ARCHIVAL REPORT

Preference for Distinct Functional Conformations ofthe Dopamine Transporter Alters the Relationshipbetween Subjective Effects of Cocaine andStimulation of Mesolimbic Dopamine

Stephen J. Kohut, Takato Hiranita, Soo-Kyung Hong, Aaron L. Ebbs, Valeria Tronci,Jennifer Green, Linda Garcés-Ramírez, Lauren E. Chun, Maddalena Mereu, Amy H. Newman,Jonathan L. Katz, and Gianluigi Tanda

Background: Subjective effects of cocaine are mediated primarily by dopamine (DA) transporter (DAT) blockade. The present studyassessed the hypothesis that different DAT conformational equilibria regulate differences in cocaine-like subjective effects andextracellular DA induced by diverse DA-uptake inhibitors (DUIs).

Methods: The relationship between cocaine-like subjective effects and stimulation of mesolimbic DA levels by standard DUIs (cocaine,methylphenidate, WIN35,428) and atypical DUIs (benztropine analogs: AHN1-055, AHN2-005, JHW007) was investigated using cocainediscrimination and DA microdialysis procedures in rats.

Results: All drugs stimulated DA levels with different maxima and time courses. Standard DUIs, which preferentially bind outward-facingDAT conformations, fully substituted for cocaine, consistently producing cocaine-like subjective effects at DA levels of 100–125% overbasal values, regardless of dose or pretreatment time. The atypical DUIs, with DAT binding minimally affected by DAT conformation,produced inconsistent cocaine-like subjective effects. Full effects were obtained, if at all, only at a few doses and pretreatment times andat DA levels 600–700% greater than basal values. Importantly, the linear, time-independent, relationship between cocaine-like subjectiveeffects and DA stimulation obtained with standard DUIs was not obtained with the atypical DUIs.

Conclusions: These results suggest a time-related desensitization process underlying the reduced cocaine subjective effects of atypicalDUIs that may be differentially induced by the binding modalities identified using molecular approaches. Since the DAT is the target ofseveral drugs for treating neuropsychiatric disorders, such as attention-deficit/hyperactivity disorder, these results help to identify safeand effective medications with minimal cocaine-like subjective effects that contribute to abuse liability.

Key Words: ADHD, benztropine analogs, cocaine discrimination,dopamine microdialysis, drug abuse and addiction, nucleusaccumbens shell

Most illicit drugs produce effects described as euphoric,and these effects are assumed to underlie substanceabuse disorders (1). The study of these subjective effects

in the laboratory is often accomplished with drug discriminationprocedures, and several previous studies have indicated that thediscriminative stimulus effects of drugs are closely related to theirsubjective effects in humans (2,3). The subjective, reinforcing, and

Address correspondence to Gianluigi Tanda, Ph.D., National Institute onDrug Abuse/Intramural Research Program/National Institutes of Health/Department of Health and Human Services, Medication DevelopmentProgram, NIDA Suite 200, 251 Bayview Blvd., Baltimore, MD 21224;E-mail: gtanda@mail.nih.gov.

From the Psychobiology (SJK, TH, S-KH, ALE, VT, JG, LG-R, LEC, MM, JLK,GT) and Medicinal Chemistry Sections (AHN) and Medications Devel-opment Program (AHN, GT), Molecular Targets and MedicationsDiscovery Branch, National Institute on Drug Abuse/IntramuralResearch Program/National Institutes of Health/Department of Healthand Human Services, Baltimore, Maryland; Department of BiologicalSciences (S-KH), Korea Advanced Institute of Science and Technology,Daejeon, South Korea; and Escuela Nacional de Ciencias Biológicas(LG-R), Instituto Politécnico Nacional, Mexico City, México.

Received Aug 7, 2013; revised Mar 8, 2014; accepted Mar 13, 2014.

0006-3223/$36.00http://dx.doi.org/10.1016/j.biopsych.2014.03.031 Publis

discriminative stimulus effects of psychomotor stimulants inhumans and laboratory animals are thought to involve increaseddopamine (DA) transmission in the mesolimbic system (4–8).In human subjects, a relationship between the subjective effectsof cocaine and DA transporter (DAT) occupancy in striatal areashas been demonstrated (9). In animals, several studies havedemonstrated a relationship between the reinforcing and dis-criminative stimulus effects of stimulants and their binding to theDAT (7,8,10–12), and these effects are thought to involveincreased DA transmission in the nucleus accumbens (4–8).

Certain structural classes of DA uptake inhibitors (DUIs), inclu-ding benztropine (BZT) analogs (13), are atypical with regard totheir DAT binding and are less efficacious than cocaine in produ-cing many of its characteristic effects (14–17). Standard DUIs, suchas cocaine, bind more potently to the DAT when it is in aconformation open to the extracellular space compared with aconformation open to the cytosol. In contrast, atypical DUIs aremore tolerant to DAT conformational status; potency of thesedrugs is decreased less compared with cocaine with the DAT inan inward, cytosol-facing conformation (13). Further, it has beensuggested that differences in the inward/outward conformationalequilibrium of the DAT might underlie the slower DAT associationand dissociation rates for atypical DUIs (13), which could contrib-ute to their atypical effects (17,18).

The present study explored the possibility that a preference forbinding the outward-facing versus inward-facing DAT conformationcould influence the potential of DUIs to produce cocaine-likesubjective effects. We assessed the cocaine-like discriminative

BIOL PSYCHIATRY 2014;76:802–809hed by Elsevier Inc on behalf of Society of Biological Psychiatry

S.J. Kohut et al. BIOL PSYCHIATRY 2014;76:802–809 803

stimulus effects of three cocaine-like DUIs: cocaine, its analogWIN35,428, and methylphenidate (a Food and Drug Administra-tion approved medication for treatment of attention-deficit/hyperactivity disorder [ADHD]) (19). Each of these compoundshas been shown to preferentially bind the outward-facing con-formation of the DAT (13,20,21). Additionally, three atypical DUIs(17), analogs of BZT (22,23), AHN1-055, AHN2-005 (a potentialADHD medication) (24,25), and JHW007, were assessed. Thesecompounds are relatively insensitive to a change in DAT con-formation (13,20). We compared these drugs across a wide rangeof doses and times after injection so that their effects could becompared under conditions in which the drugs similarly stimu-lated DA levels in the nucleus accumbens shell (NAS), assessed bybrain microdialysis (26–29), as previous studies have implicatedthis area as critically involved in the discriminative stimuluseffects of stimulant drugs (4,5,8).

Methods and Materials

SubjectsExperimentally naïve male Sprague-Dawley rats, 275 g to 350 g,

were habituated for at least 1 week before experiments and main-tained in an environmentally controlled vivarium. Experimentswere conducted during the light phase of a 12-hour cycle (seeSupplement 1 for detailed procedures).

In Vivo Brain MicrodialysisProbes had an active dialyzing surface of 1.8 mm to 2.0 mm

and were implanted during surgical procedures [uncorrectedcoordinates (30): anterior ¼ �2.0 mm and lateral ¼ �1.0 mmfrom bregma; vertical ¼ �7.9 mm from dura (see Figure S1 inSupplement 1 for probe placements)] under a mixture of ket-amine and xylazine anesthesia, 60.0 and 12.0 mg/kg intraperito-neally (IP), respectively, as described previously (31–33).

Experiments were performed on freely moving rats approximately22 to 24 hours after probe implant. Dialysates were sampled every10 minutes and immediately analyzed. After reaching stable DAvalues (three consecutive samples, �10% variability), rats weretreated with drugs. Cerebrospinal fluid (CSF) (10 mL) was sampledevery 10 minutes for the first 2 hours and every 20 minutesthereafter for 4 hours, after which CSF was sampled every 30minutes. Because some BZT analogs had long-lasting effects, CSF wassampled as necessary up to 27 hours after drug or saline injection.

Dopamine was detected in dialysate samples by high-performance liquid chromatography coupled with a coulometricdetector (5200a Coulochem II or III; ESA, Chelmsford, Massachu-setts). The average basal DA values in dialysates in the presentexperiments were 51.5 � 3.1 fmol (�SEM) in a 10 mL sample,n = 161. No significant differences were found in basal DAconcentrations from the different experimental groups (analysisof variance [ANOVA], F29,131 = 1.066, p = .39).

Drug-Discrimination StudiesRats were trained during daily sessions in operant conditioning

chambers to press one lever after cocaine (10 mg/kg, IP) and theother after saline (IP) injection, both administered 5 minutesbefore the session. The 20th consecutive response produced afood pellet (fixed ratio [FR]), and the right versus left assignmentof cocaine- and saline-associated levers was counterbalancedamong subjects.

Sessions started at various times after injection with a5-minute time-out during which responses had no consequences.

The time-out was followed by illumination of chamber lights untilthe completion of the FR requirement and delivery of a foodpellet. A 20-second time-out followed each food pellet, andsessions ended after 20 food presentations or 15 minutes,whichever comes first. Cocaine or saline training sessions werescheduled in a mixed sequence and continued until subjects metthe criteria of four consecutive sessions with �85% cocaine-appropriate or saline-appropriate responses during the entiresession and the first FR. After meeting these criteria, testingbegan with different doses of cocaine, standard DUIs (methyl-phenidate, WIN35,428), or atypical DUIs (BZT analogs: AHN1-055,AHN2-005, JHW007) each administered at various times beforesessions. Test sessions were identical to training sessions with theexception that completion of the FR requirement on either keywas reinforced.

DrugsThe drugs tested were (-)-cocaine HCl (1–30 mg/kg), methyl-

phenidate (.3–17 mg/kg), WIN35,428 (.03–3 mg/kg), AHN1-055(.3–10 mg/kg), AHN2-005 (1–30 mg/kg), and JHW007 [synthesizedas per Agoston et al. (22) and Newman et al. (23), 1–17 mg/kg].Drugs were dissolved in .9% sodium chloride or sterile water withheat and sonication, as necessary, and were injected IP (1.0 mL/kgof body weight).

Data AnalysesMicrodialysis results were expressed as a percentage of basal

DA values. Statistical analysis was carried out using one-way ortwo-way ANOVA (factors: time, dose) for repeated measures overtime with significant results subjected to post hoc Tukey's test.For drug discrimination, percentages of cocaine lever responseswere calculated by dividing the total responses on the cocaine-associated lever by total number of responses on both levers. Fullsubstitution was defined as �85% and partial substitution wasdefined as between 85% and 15% cocaine-appropriate respond-ing. Dose-effect curves for percent cocaine lever responding wereanalyzed using standard linear regression techniques, from whicheffective dose 50 (ED50) values with 95% confidence limits werecalculated (34).

To examine the relation between levels of extracellular DA andcocaine-like discriminative stimulus effects, linear regressionanalyses were performed on the percent cocaine-lever respond-ing as a function of the percent change in extracellular DA. Thepercentage of basal DA levels were calculated from levels in10-minute bins during the time period most closely correspondingto the duration of the drug discrimination sessions (15 minutesplus the 5-minute presession time-out) for each subject. Data from0% to the lowest dose producing $85% cocaine-lever respondingwere regressed on percent change in DA levels and the slopesof the best fit lines were compared across pretreatment timesfor each drug and between cocaine and each of the drugs usingthe extra sum of squares F-test. Further details are presented inSupplement 1.

Results

Microdialysis StudiesEffects of the Standard DUIs. Cocaine (1.0–30 mg/kg, IP)

dose-dependently and significantly stimulated DA levels indialysates from the NAS (Figure 1A). Two-way ANOVA indicatedsignificant (p � .001) main effects of dose (F5,31 ¼ 38.0), time(F18,558 ¼ 38.6), and their interaction (F90,558 ¼ 11.0). At the

www.sobp.org/journal

Figure 1. Top panels show the effects of standard dopamine transporter blockers, cocaine (1.0, 3, 10, 17, and 30 mg/kg intraperitoneal [IP])(A), methylphenidate (1.0, 3, 5.6, and 10 mg/kg IP) (B), and WIN35,428 (.1, .3, 1.0, and 3.0 mg/kg IP) (C), and bottom panels show the effects ofatypical dopamine transporter blockers, AHN1-055 (1.0, 3, and 10 mg/kg IP) (D), AHN2-005 (1.0, 3, 10, 17, and 30 mg/kg IP) (E), and JHW007 (1.0, 3, 10, and17 mg/kg IP) (F) on extracellular dopamine levels in dialysates from rats implanted with microdialysis probes in the nucleus accumbens shell. Results aremeans, with vertical bars representing SEM, of the amount of dopamine in 10-minute dialysate samples, expressed as percentage of basal values. Groupsizes (n) were 5, 8, 6, 6, 9 for cocaine doses 1.0, 3, 10, 17, and 30 mg/kg, respectively; 5, 4, 4, 7 for methylphenidate doses 1.0, 3, 5.6, and 10 mg/kg,respectively; 5, 5, 5, 5 for WIN35,428 doses .1, .3, 1.0, and 3.0 mg/kg, respectively; 4, 4, 6 for AHN1-055 doses 1.0, 3, and 10 mg/kg, respectively; 5, 7, 5, 6, 6 forAHN2-005 doses 1.0, 3, 10, 17, and 30 mg/kg, respectively; 6, 5, 4, 6 for JHW007 doses 1.0, 3, 10, and 17 mg/kg, respectively; and n ¼ 5 for saline, 1 mL/kg IP.

804 BIOL PSYCHIATRY 2014;76:802–809 S.J. Kohut et al.

highest dose, DA levels increased rapidly at a rate of about 40%per minute, reached a �800% maximum after 20 minutes, andapproached basal values at about 4 hours (Figure 1A). Figure S2 inSupplement 1 shows that stimulation of NAS DA levels by cocainewas not significantly different in naïve rats and those discrim-inating cocaine under these different conditions.

Similar to cocaine, methylphenidate (1.0–10 mg/kg) andWIN35,428 (.1–3.0 mg/kg) significantly and dose-dependentlyincreased DA levels (Figure 1B,C). Two-way ANOVA revealedsignificant (p � .001) main effects of dose (F4,20 ¼ 13.9), time(F15,300 ¼ 17.0), and their interaction (F60,300 ¼ 4.3); there weresimilarly significant (p � .001) main effects of WIN35,428 dose(F4,20 ¼ 183.9), time (F18,360 ¼ 69.0) and their interaction (F72,360 ¼35.8). Methylphenidate at the highest dose produced a rapidincrease in DA levels at a rate of about 30% per minute to amaximum of �630% at 20 minutes and a return to basal values in�140 minutes (Figure 1B). WIN35,428 was the most potent andeffective of the standard DUIs. At the highest dose of WIN35,428,DA levels increased at a rate of over 30% per minute and reacheda maximum of 980% at �30 minutes after injection, with effectslasting up to about 5 hours (Figure 1C).

Effects of Atypical DUIs. AHN1-055 (1.0–10 mg/kg, IP) dose-dependently increased DA levels (Figure 1D). Two-way ANOVAindicated significant (p � .001) main effects of dose (F3,15 ¼ 31.4),time (F18,270 ¼ 12.0), and their interaction (F54,270 ¼ 7.22). At 10mg/kg of AHN1-055, DA levels initially increased at a rate of�16% per minute and reached a maximum of 800% at �90

www.sobp.org/journal

minutes after injection, but in contrast to the standard DUIs, itseffects lasted for up to 7 hours. Levels of DA returned topretreatment levels within 24 hours (p � .05; Figure 1D).

AHN2-005 (1.0–30 mg/kg) dose-dependently increasedDA levels (Figure 1E). Two-way ANOVA showed significant(p � .001) main effects of dose (F5,28 ¼ 27.3), time (F18,504 ¼ 30.5),and their interaction (F90,504 ¼ 7.44). DA values increased at a rate ofabout 27% per minute and reached a maximum of �700% at about20 to 30 minutes after the highest dose, with effects lasting longerthan 24 hours after injection (Figure 1E).

JHW007 (1.0–17 mg/kg) also dose-dependently increasedDA levels (Figure 1F). Two-way ANOVA indicated significant(p � .001) main effects of dose (F4,21 ¼ 157), time (F18,378 ¼ 61.4),and their interaction (F2,378 ¼ 18.2). The onset of effects ofJHW007 was appreciably slower compared with the other com-pounds; the rate of increase at the highest dose over the first 50minutes was 7.0% per minute. Maximal stimulation of DA levels to�700% at the highest dose was obtained at 120 to 240 minutesafter injection. DA levels remained relatively higher than basalvalues and comparable with those obtained at the 6- to 7-hourtime points for at least 27 hours (Figure 1F). Control salineinjections (1 ml/kg, IP) did not affect DA levels (Figure 1F, filledcircles, p � .05).

Drug Discrimination StudiesSubstitution of Standard DUIs. Cocaine, 5 minutes before

sessions, dose-dependently increased the percentage of responses

Figure 2. Substitution for the training dose of cocaine (10 mg/kg, intraperitoneal, 5 minutes before testing) obtained after intraperitoneal administrationof various doses of cocaine (A), methylphenidate (B), WIN35,428 (C), AHN1-055 (D), AHN2-005 (E), and JHW007 (F) injected at different pretreatmenttimes, on the cocaine-associated lever; data are expressed as percentage of cocaine-appropriate lever pressing as a function of dose. Various doses ofdrugs were administered at various pretreatment intervals before the start of the session. Each point represents the mean of at least 6 (n ¼ 6–18) subjects� SEM. Saline administration produced an average of 1.74% (�.469) cocaine-appropriate responding when administered immediately before sessions,with little or no variation over the course of these experiments.

S.J. Kohut et al. BIOL PSYCHIATRY 2014;76:802–809 805

on the cocaine-appropriate lever (Figure 2A), whereas salineproduced 2.37 � .4% cocaine-appropriate responding (data notshown). Similar effects were obtained at various cocaine pretreat-ment times up to 120 minutes. At 90 minutes after injection, thedose-effect curve was shifted four-fold to the right of thatobtained with a 5-minute pretreatment time (Figure 2A; TableS1 in Supplement 1). Decreases in the maximum percentage ofcocaine-appropriate responding were found at the longest pre-treatment times. Figure S3 in Supplement 1 shows dose-relateddecreases in rates of responding that were most pronounced atthe earliest times after injections.

Methylphenidate and WIN35,428 (Figure 2B and C, respec-tively) 5 minutes before sessions also dose-dependently increasedthe percentage of cocaine-appropriate responding. Increases inpretreatment times produced shifts to the right in dose-effectcurves and increases in ED50 values (Figure 2B,C; Table S1 inSupplement 1). For methylphenidate, the shifts were relativelyabrupt with the dose-effect curve shifted to the right and down at120 minutes (Figure 2B). With WIN35,428, there was an increase inED50 value with a 60-minute compared with a 5-minute pretreat-ment time and rightward shifts thereafter up to three-fold to five-fold compared with the 60-minute assessment (Figure 2C; Table S1in Supplement 1).

Substitution of Atypical DUIs. AHN1-055 produced a dose-related increase in the percentage of cocaine-appropriateresponding (Figure 2D). Full substitution for cocaine was obtainedonly when the 10 mg/kg dose was administered 90 minutesbefore sessions. At shorter or longer pretreatment times, incom-plete substitution for cocaine was obtained (the percentage ofresponses on the cocaine-appropriate lever was greater thanvehicle levels). There was no systematic relationship betweenpretreatment time and potency of AHN1-055 (Figure 2D; Table S1in Supplement 1).

AHN2-005 produced dose- and time-related increases incocaine-appropriate responding. Full substitution for cocainewas obtained with AHN2-005 only at 30 mg/kg dose given at30 minutes or 90 minutes before sessions (Figure 2E). AHN2-005(30 mg/kg) injected 60 minutes before sessions substantiallydecreased response rates (Figure S3E in Supplement 1) to theextent that percent cocaine lever responding data were excluded(as per Supplemental Methods in Supplement 1). Inclusion ofthese values would have shown an approximate 40% drug-appropriate responding. The highest AHN2-005 dose injected120 minutes and 360 minutes before sessions produced amaximum of �60% substitution for cocaine. The potency ofAHN2-005 modestly and nonsignificantly decreased 2.6-fold at

www.sobp.org/journal

806 BIOL PSYCHIATRY 2014;76:802–809 S.J. Kohut et al.

360 minutes compared with its highest potency at 60 minutesafter injection (Table S1 in Supplement 1).

JHW007 never fully substituted for cocaine across all pretreat-ment times and doses studied (Figure 2F). The maximal percent-age of cocaine-appropriate responding was obtained at 17 mg/kgadministered 180 minutes before sessions. For those times atwhich there was a significant linear regression, there were nochanges in ED50 value as a function of pretreatment time(Figure 2; Table S1 in Supplement 1). Higher doses of JHW007were not assessed, as the highest dose produced substantialdecreases in response rates (Figure S3F in Supplement 1).

Relation of Cocaine-Appropriate Responding to Stimulationof DA

Effects of Standard DAT Inhibitors. There was a linearrelation between percent cocaine-appropriate responding andpercent stimulation of DA concentration at each cocaine pre-treatment time (Figure 3A). Linear regressions at individual timepoints all had slopes that significantly differed from zero (t values� 4.93; p values � .018). Additionally, the regressions did notsignificantly differ from that for a single linear model thatincluded all pretreatment times (F5,239 ¼ 1.875; p ¼ .099),with a slope of .492 percent drug/percent DA increase. Fullsubstitution was almost uniformly obtained at DA levelsthat were �100% to 125% greater than basal values. Largerincreases were uniformly associated with full substitution regard-less of dose or pretreatment time, with the one exception of the90-minute pretreatment time at which there was an increase�173% of DA levels and 66% substitution for cocaine (data notshown).

Figure 3. Percentage of cocaine-appropriate responding as a function of cocaor JHW007 (F) induced stimulation of dopamine (DA) levels, respectively. Extraclevels measured in 10-minute bins during the time period most closely correspothe 5-minute presession time-out) for each subject. Each point represents the

www.sobp.org/journal

Like cocaine, there was a significant linear relation betweencocaine-appropriate responding and stimulation of DA concen-tration for methylphenidate at all pretreatment times (Figure 3B;t values � 5.049; p values � .035) except 30 minutes (t value ¼4.186; p ¼ .149). As with cocaine, the regressions for theindividual times did not have slopes significantly different fromthat for a single linear model (F3,115 ¼ 2.04; p ¼ .112) with a slopeof .456. Further, none of the linear models for methylphenidatesignificantly differed from the linear model for cocaine (F4,364 ¼2.199; p ¼ .06). Full substitution for cocaine was obtained at dosesthat increased DA concentrations to �125% greater than basallevels (Figure 3B), with larger increases in DA also associated withfull substitution for cocaine.

Results from the linear regressions of cocaine-appropriateresponding and stimulation of DA concentration for WIN35,428show that at the 5-minute, 30-minute, and 240-minute pretreat-ment times slopes were significantly different from zero(Figure 3C; t values � 4.32; p values � .05), though the slopesat 60, 120, and 180 minutes approached but did not achievesignificance in relation to zero (t values � 4.261; p � .051). Slopesof the regression lines for the different WIN35,428 pretreatmenttimes were significantly different from each other (F5,93 ¼ 7.59;p � .0001), and post hoc comparisons indicated that the 120-minute and 180-minute pretreatment times differed from theothers. Doses of WIN35,428 that produced stimulation of DAgreater than �100% to 125% above basal levels were uniformlyrelated to full substitution for cocaine regardless of time pointor dose.

Effects of Atypical DUIs. With AHN1-055, the slopes of thelinear regressions of cocaine-appropriate responding and

ine (A), methylphenidate (B), WIN35,428 (C), AHN1-055 (D), AHN2-005 (E),ellular DA levels (as a percentage of basal levels) were calculated from DAnding to the duration of the drug discrimination sessions (15 minutes plusmean of at least 6 subjects � SEM.

S.J. Kohut et al. BIOL PSYCHIATRY 2014;76:802–809 807

stimulation of DA concentration were significantly different fromzero at the 5-minute, 90-minute, and 120-minute pretreatmenttimes (Figure 3D; t values � 5.88; p values � .028), though theslopes were not different from zero at 30 and 60 minutes (t values� 9.32; p � .068). Further, the slopes of the linear regressionswere not significantly different from each other (F4,79 ¼ 1.31;p ¼ .272). However, the slopes of the regressions for AHN1-055 atthe various times were significantly different from that for cocaine(F5,334 ¼ 67.22; p � .0001). Post hoc comparisons indicated thatregression lines for percent cocaine response and DA concen-tration produced by AHN1-055 had slopes that were significantlyless steep than that for cocaine at each time point studied (allcorrected p values � .043).

The linear regressions of cocaine-appropriate responding andstimulation of DA concentration for AHN2-005 had slopes signifi-cantly different from zero at all pretreatment times (Figure 3E;t values � 3.62; p values � .036), though not significantly differentfrom each other (F5,173 ¼ 1.59; p ¼ .164). However, the slopes of theregression lines for AHN2-005 were significantly different from thatfor cocaine (F6,422 ¼ 24.73; p � .0001). Post hoc comparisonsindicated that slopes of the regression lines at all pretreatmenttimes were significantly less steep than that for cocaine (allcorrected p values # .0002).

With JHW007, the linear regressions were significantly differentfrom zero at all times (Figure 3F; t values � 8.79; p values � .044)except 24 hours (1440 minutes) after injection (t ¼ 2.10; p ¼ .283).Further, the data were better fit with a single regression line thanwith each of the pretreatment times having its own slope (F5,159 ¼.688; p ¼ .633). Additionally, the regression lines for the individualtime points for JHW007 had slopes that were different from thatfor cocaine (F6,408 ¼ 30.7; p � .0001) and, except that for 24 hours,were significantly less steep than that for cocaine (all correctedp values # .0006).

Discussion

In the present study, standard and atypical DUIs dose-dependently and significantly stimulated extracellular DA levelsin dialysates from the NAS. An extensive literature has indicatedthat cocaine-like behavioral effects of monoamine uptake inhib-itors, including those with low selectivity (35,36), are due tothe enhancement of DA neurotransmission (12,27,29). Further,6-OHDA lesions of the nucleus accumbens decrease cocaine-likediscriminative stimulus effects (4,5), indicating the importance ofthis region for the subjective effects of cocaine.

Supporting the importance of DA in the NAS, we found alinear relationship between stimulation of DA levels and produc-tion of cocaine-like subjective effects with the standard DUIs.Similar relationships for methamphetamine-like subjective effectshave been shown in rats (8) and in squirrel monkeys (7). We haveexpanded those findings to cocaine-like discriminative effects andfurther systematically documented the full dose- and time-relatedsimilarity of DA-enhancing and cocaine-like subjective effects ofstandard DUIs. Importantly, the linear relationship betweenchange in DA levels and discriminative effects was independentof dose and presession treatment time at which the DA levelswere obtained and similar for the standard DUIs.

It is noteworthy that the standard DUIs fully substituted forcocaine at DA levels from �100% to 125 % over basal values,regardless of which dose or pretreatment time produced that DAlevel. In contrast, full cocaine-like subjective effects were notproduced by atypical DUIs at any dose or pretreatment time

producing that same increase. A previous study similarly foundthat increases in DA levels of 65% over basal DA values aresufficient for place conditioning by cocaine, though insufficientfor several atypical DUIs (37), which produced maximal increasesin DA levels from 138% to 277% over basal values. In the presentstudy, AHN1-055 and AHN2-005 produced cocaine-like subjectiveeffects in rats only sporadically at some presession treatmenttimes at the highest doses, whereas JHW007 did not substitutefor cocaine at any dose or pretreatment time. Each of theseatypical DUIs produced increases in DA levels that were elevatedmore than 700% above baseline, well in excess of the �100% to125% above baseline levels associated with complete substitutionfor cocaine by the standard DUIs.

Among the drugs tested, methylphenidate is currently Foodand Drug Administration approved as a medication to treat ADHD,while AHN2-005 is being developed for this indication (24,25).Previously published articles (38,39) have indicated some liabilityfor abuse of some ADHD medications. The present study suggestsa safer pharmacologic profile for AHN2-005 compared withmethylphenidate due to its reduced cocaine-like effects. The lowerabuse potential of BZT analogs has been reported previously(16,37,40–45), which is consistent with the present findings.

Various explanations for the reduced cocaine-like effects of BZTanalogs have been proposed (14,16,17). As mentioned above,in vitro studies have suggested their interaction with the DATdifferently from cocaine and cocaine-like drugs (46–48). Further,the decreases in uptake inhibition potency of DUIs in cells trans-fected with a mutant DAT that assumes an inward-facing con-formation compared with the wild-type DAT were directly relatedto the cocaine-like effectiveness of a variety of DUIs, includingseveral BZT analogs (13). In the present study, compounds intole-rant to DAT conformation (cocaine, methylphenidate, WIN35,428)all substituted fully for cocaine when NAS DA levels were elevatedby �100% to 125% over basal levels, whereas those tolerant toDAT conformation (BZT analogs) did not reliably substitute fully forcocaine at NAS DA levels several times greater.

Several studies have unsuccessfully attempted to identify anoffsite target involved in the atypical effects of BZTs (16,49,50).Recent studies, however, have suggested that sR antagonisteffects may contribute to the atypical effects of BZT analogs(42,43,51), and several studies have indicated that sR antagonistactivity coupled with DAT inhibition can block the self-administration of stimulants (43,52). Further, the presently studiedcompounds have affinities for sRs that are comparable with orexceed those for the DAT, further suggesting this action as onethat contributes to the atypical DUI effects observed. However, arecent study has found an atypical DUI (53) that does not bindsRs (54), suggesting that multiple mechanisms may contribute toatypical effects of DUIs.

Other studies have suggested that differences in the onset/offset, duration, and pattern of DA stimulation may be key factorsin the reduced cocaine-like effects of atypical DUIs (14,16,17). Forexample, binding to the inward-facing DAT conformation mayentail a slow onset/offset of DAT-mediated action (13) thatcontributes to the reduced cocaine-like effects of JHW007 andother atypical DUIs (18,55). Several studies have indicated in vivoapparent association rates at DAT that are slower than those forcocaine-like compounds, as well as slower in vitro associationrates (18,47,55). Other studies have shown that at times right afterinjection, substantive levels of several BZTs are available in thebrain (56,57) at concentrations greater than those necessary forDAT binding (58). In the present study, standard DUIs producedrapid and transient increases in DA levels up to 3 to 5 hours after

www.sobp.org/journal

808 BIOL PSYCHIATRY 2014;76:802–809 S.J. Kohut et al.

administration of the highest doses, depending on the drug,whereas each of the atypical DUIs showed slower onsets andlonger durations of action. However, a lack of cocaine-like effectsof several fast-acting BZT analogs has been recently reported (42).Further, the rate of increase in DA concentrations at the highestdose of AHN2-005 in the present study approached that ofmethylphenidate and WIN35,428. Thus, onset of effects alonecannot completely account for the differences in effects betweenthe standard and atypical DUIs. Nonetheless, onset might con-tribute to atypical effects of DUIs combined with actions at an off-target site, if the onset of effect at the off-target site is faster thanthat for DAT-mediated effects.

Slower and longer-lasting increases in DA concentrationsproduced by atypical DUIs may create conditions favorable todesensitization of postsynaptic DA receptors that, in turn, wouldbe less effective in transmitting DA signals. In contrast, the fasterand shorter-acting cocaine-like drugs may not allow a rapiddesensitization. Such a desensitization process agrees with thelow efficacy of BZT analogs in producing cocaine-like subjectiveeffects (13), as well as the previously described failure of thesecompounds to produce place conditioning (37,42). As shown fortypical DUIs after long-term administration (59), we furtherhypothesize significant acute alterations in the feedback signalson the firing of DA neurons modulated by presynaptic andpostsynaptic DA receptors, in nerve terminal release, and in DAsynthesis right after acute administration of atypical DUIs (59).

In summary, the present study showed that atypical DUIs wereable to increase DA levels in the NAS, a brain area related to thereinforcing and subjective effects of cocaine. However, thetemporal patterns of stimulation of DA levels, possibly combinedwith off-target actions or compensatory desensitization, may playa fundamental role in their reduced cocaine-like subjectiveeffects. These features of the atypical DUIs are those preferredfor the development of cocaine abuse medications (16,53,60) andmay provide an improved margin of safety beyond that for drugspresently used in the treatment of ADHD. Therefore, theseexperiments advance specific hypotheses of how the differencesin molecular mechanisms of action and the resulting time coursesof standard and atypical DUIs are related to their subjectiveeffects and also provide avenues for the discovery of DAT-basedmedications with safe therapeutic profiles.

This work was funded by the Intramural Research Program of theNational Institute on Drug Abuse, National Institutes of Health,Department of Health and Human Services.

Linda Garcés-Ramírez was recipient of a scholarship fromConsejo Nacional de Ciencia y Tecnología, México.

Takato Hiranita was supported in part by the Japanese Societyfor the Promotion of Science, Research Fellowship for JapaneseBiomedical and Behavioral Researchers at the National Institutes ofHealth.

Subjects used in this study were maintained in facilities fullyaccredited by the American Association for the Accreditation ofLaboratory Animal Care and all experiments were conducted inaccordance with the guidelines of the Animal Care and Use Committeeof the Intramural Research Program, National Institute on Drug Abuse,National Institutes of Health, and the Guide for Care and Use ofLaboratory Animals (National Research Council, 2003). In accordancewith these guidelines, the experiments have been performed trying tominimize animal suffering, reducing the number of animals used, andsearching alternatives to in vivo techniques, if available.

We thank Patty Ballerstadt and Maryann Carrigan for admin-istrative assistance, J.J. Cao for the synthesis of benztropine analogs,

www.sobp.org/journal

and Dr. Ronald J. Tallarida for advice on the linear regressionanalyses.

AHN and JLK are inventors on a patent describing the benz-tropine analogs, including AHN2-005 and JHW007, the rights ofwhich are owned by the National Institutes of Health. The rights tocommercial development of the compounds in the patent have beenlicensed to Phase 2 Discovery, Inc., of Cincinnati, Ohio.

All other authors declare no biomedical financial interests orpotential conflicts of interest.

Supplementary material cited in this article is available online athttp://dx.doi.org/10.1016/j.biopsych.2014.03.031.

1. Haertzen CA, Hickey JE (1987): Addiction Research Center Inventory(ARCI): Measurement of euphoria and other drug effects. In: BozarthMA, editor. Methods of Assessing the Reinforcing Properties of AbusedDrugs. New York: Springer-Verlag.

2. Holtzman SG (1985): Drug discrimination studies. Drug Alcohol Depend14:263–282.

3. Schuster CR, Johanson CE (1988): Relationship between the discrim-inative stimulus properties and subjective effects of drugs. Psycho-pharmacol Ser 4:161–175.

4. Callahan PM, De La Garza R 2nd, Cunningham KA (1997): Mediation ofthe discriminative stimulus properties of cocaine by mesocorticolim-bic dopamine systems. Pharmacol Biochem Behav 57:601–607.

5. Dworkin SI, Smith JE (1988): Neurobehavioral pharmacology ofcocaine. NIDA Res Monogr 88:185–198.

6. Wise RA (2005): Forebrain substrates of reward and motivation.J Comp Neurol 493:115–121.

7. Czoty PW, Makriyannis A, Bergman J (2004): Methamphetaminediscrimination and in vivo microdialysis in squirrel monkeys. Psycho-pharmacology (Berl) 175:170–178.

8. Desai RI, Paronis CA, Martin J, Desai R, Bergman J (2010): Mono-aminergic psychomotor stimulants: Discriminative stimulus effectsand dopamine efflux. J Pharmacol Exp Ther 333:834–843.

9. Volkow ND, Wang GJ, Fischman MW, Foltin RW, Fowler JS, AbumradNN, et al. (1997): Relationship between subjective effects of cocaineand dopamine transporter occupancy. Nature 386:827–830.

10. Bergman J, Madras BK, Johnson SE, Spealman RD (1989): Effectsof cocaine and related drugs in nonhuman primates. III. Self-administration by squirrel monkeys. J Pharmacol Exp Ther 251:150–155.

11. Katz JL, Izenwasser S, Terry P (2000): Relationships among dopaminetransporter affinities and cocaine-like discriminative-stimulus effects.Psychopharmacology (Berl) 148:90–98.

12. Kuhar MJ, Ritz MC, Boja JW (1991): The dopamine hypothesis of thereinforcing properties of cocaine. Trends Neurosci 14:299–302.

13. Loland CJ, Desai RI, Zou MF, Cao J, Grundt P, Gerstbrein K, et al. (2008):Relationship between conformational changes in the dopaminetransporter and cocaine-like subjective effects of uptake inhibitors.Mol Pharmacol 73:813–823.

14. Newman AH, Katz JL (2009): Atypical dopamine uptake inhibitors thatprovide clues about cocaine's mechanism at the dopamine trans-porter. Top Med Chem 4:95–129.

15. Runyon SP, Carroll FI (2008): Tropane-based dopamine transporter-uptake inhibitors. In: Trudell ML, Izenwasser S, editors. DopamineTransporters, Chemistry, Biology, and Pharmacology. Hoboken, NJ: JohnWiley & Sons, 125–170.

16. Tanda G, Newman A, Katz JL (2009): Discovery of drugs to treatcocaine dependence: Behavioral and neurochemical effects of atypicaldopamine transport inhibitors. In: Enna SJ, Williams M, editors.Advances in Pharmacology, Contemporary Drug Discovery. San Diego:Academic Press, 253–289.

17. Schmitt KC, Rothman RB, Reith ME (2013): Nonclassical pharmacologyof the dopamine transporter: Atypical inhibitors, allosteric modulators,and partial substrates. J Pharmacol Exp Ther 346:2–10.

18. Desai RI, Kopajtic TA, French D, Newman AH, Katz JL (2005): Relation-ship between in vivo occupancy at the dopamine transporter andbehavioral effects of cocaine, GBR 12909 [1-{2-[bis-(4-fluorophenyl)methoxy]ethyl}-4-(3-phenylpropyl)piperazine], and benztropine ana-logs. J Pharmacol Exp Ther 315:397–404.

S.J. Kohut et al. BIOL PSYCHIATRY 2014;76:802–809 809

19. Punja S, Zorzela L, Hartling L, Urichuk L, Vohra S (2013): Long-actingversus short-acting methylphenidate for paediatric ADHD: A systematicreview and meta-analysis of comparative efficacy. BMJ Open 3:e002312.

20. Beuming T, Kniazeff J, Bergmann ML, Shi L, Gracia L, Raniszewska K,et al. (2008): The binding sites for cocaine and dopamine in thedopamine transporter overlap. Nat Neurosci 11:780–789.

21. Schmitt KC, Reith ME (2011): The atypical stimulant and nootropicmodafinil interacts with the dopamine transporter in a differentmanner than classical cocaine-like inhibitors. PLoS One 6:e25790.

22. Agoston GE, Wu JH, Izenwasser S, George C, Katz J, Kline RH, NewmanAH (1997): Novel N-substituted 3 alpha-[bis(4'-fluorophenyl)methoxy]tropane analogues: Selective ligands for the dopamine transporter.J Med Chem 40:4329–4339.

23. Newman AH, Kline RH, Allen AC, Izenwasser S, George C, Katz JL(1995): Novel 4’-substituted and 4’,4”-disubstituted 3 alpha-(diphenyl-methoxy)tropane analogs as potent and selective dopamine uptakeinhibitors. J Med Chem 38:3933–3940.

24. Berridge CW, Devilbiss DM, Spencer RC, Schmeichel BE, Arnsten FT(2012): Attention deficit hyperactivity disorder. In: Barrett JE, Coyle JT,Williams M, editors. Translational Neuroscience. New York: CambridgeUniversity Press, 303–320.

25. Schmeichel BE, Zemlan FP, Berridge CW (2013): A selective dopaminereuptake inhibitor improves prefrontal cortex-dependent cognitivefunction: Potential relevance to attention deficit hyperactivity disor-der. Neuropharmacology 64:321–328.

26. Tanda G, Pontieri FE, Di Chiara G (1997): Cannabinoid and heroinactivation of mesolimbic dopamine transmission by a common mu1opioid receptor mechanism. Science 276:2048–2050.

27. Tanda G, Pontieri FE, Frau R, Di Chiara G (1997): Contribution ofblockade of the noradrenaline carrier to the increase of extracellulardopamine in the rat prefrontal cortex by amphetamine and cocaine.Eur J Neurosci 9:2077–2085.

28. Zahm DS (1999): Functional-anatomical implications of the nucleusaccumbens core and shell subterritories. Ann N Y Acad Sci. 877:113–128.

29. Pontieri FE, Tanda G, Di Chiara G (1995): Intravenous cocaine,morphine, and amphetamine preferentially increase extracellulardopamine in the “shell” as compared with the “core” of the ratnucleus accumbens. Proc Natl Acad Sci U S A 92:12304–12308.

30. Paxinos G, Watson C (1998): The Rat Brain in Stereotaxic Coordinates,4th ed. San Diego: Academic Press.

31. Garces-Ramirez L, Green JL, Hiranita T, Kopajtic TA, Mereu M, Thomas A,et al. (2011): Sigma receptor agonists: Receptor binding and effects onmesolimbic dopamine neurotransmission assessed by microdialysis. BiolPsychiatry 69:208–217.

32. Tanda G, Ebbs A, Newman AH, Katz JL (2005): Effects of 4'-chloro-3alpha-(diphenylmethoxy)-tropane on mesostriatal, mesocortical, andmesolimbic dopamine transmission: Comparison with effects ofcocaine. J Pharmacol Exp Ther 313:613–620.

33. Tanda G, Ebbs AL, Kopajtic TA, Elias LM, Campbell BL, Newman AH,Katz JL (2007): Effects of muscarinic M1 receptor blockade on cocaine-induced elevations of brain dopamine levels and locomotor behaviorin rats. J Pharmacol Exp Ther 321:334–344.

34. Snedecor GW, Cochran WG (1967): Statistical Methods. 6th ed. Ames,IA: Iowa State University Press.

35. Baker LE, Riddle EE, Saunders RB, Appel JB (1993): The role ofmonoamine uptake in the discriminative stimulus effects of cocaineand related compounds. Behav Pharmacol 4:69–79.

36. Broadbent J, Michael EK, Riddle EE, Apple JB (1991): Involvement ofdopamine uptake in the discriminative stimulus effects of cocaine.Behav Pharmacol 2:187–197.

37. Tanda G, Li SM, Mereu M, Thomas AM, Ebbs AL, Chun LE, et al. (2013):Relations between stimulation of mesolimbic dopamine and placeconditioning in rats produced by cocaine or drugs that are tolerant todopamine transporter conformational change. Psychopharmacology(Berl) 229:307–321.

38. Kollins SH (2007): Abuse liability of medications used to treatattention-deficit/hyperactivity disorder (ADHD). Am J Addict 16(suppl 1):35–42; quiz 43–44.

39. Wilens TE, Adler LA, Adams J, Sgambati S, Rotrosen J, Sawtelle R, et al.(2008): Misuse and diversion of stimulants prescribed for ADHD: A system-atic review of the literature. J Am Acad Child Adolesc Psychiatry 47:21–31.

40. Ferragud A, Velazquez-Sanchez C, Hernandez-Rabaza V, Nacher A,Merino V, Carda M, et al. (2009): A dopamine transport inhibitor with

markedly low abuse liability suppresses cocaine self-administration inthe rat. Psychopharmacology (Berl) 207:281–289.

41. Hiranita T, Soto PL, Newman AH, Katz JL (2009): Assessment ofreinforcing effects of benztropine analogs and their effects on cocaineself-administration in rats: Comparisons with monoamine uptakeinhibitors. J Pharmacol Exp Ther 329:677–686.

42. Li SM, Kopajtic TA, O'Callaghan MJ, Agoston GE, Cao J, Newman AH,et al. (2011): N-substituted benztropine analogs: Selective dopaminetransporter ligands with a fast onset of action and minimal cocaine-like behavioral effects. J Pharmacol Exp Ther 336:575–585.

43. Hiranita T, Kohut SJ, Soto PL, Tanda G, Kopajtic TA, Katz JL (2014):Preclinical efficacy of N-substituted benztropine analogs as antago-nists of methamphetamine self-administration in rats. J Pharmacol ExpTher 348:174–191.

44. Velazquez-Sanchez C, Ferragud A, Murga J, Carda M, Canales JJ (2010):The high affinity dopamine uptake inhibitor, JHW 007, blocks cocaine-induced reward, locomotor stimulation and sensitization. Eur Neuro-psychopharmacol 20:501–508.

45. Velazquez-Sanchez C, Ferragud A, Renau-Piqueras J, Canales JJ (2010):Therapeutic-like properties of a dopamine uptake inhibitor in animalmodels of amphetamine addiction. Int J Neuropsychopharmacol 14:655–665.

46. Chen N, Zhen J, Reith ME (2004): Mutation of Trp84 and Asp313 of thedopamine transporter reveals similar mode of binding interaction forGBR12909 and benztropine as opposed to cocaine. J Neurochem 89:853–864.

47. Kopajtic TA, Liu Y, Surratt CK, Donovan DM, Newman AH, Katz JL(2010): Dopamine transporter-dependent and -independent striatalbinding of the benztropine analog JHW 007, a cocaine antagonistwith low abuse liability. J Pharmacol Exp Ther 335:703–714.

48. Reith ME, Berfield JL, Wang LC, Ferrer JV, Javitch JA (2001): The uptakeinhibitors cocaine and benztropine differentially alter the conformationof the human dopamine transporter. J Biol Chem 276:29012–29018.

49. Campbell VC, Kopajtic TA, Newman AH, Katz JL (2005): Assessment ofthe influence of histaminergic actions on cocaine-like effects of3alpha-diphenylmethoxytropane analogs. J Pharmacol Exp Ther 315:631–640.

50. Tanda G, Kopajtic TA, Katz JL (2008): Cocaine-like neurochemicaleffects of antihistaminic medications. J Neurochem 106:147–157.

51. Katz JL, Kopajtic TA, Agoston GE, Newman AH (2004): Effects ofN-substituted analogs of benztropine: Diminished cocaine-like effectsin dopamine transporter ligands. J Pharmacol Exp Ther 309:650–660.

52. Hiranita T, Soto PL, Kohut SJ, Kopajtic T, Cao J, Newman AH, et al. (2011):Decreases in cocaine self-administration with dual inhibition of the dopa-mine transporter and sigma receptors. J Pharmacol Exp Ther 339:662–677.

53. Carroll FI, Howell LL, Kuhar MJ (1999): Pharmacotherapies for treat-ment of cocaine abuse: Preclinical aspects. J Med Chem 42:2721–2736.

54. Hiranita T, Wilkinson DS, Hong WC, Zou MF, Kopajtic TA, Soto PL, et al.(2014): 2-Isoxazol-3-phenyltropane derivatives of cocaine: molecularand atypical system effects at the dopamine transporter. J PharmacolExp Ther 349:297–309. doi:10.1124/jpet.113.212738.

55. Desai RI, Kopajtic TA, Koffarnus M, Newman AH, Katz JL (2005):Identification of a dopamine transporter ligand that blocks thestimulant effects of cocaine. J Neurosci 25:1889–1893.

56. Raje S, Cao J, Newman AH, Gao H, Eddington ND (2003): Evaluation ofthe blood-brain barrier transport, population pharmacokinetics, andbrain distribution of benztropine analogs and cocaine using in vitroand in vivo techniques. J Pharmacol Exp Ther 307:801–808.

57. Raje S, Cornish J, Newman AH, Cao J, Katz JL, Eddington ND (2005):Pharmacodynamic assessment of the benztropine analogues AHN-1055 and AHN-2005 using intracerebral microdialysis to evaluate braindopamine levels and pharmacokinetic/pharmacodynamic modeling.Pharm Res 22:603–612.

58. Tanda G, Newman A, Ebbs AL, Tronci V, Green J, Tallarida RJ, Katz JL(2009): Combinations of cocaine with other dopamine uptake inhib-itors: Assessment of additivity. J Pharmacol Exp Ther 330:802–809.

59. Katz NS, Guiard BP, El Mansari M, Blier P (2010): Effects of acute andsustained administration of the catecholamine reuptake inhibitornomifensine on the firing activity of monoaminergic neurons.J Psychopharmacol 24:1223–1235.

60. Newman AH, Kulkarni S (2002): Probes for the dopamine transporter:New leads toward a cocaine-abuse therapeutic–A focus on analoguesof benztropine and rimcazole. Med Res Rev 22:429–464.

www.sobp.org/journal

Kohut et al.

1

Preference for Distinct Functional Conformations of the Dopamine Transporter Alters the Relationship between Subjective Effects of Cocaine

and Stimulation of Mesolimbic Dopamine

Supplemental Information

Supplemental Methods

Subjects

Male Sprague–Dawley rats (Charles River, Germantown, MD, USA), weighing 275 to

350 g served as subjects. They were doubly housed in a temperature- and humidity-controlled

room maintained on a 12-h light/dark cycle (lights on: 0700–1900 hour) and had free access to

food and water. Rats assigned to the microdialysis experiments had free access to food and water

except during the microdialysis tests. Rats assigned to drug-discrimination tests were always

kept at 85% of their initial free feeding weight. Experiments were conducted during the light

phase. Animals used in the study were experimentally naive at the start. The animal housing

facilities were fully accredited by AAALAC International and all procedures were conducted in

accordance with the guidelines of the Institutional Care and Use Committee of the NIDA

Intramural Research Program and the 1996 National Research Council and Guide for Care and

Use of Laboratory Animals.

In-vivo Microdialysis Studies

Surgery. Surgical procedures were conducted under a mixture of ketamine and xylazine

(60.0 and 12.0 mg/kg intraperitoneal (i.p.), respectively) anesthesia. Rats were placed in a

stereotaxic apparatus, the skull was exposed, and a small hole was drilled to expose the dura.

Rats were then implanted with a concentric dialysis probe (see description below) aimed at the

Kohut et al.

2

nucleus accumbens shell (NAS) either in the right or in the left brain side, as described

previously (1, 2). Coordinates were from the rat brain atlas of Paxinos and Watson (3)

[uncorrected coordinates: anterior = + 2.0 mm from bregma, lateral = ± 1.0 mm from bregma,

vertical = -7.9 mm from dura]. After the surgery, rats were allowed to recover overnight in

hemispherical CMA-120 cages (CMA Microdialysis AB, Solna, Sweden). Concentric dialysis

probes were prepared with AN69 hollow fibers (Hospal Dasco, Bologna, Italy). Briefly, two 4-

cm pieces of silica fused capillary tubes (the inlet and outlet tubing of the probes) were inserted

into a 6-mm capillary (0.25 mm external diameter) dialyzing fiber (closed by a drop of glue on

the other side) with the inlet tubing set at about 0.1 mm from the closed end of the fiber and the

outlet set at about 1.8 mm from the inlet tip. The open end of the dialysis membrane was then

glued, and the protruding two silica fused capillary tubes were inserted and glued into a 22-G

stainless steel needle (2.4 mm length). The needle was attached to a CMA/10 clip

(CMA/Microdialysis AB) and mounted in a stereotaxic holder. The exposed dialyzing surface of

the fibers, i.e., the portion not covered by glue, was limited to the lowest 1.8 mm of the probes.

Drugs. (-)-Cocaine hydrochloride (Sigma-Aldrich, St. Louis, MO); methylphenidate

(NIDA/IRP Drug Supply Program); WIN35,428 (Sigma-Aldrich); AHN1-055, AHN2-005, and

JHW007 were synthesized in the Medicinal Chemistry Section, NIDA/IRP, as per (4, 5). Drugs

were dissolved in 0.9% NaCl or sterile water with heat and sonication, as necessary, and were

injected i.p. (1.0 ml/kg of body weight).

Analytical and Experimental Procedures. Experiments were performed on freely

moving rats in the same hemispherical cages in which they recovered from surgery. About 22

hours after implant, probes were connected to fluid swivels (375/ D/22QM; Instech, Plymouth

Meeting, PA, USA), and Ringer’s solution (147.0 mM NaCl, 2.2 mM CaCl2, and 4.0 mM KCl)

Kohut et al.

3

was delivered by a 1.0 ml syringe operated by a BAS Bee Syringe Pump Controller (BAS, West

Lafayette, IN, USA) through the dialysis probes at a constant flow rate of 1 μl/min. Collection of

dialysate samples started after 30 min. Ten μl samples were taken every 10 min and immediately

analyzed, as detailed below. After stable dopamine (DA) values (less than 10% variability) were

obtained for at least three consecutive samples (after about 1 hour), rats were treated with one

dose of one of the test drugs (cocaine, or one of the other dopamine uptake inhibitors) or saline.

A separate group of 12 subjects was trained to discriminate cocaine (10 mg/kg i.p.) from

saline (1 ml/kg i.p.) injections as described in the behavioral section below. After these rats

reached criteria for discrimination of cocaine from saline injections, they were implanted with

microdialysis probes as described above. Approximately 24 hours after probe implantation,

subjects were treated with one of two cocaine doses, 3 (n = 6) or 10 (n = 6) mg/kg i.p., and DA

levels were monitored for 2 hours after injection. These subjects were tested again about 24

hours after the first test, and were injected with the dose of cocaine that they did not receive in

the first test. Again, DA levels were monitored for 2 hours after the injection. This group of

subjects served as a control for the effects of cocaine on stimulation of NAS DA levels in

subjects trained in the drug-discrimination procedure (See Figure S2).

Dialysate samples were injected without purification into a HPLC apparatus equipped

with an MD 150 mm X 3.2 mm column, particle size 3.0 μm (ESA, Chelmsford, MA, USA), and

a coulometric detector (5200a Coulochem II, or Coulochem III, ESA) to quantify DA. The

oxidation and reduction electrodes of the analytical cell (5014B; ESA) were set at +125 mV and

-125 mV, respectively. The mobile phase, containing 100 mM NaH2PO4, 0.1 mM Na2EDTA, 0.5

mM n-octyl sulfate, and 18% (v/v) methanol (pH adjusted to 5.5 with Na2HPO4), was pumped by

an ESA 582 (ESA, Chelmsford, MA, USA) solvent delivery module at 0.50 ml/min. Assay

Kohut et al.

4

sensitivity for DA was 2 fmoles/sample. The average basal DA values in dialysates from the

NAS in the present experiments were 42.3 ± 1.0 fmoles (± SEM) in a 10 µl sample, n = 171, and

no significant differences were found in the concentration of basal DA in dialysate samples from

different experimental groups (ANOVA, F(30,140) = 0.516, p = 0.98).

Histology. At the end of each experiment, rats were sacrificed by pentobarbital overdose.

Brains were removed and left to fix in 4% formaldehyde in saline solution. Brains were then cut

on a vibratome 1000 Plus (The Vibratome Company, St. Louis, MO, USA) in serial coronal

slices. Figure S1 schematically shows typical locations of the dialyzing portion of the probes

implanted in the NAS. Sections are redrawn from Paxinos and Watson (3) and the anterior

coordinates, from the bregma, are indicated in each section. Only the experiments in which the

probes were located inside the NAS within these sections have been used in the present paper.

Behavioral Tests

Drug Discrimination Studies. Experiments were conducted with subjects placed in a

23.0 cm L x 29.0 cm W x 21.0 cm H operant-conditioning chamber (modified ENV-001, Med

Associates, St. Albans, VT). On the front wall of the chamber were two response keys (levers

requiring a force of 0.4 N through 1 mm to register a response) and a dispenser for delivery of

45-mg food pellets to a centrally-located, recessed tray. A pair of green and a pair of yellow

light-emitting diodes were arranged horizontally above each lever. The chamber was housed in a

ventilated enclosure which provided light and sound attenuation. White noise was delivered to

the chamber at all times to mask extraneous noise.

Rats were initially trained with food reinforcement to press both keys, and were

eventually trained to press one after cocaine (10 mg/kg, i.p.), and the other after saline (i.p.)

Kohut et al.

5

injection. Each response produced an audible click. The ratio of responses to obtain food pellets

(fixed ratio or FR) was gradually increased until, under the final conditions, the completion of 20

consecutive responses on the cocaine- or saline-appropriate key produced a food pellet. Incorrect

responses reset the FR response requirement. The right vs. left assignment of cocaine- and

saline-associated keys was counterbalanced among subjects.

Subjects were placed in chambers immediately after injection. There was a five-min

timeout period during which lights were off and responses produced an audible click, but no

other scheduled consequences. Following the timeout, the house and lever lights were turned on

until the completion of the FR 20-response requirement. Food presentation was followed by a

20-sec timeout during which only the lever lights were off and responses had no scheduled

consequences other than the feedback click. Sessions ended after 20 food presentations or 15

min, whichever occurred first. Sessions were conducted 5 days per week, and cocaine or saline

sessions were scheduled in a double-alternation sequence (i.e., cocaine-saline-saline-cocaine).

Training continued until subjects met the criteria of at least 85% cocaine-appropriate or saline-

appropriate responding over the entire session, and during the first FR of the session, for four

consecutive sessions (two each with cocaine or saline injection).

Once these criteria were met testing began. Test sessions were conducted with the

administration of different doses of cocaine, typical DA transporter-blockers (methylphenidate

and WIN35,428), or atypical DA transporter-blockers (the BTZ analogs AHN1-055, AHN2-005,

and JHW007), before sessions. In order to study a time-course for the effect of their substitution

for the training dose of cocaine, some of the test sessions were performed by administering the

drugs at various pretreatment times before sessions. Test sessions were identical to training

sessions with the exception that completion of the FR requirement on either key was reinforced.

Kohut et al.

6

All data collection and programming of behavioral contingencies was accomplished with

software from Med Associates, Inc (St. Albans, VT).

Data Analyses. Microdialysis results were expressed as a percentage of basal DA values,

calculated as the average of 3 consecutive samples, differing by no more than 10 percent,

collected right before the administration of drug or saline. Statistical analysis was carried out

using one- or two-way ANOVA (factors: time, dose) for repeated measures over time with

significant results subjected to post-hoc Tukey`s test. For drug discrimination, percentages of

cocaine-lever responses were calculated by dividing the total number of responses on the

cocaine-associated lever by the total number of responses on both levers as detailed here. Full

substitution was defined as >85%, and partial substitution as between 85% and 15% cocaine-

appropriate responding. Dose-effect curves for percent cocaine-lever responding were analyzed

using standard linear regression techniques, from which ED50 values with 95% confidence

limits were calculated (6).

To examine the relation between levels of extracellular DA stimulation and cocaine-like

discriminative-stimulus effects, linear regression analyses were performed on the percentage of

cocaine-lever responding as a function of the percent change in extracellular DA. Extracellular

DA levels (as a percentage of basal levels) were calculated from DA levels measured in 10-min

bins during the time period most closely corresponding to the duration of the drug-discrimination

sessions (15 min plus the 5 min presession timeout), for each subject. Data from 0% to the

lowest dose producing ≥85% cocaine-lever responding were regressed on percent change in DA

levels, and the slopes of the best fit lines compared across pretreatment times for each drug and

between cocaine and each of the drugs using the extra sum of squares F-test.

Kohut et al.

7

Figure S1. Forebrain sections show the limits of the positions of the dialyzing portions of the microdialysis probes (superimposed rectangles) resulting from histological analysis of brain slices. On each section, redrawn from Paxinos and Watson (3), the anterior coordinate (A) (measured from Bregma) is indicated. CPu, caudate putamen; shell, nucleus accumbens shell.

Kohut et al.

8

Figure S2. Effects of cocaine administration on stimulation of dopamine (DA) levels in dialysates from the nucleus accumbens shell (NAS) in naïve rats and in rats that were successfully trained to discriminate injections of cocaine 10 mg/kg intraperitoneal (i.p.) from saline i.p. The top panel shows the effect of 3 and 10 mg/kg doses of cocaine on extracellular DA levels in dialysates from drug-naïve rats and from rats trained in drug-discrimination (DD) procedures to discriminate 10 mg/kg i.p. cocaine from 1 ml/kg i.p. saline. The bottom panel shows the effect of administration on day 1 and 2 of 3.0 and 10.0 mg/kg doses of cocaine on extracellular DA levels in NAS dialysates from rats trained in drug-discrimination procedures to discriminate 10 mg/kg i.p. cocaine from 1 ml/kg i.p. saline. Rats were treated with one of the 2 doses of cocaine on day one and with the other dose of cocaine on day two. Data in the top panel for the drug-discrimination group of rats is the average of the effects shown in the bottom panel for each dose of cocaine. Results are means, with vertical bars representing SEM, of the amount of DA in 10-min dialysate samples, expressed as percentage of basal values.

Kohut et al.

9

Figure S3. Each panel shows the overall rate of lever pressing expressed as percentage of rate of responding for saline (0.793 ± 0.037 responses per sec), averaged over the entire session, obtained from the experiments shown in the main text, Figure 2. Abscissae: dose in mg/kg (log scale). Results represent means ± SEM from at least 6 (n = 6-18) subjects, ± SEM.

Kohut et al.

10

Table S1. ED50 values with 95% confidence limits were calculated by standard linear regression analysis of each dose-effect curve (percent cocaine-lever responding as a function of dose).

Pretreatment (min) Cocaine Methylphenidate WIN35,428 AHN1-055 AHN2-005 JHW007

5 3.59 (2.88 - 4.18)

1.67 (1.45 - 1.92)

0.52 (0.37 - 0.75) N.S. N.S. 4.80

(2.19 - 393)

20 4.45 (3.57 - 5.26) - - - - -

30 - 1.67 (1.31 - 2.10)

0.28 (0.19 - 0.39) N.S. 11.4

(8.32 - 19.9) -

40 6.14 (4.93 - 7.62) - - - - -

60 12.98 (11.3 - 15.3)

1.82 (1.20 - 2.58)

0.21 (0.13 - 0.31) N.S. 8.63

(4.20 - 165) N.S.

90 21.9 (11.9 – 69.3) - - 5.57

(4.22 - 7.60) 14.6

(8.64 - 43.6) -

120 14.05 (9.11 - 26.5)

3.15 (1.57 - 6.47)

0.47 (0.36 - 0.59)

3.70 (1.66 - 25.2)

13.1 (7.84 - 34.8) N.S.

180 - - 0.47 (0.28 - 0.79) - N.S. 4.18

(2.53 - 7.13)

240 - - 0.74 (0.49 - 1.26) - - -

360 - - 1.02 (0.75 - 1.43) - 22.5

(14.1 - 34.9) 4.41

(0.001 - 411)

720 - - - - - 6.49 (0.09 - 605)

1440 - - - - - N.S.

N.S., not significant.

Kohut et al.

11

Supplemental References

1. Tanda G, Ebbs A, Newman AH, Katz JL (2005): Effects of 4'-chloro-3 alpha-(diphenylmethoxy)-tropane on mesostriatal, mesocortical, and mesolimbic dopamine transmission: comparison with effects of cocaine. J Pharmacol Exp Ther 313:613-620.

2. Tanda G, Ebbs AL, Kopajtic TA, Elias LM, Campbell BL, Newman AH, et al. (2007): Effects of muscarinic M1 receptor blockade on cocaine-induced elevations of brain dopamine levels and locomotor behavior in rats. J Pharmacol Exp Ther 321:334-344.

3. Paxinos G, Watson C (1998): The Rat Brain in Stereotaxic Coordinates. IV ed. Sydney: Academic Press.

4. Agoston GE, Wu JH, Izenwasser S, George C, Katz J, Kline RH, et al. (1997): Novel N-substituted 3 alpha-[bis(4'-fluorophenyl)methoxy]tropane analogues: selective ligands for the dopamine transporter. J Med Chem 40:4329-4339.

5. Newman AH, Kline RH, Allen AC, Izenwasser S, George C, Katz JL (1995): Novel 4'-substituted and 4',4"-disubstituted 3 alpha-(diphenylmethoxy)tropane analogs as potent and selective dopamine uptake inhibitors. J Med Chem 38:3933-3940.

6. Snedecor GW, Cochran WG (1967): Statistical Methods. 6th ed. Iowa State University Press.