[Opioides]Evidence From Rats That Morphine Tolerance is a Learned Responde - Siegel 1975

8/19/2019 [Opioides]Evidence From Rats That Morphine Tolerance is a Learned Responde - Siegel 1975

1/9

Journal of Compamtht an d Physiological Psychology1975, Vol. 89, No. 5, 468-506

Evidence from Rats That Morphine ToleranceIs a Learned Response

Shepard SiegelMcMaster University, Hamilton, Ontario, Canada

It is proposed that th e direc t analgesic effect of m orphine becomes a t tenu-ated over the course of successive adm inistra tions o f the narcotic by a con-ditioned, compensatory, hyp eralgesic response elicited by the administra-tion procedure, the net result being analgesic tolerance. Using the ho tplate analgesia assessment situation with rats, this conditioning view oftolerance is supported by several findings: (a) It is necessary to have relia-ble environm ental cues predicting the systemic effects of m orphine if toler-ance is to be observed, (b) a hyperalgesic conditioned response m ay be ob-served in m orphine-tolerant subjects when drug adm inistration cues arefollowed by a placebo, and (c) m ere ly b y repeatedly presenting environ-mental cues previously associated wit h m orph ine (but no w presented with aplacebo), morphine tolerance can be extinguished.

Tolerance is said to have developed when,•after repeated administrations, the effectof a given dose of a drug is less than it wasoriginally. Tolerance to m any of the effectsof narcotics (especially opiates), such asanalgesia, develops rapidly and reliably,and numerous hypotheses have been pre-

sented to account for the phenomenon. Insummary, (a) the relevant effect of earlyexperience with the drug may be to alter theorganism's metabolism such that the drugis subsequently m ore efficiently metabo-lized (Mute & Woods, 1969); (b ) after thedrug molecules exert their action on centralreceptor sites, they may continue to occupythese sites thereby decreasing the popula-tion of receptor sites that can be stimu-lated by the same drug on a later occasion

(Schm idt & Livingston, 1933); (c) the ini-tial drug administrations may induce theformation of silent receptors, which func-tionally reduce the effects of later drug ad-ministrations by serving as dead spotreceptors for drug molecules that would

This research was supported b y Gran t MT-3577from th e Medical Research Council of Canada.The assistance of Doreen Mitchell is gratefullyacknowledged. The m orphine sulfate used in theseexperiments was supplied by Health and WelfareCanada (Scientific Services, Health ProtectionBranch) .

Requests for reprints should b e sent to ShepardSiegel, Department of Psychology, McMasterUniversi ty, H am il ton, Ontario, L8S 4K1, Canada .

otherwise stimu late active pharm acologicalreceptor s (Collier, 1965); or (d) narcoticsmay be conceived of as antigens, with toler-ance reflecting an im m unitylike process(Cochin & K ornetsky , 1968). A ll thesetheories of tolerance, hereinafter groupedtogether as physiological theories (for a re-

view, see Cochin, 1970), postulate some sys-temic change within the organism as a re-sult of the initial drug experience that de-creases receptor sensitivity to the drug,allows the drug to be metabolized morequickly, or serves to bind the drug beforeit can exert its action.

An alternative approach, proposed here,might be termed a conditioning theory oftolerance. According to this view, narcotictolerance is the r esult of the lear ning of an

association between the systemic effects ofthe drug and those environmental cues thatreliably precede these systemic effects.Pavlov (1927, pp. 35ff) suggested that theadm inistration of a dru g could be viewed asa conditioning trial, with the actual phar-macological assault constituting the un-conditioned stimulus (UCS) and the im-mediately antecedent environmental cuesserving as the conditioned stimulus (CS).The developm ent of the association betweenthese stimuli may be revealed if the subject,after a history of adm inistration of the drug,is presented with the drug administrationprocedure not followed by the systemic ef -

498


2/9

MORPHINE TOLERANCE AS A LEARNED RESPONSE

fects of the drug—rather, for such a con-ditioned response (CR) test session, aplacebo is administered.

There has been a considerable amount ofinterest in the theoretical and practical im-portance of conditioned pharmacologicalresponses (Gantt, 1957; Loucks, 1937; Sie-gel & Nettleton, 1970). Although manyfo rms of drug CRs can be conceptualizedand have been reported, conditioned drugresponses are commonly opposite in directionto the unconditioned effects of the drug.Thus, it has been reported that in animalswith a history of administration of an anti-cholinergic drug, such as atropine or Ditran,

which induces antisialosis, the administra-tion procedure without the drug leads toexcessive salivation (Lang, Brown, Ger-shon,& Korol, 1966; Wikler, 1948); in ani-mals in which tachycardia has been repeat-edly induced by injections of epinephrine,the injection procedure alone causes a de-crease in heart rate (Subkov & Zilov, 1937);subjects repeatedly made to evidence allergicreactions by allergen injections evidenceimmune reactions when confronted with theinjection procedure (for a review, see Hull,1934, pp. 413-416); in animals with a historyof administration of a hyperglycemic agent,such as glucose or epinephrine, the adminis-tration procedure alone leads to a decreasein blood sugar (Deutsch, 1974; Mityushov,1954; Russek & Pina, 1962); in organismsrepeatedly experiencing the hypoglycemiceffects of injected insulin, th e injection pro-cedure alone leads to an elevation in blood

sugar (Lichko, 1959; Siegcl, 1972, 1975).1

These anticipatory responses, being com-pensatory in nature, should serve to at-tenuate the drug-induced unconditionedresponse (UCR), therefore the net effect ofthe drug should decrease over successivedrug administrations. Such a decreasedresponse to a drug as a function of successiveexperiences with the drug defines tolerance.

According to the present conditioningtheory, tolerance to the analgesic effects of

1 Although the CR to physiological doses ofinsulin appears to be a compensatory hyperglyce-m ic response, conflicting findings have been re-ported when the UCS consists of very large dosesof the hormone (see Siegel, 1975).

morphine results because environmentalcues regularly paired with the administra-tion of the drug come to elicit a compensa-tory CR, hyperalgesia, which algebraicallysummates with the stable, unconditionedanalgesic effects of the narcotic. Thus, en-vironmental cues consistently predictingthe systemic effects of the drug should becrucial to the development of tolerance sincethey enable the subject to make timely com-pensatory CRs in anticipation of the anal-gesic UCR. Several experiments by Mitchelland his colleagues (Adams, Yeh, Woods, &Mitchell, 1969; Kayan, Woods, & Mitchell,1969) indicate that the rate of development

of analgesic tolerance to morphine is highlydependent upon the availability of environ-mental cues uniquely present at the time ofdrug administration. Using the standard hot plate assessment situation (Johan-nesson & Woods, 1964), in which pain sensi-tivity in the rat is assessed by observing itslatency to lick a paw when placed on a warmsurface, these investigators reported thatanalgesic tolerance to morphine developedmuch more rapidly when subjects were con-fronted with the distinctive analgesia as-sessment environment on each of the fiveoccasions that the drug was administered(even if the nociceptive stimulation was notapplied until the fifth occasion) than if theywere introduced into this environment forthe first time on the fifth occasion that thedrug was administered. Experiments 1Aand IB were designed to assess the reliabilityof these reports of the importance of drug-

associated environmental cues in the acqui-sition of tolerance.

EXPERIMENT 1A

MethodSubjects and apparatus. The subjects, 29 experi-

mentally naive, male, 90-110 day-old , Wistar-de-rived rats (obtained f rom Quebec Breeding Farms,St. Constant, Quebec, Canada), were housed inindividual cages with food and water freely avail-able. Responsivity to pain was assessed using re-

cent modifications of the hot plate technique(Eddy & Leimbach, 1963). Briefly, a 1,200-mlm i x t u r e consisting of equal parts by vo lume ofacetone and ethyl formate was boiled in a rec-tangular copper container (19 X 19 X 15 cm). Th econtainer w as completely enclosed with the ex-


3/9

SHEPARD SIEGEL

ception of provision for a condenser coil to liquifyth e vapor and re turn it to the vessel. The tem-perat ure of the vapor and the top surface of thevessel were constantly monitored and remainedat 54.2° C (±.2° C). Pain sensitivity was

assessed by placing the rat on the surface of thecontainer for 1 min and n oting the num ber of sec-onds that elapsed until it first licked a paw (here-inafter referred to as the paw-lick latency ). Thus,analgesic responses ar e indicated by relativelylong paw-lick latencies an d hyperalgesic responsesby relatively short paw-lick latencies.

Procedure. Three groups of rats received equiva-lent morphine injections on four occasions, theinterval between inj ections being 48 hr . The fourthoccasion was the test session, during which th epain sensitivity of all the rats was assessed on thehot plate in the test environment subsequent to

the d rug inject ion. The groups differed only withrespect to the cues associated with the m orphineadministrations on the three prior sessions.

One morphine group was used to demonstratethe initial analgesic UCR of the drug and the de-velopment of tolerance over the successive drugadministrations. This group was treated identi-cally on Sessions 1-3 and Session 4, i .e. , m orphine-induced analgesia was assessed on the hot plateapparatus in the test environment each time the•drug was adm inistered (Grou p M-HP, i.e., mor-phine-hot plate; n = 8). For each session, rats inthis group were transported in their home cages

from the colony room to a different room, whichcontained the hot plate apparatus, subcutaneouslyinjected with a 5 mg/kg dose (of a 5 mg/cc solu-tion) of morphine sulfate, and .5 hr later, placedon the hot plate apparatus.

A second m orph ine group was included to assesswhether an y apparent drug tolerance observed inGroup M-HP, as revealed by decreasing paw-licklatencies across sessions, may be attributable tomerely increasing practice in m aking the possiblypain-ameliorating, paw-licking response whiledrugged rather than to any functional decrease inth e n arco tic's analgesic properties. This group was

treated like Group M-HP except that the hotplate apparatus was not turned on until the four thtest session; for Sessions 1-3, these rats wereplaced on the surface of the vessel when it was atroom temperature (21.2°-22.2° C). For this secondmorphine group (Group M-CP, i.e., morphine-cold plate; n = 7), the environmental cues preced-ing th e morphine effects an d analgesia assessm entwere the same on Sessions 1-3 as on Session 4.However, these rats never pr acticed the paw-lick-ing response on the hot surface until th e testsession; thus, relatively rapid reactions on Session4 by Group M-CP rats would be attributable to

drug tolerance rather than to acquired proficiencyin paw-licking while narcot ized.A third group of rats suffered the same systemic

effects of the mo rphine equally as often and at thesame intervals as rats in Groups M-HP and M-CP except that a different set of cues was associ-ated with the systemic effects of the drug nn th e

first three sessions than on the four th , hot platesession. For this group, for Sessions 1-3, the m or-phine w as administered in the colony room simplyby removing the rat from it s cage, injecting thedrug, and r e turn ing the rat to its cage. Thus, this

group (Group M-CAGE; n = 8), like Group M-CP, had its first morphine-induced analgesia as-sessment on Session 4, and differed from GroupM-CP only because the environmental cues sur-rounding the m orphine adm inistration were dif -ferent on the three prior occasions tha t the drugwas administered.

Finally, a four th group also received four hotplate analgesia-assessment sessions in the testenvironment , but all the injections were physio-logical saline rather than morphine. This group(Group S, i.e., saline; n = 6) provided an un-drugged baseline, indicating th e effects of the re-

peated injections and hot plate experiences per se.Results and Discussion

The mean paw-lick latency for each groupon each session is shown in Figure 1 (forGroups M-CP and M-CAGE, of course,the response was assessed only on the lastsession). As indicated in Figure 1, the char-acteristic analgesic effect of morphine wasobserved on the first session—rats receivingthe narcotic for the first time (Group M-

HP) had significantly longer paw-lick la-tencies than rats receiving saline U = 7.5,p < .02). As was also expected, the analgesiceffects of morphine became less and lesspronounced on the successive sessions inwhich the drug was administered; a Wil-coxon matched-pairs signed-rank test indi-cated that Group M-HP rats had signifi-cantly shorter paw-lick latencies on Session4 than Session 1 T = 0, p = .01). As is alsoapparent in Figure 1, on Session 4 GroupM-CP rats responded to the hot plate asdid Group M-HP rats, the small differencebetween the two groups not approachingstatistical significance, that is, Group M--CP rats evidenced the short-latency paw-licking response indicative of morphinetolerance on the fourth occasion that theyreceived the drug despite the fact that theynever experienced the heat stimulation anddid not practice the response on Sessions

1-3, indicating that such practice (cf. Kay anet al., 1969) or repeated experience withthe morphine while nociceptively stressed(cf. Adams et al., 1969) is irrelevant to thedemonstration of analgesia tolerance inthis test situation.


4/9

MORPHINE TOLERANCE AS A LEARNED RESPONSE 501

Both groups of morphine-tolerant rats,M-HP and M-CP, evidenced Session 4paw-lick latencies not significantly differentfrom that of Group S rats, despite the factthat Groups M-HP and M-CP were re-

ceiving morphine and Group S physio-logical saline prior to analgesia assessment.As is obvious in Figure 1, Group M-

CAGE, in marked contrast to Groups M-HP and M-CP, was not tolerant to theanalgesic effects of morphine. The differ-ences between Group M-CAGE and eachof Groups M-HP and M-CP were statis-tically significant (U = 5.5, p < .002 an dU = 12, p < .04, respectively). Indeed, theSession 4 response latency of Group M-CAGE was not significantly different fromthat which would be expected the very firsttime rats receive the drug and analgesiaassessment (i.e., the Session 1 value for M-HP rats). Thus, these results confirm theearlier reports of the importance of drug-as-sociated environmental cues in the develop-m ent of tolerance (A dam s et al., 1969;Kayan et al., 1969).

Prior to the test session, all three mor-

phine groups suffered the same morphine-induced systemic effects equally as oftenand at the same intervals. The rats in thesegroups should have been subjected to thesame metabolic, cellular, or immunifacientmodifications presumed to be responsible fortolerance (see Cochin, 1970). It would beexpected, according to any of the physio-logical theories of tolerance, that the threegroups should be equally tolerant to theanalgesic effects of the narcotic. That GroupM-CAGE, the group that received the drugin a distinctly different environment for thepretest sessions, does not evidence any indi-cation of tolerance to morphine on the testsession suggests that reliable cues associatedwith the drug administration are importantin affecting the development of tolerance.

EXPERIMENT IB

Method

Subsequent to the test session, Group M-CAGE rats continue d to be injected with m orphineat the same 5 mg /kg dose and pain sensitivitywas assessed for three addition al sessions (theinterval between sessions re m aining at 48 hr ) sothat the course of the development of tolerance

FIGURE 1. M ean paw-lick latency on hot platefor four sessions for groups in Exper iment 1A,(Abbreviations used in group names :S = saline;M = m orphine ; HP = hot plate; CP = cold p late . )

over four drug administrat ions in these animalscould be com pared with that previously displayedby Group M-HP rats.

Results and DiscussionThe mean paw-lick latencies for the four

sessions for Group M-CAGE rats (subse-quent to the three sessions in which theyreceived the narcotic in their home cages)and the mean latencies previously demon-strated by Group M-HP rats on four ses-sions are shown in Figure 2. As is clear inFigure 2, the development of tolerance inM-CAGE rats was not at all facilitated bytheir earlier experiences wit h m orphine intheir hom e cages, i.e., the previous experiencewith the drug did not lead to any savings

in the acquisition of tolerance. Toleranceacquisition appears to depend upon a num-ber of pairings of a distinct drug adminis-tration/assessment ritual with the directeffects of that drug rather than upon merelythe fr equency of drug insults. Such pairings,according to a conditioning theory of drugtolerance, are necessary for the organism toassociate predrug cues with the physiologicaleffects of the drug and to make the compen-satory CR that functionally attenuates thedrug UCR.

EXPERIMENT 2AAs can be seen in Figure 1, by comparing

the Groups M-HP and S reaction times over


5/9

502 SHEPARD SIEGEL

2 20

MORP IN SESSIONS

F I G U R 2. Comparison of the acquisition ofmorphine-analgesia tolerance (decrease in meanpaw-lick latency after successive administrationsof the d rug) for Groups M-HP and M-CAGE ofExperiment IB. (Abbreviations: M = morphine;HP = hot plate.)

the four sessions, morphine-injected ratstend to respond increasingly like saline-in-jected control rats as they have more andmore experience with the drug. Indeed,

Group M-HP rats rapidly became so tol-erant to the analgesic effects of the drugthat they responded on the hot plate asrapidly as saline-injected animals. Accord-ing to a conditioning theory of morphinetolerance, the morphine-tolerant, short-latency response of Group M-HP rats resultsfrom a preparatory hyperalgesic CR sum-mating with the narcotic's analgesic UCR,and it should be possible to observe thisconditioned increased pain sensitivity inresponse to those cues that have been pre-dictors of systemic morphine. Experiment2A was designed to demonstrate this hyper-algesic CR directly.

MethodFollowing the usual 48-hr intersession interval,

the morphine-tolerant Group M-HP rats of Ex-per im ent 1A received a fifth session w hich wasconducted like the previous four sessions exceptth e substance injected was physiological saline

rather than morphine.The Group M-HP placebo-elicited hot plateresponse was compared with the placebo-elicitedresponses of two control groups. One was providedby Group S, which simply received a fur therplacebo session. Thus, on Session 5, Group S rats

had the same amount of experience with the in-jection procedure and assessment situation asGroup M-HP rats, but never received morphine.A second control group was included to assesswhether any unusual placebo-elicited hot platesensitivity of Group M-HP could be attributed

to residual systemic effects of previously injectedmorphine or to withdrawal from dependence uponthe drug (see Tilson, Rech, & Stolman, 1973).This group received four m orphine injections priorto the placebo test, of the same dose and at thesame time as Group M-HP, b ut always in the col-ony room. Thus, this second control group, here-inafter called Group M-CAGE:4 (n = 8), wastreated in the same manner as Group M-CAGEo f the earlier experim ent, but received four, ratherthan three morphine injections in the colonyroom. Group M-CAGE:4 received its first ex-perience with the hot plate environment andanalgesia-assessment situation when i t received aplacebo on Session 5.

Results and Discussion

The mean Session 5 paw-lick latencies,after the placebo, for Groups S, M-CAGE:4, and M-HP were, respectively, 10.3 sec,9.1 sec, and 4.4 sec. The reaction latencywas significantly shorter for Group M-HPthan for Groups S or M-CAGE:4 (bothUs = 2.5, both ps < .002), and these lattertwo control groups did not differ signifi-cantly from each other. Thus, in responseto a ritual that had been associated withmorphine administration but now not fol-lowed by the central effects of the drug,morphine-tolerant Group M-HP rats dis-played hyperalgesia. Rats with equivalentexperience with the ritual without associ-ation with the narcotic (Group S) or with thenarcotic without association with the ritual

(Group M-CAGE: 4) did not evidence such

hypersensitivity to pain.

EXPERIMENT 2B

A further experiment was conducted inan attempt to demonstrate in a within-sub-ject rather than between-subject design thathyperalgesia follows morphine administra-tion cues in morphine-tolerant rats.

Method

A ho t plate response-latency baseline was es-tablished for each of the six Group S rats by cal-culating their mean response latency for a totalo f seven consecutive sessions in which they re-ceived physiological saline prio r to pain sensitivityassessment. Longer-than-baseline response laten-


6/9

M O R P H I N E TOLERANCE AS A L E A RN E D RESPONSE 503

cies would he evidence of analgesia, and shorter-than-baseline latencies evidence of hyperalgesia.For the four sessions following baseline determi-nation, these rats were injected wi th morphine(5 mg/kg) rather than saline prior to each hotplate p lacement . By comparing th e rats' paw-lick

latencies following each morphine injection withtheir baseline, the initial analgesic response andthe development of tolerance could be evaluated.These now-morphine- to le ran t rats were left un-disturbed in their home cages for 2 wk, when theyagain received four physiological saline-hot platesessions. With the exception of this 2-wk delaybetween th e last morphine session and the firstplacebo test session, the intersession interval was48 hr .


As would

be expected

from earlier -work

(e.g., Cochin & Kornetsky, 1964), the paw-lick latency did not vary much over thecourse of the initial baseline sessions (whensubjects all received physiological salineprior to analgesia assessment). The overallmean baseline paw-lick latency was 12.9sec; it was 11.7 sec on the first baseline ses-sion and 14.5 sec on the last baseline session,there being no significant trend across base-line sessions.

Figure 3 presents the mean percent changein paw-lick latency from baseline levelsfollowing each of the four morphine injec-tions and, two wk later, four physiologicalsaline injections. As can be seen in Figure3, the analgesic effect of the initial injectionof morphine is clear; paw-lick latency al-most doubled from baseline levels, i.e., in-creased by 100%. Reaction time decreasedfollowing subsequent morphine injectionsuntil by the fourth injection of the narcoticthese rats were responding on the hot platewith about the same latency as their pre-drug baseline levels. When tested on the hotplate 2 wk later (after any residual systemiceffects of the drug had dissipated; Tilsonet al., 1973) after a placebo, these morphine-tolerant animals were clearly hypersensitiveto the heat. Their reaction times were al-most 40 % below their baseline levels, a sig-nificantly shorter response latency (Wilcoxon

matched-pairs signed-rank test, T = 1.0,p < .025). As is also clear in Figure 3, asthese subjects were successively presentedwith the placebo test sessions the magnitudeof the hyperalgesic response tended to de-

SESSION

F I G U R 3. Percent change from baseline paw-lick latency after each o f four morphine injectionsand, 2 w k later, after four physiological saline in -jections (Experiment 2B).

crease, i.e., their response latency returnedto baseline levels. It appears that as the drugadministration procedure is successivelypresented without the systemic effects of thedrug, the hyperalgesic response in morphinetolerant rats is subject to extinction, sug-gesting that it is indeed a CR. Inasmuch asit is proposed that it is this hyperalgesic CRthat is responsible for observed analgesiatolerance, extinction should be an effectiveprocedure for eliminating tolerance.

EXPERIMENT 3

If in the morphine-tolerant animal thoseenvironmental procedures associated with

the central effects of the drug elicit a com-pensatory CR, presenting these environ-mental procedures unaccompanied by thecentral effects of the narcotic should serveto extinguish these learned responses andmorphine tolerance. In other words, placebotest sessions should constitute an effectiveprocedure for attenuating established toler-ance. This prediction of a conditioningtheory of tolerance was assessed in this ex-periment.

MethodTwo groups, each containing six experim entally

naive rats of the sam e sex, strain, and age as thoseused in the previous experiments, were each given


7/9

504 SHEPARD SIEGEL

—• GROUP M P M

GROUP M R E 5 T M

MORP IN I N J E C T I O N S

F I G U R 4. Mean paw-lick latencies after eacho f six dai ly m orphine injections for groups receiv-ing either nine placebo sessions (Group M-P-M)or a 9-day rest interval (Group M-REST-M) in -terpolated between morphine Sessions 3 and 4(Experiment 3).

a total of six m orphine-analgesia assessment ses-sions, using the same procedures as describedearlier. The interval between sessions was 24 hrwith th e exception of the pr otracted interval be-tween Sessions 3 and 4, which was 9 days. The twogroups differed only with respect to their treat-ment between these third and fourth sessions.One group was simply left undisturbed in its homecage (Group M-REST-M). The second g roup r e-ceived daily placebo test sessions, i.e., they weretreated in the same manner as on m orphine ses-sions except the substance injected was physio-logical saline rather than morphine (Group M-P-M ).


The mean paw-lick latencies of bothgroups on each occasion that they received

morphine are shown in Figure 4. Both groupsevidenced tolerance to the analgesic effectsof morphine over the course of the threeinitial adm inistrations of the drug. Grou pM-REST-M continued to evidence mor-phine-tolerant, short-latency responses w henagain tested with the narcotic after the de-lay interval, as would be expected from pre-vious work demonstrating that morphinetolerance dissipates little simply with thepassage of time (Cochin & Kornetsky,1964). However , whe n tested with m orphineafter the same delay interval, Group M-P-M evidenced a nontolerant, long-latencyresponse. There was no overlap in the Ses-

sion 4 paw-lick latencies of Groups M-REST-M and M-P-M U = 0, p = .001).As can be seen in Figure 4, Group M-P-Mrats had to be retolerated to morphine,despite the fact that they had suffered the

systemic effects of the narcotic equally asoften as Group M-REST-M rats.The finding that mere repeated presenta-

tions of a drug administration procedureunaccompanied by the central effects of thedrug effectively obliterated morphine-anal-gesia tolerance is a unique prediction of aconditioning theory of tolerance, and is notexplicable by theories of tolerance that donot emphasize the role of drug-associatedenvironmental cues in the development oftolerance.

GENERAL DISCUSSION

It has been previously suggested thatlearning can influence responsivity to drugs(see Thompson & Picking, 1971) and that . . . a drug-test interaction occurs wi thmorphine and can play a role in the develop-ment of tolerance to the analgesic effect ofthis drug (Adams e t al., 1969, p. 251). Thepresent experiments were designed to assessa specific Pavlovian conditioning interpreta-tion of the phenomenon of morphine toler-ance. Based on earlier reports that the CRto a variety of pharmacological agents iscompensatory in nature, it seems reasonablethat the direct, unconditioned analgesiceffect of morphine is normally modulatedby a m orphine-anticipatory hyperalgesicCR, the net result being reflected by thedevelopm ent of m orphine tolerance. This

conditioning analysis of morphine toleranceis supported by several findings: (a) It isnecessary to have a consistent set of en-vironmental cues reliably predicting thesystemic effects of morphine if rapid toler-ance is to be observed (Adams e t al., 1969;Kay an et al., 1969; Experiment 1A of thepresent report); (b ) experience with mor-phine in one environm ent does not facilitatethe acquisition of morphine tolerance inanother environment (Experiment IB);(c) the compensatory hyperalgesic CR maybe directly observed in morphine-tolerantanimals when they are confronted by thedrug administration ritual not followed by


8/9

MORPHINE TOLERANCE AS A LEARNED RESPONSE 505

the central effects of the drug (Experiments2A and 2B), this hyperalgesic CR beingsubject to experimental extinction (Experi-ment 2B); and (d) mere presentation of

those environmental cues previously as-sociated with the narcotic, when presentedin conjunction with a placebo, is an effectiveprocedure for extinguishing established mor-phine tolerance (Experiment 3). The con-clusions concerning the mechanism of mor-phine tolerance in the rat are, of course,limited to the relatively small dose of thedrug (5 mg/kg) and to the analgesia-evalua-tion situation used in these experiments(although the hot plate procedure is per-haps the most commonly used of the simpleassessment techniques for pharmacologicallyinduced analgesia; see Evans, 1964).

The present findings concerning the im-portance of the interaction between condi-tioned and unconditioned responses incontributing to the observed effect of a drugparallel Pavlov's (1910) discussion of thesignificance of his original psychic secre-tion observations, i.e., that digestive re-

sponses in anticipation of

feeding make a

significant contribution to normally ob-served patterns of digestive functioning.Subsequent research has demonstrated theimportance of conditional responses in thenormal and pathological functioning of avariety of physiological systems in manyspecies including humans (e.g., Adam, 1967;Bykov, 1959). This work on the interactionof learning and physiological processes,conducted mostly by Eastern European and

Soviet physiologists, is the foundation of a synthetic physiology , .. . a science ofthe course of vital processes in an integralorganism during its various natural relationswith the surrounding medium (Bykov, 1960,p. 25; emphasis added). It would appearthat inasmuch as drug administration isalmost invariably predicted by a set of cues(the administration procedure, or ritual), theresponse of an integral organism to a drug

can be best understood as a combination ofthe direct reflexive effects of the drug as itacts on central receptor sites and the effectsconditioned to the drug administration pro-cedure.

R F R N S

Adam, G. Inleroception and behaviour. Budapest:Publishing House of the Hungarian Academy ofSciences, 1967.

Adam s, W. H., Yeh, S. Y., Woods, L. A ., & Mitch-ell, C. L. Drug-test interaction as a fac tor in thedevelopment of tolerance to the analgesic effectof morphine. Journal of Pharmacology and Ex-perimental Therapeutics, 1969, 168, 251-257.

Bykov, K. M. The cerebral cortex and the internalorgans. Moscow: Foreign Languages PublishingHouse, 1959.

Bykov, K. M. (Ed.), Text-book of physiology.Moscow: Foreign Languages Publishing House,1960.

Cochin, J. Possible mechanisms in development oftolerance. Federation Proceedings, 1970, 29, 19 -27.

Cochin, J., & Kornetsky, C. Development andloss of tolerance to morphine in the rat a f tersingle an d multiple injections. Journal of Phar-macology and Experimental Therapeutics, 1964,145, HO.

Cochin, J. , & Kornetsky, C. Factors in blood ofmorphine-tolerant animals that attenuate orenhance effects of morphine in nontolerant ani-mals. Proceedings of the Association for Researchin Nervous an d Mental Disease, 1968, 46 , 268-279.

Collier, H. O. J. A general theory of the genesis ofdrug dependence by induction of receptors.Nature, 1965, 20B, 181-182.

Deutsch, R . Con di t ioned hypoglycemia: A mecha-nism for saccharin-induced sensitivity to in-sulin in the rat. Journal of Comparative andPhysiological Psychology, 1974, 86 , 350-358.

Eddy, N . B. , & Leimbach, D . Synthetic anal-gesics: II. dithienylbutenyl- and dithienylbutyl-amines. Journal of Pharmacology and Experi-mental Therapeutics, 1953, 107, 385-393.

Evans, W. 0. A crit ical review of some new meth-ods in animal analgesiometry. Journal of Clin-ical Pharmacology and New Drugs, 1964, 4 , 179-187.

Gantt, W . H. Pharmacological agents in study ofhigher nervous activity. Diseases of the NervousSystem, 1957, 18 , 339-341.

Hull, C. L. Learning: II. The fac tor of the condi-tioned reflex. In C. M urchinson (Ed.), A hand-book of general experimental psychology. Wor-cester, Mass.: Clark University Press, 1934.

J6hannesson, T., & Woods, L. A. Analgesic actionand brain and plasma levels of morphine andcodeine in morphine tolerant, codeine tolerantand non-tolerant rats. Acta Pharmacologica etToxicologica, 1964, 21 , 381-396.

Kayan, S., Woods, L . A ., Mitchell, C. L. Ex-

perience as a fac tor in the development of toler-ance to the analgesic effect of morphine. Euro-pean Journal of Pharma cology, 1969, 6, 333-339.

Lang, W. J., Brown, M. L., Gershon, S., & Korol,B . Classical and physiologic adaptive condi-tioned responses to anticholinergic drugs in


9/9

506 SHEPARD SIEGEL

conscious dogs. International Journal of Neuro-pharmacology, 1966, 5, 311-315.

Lichko, A. E. Conditioned reflex hypoglycaemiain m a n. Pavlov Journal of Higher Nervous Activ-ity, 1959, 9, 731-737.

Loucks, R. B . Hum oral condi t ioning in m amm als.Journal of Psychology, 1937, 4, 295-307.

Mi tyushov, M. I. Uslovnorleflektornaya inkret-siya insulina (The conditional-reflex incret ionof insul in) . Zhurnal Vysshei Nervnoi Deiatel(Journal of Higher Nervous Activity), 1954, 4,206-212. (Read in t ranslat ion prepared by L. J.Shein, Department of Russian, McM asterUniv. )

Mute, S . J., & Woods, L. A. Distribution of N -C14-methyl labeled m orphine: I. In central nervoussystem of nontolerant and tolerant dogs. Jour-nal of Pharmacology and Exp erimental Thera-peutics, 1969, 188, 251-257.

Pavlov, I. P. The work of the digestive glands (W .H. Thom pson, trans.). Charles Griffin and Com-pany: London, 1910.

Pavlov, I. P. Conditioned reflexes (G. V. Anrep,trans.). London: Oxford University Press,1927.

Russek, M ., & Pifia, S. Conditioning of adrenalinanorexia. Nature, 1962,193,1296-1297.

Schmidt , C. F., & Livingston, A. E. The relationof dosage to the development of tolerance to

morphine in dogs. Journal of Pharmacology a ndExperimental Therapeutics, 1933, 47 , 443-471.

Siegel, S. Condi t ioning of insulin-induced gly-cemia. Journal of C omparative and PhysiologicalPsychology, 1972, 78, 233-241.

Siegel, S. Cond itioning insulin effects. Journal ofComparative and Physiological Psychology, 1975,89, 189-199.

Siegel, S., Net t le ton , N . Condi t ioning o f insu-l in-induced hyperphagia. Journal of Comparativeand Physiological Psychology, 1970, 7 , 390-393.

Subkov, A. A ., & Zilov, G. N. The role of condi-tioned reflex adaptation in the origin of hyper-ergic reactions. Bulletin de Biologie et de Mede-cine Experimental^, 1937, 4, 294-296.

Thompson, T., & Pickens, R. (Eds.). Stimulusproperties of drugs. N ew York : Appleton-Cen-tury-Crofts , 1971.

Tilson, H . A., Rech, R. H., & Stolm an, S. Hyper-algesia during withdrawal as a means of measur-ing the degree of dependence in morphine de-pendent rats. Psychopharmacologia, 1973, 28,287-300.

Wikler, A . Rece nt progress in research o n the neu-rophysiologic basis of morphine addict ion.American Journal of Psychiatry, 1948, 105, 329-338.

(Received July 5, 1974)

Documents

[Opioides]Evidence From Rats That Morphine Tolerance is a Learned Responde - Siegel 1975