Dissociating learning and performance: Drug and hormone enhancement of memory storage

0361-9230189 $3.00 + .oO

Brain Research BuNefin, Vol. 23, pp. 339-345. 0 Pergamon Press pk. 1989. Printed in the U.S.A

MEETING REPORT

Dissociating Learning and Performance: Drug and Hormone Enhancement of Memory Storage JAMES L. McGAUGH

Center for the Neurobiology of Learning and Memory and Department of Psychobiology, University of California, Irvine, CA 92717

McGAUGH, J. L. Dissociating learning andperformance: Drug and hormone enhancement of memory storage. BRAIN RES BULL 23(4/5) 339-345, 1989.-This paper reviews selected studies examining the enhancing effects of drugs and hormoneson learning and memory. Many strategies have been used in an effort to dissociate drug effects on learning from drug effects on other processes affecting the performance of responses. These strategies include the use of tasks with various motivational and response requirements, the use of studies explicitly examining drug influences on performance, the use of posttraining drug administration and the use of various forms of latent learning tasks. It seems clear from these studies that the dissociation of learning and performance effects of drugs cannot rest on one task or one experiment. Overall, the evidence summarized in this paper provides strong support for the conclusion that drugs can and do enhance retention and that the effects are due to influences on memory storage rather than to other factors that influence performance. Drugs CNS stimulants Adrenergic antagonists Learning Latent extinction Latent inhibition State-dependency Performance

GABA antagonists Memory Posttraining

Hormones treatments

Latent learning Opiate antagonists

difficult in studies examining the effects of treatments that alter neurobiological systems. If a drug, for example, alters performance on a learning or memory task, it is, at the very least, risky to conclude that the drug affects performance because it alters the neurobiological systems underlying learning and memory. In such studies controls are obviously needed to rule out the contribution of other influences of the drug that might affect performance on the task. The problem of determining the basis or bases of drug enhancement of performance on a learning task is illustrated by the first (to my knowledge) study reporting drug enhancement of learning (31). In this study, Lashley injected rats with strychnine sulphate each day before they were trained on a maze and found that they made fewer errors than control rats, as they acquired the task. If one is only interested in concluding that acquisition performance can be improved on a cognitive task, then perhaps Lashley’s findings constitute acceptable evidence. But, the findings are clearly incomplete if one is, as most investigators are, interested in knowing the basis of the improvement. Many additional questions need to be asked. Does the drug alter sensory, motivational or motor processes in ways that would produce improved maze performance? For example, the drug might increase hunger or alter taste (the rats were rewarded with food). The drug might also increase alertness or alter motor performance (44,

IS it possible to improve learning and memory with drugs? This question is the focus of much current research investigating the effects, on learning and memory, of treatments affecting neurochemical systems. Much of the research stems from an interest, either explicit or implicit, in finding treatments for disorders of learning and memory-particularly those associated with aging. Research addressing this question is also guided by a more general interest in understanding the neurochemical systems involved in learning and memory. Regardless of the basis of the interest, all studies of drug influences on learning and memory face the difficult task of determining how drugs act to improve performance on tasks that are used to measure learning and memory. The distinction between learning and performance originally offered by Tolman (54) is accepted by most investigators as essential for understanding cognitive processes. Learning and memory are inferred from experience-induced changes in behavior. The problem faced by the investigator is that of determining that the changes in performance are due to learning and memory rather than to other influences. As performance is influenced by many conditions, including arousal, sensitization, fatigue, illness, and so forth, distinguishing changes resulting from the acquisition of information from those due to other influences is not a simple task for any study of learning and memory. The difficulty of distinguishing between learning and performance becomes more

339

310

53). If drugs are administered prior to training. controls are needed for each of these possible influences on maze performance if one is to be comfortable (as well as correct) concluding that the drug improves performance by enhancing neural processes underlying learning. CAN

DRI!G

EFFECTS

ON LEARNING

AND

MEMORY

BE DISSOCIATED’?

The questions raised by Lashley’s study were addressed by many subsequent studies of the effects of strychnine on learning (35.45). I will briefly review some of the major findings of these early experiments as the issues addressed by these studies, as well as the experimental approaches used. remain relevant for current studies of drug influences on learning and memory. Task Generali~ One approach to the problem of determining the basis of drug effects on performance in learning tasks is that of examining the effects of the drug in tasks that differ in motivational and response requirements. If it is proposed, for example, that the performance enhancement is due specifically to an increase in hunger or to a reduction in locomotor activity then the drug would not be expected to enhance performance learning tasks using shock motivation or tasks in which learning requires increases in locomotor activity. Thus, an examination of task generality of the effect is essential for studies of drug influences on learning. The findings of studies examining the effects of strychnine administered prior to training indicate that strychnine enhances learning performance in a wide variety of tasks including maze learning (43), aversively (46) as well as appetitively motivated (41,51) discrimination learning (both successive and simultaneous discrimination), as well as escape learning (29) and habituation (1). Such findings are consistent with the interpretation that strychnine improves performance by enhancing learning. Or, more conservatively. the findings argue against many. but perhaps not all alternative interpretations stressing specific motivational or performance effects. The problem facing the experimenter is that learning and memory are defined by exclusion. And it is difficult. at best, to determine whether all possible, or even reasonable. alternative interpretations have been excluded. Drug hfluences

on Performance

Another approach to the problem is that of examining the effect of drug and control injections on performance of the task during and after acquisition. For example, after rats have learned a maze while under the influence of strychnine do they still perform well when tested without the drug? And, is the performance of trained control animals influenced by administering the drug prior to being tested in the maze? In his original study Lashley (3 1) reported that when strychnine and a control solution were administered to trained rats on alternate days, maze performance was better on days when the animals received strychnine. However, considered alone, the finding that strychnine enhanced performance does not allow the conclusion that strychnine acts only through effects on performance. Strychnine might affect both learning and performance. The critical question is whether the difference between the performance of drug and control animals during training is due entirely to drug influences on performance. That question could, of course, be addressed. Studies addressing the question would need to include controls for state-dependency as differences in performance might result simply from changes in drug state during training and testing. Latent Learning The distinction

between

learning and performance

was first

addressed experimentally in the well-known “latent learning’ studies conducted by Tolman and his students (55,55). In these studies. errors on a maze were significantly reduced when a reward was introduced into the goal box following several nonrewarded trials. Clearly the animals were acquiring information during the earlier trials. But the learning was latent until the reward was introduced. Latent learning tasks provide a powerful means of distinguishing the effects of drugs on learning from their effects on performance. As is discussed below, several varieties of latent learning tasks have, in fact, been used. And, in each case, they have provided evidence that drugs can influence the storage of information independently of their direct effects on performance. For example, in one study (56) rats were injected with a strychnine-like drug each day for several days immediately after either a rewarded or nonrewarded training trial in a maze (see below for a discussion of the rationale underlying the use of posttraining injections). The drug enhanced the learning of rewarded groups. However. the drug did not affect the performance of the nonrewarded groups. Then, for several additional trials, the drug injections were discontinued and all animals were rewarded. On these trials the errors of both groups that had received the drug immediately after each of the earlier trials were lower than those of both control groups. Thus, the drug enhanced the latent learning that occurred on the early nonrewarded trials. But. the fact that the drug did not improve performance when administered after nonrewarded trials indicates that the drug injections were not rewarding. These findings clearly argue that this drug improves ieaming by enhancing the storage of information. Of course. this drug might also influence performance when administered prior to testing. But, by using latent learning procedures the effects on learning and performance can be experimentally distinguished. Posttraining Administrution The experiment discussed above combined two techniques that are used to distinguish learning and performance in studies of drug effects on learning: latent learning and posttraining drug administration. The use of posttraining administration in studies of drug enhancement of learning and memory is based on the general assumption that the processes underlying the storage of information are initiated by training and continue for some period following the completion of a training experience (39.50). Thus, according to this assumption it should be possible to modulate memory storage by administering treatments shortly after training. Further, treatments administered several hours following training should be ineffective even though they are closer in time to the subsequent retention test. The first studies examining the effects of posttraining injections on memory used strychnine and other central nervous system stimulants (2. 3. 47, 48). There is extensive evidence indicating that such treatments enhance memory in a time-dependent manner-that is, delayed injections are ineffective (41). Furthermore, posttraining injections of such treatments have been found to enhance retention of a wide variety of training tasks (36, 39, 45). The findings of studies using posttraining injections provide strong evidence that drugs can affect memory without directly affecting performance during acquisition and retention testing. Drug effects on performance are excluded as possible contributions to the enhanced retention. The use of posttraining injections together with latent learning tasks has provided the clearest evidence that drugs can enhance the storage of information (22.56). The central feature of a latent learning task is that a first phase of training provides an opportunity for learning without the influence of reinforcement on performance followed by a test to determine the degree of learning. In a sensory preconditioning

DRUG EFFECTS ON PERFORMANCE

task, for example, two stimuli (e.g., a light and sound) are presented together on several trials. Then, one of the stimuli (e.g., the tone) is paired with a reinforcing stimulus (e.g., a shock) using classical conditioning procedures. Finally, the other stimulus (e.g., the light) is presented to see if it will elicit the conditioned response. Using such procedures, Humphrey (22) found that strychnine administered immediately after the initial pairing of the two stimuli enhanced the association (sensory preconditioning) between the two stimuli as indicated by the effectiveness of the light in eliciting the conditioned response on the final test. These findings are highly comparable to those obtained in the latent learning maze study summarized above. Clearly, drug enhancement of memory does not require the presence of the drug during either training or retention testing. The use of posttraining drug administration does not, however, eliminate all possible nonassociative effects of drugs on retention performance. As I noted above, it is possible that a drug might have rewarding or punishing effects independently of its effects on information storage. The findings of the latent learning study discussed above clearly indicated that the drug enhanced learning but was not rewarding. But, the issue must be addressed with each drug examined. Thus, studies using posttraining injections typically include nomeinforced controls to assess possible rewarding and punishing effects. It has also been suggested that posttraining injections may affect performance by altering perceptual states following training (6). Furthermore, it has been argued that recently acquired information may be stored in a brain state induced by the posttraining injection (28). These issues are discussed further below. INVOLVEMENT

OF NEUROMODULATORY

341

AND MEMORY

SYSTEMS

IN MEMORY

STORAGE

The early experiments examining the effects of strychnine and other CNS stimulants on learning provided essential evidence arguing that it is possible to distinguish between learning and performance effects of drugs. Studies using different tasks, posttraining injections, and various forms of latent learning have provided rather convincing evidence in support of the conclusion that drugs can enhance memory storage. More importantly, however, the findings of these early experiments encouraged the use of drugs and other posttraining treatments in the investigation of the involvement of neuromodulatory systems in memory storage. There is now extensive evidence indicating that retention can be enhanced by drugs affecting a number of neurochemical systems, including catecholaminergic, opioid peptidergic, cholinergic and GABAergic systems (38). The general assumption motivating this research is that drugs can be used to understand the involvement of these systems in learning and memory. Thus, in this research dissociation of learning and performance effects is absolutely critical to the interpretation of the findings. Involvement of GABAergic Systems An early study from my laboratory (3) was the first to suggest the possible involvement of the GABAergic system in memory. In that study we found that posttraining injections of the GABAergic antagonist picrotoxin enhanced rats’ maze learning. This basic finding was confirmed and extended by findings from other laboratories (4, 10, 13, 18, 36, 57) as well as by more recent findings from my laboratory. Briefly, the findings of studies of the effects of GABAergic antagonists indicate that: 1) time-dependent memory enhancement is produced by posttraining injections (12); 2) enhancement is obtained in several types of tasks (24); 3) the memory-enhancing effects are not due to rewarding or punishing effects (4); 4) the posttraining memory-enhancing effects are seen

1

-

3.25

Picrotoxin

(mglkg)

PIG. 1. Dose-dependent and time-dependent effects of p&training injections of picrotoxin (II’) on retention, by mice, of a one-trial inhibitory avoidance task. Each bar indicates median response latency ( 2 interquartile range) on the retention test trial. From (12).

in a type of latent learning task as well as in conventional learning tasks (unpublished findings); and 5) the effects are not due to induction of posttraining state-dependency (12). Thus, by exclusion, the findings provide strong support for the view that GABAergic antagonists enhance memory storage. As a peripherally acting GABAergic antagonist is ineffective (4), while direct injections into a region of the brain (amygdaloid complex) are effective (5) the findings also argue that the drugs enhance memory by directly modulating brain systems. Figure 1 shows the findings of a study, using mice, of the effects of posttraining injections of picrotoxin on retention of a one-trial inhibitory avoidance task. The effects are both dose dependent and time dependent. And posttraining injections do not affect the retention performance of animals unless they received footshock on the training trial (12). Similar results have been obtained with the GABAergic antagonist bicuculline (4). And, while it is not the focus of this paper, retention is impaired by posttraining administration of the GABAergic agonists baclofen and muscimol (9,lO). Figure 2 shows the effects of posttraining injections’ of bicuculline on retention of a Y-maze discrimination task (4). In this task mice were trained on several trials to escape from footshock by entering one arm of the Y-maze. On the retention test the position of the safe arm was reversed and retention was assessed by the number of errors made on 8 reversal training trials. The experiment was based on other evidence indicating that errors on the discrimination reversal test increased directly with the amount of original training. Thus, if posttraining administration enhances learning of the discrimination, the effects should be comparable to those produced by additional training. Clearly, the posttraining injections produced dose-dependent enhancement of retention of the original training, as indicated by errors on the retention test. With this task, as well as with the inhibitory avoidance task, retention was not affected by posttraining intraperitoneal injections of bicuculline methiodide, a GABAergic antagonist that does not pass the blood-brain barrier (4). Izquierdo (27,28) has reported evidence suggesting that the information acquired on training may be stored in the brain state induced by the posttraining treatment. Under some conditions, the retrograde amnesia induced by posttraining treatments can be attenuated by administration of the same treatments prior to the retention test (27,28). While state-dependency is not seen with all posttraining treatments that produce amnesia (11,42) the finding of

ElCUCULLlNE Y-MAZE

-

_

FIG. 2. Effects of posttraining injections of bicuculline (IP) on retention, by mice, of the Y-maze ~sc~~nation task. Each bar indicates the median errors ( A interquartile range) on the disc~mination reversat retention test. *p
state-dependency under at least some conditions suggests the possibility that retention enhancement produced by posttraining treatments might result from state-dependency rather than from e~~~ernent of memory storage processes. The basic assumption

underlying the state-dependent hypothesis is that retention performance reflects the degree to which the brain state at the time of retention testing is congruent with the state that normally occurs or is induced following training. In studies of retention enhancement produced by posttraining treatments animals are usually not given additional treatments prior to the retention test. Thus a statedependent inte~~tation of the retention e~~cement produced by a posttraining treatment would require an ad hoc assumption that the animal’s state normally occurring at the time of the retention test is at least somewhat congruent with the state induced by the posttraining treatment. According to this interpretation, administration of the same retention-enhancing treatment prior to the retention test would be expected to decrease the congruence and, thus, attenuate the effect of the post~aining treatment. Figure 3 shows the results of an experiment (12) in which mice received either saline or picrotoxin immediately posttraining and either saline or picrotoxin, in different doses, at different times prior to the retention test. The results are clear-cut. Retention is enhanced by posttraining picrotoxin but is not influenced by picrotoxin administered prior to the retention test. Thus, the findings argue against a state-dependent inte~retation of the effects of picrotoxin on memory. And, furthermore, they indicate that picrotoxin does not affect retention performance when administered prior to the retention test, Such findings are consisent with those of other findings in supporting the view that picrotoxin, as well as other GABAergic antagonists, enhance retention by modulating memory storage processes. This inte~retation is also supported by a recent study in which we examined the effect of posttraining administration of picrotoxin on the latent extinction of a conditioned emotional response (CER) (unpublished findings). In this study (see Fig. 4) mice were enclosed in one arm of a Y-maze and received 20 sequences of tone paired with shock. On the following day some of the animals received 20 latent extinction trials; that is they were again placed in the apparatus and were presented with tones but no shock. Picrotoxin or saline was administered immediately following the latent extinction training. Animals that were not given extinction trials also received either picrotoxin or saline on day 2. On the third day, all animals were placed in the maze and locomotor

FIG. 3. Effects of two doses of picrotoxin (IP) or saline administered posttraining and at three time intervals prior to retention testing on retention performance of mice on a one-trial inhibitory avoidance task. Each bar indicates the median response latency ( ? interquartile range) on the retention test trial. From (12).

activity was measured in the presence of the tone. As can be seen: 1) the CER training reduced locomotor activity; 2) the latent extinction training attenuated the effect of the CER training: and 3)

posttraining picrotoxin enhanced latent extinction of the CER; 4) picrotoxin administered on day 2 had no effect in the absence of extinction training. In recent studies we have also found that retention is modulated by injections of GABAergic agonists and

q NFS @ FS

* cr

_L-

_.L-

/ t*

l

i

S&L

PIG

)N

NO EXTINC TIC

10

EXTINCTION

FIG. 4. Effect of picrotoxin on extinction of a conditioned emotional response (CER). The mice were given 20 pairings of tone + shock (shaded bars) or tone without shock (open bars) in one alley of a Y-maze) on day one, and given either extinction training (tone alone while placed in the maze alley) or no extinction training, followed by an injection of picrotoxin (IP), on day two. Bars indicate mean alley entries (+ standard errors) when allowed to explore the maze on day three. On day three, the exploratory responses of mice given picrotoxin following the extinction training on day 2 was comparable to that of nonshocked controls. FS=footshock, NFS= no footshock. **p
DRUG EFFECTS ON PERFORMANCE

343

AND MEMORY

antagonists into the amygdaloid complex (5). These latter findings are consistent with evidence from other recent experiments summarized below suggesting that neuromodulatory systems within the amygdaloid complex are involved in the regulation of memory storage. Involvement of Adrenergic Systems The evidence of time-dependency in memory storage has suggested that memory storage may normally be regulated by endogenous systems activated by learning experiences (19.30). Extensive evidence suggests that hormones released in response to stimulation and stress modulate memory storage. Experiments examining the effects of hormones on learning and memory have examined a number of hormones, including epinephrine, ACTH, vasopressin, opioid peptides, CCK, substance P as well as other peptide hormones (37,38). Most of the work in my laboratory has focussed on adrenergic and opioid peptide systems. In this section and the following section I will briefly review some of the recent findings concerning the involvement of these systems in learning and memory. Gold and van Buskirk’s report (20) that posttraining injections of epinephrine produced dose-dependent and time-dependent enhancement of memory in an inhibitory avoidance task stimulated extensive research examining the involvement of adrenergic systems in memory (37). Such a quest assumes, of course, that the effects of epinephrine seen in such studies are due to influences on learning rather than to other nonassociative processes that affect retention performance. In general, the findings of studies that have addressed this issue have provided strong evidence that epinephrine influences memory storage. The findings indicate that: 1) posttraining injections of epinephrine produce dose-dependent and time-dependent enhancement of retention (20); 2) retention enhancement is seen in a variety of aversively motivated training tasks, including inhibitory avoidance, active avoidance, discrimination learning, as well as appetitively motivated tasks (20, 21, 24, 32, 52); 3) epinephrine injections alone are neither rewarding or punishing (34); 4) the memory-enhancing effects are longlasting (24). The long-lasting effects of a single posttraining injection of epinephrine are shown in Fig. 5. Mice in this study (24) were trained on the Y-maze discrimination task described above and injected immediately following training with either saline or a low dose or high dose of epinephrine. Independent groups were then tested for retention, using performance on discrimination reversal as the index of retention, one day, one week, or one month later. Comparable results were obtained at all retention intervals: The low dose of epinephrine enhanced learning while the high dose impaired retention. The findings obtained on the one day and one week tests are highly comparable to those obtained with other types of training tasks. This is the only experiment, to date, which has examined the retention-modulating effects of epinephrine at a retention interval of greater than one week. The finding of comparable effects at all retention intervals provides additional support for the view that epinephrine affects retention performance by strengthening processes underlying the long-term storage of information. As I indicated earlier, studies of the effects of epinephrine on learning are guided by an interest in understanding how this adrenal medullary hormone acts to influence memory storage. While this issue is not the focus of the present discussion, some recent findings are relevant to the issues raised here, The effects of peripherally administered epinephrine are blocked by intra-amygdala injections of the adrenergic antagonist propranolol. Further, retention of an inhibitory avoidance task is enhanced by intra-amygdala

Legend 0

SALINE

rZa EPI 0.3 mg/kg iZZ EPI 1.0 mg/kg

24 HR RETENTION

INTERVAL

FIG.

5. Effects of posttraining epinepbrine (IP) on retention, in mice, of the Y-maze discrimination task one day, one week, and one month following original training. Each bar indicates mean errors ( 2 SEM) on the

discrimination

reversal retention test. From (24).

injections of norepinephrine in a dose-dependent and time-dependent manner (33). These findings, considered together with other findings suggest that epinephrine effects on memory involve the release of central norepinephrine (25, 33,40). More generally, the findings are consistent with evidence that GABAergic influences on memory, as discussed above, as well as opioid peptidergic influences, as discussed below, involve activation of systems within the amygdaloid complex. Involvement of Opioid Peptidergic Systems Studies of the memory-modulating effects of posttraining injections of opioid peptide agonists and antagonists have provided additional evidence that memory storage is regulated by endogenous neuromodulatory systems (7, 8, 14, 26, 49). In general, posttraining administration of opioid agonists such as morphine and P-endorphin impair retention while opioid antagonists such as naloxone and naltrexone enhance memory. The effects of opiate antagonists are generally comparable to those seen with strychnine, picrotoxin and epinephrine: 1) the memory-enhancing effects of posttraining injections are dose dependent and time dependent (23); 2) Evidence of retention enhancement with posttraining treatments has been obtained with a wide variety of tasks, including inhibitory avoidance, active avoidance, discrimination learning, habituation and spatial learning (17, 26, 40); 3) In doses that enhance retention, opiate antagonists are neither rewarding nor punishing when administered alone (23); 4) Retention enhancement is obtained with a form of latent learning as well as explicit learning tasks. This latter effect was shown in a study by Gallagher and her colleagues (14) using a latent inhibition procedure. The term latent inhibition refers to the decreased effectiveness of a stimulus in conditioning if the stimulus has previously been presented without the unconditioned stimulus. That is, the previous learning that the stimulus lacks significance lessens its effectiveness as a CS. In the Gallagher et al. study naloxone or saline was administered to rabbits immediately following the preexposure of a tone which was later to be used as a CS in a

344

heart-rate conditioning experiment. Latent inhibition was evidenced by retarded rate of conditioning to the preexposed tone, in comparison with conditioning obtained with animals not preexposed to the tone. Latent inhibition was enhanced by posttraining naloxone: no conditioning was obtained in animals given naloxone following the preexposure trials. The major focus of studies investigating the effects of opiate antagonists on memory concerns the brain systems through which the drugs work to influence memory. The findings of a number of studies suggest that these drugs work through influences involving. at least in part. the amygdaloid complex. Gallagher and her colleagues have found that retention is enhanced by posttraining intra-amygdala injections of opiate antagonists and impaired by noradrenergic antagonists (15.16). Findings from my laboratory suggest that naloxone effects on memory involve the release of NE within the amygdala. The memory-enhancing effects of systemically as well as intra-amygdally injected naloxone are blocked by B-noradrenergic antagonists in doses that are ineffective when administered alone (25,40). But, such conclusions require the assumption that opiate antagonist effects on memory are due to influences on the storage of information rather than to nonassociative processes affecting the performance of learned responses. Thus. the experiments examining alternative interpretations are clearly critical to the neurobiological inquiry. DISSOCIATING LEARNING AND PERFORMANCE: WHAT CAN WE CONCLUDE’

As I noted in the introduction, studies of drug enhancement of learning and memory are generally motivated by an interest in improving and/or understanding brain processes underlying the storage of information. Thus, it is essential that learning and memory be dissociated from performance. But, it is clear from the studies summarized in this paper that such an effort, at best, is not simple. For each drug of interest, experiments are required to examine dose and time dependency (the latter if posttraining treatments are to be used) as well as possible rewarding and state-dependency effects. Furthermore, the use of a variety of tasks is essential in order to eliminate interpretations restricted to one type of learning task. Novel hypotheses concerning drug effects on learning do appear from time to time. The proposal that drugs may act to

induce posttraining state-dependency is one recent example (27). As the findings of our study using picrotoxin discussed above indicate, this hypothesis is clearly inadequate as an explanation of drug enhancement of learning. In a recent paper Carey (6) proposed that posttraining treatments may affect subsequent retention simply because they distort animals’ perception of the stimuli used in training. Thus. according to this vie&. posttraining treatments may attenuate or enhance the perceived magnitude of the noxious stimulus used in training. This interpretation was offered as an explanation of some of the evidence from studies using inhibitory avoidance tasks. In such studies. for example retention can be increased either by increasing the footshock intensity used in training or by administering a posttraining injection of a drug or hormone. However. this example serves to illustrate the importance of the various types of experiments that are used to differentiate learning and performance effects of drugs. While the perceptual distortion hypothesis of posttraining enhancement effects might fit well with a limited set of observations based on inhibitory avoidance training, it clearly is inadequate as an explanation of drug enhancement of learning seen in various forms of latent learning tasks. How would posttraining perceptual distortion provide an account of latent learning in a maze, latent inhibition, and latent extinction’? .4nd, if the posttraining treatments ‘I simulate the physiological consequences of the noxious stimulation used to induce passive avoidance behavior” (6). then why do such treatments enhance retention in appetitively motivated task.j? In summary, it is clear that the dissociation of learning and performance requires a variety of experimental approaches. It cannot be accomplished by any one learning task or by any one experiment alone, as learning and memory arc inferred. not observed, and are based on exclusionary evidence. With this caveat, the evidence summarized in this paper provides strong support for the conclusion that drugs and homrones can and do enhance retention and that the effects are due to influences on memory storage rather than to other factors that influence performance. ACKNOWLEDGEMENTS

Supported in part by USPHS Grant MH12526 from National Institute of Mental Health and National Institute of Drug Abuse, and Contract NO00 14-87-K-05 18 from the Office of Naval Research.

REFERENCES Andry, D. K.; Luttges, M. W. Facilitated habituation: Strychnine dose-response effects on neural and behavioral habituation. Agents Actions 2/3:103-l 17; 1971. Bovet, D.; McGaugh, .I. L.; Oliverio, A. Effects of posttrial administration of drugs on avoidance learning of mice. Life Sci. 5: 1309-131s; 1966. Breen, R. A.; McGaugh, .I. L. Facilitation of maze learning with posttrial injections of picrotoxin. J. Comp. Physiol. Psychol. 54: 495-501; 1961. Brioni, J. D.; McGaugh, .I. L. Posttraining administration of GABAergic antagonists enhance retention of aversively motivated tasks. Psychopharmacology 96:505-510; 1988. Brioni, J. D.; Nagahara, A. H.; McGaugh, J. L. Involvement of the amygdala GABAergic system in the modulation of memory storage. Brain Res. 487:105-l 12; 1989. Carey, R. J. Post-trial hormonal treatment effects: memory modulation or perceptual distortion? J. Neurosci. Methods 22:27-30; 1987. Castellano, C. Effects of morphine and heroin on discrimination learning and consolidation in mice. Psychopharmacologia 42:235242; 1975. Castellano, C. Strain-dependent effects of naloxone on discrimination learning in mice. Psychopharmacology (Berlin) 73:152-156; 1981. Castellano. C.; Brioni, J. D.; Nagahara, A. H.; McGaugh, J. L.

IO

II

12

13

14

15 16

Post-training systemic and intra-amygdala administration of the GABAb agonist baclofen impair retention. Behav. Neural Biol. 52:170-179: 1989. Castellano, C.; Introini-Collison, I. B.; Pavone, F.; McGaugh, J. L. Effects of naloxone and naltrexone on memory consolidation in CD1 mice: Involvement of GABAergic mechanisms. Pharmacol. Biochem. Behav. 32:563-567; 1989. Castellano, C.; McGaugh, J. L. Effect of morphine on one-trial inhibitory avoidance in mice: Lack of state dependency. Psychobiology 17:89-92; 1989. Castellano, C.: McGaugh, J. L. Retention enhancement with posttraining picrotoxin: Lack of state dependency. Behav. Neural Biol. 51:165-170; 1989. Castellano, C.; Pavone, F. Effects of ethanol on passive avoidance behavior in the mouse: Involvement of GABAergic mechanisms. Pharmacol. Biochem. Behav. 29:321-324; 1988. Gallagher, M.; Fanelli, R. J.; Bostock, E. Opioid peptides: Their position among other neuroregulators of memory. In: McGaugh, J. L., ed. Contemporary psychology: Biological processes and theoretical issues. Amsterdam: Elsevier/North Holland; 1985:69-93. Gallagher, M.; Kapp, B. S. Manipulation of opiate activity in the amygdala alters memory processes. Life Sci. 23:1973-1978; 1978. Gallagher, M.; Kapp, B. S.; Pascoe, J. P.; Rapp. P. R. A neuro-

DRUG EFFECTS ON PERFORMANCE

17. 18.

19.

20. 21.

22.

23,

24.

25.

26.

27,

28. 29. 30 31. 32. 33.

34. 35.

AND MEMORY

pharmacology of amygdaloid systems which contribute to learning and memory. In: Ben-Ari, Y,, ed. The amygdaloid complex. Amsterdam: Elsevier/Nortb Holland; 1981:343-354. Gallagher, M.; King, R. A.; Young, N. B. Opiate antagonists improve spatial memory. Science 221:975-976; 1983. Garg, M.; Holland, H. C. Consoli~tion and maze learning: The effects of posttrial injections of a stimulant drug (picmtoxin). Psychopharmacologia 12:96-103; 1968. Gold, P. E.; McGaugh, J. L. A single-trace, two-process view of memory storage processes. In: Deutsch, D.; Deutsch, J. A., eds. Short-term memory. New York Academic Press; 1975:355-378. Gold, P. E.; van Buskirk, R. Facilitation of time-dependent memory processes with posttrial epinephrine injections. Behav. Biol. 13: 145-153; 1975. Gold, P. E.; van Buskirk, R.; Haycock, J. Effects of post~aining epinephrine injections on retention of avoidance training in mice. Behav. Biol. 20:197-204; 1977. Humphrey, G. L. Effects of post-training strychnine on memory of stage I and stage II of sensory preconditioning in rats. Ph.D. thesis, University of California, Los Angeles, 1968. Intmini, I. B. Participation de peptidos opioides endogenos en el proceso de consolidation de1 la memoria. Su posibie interaction con otros sistemas neuronales. Ph.D. thesis, University of Buenos Aires, 1984. Introini-Collison, I. B.; McGaugh, J. L. Epinephrine modulates long-term retention of an aversively-motivated discrimination task. Behav. Neural Biol. 45:358-365; 1986. Introini-Colhson, I. B.; Nagahara, A. H.; McGaugh, J. L. Memoryenhancement with Inca-amygdala ~s~aining naloxone is blocked by concurrent administration of propranolol. Brain Res. 476:94-101; 1989. Izquierdo, I. Effect of naloxone and morphine on various forms of memory in the rat: Possible role of endogenous opiate mechanisms in memory consolidation. Psychopharmacology (Berlin) 66: 199-203; 1979. Izquierdo, I. Memory depends on me relation between the neurohumoral and hormonal states present after training and at the time of testing. In: Lynch, G.; McGaugh, J. L.; Weinberger, N. M., eds. Neurobiology of learning and memory. New York: Guilford Press; 1984:333-350. Izquierdo, I. Different forms of posttraining memory processing. Behav. Neural Biol.; in press. Kelemen, K.; Bovet, D. Effect of drugs upon the defensive behavior of rats. Acta Physiol. Hung. 19:143-154; 1961. Kety, S. Brain catecholamines, affective states and memory. In: McGaugh, J. L., ed. The chemistry of mood, motivation and memory. New York: Raven Press; 1972:65-80. Lashley, K. S. The effect of strychnine and caffeine upon rate of learning. Psychobiology 1:141-170; 1917. Liang, K. C.; Bennett, C.; McGaugh, J. L. Peripheral epinephrine modulates the effects of posttraining amygdala stimulation on memory. Behav. Brain Res. 15:93-100; 1985. Liang, K. C.; Juler, R. G.; McGaugh, J. L. Modulating effects of posttraining epinephrine on memory: Involvement of the amygdala noradrenergic system. Brain Res. 368:125-133; 1986. Liang, K. C.; McGaugh, J. L. Lesions of the stria terminalis attenuate the enhancing effect of post-training epinephrine on retention of an inhibitory avoidance response. Behav. Brain Res. 9:49-58; 1983. McGaugh, J. L. Drug facilitation of memory and learning. In: Efron, D. H., ed. Psychoph~acology: A review of progress. Washington, DC: U.S. Government Printing Office, PHS Publ. No. 1836; 1968:

345

891-904. 36. McGaugh, J. L. Drug facilitation of learning and memory. Annu. Rev. Pharmacol. 13:229-241; 1973. 37. McGaugh, J. L. Hormonal influences on memory. Annu. Rev. Psychol. 34297-323; 1983. 38. M&laugh, J. L. Involvement of hormonal and ne~m~ulato~ systems in the regulation of memory storage. Annu. Rev. Neurosci. 12:255-287; 1989. 39. McGaugh, J. L.; Hem, M. J. Memory consolidation. San Francisco: Albion Publ.; 1972. 40. McGaugh, J. L.; Introini-Collison, I. B.; Nagahara, A. H. Memoryenhancing effects of posttraining naloxone: Involvement of b-noradrenergic influences in the amygdaloid complex. Brain Res. 446: 37-49; 1988. 41. McGaugh, J. L.; Krivanek, J. A. Strychnine effects on disc~~nation learning in mice: Effects of dose and time of administration. Physiol. Behav. 5:1437-1442; 1970. 42. McGaugh, J. L.; Landfield, P. W. Delayed development of amnesia following electroconvulsive shock. Physiol. Behav. 5: 1109-l 113; 1970. 43. McGaugh, J. L.; Petrinovich, L. The effect of strychnine sulphate on maze-learning. Am. J. Psychol. 72:99-102; 1959. 44. McGaugh, J. L.; Petrinovich, L. F. Comments concerning the basis of learning enhancement with central nervous system stimulants. Psychol. Rep. 12:211-214; 1963. 45. McGaugh, J. L.; Petrinovich, L. F. Effects of drugs on learning and memory. Int. Rev. Neurobiol. 8:139-196; 1965. 46. McGaugh, J. L.; Thomson, C. W. Facilitation of simultaneous disc~mination learning with strychnine sulphate. Psychoph~acologia 3:166-172; 1962. 47. McGaugh, J. L.; Thomson, C. W.; Westbrook, W. H.; Hudspeth, W. J. A further study of learning facilitation with strychnine sulphate. Psychopharmacologia 3:352-360; 1962. 48. McGaugh, J. L.: Westbrook, W. H.; Thomson, C. W. Facilitation of maze learning wih posttrial injections of 5-7 diphenyl-1-3-diazadamantand-ol (1757 I.S.). J. Comp. Physiol. Psychol. 55710-713; 1962. 49. Messing, R. B.; Jensen, R. A.; Martinez, J. L., Jr.; Spiehier, V. R.; Vasquez, B. J.; Soumireu-Mourat, B.; Liang, K. C.; McGaugh, J. L. Naloxone enhancement of memory. Behav. Neural Biol. 27:266-275; 1979. 50. Mueller, G. E.; Pilzecker, A. Exp. Beitr. Lehre Gedach. Z. Psychol. i:l-288; 1900. 51. Pe~novi~h, L. Facilitation of successive dis~~iMtion learning by strychnine sulphate. Psychoph~acoIogia 4: 103-l 13; 1963. 52. Stemberg, D. B.; Isaacs, K.; Gold, P. E.; McGaugh, J. L. Epinephtine facilitation of appetitive learning: Attenuation with adrenergic receptor antagonists. Behav. Neural Biol. 44:447453; 1985. 53. Thiessen, D. D.; Schlesinger, K.; Calhoun, W. H. Better learning: neural enhancement or reduced interference? Psychol. Rep. 9:493496; 1961. 54. Tolman, E. C. Purposive behavior in animals and men. New York: Appleton-Century Crofts; 1932. 55. Tolman, E. C.; Honzik, C. H. Introduction and removal of reward and maze performance in rats. Univ. Calif. PubI. Psychol. 4:257-275; 1930. 56. Westbrook, W. H.; McGaugh, J. L. Drug facilitation of latent learning. Psychopharmacologia 5440-446; 1964. 57. Yonkov, D. I.; Georgiev, V. P. Memory effect of GABAergic an~gonists in rats trained with two-way active avoidance tasks. A& Physiol. Pharmacol. Bulg. 11:44-49; 1985.