Association of the dopamine b-hydroxylase 19 bp insertion/deletionpolymorphism with positive symptoms but not tardive dyskinesiain schizophrenia

Na Zhou1,2,†, Qiong Yu3,†, Xiaokun Li1*, Yaqin Yu3*, Changgui Kou3, Wenjun Li3, Hongqin Xu3, XingguangLuo4, Lingjun Zuo4, Thomas R. Kosten5 and Xiang Yang Zhang5*1School of Basic Medicine, Jilin University, Changchun, China2School of Stomatology, Jilin University, Changchun, China3Center of Medical Genomics, School of Public Health, Jilin University, Changchun, China4Department of Psychiatry, Yale University School of Medicine, New Haven, CT, USA5Menninger Department of Psychiatry and Behavioral Sciences, Baylor College of Medicine, and Michael E. DeBakey VA Medical Center,Houston, Texas, USA

Objective Overactivity of dopaminergic neurotransmission is a putative mechanism of tardive dyskinesia (TD). Dopamine beta-hydroxy-lase (DBH) is a key enzyme in the conversion of dopamine to norepinephrine, and plasma DBH activity is altered in TD patients. This studyexamined whether the functional DBH 5’-Ins/Del polymorphism was associated with TD severity in Chinese patients with schizophrenia.Methods We compared the rate of this polymorphism in patients with (n= 312) and without TD (n= 435), and healthy controls (n= 625).The severity of TD was assessed using the Abnormal Involuntary Movement Scale (AIMS) and psychopathology using the Positive and Neg-ative Syndrome Scale (PANSS).Results There were no significant differences in the distribution of the allele and genotype frequencies between the patients and controls, orbetween the patients with and without TD. Also, there was no significant difference in the AIMS total score between the three genotypegroups. However, the PANSS positive symptom subscore was significantly higher in patients with Del/Del genotype (13.2� 5.2) than thosewith Ins/Del (11.2� 4.9) and Ins/Ins (11.1� 3.1) genotypes (both p< 0.05).Conclusion These results suggest that although the DBH 5’-Ins/Del polymorphism was not associated with susceptibility to TD in patientswith schizophrenia, it might be related to positive symptoms of schizophrenia. Copyright © 2013 John Wiley & Sons, Ltd.

key words—schizophrenia; tardive dyskinesia; dopamine beta-hydroxylase; polymorphism; association

INTRODUCTION

Tardive dyskinesia (TD) is a potentially irreversibleinvoluntary movement that develops in approximately25–40% of schizophrenia patients on long-term typicalantipsychotic medication (Correll and Malhotra, 2004;Correll and Schenk, 2008; Jeste, 2004; Kane, 2004;Margolese et al., 2005; Remington et al., 2007; Tarsyand Baldessarini, 2006). TD is associated with poor qual-ity of life, non-adherencewithmedications, and increased

medical morbidity and mortality, but it has no well-accepted treatments, and its pathophysiology remainspoorly understood (Youssef and Waddington, 1987;Browne et al., 1996; Soares and McGrath, 1999;Ballesteros et al., 2000; Marsalek, 2000; Gerlach, 2002;Lohr et al., 2003; Remington et al., 2007; Zai et al.,2010). Several risk factors have been identified fordevelopment of TD, including genetic predisposition toTD (Rosengarten et al., 1994; Glazer, 2000; Lereret al., 2002; Pae et al., 2004; Patterson et al., 2005; Pae,2008). For example, concordance for the presence orabsence of TD among first-degree relatives (Mulleret al., 2001) and two recent meta-analyses showingthat polymorphisms in the genes coding for the dopamineD3 receptor, catechol-O-methyltransferase (COMT),dopamine D2 receptor, and manganese superoxidedismutase appear to make modest but significantcontributions to the overall variability in TD risk

*Correspondence to: X. Li, School of Basic Medicine, Jilin University,Changchun 130021, China. Tel: +86-431-85619437 E-mail: yuyq@jlu.edu.cn; Y. Yu, Department of Epidemiology and Statistics, Center of MedicalGenomics School of Public Health, Jilin University, 1163 Xinmin Street,Changchun 130021, China. Tel: +86-0431-85619437 E-mail:yuyaqin5540@sina.com; X. Y. Zhang, Department of Psychiatry andBehavioral Sciences, Baylor College of Medicine; VA Medical Center,2002 Holcombe Boulevard, Houston, Texas, 77030, USA. Tel: 713-791-1414 ext. 5824; Fax: 713-794-7938 E-mail: xyzhang@bcm.edu†These two authors contributed equally to this work.

Received 3 December 2012Accepted 28 February 2013Copyright © 2013 John Wiley & Sons, Ltd.

human psychopharmacologyHum. Psychopharmacol Clin Exp 2013; 28: 230–237.Published online 5 April 2013 in Wiley Online Library(wileyonlinelibrary.com) DOI: 10.1002/hup.2311

(Bakker et al., 2008). However, the results of thesegenetic association studies are often in conflict (Bakkeret al., 2008); for example, a most recent study thatexamined a total of 128 candidate genes selected fromthe literature (e.g., dopamine, serotonin, glutamate, andGABA pathways) and composite genotypes for 10drug-metabolizing enzymes. No single marker or haplo-type association reached statistical significance with theTD association after adjustment for multiple comparisons(Tsai et al., 2010). More recently, several genome-wideassociation studies have reported that some emergingtop candidate genes, such as heparan sulfate proteoglycan(HSPG2) (Syu et al., 2010), glioma-associated oncogenefamily zinc finger (GLI2) (Greenbaum et al., 2010), ordipeptidyl peptidase-like protein-6 (DPP6) (Tanakaet al., 2011) may contribute to TD susceptibility. Aberget al. (2010) reported that one polymorphism rs7669317reached genomewide significance for the AbnormalInvoluntary Movement Scale (AIMS) score, whichis located in intergenic regions between thepyrophosphatase (inorganic) 2 (PPA2) and the locusencoding hypothetical protein FLJ20184 on chromo-some 4q24. Thus, further studies are necessary to de-lineate the genetic basis of the complex TDphenotype (Basile et al., 2001).One of the most popular hypotheses about the

pathophysiology of TD is the overactivity of dopamineneurotransmission in the basal ganglia and up-regulationof dopamine receptors, due to chronic blockade ofdopamine D2-like receptors by typical antipsychotics(Casey, 2000; Lohr et al., 2003). Also, it has beenhypothesized that oxidative damage, due to an increaseddopamine turnover, may play a role in the developmentof TD (Wei et al., 1998). The generation of excessivereactive oxygen species in vivo resulting from animbalance between free-radical metabolism and theantioxidant defense system can produce interactionswith lipid, proteins, and nucleic acids, resulting in cellu-lar dysfunction or cell death (Lohr et al., 2003). Thus,genes involved in the biosynthesis and metabolism ofthe dopamine have been considered candidates for TD.Dopamine beta-hydroxylase (DBH) is a key enzyme inthe conversion of dopamine to norepinephrine innoradrenergic neurons, and variations in DBH activitymay be important in the development of various nervousand mental disorders (Wei et al., 1998). Several earlierstudies found higher plasma DBH activity in schizo-phrenic patients with TD than without TD (Jeste et al.,1981; Wagner et al., 1982; Kaufmann et al., 1986),although some did not replicate this association (Glazeret al., 1987).DBH activity in serum and cerebrospinal fluid is highly

correlated with genetically regulated levels of DBH

protein (Wilson et al., 1988; O’Connor et al., 1994;Cubells et al., 1997). Upstream of the gene encodingthis enzyme is a biallelic 19 bp insertion/deletionpolymorphism (DBH 5’-Ins/Del), which is located4.5 kilobases upstream of the transcriptional start site forthe DBH gene. The minor homozygous allele for DBH5’-Ins/Del, the deletion or “Del” allele is associated withdecreased promoter activity causing lower DBH levelsand enzyme activity in plasma and cerebrospinal fluid(CSF) (Cubells et al., 1998, 2000; Tang et al., 2007).The humanDBH gene has been localized to chromosome9q34 (Perry et al., 1991). Two studies of multiplex fami-lies affected with schizophrenia reported modest positivelinkage scores in this region (Riley et al., 1997;Kaufmann et al., 1998), whereas an Austrian researchteam, directly assessing the DBH locus, did not find evi-dence for linkage (Meszaros et al., 1996). Little evidencefor association between DBH variants and schizophreniahas been reported (Wei et al., 1996, 1998;Williams et al.,1999; Arrufat et al., 2000; Jonsson et al., 2003; Parket al., 2007), but the DBH gene may be involved withthe emergence of psychotic symptoms in schizophreniaor in cocaine induced paranoia (Cubells et al., 2000;Yamamoto et al., 2003). More recently, Park et al.(2007) found that the DBH-444G/G genotype wasassociated with psychotic symptoms in schizophrenia.Although numerous studies have shown altered

biochemical activity of DBH in TD, to our knowledge,no genetic study has examined the role of genetic DBHvariants in the pathogenesis of TD. In this investigation,we tested for associations of TD and the severity ofpsychotic symptoms with the functional polymorphismof the DBH 5’-Ins/Del. TD was assessed by means ofboth continuous (AIMS) and dichotomous (TDoccurrence) measures in a sample of Chinese Hanschizophrenic patients with TD.

METHODS

Subjects

Seven hundred and forty seven schizophrenic inpatients(male/female = 619/128) were recruited from BeijingHui-Long-Guan hospital, a Beijing-city-ownedpsychiatric hospital; and HeBei Province VeteranPsychiatric Hospital in BaoDing city, a middle-size citythat is about 50miles away from Beijing. All patientsmet the following inclusion criteria: (1) aged 25–75years;(2) confirmed DSM-IV diagnosis of schizophrenia;(3) with at least 5 years of illness; and (4) had beenreceiving stable doses of oral antipsychotic drugs for atleast 6months before entry into the study. All patientswere chronic, with a mean duration of illness of

231dbh gene and td

Copyright © 2013 John Wiley & Sons, Ltd. Hum. Psychopharmacol Clin Exp 2013; 28: 230–237.DOI: 10.1002/hup

24.6� 9.9 years. The mean duration of hospitalizationwas 9.8� 8.7 years. Antipsychotic drug treatmentconsisted mainly of monotherapy with the atypicalantipsychotics including clozapine (n=339), risperidone(n=161), quetiapine (n=31), aripirazole (n=25),olanzapine (n=10), and loxapine (n=9); and the typicalantipsychotics including chlorpromazine (n = 56),sulpiride (n = 40), perphenazine (n = 38), haloperidol(n=23), pipotiazine palmitate injection (n=10), andothers (n=5). Mean antipsychotic dose (inchlorpromazine equivalents) was 459� 430mg per day.We recruited 625 healthy controls (male/female =

404/221) matched for age and gender from the localcommunity. All healthy controls were interviewed bytrained investigators, who were supervised by one ofthe research psychiatrists. Psychiatric disorders wereruled out among healthy controls by a psychiatricevaluation. None of them had a personal or familyhistory of psychiatric disorder, and all were HanChinese from the Beijing area.A complete medical history, physical examination,

and laboratory tests were obtained from patients andcontrol subjects. All subjects were physically healthy,and any subjects with major medical illnesses or drugand alcohol abuse/dependence were excluded. Allsubjects gave written informed consent, which wasapproved by the Institutional Review Board of BeijingHui-Long-Guan hospital.

Clinical measures

A detailed questionnaire including general informa-tion, sociodemographic characteristics, smoking be-havior, and medical and psychological conditionswas administered to each subject by a member ofthe research staff. Additional information was col-lected from available medical records and collateraldata. Additional visits were requested for subjectswith missing or ambiguous data.Four experienced psychiatrists, who were blinded to

the clinical status of the patients and to the charted clinicaldata, assessed the severity of TD using AIMS. The firstseven AIMS items determine the severity of abnormalmovements for each of seven respective regions of thebody. When a patient had an AIMS score of at least 3(moderate degree) in any body part or with at least 2(mild degree) in two or more body parts among the sevenAIMS items, he or she was considered to have TD. Theothers were assumed not to have TD, using the criteriaof Schooler and Kane (Schooler and Kane, 1982). Atthe same time, the same four psychiatrists assessed thepatients’ psychopathology using the Positive andNegative Syndrome Scale (PANSS).

All these raters had simultaneously attended atraining session in the use of the AIMS and thePANSS before this study start. After training, a corre-lation coefficient greater than 0.8 was maintained forthe AIMS or PANSS total score by repeated assess-ments. All evaluations were carried out prior to thelaboratory procedures.

Genetic analysis

We extracted DNA by using standard protocols. Thegenotypes of the DBH 5’-Ins/Del were identified onthe basis of an earlier report (Cubells et al., 2000).The following primers (Genbank Accession No.X63418) were used to generate PCR products of144 bp (DBH 5’-Del) or 163 bp (DBH 5’-Ins): sense:5’-GCA AAA GTC AGG CAC ATG CAC C-3’; anti-sense: 5’-CAA TAA TTT GGC CTCAAT CTTGG-3’.The resulting fragments were resolved on 2% agarosegels, stained with ethidium bromide, trans-illuminatedwith UV light, and digitally imaged for genotyping.Genotyping was duplicated and carried out blind tothe clinical status.

Statistical analysis

Deviations from Hardy–Weinberg equilibrium wereassessed using the Hardy-Weinberg simple program(HWSIM) program (Cubells et al., 1997). Differencesof the DBH 5’-Ins/Del allele and genotype frequenciesbetween patients with TD and without TD were evalu-ated using w2 tests. Furthermore, a linear regression(for PANSS total score) or logistic regression (for TDand schizophrenia) was employed to investigate theassociation with DBH 5’-Ins/Del genotype.Group differences were compared using Student’s

two-sample t-test or one-way analysis of variance forcontinuous variables and chi squared for categoricalvariables. Bonferroni corrections were applied to eachtest to adjust for multiple testing. A univariate analysiswas performed to examine the relationship betweendemographic variable (independent variable) and TDas the dependent variables. Only those variablesthat were statistically significantly associated withTD were included in the following logistic regression.Then, a logistic regression analysis was performedto adjust potential confounding factors for TD usingTD as a dependent variable and those variables show-ing significant association with AIMS total score asindependent variables.The power (defined as the chance that true

differences will actually be detected) of the samplewas calculated with Quanto Software, with knownrisk allele frequencies, a schizophrenia populationprevalence of 0.01, or a TD prevalence of 0.33 among

232 n. zhou ET AL.

Copyright © 2013 John Wiley & Sons, Ltd. Hum. Psychopharmacol Clin Exp 2013; 28: 230–237.DOI: 10.1002/hup

schizophrenia and examining log additive, recessive,and dominant models.

RESULTS

Allele and genotype frequencies of the DBH 5’-Ins/Delpolymorphism in schizophrenic patients with andwithout TD and healthy controls

Allele and genotype frequencies of samples for theDBH 5’-Ins/Del polymorphism are given in Table 1.Distributions of the DBH 5’-Ins/Del genotypes wereconsistent with Hardy–Weinberg equilibrium in bothpatients with schizophrenia and healthy controls (bothp> 0.05). No significant differences were found inDBH 5’-Ins/Del genotype and allele distributions betweenthe patients and healthy controls (w2=0.08, d.f. = 2,p=0.96; and w2=0.04, d.f. = 1, p=0.85, respectively).No significant sex differences in DBH 5’-Ins/Delgenotype and allele distributions were observed in thetotal subjects or when the patients and healthy controlswere analyzed separately (all p> 0.05).Out of the 747 patients, 312 (42%) were diagnosed as

having TD. The distribution of the allele and genotypefrequencies did not differ between the patients with andwithout TD (w2 = 0.8, d.f. = 2, p = 0.67; and w2 = 0.78,d.f. = 1, p= 0.38, respectively) (Table 1). Furthermore,the frequencies of genotype and alleles between theschizophrenia patients with TD and healthy controlsdid not differ (both p> 0.05). Furthermore, a logisticregression for schizophrenia or TD after controlling forpotential confounding factors still showed no associa-tion with DBH 5’-Ins/Del genotype (both p> 0.05).This total sample had 0.98–0.99 power to detect DBH

5’-Ins/Del polymorphism associated with schizophreniaat the mode of dominant, recessive, and log additiveinheritance, with an odds ratio (OR) of 2 (alpha = 0.05,two-tailed test). Furthermore, the TD sample had0.94–0.99 power to detect this polymorphism associatedwith TD at the mode of dominant, recessive, and logadditive inheritance, with an OR of 2 (alpha = 0.05,two-tailed test).

Demographic, clinical variables and AIMS scores ofpatients in the DBH 5’-Ins/Del genotype groups

Demographic variables, clinical variables, and AIMSscores in the DBH 5’-Ins/Del genotype groups areshown in Table 2. The PANSS total score andits subscale scores were normally distributed acrossthe patients and among the genotype categories(all p> 0.05). There was a significant difference inthe PANSS positive symptom subscore among thethree genotypes (F = 3.36, p< 0.02) with higherpositive symptom score in patients with Del/Delgenotype than those with Ins/Ins and Ins/Del geno-types (p< 0.02 and p< 0.03, respectively). However,these results did not pass Bonferroni test.In addition, there was no significant difference between

these three groups in terms of age, sex, duration ofillness, schizophrenia subtype, current neurolepticdoses, duration of antipsychotic treatment, number ofhospitalization, or the PANSS total and the othersubscale scores (all p> 0.05). Furthermore, the AIMStotal score did not differ among the three genotypegroups: 3.2� 3.5 (Del/Del), 3.1� 3.3 (Del/Ins), and3.3� 3.6 (Ins/Ins) (p> 0.05).

Clinical characteristics and DBH 5’-Ins/Del genotypeassociated with tardive dyskinesia

The univariate analyses are presented in Table 3. Thecharacteristics that were significantly different betweenpatients with and without TD were as follows: sex,age, education, the number of hospitalizations, and thenegative symptom subscore of PANSS (all p< 0.01).Thus, these variables entered into the following logisticregression analysis. There were no significant differ-ences in the antipsychotic type or dose between TDand non-TD groups (both p> 0.05).We found that the old age [p= 0.001, OR=1.05, 95%

confidence interval (CI) = 1.02–1.08], male (p=0.004,OR=2.15, CI = 1.27–3.65), and greater negative symp-tom (p= 0.04, OR=1.04, CI = 1.01–1.06) were the riskfactors for TD. This logistic regression model, however,

Table 1. DBH 5’-ins/del allele and genotype frequencies in schizophrenia and healthy controls

Genotype distribution Allele frequency

Ins/Ins Ins/Del Del/Del Ins Del

Control (n= 625) 216(34.6%) 297(47.5%) 112(17.9%) 729(58.3%) 521(41.7%)Schizophrenia 253(33.9%) 360(48.2%) 134(17.9%) 866(58.0%) 628(42.0%)TD (n= 312) 111(35.6%) 148(47.4%) 53(17.0%) 370(59.3%) 254(40.7%)Non-TD (n= 435) 142(32.6%) 212(48.7%) 81(18.6%) 496(57.0%) 374(43.0%)

Note: No significant differences were found in DBH 5’-Ins/Del genotype and allele distributions between the patients and healthy controls (w2 = 0.08, d.f. = 2, p=0.96;and w2 = 0.04, d.f. = 1, p=0.85, respectively), or between the patients with and without TD (w2 = 0.80, d.f. = 2, p=0.67; and w2 = 0.78, d.f. = 1, p=0.38, respectively).

233dbh gene and td

Copyright © 2013 John Wiley & Sons, Ltd. Hum. Psychopharmacol Clin Exp 2013; 28: 230–237.DOI: 10.1002/hup

showed no association between DBH 5’-Ins/Delgenotype single-nucleotide polymorphism (SNP) andTD (p= 0.39).

DISCUSSION

This study had two major findings. (1) The DBH 5’-Ins/Del gene polymorphism was not associated withTD. (2) There were significantly higher positive symp-tom scores in patients carrying the DBH Del/Del geno-type than patients with Ins/Ins and Ins/Del genotypes;however, these results did not pass Bonferroni test.Although earlier biochemical studies had established

that TD was related to low DBH activity (Jeste et al.,1981; Wagner et al., 1982; Kaufmann et al., 1986), wedid not find significant association between the DBH 5’-Ins/Del gene polymorphism and TD. Lack of associationis hardly uncommon in human genetics, and some recentreports show that there is always genetic heterogeneity(Lerer et al., 2002; Hallock and Thomas, 2012; Lvet al., 2012). Some recent meta-analyses also showedgenetic heterogeneity for TD (Lerer and Segman, 2006;Bakker et al., 2008). Our negative finding could be addedto future meta-analysis across studies to integrate the data.

Our results suggest that the DBH 5’-Ins/Del genepolymorphism may not play a major role in thesusceptibility to development of TD in patients withschizophrenia. However, no study has been conductedto ascertain whether the DBH genetic variants affectthe DBH activity in schizophrenia. Whether thealteration of DBH activity in TD is caused by DBHgenetic variants or by other risk factors for TD is still un-known. In addition, the relationship between otherpolymorphisms of DBH and TD deserves furtherinvestigation.Moreover, antipsychotic-induced abnormalmovements could result from the combined influenceof several genes and clinical features (Lee et al., 2007).Therefore, the relationship between DBH genetic effectsand antipsychotic-induced TD requires further studiesincluding gene–gene interactions and haplotype analysisof the DBH gene.Our finding of no association between the DBH 5’-

Ins/Del polymorphism and schizophrenia accords withprevious reports (Yamamoto et al., 2003; Park et al.,2007). However, the patients with schizophrenia whohad the Del/Del genotype showed higher PANSS scoresthan patients with the Ins/Ins genotype, particularly forpositive symptoms. This association is consistent with

Table 2. Demographic and clinical variables in schizophrenia patients divided with regard to DBH 5’-Ins/Del genotypes

Del/Del Ins/Del Ins/Ins F or w2 p

Sex (male/female) 115/19 293/67 211/42 0.33 0.85Age (years) 46.6� 9.2 48.0� 9.6 48.5� 10.2 1.69 0.19Education (years) 9.4� 8.2 9.1� 5.5 8.7� 2.5 0.86 0.42Age of onset (years) 23.3� 4.2 23.4� 5.3 23.3� 5.8 0.03 0.97Hospitalization times 3.9� 2.3 4.3� 2.7 4.2� 3.2 0.98 0.38Dose of antipsychotics(chlorpromazine equivalents)

420� 270 458� 389 479� 499 0.87 0.42

PANSS scorePositive symptoms 13.2� 5.2 11.2� 4.9 11.1� 3.1 3.36 <0.05Negative symptoms 23.5� 7.6 22.9� 8.5 22.8� 8.4 0.35 0.71General psychopathology 25.1� 5.6 25.1� 5.5 25.7� 6.5 0.79 0.46Total score 61.8� 13.9 59.2� 14.8 59.6� 15.4 1.91 0.16AIMS total score 3.2� 3.5 3.1� 3.3 3.3� 3.6 0.14 0.84

Table 3. Demographics of TD and non-TD patients and healthy controls

Non-TD (n= 435) TD (n= 312) NC (n= 625) Adjusted p-value

Sex (male/female) 342/93 277/35 404/221 <0.0001Age (years) 46.8� 9.8 49.5� 9.3 46.1� 13.7 <0.0001Education (years) 9.4� 6.6 8.4� 2.6 9.2� 5.4 <0.01Age of onset (years) 23.3� 5.5 23.5� 5.1Hospitalization times 3.9� 2.7 4.5� 2.9 ***Dose of antipsychotics (chlorpromazine equivalents) 475� 473 436� 372PANSS Total score 59.3� 15.3 61.2� 13.9Positive subscore 11.9� 5.3 11.5� 4.5Negative subscore 21.9� 8.4 24.6� 7.8***General subscore 25.5� 6.1 25.1� 5.5

Note: ***p<0.005

234 n. zhou ET AL.

Copyright © 2013 John Wiley & Sons, Ltd. Hum. Psychopharmacol Clin Exp 2013; 28: 230–237.DOI: 10.1002/hup

two previous reports indicating that a haplotype,including the Del allele of DBH 5’-Ins/Del, isassociated with increased vulnerability to paranoidsymptoms in cocaine abusers (Cubells et al., 2000)and psychotic symptoms in schizophrenia (Yamamotoet al., 2003). However, a more recent study showed thatwhile the G/G genotype of the DBH 444G/A wasassociated with increased psychotic symptoms inschizophrenia, the Del/Del genotype of DBH 5’-Ins/Del polymorphism did not show such an association(Park et al., 2007). Thus, the DBH gene appearsto play a role in increasing the vulnerability topsychotic symptoms, but specific polymorphismsmay differ across ethnicity or other factors. However,it is noteworthy that our findings of associationbetween DBH 5’-Ins/Del polymorphism and thepsychotic symptoms did not withstand multiple testing,suggesting that this polymorphism may only play asmall effect.It was reported that the “Del” alleles of the DBH 5’-

Ins/Del gene polymorphism is associated with lowplasma DBH activity (Cubells et al., 1998), and lowDBH activity may indicate a slower conversion of dopa-mine to norepinephrine, with an elevated dopamine/norepinephrine ratio in the brain (Stanley et al., 1996),supporting the hypothesis of a hyperdopaminergicstate in schizophrenia (Carlsson et al., 2001). It isknown that the positive symptoms are associated withsubcortical hyperdopaminergic state in schizophrenia(Davis et al., 1991; Moncrieff, 2009). Taken together,it is possible that DBH gene variants by affecting theDBH enzyme activity contribute to increased dopaminetransmission, which in turn increase vulnerability topsychotic symptoms.This study has several limitations. First, only one

polymorphism was studied, which does not fully tagall the genetic variants in DBH gene. There is evidencethat other functional polymorphisms in the DBH geneor regulatory regions may also influence the enzymeactivity, such as DBH-1021C/T polymorphism thatis widely considered as a functional one with minoreffect (Cubells et al., 2000), which could have beeninvestigated to provide a more comprehensive analy-sis of the effects of variation within the gene. Second,the sample size was limited. Thus, the results of anassociation between the DBH polymorphism andpsychotic symptoms are tentative and need to bereplicated in independent samples. Third, hiddenpopulation stratification of our samples could be con-founders, although the Han Chinese in Beijing areethnically relatively homogenous. A replication studywith genomic controls or a family-based populationstudy would help address this limitation. Fourth, it

would be helpful if plasma levels of DBH activity inthe patients and healthy controls had been providedto interpret the functionality of this polymorphismin the Chinese population, and could help explainthe negative results. Also, it would be fruitful toreport if DBH activity might be associates with TDsymptomatology. Fifth, all schizophrenic patients inthe present study were inpatients, with long-termduration of illness. Thus, generalizing our study islimited by our sample of chronically hospitalizedpatients, who had more severe psychopathology anda longer duration of illness than typical psychoticoutpatients. Sixth, the prevalence of TD in our sampleof 42% is in the middle range (15–70%) of existingworldwide prevalence estimates (Correll and Schenk,2008), but higher than the estimates from Asia(Zhang et al., 2009). We currently could not offer areasonable explanation regarding these differencesin TD prevalence.In summary, the DBH 5’-Ins/Del polymorphism

does not appear to be a risk factor for TD, but may playa role by increasing the vulnerability to psychoticsymptoms in schizophrenia. However, it is noteworthythat this is essentially a negative study. Although theprimary hypothesis was a strong one, it was notsupported by our current results. The report of an asso-ciation to positive symptoms was a post-hoc result anddid not withstand the Bonferroni correction for multi-ple comparisons. Moreover, as there were multipleconfounds in the cross-sectional assessment of positivesymptoms, our findings should be viewed as evenmore exploratory. Further adequately powered studiesinvolving gene–gene interactions and other polymor-phisms of DBH gene (e.g. �1021C/T) in a largersample will be needed. In addition, the positive findingsin the present study should be followed-up withtranscriptomics studies, with a broader focus on geneexpression using the recent integrative platforms fortranscriptomics (Zhao et al., 2012).

CONFLICT OF INTEREST

The authors have declared that there is no conflict ofinterest.

ACKNOWLEDGEMENTS

This study has been supported by grants from the StanleyMedical Research Institute (03T-459 and 05T-726) and theDepartment of Veterans Affairs, VISN 16, Mental IllnessResearch, Education and Clinical Center (MIRECC),United States National Institute of Health K05-DA0454,P50-DA18827, and U01-MH79639.

235dbh gene and td

Copyright © 2013 John Wiley & Sons, Ltd. Hum. Psychopharmacol Clin Exp 2013; 28: 230–237.DOI: 10.1002/hup

REFERENCES

Aberg K, Adkins DE, Bukszár J, et al. 2010. Genomewide associationstudy of movement-related adverse antipsychotic effects. Biol Psychiatry67(3): 279–282.

Arrufat FJ, Diaz R, Queralt R, et al. 2000. Analysis of the polymorphic(GT)(n) repeat at the dopamine beta-hydroxylase gene in Spanishpatients affected by schizophrenia. Am J Med Genet 96: 88–92.

Bakker PR, van Harten PN, van Os J. 2008. Antipsychotic-induced tardivedyskinesia and polymorphic variations in COMT, DRD2, CYP1A2 andMnSOD genes: a meta-analysis of pharmacogenetic interactions. MolPsychiatry 13: 544–556.

Ballesteros J, González-Pinto A, Bulbena A. 2000. Tardive dyskinesiaassociated with higher mortality in psychiatric patients: results of ameta-analysis of seven independent studies. J Clin Psychopharmacol20: 188–194.

Basile VS, Ozdemir V, Masellis M, et al. 2001. Lack of associationbetween serotonin-2A receptor gene (HTR2A) polymorphisms and tardivedyskinesia in schizophrenia. Mol Psychiatry 6: 230–234.

Browne S, Roe M, Lane A, et al. 1996. A preliminary report on theeffect of a psychosocial and educative rehabilitation programme onquality of life and symptomatology in schizophrenia. Eur Psychiatry11: 386–389.

Carlsson A, Waters N, Holm-Waters S, Tedroff J, Nilsson M, Carlsson ML,2001. Interactions between monoamines, glutamate, and GABA inschizophrenia: new evidence. Annu Rev Pharmacol Toxicol 41: 237–260.

Casey DE. 2000. Tardive dyskinesia: pathophysiology and animal models.J Clin Psychiatry 61(Suppl 4): 5–9.

Correll CU, Malhotra AK. 2004. Pharmacogenetics of antipsychotic-inducedweight gain. Psychopharmacology 174: 477–489.

Correll CU, Schenk EM. 2008. Tardive dyskinesia and new antipsychotics.Curr Opin Psychiatry 21: 151–156.

Cubells JF, Kobayashi K, Nagatsu T, et al. 1997. Population genetics ofa functional variant of the dopamine beta-hydroxylase gene (DBH). AmJ Med Genet 74: 374–379.

Cubells JF, van Kammen DP, Kelley ME, et al. 1998. Dopamine beta-hydroxylase: two polymorphisms in linkage disequilibrium at thestructural gene DBH associate with biochemical phenotypic variation.Hum Genet 102: 533–540.

Cubells JF, Kranzler HR,McCance-Katz E, et al. 2000. A haplotype at the DBHlocus, associated with low plasma dopamine beta-hydroxylase activity, alsoassociates with cocaine-induced paranoia. Mol Psychiatry 5: 56–63.

Davis KL, Kahn RS, Ko G, Davidson M. 1991. Dopamine in schizophre-nia: a review and reconceptualization. Am J Psychiatry 148: 1474–1486.

Gerlach J. 2002. Improving outcome in schizophrenia: the potential importanceof EPS and neuroleptic dysphoria. Ann Clin Psychiatry 14: 47–57.

Glazer WM. 2000. Expected incidence of tardive dyskinesia associatedwith atypical antipsychotics. J Clin Psychiatry 61(Suppl 4): 21–26.

Glazer WM, Morgenstern H, Jeste DV, Zahner G, Hafez HM, Benarroche CL.1987. Serum dopamine beta hydroxylase activity and tardive dyskinesia.Psychoneuroendocrinology 12: 289–294.

Greenbaum L, Alkelai A, Rigbi A, Kohn Y, Lerer B. 2010. Evidence forassociation of the GLI2 gene with tardive dyskinesia in patients with chronicschizophrenia.Mov Disord 25(16): 2809–2817.

Hallock P, Thomas MA. 2012. Integrating the Alzheimer’s disease proteomeand transcriptome: a comprehensive network model of a complex disease.OMICS 16(1-2): 37–49.

Jeste DV. 2004. Tardive dyskinesia rates with atypical antipsychotics inolder adults. J Clin Psychiatry 65(Suppl 9): 21–24.

Jeste DV, Neckers LM, Wagner RL, et al. 1981. Lymphocyte monoamineoxidase and plasma prolactin and growth hormone in tardive dyskinesia.J Clin Psychiatry 42: 75–77.

Jönsson EG, Abou Jamra R, Schumacher J, et al. 2003. No association betweena putative functional promoter variant in the dopamine beta-hydroxylase geneand schizophrenia. Psychiatr Genet 13: 175–178.

Kane JM. 2004. Tardive dyskinesia rates with atypical antipsychotics inadults: prevalence and incidence. J Clin Psychiatry 65(Suppl 9): 16–20.

Kaufmann CA, Jeste DV, Shelton RC, Linnoila M, Kafka MS, Wyatt RJ. 1986.Noradrenergic and neuroradiological abnormalities in tardive dyskinesia. BiolPsychiatry 21(8-9): 799–812.

Kaufmann CA, Suarez B, Malaspina D, et al. 1998. NIMH Genetics InitiativeMillenium Schizophrenia Consortium: linkage analysis of African–Americanpedigrees. Am J Med Genet 81: 282–289.

Lee HJ, Kang SG, Choi JE, et al. 2007. No association between dopamine D4receptor gene-521 C/T polymorphism and tardive dyskinesia in schizophre-nia. Neuropsychobiology 55: 47–51.

Lerer B, Segman RH. 2006. Pharmacogenetics of antipsychotic therapy:pivotal research issues and the prospects for clinical implementation.Dialogues Clin Neurosci 8(1): 85–94.

Lerer B, Segman RH, Fangerau H, et al. 2002. Pharmacogenetics of tardive dys-kinesia: combined analysis of 780 patients supports association with dopa-mine D3 receptor gene Ser9Gly polymorphism. Neuropsychopharmacology27: 105–119.

Lohr JB, Kuczenski R, Niculescu AB, 2003. Oxidative mechanisms andtardive dyskinesia. CNS Drugs 17: 47–62.

Lv S, XuY, ChenX, et al. 2012. Prioritizing cancer therapeutic small moleculesby integrating multiple OMICS datasets. OMICS 16(10): 552–559.

Margolese HC, Chouinard G, Kolivakis TT, Beauclair L, Miller R, AnnableL. 2005. Tardive dyskinesia in the era of typical and atypicalantipsychotics. Part 2: Incidence and management strategies in patientswith schizophrenia. Can J Psychiatry 50: 703–714.

Marsálek M. 2000. Tardive drug-induced extrapyramidal syndromes.Pharmacopsychiatry 33(Suppl 1): 14–33.

Meszaros K, Lenzinger E, Füreder T, et al. 1996. Schizophrenia and thedopamine-beta-hydroxylase gene: results of a linkage and associationstudy. Psychiatr Genet 6: 17–22.

Moncrieff J. 2009. A critique of the dopamine hypothesis of schizophreniaand psychosis. Harv Rev Psychiatry 17: 214–225

Müller DJ, Schulze TG, Knapp M, et al. 2001. Familial occurrence oftardive dyskinesia. Acta Psychiatrica Scandinavia 104: 375–379.

O’Connor DT, Cervenka JH, Stone RA, et al. 1994. Dopamine beta-hydroxylase immunoreactivity in human cerebrospinal fluid: properties,relationship to central noradrenergic neuronal activity and variation inParkinson’s disease and congenital dopamine beta-hydroxylase defi-ciency. Clin Sci 86: 149–158.

Pae CU. 2008. Additive effect between quinine oxidoreductase gene(NQO1: Pro187Ser) and manganese superoxide dismutase gene(MnSOD: Ala-9Val) polymorphisms on tardive dyskinesia in patientswith schizophrenia. Psychiatry Res 161(3): 336–338.

Pae CU, Yu HS, Kim JJ, et al. 2004. Quinone oxidoreductase (NQO1) genepolymorphism (609C/T) may be associated with tardive dyskinesia, butnot with the development of schizophrenia. Int J Neuropsychopharmacol7(4): 495–500.

Park JK, Kim JW, Lee HJ, et al. 2007. Dopamine beta-hydroxylase genepolymorphisms and psychotic symptoms in schizophrenia. Am J MedGenet B Neuropsychiatr Genet 144B: 944–945.

Patterson TL, Bucardo J, McKibbin CL, et al. 2005. Development and pilottesting of a new psychosocial intervention for older Latinos with chronicpsychosis. Schizophr Bull 31: 922–930.

Perry SE, SummarML, Phillips JA 3rd, Robertson D. 1991. Linkage analysis ofthe human dopamine beta-hydroxylase gene. Genomics 10: 493–495.

Remington G, Kwon J, Collins A, Laporte D, Mann S, Christensen B. 2007.The use of electronic monitoring (MEMS) to evaluate antipsychoticcompliance in outpatients with schizophrenia. Schizophr Res 90: 229–237.

Riley BP, Tahir E, Rajagopalan S, et al. 1997. A linkage study of theN-methyl-D-aspartate receptor subunit gene loci and schizophrenia insouthern African Bantu-speaking families. Psychiatr Genet 7: 57–74.

Rosengarten H, Schweitzer JW, Friedhoff AJ. 1994. Possible geneticfactors underlying the pathophysiology of tardive dyskinesia. PharmacolBiochem Behav 49: 663–667.

Schooler NR, Kane JM. 1982. Research diagnoses for tardive dyskinesia.Arch Gen Psychiatry 39: 486–487.

Soares KV, McGrath JJ. 1999. The treatment of tardive dyskinesia—asystematic review and meta-analysis. Schizophr Res 39: 1–16.

Stanley JA, Williamson PC, Drost DJ, et al. 1996. An in vivo proton mag-netic resonance spectroscopy study of schizophrenia patients. SchizophrBull 22: 597–609.

Syu A, Ishiguro H, Inada T, et al. 2010. Association of the HSPG2 genewith neuroleptic-induced tardive dyskinesia. Neuropsychopharmacology35(5): 1155–1164.

236 n. zhou ET AL.

Copyright © 2013 John Wiley & Sons, Ltd. Hum. Psychopharmacol Clin Exp 2013; 28: 230–237.DOI: 10.1002/hup

Tanaka S, Syu A, Ishiguro H, et al. 2011. DPP6 as a candidate genefor neuroleptic-induced tardive dyskinesia. Pharmacogenomics J.doi:10.1038/tpj.2011.36. [Epub ahead of print]

Tang YL, Epstein MP, Anderson GM, Zabetian CP, Cubells JF. 2007. Genotypicand haplotypic associations of the DBH gene with plasma dopamine beta-hydroxylase activity in African Americans. Eur J Hum Genet 15: 878–883.

Tarsy D, Baldessarini RJ. 2006. Epidemiology of tardive dyskinesia: is riskdeclining with modern antipsychotics? Movement Disorder 21: 589–598.

Tsai HT, Caroff SN, Miller del D, et al. 2010. A candidate gene study ofTardive dyskinesia in the CATIE schizophrenia trial. Am J Med GenetB Neuropsychiatr Gene 153B(1): 336–340.

Wagner RL, JesteDV, Phelps BH,Wyatt RD. 1982. Enzyme studies in tardive dys-kinesia. I. One-year biochemical follow-up. JClin Psychopharmacol 2: 312–314.

Wei J, Ramchand CN, Hemmings GP. 1996. A study of the relationshipbetween the DBH activity in serum and a MspI polymorphic site in intron9 of the human DBH gene in schizophrenia. Schizophr Res 22: 77–80.

Wei J, Ramchand CN, Hemmings GP. 1998. TaqI polymorphic sites at thehuman dopamine beta-hydroxylase gene possibly associated withbiochemical alterations of the catecholamine pathway in schizophrenia.Psychiatr Genet 8: 19–24.

Williams HJ, Bray N, Murphy KC, Cardno AG, Jones LA, Owen MJ. 1999.No evidence for allelic association between schizophrenia and a

functional variant of the human dopamine beta-hydroxylase gene(DBH). Am J Med Genet 88: 557–559.

Wilson AF, Elston RC, Siervogel RM, Tran LD. 1988. Linkage of a generegulating dopamine-beta-hydroxylase activity and the ABO blood grouplocus. Am J Hum Genet 42: 160–166.

Yamamoto K, Cubells JF, Gelernter J, et al. 2003. Dopamine beta-hydroxylase (DBH) gene and schizophrenia phenotypic variability:a genetic association study. Am J Med Genet B Neuropsychiatr Genet117B: 33–38.

Youssef HA, Waddington JL. 1987. Morbidity and mortality in tardivedyskinesia: associations in chronic schizophrenia. Acta PsychiatricaScandinavia 75: 74–77.

Zai CC, Tiwari AK, Basile V, et al. 2010. Oxidative stress in tardivedyskinesia: genetic association study and meta-analysis of NADPHquinine oxidoreductase 1 (NQO1) and superoxide dismutase 2(SOD2, MnSOD) genes. Prog Neuropsychopharmacolog BiolPsychiatry 34(1): 50–56.

Zhang XY, Chen da C, Qi LY, et al. 2009. Gender differences in theprevalence, risk and clinical correlates of tardive dyskinesia in Chineseschizophrenia. Psychopharmacology 205(4): 647–654.

ZhaoD,Wu J, ZhouY,GongW,Xiao J, Yu J. 2012.WikiCell: a unified resourceplatform for human transcriptomics research. OMICS 16(6): 357–362.

237dbh gene and td

Copyright © 2013 John Wiley & Sons, Ltd. Hum. Psychopharmacol Clin Exp 2013; 28: 230–237.DOI: 10.1002/hup