effect of dopamine d3 receptor gene polymorphisms and clozapine treatment response: exploratory...

ORIGINAL ARTICLE

Effect of dopamine D3 receptor gene

polymorphisms and clozapine treatment

response: exploratory analysis of nine

polymorphisms and meta-analysis of the

Ser9Gly variant

R Hwang1, C Zai1, A Tiwari1,DJ Muller1, MJ Arranz2,AG Morris3, PJ McKenna4,J Munro2, SG Potkin5,JA Lieberman6, HY Meltzer7

and JL Kennedy1

1Neurogenetics Section, Centre for Addiction andMental Health, University of Toronto, Toronto,Ontario, Canada; 2King’s College London,Department of Psychological Medicine, Instituteof Psychiatry, Denmark Hill, London, UK;3Neurogenetics Group, Division of Neuroscienceand Mental Health, Department of Cellular andMolecular Neuroscience, Faculty of Medicine,Imperial College London, Hammersmith Hospital,London, UK; 4Benito Menni Complex Assitencialen Salut Mental, Germanes Hospitalaries delSagrat Cor de Jesus, Sant Boi de Llogregat,Barcelona, Spain; 5Brain Imaging Center,University of California, Irvine, Irvine, CA, USA;6University of North Carolina Medical School,Neurosciences Hospital, Chapel Hill, NC, USA and7Psychiatric Hospital at Vanderbilt University,Nashville, TN, USA

Correspondence:Dr JL Kennedy, Neurogenetics Section, Centrefor Addiction and Mental Health, University ofToronto, 250 College Street, R-76, Toronto,Ontario, Canada M5T 1R8.E-mail: [email protected]

Received 28 July 2009; revised 4 November2009; accepted 7 November 2009; publishedonline 22 December 2009

D2 blockade has been implicated in having a central role in antipsychoticresponse. However, treatment refractoriness, in spite of complete D2 blockade,as well as the efficacy of clozapine (CLZ) in a portion of this patient population,indicates the involvement of other factors as well. Several lines of evidencesuggest a role for D3. Furthermore, an earlier meta-analysis by Jonsson et al.(2003) (n¼233) suggested a role for genetic variation in the D3 gene. Relevantto this study, Jonsson et al. found the Ser allele of the D3 serine-to-glycinesubstitution at amino acid position 9 (Ser9Gly) polymorphism to be associatedwith worse CLZ response compared with the Gly allele. In this study, weattempt to validate these findings by performing a meta-analysis in a muchlarger sample (n¼758). Eight other variants were also tested in our own sampleto explore the possible effect of other regions of the gene. We report a negativebut consistent trend across individual studies in our meta-analysis for the DRD3Ser allele and poor CLZ response. A possible minor role for this single-nucleotidepolymorphism cannot be disregarded, as our sample size may have beeninsufficient. Other DRD3 variants and haplotypes of possible interest were alsoidentified for replication in future studies.The Pharmacogenomics Journal (2010) 10, 200–218; doi:10.1038/tpj.2009.65;published online 22 December 2009

Keywords: DRD3; gene; clozapine; antipsychotics; response; pharmacogenetics

Introduction

Treatment response in schizophrenia (SCZ) is heterogeneous.1 A recent multi-center study found that more than 70% of patients with chronic SCZdiscontinued their antipsychotic drug (APD) treatment because of pooreffectiveness or tolerability.2 Evidence from twin case studies has indicated thatthere may be a genetic basis for this interpatient variability.3,4 As the dopamine(DA) D2 receptor (D2) has been implicated in having a central role in the actionof APDs,5 several studies in the past have examined the possible associationbetween polymorphisms in the D2 gene (DRD2) and APD response. Results fromthese studies, however, cannot satisfactorily explain the observed variability in

The Pharmacogenomics Journal (2010) 10, 200–218& 2010 Nature Publishing Group All rights reserved 1470-269X/10

www.nature.com/tpj

mailto:[email protected]

http://dx.doi.org/10.1038/tpj.2009.65

http://www.nature.com/tpj

patient response to APDs.1,6 This fact, along with the efficacyof clozapine (CLZ) in some treatment-resistant patients, inspite of a low D2 affinity,7 suggests that other candidatesmay also be contributing factors to APD response.8

The DA D3 receptor (D3) is a candidate for consideration,as most APDs have binding affinities for D3 that are similarto those for D2 (reviewed in Sokoloff et al.9). Furthermore,postsynaptic D3, which is concentrated in the ventralstriatum,10 was observed to be elevated in SCZ patientswho had been removed from their APDs. In contrast, D3was downregulated in those patients who had remainedmedicated.11,12

This finding is especially interesting in the light of theproposed role of the ventral striatum in acting as a gate formodulating efferents from a number of other regions in thebrain that are disturbed in SCZ.13–15 In addition, this area isunder the control of the mesolimbic DA system. Therefore,the elevation of D3 levels in the ventral striatum observedby Gurevich et al.11 may reflect a hyperdopaminergic stateof the mesolimbic DA system that is well documented inSCZ.5,16 This might result in an altered neural processingin the striato-pallidal-thalamo-cortical limbic loop17,18 andultimately contribute to some of the symptoms seen inSCZ. The observed downregulation of D3 by APDs11,12 mayprovide a means of resolving the imbalance in this loop.19

Although findings from preclinical animal models of APDefficacy with selective D3 antagonists have not been entirelyconsistent,20–22 there is still plenty of evidence to suggest apossible role for D3 in APD response. D3 antagonists, highlyselective over D2, were shown to be effective in behaviouralmodels of positive symptoms of SCZ.23–26 Other studieshave shown a beneficial effect of D3 antagonism in modelsof negative symptoms,22,27,28 cognitive deficits29,30 andextrapyramidal symptoms.21,22,31,32 More relevant to thisstudy, several lines of evidence suggest that D3 blockademay be of particular importance for the efficacy of CLZ.These include studies that have observed the reversal ofphencyclidine- and neonatal ventral hippocampal lesion-induced prepulse inhibition (behavioural models of CLZresponse) by selective D3 antagonism,26,27 and similaritiesbetween selective D3 antagonists and CLZ in their effectson patterns of immediate early gene response33–36 anddepolarization inactivation of DA neurons.37–39

Genetic association studies on the D3 gene (DRD3) havefocused on a single-nucleotide polymorphism (SNP) thatcauses a serine-to-glycine substitution at amino acid posi-tion 9 (Ser9Gly) in the extracellular N-terminal domainof D3. Functional studies have found a significant increasein binding affinity for DA and D3 selective antagonists incell lines transfected with the Gly allele compared withthose transfected with the Ser allele.40,41 The Gly allele hasalso been observed to affect more robust DA responses inD3-mediated signal transduction pathways. These includemitogen-associated protein kinase, prostaglandin E2 (PGE2)and forskolin-stimulated cyclic adenosine monophosphate(cAMP) signalling pathways.40,42 One study, however,observed the opposite effect on forskolin-stimulated cAMPformation.42

Results from the association studies between DRD3 Ser9Glyand SCZ have been inconclusive. Although initial meta-analyses have provided some evidence for a marginalassociation between the Ser allele, Ser-containing genotypesand Ser9Gly homozygosity with SCZ,43–45 these findings havenot been supported by larger, more recent meta-analyses.46–48

These observations, however, do not preclude a variation inDRD3 to be irrelevant for APD response. In fact, to date,association studies between DRD3 Ser9Gly and APD responsehave generated promising findings. In an earlier meta-analysis performed by Jonsson et al.,44 responders to tradi-tional APDs were observed to have higher Ser allele, Ser/Serand homozygous genotype frequencies compared with non-responders, whereas the reverse was true for CLZ treatmentstudies. With a focus on just CLZ treatment studies, weattempt to validate the findings of Jonsson et al.44 in asubstantially larger sample (Part B) by performing compre-hensive meta-analyses containing all studies published todate. In this study, we were able to increase the total samplesize from 233 in the Jonsson et al. study44 to 758

To explore the possible effect of genetic variation in otherDRD3 regions on CLZ response, a separate component ofthis study includes the association analyses of eight otherSNPs in and around the DRD3 region in our own sample(Part A). Haplotype analyses were also performed.

Materials and methods

PART A: Association analyses in our sampleClinical samples. (i) Categorical response measure sample. Oursample consisted of 232 unrelated patients obtained fromthree research clinics: Case Western Reserve University inCleveland, Ohio (HY Meltzer, n¼ 105); Hillside Hospital inGlen Oaks, New York (JA Lieberman, n¼92); and Universityof California at Irvine (SG Potkin, n¼35). These subjectshad Diagnostic and Statistical Manual of Mental Disorders(DSM)-III-R or DSM-IV diagnoses of SCZ49 and almost allof whom met the criteria for treatment refractoriness orintolerance to typical APD therapy.50 Patients were treatedwith CLZ and evaluated prospectively for a minimum of 6months after a 2- to 4-week washout period, during whichthey did not receive any medication unless clinicallynecessary. CLZ blood levels were monitored throughout thecourse of treatment to ascertain compliance. Treatmentresponse was evaluated using the 18-item Brief PsychiatricRating Scale (BPRS) at the time of enrolment into the study(baseline) and after 6 months of CLZ treatment. These datawere dichotomized into the categorical responder/non-responder measure used in this study, on the basis of criteriadescribed by Kane et al.:50 patients with X20% reduction inthe overall BPRS score from baseline were categorized asresponders. This sample consisted of 183 Caucasians and 49African Americans. A detailed description of the demographiccharacteristics of each clinical and ethnic group in this samplecan be found in Table 1. Differences in response rates betweenclinical sites were not observed (P¼ 0.805). Therefore, datafrom the three clinical sites were analysed together

DRD3 polymorphisms and clozapine responseR Hwang et al

201

The Pharmacogenomics Journal

(ii) Continuous response measure sample. Quantitativetreatment response measures were available for a subsample(n¼132) of the patient population described in Part A(i). Thesedata include the overall 18-item BPRS score, a 4-item positivesymptom subscale (BPOS) score (includes conceptual disorga-nization, suspiciousness, hallucinations and unusual thoughtcontent) and a 3-item negative symptom subscale (BNEG)score (includes emotional withdrawal, motor retardation andblunted affect) assessed at baseline and after 6 months of CLZtreatment. For the individual SNP analyses, the score change atthe 6-month assessment from baseline was used as theresponse measure. For haplotype analyses, a percent (%)change score measure was used: % change score¼ (6-monthscore–baseline score)/(baseline score). This measure was usedbecause it is the only way to account for baseline effects in ahaplotype analysis. Our subsample consisted of 97 Caucasiansand 35 African Americans. A detailed description of the demo-graphic characteristics of each clinical and ethnic group inthis subsample can be found in Table 1. Differences betweenclinical sites in 6-month change scores (BPRS (P¼ 0.316);BPOS (P¼0.316); BNEG (P¼0.332)) and 6-month percentagechange scores (BPRS (P¼0.915); BPOS (P¼0.729); BNEG(P¼0.658)) were not observed. Therefore, data from the threeclinical sites were analysed together.

Gene polymorphism analyses. Blood samples were collectedfrom clinical sites and sent to the Centre for Addictionand Mental Health (CAMH), College Street site (Toronto,Ontario, Canada). Genomic DNA was extracted from whole-blood samples using the high-salt method described byLahiri and Nurnberger.51 Nine DRD3 SNPs were genotyped.From the 50- to 30-end of the DRD3 region, these SNPswere (1) rs905568 C/G (C¼Allele 1, G¼Allele 2) (outsideof DRD3 50-untranslated region (UTR), in the downstreamregion of the zinc-finger protein 80 gene, relative SNP position�60457); (2) rs2399504 G/A (G¼Allele 1, A¼Allele 2) (in the50-UTR, relative SNP position �47396); (3) rs7611535 A/G(A¼Allele 1, G¼Allele 2) (in the 50-UTR, relative SNP position�33304); (4) rs6762200 A/G (A¼Allele 1, G¼Allele 2) (in the50-UTR, relative SNP position �27794); (5) rs1394016 C/T(C¼Allele 1, T¼Allele 2) (in the 50-UTR, relative SNP position�19050); (6) rs6280 A/G (Ser9Gly) (A (Ser)¼Allele 1, G(Gly)¼Allele 2) (in exon 2, relative SNP position þ24); (7)rs167770 C/T (C¼Allele 1, T¼Allele 2) (in intron 2, relativeSNP position þ 11277); (8) rs2134655 A/G (A¼Allele 1,G¼Allele 2) (in intron 5, relative SNP position þ32638);and (9) rs2087017 C/T (C¼Allele 1, T¼Allele 2) (in the30-UTR, relative SNP position þ48826). The relative SNPlocations are shown in Figure 1. These SNPs were tag SNPs,chosen to minimize redundant information (high linkagedisequilibrium (LD)) and maximize coverage across the DRD3region. On the basis of data from the International HapMapProject (IHMP),52 we would have only needed to include threeadditional tag SNPs to have full coverage of DRD3 in the IHMPCaucasian sample (Utah residents with ancestry from northernand western parts of Europe) at a pairwise SNP r2 thresholdof 0.80. In the African sample (Yoruban in Ibadan, Nigeria),23 additional tag SNPs would have been required to fulfil thisT

ab

le1

Dem

og

rap

hic

dis

trib

uti

on

so

fcl

inic

al

site

san

deth

nic

gro

up

s:sa

mp

les

(i)

an

d(i

i)

Sam

ple

(i):

Cate

gorica

l(r

esp

on

der/

non

-resp

on

der)

data

Gro

up

Res

ponder

s/non-r

espond

ers

(n(%

))C

auca

sian/A

fric

an

Am

eric

an

(n(%

))M

ale

s/fe

male

s(n

(%))

Mea

nage±

s.d.

Meltze

r(n¼

105)

51/5

4(4

9/5

1)

79/2

6(7

5/2

5)

75/3

0(7

1/2

9)

32.5

±8.9

Lieb

erm

an

(n¼

92)

56/3

6(6

1/3

9)

76/1

6(8

3/1

7)

67/2

5(7

3/2

7)

34.9

±8.5

Potk

in(n¼

35)

15/2

0(4

3/5

7)

28/7

(80/2

0)

31/4

(89/1

1)

39.1

±6.5

Cauca

sian

(n¼

183)

97/8

6(5

3/4

7)

139/4

4(7

6/2

4)

34.4

±8.2

Afr

ican

Am

erica

n(n¼

49)

25/2

4(5

1/4

9)

34/1

5(6

9/3

1)

34.9

±10.2

Sam

ple

(ii)

:Q

uan

tita

tive

(%ch

an

ge

score

)d

ata

Gro

up

BPRS

BPO

SBN

EGC

auca

sian/A

fric

an

Am

eric

an

(n(%

))M

ale

s/fe

male

s(n

(%))

Mea

nage±

s.d.

Mea

nch

ange±

s.d.

(n)

Mea

n%

change±

s.d.

(n)

Mea

nch

ange±

s.d.

(n)

Mea

n%

change±

s.d.

(n)

Mea

nch

ange±

s.d.

(n)

Mea

n%

change±

s.d.

(n)

Meltze

r(n¼

97)

�9.0

9±

12.3

2(9

5)�

0.1

3±

1.1

5(9

5)�

3.1

5±

5.0

3(8

8)�

0.1

1±

0.9

4(8

4)�

1.1

8±

3.3

4(9

0)

0.0

1±

0.9

9(8

5)

73/2

4(7

5/2

5)

69/2

8(7

1/2

9)

32.2

±8.7

Potk

in(n¼

35)

�6.6

5±

12.0

2(3

5)�

0.1

0±

0.2

2(3

5)�

3.3

4±

6.2

3(3

5)�

0.1

7±

0.3

8(3

5)�

0.4

6±

4.5

4(3

5)

0.0

9±

0.4

1(3

5)

28/7

(80/2

0)

31/4

(89/1

1)

39.7

±7.3

Cauca

sian

(n¼

101)

�9.0

1±

12.6

3(9

9)�

0.1

1±

1.1

2(9

9)�

3.2

7±

5.8

0(9

6)�

0.0

9±

0.9

1(9

2)�

1.0

8±

3.5

7(9

8)�

0.0

2±

0.6

9(9

5)

79/2

2(7

8/2

2)

33.8

±10.2

Afr

ican

Am

erica

n(n¼

31)�

6.6

1±

10.9

0(3

1)�

0.1

5±

0.3

0(3

1)�

2.9

6±

3.5

1(2

7)�

0.2

4±

0.3

0(2

7)�

0.5

9±

4.2

2(2

7)

0.2

3±

1.3

3(2

5)

21/1

0(6

8/3

2)

34.9

±10.9

Ab

bre

viati

on

s:BN

EG

,BPRS

neg

ati

ve;

BPO

S,

BPRS

posi

tive

;BPRS,

Bri

ef

Psy

chia

tric

Rati

ng

Sca

le.


202


requirement. PCRs (10ml) on 20 ng genomic DNA wereperformed using TaqMan allele-specific assays with thefollowing conditions for all SNPs (with the exception ofrs905568): 95 1C for 10 min, followed by 40 cycles of 92 1C for15 s and 60 1C for 1 min. The following conditions were usedfor rs905568: 95 1C for 10min, followed by 60 cycles of 92 1Cfor 15 s and 60 1C for 1 min. Genotypes were determined usingthe ABI Prism 7000 Sequence Detection System with the allelicdiscrimination programme within the ABI software (AppliedBiosystems, Foster City, CA, USA). Results were confirmed bytwo researchers. All ambiguous genotypes were retyped. Geno-types that remained ambiguous were excluded from furtheranalyses. Of our samples, 10% were randomly regenotyped.The accuracy calls for all our SNPs in the Caucasian samplewere above 98.9%. In the African-American sample, allaccuracy calls were above 96.0%. All genotyping procedureswere performed at CAMH, College Street site. Laboratory staffmembers were blinded to psychiatric ratings.

Statistical analyses. (i) Individual SNP analyses. For thecategorical response data, genotype and allele groups acrosseach of the nine DRD3 SNPs were compared for responder/non-responder frequencies using the w2-test. Quantitative6-month change score (BPRS, BPOS and BNEG) distributionswere compared between genotype groups while taking intoaccount the effects of possible confounding factors usinganalysis of covariance. The statistical programme used toperform these tests was the Statistical Package for the SocialSciences for Windows (Release 15.0.1.1).53

(ii) LD analyses. These were performed using Haploview(version 4.1).54 LD blocks were defined on the basis of solidspine of LD criteria: D040.8.

(iii) Haplotype analyses. These were performed in bothCaucasian and African-American samples using UNPHASEDsoftware (Cambridge, UK) for genetic association analysis(version 2.402).55 The COCAPHASE programme (Cambridge,UK) within UNPHASED was used to assess the categoricalresponder/non-responder data, whereas QTPHASE withinUNPHASED was used to examine the quantitative percentagechange score data (BPRS, BPOS and BNEG). Global haplotypetests compared responder/non-responder proportions andmean percentage change scores across all haplotypes withina specific haplotype block. Individual haplotype tests comparedthe values of a specific haplotype against the pooled valuesof all other haplotypes with the block. Odds ratios (ORs)for significant findings from categorical data were generated

using calculator for confidence intervals of OR in anunmatched case–control Study.56

Part B: Meta-analyses of DRD3 Ser9Gly/CLZ response studiesClinical samples for meta-analyses. A computer search on theUnited States National Library of Medicine and NationalInstitutes of Health online search engine database, PubMed,was performed in order to identify all papers published up to15 October 2009 on association studies related to the DRD3Ser9Gly polymorphism and CLZ response. Six studies wereidentified: Shaikh et al.,57 Gaitonde et al.,58 Malhotra et al.,59

Scharfetter et al.,60 Arranz et al.61 and Barlas et al.62 However,Ser9Gly genotype data for Gaitonde et al.58 and Arranzet al.61 were not available from their respective publications.These data were obtained directly from the authors. In addi-tion to the 49 subjects previously reported on in Gaitondeet al.,58 we were provided with eight more subjects who wereadded to the sample after publication. Similarly, Arranzet al.61 provided data for 84 more subjects who were added tothe sample of 200 previously reported on. Table 2 provides adescription of these six samples. It should be noted, however,that Arranz et al.61 had incorporated the data of Shaikhet al.57 as a part of their study sample. Therefore, the studyby Shaikh et al.57 was excluded from our meta-analyses.Including the two samples from the association analyses inPart A of this study, our meta-analyses consisted of sevensamples in total, with a sample size of 758.

Statistical analysis. Meta-analyses of ORs on the effects ofSer allele vs Gly allele, Ser/Ser genotype vs Gly-containinggenotypes (Ser/Gly, Gly/Gly) and homozygote genotypes(Ser/Ser, Gly/Gly) vs the heterozygote genotype (Ser/Gly)were performed using the Stata statistical software package.63

OR and standard errors were calculated for individual studiesusing the ‘metan’ command, with the pooled OR and standarderror calculated under the random effects model. Ethnicity asa possible source of heterogeneity among studies was assessedby meta-regression analysis using the ‘metareg’ command. Wetested for publication bias using the ‘metabias’ command.

Results

PART A: Association analyses in our sampleIndividual SNP analyses. Genotype distributions for six of thenine DRD3 SNPs examined in this study were found to besignificantly different between Caucasians and African

6) rs6280 (Ser9Gly)

4) rs6762200

3) rs76115358) rs2134655

5) rs1394016

7) rs167770

10) rs1025398

9) rs2087017

1) rs905568

2) rs23995042 31 4 5 6 7

5’ 3’

Figure 1 Human dopamine D3 receptor gene structure and marker sites studied.


203


Americans (rs2399504, rs7611535 and rs1394016 werenot significantly different) (data not presented). In thelight of these differences, we decided to analyse the twopatient populations separately. No significant deviationfrom Hardy–Weinberg equilibrium was observed for anyof the nine SNPs in either Caucasian or African-Americansamples. The quantitative 6-month change responsemeasures (BPRS, BPOS and BNEG) were normally distri-buted and found to be significantly correlated with base-line scores. Baseline scores therefore were included as acovariate in the analysis of quantitative response measures.Possible confounding factors of age and gender did nothave a significant influence on any of these measures ineither population and therefore were not included inthe analysis.

No significant differences between genotypes or alleles ofany of the nine SNPs were observed for response to CLZ, asmeasured by the dichotomous responder/non-respondervariable, in either Caucasian or African-American samples(see Tables 3 and 4).

However, there were a few findings of possible interest incomparison with quantitative response measures (see Tables2 and 3). In the Caucasian sample, the A allele (allele 1) ofrs2134655 was associated with better response on the BPOSrating scale (P¼0.007), with perhaps a dose-dependenteffect. In African Americans, heterozygotes of rs2134655also responded better on BPOS than did GG homozygotes(no AA homozygotes were observed); however, this compari-son did not reach statistical significance (P¼0.089). A dose-dependent relationship between the T allele of rs1394016and better response on the BNEG scale (P¼0.018), as wellas a trend toward significance for the GG genotype of

rs2399504 and overall BPRS response (P¼0.058), was alsoobserved in African Americans.

No association was observed for the Ser9Gly variantwith any of the comparisons performed in either patientpopulation.

LD analyses. Pairwise LD analyses between the nine SNPs inboth Caucasian and African-American samples are presentedin Figure 2. As expected, a higher degree of LD was observedin the Caucasian sample than in the African-Americansample. This was also observed in the Caucasian and Africanpopulation data from the IHMP.52 The Caucasian sampledata from the IHMP, however, indicated a higher degreeof LD compared with data from the Caucasian sample inthis study. This may be because of a higher degree ofhomogeneity within the population sampled in the IHMPthan in our sample. The degrees of LD within our African-American sample and the IHMP African sample, however,were comparable.

Using a D040.8 criteria for pairwise LD, we observed two3-SNP LD blocks in our Caucasian sample: SNP blocks 2–3–4and 6–7–8. SNPs 8 and 9 also formed a two-SNP haplotypeblock. In our African-American sample, there was alsoevidence for LD between SNPs 2, 3 and 4 and between SNPs8 and 9. In addition, there were other two-SNP LD blocksobserved in the African-American sample: SNP blocks 1 and2 and 6 and 7.

Haplotype analyses. Two-marker and three-marker slidingwindow haplotype analyses across DRD3 were performed.These window sizes were chosen to appropriately examinehaplotypes within the specific LD blocks identified by

Table 2 Characteristics of studies included in meta-analyses

Study Sample location(ethnicity)

Sample size Diagnosis Duration ofCLZ treatment

Criteria for response

Shaikh et al.57 United Kingdom(Caucasian)

133 DSM-III-R SCZ;treatment refractoryor intolerant

43 months X20-point improvement on GAS

Gaitonde et al.58 United Kingdom(Caucasian)

57 (49+8a) DSM-III-R SCZ Not reported Not reported; used Present StateExamination and High RoydsEvaluation of Negativity

Malhotra et al.59 United States(Caucasian)

68 DSM-III-R SCZ or SAD 10 weeks X20% improvement on BPRS

Scharfetter et al.60 Pakistan(Caucasian)

32 DSM-IV SCZ;treatment refractory

6 months X50% improvement on BPRS

Arranz et al.61 United Kingdom(Caucasian)

284 (200+84a) DSM-III-R SCZ 43 months X20-point improvement on GAS

Barlas et al.62 Turkey (Caucasian) 89 DSM-IV SCZ 16 weeks X30% improvement on BPRSPresent sample(Caucasian)

UnitedStates(Caucasian)

180 DSM-III-R or DSM-IVSCZ


Present sample(African American)

United States(African American)

48 DSM-III-R or DSM-IVSCZ


Abbreviations: BPRS, Brief Psychiatric Rating Scale; CLZ, clozapine; DSM-III-R, Diagnostic and Statistical Manual of Mental Disorders III-R; GAS, Global Assessment Scale;

SAD, Schizoaffective Disorder; SCZ, schizophrenia.aAdditional, previously unpublished samples.


204


Tab

le3

Ind

ivid

ual

SN

Pan

aly

sis

of

cate

go

rica

l(r

esp

on

der/

no

n-r

esp

on

der

data

)(s

am

ple

i))

an

d6-m

on

thch

an

ge

sco

red

ata

(sam

ple

ii))

Cauca

sian

SN

PG

enoty

pe

Alle

le

11

12

22

P-v

alu

e1

2P-v

alu

e

N(f

req

uen

cy)

Resp

on

ders

27

(0.3

0)

46

(0.5

1)

17

(0.1

9)

0.4

89

100

(0.5

6)

80

(0.4

4)

0.5

01

Non

-resp

on

ders

31

(0.3

8)

35

(0.4

3)

16

(0.1

9)

97

(0.5

9)

67

(0.4

1)

rs905568

Mean

(N),

95%

CI,

s.d

.BPRS

Base

37.7

(27),

31.2

/44.1

,16.3

39.2

(51),

34.6

/43.8

,16.3

33.0

(13),

25.3

/40.7

,12.7

0.1

34

1¼

CD

�12.5

(27),�

18.1

/�6.9

,14.2

�9.1

(51),�

12.4

/�5.8

,11.8

�3.7

(13),�

10.2

/2.8

,10.8

2¼

GBPO

SBase

13.3

(28),

10.7

/15.9

,6.7

13.1

(49),

11.1

/15.2

,7.0

12.2

(12),

9.1

/15.2

,4.9

0.0

86

D�

5.3

(28),�

7.8

/�2.8

,6.4

�2.9

(49),�

4.3

/�1.5

,4.9

�1.6

(12),�

6.0

/2.8

,6.9

BN

EG

Base

7.1

(27),

5.6

/8.6

,3.8

7.0

(51),

5.6

/8.4

,5.1

7.9

(12),

3.5

/12.3

,6.9

0.4

66

D�

0.9

(27),�

2.4

/0.5

,3.6

�1.3

(51),�

2.2

/�0.3

,3.4

�0.4

(12),�

3.4

/2.6

,4.7

N(f

req

uen

cy)

Resp

on

ders

58

(0.6

5)

30

(0.3

3)

2(0

.02)

0.2

46

146

(0.8

1)

34

(0.1

9)

0.1

70

Non

-resp

on

ders

47

(0.5

7)

29

(0.3

6)

6(0

.07)

123

(0.7

5)

41

(0.2

5)

rs2399504

Mean

(N),

95%

CI,

s.d

.BPRS

Base

38.2

(61),

34.2

/42.1

,15.5

38.9

(27),

32.4

/45.4

,16.5

22.7

(3),�

8.4

/53.7

,12.5

0.8

38

1¼

GD

�9.9

(61),�

13.4

/�6.4

,13.6

�8.9

(27),�

13.0

/�4.7

,10.6

�1.7

(3),�

5.5

/2.1

,15.3

2¼

ABPO

SBase

13.3

(59),

11.7

/14.8

,6.0

13.2

(27),

10.1

/16.2

,7.7

7.7

(3),�

10.0

/25.3

,7.1

0.8

94

D�

3.6

(59),�

5.3

/�2.0

,6.3

�3.6

(27),�

5.5

/�1.7

,4.8

0(3

),�

4.3

/4.3

,1.7

BN

EG

Base

7.4

(60),

6.1

/8.7

,5.1

6.7

(27),

4.8

/8.6

,4.9

6.3

(3),�

6.4

/19.1

,5.1

0.3

45

D�

1.4

(60),�

2.3

/�0.4

,3.8

�0.3

(27),�

1.6

/1.1

,3.3

�2.3

(3),�

5.2

/0.5

,1.2

N(f

req

uen

cy)

Resp

on

ders

2(0

.02)

45

(0.5

0)

44

(0.4

8)

0.0

88

49

(0.2

7)

133

(0.7

3)

0.4

64

Non

-resp

on

ders

8(0

.10)

34

(0.4

1)

40

(0.4

9)

50

(0.3

0)

114

(0.7

0)

rs7611535

Mean

(N),

95%

CI,

s.d

.BPRS

Base

23.0

(3),�

7.5

/53.5

,12.3

39.4

(40),

34.0

/44.9

,17.0

37.7

(47),

33.3

/42.0

,14.9

0.7

32

1¼

AD

�3.0

(3),�

10.5

/4.5

,3.0

�8.8

(40),�

12.5

/�5.0

,11.8

�10.0

(47),�

14.0

/�6.0

,13.6

2¼

GBPO

SBase

8.3

(3),�

6.6

/23.3

,6.0

13.3

(40),

10.8

/15.4

,7.3

13.3

(45),

11.4

/15.1

,6.1

0.4

88

D�

0.3

3(3

),�

4.1

/3.5

,1.5

�2.8

(40),�

4.4

/�1.1

,5.2

�4.1

(45),�

5.9

/�2.2

,6.2

BN

EG

Base

5.3

(3),�

3.4

/14.1

,3.5

7.7

(40),

5.8

/9.5

,5.8

6.9

(46),

5.6

/8.2

,4.4

0.3

52

D�

2.0

(3),�

4.5

/0.5

,1.0

�0.8

(40),�

1.9

/0.3

,3.5

�1.3

(46),�

2.4

/�0.1

,3.9

N(f

req

uen

cy)

Resp

on

ders

9(0

.10)

53

(0.6

0)

27

(0.3

0)

0.3

11

71

(0.4

0)

107

(0.6

0)

0.6

81

Non

-resp

on

ders

14

(0.1

7)

41

(0.5

0)

27

(0.3

3)

69

(0.4

2)

95

(0.5

8)

rs6762200

Mean

(N),

95%

CI,

s.d

.BPRS

Base

32.9

(9),

22.6

/43.2

,13.4

39.8

(52),

35.4

/44.2

,15.8

36.2

(29),

29.9

/42.6

,16.7

0.1

23

1¼

AD

�4.2

(9),�

10.1

/1.6

,7.6

�8.5

(52),�

11.9

/�5.2

,12.1

�11.9

(29),�

17.3

/�6.4

,14.3

2¼

GBPO

SBase

12.1

(9),

8.1

/16.1

,5.2

13.3

(49),

11.3

/15.4

,7.1

12.8

(30),

10.3

/15.2

,6.5

0.1

11

D�

2.2

(9),�

5.5

/1.1

,4.3

�2.6

(49),�

4.3

/�1.0

,5.7

�4.8

(30),�

7.0

/�2.6

,5.9

BN

EG

Base

7.0

(9),

2.4

/11.6

,6.0

7.6

(51),

6.1

/9.1

,5.4

6.6

(29),

5.1

/8.1

,3.9

0.3

90

D0

(9),�

3.1

/3.1

,4.1

�1.1

(51),�

2.1

/�0.1

,3.6

�1.4

(29),�

2.8

/0,

3.7

N(f

req

uen

cy)

Resp

on

ders

29

(0.3

2)

47

(0.5

2)

14

(0.1

6)

0.4

65

105

(0.5

8)

75

(0.4

2)

0.5

18

Non

-resp

on

ders

20

(0.2

4)

50

(0.6

1)

12

(0.1

5)

90

(0.5

5)

74

(0.4

5)

rs1394016

Mean

(N),

95%

CI,

s.d

.BPRS

Base

39.1

(27),

32.8

/45.3

,15.8

37.9

(48)

33.1

/42.6

,16.5

35.9

(16),

28.1

/43.6

,14.5

0.8

65

1¼

CD

�9.4

(27),�

14.5

/�4.2

,13.0

�9.9

(48),�

13.7

/�6.1

,13.1

�7.6

(16),�

13.3

/�1.9

,10.7

2¼

TBPO

SBase

14.5

(28),

11.8

/17.3

,7.1

12.0

(47),

10.1

/13.9

,6.4

13.7

(14),

10.2

/17.2

,6.1

0.9

87

D�

4.1

(28),�

6.8

/�1.4

,7.0

�3.0

0(4

7),�

4.4

/�1.6

,4.8

�3.9

(14),�

7.7

/�0.2

,6.5

BN

EG

Base

7.1

(27),

5.4

/8.8

,4.3

7.4

(47),

5.9

/8.9

,5.4

6.6

(16),

3.5

/9.6

,5.8

0.5

48

D�

0.7

(9),�

2.1

/0.7

,3.5

�1.1

(51),�

2.1

/�0.1

,3.4

�1.6

(29),�

4.1

/�0.9

,4.7


205


Tab

le3

Conti

nued

Cauca

sian

SN

PG

enoty

pe

Alle

le

11

12

22

P-v

alu

e1

2P-v

alu

e

N(f

req

uen

cy)

Resp

on

ders

34

(0.3

6)

50

(0.5

2)

11

(0.1

2)

0.9

57

118

(0.6

2)

72

(0.3

8)

0.7

80

Non

-resp

on

ders

32

(0.3

8)

44

(0.5

2)

9(0

.10)

108

(0.6

4)

62

(0.3

6)

Ser9

Gly

Mean

(N),

95%

CI,

s.d

.BPRS

Base

38.2

(34),

32.8

/43.6

,15.5

36.6

(54),

32.0

/41.2

,16.8

34.1

(8)

23.2

/45.0

,13.0

0.6

72

1¼

A(S

er)

D�

10.6

(34),�

16.0

/�5.3

,15.3

�7.9

(54),�

10.9

/�4.9

,11.0

�6.8

(8),�

15.9

/2.4

,11.0

2¼

G(G

ly)

BPO

SBase

13.5

(35),

11.3

/15.6

,6.2

12.0

(51),

9.9

/14.1

,7.5

13.4

(8),

10.2

/16.5

,3.8

0.3

20

D�

4.6

(35),�

6.6

/�2.5

,5.9

�2.3

(51),�

3.9

/0.6

,5.8

�3.6

(8),�

8.2

/0.9

,5.4

BN

EG

Base

7.4

(34),

5.9

/9.0

,4.3

6.8

(53),

5.3

/8.2

,5.2

6.9

(8),

1.6

/12.1

,6.3

0.4

29

D�

1.1

(34),�

2.4

/0.1

,3.6

�1.1

(53),�

2.0

/�0.1

,3.5

0.4

(8),�

3.1

/3.9

,4.2

N(f

req

uen

cy)

Resp

on

ders

7(0

.08)

43

(0.4

8)

40

(0.4

4)

0.8

77

57

(0.3

2)

123

(0.6

8)

0.8

97

Non

-resp

on

ders

8(0

.10)

37

(0.4

5)

37

(0.4

5)

53

(0.3

2)

111

(0.6

8)

rs167770

Mean

(N),

95%

CI,

s.d

.BPRS

Base

35.6

(5),

17.6

/53.6

,14.5

37.6

(43),

32.3

/42.8

,17.2

38.4

(43),

33.9

/43.0

,14.8

0.3

66

1¼

CD

�3.2

(5),�

14.4

/8.0

,9.0

�8.4

(43),�

12.0

/�8.3

,11.5

�11.0

(43),�

15.2

/�6.7

,13.9

2¼

TBPO

SBase

12.4

(5),

6.7

/18.1

,4.6

12.8

(42),

10.5

/15.1

,7.4

13.4

(42),

11.4

/15.2

,6.0

0.4

12

D�

0.6

(5),�

6.1

/4.9

,4.4

�3.1

(42),�

5.0

/�1.3

,5.8

�4.1

(42),�

6.0

/�2.3

,5.9

BN

EG

Base

8.8

(5),�

0.6

/18.2

,7.6

6.6

(43),

5.1

/8.1

,4.9

7.5

(42),

6.0

/9.0

,4.9

0.1

64

D0.2

(5),�

4.7

/5.1

,4.0

�0.5

(43),�

1.5

/0.5

,3.3

�1.7

(42),�

3.0

/�0.5

,3.9

N(f

req

uen

cy)

Resp

on

ders

6(0

.07)

32

(0.3

5)

52

(0.5

8)

0.8

18

44

(0.2

4)

136

(0.7

6)

0.7

05

Non

-resp

on

ders

5(0

.06)

33

(0.4

0)

44

(0.5

4)

43

(0.2

6)

121

(0.7

4)

rs2134655

Mean

(N),

95%

CI,

s.d

.BPRS

Base

37.3

(4),

11.7

/62.8

,16.1

39.6

(33),

34.3

/44.9

,14.8

37.0

(53),

32.4

/41.6

,16.7

0.3

25

1¼

AD

�16.8

(4),�

31.1

/�2.4

,9.0

�10.1

(33),�

15.1

/�5.0

,14.2

�8.1

(53),�

11.3

/�4.9

,11.7

2¼

GBPO

SBase

14.5

(4),

5.3

/23.7

,5.8

14.0

(34),

11.8

/16.2

,6.4

12.2

(50),

10.3

/14.2

,6.8

0.0

07

D�

11.5

(4),�

19.8

/�3.2

,5.2

�3.6

(34),�

5.7

/�1.6

,5.7

�2.5

(50),�

4.0

/�1.0

,5.2

BN

EG

Base

8.8

(4),

2.9

/14.6

,3.7

6.4

(34),

5.2

/7.7

,3.5

7.6

(51),

6.0

/9.2

,5.9

0.5

40

D0

(4),�

4.3

/4.3

,2.7

�0.9

(34),�

2.2

/0.4

,3.6

�1.3

(51),�

2.3

/�0.2

,3.8

N(f

req

uen

cy)

Resp

on

ders

17

(0.1

9)

47

(0.5

1)

27

(0.3

0)

0.5

56

81

(0.4

5)

101

(0.5

5)

0.3

02

Non

-resp

on

ders

11

(0.1

3)

42

(0.5

1)

29

(0.3

6)

64

(0.3

8)

100

(0.6

2)

rs2087017

Mean

(N),

95%

CI,

s.d

.BPRS

Base

36.9

(17),

29.1

/44.7

,15.2

38.9

(49),

33.9

/43.8

,17.4

36.6

(25),

31.1

/42.1

,13.3

0.5

75

1¼

CD

�9.7

(17),�

17.3

/�2.1

,14.7

�10.5

(49),�

14.2

/�6.7

,13.2

�6.8

4(2

5),�

10.9

/�2.8

,9.8

2¼

TBPO

SBase

12.7

(18),

9.5

/15.9

,6.5

13.0

(46),

10.8

/15.2

,7.4

13.4

(25),

11.3

/15.6

,5.2

0.8

65

D�

3.9

(18),�

7.5

/�0.4

,7.2

�3.3

(46),�

5.0

/�1.7

,5.7

�3.4

(25),�

5.5

/�1.3

,5.1

BN

EG

Base

8.3

(17),

5.1

/11.5

,6.2

6.9

(47),

5.6

/8.2

,4.3

6.8

(26),

4.6

/9.1

,5.5

0.6

23

D�

1.1

(17),�

3.0

/0.9

,3.7

�1.3

(47),�

2.3

/�0.1

,3.6

�0.7

(26),�

2.3

/0.9

,3.9

Ab

bre

viation

s:BN

EG

,BPRS

neg

ative

;BPO

S,

BPRS

posi

tive

;BPRS,

Brief

Psy

chia

tric

Ratin

gSca

le;

CI,

con

fid

en

cein

terv

al;

Ser9

Gly

,se

rin

e-t

o-g

lyci

ne

sub

stit

uti

on

at

am

ino

aci

dp

osi

tion

9;

SN

P,

sin

gle

-nucl

eoti

de

poly

morp

his

m.

Bold

ed

valu

es

ind

icate

aP-v

alu

eo

0.0

5.


206


Tab

le4

Ind

ivid

ual

SN

Pan

aly

sis

of

cate

go

rica

l(r

esp

on

der/

no

n-r

esp

on

der

data

)(s

am

ple

i))

an

d6-m

on

thch

an

ge

sco

red

ata

(sam

ple

ii))

Afr

ican

Am

eric

an

SN

PG

enoty

pe

Alle

le

11

12

22

P-v

alu

e1

2P-v

alu

e

N(f

req

uen

cy)

Resp

on

ders

2(0

.08)

8(0

.32)

15

(0.6

0)

0.9

13

12

(0.2

4)

38

(0.7

6)

0.7

26

Non

-resp

on

ders

2(0

.08)

9(0

.38)

13

(0.5

4)

13

(0.2

7)

35

(0.7

3)

rs905568

Mean

(N),

95%

CI,

s.d

.BPRS

Base

65.0

(2),�

62/1

92,

14.1

35.8

(12),

29.3

/42.2

,10.2

33.8

(17),

26.4

/41.2

,14.4

0.7

62

1¼

CD

�19.5

(2),�

140/1

01,

13.4

�5.0

(12),�

11.7

/1.7

,10.5

�6.2

(17),�

11.7

/�0.8

,10.6

2¼

GBPO

SBase

21.5

(2),

2.4

/40.6

,2.1

13.7

(9),

10.5

/16.8

,4.1

13.2

(16),

10.1

/16.2

,5.7

0.9

42

D�

3.5

(2),�

35.3

/28.3

,3.5

�3.2

(9),�

6.1

/�0.3

,3.8

�2.8

(16),�

4.7

/�0.8

,12.7

BN

EG

Base

13.0

(2),�

37.8

/63.8

,5.7

8.3

(9),

4.8

/11.9

,4.6

5.6

(16),

3.5

/7.8

,4.1

0.3

92

D�

6.5

(2),�

51.0

/38.0

,4.9

�0.6

(9),�

4.3

/3.2

,4.8

0.1

(16),�

1.7

/1.9

,3.4

N(f

req

uen

cy)

Resp

on

ders

20

(0.8

3)

3(0

.13)

1(0

.04)

0.0

80

43

(0.9

0)

5(0

.10)

0.2

13

Non

-resp

on

ders

14

(0.6

1)

9(0

.39)

0(0

.00)

37

(0.8

0)

9(0

.20)

rs2399504

Mean

(N),

95%

CI,

s.d

.BPRS

Base

34.9

(19),

28.8

/41.1

,12.8

33.9

(10),

24.6

/43.2

,13.0

—0.0

58

1¼

GD

�8.4

(19),�

13.1

/�3.7

,9.7

�0.7

(10),�

8.0

/6.6

,10.2

—2¼

ABPO

SBase

13.1

(16),

10.2

/15.9

,5.3

13.9

(9),

10.1

/17.7

,5.0

—0.6

55

D�

3.1

(16),�

5.2

/�1.0

,3.9

�2.6

(9),�

5.0

/�0.2

,3.1

—BN

EG

Base

6.1

(16),

3.9

/8.3

,4.2

7.4

(9),

3.7

/11.2

,4.9

—0.7

01

D0.3

(16),�

1.7

/2.2

,3.6

�0.8

(9),�

4.3

/2.7

,4.6

—

N(f

req

uen

cy)

Resp

on

ders

1(0

.04)

7(0

.29)

16

(0.6

7)

0.5

83

9(0

.19)

39

(0.8

1)

0.3

93

Non

-resp

on

ders

1(0

.04)

10

(0.4

4)

12

(0.5

2)

12

(0.2

6)

34

(0.7

4)

rs7611535

Mean

(N),

95%

CI,

s.d

.BPRS

Base

35.0

(1)

37.2

(1),

26.4

/47.9

,17.8

36.2

(13),

29.6

/42.7

,12.7

0.1

85

1¼

AD

1.0

(1)

�3.5

(13),�

11.5

/4.4

,13.2

�9.4

(17),�

13.8

/�5.0

,8.5

2¼

GBPO

SBase

15.0

(1)

14.8

(12),

11.4

/18.3

,5.4

13.1

(14),

9.9

/16.4

,5.7

0.6

20

D�

4.0

(1)

�2.3

(12),�

4.6

/0,

3.6

�3.4

(14),�

5.5

/�1.4

,3.6

BN

EG

Base

6.0

(1)

8.0

(12),

4.5

/11.5

,5.5

6.4

(14),

4.0

/8.7

,4.0

0.5

26

D0

(1)

�1.9

(12),�

5.1

/1.3

,5.1

0.5

(14),�

1.4

/2.4

,3.3

N(f

req

uen

cy)

Resp

on

ders

11

(0.4

4)

10

(0.4

0)

4(0

.16)

0.3

75

32

(0.6

4)

18

(0.3

6)

0.4

71

Non

-resp

on

ders

11

(0.4

6)

12

(0.5

0)

1(0

.04)

34

(0.7

1)

14

(0.2

9)

rs6762200

Mean

(N),

95%

CI,

s.d

.BPRS

Base

39.3

(15),

30.2

/48.3

,16.4

32.6

(14),

24.3

/40.0

,12.7

43.5

(2),�

64.5

/151,

12.0

0.9

37

1¼

AD

�7.7

(15),�

14.8

/�0.7

,12.8

�4.8

(14),�

10.2

/0.6

,9.4

�11.0

(2),�

61.8

/39.8

,5.7

2¼

GBPO

SBase

14.9

(14),

11.7

/18.1

,5.5

13.0

(11),

9.2

/16.8

,5.6

12.5

(2),�

32.0

/57.0

,4.9

0.8

76

D�

3.1

(14),�

5.2

/�0.9

,3.7

�2.6

(11),�

5.0

/�0.3

,3.5

�4.0

(2),�

42.1

/34.1

,4.2

BN

EG

Base

6.5

(14),

3.4

/9.6

,5.3

7.5

(11),

4.6

/10.5

,4.4

8.5

(2),

2.1

/14.9

,0.7

0.5

69

D�

0.1

(14),�

2.6

/2.3

,4.3

�0.5

(11),�

3.6

/2.5

,4.5

�4.0

(2)

N(f

req

uen

cy)

Resp

on

ders

11

(0.4

4)

9(0

.36)

5(0

.20)

0.4

18

31

(0.6

2)

19

(0.3

8)

0.6

30

Non

-resp

on

ders

10

(0.4

2)

12

(0.5

0)

2(0

.08)

32

(0.6

7)

16

(0.3

3)

rs1394016

Mean

(N),

95%

CI,

s.d

.BPRS

Base

34.7

(15),

27.8

/41.5

,12.4

33.3

(13),

25.4

/41.3

,13.2

60.0

(3),

27.7

/92.3

,13.0

0.3

33

1¼

CD

�6.7

(15),�

12.5

/�0.9

,10.5

�3.4

(13),�

9.1

/2.3

,9.4

�20.3

(3),�

49.4

/8.8

,11.7

2¼

TBPO

SBase

14.1

(13),

14.1

/17.3

,5.3

12.8

(11),

9.0

/16.6

,5.7

17.7

(3),

5.9

/29.4

,4.7

0.8

99

D�

3.3

(13),�

5.7

/�0.9

,3.9

�2.5

(11),�

4.7

/�0.2

,3.4

�3.3

(3),�

9.6

/2.9

,2.5

BN

EG

Base

4.8

(13),

2.3

/7.4

,4.3

8.3

(11),

5.7

/10.8

,3.8

12.3

(3),

1.1

/23.5

,4.5

0.0

18

D1.3

(13),�

0.5

/3.1

,3.0

�0.9

(11),�

3.4

/1.6

,3.7

�7.7

(3),�

15.7

/0.3

,3.2


207


Tab

le4

Conti

nued

Afr

ican

Am

eric

an

SN

PG

enoty

pe

Alle

le

11

12

22

P-v

alu

e1

2P-v

alu

e

N(f

req

uen

cy)

Resp

on

ders

1(0

.04)

9(0

.36)

15

(0.6

0)

0.7

98

11

(0.2

2)

39

(0.7

8)

0.6

39

Non

-resp

on

ders

2(0

.09)

8(0

.35)

13

(0.5

6)

12

(0.2

6)

34

(0.7

4)

Ser9

Gly

Mean

(N),

95%

CI,

s.d

.BPRS

Base

—34.3

(13),

27.6

/41.0

,11.0

37.2

(17),

28.5

/45.8

,16.8

0.6

60

1¼

A(S

er)

D—

�5.0

(13),�

11.1

/1.1

,10.0

�7.6

(17),�

13.8

/�1.5

,12.0

2¼

G(G

ly)

BPO

SBase

—13.4

(10),

10.6

/16.2

,4.0

13.9

(16),

10.6

/17.2

,6.2

0.9

25

D—

�2.9

(10),�

5.6

/�0.2

,3.7

�3.1

(16),�

5.0

/�1.2

,3.6

BN

EG

Base

—7.9

(10),

4.7

/11.1

,4.5

6.4

(16),

3.8

/9.1

,4.9

0.9

24

D—

�0.8

(10),�

4.1

/2.5

,4.6

�0.3

(16),�

2.5

/1.9

,4.2

N(f

req

uen

cy)

Resp

on

ders

11

(0.4

6)

10

(0.4

2)

3(0

.12)

0.7

25

32

(0.6

7)

16

(0.3

3)

0.4

24

Non

-resp

on

ders

8(0

.35)

11

(0.4

8)

4(0

.17)

27

(0.5

9)

19

(0.4

1)

rs167770

Mean

(N),

95%

CI,

s.d

.BPRS

Base

41.9

(9),

29.9

/53.9

,15.7

33.8

(17),

26.4

/41.1

,14.2

36.4

(5),

18.7

/54.1

,14.2

0.8

16

1¼

CD

�9.6

(9),�

18.5

/�0.6

,11.6

�4.6

(17),�

9.2

/0,

8.9

�8.0

(5),

28.1

/12.1

,16.2

2¼

TBPO

SBase

14.2

(9),

10.4

/18.0

,4.9

14.0

(13),

10.0

/18.0

,6.5

13.4

(5),

9.1

/17.7

,3.4

0.7

15

D�

3.8

(9),�

6.2

/�1.4

,3.2

�2.7

(13),�

4.6

/�0.8

,3.2

�2.2

(5),�

8.7

/4.3

,5.3

BN

EG

Base

7.9

(9),

3.8

/12.0

,5.3

7.9

(13),

5.4

/10.5

,4.2

3.4

(5),�

0.9

/7.7

,3.4

0.9

67

D�

1.2

(9),�

5.0

/2.6

,5.0

�0.8

(13),�

3.2

/1.5

,3.8

1.2

(5),�

3.9

/6.3

,4.1

N(f

req

uen

cy)

Resp

on

ders

0(0

)2

(0.0

8)

22

(0.9

2)

0.3

52

2(0

.04)

46

(0.9

6)

0.3

69

Non

-resp

on

ders

0(0

)4

(0.1

7)

19

(0.8

3)

4(0

.09)

42

(0.9

1)

rs2134655

Mean

(N),

95%

CI,

s.d

.BPRS

Base

—31.0

(6),

24.9

/37.1

,5.8

37.9

(25),

31.4

/44.4

,15.8

0.3

95

1¼

AD

—�

1.7

(6),�

10.8

/7.5

,8.7

�7.8

(25),�

12.4

/�3.2

,11.2

2¼

GBPO

SBase

—14.2

(5),

10.8

/17.6

,2.8

13.9

(22),

11.3

/16.5

,5.9

0.0

89

D—

�5.4

(5),�

8.6

/�2.2

,2.6

�2.4

(22),�

4.0

/�0.9

,3.5

BN

EG

Base

—5.8

(5),

2.1

/9.5

,2.9

7.4

(22),

5.2

/9.6

,5.0

0.2

87

D—

1.6

(5),�

5.3

/8.5

,5.5

�1.1

(22),�

2.8

/0.6

,3.8

N(f

req

uen

cy)

Resp

on

ders

10

(0.4

0)

9(0

.36)

6(0

.24)

0.4

41

29

(0.5

8)

21

(0.4

2)

0.8

61

Non

-resp

on

ders

7(0

.29)

13

(0.5

4)

4(0

.17)

27

(0.5

6)

21

(0.4

4)

rs2087017

Mean

(N),

95%

CI,

s.d

.BPRS

Base

30.3

(9),

20.9

/39.8

,12.3

38.9

(14),

31.0

/46.7

,13.6

39.5

(8),

24.4

/54.6

,18.1

0.1

01

1¼

CD

�8.3

(9),�

14.0

/�2.7

,7.3

�3.3

(14),�

9.0

/2.4

,9.9

�10.5

(8),�

22.9

/1.9

,14.9

2¼

TBPO

SBase

10.3

(7),

4.7

/15.9

,6.1

15.6

(12),

12.4

/18.8

,5.0

14.8

(8),

11.2

/18.3

,4.3

0.6

49

D�

2.4

(7),�

6.0

/1.1

,3.8

�2.6

(12),�

4.5

/�0.6

,3.1

�4.0

(8),�

7.4

/�0.6

,4.1

BN

EG

Base

6.6

(7),

4.0

/9.2

,2.8

8.1

(12),

4.8

/11.3

,5.1

6.0

(8),

1.4

/10.6

,5.5

0.7

87

D�

0.6

(7),�

3.5

/2.4

,3.2

�0.5

(12),�

3.2

/2.2

,4.2

�0.8

(8),�

5.3

/3.8

,5.4

Ab

bre

viation

s:BN

EG

,BPRS

neg

ative

;BPO

S,

BPRS

posi

tive

;BPRS,

Brief

Psy

chia

tric

Ratin

gSca

le;

CI,

con

fid

en

cein

terv

al;

Ser9

Gly

,se

rin

e-t

o-g

lyci

ne

sub

stitution

at

am

ino

aci

dp

osi

tion

9;

SN

P,

sin

gle

-nucl

eotid

e

poly

morp

his

m.

Bold

ed

valu

es

ind

icate

aP-v

alu

eo

0.0

5.


208


LD analyses (two-marker and three-marker LD blocks, seeFigure 2). Moreover, the sliding window approach waschosen so that we could systematically explore otherhaplotypes in areas of lower LD.

In the Caucasian sample, haplotype analyses within theblocks identified from the LD analyses (SNP blocks 2–3–4,6–7–8 and 8–9) found a rare haplotype in the SNP block6–7–8 (haplotype 1–1–2) to be associated with response(P¼0.037) (see Table 5). This haplotype was associated withnon-responder CLZ treatment outcome status. This findingseems to be because of the effect of haplotype 1–1 in SNPwindow 6–7, as both haplotypes were observed to havesimilar degrees of association with responder/non-responderstatus.

Other significant results from the Caucasian sampleinclude the association of Ser9Gly containing haplotypes:haplotype 1–2 in SNP window 6–7 with good BNEG response(P¼0.045), haplotype 1–1 from SNP block 6–7 that wasassociated with non-responder status (P¼0.035) and haplo-type 1–2–1 from SNP block 5–6–7 that was associated withBNEG response (P¼0.044).

In addition, haplotype 1–2 in SNP window 7–8 wasassociated with poor BNEG response (P¼0.042). Haplotype1–2–1 from SNP block 1–2–3 was associated with non-responder status (P¼0.010). This was most likely because ofhaplotype 1–2 from SNP block 1–2, as it had the same effect.

Finally, haplotype 2–2–2 from SNP window 3–4–5 wasassociated with poor BPRS response as measured by thepercentage change score (P¼0.033).

In the African-American sample, no haplotypes withinthe LD blocks identified from the LD analyses (SNP blocks2–3–4, 1–2, 6–7 and 8–9) were found to be significant (seeTable 6). Instead, haplotypes that contained either haplo-type 1–2 or 2–1 from SNP window 4–5 were associatedwith good overall CLZ response as measured by thecategorical responder/non-responder variable (P¼ 0.017)and overall BPRS percentage change (P¼0.049). Theseinclude haplotype 2–2–1 from SNP window 3–4–5(P¼0.021), and haplotypes 1–2–2 (P¼0.035) and 2–1–2(P¼0.012) from SNP window 4–5–6. In addition, haplotype1–1–1 from SNP window 3–4–5, which included the 1–1SNP 4–5 haplotype, was associated with poor overall BPRSresponse (P¼0.010); further, haplotype 2–2–2 from SNPwindow 4–5–6 was associated with good BPOS response(P¼0.027).

Other significant findings in the African-American sampleinclude haplotype 1–2 from SNP window 2–3 that wasassociated with better overall BPRS response (P¼0.029) andhaplotype 1–1 from SNP window 3–4 that was associatedwith worse overall BPRS response (P¼0.049).

Part B: Meta-analyses of DRD3 Ser9Gly/CLZ response studies

DRD3 Ser9Gly data from each of the studies included in themeta-analyses are presented in Table 2.

We did not find significant differences in the Ser vs Glyallele, Ser/Ser vs Gly-carrier genotypes or homozygote vsheterozygote genotypes in responders compared with non-responders.

The Ser allele (OR¼ 0.82, 95% confidence interval (CI):0.65–1.04; P¼0.100) (Figure 3), Ser/Ser genotype (OR¼0.75,95% CI: 0.53–1.06; P¼0.108) (Figure 4) and homo-zygote genotypes (OR¼0.87, 95% CI: 0.64–1.17; P¼0.345)(Figure 5), however, occurred more frequently in non-responders than in responders.

There was no evidence of heterogeneity among studiesbecause of ethnicity or publication bias for any of thecomparisons (data not shown).

Discussion

One of the objectives of this study was to replicate andextend the findings of the Jonsson et al. meta-analyses.44 Intheir study, Jonsson et al. found the DRD3 Ser allele, theSer/Ser genotype and homozygote genotypes (Ser/Ser andGly/Gly genotypes combined) to occur significantly morefrequently in non-responders to CLZ compared with the Glyallele (OR¼0.64, 95% CI: 0.43–0.94; P¼ 0.021), the Ser/Glyand Gly/Gly genotypes combined (OR¼ 0.45, 95% CI: 0.27–0.77; P¼0.003) and the heterozygote genotype (Ser/Gly)(OR¼0.49, 95% CI: 0.29–0.83; P¼0.007), respectively. Theeffect sizes reported by Jonsson et al. were quite strong fora genetic association study in a complex phenotype. One ofthe weaknesses of that study, however, was that it includeddata from just three of the five DRD3 Ser9Gly/CLZ response

Caucasian (D’; 95% Confidence Interval; r2)

0.440.27-0.5

0.650.39-0.81

0.370.11-0.57

0.490.31-0.63

0.080.00-0.23

0.650.47-0.77

0.500.28-0.67

0.860.66-0.95

9

0.11 0.09 0.04 0.12 0.01 0.18 0.09 0.18 rs2087017

0.920.75-0.98

0.490.12-0.74

0.660.31-0.84

0.720.48-0.86

0.810.62-0.91

1.000.84-1.00

1.000.80-1.00

8 1.000.23-1.00

0.120.060.020.21 0.17 0.19 0.15 rs2134655 0.09

0.800.69-0.88

0.780.65-0.87

0.760.66-0.84

0.870.77-0.93

0.090.00-0.34

0.960.89-0.99

7 0.570.06-0.88

0.180.01-0.44

0.40 0.36 0.50 0.52 0.00 0.72 rs167770 0.04

0.870.79-0.93

0.750.60-0.85

0.760.64-0.85

0.910.84-0.95

0.140.01-0.30

6 1.000.77-1.00

0.730.18-0.92

0.020.00-0.50

0.61 0.26 0.73 rs6280 (Ser9Gly) 0.48 0.14

0.06-.01-0.20

0.100.00-0.35

0.100.00-0.31

0.080.00-0.23

5 0.560.22-0.77

0.130.01-0.59

1.000.05-0.97

0.060.00-0.35

0.00 0.00 0.01 0.01 rs1394016 0.17 0.01 0.04 0.00

0.810.73-0.87

0.890.76-0.96

1.000.94-1.00

4 0.460.18-0.64

0.860.63-0.95

0.500.24-0.68

0.600.06-0.89

0.010.00-0.42

0.60 0.33 0.59 rs6762200 0.18 0.49 0.20 0.05 0.00

0.790.66-0.87

0.880.79-0.94

3 1.000.37-1.00

0.140.01-0.44

1.000.13-0.99

0.780.21-0.94

0.730.18-0.92

0.490.14-0.73

0.33 0.56 rs7611535 0.14 0.01 0.08 0.11 0.11 0.100.58

0.38-0.722 1.00

0.81-1.00

1.000.20-1.00

0.200.02-0.56

1.000.09-0.98

1.000.33-1.00

0.520.09-0.82

0.360.04-0.69

0.12 rs2399504 0.71 0.10 0.02 0.06 0.12 0.040.09

1 1.000.11-0.99

0.320.03-0.77

0.660.41-0.82

0.710.44-0.86

0.760.57-087

0.560.23-0.77

0.710.14-0.92

0.060.00-0.40

rs905568 0.06 0.01 0.31 0.30 0.58 0.18 0.11 0.00

African American (D’; 95% Confidence Interval; r2)

*shaded pair-wise comparisons indicate D’ > 0.80

0.02

0.000.010.39

Figure 2 Pairwise single-nucleotide polymorphism (SNP) linkage

disequilibrium (LD) analyses in the Caucasian and African-Americansamples. *Shaded pairwise comparisons indicate D0 40.80.


209


Tab

le5

Slid

ing

win

do

wh

ap

loty

pe

an

aly

ses

acr

oss

DR

D3

for

cate

go

rica

lre

spo

nse

data

inb

oth

Cau

casi

an

an

dA

fric

an

-Am

eri

can

sam

ple

s

Glo

balP-v

alu

eH

aplo

type

Res

ponder

(n(e

stim

ate

dfr

equen

cy))

Non-r

esponder

(n(e

stim

ate

dfr

equen

cy))

Odds

rati

o(9

5%

CI)

w2P-v

alu

e

Cauca

sian

3-S

NP

win

dow

1–2–3

0.0

74

1–2–1

1.2

(0.0

1)

9.4

(0.0

6)

8.6

3(1

.30–57.1

2)

6.5

80.0

10

2–3–4

0.5

11

3–4–5

0.4

47

4–5–6

0.5

64

5–6–7

0.1

35

1–1–1

3.8�

10�

24

(2.1�

10�

26)

3.3

(0.0

2)

9.6

0(0

–N

)4.4

80.0

34

6–7–8

0.3

19

1–1–2

3.7�

10�

12

(2.1�

10�

14)

3.2

(0.0

2)

9.4

5�

10

11

(0–N

)4.3

50.0

37

7–8–9

0.7

44

2-S

NP

win

dow

1–2

0.0

39

1–2

2.8

(0.0

2)

15.2

(0.0

9)

6.3

4(1

.74–22.6

9)

8.3

40.0

04

2–3

0.5

55

3–4

0.8

03

4–5

0.4

75

5–6

0.5

72

6–7

0.2

09

1–1

1.3�

10�

11

(7.1�

10�

14)

3.2

(0.0

2)

2.8

2�

10

11

(0–N

)4.4

20.0

35

7–8

0.9

02

8–9

0.8

23

Afr

ican

Am

eric

an

3-S

NP

win

dow

1–2–3

0.8

01

2–3–4

0.5

35

3–4–5

0.0

06

2–2–1

10.6

(0.2

2)

1.0

(0.0

2)

0.0

8(0

.01–0.6

4)

5.3

00.0

21

4–5–6

0.0

04

1–2–2

12.0

(0.2

4)

2.1

(0.0

4)

0.1

5(0

.03–0.6

9)

4.4

20.0

35

2–1–2

7.0

(0.1

4)

0.0

(0.0

0)

06.2

50.0

12

5–6–7

0.4

06

6–7–8

0.5

85

7–8–9

0.2

79

2-S

NP

win

dow

1–2

0.5

58

2–3

0.4

34

3–4

0.6

23

4–5

0.0

06

2–1

10.8

(0.2

2)

1.1

(0.0

3)

0.0

8(0

.01–0.6

3)

5.7

20.0

17

5–6

0.3

95

6–7

0.6

93

7–8

0.2

94

8–9

0.6

61

Ab

bre

viation

s:C

I,co

nfid

en

cein

terv

al;

DRD

3,

D3

gen

e;

SN

P,

sin

gle

-nucl

eotid

ep

oly

morp

his

m.

Sp

eci

fic

nucl

eotid

eco

din

gfo

ralle

les

1or

2fo

reach

SN

Pca

nb

efo

un

din

Tab

le1.

Bold

ed

valu

es

ind

icate

aP-v

alu

eo

0.0

5.


210


Tab

le6

Slid

ing

win

do

wh

ap

loty

pe

an

aly

ses

acr

oss

DR

D3

for

qu

an

tita

tive

resp

on

sed

ata

inb

oth

Cau

casi

an

an

dA

fric

an

-Am

eri

can

sam

ple

s

BPRS

BPO

SBN

EG

Glo

bal

P-v

alu

eH

aplo

type

Mea

nch

ange

(N,

freq

uen

cy)

Haplo

type

P-v

alu

eG

lobal

P-v

alu

eH

aplo

type

Mea

nch

ange

(N,

freq

uen

cy)

Haplo

type

P-v

alu

eG

lobal

P-v

alu

eH

aplo

type

Mea

nch

ange

(N,

freq

uen

cy)

Haplo

type

P-v

alu

e

�0.1

1a

�0.0

9a

�0.0

2a

Cauca

sian

3-S

NP

win

dow

1–2–3

0.9

66

0.9

90

0.2

80

2–3–4

0.9

58

0.9

96

0.3

70

3–4–5

0.3

89

2–2–2

0.2

6(4

4.6

,0.2

5)

0.0

33

0.5

60

1.0

00

4–5–6

0.3

53

1.0

00

0.2

94

5–6–7

0.5

01

0.5

44

1.0

00

1–2–1

�0.0

1(2

5,

0.1

5)

0.0

44

6–7–8

0.7

14

1.0

00

0.3

14

7–8–9

0.7

86

0.8

16

0.4

74

2-S

NP

win

dow

1–2

0.9

19

0.9

98

0.2

21

2–3

0.9

12

0.9

95

0.7

03

3–4

0.7

31

0.9

45

0.3

88

4–5

0.1

67

0.2

98

0.3

18

5–6

0.2

35

0.3

22

0.1

82

6–7

0.9

33

0.9

86

0.2

43

1–2

�0.0

9(1

10,

0.6

4)

0.0

45

7–8

0.3

90

1.0

00

0.1

08

1–2

0.1

1(5

1,

0.3

0)

0.0

42

8–9

0.5

29

0.3

44

0.7

29

�0.1

5a

�0.2

4a

0.2

3a

Afr

ican

Am

eric

an

3-S

NP

win

dow

1–2–3

0.1

57

0.6

32

0.6

76

2–3–4

0.2

08

0.6

54

0.6

17

3–4–5

0.0

21

1–1–1

0.0

7(1

2.3

,0.2

)0.0

10

0.2

70

0.2

09

4–5–6

0.0

14

0.0

25

2–2–2

�0.5

8(4

,0.0

8)

0.0

27

o0.0

01

5–6–7

0.0

59

0.3

46

0.1

17

6–7–8

0.8

10

0.5

96

1.0

00

7–8–9

0.4

86

0.7

30

1.0

00

2-S

NP

win

dow

1–2

0.2

69

0.8

98

0.8

19

2–3

0.0

81

1–2

�0.1

9(4

5,

0.7

8)

0.0

29

0.0

87

0.4

97

3–4

0.1

43

1–1

�0.0

2(1

5,

0.2

4)

0.0

49

0.3

48

0.3

68

4–5

0.0

31

2–1

�0.4

3(5

.1,

0.0

8)

0.0

49

0.1

49

0.2

68

5–6

0.0

76

0.2

78

1.0

00

6–7

0.8

39

0.7

65

0.2

07

7–8

0.6

55

0.4

85

1.0

00

8–9

0.5

45

0.4

26

0.3

23

Ab

bre

viati

on

s:BN

EG

,BPRS

neg

ati

ve;

BPO

S,

BPRS

posi

tive

;BPRS,

Bri

ef

Psy

chia

tric

Rati

ng

Sca

le;

DRD

3,

D3

gen

e;

SN

P,

sin

gle

-nucl

eoti

de

poly

morp

his

m.

aN

um

bers

ind

icate

calc

ula

ted

valu

es

for

the

en

tire

Cauca

sian

or

Afr

ican

-Am

erica

nsa

mp

le.

Sp

eci

fic

nucl

eotid

eco

din

gfo

ralle

les

1or

2fo

reach

SN

Pca

nb

efo

un

din

Tab

le1.

Bold

ed

valu

es

ind

icate

aP-v

alu

eo

0.0

5.


211


association studies published at the time.57,59,60 The Jonssonet al. meta-analyses did not contain data from either Arranzet al.61 or Gaitonde et al.,58 most likely because those rawdata had not been published. In the present meta-analyses,these data have been included and were obtained directlyfrom the authors. Gaitonde et al. and Arranz et al. alsoprovided us with additional subjects who were added to

their respective samples subsequent to publication. Datafrom the study by Shaikh et al.,57 however, had to beexcluded from the current meta-analyses, as their data hadbeen incorporated into the Arranz et al. study already. CLZdata from three other samples were also included in thepresent meta-analyses: the two samples from the presentstudy and the sample from the study by Barlas et al.62

Odds ratio

0.044582 1 22.4306

Study % WeightOdds ratio(95% CI)

1.08 (0.49,2.37)Gaitonde96 8.5

0.72 (0.34,1.53)Malhotra98 9.1

0.17 (0.04,0.68)Scharfetter99 2.9

0.78 (0.53,1.14)Arranz00 34.1

0.94 (0.49,1.80)Barlas09 12.4

0.94 (0.61,1.44)Present (Cauc.) 27.0

0.82 (0.32,2.10)Present (Af. Am.) 6.0

0.82 (0.65,1.04), p = 0.100Overall (95% CI)

Figure 3 Meta-analysis of clozapine (CLZ) response studies comparing effects of the DRD3 Ser allele vs the Gly allele.

Odds ratio0.022192 1 45.0614

Study– % Weight Odds ratio(95% CI)

1.13 (0.39,3.30)Gaitonde96 9.6

0.48 (0.15,1.53)Malhotra98 8.2

0.12 (0.02,0.62)Scharfetter99 4.2

0.78 (0.47,1.28)Arranz00 35.1

0.87 (0.38,2.00)Barlas09 15.1

0.92 (0.50,1.69)Present (Cauc.) 25.9

0.44 (0.04,5.18)Present (Af. Am.) 1.9

0.75 (0.53,1.06), p = 0.108Overall (95% CI)

Figure 4 Meta-analysis of clozapine (CLZ) response studies comparing effects of the DRD3 Ser/Ser genotype vs Gly carrier genotypes

(Ser/GlyþGly/Gly).


212


In total, we increased the sample size from 233 in theJonsson et al. study to 758 in the present, comprehensivemeta-analyses (more than three times larger).

In contrast to Jonsson et al., we did not observe anassociation between the DRD3 Ser allele (P¼ 0.100), Ser/Ser(P¼0.108) or homozygote genotypes (P¼0.345) with pooroverall CLZ response in our meta-analyses. However, aninteresting feature observed by us was the consistency acrossthe individual studies: in six of the seven studies, the Serallele, Ser/Ser genotype and homozygote genotypes sharedthe same direction of effect (ORo1) towards poor response(with Gaitonde et al.58 being the exception). Although mostof the effect sizes are small, this feature indicates that wecannot rule out the possibility of a minor role of the Ser9Glyvariant in CLZ response on the basis of this negative result.Moreover, we should consider that the true magnitude ofgenetic effects is usually less than that described in theinitial reports of association, owing to the phenomenon ofthe ‘winner’s curse’ (discussed in Ioannidis et al.64 andLohmueller et al.65). Therefore, the initial positive findingsobserved in Jonsson et al. and the results in the presentmeta-analyses could be interpreted in either of the twofollowing ways: (1) the Jonsson et al. finding reflects a falsepositive that will continue to lose statistical significance asmore data are accumulated; or (2) there is indeed a trueassociation between DRD3 Ser9Gly and overall CLZ responsethat was exaggerated in Jonsson et al., but is finding a moreappropriate, smaller effect size in this larger study. If there isa real effect of DRD3 Ser9Gly on overall CLZ response, and itis in fact a small one, as would be expected for a complexphenotype, then the sample size may need to be increasedsignificantly to convincingly identify this variant as a

predictor of APD response. Newton-Cheh and Hirschhorn66

offered a sample size range of 1000–10 000 patients thatwould be needed to achieve significance (Po0.05) forcommon variants of modest effect.

In addition to insufficient sample size, our negativeresult in the meta-analyses, in spite of the consistencyobserved across individual studies, may also be explainedby an unfavourable signal-to-noise ratio. This may be thecase, especially considering the heterogeneous nature ofmeta-analytic studies. In other words, confounding vari-ables may have prevented the observed trend from becom-ing statistically significant. Possible factors that could haveincreased the ‘noise’ in our study include ethnicity, age,gender, duration of illness, co-medication, compliance,plasma levels, smoking, pharmacokinetic factors, and soon. In our meta-analyses, we were able to test but did notfind ethnicity to be a confounding factor. Unfortunately,because of differences in study design and availability of rawdata, we were unable to test and account for the otherpossible confounding variables mentioned.

Considering the possibility that the DRD3 Ser9Glyvariant has an effect on CLZ response, its possible modesof action are unknown. The Gly allele, however, in additionto being associated with increased binding affinity toDA- and D3-selective antagonists,40,41 has also been linkedto having more robust effects on various signal trans-duction systems.40,42 Signal transduction systems translatethe short-term interactions between neurotransmitters andreceptors into long-term adaptive changes in neural activity.This may explain why repeated short-term receptor block-ade can result in progressive effectiveness over several weeksof therapy. It is therefore important to consider any effects

Odds ratio

0.057725 1 17.3234

Study– % WeightOdds ratio(95% CI)

1.11 (0.39,3.19)Gaitonde96 8.0

0.50 (0.17,1.47)Malhotra98 7.7

0.28 (0.06,1.37)Scharfetter99 3.5

0.95 (0.58,1.55)Arranz00 36.4

0.83 (0.36,1.94)Barlas09 12.3

0.97 (0.54,1.74)Present (Cauc.) 25.8

0.95 (0.29,3.10)Present (Af. Am.) 6.3

0.87 (0.64,1.17), p = 0.345Overall (95% CI)

Figure 5 Meta-analysis of clozapine (CLZ) response studies comparing effects of DRD3 homozygote genotypes (Ser/SerþGly/Gly) vs theheterozygote genotype (Ser/Gly).


213


that variation in D3 may have on signal transductionfunction in order to understand its possible contributionto CLZ response, and to APD response in general.

One signal transduction system to consider is the cAMPpathway. This system is of particular relevance, as all DAreceptors are either positively or negatively linked to it.Increased cerebrospinal fluid cAMP levels have beenobserved in SCZ,67 whereas normalization of these levelshas been associated with AP treatment67,68 and response toAP treatment.69 The effect of D3 on this system hasbeen derived through in vitro experiments. Agonism at D3was shown to inhibit forskolin-stimulated cAMP produc-tion,40,42,70,71 whereas selective D3 antagonism blockedthis effect.70 In addition, the effect of DA on inhibitingforskolin-stimulated cAMP production was found to bealtered by DRD3 Ser9Gly variation.40,42

Prostaglandin signalling is another system to consider inview of the immune imbalance and inflammatory processhypothesis of SCZ (reviewed in Muller et al.72). To elaborate,elevated levels of PGE2, a signalling molecule, were asso-ciated with SCZ and symptom severity.73 In addition, abeneficial effect on treatment outcome of cyclooxygenase-2(COX-2) inhibitor add-on treatment to risperidone wasobserved.74 Inhibition of COX-2, an upstream regulator ofPGE2, is expected to reduce PGE2 levels. Further evidence fora possible role of PGE2 in SCZ and AP response is the anti-inflammatory modulatory effect of APs, including CLZ, oncytokines upstream (tumour necrosis factor-a) and down-stream (interleukin-2, -6, -10) of PGE2 (reviewed in Suginoet al.75 and Monteleone et al.76). These cytokines havealso been associated with SCZ (reviewed in Muller et al.72).Of interest, D3 has been linked to PGE2 signalling throughexperiments in vitro. It was shown that D3 mediates DAand D3 antagonist inhibition of stimulus-evoked releaseof PGE2

42 and arachidonic acid.77 Furthermore, Hellstrandet al.42 showed that the effect of DA and D3 antagonists oninhibition of PGE2 release only occurs in cells transfectedwith the Gly variant, not the Ser variant. This may be whythe Gly allele, in particular, was observed more frequently,although non-significantly, in patients who respondedmore favourably to CLZ treatment in the meta-analysesconducted in this study.

Another pathway that D3 may affect is the mitogen-activated protein kinase (MAPK) pathway. This system isimplicated in the pathophysiology of SCZ because post-mortem studies have found MAPK protein and mRNAlevels to be increased in SCZ brains.78 Furthermore, variousAPs, both typical and atypical, have been shown to alterthe MAPK pathway (reviewed in Browning et al.79). Particu-larly relevant to this study, CLZ selectively activatedthe MAPK/extracellular signal-regulated kinase (a specificMAPK pathway) pathway in vitro,79,80 whereas inhibition ofthis pathway reversed the actions of CLZ in the conditionedavoidance response mouse model of psychosis,79 providingstronger evidence that this pathway may have a role inthe AP effect of CLZ. It has been shown in vitro that MAPKactivation by DA is a D3-mediated response.70,81 More-over, DRD3 Ser9Gly variation has previously been shown

to have differential effects on this response, with the Glyallele having a more prolonged response than the Serallele.40

Exactly what effect DRD3 Ser9Gly variation might haveon these pathways or in ligand binding to D3, however, isunclear. Although it has been suggested that this poly-morphism may alter transduction properties by affectingposttranslational glycosylation,42 Jeanneteau et al.40 pro-vided evidence to the contrary. This polymorphism, locatedin the extracellular tail of the receptor, does not lie inregions of the receptor protein that would usually beexpected to be involved in signal transduction or ligandbinding. However, Jeanneteau et al.40 proposed that theSer9Gly variant may affect these properties by changing theoverall conformation of the receptor and influencingthe ability of D3 to form oligomers. Indeed, there is abun-dant evidence that the formation of oligomers is importantfor the signalling functions of G-protein-coupled receptors(reviewed in Bouvier82). Furthermore, it has been shownin vivo that D3 forms dimers83 and that hetero-oligomeriza-tion of wild-type D3 with mutant receptors alters ligand-binding affinity.84 However, as the functional effects of thisSNP have been mostly obtained in vitro, it is difficult tospeculate on its possible mode of action in vivo. Anotherquestion that remains is whether and how the possiblefunctional effects that this SNP has on signal transductionpathways may translate into CLZ/AP treatment outcomeeffect. One possibility may be through downregulation ofD3 levels in the ventral striatum as previously described.11,12

A further examination of these issues is warranted.Another item to consider is why the effect of the Ser9Gly

variant in Gaitonde et al. was again, although non-significantly, in the opposite direction to the other sixstudies included in the meta-analyses. Their results wereunlikely to be because of population differences, as thesample consisted of Caucasian patients (as did five of the sixother studies) and was recruited from the United Kingdom(as were two other studies) (see Table 2). The discrepanciesobserved in Gaitonde et al. were also unlikely to be becauseof the response criteria used in that study. Although thecriteria used were not reported, the response rate of 53% waswell within the range of most of the other studies in themeta-analyses (see Table 7). Alternative explanationsinvolving these particular confounding variables are there-fore unlikely, although other factors are still possible, asdiscussed earlier. If the Ser9Gly variant is not itself a truecausal variant, then it may be in incomplete LD with adifferent variant of true causal effect instead. If this is thecase, then it may be worthwhile to examine other SNPs thatare in tight LD with Ser9Gly. These SNPs should also beanalysed in combination as haplotypes, as this approachmay be more revealing than considering each SNP indivi-dually.85 Although the same haplotypes were not tested inour own samples in this study, previous association studieshave found specific haplotypes, containing Ser9Gly andtightly linked SNPs from the 50 region,86,87 as well as fromthe 30 region88 of the gene, to be associated with SCZ, inspite of the Ser9Gly variant not being significant by itself.


214


In this study, although the Ser9Gly SNP was negative,haplotype analyses showed some haplotypes that wereassociated with CLZ response in both the Caucasian andAfrican-American samples. Positive reports on haplotypesspanning non-coding regions of the gene indicate that theSer9Gly variant may be linked to elements that regulatetranscription. These results indicate that in addition to apossible effect on signal transduction or ligand binding, oneof the functional consequences of DRD3 variation may bealtered levels of gene expression.

In addition to the Ser9Gly meta-analyses, we alsoexamined nine tag SNPs (including Ser9Gly) spanning theDRD3 region from 50-UTR to 30-UTR in our own samples.The purpose was not to study closely linked variants toSer9Gly, but rather to explore other regions of the gene. Wefound SNP 8 (rs2134655, intron 5) to be associated withpositive symptoms in Caucasians, as well as to have apossible trend in African Americans. In turns out, in fact,that this SNP was observed to be in strong LD with Ser9Gly(SNP 6) in Caucasians and moderate LD in African Amer-icans. Therefore, rs2134655 could be contributing to or evenresponsible for the positive findings of Ser9Gly in previousstudies, although it is not clear what function rs2134655might have. In any case, rs2134655 could be a marker forconsideration in future DRD3 studies.

Another finding of interest is the association, beforecorrection for multiple testing, of rs1394016 with negativesymptoms in African Americans. Although not observedin Caucasians, it should be noted that the most promisingindication of D3 receptor blockade may be the ameliorationof negative symptoms (and cognitive symptoms) (reviewedin Joyce and Millan89). Overall, our findings are in linewith previous studies that have implicated D3 antagonismnot only for general APD effect but also specifically for themechanism of action of CLZ, and its resulting uniquetherapeutic profile (that is, an effect on negative symptoms(reviewed in Naheed and Green90; see Introduction)).

An issue relating to our exploratory analysis is correctionfor multiple testing. As we explored the effect of nine SNPsacross DRD3, an adjusted P-value threshold for statisticalsignificance was calculated using the SNPSpD programme(Brisbane, Queensland, Australia).91 This programme correctsfor testing of multiple markers by generating an a-thresholdthat takes into account the LD between the SNPs studied.The corrected significance thresholds were P-value of 0.007in the Caucasian sample and P-value of 0.006 in the African-American sample. When applying these thresholds to ourfindings, the rs2134655 result in the Caucasian sample stillremains significant, whereas results in the African-Americansample become negative. As the number of SNPs in ourstudy is not large, to avoid the risk of false negatives, wechose to highlight our negative trending results. We willleave it to the discretion of others whether or not toexamine these SNPs in independent samples in order toexplore their potential significance.

With regard to statistical power, on the basis of the lowestminor allele frequencies (MAFs) (see Tables 1 and 2), ourCaucasian sample (lowest MAFE20%; 167 non-responders;T

ab

le7

DR

D3

Ser9

Gly

data

fro

meach

stu

dy

incl

ud

ed

inm

eta

-an

aly

ses

Stu

dy

Res

ponse

statu

sN

(%of

study

sam

ple

)G

enoty

pe

freq

uen

cies

Ser

vsG

lySer

/Ser

vsG

lyca

rrie

rsH

om

ozy

gote

svs

het

erozy

gote

s

S/S

S/G

G/G

SG

S/S

S/G

+G

/GS/S

+G

/GS/G

Gait

on

de

etal.

58

R30

(53)

12

17

141

19

12

18

13

17

NR

27

(47)

10

16

136

18

10

17

11

16

Malh

otr

aet

al.

59

R19

(28)

511

321

17

514

811

NR

49

(72)

21

20

862

36

21

28

29

20

Sch

arf

ett

er

etal.

60

R21

(66)

512

422

20

516

912

NR

11

(34)

83

019

38

38

3A

rran

zet

al.

61

R189

(67)

76

94

19

246

132

76

113

95

94

NR

95

(33)

44

46

5134

56

44

51

49

46

Barl

as

etal.

62

R47

(53)

23

20

466

28

23

24

27

20

NR

42

(47)

22

16

460

24

22

20

26

16

Pre

sen

tsa

mp

le(C

auca

sian

)R

95

(53)

34

50

11

118

72

34

61

45

50

NR

85

(47)

32

44

9108

62

32

53

41

44

Pre

sen

tsa

mp

le(A

fric

an

Am

eri

can

)R

25

(52)

19

15

11

39

124

16

9N

R23

(48)

28

13

12

34

221

15

8

Ab

bre

viati

on

s:D

RD

3,

D3

gen

e;

NR,

non

-resp

on

ders

;R,

resp

on

ders

;Ser9

Gly

,se

rin

e-t

o-g

lyci

ne

sub

stit

uti

on

at

am

ino

aci

dp

osi

tion

9;

SN

P,

sin

gle

-nucl

eoti

de

poly

morp

his

m.


215


1.34 responder/non-responder ratio) had at least 80% powerto detect an effect size of as high as 1.65/0.61 at a¼0.05.Using the same criteria, our African-American sample(49 non-responders; 1.04 responder/non-responder ratio)has at least 80% power to detect an effect size as high as2.8/0.36.92

In spite of studies showing questionable efficacy ofD3-preferring ligands in clinical trials,93,94 the possibilitythat D3 may still be important for modulating AP responseremains. In this study, we provide evidence suggestingthat a minor role for DRD3 variation in CLZ response cannotbe dismissed. Additional studies on Ser9Gly need to beperformed in order to increase the collective sample size inthe literature so that the possible role of this variant can bemore clearly defined. Furthermore, functional studies onSNPs in regulatory elements in LD with the Ser9Glypolymorphism should be examined.

Conflict of interest

JLK has been a consultant for GlaxoSmithKline, Johnsonand Johnson and for TheraGenetics. JLK and RH are co-authors on a patent application for genetic prediction ofantipsychotic response.

Acknowledgments

RH was supported by the Canadian Institute of the Health Research(CIHR) Molecular Medicine Training Program, JLK was supportedby the CIHR and GlaxoSmithKline, and AM was supported bythe Instituto de Salud Carlos III, Centro de Investigacion en Red deSalud Mental, CIBERSAM.

References

1 Malhotra AK, Murphy Jr GM, Kennedy JL. Pharmacogenetics ofpsychotropic drug response. Am J Psychiatry 2004; 161: 780–796.

2 Lieberman JA, Stroup TS, McEvoy JP, Swartz MS, Rosenheck RA, PerkinsDO et al. Effectiveness of antipsychotic drugs in patients with chronicschizophrenia. N Engl J Med 2005; 353: 1209–1223.

3 Mata I, Madoz V, Arranz MJ, Sham P, Murray RM. Olanzapine:concordant response in monozygotic twins with schizophrenia. Br JPsychiatry 2001; 178: 86.

4 Vojvoda D, Grimmell K, Sernyak M, Mazure CM. Monozygotic twinsconcordant for response to clozapine. Lancet 1996; 347: 61.

5 Kapur S, Mamo D. Half a century of antipsychotics and still a central rolefor dopamine D2 receptors. Prog Neuropsychopharmacol Biol Psychiatry2003; 27: 1081–1090.

6 Hwang R, Shinkai T, De Luca V, Muller DJ, Ni X, Macciardi F et al.Association study of 12 polymorphisms spanning the dopamine D(2)receptor gene and clozapine treatment response in two treatmentrefractory/intolerant populations. Psychopharmacology (Berl) 2005; 181:179–187.

7 Nordstrom AL, Farde L, Nyberg S, Karlsson P, Halldin C, Sedvall G. D1,D2, and 5-HT2 receptor occupancy in relation to clozapine serumconcentration: a PET study of schizophrenic patients. Am J Psychiatry1995; 152: 1444–1449.

8 Meltzer HY. The mechanism of action of novel antipsychotic drugs.Schizophr Bull 1991; 17: 263–287.

9 Sokoloff P, Diaz J, Le Foll B, Guillin O, Leriche L, Bezard E et al. Thedopamine D3 receptor: a therapeutic target for the treatment ofneuropsychiatric disorders. CNS Neurol Disord Drug Targets 2006; 5:25–43.

10 Gurevich EV, Joyce JN. Distribution of dopamine D3 receptor expressingneurons in the human forebrain: comparison with D2 receptorexpressing neurons. Neuropsychopharmacology 1999; 20: 60–80.

11 Gurevich EV, Bordelon Y, Shapiro RM, Arnold SE, Gur RE, Joyce JN.Mesolimbic dopamine D3 receptors and use of antipsychotics inpatients with schizophrenia. A postmortem study. Arch Gen Psychiatry1997; 54: 225–232.

12 Joyce JN, Gurevich EV. D3 receptors and the actions of neuroleptics inthe ventral striatopallidal system of schizophrenics. Ann N Y Acad Sci1999; 877: 595–613.

13 Amalric M, Koob GF. Functionally selective neurochemical afferents andefferents of the mesocorticolimbic and nigrostriatal dopamine system.Prog Brain Res 1993; 99: 209–226.

14 Csernansky JG, Murphy GM, Faustman WO. Limbic/mesolimbicconnections and the pathogenesis of schizophrenia. Biol Psychiatry1991; 30: 383–400.

15 Mogenson GJ, Yang CR, Yim CY. Influence of dopamine on limbicinputs to the nucleus accumbens. Ann N Y Acad Sci 1988; 537:86–100.

16 Abi-Dargham A, Laruelle M. Mechanisms of action of second generationantipsychotic drugs in schizophrenia: insights from brain imagingstudies. Eur Psychiatry 2005; 20: 15–27.

17 Damask SP, Bovenkerk KA, de la Pena G, Hoversten KM, Peters DB,Valentine AM et al. Differential effects of clozapine and haloperidolon dopamine receptor mRNA expression in rat striatum and cortex.Brain Res Mol Brain Res 1996; 41: 241–249.

18 Joyce JN, Lexow N, Bird E, Winokur A. Organization of dopamineD1 and D2 receptors in human striatum: receptor autoradiographicstudies in Huntington’s disease and schizophrenia. Synapse 1988; 2:546–557.

19 Joyce JN, Goldsmith SG, Gurevich EV. Limbic circuits and monoaminereceptors: dissecting the effects of antipsychotics from diseaseprocesses. J Psychiatr Res 1997; 31: 197–217.

20 Gyertyan I, Saghy K. Effects of dopamine D3 receptor antagonists onspontaneous and agonist-reduced motor activity in NMRI mice andWistar rats: comparative study with nafadotride, U 99194A and SB277011. Behav Pharmacol 2004; 15: 253–262.

21 Millan MJ, Dekeyne A, Rivet JM, Dubuffet T, Lavielle G, Brocco M.S33084, a novel, potent, selective, and competitive antagonist atdopamine D(3)-receptors: II. Functional and behavioral profile com-pared with GR218,231 and L741,626. J Pharmacol Exp Ther 2000; 293:1063–1073.

22 Reavill C, Taylor SG, Wood MD, Ashmeade T, Austin NE, Avenell KYet al. Pharmacological actions of a novel, high-affinity, and selectivehuman dopamine D(3) receptor antagonist, SB-277011-A. J PharmacolExp Ther 2000; 294: 1154–1165.

23 Park WK, Jeong D, Cho H, Lee SJ, Cha MY, Pae AN et al. KKHA-761, apotent D3 receptor antagonist with high 5-HT1A receptor affinity,exhibits antipsychotic properties in animal models of schizophrenia.Pharmacol Biochem Behav 2005; 82: 361–372.

24 Park WK, Jeong D, Yun CW, Lee S, Cho H, Kim GD et al.Pharmacological actions of a novel and selective dopamine D3 receptorantagonist, KCH-1110. Pharmacol Res 2003; 48: 615–622.

25 Swain SN, Beuk J, Heidbreder CA, Beninger RJ. Role of dopamine D3receptors in the expression of conditioned fear in rats. Eur J Pharmacol2008; 579: 167–176.

26 Zhang M, Ballard ME, Kohlhaas KL, Browman KE, Jongen-Relo AL,Unger LV et al. Effect of dopamine D3 antagonists on PPI in DBA/2J miceor PPI deficit induced by neonatal ventral hippocampal lesions in rats.Neuropsychopharmacology 2006; 31: 1382–1392.

27 Brocco M, Gobert A, Dekeyne A, Lavielle G, Millan M. S33138, a noveland preferential antagonist at dopamine D3 receptors: functional profilein rodent models of the control of negative-cognitive symptoms ofschizophrenia. Int J Neuropsychopharmacol 2004; 7S2: S265.

28 Rodriguez-Arias M, Felip CM, Broseta I, Minarro J. The dopamine D3antagonist U-99194A maleate increases social behaviors of isolation-induced aggressive male mice. Psychopharmacology (Berl) 1999; 144:90–94.

29 Glickstein SB, Desteno DA, Hof PR, Schmauss C. Mice lacking dopamineD2 and D3 receptors exhibit differential activation of prefrontalcortical neurons during tasks requiring attention. Cereb Cortex 2005;15: 1016–1024.


216


30 Laszy J, Laszlovszky I, Gyertyan I. Dopamine D3 receptor antagonistsimprove the learning performance in memory-impaired rats. Psycho-pharmacology (Berl) 2005; 179: 567–575.

31 Audinot V, Newman-Tancredi A, Gobert A, Rivet JM, Brocco M,Lejeune F et al. A comparative in vitro and in vivo pharmacologicalcharacterization of the novel dopamine D3 receptor antagonists (+)-S14297, nafadotride, GR 103,691 and U 99194. J Pharmacol Exp Ther1998; 287: 187–197.

32 Gyertyan I, Saghy K. The selective dopamine D3 receptor antagonists,SB 277011-A and S 33084 block haloperidol-induced catalepsy in rats.Eur J Pharmacol 2007; 572: 171–174.

33 Guo N, Klitenick MA, Tham CS, Fibiger HC. Receptor mechanismsmediating clozapine-induced c-fos expression in the forebrain. Neuro-science 1995; 65: 747–756.

34 Robertson GS, Fibiger HC. Neuroleptics increase c-fos expression in theforebrain: contrasting effects of haloperidol and clozapine. Neuroscience1992; 46: 315–328.

35 Robertson GS, Lee CJ, Sridhar K, Nakabeppu Y, Cheng M, Wang YMet al. Clozapine-, but not haloperidol-, induced increases in deltaFosB-like immunoreactivity are completely blocked in the striatum of micelacking D3 dopamine receptors. Eur J Neurosci 2004; 20: 3189–3194.

36 Southam E, Lloyd A, Jennings CA, Cluderay JE, Cilia J, Gartlon JE et al.Effect of the selective dopamine D3 receptor antagonist SB-277011-Aon regional c-Fos-like expression in rat forebrain. Brain Res 2007; 1149:50–57.

37 Ashby Jr CR, Minabe Y, Stemp G, Hagan JJ, Middlemiss DN. Acute andchronic administration of the selective D(3) receptor antagonistSB-277011-A alters activity of midbrain dopamine neurons in rats: anin vivo electrophysiological study. J Pharmacol Exp Ther 2000; 294:1166–1174.

38 Chiodo LA, Bunney BS. Typical and atypical neuroleptics: differentialeffects of chronic administration on the activity of A9 and A10 midbraindopaminergic neurons. J Neurosci 1983; 3: 1607–1619.

39 Macdonald GJ, Branch CL, Hadley MS, Johnson CN, Nash DJ, Smith ABet al. Design and synthesis of trans-3-(2-(4-((3-(3-(5-methyl-1,2,4-oxadiazolyl))- phenyl)carboxamido)cyclohexyl)ethyl)-7-methylsulfonyl-2,3,4,5-tetrahydro-1 H-3-benzazepine (SB-414796): a potent andselective dopamine D3 receptor antagonist. J Med Chem 2003; 46:4952–4964.

40 Jeanneteau F, Funalot B, Jankovic J, Deng H, Lagarde JP, Lucotte G et al.A functional variant of the dopamine D3 receptor is associated with riskand age-at-onset of essential tremor. Proc Natl Acad Sci USA 2006; 103:10753–10758.

41 Lundstrom K, Turpin MP. Proposed schizophrenia-related gene poly-morphism: expression of the Ser9Gly mutant human dopamine D3receptor with the Semliki Forest virus system. Biochem Biophys ResCommun 1996; 225: 1068–1072.

42 Hellstrand M, Danielsen EA, Steen VM, Ekman A, Eriksson E, Nilsson CL.The ser9gly SNP in the dopamine D3 receptor causes a shift from cAMPrelated to PGE2 related signal transduction mechanisms in transfectedCHO cells. J Med Genet 2004; 41: 867–871.

43 Dubertret C, Gorwood P, Ades J, Feingold J, Schwartz JC, Sokoloff P.Meta-analysis of DRD3 gene and schizophrenia: ethnic heterogeneityand significant association in Caucasians. Am J Med Genet 1998; 81:318–322.

44 Jonsson EG, Flyckt L, Burgert E, Crocq MA, Forslund K, Mattila-Evenden Met al. Dopamine D3 receptor gene Ser9Gly variant and schizo-phrenia: association study and meta-analysis. Psychiatr Genet 2003;13: 1–12.

45 Williams J, Spurlock G, Holmans P, Mant R, Murphy K, Jones L et al.A meta-analysis and transmission disequilibrium study of associationbetween the dopamine D3 receptor gene and schizophrenia. MolPsychiatry 1998; 3: 141–149.

46 Jonsson EG, Kaiser R, Brockmoller J, Nimgaonkar VL, Crocq MA. Meta-analysis of the dopamine D3 receptor gene (DRD3) Ser9Gly variant andschizophrenia. Psychiatr Genet 2004; 14: 9–12.

47 Ma G, He Z, Fang W, Tang W, Huang K, Li Z et al. The Ser9Glypolymorphism of the dopamine D3 receptor gene and risk ofschizophrenia: an association study and a large meta-analysis. SchizophrRes 2008; 101: 26–35.

48 Utsunomiya K, Shinkai T, De Luca V, Hwang R, Sakata S, Fukunaka Yet al. Genetic association between the dopamine D3 gene polymorphism

(Ser9Gly) and schizophrenia in Japanese populations: evidence froma case–control study and meta-analysis. Neurosci Lett 2008; 444:161–165.

49 APA. Diagnostic and Statistical Manual of Mental Disorders (DSM-IV) 4thedn. American Psychiatric Association: Washington, DC, 1994.

50 Kane J, Honigfeld G, Singer J, Meltzer H. Clozapine for the treatment-resistant schizophrenic. A double-blind comparison with chlorproma-zine. Arch Gen Psychiatry 1988; 45: 789–796.

51 Lahiri DK, Nurnberger Jr JI. A rapid non-enzymatic method for thepreparation of HMW DNA from blood for RFLP studies. Nucleic Acids Res1991; 19: 5444.

52 International HapMap Consortium. The International Hapmap Project.Nature 2003; 426: 789–796.

53 SPSS (2007). Statistical Package for the Social Sciences for Windows,release 15.0.1.1. SPSS Inc.: Chicago.

54 Barrett JC, Fry B, Maller J, Daly MJ. Haploview: analysis and visualizationof LD and haplotype maps. Bioinformatics 2005; 21: 263–265.

55 Dudbridge F. Pedigree disequilibrium tests for multilocus haplotypes.Genet Epidemiol 2003; 25: 115–121.

56 Bland JM, Altman DG. Statistics notes. The odds ratio. BMJ 2000;320: 1468.

57 Shaikh S, Collier DA, Sham PC, Ball D, Aitchison K, Vallada H et al. Allelicassociation between a Ser-9-Gly polymorphism in the dopamine D3receptor gene and schizophrenia. Hum Genet 1996; 97: 714–719.

58 Gaitonde EJ, Morris A, Sivagnanasundaram S, McKenna PJ, Hunt DM,Mollon JD. Assessment of association of D3 dopamine receptor MscIpolymorphism with schizophrenia: analysis of symptom ratings, familyhistory, age at onset, and movement disorders. Am J Med Genet 1996;67: 455–458.

59 Malhotra AK, Goldman D, Buchanan RW, Rooney W, Clifton A,Kosmidis MH et al. The dopamine D3 receptor (DRD3) Ser9Glypolymorphism and schizophrenia: a haplotype relative risk study andassociation with clozapine response. Mol Psychiatry 1998; 3: 72–75.

60 Scharfetter J, Chaudhry HR, Hornik K, Fuchs K, Sieghart W, Kasper Set al. Dopamine D3 receptor gene polymorphism and response toclozapine in schizophrenic Pakistani patients. Eur Neuropsychopharmacol1999; 10: 17–20.

61 Arranz MJ, Munro J, Birkett J, Bolonna A, Mancama D, Sodhi M et al.Pharmacogenetic prediction of clozapine response. Lancet 2000; 355:1615–1616.

62 Barlas IO, Cetin M, Erdal ME, Semiz UB, Basoglu C, Ay ME et al. Lackof association between DRD3 gene polymorphism and response toclozapine in Turkish schizoprenia patients. Am J Med Genet B Neuro-psychiatr Genet 2009; 150B: 56–60.

63 StataCorp (2003). Stata Statistical Software, release 8. StataCorp LP:College Station, TX.

64 Ioannidis JP, Ntzani EE, Trikalinos TA, Contopoulos-Ioannidis DG.Replication validity of genetic association studies. Nat Genet 2001; 29:306–309.

65 Lohmueller KE, Pearce CL, Pike M, Lander ES, Hirschhorn JN. Meta-analysis of genetic association studies supports a contribution ofcommon variants to susceptibility to common disease. Nat Genet2003; 33: 177–182.

66 Newton-Cheh C, Hirschhorn JN. Genetic association studies of complextraits: design and analysis issues. Mutat Res 2005; 573: 54–69.

67 Bowers Jr MB, Study RE. Cerebrospinal fluid cyclic AMP and acidmonoamine metabolites following probenecid: studies in psychiatricpatients. Psychopharmacology (Berl) 1979; 62: 17–22.

68 Hatta Y, Hatta S, Saito T. Effects of ceruletide on the dopamine receptor-adenylate cyclase system in striatum and frontal cortex of ratschronically treated with haloperidol. Psychopharmacology (Berl) 1993;110: 383–389.

69 Biederman J, Rimon R, Ebstein R, Zohar J, Belmaker R. Neurolepticsreduce spinal fluid cyclic AMP in schizophrenic patient. Neuropsycho-biology 1976; 2: 324–327.

70 Millan MJ, Mannoury la Cour C, Novi F, Maggio R, Audinot V, Newman-Tancredi A et al. S33138 [N-[4-[2-[(3aS,9bR)-8-cyano-1,3a,4,9b-tetra-hydro[1]-benzopyrano[3,4-c]pyrr ol-2(3H)-yl)-ethyl]phenylacetamide],a preferential dopamine D3 versus D2 receptor antagonist and potentialantipsychotic agent: I. Receptor-binding profile and functional actionsat G-protein-coupled receptors. J Pharmacol Exp Ther 2008; 324:587–599.


217


71 Nishi A, Snyder GL, Greengard P. Bidirectional regulation of DARPP-32phosphorylation by dopamine. J Neurosci 1997; 17: 8147–8155.

72 Muller N, Strassnig M, Schwarz MJ, Ulmschneider M, Riedel M. COX-2inhibitors as adjunctive therapy in schizophrenia. Expert Opin InvestigDrugs 2004; 13: 1033–1044.

73 Kaiya H, Uematsu M, Ofuji M, Nishida A, Takeuchi K, Nozaki M et al.Elevated plasma prostaglandin E2 levels in schizophrenia. J NeuralTransm 1989; 77: 39–46.

74 Muller N, Riedel M, Scheppach C, Brandstatter B, Sokullu S, Krampe Ket al. Beneficial antipsychotic effects of celecoxib add-on therapycompared to risperidone alone in schizophrenia. Am J Psychiatry 2002;159: 1029–1034.

75 Sugino H, Futamura T, Mitsumoto Y, Maeda K, Marunaka Y. Atypicalantipsychotics suppress production of proinflammatory cytokines andup-regulate interleukin-10 in lipopolysaccharide-treated mice. ProgNeuropsychopharmacol Biol Psychiatry 2009; 33: 303–307.

76 Monteleone P, Fabrazzo M, Tortorella A, Maj M. Plasma levels ofinterleukin-6 and tumor necrosis factor alpha in chronic schizophrenia:effects of clozapine treatment. Psychiatry Res 1997; 71: 11–17.

77 Nilsson CL, Hellstrand M, Ekman A, Eriksson E. Both dopamine and theputative dopamine D3 receptor antagonist PNU-99194A induce abiphasic inhibition of phorbol ester-stimulated arachidonic acid releasefrom CHO cells transfected with the dopamine D3 receptor. Life Sci1999; 64: 939–951.

78 Kyosseva SV. The role of the extracellular signal-regulated kinase pathwayin cerebellar abnormalities in schizophrenia. Cerebellum 2004; 3: 94–99.

79 Browning JL, Patel T, Brandt PC, Young KA, Holcomb LA, Hicks PB.Clozapine and the mitogen-activated protein kinase signal transductionpathway: implications for antipsychotic actions. Biol Psychiatry 2005;57: 617–623.

80 Lu XH, Dwyer DS. Second-generation antipsychotic drugs, olanzapine,quetiapine, and clozapine enhance neurite outgrowth in PC12 cells viaPI3K/AKT, ERK, and pertussis toxin-sensitive pathways. J Mol Neurosci2005; 27: 43–64.

81 Beom S, Cheong D, Torres G, Caron MG, Kim KM. Comparative studiesof molecular mechanisms of dopamine D2 and D3 receptors for theactivation of extracellular signal-regulated kinase. J Biol Chem 2004;279: 28304–28314.

82 Bouvier M. Oligomerization of G-protein-coupled transmitter receptors.Nat Rev Neurosci 2001; 2: 274–286.

83 Nimchinsky EA, Hof PR, Janssen WG, Morrison JH, Schmauss C.Expression of dopamine D3 receptor dimers and tetramers in brainand in transfected cells. J Biol Chem 1997; 272: 29229–29237.

84 Lee SP, Xie Z, Varghese G, Nguyen T, O’Dowd BF, George SR.Oligomerization of dopamine and serotonin receptors. Neuropsycho-pharmacology 2000; 23(4 Suppl): S32–S40.

85 Liu N, Zhang K, Zhao H. Haplotype-association analysis. Adv Genet2008; 60: 335–405.

86 Ishiguro H, Okuyama Y, Toru M, Arinami T. Mutation and associationanalysis of the 50 region of the dopamine D3 receptor gene inschizophrenia patients: identification of the Ala38Thr polymorphismand suggested association between DRD3 haplotypes and schizo-phrenia. Mol Psychiatry 2000; 5: 433–438.

87 Sivagnanasundaram S, Morris AG, Gaitonde EJ, McKenna PJ, Mollon JD,Hunt DM. A cluster of single nucleotide polymorphisms in the 50-leaderof the human dopamine D3 receptor gene (DRD3) and its relationshipto schizophrenia. Neurosci Lett 2000; 279: 13–16.

88 Talkowski ME, Mansour H, Chowdari KV, Wood J, Butler A, Varma PGet al. Novel, replicated associations between dopamine D3 receptorgene polymorphisms and schizophrenia in two independent samples.Biol Psychiatry 2006; 60: 570–577.

89 Joyce JN, Millan MJ. Dopamine D3 receptor antagonists as therapeuticagents. Drug Discov Today 2005; 10: 917–925.

90 Naheed M, Green B. Focus on clozapine. Curr Med Res Opin 2001; 17:223–229.

91 Nyholt DR. A simple correction for multiple testing for single-nucleotidepolymorphisms in linkage disequilibrium with each other. Am J HumGenet 2004; 74: 765–769.

92 Gauderman WJ. Sample size requirements for matched case–controlstudies of gene–environment interaction. Stat Med 2002; 21: 35–50.

93 Lahti AC, Weiler M, Carlsson A, Tamminga CA. Effects of the D3 andautoreceptor-preferring dopamine antagonist (+)-UH232 in schizo-phrenia. J Neural Transm 1998; 105: 719–734.

94 Lecrubier Y. A partial D3 receptor agonist in schizophrenia. EurNeuropsychopharmacol 2003; 13(S4): S167–S168.


218


effect of dopamine d3 receptor gene polymorphisms and clozapine treatment response: exploratory...

Documents