A genome-wide scan for attention-deficit/hyperactivity disorder in 155 German sib-pairs

ORIGINAL ARTICLE

A genome-wide scan for attention-deficit/hyperactivitydisorder in 155 German sib-pairsJ Hebebrand1, A Dempfle2, K Saar3, H Thiele4, B Herpertz-Dahlmann5, M Linder6, H Kiefl6,

H Remschmidt7, U Hemminger8, A Warnke8, U Knolker9, P Heiser10, S Friedel1, A Hinney1, H Schafer2,

P Nurnberg4 and K Konrad5

1Department of Child and Adolescent Psychiatry, University of Duisburg-Essen, Essen, Germany; 2Institute of MedicalBiometry and Epidemiology, University of Marburg, Marburg, Germany; 3Molecular Genetics and Gene Mapping Center, MaxDelbruck Center Berlin-Buch, Berlin, Germany; 4Cologne Center for Genomics, University of Cologne, Cologne, Germany;5Department of Child and Adolescent Psychiatry, RWTH Aachen, Aachen, Germany; 6Psychiatric Clinic for Children andAdolescents, Regensburg, Germany; 7Department of Child and Adolescent Psychiatry, University of Marburg, Marburg,Germany; 8Department of Child and Adolescent Psychiatry, University of Wurzburg, Wurzburg, Germany; 9Department of Childand Adolescent Psychiatry, University of Lubeck, Lubeck, Germany and 10Department of Psychiatry, University of Marburg,Marburg, Germany

Three groups have previously performed genome scans in attention-deficit/hyperactivitydisorder (ADHD); linkage to chromosome 5p13 was detected in all of the respective studies. Inthe current study, we performed a whole-genome scan with 102 German families with two ormore offspring who currently fulfilled the diagnostic criteria for ADHD. Including subsequentfine mapping on chromosome 5p, a total of 523 markers were genotyped. The highestnonparametric multipoint LOD score of 2.59 (empirical genome-wide significance 0.1) wasobtained for chromosome 5p at 17 cM (according to the Marshfield map). Subsequent analysesrevealed (a) a higher LOD score of 3.37 at 39 cM for a quantitative severity score based onsymptoms of inattention than for hyperactivity/impulsivity (LOD score of 1.11 at 59 cM), and (b)an HLOD of 4.75 (empirical genome-wide significance 0.001) based on a parametric modelassuming dominant inheritance. The locus of the solute carrier 6A3 (SLC6A3; dopaminetransporter 1; DAT1) localizes to 5p15.33; the gene has repeatedly been implicated in theetiology of ADHD. However, in our sample the DAT1 VNTR did not show association withADHD. We additionally identified nominal evidence for linkage to chromosomes 6q, 7p, 9q, 11q, 12q and 17p, which had also been identified in previous scans. Despite differences inethnicity, ascertainment and phenotyping schemes, linkage results in ADHD appearremarkably consistent.Molecular Psychiatry (2006) 11, 196–205. doi:10.1038/sj.mp.4001761; published online 11 October 2005

Keywords: ADHD; 5p13; dopamine transporter

Introduction

Attention-deficit/hyperactivity disorder (ADHD) isone of the most heritable disorders in child andadolescent psychiatry; heritability is estimated atapproximately 0.81 and the recurrence risk in siblingsis in the magnitude of 5.2 A number of candidategenes have been investigated using both case–controland family-based studies in order to identify geneticvariation underlying the disorder.3 The most fre-quently analyzed genes include the dopamine D4receptor gene (DRD4) and the gene for the dopaminetransporter (DAT1, SLC6A3); meta-analyses have

yielded significant (Po0.02) results for DRD44,5 andnegative results (P¼0.06 and 0.21) for the DAT1VNTR.5,6

The first whole-genome scan was performed in aUS sample encompassing 126 affected sib-pairs from104 families,7 which was subsequently extended to204 families with a total of 270 sib-pairs.8,9 Anindependent genome scan based on 164 sib-pairsfrom 106 families was performed in a Dutch sample.10

A third group performed a genome-wide scan of 16extended and multigenerational families with ADHDfrom a genetic isolate located in Columbia.11 Thesegenome scans and subsequent fine mapping studies ofpeak regions have yielded significant evidence oflinkage12 on 4q13.2, 5q33.3,11 6q12,9 11q22,11 16p13,13

and 17p11.9,11 Chromosomal regions that yielded atleast nominal evidence of linkage in two of the threegenome scans include 11q22, 17p11, 20q13,8,9,11 3q13and 9q33.10,11 Nominal evidence of linkage was

Received 17 May 2005; revised 9 August 2005; accepted 1September 2005; published online 11 October 2005

Correspondence: Professor J Hebebrand, Department of Child andAdolescent Psychiatry and Psychotherapy, University of Duisburg-Essen, Virchowstr. 174, D-45147 Essen, Germany.E-mail: [email protected]

Molecular Psychiatry (2006) 11, 196–205& 2006 Nature Publishing Group All rights reserved 1359-4184/06 $30.00

www.nature.com/mp

detected to 5p13 in all three scans.8–11 Within thisregion, the highest multipoint MLS scores werelocalized at markers D5S418 (MLS¼2.559) andD5S2500 (MLS¼ 1.43 for the broad phenotype defini-tion of ADHD10); the distance between the twomarkers is 18.7 Mb.

In the current study, we report the results of afourth independent genome-wide scan performed in aGerman sample of 102 families encompassing a totalof 229 affected children.

Materials and methods

The 102 families were ascertained and phenotypicallycharacterized by physicians either in the outpatientunits of the Departments of Child and AdolescentPsychiatry of the Universities of Aachen, Lubeck,Marburg and Wurzburg and of the district Oberpfalzin Regensburg for ADHD. The physicians were in theprocess of specialization in child and adolescentpsychiatry; they were specifically trained for thepurpose of the study. Phenotypical assessments wereperformed either within the respective outpatientunits or at the families’ homes. Informed parentalconsent was obtained for all participants and thestudy was approved by the ethics committees of allparticipating hospitals.

Families were included if they had at least twoaffected children with ADHD, one of whom werequired to be older than 8 years and the other(s) tobe older than 6 years. The children all currentlyfulfilled diagnostic criteria for ADHD according toDSM-IV;14 index patients were required to fulfilllifetime criteria for the combined type (see Table 1for clinical features of our sample). In 80, 19 and threefamilies two, three, or four affected children wereascertained, respectively. Of the 155 sib-pairs, 127were independent; four of the sib-pairs were based onhalf-sibs. In 94 families, both parents of the indexpatient were ascertained, and in the remaining eightonly the mother was included. The vast majority ofparents were of German origin; only 12 parents camefrom other European countries. All were of Caucasianethnicity. Socio-economic status (SES) was deter-mined for the whole family based on reportedpaternal profession and categorized into three groups.A total of 26, 61 and two families had a high, mediumand low SES, respectively. For 13 families, the SEScould not be determined.

Exclusion criteria (which applied both to indexpatients and siblings) were (a) general IQp75 asassessed with one of the following IQ tests: theWechsler Intelligence Scale for Children (WISC-III;15

German version16), the Kaufman Assessment Batteryfor Children (K-ABC;17 German version18) or theCulture Fair Test (CFT)19,20, (b) potentially confound-ing psychiatric diagnoses such as schizophrenia, anypervasive developmental disorder, Tourette’s disor-der, and primary mood or anxiety disorder (patientswith comorbid mood or anxiety disorders were notexcluded; in these cases the disorder clearly devel-

oped after onset of ADHD), (c) neurological disorderssuch as epilepsy, (d) a history of any acquired braindamage or evidence of the fetal alcohol syndrome, (e)very preterm birth and/or (f) maternal reports ofsevere prenatal, perinatal or postnatal complications.

Psychiatric classification was based on the Sche-dule for Affective Disorders and Schizophrenia forSchool-Age Children Present and Lifetime version(K-SADS-PL;20 German translation21). The K-SADS-PLis a highly reliable semi-structured interview used toassess a wide range of psychiatric disorders accordingto DSM III-R,22 DSM-IV14 or ICD-10.23 Mothersreceived (1) the unstructured introductory interview,

Table 1 Clinical features of ascertained ADHD children

Index patients SiblingsN¼ 102 N¼ 127

SexMale 86 (84.3%) 79 (62.2%)Female 16 (15.7%) 48 (37.8%)

Mean age 11.1 12.0Mean IQ 102.4 104.0

ADHD subtypea

Inattentive 0 61 (48.0%)Hyperactive/impulsive 0 10 (7.9%)Combined 102 (100%) 56 (44.1%)

Comorbiditiesb,c

Oppositional defiantdisorder

30 (29.4%) 25 (19.7%)

Conduct disorder 6 (5.9%) 2 (1.6%)Mood disorder 1 (1%) 3 (2.4%)Anxiety disorder 10 (9.8%) 8 (6.3%)

Number of ADHD symptomsd (parental rating)Mean number ofinattentive symptoms

8.0 7.3

Mean number ofhyperactive/impulsivesymptoms

7.9 5.3

ADHD total severity 79 76score (teacher rating)e,f

CBCL (parent rating)g

Externalizingh 68.1 64.7Internalizingh 66.9 66.4Inattentionh 71.4 68.2

aAccording to current best estimate diagnoses.bMultiple scoring possible.cCurrent diagnoses according to K-SADS.20,21

dRefers to child in unmedicated state, according to K-SADS.eRefers to the last 6 months, according to the FBB–HKS¼German teacher rating scale for ADHD,27 percentrange.fNote that the majority of patients were on stimulantmedication.gCBCL¼German translation of Child Behaviour Check-list.25,26

hT score.

Genome-wide scan in 155 German ADHD sib-pairsJ Hebebrand et al

197

Molecular Psychiatry

(2) the diagnostic screening interview, (3) the supple-ment completion checklist and upon fulfillment ofscreening criteria the appropriate diagnostic supple-ments. Mothers were asked to report the symptomsapplicable to their children in the unmedicated state.The child was interviewed with the screening inter-view of the K-SADS and in case of positive screeningfor mood or anxiety disorders with the respectivesupplements of the K-SADS-PL. The rationale for thisapproach was that internalizing disorders are morereliably assessed by self-reports than by parentalratings, while it turns out to be the other way roundfor externalizing childhood behavioral disorders.24

Additionally, we employed the Child BehaviourChecklist (CBCL,25 German translation26) and a Ger-man teacher rating scale for ADHD (FBB-HKS);27 thequestions of the latter instrument directly address theDSM-IV items.

The K-SADS-PL (as based on the maternal inter-view) was used to assess ADHD both categoricallyand dimensionally. The sum of nine possible inatten-tive, nine hyperactive/impulsive or all 18 possiblesymptoms yielded a severity score of inattention,hyperactivity/impulsivity and a total score, respec-tively.

‘Best estimate’ diagnoses, which totally overlappedwith the diagnosis of ADHD in the K-SADS-PL, weredetermined for all children after individual review ofdiagnoses, symptoms and impairment level by seniorclinicians (JH, KK, HK, UH) in regular case confer-ences of participating clinicians scheduled approxi-mately every 6 months throughout the 2-yearascertainment scheme. We consider our index pa-tients as representative of ADHD patients of Germanchild and adolescent psychiatric units. Thus, 161 ofthe total of 229 affected offspring were currently orhad previously been treated with stimulant medica-tion. In 202 children, ADHD had clinically beendiagnosed previously by a child and adolescentpsychiatrist (155), a pediatrician (31), a psychologist(13) or other physician (3).

GenotypingWe performed a genome-wide scan based on a total of425 individuals with initially 475 autosomal, 24 X-chromosomal and two pseudoautosomal STR markersspanning the entire genome with an average (max-imum) distance of 7(13)cm (adapted from Saar etal.28). Briefly, markers were amplified in multiplexreactions in 384-well microtiter plates on ABIGeneAmp PCR 9700 machines (Applied Biosystems,Darmstadt, Germany). An aliquot of the PCR reactionwas submitted to analysis on ABI 3730 sequencers.Semi-automated genotyping was performed with thehelp of the GeneMapper software version 3.0 (ABI).All genotypes were scored independently by anexperienced lab technician and subsequently by onescientist. In all, 16 additional markers (see Table 2) inthe major region of interest on chromosome 5 (mostlyaround the DAT1 (SLC6A3) locus) were subsequentlygenotyped, including the DAT1 VNTR.5,6

Statistical analysesFamilial relationships were verified using the GRRprogram29 on the genome scan markers, which led tothe identification of two pairs of identical twins,which were excluded from further analysis (notincluded in the number of sib-pairs stated above)and two pairs of half-sibs which had been reported tobe full-sibs. We checked all markers for Mendelianinconsistencies using PedCheck30 and in case ofinconsistencies set the genotypes of all familymembers for the respective marker and family onmissing. Hardy–Weinberg equilibrium was checkedfor all genetic markers on parental genotypes by w2

tests as implemented in Mega2 version 3.0,31 andMerlin version 1.0 alpha.32,33 We checked for likelygenotyping errors (e.g., manifesting as close doublerecombinants) using Merlin32,33 and set unlikelygenotypes on missing for the respective individuals.

Allele frequencies for linkage analysis (relevantonly for the few missing parents) were estimated fromall parental alleles. Data handling and preparation offiles was done with Mega2 version 3.0.31 Our primary

Table 2 Linkage results based on fine mapping of chromo-some 5p: genotyped markers and nonparametric multipointLOD scores in the chromosome 5 region between 0 and 65 cM(initial markers are in bold, their LOD scores in the initialanalysis are in parenthesis)

Marker Position(Kosambi cM)a

Position(Mb)b

LOD score

D5S1981 1 1.21 1.32D5S2005 1.72 1.39 1.53 (2.66)D5S678 1.8 1.41 1.65DAT1 VNTR 2 1.45 1.71D5S1970 5.43 2.49 2.20D5S417 6.67 3.17 1.76D5S2849 7.77 3.47 1.85 (1.96)D5S406 11.85 5.04 1.76 (1.77)D5S1953 16.72 7.71 2.58D5S807 19.02 9.26 2.24 (2.23)D5S667 22.34 11.13 1.43D5S817 22.88 11.63 1.61D5S1991 26.73 14.92 1.67 (1.96)D5S416 28.76 16.77 1.81D5S486 31.78 17.22 1.80D5S2845 36.25 22.42 1.73 (2.18)D5S813 37.32 23.63 1.98D5S502 39.46 25.72 1.93D5S2848 39.99 26.81 1.69D5S661 41.06 29.98 1.63D5S1470 45.34 32.53 1.86 (1.84)D5S674 47.09 33.63 2.13D5S663 51.99 34.84 1.66 (1.86)D5S1457 59.3 41.07 1.11 (1.35)D5S461 64.67 55.88 0.66

aAccording to the Marshfield map.bAccording to Golden Path (hg17); markers separated byp1 cM and markers not included in the Marshfield mapwere placed according to the deCode map46 or Golden Path(hg17).


198


linkage analysis was on the qualitative affectionstatus ADHD using the nonparametric Sall statistic.34

For easier comparison of our results with previouslypublished scans, we give both the NPL score and thenonparametric LOD score, using the Kong–Coxtransformation35 as implemented in Merlin.32,33 Thistransformation is not monotone; NPL scores tend tobe highest at genotyped markers, while the Kong–CoxLOD scores usually drop at markers and are higherbetween markers,35 as parametric LOD scores do aswell. We also performed linkage analysis on thequantitative DSM-IV scores (total score, inattentiveand hyperactive/impulsive subscores defined as thenumber of DSM-IV ADHD symptoms present, fromthe K-SADS-PL) using Merlin-regress36 assuming aheritability of 0.6 and a population mean (variance) of3(2.5), 1.6 (1.5) and 1.4 (1.5), respectively, asdetermined previously in Germany.37 This analysisdoes not require that the phenotypes are normallydistributed. Finally, we also performed parametricHLOD linkage analysis with Merlin32,33 under adominant model assuming a penetrance of 0.6(phenocopy rate 0.05) and an allele frequency of 0.01.The empirical genome-wide significance of all linkageresults was assessed by simulating 1000 replicates ofour data under the null hypothesis of no linkage withthe same characteristics regarding family structures,phenotypes, marker spacing and informativity. Thesesimulated scans were analyzed analogously to the realdata and the highest nonparametric LOD score andparametric HLOD score in each scan were recorded.The percentage of times that the simulated maximumscore exceeds the score obtained in the real data givesthe empirical genome-wide P-value.

Post-hoc analyses as to whether the evidence fornonparametric linkage on chromosome 5p could beexplained by the investigated DAT1 VNTR was doneusing the genotype-IBD sharing test GIST,38 whichinvestigates whether the allelic distribution of apolymorphism might be responsible for an observedlinkage by testing whether probands with a certaingenotype contributed more than expected to thelinkage signal. Additionally, we used the pedigreedisequilibrium test (PDT),39 a test for allelic associa-tion in pedigrees.

Results

The results of the nonparametric linkage analysis areillustrated in Figure 1 and Tables 2 and 3; the highestmultipoint LOD score of 2.74 (which corresponds toan NPL score of 3.25) in the initial analysis based onthe original 501 STR markers was obtained onchromosome 5p at 3 cM next to the marker D5S2005(1.72 cM on the Marshfield map, with a LOD score of2.66) in a broad peak extending beyond D5S1457(59.3 cM, LOD score 1.35). Other multipoint LODscores exceeding 1 were obtained for chromosomes 8,12, 17 and in the pseudoautosomal region of the Xand Y chromosomes. Upon genotyping of an addi-tional 16 markers on chromosome 5p, the maximummultipoint LOD score in this region was 2.59 (whichcorresponds to an NPL score of 3.52; empiricalgenome-wide P-value 0.1) at 17 cM, close to themarker D5S1953 with a LOD score of 2.58.

The highest multipoint LOD scores for the quanti-tative DSM-IV scores (obtained upon inclusion of all521 STR markers) were observed on chromosome 5p

0

1

2

3

Chromosome

Mu

ltip

oin

t L

OD

sco

re

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 2122 X XY

Figure 1 Nonparametric multipoint LOD scores for ADHD.


199


for the total score and inattentive subscore (maximumLOD scores of 2.31 at D5S674 (47.09 cM) and 3.37 atD5S502 (39.46 cM), respectively). For the inattentivesubscore, the second highest LOD score of 3.26 wasobtained on chromosome 12 (D12S1679 at 153.19 cM).The highest linkage peak for the hyperactive/impul-sive subscore was on chromosome 8 (maximum LODscore of 1.26 at D8S1778 (110.2 cM)) (Figures 2–4).

In the parametric HLOD analysis under a dominantmodel, we obtained the highest multipoint LOD score

of 4.75 again on chromosome 5p, with all familiescontributing to this LOD score (empirical genome-wide P-value 0.001). HLODs above 1 were also obser-ved on chromosomes 7, 8, 12 and in the pseudo-autosomal region on the X and Y chromosomes(Table 4).

The DAT1 VNTR did not show an association withADHD (multiallelic PDT P-value of 0.97). In particu-lar, the 480 bp allele, which had been associated withADHD in several previous studies (see meta-analyses

Table 3 Nonparametric multipoint LOD scores above 0.5 and adjacent markers

Adjacentmarkers

Chromosome Position ofmaximumLOD score(Kosambi

cM)a

Position ofmaximum LOD

score (Mb)b

MultipointLOD score

D1S2667 1 30 11 0.65D1S2703 1 236 209 0.71D5S1953 5 17 8 2.59c

D6S1053 6 75 65 0.58D7S2490 7 88 75 0.92D8S277 8 8 65 1.29D9S1851 9 104 17 0.68D10S1432 10 94 74 0.67D10S1651 10 169 133 0.76D12S392 12 166 130 2.10D14S306 14 46 37 0.55D17S1308 17 1 0.6 1.39DXYS233 X pseudoautosomal

region1 0.9 1.01

aAccording to the Marshfield map.bAccording to Golden Path (hg17).cBased on fine-mapping of chromosome 5p.

0

1

2

Chromosome

Mu

ltip

oin

t L

OD

sco

re

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 2122XY

Figure 2 Nonparametric multipoint LOD scores for total number of DSM-IV ADHD symptoms.


200


5, 6), did not show any relevant over-transmission inour sample (298 times transmitted vs 292 times nottransmitted). The linkage signal also did not stemfrom the VNTR (GIST P-values 40.7 under dominant,additive or recessive models).

Discussion

Our genome-wide scan in ADHD and all threeprevious scans7–11 have detected linkage signals in

the distal region of chromosome 5p; no other linkagesignal has consistently emerged in all scans. The LODscore obtained by Ogdie et al.9 was slightly lower thanour score; in both studies the Lander and Krugylak12

criterion for suggestive evidence for linkage wasfulfilled; the Bakker et al.10 LOD score of 1.43 wasbelow this criterion. Our HLOD score of 4.75 obtainedin the post-hoc parametric analysis even fulfilled thecriterion for evidence of linkage for a complexdisorder (genome-wide empirical P-value 0.001).

0

1

2

3

Chromosome

Mu

ltip

oin

t L

OD

sco

re

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22XY

Figure 3 Nonparametric multipoint LOD scores for number of DSM-IV inattentive ADHD symptoms.

0

1

2

Chromosome

Mu

ltip

oin

t L

OD

sco

re

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 2122

XY

Figure 4 Nonparametric multipoint LOD scores for number of DSM-IV hyperactive/impulsive ADHD symptoms.


201


Our initial nonparametric LOD score value of 2.74next to the marker D5S2005 (1.72 cM on the Marsh-field map) was obtained within a broad peak extend-ing from the distal end of the short arm to beyond60 cM; our subsequent fine-mapping efforts resultedin a nonparametric LOD score of 2.59 at 17 cM. Ogdieet al.9 localized their 5p peak to 59 cM, Bakker et al.10

to 69 cM. In light of the small sample sizes, whichlead to considerable stochastic variation of thelocation of linkage peaks,40 it seems plausible toassume that genetic variation at the same locusunderlies the chromosome 5p linkage peaks. Thecommon identification of this region is even morenoteworthy in light of the ethnic differences anddifferences in ascertainment and phenotyping.

Our post-hoc parametric HLOD analysis based on asingle model assuming autosomal dominant inheri-tance with a penetrance of 0.6, a phenocopy rate of0.05 and an allele frequency of 0.01 resulted in aconsiderably higher multipoint LOD score of 4.75 at17 cM. We chose this particular model because wethought that the specified parameters might comeclose to the real situation in ADHD. Indeed, the HLODwas considerably higher than the nonparametricmultipoint LOD score and was significant with anempirical genome-wide P-value of 0.001.

Our analyses based on the quantitative DSM-IVscores again picked up the chromosome 5p region;interestingly, the highest multipoint LOD scores onchromosome 5p differed considerably between theinattentive (3.37 at D5S502, 39.5 cM) and the hyper-

activity/impulsivity (1.11 at D5S1457, 59.3 cM) sub-scores, possibly suggesting that inattention is thephenotype more strongly influenced by the gene(s) inthis linkage region (Figure 5). Another likely explana-tion for this finding is chance, further possibilitiesinclude a potentially better temporal stability ofsymptoms of inattention, their more valid rating bythe mothers of our probands and/or a less pronouncedcontrast (rater) effect (reviewed by Heiser et al.,3) incomparison to symptoms of hyperactivity/impulsiv-ity. Also note that due to our assessment scheme(where the index case is required to have combinedsubtype), variance for inattentive symptoms wasslightly larger than for hyperactive–impulsive symp-toms which might have contributed to higher LODscores for inattentive symptoms. The LOD scoredifferences also need to be interpreted in light of thefact that we had requested the mothers to rate currentsymptoms as applicable to their unmedicated chil-dren.

The DAT1 located at the distal end of 5p is anobvious candidate gene within the linkage peaksobserved in the different genome scans.7 Stimulantmedications that are efficacious in ADHD blockbinding of dopamine to the dopamine transporter,41

and imaging studies have found reduced dopaminetransporter binding in adults with ADHD.42–44 We hadtherefore included the DAT1 VNTR and the STRmarkers D5S2005, D5S1981 and D5S678 in our finemapping efforts in an attempt to assess the contribu-tion of genetic variation at the DAT1 locus to the

Table 4 Results of parametric linkage analysis of ADHD under a dominant model

Adjacentmarkers

Chromosome Position ofmaximum

HLOD score(Kosambi cM)a

Position ofmaximumLOD score

(Mb)b

MaximumHLOD score

% familieslinked in

HLODanalysis

D4S403 4 28 13 0.62 41D4S1652 4 209 190 0.77 47D5S1953 5 17 8 4.75c 100D7S1802 7 32 20 0.56 37D7S2490 7 88 75 1.24 62D8S277 8 8 65 1.49 65D9S1851 9 104 17 0.56 34GATA121A08 10 94 70 1.09 53D10S1651 10 169 133 0.63 42D11S4176 11 94 94 0.57 38D12S392 12 165 130 1.61 68D14S306 14 46 3 0.87 58D15S822 15 13 25 0.52 40D15S1025 15 69 67 0.56 31D17S1308 17 1 0.6 0.76 50D19S254 19 101 62 0.64 39DXYS233 X pseudo-autosomal

region1 0.9 1.01 47

aAccording to the Marshfield map.bAccording to Golden Path (hg17).cBased on fine mapping of chromosome 5p.


202


linkage signal. Interestingly, in our study allelicvariation at the DAT1 VNTR was not responsible forthe linkage signal. Previous association and func-tional studies pertaining to DAT1 have mainlyfocused on the 30-UTR VNTR.3 A meta-analysis basedon 11 studies with a total of 824 informative meioses

yielded an almost significant (P¼ 0.06) pooled oddsratio of 1.27,5 whereas the meta-analysis of Purper-Ouakil et al.,6 based on 13 published family-basedassociation studies, showed that the 10-repeat allelehas no significant impact on the risk for ADHD. Arecent functional study did not confirm earlier

0

1

2

3

4

5

0 20 40 60 8 100 120 140 160 180 200cM

Mu

ltip

oin

t L

OD

sco

res

nonparametric LOD DSM IV totalDSM IV inattentionDSM IV hyperactiveparametric HLOD

Figure 5 LOD scores for ADHD on chromosome 5.

Table 5 Overlap of linkage results of current study, Bakker et al.,10 Fisher et al.,7 Ogdie et al.,8,9 and Arcos-Burgos et al.11

Chromosome Current study Bakker et al. (2003) Fisher et al. (2002)Ogdie et al. (2003, 2004)

Arcos-Burgos et al. (2004)

Location LOD Location MLSa Location MLSa Location NPLb

5p 17 cM 2.59 69 cM 1.43c 59 cM 2.55 5p13.3d NR6q 75 cM 0.58 166 cM 1.19e 89 cM 3.30 NR NR7p 88 cM 0.92 70 cM 3.04e NR NR NR NR8 8 cM 1.29 NR NR NR NR 67 cM 1.99q 104 cM 0.68 137 cM 2.05e NR NR NR NR11q 95 cM 0.41 NR NR 133 cM 1.27 113 cM 4.012q 166 2.10 NR NR 165 cM 1.09 NR NR17p 1 cM 1.39 NR NR 46 cM 3.63 12 cM 1.42

MLS, maximum likelihood score; NR, not reported.aCalculated with Mapmaker/sibs.bTwo-point, estimated with Simwalk2.cBroad diagnostic criteria.d‘Other nominal regions exhibiting concordance among different analyses were detected at chromosomes 5p13.3,y’11

eNarrow diagnostic criteria.


203


findings that the VNTR has an effect on DAT1expression levels.45 Instead, it was argued that thepreviously reported associations with psychiatricphenotypes and in particular ADHD might bemediated via linkage disequilibrium with function-ally relevant polymorphisms. This is consistent withour results that show no allelic association of theDAT1 VNTR but imply that other genetic variation inDAT1 or another gene in the 5p region is involved insusceptibility to ADHD.

Quite evidently, further studies are required in anattempt to uncover the genetic variation, that infunctional terms underlies our and previous chromo-some 5p linkage results and positive DAT1 associa-tion studies.3,5,6 Efforts should be undertaken togenotype the same DAT1 markers in those samplesthat have been analyzed for linkage and/or associa-tion, thus enabling meta-analyses based on singlemarkers and haplotypes.

It deserves to be pointed out that in addition tolinkage to chromosome 5p we identified furtherregions that potentially coincide with some of thosedetected in the three previous genome scans forADHD (Table 5). In particular, our nonparametricmultipoint LOD score of 1.39 on chromosome 17 at1 cM (0.57 Mb) is in the vicinity of peaks detected byboth Ogdie et al.8,9 and Arcos-Burgos et al.11 Simi-larly, our multipoint LOD score of 0.92 on chromo-some 7 at 88 cM (75 Mb) is in the same region as thehighest LOD score of 3.04 detected by Bakker et al.10

at 70 cM (49 Mb). Our nonparametric multipoint LODscore of 2.1 on chromosome 12 (166 cM, 130 Mb)coincides with a peak (MLS of 1.09) detected byFisher et al.7 (165 cM, 125 Mb). These overlappinglinkage results are extremely promising. They suggestthat specific ADHD genes are common to differentpopulations and that they can be detected in variousADHD samples despite differences in ascertainment,phenotyping and statistical analyses.

Acknowledgments

We thank the families for their participation in thisstudy. Seven physicians ascertained the families andperformed the phenotypical assessments; we greatlyacknowledge their substantial contribution to thisstudy. We thank Regina Pospiech and Inka Szangoliesfor excellent technical assistance and Gundula Ring-ler for exceptional data management. The GermanMinistry for Education and Research (National Gen-ome Research Net; grants 01GS0118, 01GS0482,01GS0483, 01GR0460) and the Deutsche Forschungs-gemeinschaft (grant SCHA 542/10-2) financially sup-ported this study.

Electronic-Database Information

The URLs for data presented herein are as follows:Mega2 version 3.0: http://watson.hgen.pitt.edu

Marshfield map: http://research.marshfieldclinic.org

Golden Path (hg17): http://genome.ucsc.edu

References

1 Faraone SV, Biederman J. Neurobiology of attention-deficithyperactivity disorder. Biol Psychiatry 1998; 44: 951–958.

2 Biederman J, Faraone SV, Keenan K, Benjamin J, Krifcher B, MooreC et al. Further evidence for family-genetic risk factors in attentiondeficit hyperactivity disorder. Patterns of comorbidity in probandsand relatives psychiatrically and pediatrically referred samples.Arch Gen Psychiatry 1992; 49: 728–738.

3 Heiser P, Friedel S, Dempfle A, Konrad K, Smidt J, Grabarkiewicz Jet al. Molecular Genetic Aspects of Attention-Deficit/Hyperactiv-ity Disorder. Neurosci Biobehav Rev 2004; 28: 625–641.

4 Faraone SV, Doyle AE. The nature and heritability of attention-deficit/hyperactivity disorder. Child Adolesc Psychiatr Clin N Am2001; 10: 299–316, viii–ix.

5 Maher BS, Marazita ML, Ferrell RE, Vanyukov MM. Dopaminesystem genes and attention deficit hyperactivity disorder: a meta-analysis. Psychiatr Genet 2002; 12: 207–215.

6 Purper-Ouakil D, Wohl M, Mouren MC, Verpillat P, Ades J,Gorwood P. Meta- analysis of family-based association studiesbetween the dopamine transporter gene and attention deficithyperactivity disorder. Psychiatr Genet 2005; 15: 53–59.

7 Fisher SE, Francks C, McCracken JT, McGough JJ, Marlow AJ,MacPhie IL et al. A genome-wide scan for loci involved inattention-deficit/hyperactivity disorder. Am J Hum Genet 2002; 70:1183–1196.

8 Ogdie MN, Macphie IL, Minassian SL, Yang M, Fisher SE, FrancksC et al. A genomewide scan for attention-deficit/hyperactivitydisorder in an extended sample: suggestive linkage on 17p11. Am JHum Genet 2003; 72: 1268–1279.

9 Ogdie MN, Fisher SE, Yang M, Ishii J, Francks C, Loo SK et al.Attention deficit hyperactivity disorder: fine mapping supportslinkage to 5p13, 6q12, 16p13, and 17p11. Am J Hum Genet 2004;75: 661–668.

10 Bakker SC, Meulen EM, Buitelaar JK, Sandkuijl LA, Pauls DL,Monsuur AJ et al. A whole-genome scan in 164 Dutch sib pairswith attention-deficit/hyperactivity disorder: suggestive evidencefor linkage on chromosomes 7p and 15q. Am J Hum Genet 2003;72: 1251–1260.

11 Arcos-Burgos M, Castellanos FX, Pineda D, Lopera F, Palacio JD,Palacio LG et al. Attention-deficit/hyperactivity disorder in apopulation isolate: linkage to loci at 4q13.2, 5q33.3, 11q22 and17p11. Am J Hum Genet 2004; 75: 998–1014.

12 Lander E, Kruglyak L. dissection of complex traits: guidelines forinterpreting and reporting linkage results. Nat Genet 1995; 11:241–247.

13 Smalley SL, Kustanovich V, Minassian SL, Stone JL, Ogdie MN,McGough JJ. Genetic linkage of attention-deficit/hyperactivitydisorder on chromosome 16p13, in a region implicated in autism.Am J Hum Genet 2002; 71: 959–963.

14 American Psychiatric Association. Diagnostic and StatisticalManual of Mental Disorders, 4th edn. American Psychiatric Press:Washington, DC, 1994.

15 Wechsler D. Examiner’s Manual: Wechsler Intelligence Scale forChildren-Third Edition. Psychological Corporation: New York,1991.

16 Tewes U, Rossmann R, Schallberger U. Der Hamburg-Wechsler-Intelligenztest fuer Kinder (HAWIK-III). Huber-Verlag: Bern, 1999.

17 Kaufman AS, Kaufman NL. K-ABC: Kaufman – Assessment Batteryfor Children. AGS Publishing: Minnesota, 1994.

18 Melchers P, Preuss U. K-ABC: Kaufman – Assessment Battery forChildren. Deutsche Bearbeitung. Swets & Zeitlinger: Amsterdam,1994.

19 Weiss RH. Grundintelligenztest Skala 2 (CFT 20) mit Wortschatz-test (WS) und Zahlenfolgentest (ZF). 4., ueberarbeitete Auflage.Westermann: Braunschweig, 1998.

20 Kaufman J, Birmaher B, Brent D, Rao U, Flynn C, Moreci P et al.Schedule for Affective Disorders and Schizophrenia for School-Age Children-Present and Lifetime Version (K-SADS-PL-D): initialreliability and validity data. J Am Acad Child Adolesc Psychiatry1997; 36: 980–988.

21 Delmo C, Weiffenbach O, Gabriel M, Poustka F. 3. Auflage derdeutschen Forschungsversion des K-SADS-PL-D, erweitert umICD-10-Diagnostik. 2002; 22.


204


22 American Psychiatric Association. Diagnostic and statisticalmanual of mental disorders, 3rd edn, rev. edn. AmericanPsychiatric Association: Washington, DC, 1987.

23 World Health Organization. The ICD-10 classification of mental andbehavioral disorders: Diagnostic criteria for research. Genf 1993.

24 Achenbach TM, Edelbrock CS. Psychopathology of childhood.Annu Rev Psychol 1984; 35: 227–256.

25 Achenbach TM. Empirically Based Taxonomy: How to UseSyndromes and Profile Types Derived From the CBCL From 4 to18, TRF, and WSR. Department of Psychiatry, University ofVermont, Burlington, 1993.

26 Remschmidt H, Walter R. Psychological symptoms in schoolchildren. An epidemiologic study. Z Kinder Jugendpsychiatr 1990;18: 121–132.

27 Doepfner M, Lehmkuhl G. Diagnostik-System fur psychischeStorungen im Kindes- und Jugendalter nach ICD-10 und DSM-IV.Huber: Bern, 1998.

28 Saar K, Geller F, Ruschendorf F, Reis A, Friedel S, Schauble N etal. Genome scan for childhood and adolescent obesity in Germanfamilies. Pediatrics 2003; 111: 321–327.

29 Abecasis GR, Cherny SS, Cookson WO, Cardon LR. GRR: graphicalrepresentation of relationship errors. Bioinformatics 2001; 17:742–743.

30 O’Connell JR, Weeks DE. PedCheck: a program for identification ofgenotype incompatibilities in linkage analysis. Am J Hum Genet1998; 63: 259–266.

31 Mukhopadhyay N, Almasy L, Schroeder M, Mulvihill WP, WeeksDE. Mega2, a data-handling program for facilitating genetic linkageand association analyses. Am J Hum Genet 1999; 65: A43632.

32 Abecasis GR, Cherny SS, Cookson WO, Cardon LR. Merlin-rapidanalysis of dense genetic maps using sparse gene flow trees. NatGenet 2002; 30: 97–101.

33 Abecasis GR, Wigginton JE. Handling marker–marker linkagedisequilibrium: pedigree analysis with clustered markers. Am JHum Genet, in press.

34 Whittemore AS, Halpern J. A class of tests for linkage usingaffected pedigree members. Biometrics 1994; 50: 118–127.

35 Kong A, Cox NJ. Allele-sharing models: LOD scores and accuratelinkage tests. Am J Hum Genet 1997; 61: 1179–1188.

36 Sham PC, Purcell S, Cherny SS, Abecasis GR. Powerful regression-based quantitative-trait linkage analysis of general pedigrees. Am JHum Genet 2002; 71: 238–253.

37 Bruehl B, Doepfner M, Lehmkuhl G. Der Fremdbeurteilungsbogenfur hyperkinetische Storungen (FBB-HKS) - Pravalenz hyperkine-tischer Storungen im Elternurteil und psychometrische Kriterien.Kindheit und Entwicklung 2000; 9: 115–125.

38 Li C, Scott LJ, Boehnke M. Assessing whether an allele canaccount in part for a linkage signal: the genotype-IBD sharing test(GIST). Am J Hum Genet 2004; 74: 418–431.

39 Martin ER, Monks SA, Warren LL, Kaplan NL. A test for linkageand association in general pedigrees: the pedigree disequilibriumtest. Am J Hum Genet 2000; 67: 146–154.

40 Cordell HJ. Sample size requirements to control for stochasticvariation in magnitude and location of allele-sharing linkagestatistics in affected sibling pairs. Ann Hum Genet 2001; 65:491–502.

41 Spencer T, Biederman J, Wilens T. Pharmacotherapy of attentiondeficit hyperactivity disorder. Child Adolesc Psychiatric ClinNorth Am 2000; 9: 77–97.

42 Dougherty DD, Bonab AA, Spencer TJ, Rauch SL, Madras BK,Fischman AJ. Dopamine transporter density in patients withattention deficit hyperactivity disorder. Lancet 1999; 354:2132–2133.

43 Krause KH, Dresel SH, Krause J, Kung HF, Tatsch K. Increasedstriatal dopamine transporter in adult patients with attentiondeficit hyperactivity disorder: effects of methylphenidate asmeasured by single photon emissioncomputed tomography.Neurosci Lett 2000; 285: 107–110.

44 Vles JS, Feron FJ, Hendriksen JG, Jolles J, van Kroonenburgh MJ,Weber WE. Methylphenidate down-regulates the dopamine re-ceptor and transporter system in children with attention deficithyperkinetic disorder (ADHD). Neuropediatrics 2003; 34: 77–80.

45 Mill J, Asherson P, Craig I, D’Souza U. Transient expression ofallelic variants of a VNTR in the dopamine transporter gene(DAT1). BMC Genetics 2005; 6: 3.

46 Kong A, Gudbjartsson DF, Sainz J, Jonsdottir GM, Gudjonsson SA,Richardsson B et al. A high-resolution recombination map of thehuman genome. Nat Genet 2002; 31: 241–247.


205


A genome-wide scan for attention-deficit/hyperactivity disorder in 155 German sib-pairs

Documents

Transcript of A genome-wide scan for attention-deficit/hyperactivity disorder in 155 German sib-pairs