Behavioral Neuroscience1983. Vol 97, No 4, 658-662

Copyright 1983 by theAmerican Psychological Association, Inc

Overshadowing Effects in the Stimulus Controlof Morphine Analgesic ToleranceTimm A. Walter and David C. Riccio

Kent State University

This study attempted to replicate previous demonstrations of classical condi-tioning of morphine analgesic tolerance, with the additional aim of determin-ing whether stimulus overshadowing effects might explain previous conflictingfindings. Eight groups of rats received a series of 10 morphine (5 mg/kg) and/or saline injections, differing only with respect to the contingency between acompound visual-auditory conditioned stimulus (CS) and the substance in-jected. When tested for analgesic responding to morphine in the presence ofthe compound CS, only those groups for which the CS and morphine injectionswere paired during the acquisition sequence evidenced tolerance. In a secondexperiment, tolerant animals were tested in the presence of one component ofthe compound CS. When a loud (85 dB) tone was used in the compound, lessanalgesic tolerance was elicited later by the weaker visual stimulus alone. Thisdifferential stimulus control of the analgesic response suggests that over-shadowing may contribute to failures to replicate conditioned morphine tol-erance. That internal morphine-produced stimuli might overshadow externalcues is considered.

There is a growing body of literature which. suggests that the acquisition of morphine anal-gesic tolerance, once considered to be a purely"pharmacological" phenomenon, may be underthe control of classical conditioning processes.More specifically, Siegel (1975, 1977) proposedthat tolerance reflects the conditioning of a com-pensatory hyperalgesic response to environmen-tal cues associated with the administration ofthe drug; presumably, the conditioned responsethen algebraically summates with the stable un-conditioned analgesic effect of morphine.

Consistent with this analysis, it has beendemonstrated that presentation of previouslyconditioned environmental stimuli unaccom-panied by morphine injection elicits a hyperal-gesic response to hot-plate testing (Siegel, 1975);that tolerance developed in the presence of oneset of environmental cues is not observed in thepresence of other environmental stimuli (Advo-kat, 1980; Siegel, 1976; Siegel, Hinson, & Krank,.1978); and that tolerance is subject to experi-mental extinction (Siegel, 1975, 1977; Siegel,Sherman, & Mitchell, 1980; Cappell & Poulos,Note 1), latent inhibition, and partial reinforce-ment effects (Siegel, 1975,1977).

This research was submitted by the first author inpartial fulfillment of the requirements for a master'sdegree at Kent State University. It was supported byNational Institute of Mental Health Grants MH30223and MH37535 awarded to the second author.

Requests for reprints should be sent to Timm A.Walter, Department of Psychology, Kent State Uni-versity, Kent, Ohio 44242.

Although this model seems to provide a rela-tively consistent and parsimonious account ofthese data, other investigators have failed toreplicate these findings and have advanced al-ternative explanations of the data (Bardo &Hughes, 1979; Sherman, 1979; Hughes & Bardo,Note 2). More definitive interpretation of theseinconsistencies is complicated, however, by anumber of procedural differences among thesestudies and, more generally, by a lack of moresystematic specification of the putative control-ling environmental stimuli.

Hutton, Woods, and Makous (1970) and Pou-los and Cappell (1979) documented the acquisi-tion of control of conditioned drug responses bystimuli other than those designated as the nom-inal (environmental) conditioned stimuli (CS),and in both cases these "unauthorized" stimuliobscured any evidence of conditioning beforepotentially relevant procedural and environ-mental stimuli were more systematically con-trolled. The "anomalous" findings in these stud-ies essentially constituted a demonstration ofovershadowing (Kamin, 1968), and it is possiblethat similar findings may be operative in themodulation of morphine analgesic tolerance.

The present investigation was an attempt toreplicate previous demonstrations of the stimu-lus control of morphine analgesic tolerance byusing procedures that would permit both a moreprecise specification of the controlling stimuliand a more systematic examination of over-shadowing effects as a possible explanation forthe conflicting findings reported thus far. Amodified discrimination design (cf. Siegel, 1979),incorporating a discrete visual-auditory com-

658

BRIEF COMMUNICATIONS 659

pound CS consisting of a fixed-intensity lightand one of two levels of white noise, was em-ployed during the tolerance acquisition phase inorder to dissociate other potentially relevantenvironmental and procedural stimuli from thesystemic effects of morphine. In the first exper-iment, all analgesic assessments were conductedin the presence of the compound CS. Thosegroups that displayed significant analgesic tol-erance were used as subjects in a second exper-iment in which analgesic assessments were con-ducted in the presence of one or the other com-ponent of the compound CS. Overshadowingwas indicated by significant differences betweenthe levels of analgesic responding observed inthe presence of each component of the com-pound CS.

MethodEighty experimentally naive Sprague-Daw-

ley-derived male rats weighing 415-575 g wereindividually housed in wire mesh cages on a14:10 hr light/dark cycle with ad lib access tofood and water for 2 wk prior to and throughoutthe duration of the experiment.

Pain sensitivity was assessed by a standardhot-plate procedure in which analgesia is in-dexed by an animal's latency to respond with apaw-lick or a jumping response when placed ona heated surface. Due to the relatively small sizeof the containment chambers used during train-ing/testing, which prevented animals from emit-ting a jumping response, all analgesic assess-ments in this investigation were indexed by paw-lick latency. The surface of the hot plate wasmaintained at 49 .5 C, as a pilot study hadestablished more reliable assessment of analge-sia at this temperature.

All injections were administered sc on thedorsal surface of the neck in a volume of 1 ml/kg of body weight. Morphine sulfate was dis-solved in a .9% saline vehicle at a concentrationof 5 mg/ml. All other injections were of anequivalent volume of the saline vehicle.

During analgesic assessments, animals wereconfined on the hot-plate surface in a squarePlexiglas chamber identical in construction toeight other chambers used for confinement andstimulus presentations during the tolerance ac-quisition phase of the experiment. White noiselevels were calibrated at 50 or 85 dB (SPL) andvisual stimuli consisted of two 40-W fluorescentlights suspended 11 cm over the top of two rowsof four conditioning chambers.

The experimental design and procedures de-scribed below are summarized in Table 1. Duringthe tolerance acquisition phase of the first ex-periment, eight groups of rats (n = 10) wereadministered daily injections of either saline or

morphine, paired with one of two different stim-ulus conditions. On each day during this phase,animals were transported from the colony roomto an adjacent laboratory where they were in-jected and immediately placed in the condition-ing chamber in the presence of the appropriateauditory and visual stimuli. Subjects were re-moved after 30 min and transported immedi-ately to the colony room. Beginning on the firstday of this phase, four groups received morphineinjections paired with a 30-min presentation ofthe compound CS every other day for a total offive morphine injections. Groups P50L andP50N were presented with a 50-dB white noisein compound with the light cue; Groups P85Land P85N were treated identically but withwhite noise at 85 dB. On intervening days thesegroups were given the same treatment exceptthat the substance injected was physiologicalsaline and the compound CS was not presented.Illumination of the training/testing room wasprovided by a 40-W red-filtered bulb suspendedfrom the ceiling in an opposite corner of theroom.

As a control for nonassociative treatment ef-fects in an "unpaired" condition, Groups U50and U85 received an equivalent regimen of mor-phine injections except that the injections wereadministered following CS exposure and sub-jects were then returned to the colony room.Groups S-S and S-M were subjected to proce-dures identical to these for the paired groups(P50L, P50N, P85L, P85N) except that all in-jections were of physiological saline. Half of theanimals in each of Groups S-S and S-M wereconditioned with one or the other compound CS.

Analgesic assessments were conducted 24 hrafter the final day of the acquisition phase, 48hr after the last injection of morphine. Eachgroup of animals was transported to the train-ing/testing room and injected with a 5 mg/kgtest dose of morphine before being placed in theconditioning chambers. Animals were exposedto the same compound CS as during the acqui-sition phase. Thirty minutes later, each animalwas removed from its chamber and placed indi-vidually in an identical chamber that was usedfor analgesic testing. The compound CS wascontinued for the duration of each animal's paw-lick latency assessment. Animals in Groups S-Swere subjected to the same test procedures ex-cept that the substance injected was physiolog-ical saline.

The second experiment was conducted 48 hrafter the analgesic assessments in Experiment 1and included only those groups for which mor-phine injections had been paired with compoundCS presentations. Groups P50L and P85L weretested in the presence of the light alone; GroupsP50N and P85N were tested in the presence of

660 BRIEF COMMUNICATIONS

0) C2 S 3H

BRIEF COMMUNICATIONS 661

HI

15

10Otf)

I -2 5

EXPERIMENT 1

II

EXPERIMENT 2

iiiU50 P50N P50L U85 P85N P8SL S-S S-M P50N P50L P8SN P85L

Figure 1. Mean paw-lick latencies (1 SD) in Experiments 1 and 2. (U groups [cross-hatched bars] denote unpaired morphine groups; P groups [solid bars] denote the pairedmorphine condition; S groups [unfilled bars] denote saline control groups. Group S-M wasgiven a single test dose of morphine on the final day of Experiment 1; S-S received onlysaline.)

those in Groups S-M, which had received mor-phine for the first time, both in terms of therelative magnitude and variability of the anal-gesic response. The paired groups, however, allevidenced a reliably smaller and less variableresponse to the test dose of morphine, a resultpresumably reflecting the conditioning of a com-pensatory hyperalgesic response to the com-pound CS.

Planned orthogonal comparisons on the la-tency data in Experiment 2 indicated significantdifferences in mean analgesic response (p < .05)and variance (p < .001) between Group P85Land Group P85N but not between Group P50Nand Group P50L. For animals conditioned incompound with the 85-dB noise, it appears thattolerance was controlled to a large extent by thenoise component of the compound CS, whereastolerance seems to have been controlled to anequal extent by both stimuli in the 50-dB groups.Inspection of the performance of Group P58Lin each experiment suggests that some tolerancewas displayed during Experiment 2 which atten-uated the magnitude of the overshadowing effectbetween Group P85N and Group P85L. Thattolerance was still controlled to some extent bythe light component in this case is similar tofindings reported by Mackintosh (1976) usingthe same compound CS in a conditionedsuppression paradigm.

These findings extend the range of classicalconditioning procedures that have been dem-onstrated to modulate the development and/ordisplay of morphine analgesic tolerance. More-over, they suggest a possible mechanism to ac-count for some of the conflicting findings in theliterature. It is possible, as Sherman (1979) ac-knowledged, that unauthorized environmentalor procedural stimuli common to the trainingand testing environments overshadowed controlby the nominal CS in some of these studies and

thus obscured any evidence of conditioning. Yetanother more intriguing possibility has to dowith the specific nature of the controlling stim-uli. Inasmuch as morphine-produced internalstimuli have been shown to be capable of sup-porting bar-press (Belleville, 1964; Hirschhorn& Rosecrans, 1971) and T-maze discriminationlearning (Hill et al., 1971; Overton, 1971; Win-ter, 1975), we suggest that these stimuli mightalso be capable of acquiring control over com-pensatory responding and, hence, tolerance, par-ticularly under conditions in which the relativesalience or predictive value of external stimulihas been degraded in some manner. Tolerancecontrolled by internal, morphine-produced stim-uli, unlike that mediated by environmental stim-uli, would be expected to be relatively transsi-tuational or "pharmacological" in nature, eventhough the same underlying conditioning mech-anisms would be involved. The question thenbecomes one of establishing the extent to whichtolerance, in any given case, is controlled by oneor the other of these two general classes ofstimuli, rather than one of making a distinctionbetween two different "kinds" of tolerance.

Reference Notes1. Cappell, H., & Poulos, C. X. Associative factors in

tolerance to morphine Dose response determi-nation. Paper presented at the meeting of theAmerican Psychological Association, Toronto,August-September 1978.

2 Hughes, R. A., & Bardo, M. T Morphine analgesictolerance in rats A search for hyperalgesia. Pa-per presented at the meeting of the Psycho-nomic Society, November 1978.

ReferencesAdvokat, C. Evidence for conditioned tolerance to the

tail-flick reflex Behavioral and Neural Biology,1980,29, 385-389.

662 BRIEF COMMUNICATIONS

Bardo, M. T., & Hughes, R. A. Exposure to a non-functional hotplate as a factor in the assessment ofmorphine analgesia in rats. Pharmacology Biochem-istry and Behavior, 1979, 10, 481-485.

Belleville, R. E. Control of behavior by drug-producedinternal stimuli. Psychopharmacologia, 1964, 5, 95-105.

Hill, H. E., Jones, B. E., & Bell, E C. State dependentcontrol of discrimination by morphine and pento-barbital. Psychopharmacologia, 1971, 22, 305-313.

Hirachhom, I. D., & Rosecrans, J. A. A comparison ofthe stimulus effects of morphine and lysergic aciddiethylamide. Pharmacology Biochemistry and Be-havior, 1971,2, 361-366.

Hutton, R. A., Woods, S. C, & Makous, W. Condi-tioned hypoglycemia: Pseudoconditioning controls.Journal of Comparative and Physiological Psychol-ogy, 1970, 71, 198-201

Kamin, L. J. Attention-like processes in classical con-ditioning. In M. R. Jones (Ed.), Miami Symposiumon the Prediction of Behavior. Coral Gables, Fla.:University of Miami Press, 1968.

Mackintosh, M. J. Overshadowing and stimulus inten-sity. Animal Learning & Behavior, 1976,4, 186-192.

Overton, D. A. Discriminative control of behavior bydrug states. In T. I. Thompson & R. Pickens (Eds.),Stimulus properties of drugs New York: Appleton-Century-Croft, 1971.

Poulos, C. X., & Cappell, H. An associative analysisof pretreatment effects in gustatory conditioning by

amphetamine. Psychopharmacology, 1979, 64, 201-207.

Sherman, J. E. The effects of conditioning and noveltyon the rat's analgesic and pyretic responses to mor-phine. Learning and Motivation, 1979,10, 383-418.

Siegel, S. Evidence from rats that morphine toleranceis a learned response. Journal of Comparative andPhysiological Psychology, 1975, 89, 498-506.

Siegel, S. Morphine analgesic tolerance: Its situationalspecificity supports a Pavlovian conditioning model.Science, 1976,193, 323-325.

Siegel, S. Morphine tolerance acquisition as an asso-ciative process. Journal of Experimental PsychologyAnimal Behavior Processes, 1977, 3, 1-13.

Siegel, S. The role of conditioning in drug toleranceand addiction. In J. D. Keehn (Ed.), Psychopathol-ogy in animals' Research and clinical applicationsNew York: Academic Press, 1979.

Siegel, S., Hinson, R. E., & Krank, M. D. The role ofpredrug signals in morphine analgesic toleranceSupport for a Pavlovian conditioning model of tol-erance. Journal of Experimental Psychology. AnimalBehavior Processes, 1978, 4, 188-196.

Siegel, S., Sherman, J. E., & Mitchell, D. Extinctionof morphine analgesic tolerance. Learning and Mo-tivation, 1980, 11, 289-301.

Winter, J. C. The stimulus properties of morphineand ethanol. Psychopharmacologia, 1975, 44, 209-214.

Received January 24,1983