Gene–gene effects on central processing of aversive stimuli

ORIGINAL ARTICLE

Gene–gene effects on central processing of aversivestimuliMN Smolka1,2, M Buhler2,3, G Schumann2,4, S Klein2, X-Z Hu5, M Moayer2, A Zimmer2, J Wrase2,6,

H Flor2, K Mann2, DF Braus7, D Goldman5 and A Heinz2,6

1Department of Psychiatry, Technische Universitat Dresden, Dresden, Germany; 2Central Institute of Mental Health,Mannheim, Germany; 3Graduate School of Neural and Behavioural Sciences, University of Tubingen, Tubingen, Germany;4King’s College London, Section of Addiction Biology, Institute of Psychiatry, London, UK; 5Laboratory of Neurogenetics,National Institute on Alcohol Abuse and Alcoholism, NIH, Bethesda, MD, USA; 6Department of Psychiatry of the Charite,Charite Campus Mitte, Berlin, Germany and 7Department of Psychiatry, University of Hamburg, Hamburg, Germany

Emotional reactivity and regulation are fundamental to human behavior. As inter-individualbehavioral variation is affected by a multitude of different genes, there is intense interest toinvestigate gene–gene effects. Functional sequence variation at two genes has beenassociated with response and resiliency to emotionally unpleasant stimuli. These genesare the catechol-O-methyltransferase gene (COMT Val158Met) and the regulatory region(5-HTTLPR) of the serotonin transporter gene. Recently, it has been proposed that 5-HTTexpression is not only affected by the common S/L variant of 5-HTTLPR but also by an A to Gsubstitution. Using functional magnetic resonance imaging, we assessed the effects of COMTVal158Met and both 5-HTT genotypes on brain activation by standardized affective visualstimuli (unpleasant, pleasant, and neutral) in 48 healthy subjects. Based on previous studies,the analysis of genotype effects was restricted to limbic brain areas. To determine allele-doseeffects, the number of COMT Met158 alleles (i.e., lower activity of COMT) and the number of5-HTT low expressing alleles (S and G) was correlated with the blood oxygen level-dependent(BOLD) response to pleasant or unpleasant stimuli compared to neutral stimuli. We observedan additive effect of COMT and both 5-HTT polymorphisms, accounting for 40% of the inter-individual variance in the averaged BOLD response of amygdala, hippocampal and limbiccortical regions elicited by unpleasant stimuli. Effects of 5-HTT and COMT genotypes did notaffect brain processing of pleasant stimuli. These data indicate that functional brain imagingmay be used to assess the interaction of multiple genes on the function of neuronal networks.Molecular Psychiatry (2007) 12, 307–317. doi:10.1038/sj.mp.4001946; published online 9 January 2007

Keywords: fMRI; COMT; 5-HTT; genetics; emotion; amygdala

Introduction

Most common behavioral traits and disorders havecomplex patterns of inheritance, which are due toenvironment� gene interactions, genetic heterogene-ity and combined gene effects (including gene� geneinteraction), among other factors.1 Seemingly distincttraits can be causally related at the level of sharedintermediate phenotypes, including heritable ‘endo-phenotypes’,2 which may be controlled by subsets ofgenes that determine the genetically more complexclinical phenotypes.3 To understand the importanceof combined gene effects in behavior, the mostpowerful level of analysis would appear to be theintermediate phenotype, because at this level other

origins of complexity have been minimized. Emo-tional reactivity and the associated brain circuits mayrepresent such an intermediate phenotype, because ofits fundamental role in human behavior.

Genes have been identified that affect emotionalreactivity. Catechol-O-methyltransferase (COMT),which metabolizes dopamine (DA), norepinephrine(NE) and epinephrine (EPI) contains a Val158Metpolymorphism4 with a codominant effect.5 Whereasancient in humans, Met158 is not seen in otherprimates.6 The evolutionarily more recent Met158

allele is unstable at 371C and has 13 to 1

4 of the activityof the enzyme containing Val158.7,8 While carriers ofthe Val158 allele showed a small but significantincrease in perseverative errors in an executivefunction task9–11 and reduced working memory func-tion,12 recent findings indicate that a compensatingadvantage of the Val158 allele may be an increase inemotional resilience against anxiety, affective dis-orders and sustained pain.13–18 Behavioral geneticshas revealed an effect of the COMT Val158Met

Received 17 April 2006; revised 30 October 2006; accepted 6November 2006; published online 9 January 2007

Correspondence: Dr A Heinz, Department of Psychiatry andPsychotherapy, Charite Campus Mitte, Charite – Universitatsme-dizin Berlin, Schumannstr. 20-21, 10117 Berlin, Germany.E-mail: [email protected]

Molecular Psychiatry (2007) 12, 307–317& 2007 Nature Publishing Group All rights reserved 1359-4184/07 $30.00

www.nature.com/mp

genotype4 on negative mood states17,19 and on anotherintermediate phenotype, executive cognitive func-tion.9 A functional20,21 22–23-bp imperfect repeatpolymorphism in the regulatory region (5-HTTLPR)of the serotonin transporter gene (SCL6A4) creates ashort (S) allele and a long (L) allele (14- and 16-repeatalleles) and alters promoter activity. The S varianthas been reported to be associated with lower 5-HTTexpression levels, resulting in lower serotoninuptake activity when compared with the L variant.20

Subsequently, different groups conducted sequencingscreens and identified further variants of SLC6A4.22–24

Of special interest is a single-nucleotide polymorph-ism (SNP) with an A to G substitution (dbSNP:rs25531) and with a minor allele frequency > 10%.Although it is controversial whether rs25531 islocated just upstream of the 5-HTTLPR24 or in oneof the extra repeats of the L allele,22,25 this SNP wasshown to be functional, altering expression and AP2transcription factor binding.25 Like the common Sand L alleles, the rs25531 G allele was found to beassociated with lower levels of 5-HTT mRNA expres-sion compared with the A allele.25 The deletion/insertion in the 5-HTTLPR seems to modulate traitanxiety20 and fear conditioning.26 Moreover, a gene�environment interaction between 5-HTTLPR andstressful life events in the development of mooddisorder and suicide attempts was observed,27 and agene� environment interaction for an orthologousserotonin transporter polymorphism in the Rhesusmacaque was also noted.28 Recent studies29–31 haveshown that including both genotypes that affect5-HTT transcription (i.e., 5-HTTLPR and rs25531)may help to analyze the association of 5-HTT functionand liability for different psychiatric disorders moreadequately.

Nevertheless effects of both genes (SLC6A4 andCOMT) on behavior are moderate: for example,5-HTTLPR explains no more than 4% of the variancein trait anxiety.20 A meta-analysis of studies linking5-HTTLPR to anxiety revealed that the overall effectof the locus is highly significant, however, on averageeach copy of the S allele led only to a 0.1 s.d.increment in trait anxiety. Such weak associationsmay easily escape replication32 and elude attempts toquantify combined gene effects or the impact ofgene� gene interactions33–35 on personality.

However, both the COMT and 5-HTT gene loci exertsubstantially stronger effects on central processingof unpleasant stimuli: in paradigms that assess thecentral processing of affective visual stimuli withfunctional magnetic resonance imaging (fMRI). Weand others have shown that both 5-HTTLPR36,37 andCOMT Val158Met38 account for substantial portionsof inter-individual variance in amygdala responseduring the presentation of aversive visual stimuli.The low expression 5-HTTLPR S allele was associatedwith higher responsiveness to unpleasant stimulicompared with the L allele. Interestingly, 5-HTTLPRpredicted not only amygdala activity but also func-tional connectivity with paralimbic areas and

morphometric differences in the amygdala and peri-genual cingulate.37,39 The number of low-activityCOMT-Met158 alleles was positively correlated withbrain activation elicited by aversive stimuli in the lefthippocampus and the right amygdala38 and predictedopioid system activations in the amygdala and else-where after painful stimuli.17 However, although 5-HTTLPR and COMT Val158Met have both been linkedto brain intermediate phenotypes for anxiety, com-bined gene–gene effects have so far not been reported.Moreover, it has not yet been explored whether thers25531 genotype, which affects 5-HTT expression,moderates central processing of emotional stimuli.

Therefore, we tested the hypothesis that process-ing of unpleasant stimuli in the limbic system isincreased with low expression of 5-HTT, that is,higher number of S and G alleles of 5-HTTLPR andrs25531, respectively, as well as with low COMTactivity, that is, the dose of met158 alleles. As 5-HTTand COMT activity should independently affect 5-HT,DA and NE levels, we hypothesized additive effectsof genotypes (i.e., no significant interactions).

Materials and methods

SubjectsForty-eight right-handed healthy male volunteers(age 41.279.4 (M7s.d.) years) participated in thestudy after providing informed, written consentaccording to the Declaration of Helsinki. The EthicsCommittee of the University of Heidelberg approvedthe study. All subjects were ‘Caucasians’ of centralEuropean descent. Twenty-nine of these subjectshave been published in previous studies,37,38 whilean additional number of subjects (n = 20) wererecruited for the current study. Standardized clinicalassessment with the Structured Clinical Interview Iand II40,41 was performed to exclude subjects with alifetime history of axis I or II psychiatric disorders,according to DSM IV and ICD 10. Present drug abusewas excluded with urine tests. Only subjects free ofany medication were included. Additionally, weassessed the level of education, anxiety (SymptomChecklist 90 Revised (SCL-90-R);42 State Trait AnxietyInventory (STAI)43) and depression (SCL-90-R;42

Center for Epidemiological Studies Depression Scale(CES-D); and 44 Hamilton Depression Rating Scale(HAMD)45).

Imaging studyFor emotion induction, we used affectively unplea-sant, pleasant and neutral pictures. Each category wasrepresented by 18 pictures. Pleasant and unpleasantcues were taken from the International AffectivePicture System (IAPS),46 in which images are stan-dardized across the dimensions of emotional valenceand arousal.47 IAPS catalog numbers were 1440, 2340,2391, 4220, 4250, 4680, 5260, 5450, 5470, 5480, 5623,5660, 5830, 7580, 8120, 8190, 8300 and 8510 forpositive pictures and 2800, 3000, 3015, 3080, 3102,3140, 3180, 3230, 3261, 3350, 6360, 6570, 9040, 9340,

Gene–gene effects on aversive stimuliMN Smolka et al

308

Molecular Psychiatry

9520, 9570, 9910 and 9921 for negative pictures.Participants were instructed to passively view thestimuli, because even simple rating tasks can alterbrain activation pattern.48

The stimuli were presented for 750 ms using anevent-related design and were arranged in an indivi-dually randomized order for each subject. To recon-struct the blood oxygen level-dependent (BOLD)event-related time course, it is necessary to sampledata points at different peristimulus time points. Thiswas achieved by a random jitter between intertrialinterval and acquisition time, resulting in an equaldistribution of data points after each single stimulus.The intertrial interval was randomized between 3and 6 acquisition times (i.e., 9.9–19.8 s). During theintertrial interval, a fixation cross was presented.

Scanning was performed with a 1.5 T whole-bodytomograph (Magnetom VISION; Siemens, Erlangen,Germany) equipped with a standard quadrature headcoil. For fMRI, 24 slices were acquired every 3.3 s(4 mm thickness, 1 mm gap) using an EPI-Sequence(TR = 1.8 ms, TE = 66 ms, a= 901) with in-plane resolu-tion of 64�64 pixels (FOV 220 mm), resulting in avoxel size of 3.4�3.4�5 mm3. fMRI slices wereoriented axially parallel to the AC-PC line accordingto Talairach and Tournoux.

For anatomical reference we acquired a morpholo-gical 3D T1-weighted magnetization prepared rapidgradient echo (MPRAGE) image data set (1�1�1 mm3 voxel size, FOV 256 mm, 162 slices, TR =11.4 ms, TE = 4.4 ms and a= 121) covering the wholehead.

After the MR scan, a subset of the stimuli (8/18 percategory) were again presented on a computermonitor for 6 s and assessed for arousal and valenceaccording to the standardized procedure described byBradley and Lang.49

Data analysisData were analyzed with Statistical ParametricMapping (SPM2; Wellcome Department of ImagingNeuroscience, London, UK). After temporal andspatial realignment, the structural 3D data set was co-registered to the first T2* image. The structural imagewas spatially normalized to a standard template usinga 12-parameter affine transformation with additionalnonlinear components. A nonlinear transformationwas subsequently applied to the T2* data and voxelswere re-sampled at a resolution of 3�3� 3 mm3. Thefunctional data were smoothed using an isotropicGaussian kernel for group analysis (12 mm FWHM).

Statistical analysis on the first (individual) levelwas performed by modeling the different conditions(unpleasant, pleasant and neutral pictures; deltafunctions convolved with a synthetic hemodynamicresponse function and its time derivative) as expla-natory variables within the context of the generallinear model (GLM) on a voxel-by-voxel basis withSPM2.

To detect associations between the genotypes andfMRI activation on a voxel by voxel basis, the contrast

images of all subjects (signal change of unpleasantversus neutral pictures and pleasant versus neutralpictures) were included in a second-level multipleregression analysis with SPM2. To model theassumed gene dose effects, genotypes were coded asthree covariates: COMT genotype was coded by thenumber of low-activity met158 alleles (0, 1 or 2);5-HTTPLR S/L genotype was coded as the number oflow-expression S alleles (0, 1 or 2); rs25531 genotypewas coded as number of low-expression G alleles (0, 1or 2). We decided to primarily model the effects of 5-HTTLPR and rs25531 independently owing to find-ings that the rs25531 SNP can be found in the contextof both S and L alleles,50 and that rs25531, but not5-HTTLPR, was associated with selective 5-HT reup-take inhibitor response.24 Because carriers of both along (L) allele and the G substitution (LG) in the 5-HTT gene have been suggested to functionally mimicS-carriers25 and no SG carriers were found, we addi-tionally modeled the data using a triallelic approach:5-HTTLPR LA/LG/S genotype was coded as thenumber of low-expression S or LG alleles (0, 1 or 2).

To assess how much variance is explained bygenetic variation (effects of single genotypes as wellas combined genotype effects) and to test for gene–gene interactions, we extracted the BOLD responsedata of single voxels from the SPM Beta images usingMatlab (The MathWorks, Natick, MA, USA). Afterexport from Matlab, data were analyzed with SPSS12.0 for Windows (SPSS Inc., Chicago, IL, USA) usingthe GLM Univariate procedure. The three genotypeswere entered as covariates and the BOLD responsedata of single voxels were used as the dependentvariable. To estimate the effect size of single geno-types (partial Eta2) and the combined gene effects (R2)on brain activation, only main effects were modeled.To test for interactions between genotypes, all two-way interactions were additionally included in themodel.

To estimate the global effect of genotype on limbicreactivity, we computed the mean BOLD response ofthe limbic system to pleasant and unpleasant stimuli,respectively. For this purpose, we extracted the BOLDresponse data of all voxels of the limbic volume ofinterest (VOI) (for details on the VOI, see below) fromthe SPM Beta images and averaged them usingMatlab. After export from Matlab, data were analyzedwith SPSS 12.0 using again the GLM Univariateprocedure, as described above.

The analysis of the genotype� task interaction wasrestricted to limbic areas (limbic VOI: amygdala,hippocampus, parahippocampus and cingulatedgyrus). We generated anatomical masks with theWFU PickAtlas toolbox51 (Department of Radiology,Wake Forest University School of Medicine, Winston-Salem, NC, USA) using the MNI-based aal atlas.52 Themask for the entire limbic system was generated bymerging bilateral masks for amygdala, hippocampus,parahippocampal gyrus, anterior, middle and poster-ior cingulum. These masks were also used to assessthe regional effects of genotype. The applied mask for


309


the entire limbic VOI consisted of 3107 voxels(3�3� 3 mm3), i.e. 84 ml. A statistical threshold ofP < 0.05 corrected for the VOI was used for theregression analyses conducted with SPM2. Correctionfor multiple testing was performed with the FalseDiscovery Rate (FDR) procedure. The FDR is theproportion of supra-threshold voxels that are falsepositives. In comparison, a traditional multiplecomparisons procedure (e.g., Bonferroni or randomfield correction) controls Family-Wise Error Rate(FWER) at or below alpha. FWER is the chance ofone or more false positives anywhere (not just amongsupra-threshold voxels). If there is truly no signal inthe image anywhere, then an FDR procedure controlsFWER, just as Bonferroni and random field methodsdo. If there is some signal in the image, an FDRmethod will be more powerful than the traditionalmethods.53

GenotypingThe genotype information and flanking sequences ofthe COMT polymorphism were collected using publicSNP databases and analyzed using the followingprocedure: DNA was isolated from venous bloodwith QIAamp DNA Mini Kit (Qiagen, 40724 Hilden,Germany). Polymerase chain reaction (PCR) was per-formed with HotStarTaq DNA polymerase (Qiagen),4 ng template DNA in a total volume of 25ml PCRreaction and the primers COMTfor: 50-CAC CTG TGCTCA CCT CTC CT-30 and COMTrev: 50-GGG TTT TCAGTG AAC GTG GT-30. PCR conditions: 951C – 15 mininitial heating, then 35 cycles of 941C – 60 s, 581C– 60 s, 721C – 60 s and final extension for 10 min at721C. After purification of the amplified products(QIAquick PCR purification kit, Qiagen), the oligo-nucleotide primer 50-CGG ATG GTG GAT TTC GCTGGC-30 was extended according to the manufacturer’sinstructions (AcycloPrime – FP SNP detection kit;Perkin Elmer, 63110 Rodgau, Germany) and floures-cence polarization measured using a Wallac Victor21420 Multilabel Counter (Perkin Elmer).

We assessed the genotype of the promoter of the5-HTTLPR gene with PCR using oligonucleotideprimers (stpr5, 50-GGCGTTGCCGCTCTGAATGC;int1, 50-CAGGGGAGATCCTGGGAGGA). PCR ampli-fication was carried out in a final volume of 30mlconsisting of 50 ng genomic DNA, 2.5 mM deoxyribo-nucleotides (dGTP/7-deaza-20-dGTP = 1/1), 0.1 mg ofsense and antisense primers, 10 mM Tris-HCl (pH 8.3),50 mM KCl, 1.5 mM MgCl2 and 1 U of Taq DNA poly-merase. Annealing was carried out at 611C for 30 s,extension at 721C for 1 min and denaturation at 951Cfor 30 s for 35 cycles. The resulting PCR product waselectrophoretically separated on a 1.5% agarose gel.

In the case of 50 nuclease genotyping of SCL6A4 Aand G alleles, oligonucleotide primers (F: GCAACCTCCCAGCAACTCCCTGTA; R: GAGGTGCAGGGGGATGCTGGAA) and dye-labeled probes were desig-ned to optimize allele discrimination using PrimerExpress software (ABI, Foster City, CA 94404, USA).Probes were designed that were specific for the A and

G alleles. The fluorogenic probes were labeled at the50 end with either FAM or VIC (A: 6FAM-CCCCCCTGCACCCCCAGCATCCC MGB; G: VIC-CCCCTGCACCCCCGGCATCCCC MGB).

PCR assay conditions were optimized as follows:DNA 1 ml (25–50 ng) and 24 ml PCR mastermix includ-ing both probes (120 nmol ADP and 60 nmol ICP) andPCR primers (200 nmol for each). PCR buffer included5 mmol/l MgCl, 1� Corebuffer, 0.2 mmol/l dATP,dGTP, dCTP and 0.4 mmol/l dUTP, 0.25 U/ml TaqGold and 0.01 U/ml Amperase UNG. A 25 ml volumewas added to each well of the 96-well optical plates(4306737, ABI) and sealed using optical caps(4323032, ABI). DNA amplification was achievedunder the following conditions: 2 min at 501C,10 min at 951C, then 40 cycles at 961C for 15 s and62.51C for 90 s. ABIPRISM 7700 Sequence detectionsystem software processed the fluorescence data andgenerated genotypes.

Results

Genotyping

For COMT, nine subjects were homozygous Met158/Met158 and 14 subjects were homozygous Val158/Val158;25 subjects were heterozygous. For 5-HTTLPR, 15subjects were homozygous L/L and 11 were homo-zygous S/S; 22 were heterozygous (Table 1). SCL6A4genotyping of the A > G SNP revealed one subjecthomozygous G/G and five subjects heterozygousG/A and 42 individuals homozygous A/A (Table 1).No A > G exchange was detected on S-alleles usingboth sequencing analysis and 50 nuclease genotyping.COMT and 5-HTT genotypes were uncorrelated(P > 0.5).

Behavioral data

Genotype groups did not differ with respect to age,education, ratings of anxiety or depression, whichwere all in the normal range (Table 2). By subjectiverating scales of valence and arousal, subjects experi-enced the positive pictures as more arousing(P < 0.001) and affectively more pleasant (P < 0.001)than the neutral pictures. Negative pictures were alsorated as more arousing (P < 0.001) and less pleasant

Table 1 Genotype frequencies

COMT SCL6A4

Val/Val Val/Met Met/Met14 25 9 48 rs25531 5-HTTLPR

3 8 2 13 A/A 15 L/L1 0 0 1 G/A0 1 0 1 G/G7 8 3 18 A/A 22 S/L1 2 1 4 G/A2 6 3 11 A/A 11 S/S


310


(P < 0.001) than the neutral slides. Arousal ratings ofpleasant and unpleasant stimuli were similar (P = 0.9)(Table 3). Ratings of valence and arousal were notrelated to 5-HTT or COMT Val158Met genotypes.

Global effects of COMT Val158Met and 5-HTT genotypeson processing of unpleasant stimuliRegression analyses revealed a significant (PFDR <0.05, T > 1.735) combined effect of COMT and 5-HTTgenotypes on activation by unpleasant stimuli ofmultiple limbic brain regions including the amygdala,hippocampus, parahippocampal gyrus and cingulategyrus: all of these combined effects were bilateral(Figure 1). In the entire limbic VOI, 90% of all voxels(i.e., 2794/3107 voxels) revealed the combined geno-type effect. Combined effects of 5-HTT and COMTgenotypes were restricted to unpleasant stimuli and

did not affect processing of pleasant stimuli. Fortypercent (P < 0.0001) of inter-individual variance infMRI BOLD response averaged across the entire VOIwas explained by the combined genotype effect.

The combined effect of COMT and both 5-HTTgenotypes on brain activity in the entire limbic VOIwas additive (Figure 2). There were no significanttwo-way interactions between genotypes (Pmin = 0.23)and the effects were also numerically not greaterthan the additive effect: After entering all two-wayinteractions to the model, the amount of varianceexplained was 42% and only slightly higher com-pared with the 40% when we modeled the maineffects of genotype only.

Using the same threshold as for the combined geneeffect (T > 1.735), COMT genotype alone predictedbrain activity in 1996 voxels when controlling for

Table 2 Socio-demographic and clinical data

All subjects (M7s.d.) Association with genotype (Pearson’s R)

COMTa 5-HTTLPRb rs25531c

Age 41.279.4 0.16 �0.01 �0.29Education years 11.772.7 0.05 0.13 �0.19Cigarettes per day 4.078.9 0.36 0.10 �0.09STAI state 33.577.1 0.02 0.08 �0.17STAI trait (t-value) 44.978.4 �0.24 0.03 �0.10SCL-90-R anxiety (t-value) 41.375.1 0.04 �0.05 �0.02HAMD 0.1470.52 0.24 0.09 �0.10CES-D (t-value) 40.776.7 0.04 0.08 �0.11SCL-90-R depression (t-value) 40.876.0 �0.09 �0.18 �0.12

Abbreviations: CES-D, Center for Epidemiological Studies Depression Scale; COMT, catechol-O-methyltransferase; HAMD,Hamilton Depression Rating Scale; SCL-90-R, Symptom Checklist 90 Revised; STAI, State Trait Anxiety Inventory.aNumber of Met158 alleles.bNumber of S alleles.cNumber of G alleles.

Table 3 Ratings of valence and arousal of the visual stimuli

All subjects (M7s.d.) Association with genotype (Pearson’s R)

COMTa 5-HTTLPRb rs25531c

ValenceUnpleasant 2.1171.24 �0.19 0.04 0.09Neutral 6.0171.15 0.04 0.17 0.01Pleasant 7.3770.94 0.02 0.24 0.08

ArousalUnpleasant 4.6972.56 0.05 0.07 �0.22Neutral 2.7771.75 �0.38 �0.11 0.18Pleasant 4.7671.45 �0.19 �0.05 0.13

Valence scale: from 1 = unpleasant to 9 = pleasant.Arousal scale: from 1 = not arousing to 9 = very arousing.aNumber of Met158 alleles.bNumber of S alleles.cNumber of G alleles.


311


Figure 1 Effect of COMT and 5-HTT genotypes on limbic reactivity to aversive stimuli. Combined effect of COMT and 5-HTT (5-HTTLPR and rs25531) genotypes on limbic activation elicited by unpleasant stimuli in the amygdala, hippocampalformation and cingulate cortex. The combined genotype effect was significantly (PFDR < 0.05 corrected for VOI) associatedwith stimulus-induced activation in 90% of all voxels within the VOI and accounted for 40% of inter-individual variance infMRI BOLD response averaged across the entire VOI. Partial Eta2 coefficients (proportion of variance attributable to a factor)at the right panel were estimated for the voxel with the maximal combined gene effect of the respective region.


312


both other genotypes. Analysis of variance (ANOVA)statistics revealed that 22% of variance (P = 0.0005) inthe mean fMRI BOLD response across the ROI weredue to COMT. As posited, the BOLD response tounpleasant stimuli increased with the dose of met158

alleles. The traditional 5-HTTLPR S/L genotypingpredicted activity of 660 voxels (and 7% of variancein the mean fMRI BOLD response across the ROI;P = 0.034). Interestingly, rs25531 genotype had thehighest impact and predicted brain activity in 2102voxels (and 24% of variance in the mean fMRI BOLDresponse across the ROI; P = 0.0002). For both 5-HTTgenotypes, a higher limbic responsiveness was asso-ciated with the lowexpression variants, S and G,respectively. Together, both 5-HTT genotype variantswere associated with the BOLD response to unplea-sant stimuli in 2218 voxels (26% of variance; P =0.0003).

Regional effects of COMT Val158Met and 5-HTTgenotype on processing of unpleasant stimuliAnalyses of regional effects inside the VOI wereadditionally performed on a voxel-by-voxel basis(Table 4). These analyses revealed highly significantand substantial combined genotype effects on inter-

individual variance in activity of amygdala (left:R2 = 29%; right: R2 = 39%), hippocampal formation(left: R2 = 40%; right: R2 = 44%), posterior cingulate(left: R2 = 31%; right: R2 = 29%), middle cingulate(left: R2 = 32%; right: R2 = 33%) and anterior cingulate(left: R2 = 35%; right: R2 = 41%). The effects of 5-HTTand COMT genotypes on regional activations werealways additive, as no significant two-way inter-actions between genes were observed.

COMT alone had a significant effect in all theselimbic brain regions except the posterior cingulate(Table 4). The traditional 5-HTTLPR genotype alonewas significantly related to the BOLD response tounpleasant stimuli in the amygdala, hippocampalformation and the anterior cingulate. As in theanalysis of global effects, rs25531 genotype stronglyaffected processing of unpleasant stimuli and the lowexpression G allele was significantly associated withhigher brain activity in all investigated limbicregions.

When considering 5-HTTLPR as triallelic (S/LG/LA)with equivalent effects of S and LG, COMT and5-HTTLPR explained a comparable part of limbicactivation elicited by aversive stimuli in the amygda-la, hippocampus, anterior, medial and posterior

Figure 2 Effect of genotype on BOLD signal change to aversive stimuli in amygdala and anterior cingulate (ACC). Forclarity, only groups homozygous for low (Met) or high activity (Val) COMT alleles and high (LA) or low (S*, i.e. LG or S)transcribing 5-HTT genotypes are shown. Bars denote difference (mean7s.e.m.) of the BOLD signal during presentation ofunpleasant compared with neutral pictures. For genotype frequencies, see Table 1 and for MNI coordinates and statistics, seeTable 5.


313


cingulate cortex (Table 5). Like in the model withindependent effects of rs25531 and 5-HTTLPR, theeffects of 5-HTT and COMT genotypes on regionalactivations were always additive, as no significanttwo-way interactions between genes were observed.

Discussion

Our data indicate that processing of aversive emo-tional stimuli in the amygdala, hippocampus andlimbic cortex is additively affected by functionalpolymorphisms of COMT and the 5-HT transporter,and that the combined effect of these three genotypes

is greater than either individually. COMT and 5-HTTgenotypes substantially affected the processing ofunpleasant but not pleasant emotional stimuli andexplained 40% of the inter-individual variance in thefMRI BOLD response elicited by these stimuli inthe limbic system. Considering the limitations ofthis technology, 40% is a large proportion of inter-individual variation in aversive signal processing.These gene effects on emotional brain activation (asa candidate for an intermediate phenotype) aresubstantially larger than effects on personality traitsrelevant to anxiety or affective disorders, such as the4% effect of 5-HTTLPR on anxiety.20 Whereas the

Table 4 Effect of COMT Val158Met and both SCL6A4 polymorphisms on the activation of limbic brain areas by unpleasantstimuli

Region Side BA MNI coordinates T PFDR Partial Eta2 All R2

(%)

x y z COMT(%)

5-HTTLPR(%)

rs25531(%)

Amygdala L �27 0 �21 4.20 0.001 9 11 17 29R 27 � 9 �12 5.17 0.001 11 5 33 39

Hippocampalformation

L �21 �15 �18 5.27 0.001 12 19 24 40

R 30 � 9 �12 5.56 0.001 13 3 39 44Post. cingulate L 30 �3 �45 21 4.15 0.001 4 2 28 31

R 30 3 �45 18 4.17 0.001 5 4 26 29Mid. cingulate L 24 �12 6 48 4.45 0.001 10 4 26 32

R 24 12 12 33 4.13 0.001 26 2 10 33Ant. cingulate L 32 �9 33 27 4.78 0.001 8 18 22 35

R 32 12 33 30 5.32 0.001 20 8 27 41

Abbreviations: COMT, catechol-O-methyltransferase; MNI: Montreal Neurological Institute.Partial Eta2: proportion of variance attributable to a factor.COMT genotype was coded as the number of Met158 alleles (0, 1 or 2).5-HTTLPR genotype was coded as the number of S alleles (0, 1 or 2).rs25531 genotype was coded as the number of G alleles (0, 1 or 2).

Table 5 Effect of COMT val158met and 5-HTTPLR S/LG/LA triallelic genotype on the activation of limbic brain areas byunpleasant stimuli

Region Side BA MNI coordinates T PFDR Partial Eta2 All R2 (%)

x y z COMT (%) 5-HTTLPR (%)

Amygdala L �27 0 �18 3.83 0.004 10 17 25R 21 �9 �12 4.31 0.004 18 13 29

Hippocampal formation L �27 �15 �21 4.71 0.004 15 23 34R 30 �21 �12 3.18 0.008 6 13 19

Post. cingulate L 30 �3 �48 15 2.99 0.008 5 13 18R 29 3 �45 15 2.96 0.008 6 12 17

Med. cingulate L 23 �3 �27 33 4.04 0.004 12 17 27R 24 15 12 30 5.06 0.004 29 12 37

Ant. cingulate L 32 �9 33 27 4.18 0.004 7 26 31R 32 6 36 33 4.50 0.004 16 18 31

Abbreviations: COMT, catechol-O-methyltransferase; MNI: Montreal Neurological Institute.Partial Eta2: proportion of variance attributable to a factor.COMT genotype was coded as the number of met158 alleles (0, 1 or 2).5-HTTPLR S/LG/LA triallelic genotype was coded as the number of S and LG alleles (0, 1 or 2).


314


limited effect of genetic variants in previous studiesassessing behavioral traits only allowed for an assum-ption of additive genotype effects, the strong effect ofindividual genes on stimulus-induced fMRI activa-tion permits modeling and a clear demonstration ofsuch additive effects. The resulting additional valid-ity supports the usefulness of functional imaging forthe assessment of gene–gene effects on intermediatephenotypes in humans.54

The Val158Met polymorphism of COMT substan-tially affects brain catabolism of DA and NE: theCOMT variant containing Met158 has 25–33% of theactivity of the Val158 enzyme.7,8 In heterozygousindividuals, enzyme activity is midway betweenthose homozygous for COMT Val158Met.5 Therefore,the observed positive correlation between the numberof Met158 alleles and the BOLD response suggests thatthe amplified activation of limbic areas elicited byunpleasant stimuli may be related to increased levelsof DA and/or NE in these brain regions or connectedbrain areas.

Processing of unpleasant stimuli also increasedwith the number of 5-HTTLPR S alleles, that isreduced 5-HTT activity. Recently, it has been pro-posed that an A to G substitution (rs25531) inside orjust upstream of the 5-HTTLPR22–24 is functional andalters expression and AP2 transcription factor bind-ing. The G allele of SCL6A4 rs25531 was supposedto reduce 5-HTT expression in a similar way as theS allele of 5-HTTLPR.25 Together, both SCL6A4genotypes conferred four-fold differences in 5-HTTtranscription.25 To the best of our knowledge, our datashow for the first time that (1) the dosage of the lowexpressing G alleles is associated with amplifiedlimbic processing of unpleasant stimuli and (2)that modeling both 5-HTT genotypes yields betterresults compared to only modeling the traditionally5-HTTLPR genotype. At the neurochemical level, thecorrelation between the number of low expressing Sor G alleles and the BOLD response suggests that theamplified activation of limbic areas elicited byunpleasant stimuli may be due to actual alterationsin 5-HT neurotransmission and/or differences inbrain development of limbic and related circuitsassociated with SCL6A4 genotype.39

There are a series of internalizing disorders, inwhich COMT and SCL6A4 have been variablyimplicated and in which DA, NE and 5-HT play avital role.1 Several of these disorders, major depres-sion, generalized anxiety and phobia, are commonand have enormous impact on morbidity and mortal-ity.55 Depression is worsened or triggered by drugsthat deplete DA and 5-HT, and a relationship of 5-HTdepletion to irritability and anxiety is also known.56

Serotonergic neurotransmission in the amygdala hasbeen implicated in fear conditioning,26 and increasedamygdala activation during the presentation of aver-sive stimuli was associated with an increased risk fordisorders associated with negative mood states.27,39,57

Moreover, DA potentiates the response of the amyg-dala by attenuating the effect of inhibitory input from

the PFC and by augmenting the effect of excitatoryinput from sensory cortices.58 Consistent with this,fMRI studies in humans showed that DA modulatesthe responsivity of the amygdala to emotionalstimuli,59,60 and that COMT Val158Met affects DAsynthesis in midbrain.61,62 The potentiated amygdalaresponse in carriers of the Met158 alleles may thusreflect DA gating of amygdala inputs. Additionally,afferent NE neurons modulate affective responses andmemory in limbic structures such as the amygdalaand the hippocampus.63 The hippocampus has beenimplicated in the inhibition of stress responses viainhibitory connections with many of the subcorticalstructures, and is activated during acute stress64 andmay be crucial for the regulation of arousal andaffective states.65 A strong impact of COMT and 5-HTT genotype on processing of aversive visualstimuli was also found in the anterior cingulate, aregion that has previously been linked with theregulation of affective states.66

The effect of 5-HTTLPR on extra-limbic corticalactivation by affective stimuli was recently reported.67

Therefore, we explored whether the assessed geno-types also affect processing of affective stimulioutside the limbic system and observed that proces-sing of aversive stimuli was affected in the PFC,insula, and temporal, parietal and occipital cortices(data not shown). Further studies should investigateCOMT and 5-HTT genotype effects on cognitivefunction and extra-limbic brain areas.

Remarkably enough, the mainstay of pharmacother-apy of depression remains drugs that augmentmonoaminergic neurotransmitters.56 The mechanismsof both the development and treatment response ofnegative mood states involve more than the simplebalance or sum total of monoamine neurotransmitterexpression – for example, neurogenesis appears to becritical, especially in treatment response.68 Furtherstudies should assess combined genotype effects onactual expression of 5-HTT availability and COMTactivity during different stages of development.

Owing to the limited number of subjects with the G-allele, the regression analysis modeling linear effectsof rs25531 is vulnerable to effects of outliers. How-ever, using a triallelic approach instead of separatelymodeling the effects of both 5-HTT polymorphismsyielded comparable results, which again explainedmore variance than only using the traditional5-HTTLPR.

In summary, our data show that the processing ofaversive visual stimuli in the amygdala and otherlimbic brain areas is additively increased in carriersof the low-activity COMT Met158 allele and low 5-HTTtranscribing 5-HTTLPR S and rs25531 G alleles.Confrontation with affective cues may thus induceexcessive activation of brain networks associated withemotional and cognitive behavior control and con-tribute to the lowered resilience against anxiety anddysphoric mood states in subjects with geneticallydetermined low COMT activity and/or low 5-HTTtranscription.


315


Acknowledgments

This study was supported by the Deutsche For-schungsgemeinschaft (He 2597/7-3, Sm 80/1-1 &SFB 636/C1) and the German Ministry of Educationand Research (NGFN-2 No. 01 GS 0475, Baden-Wurttemberg Consortium for Addiction ResearchNo. 01 EB 0110/0410).

References

1 Goldman D, Oroszi G, Ducci F. The genetics of addictions:uncovering the genes. Nat Rev Genet 2005; 6: 521–532.

2 Gottesman II, Gould TD. The endophenotype concept in psychia-try: etymology and strategic intentions. Am J Psychiatry 2003; 160:636–645.

3 Schumann G, Spanagel R, Mann K. Candidate genes for alcoholdependence: animal studies. Alcohol Clin Exp Res 2003; 27:880–888.

4 Lachman HM, Papolos DF, Saito T, Yu YM, Szumlanski CL,Weinshilboum RM. Human catechol-O-methyltransferase pharma-cogenetics: description of a functional polymorphism and itspotential application to neuropsychiatric disorders. Pharmaco-genetics 1996; 6: 243–250.

5 Weinshilboum RM, Otterness DM, Szumlanski CL. Methylationpharmacogenetics: catechol O-methyltransferase, thiopurinemethyltransferase, and histamine N-methyltransferase. Annu RevPharmacol Toxicol 1999; 39: 19–52.

6 Palmatier MA, Kang AM, Kidd KK. Global variation in thefrequencies of functionally different catechol-O-methyltransferasealleles. Biol Psychiatry 1999; 46: 557–567.

7 Lotta T, Vidgren J, Tilgmann C, Ulmanen I, Melen K, Julkunen Iet al. Kinetics of human soluble and membrane-bound catecholO-methyltransferase: a revised mechanism and description ofthe thermolabile variant of the enzyme. Biochemistry 1995; 34:4202–4210.

8 Spielman RS, Weinshilboum RM. Genetics of red cell COMTactivity: analysis of thermal stability and family data. Am J MedGenet 1981; 10: 279–290.

9 Egan MF, Goldberg TE, Kolachana BS, Callicott JH, Mazzanti CM,Straub RE et al. Effect of COMT Val108/158 Met genotype onfrontal lobe function and risk for schizophrenia. Proc Natl AcadSci USA 2001; 98: 6917–6922.

10 Joober R, Gauthier J, Lal S, Bloom D, Lalonde P, Rouleau G et al.Catechol-O-methyltransferase Val-108/158-Met gene variants asso-ciated with performance on the Wisconsin Card Sorting Test. ArchGen Psychiatry 2002; 59: 662–663.

11 Malhotra AK, Kestler LJ, Mazzanti C, Bates JA, Goldberg T,Goldman D. A functional polymorphism in the COMT gene andperformance on a test of prefrontal cognition. Am J Psychiatry2002; 159: 652–654.

12 Goldberg TE, Egan MF, Gscheidle T, Coppola R, Weickert T,Kolachana BS et al. Executive subprocesses in working memory:relationship to catechol-O-methyltransferase Val158Met genotypeand schizophrenia. Arch Gen Psychiatry 2003; 60: 889–896.

13 Mynett-Johnson LA, Murphy VE, Claffey E, Shields DC, McKeon P.Preliminary evidence of an association between bipolar disorderin females and the catechol-O-methyltransferase gene. PsychiatrGenet 1998; 8: 221–225.

14 Ohara K, Nagai M, Suzuki Y, Ohara K. Low activity allele ofcatechol-o-methyltransferase gene and Japanese unipolar depres-sion. Neuroreport 1998; 9: 1305–1308.

15 Papolos DF, Veit S, Faedda GL, Saito T, Lachman HM. Ultra-ultrarapid cycling bipolar disorder is associated with the low activitycatecholamine-O-methyltransferase allele. Mol Psychiatry 1998; 3:346–349.

16 Enoch MA, Schuckit MA, Johnson And BA, Goldman D. Geneticsof alcoholism using intermediate phenotypes. Alcohol Clin ExpRes 2003; 27: 169–176.

17 Zubieta JK, Heitzeg MM, Smith YR, Bueller JA, Xu K, Xu Y et al.COMT val158met genotype affects mu-opioid neurotransmitterresponses to a pain stressor. Science 2003; 299: 1240–1243.

18 Diatchenko L, Slade GD, Nackley AG, Bhalang K, Sigurdsson A,Belfer I et al. Genetic basis for individual variations in painperception and the development of a chronic pain condition. HumMol Genet 2005; 14: 135–143.

19 Enoch MA, Xu K, Ferro E, Harris CR, Goldman D. Genetic originsof anxiety in women: a role for a functional catechol-O-methyltransferase polymorphism. Psychiatr Genet 2003; 13:33–41.

20 Lesch KP, Bengel D, Heils A, Sabol SZ, Greenberg BD, Petri S et al.Association of anxiety-related traits with a polymorphism in theserotonin transporter gene regulatory region. Science 1996; 274:1527–1531.

21 Heinz A, Jones DW, Mazzanti C, Goldman D, Ragan P, Hommer Det al. A relationship between serotonin transporter genotype andin vivo protein expression and alcohol neurotoxicity. BiolPsychiatry 2000; 47: 643–649.

22 Nakamura M, Ueno S, Sano A, Tanabe H. The human serotonintransporter gene linked polymorphism (5-HTTLPR) shows tennovel allelic variants. Mol Psychiatry 2000; 5: 32–38.

23 Kim SJ, Cox N, Courchesne R, Lord C, Corsello C, Akshoomoff N etal. Transmission disequilibrium mapping at the serotonin trans-porter gene (SLC6A4) region in autistic disorder. Mol Psychiatry2002; 7: 278–288.

24 Kraft JB, Slager SL, McGrath PJ, Hamilton SP. Sequence analysis ofthe serotonin transporter and associations with antidepressantresponse. Biol Psychiatry 2005; 58: 374–381.

25 Hu XZ, Lipsky RH, Zhu G, Akhtar LA, Taubman J, Greenberg BD etal. Serotonin transporter promoter gain-of-function genotypes arelinked to obsessive-compulsive disorder. Am J Hum Genet 2006;78: 815–826.

26 Garpenstrand H, Annas P, Ekblom J, Oreland L, Fredrikson M.Human fear conditioning is related to dopaminergic and seroto-nergic biological markers. Behav Neurosci 2001; 115: 358–364.

27 Caspi A, Sugden K, Moffitt TE, Taylor A, Craig IW, Harrington Het al. Influence of life stress on depression: moderationby a polymorphism in the 5-HTT gene. Science 2003; 301:386–389.

28 Barr CS, Newman TK, Lindell S, Shannon C, Champoux M, LeschKP et al. Interaction between serotonin transporter gene variationand rearing condition in alcohol preference and consumption infemale primates. Arch Gen Psychiatry 2004; 61: 1146–1152.

29 Beitchman JH, Baldassarra L, Mik H, De LV, King N, Bender Det al. Serotonin transporter polymorphisms and persistent,pervasive childhood aggression. Am J Psychiatry 2006; 163:1103–1105.

30 Hu X, Oroszi G, Chun J, Smith TL, Goldman D, Schuckit MA. Anexpanded evaluation of the relationship of four alleles to the levelof response to alcohol and the alcoholism risk. Alcohol Clin ExpRes 2005; 29: 8–16.

31 Stein MB, Seedat S, Gelernter J. Serotonin transporter genepromoter polymorphism predicts SSRI response in generalizedsocial anxiety disorder. Psychopharmacology (Berlin) 2006; 187:68–72.

32 Gelernter J, Kranzler H, Cubells JF. Serotonin transporter protein(SLC6A4) allele and haplotype frequencies and linkage disequili-bria in African- and European-American and Japanese populationsand in alcohol-dependent subjects. Hum Genet 1997; 101:243–246.

33 Strobel A, Lesch KP, Jatzke S, Paetzold F, Brocke B. Furtherevidence for a modulation of novelty seeking by DRD4 exon III,5-HTTLPR, and COMT val/met variants. Mol Psychiatry 2003;8: 371–372.

34 Benjamin J, Osher Y, Lichtenberg P, Bachner-Melman R, GritsenkoI, Kotler M et al. An interaction between the catechol O-methyltransferase and serotonin transporter promoter regionpolymorphisms contributes to tridimensional personality ques-tionnaire persistence scores in normal subjects. Neuropsychobio-logy 2000; 41: 48–53.

35 Benjamin J, Osher Y, Kotler M, Gritsenko I, Nemanov L, BelmakerRH et al. Association between tridimensional personality ques-tionnaire (TPQ) traits and three functional polymorphisms:dopamine receptor D4 (DRD4), serotonin transporter promoterregion (5-HTTLPR) and catechol O-methyltransferase (COMT).Mol Psychiatry 2000; 5: 96–100.


316


36 Hariri AR, Mattay VS, Tessitore A, Kolachana B, Fera F, GoldmanD et al. Serotonin transporter genetic variation and the response ofthe human amygdala. Science 2002; 297: 400–403.

37 Heinz A, Braus DF, Smolka MN, Wrase J, Puls I, Hermann D et al.Amygdala-prefrontal coupling depends on a genetic variation ofthe serotonin transporter. Nat Neurosci 2005; 8: 20–21.

38 Smolka MN, Schumann G, Wrase J, Grusser SM, Flor H, Mann Ket al. Catechol O-methyltransferase val158-met genotype affectsprocessing of emotional stimuli in the amygdala and prefrontalcortex. J Neurosci 2005; 25: 836–842.

39 Pezawas L, Meyer-Lindenberg A, Drabant EM, Verchinski BA,Munoz KE, Kolachana BS et al. 5-HTTLPR polymorphism impactshuman cingulate–amygdala interactions: a genetic susceptibilitymechanism for depression. Nat Neurosci 2005; 8: 828–834.

40 First MB, Spitzer RL, Gibbon M, Williams JBW. Structured ClinicalInterview for DSM-IV-TR Axis I Disorders, Research Version,Patient Edition (SCID-I/P). Biometrics Research, New York StatePsychiatric Institute: New York, 2001.

41 First MB, Spitzer RL, Gibbon M, Williams JBW. Structured ClinicalInterview for DSM-IV Personality Disorders, (SCID-II). AmericanPsychiatric Press, Inc.: Washington, DC, 1997.

42 Derogatis LR. Symptom Checklist 90-Revised/SCL-90-R: Adminis-tration, Scoring and Procedures Manual II. Clinical PsychometricResearch: Baltimore, MD, 1983.

43 Spielberger CD, Gorsuch RL, Lushene RE. Manual for the State-Trait-Anxiety Inventory. Consulting Psychologists Press: Palo Alto,CA, USA, 1970.

44 Radloff LS. The CES-D scale: a self report depression scale forresearch in the general population. Appl Psychol Meas 1997; 1:385–401.

45 Hamilton M. The Hamilton Rating Scale for Depression inAssessment of Depression. Springer: Berlin, 1986, pp 143–152.

46 Lang PJ, Bradley MM, Cuthbert BN. The International AffectivePicture System (IAPS). Center for Research in Psychophysiology,University of Florida: Gainsville, FL, 1999.

47 Lang PJ. The emotion probe. Studies of motivation and attention.Am Psychol 1995; 50: 372–385.

48 Taylor SF, Phan KL, Decker LR, Liberzon I. Subjective rating ofemotionally salient stimuli modulates neural activity. Neuroimage2003; 18: 650–659.

49 Bradley MM, Lang PJ. Measuring emotion: the self-assessmentmanikin and the semantic differential. J Behav Ther ExpPsychiatry 1994; 25: 49–59.

50 Wendland JR, Martin BJ, Kruse MR, Lesch KP, Murphy DL.Simultaneous genotyping of four functional loci of humanSLC6A4, with a reappraisal of 5-HTTLPR and rs25531. MolPsychiatry 2006; 11: 224–226.

51 Maldjian JA, Laurienti PJ, Kraft RA, Burdette JH. An automatedmethod for neuroanatomic and cytoarchitectonic atlas-basedinterrogation of fMRI data sets. Neuroimage 2003; 19: 1233–1239.

52 Tzourio-Mazoyer N, Landeau B, Papathanassiou D, Crivello F,Etard O, Delcroix N et al. Automated anatomical labeling of

activations in SPM using a macroscopic anatomical parcellationof the MNI MRI single-subject brain. Neuroimage 2002; 15:273–289.

53 Genovese CR, Lazar NA, Nichols T. Thresholding of statisticalmaps in functional neuroimaging using the false discovery rate.Neuroimage 2002; 15: 870–878.

54 Hariri AR, Weinberger DR. Imaging genomics. Br Med Bull 2003;65: 259–270.

55 Murray CJ, Lopez AD. Global mortality, disability, and thecontribution of risk factors: Global Burden of Disease Study.Lancet 1997; 349: 1436–1442.

56 Delgado PL. Depression: the case for a monoamine deficiency.J Clin Psychiatry 2000; 61(Suppl 6): 7–11.

57 Hariri AR, Drabant EM, Munoz KE, Kolachana BS, Mattay VS,Egan MF et al. A susceptibility gene for affective disorders and theresponse of the human amygdala. Arch Gen Psychiatry 2005; 62:146–152.

58 Rosenkranz JA, Grace AA. Modulation of basolateral amygdalaneuronal firing and afferent drive by dopamine receptor activationin vivo. J Neurosci 1999; 19: 11027–11039.

59 Hariri AR, Mattay VS, Tessitore A, Fera F, Smith WG, WeinbergerDR. Dextroamphetamine modulates the response of the humanamygdala. Neuropsychopharmacology 2002; 27: 1036–1040.

60 Tessitore A, Hariri AR, Fera F, Smith WG, Chase TN, Hyde TMet al. Dopamine modulates the response of the human amygdala:a study in Parkinson’s disease. J Neurosci 2002; 22: 9099–9103.

61 Akil M, Kolachana BS, Rothmond DA, Hyde TM, Weinberger DR,Kleinman JE. Catechol-O-methyltransferase genotype anddopamine regulation in the human brain. J Neurosci 2003; 23:2008–2013.

62 Meyer-Lindenberg A, Kohn PD, Kolachana B, Kippenhan S,McInerney-Leo A, Nussbaum R et al. Midbrain dopamine andprefrontal function in humans: interaction and modulation byCOMT genotype. Nat Neurosci 2005; 8: 594–596.

63 Bremner JD, Krystal JH, Southwick SM, Charney DS. Noradrener-gic mechanisms in stress and anxiety: I. Preclinical studies.Synapse 1996; 23: 28–38.

64 Lopez JF, Akil H, Watson SJ. Neural circuits mediating stress. BiolPsychiatry 1999; 46: 1461–1471.

65 McNaughton N, Gray JA. Anxiolytic action on the behaviouralinhibition system implies multiple types of arousal contribute toanxiety. J Affect Disord 2000; 61: 161–176.

66 Phillips ML, Drevets WC, Rauch SL, Lane R. Neurobiology ofemotion perception I: The neural basis of normal emotionperception. Biol Psychiatry 2003; 54: 504–514.

67 Canli T, Omura K, Haas BW, Fallgatter A, Constable RT, Lesch KP.Beyond affect: a role for genetic variation of the serotonintransporter in neural activation during a cognitive attention task.Proc Natl Acad Sci USA 2005; 102: 12224–12229.

68 Santarelli L, Saxe M, Gross C, Surget A, Battaglia F, Dulawa S et al.Requirement of hippocampal neurogenesis for the behavioraleffects of antidepressants. Science 2003; 301: 805–809.


317


Gene–gene effects on central processing of aversive stimuli

Documents

Transcript of Gene–gene effects on central processing of aversive stimuli