Development of Context-Specific Tolerance to

Beha~ Ioral Neurosdcnce 19S6. ~,'OI I(ll'l. No 5, Ol 1-023

Cop', nghl I q86 b', lhe American P',~.cholo~cal Association. In,: fl735.7044,'86,$00.~5

Development of Context-Specific Tolerance to Morphine: Support for a Dual-Process Interpretation Michael S. Paletta and Allan R. Wagner Yale University Four experiments were concerned x~ith the dex elopment in rats o f context-specific tolerance to the sedating and analgesic effects of morphine. Experiment I ~as conducted to assess the temporal course of acti~ it.~changes and analgesia consequent to acute morphine administration. Experiments 2, 3, and 4 x~ereconducted to assess the development of context-specific morphine tolerance in the t~o measures under different conditions of pairing of morphine x~itha distinctive enx~ronment. In support of a dual-process model (postulating both a general tendency for eondiuoned diminution of unconditioned responding and a more restricted influence of the development of specific conditioned compensator3 responses), tolerance ~as observed in both measures, but e~idence of conditioned compensatory response ~as found only in the aetivit.~ measure. The dill'erent~alevidence of conditioned compensator3 response in the t~o measures x~as interpreted as consistent ~uh the fact that the activit) measure shox~ed a b~phas~c uncondiuoned response in Experiment I ~ hereas the analgesic measure d~d not.

The tolerance that a subject develops to certain drugs with repeated experience is. at least in part, an associative consequence (Siegel, 1975, 1977). The experiments reported here are concerned with elucidating the kind of associative processes that may be involved in morphine tolerance and ~ith evaluating the potential usefulness of a recent proposal of Wagner (1981) as to the circumstances under which "'conditioned compensator3.' responses," antagonistic to the prominent unconditioned response (Siegel, 1981), ma~ be implicated. Mitchell and his collegues (.Adams, Yeh, Woods, & Mitchell, 1969: Ferguson, Adams, & Mitchell, 1969) observed that morphine tolerance, as exhibited in a decreased analgesic effect, was greater when morphine had been experienced in the same situation in which the analgesic effect was eventuall.~ assessed. Siegel (1976: Siegel, Hinson, & Krank, 1978) subsequently made clear that morphine tolerance can be developed specific to an arbitrar3 environmental context in which the drug is experienced just as a Pavlovian conditioned response (CR) can be acquired to an arbitrar3 conditioned stimulus (CS). Siegel, Sherman, and Mitchell (1980) furthered an associative interpretation by demonstrating that contextspecific tolerance can be "'extinguished" by exposing animals to the context in the absence of morphine. And it is nox~ known that morphine tolerance evidences other characteristics of associative learning such as latent inhibition r 1977). retardation in acquisition with partial as compared with continuous reinforcement (Siegel, 1977, 1978), and conditioned inhibition (Siegel, Hinson, & Krank, 1981).

There are several explanations of this associative phenomenon. Siegel based his interpretation on the development of "conditioned compensator)' responses." Siegel (e.g., 1979, 1981J assumed that with repeated pairings of a distinctive environment (the CS) and the administration of morphine (the unconditioned stimulus [US]), the environment can come to produce a learned response (a CR) that is antagonistic to the usual physiological response to morphine (the unconditioned response [UR]), so that when the US is given in that environment the UR will appear to be reduced. The critical obserxation tbr this interpretation is that the conditioned compensatory response can be observed in the absence of the morphine-elicited UR, for example, the drug-experienced animal may evidence h.~pothennia when injected with saline as well as diminished hyperthermia when injected with morphine (Siegel, 1978). Baker and Tiffan.v (1985) presented an alternative interpretation based on Wagner's "'priming" theory (Wagner. 1976, 1978). This theor), which was developed to account for the phenomena of habituation (Davis, 1970) and "'blocking" (Kamin, 1969), presumes that a "'signaled" US produces less processing and a diminished UR, apart from any overt CR generated by the signal. The prototypical phenomenon is seen to be the "'conditioned diminution of the UR" (Kimble & Ost, 1961 ), rather than the development of a (compensator)') CR. According to this reasoning, an effect of morphine experience should be a context-specific tolerance to morphine, but not an altered behavior in the absence of morphine. More recentl.v, Wagner ( 1981 ) presented a comprehensive model (termed SOP) of associative learning and performance that allows for both the manner of effect emphasized by Siegel ( 1981) and that emphasized by Baker and Tiffan~ (1985). In essence, it supposes that when an environmental stimulus becomes a signal Ibr a US, it will lead to a "'conditioned diminution of the UR" per se but may in addition elicit CRs that variously mimic or are antagonistic to the characteristic UR. Evidence tbr this dual-process model was marshaled in analyses of the conditioned modulation of the response to nonpharmacological USs (Donegan, 198 I; Donegan & Wag-

The research reported and preparation of this manuscript ~ere supported in part b.~ National Science Foundation Grant BNS8023399 to A. R. Wagner. The authors acknowledge the contributions of Scott Fukomoto. who collaborated in prelimmar3 studies prior to the investigations reported. Correspondence concerning this article should be addressed to Allan R. Wagner, Department of Ps)cholog3, Yale Uni~ersit.~.Ne~ Ha~en, Connecticut 06520. 611

612

M I C H A E L S. P A L E T T A A N D A L L A N R. W A G N E R

ner, in press), and the argument was extended by analogy to cases of drug tolerance. Of special interest lbr the studies reported here is the suggestion of SOP concerning when drug tolerance should reflect the involvement of a " ' c o m p e n s a t o r " CR in addition to conditioned diminution of the UR. According to SOP, when a CR is acquired to an environmental signal, or CS, the CR can consistentl.v be expected to mimic a "'secondary" component of the UR, which ma.~ in turn be either similar to or antagonistic to the prominent, primary response to the US (see Wagner, 1981 ). Thus, in studies o f d r u g tolerance, it would be expected that there ~ould be evidence of a conditioned compensator), response, as emphasized by Siegel (e.g., 1981), only on those response measures in which the drugelicited UR also evidences a "compensator"" secondaD' phase. Indications of a conditioned compensator3.' response should otherx~ise be absent (or, indeed, be replaced by a conditioned mimicking response), and any associative tolerance should be a consequence only of the tendency for conditioned diminution of the UR. in order to evaluate this reasoning, morphine tolerance ~as investigated in rats b.~ using t~o measures that show different patterns as LIRs, namely, activit.~ and analgesia. Activity has been reported (Babbini & Davis, 1972: Fog, 1969) to show a biphasic UR, with an initial phase of hypoactivity being followed b v a s e c o n d a r ) p h a s e of h~peractivit.v. In contrast, analgesia, as measured b.v the tail-flick technique (D'Amour & Smith, 1941 ~, did not appear in exploratoD' studies to show a biphasic UR, but rather a monophasic hypoalgesia. It would thus be expected that tolerance would involve evidence of a conditioned compensato~ response in the case of the activib measure, but not in the case of the analgesia measure. Experiment I documented the different UR patterns in the two response measures under the general conditions of the subsequent studies. Experiments 2, 3, and 4 then evaluated the predicted pattern of context-specific tolerance and compensator3, responding. Experiment I As has been noted, in measures of general activity the unconditioned response to morphine has previously been demonstrated to show a biphasie course, with initial behavioral sedation being followed by hyperactivity (Babbini & Davis, 1972: Fog, 1969: Kamat, Durra, & Pradham, 1974). And using such measure, Mucha, Volkovskis, and Kalant (1981) presented evidence of a "'conditioned compensator)' response" (in the form of h.~peractivit.v, specific to the drug environment, anticipatoD to the receipt of morphine or subsequent to the receipt of saline). The existing data with respect to analgesia are less clear. Although the analgesia response has received more attention than any other measure in studies of context-specific morphine tolerance (Siegel, 1975: Siegel et al., 1981: Tiffany & Baker, 1981), the course of the UR has not been well documented. And the CR in man.v of the studies may not have been what it appeared to be (Baker & Tiffany, 1985: Tiffany, Petrie, Baker, & Dahl, 1983). Consider the so-called hot-plate test, in which animals respond by lifting the paw from a

heated floor and the analgesic effect of morphine is indicated by a longer latency to respond. Siegel (1975) presented considerable evidence that context-specific morphine tolerance is accompanied by a conditioned compensator, response, which ma.~ be seen in a faster response than prior to training, when saline is gixen in the environment where morphine had previousl.v been received. What must be entertained, however, is that this apparent conditioned hypera[gesia may be a secondary consequence of conditioned hyperactivit.~--the more active the animal, the more likely it is to lift its paw regardless of its sensitivity to the noxious stimulus. It may be impossible to obviate completely such reasoning in regard to analgesia assessment, but it is possible to use a measure that would appear to be less open to influence by general activity than the hot-plate test. One such measure invoh'es the tail-flick response to a noxious stimulus to the tail (D'Amour & Smith, 1941 : Lewis, Sherman, & Liebeskind, 1981 ). The response has a characteristic form that allows it to be distinguished from other ongoing activity. And it may be of interest that whereas the tail-flick response has been reported (e.g., Advocat. 1980) to evidence context-specific morphine tolerance, in one study (LaHoste, Olson, Olson, & Kastin, 1980) in which relevant assessment was made, there ~as no indiciation of conditioned c o m p e n s a t o r ' responding. However, in this study the context-specific tolerance was itself relatively uncertain. The tail-flick measure was used in the experiments reported here to tarot relatively independent assessments of analgesia and activity. Experiment 1 evaluated the activity and analgesia UR to a 5 mg/kg dose of morphine, which has been commonly used (e.g., Siegel, 1975, 1976, 1977, 1978: Tiffany & Baker, 1981) in studies of context-specific tolerance. The responses were monitored at intervals over 24 hr following drug administration.

Method Subjects The subjects ~ere 16 male Sprague-Da~le~ rats, approximatel.~ 50 da.~s of age, obtained from the Charles Ri~er Breeding Laboratories. They were maintained with tbod and water continuously a~ailable except during experimental sessions. Four testing squads of 4 animals each were formed on the basis of matched bod.~ weights l r a n g m g between 100 and 125 g at the start of the experiment).

Behavioral Apparatus Certain of the m a n i p u l a n o n s in this and subsequent experiments were conducted in the area of the animal's h o m e cage. [n addition, there ~ere two m e a s u r e m e n t de~ ices located in a separate laborator3. foonl.

Home rage. The animal's home cage ~as a 14 x 18 x 17 cm, dra~er-mounted individual unit, with one side of the cage and the floor constructed of ~ire mesh. The cages ~xere located in an animal colon3 room that was continuoush illuminated and maintained at 21 (+ 21 ~ throughout the experiment. .~ctivit.v-te~lmg em'iromntqtt. Actixit.~ ~ a s measured in a 19 x 14 • 9 cm, ventilated clear plastic box within a 47 x 47 x 67 cm

613

C O N T E X T - S P E C I F I C TOLERANCE isolation chamber. The plastic box rested on a spring mounting, ~hich allotted for ~ertical displacement (maximum of 7 mnt'~ as a function of the animal's movement, and a loam-rubber cushion, which isolated the apparatus from environmental vibration. Movement ~as monitored via a phonograph cartridge IAstatic Model 16J assembl.~ attached to the bottom of the animal's box Isee Donegan, 19811 and electricall 3 coupled to a Beckman T.~ pe RP Dynograph. Displacement produced a rapidl.~ damped, sinusoidal graphic record. The sensitivit3 of tile s.~stem was calibrated such that the dropping of a 3-g v.eight from a height of 10 cm upon the center of the box Iloaded ~ith the equivalent of an animal's v.eightk produced a peak graphic deflection o f 13 cm. There were four equivalent environments, as described. ~hich allotted 4 animals to be run concurrentl.~. For purposes of subsequent experiments (see Experiments 2, 3, and 41, the acti~ it.~ environment was made distinctive from the home cage in the Ibllo~ing ~a.~s. Fifteen beaded metal chains ~approximatel} 2.5 mm in diameter and 8.5 cm in length~ ~ere suspended in fi~e ro~s of three from the ceding of the box so as to interrupt e~enl3 the interior space and to provide a complex tactual stimulus. A 2-in. gat.ze pad, co~ered b.~ wire screening and saturated x~ith am.',l acetate, v, as placed in the plasuc chamber 10 rain before each experimental session to provide a distinctive odor. The isolation chamber was not dlumlnated but contained a salient auditor), surround (70-dB SPL, A scalel emanating from a ~entilation fan and a broad-spectrum noise source. The temperature of the experimental environment, as monitored, ~as 20 (___2~ *C. .,tlllll~C~l~l-tll~'r tl/~lltll21lll~. The analgesia-assessment dexice v.as a modification of the tail-flick apparatus dexeloped b~ D'Amour and Smith 11941). It consisted of an animal-restraining tube. 18 cm long and 6.5 cm in dianteter, adjoining a metal plate containing a 1.5-cm deep groove into ~hich the a m m a r s tail could be rested. A Graflex Compact Projector, modified b.~ inversion of the condenser lens, was mounted 2 cm above the tail plate so that the light from a 150-W, 110-V, General Electric projection lamp could be focused on the animal's tail at a point appro\imately mid~a.~ between the base and the tip. A stopwatch ~,as t, sed to measure the latenc3 from application of the thermal stimulus until the occurrence of a tail flick (see belo~). Four identical restraining tubes were emplo.~ed so that 4 animals could be tested sequentlall~ b3 positioning each in turn next to the tail plate and light source. The apparatus was located on a laboratou,, bench immediately adjacent to the chamber that housed the act~ it.~-testing apparatus.

Procedztre Following a period of habituation to an injection routine and to conlinement in the restraining tubes, all animals were tested in the activlt) IAc~ de~ ice and in the analgesia (Anl de~ ice, one in assocation ~ith a morphine (]kl) injection and the other in association ~ith a saline IS) injection, in different ordered combinations. Specificall.~, the lour squads of 4 animals received the separate testing sequences AcM-AnS. AcS-AnM. AnM-AcS. and AnS-AcM. ~hich provided o~erall counterbalancing in terms of the order of the response measures as ~ell as of the substance administered. Hat, ituation. ,Xll animals received 6 days of habituation experience, designed to reduce an} stress associated ~ t h injections and confinement in the restra.ning tube. On each day, ammals ~ere weighed at appro\imately 0830 and ~ere subsequentl} gixen two injections of ph.~siological saline 0.5 and 5.0 hr later. Injections ~ere administered sc in the dorsal neck area in a xok, me of I ml.."kg. Two hours after ~elghing and separate from the injection experience, each animal was removed from the home cage and placed in a restraining tube for 5 rain. All hab~tuauon and manipulations ~ere conducted in the area of the subject's home cage.

.4.~t,s~mcnt otacttvitr and ana/gesta. Follo~ing the completion of habitt, ation, each animal's acti~ it3 and analgesia v, ere assessed in relat,m to the adrninistration of morphine or saline in one of the fot, r orders pre~ iousl~ described. Each assessment, m order, occupied 2 experimental days and invoked a series of nine separate measures, the first taken 0.5-1.0 hr after the usual dail.~ v, eighing, and the remainder distributed over the next 24 hr. On each measurement occasion, animals ~ere transported in their designated running squads of 4 to the appropriate laborator3, environment, the measure ~as promptl3 taken, and animals ~ere returned to their home cages. Immediatel.~ prior to placing animals in the apparatus for the second measure in the series, they ~ere injected either with morphine sulfate {5 mg/kg in a 5 mg/ml solution ,,ith 0.9% saline solution} or with an equal xolume of the saline xehicle. Designating this test as beginning at the time of injection (0), the remaining tests were initiated 30 rain before and 30. 60, 120, 24{), 480, 960. and 1,441) min after injection. On the occasions of activity measurement, the 4 animals were placed in the four a~ailable chambers for a concurrent 5-min observation period. The activity score for an3 sample period was the number of graphic deflections greater than 13 mm r to the calibration unitl that occurred during the sampling time. On the occasions of analgesia measurement, the 4 animals were placed in the a,.ailable restraining tubes and then tested in rotation until four tests on each animal were completed. In each case, an animal's tail was placed in the grooved tail plate, and the light was turned on and was terminated by either a tail displacement or a 20-s maximum. The tail-flick response to the thermal stimulus v.as easily recognized as la) an isolated mo~,ement of the tail, (bl involving an abrupt curling motion that (c) vigorousl.~ displaced the tail from the groove. The animal could potentiall.', mo~e the tail from the groove as an accompaniment of gross bod3 movement. Howe',er, gross bod.~ movement rarel.~ occurred during the final three tests. These tests ~ere consequentl.~ selected for reporting. And m the unusual e~ent of a tall displacement during these tests that did not satisl3 the abodementioned characteristics of an isolated tail-fhck response, that test ~as discot.nted and replaced b~ an additional test alter tile customar3 interlest inter~al which ranged between I and 2 rain. The 20-s ina\imunl ~as set to preclude tissue damage. And as further precaution, the point of incidence of the hght on the tail ~as shifted after each test. The inean latency from light onset to the response ~or the 20-s ma,,imuml o~er the last three test occasions was used as the analgesia score for each observation period.

Data .4 nal.~'sis Because there ~ere no marked differences apparent in the pattern of activity or analgesia from those animals tested first or second on the respective measures, the data Ibr each response measure were combined oxer test orders. The 16 animals thus proxided eight obserxations of the activity time course and eight of the analgesia ume course, in conjunction ~ith morphine injecuon, each in comparison ~lth eight similar observations m conjunction with a saline i~oection.

Results and Discussion Figure 1 p r e s e n t s the m e a n activity c o u n t s a n d m e a n tailflick latencies in t h e successive o b s e r v a t i o n p e r i o d s in relation to e i t h e r a m o r p h i n e i n j e c t i o n o r a saline i n j e c t i o n . C o n s i d ering first the activity m e a s u r e , it c a n be o b s e r v e d t h a t t h e r e was little variation in t h e animals" activity o n t h e d i f f e r e n t m e a s u r e m e n t o c c a s i o n s in the series f o l l o w i n g t h e saline

614

MICHAEL S. P~LETTA AND ALLAN R. ~,~,AGNER

Fi,ez~re/. Temporal course ofacti~ it.~ changesand analgesia following morphine administration in Experiment I. ITop panel: Mean acti~ it~ counts in 5-rain measurement periods starting at the designated times preceding and follo~ing morphine or saline injection. Bottom panel: Mean latenc} of the tail-ilick response to a thermal stimulus under comparable conditions. injection. In contrast, the morphine injection produced an initial decrease in activit3 which was observed in the first 5 min alter injection and was still prominent on the third postinjection measure beginning 60 rain later. This effect was followed by hyperactivity, observed on the measure beginning 240 min alter injection, and eventual return to the saline baseline. Analysis of variance of the successive activity observations in the t~o series associated with morphine versus saline revealed a highly reliable Drug X Observation interaction, F~8, 112) = 6.9, p < .001. Subsequent comparisons of the acti~it.~ following morphine in relation to that following saline, at each of the observation periods, indicated a reliable hypoactivity during the first 5 rain alter morphine injection and in the similar 5-rain periods beginning 30 and 60 rain postinjection, ts(l 4) .05. Anabsis of variance revealed that whereas there was a reliable difl~erence anaong the tbur groups, F(3. 24) = 9.5, p < .001, the onl.v painvise comparison to reach statistical significance was the aforementioned shorter latencies in Group M-E30 than in Group M-HC30. There were no apparent differences among the four groups on the saline test day. The indication of less context-specific tolerance on both the activity and the analgesia measure in Group M-E90 compared ~.ith Group M-E30 is in agreement with the predictions from SOP (Wagner, 1981), as previously outlined. The data from Group M - E l 0 ~ere less consistent, which sho~s diminished tolerance in relation to Group M-E30 on the analgesic measure but substantial tolerance and conditioned compensatory responding on the activity measure. In combination, the data from Groups M-E90 and M-E 10 compared with those from Group M-E30 are at least encouraging to the suppositions of SOP concerning the opportunities for excitatoD and inhibitor3, learning that ma.~ take place.

Fig,re 6. Mean tail-flick latencies o11 the morphine and saline test days for the lbur groups in Experiment 4 for ~ hich activity scores are presented in Figure 5. lThe groups are identified in Figure 5.)

621 General Discussion

Donegan and Wagner (in press) reviewed the available literature on the development of context-specific modulation of the response to nonpharmacological USs and concluded that changes in the measured response reflected two processes, a diminished effect of a signaled compared with an unsignaled US (a "'conditioned diminution of the UR") and whatever overlying CR tendencies might be generated by the signal. They argued that the conditioned diminution of the UR is a perfectly general effect ahvays working to produce tolerance to the US whereas the CR process may. variously', add further to the appearance of tolerance (if the CR is antagonistic or "'compensator3'" to the measured UR), may have no role (if the CR is onhogonal to the measured UR), or may' offset or reverse the appearance of tolerance (if the CR mimics the measured response). In evaluation of the appropriateness of this reasoning, Donegan (1981) demonstrated among other things that one measured response to a signaled US could show potentiation while another concurrently showed diminution, when the tbnner included a substantial, mimicking CR and the latter did not. The point of the presently' reported studies is that the same manner of interpretation, formalized in Wagner's ( 1981 ) SOP model, appears to be useful in understanding the contextspecific tolerance to morphine. Both of the response measures employed, activity' and analgesia, showed a context-specific tolerance. In addition, the activity measure, but not the analgesia measure, showed a conditioned compensator3, response in the absence of morphine. An interpretation that appeals only' to the conditioned diminution of the UR (Baker & Tiffany, 1985). or only to a conditioned compensator)' response (Siegel, 1979), can not embrace the total pattern. The dual-process interpretation of SOP is calculated to do so. A challenge for any theor)' of tolerance that would appeal to the influence of CR tendencies is to specify when such tendencies should be expected to be implicated and, if implicated, when they should be expected to involve either a response that is compensator)' to the measured UR or a response that mimics the measured UR. Wagner (1981) has supposed that the form of the CR is always the same as the secondary response to the US. And although the theor)' is silent as to why the secondar)' response may have various relations to the prominent initial response to the US, it provides a potentially useful empirical rule. Thus, one should not expect a compensator3 CR, and the added evidence of tolerance it would afford, unless there is also a seconda~' UR of the same form. This rule was useful in the studies reported insomuch as the response measure (activity) that evidenced a secondar)' UR tendency antagonistic to the primary UR tendency also evidenced a compensator)' CR. The other response measure (analgesia) did not show the one and, predictably, did not show the other. One need not, of course, be persuaded b3 the present data to accept all of the relexant tenets of SOP. One could, for example, be impressed that the h.vperactivitv CR that was observed was foreshadowed by a hyperactivit3 component to the UR and be encouraged to suppose that some such mimicr) ma.v generalh obtain in instances of apparent "'compensa-

622

MICHAEL S. PALETTA AND ALLAN R. WAGNER

tory" CRs. But one could do so without assigning necessar3. significance to the fact that the h.vperactivity component of the UR ~as secondar) in time, or without assuming that it was in such position as a characteristic part of a stimulusprocessing cascade. Indeed, regarding the latter point, morphine produces so man) effects in parallel and in sequence that one must be uncomfortable with any provisional generalization that would treat its administration as though it involved a single stimulus. In fact, there are several classes of opiate receptors (e.g., Lord, Waterfield, Hughes, & Kosterlitz, 1976) and numerous receptor fields in the nervous system (e.g., Pert, DeWald, kiao, & Sivit, 1979) that ma.v be responsible for dissociable responses to morphine (e.g., Bozarth & Wise, 1984: Jacquet. 1979}. Thus, one could suppose that morphine, perhaps through different receptor mechanisms, produces separable hypoactivity and hyperactivity URs, with the former initially more profound but the latter more persistent, leading to a biphasic o~erall response course. And one could entertain various reasons ~h) the CR mimicked one feature rather than the other. Perhaps the mechanisms of conditioning favor more persistent components. Or perhaps the neuronal processes underlying the separate components are differentially conditionable. Eikelboom and Stewart (I 982) suggested that the CR ~ill mimic UR components that are a result of"afferent'" impact of the US or its sequelae, but not UR components that are a result of"efferent +. impact. At the present time one can only remain open to the rich possibilities that need to be evaluated. it should be important, however, to an) theoretical analysis of morphine tolerance and the development of conditioned compensator3' responding that these effects were not found to be increased in Experiment 4 b3 extending the animals' confinement in the experintental environment postinjection so as to overlap better in time the period of hyperactivit) observed in Experiment 1 (i.e., Group M-E90 compared with Group M-E30). Rather. evidence of tolerance and the development of conditioned compensatory responding was reduced. This finding is congruent with the tenets of SOP, which suppose that processing of the experimental environment (CS) in conjunction with the seconda~' (A21 processing of the US should lead to inhibition and decreased net associative tendencies. Before concluding, it must be emphasized again that there are some relatively uninteresting constructions that ma.~ be placed on certain of the essential findings. It has been taken as important that the activity measure gave evidence of conditioned hyperactivity in the absence of morphine (consistent with the occurrence of a secondar), hyperactivity phase of the UR) whereas the analgesia measure offered no indication of a conditioned hyperalgesia in the absence of morphine {consistent with the absence of a secondary hyperalgesia phase of the UR). But is is possible to raise questions with respect to both of these measures. We have previously acknowledged that it is possible that the analgesia measure was insensitive to detecting conditioned hyperalgesia that may have been present. The results of Experiment 3 did not encourage this view, but it is still possible. Siegel and MacRae (1984), in commenting on other reports of context-specific tolerance in which compensator)., CRs have

not been observed, suggested that in some response measures compensator3.' CRs may not be readil.v detectable in isoLation because of the counteraction of "'regulator3.,, homeostatic influences." And they suggested that in such case a tendency to manifest a CR may be more observable if the response s.vstem is "challenged" or "'primed. `+Although it is unclear how much credence to give to this reasoning or the evidence that inspired it (see, e.g., Goudie & Griffiths, 1984k it is surel), conceivable that conditioned hyperalgesia ~ould have been observed in the present studies had analgesia been evaluated under other pharmacological conditions. The alternative construction that must be noted with respect to the activity measure is that the evidence of general hyperactivity in the experimental environment in the several comparisons of Group M-E with Group M-HC may have been the result of differential habituation rather than the development of conditioned compensator), responding. Animals recei~ ing saline in the experimental environment showed less activity over successive experiences. Notice, for example, in Figure 2 the diminished preinjection activity levels of Groups M-HC and S on the 2 test days of Experiment I in relation to their levels on Day I. In comparison, the preinjection activity level of Group M-E in the same environment was maintained at nearly the same level on the test days as on Day I. Thus. one could suppose that morphine injection simply interfered with habituation. We know of no demonstration of this particular effect of morphine+ but it is a reasonable possibility given the fact that low doses of morphine have been reported (e.g., Izquierdo et al., 1980) to produce a retrograde amnesic effect in studies of associative learning. The present studies are not uniquely subject to this possibility. It will simply be necessa~' eventually, as it has been in other circumstances (e.g., Pfautz, Donegan, & Wagner, 1978). to distinguish c a r e f u l l ) t h e differential development of a CR from any differential "'protection from habituation."

References Adams, W. H., Yeh, S. Y., Woods, k. A., & Mitchell, C. L. (1969). Drug-test interaction as a factor in the development of tolerance to the analgesic effect of morphine. Journal o/Pharmacolog)" and Experimental TherapettttL~, 168. 251-257. Advocat, C. (1980). E~idence for conditioned tolerance of the tailflick reflex. Behavioral and Neural Btolo.~y. 29. 385-389. Babbini, M., & Davis, W. M. (1072). Time-dose relationships for locomotor activit) effects of morphine after acute or repeated treatment. Briti,~h Jottrnal t?l Pharmacolo.~y, 46. 213-224. Baker. T. B., & Tiffan.~, S. T. (1985k Morphine tolerance as habituation. P~ychoh~gical Revtew, 92. 78-108. Bozarth, M. A., & Wise, R. A. (1984~. Anatomically distinct opiate receptor fields mediate reward and ph)sical dependence. Science. 224. 516-517. D'Amour. F. E., & Smith, D. L. 11941 p. A method for determining loss of pat n sensation. Journal ~!tP/larma~ oh:e)' arid Exp~'r/mctlta/ Theral,elttic~, "2. 74-79. Da~ is, M. (I 970k Effects of interstimulus interval length and variabdity on startle response habituation. Jottrnal o lComparative and Phl'Molo.eical Ps.|,chology. "2. 177- [92.

CONTEXT-SPECIFIC TOLERANCE Donegan, N. H. (I 9811. Priming-produced l~cilitatton or diminution of responding to a Pavlovian unconditioned stimulus. Jour,al of E~perime,/al Pwchok~,ew Animal B~qmvtor Proce~ ~e~, ". 295-312. Donegan, N. H., & Wagner, A. R. (in press). Conditioned diminution and thcilitat=on of the UR: A sometimes opponent process interpretation. Iq 1. Gormezano, W. F. Prokas~, & R. F. Thompson ( Eds. 1, ("ht~/~ a/~'omlitton/,g IlL Behavior. neuroph.l'~/oh~glcal, am/ ttettror heroical sttldtr m / h e rabbit Hillsdale, N J: Erlbaum. E=kelboom, R., & Ste~arl, J. 11982). Conditioning of drug-induced ph.~siolog=cal response. Pwchohteical Ro'tew, 89. 518-527. Fanselo~, M. S., & German, C. (1982). Explicitl.~ unpaired deh~er3 of morphine and the test situation: Extinction and retardation of the suppressing effects of morphine on locomotor activit.~. Behavioral and ,Vettral Bloh)gv, 35. 231-242. Ferguson, R. K., Adams, W. J., & Mitchell. C. L. (1969). Studies of tolerance development to morphine analgesia in rats tested on the hot plate. Eltlopeul/ Jottr/lal Q/'P//ttr//la~'oh)ff.l', 8. 83-92. Fog, R. ( 1969L Behavioral effects in rats of morphine and amphetam i ne and of a corn bi nation of the t~ o drugs. P~.vchopharmacolo.ffla, I0. 305-312. Goudie, A. J., & Griffiths, J. W. (1984). Environmental spec~fic~t.~ of tolerance. Trt,/ids ill .%k,l/ro~c/Olce. ". 310-311. Hinson, R. E., & Siegel, S. 11983). Anticipator3 hyperexcilabilit.~ and tolerance to the narcotizing el'l~ct of morphine in the rat. Behavioral .Vt,lo'o~cte,ce, 9". 759-767. Izquierdo, I., Dias, R., Souza D., Carrasco, M., Elisabetsk3, E., & Perry. M. (1980). The role of peptides in memory and learning. Behavioral Biz//n Re~ct,arch, I. 451-468. Jacquet. Y. F. (19791. d-endorphin and ACTH-opmte peptides with coordinated roles in the regulation of behavior. Trettds tit Neurost ience, 2. 140-143. Kamat, K. A., Dutta, S. N., & Pradham, S. N. i' 1974). Conditioning of morphine-induced enhancement of motor activit.~. Re~earch in Otem/cal Patholo.er a,d Pharmacohl~y, ". 367-373. Kamin, L. J. 11969). Predictability. surprise, attention, and conditioning. In J. B. Campbell & R. Church (Eds.), Pu,i~hment and aver~ive ~olldtt/O/ltl/g ( pp. 279-2961. Ne~ York: Appleton-Centur3Crofts. K=mble. G. A., & Ost, J. W. P. (19611. A conditioned inhibitory process in e3elid conditioning. Journal of E~pert,w,tal P~,.vchology, 6/. 150-156. Krank, M. P., Hmson, R. E., & Siegel, S. (19811. Conditioned h.~peralgesia is elicited b.~ environmental s~gnals of morphine. Behavh,zd and Ne,ral Biolo,ev, 32. 148-157. LaHoste. G. L., Olson, R. D., Olson, G. A., & Kastm. A. J. 11981)). Effects of Pa~ Io~ ian conditioning and MIF-I on the development of morphine tolerance in rats. P/larl/lttco[o,ffl' Biochemis/rr told Behavior, 13. 799-804. Levels, J. W., Sherman, J. E.. & Liebeskind. J. C. 11981 ). Opioid and non-opio=d analgesia: Assessment of tolerance and cross-toleranee. Jotlrtltll O/',\ellrO~t'lellt ('. 1. 358-363. Lord, J. A. H.. Waterfield, A. A., Hughes, J., & Kosterlitz, H. ~.~,. (1976). Multiple opiate receptors. In H. W. Kosterlitz (Ed.), Oplate~ and emh~,eenou~ opiotdpept/de~ Ipp. 275-28()L Amsterdam: Elsevier/North-Holland B=omedical Press. Mucha, R. F., Volko~skis, C., & Kalant, H. (1981k Conditioned increases in locomotor acti~it.~ produced with morphine as an unconditioned stimulus, and the relation of conditioning to acute morphine elt~ct and tolerance. Jottr/tal olComparattve al/d P/I)'~lohtgtcal Psychoh~gy, 95, 351-362. Pert, A., DeWald, L. A., Liao, H.. & Si~ it. C. 11979). Effects ofopmtes

623

and opioid peptides in motor behaviors: Sites and mechanisms of action. In E. U ~ i n , W. E. Bunny. & N. S. Kline (Eds.), Emh)rphmes m mental heahh re~ealz'h (pp. 45-61 k New York: Oxlbrd LJniversity Press. Pfautz, P. L., Donegan, N. H., & Wagner, A. R. (1978). Sensor). precondit=oning xersus protection from hab=tuation. Jol/rnal of E.\~erlmell/al P~)'thoh).~|' .4ntmal Behavior Processes, 4, 286-295. Sherman, J. E. 11979). The effects of conditioning and novelty on rat's analgesic and pyretic responses to morphine. Lear/ling a/iN ,th~tiva/ion, 10. 383-418. Siegel, S. (1975). E~idence from rats that morphine tolerance is a learned response. Jo,rnal of Comparative and Ph)'siolo,~itYtl P,~.I'cholo,ey, 89. 498-506. Siegel, S. 11976). Morphine analgesic tolerance: Its situation specificit3 supports a Pa~lovian conditioning model. Scl~qlce, 193. 323325. Siegel, S. (1977). Morphine tolerance acquisition as an associative process. Jottrllal of E.~peri/ne/l/al Psychoh)g.t'. .4tlimal Behavior Proces~e~, 3. I- 13. Siegel, S. (1978). Tolerance to the h~perthermic effect of morphine in the rat is a learned response. Journal of Comparattve atut Physioh)gical PsTchology, 92, I137-1149. Siegel. S. (1979). The role of conditioning in drug tolerance and addiction. In J. D. Keehn (Ed.), P~ychopathology i, a,imal~ Resealz'h apphta/ions (pp. 143-168). Ne~ York: Academic Press. Siegel, S. 11981). Classical conditioning, drug tolerance, and drug dependence. In Y. Israel, F. B. Sloser, H. Kalant, R. E. Popham, ~,~,. Schmidt, & R. G. Smart (Eds.), Rest,arch adva/lce~ ill alcohol atld dru.e at, use ( Vol. 7, pp. 2(17-2461. Nex~ York: Plenum Press. Siegel, S., Hinson, R. E., & Krank. M. D. 11978). The role of predrug signals in morphine analgesic tolerance: Support for a Pavloxian conditioning model of tolerance. Journal qf E,x7~erimental P~ychology. .~ntmal Behavior Processes, 4. 188-196. Siegel, S., Hinson, R. E., & Krank, M. D. (1981). Morphine-induced attenuation of morphine tolerance. & iotce, 212. 1533-1534. Siegel, S., & MacRae, J. 11984). Environmental specificit3 of tolerance. Trem# tit Neuroscwnce, ~. 140-143. Siegel, S., Sherman, J. E., & Mitchell. D. (19801. Extinction of morphine analgesia. Learni/tg altd .lh)/ivatt~m, I I. 289-293. Tiffan.~, S. T., & Baker, T. B. (1981). Morphine tolerance in rats: Congruence ~ ith a Pa~ 1o~ian paradigm. Jollrllal r Compalzztive and Ph.t'sioh)gical P~'.rchology, 95. 747-762. Tiffan.~, S. T., Petrie, E. C., Baker, T. B., & Dahl, J. L. (1983). Conditioned morphine tolerance in the rat: Absence of a compensator3 response and cross-tolerance with stress. Behavioral Neurost'le/lce, 9". 335-353. %agner, A. R. (1976). Printing in STM: An information processing mechanism for self generated depression in performance. In T. J. Tighe & R. N. Leaton (Eds.), Habi/uatton: PervlecttvesJh)m chifll deveh)pmem, animal behavior, and neuroph)'~ioh)g.v (pp. 95-128). Hillsdale, N J: Erlbaum. Wagner. A. R. (1978). Expectancies and the priming of STM. In S. H. Hulse, H. Fowler, & K. W. Homg (Eds.), Cognitiveprocesse3 ill animal behavior [pp. 177-2091. Hillsdale, N J: Erlbaum. Wagner, A. R. I 1981 I. SOP: A model ofautomatic memor3, processing in animal behavior. In N. E. Spear & R. R. Miller (Eds.), h!lbrmalio, pro(es~mg in allimal~ Memoo" mechanisms (pp. 5-471. Hillsdale, N J: Erlbaum. Received N o v e m b e r 14, 1984 Revision received S e p t e m b e r 23, 1985 9