage. For example, Kubanis, Gobbel, and Zornetzer (9) have shown that retention performance of old and young mice is similar 2 h after the training trial, but at 24 h the old animals exhibit a profound deficit. This finding indicates that the deficit in the aged rats is the result of abnormally rapid forgetting rather than impaired acquisition. The intention of the 10-min test in Experiment 2 was to provide information on the effects of halothane on acquisition processes, but because of the poor performance by both halothane- and air-treated mice on this test, the interpretation of these data is ambiguous. The objective of Experiment 3 was to examine the effects of halothane on the acquisition of step-through inhibitory avoidance. Procedure Twenty-five mice were the subjects for this experiment. Thirteen mice were exposed to halothane for 30 min as previously described and the 12 control animals were left in room air. As in Experiment 2, training was begun 15 min following recovery. The training session was begun by placing mice in the white compartment and raising the door. When the mouse entered the dark compartment the shock (0,2 mA) was initiated and remained on for the remainder of the training session. The mouse could escape the shock by running back into the white compartment. The door remained raised and mice received foot shock each time they stepped onto the bars of the dark compartment. Training was continued until mice remained in the white compartment for 100 s without a foot shock. The number of shocks taken until the learning criterion was attained was recorded for all animals. Retention was tested 24 h later. Results Mean number of shocks taken to attain criterion was 12.75 for the air-exposed mice and 12.41 for the halothane-treated subjects. Twenty-four h retention test scores were 207.6 + 27.2 s for the air controls and 112.7 + 26.7 s for the halothane-exposed subjects. This difference was statistically significant, t(23) = 2.488, p = 0.021. These data indicate that mice anesthetized
300
LATENCY(SEC) ~lB
250
TRAINING TESTING
200
150
T
100
50
0 BEFORE TRAINING BEFORE TESTING
AIR AIR
HALOTHANE AIR HALOTHANE AIR HALOTHANE HALOTHANE
FIG. 3. State dependent learning. Mice were trained after exposure to halothane or air. Half of each group received halothane or air before testing.
ROSMAN ET AL.
452 with halothane learned the avoidance response normally but forgot that learning significantly faster than the nonexposed control subjects.
TEST LATENCY (SEC) 140 120
EXPERIMENT 4
The purpose of this experiment was to determine the temporal gradient ofhalothane-induced anterograde amnesia. To accomplish this, we trained different groups of mice at varying times after recovery from anesthesia and tested for retention 24 h after training. Procedure
tO0 80 60 40
Forty mice were anesthetized in batches of 10 as described above and randomly assigned to one of four groups (N = 10 per group) which were trained either 15, 120, 240, or 1440 rain after recovery from halothane. A control group (N = 10) was not anesthetized. Retention was tested 24 h following training. Results The test latencies of the five treatment groups are shown in Fig. 4. A one-way ANOVA applied to these data revealed a significant difference among the test latencies of the five groups, F(4, 45) = 3.07, p = 0.025. Inspection of Fig. 4 shows that amnesia occurs if mice are trained up to 2 h after recovery from halothane, but retention is normal if training is delayed for 4 h, 120 vs. 240 min, t(18) = 2.14, p -- 0,044. DISCUSSION
The principal finding of this study is that halothane anesthesia causes state-dependent anterograde amnesia. If training is initiated within 2 h following anesthesia, learned information is not accessible to normal retrieval cues at the retention test. However, if mice are reexposed to the anesthesia prior to the retention test, restoration of the memory of the avoidance response occurs. This finding, in conjunction with the observation that unanesthetized mice become amnestic if exposed to halothane before the test, is evidence that halothane causes memory failure by disrupting retrieval processes. The symmetrical state dependence can be most readily explained by supposing that halothane anesthesia produces an altered central state accompanied by novel interoceptive stimuli which changes the way the animal perceives the training environment. These altered stimulus conditions form the context within which the avoidance response is learned and they become incorporated, along with other contextual features, as attributes of the memory. These contextual stimuli are known to function as retrieval cues, and a sufficient number have to be available at testing in order for the event to be remembered. Amnesia occurs 24 h after halothane because the central state has changed and the stimulus conditions no longer match those which existed at the time of training. Reexposing mice to halothane prior to testing restores remembering because the anesthesia reinstates a significant part of the context within which the original learning took place. The demonstration that retrieval is disrupted by halothane should not be taken to imply that encoding processes are unaffected by the anesthetic. The results of the present experiment make it clear that it is the interaction between encoding and retrieval which determines retention performance. Whether mice remember being shocked in the dark compartment depends on the conditions (halothane or air cues) which exist at the time of
20 0 tSmin
2 Hrs
4 hra
24 hra
Control
Recovery-Training Interval FIG. 4. Temporal gradient of anterograde amnesia, Mice were anesthetized with halothane and trained either 15 rain, 2, 4 or 24 h after recovery from anesthesia. Retention was tested 24 h after training.
testing. Furthermore, the effectiveness of cues in the test environment in retrieving the memory of shock depends on the conditions (halothane or air) under which original learning took place. A similar phenomenon has been reported in anestheticinduced amnesia in man (1). In this study subjects were read a list of words during recovery from a period of anesthesia. Retention was tested by a forced-choice recognition test in which errors could be categorized as acoustically similar to the item, semantically similar, or unrelated. Analysis of errors showed that subjects selected significantly more acoustically similar items than would be expected under undrugged conditions, suggesting that they had been encoding information in terms of the acoustical rather than the semantic features of the material. Amnesia occurred in this study because the cues used during encoding did not match the cues used at the retrieval test. Taken together, these two studies suggest that explanation of the disrupting effects of anesthetics on remembering should be sought in the interaction of encoding and retrieval processes rather in either process alone (13). The clear absence of retrograde amnesia in this study is noteworthy. This finding confirms a similar experiment using mice as subjects (4), but conflicts with a rat experiment which demonstrated a gradient of retrograde amnesia (12). The demonstration of a retrograde disruption of performance with halothane is not strong evidence for amnesia without the inclusion of control groups for the aversive effects of anesthetics which are known to occur in rats (3,7). Overall, it does not appear that halothaneinduced retrograde amnesia is a robust phenomenon. The alterations in neural functioning which underlie the halothane-induced retrieval deficit remain to be determined. A research strategy which may reveal which neurochemical systems are involved in the retrieval failure is to attempt to block or reverse the amnesia by administering selected pharmacological agents (e.g., cholinergic agonists, calcium channel blockers, GABA antagonists) before anesthesia, after learning or before testing. This line of investigation is currently being pursued in our laboratory.
HALOTHANE-INDUCED
AMNESIA IN MICE
453
REFERENCES 1. Adam, N. Effects of general anesthetics on memory functions in man. J. Comp. Physiol. Psychol. 83:294-305; 1973. 2. Adam, N.; Castro, A. D.; Clark, D. L. State-dependent learning with a general anesthetic (Isoflurane) in man. Life Sci. 4:125134; 1974. 3. Alexinsky, T.; Chapouthier, G. Halothane anesthesia and DMS performance in rats: Memory impairment or avoidance behavior? Physiol. Behav. 22:99-105; 1979. 4. Angel, C.; Bounds, H. M., Jr.; Perry, A. A comparison of the effects of halothane on blood-brain barrier and memory consolidation. Dis. Nerv. Syst. 33(2):87-93; 1972. 5. Cherkin, A.; Lee-Teng, E. Interruption by halothane of memory consolidation in chicks. Fed. Proc. 24:328; 1965. 6. Deutsch, J. A.; Rogers, J. B. Cholinergic excitability and memory: Animal studies and their clinical implications. In: Davis, K. L.; Berger, P. A., eds. Brain acetylcholine and neuropsychiatric disease. New York: Plenum Press; 1979:175-204, 7. Gisquest-Verrier, P. Accelerated extinction after post trial halothane anesthesia: An aversive effect. Physiol. Behav. 26:223-231 ; 1981.
8. Gutekunst, R.; Youniss, J. Interruption of imprinting following anesthesia. Percept. Mot. Skills 16:340; 1963. 9. Kubanis, P.; Gobbel, G.; Zornetzer, S. F. Age related memory deficits in swiss mice. Behav. Neural Biol. 32:241-247; 1981. 10. Lecanuet, J. P.; Alexinsky, T.; Cbapouthier, G. The following response in chicks: Conditions for the resistance of consolidation to a disruptive agent. Behav. Biol. 12:365-372; 1974. 11. Lecanuet, J. P.; Deweer, B.; Bloch, V. Effect of post-exposure anesthesia on the retention of imprinting. Behav. Biol. 12:365-372; 1974. 12. Penrod, W. C.; Boice, R. Effects of halothane anesthesia on the retention of passive avoidance task in rats. Psychon. Sci. 23:205207; 1971. 13. Sehacter, D. L.; Tulving, E. Amnesia and memory research. In: Cermak, L. S., ed. Human memory and amnesia. Hillsdale, N J: Lawrence Erlbaum Assoc. Inc.; 1982:1-32. 14. Schmaltz, G. Can halothane anesthesia have any aversive effects on the rat? Physiol. Behav. 22:25-29; 1979. 15. Zinkin, S.; Lecanuet, J. P.; Deweer, B. Retroactive and proactive effects of anesthesia on following in chicks. Physiol. Behav. 16:185189; 1976.
