Psychopharmacology (1985) 87:101-106 Psychopharmacology 9 Springer-Verlag 1985

Absence of environment-specificity in morphine tolerance acquired in non-distinctive environments: habituation or stimulus overshadowing ? Richard Dafters and Laura Bach

Psychology Department, Adam Smith Building, Glasgow University, Glasgow Gt2, Scotland

Abstract. The experiments reported here investigated the mechanisms of drug tolerance acquisition in environments differing in distinctiveness. Specifically, they examined the hypothesis that tolerance acquired in non-distinctive envi- ronments might involve habituation while tolerance ac- quired in distinctive environments involves a classical con- ditioning or associative learning mechanism. In Experi- ment 1, rats pre-exposed to injection-ritual cues (placebo injections) prior to acquisition of tolerance to morphine analgesia in distinctive or non-distinctive environments showed a typically attenuated tolerance response on an en- vironment-change test. The magnitude of the attentuation was not affected by the distinctiveness of the acquisition environment. In Experiment 2, rats acquiring tolerance in distinctive or non-distinctive environments, but without prior injection-ritual pre-exposnre, did not demonstrate an attenuation of tolerance on an environment-change test. Tolerance acquired in either environment was unaffected by a subsequent rest period in the colony room, but was attenuated by a subsequent period of daily placebo injec- tions in the colony room. It is argued that failure to observe environment-specific tolerance, as in Experiment 2 and in previous reports in the literature, may reflect overshadow- ing of environmental stimuli by injection-ritual stimuli, and are not indicative of a fundamental difference between the mechanisms of tolerance acquisition in environments vary- ing in distinctiveness.

Key words: Morphine analgesic tolerance - Environment distinctiveness - Habituation - Compensatory conditioning - Overshadowing - Rat

A substantial body of data supports the view that morphine tolerance reflects the operation of a classical conditioning mechanism (see Siegel 1983 for review). According to this view, environmental stimuli associated with the administra- tion of the drug elicit compensatory conditioned responses which attenuate the unconditioned effects of the drug and thereby produce tolerance.

Recently, however, it has been claimed that the applica- bility of the conditioning account is limited to those situa- tions where tolerance develops in the presence of distinctive environmental stimuli - the putative conditioned stimulus (Kesner and Baker 1981 ; Kesner and Cook 1983). In non-

Offprint requests to .' R. Dafters

distinctive environments tolerance is said to occur through a non-associative habituation process - that is, the response to a repeated drug administration will decrease in the same way, and for the same reasons, that the response to a re- peated exteroceptive stimulus decreases (e.g. Groves and Thompson 1970). The mechanism of this habituation pro- cess is unknown, but there are suggestions that it may be subserved by alterations in specific neurotransmitter sys- tems (Kesner and Baker 1981). The important difference between the suggested conditioning and habituation mecha- nisms of tolerance is that only the former involves associa- tive learning and the elicitation of compensatory condi- tioned responses.

This dual-process view of tolerance makes very clear predictions about the effects of changing the environmental context on the manifestation of tolerance. To the extent that tolerance is the result of responses conditioned to envi- ronmental cues, as in distinctive environments, changing those cues during the administration of the drug should abolish, or at least attenuate, tolerance. Such an environ- ment-specific tolerance effect has been reported in rats tol- erant to morphine (Siegel et al. /978; Tiffany and Baker 1981), alcohol (L6 et al. 1979; Crowell et al. 1981), and amphetamine (Demellweek and Goudie 1983), and in hu- mans tolerant to alcohol (Dafters and Anderson 1982). In non-distinctive environments, on the other hand, where pre- sumably only the drug-habituation process is permitted, a change in the environmental context would not be expected to significantly affect the manifestation of tolerance. This prediction has recently been confirmed (Kesner and Cook 1983).

However, it may be possible to account for the differen- tial effects of tolerance acquired in distinctive and non- distinctive environments without appealing to an additional habituation process. The development of a compensatory conditioned response does not require the presence of a nominal conditioned stimulus (in the form of distinctive environmental cues), only that stimuli are present which reliably predict the occurrence of the drug. In experiments in which the drug is injected such stimuli are invariably present in the form of injection-ritual cues (handling, inser- tion of the hypodermic needle, etc.). It is possible that such cues may be sufficiently salient to overshadow the environ- mental cues in non-distinctive environments but not in dis- tinctive ones (see Kamin 1969 for a discussion of over- shadowing in traditional conditioning paradigms). Such a selective overshadowing effect could thus account for the

102

differential effects of environment-change observed by Kesner and Cook (1983), and indeed, an explicit demon- stration of overshadowing in the morphine analgesia para- digm has recently been reported (Walter and Riccio 1983).

The first experiment to be reported here sought to evalu- ate the overshadowing interpretation by attempting to di- minish the ability of injection-ritual cues to act as condi- tioned stimuli and hence to overshadow environmental stimuli during tolerance development. Rats were pre-ex- posed to injection-ritual cues (placebo injections) prior to tolerance acquisition in distinctive or non-distinctive envi- ronments and a subsequent environment-change test. Such pre-exposure to the conditioned stimulus normally retards subsequent development of the conditioned response - a phenomenon known as latent inhibition (Lubow 1973). To the extent that this procedure is effective in preventing over- shadowing, and in contrast to the results of Kesner and Cook (1983), attenuation of tolerance should occur even in that group which acquired tolerance in a non-distinctive environment. The experiment also examined the effect on tolerance of a 14-day "rest" period in the colony room during which all experimental treatments were discontin- ued. If tolerance reflects habituation of the drug effect it might be expected, on the basis of previous habituation research (Groves and Thompson 1970), that some sponta- neous recovery of the analgesic response would occur on the subsequent drug administration test. If tolerance reflects the development of conditioned responses, no recovery of the drug effect is to be expected in the absence of explicit extinction training.

Experiment 2 sought to confirm the finding of Kesner and Cook (1983) that under standard conditions, where no pre-exposure to the injection ritual is given, the respon- siveness of tolerance to an environment change depends upon the distinctiveness of the acquisition environment. In addition, it compared the effect on tolerance of a 14-day rest period in the colony room with that of an explicit ex- tinction procedure in which injection-ritual stimuli (the pu- tative conditioned stimulus), in the form of placebo injec- tions, were administered unaccompanied by drug (the puta- tive unconditioned stimulus). To the extent that injection- ritual stimuli have control over the tolerance response, as the conditioning theory maintains, tolerance will be atten- uated in the extinction group but not in the rested group. The use of extinction tests in this way to confirm the ex- istence of an environment-drug association has previously been reported in studies of morphine (Siegel 1975; Siegel 1977) and alcohol tolerance (Crowell et al. 1981).

Materials and method

Exper iment 1 - injection ritual pre-exposure. Twelve male hooded Lister rats obtained from Bantin and Kingman Ltd., Aldbrough, Hull, England and weighing 400-580 g at the beginning of the experiment were used. They were maintained in individual cages on a freely available food and water schedule. The rats were assigned at random to two groups (N= 6) at the start of the experiment. The anal- gesia-testing apparatus consisted of a standard rat operant chamber (30 x 30 x 30 cm) with a grid floor composed of 0.32-cm diameter metal rods, through which electric shock could be delivered from a shock generator (Model 521C, Campden Instruments Ltd., London) via a constant current scrambler (Model 521S, Campden Instruments Ltd., Lon-

don) for 0.5 s. The apparatus was located in a room imme- diately adjacent to the colony room. All injections were subcutaneous in the dorsal surface of the neck. The dose used throughout the experiment was 10 mg/kg morphine sulphate in 10 mg/ml solution of 0.9% physiological saline. The volume of the physiological saline (placebo) injection was I ml/kg.

The design of the experiment is shown in Table 1. Dur- ing the initial pre-exposure phase (14 sessions) all rats were weighed daily in the colony room, administered a placebo injection, and immediately returned to the home cage. On each daily session of the tolerance acquisition phase (14 ses- sions) rats were weighed and injected in the colony room and then immediately transported in a mobile cage-rack into the adjacent room containing the testing apparatus. On alternate sessions this room constituted the distinctive and non-distinctive environment. On distinctive-environ- ment sessions the room was illuminated with flashing light from a Griffin xenon stroboscope (2 flashes/s) and a white noise stimulus of 70 dB was continuous; on non-distinctive sessions the room was illuminated with white light from an overhead 150-W light source identical to that in the colony room, and the white noise was absent.

The rats remained undisturbed in the test room for 30 min before being placed in the analgesia-testing appara- tus. During the test each rat was placed on the shock grid, and after a 60-s adaptation period, received an ascending series of 0.5-s footshocks. The intershock interval was usu- ally 6 s but shocks were only administered when all four paws were in contact with the grid. Shocks started at 0.5-mA intensity and increased in 0.5-mA steps until a vo- calisation response (defined as any detectable vocalisation) was obtained. At this point the rat was removed from the test box and returned to its cage and the shock intensity recorded. If a rat failed to make the criterion response when a current of 2.5 mA was delivered, the trial was terminated and an intensity of 2.5 mA recorded. Vocalisation threshold was used because it has been reported as being a more sensitive measure in this apparatus than either flinch or jump thresholds (Kesner and Cook 1983). After testing, the rats remained in the test room until a total of 60 rain had elapsed since they first entered it, and were then re- turned to the colony room. The only difference between the treatments of groups M-DIST and M-NONDIST in this phase of the experiment concerned the relationship of drug and environment. Thus, on distinctive-environment sessions group M-DIST received morphine and group M- NONDIST received placebo; on non-distinctive-environ- ment sessions, group M-DIST received placebo and group M-NONDIST received morphine.

On successive days following the tolerance acquisition phase, each group experienced an environment-change test. On the first test (T1) both groups were adinistered mor- phine and tested in the distinctive environment - this con- stituted an environment-change condition for group M- NONDIST; on the second test (T2) both groups were ad- ministered morphine and tested in the non-distinctive envi- ronment - this constituted an environment-change condi- tion for group M-DIST. For the six sessions following these tests the subjects were again exposed to the regimen of alternating distinctive and non-distinctive sessions which they had experienced during tolerance acquisition. This reacquisition phase was intended to counteract any distur- bances to the established environment-drug associations

103

Table 1. Experimental design

Group Injection-ritual Tolerance pre-exposure acquisition (A) (14 sessions) (14 sessions)

Environment-specific tests

(T1) (T2)

Reacquisition (six sessions)

Post-rest tolerance test (PRT)

M-DIST P-COLONY ROOM M-DIST/ M-DIST M-NONDIST P-NONDIST

M-NONDIST P-COLONY ROOM M-NONDIST/ M-DIST M-NONDIST P-DIST

M-DIST/ P-NONDIST M-NONDIST/ P-DIST

M-DIST

M-NONDIST

M = morphine; P = placebo; DIST = distinctive environment; NONDIST = non-distinctive environment

which may have been caused by the environment-change tests. Finally, both groups of rats were left undisturbed in the colony room for 14 days before receiving a post-rest tolerance test (PRT) - that is, an injection of morphine and analgesia assessment in the appropriate environment (distinctive-environment for group M-DIST; non-distinc- tive-environment for group M-NONDIST). 7-.5

Exper iment 2 - no pre-exposure. For the 12 rats (weighing 430-610 g) used in this experiment details concerning sub- jects, drugs and apparatus were identical to those described ~o for Experiment 1. The two groups of subjects (N= 6) also received identical treatments to those shown for M-DIST and M-NONDIST groups in Table l, except for the om-

1.5 ission of the injection-ritual pre-exposure phase and the addition of an extinction phase following the post-rest toler- ance test (PRT). In this final phase of the experiment, three subjects from each group were randomly selected and as- ~" signed to an Extinction group; the remaining subjects z %0 formed a Rest control group. Rats in the latter group were .- o treated exactly as they had been during the earlier rest phase of the experiment - that is, they were left undisturbed in N the colony room for 14 days; rats in the Extinction group ~ o,5 also remained in the colony room for 14 days but were o weighed and injected with placebo daily under identical conditions to those experienced in the earlier phases of the o.o experiment. On the day following the final extinction ses- sion each rat was injected with morphine and given an anal- gesia test in its appropriate environment, i.e. that which had been paired with morphine during tolerance acquisi- tion.

Results

Exper iment I - injection-ritual pre-exposure. As Fig. 1 shows, tolerance (a decrease in vocalisation threshold) clearly develops in both groups during the morphine ses- sions of tolerance acquisition. There appears to be no con- sistent change in either group in placebo sessions. Analysis of variance of vocalisation scores with Group (M-DIST vs M-NONDIST), Treatment (morphine vs placebo), and Sessions as factors revealed significant main effects of Treatment [F(1,10)=93.0, P0.05 in each case).

To check that tolerance levels in the two groups had not been differentially affected by the test sessions and were therefore equal prior to the final phase of the experiment, analysis of variance was conducted on the morphine ses- sions of the reacquisition phase with Group and Sessions as factors. Neither main effect nor their interaction was

104

9 EXT INCT ION

A REST

2.5

A

~2.0

O

I 11

n.

~1.0

N

< 0.5 (J O >

0.G

[] N I -D IST

O N I -NONDIST

N~I ! Nl- D I ST MORPHINE

N- - I IM-D IST PLACEBO

~Nl - NONDIST MORPHINE

O- OM- NONDISsT PLACEBO

q I

I i

I I I I I I ] .L J. I I I J . . L

A T1 T2 R PRT PET

Fig. 2. Mean vocalization threshold scores in Experiment 2. A ac- quisition sessions; T1 environment-change test 1 ; T2 environment- change test 2; R reacquisition sessions; PRT post-rest rest; PET post-extinction test. Test data shown_+ 1 SEM

significant (P>0A in all cases). The effect of the final 14-day "rest" phase was assessed by analysis of variance with Group and Stage as factors (the latter comparing scores on the post-rest tolerance test (PRT) with those on the last morphine session of reacquisition). No main effects or interactions were significant (P> 0.1 in all cases).

Thus, in Experiment t the groups differed neither in the acquisition of tolerance nor in the degree to which toler- ance was attenuated by a change in the environmental stim- uli accompanying the drug administration. There was no evidence of spontaneous recovery of the analgesic response after the 14-day rest period in either group.

Exper iment 2 - no pre-exposure. As Fig. 2 shows, tolerance developed in both groups and analysis of variance with Group (M-DIST vs M-NONDIST), Treatment (morphine vs placebo), and Sessions as factors revealed significant main effects of Treatment [F(1,10)=81.8, P 0.2 in all cases).

The difference between the effects of the environment- change in Experiments 1 and 2 attests to the efficacy of injection-ritual pre-exposure in modifying stimulus control of tolerance. However, this conclusion is weakened some- what by the fact that it requires an interexperiment compar- ison. Stronger evidence would be provided by combining the environment-change data from the two experiments and performing a single, factorial analysis - a legitimate proce- dure since the Subjects in the two experiments were treated identically except for the variable under examination (the presence or absence of injection-ritual pre-exposure). Thus, an Analysis of Variance was performed on the combined data with Pre-exposure condition (pre-exposure/no pre-ex- posure), Drug-environment (M-DIST/M-NONDIST), and Test-environment (same/changed) as factors. There was a significant main effect of Test-environment IF(l,20) = 5.07, P < 0.05], and a significant interaction of Pre-exposure con- dition Test-environment IF(I,20)= 14.1, P 0.2) in all cases. The significant interaction was examined further by means of Simple Main Effects analysis, which showed that the environment-change manipulation was ef- fective in reducing tolerance only in the injection-ritual pre- exposed Subjects (F(1,20) = 9.13, P< 0.01].

As in Experiment 1, analysis of morphine session scores during reacquisition in Experiment 2 showed that in neither group had tolerance been affected by the environment- change tests, and that both groups were demonstrating equivalent levels of tolerance prior to the rest phase [Group, Sessions, and Group Sessions effects were all non-signifi- cant (P>0.05 in each case)]. Again, as in Experiment 1, a 14-day rest period in the colony room did not affect toler- ance levels in either group - Group and Stage main effects were non-significant, as was the Group Stage interaction (P>0.1 in all cases). However, as is clear from Fig. 2, a difference between the Extinction and Rest groups emerged on the final post-extinction tolerance test (PET). Analysis of variance of the extinction phase data, with Group of origin (M-DIST/M-NONDIST), Treatment (extinction/ rest), and Stage (pre/post Treatment) as factors, revealed significant main effects of Treatment [F(1,8)=5.26, P< 0.05] and Stage [F(1,8)=7.69, P

105

ment irrespective of its distinctiveness. This lends support to the view that, in studies where pre-exposure to injection- ritual cues has not been given (e.g. Kesner and Cook 1983), failures to demonstrate environment-specific tolerance may be due to the overshadowing of relatively non-distinctive environmental stimuli by injection-ritual stimuli, rather than to the operation of an independent habituation pro- cess. The same experiment found that a 14-day rest period in the colony room did not affect tolerance levels in groups trained in either distinctive or non-distinctive environments - a result in accord with a classical conditioning account of tolerance and which argues against an habituation inter- pretation. According to the latter, loss of tolerance should occur in the non-distinctive group due to spontaneous re- covery of the habituated analgesic response; according to the former, loss of tolerance would not be expected in either group in the absence of explicit extinction procedures (i.e. exposure to the drug-associated stimuli in the absence of drug).

Experiment 2 provided further confirmation of the con- ditioning/overshadowing analysis by showing that where the salience of injection-ritual stimuli is not reduced by a pre-exposure procedure such stimuli do indeed gain control of the tolerance response and overshadow drug-associated environmental stimuli - a conclusion derived from the inef- fectiveness of the environment-change manipulation ob- served in this experiment. Although it had been predicted, from the results of Kesner and Cook (1983), that such over- shadowing might only occur in the group made tolerant in the non-distinctive environment, in the present experi- ment it occurred in both groups. This discrepancy may be due to differences in the relative salience of environmental and injection-ritual stimuli between the experiment of Kesner and Cook (1983) and the present experiments - it is likely, for example, that the salience of injection-ritual stimuli will be affected by Experiment-specific factors such as the roughness of handling and the degree of discomfort caused by the injection itself. Such factors were, of course, held constant in the present experiments, and the only vari- able which can plausibly explain the differential effective- ness of the environment-change in Experiments 1 and 2 is the presence or absence of injection-ritual pre-exposure.

Experiment 2 confirmed the finding of the first experi- ment that 14 days' rest in the colony room, in the absence of all experimental treatments, did not attenuate tolerance in either group, but showed that subjects receiving an addi- tional 14 days of extinction (placebo injections) did show reduced tolerance, unlike subjects receiving a further period of rest. This finding also lends support to the conditioning/ overshadowing view - the compensatory conditioned re- sponse thought to mediate tolerance, like all conditioned responses, should be reduced by an extinction procedure but not by mere passage of time.

The results reported here should be taken as indicative of an associative mechanism in tolerance (as opposed to a non-associative one), and not as evidence for classical conditioning as opposed to habituation. One reason for emphasizing this distinction is that the classical condition- ing model requires evidence, independent of a reduced drug effect per se, of a learned response that is compensatory for the initial drug effect. Such evidence was not obtained or sought in the present experiments and has proved to be elusive in the literature generally (Hughes and Bardo 1978; Morris et al. 1981 ; La Hoste et al. 1980; Tiffany et al.

1983): A second reason for emphasizing an associative ver- sus non-associative distinction as opposed to a conditioning versus habituation one is that a recently proposed theory of tolerance suggests that associative tolerance phenomena, such as the environment-specificity effect and the oversha- dowing effect reported here, which have traditionally been taken as evidence for the classical conditioning theory, can be explained by an habituation model (Baker and Tiffany 1985). This model also emphasizes the difference between tolerance acquired in distinctive and non-distinctive envi- ronments but accounts for these differences in terms derived from a powerful theory of habituation proposed by Wagner (1976, 1979, 1981). According to this theory, habituation occurs to a stimulus which is already represented or "primed" in short-term memory, and an important feature of the theory is that a stimulus may be primed in memory either non-associatively, by a prior recent presentation of the stimulus (drug) itself (self-generated priming), or asso- datively, by presentation of stimuli (such as a distinctive context) previously paired with the stimulus (associative priming). According to this analysis, then, tolerance ac- quired in a non-distinctive environment primarily reflects a non-associative habituation process; tolerance in a dis- tinctive environment primarily reflects an associative habit- uation process. The advantages of this theory over the clas- sical conditioning account are that it does not depend on the demonstration of compensatory conditioned responses, and it can account for some features of tolerance which are not predicted by the conditioning account (Baker and Tiffany 1985).

Whether or not the habituation theory proves to be a more satisfactory account of tolerance than the classical conditioning model, an important implication of the experi- ments reported here is that no conclusion regarding the associative or non-associative properties of tolerance may be drawn from the outcome of an environment-change test unless precautions are taken to prevent unauthorised proce- dural stimuli (such as injection ritual cues) from entering into an association with the drug.

References

Baker TB, Tiffany ST (1985) Morphine tolerance as habituation. Psychol Rev 92:78-108

Crowell CR, Hinson RE, Siegel S (1981) The role of conditional drug responses in tolerance to the hypothermic effects of etha- nol. Psychopharmacology 73 : 51 54

Dafters RI, Anderson G (1982) conditioned tolerance to the tachy- cardia effect of ethanol in humans. Psyehopharmacology 78 : 365-367

Demellweek C, Goudie AJ (1983) An analysis of behavioural mech- anisms involved in amphetamine anorectic tolerance. Psycho- pharmacology 79 : 58-66

Groves PM, Thompson RF (1970) Habituation: a dual-process theory. Psychol Rev 77: 419-450

Hughes RA, Bardo MT (1978) Morphine analgesic tolerance in rats: a search for hyperalgesia. Paper presented to the Psychon- omic Society Meeting, San Antonio, USA

Kamin LJ (1969) Predictability, surprise, attention and condition- ing. In Campbell BA, Church RM (eds) Punishment and behav- ior. New York: Appleton-Century-Crofts, pp 279-296

Kesner RP, Baker TB (1981) A two-process model of opiate toler- ance. In:Martinez JL, Jensen RA, Messing RB, Rigter H, McGaugh JL (eds) Endogneous peptides and learning and memory processes. Academic Press, New York

106

Kesner RP, Cook DG (1983) Role of habituation and conditioning in the development of morphine tolerance. Behav Neurosci 97:4~12

La Hoste GA, Olson RA, Olson GA, Kastin AJ (1980) Effects of Pavlovian conditioning and MIF-I on the development of morphine tolerance in rats. Pharmacol Biochem Behav 13 : 799-804

La AD, Poulos CX, Cappell HD (1979) Conditioned tolerance to the hypothermic effect of ethyl alcohol. Science 206:1109-1110

Lubow RE (1973) Latent Inhibition. Psychol Bull 79:398-407 Morris RGM, Jonzen RAI, Welsh B, Cahusac PMB (1981) Paper

presented to Experimental Psychology Society Meeting, Liver- pool, UK

Siegel S (1975) Evidence from rats that morphine tolerance is a learned response. J Comp Physiol Psychol 5:498-506

Siegel S (1977) Morphine tolerance acquisition as an associative process. J Exp Psychol: Anim Behav 3:1-13

Siegel S (1983) Classical conditioning, drug tolerance, and drug dependence. In: Israel Y, Glaser FB, Kalant H, Popham RE, Schmidt W, Smart RG (eds) Research advances in alcohol and drug problems, Vol 7. Plenum Press, New York

Siegel S, Hinson RE, Krank MD (1978) The role of predrug signals in morphine analgesic tolerance: support for a Pavlovian condi- tioning model fo tolerance. J Exp Psychol (Anim Behav) 4:188

Tiffany ST, Baker TB (1981) Morphine tolerance in the rat: con- gruence with a Pavlovian paradigm. J Comp Physiol Psychol 95 : 747-762

Tiffany ST, Petrie EC, Baker TB, Dahl J (1983) Conditioned mor- phine tolerance in the rat: absence of a compensatory response and cross tolerance with stress. Behav Neurosci 97:335-353

Wagner AR (1976) Priming in STM: an information processing mechanism for self-generated or retrieval-generated depression in performance. In: Tighe TJ, Leaton RN (eds) Habituation: Perspectives from child development, animal behavior and neu- rophysiology. Hillsdale NJ : Erlbaum

Wagner AR (1979) Habituation and memory. In: Dickinson A, Boakes RA (eds) Mechanisms of learning and motivation. Hillsdale NJ: Erlbaum

Wagner AR (1981) SOP: a model of automatic memory processing in animal behavior. In Spear NE, Miller RR (eds) Information processing in animals: Memory mechanisms, Hillsdale, NJ : Erl- baum

Walter TA, Riccio DC (1983) Overshadowing effects in the stimu- lus control of morphine analgesic tolerance. Behav Neurosci 97 : 658-662

Received January 22, 1985; Final version April 10, 1985