Brain Research 946 (2002) 214–220www.elsevier.com/ locate/bres

Research report

C onditioned saccharin aversion induced by self-administered cocainenegatively correlates with the rate of cocaine self-administration in

rats*Felipe Gomez

Department of Pharmacology, University of Michigan, Ann Arbor, MI 48109-0632,USA

Accepted 4 April 2002

Abstract

Previous studies showed that the degree of saccharin aversion induced by a self-administered drug correlated with the amount of drugtaken. The present paradigm evaluated the relationship between conditioned saccharin (0.15% w/w) aversion and cocaine self-administration (0.15 mg/kg per injection) when drug intake was limited to 1 mg/kg per session over five taste–drug pairings. The lowdose of self-administered cocaine induced a significant decrease in saccharin intake as a result of the conditioning pairings. In addition,conditioned saccharin intake was significantly correlated with the rate of cocaine self-administration (r520.54; P,0.05). A secondexperiment was designed to control for the effect of cocaine concentration in blood per unit of time. Saccharin was paired with one ofthree doses of non-contingent IV cocaine (0, 1, 3, 5 mg/kg) over five taste–drug pairings. Although, 3 or 5 mg/kg of non-contingent IVcocaine induced a dose-dependent decrease in saccharin intake after five taste–drug pairings, the low dose of 1 mg/kg was not effectivein inducing the effect. Taken together, these results suggest that 1 mg/kg per session of cocaine was effective in inducing saccharinaversion only in rats that self-administered the drug at a high rate. It is suggested that a Pavlovian effect such as conditioned saccharinaversion induced by a self-administered drug may be an alternative tool to investigate the anticipation of addictive drugs. 2002 Elsevier Science B.V. All rights reserved.

Theme: Neural basis of behavior

Topic: Drugs of abuse: cocaine

Keywords: Cocaine; Conditioned taste aversion; Drug self-administration

1 . Introduction havior, seeking behavior has been expressed by the numberof cycle of response on the first lever when cocaine was no

Drug self-administration alone is a limited tool for longer delivered [13]. Although the direct influence of theinvestigating anticipatory phenomena such as craving since drug was eliminated, a disadvantage of this procedure isthe direct effect of the drug influences the response after that operant behavior must be recorded following dis-the first injection. Under such circumstances, the direct continuation of drug delivery (i.e. in extinction). Operanteffect of the drug cannot be dissociated from spontaneous response in the absence of the reinforcer may reflectdrug-seeking behavior. Second-order schedules of rein- resistance to extinguish drug-maintained behavior. Inforcement have been used as an alternative for studying addition, this observation requires the interruption of self-drug-seeking behavior (see Ref. [7] for review). Drug administration sessions during testing.seeking has been evaluated by the response on a ‘seeking In a different approach, Pavlovian conditioned effectslever’ that gives access to a ‘taking lever’ that, in turn, such as increase in heart rate or decrease in skin tempera-controls drug delivery on a fixed ratio 1. In an attempt to ture induced by exposure to drug-associated paraphernaliacompletely isolate drug-seeking from drug-taking be- has been used to estimate human craving [5,19]. There are

few examples where a classical conditioning paradigm hasbeen combined with drug self-administration in animal*Tel.: 11-734-764-3172; fax:11-734-764-7118.

E-mail address: felipeg@umich.edu(F. Gomez). models. Previous studies showed that the degree of sac-

0006-8993/02/$ – see front matter 2002 Elsevier Science B.V. All rights reserved.PI I : S0006-8993( 02 )02886-X

F. Gomez / Brain Research 946 (2002) 214–220 215

charin aversion induced by self-administered cocaine [9] or saccharin was presented twice a day for 10 min prior toapomorphine [20] correlated with the amount of drug every drug self-administration session.taken. Rats that self-administered the most drugs alsoshowed the greatest aversion to saccharin, whereas animals2 .1.2. Apparatuswith low drug intake did not avoid the sweet solution Training was conducted in six operant chambers (Habit-[9,20]. These findings supported the theory that saccharin est, Coulbourn Instruments) equipped with a house light, aaversion induced by self-administered drugs is related to retractable lever, a dipper mechanism, and a pneumaticthe reinforcing properties of these drugs. However, the syringe pump apparatus (IITC Life Science, Inc., Chicago,degree of saccharin aversion could have been simply IL) for drug delivery. Operant chambers were interfacedrelated to the amount of drug administered. The present with IBM computers. Self-administration programs wereparadigm evaluated the relationship between conditioned customized to control stimulus arrays within the operantsaccharin aversion and cocaine self-administration when chamber. They were written using BASIC programmingdrug intake was limited to 1 mg/kg per session across five language for DOS.taste–drug self-administration pairings. In the previousstudy rats that self-administered 2.5 mg/kg per session of2 .1.3. Drugs and solutionscocaine or less (1-h sessions of unlimited access) did not The sucrose-sweetened milk solution used during pre-show saccharin aversion [9]. If the magnitude of saccharin training was prepared with 200 g of Eagle Brand con-aversion were simply dependent on the self-administered densed milk (Eagle Family Foods Inc., Tarrytown, NY) inamount, 1 mg/kg per session would be below the level of 1 l of tap water. When indicated, every operant responseeffectiveness. resulted in a 2 s access to the sweet solution (0.1 ml /

A second experiment was designed to control for the delivery). Cocaine hydrochloride was dissolved in physio-effect of cocaine concentration in blood per unit of time. logical saline at a concentration of 2 mg/ml. TheThe conditioned effect of non-contingent intravenous (IV) pneumatic syringes were set to inject a dose of 0.15 mg/kgcocaine (injected in less than 1 min) on saccharin intake per infusion in less than 1 s. The cocaine was provided bywas assessed. The conditioned effect of 1 mg/kg non- the National Institute on Drug Abuse and saccharin wascontingent cocaine was then compared with the effect obtained from the Sigma Chemical Company (St. Louis,induced by the same amount of self-administered drug. MO). Saccharin solution was prepared in distilled water

By determining the relationship between the reinforcing (0.15% w/w) and presented in the home cages.effects of cocaine and conditioned saccharin aversion,these experiments are intended to demonstrate that a2 .1.4. Procedureclassical conditioned response may reflect the anticipationfor cocaine when the drug is self-administered. 2 .1.4.1. Pre-conditioning training. Operant training and

catheter implantation were conducted as described in arecent publication [9]. The rats were trained to press alever for access to a sweetened milk solution and then

2 . Experiment 1implanted with intra-jugular catheters (Micro-Renathanecatheter 040, Braintree, MA). Two to 3 days after surgery,

2 .1. Methodsthe rats were allowed to respond for sweetened milk in theoperant chamber. The duration of all pre-conditioning

2 .1.1. Subjects sessions was 1 h, and two sessions were scheduled eachTwenty-four male Sprague–Dawley rats (Harlan) weigh- day (7 days per week). All instrumental sessions were

ing 250–350 g were used in this experiment. Animals were conducted twice a day. Prior to every session, the intra-housed in individual cages in a room with a temperature jugular catheters were flushed with a solution of heparin-ranging between 23 and 258C. Lighting was maintained on ized (50 UI) saline. On the first three sessions the flusha 12:12 h light–dark cycle and operant sessions were was followed by an IV administration of 3.5 mg/kgconducted at 1 and 7 h after the onset of the light. Food Chloramphenicol in order to prevent infection and eventualwas restricted to four pellets of dry rodent diet (Lab Diet; collapse of the jugular vein. During these first threeBrendwood, MO) per day in order to maintain a mild state sessions, the fixed ratio was increased by one after everyof deprivation in addition to keeping rats’ weights around 50 milk-reinforced responses until a value of five was300 g. Food was given after the operant session and before reached. This procedure facilitates the acquisition of drugthe offset of the lights. Fluids were presented in inverted self-administration on an FR1 schedule [1]. On the fourthglass bottles with rubber stoppers and stainless steel spouts session, rats responded on a fixed ratio of one lever pressplaced on the top of the cage. Animals had free access to per reinforcer delivery. This fixed ratio was maintained forwater during the pre-conditioning phase. During condition- the rest of the study. During the fifth session, everying, water was made available every day in the afternoon response resulted in the simultaneous delivery of sweetfor a period of 1–2 h (before the offset of the lights) and milk and an injection of 75ml /kg of saline. The sixth

216 F. Gomez / Brain Research 946 (2002) 214–220

pre-training session consisted of the simultaneous deliveryP,0.001) and sessions (F (4,84)52.62, P50.04) wereof sweet milk and an injection of 0.15 m/kg of cocaine significant. The group effect does not provide additional(n519) or 75ml of saline (n55) for every lever press. On information about the difference between these groupsthe following two sessions, rats were allowed to self- since they were created based on that distinction. However,administer cocaine or saline for 1 h in order to determine the effect of the successive self-administration sessionthat cocaine but not saline maintained the instrumental reflects the tendency for the SR and control groups toresponse in absence of milk. decrease operant behavior (Fig. 1). In addition, the absence

of interaction between group and session effects means2 .1.4.2. Conditioning. After the two 1-h sessions of un- that the difference in self-administration rate between FRslimited cocaine self-administration water was removed and SRs cannot be attributed to the conditioning. Finally,overnight prior to taste–drug pairings. Taste–drug con- by the end of the conditioning pairings, none of the ratsditioning consisted of five consecutive pairings of 10 min that received contingent saline completed seven responses.access to saccharin (0.15% w/w) followed by access toseven injections of cocaine (0.15 mg/kg per injection) or 2 .2.2. Saccharin intakesaline (75ml /kg per inj). The first saccharin presentation Saccharin intake in the whole group of cocaine-injectedwas combined with a morning self-administration session. rats decreased (2868%) whereas the saline injected groupSaccharin intake during this first presentation is referred to drank fairly constant amounts of saccharin from theas unconditioned intake. Rats were removed from the beginning (760.7 ml) to the end of the conditioningoperant chamber after having completed the total number (7.460.7 ml). Five pairings of saccharin and 1 mg/kg perof responses. Six hours after the last saccharin–cocaine/ session of self-administered cocaine resulted in a signifi-saline pairing, conditioned intake was recorded during 10 cant decrease in saccharin intake (F (1,22)54.5, P,0.05).min access to saccharin. Animals were maintained on the However there was much variability; the standard devia-schedule of fluid restriction through out the conditioning tion of the conditioned saccharin intake was 48% of thethat is described in the Subjects section. average. When FR, SR and control groups were compared,

rats that showed rapid rates of cocaine self-administration2 .1.5. Statistical analysis decreased saccharin intake following saccharin–drug pair-

Based on the rate of drug self-administration on the last ings. However, no changes in saccharin intake wereconditioning pairing, cocaine-exposed rats were divided observed in rats that responded slowly for cocaine (Fig. 2).into ‘Fast Responding’ (FR) and ‘Slow Responding’ (SR) These effects were assessed by a 332 ANOVA analysisgroups. Subjects that had equal or higher self-administra- that compared the groups on the initial and final pre-tion rates than the median value of the cocaine-exposed sentations. The statistical analysis revealed that the con-population were considered FRs. Those that showed rates ditioning was effective in inducing changes in saccharinlower than the median were considered SRs. Two-way intake among the three groups (F (2,21)59.96,P,0.001).ANOVA was used to compare the rates of operant responseand saccharin intake among FR, SR, and control groups.Total milk and cocaine intake was also compared. Post-hocanalysis using Newman–Keuls tests was applied to de-termine statistical differences among groups. Aversion tosaccharin was established by comparing conditioned intakewith both unconditioned intake and post-training saccharindrinking in vehicle-injected controls. A linear correlationwas calculated between the rate of cocaine self-administra-tion on the final pairing and the conditioned saccharinintake among the cocaine exposed rats. The relationshipbetween unconditioned saccharin intake and both totalcocaine intake and rate of self-administration were alsoevaluated.

2 .2. Results

2 .2.1. Rate of cocaine self-administrationThe rates of operant response of the FR and SR and

Fig. 1. Mean (6S.E.M.) rate of cocaine (0.15 mg/kg per inj) or salinecontrol groups were compared across five conditioning(0.75 ml /kg per inj) self-administration in control, SR and SR groups

sessions. A 335 ANOVA analysis varying groups (FR, SR across five conditioning sessions conducted following saccharin pre-and controls) and self-administration conditioning sessions sentations. Sessions were limited to seven operant responses that allowedrevealed that both the effects of group (F (2,21)513.2, intake of 1 mg/kg of cocaine or 525ml /kg of saline per session.

F. Gomez / Brain Research 946 (2002) 214–220 217

Fig. 3. Conditioned saccharin intake as a function of the rate of cocaineFig. 2. Mean (6S.E.M.) unconditioned (open blocs) vs. conditioned self-administration on the final conditioning session (r520.54;P,0.05).0.15% w/w saccharin intake (gray blocs) in Control, SR and FR groups. Cocaine-exposed rats were divided in FR and SR groups based on the rateOnly the FR group showed a significant cocaine-induced decrease in of self-administration. The median value (dashed line) was the referencesaccharin intake relative to both unconditioned intake and saccharin to separate the groups.intake by saline injected controls. *P,0.05, ** P,0.01, *** P,0.001.

2 .3. DiscussionOnly the FR group decreased saccharin intake significantlyrelative to unconditioned saccharin intake (P,0.01). Con- Five pairings of saccharin with 1 mg/kg per session ofditioned intake in the FR (P,0.001) but not in the SR self-administered cocaine (0.15 mg/kg per inj) weregroup was lower than the amount of saccharin consumed effective at inducing conditioned saccharin aversion. Inby controls. Finally, conditioned saccharin intake was addition, when conditioned saccharin intake was comparedlower in the FR than in the SR group (P,0.05). between rats that took cocaine rapidly (FR) and those that

Conditioned saccharin intake was significantly corre- showed slow rates of self-administration (SR), only the FRlated with the rate of cocaine self-administration on the animals showed saccharin aversion. This observation isfinal saccharin–cocaine pairing. The relationship is defined consistent with our previous report describing individualby a negative linear correlation (r520.54; P,0.05) (Fig. differences in the conditioned effects of self-administered3). No significant correlation was found between un- cocaine on saccharin intake [9]. In that experiment,conditioned saccharin intake and either rate of cocaine cocaine intake was proportional to saccharin aversion. Inself-administration or total cocaine intake (1-h session the present experiment, conditioned saccharin intake wasprior to conditioning). negatively correlated with the rate of cocaine self-adminis-

tration. Considering that in the current paradigm, theamount of cocaine administered by all rats was equal,

2 .2.3. Milk maintained response conditioned effects of the self-administrated drug did notSweet milk available during training of the operant depend on the total intake. Thus, saccharin aversion is

response maintained high rates of responding in all rats. most likely related to the reinforcing effect of the drug asThere was no significant difference in milk intake between indicated by the rate of self-administration.FR and SR groups. A 0.15 mg/kg injection of cocaine The lack of difference in sweet milk intake between FRdelivered with each milk presentation dramatic decreased and SR groups does not support the hypothesis that self-the response (F (4,42)58.11,P50.0001). Newman–Keuls administration of stimulant drugs such as cocaine [10] ortests showed that cocaine induced a significant decrease in amphetamine [4] can be predicted by the avidity forthe milk-maintained response in both FR and SR groups sucrose. The studies that argued in favor of this hypothesiswhen both reinforcers were presented simultaneously (P, fail to establish a direct correlation between cocaine self-0.001). There was no significant difference in the degree of administration and sucrose intake. Similarly, the effect ofdecrease in milk-maintained response between FR and SR. i.v. cocaine on milk-maintained response was not differentIt was observed that during the initial simultaneous between FR and SR groups. I contrast to the conditioneddelivery of cocaine and milk, rats failed to drink the milk. effect of self-administered cocaine on saccharin intake the

218 F. Gomez / Brain Research 946 (2002) 214–220

decrease in milk-maintained responding may reflect a as in Experiment 1 except that the drug was delivereddisruptive effect of novel cocaine on operant behavior. An non-contingently. Ten minutes access to saccharin wasalternative possibility is that cocaine contributed to reach a paired with administration of either cocaine (1 mg/kg,maximal level of reinforcement after a much lower level of n56; 3 mg/kg, n55; 5 mg/kg, n56) or saline (n55).responding compared to the amount of responding for milk Conditioned intake was recorded 6 h after the last taste–than would be necessary to obtain a similar level of drug pairing during a final saccharin presentation.reinforcement. On the other hand, Blavet et al. [2] showedthat i.v. cocaine reduced food intake during the hour that 3 .1.4. Statistical analysisfollowed the injection [2]. Another study demonstrated that Saccharin aversion was determined by comparing con-when a sweet solution and cocaine are concurrently ditioned intake with both unconditioned intake of everyavailable rats do not readily consume the solution during group and the post-training intake of the saline injectedacquisition of cocaine self-administration. However, after a controls. A two-way ANOVA was conducted to evaluate5-day saline substitution period they were able to readily the conditioned effect of the drug. A post-hoc assessmentdrink the sweet solution concurrently with cocaine [3]. of the significant interactions resulting from that analysisTaken together, these findings suggest that novelty to thewas performed in order to determine statistical differencesdrug is at least one of the mechanisms that mediate among the groups.cocaine-induced reduction of milk-reinforced response.

3 .2. Results

3 . Experiment 23 .2.1. Saccharin intake

After four saccharin–cocaine pairings, rats injected withThe above experiment was designed to evaluate the3- and 5-, but not 1-mg/kg IV doses of cocaine decreasedeffect of self-administered cocaine on conditioned sac-their saccharin intake (Fig. 1). A 432 ANOVA reveled acharin intake when all experimental subjects consumed thesignificant interaction between the effects of dose andsame amount of drug. However, the amount of drug thatconditioning (F (3,18)53.55, P,0.05). Newman–Keulsreaches the blood stream in a given time depends on thetests showed that only the groups injected with 3 and 5rate of self-administration. Animals with faster self-ad-mg/kg of cocaine significantly decreased saccharin intakeministration rates should reach a higher plasma concen-compared with saline injected controls (P,0.05). Whentration of drug per unit of time. In order to rule out thecompared with their unconditioned intake, only the ani-possibility that 1 mg/kg of cocaine would induce saccharinmals treated with the highest dose (5 mg/kg) showed aaversion only when it was administered rapidly, saccharinsignificant decrease (P,0.05). Five doses of 1 mg/kgwas paired with three doses of acutely administeredinjected in approximately 3 s were not effective incocaine (delivered in less than 1 min).suppressing saccharin intake either relative to precon-ditioned intake or compared with saline-injected controls at3 .1. Methodstest. Standard deviation in conditioned saccharin intakewas less than 14% of the average when the lower dose of 13 .1.1. Subjects and apparatusmg/kg was used (Fig. 4).Twenty-two male Sprague–Dawley rats (Harlan) weigh-

ing 250–350 g were used in this experiment. They werehoused and fed as described in Experiment 1. Fluids were3 .3. Discussionalso presented in the same manner as described in Experi-ment 1. The effect of non-contingent IV cocaine on conditioned

saccharin intake showed a dose–dependent response in the3 .1.2. Drugs and solutions conditioned taste aversion (CTA) paradigm. The present

Cocaine hydrochloride and saccharin solutions were result is the first report of such a relationship obtained byprepared as described above. The drug injections were intravenous administration of cocaine. Early reportsperformed manually with a 1 ml plastic syringe. All showed a dose-dependent effect of intra-peritoneal (IP)injections were completed in less than 1 min at a rate of injections of cocaine and the resulting conditioned sac-approximately 0.05 ml /s. charin aversion [12]. Ferrari et al. [8] compared the effects

IP and subcutaneous (SC) administration of cocaine after3 .1.3. Procedure four saccharin–drug pairings. Their study showed that 32

Intra jugular catheters were implanted as the first step of and 50 mg/kg of SC cocaine induced a 95 and 98%the protocol. The implantation technique was previously decrease in saccharin intake, respectively. IP administra-described [9]. Following recovery from the surgery, the tion of the same doses was not effective in suppressingrats were placed on water-deprivation as described above. saccharin intake. The authors hypothesized that the effectConditioning consisted of five saccharin–cocaine pairings of the route of administration on conditioned aversion was

F. Gomez / Brain Research 946 (2002) 214–220 219

native hypothesis is that animals having more avidity forthe drug show larger conditioned effects. I have previouslyreported that intake of self-administered cocaine correlatedwith conditioned saccharin aversion [9]. The relationshipbetween saccharin aversion and the reinforcing propertiesof cocaine was not clearly established since animals thattook larger amounts of the drug also showed largerconditioned effects. In the present experiment, all subjectsreceived the same amount of drug and yet the conditionedeffects not only differed across individuals but also corre-lated with rates of self-administration. Taken together thepresent and previous findings [9] suggest that self-adminis-tered cocaine was effective of inducing saccharin aversionin rats that not only would self-administered the drug fasterbut also would take larger amounts of it during unlimited1-h sessions.

Most studies on conditioned taste aversion have beendone using non-contingent administration of the uncon-ditioned stimulus (e.g. drug injections) (see Ref. [18] for

Fig. 4. Mean (6S.E.M.) unconditioned (open blocs) vs. conditioned review). The fact that addictive drugs are capable of0.15% w/w saccharin intake (gray blocs) in rats injected with IV cocaine

inducing CTA has been described as a paradox [11,14].(0, 1, 3, 5 mg/kg per session). Only the highest dose of cocaine (5 mg/kgThere is no consensus among the theories that have beenper s) induced saccharin aversion relative to the unconditioned intake ofadvanced to explain this paradox. On one hand, saccharinthe group and post-training intake in saline injected controls. *P,0.05.

aversion induced by cocaine has been attributed to theanxiogenic effect of the drug [6]. In contrast, several

due to the longer duration of action following SC adminis- studies showed that when saccharin aversion was inducedtration [8]. by cocaine, the avoidance was accompanied by a ‘consum-

Although there are several studies that have addressed matory-like’ pattern of behavior whereas when aversionthe effects of cocaine on conditioned saccharin intake, was induced by a toxic substance (lithium chloride) anthese paradigms do not provide information about the ‘aversive-like’ pattern was observed. This characterizationconcentration of the drug in the blood per unit of time. The was defined according to the criteria of ‘taste reactivity’present results suggest that five IV injections of non- [16,17]. The current results do not provide a solution forcontingent cocaine (1 mg/kg) were not sufficient to cause the paradox. Although, they show that cocaine induced anconditioned saccharin aversion. The poor effectiveness of aversion to saccharin, it is clear that the conditionedthis low dose on every cocaine treated subject is supported behavior does not result from an aversion to cocaine sinceby the fact that the standard deviation of the average the drug is readily self-administered. In contrast, there isconditioned saccharin intake was very narrow. Rats that evidence to suggest that in some cases CTA induced byself-administered 1 mg/kg per session cocaine at high non-contingent administration of an addictive drug mayrates developed taste aversion whereas those that received indeed reflect an aversion to the drug. For example, Fisherthe same non-contingent amount at a much higher rate did rats (the F344/N strain) that do not readily acquirenot. morphine self-administration do develop a robust aversion

to saccharin when paired with non-contingent morphineinjections [15].

4 . General discussion The fact that a conditioned effect correlates with the rateof cocaine self-administration means that the magnitude of

The results of this study showed that a small amount of the conditioned behavior could be used to estimate theself-administered cocaine was capable of inducing sac- reinforcing effect of the drug when it is not contaminatedcharin aversion after five saccharin–cocaine pairings in by its direct effect. As described in the Introduction, therats that showed high rates of self-administration. In reaction to certain objects associated with drug consump-addition, rate of self-administration significantly correlated tion has been used to estimate craving in human patients.with conditioned saccharin intake. Although this observa- In those patients conditioned effects were expressed astion suggests that the conditioned effect may depends on physiological responses such as increase in heart rate orthe rate of administration, five non-contingent ‘bolus’ decrease in skin temperature [5,19]. The present studycocaine injections (1 mg/kg administered approximately opens the possibility to explore a similar model in rats. It is56 times faster than the average rate of the FR group) did likely that other conditioned effects in addition to sac-not induce saccharin aversion in any subject. An alter- charin aversion induced by self-administered cocaine can

220 F. Gomez / Brain Research 946 (2002) 214–220

[7] B.J. Everitt, T.W. Robbins, Second-order schedules of drug re-also be identified in rats. Thus, the combination of classicalinforcement in rats and monkeys: measurement of reinforcingconditioning and drug self-administration paradigms mayefficacy and drug-seeking behaviour, Psychopharmacology 153

be a suitable way to further explore the reinforcing effects (2000) 17–30.of addictive drugs. [8] C.M. Ferrari, B.F. O’Connor, A.L. Riley, Cocaine-induced taste

aversion: effects of route of administration, Pharmacol. Biochem.Behav. 38 (1991) 267–271.

[9] F. Gomez, Induction of conditioned taste aversion with a self-A cknowledgementsadministered substance in rats, Brain. Res. Prot. 8 (2001) 137–142.

[10] B.A. Gosnell, Sucrose intake predicts rate of acquisition of cocaineThis research was supported by the US Public Health self-administration, Psychopharmacology 149 (2000) 286–292.

Service Grant DA 12933-01. I thank the National Institute [11] A.J. Goudie, Aversive stimulus properties of drugs, Neurophar-macology 18 (1979) 971–979.on Drug Abuse for generously providing the cocaine

[12] A.J. Goudie, D.W. Dickins, E.W. Thornton, Cocaine-induced con-hydrochloride. I am also grateful for the valuable helpditioned taste aversions in rats, Pharmacol. Biochem. Behav. 8provided by Dr Gail Winger and Graham Flory during the(1978) 757–761.

writing of this manuscript. [13] D.M. Hutcheson, B.J. Everitt, T.W. Robbins, A. Dickinson, The roleof withdrawal in heroin addiction: enhances reward or promotesavoidance?, Nat. Neurosci. 4 (2001) 943–947.

[14] T. Hunt, Z. Amit, Conditioned taste aversion induced by self-R eferencesadministered drugs: paradox revisited, Neurosci. Biobehav. Rev. 11(1987) 107–130.

[1] T.J. Baird, S.X. Deng, D.W. Landry, G. Winger, J.H. Woods, Natural [15] D. Lancellotti, B.M. Bayer, J.R. Glowa, R.A. Houghtling, A.L.and artificial enzymes against cocaine. I. Monoclonal antibody Riley, Morphine-induced conditioned taste aversions in the LEW/N15A10 and the reinforcing effects of cocaine in rats, J. Pharmacol. and F344/N rat strains, Pharmacol. Biochem. Behav. 68 (2001)Exp. Ther. 295 (2000) 1127–1134. 603–610.

[2] N. Blavet, F.V. DeFeudis, F. Clostre, Studies on food intake in the [16] L.A. Parker, Taste reactivity responses induced by cocaine-,fasted rat, Gen. Pharmacol. 13 (1982) 293–297. phencyclidine-, and methamphetamine-paired sucrose solutions,

[3] M.E. Carroll, S.T. Lac, S.L. Nygaard, A concurrent available non- Pharm. Biochem. Behav. 40 (1993) 983–986.drug reinforcer prevents the acquisition or decreases the mainte-

[17] L.A. Parker, Rewarding drugs produce taste avoidance, but not tastenance of cocaine-reinforced behavior, Psychopharmacology 97

aversion, Neurosci. Behav. Rev. 19 (1995) 143–151.(1989) 23–29.

[18] A.L. Riley, D.L. Tuck, Conditioned taste aversions: a behavioral[4] N.J. De Sousa, D.E.A. Bush, Vaccarino, individual differences in

index of toxicity, Ann. NY Acad. Sci. 443 (1985) 100–109.sucrose intake are predictive of intravenous amphetamine self-

[19] S.J. Robbins, R.N. Ehrman, A.R. Childress, C.P. O’Brien, Using cueadministration levels, Soc. Neurosci. Abstr. 22 (1996) 705.

reactivity to screen medications for cocaine abuse: a test of[5] R.N. Ehrman, S.J. Robbins, A.R. Childress, C.P. O’Brian, Con-

amantadine hydrochloride, Addict. Behav. 17 (1992) 491–499.ditioned responses to cocaine-related stimuli in cocaine abuse

[20] R.A. Wise, R.A. Yokel, H. DeWit, Both positive reinforcement andpatients, Psychopharmacology 107 (1992) 523–529.

conditioned aversion from amphetamine and from apomorphine in[6] A. Ettenberg, T.D. Geist, Animal model for investigating the

rats, Science 191 (1976) 1273–1275.anxiogenic effects of self-administered cocaine, Psychopharmacolo-gy 103 (1991) 455–461.