Olfactory Repeated Discrimination Reversal in Rats: Effects of Chlordiazepoxide, Dizocilpine and Morphine

Mark Galizio, Laurence Miller, Adam Ferguson, Patrick McKinney & Raymond C. Pitts

Department of Psychology

University of North Carolina at Wilmington

Running Head: Drug Effects on Repeated Reversals

Address Correspondence to:

Mark Galizio

Department of Psychology

University of North Carolina at Wilmington

Wilmington, North Carolina 28403

Fax: 910-962-7010

e-mail: galizio@uncwil.edu

Abstract

A new procedure designed to assess effects of drugs and other manipulations on non-spatial learning in rats is described. Effects of a benzodiazepine, chlordiazepoxide, a non-competitive NMDA receptor antagonist, dizocilpine (MK-801), and an opiate agonist, morphine, were assessed. In each session rats were exposed to two different 2-choice discrimination problems with food reinforcement for correct responses. One problem (performance discrimination) remained the same throughout the study. That is, one odor was always correct (S+) and the other was never correct (S-). For the other problem (reversal discrimination), stimuli changed every session. Six different odors were used to program the reversal discrimination and on any given session, S+ was a stimulus that had served as S- the last time it had appeared and S- was a stimulus that had been S+ on its last appearance. Thus, in each session the acquisition of a discrimination reversal problem could be studied along with the performance of a comparable, but previously learned, discrimination. Chlordiazepoxide interfered with reversal learning at doses that had no effect on the performance discrimination. Morphine and dizocilpine also impaired reversal learning, but only at doses that also affected performance of the well-learned performance discrimination.

Keywords: Benzodiazepine, NMDA antagonist, learning, repeated acquisition, chlordiazepoxide, morphine, dizocilpine
The therapeutic potential of benzodiazepine and opioid agonists and N-methyl-D-aspartate (NMDA) receptor antagonists is limited by a number of adverse side effects including concerns that they disrupt learning and memory, and these have led to a considerable literature (Hardman, Limbird & Gilman, 2001; Mintzer, Copersino & Stitzer, 2005; Mintzer & Griffiths, 2003; Willets, Balster & Leander, 1990). Preclinical studies with nonhuman subjects that model the complex cognitive demands evident in humans are of considerable importance for the development of suitable treatments with such compounds and also have the potential of increasing understanding of the neurobiology of learning and memory processes.

Studies of spatial learning are perhaps the most commonly used procedures that model the rapid, often one-trial, learning that characterizes human cognitive performance. For example, the Morris (1981) Swim Task (MST), one of the most popular procedures in the study of spatial learning, requires rats to swim in a large, circular pool to a platform located in a fixed position, and submerged just beneath the surface of clouded water (hidden-platform task). A procedure to study repeated acquisitions in the MST was developed in our laboratory (Keith & Galizio, 1997) in which rats were trained to swim to a different platform location every session. Under these conditions, rats come to display rapid, within-session learning of the platform position, with acquisition often evident in a single trial (Galizio, Keith, Mansfield & Pitts, 2003; Keith & Galizio, 1997; Keith, Pitts, Pezzuti & Galizio, 2003).

In order to determine the effects of drugs and other neurobiological manipulations on this rapid spatial learning, Keith and Galizio (1997) also developed an adaptation of the multiple-component repeated acquisitions/performance procedure. These procedures allow direct, within-subject comparisons of drug effects on a well-learned sequence of behavior with the acquisition of a new sequence in a single experiment session and have a long history of use in of distinguishing between drug effects on learning from influences on other aspects of performance, e.g., sensory-motor effects, motivation, etc. (Thompson & Moerschbaecher, 1979). The spatial-learning adaptation of the repeated-acquisitions/performance procedure was developed by exposing rats to two pools (or to two different background stimuli). As above, in one pool, rats learned a new platform position in each session (acquisition component), while in another pool the platform was located in the same position on every session so that performance on a previously learned position could be assessed (performance component). Using this procedure, benzodiazepine and opiate agonists consistently impaired spatial acquisition at doses that had no effect on swimming to the well-learned component (Galizio et al., 2003; Keith & Galizio, 1997; Keith et al., 2003; Padlubnaya, Galizio, Pitts & Keith, in press). However, effects of NMDA antagonists have generally not been selective. For example, Keith and Galizio found that a non-competitive NMDA antagonist, dizocilpine (DZP), impaired acquisition in a dose-dependent manner, but the same doses that increased acquisition escape latencies also produced clear evidence of impairment in the performance component. Similar results were obtained with a competitive NMDA antagonist, LY235959 (Galizio, et al., 2003). These results were consistent with conclusions reached by Cain and his colleagues (e.g., Saucier et al., 1996) that NMDA antagonists do not selectively affect spatial learning in the MST when appropriate control conditions are used.

One question raised by these findings is whether they are specific to spatial learning tasks. There have been a number of investigations of the effects of NMDA antagonists on repeated acquisition of response chains (Moerschbaecher, Thompson & Winsauer, 1985; Thompson, Winsauer & Mastrapaolo, 1987) and non-spatial conditional discriminations (France, Moerschbaecher & Woods, 1991) in monkeys, and in each case selective impairments of learning were obtained. That is, NMDA antagonists (DZP, ketamine and phencyclidine) impaired acquisition of a new response chain or conditional discrimination at doses that did not affect performance of a well-learned one. Similar results have been obtained with benzodiazepines, which consistently produce selective impairments of non-spatial learning in repeated acquisition/performance procedures (Auta, Faust, Lambert, Guidotti, Costa & Moerschbaecher, 1995; Auta, Winsauer, Faust, Lambert & Moerschbaecher, 1997; Winsauer, Bixler, & Mele, 1996) in monkeys. Interestingly, opiate agonists have generally produced only non-selective effects with non-spatial repeated acquisition procedures (Moerschbaecher & Thompson, 1983; Moerschbaecher, Thompson & Winsauer, 1985) in marked contrast to the selective effects noted on spatial learning (Galizio, et al, 2003). Most of the studies noted above using non-spatial procedures to study repeated acquisition have used primates as subjects, and thus, the discrepant outcomes among the various studies reviewed above might be attributed to species differences (primates vs. rats) or to type of learning task (spatial vs. non-spatial). The present study was an attempt to clarify some of these possibilities by determining drug effects using a non-spatial, but rapidly acquired, learning task in rats: olfactory discrimination learning.

In contrast to the slow rates of visual discrimination learning shown by rats, rapid within-session olfactory discrimination can be readily demonstrated in rats (Nigrosh, Slotnick & Nevin, 1975; Slotnick, Hanford & Hodos, 2000). Simple methods of studying olfactory discriminations have been developed in which rodents are trained to dig in cups containing a mixture of sand and spices or other odorants to obtain a buried food pellet (Bunsey & Eichenbaum, 1996; Mihalick, Langlois, Krienke & Dube, 2000). In the present study, these procedures were adapted to develop a repeated acquisition/performance methodology. In each session, rats were exposed to two different odor pairs. In each pair, the stimulus cup associated with one odor contained a food pellet (S+) and the other odor did not (S-). For one discrimination, stimulus pairs were constructed such that in each session food was produced by digging in a stimulus that had served as S- in the last session it had been used but not for digging in the other member of the pair, which had served as S+ in the last session it had been used. Thus, a discrimination reversal was required in each session. The other stimulus pair, intermixed with reversal trials, served as a control with one stimulus designated as S+ and the other S- in every session throughout the experiment. Rats were trained under these conditions until asymptotic performances were reached, after which the effects of a benzodiazepine agonist (chlordiazepoxide, CDZ), an opiate agonist, morphine, and a non-competitive NMDA antagonist (dizocilpine, DZP) were assessed.

Method

Subjects

Ten male Holtzman Sprague-Dawley albino rats, 120-180 days old at the start of the study, served as subjects. The subjects were housed individually in hanging steel cages with a reversed 12 h light, 12 h dark cycle. The subjects’ diet was restricted to the sucrose pellets used as reinforcement and to standard rat chow pellets available for one hour each day (usually half an hour after they completed a testing session). The subjects had unlimited access to water in their home cage.

Apparatus

The apparatus was a modified operant chamber with interior dimensions 28 cm long x 26 cm wide x 30 cm high. The floor of the apparatus was constructed of stainless steel bars 0.5 cm in diameter and spaced 1.3 cm apart. The apparatus was modified for the experiment by removing a 5 cm section of the front wall permitting a stimulus presentation tray to be inserted. The plastic tray was 28 cm long x 12 cm wide x 4 cm high and contained two symmetrical circular holes 5 cm in diameter, 8 cm from the wall and 5 cm from each other. Plastic cups (2 oz.) were placed in these holes for stimulus presentation. When the tray was completely inserted, the two comparison cups (eight cm apart) were accessible to the rat (see Figure 1). A small speaker adjacent to the operant box emitted constant white masking noise (70 dB).

Stimuli

Olfactory stimuli were generated by mixing household spices (Great American Spice Co.-- celery, cinnamon, garlic, ginger, mustard, onion, paprika, and sage) or coffee (Folgers) with sterilized play sand. Stimulus cups were filled to approximately 1 cm below the rim with scented sand and the cup designated as correct (S+) on a particular trial was baited by placing a sucrose pellet 1 cm below the surface of sand with tweezers. A pellet was inserted and removed from the stimulus cup that was designated as incorrect on a trial (S-) to insure that neither displacement of the sand nor the scent of the pellet or tweezers would serve as potential cues. A ratio of 1 g of spice per 100 g of sand was used because pilot research suggested that these spice levels were sufficient to mask the scent of the sucrose pellet. In these pilot studies, the rats of the present study as well as additional rats were exposed to multiple trials on which they could choose between two cups filled with sand with a specified concentration of one of the above spices, one baited with a sucrose pellet and the other not. No evidence of above-chance pellet detection was obtained at the 1 g/100 g concentrations with any of spices used in the present study. Within each experimental session, different cups were used on each trial.

Procedure

Pre-training. The first behavioral session consisted of allowing the subject to adapt to the apparatus for 30 min. In the next few sessions, 45 mg sucrose pellets were placed in empty stimulus cups in the tray, which was inserted into the apparatus until the pellets were consumed. These sessions were continued until the rat consumed 50 pellets. Subsequent sessions involved filling the cups with unscented sand with a food pellet on top, and gradually moving the pellet from the surface to a depth of 1 cm. Each session involved 20 trials and this phase was continued until the subject consumed all of the pellets within 20 min.

Repeated Reversal Training. In each session of this phase, rats were exposed to a 20-trial simultaneous odor discrimination with two different spices—one designated as the correct stimulus (S+) and the other as incorrect (S-) for that session. Each trial began with the insertion of the stimulus tray with one S+ and one S- cup. The left cup contained S+ on half the trials (10) and the order was random with the constraint the S+ was not presented on either side more than twice in succession. A response was defined as paws or snout in contact with the sand plus displacement of sand by digging motions with the paws or penetration of sand by the snout. The trial continued until a response was made and the animal consumed the pellet or after 30 s elapsed, whichever came first. The experimenter retracted the tray, recorded the response, removed the stimulus cups and replaced them with the cups programmed for the next trial during the inter-trial interval (approximately 5 s).

Stimuli were chosen to serve as S+ and S- from a pool of eight before each session. In most cases, the stimuli were cinnamon, coffee, garlic, ginger, mustard, onion, paprika, and sage. One rat (M10) would not approach cups containing ginger, so this spice was replaced by celery for him. Across sessions the order in which these stimuli appeared as S+ or S- was randomized with the following constraints: once an odor served as S+ on a session it had to serve as S- before appearing again as S+; once a stimulus was used in a session, it was not used again until all other stimuli had been used at least once; finally, particular stimulus pairs were not used consecutively. Thus, in each session the acquisition of a discrimination problem could be studied because the rat had to learn to respond to a stimulus that had most recently been S- and to withhold responding to a stimulus that most recently served as S+ (reversal learning). Rats were tested under these conditions until they met a criterion of 90% correct or better on the final ten trials of each of five consecutive sessions. One rat (T2) developed a position bias and, after performing at chance levels for 25 sessions, a timeout punishment procedure was introduced such that the stimulus tray was removed for 5 s after errors. This resulted in greatly improved performance and after a few sessions the punishment procedure was discontinued and this animal was treated like the others for the remainder of the experiment.

Repeated Reversal/Performance Phase. In this phase a second simultaneous discrimination was added to the repeated acquisition sequence each session, but this second discrimination was invariant across sessions. That is, one of the stimuli was designated as S+ throughout the experiment, while the other was designated as S-. Thus, this discrimination (performance discrimination) did not have to be learned within each session. Stimuli for the performance discrimination were chosen arbitrarily for each rat. During this phase, the number of trials per session was increased to 24, with 16 Acquisition trials (that is, trials with the reversed discrimination) and 8 performance trials. The performance trials were randomly interspersed between the acquisition trials, with the only rule being that no more than two performance trials could appear consecutively. For analysis, the acquisition trials were divided into four bins of four trials and the performance trials were broken down in four bins of two trials for analysis. When criteria were met of 100% correct throughout the session for performance and 87.5% correct for the last two bins of acquisition trials for eight consecutive sessions, the drug phase began. As the experiment progressed, twenty sessions were arbitrarily selected to test inter-rater reliability with 8 of the 10 rats tested on at least one such session. In these sessions, one of the raters was blind as to which stimulus was correct or incorrect and there was agreement on 99.1% of the trials.

Drug Phase. Sessions were conducted five days/wk and injections were given on Tuesdays and Fridays. Five rats were studied with CDZ, five with DZP, and seven with morphine. Most rats were tested under more than one drug, and when this occurred, two weeks of baseline training without injections intervened between one drug study and the next. Chlordiazepoxide hydrochloride (Sigma), dizocilpine (MK-801) maleate (Tocris), and morphine sulfate (NIDA) were dissolved in 0.9% saline and administered in a volume of 1 ml/kg. Doses were calculated as the salts. CDZ and morphine injections were given 15 min prior to the session. DZP was administered 30 min prior to the session. Doses were determined two to four times, in ascending order for the initial determination, and semi-randomly thereafter. Doses tested depended on the sensitivity of individual rats with the goal to determine at least one dose that was low enough to be without effect and at least one dose that was high enough to affect the performance discrimination.

Results

The pre-training phase of the study required from 7 to 14 sessions. It required considerably more sessions for rats to reach criterion for the Repeated Reversal Phase, with a range of 15-39 sessions (M = 23.9). Introduction of the performance discrimination in the Repeated Reversal/Performance Phase required 12 to 51 additional training sessions before the criterion was reached (M =26.5).

Figures 2-4 shows the effects of the various drugs on the performance (white triangles) and reversal (black circles) discriminations with mean percent correct on the vertical axis and within-session trials blocks on the horizontal axis. The first two panels show percent correct on baseline and saline control sessions for all four studies and reveal that rats were learning the reversed discrimination to high levels of accuracy within the session while maintaining virtually perfect accuracy on the well-learned performance discrimination.

Figure 2 shows the effects of CDZ and reveals that low doses of CDZ (1.0 and 3.0 mg/kg) did not affect accuracy in either condition, but the 10.0 mg/kg dose appeared to produce selective effects: that is, it impaired learning while having no effect on the performance discrimination. Reversal learning was also clearly impaired at 17.0 mg/kg, but at this dose a decline in performance was also noted. A Dose X Component X Block repeated measures ANOVA confirmed the dose-dependent impairment of accuracy with a significant main effect of CDZ Dose, F (5, 20) = 12.10, p < .01, and the reliability of the selective effects on reversal learning observed at the 10 mg/kg dose by a significant Dose X Component interaction, F (5, 20) = 5.64, p < .01.

Figure 3 shows that morphine also disrupted reversal learning in a dose-dependent fashion, but the doses that produced declines in reversal accuracies also tended to disrupt performance accuracies as well. Reversal learning was impaired at the 5.6, 10.0 mg/kg doses, but these were accompanied by small disruptions in performance. Both reversal and performance discriminations were substantially impaired at the 17.0 mg/kg dose. Consistent with such an interpretation, the main effect of Morphine Dose was statistically significant, F (5, 30) = 41.88, p < .01, but the Dose X Component interaction term failed to reach significance (p > .05).

Figure 4 shows mean within-session accuracy for the DZP doses tested in all five rats. While 0.03 mg/kg DZP appeared to be without effect on accuracy, the 0.1 mg/kg dose produced small impairments on both the performance and reversed discriminations. At 0.3 mg/kg DZP performance declined to well below chance levels due to a lack of responding by most rats. The non-selective impairment produced by DZP was confirmed statistically with a Dose X Component X Bin repeated measures ANOVA with a significant main effect of Dose, F (4, 16) = 11.86, p < .01 and the absence of a significant Dose X Component interaction (p > .05).

Discussion

The present study demonstrated the feasibility and utility of extending the repeated acquisition/performance procedure to the analysis of olfactory discrimination learning in rats. In the Repeated Reversal Phase of the experiment, all rats eventually showed rapid, within-session acquisition of a discrimination in which the S+ and S- were stimuli whose functions were reversed on each appearance. Average accuracy levels of 85% or higher were generally observed by the second block of four trials, making the procedure a potentially useful model for the rapid learning that characterizes human performances (see also Eichenbaum, 1998; Slotnick, Hanford & Hodos, 2000). In addition to learning a discrimination reversal in each session of the present study, rats also were tested on a different, well-learned discrimination problem (performance). The high levels of accuracy maintained on the performance discrimination throughout the session made it possible to evaluate drug effects on reversal learning while simultaneously evaluating on performance of a previously learned problem.

Intermediate doses of CDZ (10 mg/kg) interfered with reversal learning without affecting performance. These results replicated the effects of benzodiazepines on rapid spatial learning by rats in the Morris Swim Task (Keith & Galizio, 1997; Keith, Pitts, Pezzuti & Galizio, 2003; Padlubnaya, Galizio, Pitts & Keith, in press) and extend these findings to a non-spatial learning problem. The selective disruption of learning by CDZ in the present study was also consistent with findings of selective learning impairments produced by benzodiazepines using repeated acquisition procedures in other species, including pigeons (Thompson, 1975), monkeys (Auta, Faust, Lambert, Guidotti, Costa & Moerschbaecher, 1995; Auta, Winsauer, Faust, Lambert & Moerschbaecher, 1997; Winsauer, Bixler & Mele, 1996) and humans Bickel, Higgins, & Griffith,1989; Desjardins, Moerschbaecher, & Thompson,1982).

Morphine also impaired reversal learning in the present study, but only at relatively high doses that also interfered with performance. These results were consistent with most other studies of opiate effects using non-spatial repeated acquisition (Moerschbaecher & Thompson, 1983; Moerschbaecher, Thompson & Winsauer, 1985) which have generally reported only non-selective effects of opiate drugs. Interestingly, the present findings with respect to morphine differed from the selective effects on spatial learning in the MST noted by Galizio et al (2003). In that study, doses of morphine as low as 3.0 mg/kg interfered with place learning but did not disrupt rat’s ability to navigate to a well-learned hidden platform location. These outcomes suggest the possibility that the opioid system may play a more critical role in spatial than in non-spatial learning, but other explanations are possible. For example, the MST involves different reinforcers than the repeated reversal procedure (escape from the pool vs. food) and different sensory modalities are critical (visual vs. olfactory). Further explorations of opiate effects on selected learning models will be necessary to further determine the generality of spatial vs. non-spatial differences with respect to opiate drugs.

Like morphine, NMDA receptor antagonist DZP also produced dose-dependent impairment of accuracy on both discriminations that were non-selective. Disruption of reversal learning did not reliably occur until doses that were high enough to also impair performance were reached. These results were also consistent with previous findings of non-selective effects of NMDA receptor antagonists on spatial learning by rats (Galizio, et al., 2003; Keith & Galizio, 1997). However, studies of NMDA receptor antagonists on repeated acquisition of non-spatial problems in monkeys and pigeons have generally yielded selective effects (France, Moerschbaecher & Woods, 1991; Moerschbaecher & Thompson, 1980; Moerschbaecher, Thompson & Winsauer, 1985; Thompson, Winsauer & Mastropaolo, 1987). There are few studies of NMDA receptor antagonists on non-spatial repeated acquisition in rats. Clissold, Ferkany and Pontecorvo (1991) found that NMDA antagonists interfered with acquisition of a repeated discrimination task involving visual and auditory stimuli, but because a performance discrimination task was not included, the selectivity of these effects cannot be determined. Cohn and Cory-Slechta (1993) used a repeated acquisition of response chains procedure and found only very slight effects of DZP on acquisition until doses that also affected performance were reached and using similar procedures, Gerak, Stevenson, Winsauer and Moerschbaecher (2004) found only non-selective effects of ketamine. Because NMDA antagonists block hippocampal LTP, and thus would be expected to disrupt learning, further analysis of these compounds is clearly warranted.

In conclusion, the repeated reversal procedure with olfactory stimuli has much to commend it as a baseline to study drug effects on learning in rats. Consistently rapid acquisition rates with high levels of accuracy on the performance discrimination were observed after an average of about 60 sessions of training. In contrast, repeated acquisition of response chains in rats is often characterized by high levels of between- and within-subject variability and very slow acquisition (Pontecorvo & Clissold, 1993). For example, Winsauer, Rodriguez, Cha, & Moerschbaecher (1999) reported that a mean of 110 training sessions was required in one of the few available studies of repeated acquisition of response chains in rats. In addition to these practical concerns, the rapid learning and high level of stimuli control obtained in rats with olfactory stimuli makes it possible to evaluate drug effects on processes that may provide a useful model of human learning.

References

Auta, J., Faust, W. B., Lambert, P., Guidotti, A., Costa, E. and Moerschbaecher, J. M. (1995). Comparison of the effects of full and partial allosteric modulators of GABAa receptors on complex behavior processes in monkeys. Behavioural Pharmacology, 6 323-332.

Auta, J., Winsauer, P. J., Faust, W. B., Lambert, P. & Moerschbaecher, J. M. (1997). Effects of negative allosteric modulators of gamma-aminobutyric acid A receptors on complex behavioral processes in monkeys. The Journal of Pharmacology and Experimental Therapeutics. 280, 316-325.

Bickel, W., Higgins S., Griffiths R. (1989). Repeated diazepam administration: effects on the acquisition and performance of response chains in humans. Journal of the Experimental Analysis of Behavior, 52, 47-56.
Bunsey, M. & Eichenbaum, H. (1996). Conservation of hippocampal memory function in rats and humans. Nature, 379, 255-257.
Clissold, D. B., Ferkany, J. W. & Pontecorvo, M. J. (1991). Competitive and noncompetitive N-methyl-D-aspartate (NMDA) antagonists, haloperidol, and scopolamine impair performance in a nonspatial operant discrimination task. Psychobiology, 19, 332-338. 

Cohn, J. and Cory-Slechta, D.A. (1992). Differential effects of MK-801, NMDA and scopolamine on rats learning a four-member repeated acquisition paradigm. Behavioral Pharmacology, 3, 403-413.

Desjardins, P. J., Moerschbaecher, J.M. and Thompson, D.M. (1982). Intravenous diazepam in humans: Effects of acquisition and performance of response chains. Pharmacology, Biochemistry, and Behavior, 17, 1055-1059.

Eichenbaum, H. (1998). Using olfaction to study memory. In C. Murphy (Ed.) Olfaction and taste XII: An international symposium, Annals of the New York Academy of Sceinces, pp. 657-669, NY: New York Academy of Sciences.

France, C. P., Moerschbaecher, J. M. & Woods, J. H. (1991). MK-801 and related compounds in monkeys: Discriminative stimulus effects and effects on a conditional discrimination. . The Journal of Pharmacology and Experimental Therapeutics. 257, 727-734.

Galizio, M., Keith, J. R., Mansfield, W. and Pitts, R. C. (2003). Repeated spatial acquisition: Effects of NMDA antagonists and morphine. Experimental and Clinical Psychopharmacology, 11, 79-90.

Gerak, L. R., Stevenson, M. W., Winsauer, P. J. & Moerschbaecher, J. M. (2004). Effects of pregnanolonealone and in combination with other positive GABA-A modulators on complex behavior in rats. Psychopharmacology, 173, 195-202.

Hardman, J. G., Limbird, L. E. & Gilman, A. G. (eds.) (2001). Goodman & Gilman’s the pharmacological basis of therapeutics, 10^th edition. McGraw-Hill: New York.

Harlow, H. F. (1949). The formation of learning sets. Psychological Review, 56, 51-65.

Keith, J.R. and Galizio, M. (1997). Acquisition in the Morris swim task is impaired by a benzodiazepine but not an NMDA antagonist: a new procedure for distinguishing acquisition and performance effects. Psychobiology, 25, 217-228.

Keith, J. R., Pitts, R. C., Pezzuti, T. & Galizio, M. (2003). GABA-A modulator effects on a multiple-component, repeated-acquisition test of spatial learning. Behavioral Pharmacology, 14, 67-76.

Mihalick, S. M., Langlois, J. C., Krienke, J. D. & Dube, W. V. (2000). An olfactory discrimination procedure for mice. Journal of the Experimental Analysis of Behavior, 73, 305-318.

Mintzer, M. Z., Copersino, M. L. & Stitzer, M. L. (2005). Opioid abuse and cognitive performance. Drug and Alcohol Dependence, 78, 225-230.

Mintzer, M. Z. & Griffiths, R. R. (2003). Triazolam-amphetamine interaction: Dissociation of effects on memory versus arousal. Journal of Psychopharmacology, 17, 17-29.

Moerschbaecher, J.M. & Thompson, D.M. (1980). Effects of phencyclidine, pentobarbital, and d-amphetamine on the acquisition and performance of conditional discriminations in monkeys. Pharmacology, Biochemistry, and Behavior, 13, 887-894.

Moerschbaecher, J.M. & Thompson, D.M. (1983). Differential effects of prototype opoid agonists on the acquisition and performance of conditional discriminations in monkeys. Journal of Pharmacology and Experimental Therapeutics, 226, 738-748.

Moerschbaecher, J. M., Thompson, D. M. & Winsauer, P. J. (1985). Effects of opioids and phencyclidine in combination with naltrexone on the acquisition and performance of response sequences in monkeys. Pharmacology, Biochemistry and Behavior, 22, 1061-1069.

Morris, R.G.M. (1981). Spatial localization does not require the presence of local cues. Learning and Motivation, 12, 239.260.

Nigrosh, B., Slotnick, B. M. & Nevin, J. (1975). Reversal learning and olfactory stimulus control in rats. Journal of Comparative and Physiological Psychology, 80, 285-294.

Pontercorvo, M. J. & Clissold, D. B. (1993). Complex and delayed discriminations: automated repeated measures techniques. In F. van Haaren (Ed.) Methods in behavioral pharmacology, pp. 147-194, Amsterdam: Elsevier.

Saucier, D., Hargreaves, E.L., Boon, F., Vanderwolf, C.H., & Cain, D.P. (1996). Detailed behavioral analysis of water maze acquisition under systemic NMDA or muscarinic antagonism: Nonspatial pretraining eliminates spatial learning deficits. Behavioral Neuroscience, 110, 103-116.

Slotnick, B.M., Hanford, L., & Hodos, W. (2000). Can rats acquire an olfactory learning set? Journal of Experimental Psychology: Animal Behavior Processes, 26, 399-415.

Thompson, D.M. (1975). Repeated acquisition of response sequences: stimulus control and drugs. Journal of the Experimental Analysis of Behavior, 23, 429-436.

Thompson, D.M. and Moerschbaecher, J.M. (1979). An experimental analysis of the effects of d-amphetamine and cocaine on the acquisition and performance of response chains in monkeys. Journal of the Experimental Analysis of Behavior, 32, 433-44.

Thompson, D.M. and Moerschbaecher, J.M. (1982). Phencyclidine in combination with pentobarbital: Supra-additive effects on complex operant behavior in pigeons. Pharmacology Biochemistry and Behavior, 17, 353-357.

Thompson, D. M., Winsauer, P. J. & Mastrapaolo, J. (1987). Effects of phencyclidine, ketamine and MDMA on complex operant behavior in monkeys. Pharmacology, Biochemistry & Behavior, 26, 401-405.

Willets, J., Balster, R. L. & Leander, J. D. (1990). The behavioral pharmacology of NMDA receptor antagonists. Trends in Pharmacological Science, 11, 423-428.

Winsauer, P.J., Bixler, M.A., and Mele, P.C. (1996). Comparison of the effects of typical and atypical anxiolytics on learning in monkeys and rats. The Journal of Pharmacology and Experimental Therapeutics, 276, 1111-1127.

Winsauer, P. J., Rodriguez, F. H., Cha, A. E. and Moerschbaecher, J.M. (1999). Full and partial 5-HT-1A receptor agonists disrupt learning and performance in rats. Journal of Pharmacology and Experimental Therapeutics, 288, 335-347.

Author Notes

The authors thank Brendan Curl for constructing the apparatus and John Wagner for assisting with data collection. The gifts of morphine from National Institute on Drug Abuse is gratefully acknowledged. The procedures were approved by the UNCW Animal Care and Use Committee. This research was supported by a grant from the National Institute on Drug Abuse (DA12879).

Figure Captions

Figure 1. Photograph of the apparatus

Figure 2. Mean percent correct is plotted as a function of CDZ dose on repeated reversal learning (black circles) and performance (white triangles) discriminations for the doses tested in all five rats. Each session was divided into four bins of 4 trials (repeated reversal) or 2 trials (performance) and points represent the mean of all subjects for that bin. The panel labeled BL presents means obtained on the eight criterion sessions that preceded the drug administration phase. The panel labeled Sal shows means obtained when saline was injected. Vertical bars indicate standard error of the mean. When error bars are not visible, they are smaller than the symbol used to represent the data point.

Figure 3. Mean percent correct is plotted as a function of morphine dose. Other features are as described in Figure 2.

Figure 4. Mean percent correct is plotted as a function of DZP dose. Other features are as described in Figure 2.

Figure 1

Figure 2

Figure 3