ARCHIVAL REPORT

Monoamine Levels Within thand Putamen Interact to PrePerformanceStephanie M. Groman, Alex S. James, Emanuele Seu

sendysbitorocsenaria

alont

anf asoorigh

al aysfu

Key Words: Cognitive control, dopamine, individual differences,

lity in

tonin

in these brain regions can alter reversal learning performance, itremains unknown whether naturally occurring variation in mono-amine levels explains individual differences in behavioral flex-ibility. Because the deficits in behavioral flexibility associated withneuropsychiatric disorders are likely linked to subtle, heritablepatterns of neurochemical anomalies, it remains crucial to dentify

From the Departments of Psychology (SMG, ASJ, ES, MAC, SH, JDS) and

Psychiatry and Biobehavioral Sciences (JDS), University of California,

Los Angeles, California.

Address correspondence to J. David Jentsch, Ph.D., UCLA, Department of

Psychology, PO Box 951563, Los Angeles, CA 90095-1563; E-mail:

Jentsch@psych.ucla.edu.

Received Sep 21, 2012; revised Dec 1, 2012; accepted Dec 6, 2012.

0006-3223/$36.00 BIOL PSYCHIATRY 2013;73:756762http://dx.doi.org/10.1016/j.biopsych.2012.12.002 & 2013 Society of Biological Psychiatrypreviously reinforced response (6) and the ability to modifybehaviors in a reversal learning task (7,8). In addition, efficient finding consistent with reported deficits in behavioral flexibi

Parkinsons disease (20).Although experimental depletion of dopamine and seroprefrontal cortex, response inhibition, serotonin, striatum

Adaptive modulation of behavior in response to environ-mental change is necessary for individuals to behave inboth a flexible and goal-directed manner. Conversely, the

compulsive and rigid behaviors that are present in someindividuals with behavioral addictions or obsessive-compulsivedisorder may result from impairments in the neural circuitry thatnormally allows individuals to flexibly update their behaviors.Consistent with this hypothesis, affected individuals exhibitdeficits in laboratory tasks of behavioral flexibility (13), as wellas abnormalities in brain regions that are critical for flexiblebehavior (4,5).

Behavioral flexibility can be assessed across species usingreversal learning procedures, which index an individuals abilityto adaptively modify behavior when previously learned rewardassociations are altered. The orbitofrontal cortex has been wellestablished as a region critical to flexible, adaptive responding, aslesions to the orbitofrontal cortex impair the extinction of a

associated with event-related recruitment of the lateral orbito-frontal cortex (9).

The orbitofrontal cortex is connected with both the caudatenucleus and putamen (10,11), through which it likely acts to altercognitive and motor behaviors. The striatum has long beenrecognized as an important brain region for simple forms of goal-directed behaviors (12,13), and recent evidence has indicatedthat the dorsal striatum is also involved in reversal learning(1416). Moreover, functional neuroimaging studies in humanshave demonstrated activation within the caudate nucleus andputamen during performance of a reversal learning task (9,17),suggesting that the ability to adaptively modify behaviors maybe governed by an integrative network involving both corticaland subcortical brain nuclei.

Within each of the brain regions implicated in reversallearning, a variety of neurotransmitters can participate in differ-ent ways to facilitate behavior. Neurotransmitter depletionstudies indicate that within the orbitofrontal cortex, serotoninplays an essential role, while dopamine does not (18). On theother hand, within the dorsal striatum, depletion of dopamine,but not serotonin, impairs reversal learning performance (19), aand James David Jentsch

Background: The compulsive and inflexible behaviors that are preand obsessive-compulsive disorder, may be due to neurochemicalExperimental removal of serotonin or dopamine within the orresponding in a reversal learning test, suggesting that these neudirected behaviors. Nevertheless, the behavioral impairments predifferences, and it remains unknown whether naturally occurring vin flexible, reward-directed behaviors.

Methods: The current study assessed the ability of 24 individuproblems and examined whether monoamine levels in the orbitofrbehavior.

Results: The interaction between dopamine levels in the putamenvariance in a measure of behavioral flexibility but not measures ohyperbolic function, with reversal learning performance being pserotonin and putamen dopamine or in monkeys with relatively h

Conclusions: These results support the hypothesis that subcorticgoal-directed behavior, providing insight into the neurochemical dpresent in psychiatric disorders.e Orbitofrontal Cortexdict Reversal Learning

, Maverick A. Crawford, Sandra N. Harpster,

t in many psychiatric disorders, particularly behavioral addictionsfunction within the circuitry that enables goal-directed behaviors.frontal cortex or dorsal striatum, respectively, impairs flexiblehemical systems exert important modulatory influences on goal-t in psychiatric disorders are likely due to subtle neurochemicaltion in neurochemical levels associate with individual differences

juvenile monkeys to acquire, retain, and reverse discriminational cortex, caudate nucleus, and putamen could explain variance in

d serotonin levels in the orbitofrontal cortex explained 61% of thesociative learning or memory. The interaction mirrored that of aest in either monkeys with relatively low levels of orbitofrontallevels of orbitofrontal serotonin and putamen dopamine levels.

nd cortical neuromodulatory systems interact to guide aspects ofnction that may underlie the inflexible and compulsive behaviors

updating of behavior during reversal in normal humans is

all subjects were housed in peer groups during the period

S.M. Groman et al. BIOL PSYCHIATRY 2013;73:756762 757of testing, which began at 2 years of age and ended by 3 yearsof age. Subjects had unlimited access to water and receivedtwice daily portions of standard monkey chow (Teklad; HarlanLaboratories, Indianapolis, Indiana). Chow was never withheld orreduced to motivate performance of the tasks.

All monkeys were maintained in accordance with the Guide forthe Care and Use of Laboratory Animals of the Institute ofLaboratory Animal Resources, National Research Council, Depart-ment of Health, Education and Welfare Publication No. (NIH)85-23, revised 1996. Research protocols were approved by theUniversity of California, Los Angeles, Chancellors Animal ResearchCommittee.

Discrimination Acquisition, Retention, and Reversal LearningSubjects were assessed for their ability to acquire, retain, and

reverse visual discrimination problems, using procedures similarto those previously described (22,23). Testing was conducted 6 to7 days per week. Monkeys were trained to enter into a metaltunnel that was fitted with a modified Wisconsin General TestingApparatus. This apparatus was equipped with an opaque screenthat could be raised or lowered to present three equally spacedopaque boxes to the monkey. These boxes were fitted withhinged lids that allowed for small pieces of food reward (i.e., apiece of banana, apple, or grape) to be concealed within andinserts on the lid where unique visual stimuli could be displayed(clip art).whether normal variation in behavioral abilities are linked withnormal variation in neurotransmitter function. One approach fordetermining the neurochemical basis of cognition is to relatedifferences in neurochemical measurements with differences inbehavior at an individual level (21). We have previously shownthat naturally occurring variation in striatal D2-like receptoravailability is associated with individual differences in reversallearning performance (22), suggesting that slight differences inneurotransmitter function may contribute to individual variabilityin behavioral flexibility. Furthermore, the experimental depletionstudies were conducted in independent studies and did notexamine the possibility that neuromodulatory systems withindiscrete brain regions functionally interact with each other toinfluence the neural circuitry that guides behavior.

The current study assessed a cohort of juvenile monkeys fortheir ability to acquire, retain, and reverse novel visual discrimi-nation problems, and their brain monoamine levels were subse-quently ascertained. We used these data to test a specifichypothesis: namely, whether the interaction between serotoninlevels in the orbitofrontal cortex and dopamine levels in theputamen and caudate nucleus predicted behavioral performance.Specifically, we hypothesized that the interaction between levelsof dopamine in the dorsal striatum and serotonin in theorbitofrontal cortex would predict performance during reversallearning but not the ability to acquire or retain visual discrimina-tion problems.

Methods and Materials

SubjectsTwenty-four male juvenile vervet monkeys (Chlorocebus

aethiops sabaeus) born at the University of California, LosAngeles, Vervet Research Colony were involved in this study.The subjects were recruited from two successive annual birthcohorts (12 from the 2004 cohort and 12 from the 2005 cohort);Testing sessions began when the opaque screen was raised topresent the three boxes (each fitted with a unique visualstimulus) to the monkey. One of the three visual stimuli wasuniformly associated with food reward, while the others werenot. Monkeys were only allowed to open one box per trial. A trialended after a correct choice (opening the box containing thefood reward), an incorrect choice (opening an empty box), or anomission (no response for 30 sec) occurred. The position of thevisual stimuli was pseudorandomly varied across trials. Subjectsreceived up to 30 trials per day.

Initially, subjects were trained to acquire four consecutivediscrimination problems, each involving three novel visual sti-muli. Subjects were required to learn which one of the threestimuli was associated with food reward by achieving a perfor-mance criterion (8 correct choices within 10 consecutive trials). Ifthe performance criterion was not achieved in 30 trials, thesession was terminated and the monkey was returned to thesocial enclosure. The same discrimination problem was presentedthe following day(s) until the performance criterion was met.

After completing the four discrimination-acquisition problems,monkeys were then trained to acquire, retain, and reverse noveldiscrimination problems. The first phase of each discriminationproblem was an acquisition phase that was identical to theacquisition-only training the monkeys had previously received. Oncethe same performance criterion had been met, the session wasterminated and the monkey was returned to the social enclosure.

One day after reaching the criterion for acquisition, the abilityof subjects to remember the stimulus-reward association fromthe day before was assessed in a retention phase, where thestimulus-reward contingencies remained unchanged relative tothe previous training day. This phase persisted until subjects metthe performance criterion of four correct responses in fiveconsecutive trials. Immediately after reaching the criterion inthe retention phase, the reversal phase of the discriminationproblem began; there were no environmental events that cuedthis transition, other than a change in reinforcement contingen-cies. In the reversal phase, the stimulus that was previouslyrewarded was no longer rewarded, and one of the two previouslyunrewarded stimuli was now rewarded. The reversal phasepersisted until monkeys achieved criterion (8 correct choices in10 consecutive trials) or until 30 trials had been completed,whichever occurred first. If monkeys did not reach the perfor-mance criterion within 30 trials, the reversal phase continued thefollowing day(s) until the performance criterion was met.Monkeys were assessed on their ability to acquire, retain, andreverse three novel discrimination problems. Due to inclementweather that prevented testing, one reversal session was notcompleted by four different monkeys, and these data wereexcluded from the analysis.

The primary dependent measures were the total number oftrials required to reach criterion and the number of correctresponses made in the acquisition, retention, and reversal phases.For the reversal phase, the total number of responses directed atthe previously rewarded stimulus (a perseverative response) andthe never rewarded stimulus (a neutral response) were alsorecorded. The proportion of each response type (correct, perse-verative, or neutral) was determined by dividing the total numberof each response by the number of trials required to reachcriterion for each phase.

Quantification of Monoamine LevelsApproximately 2 to 3 months after completing these beha-

vioral assessments, brain tissue was collected at necropsy, usingwww.sobp.org/journal

Figure 1. A depiction of the brain regions collected

758 BIOL PSYCHIATRY 2013;73:756762 S.M. Groman et al.methods previously described (23,24). Tissue homogenates wereassayed using high pressure liquid chromatography similar toprocedures described elsewhere (24) and protein content inhomogenates was quantified using the Lowry method (25).

Tissue was acquired from a wide range of brain regions.However, we restricted our analyses here to monoamine levels inthe orbitofrontal cortex, putamen, and caudate nucleus collectedfrom regions identified in Figure 1. Based on previous evidence(11), tissue acquired from the caudate nucleus and putamen werewithin the terminal fields of the orbitofrontal and dorsal anteriorcingulate cortex. Although behavioral data were collected in 24monkeys, brain tissue was only collected from 20 monkeys, as 4subjects were transferred to another behavioral protocol.

Statistical AnalysesAll statistical analyses were conducted using SPSS 15.0 (SPSS

Inc, Chicago, Illinois). The reliability of the behavioral measure-ments was determined using Cronbachs a, a coefficient ofreliability. Linear regressions using cohort (2004 vs. 2005) as anindependent variable were then conducted on all measurementsbefore any further analysis to remove extraneous variance thatwas attributable to cohort. These residuals were then centered

for further statistical analysis. To examine the relationship bet-ween monoamine levels and behavior measurements, multiplelinear regression was conducted with individual monoaminelevels and their interaction term(s) as the independent variable(s)and the behavioral measurement(s) as the dependent variableusing a forced entry model:

Behavior b0 b1putamen or caudate dopamine levels b2 orbitofrontal serotonin levels b3putamen or caudate dopamine levels orbitofrontal serotonin levels e

Results

Acquisition, Retention, and Reversal LearningReliability of the number of trials required to reach criterion

across the acquisition and reversal sessions was moderately high(acquisition: Cronbachs a .72, reversal: Cronbachs a .62).However, contrary to our previous observations, reliability of thenumber of trials required to reach criterion across the three

www.sobp.org/journalretention sessions was surprisingly low (Cronbachs a .33). Thesubjects required fewer trials to reach criterion in the acquisitionphase (56.2 28.6) compared with the reversal phase(78.2 38.7) (t22 4.93, p .001), indicating that they foundthe reversal of the stimulus-reward contingencies more difficultthan their initial acquisition. Descriptive statistics for the errortype in the reversal phase revealed that the probability of makinga response to the initially reinforced stimulus was significantlygreater than the probability of making a response to the never-rewarded stimulus (t23 15.58, p .001), showing that theincrease in the number of trials required to reach criterion duringreversal was driven by an increased probability of making aperseverative response, rather than neutral responses.

Monoamine LevelsThe average levels of dopamine and standard error, normalized

to protein content, in the putamen and caudate nucleus were101 13.2 ng/mg and 110 5.58 ng/mg, respectively. The aver-age level of serotonin and standard error within the orbitofrontalcortex, normalized to protein content, was 1.94 .29 ng/mg.

Neurochemical Correlates of Behavior

for quantification of monoamine levels. The shadedarea in (A) (predominately Brodmann area 47)represents the area of the orbitofrontal cortexextracted (Bregma 13.50 mm; interaural 35.40 mm).Tissue punches from the caudate nucleus and puta-men (represented by the gray circles) were collectedfrom tissue slices similar to that presented in (B)(Bregma 2.70 mm; interaural 24.60 mm). (Reprintedfrom Paxinos et al. [37], with permission from Else-vier, copyright 2009.)Multiple linear regression was used to examine the relation-ship between serotonin and dopamine levels in the orbitofrontalcortex and putamen, respectively, and the behavioral measurescollected. First, the relationship between orbitofrontal serotoninlevels, putamen dopamine levels, and their interaction wasexamined with respect to the average number of trials requiredto reach criterion in the acquisition, retention, and reversalphases. Neither levels of dopamine in the putamen or serotoninin the orbitofrontal cortex nor the interaction of the twopredicted a significant amount of variance in either acquisitionor retention performance (acquisition: F3,14 1.97, p .16; reten-tion: F3,11 .079, p .97) (Figure 2A,B). Moreover, the maineffects of orbitofrontal serotonin and putamen dopamine levelsdid not predict a significant proportion of variance in the numberof trials required to reach criterion in the reversal phase [F(2,15) .74, p .49]. However, as hypothesized, the regression model,including the interaction term, was significant (R2 changeF1,14 18.82, p .001), accounting for 61% of the variance inthe average number of trials required to reach criterion inthe reversal phase (F3,11 7.36; p .003). In this model, the

S.M. Groman et al. BIOL PSYCHIATRY 2013;73:756762 759regression coefficients for orbitofrontal serotonin and putamendopamine levels remained nonsignificant (all 1.5), whilethe regression coefficient for the interaction was significant(t11 4.34, p .001), explaining 52% more of the variance thandid the main effects of dopamine and serotonin levels. Theinteraction regression coefficient was positive, such that as levels

Figure 2. A three-dimensional plot of the relationship between standar-dized levels of dopamine in the putamen, standardized levels of serotoninin the orbitofrontal cortex, and standardized behavioral performancemeasures in the acquisition (A), retention (B), and reversal (C) phases.Values for individual monkeys are presented in closed circles. Thetransparent plane overlaid on the individual data points represents howthe interaction between levels of dopamine in the putamen andserotonin in the orbitofrontal cortex predicts the number of trials requiredto reach criterion in each of the phases of the task. Relatively high trials toreach criterion is indicative of poorer performance in each stage.of dopamine in the putamen increased by one standardized unit,the slope of the relationship between orbitofrontal serotoninand number of trials required to reach criterion during reversallearning increased by a factor of .74. Monkeys with the lowestdopamine and serotonin levels required the greatest number oftrials to reach criterion in the reversal. Among monkeys with lowdopamine levels in the putamen (i.e., one standard deviationbelow the mean), relatively higher levels of orbitofrontal corticalserotonin were associated with fewer trials required to reachcriterion (better performance). However, high putamen dopa-mine levels (i.e., one standard deviation above the mean) wereassociated with a positive relationship between serotonin andreversal phase trials: for monkeys with the highest putamendopamine levels, low serotonin was associated with goodreversal learning performance (Figure 2C). Therefore, the inter-action between orbitofrontal serotonin and putamen dopaminelevels on reversal performance mirrored that of a hyperbolicfunction, with reversal learning performance being poorest ineither monkeys with relatively low levels of orbitofrontal ser-otonin and putamen dopamine or in monkeys with relativelyhigh levels of orbitofrontal serotonin and putamen dopaminelevels.

Next, we sought to determine the neurochemical and anato-mical specificity of the relationships mentioned above withreversal performance. To do this, levels of serotonin in theputamen and dopamine in the orbitofrontal cortex and theinteraction of the two were regressed against the averagenumber of trials required to reach criterion in the reversal phase.This model did not account for a significant proportion ofvariance in reversal performance (F3,14 .997, p .42)(Figure 3A), providing evidence for the selective neurochemicalinvolvement of dopamine in the putamen and serotonin in theorbitofrontal cortex with reversal learning.

Finally, levels of dopamine in the caudate nucleus andserotonin in the orbitofrontal cortex and their interaction wereregressed against reversal learning performance. Contrary to ourhypothesis, the regression model failed to account for a sig-nificant proportion of variance in reversal learning performance(F3,14 .34; p .80) (Figure 3B).

Discussion

The results of the current study demonstrate that individualdifferences in monoaminergic systems within distinct brain nucleiinteract to predict flexibility of behavior during a reversal learningtask. By contrast, these neurochemical measures did not predictbasic associative learning or memory (acquisition or retention)performance. Importantly, the relationship between serotonin inthe orbitofrontal cortex, dopamine in the putamen, and reversallearning was highly specific because a variant model, whichincluded dopamine in the orbitofrontal cortex and serotoninin the putamen as predictors, did not explain a significantamount of variance in reversal performance. Finally, the findingthat levels of dopamine in the putamen, and not in a region ofanatomical proximity (the caudate nucleus), modified the rela-tionship between orbitofrontal serotonin and reversal learningperformance implicates anatomical specificity of the observedrelationships.

Neurochemical Interactions Within Discrete Brain RegionsPredict Reversal Learning Performance

Previous studies have found that depletion of serotonin in theorbitofrontal cortex or dopamine in the dorsal striatum canwww.sobp.org/journal

760 BIOL PSYCHIATRY 2013;73:756762 S.M. Groman et al.cause reversal learning deficits (18,19,26). In these studies, largeand systematic reductions in local neurotransmitter levelswere produced. However, recently, individual differences inmethylphenidate-induced dopamine release were reported tocorrelate with methylphenidate-induced changes in reversallearning performance, indicating that dopaminergic differences,an order of magnitude smaller than depletion studies, within thestriatum may directly influence reversal learning performance(27). Building upon these results, the current study providesevidence that orbitofrontal serotonin levels and putamen dopa-mine levels functionally interact, explaining 61% of the variancein individual reversal learning performance. The interactionfollowed a hyperbolic function: in monkeys with low putamendopamine levels, poor reversal learning was associated withrelatively low orbitofrontal cortex serotonin; however, as dopa-mine levels in the putamen increased, the relationship betweenserotonin and reversal learning reversed, with low levels ofserotonin predicting relatively better reversal learning perfor-mance. The selective relationship between levels of orbitofrontal

Figure 3. A three-dimensional plot of the relationship between pre-frontal and striatal neurochemical levels on reversal learning performance.Values for individual monkeys are presented in closed circles. Thetransparent plane overlaid on the individual data points represents howthe interaction between levels of monoamines predict the number oftrials required to reach criterion in the reversal phase. (A) presents therelationship between standardized levels of serotonin in the putamen,dopamine in the orbitofrontal cortex, and reversal learning performance.(B) presents the relationship between standardized levels of dopamine inthe caudate nucleus, serotonin in the orbitofrontal cortex, and reversallearning performance. Relatively high trials to reach criterion is indicativeof poorer performance in each stage.

www.sobp.org/journalserotonin and putamen dopamine on reversal learning perfor-mance is likely linked to the increases in perseverative respond-ing that occur during the reversal phase of the task, suggestingthat a functional circuit between these monoamine systems mayunderlie the ability of individuals to inhibit habitual, prepotentresponses.

The current study did not detect a significant interactionbetween dopamine levels in the caudate nucleus, orbitofrontalserotonin, and reversal learning performance, which apparentlycontrasts with the effects of depletion of dopamine in thedorsomedial striatum on reversal learning (19,26). Neuroimagingstudies in humans have found increased activation duringreversal learning in both the putamen (17) and in the caudate(9). Moreover, individual differences in reversal learning perfor-mance in monkeys has been associated with positron emissiontomography based measurements of dopamine D2-like receptorsin both the caudate nucleus and putamen (22), indicating thatdopaminergic signaling in both striatal regions is important forflexible, goal-directed behaviors. Because the prefrontal cortexbroadly innervates the striatum (11,28), dopamine within thecaudate nucleus may interact with serotonin tone in otherprefrontal regions, not investigated here, that are involved ingoal-directed behaviors, such as the anterior cingulate cortex(29), and these cortical-striatal circuits work in parallel to encodediscrete types of information that together guide goal-directedbehavior. Of additional note, our neurochemical assessmentswere made in a specific orbitofrontal cortex region (area 47),though past lesion studies examining effects on reversal learninghave involved other or additional orbitofrontal cortex regions.Given the potential functional heterogeneity of the orbitofrontalcortex, it is possible that the pattern of results we report here arespecific to area 47, with different results found, for example, inmore ventromedial orbitofrontal cortex regions often implicatedin reversal performance (6,30,31). Systematic investigation intothe neurochemical mechanisms within the various orbitofrontalcortex subregions is needed and may provide insight into howthese neurochemical systems interact with discrete striatalcompartments to guide behavior.

Potential Links Between Putamen Dopamine andOrbitofrontal Serotonin

Although the mechanistic nature of the interaction betweenorbitofrontal serotonin and putamen dopamine systems onreversal learning is unknown, there is evidence to support theidea that striatal dopamine signaling, working through D2-likereceptors, mediates evoked prefrontal serotonin release. Specifi-cally, pharmacologic blockade of striatal D2-like receptors sig-nificantly attenuates tail-pinch evoked increases in prefrontal andstriatal serotonin efflux (32). Therefore, dopaminergic stimulationof D2-like receptors within the striatum may directly influenceserotonin release within the orbitofrontal cortex, potentiallyexplaining the interaction between these two neurochemicalsystems on tasks of behavioral flexibility.

Conversely, the prefrontal serotonin system may exert a top-down influence over striatal dopamine release. Pharmacologicactivation of serotonin 2C receptors within the prefrontal cortexpotentiates cocaine-induced striatal dopamine release (33), impli-cating a facilitatory pathway between prefrontal serotonin andstriatal dopamine. Based on these data, serotonin within theorbitofrontal cortex and dopamine within the putamen maycoalesce at the network level, directly influencing one another,and explaining the phenomena reported here. However, thedirection of the interaction between dopamine tone in the

striatum and serotonin levels in the orbitofrontal cortex on 5. Chamberlain SR, Menzies L, Hampshire A, Suckling J, Fineberg NA, del

S.M. Groman et al. BIOL PSYCHIATRY 2013;73:756762 761reversal learning performance remains unknown.

Implications for Psychiatric DisordersNeurochemical dysfunction is linked with many of the

behavioral impairments that are present in individuals diag-nosed with addictions. For example, striatal dopaminergicdysfunction in substance-dependent individuals has been foundto associate with self-report measures of impulsivity, levels ofself-reported craving for drugs, and cognitive abilities (34).Notably, dopaminergic alterations within the striatum that occurin response to chronic drug exposure have been found tocorrelate with drug-induced changes in reversal learning per-formance (23). Variation in central serotonin systems alsoassociates with impulsiveness, and pharmacologic manipulationof cortical serotonin transmission signaling alters reversal learn-ing performance (35,36).

The results of the current study indicate that levels oforbitofrontal serotonin interact with dopaminergic tone in theputamen to influence reversal learning performance. Therefore,in different individuals, the neurochemical etiology of reversallearning deficits may be slightly different, despite the revealedbehavioral deficit being identical, and correspondingly, thepharmacologic treatments targeted to improve behavioral flex-ibility would therefore differ. These data support the notion thatpharmacologic treatment strategies may achieve greater efficacyif they take into account, and are targeted at remediating, anindividuals particular set of regionally specific neurotransmittersystem abnormalities, each of which may exhibit complex cross-system interactions that moderate both the behavior itself and anindividuals response to pharmacologic intervention.

These data also support the notion that, while experimentalmanipulations that produce extreme changes in a neurochemicalsystem (e.g., depletion) can reveal the necessary influences ofindividual transmitters on behavioral processes, additional meth-ods may be required to reveal the subtle interactions betweenneurotransmitters that emerge at a network level, each of whichmay be both complex and multifaceted. Further studies of therelationship between individual differences in brain neurotrans-mitter phenotypes and behavior may help to decompose the rolefor distinct and interacting neuromodulatory systems in aspectsof cognition.

These studies were supported by Public Health Service GrantsP20-DA022539, P50-MH077248, and RL1-MH083270 (JDJ). Addi-tional support was derived from Public Health Service Grants F31-DA028812 and T32-DA024635.

The authors report no biomedical financial interests or potentialconflicts of interest to disclose.

1. Ghahremani DG, Tabibnia G, Monterosso J, Hellemann G, Poldrack RA,London ED (2011): Effect of modafinil on learning and task-relatedbrain activity in methamphetamine-dependent and healthy indivi-duals. Neuropsychopharmacology 36:950359.

2. Ersche KD, Roiser JP, Robbins TW, Sahakian BJ (2008): Chronic cocainebut not chronic amphetamine use is associated with perseverativeresponding in humans. Psychopharmacology (Berl) 197:421431.

3. Valerius G, Lumpp A, Kuelz AK, Freyer T, Voderholzer U (2008):Reversal learning as a neuropsychological indicator for the neuro-pathology of obsessive compulsive disorder? A behavioral study.J Neuropsychiatry Clin Neurosci 20:210218.

4. Thompson PM, Hayashi KM, Simon SL, Geaga JA, Hong MS, Sui Y, et al.(2004): Structural abnormalities in the brains of human subjects whouse methamphetamine. J Neurosci 24:60286036.Campo N, et al. (2008): Orbitofrontal dysfunction in patients withobsessive-compulsive disorder and their unaffected relatives. Science321:421422.

6. Butter CM, Mishkin M, Rosvold HE (1963): Conditioning and extinctionof a food-rewarded response after selective ablations of frontal cortexin rhesus monkeys. Exp Neurol 7:6575.

7. Dias R, Robbins TW, Roberts AC (1996): Dissociation in prefrontalcortex of affective and attentional shifts. Nature 380:6972.

8. Fellows LK, Farah MJ (2003): Ventromedial frontal cortex mediatesaffective shifting in humans: Evidence from a reversal learningparadigm. Brain 126:18301837.

9. Ghahremani DG, Monterosso J, Jentsch JD, Bilder RM, Poldrack RA(2010): Neural components underlying behavioral flexibility in humanreversal learning. Cereb Cortex 20:18431852.

10. Selemon LD, Goldman-Rakic PS (1985): Longitudinal topography andinterdigitation of corticostriatal projections in the rhesus monkey. JNeurosci 5:776794.

11. Haber SN, Kim KS, Mailly P, Calzavara R (2006): Reward-related corticalinputs define a large striatal region in primates that interface withassociative cortical connections, providing a substrate for incentive-based learning. J Neurosci 26:83688376.

12. Yin HH, Ostlund SB, Knowlton BJ, Balleine BW (2005): The role of thedorsomedial striatum in instrumental conditioning. Eur J Neurosci 22:513523.

13. Yin HH, Knowlton BJ, Balleine BW (2006): Inactivation of dorsolateralstriatum enhances sensitivity to changes in the action-outcome con-tingency in instrumental conditioning. Behav Brain Res 166:189196.

14. Castane A, Theobald DE, Robbins TW (2010): Selective lesions of thedorsomedial striatum impair serial spatial reversal learning in rats.Behav Brain Res 210:7483.

15. Divac I, Rosvold HE, Szwarcbart MK (1967): Behavioral effects ofselective ablation of the caudate nucleus. J Comp Physiol Psychol 63:184190.

16. Clarke HF, Robbins TW, Roberts AC (2008): Lesions of the medialstriatum in monkeys produce perseverative impairments duringreversal learning similar to those produced by lesions of theorbitofrontal cortex. J Neurosci 28:1097210982.

17. Freyer T, Valerius G, Kuelz AK, Speck O, Glauche V, Hull M,Voderholzer U (2009): Test-retest reliability of event-related functionalMRI in a probabilistic reversal learning task. Psychiatry Res 174:4046.

18. Clarke HF, Dalley JW, Crofts HS, Robbins TW, Roberts AC (2004):Cognitive inflexibility after prefrontal serotonin depletion. Science 304:878880.

19. Clarke HF, Hill GJ, Robbins TW, Roberts AC (2011): Dopamine, but notserotonin, regulates reversal learning in the marmoset caudatenucleus. J Neurosci 31:42904297.

20. Peterson DA, Elliott C, Song DD, Makeig S, Sejnowski TJ, Poizner H(2009): Probabilistic reversal learning is impaired in Parkinsonsdisease. Neuroscience 163:10921101.

21. Sahakian BJ, Sarna GS, Kantamaneni BD, Jackson A, Hutson PH,Curzon G (1985): Association between learning and cortical catecho-lamines in non-drug-treated rats. Psychopharmacology (Berl) 86:339343.

22. Groman SM, Lee B, London ED, Mandelkern MA, James AS, Feiler K,et al. (2011): Dorsal striatal D2-like receptor availability covaries withsensitivity to positive reinforcement during discrimination learning. JNeurosci 31:72917299.

23. Groman SM, Lee B, Seu E, James AS, Feiler K, Mandelkern MA, et al.(2012): Dysregulation of d2-mediated dopamine transmission inmonkeys after chronic escalating methamphetamine exposure. JNeurosci 32:58435852.

24. Jentsch JD, Tran A, Le D, Youngren KD, Roth RH (1997): Subchronicphencyclidine administration reduces mesoprefrontal dopamine uti-lization and impairs prefrontal cortical-dependent cognition in the rat.Neuropsychopharmacology 17:9299.

25. Lowry OH, Rosebrough NJ, Farr AL, Randall RJ (1951): Proteinmeasurement with the Folin phenol reagent. J Biol Chem 193:265275.

26. ONeill M, Brown VJ (2007): The effect of striatal dopamine depletionand the adenosine A2A antagonist KW-6002 on reversal learning inrats. Neurobiol Learn Mem 88:7581.

27. Clatworthy PL, Lewis SJ, Brichard L, Hong YT, Izquierdo D, Clark L,et al. (2009): Dopamine release in dissociable striatal subregionswww.sobp.org/journal

predicts the different effects of oral methylphenidate on reversallearning and spatial working memory. J Neurosci 29:46904696.

28. Haber SN (2003): The primate basal ganglia: Parallel and integrativenetworks. J Chem Neuroanat 26:317330.

29. Chudasama Y, Daniels TE, Gorrin DP, Rhodes SE, Rudebeck PH, MurrayEA (2012): The role of the anterior cingulate cortex in choices basedon reward value and reward contingency [published online ahead ofprint September 3]. Cereb Cortex

30. Iversen SD, Mishkin M (1970): Perseverative interference in monkeysfollowing selective lesions of the inferior prefrontal convexity. ExpBrain Res 11:376386.

31. Rudebeck PH, Murray EA (2008): Amygdala and orbitofrontal cortexlesions differentially influence choices during object reversal learning.J Neurosci 28:83388343.

32. Mendlin A, Martin FJ, Jacobs BL (1999): Dopaminergic input isrequired for increases in serotonin output produced by behavioralactivation: An in vivo microdialysis study in rat forebrain. Neuroscience93:897905.

33. Leggio GM, Cathala A, Moison D, Cunningham KA, Piazza PV,Spampinato U (2009): Serotonin 2C receptors in the medial prefrontalcortex facilitate cocaine-induced dopamine release in the rat nucleusaccumbens. Neuropharmacology 56:507513.

34. Lee B, London ED, Poldrack RA, Farahi J, Nacca A, Monterosso JR, et al.(2009): Striatal dopamine d2/d3 receptor availability is reduced inmethamphetamine dependence and is linked to impulsivity. JNeurosci 29:1473414740.

35. Boulougouris V, Robbins TW (2010): Enhancement of spatial reversallearning by 5-HT2C receptor antagonism is neuroanatomically spe-cific. J Neurosci 30:930938.

36. Brigman JL, Mathur P, Harvey-White J, Izquierdo A, Saksida LM, BusseyTJ, et al. (2010): Pharmacological or genetic inactivation of theserotonin transporter improves reversal learning in mice. Cereb Cortex20:19551963.

37. Paxinos G, Huang X-F, Toga AW (2009): The Rhesus Monkey Brain inStereotaxic Coordinates. San Diego: Academic Press.

762 BIOL PSYCHIATRY 2013;73:756762 S.M. Groman et al.www.sobp.org/journal

Monoamine Levels Within the Orbitofrontal Cortex and Putamen Interact to Predict Reversal Learning PerformanceMethods and MaterialsSubjectsDiscrimination Acquisition, Retention, and Reversal LearningQuantification of Monoamine LevelsStatistical Analyses

ResultsAcquisition, Retention, and Reversal LearningMonoamine LevelsNeurochemical Correlates of Behavior

DiscussionNeurochemical Interactions Within Discrete Brain Regions Predict Reversal Learning PerformancePotential Links Between Putamen Dopamine and Orbitofrontal SerotoninImplications for Psychiatric Disorders

Implications for Psychiatric Disorders