C H A P T E R

28

Animal Models of Drug Abuse:Place and Taste Conditioning

Catherine M. DavisDepartment of Psychiatry and Behavioral Sciences, Division of Behavioral Biology, Johns Hopkins

University School of Medicine, Baltimore, MD, USA

A

h

O U T L I N E

What is Drug Addiction and why Shouldwe Study it?

681

Reward and Reinforcement

682

Aversive Drug Effects

683

The Place-Conditioning Procedure

684

nima

ttp://

What are Conditioned Place Preferences? 6

84 A Brief History of Place Conditioning 6 85 Advantages of the Place-Conditioning Procedure 6 87 Place Conditioning: Experimental Protocol

and Conditioning Apparatus 6

89

68l Models for the Study of Human Disease

dx.doi.org/10.1016/B978-0-12-415894-8.00028-2

The Flavor-Conditioning Procedure

693

1

What are Conditioned Taste Aversions? 6

93 A Brief History of Taste Aversion Learning 6 94 Advantages of the CTA Procedure 6 96 Conditioned Taste Aversion: Experimental

Protocol and Conditioning Apparatus 6

99

Conclusion

703

Acknowledgments

703

References

703

WHAT IS DRUG ADDICTION AND WHYSHOULD WE STUDY IT?

Drug addiction is defined as a chronically relapsingdisorder, characterized by (1) the compulsion to seekout and take the drug, (2) loss of control in limitingdrug intake, and (3) the emergence of a negativeemotional state, reflecting a withdrawal syndromewhen drug access is prevented.1 Addictions to licit(e.g. alcohol and nicotine) and illicit (e.g. marijuana,cocaine, and heroin) substances are pervasive societalproblems resulting in severe emotional, financial, andphysical costs to the addicted individual, the indi-vidual’s family, and the community.2 The economiccosts (e.g. healthcare costs, crime-related costs, andloss of productivity) of drug dependence in the UnitedStates exceed $600 billion annually; interestingly,

approximately two-thirds of these costs are associatedwith use of the licit drugs nicotine and alcohol (citedin Ref. 2). Substance use and abuse are also associatedwith other public health issues, including the spreadof infectious diseases (e.g. HIV and hepatitis C) andimpaired driving.3,4 Despite the myriad costs to individ-uals and society, justifying the need for animal models ofsubstance abuse and dependence can be a difficult taskfor researchers given that drug addiction is perceivedby a large portion of the general public to result frommoral weakness, a lack of willpower, and/or problemswith self-discipline.2,5 For example, a federally fundedstudy examining cocaine’s influence on risky sexualbehavior in quail was recently included in a senator’s“Wastebook,” a document detailing how the US federalgovernment “wastes” tax dollars on research on drugaddiction.6

Copyright � 2013 Elsevier Inc. All rights reserved.

28. ANIMAL MODELS OF DRUG ABUSE: CPP AND CTA682

In the scientific community, however, substance abuseand dependence are viewed as a disease of the brain thatwarrants the same type of investments in research andtreatment strategies as those given to diseases likeautism, schizophrenia, andAlzheimer’s and Parkinson’sdiseases.5 Furthermore, drug abuse researchers employthe same strategies as researchers investigating theseother diseases: animalmodels are developed to representsomecharacteristic of the disorder and then studied,withthe understanding that every detail of the disorder ismost often not completely reproducible in a laboratoryanimal, but that an understanding the disease processandwhat factors contribute to this process can be gained.For example, “schizophrenic” or “autistic” rats and micedo not exist,7 although there are rats andmice that exhibitbehaviors characteristic of schizophrenia, autism, oreven obsessive-compulsive disorder8,9 these animalmodels are used to further our understanding of themolecular, cellular, and genetic components of a specificdisorder and how these deficiencies translate into quan-tifiable behavioral symptoms.

Similar to many of these other human disease states,drug addiction is a uniquely human phenomenon;thus, any attempt to study it in the laboratory inevitablyconstrains its reproducibility. Investigators must keepthis in mind when designing and employing animalmodels of drug addiction. Some models have verygood face validity, such as the drug self-administrationprocedure in which an animal emits a response (e.g.lever press) to receive a drug injection reinforcer. Elegantexperiments using the self-administration procedure tostudy in rats the diagnostic criteria of addiction (out-lined in the Diagnostic and Statistical Manual of MentalDisorders, 4th ed., or DSM-IV-TR) have been reportedand arguably provide new, more “realistic” approachesto study the neurobiology of addiction.10–13 For othermethods, the connection to human drug abusers is notas evident, but the procedures nonetheless allowresearchers to examine some aspect of addiction that isimportant to its overall etiology. Given these issues,the specific question under investigation undoubtedlyinfluences the procedure employed; great care must betaken to understand each procedure’s strengths andlimitations and how these factors affect the interpreta-tion of the results of the study. Even with these caveats,preclinical animal models have been instrumental infurthering our understanding of the characteristics ofaddiction, including drug taking, physical dependence,tolerance, drug craving, and the factors that mightincrease the likelihood of relapse. These models haveadded greatly to our understanding of drug addiction;however, there is still much to learn and numerousreasons to continue refining our current models, in addi-tion to developing new models as various techniquesbecome even more sophisticated.

VIII. BEH

In the current chapter, two popular methods used inthe drug abuse field, place conditioning (primarilyfocused on conditioned place preference, or CPP) andconditioned taste aversion (CTA), will be discussed.Both procedures have been used extensively to studythe rewarding and aversive effects associated with drugexposure. With each of these procedures comes a ratherlarge literature base that includes, but is not limited to,variations in procedural details, equipment, and studydesign; debates on data analysis and interpretation; andexcellent reviews commenting on the strengths, weak-nesses, and controversies surrounding each procedure.Although many of these factors will be discussed in thecurrent chapter, readers are encouraged to accessthese resources for additional information.14–23

REWARD AND REINFORCEMENT

When beginning any discussion of animal models ofdrug addiction, the distinction between reward and rein-forcement should be made and include what these termsmean in the context of drug abuse research. Drugs ofabuse are considered reinforcers and, in a Skinnerianview, are thus defined as events that follow an instru-mental response and influence the probability of thatresponse occurring in the future. Regardless of thevalence (i.e. positive or negative), reinforcement willincrease the probability of behavioral output in anorganism, in order for the organism to earn an appetitivereinforcer or to avoid experiencing an aversive one. Inter-estingly at different stages of the addictionprocess, drugsof abuse are thought to differently reinforce drug-takingbehavior, with positive reinforcement occurring in theearly stages of drug use and negative reinforcementinvolved predominantly following more chronic drugtaking. Sanchis-Sugura and Spanagel24 argues that, inthis context, drugs of abuse serve as reinforcers in threedistinct ways: (1) Drugs serve as primary motivators(i.e. positive reinforcement) where behavioral contin-gencies are positively reinforced; in other words, theyserve as appetitive stimuli that increase the emission ofthe instrumental response; stimuli associated with drugintake in these situations can also become “conditionedreinforcers,” a process by which they acquire incentive-motivational properties (which are considered to becomepathological in addiction), and when these stimuliare presented in the absence of the drug, they elicit“wanting” or craving responses (for a review, seeRef. 25; see also Refs 26–28). (2) Drugs acting as rein-forcers increase the associative strength of specificstimulus–response contingencies, which results in anincreased probability of the instrumental responseoccurring; reinforcers thus serve to enhance informationstorage about the situations in which they occur,

AVIOR

FIGURE 28.1 Schematic depicting the rewarding effects (green

line), aversive effects (red line), and self-administration (blue line)

of a drug of abuse and how the balance of these two effects impacts

drug intake. As the dose increases, the rewarding effects and self-administration increase as well; in this instance, the rewardingeffects are of a greater magnitude than the aversive effects. As theaversive effects increase, self-administration of the drug is negativelyimpacted and decreases. Importantly, a change in the rewardingeffects of the drug is not necessary for the aversive effects to nega-tively impact drug intake. Given that self-administration is the balanceof these effects, a drug can have robust rewarding effects, but self-administration could decrease because the aversive effects are toosevere. Reprinted from Ref. 32 with permission from the New York Academyof the Sciences.

AVERSIVE DRUG EFFECTS 683

a process that can bias response choiceswhen those situ-ations occur in the future. (3) Acting as a negative rein-forcer, drugs can reduce specific needs or drives; thisprocess is most important after prolonged drug taking,when abstinence can result in withdrawal symptomsthat are alleviated following drug intake.24,29

Although negative reinforcement by drugs of abuse isan important aspect of the addiction process,29 drugshave traditionally been viewed as primary appetitivestimuli in the drug abuse field, for which respondingis thought to be due to positive reinforcement. Anothercommon term used to describe positive reinforcementin this regard is reward. This term is often mistakenlyused as a synonym for reinforcement, without anydistinction between the valences used to distinguishthe type of reinforcement. When describing positivereinforcement, rewards are those appetitive reinforcersthat increase the probability of emission of the instru-mental response. Reward is also used to describehedonia, which is defined as a hypothetical internal stateof pleasure associated with acquiring, using, and/orconsuming appetitive stimuli.24,30 Given these defi-nitions of reward, it is clear that this term can be usedto describe positive reinforcement contingencies orappetitive stimuli, but is distinct from the broaderconcept of reinforcement, which encompasses mecha-nisms that are reward independent.

AVERSIVE DRUG EFFECTS

Although drugs of abuse are typically viewed asappetitive stimuli, it is argued that these compoundshave multiple behavioral effects, including aversiveproperties that might impact the overall acceptabilityof a drug and its subsequent self-administration.32,33

For example, Riley33 argues that the “overall hedoniceffect of a drug” and the likelihood of its use and abuseare dependent on the balance of a drug’s rewardingand aversive effects. When examining patterns of self-administration, an inverted U-shaped function typicallyemerges, such that drug intake increases with dose untilsome asymptotic level is achieved; as the dose continuesto increase, self-administration begins to decrease(Fig. 28.1).

The rewarding effects are thought to increase withdose and most likely reach some maximal level wherefewer self-injections are needed to achieve the desiredeffect, in addition to possible receptor saturation.34

During this time, a drug’s aversive effects are alsoincreasing and the overall balance of these effectspresumably results in the observed pattern of drugintake.33 Describing use and abuse as a balance of theseaffective properties suggests that reward and aversionare on a continuumwhere a drug with robust rewarding

VIII. BEH

effects must also be weakly aversive (or the opposite, inwhich an aversive drug must be weakly rewarding). It ismore likely, however, that each of these effects is medi-ated independently, such that a drug has some levelof rewarding and/or aversive effects and each effectfunctions to impact overall drug acceptability (datadescribed in this chapter argue for these independenteffects). The perceived balance of these effects thendictates drug intake, where a drug perceived as “moreaversive” is self-administered less than a drug that isperceived as “more rewarding.” Importantly, these aver-sive drug effects are considered distinct from those thatnegatively reinforce drug taking following withdrawal;aversive drug effects in this context are experiencedfollowing acute administration and could serve a protec-tive function by limiting (or eliminating) future intake.Thus, aversive drug effects could decrease the likelihoodthat a drug will be abused. Interestingly, the two proce-dures reviewed in this chapter have been used indepen-dently and in combination to demonstrate that a drug’srewarding and aversive effects are measurable simulta-neously in individual subjects following a single drugadministration. Furthermore, these procedures haveshown that not all manipulations and subject variables,including drug history, strain and other genetic differ-ences, sex, and stress, alter these effects in the samemanner (e.g. decreased aversion and increased reward),which makes these procedures useful for continued

AVIOR

28. ANIMAL MODELS OF DRUG ABUSE: CPP AND CTA684

investigations focused on understanding the factors thatimpact drug use and addiction.32,33,35

FIGURE 28.2 Various place-conditioning apparatuses have been

reported in the literature, including two- and three-compartment

designs with additional alleys and/or goal boxes. Sources: (a) Refs

36,38,135; (b) Refs 61,62; (c) Ref. 136; (d) Refs 17,43,52,69, and (e) Refs 39.

Some examples have been redrawn from Ref. 55

THE PLACE-CONDITIONINGPROCEDURE

What are Conditioned Place Preferences?

Place conditioning is a simple procedure used toassess the positive (rewarding) or negative (aversive)motivational effects of exposure to various stimuli,including drugs of abuse, food, sexual behavior, novelenvironments, or foot shock.36 Conditioned place prefer-ence (CPP) is a procedure commonly used to assess therewarding effects of a stimulus by measuring increasedapproach and contact behaviors, whereas conditionedplace aversion (CPA) measures increased avoidanceand/or escapebehaviors indicative of the aversive effectsof a specific stimulus. Although these procedures aresimple and effective methods for assessing the motiva-tional effects of various stimuli in different species,the CPP procedure is most commonly used in rodents(rats andmice) as an indirect assessment of the rewardingeffects of various classes of drugs, especially drugs ofabuse. Recently, conditioning of an amphetamine-induced CPP was reported in human subjects; theseresults further support the use ofCPP in animals to assessthe rewarding effects of abused drugs.37 In a standardCPPexperiment, subjects undergo conditioning sessions,in which a drug injection is pairedwith specific cues (e.g.tactile, visual, and olfactory) while the subject is confinedto one compartment of a conditioning apparatus, typi-cally a shuttle boxlike apparatus (Fig. 28.2); subjectsexperience different cues paired with a saline injectionwhile confined to the opposite compartment. Afterseveral drug–environment and saline–environmentpairings (i.e. conditioning trials), subjects are tested forthe expression of a CPP during a drug-free test sessionin which subjects are given access to the entire apparatusand can move freely between the conditioning compart-ments for a specified time period (typically 15–30 min).Subjects that spend significantly more time in thecompartment paired with drug administration are saidto express a conditioned preference. In contrast, avoid-ance of the drug-paired compartment, by spendingmore time in the saline-paired side, is often interpretedas a CPA.17,36

This procedure is based on principles of Pavlovianclassical conditioning, where the effect of the drugadministered serves as the unconditioned stimulus(US) and is paired with the initially neutral environ-mental cues of the shuttle box compartmentdafterseveral pairings as stated in this section, these neutralenvironmental cues become associated with the

VIII. BEH

rewarding effects of the drug and acquire secondarymotivational properties. The drug-paired compartmentthen becomes the positive conditioned stimulus (CSþ)that evokes a conditioned motivational response andelicits approach and contact behavior during the subse-quent drug-free test session. The other shuttle boxcompartment, which is usually paired with salineinjections (or no injection at all), contains different cuesfrom those of the CSþ. These environmental cuesremain neutral in that they do not become associatedwith the drug’s effects (US) and thus become the nega-tive conditioned stimulus (CS�). The fact that an animalwill approach and/or contact stimuli that have beenpreviously paired with rewarding drug effects is funda-mental to the CPP procedure. When an animal receivesthese drug–environment pairings and subsequentlyapproaches and spends time in contact with the drug-paired side of the conditioning apparatus or shuttlebox, it is inferred that the drug administered hadrewarding effects and that these effects have becomeassociated with the specific compartment in which theconditioning trials occurred. In a similar manner, avoid-ance and/or escape behaviors elicited by a CSþ paired

AVIOR

THE PLACE-CONDITIONING PROCEDURE 685

with an aversive stimulus are essential to the CPA proce-dure as well. Although most drugs of abuse result inconditioned preferences for the CSþ, these same drugscan result in a CPA in different strains of rodents orfollowing variations in the route of drug administrationor temporal parameters used when pairing the drug’seffects with the environmental cues available in theconditioning apparatus (discussed further in thischapter). An understanding of the procedure and care-ful selection of the parameters used will help to reducemany of these confounding factors and provide for aneasier, more straightforward interpretation of theresults.

A Brief History of Place Conditioning

Garcia et al.38 first used the place-conditioning proce-dure to assess the motivational effects of whole-bodygamma or X-ray irradiation in male rats. A two-compartment apparatus was employed, which includedone compartment that was painted black and containeda grid floor; the opposite compartment was paintedwhite and contained a wire mesh floor. Rats experiencedfour irradiation-compartment pairings in either theblack or the white compartment; the radiation-pairedcompartment (or sham irradiation) was divided acrosssubjects such that baseline preference for the blackcompartment was used to divide rats into groups withcomparable baselines (i.e. time spent in the blackcompartment). Half of the rats were irradiated in theblack compartment, and half received the exposures inthe white compartment. Following conditioning, allanimals displayed significant decreases in the amountof time spent in the compartment paired with irradia-tion, indicating a conditioned place aversion. Althoughthe groups conditioned with gamma irradiation in theblack compartment initially spent more time thereduring the pretest session compared to rats conditionedin the white compartment, there were no significantdifferences in the radiation-induced CPA between thegroups.

Although this initial study was assessing aversioninduced by irradiation, Garcia et al.38 employed severaltechniques that are currently used in place-conditioningstudies with drugs of abuse. For instance, the baselinepreference for each conditioning compartment wasdetermined on the day prior to the first conditioningsession during a pretest session, which consisted ofa test period (150 min) where the rats could freelyexplore both the black and white compartments. Thisstudy was one of the first to use the pretest session inorder to determine each subject’s baseline preferencefor the compartment subsequently paired with the US(in this case, irradiation). In this way, these investigatorscompared the shift in the amount of time spent in this

VIII. BEH

compartment during the pretest to the amount of timespent in the same compartment following the condi-tioning sessions for each subject. Although the ratsconditioned with irradiation in the black or whitecompartment were treated as separate experimentalgroups, a procedure similar to this is commonly usedin place conditioning with drugs of abuse to specificallyassign the CS–US pairings across the experimentalgroups (see “Place Conditioning: Experimental Protocoland Conditioning Apparatus”). Garcia et al.38 condi-tioned some rats to their naturally preferred sides (i.e.pairings of groups receiving gamma irradiation andblack compartments) and others to their naturally lesspreferred side (i.e. pairings of groups receiving gammairradiation and white compartments); the former proce-dure is more common in place aversion work in order toavoid a floor effect, that is, an aversive stimulus is pairedwith the naturally preferred side of the conditioningapparatus because subjects are most likely to decreasethe amount of time spent in the compartment pairedwith an aversive stimulus. Both groups conditionedwith X-ray irradiation paired with the white or blackcompartment were conditioned to their naturallypreferred side (e.g. the group of rats receivingirradiation–black compartment pairings preferred theblack compartment during the pretest, and the groupof rats receiving irradiation–white compartment pair-ings preferred the white compartment during thepretest). Interestingly, the irradiation-induced aversionswere similar in strength regardless of what compart-ment the animals received the exposures. This is notalways the case when assessing drugs of abuse in theCPP design, and the choice of the drug-paired compart-ment is an important one.15

Whereas Garcia et al. were interested in examining theeffects of irradiation in the place-conditioningprocedure,Beach39 published the first paper assessing a drug-induced CPP in rats. Beach hypothesized that rats wouldassociate a specific place with the drug’s effects if theywere confined to this place after each drug administrationand remained in this place long enough to adequatelyexperience the drug’s effects. This work was to demon-strate that rats could learn to associate a specific environ-ment with a drug’s effects and would subsequentlyprefer this environment when tested while “needing”morphine (tested 2–4 h following saline administration)or while morphine “sated” (tested 2–4 h after beinginjected with morphine).39 Rats were chosen as subjectsfor this study to determine if this species could learndrug effect–environment associations in a mannersimilar to the morphine–environment associationsdisplayed by morphine-dependent chimpanzees in anearlier report by Spragg.40 Beach successfully demon-strated that morphine-dependent and nondependentrats both learned to associate the effects of morphine

AVIOR

28. ANIMAL MODELS OF DRUG ABUSE: CPP AND CTA686

administration with a specific environment and subse-quently preferred this environment to the one in whichthe rats received saline injections (or no injection at all).

This studynot only is the first report of a drug-inducedCPP, but also is an example of simple changes to theplace-conditioning procedure and apparatus from thosereported by Garcia and colleagues. For example, Beachemployed a Y-shaped discrimination box in which twodistinct arms connected a starting box to two separate“goal boxes” (i.e. conditioning compartments).

The start box, left goal box, and right goal box wereeach painted a different color, and each contained addi-tional visual and tactile cues. The left goal box had graywalls with square black insets and a flyscreen-coveredfloor. The right goal box had black walls with circularwhite insets and a wood floor. Additionally, theY-shaped discrimination box had four doors that couldbeused to close off the start box or eitherof the goal boxes.In 16 pretest choice sessions, rats were placed in the startbox and allowed to access either of the goal boxes. Oncethe rats reached a goal box, a choice for that goal boxwas recorded; after 5 s in the chosen goal box, rats wereremoved to prevent them from moving back down thearm and into the start box and/or the other goal box.From these pretest choice sessions, preferred and non-preferred goal boxes were determined for each rat. Ratswere subsequently conditioned to their nonpreferredgoal box (i.e. rats were injected with morphine, placedin the start box, and run to the goal box that was chosenless during the pretest sessions). In this way, some of therats experienced morphine in the gray goal box andothers experienced morphine in the black goal box, butall rats were conditioned to the nonpreferred goal box.This study can be considered the first use of the “biased”design in place conditioning with drugs of abuse, a tech-nique where subjects are conditioned to their initiallynonpreferred side. A significant change in the amountof time that subjects spend in the nonpreferred sidefollowing conditioning with morphine, for example,indicates a preference for the drug-paired side. This isin contrast to the “unbiased” design in which the drug-paired compartment is counterbalanced across subjectsregardless of their initial compartment preference(more details on biased vs unbiased designs arediscussed further in this chapter).

Following Beach’s report of morphine CPP inmorphine-dependent and nondependent rats, Kumar134

reported conditioned preferences for morphine-pairedenvironments in morphine-dependent rats testedduring withdrawal (i.e. 48 h after the last morphineinjection) using a two-compartment apparatus. Further-more, two additional studies included interestingchanges to the CPP procedure that have since receivedconsiderable attention in the CPP literature.41,42 Morespecifically, Rossi and Reid reported “acquisition” of a

VIII. BEH

morphine CPP by conditioning and testing rats over a4-day cycle where rats received morphine–environmentor saline–environment pairings over 3 days and werethen tested for CPP on the fourth day of the cycle;four conditioning cycles were completed, whichallowed for analysis of the CPP after additionalmorphine–environment pairings. Although initial sidepreferences influenced the results, morphine did condi-tion a significant CPP; however, a clear increase in theconditioned preference over test trials was not evident(for a more recent study of morphine CPP acquisition,see Ref. 43). In the study by Reicher and Holman,41 CPPprocedures were combined with the CTA procedureto simultaneously test the rewarding and aversiveeffects of amphetamine administration. In this combinedCTA–CPP procedure, Reicher and Holman reportedamphetamine-induced taste aversions and place prefer-ences conditioned in rats that consumed a distinctivelyflavored solution in one shuttle box compartment anda differently flavored solution on the other side.44,45

Since these initial studies, the use of the place-conditioning procedure has become widespread in thefield of drug abuse research and has amassed an exten-sive literature containing the use of different species,classes of drug, procedures, and methodologies. Forexample, drug-induced place conditioning has beenreported in rats and mice, hamsters, birds, zebrafish,nonhuman primates, and humans. These studies havealso demonstrated that most drugs serving as reinforcersin self-administration procedures also condition placepreferences, including cocaine, amphetamine, nicotine,morphine, and ethanol. Given the variation in place-conditioning procedures in the early studies mentionedin this chapter, it is not surprising that variations in theplace-conditioning procedure, experimental design,and conditioning apparatus and environmental cuesemployed have evolved considerably since these earlystudies. Given these variations, no single method forconducting place-conditioning studies exists. Nonethe-less, generalities among most studies are apparent,but substantial variability still exists in these itemsthroughout the place-conditioning literature. Whilemany of these variations impact the interpretation ofthe results of place-conditioning studies, others havedemonstrated interesting interactions between the routeand timing of drug administration, strain of animalsemployed, and strength of preference or aversion condi-tioning.15,46 For example, intraperitoneal (IP) or intrave-nous (IV) injection of ethanol in mice results in a CPP,whereas intragastric ethanol administration conditionsa place preference when a 5 min delay occurs followingdrug administration and prior to placement into theethanol-paired compartment. IP or intragastric ethanoladministration interestingly results in place aversionswhen exposure to the conditioning compartment

AVIOR

FIGURE 28.3 Illustration of a three-compartment apparatus

used for place-conditioning studies with rats. Each individualconditioning compartment measures 30 cm� 30 cm� 39 cm, and thecenter compartment measures 10 cm� 30 cm� 39 cm (total size:70 cm� 30 cm� 39 cm). The “white” compartment consists of a black-and-white striped floor and solid white walls (in addition toa textured floor surface; see Ref. 43). The “black” compartmentcontains a solid-colored floor and walls (in addition to a smooth floorsurface). During the pretest or the posttest, rats have free access to allthree compartments. During the conditioning trials, however, guillo-tine doors are inserted at the border of the white and centercompartments and at the border of the black and center compart-ments. Picture courtesy of Peter G. Roma.

THE PLACE-CONDITIONING PROCEDURE 687

(i.e. CSþ) occurs immediately prior to or immediatelyfollowing ethanol administration, respectively.46

Advantages of the Place-ConditioningProcedure

Place conditioning is a methodologically simple andrelatively inexpensive procedure: Place conditioning isa popular method in the drug abuse field because itprovides numerous advantages compared to othermethods used to assess the conditioned motivationaleffects of drugs. First, the procedure is methodologicallysimple in terms of (1) experimenter training and(2) expensiveness of the necessary experimental equip-ment. More specifically, extensive training in survivalsurgery, jugular catheter implantation, and the properanesthesia and analgesia procedures needed for IVself-administration protocols are not required for ageneral place-conditioning study. Indeed, IP or subcuta-neous (SC) drug injections are the most widely usedmethods to administer drugs in the place-conditioningprocedure, although oral gavage and intragastriccannulaehave beenused for ethanol administrations.46,47

While IV drug administration via indwelling jugularcatheters is not a requirement for a standard place-conditioning study, this drug administration techniquehas been used in the CPP procedure for several differentdrugs of abuse, including amphetamine, morphine, andethanol.48–51

In terms of the equipment needed for a place-conditioning study, the conditioning apparatus can bepurchased from a commercial vendor (e.g. Med Associ-ates, Harvard Apparatus, or San Diego Instruments) orhandmade with some common materials found athome improvement stores, for example. While commer-cially available place-conditioning systems commonlyrequire the added expense of a computer and softwarepackage for data acquisition and analysis, these systemsare fully automated in terms of data collection and resultin immediate access to each animal’s data once the testsession has ended. For example, the rodents’ movementsare typically tracked by a series of infrared light sourcesand detectors (i.e. photobeams) located on oppositesides of the apparatus that form a grid to monitor bodypositions in each chamber. Beam breaks that occurwhen an animal traverses the apparatus are recordedby a computer software package and are easily trans-formed into “time spent in compartment” measures.That is, the breaking of specific photobeams is used todetermine the animal’s position in the apparatus and todetermine in what compartment the animal is located.Furthermore, a break in one or several photobeams, inaddition to photobeams in another area being released,is used to determine when the animal has entereda specific compartment. Some systems also monitor

VIII. BEH

vertical movement (rearing) in addition to horizontalmovement throughout the apparatus. Test sessions inan automated apparatus can be video recorded for lateranalysis, but this is not necessary for data acquisitionlike it is in a system without photobeams.

Although the commercially available systems confermany advantages in terms of ease of data collection,purchasing an expensive system is not required toperform place-conditioning studies because the place-conditioning apparatus can be constructed of readilyavailable materials, such as wood or Plexiglas� sheets.Size measurements for rat or mouse place-conditioningchambers are easily acquired through a brief literaturesearch (Fig. 28.3), in addition to examples of visualand tactile cues used in each compartment and varia-tions in lighting schemes that can help to reduce biasfor one compartment over another in experimentalsubjects.

When constructing an apparatus out of materials likewood, it is imperative that the apparatus can be sani-tized easily, so painting of the wood surface is necessary;this is not the case for materials such as Plexiglas�,which will not absorb liquids on their surfaces

AVIOR

28. ANIMAL MODELS OF DRUG ABUSE: CPP AND CTA688

(e.g. urine). Custom-made conditioning chambers allowthe experimenter to fully control the appearance of eachcompartment of the apparatus and the ease with whicha compartment’s appearance can be altered, if necessary.In this way, an experimenter can design an apparatuswith floor textures and/or wall patterns that are easilyinterchangeable.36,52 The advantage of an easily alter-able apparatus is that the cues within each compartmentof the apparatus can be changed tomanipulate the initialside preference that is typically found in rodents.For example, one strain of rats or mice might show nobias when a wire mesh floor, a grid floor, or differentlighting conditions are used in the opposing compart-ments;52 however, this might not be the case fora different strain of rodent or for the opposite sex ofthe same strain of rodent.

In a handmade place-conditioning apparatus, theinfrared light sensors and detectors can be added toautomate data collection during the test sessions ina manner similar to that described in this chapter; thiswill require electronic circuits and computer softwarethat can be made in the laboratory. If constructing thesetypes of circuits and software is not possible, othermethods are available for collecting data in this type ofapparatus, and although they can be labor-intensive,most require only a video camera (or cameras, depend-ing on the number and/or placement of the apparatus inthe laboratory), a stopwatch (or some other way to keeptrack of the time spent in each compartment), and anattentive experimenter or technician (or several attentiveexperimenters or technicians). When “scoring” place-conditioning test sessions (e.g. pretest and posttest) inthis manner, each shuttle box apparatus needs to be infull view of a video camera so that the entire apparatuscan be seen easily on playback. To achieve this, thecamera needs to be positioned above the apparatus,looking down at the conditioning compartments suchthat an animal’s body position is easily discerned abovethe border for each compartment. Furthermore, allexperimenters responsible for scoring the data need tobe adequately trained on what criteria will be used todetermine when an animal is in one compartment ofthe shuttle box and what criteria will be used to deter-mine when that animal is no longer considered to bein that specific compartment. These criteria should bedetermined ahead of time, and reliability testingbetween experimenters should be conducted. Withreliably trained experimenters and consistent method-ology in the laboratory, hand-scoring recorded place-conditioning test sessions provide an inexpensivealternative to purchasing a fully automated systemfrom a commercial vendor.

Place-conditioning studies are short in duration and havethe potential for a high throughput of subjects: Anotheradvantage to the standard place-conditioning procedure

VIII. BEH

is the large number of animals that can be conditionedand tested in a relatively short period of time. If no otherexperimental manipulations are required, a standardplace-conditioning study can run anywhere from 8to 14 days. More specifically, the standard place-conditioning procedure usually consists of a pretestsession or a habituation day (Day 1), four drug and fourvehicle-conditioning trials (Days 2–9), and a test session(Day 10). If all sessions occur on consecutive days,including on the weekends, then the procedure requires10 days; if weekends are excluded, the procedure takestwo 5-day weeks (or 12 total days), whereWeek 1 wouldinclude Days 1–5 (Days 6–7 would be the weekend dayswhere no sessions would occur) and Week 2 wouldinclude Days 8–12. Some investigators recommendinserting a 2-day break between the end of the condi-tioning trials and the place-conditioning test,17 giventhat they have acquired a greater drug-induced CPPfollowing this protocol compared to testing immediatelyafter conditioning (i.e. the day following the last condi-tioning trial).

It should be noted that when conditioning withdifferent drugs or routes of administration, only one ortwo CSþ drug trials are sometimes necessary to condi-tion a preference, which subsequently results in a shorterexperimental period. For example, Bardo et al.49 useda one-trial (a single exposure to the CSþ and CS�)place-conditioning procedure in which rats receiveda single IV injection of amphetamine during one CSþconditioning session; doses of 1 or 3 mg/kg amphet-amine induced a significant CPP in this design (seealso Refs 48,49,51). Ethanol-induced place-conditioningstudies with a reduced number of conditioning sessions(i.e. two CSþ trials and two CS� trials) have beenreported as well.46 While these shorter conditioningperiods have been reported in other laboratories,adhering to a more common conditioning approach isbeneficial to a laboratory that lacks substantial experi-ence with the place-conditioning procedure. Once theparameters within the laboratory are established andplace conditioning is produced with known drug(s)and doses, then alterations to this standard timelinecan be made; these changes, however, could requireadditional control groups to demonstrate the reliabilityof the findings.

In a standard place-conditioning experiment, a periodof at least 24 h commonly separates the conditioningtrials, and this results in either one CSþ or one CS� trialbeing run per day in each animal. The studiesmentioned here decreased the total number of condi-tioning trials, but still maintained this 24 h interval(i.e. an intertrial interval). Several investigators havereported place conditioning following two conditioningtrials in one day (e.g. a drug trial in the a.m. followed bya vehicle trial in the p.m.). Use of an intertrial interval

AVIOR

THE PLACE-CONDITIONING PROCEDURE 689

shorter than 24 h should be carefully considered; seeRef. 53 for a discussion of the advantages and disadvan-tages of this procedure.

Other advantages of the place-conditioning procedure: Inaddition to the advantages discussed in this chapter,there are several additional items that make the place-conditioning procedure an excellent preparation withwhich to study the conditioned motivational effects ofdrugs of abuse. First, test sessions in this procedureare drug-free, which eliminates the confounding effectsof drug-induced motor and/or sensory impairments(or enhancements) that occur in other procedureswhere subjects are tested shortly following drugadministration.36 Second, this procedure is sensitiveto the effects of low drug doses, and when a varietyof doses are tested in separate groups of subjects withinthe procedure, monophasic dose–response curves havebeen reported. It is important to note that this is notalways the case, and a common criticism of the CPPdesign is the difficulty in obtaining well-defined dose-dependent responses using the between-groupsdesign.54 Furthermore, many studies report an “all-or-none” effect in which, after achieving a thresholddose and inducing CPP, no further increase in CPPoccurs at higher doses (for discussions of this effect,see Refs 17,55).

The use of drug-naı̈ve subjects is an additional advan-tage to the place-conditioning procedure. Severaldifferent drugs of abuse induce CPP in a variety ofdrug-naı̈ve subjects, which provides a reliable methodto study the rewarding effects of a specific drug withoutthe added effects of tolerance (or sensitization) fromprevious drug exposures. Generally speaking, the factthat a lengthy drug history is not necessary to induceplace conditioning and, as mentioned in this chapter,specific routes of administration can reduce the numberof conditioning trials needed to induce CPP all add tothis procedure’s flexibility and adaptability. Althougha drug history is most often not needed for drug-induced CPP, there are examples of rats needingprevious drug exposures (with the US or a differentdrug) or stressful experiences to induce an ethanolCPP.56–58 Thus, experimenters should research exten-sively the specific species and/or strain to be used inthe place-conditioning procedure, understand thesedifferences, and adapt the general procedure and/orinterpretation of the results to account for these factors.

Finally, the neurobiological underpinnings of drugreward can be investigated in the place-conditioningprocedure, and recent work in the drug abuse fielddemonstrates the utility of this procedure for thesetypes of studies. This procedure has been used exten-sively to investigate differences in drug-induced rewardamong genetically modified rodents, different rodentstrains, and even different animal species with the

VIII. BEH

goal of understanding the neurobiological mechanismsresponsible for the rewarding effects of abuse drugsand how these effects are altered following chronicdrug exposure or withdrawal. For instance, inhibitionof the IRS2-Akt (protein kinase B) pathway, which isthought to mediate opiate-induced changes in dopami-nergic neurons of the ventral tegmental area, results ina diminished morphine-induced CPP.59 Cocaine CPP,interestingly, is apparent at a lower dose in fosB mutantmice compared to wild-type mice,60 which demon-strates the sensitivity of the place-conditioning proce-dure to molecular changes related to chronic cocaineuse as well.

Place Conditioning: Experimental Protocoland Conditioning Apparatus

Although the characteristics of a place-conditioningstudy have been mentioned briefly in this chapter, thecurrent section aims to describe the place-conditioningstudy in detail, including a discussion of some generalparameters and characteristics of the experimental designthat should be taken into consideration when designingand executing a place-conditioning experiment.

Phases of a place-conditioning study: There are severalgeneral phases to a place-conditioning study, and thesewill be discussed here. These phases are includedbecause they are commonly reported in place-conditioning studies; however, some can be altered oreliminated depending on the goals of the experimentand the extent to which the laboratory has experiencewith the procedure and/or apparatus.

The Pretest: The pretest or preconditioning test isperformed at least one day prior to the start of theplace-conditioning trials and consists of a drug-freetest period (typically 15 min) where subjects are allowedto freely explore all parts of the apparatus. During thistest, the amount of time spent in the to-be-conditionedcompartments is recorded. This test is used to accom-plish several important experimental goals, including(1) determining whether the subjects have an inherentbias for one set of cues within one chamber over another,(2) giving the subjects a brief exposure to the novel envi-ronment of the place-conditioning apparatus, and (3)providing a data point to which posttest changes inpreference are compared. Following the pretest, thebaseline preference score for each compartment foreach individual animal is used to assign drug- andvehicle-paired environments. If an unconditionedpreference for one of the compartments exists (i.e. aninherent or natural bias or a side preference), subjectsthen receive drug injections in their least-preferredcompartment (i.e. the compartment in which they spentthe least amount of time), and this is termed a “biasedprocedure” (Fig. 28.4).

AVIOR

FIGURE 28.4 Morphine-induced place preference in adult male Fisher 344 inbred rats. (a) All rats received one 15 min pretest on the dayprior to the start of the conditioning phase. There was a significant preference for the white textured side of the apparatus; thus, a “biaseddesign” was used and rats were conditioned with morphine in the black smooth side (i.e. the least-preferred compartment was the CSþ). Ratsthen received four morphine (5 mg/kg, SC) and CSþ pairings and four vehicle and CS– pairings. The control group received vehicle paired withboth compartments. (b) Mean time in seconds spent in the CSþ compartment on the pretest and posttest by the vehicle-conditionedand morphine-conditioned rats. Morphine-conditioned rats displayed a significant preference for the CSþ compartment compared to ratsconditioned with vehicle. (c) This figure presents the same data as a difference score (i.e. posttest seconds in CSþ minus pretest seconds in CSþ)to demonstrate the significant morphine-induced CPP. (d) The same CPP data presented as mean percentage of time spent in the CSþcompartment (i.e. CSþ compartment seconds divided by the total time spent in both the CSþ and CS� compartments multiplied by 100).Although there was a pretest in this study, there are numerous ways to present the data, and not all require a comparison to the pretestperformance. Data redrawn from Ref. 31.

28. ANIMAL MODELS OF DRUG ABUSE: CPP AND CTA690

If no inherent side preferences exist, the drug-pairedcompartment is counterbalanced across animals, wherehalf of the subjects receive drug injections in onecompartment and the other half receive drug injectionsin the opposite compartment; this is termed the “unbi-ased procedure.”

The pretest is not always necessary, and in laborato-ries where the place-conditioning parameters havebeen optimized (see Ref. 36 for a specific procedure,drug, and/or rodent strain or strains), the pretest is notregularly used. In these cases, a short (e.g. 5 min) habit-uation session is used to reduce the effect of noveltydescribed in this section. However, the pretest is recom-mended if the apparatus (or cues within either compart-ment) or rodent strain is new to the laboratory. Inaddition, acquiring the pretest data allows experi-menters the opportunity to alter the cues within each

VIII. BEH

compartment in order to eliminate any inherentside preferences and use the “unbiased procedure”(described below; see also Ref. 52).

Conditioning Trials: In a standard place-conditioningstudy, there are typically four conditioning trials inwhichthe drug is administered prior to placement into the CSþcompartment and four additional trials where a vehicle(or no injection) is administered prior to placement intothe CS� compartment (for a vehicle-only control group,vehicle injections are administered prior to placementinto both the CSþ or CS� compartments). More speci-fically, for each drug-conditioning trial, animals areremoved from their home cages, injected with the appro-priate dose of drug (or vehicle), and immediatelyplaced into the CSþ compartment (or the CS� compart-ment if receiving a vehicle). In order to counterbalancethe drug and vehicle pairings across subjects on each

AVIOR

THE PLACE-CONDITIONING PROCEDURE 691

conditioning trial, half of the experimental animals areinjected with drug and the other half are injected withvehicle; this arrangement is reversed on the next condi-tioning trial. A vehicle-only control group is injectedwith vehicle prior to placement in either the CSþ orCS� compartments; these vehicle–compartment pair-ings should also be counterbalanced, such that half ofthe control group receives vehicle paired with the CSþand the other receives vehicle paired with the CS� oneach conditioning trial.

Conditioning trials typically occur once every 24 h,with the same type of trial occurring at 48 h intervals;for example, a conditioning trial in which drug wasadministered is followed 24 h later by a conditioning trialin which vehicle is administered. Thus, the conditioningperiod typically requires 8 days of altering drug andvehicle trials. As mentioned in this chapter, changes inroute of drug administration or other parameters providea means by which the number of conditioning trials isreduced or otherwise altered from this common arrange-ment (e.g. IV administration and one-trial conditioning).Furthermore, greater CPP has been reported for thesame drug with an increasing number of conditioningtrials, a finding that encourages the use of more thanone or two conditioning trials. For example, morphineCPP increases with the number of conditioning trials,and a significant CPP is apparent after three or fourconditioning trials, but not after one or two61 (but seeRef. 62). However, in strains that show differential sensi-tivity to opioids, a significant morphine-induced CPPis apparent after only two conditioning trials fora 1 mg/kg dose of morphine, but is not apparent afterfour conditioning trials with a higher 10 mg/kg dose43

(but see Ref. 63).Although the duration of the conditioning trials is

commonly 30 min, this parameter varies considerablyacross studies with reports of conditioning trials as shortas 4 min and as long as 120 min.36,64,65 One of the mostimportant factors used to determine the length of theconditioning sessions is the duration of the drug effect,such that longer conditioning trials are typically usedfor drugs with longer half-lives. For example, studiesassessing cocaine-induced CPP in rats commonly report20–30 min conditioning trials,51,66–69 whereas morphine-conditioning trials are usually 30 min or longer (see alsoRef. 15; for up to 120 min, see Ref. 64). Conditioning-trialduration is also dependent on the strain of rodent used assubjects. More specifically, 60 min cocaine-conditioningtrials are most effective when DBA/2J mice are used assubjects, but in C57BL/6J inbred mice, cocaine-inducedCPP occurs following conditioning trials between 15and 60 min in length.70 Additionally, ethanol is mosteffective in the CPP procedure in various strains ofmice when 5 min conditioning trials are used.71 Thesedifferences emphasize the importance of selecting the

VIII. BEH

appropriate trial duration for the drug of interest andsubjects used.

The Posttest: Following the conditioning period,a CPP test is run to determine if subjects spend moretime in the drug-paired compartment. The posttest isidentical to the pretest: in a drug-free state, subjectsare allowed to move through all compartments of theapparatus for a specific amount of time (usually15 min). The amount of time spent in each compartmentis recorded. A drug-induced CPP is apparent whena subject spends more time in the drug-paired compart-ment compared to the time spent in that same compart-ment during the pretest or compared to the time spentin the CS� compartment. Thus, CPP data are expressedeither as an increase in preference for the CSþ compart-ment from pretest to posttest or as an increase in prefer-ence for the CSþ compartment compared to the CS�compartment (see Ref. 55). As mentioned here, the post-test occurs at some point following the final conditioningtrial; some studies assess CPP 24 h following the lastconditioning trial, and others assess CPP 2 daysfollowing the last conditioning trial. In the latter instance,no drug injections and/or exposure to the compartmentsare given during these 2 “rest” days.17

The Place-Conditioning Apparatus: The mostcommon apparatus used in place-conditioning studiesis a two- or three-compartment shuttle box, but severaldifferent arrangements have been reported and areshown in Fig. 28.2 (see also Ref. 55). While these shuttleboxes vary in size, lighting conditions, and environ-mental cues, there are several characteristics commonto the experimental apparatuses of most place-conditioning studies with rodents. For mice, a stan-dard-sized conditioning apparatus is approximately30 cm long� 15 cm high� 15 cm deep;36 for rats,a commercially available apparatus from San DiegoInstruments, for example, is approximately 68.5 cmlong� 34.5 cm high� 21 cm deep (Fig. 28.3); these sizescan vary slightly with the addition of a third compart-ment. Importantly, each compartment is distinct fromthe other, regardless of whether a two- or three-compartment apparatus is used. The differences mostoften include the type and texture of flooring materialand the color and appearance of the walls of eachcompartment. The floors are typically made of variousmaterials including wire mesh, steel panels with holesor rods, textured or smooth Plexiglas, textured orsmooth Kydex plastic, and rodent bedding. If the floorcues (i.e. tactile cues) are the to-be-conditioned cuesassociated with the CSþ compartment, the walls of theapparatus do not necessarily need to include any visualcues. However, if visual cues, in addition to the tactilecues, are to be used, a common practice is for onecompartment to be a solid color (e.g. black) and anothercompartment to contain some type of pattern (e.g. black

AVIOR

28. ANIMAL MODELS OF DRUG ABUSE: CPP AND CTA692

and white stripes) or for both compartments to be solidcolors (e.g. black or white) with different texture and/orcolored floors (for an example, see Ref. 43). Regardless ofthe floor and/or wall materials chosen, it is imperativethat all surfaces of the apparatus are constructed ofmaterials that are easy to sanitize (in accordance withinstitutional animal care guidelines) during a place-conditioning study in order to eliminate olfactory cuesfrom each subject.

The use of a two- or three-compartment apparatushas received some attention in the place-conditioningliterature as an important design characteristic thatprovides an advantage in terms of a control for novelty,but has been shown to impact the strength of condi-tioning with certain drugs. More specifically, a thirdcompartment in the conditioning apparatus is typicallysituated in between each conditioning compartmentand is available for exploration only during the test trials(e.g. pretest and posttest); since this compartment is notaccessed during conditioning, it is considered a novelenvironment compared to the other compartments.However, the number of pretest sessions where subjectshave access to this third compartment can impact thenovelty of this area and the amount of time spent inthis area during a test session.62

Rats, for example, exhibit preferences for novelenvironments,62,72–75 and it has been argued thatdrug-induced place conditioning actually representsavoidance of the saline-paired compartment,75 sincethe effects of the drug could interfere with habituationto the drug-paired compartment and is perceived as themore novel compartment during the drug-free testsession. In a series of studies, however, Parker62

demonstrated that for several drugs, including apomor-phine, amphetamine, and morphine, rats displayedsignificant preferences for the drug-paired compart-ment compared to both the saline-paired and novelcompartments of a three-compartment, Y-shaped appa-ratus. Interestingly, rats’ preferences for the novelcompartment were intermediate to those of the drug-and saline-paired compartments, with rats preferringthe novel compartment more than the saline-pairedcompartment, but less than the drug-paired one.62

Further, in a meta-analysis by Bardo et al.15 a three-compartment apparatus was associated with a largereffect size in heroin and cocaine place-conditioningstudies, but no association was found when morphineor amphetamine was used. These authors argue furtherthat drug-induced CPP might be enhanced in the three-compartment apparatus because the availability of thenovel compartment during the test session reducesthe amount of time spent in the saline-pairedchamber.15 Thus, these data demonstrate that the choiceof using a two- or three-compartment apparatus is animportant one that can impact the results of the study

VIII. BEH

and should be made only after careful considerationof these issues and the benefits (or limitations) associ-ated with the use of each type of apparatus.

General Rodent Place-Conditioning Procedure:

Pretest Session (Day 1)

1. Clean all conditioning chambers thoroughly withsoap and water; allow to air dry before beginningthe pretest. If changing of floor materials isrequired, make certain that the appropriateconditioning floors and cues are inserted intothe appropriate chambers. These cues will be thesame ones used during conditioning and on theposttest.

2. Weigh rodents. Place each subject into the middle ofthe two-compartment apparatus, and recordmovements for 15 min. If the apparatus has a third(middle) compartment, place the rodent into themiddle compartment, remove the barriers, andallow the animal to freely explore all compartmentsfor 15 min while recording the animal’s movements.Note: If the home cage area and the conditioningapparatus are not located in the same location, someinvestigators recommend moving all animals toa quiet area closer to the conditioning location for anhour prior to the start of the pretest (or conditioningsessions or posttest).

3. Using the pretest data, determine each subject’sinitial side preferences. If an unconditionedpreference exists, the least-preferred compartment isthe drug-paired compartment (CSþ) and a biasedprocedure should be used. If there is nounconditioned preference, drug- and vehicle-pairedcompartments are counterbalanced across subjects,such that half receive drug (or vehicle) in onecompartment and half receive drug (or vehicle) inthe other compartment (i.e. an unbiased procedure).Once subjects are assigned a drug-pairedcompartment, the presentation of CSþ and CS�trials must be counterbalanced as well. That is, halfof the subjects receive drug–compartment pairings(or CSþ trials) on odd-numbered conditioning trials(i.e. drug on Days 3, 5, 7, and 9), and half receivethese pairings on even-numbered conditioning trials(i.e. drug on Days 2, 4, 6, and 8).

4. Thoroughly clean all compartments betweenanimals and at the end of the day.

Conditioning trials (Days 2–9)

5. Twenty-four hours following the pretest, theconditioning trials commence. Begin by cleaningeach compartment and placing the appropriate floormaterials and cues in each conditioningcompartment; insert barriers to keep animalsconfined to a specific compartment.

AVIOR

THE FLAVOR-CONDITIONING PROCEDURE 693

Note: Some laboratories use the entire conditioningapparatus during each conditioning trial, such thatthe floor material for the CSþ or CS� trial covers theentire apparatus. Others insert barriers between thetwo compartments such that the apparatus is splitinto its conditioning compartments and the barrier ispresent throughout the entire conditioning phase.Regardless of what option is used, the conditioningapparatus must contain both conditioning floors andcues during the posttest session.

6. Weigh each subject, inject with the appropriatesolution (i.e. drug or vehicle), and immediately placethe subject into the assigned compartment fora specified time periodda general duration is30 min, but this parameter varies with different drugUSs and should be chosen for the specific drug beingstudied. Subjects’ activity can be recorded duringthe conditioning trials, but it is not required.Note: Subjects should be run at the same time eachday throughout the entire procedure (i.e. frompretest through conditioning and the posttest). Thus,drug solutions should be made in advance to allowenough time for the equipment setup and syringepreparation each day.

7. Once the trial is complete, remove the animal andreturn it to its home cage. Clean each compartment,and prepare the equipment for conditioning trialswith additional subjects, if necessary.

8. Twenty-four hours after the first conditioning trial,steps 5–7 are repeated. Animals receive an injectionof the opposite solution, followed by confinement tothe other compartment. More specifically, an animalthat received a drug–compartment pairing (or CSþtrial) on Day 2 now receives a vehicle–compartmentpairing (or CS� trial) on Day 3. This process isreversed for animals that received CS� trials onDay 2. Conditioning trials continue in an alternatingmanner every 24 h until Day 9, which results in fourCSþ trials and four CS� trials.

Posttest (Preference Test; Day 10)

9. Twenty-four hours after the final conditioning trial,clean all compartments. Prepare each conditioningapparatus with the appropriate floor materials andcues. If using a two-compartment apparatus,remove the barrier between the compartments andplace the animal in the middle of the apparatus.Record activity. If using a three-compartmentapparatus, place the animal in the middlecompartment of the apparatus; remove the barriersand record activity.

10. Once the test is complete, remove the animal fromthe conditioning apparatus and return it to its homecage. Clean each compartment thoroughly, andprepare for additional animals, if necessary.

VIII. BEH

THE FLAVOR-CONDITIONINGPROCEDURE

What are Conditioned Taste Aversions?

Flavor conditioning is a simple procedure mostcommonly used to assess the positive (appetitive) ornegative (aversive) effects of numerous compounds,including various nutrients, food sources, toxins,poisons, ionizing radiation, and drugs of abuse (seeRef. 76). Conditioned flavor preference (CFP) is a proce-dure used to assess the positive effects of a stimulus bymeasuring increased consumption of a solution,whereas conditioned taste aversion (CTA) is used toassess the aversive effects of different stimuli bymeasuring decreases in consumption of a palatable solu-tion. While CFP is a common procedure used to assessthe positive effects of oral and postoral associations,77,78

the current review will focus on the use of the taste aver-sion procedure in measuring the aversive effects ofdrugs of abuse. In a standard CTA experiment, subjectsundergo conditioning trials in which consumption ofa novel, palatable solution (typically saccharin) is pairedwith the injection of a drug. Interspersed between con-ditioning trials are water recovery days, where subjectsreceive water (or a solution with another distinct flavor)to drink but do not receive any drug injections. Afterseveral saccharin–drug pairings, subjects are tested forthe expression of the CTA during a final test trial inwhich subjects are given access to the saccharin solutionand consumption is recorded; no injections followsaccharin access on this test trial. Subjects that decreaseconsumption of the saccharin solution following thesaccharin–drug pairings are said to express a condi-tioned taste aversion. While the methodology is slightlydifferent, procedures similar to this that result in greaterconsumption of the novel fluid are indicative of condi-tioned flavor preferences.

Similar to the CPP procedure discussed in thischapter, the CTA procedure is also based on principlesof Pavlovian classical conditioning, where the effect ofthe drug administered serves as the US and is subse-quently paired with the novel saccharin solutiondafterseveral saccharin–drug pairings, consumption of thesaccharin solution (CS) decreases due to its associationwith the aversive effects of drug administration. Inter-estingly, most drugs that serve as reinforcers in theself-administration procedure and that condition placepreferences also induce taste aversions, which suggeststhat reinforcing drugs of abuse have both rewardingand aversive effects, and that the balance of these effects(i.e. reward and aversion, as discussed in this chapter)could impact overall acceptability, which could influ-ence subsequent use and abuse of these compounds(see Ref. 79).

AVIOR

28. ANIMAL MODELS OF DRUG ABUSE: CPP AND CTA694

A Brief History of Taste Aversion Learning

One of the first demonstrations of taste aversionlearning was in 1951, when John Garcia and colleaguesobserved that rats decreased consumption of waterfrom bottles that had been present during irradiation(for a review of the history of CTA, see Ref. 76). Garciaargued that the decrease in consumption followingexposure to radiation was due to the acquired associa-tion of the taste (imparted on the water from the plasticbottles in which it was contained) with the aversiveeffects of gamma irradiation. This explanation appearedto be valid, given that the same rats would ingest waterfrom glass bottles in their home cage. In 1955, Garcia andcolleagues sought to systematically test this hypothesisby pairing a novel saccharin solution with gammairradiation.80 Consistent with his earlier findings, ratsreceiving saccharin–irradiation pairings decreasedconsumption of this palatable solution, whereas ratsreceiving saccharin paired with sham irradiation dis-played no change in saccharin consumption. Thus, theeffects of gamma radiation were causing a decrease inconsumption of the novel fluid, demonstrating that radi-ation exposure had identifiable effects that could bereadily associated with other stimuli, such as taste.Importantly, this effect was not simply a direct effect ofradiation-induced damage impairing the ability ofthese animals to consume the sweetened solution, giventhat rats receiving exposure to gamma radiation thatwas not explicitly paired with the saccharin solutioncontinued to consume the saccharin solution at controllevels.80,81 The effects of gamma irradiation were thusassociated with the novel-tasting saccharin solution,and this association created the decrease in consumption

FIGURE 28.5 Median saccharin preference in rats conditioned

with a single exposure to gamma irradiation at 30 or 57 rad. Theirradiation-induced CTA was apparent for approximately 30 dayspostirradiation; after that time, some extinction of the aversion wasapparent. Redrawn from Ref. 80.

VIII. BEH

on successive saccharin presentations. Garcia et al.determined that irradiation could condition a decreasein saccharin consumption indicative of a taste aversion(Fig. 28.5).

Following these initial studies, numerous reportsfrom Garcia and his colleagues and other investigatorsdetailed several important findings that now charac-terize the phenomenon of taste aversion learning as a“specialized form of learning.” In his initial study in1955, Garcia et al. demonstrated that radiation-induced taste aversion was evident following onlyone saccharin–radiation pairing. Furthermore, theseaversions were still evident at 30 days postirradia-tion.80 Shortly following this report of one-triallearning of CTA, it was shown that taste aversionscould be acquired over long time delays betweenconsumption of the saccharin CS and presentation ofthe US, which in these early studies was ionizingradiation82–84 but this has since been shown to occurwith other compounds serving as the US, such aslithium chloride, an emetic, and cocaine, a drug ofabuse.85–88 Lastly, Garcia and Koelling89 reported thattaste aversions appeared to be selectively acquired togustatory stimuli, whereas stimuli like audiovisualcues did not serve as a CS in this design becausethey were not readily associated with the X-ray orlithium chloride USs (see also Refs 90,91). Takentogether, one-trial learning, long-delay learning, andthe selective associations of this procedure tookaversion learning beyond an interesting radiation-induced effect to a learning example that was some-what at odds with the traditional views of associativelearning. More specifically, traditional associativemodels generally assumed that control could be estab-lished only after many conditioning trials with short-delay and/or trace-conditioning techniques and thatconditioning could be achieved equally well withmost CS–US combinations.92,93 Conditioned tasteaversion learning was thus described as a specializedform of learning that facilitated specific associationsimperative to an animal’s survival, such as thosebetween taste and the postingestive consequences ofconsumption (e.g. illness). These associations areargued to be highly adaptive because they providean animal a means by which rapid acquisition oftaste–illness associations can occur following consum-ption of natural toxins in the environment (e.g. thosemost likely found in food sources) followed sometimelater (e.g. after the natural delay imposed by digestion)by illness.

After these initial assessments in which radiation andlithium chloride induced taste aversions to palatablesaccharin solutions, investigations began to assesswhat other compounds could induce a CTA. Giventhat aversion learning is viewed as a unique form of

AVIOR

THE FLAVOR-CONDITIONING PROCEDURE 695

learning that enables an animal to learn about andsubsequently avoid food sources containing poisons ortoxins, many of these early studies focused on assessingtaste aversion learning induced by compounds orsubstances with known aversive effects, includingpoisons, classical toxins, and common emetics. One ofthe first emetic compounds examined in this contextwas apomorphine, a nonselective dopamine agonist.Apomorphine-induced aversions were robust, werereadily acquired after only one CS–US pairing, andcould withstand long delays during conditioning,much like the taste aversions reported earlier followingirradiation.94–96 Furthermore, lithium chloride gainedpopularity as a tool to investigate how manipulationsto the standard taste aversion procedure would impactacquisition of a taste aversion.97,98 Work assessing otheraversive compounds in the CTA design steadilyincreased throughout the 1970s and into the 1980s:various well-known toxins, including strychnine sulfate,acetaldehyde, sodium fluoride, paraquat, and physo-stigmine sulfate, all reportedly induced CTAs (fora review, see Ref. 99). Clearly, taste aversion learningand the conditions under which it is acquired are notsimply an effect of exposure to irradiation, but arecommon characteristics of a procedure that has generalutility as a measure of drug toxicity; more specifically,the fact that toxin- and poison-induced CTAs areacquired at doses much lower than those needed toadversely impact other behaviors, including food andwater consumption, demonstrates the sensitivity ofthis procedure for measuring the aversive effects ofvarious compounds.

While toxins, poisons, and emetics are common USsin the CTA procedure, drugs of abuse have receivedconsiderable attention in taste aversion experiments aswell. Further, the taste aversion procedure was viewedas a sensitive assay with which to investigate the poten-tial aversive effects of these drugs, given that thesecompounds were viewed as stimuli with numerousbehavioral effects because they were reinforcers in self-administration and discriminative stimuli in drugdiscrimination studies and could alter learning andmemory in other procedures.100–102 One of the firstreports of drug-induced taste aversion learning waspublished in 1971, in which Cappell and LeBlancassessed the aversive effects of amphetamine or mesca-line in rats.103 In this study, both compounds induceddose-dependent aversions to the saccharin solutionfollowing one conditioning trial in which a drug adminis-tration was paired with saccharin to drink. These authorsargued that the amphetamine- and mescaline-inducedCTAs support the idea that these two drugs, at the dosesused, had aversive effects that resulted in significantdecreases in saccharin consumption on the test day.They argued further that the CTA results also support

VIII. BEH

the lack of self-administration ofmescaline inmonkeys104

and the decrease in amphetamine self-administration inrats at 1 and 2 mg/kg (doses that conditioned a tasteaversion; see Ref. 105).

In a relatively short period of time, most major drugsof abuse were shown to induce CTAs, includingalcohol, morphine, cocaine, and D9-tetrahydro-cannabinol, under the same parametric conditions asirradiation, emetics, and toxins (e.g. one-trial learningor long-delay learning).86 The fact that drugs of abusewere found to have aversive effects is not surprisingsince drug effects are generally dose dependent, withtoxicity occurring at greater drug doses. Yet the factthat these effects are evident at the same drug dose isless intuitive. In a combined CPP–CTA study, Reicherand Holman41 presented data demonstrating thatthe same injection of amphetamine could conditiona taste aversion while conditioning a place preferenceto the environment in which the saccharin wasconsumed. More specifically, rats received an injectionof amphetamine prior to placement into one compart-ment of a place-conditioning apparatus. Within thisamphetamine-paired (CSþ) compartment, rats hadaccess to a distinctly flavored solution (almond orbanana flavor). In the CS� compartment, rats hadaccess to a different solution; if almond was the flavorused in the CSþ compartment, banana was the flavorused in the CS� compartment. Interestingly, amphet-amine administration in this design conditioneda significant place preference while also conditioninga significant taste aversion. Furthermore, these sameresults were evident after a 20 min delay betweenamphetamine administration and placement into theCSþ compartment. Reicher and Holman argued thatamphetamine had rewarding and aversive effectsthat occurred simultaneously in the same animal andthat each of these effects could condition differentbehaviors (i.e. approach and avoidance).

Since these initials studies with drugs of abuse in thetaste aversion procedure, a wealth of literature has accu-mulated investigating drug-induced CTAs and factorsthat influence drug taking, including drug history,genetic manipulations (e.g. inbred rats and mice, andgenetically modified animals), stress, age, and thematernal environment. In addition, the taste aversionprocedure is used in behavioral pharmacology studiesaimed at determining what receptor systems are respon-sible for drug-induced CTAs (see Refs 106,107) and hasbeen adapted for use as a procedure to assess thediscriminative stimulus effects of various drugs, espe-cially low doses of opioid antagonists in drug-naı̈verats.108 Thus, the CTA procedure is commonly used toassess the aversive effects of drugs of abuse, and oncethe general procedure is instituted, adaptations tothe procedure can be made to enable subsequent

AVIOR

28. ANIMAL MODELS OF DRUG ABUSE: CPP AND CTA696

investigations in a manner similar to those mentioned inthis chapter.

Advantages of the CTA Procedure

The advantages of the CTA procedure are similar tothose listed in this chapter for the CPP procedure; none-theless, this section will outline these advantages andfocus on specific equipment and techniques that areuseful for taste aversion studies.

The CTA procedure is methodologically simple and doesnot require special and/or expensive equipment: The tasteaversion procedure is a popular method to study theaversive effects of drugs of abuse for several reasons,including (1) the procedure’s simplicity in terms ofexperimenter training and (2) the fact that little special-ized equipment is needed; most CTA studies are con-ducted in an animal’s home cage environment. Muchlike that described in this chapter for the place-conditioning procedure, taste aversion conditioningdoes not require any specialized training like jugularcatheter surgery for self-administration studies. For theCTA procedure, drug is most commonly administeredvia the IP or SC route. While these drug administrationsare relatively simple, the route of administration used intaste aversion conditioning is an important factor andhas been shown to differentially impact CTAs inducedby different drugs of abuse. Cocaine-induced aversions,for instance, are weaker when drug administrations areIP compared to aversions induced by SC administeredcocaine. Ferrari et al.109 demonstrated that IP cocaine-induced weak dose-independent aversions, such thatall three dose groups (18, 32, and 50 mg/kg) signifi-cantly decreased saccharin consumption on the finaltest trial when compared to consumption of the vehiclecontrol group. In contrast, rats injected SC with cocaineexhibited robust taste aversions at 32 and 50 mg/kgdoses; decreases in saccharin consumption in thesetwo groups were evident on the second conditioningtrial (see also Refs 109–112). In addition, intracerebro-ventricular drug administrations are also used in tasteaversion learning, primarily in studies interested indetermining whether a drug’s aversive effects are cen-trally or peripherally mediated or what brain areas areinvolved in a drug-induced CTA.113–116

The equipment needed for a taste aversion study isrelatively simple and typically includes two differentsets of water bottles with rubber stoppers and sippertubes, in addition to some type of sweetener, whichwill be prepared as a solution in tap water and pairedwith drug administrations. Two sets of water bottlesare needed because one set is used to provide theanimals with plain tap water (or plain filtered water),and the other set is used for the saccharin solutionduring the conditioning trials. The use of different

VIII. BEH

water and saccharin–water bottles helps to ensurethat the taste of the saccharin remains novel andprohibits subjects’ exposure to this taste on habituationand/or water recovery days. For this reason, two sets ofrubber stoppers with sipper tubes are needed as well.Different types of drinking tubes are used in tasteaversion studies, and the type of tube chosen usuallydetermines how fluid consumption is measured(Fig. 28.6).

For example, graduated 50 ml Nalgene drinkingtubes are commonly used for taste aversion studieswith rats, and fluid consumption is easily recorded in0.5 ml increments. Fluid measurements are taken priorto placement of the drinking tube onto each subject’scage and are taken again after the drinking periodhas ended. Given that movement of the drinking tubescan lead to some fluid loss, care must be taken to placedrinking tubes on the cages while minimizing drippingor spilling. To determine the amount of fluid lossthat occurs during placement of the drinking tube,a full drinking tube can be placed on an empty cageand pre- and postfluid access measurements can berecorded. Nongraduated drinking bottles are commonlyused as well; however, bottles are weighed prior to eachsubject’s fluid access period and then weighed againimmediately following the end of the drinking period.

A review of the taste aversion literature demonstratesthe variety of gustatory stimuli that have been used intaste aversion studies. For example, numerous studieshave used condensed milk or other flavored solutions,including Kool-Aid�, flavor extracts (e.g. almond,banana, coconut, or orange), vanilla, sucrose, or poly-cose. One of the most common gustatory stimuli isa 0.1% sodium saccharin solution, a relatively inexpen-sive noncaloric sweetener dissolved in tap water.Noncaloric sweeteners are typically used to decreasethe likelihood that changes in consumption result fromthe high caloric value of a solution, such as sucrose,glucose, or polycose;117 nonetheless, caloric sweetenersare used in drug-induced aversion learning111,118,119

and to examine the specificity of the tastant–CS associa-tion.120 In addition, strain differences in preference forvarious gustatory stimuli exist, with most rats andmice displaying a preference for saccharin solutionsover water; however, the amount of saccharin (or otherflavored solution) consumed does differ betweenmany of these strains, and this can impact the resultsof a taste aversion study. For example, Tordoff et al.examined two-bottle-choices tests (i.e. water vs solu-tions containing sucrose, NaCl, saccharin, etc.) in 14common laboratory rat strains that included threeoutbred and 11 inbred strains. All of these strains dis-played a preference for various concentrations ofsaccharin, but the amount of saccharin consumed diddiffer among rat strains.121 Similar results have been

AVIOR

FIGURE 28.6 (a) 50 ml Nalgene tube for fluid access during taste aversion conditioning. This volume of fluid is adequate for a 20–30 minfluid access period. The tube is inverted when it is attached to the rat’s cage. (b) Curved steel sipper tube. If rats are housed in wire mesh cages,these curved sipper tubes are needed for fluid access. If rats are housed in polycarbonate cages with a wire top, curved or straight sipper tubescan be used. (c) Size #7 rubber stopper. This size fits snugly in the 50 ml Nalgene tube. When putting the stoppers into the tubes, wet the stopperand twist it once it has been inserted. This provides a tight seal so that the stoppers are not easily knocked out of the tube. (d) The tube, withinserted stopper, inverted and attached to a wire mesh rat cage.

THE FLAVOR-CONDITIONING PROCEDURE 697

reported for numerous strains of inbred mice as well,and most strains display a preference for at least oneconcentration of saccharin in solution122 (see alsoRefs 123–125). Selection of the appropriate gustatorystimuli for aversion learning can depend on the strainof rodents used in the study; thus, experimenters shouldbe aware of any possible strain differences in tastantpreferences prior to the start of conditioning. If straindifferences are a focus of the aversion study, the US tast-ant chosen should be equally preferred between thestrains. Finally, while drinking tubes or bottles anda novel tastant such as saccharin are required for anaversion study, a specific conditioning apparatus is notnecessary and most CTA studies condition subjects intheir home cages when animals are housed individually;a conditioning apparatus, such as a test cage, has beenused when taste aversion conditioning is done inanimals that are group housed.126

The CTA procedure is relatively short in duration and hasthe potential for a high throughput of subjects: Anotheradvantage to the CTA procedure is the fact that largenumbers of subjects can be conditioned in a relativelyshort period of time. For example, taste aversionstudies commonly employ 4-day conditioning cyclesthat begin with a conditioning trial where 20 min ofaccess to saccharin is followed by an injection ofa drug of abuse; 3 water recovery days follow in which

VIII. BEH

subjects have access to only plain water during the20 min fluid access period and do not receive postcon-sumption drug injections. This 4-day conditioningcycle is typically repeated four times and is thenfollowed by a test day in which subjects receive accessto saccharin, but do not receive any injections followingthe consumption period. When a taste aversion study isconducted using these common parameters, the entireconditioning period requires only 17 consecutive days(including weekends). The conditioning cycles are,however, preceded by a habituation phase in whichsubjects are given 20 min of access to plain tap waterat the same time each day. The length of this habitua-tion phase is determined by the variability in waterconsumption: once subjects display stability in theamount of water consumed each day (e.g. for rats, theamount consumed does not differ by more than 2 mlfor 3 consecutive days)32 and are approaching anddrinking from the water bottle almost immediatelyafter its placement on the cage (i.e. within 2 s of itspresentation), the conditioning phase begins. Whilethere is no standard method for determining stability,these criteria result in relatively stable levels of waterconsumption in approximately 7–14 days, dependingon the strain of rodent (Fig. 28.7; see Ref. 127). If noother manipulations are included, a taste aversionstudy requires a little more than a month (i.e. 31

AVIOR

FIGURE 28.7 Mean water consumption in three different rat

strains during the habituation phase of six different taste aversion

studies (A or B). Although some rat strains consume more watercompared to others, all strains display the same pattern of waterconsumption during the habituation phase. More specifically, waterconsumption increases over the habituation period to an asymptoticvolume from which most rats do not vary (note the errors bars arewithin the symbol for each point on the graph). The SD rats (pinkand yellow squares) needed only a 12 day habituation phase, whereasthe F344 and LEW rats required 12–15 days, depending on the study(A or B, respectively). F344¼ inbred Fischer 344; LEW ¼ inbred Lewis;and SD¼ outbred Sprague Dawley. Data redrawn from Refs 31,79,131.

28. ANIMAL MODELS OF DRUG ABUSE: CPP AND CTA698

consecutive days) to complete, assuming a 2-weekhabituation phase. Obviously, additional manipula-tions, such as a drug preexposure phase, increase thenumber of days needed to complete the taste aversionstudy.79,128,129

Although the 4-day conditioning cycle is common inthe taste aversion literature, shorter conditioning cycleshave been reported and, depending on the manipula-tion, are necessary for the specific experimentalprotocol. For instance, in the combined CTA–CPP proce-dure, the taste aversion procedure is adapted to fit theconditioning phase of the CPP procedure, such thatsaccharin access occurs on the same day as the druginjection and placement into the CSþ compartment.Water access occurs on the intervening days whenvehicle injections precede placement into the CS�compartment. In this combined design, the taste aver-sion conditioning cycles are 2 days and include a condi-tioning trial followed by one water recovery day, insteadof the 3 water recovery days described above. Theconcern with this procedure is that subjects displayinga CTA will not be adequately rehydrated during theone water recovery day, which could negatively impactthe subsequent conditioning trial. Using this combineddesign, however, Roma et al.130 demonstrated thatinbred Fischer and Lewis rats both maintained relativelystable levels of fluid consumption on each waterrecovery day separating each saccharin–IP ethanol–CSþconditioning trial, regardless of the strength of theethanol-induced aversion. Further, the body weights ofboth strains remained stable over the 13-day habituation

VIII. BEH

phase and the 8-day CTA–CPP conditioning phase130

(see also Ref. 44) in all rats, including those subjects dis-playing robust ethanol-induced aversions, that is,subjects that were only consuming water every otherday due to the aversion to the saccharin solution (forsimilar conditioning cycles with cocaine, see Ref. 131).Regardless of the length of the conditioning phase,almost all drug-induced taste aversion studies reportdose–response assessments in which different groupsof subjects (8–10 per dose group) receive saccharinpaired with a specific dose of the drug of interest, inaddition to a control group that receives saccharin–-vehicle injection pairings. Dose–response assessmentstypically include two or more drug doses with separategroups of animals for each dose used; in this way, largenumbers of animals can be conditioned in a relativelyshort period of time in a single taste aversion study.

Other advantages of the taste aversion procedure: In addi-tion to the advantages described in this section, there areseveral additional advantages of the taste aversionprocedure that add to its utility and the ease with whichit can be performed in the laboratory. First, consumptionof the fluid CS is always done under drug-free condi-tions. More specifically, subjects receive the fluid CS todrink for a specified amount of time (e.g. 20 min); oncethis period is over, subjects are then injected with thedrug US. Any changes in consumption then are notsimply a function of drug-induced motor impairments,for example, because the subjects receive drug only afterthe fluid access period has ended. Similar to the drug-free posttest in the place-conditioning proceduredescribed in this chapter, the taste aversion test day(i.e. the day after the final conditioning cycle) isa drug-free test as well.

A second additional advantage is the fact that tasteaversion studies routinely report dose–response func-tions where conditioning day saccharin consumptiondecreases over trials, such that increasing drug dosesare typically associated with a more robust decrease insaccharin consumption. While these functions arecommon in taste aversion learning, floor effects (i.e. nosaccharin consumption on subsequent conditioningdays) do occur and are influenced not only by the condi-tioning dose but also by other factors including baselineconsumption of the novel fluid CS and the strain ofrodent used as subjects. For this reason, choosing a novelfluid CS that is readily consumed by the strain of rodentserving as subjects in a taste aversion study is important.Nonetheless, most drug-induced CTAs are conditionedat relatively low doses where toxicity is not an issueand conditioning of other behaviors is supported. Forexample, morphine CTAs are readily acquired at5 mg/kg, a dose that also conditions a significant placepreference in the same animals.44,45,132 These findingsare true for most drugs of abuse, such that the range of

AVIOR

THE FLAVOR-CONDITIONING PROCEDURE 699

drug doses supporting taste aversion learning alsosupports conditioned place preferences.

Conditioned Taste Aversion: ExperimentalProtocol and Conditioning Apparatus

Although the characteristics of a taste aversion studywere mentioned briefly in this chapter, the currentsection aims to describe taste aversion conditioning indetail, including a discussion of some general parame-ters and characteristics of the experimental design thatshould be taken into consideration when designingand executing a CTA experiment.

Phases of a CTA study: There are several general phasesto a CTA study, and these will be discussed in detail here.These phases are included because they are commonlyreported in taste aversion studies; however, some phasescan be altered depending on the goals of the experiment.

The Habituation Phase: When beginning a tasteaversion study, subjects are typically water restrictedand habituated to drinking during a specified period(i.e. the fluid access period) each day. In order to achievereliable drinking during this period, rats are presentedwith plain tap water following 23.5 h of water depriva-tion. More specifically, rats are maintained on ad libitumaccess to water up to the day prior to the start of thehabituation phase. On this day, ad libitum water bottlesare removed from the rats’ home cages at a specifiedtime (e.g. 10.00 h). On the following day, which is thestart of the habituation phase, subjects receive waterbottles containing plain water to drink at this sametime; rats are given 20 min of access to these waterbottles, and consumption is recorded. If presentationof the drinking bottles occurs earlier or later than thepredetermined fluid access period, the amount of waterconsumed can vary substantially between individualsubjects. For this reason, the water bottles must bepresented at the same time each day to maintain thesubjects’ level of fluid restriction throughout the study.

It is common for subjects to increase water consum-ption throughout the habituation period; thus, thisphase continues until all subjects display stable day-to-day consumption levels.

Using each subject’s mean water consumption,stability is commonly defined as a subject consumingwithin 2 ml of its mean fluid consumption over 3 consec-utive days.79 Beginning the conditioning phase aftersubjects’ water consumption has stabilized helps toensure that changes in consumption are a function ofconditioning and are not due to continued habituationto the fluid restriction procedure. Further, subjectsshould be weighed each day prior to the fluid accessperiod, and weights should be closely monitoredthroughout this phase and the conditioning phasedescribed in this section.

VIII. BEH

Although a single, daily fluid access period iscommonly used in taste aversion conditioning, someinvestigators also include an additional drinking periodduring which subjects have access to only water. Forexample, Grigson and Freet133 employed a 5 min fluidaccess period in the morning followed by an hour-longfluid access period in the afternoon (see also Refs111,116). During the habituation phase, rats had accessto distilled water during the 5 min and 1 h fluid accessperiods, but on conditioning trials saccharin was avail-able during the 5 min period only; distilled water wasalways available during the 1 h afternoon fluid accessperiod. This additional fluid access period is includedbecause subjects were given only 5 min of access towater or saccharin during conditioning and neededthe afternoon period to rehydrate; conditioning trialsin this study occurred every other day, instead of everyfourth day (discussed further in this section).

The Conditioning Phase:Once subjects aremaintain-ing stable levels of fluid consumption during the habit-uation phase (approximately 10–14 days), theconditioning phase begins. On the conditioning daysof this phase, subjects receive saccharin during their20 min fluid access period, and consumption isrecorded. Following the fluid access period, subjectsare removed from the home cage, injected with theappropriate drug dose or vehicle, and returned to thehome cage. The water recovery days of this phase,however, are identical to the days of the habituationphase: subjects receive plain water during the 20 minfluid access period, and consumption is recorded. Nodrug injections occur on these days. Taken together,one conditioning cycle consists of a saccharin-conditioning day, in which saccharin consumptionprecedes an injection of drug or vehicle, followed by 3water recovery days. This 4-day conditioning cycle isthen repeated four times for a total of 4 conditioningdays and 12 water recovery days.

The first day of the conditioning cycle is importantfor two reasons: (1) subjects are assigned to an experi-mental or control group based on their saccharinconsumption, and (2) the CS–US interval is determinedat this time. While subjects can be grouped by meanwater consumption on the 3–5 days prior to the firstconditioning day, it is common for subjects to beassigned to a group based on total saccharin consumedon the first day of conditioning. More specifically, oncethe fluid access period is over, the consumption valuesare ranked from largest to smallest and subjects areassigned to an experimental or control group suchthat mean consumption is comparable across groups.By using this pseudorandom assignment technique,saccharin consumption on the first conditioning day isnot significantly different between the various experi-mental (and control) groups, which provides a good

AVIOR

28. ANIMAL MODELS OF DRUG ABUSE: CPP AND CTA700

baseline for subsequent saccharin consumption daycomparisons. This comparable baseline also increasesthe likelihood that data can be presented as total milli-liters consumed instead of having to be transformedto a percentage shift from or percentage of controlconsumption (Fig. 28.8).

In addition to grouping subjects on this day, theCS–US interval is determined as well, and this intervalneeds to be maintained on subsequent conditioningdays. For example, each subject receives 20 min of accessto saccharin followed by an injection sometime afterfluid consumption. The time between fluid consumptionand the subsequent drug injection should be carefullytimed on the first conditioning day, and this CS–USinterval should be maintained throughout the condi-tioning phase. Thus, if subjects are injected 5 min aftersaccharin consumption on the first conditioning day,subsequent injections should also occur 5 min aftersaccharin access.

The One-Bottle Final Aversion Test: Following thefinal water recovery day of the last conditioning cycle,the final aversion test is typically administered. Thistest consists of 20 min of access to saccharin during thefluid access period, following which no injections areadministered. When done in this manner, this test isalso referred to as a one-bottle aversion test. If thedrug-induced taste aversions are robustdsubjects areconsuming little to no saccharin on this testdthe one-bottle aversion test is typically the final day of a condi-tioned taste aversion study.

The Two-Bottle Final Aversion Test: If the condi-tioned taste aversion is not robust, does not differ by

FIGURE 28.8 Morphine-induced conditioned taste aversion in

adult male Fischer 344 rats. Following the habituation phase onConditioning Day 1, all rats received 20 min of access to saccharinfollowed by a subcutaneous injection of morphine (5 mg/kg) orvehicle. Consumption on this day is equivalent between the groupsbecause group assignments were made based on the ranking proce-dure described in the text. Over the 3 subsequent conditioning days,rats receiving morphine display a significant decrease in saccharinconsumption compared to vehicle-treated control rats, in addition toa significant decrease in saccharin intake on the one-bottle finalaversion test. Redrawn from Ref. 31.

VIII. BEH

dose, or is relatively weak, for example, a more sensitivetwo-bottle test is usually administered. This test isconsidered to be more sensitive than the one-bottle testdescribed here because subjects have access to saccharinand water concurrently during the 20 min access periodand can choose to drink the saccharin or water duringthis time. If a taste aversion has been conditioned,subjects consume a greater percentage of watercompared to saccharin during this period, with similarlevels of total fluid consumption (i.e. saccharinþwaterconsumption) between experimental and controlgroups. If the two-bottle test will be run following theone-bottle aversion test, it is important to consider inject-ing the subjects following the one-bottle test and givingthe subjects 3 water recovery days (i.e. running a fifthconditioning cycle). In this way, the subjects have alwaysexperienced saccharin paired with the drug effects andno extinction effects will be evident on the results ofthe two-bottle test.

In a recent paper examining taste aversions inducedby opioid receptor agonists, Davis and colleagues127

demonstrated the sensitivity of the two-bottle aversiontest following what appeared to be weak aversionsinduced by (�)-U50,488H and SNC80, which are kappaand delta opioid receptor agonists, respectively.

Saccharin consumption in the experimental groupsconditioned with either drug did not differ much fromtheir control group’s saccharin consumption over fiveconditioning trials. More specifically, only the subjectsconditioned with the highest dose of the kappa agonistdisplayed significant decreases in saccharin consump-tion compared to control subjects; subjects conditionedwith SNC80 displayed similar levels of saccharinconsumption throughout conditioning that did notdiffer between the conditioning doses. However, robustand dose-dependent (for the kappa agonist only) drug-induced aversions were apparent in the two-bottleaversion test, which was administered following theone-bottle test and 3 water recovery days (i.e. a fifthconditioning cycle; Fig. 28.9). These results demonstratethe usefulness of the two-bottle aversion test, and whileit is not required, administering this test is highly recom-mended, especially when aversions are weak overconsecutive conditioning trials.

General Rat Taste Aversion Conditioning Procedure:Prehabituation (Day 0):

1. Weigh rats. Remove ad libitum water bottles toprovide 23.5 h of water deprivation prior to the startof the habituation phase on the following day.

Habituation Phase (Days 1–14):

2. Fill one set of 50 ml Nalgene tubes with plain water(e.g. distilled, filtered, or tap), and allow the water tocome to room temperature.

AVIOR

FIGURE 28.9 Mean consumption during one- and two-bottle final aversion tests for Fischer 344 and Lewis conditioned with (L)-

U50,488H, a kappa opioid receptor agonist, or SNC80, a delta opioid receptor agonist. All rats received four saccharin–drug pairings and thenreceived a one-bottle final aversion test (panels (a) and (d)). The drug-induced decreases in saccharin consumption were weak and did not differby dose; thus, a two-bottle aversion test followed a fifth conditioning cycle. On the two-bottle test, taste aversions following conditioning with(�)-U50,488H were robust and differed by dose but not by strain (panel (b)). Rats conditioned with 0.90 (#) or 1.6 mg/kg (^) consumedsignificantly less saccharin than rats receiving 0.0 or 0.28 mg/kg (–)-U50, 488H. For SNC80-conditioned rats, the two-bottle test revealed a morerobust decrease in saccharin consumption that remained dose independent. *Denotes all groups consuming significantly less saccharin than ratsreceiving 0.0 mg/kg; see panel (e). Panels (c) and (f) depict total fluid consumption on the two-bottle test for both strains; no differences wereevident in overall fluid consumption, even though significant aversions to the saccharin solution were apparent. Figure redrawn from Ref. 127.

THE FLAVOR-CONDITIONING PROCEDURE 701

3. Weigh rats.4. Once the water has reached room temperature,

invert the first Nalgene tube. Allow the air bubblesto escape, and record the preconsumption volume.Attach this bottle to the cage of the first rat. Startstopwatch or timer. Invert the next Nalgene tube;allow air bubbles to escape, and recordpreconsumption volume.

5. Once 10–15 s has passed, attach the second waterbottle to the second rat’s cage. Continue in thismanner until all rats have water bottles. Exit thedrinking area with the stopwatch or timer.Note: A standard time interval should occurbetween the placements of each water bottle, suchthat each subject receives a full 20 minconsumption period. This is easily done by placingwater bottles on individual cages at 10–15 sintervals. Depending on the number of subjects, thetotal experimental time could exceed 20 min due tothe staggered drinking periods; however, thistechnique ensures that each subject receives the full

VIII. BEH

time to drink regardless of when the water bottle isplaced on the cage.

6. Once 20 min has elapsed, return to the drinking areaand gently remove the first Nalgene tube from thefirst rat’s cage; avoid spilling or dripping any waterfrom the tube. Record the postconsumption volume.Wait 10–15 s, and remove the next bottle; record thepostconsumption volume. Repeat until all bottleshave been removed.

7. Empty bottles, and rinse with fresh tap water. Storebottles inverted, and allow them to air dry.

8. Repeat steps 2–7 on each day of the habituationphase. Monitor body weights and daily waterintake. Note: If any subject is not adapting to thehabituation schedule and is showing signs ofdistress and/or illness, the animal should beremoved from the study and given ad libitumaccess to water, in addition to softened food andany other necessary veterinary care.

9. After approximately 7 days, water consumption bymost subjects will begin to stabilize. Acquire 3-day

AVIOR

28. ANIMAL MODELS OF DRUG ABUSE: CPP AND CTA702

averages for each subject, and compare them to thatday’s consumption values. When average dailyconsumption does not vary by more than �2 ml,water consumption is stable.

Conditioning Phase:Conditioning Cycle Day 1 (Day 15):

10. On the morning of the first conditioning day,dissolve saccharin in plain waterduse the same typeof water that was used during habituation (e.g.distilled, filtered, or tap). Fill the second set ofNalgene tubes with the saccharin; let the bottles sit towarm to room temperature.Note: A different set of bottles and stoppers must beused for saccharin access to avoid exposing thesubjects to the saccharin flavor on water recoverydays (or habituation days in a future study). Labelingthe saccharin bottles prior to the start of conditioningshould be done to avoid any cross-contamination.Further, once the saccharin bottles and stoppers havebeen used for saccharin access, they should never beused for plain water and should not be stored withthe water-only bottles and stoppers.

11. Weigh rats; prepare drugs and syringes.

12. Similar to the habituation phase, invert the firstNalgene tube, allow the air bubbles to escape, thenrecord preconsumption volume, and place the bottleon the cage of the first rat. Start the stopwatch ortimer.

13. Invert the second Nalgene tube and repeat, placingthe second bottle on the next rat’s cage 10–15 s afterplacement of the first bottle.

14. Continue until all rats have saccharin bottles on theircages. Exit the drinking area with the stopwatch ortimer.

15. Return to the drinking area after 20 min, and removethe first rat’s saccharin bottle. Record thepostconsumption volume. Wait 10–15 s, thenremove the next rat’s saccharin bottle and record thepostconsumption volume. Repeat every 10–15 s toremove all bottles.

16. Rank subjects by saccharin consumption fromgreatest to least. Denoting each dose group bya different letter (e.g. high dose¼ 1 and control¼ 0),assign each subject a dose value in ascending order(e.g. 0, 1, and 2); after the last dose value, repeat thevalues in descending order (e.g. 2, 1, and 0). In thisway, average saccharin consumption across allgroups on the first conditioning day will be similar.Note: This ranking procedure typically takes about3–5 min; use of Microsoft Excel is highlyrecommended. If an experimenter allots 5 min tothis ranking procedure, drug injections will notbegin until 5 min after the saccharin access period. It

VIII. BEH

is important that this time interval be maintained onsubsequent conditioning days when the rankingprocedure is not completed (because groupassignment has been completed), such that theCS–US interval is consistent across conditioningdays. If an experimenter does not allot a specifiedperiod of time to ranking, it is recommended thatthe stopwatch or timer be used to time theranking procedure and determine precisely whenthe first injection was administered followingsaccharin access.

17. After the last injection, dispose of needle andsyringes. Empty saccharin bottles, rinse with warmtap water, and invert to dry. Soap should not be usedto clean the saccharin bottles because this couldimpart a taste on the bottle and/or stopper thatcould influence conditioning.

Conditioning Cycle 1: Days 2–4 (Days 16–18):

18. During the water recovery days, no saccharin isconsumed and no injections are administered.Similar to the habituation phase, fill Nalgene tubeswith plain water and let it sit out to warm to roomtemperature.

19. Weigh rats.

20. Place Nalgene bottles on cages as described above;record pre- and postconsumption volumes. Emptyand rinse bottles.

Conditioning Cycles 2–4 (Days 19–30):

21. The second conditioning day is similar to the first inall respects except that there is no rankingprocedure.

22. Prepare saccharin as stated above. Prepare drug andsyringes.

23. Place saccharin on cages, and record thepreconsumption volumes.

24. After 20 min, return to the drinking area and removethe saccharin.

25. Maintaining the same CS–US interval employed onconditioning day 1, inject subjects after the saccharinaccess period.

26. On the following three water recovery days, prepareNalgene tubes with plain water, weigh rats, andrecord the water consumption.

One-Bottle Final Aversion Test (Day 31):

27. On the day following the last water recovery day ofthe final conditioning cycle, prepare saccharin in themorning and fill the Nalgene tubes.

AVIOR

CONCLUSION 703

28. Similar to the conditioning days, place saccharinbottles on the rats’ cages at 10–15 s intervals. Recordpre- and postconsumption volumes. If a two-bottletest will not be run, no injections are necessary andthis is the last day of the taste aversion study.

29. If a two-bottle aversion test will be run, prepare drugand syringes. Inject subjects after saccharin access.Provide plain water during the fluid access periodon the following three daysdthis constitutes a fifthconditioning cycle (Days 31–34).

Two-Bottle Final Aversion Test (Day 35):

30. Twenty-four hours following the last water recoveryday, dissolve saccharin in plain water and fill theNalgene tubes. Additionally, fill the other set ofNalgene tubes with plain water. Let the water inboth sets of tubes warm to room temperature.

31. Invert the first saccharin Nalgene tube, allowbubbles to escape, and record thepreconsumption volume. Invert the first plain-water Nalgene tube, allow bubbles to escape, andrecord the preconsumption volume. Holding onebottle in each hand, present both sipper tubes tothe first rat simultaneously. Attach both tubes tothe first cage.

32. For the second rat, place the saccharin tube on theopposite side of the cage compared to the first rat.More specifically, if the saccharin tube is on the rightside of the first rat’s cage, place the saccharin tube onthe left side of the second rat’s cage. Repeat at10–15 s intervals for the remaining subjects; continueto alter the location of the saccharin bottle from leftto right. Exit the drinking area.Note: The two-bottle test is easier to administer ifthere are two experimentersdone person to invertand place the bottles on the cages, and another torecord the consumption values and monitor thestopwatch or timer.

33. After 20 min, return to the drinking area and removeeach rat’s bottles at the same time. Record thepostconsumption volume for both bottles.

34. Rinse both sets of bottles, and invert to air dry.

CONCLUSION

The conditioned place preference and conditionedtaste aversion procedures are two common methodsfor assessing, respectively, the rewarding and aversiveeffects of drugs of abuse. Both of these procedureshave numerous advantages, including the ease with

VIII. BEH

which the procedures can be performed and the factthat both procedures require relatively inexpensiveequipment. Further, the popularity of these proceduresin the drug abuse field gives investigators a wealth ofdata that can be examined when designing new CPPand/or CTA studies to aid in the selection of parameters,methodology, and other important experimentalvariables. While these procedures might not representevery aspect of drug abuse or addiction, these proce-dures are valuable tools to enable researchers to gaina greater understanding of the effects of drugs on thebrain and behavior and to enable discovery ofpharmaco-therapeutic strategies to treat drug addiction.

Acknowledgments

Preparation of this chapter was supported by a fellowship from theNational Space Biomedical Research Institute (NSBRI) through NASANCC 9-58-PF02602.

References

1. Koob GF, Volkow ND. Neurocircuitry of addiction. Neuro-

psychopharmacology 2010;35(1):217–38.2. Nicholson KL, Ator NA. IACUC perspective on drug addiction

research. ILAR J 2011;52(3):7.3. CDC. AIDS surveilance–general epidemiology L178 slide series (slide

9). Atlanta: CDC Division of HIV/AIDS Prevention, NationalCenter for HIV, STD, and TB Prevention; 2005.

4. CDC. Targeting tobacco use: the Nation’s leading cause of death.Atlanta: CDC National Center for Chronic Disease Preventionand Health Promotion; 2005.

5. Leshner AI. Addiction is a brain disease, and it matters. Science1997;278(5335):45–7.

6. Scicurious. Cocaine and the sexual habits of quail, or, why doesNIH fund what it does? 2011; http://blogs.scientificamerican.com/scicurious-brain/2011/12/28/cocaine-and-the-sexual-habits-of-quail-or-why-does-nih-fund-what-it-does/.

7. Baker M. Animal models: inside the minds of mice and men.Nature 2011;475(7354):123–8.

8. Peca J, Feliciano C, Ting JT, Wang W, Wells MF, Venkatraman TN,et al. Shank3 mutant mice display autistic-like behaviours andstriatal dysfunction. Nature 2011;472(7344):437–42 [ResearchSupport, N.I.H., Extramural Research Support, Non-U.S. Gov’t].

9. Welch JM, Lu J, Rodriguiz RM, Trotta NC, Peca J, Ding JD, et al.Cortico-striatal synaptic defects and OCD-like behaviours inSapap3-mutant mice. Nature 2007;448(7156):894–900.

10. Belin D, Mar AC, Dalley JW, Robbins TW, Everitt BJ. Highimpulsivity predicts the switch to compulsive cocaine-taking[Research Support, Non-U.S. Gov’t]. Science 2008;320(5881):1352–5.

11. Deroche-Gamonet V, Belin D, Piazza PV. Evidence for addiction-like behavior in the rat. Science 2004;305(5686):1014–7.

12. Robinson TE. Neuroscience: addicted rats [Comment]. Science2004;305(5686):951–3.

13. Vanderschuren LJ, Everitt BJ. Drug seeking becomes compul-sive after prolonged cocaine self-administration. Science 2004;305(5686):1017–9.

14. Bardo MT, Bevins RA. Conditioned place preference: what doesit add to our preclinical understanding of drug reward?Psychopharmacology (Berlin) 2000;153(1):31–43.

AVIOR

28. ANIMAL MODELS OF DRUG ABUSE: CPP AND CTA704

15. Bardo MT, Rowlett JK, Harris MJ. Conditioned place preferenceusing opiate and stimulant drugs: a meta-analysis [Meta-Anal-ysis, Research Support, U.S. Gov’t, P.H.S.]. Neurosci Biobehav Rev1995;19(1):39–51.

16. Liu Y, Le Foll B, Wang X, Lu L. Conditioned place preferenceinduced by licit drugs: establishment, extinction, and rein-statement. Sci World J 2008;8:1228–45.

17. Mueller D, de Wit H. Conditioned place preference in rodentsand humans. In: Raber J, editor. Animal models of behavioral

analysis, Neuromethods, vol. 50. Springer; 2011. p. 133–52.18. Riley AL, King HE, Hurwitz ZE. Conditioned Taste Aversion:

An Annotated Bibliography; 2012, from http://ctalearning.com/.

19. Schechter MD, Calcagnetti DJ. Trends in place preferenceconditioning with a cross-indexed bibliography; 1957–1991[Bibliography Review]. Neurosci Biobehav Rev 1993;17(1):21–41.

20. Schechter MD, Calcagnetti DJ. Continued trends in the condi-tioned place preference literature from 1992 to 1996, inclusive,with a cross-indexed bibliography [Bibliography]. Neurosci

Biobehav Rev 1998;22(6):827–46.21. Tzschentke TM. Measuring reward with the conditioned place

preference paradigm: a comprehensive review of drug effects,recent progress and new issues [Research Support, Non-U.S.Gov’t Review]. Prog Neurobiol 1998;56(6):613–72.

22. Tzschentke TM. Measuring reward with the conditioned placepreference (CPP) paradigm: update of the last decade. AddictBiol 2007;12(3–4):227–462.

23. van der Kooy D. Place conditioning: a simple and effectivemethod for assessing the motivational properties of drugs. In:Bozarth MA, editor. Methods of assessing the reinforcing properties

of abused drugs. New York: Springer-Verlag; 1987. p. 229–40.24. Sanchis-Segura C, Spanagel R. Behavioural assessment of drug

reinforcement and addictive features in rodents: an overview[Research Support, Non-U.S. Gov’t Review]. Addict Biol

2006;11(1):2–38.25. Robinson TE, Berridge KC. The neural basis of drug craving: an

incentive-sensitization theory of addiction [Research Support,U.S. Gov’t, P.H.S. Review]. Brain Res Rev 1993;18(3):247–91.

26. Robinson TE, Berridge KC. The psychology and neurobiology ofaddiction: an incentive-sensitization view [Review]. Addiction2000;95(Suppl. 2):S91–117.

27. Robinson TE, Berridge KC. Incentive-sensitization and addic-tion [Review]. Addiction 2001;96(1):103–14.

28. Robinson TE, Berridge KC. Review. The incentive sensitizationtheory of addiction: some current issues [Research Support,N.I.H., Extramural Review]. Phil Trans Roy Soc Lond B Biol Sci

2008;363(1507):3137–46.29. Koob GF, Le Moal M. Neurobiology of addiction. New York:

Academic Press; 2006.30. Everitt BJ, Robbins TW. Neural systems of reinforcement for

drug addiction: from actions to habits to compulsion [ResearchSupport, Non-U.S. Gov’t Review]. Nat Neurosci 2005;8(11):1481–9.

31. Davis CM, Riley AL. Drug preexposure in Fischer (F344) andLewis (LEW) rats: effects on place and taste conditioning. Posterpresented at the College on the Problems of Drug Dependence(CPDD) Annual Meeting, Scottsdale, AZ; 2006.

32. Davis CM, Riley AL. Conditioned taste aversion learning:implications for animal models of drug abuse. Ann N YAcad Sci2010;1187:247–75.

33. Riley AL. The paradox of drug taking: the role of the aversiveeffects of drugs [Research Support, Non-U.S. Gov’t Review].Physiol Behav 2011;103(1):69–78.

34. Lynch WJ, Carroll ME. Regulation of drug intake [Review]. ExpClin Psychopharmacol 2001;9(2):131–43.

VIII. BEH

35. Riley AL, Davis CM, Roma PG. Strain differences in tasteaversion learning: implications for animal models of drugabuse. In: Reilly S, Schachtman TR, editors. Conditioned tasteaversion: behavioral and neural processes. New York: OxfordUniversity Press; 2009. p. 226–61.

36. Cunningham CL, Gremel CM, Groblewski PA. Drug-inducedconditioned place preference and aversion in mice. Nat Protoc2006;1(4):1662–70.

37. Childs E, de Wit H. Amphetamine-induced place preference inhumans. Biol Psychiatry 2009;65(10):900–4.

38. Garcia J, Kimeldorf DJ, Hunt EL. Spatial avoidance in the rat asa result of exposure to ionizing radiation. Br J Radiol 1957;30(354):318–21.

39. Beach HD. Morphine addiction in rats. Can J Psychol 1957;11(2):104–12.

40. Spragg SDS. Morphine addiction in chimpanzees. Comp Psychol

Monogr 1940;15:1–132.41. Reicher MA, Holman EW. Location preference and flavor

aversion reinforced by amphetamine in rats. Anim Learn Behav

1977;5(4):4.42. Rossi NA, Reid LD. Affective states associated with morphine

injections. Physiol Psychol 1976;4(3):5.43. Davis CM, Roma PG, Dominguez JM, Riley AL. Morphine-

induced place conditioning in Fischer and Lewis rats: acquisi-tion and dose-response in a fully biased procedure. Pharmacol

Biochem Behav 2007;86(3):516–23.44. Simpson GR, Riley AL. Morphine preexposure facilitates

morphine place preference and attenuates morphine tasteaversion. Pharmacol Biochem Behav 2005;80(3):471–9.

45. Verendeev A, Riley AL. Relationship between the rewardingand aversive effects of morphine and amphetamine in indi-vidual subjects. Learn Behav 2011;39(4):399–408.

46. Cunningham CL, Clemans JM, Fidler TL. Injection timingdetermines whether intragastric ethanol produces conditionedplace preference or aversion in mice. Pharmacol Biochem Behav

2002;72(3):659–68.47. Ciccocioppo R, Panocka I, Froldi R, Quitadamo E, Massi M.

Ethanol induces conditioned place preference in geneticallyselected alcohol-preferring rats. Psychopharmacology (Berlin)

1999;141(3):235–41.48. Bardo MT, Neisewander JL. Single-trial conditioned place

preference using intravenous morphine [Research Support, U.S.Gov’t, P.H.S.]. Pharmacol Biochem Behav 1986;25(5):1101–5.

49. Bardo MT, Valone JM, Bevins RA. Locomotion and conditionedplace preference produced by acute intravenous amphetamine:role of dopamine receptors and individual differences inamphetamine self-administration. Psychopharmacology (Berlin)

1999;143(1):39–46.50. Kelley BM, Bandy AL, Middaugh LD. A study examining

intravenous ethanol-conditioned place preference in C57BL/6Jmice. Alcohol Clin Exp Res 1997;21(9):1661–6.

51. Mucha RF, van der Kooy D, O’Shaughnessy M, Bucenieks P.Drug reinforcement studied by the use of place conditioning inrat. Brain Res 1982;243(1):91–105.

52. Roma PG, Riley AL. Apparatus bias and the use of light andtexture in place conditioning. Pharmacol Biochem Behav

2005;82(1):163–9.53. Bevins RA, Cunningham CL. Place conditioning: a methodo-

logical analysis. In: Anderson MJ, editor. Tasks and techniques:a sampling of the methodologies for the investigation of animal

learning, behavior, and cognition. Hauppauge, NY: Nova SciencePublishers; 2006. p. 99–110.

54. Spyraki C, Kazandjian A, Varonos D. Diazepam-induced placepreference conditioning: appetitive and antiaversive properties.Psychopharmacology (Berlin) 1985;87(2):225–32.

AVIOR

CONCLUSION 705

55. Carr GD, Fibiger HC, Phillips AG. Conditioned place preferenceas a measure of drug reward. In: Liebman JM, Cooper SJ,editors. The neuropharmacological basis of reward. New York:Oxford; 1989. p. 264–319.

56. Marglin SH, MacKechnie DK, Mattie ME, Hui YH, Reid LD.Ethanol with small doses of morphine establishes a conditionedplace preference. Alcohol 1988;5(4):309–13.

57. Matsuzawa S, Suzuki T, Misawa M. Conditioned fear stressinduces ethanol-associated place preference in rats. Eur J

Pharmacol 1998;341(2–3):127–30.58. Matsuzawa S, Suzuki T, Misawa M. Ethanol, but not the anxi-

olytic drugs buspirone and diazepam, produces a conditionedplace preference in rats exposed to conditioned fear stress.Pharmacol Biochem Behav 2000;65(2):281–8.

59. Russo SJ, Bolanos CA, Theobald DE, DeCarolis NA, Renthal W,Kumar A, et al. IRS2-Akt pathway in midbrain dopamineneurons regulates behavioral and cellular responses to opiates.Nat Neurosci 2007;10(1):93–9.

60. Hiroi N, Brown JR, Haile CN, Ye H, Greenberg ME, Nestler EJ.FosB mutant mice: loss of chronic cocaine induction of Fos-related proteins and heightened sensitivity to cocaine’spsychomotor and rewarding effects [Research Support, Non-U.S. Gov’t Research Support, U.S. Gov’t, P.H.S.]. Proc Natl

Acad Sci USA 1997;94(19):10397–402.61. Mucha RF, Iversen SD. Reinforcing properties of morphine and

naloxone revealed by conditioned place preferences: a proce-dural examination [Research Support, Non-U.S. Gov’t].Psychopharmacology 1984;82(3):241–7.

62. Parker LA. Place conditioning in a three- or four-choice appa-ratus: role of stimulus novelty in drug-induced place condi-tioning. Behav Neurosci 1992;106(2):294–306.

63. Grakalic I, Schindler CW, Baumann MH, Rice KC, Riley AL.Effects of stress modulation on morphine-induced conditionedplace preferences and plasma corticosterone levels in Fischer,Lewis, and Sprague-Dawley rat strains [Research Support,N.I.H., Intramural Research Support, Non-U.S. Gov’t]. Psycho-pharmacology 2006;189(3):277–86.

64. Parker LA, Tomlinson T, Horn D, Erb SM. Relative strength ofplace conditioning produced by cocaine and morphine assessedin a three-choice paradigm. Learn Motiv 1994;25:83–94.

65. Reid LD, Hunter GA, Beaman CM, Hubbell CL. Toward under-standing ethanol’s capacity to be reinforcing: a conditioned placepreference following injections of ethanol [Research Support,Non-U.S. Gov’t]. Pharmacol Biochem Behav 1985;22(3):483–7.

66. Guitart X, Beitner-Johnson D, Marby DW, Kosten TA, Nestler EJ.Fischer and Lewis rat strains differ in basal levels of neuro-filament proteins and their regulation by chronic morphine inthe mesolimbic dopamine system [Comparative Study ResearchSupport, Non-U.S. Gov’t Research Support, U.S. Gov’t,Non-P.H.S. Research Support, U.S. Gov’t, P.H.S.]. Synapse 1992;12(3):242–53.

67. Isaac WL, Nonneman AJ, Neisewander J, Landers T, Bardo MT.Prefrontal cortex lesions differentially disrupt cocaine-reinforced conditioned place preference but not conditionedtaste aversion. Behav Neurosci 1989;103(2):345–55.

68. Kosten TA, Miserendino MJ, Chi S, Nestler EJ. Fischer andLewis rat strains show differential cocaine effects in condi-tioned place preference and behavioral sensitization but not inlocomotor activity or conditioned taste aversion [ComparativeStudy Research Support, Non-U.S. Gov’t Research Support,U.S. Gov’t, P.H.S.]. J Pharmacol Exp Therapeut 1994;269(1):137–44.

69. Spyraki C, Fibiger HC, Phillips AG. Cocaine-induced placepreference conditioning: lack of effects of neuroleptics and6-hydroxydopamine lesions. Brain Res 1982;253(1–2):195–203.

VIII. BEH

70. Cunningham CL, Dickinson SD, Grahame NJ, Okorn DM,McMullin CS. Genetic differences in cocaine-induced condi-tioned place preference in mice depend on conditioning trialduration [Research Support, U.S. Gov’t, P.H.S.]. Psychopharma-

cology 1999;146(1):73–80.71. Cunningham CL, Prather LK. Conditioning trial duration

affects ethanol-induced conditioned place preference in mice.Anim Learn Behav 1992;20(2):187–94.

72. Bardo MT, Neisewander JL, Pierce RC. Novelty-induced placepreference behavior in rats: effects of opiate and dopaminergicdrugs. Pharmacol Biochem Behav 1989;32(3):683–9.

73. Carr GD, Phillips AG, Fibiger HC. Independence of amphetaminereward from locomotor stimulation demonstrated by conditionedplace preference. Psychopharmacology 1988;94(2):221–6.

74. Hughes RN. Behaviour of male and female rats with free choice oftwoenvironments differing innovelty.AnimBehav 1968;16(1):92–6.

75. Scoles MT, Siegel S. A potential role of saline trials in morphine-induced place-preference conditioning [Research Support, Non-U.S. Gov’t Research Support, U.S. Gov’t, P.H.S.]. Pharmacol

Biochem Behav 1986;25(6):1169–73.76. Freeman KB, Riley AL. The origins of conditioned taste aversion

learning: a historical analysis. In: Reilly S, Schachtman TR,editors. Conditioned taste aversion: behavioral and neural processes.New York: Oxford University Press; 2009. p. 9–36.

77. Sclafani A. Oral and postoral determinants of food reward[Research Support, U.S. Gov’t, P.H.S. Review]. Physiol Behav2004;81(5):773–9.

78. Touzani K, Bodnar RJ, Sclafani A. Neuropharmacology oflearned flavor preferences [Research Support, N.I.H., Extra-mural Review]. Pharmacol Biochem Behav 2010;97(1):55–62.

79. Davis CM, de Brugada I, Riley AL. The role of injection cues inthe production of the morphine preexposure effect in tasteaversion learning [Research Support, Non-U.S. Gov’t]. LearnBehav 2010;38(2):103–10.

80. Garcia J, Kimeldorf DJ, Koelling RA. Conditioned aversion tosaccharin resulting from exposure to gamma radiation. Science1955;122(3160):157–8.

81. Garcia J, McGowan BK, Ervin FR, Koelling RA. Cues: theirrelative effectiveness as a function of the reinforcer. Science

1968;160(3829):794–5.82. McLaurin WA, Scarborough BB. Extension of the interstimulus

interval in saccharin avoidance conditioning. Radiat Res

1963;20:317–24.83. Revusky SH. Aversion to sucrose produced by contingent

X-irradiation: temporal and dosage parameters. J Comp Physiol

Psychol 1968;65(1):17–22.84. Smith JC, Roll DL. Trace conditioning with X-rays as an aversive

stimulus. Psychonomic Sci 1967;9(1):11–2.85. Etscorn F, Stephens R. Establishment of conditioned taste

aversion with a 24-hour CS-UCS interval. Physiol Psychol

1973;1(3):251–3.86. Freeman KB, Riley AL. Cocaine-induced conditioned taste

avoidance over extended conditioned stimulus-unconditionedstimulus intervals. Behav Pharmacol 2005;16(7):591–5.

87. Nachman M. Learned aversion to the taste of lithium chlorideand generalization to other salts. J Comp Physiol Psychol

1963;56:343–9.88. Riley AL, Mastropaolo JP. Long-delay taste aversion learning:

effects of repeated trials and two-bottle testing. Bull Psychon Soc1989;27(2):145–8.

89. Garcia J, Koelling RA. Relation of cue to consequence inavoidance learning. Psychonomic Sci 1966;4:2.

90. Garcia J, Ervin FR. Gustatory-visceral and telereceptor-cutaneous conditioningdadapation in internal and externalmilieus. Commun Behav Biol 1968;1:389–415.

AVIOR

28. ANIMAL MODELS OF DRUG ABUSE: CPP AND CTA706

91. McGowan BK, Hankins WG, Garcia J. Limbic lesions andcontrol of the internal and external environment. Behav Biol

1972;7(6):841–52.92. Klosterhalfen S, Klosterhalfen W. Conditioned taste aversion

and traditional learning. Psychol Res 1985;47(2):71–94.93. Rescorla RA, Wagner AR. A theory of Pavlovian conditioning:

variations in the effectiveness of reinforcement and non-reinforcement. In: Black AH, Prokasy WF, editors. Classical

conditioning II: current research and theory. New York: AppletonCentury Crofts; 1972. p. 64–99.

94. Brackbill RM, Rosenbush SN, Brookshire KH. Acquisition andretention of conditioned taste aversion as a function of the tastequality of the CS. Learn Motiv 1971;2(4):341–50.

95. Garcia J, Ervin FR, Koelling RA. Learning with prolonged delayof reinforcement. Psychonomic Sci 1966;5(3):121–2.

96. Stolerman IP, D’Mello GD. Conditioned taste aversions inducedwith apomorphine and apomorphine analogue in rats. Exp

Brain Res 1979;36:R22–3.97. Nachman M, Ashe JH. Learned taste aversions in rats as

a function of dosage, concentration, and route of administrationof LiCl. Physiol Behav 1973;10(1):73–8.

98. Nachman M, Lester D, Le Magnen J. Alcohol aversion in the rat:behavioral assessment of noxious drug effects. Science

1970;168(3936):1244–6.99. Riley AL, Tuck DL. Conditioned taste aversions: a behavioral

index of toxicity [Comparative Study]. Ann N Y Acad Sci 1985;443:272–92.

100. Colpaert FC, Niemegeers CJ, Kuyps JJ, Janssen PA. Apomor-phine as a discriminative stimulus, and its antagonism byhaloperidol. Eur J Pharmacol 1975;32(02):383–6.

101. D’Mello GD, Stolerman IP. Comparison of the discriminativestimulus properties of cocaine and amphetamine in rats[Comparative Study]. Br J Pharmacol 1977;61(3):415–22.

102. Schuster CR, Thompson T. Self administration of and behavioraldependence on drugs [Review]. Annu Rev Pharmacol 1969;9:483–502.

103. Cappell H, LeBlanc AE. Conditioned aversion to saccharin bysingle administrations of mescaline and D-amphetamine.Psychopharmacologia 1971;22(4):352–6.

104. Deneau G, Yanagita T, Seevers MH. Self-administration ofpsychoactive substances by the monkey. Psychopharmacologia

1969;16(1):30–48.105. Pickens R, Harris WC. Self-administration of d-amphetamine by

rats. Psychopharmacologia 1968;12(2):158–63.106. Freeman KB, Konaklieva MI, Riley AL. Assessment of the

contributions of Naþ channel inhibition and general peripheralaction in cocaine-induced conditioned taste aversion [Compar-ative Study Research Support, Non-U.S. Gov’t]. Pharmacol Bio-chem Behav 2005;80(2):281–8.

107. Serafine KM, Riley AL. Preexposure to cocaine attenuates aver-sions induced by both cocaine and fluoxetine: implications for thebasis of cocaine-induced conditioned taste aversions. PharmacolBiochem Behav 2010;95(2):230–4 [Research Support, Non-U.S.Gov’t].

108. Riley AL. Drug discrimination learning: assessment of opioidreceptor pharmacology. In: Bouton ME, Fanselow MS, editors.Learning,motivation, and cognition: the functional behaviorismofRobert

C. Bolles. Washington, DC: American Psychological Association;1997. p. 225–54.

109. Ferrari CM, O’Connor DA, Riley AL. Cocaine-induced tasteaversions: effect of route of administration. Pharmacol Biochem

Behav 1991;38(2):267–71.110. Busse GD, Freeman KB, Riley AL. The interaction of sex

and route of drug administration in cocaine-induced conditionedtaste aversions. Pharmacol Biochem Behav 2005;81(4):814–20.

VIII. BEH

111. Grigson PS, Cornelius K, Wheeler DS. The suppressive effectsof intraperitoneal cocaine are augmented when evaluated innondeprived rats. Pharmacol Biochem Behav 2001;69(1–2):117–23.

112. Mayer LA, Parker LA. Rewarding and aversive properties of IPand SC cocaine: assessment by place and taste conditioning.Psychopharmacology (Berlin) 1993;112(2–3):189–94.

113. Amit Z, Levitan DE, Brown ZW, Rogan F. Possible involvementof central factors in the mediation of conditioned taste aversion.Neuropharmacology 1977;16(2):121–4.

114. Clegg DJ, Wortman MD, Benoit SC, McOsker CC, Seeley RJ.Comparison of central and peripheral administration of C75 onfood intake, body weight, and conditioned taste aversion[Research Support, Non-U.S. Gov’t Research Support, U.S.Gov’t, P.H.S.]. Diabetes 2002;51(11):3196–201.

115. Greenshaw AJ, Buresova O. Learned taste aversion to saccharinfollowing intraventricular or intraperitoneal administration of D,L-amphetamine. Pharmacol Biochem Behav 1982;17(6):1129–33.

116. Liu C, Sue Grigson P. Mu opioid receptor agonistDAMGO-induced suppression of saccharin intake in Lewis andFischer rats. Brain Res 2005;1064(1–2):155–60.

117. Gomez F, Grigson PS. The suppressive effects of LiCl, sucrose,and drugs of abuse are modulated by sucrose concentration infood-deprived rats. Physiol Behav 1999;67(3):351–7.

118. McDonald RV, Parker LA, Siegel S. Conditioned sucrose aversionsproduced by naloxone-precipitated withdrawal from acutelyadministeredmorphine.PharmacolBiochemBehav1997;58(4):1003–8.

119. Parker LA, Brosseau L. Apomorphine-induced flavor-drugassociations: a dose-response analysis by the taste reactivitytest and the conditioned taste avoidance test. Pharmacol Biochem

Behav 1990;35(3):583–7.120. Nissenbaum JW, Sclafani A. Qualitative differences in poly-

saccharide and sugar tastes in the rat: a two-carbohydrate tastemodel [Research Support, Non-U.S. Gov’t Research Support,U.S. Gov’t, P.H.S.]. Neurosci Biobehav Rev 1987;11(2):187–96.

121. Tordoff MG, Alarcon LK, Lawler MP. Preferences of 14 ratstrains for 17 taste compounds. Physiol Behav 2008;95(3):308–32.

122. Bachmanov AA, Tordoff MG, Beauchamp GK. Sweetenerpreference of C57BL/6ByJ and 129P3/J mice. Chem Senses

2001;26(7):905–13.123. Fuller JL. Single-locus control of saccharin preference in mice.

J Hered 1974;65(1):33–6.124. Kotlus BS, Blizard DA. Measuring gustatory variation in mice:

a short-term fluid-intake test. Physiol Behav 1998;64(1):37–47.125. Reed DR, Li S, Li X, Huang L, Tordoff MG, Starling-Roney R,

et al. Polymorphisms in the taste receptor gene (Tas1r3) regionare associated with saccharin preference in 30 mouse strains.J Neurosci 2004;24(4):938–46.

126. Smith JK, Neill JC, Costall B. The influence of postweaninghousing conditions on drug-induced conditioned taste aversion.Pharmacol Biochem Behav 1998;59(2):379–86.

127. Davis CM, Rice KC, Riley AL. Opiate-agonist induced tasteaversion learning in the Fischer 344 and Lewis inbred ratstrains: evidence for differential mu opioid receptor activation.Pharmacol Biochem Behav 2009;93(4):397–405.

128. Berman RF, Cannon DS. The effect of prior ethanol experienceon ethanol-induced saccharin aversions. Physiol Behav 1974;12(6):1041–4.

129. Risinger FO, Cunningham CL. Genetic differences inethanol-induced conditioned taste aversion after ethanol pre-exposure [Research Support, U.S. Gov’t, P.H.S.]. Alcohol 1995;12(6):535–9.

130. Roma PG, Flint WW, Higley JD, Riley AL. Assessment of theaversive and rewarding effects of alcohol in Fischer and Lewisrats. Psychopharmacology (Berlin) 2006;189(2):187–99.

AVIOR

CONCLUSION 707

131. Davis CM, Riley AL. The effects of cocaine preexposure oncocaine-induced taste aversion learning in Fischer and Lewis ratstrains. Pharmacol Biochem Behav 2007;87(1):198–202.

132. Martin GM, Bechara A, van der Kooy D. Morphine pre-exposure attenuates the aversive properties of opiateswithout preexposure to the aversive properties [ResearchSupport, Non-U.S. Gov’t]. Pharmacol Biochem Behav 1988;30(3):687–92.

133. Grigson PS, Freet CS. The suppressive effects of sucrose andcocaine, but not lithium chloride, are greater in Lewis than inFischer rats: evidence for the reward comparison hypothesis

VIII. BEH

[Comparative Study Research Support, U.S. Gov’t, P.H.S.]. BehavNeurosci 2000;114(2):353–63.

134. Kumar R. Morphine dependence in rats: secondary reinforce-ment from environmental stimuli. Psychopharmacologia 1972;25(4):332–8.

135. Barr GA, Paredes W, Bridger WH. Place conditioning withmorphine and phencyclidine: dose dependent effects. Life Sci1985;36:363–8.

136. Carr GD, White NM. Conditioned place preference from intra-accumbens but not intra-caudate amphetamine injections. LifeSci 1983;33:2551–7.

AVIOR