Comprehensive analysis of tagging sequence variants in DTNBP1 shows no association with schizophrenia or with its composite neurocognitive endophenotypes.код для вставкиСкачать
American Journal of Medical Genetics Part B (Neuropsychiatric Genetics) 147B:1159– 1166 (2008) Comprehensive Analysis of Tagging Sequence Variants in DTNBP1 Shows No Association With Schizophrenia or With Its Composite Neurocognitive Endophenotypes Kirsten Peters,1 Steven Wiltshire,2,3 Anjali K. Henders,4 Milan Dragović,5 Johanna C. Badcock,5 David Chandler,5 Sarah Howell,5 Chris Ellis,2,3 Sonja Bouwer,1 Grant W. Montgomery,4 Lyle J. Palmer,2,3,5 Luba Kalaydjieva,1 and Assen Jablensky5* 1 Laboratory for Molecular Genetics, Western Australian Institute for Medical Research, University of Western Australia, Nedlands, Western Australia, Australia 2 Laboratory for Genetic Epidemiology, Western Australian Institute for Medical Research, University of Western Australia, Nedlands, Western Australia, Australia 3 UWA Centre for Medical Research, University of Western Australia, Nedlands, Western Australia, Australia 4 Genetic Epidemiology Laboratory, Queensland Institute of Medical Research, Brisbane, Queensland, Australia 5 Centre for Clinical Research in Neuropsychiatry, School of Psychiatry and Clinical Neurosciences, University of Western Australia, Mount Claremont, Western Australia, Australia In a previous study we identified a relatively homogeneous subtype of schizophrenia characterized by pervasive cognitive deficit, which was the exclusive contributor to our findings of linkage to 6p25-p24. The 6p region contains Dysbindin (DTNBP1), considered to be one of the major schizophrenia candidate genes. While multiple studies have reported association between genetic variation in DTNBP1 and schizophrenia, the findings have been inconsistent and controversial, leading to recent calls for systematic reexamination and unambiguous evidence of association. To address this, we have undertaken a comprehensive survey of common genetic variation within DTNBP1 and its association with schizophrenia, using a HapMap-based genetagging approach. We genotyped 39 tSNPs in a sample of 336 cases and 172 controls of Anglo-Irish ancestry, with the phenotype defined as clinical schizophrenia, and as composite neurocognitive endophenotypes. Allele and haplotype frequencies, and LD structure in our control sample were similar to those in other European populations. Using multivariate generalized linear modeling, we observed no significant association between any tSNP and any outcome variable. Association with haplotypes was examined across the gene and in the previously associated 50 region. Neither global haplotype tests, nor specific analysis of the ‘‘risk’’ haplotype previously reported in an This article contains supplementary material, which may be viewed at the American Journal of Medical Genetics website at http://www.interscience.wiley.com/jpages/1552-4841/suppmat/ index.html. Grant sponsor: NHMRC; Grant numbers: 404046, 404025; Grant sponsor: Wind-Over-Water Foundation. *Correspondence to: Prof. Assen Jablensky, M.D., DMSc, School of Psychiatry and Clinical Neurosciences, MRF Building Level 3, 50 Murray Street, Perth 6000, Australia. E-mail: email@example.com Received 28 November 2007; Accepted 24 January 2008 DOI 10.1002/ajmg.b.30741 Published online 3 March 2008 in Wiley InterScience (www.interscience.wiley.com) ß 2008 Wiley-Liss, Inc. ethnically related population, the Irish highdensity schizophrenia families, showed significant evidence of association with schizophrenia or with the neurocognitive endophenotypes in our sample. The framework and results of this study should facilitate further attempts at reanalysis of DTNBP1, in terms of standardized approaches to both phenotype definition and analysis of genetic variation. ß 2008 Wiley-Liss, Inc. KEY WORDS: schizophrenia; DTNBP1; association; endophenotypes; HapMap gene-tagging Please cite this article as follows: Peters K, Wiltshire S, Henders AK, Dragović M, Badcock JC, Chandler D, Howell S, Ellis C, Bouwer S, Montgomery GW, Palmer LJ, Kalaydjieva L, Jablensky A. 2008. Comprehensive Analysis of Tagging Sequence Variants in DTNBP1 Shows No Association With Schizophrenia or With Its Composite Neurocognitive Endophenotypes. Am J Med Genet Part B 147B:1159–1166. INTRODUCTION The 6p24-p22 region is now regarded as one of the wellestablished schizophrenia loci (OMIM SCZD3 #600511). It has been linked to schizophrenia in a number of studies [Schwab et al., 1995; Straub et al., 1995; Wang et al., 1995] and was among the few supported by a genome scan meta-analysis [Lewis et al., 2003]. A candidate gene in the region, Dystrobrevin binding protein 1 (DTNBP1, Dysbindin), was first identified by Straub et al.  and Van den Oord et al.  who observed association between DTNBP1 sequence variants and schizophrenia in a sample of Irish high-density schizophrenia families. Subsequent analyses of nearly 30 independent, mostly case–control samples [Schwab et al., 2003; Van den Bogaert et al., 2003; Funke et al., 2004; Kirov et al., 2004; Numakawa et al., 2004; Gornick et al., 2005; Duan et al., 2007; Tosato et al., 2007; Turunen et al., 2007; Vilella et al., 2007 and references therein] have led to reports of positive replication findings in many (but not all) studies. A number of studies described association between DTNBP1 variation and different measures of cognitive function in patients as well as in normal subjects [Burdick et al., 2006; Fallgatter et al., 2006; 1160 Peters et al. Donohoe et al., 2007; Zinkstok et al., 2007], and an effect of ‘‘high risk’’ variants and/or haplotypes on the clinical manifestations and progression of the disease [Fanous et al., 2005; Gornick et al., 2005; DeRosse et al., 2006; Burdick et al., 2007; Tosato et al., 2007]. While the replication claims have placed Dysbindin among the major candidate genes conferring susceptibility to schizophrenia [Harrison and Weinberger, 2005; Owen et al., 2005; Karayiorgou and Gogos, 2006; Riley and Kendler, 2006], the results of these association studies have been inconsistent and controversial. Use of limited sets of partially overlapping markers, and widely divergent findings where associated alleles and haplotypes differ even between samples derived from closely related populations [Harrison and Weinberger, 2005; Karayiorgou and Gogos, 2006; Riley and Kendler, 2006] have made the findings difficult to interpret. A recent metaanalysis [Li and He, 2007] has failed to support association, and a comprehensive study [Mutsuddi et al., 2006] of normal variation and linkage disequilibrium (LD) patterns in DTNBP1 makes it unlikely that the reported disparate positive findings can be attributed to inter-population differences or to the existence of multiple rare susceptibility alleles. Dysbindin has been used as an example of fallible interpretation of the findings in replication studies [NCI-NHGRI Working Group on Replication in Association Studies, 2007] and Mutsuddi et al.  have called for an unambiguous confirmation of the association, based on a systematic approach. In a previous study [Hallmayer et al., 2005], we defined two major schizophrenia subtypes with markedly different neurocognitive profiles and tentatively established a distinct genetic basis of the cognitive deficit (CD subtype), based on its exclusive contribution to significant linkage (lod 3.5) to the 6p25-p24 region, where the remaining families with probands classified as cognitively spared yielded a negative LOD score. Although DTNBP1 falls outside that interval, it is within the more broadly defined 6p24-p22 linkage region [Schwab et al., 1995; Straub et al., 1995; Wang et al., 1995; Lewis et al., 2003] and a prominent candidate gene demanding investigation in our sample. Here we report the results of a comprehensive analysis of DTNBP1 variation in a case–control association study design. Our aims were to: (i) Derive a comprehensive set of tagSNPs (tSNPs) capturing common genetic variation in DTNBP1, and assess their analytical performance; (ii) Examine the evidence for association between DTNBP1 variants and schizophrenia, defined as a clinical diagnostic category; and (iii) Search for association between DTNBP1 variants and a composite neurocognitive endophenotype, incorporating measures of performance in multiple neurocognitive domains. MATERIALS AND METHODS Subjects The study included a total of 508 individuals (371 male and 137 female) of whom over 75% are of Anglo-Irish descent. Schizophrenia (SCZ) cases (N ¼ 336, 269 male) included the 93 probands from our previously reported family study [Hallmayer et al., 2005], whereas the remainder were newly recruited subjects. The 172 controls (102 male) were recruited by random sampling from local telephone directories, or among Red Cross blood donors, with screening for psychopathology used to exclude individuals with previous diagnosis of psychotic illness in themselves or in any of their first-degree relatives. Written informed consent was obtained from all participating subjects. The study was approved by the Human Research Ethics Committee of The University of Western Australia and the North Metropolitan Health Area Ethics Committee, Perth, Western Australia. Phenotype Characterization Clinical diagnosis of affected subjects was based on a videorecorded structured interview with the use of the Schedules for Clinical Assessment in Neuropsychiatry (SCAN), version 2.0 [Wing et al., 1990], a review of case records, and a structured developmental history obtained from a key family member. Research diagnoses were established by consensus between two senior clinicians, who reviewed independently the entire diagnostic information, including the videotape of the SCAN interview, and assigned ICD-10 and DSM-IV lifetime diagnoses. All participants (including controls) were administered a battery of tests assessing several domains of neurocognitive function, as described in detail [Hallmayer et al., 2005]. In brief, these included: General cognitive ability: prior or premorbid IQ (the National Adult Reading Test) and current IQ (the Shipley Institute of Living Scale); Sustained attention: visual Continuous Performance Task, degraded-stimulus version (involving an increased demand on visual encoding) and identical-pairs version (selectively engaging working memory); Executive function: The FAS version of the Controlled Oral Word Association Task (assessing effortful retrieval of stored lexical knowledge); Verbal memory: the Rey Auditory Verbal Learning Test (assessing immediate and delayed recall of word lists, retention after distraction, and errors, generating an index of encoding in verbal memory); Speed of information processing: Inspection Time task (measuring perceptual encoding and speed of processing, unconfounded by motor-reaction time); Neurobehavioral features: a structured examination of soft (non-localizing) neurological signs and the Edinburgh Handedness Inventory of behavioral lateralization. A form of latent structure analysis, known as grade of membership (GoM) analysis [Woodbury et al., 1978; Manton et al., 1994] was used to integrate the multiple cognitive measures into a parsimonious number of ‘‘pure types’’ and assign, to each subject, scores of affinity to each one of several pure types. This approach identified two distinct neurocognitive profiles, which comprised >90% of the schizophrenia patients and >23% of their unaffected first-degree relatives: a cognitive deficit (CD) subtype and a cognitively spared (CS) subtype. Generalized cognitive deficiency was the most salient characteristic of the CD subtype, in contrast to mild or patchy deficits in the CS subtype. Power of the Study Power to detect association was determined for each of our data sets using the Genetic Power Calculator [Purcell et al., 2003]. For our SCZ/control sample, detectable odds ratios for the homozygote ranged from 2.7 to 2.15, for allele frequencies between 0.1 and 0.4. For the CD/control sample, these odds ratios ranged from 3 to 2.45, and for the CS/control dataset 3.2– 2.6, for the same allele frequency range. These effect sizes are consistent with those reported in previous DTNBP1 association studies [Schwab et al., 2003; Van den Bogaert et al., 2003; Van den Oord et al., 2003; Funke et al., 2004; Numakawa et al., 2004; Williams et al., 2004]. Selection of Polymorphic Markers SNPs were selected to represent the common genetic variation in the DTNBP1 genomic structure and 10 kb upstream and downstream of the gene. We used HapMap phase II-listed variants with a minor allele frequency >5%, augmented with coding SNPs from dbSNP, and SNPs previously reported in association with schizophrenia. TagSNPs (tSNPs) were identified using the pair-wise option of Tagger [de Bakker et al., 2005] implemented in Haploview No Association Between DTNBP1 and Schizophrenia 2 [Barrett et al., 2005], with a threshold of r > 0.8. High Illumina designability scores served as an additional criterion for the assembly of the final SNP panel. Genotyping and Primary Data Analysis Genotyping was performed on an Illumina BeadStation using the GoldenGate technology. DNA samples from CEPH trio 1334 (obtained from the Coriell Cell Repository) served as internal controls for quality of clustering and reproducibility. The primary analysis of the genotyping data with the Illumina BeadStudio software was followed by visual inspection and assessment of data quality and clustering. Statistical Analyses Deviations from Hardy–Weinberg equilibrium (P < 0.001) were examined using PLINK [Purcell et al., 2007]. We used Haploview [Barrett et al., 2005] to determine the residual LD between the selected tSNPs and characterize in our control population the haplotype block structure [Gabriel et al., 2002] of DTNBP1 and immediate flanking sequence. Association of DTNBP1 sequence variants with clinically diagnosed schizophrenia and with neurocognitive endophenotypes was analyzed using generalized linear modeling, implemented in the statistical program SimHap [McCaskie et al., 2006]. The primary dichotomous outcome variables were SCZ, CD, and CS, in the respective case/control sample. The principal explanatory variables were the genotyped sequence variants in DTNBP1. The SNP genotypes were coded into three classes (0 ¼ major allele homozygote, 1 ¼ heterozygote, 2 ¼ minor allele homozygote) and analyzed as a linear (additive, gene-dosage) covariate. Additionally, each SNP was analyzed as a factored, or codominant, effect, where the heterozygotes and homozygotes were explicitly modeled; in instances where there were less than five homozygotes in cases or controls, the SNP in question was modeled as a dominant effect only. We corrected for the multiple testing inherent in this study using the Bonferroni method [Armitage et al., 2002]. Haplotypes were inferred for individuals with ambiguous phase and haplotype frequencies were estimated using an expectation–maximization algorithm as implemented in SimHap [McCaskie et al., 2006]. Haplotypes were recoded as independent factors into three classes (0, 1, or 2), representing the number of copies of each haplotype in an individual’s diplotype. In each analysis, we performed a global haplotype test, where each haplotype is modeled as an additive effect and examined relative to the same (most abundant) baseline haplotype. Individual haplotypes of interest were examined in haplotype-specific tests (under an additive model). RESULTS The study included a total of 508 subjects, of whom 336 were clinically affected, meeting ICD-10 and DSM-IV diagnostic criteria for schizophrenia. Using a battery of neurocognitive tests, 155 of the affected individuals were assigned to the cognitive deficit (CD) subtype of schizophrenia, and 121 to the cognitively spared (CS) subtype [Hallmayer et al., 2005]. The remainder fell into two residual (non-CD/non-CS) subtypes: elderly individuals whose cognitive deficit pattern could be age-related, and a small number of affected subjects with cognitive function indistinguishable from controls. From this classification, we constructed three case/control samples for analysis: schizophrenia (SCZ)/controls, CD/controls, and CS/controls. The non-CD/non-CS subjects were included only in the SCZ/controls sample. Our analytical approach aimed at comprehensive coverage of the genomic DTNBP1 sequence and of 10 kb upstream and 1161 downstream of the gene. Using resources and criteria likely to be applied in most future studies (please see Materials and Methods Section), we constructed a panel of 39 tag SNPs (tSNPs), capturing a total of 151 common variants across Dysbindin (Supplementary Material Table I). The resulting map (Supplementary Material Table I and Fig. 1) shows some differences with the map proposed in a recent comprehensive study of normal DTNBP1 variation [Mutsuddi et al., 2006], due to different selection criteria and genotyping technology (e.g., 18/42 tSNPs selected in the Mutsuddi et al. study  are not included in our set: 8 of these are captured by our tSNPs, 4 have MAF < 5%, and 6 are not listed in HapMap). The 37 kb region of DTNBP1 (50 region of the gene to IVS4) where a cluster of SNPs have been studied and found positive, in different combinations, in European populations [Schwab et al., 2003; Van den Bogaert et al., 2003; Van den Oord et al., 2003; Funke et al., 2004; Kirov et al., 2004; Bray et al., 2005; Gornick et al., 2005; Mutsuddi et al., 2006], was covered in our study by a total of 12 tSNPs (Fig. 1, Supplementary Material Table I): rs3213207, rs2619545, rs1011313, rs2619547, rs2619528, rs2619522, rs1018381, rs1997679, rs909706, rs9476886, rs2743852, and rs2619538 (ordered here by chromosomal position). Markers rs2619547, rs909706, and rs2743852 are not listed in HapMap, but were included in our study based on previous evidence of association [Funke et al., 2004; Williams et al., 2004; Gornick et al., 2005]. Three further SNPs, previously analyzed and found in association, namely rs2005976, rs760761, and rs1474605 [Van den Bogaert et al., 2003; Van den Oord et al., 2003], were not included in our set of tSNPs. The markers rs2005976 and rs760761 could not be included due to incompatibility with the genotyping technology used (Illumina designability scores of 0). All three markers are captured by other SNPs genotyped in our study: rs2005976 has been shown previously to be in complete LD with rs2619528 [Van den Oord et al., 2003], and rs760761 and rs1474605 is captured by rs2619545 (r2 ¼ 1). Genotyping using the Illumina GoldenGate technology resulted in an average call rate of 99.9%. Eight samples with a call rate <98% were removed from further analysis. Marker rs11558324 [Gornick et al., 2005], tagging only itself, showed poor genotyping performance and was also removed. In addition to the standard analysis with the Illumina BeadStudio software, data quality was assessed by visual inspection. SNP rs3213207, the marker most consistently analyzed in DTNBP1 association studies in Europeans [Schwab et al., 2003; Van den Bogaert et al., 2003; Van den Oord et al., 2003; Funke et al., 2004; Kirov et al., 2004; Bray et al., 2005], produced ambiguous results in the Illumina genotyping. While the BeadStudio Genotyping Module software gave it a high Gentrain score (0.754), visual inspection revealed two heterozygote clusters, possibly resulting from the presence of another polymorphic position in the same DNA fragment. To test this possibility and obtain unequivocal genotype assignments, we sequenced the fragment around rs3213207 in DNA samples representing the four clusters. This analysis confirmed our assumption, detecting the presence of rs12527496, located 20 bp away from our target SNP. It also confirmed the BeadStudio genotype assignment, with samples in both heterozygote clusters shown to be heterozygous for the targeted SNP rs3213207, but different in their rs12527496 genotypes. In our design, rs12527496 was captured by rs3829893, and sequencing analysis confirmed the genotypes predicted on the basis of LD. After data cleaning, the final sample for statistical analysis comprised 329 SCZ cases, including 152 CD and 119 CS, and 171 controls. The final number of tSNPs was 38, capturing 150 variants. No deviation from Hardy–Weinberg equilibrium, examined using PLINK [Purcell et al., 2007], was observed for any of the 38 tSNPs (P > 0.009) (Supplementary 1162 Peters et al. Fig. 1. Coverage and LD structure of the 37 kb 50 region of DTNBP1 previously reported to be associated with schizophrenia in European populations. The order of markers left to right follows nucleotide positions along the chromosome, with the DTNBP1 genomic structure (on minus strand) shown above. The 37 kb spans the region upstream of DTNBP1 to IVS4. Based on HapMap phase II-listed SNPs, MAF > 5%, and r2 > 0.8, the region was covered by the 12 tSNPs shown above. D0 values were calculated from the genotyping data obtained in the 171 control subjects, using Haploview [Barrett et al., 2005]. The region is split into three haplotype blocks, with Block 2 comprising most previously associated markers (referred to SNPs 4–8 by Mutsuddi et al. ). [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.] Material Table IIA). The minor allele frequencies observed in our control sample were very similar to those reported in HapMap (Supplementary Material Table IIA). The haplotype block structure [Gabriel et al., 2002] of DTNBP1 in our control population, as defined by Haploview [Barrett et al., 2005], comprised eight small blocks of up to 70 kb (Supplementary Material Fig. 1). Blocks 3, 4, and 5 were located within the large (122 kb, spanning exons 4–8) block which is shown in HapMap, but was not present in our control sample of 171 subjects. In the 37 kb spanning the region 50 of DTNBP1 to IVS4, where most previous association findings in Europeans cluster [Schwab et al., 2003; Van den Bogaert et al., 2003; Van den Oord et al., 2003; Funke et al., 2004; Kirov et al., 2004; Bray et al., 2005], we observed three haplotype blocks of 4 kb (IVS4), 23 kb (IVS4 to IVS1), and 1 kb (IVS1; Fig. 1). The 23 kb block is in the center of the previously associated region; it is formed by four of our tSNPs (rs2619547, rs2619528, rs2619522, and rs1018381), capturing two additional SNPs (rs2005976 and rs760761) that have been analyzed in previous studies [Schwab et al., 2003; Van den Bogaert et al., 2003; Van den Oord et al., 2003; Funke et al., 2004]. Within that region, Mutsuddi et al.  selected six SNPs—rs3213207, rs1011313, rs2005976, rs760761, rs1018381, and rs2619538 (referred to as tSNP 2, 3, 5, 6, 8, and 11 respectively in Table III and Figure 1 of Mutsuddi et al. ), representative of the associated markers and haplotypes in European populations. Mutsuddi et al.  derived common haplotypes and estimated their frequencies in the CEU sample. Analysis of the frequencies of these five haplotypes in our control population showed similar values (Fig. 2, panel A), except for haplotype 1 which was marginally less common than haplotype 2. The additional SNPs genotyped in our sample led to further differentiation, generating 11 haplotypes with frequencies >1% (Fig. 2, panel B). We also examined the 8-marker haplotypes identified in the study of Irish high-density schizophrenia families, where association No Association Between DTNBP1 and Schizophrenia 1163 Fig. 2. Haplotypes in the 37 kb 50 region of DTNBP1 in our control sample and in previous studies of European populations. LD analysis (Fig. 1) revealed three blocks in the region, therefore the data in this figure do not represent non-recombinant ancestral haplotypes but are shown for the purpose of comparing population frequencies. SNPs with genotypes inferred based on LD are shown in brackets. * The Illumina assays for these SNPs target the opposite strand; to allow comparisons with previously reported data, we present the reverse complement alleles. A: 6-marker haplotypes as constructed by Mutsuddi et al. . Phylogenetic tree of the common haplotypes as shown in Figure 1 of the Mutsuddi et al. study . The haplotype frequencies at the bottom have been obtained from the analysis of 30 CEU trios [Mutsuddi et al., 2006] (in black) and of the 171 unrelated control subjects in the present study (in red). B: Further haplotype differentiation in this study. The genotyping of additional SNPs results in the differentiation of the two most common 6-marker haplotypes. The color coding of derivative haplotypes follows that shown in panel A, and the ordering is based on the observed frequencies. The alleles defining the 6-marker haplotypes are in bold. C: 8-marker haplotypes analyzed in the Irish study of high-density schizophrenia families [Van den Oord et al., 2003; Riley and Kendler, 2006]. The haplotype frequencies are very similar in our control population, in the Irish schizophrenia families [Van den Oord et al., 2003], and in the CEU trios [Mutsuddi et al., 2006]. Haplotype 2 corresponds to the Irish high-risk haplotype [Van den Oord et al., 2003; Riley and Kendler, 2006]. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.] with Dysbindin was first reported [Van den Oord et al., 2003; Riley and Kendler, 2006]. The frequencies observed in our control sample of predominantly Anglo-Irish ancestry were very similar to those observed in the Irish families [Van den Oord et al., 2003; Riley and Kendler, 2006], as well as to those in the CEU sample [Mutsuddi et al., 2006] (Fig. 2, panel C). After establishing this framework, we analyzed the association of DTNBP1 sequence variants with clinically diagnosed schizophrenia and with neurocognitive endophenotypes. All results, reported as odds ratios and 95% confidence limits, are shown in detail in Supplementary Material Table II. None of the 38 tSNPs tested under an additive genetic model was significantly associated with SCZ (P > 0.19), the CD (P > 0.20) or the CS subtype (P > 0.23). Under a codominant model, SNPs rs2619545 and rs1997679 (both located within the previously reported associated region) were nominally associated with SCZ (P ¼ 0.05 and 0.03, respectively), but were non-significant after Bonferroni correction for multiple testing (P ¼ 0.86 and 1164 Peters et al. 0.69, respectively). None of the rare SNPs analyzed under a dominant (instead of codominant) model was significantly associated with any of the outcomes (P > 0.18). The results of the haplotype-based association analyses are presented in Supplementary Material Table III. We performed the global haplotype test for association with the SCZ, CD, and CS phenotype using the haplotypes in the eight observed LD blocks (as shown in Supplementary Material Fig. 1) across DTNBP1. Significant association was observed for only the SCZ phenotype with block 1 (P ¼ 0.03), but was non-significant after Bonferroni correction for multiple testing (P ¼ 0.22). To obtain data comparable to previous studies, we also performed the global haplotype test with the 6-marker haplotypes defined by Mutsuddi et al.  (including our genotyped tSNPs rs3213207, rs1011313, rs1018381 and rs2619538, and inferred rs2005976 and rs760761). No significant results were obtained in this analysis: P ¼ 0.72, P ¼ 0.42, and P ¼ 0.41 respectively for SCZ, CD, and CS. Finally, based on the similar ethnic composition of the two samples, we specifically tested the risk haplotype G-G-A-A-T-G-C-G identified in the Irish schizophrenia families [Van den Oord et al., 2003; Riley and Kendler, 2006] (haplotype 2 in Fig. 2, panel C) for association with schizophrenia or its subtypes in our sample. No significant association was found in the haplotype-specific test, between this haplotype relative to all other haplotypes, and either SCZ, CD, or CS (all P > 0.14). DISCUSSION In this study, based on our previously identified neurocognitive subtypes of schizophrenia and linkage to the short arm of chromosome 6 [Hallmayer et al., 2005], we aimed at comprehensive analysis of variation in DTNBP1 and its association with the disease phenotypes. We relied on widely used marker selection criteria and genotyping technology to develop and test an analytical framework that would be applicable to future studies. Our selection of markers captures common genetic variants across the gene and its flanking regions, and results in genotyping data of good quality. While some of the associated SNPs reported in the literature are not directly analyzable using the Illumina GoldenGate technology, they can be captured reliably by tSNPs at very high r2 and D0 values. The allele frequencies and LD pattern observed in our control sample are in agreement with those reported in public databases and in a recent detailed study of DTNBP1 [Mutsuddi et al., 2006]. An exception is the large LD block across exons 4– 8, present in HapMap, which is broken in our larger control population whose size is larger than the CEU sample. In the 50 region of DTNBP1, harboring most of the reported associated markers and haplotypes, the LD structure in our sample resembled closely that in the CEU trios [Mutsuddi et al., 2006]. Haplotype frequencies within the blocks, as well as extending beyond to cover the commonly associated markers, were in the range observed in other European populations [Schwab et al., 2003; Van den Oord et al., 2003; Mutsuddi et al., 2006], with further differentiation of the two most common haplotypes resulting from the inclusion of additional SNPs. Our systematic analysis of association between Dysbindin sequence variants and schizophrenia produced entirely negative results. We found no evidence of association with any of the polymorphic markers across the entire gene sequence and flanking regions or with any of the haplotypes across the gene or within the previously implicated 50 region. Specific examination of the risk haplotype found in the Irish high-density family sample [Van den Oord et al., 2003; Riley and Kendler, 2006] also failed to reveal any association in our ethnically closely related sample. A number of studies have focused on the association of Dysbindin with specific cognitive performance tasks in schizophrenia patients and normal subjects. Although many results are positive, their interpretation is problematic. Burdick et al.  found association in schizophrenia patients and normal controls between a DTNBP1 haplotype and a factor score for general cognitive ability (g) derived by principal component analysis of six cognitive measures. The same DTNBP1 haplotype was also associated with 13.5 IQ points decline from estimated premorbid to current IQ [Burdick et al., 2007]. A different risk haplotype was reported to be associated with impaired prefrontal brain function [Fallgatter et al., 2006] in 48 healthy German controls and with better intellectual functioning in a Dutch sample of 52 nuclear families and 31 controls [Zinkstok et al., 2007]. Still another haplotype was found to affect spatial working memory and explain 12% of its variance in Irish schizophrenia patients [Donohoe et al., 2007], while at the same time leading to better performance in measures of executive function [Donohoe et al., 2007]. Apart from the diversity of risk haplotypes which mirrors the general pattern of inconsistency in DTNBP1 association findings, the difficulties in interpreting these results are compounded by the use of diverse measures and conflicting findings in the assessment of neurocognitive function, failing to outline a consistent cognitive profile associated with DTNBP1. In addition to a meticulously characterized clinical sample, we also examined DTNBP1 association with schizophrenia endophenotypes derived from comprehensive neurocognitive testing and different measures of cognitive function. As most neurocognitive tasks typically engage several component processes, our endophenotype approach relies on searching for multivariate patterns of dysfunction, rather than for single isolated deficits. Our definition of the CD subtype incorporates concurrent deficits in most of the neurocognitive domains measured in the above studies. One should also note that this subtype has been shown to be more homogeneous and heritable than either the clinical category of schizophrenia or the CS subtype, and to account specifically for our 6p linkage findings [Hallmayer et al., 2005]. The results of the present study indicate that the cognitive deficit subtype of schizophrenia is not associated with variation in DTNBP1. The genetic data implicating DTNBP1 in susceptibility to schizophrenia have triggered investigations into the neuronal functions of Dysbindin. While important functions have emerged [Numakawa et al., 2004; Weickert et al., 2004; Kumamoto et al., 2006; Talbot et al., 2006; Camargo et al., 2007], the genetic evidence remains controversial. Our analytical framework and the results of this study may facilitate further attempts at systematic re-analysis of the involvement of DTNBP1 in schizophrenia susceptibility. In addition to the need for a uniform comprehensive approach to association analyses [Mutsuddi et al., 2006; NCI-NHGRI Working Group on Replication in Association Studies, 2007], we emphasize the need for a standardized testing of cognitive function, such as that adopted by the Consortium on the Genetics of Schizophrenia [Gur et al., 2007] and recently by the Australian Schizophrenia Research Bank. ACKNOWLEDGMENTS We thank all subjects participating in this study, Dr. Brenda Powell and Dr. Pamela McCaskie for help with the SNP panel design and the use of SimHap, and Padma Sivadorai for expert technical assistance. The project was supported by NHMRC grants 404046 and 404025. LJP was supported by the WindOver-Water Foundation. REFERENCES Armitage P, Berry G, Matthews JNS. 2002. Statistical methods in medical research. Clayton: Blackwell Publishing. 368 p. No Association Between DTNBP1 and Schizophrenia Barrett JC, Fry B, Maller J, Daly MJ. 2005. Haploview: Analysis and visualization of LD and haplotype maps. Bioinformatics 21:263–265. Bray NJ, Preece A, Williams NM, Moskvina V, Buckland PR, Owen MJ, O’Donovan MC. 2005. Haplotypes at the dystrobrevin binding protein 1 (DTNBP1) gene locus mediate risk for schizophrenia through reduced DTNBP1 expression. Hum Mol Genet 14:1947–1954. Burdick KE, Lencz T, Funke B, Finn CT, Szeszko PR, Kane JM, Kucherlapati R, Malhotra AK. 2006. Genetic variation in DTNBP1 influences general cognitive ability. Hum Mol Genet 15:1563–1568. Burdick KE, Goldberg TE, Funke B, Bates JA, Lencz T, Kucherlapati R, Malhotra AK. 2007. DTNBP1 genotype influences cognitive decline in schizophrenia. Schizophr Res 89:169–172. Camargo LM, Collura V, Rain JC, Mizuguchi K, Hermjakob H, Kerrien S, Bonnert TP, Whiting PJ, Brandon NJ. 2007. Disrupted in schizophrenia 1 interactome: Evidence for the close connectivity of risk genes and a potential synaptic basis for schizophrenia. Mol Psychiatry 12:74–86. de Bakker PI, Yelensky R, Pe’er I, Gabriel SB, Daly MJ, Altshuler D. 2005. Efficiency and power in genetic association studies. Nat Genet 37:1217– 1223. 1165 Lewis CM, Levinson DF, Wise LH, DeLisi LE, Straub RE, Hovatta I, Williams NM, Schwab SG, Pulver AE, Faraone SV, Brzustowicz LM, Kaufmann CA, Garver DL, Gurling HM, Lindholm E, Coon H, Moises HW, Byerley W, Shaw SH, Mesen A, Sherrington R, O’Neill FA, Walsh D, Kendler KS, Ekelund J, Paunio T, Lönnqvist J, Peltonen L, O’Donovan MC, Owen MJ, Wildenauer DB, Maier W, Nestadt G, Blouin JL, Antonarakis SE, Mowry BJ, Silverman JM, Crowe RR, Cloninger CR, Tsuang MT, Malaspina D, Harkavy-Friedman JM, Svrakic DM, Bassett AS, Holcomb J, Kalsi G, McQuillin A, Brynjolfson J, Sigmundsson T, Petursson H, Jazin E, Zoëga T, Helgason T. 2003. Genome scan metaanalysis of schizophrenia and bipolar disorder, Part II: Schizophrenia. Am J Hum Genet 73:34–48. Li D, He L. 2007. Association study between the dystrobrevin binding protein 1 gene (DTNBP1) and schizophrenia: A meta-analysis. Schizophr Res 96:112–118. Manton KG, Korten A, Woodbury MA, Anker M, Jablensky A. 1994. Symptom profiles of psychiatric disorders based on graded disease classes: An illustration using data from the WHO International Pilot Study of Schizophrenia. Psychol Med 24:133–144. DeRosse P, Funke B, Burdick KE, Lencz T, Ekholm JM, Kane JM, Kucherlapati R, Malhotra AK. 2006. Dysbindin genotype and negative symptoms in schizophrenia. Am J Psychiatry 163:532–534. McCaskie PA, Cadby G, Hung J, McQuillan BM, Chapman CM, Carter KW, Thompson PL, Palmer LJ, Beilby JP. 2006. The C-480T hepatic lipase polymorphism is associated with HDL-C but not with risk of coronary heart disease. Clin Genet 70:114–121. Donohoe G, Morris DW, Clarke S, McGhee KA, Schwaiger S, Nangle JM, Garavan H, Robertson IH, Gill M, Corvin A. 2007. Variance in neurocognitive performance is associated with dysbindin-1 in schizophrenia: A preliminary study. Neuropsychologia 45:454–458. Mutsuddi M, Morris DW, Waggoner SG, Daly MJ, Scolnick EM, Sklar P. 2006. Analysis of high-resolution HapMap of DTNBP1 (Dysbindin) suggests no consistency between reported common variant associations and schizophrenia. Am J Hum Genet 79:903–909. Duan J, Martinez M, Sanders AR, Hou C, Burrell GJ, Krasner AJ, Schwartz DB, Gejman PV. 2007. DTNBP1 (Dystrobrevin binding protein 1) and schizophrenia: Association evidence in the 30 end of the gene. Hum Hered 64:97–106. NCI-NHGRI Working Group on Replication in Association Studies. 2007. Replicating genotype-phenotype associations. Nature 447:655–660. Fallgatter AJ, Herrmann MJ, Hohoff C, Ehlis AC, Jarczok TA, Freitag CM, Deckert J. 2006. DTNBP1 (Dysbindin) gene variants modulate prefrontal brain function in healthy individuals. Neuropsychopharmacology 31:2002–2010. Fanous AH, Van den Oord EJ, Riley BP, Aggen SH, Neale MC, O’Neill FA, Walsh D, Kendler KS. 2005. Relationship between a high-risk haplotype in the DTNBP1 (dysbindin) gene and clinical features of schizophrenia. Am J Psychiatry 162:1824–1832. Funke B, Finn CT, Plocik AM, Lake S, DeRosse P, Kane JM, Kucherlapati R, Malhotra AK. 2004. Association of the DTNBP1 locus with schizophrenia in a U.S. population. Am J Hum Genet 75:891–898. Gabriel SB, Schaffner SF, Nguyen H, Moore JM, Roy J, Blumenstiel B, Higgins J, DeFelice M, Lochner A, Faggart M, Liu-Cordero SN, Rotimi C, Adeyemo A, Cooper R, Ward R, Lander ES, Daly MJ, Altshuler D. 2002. The structure of haplotype blocks in the human genome. Science 296:2225–2229. Gornick MC, Addington AM, Sporn A, Gogtay N, Greenstein D, Lenane M, Gochman P, Ordonez A, Balkissoon R, Vakkalanka R, Weinberger DR, Rapoport JL, Straub RE. 2005. Dysbindin (DTNBP1, 6p22.3) is associated with childhood-onset psychosis and endophenotypes measured by the premorbid adjustment scale (PAS). J Autism Dev Disord 35:831–838. Gur RE, Calkins ME, Gur RC, Horan WP, Nuechterlein KH, Seidman LJ, Stone WS. 2007. The consortium on the genetics of schizophrenia: Neurocognitive endophenotypes. Schizophr Bull 33:49–68. Hallmayer JF, Kalaydjieva L, Badcock J, Dragovic M, Howell S, Michie PT, Rock D, Vile D, Williams R, Corder EH, Hollingsworth K, Jablensky A. 2005. Genetic evidence for a distinct subtype of schizophrenia characterized by pervasive cognitive deficit. Am J Hum Genet 77:468–476. Harrison PJ, Weinberger DR. 2005. Schizophrenia genes, gene expression, and neuropathology: On the matter of their convergence. Mol Psychiatry 10:40–68. Karayiorgou M, Gogos JA. 2006. Schizophrenia genetics: Uncovering positional candidate genes. Eur J Hum Genet 14:512–519. Kirov G, Ivanov D, Williams NM, Preece A, Nikolov I, Milev R, Koleva S, Dimitrova A, Toncheva D, O’Donovan MC, Owen MJ. 2004. Strong evidence for association between the dystrobrevin binding protein 1 gene (DTNBP1) and schizophrenia in 488 parent-offspring trios from Bulgaria. Biol Psychiatry 55:971–975. Kumamoto N, Matsuzaki S, Inoue K, Hattori T, Shimizu S, Hashimoto R, Yamatodani A, Katayama T, Tohyama M. 2006. Hyperactivation of midbrain dopaminergic system in schizophrenia could be attributed to the down-regulation of dysbindin. Biochem Biophys Res Commun 345:904–909. Numakawa T, Yagasaki Y, Ishimoto T, Okada T, Suzuki T, Iwata N, Ozaki N, Taguchi T, Tatsumi M, Kamijima K, Straub RE, Weinberger DR, Kunugi H, Hashimoto R. 2004. Evidence of novel neuronal functions of dysbindin, a susceptibility gene for schizophrenia. Hum Mol Genet 13:2699–2708. Owen MJ, Craddock N, O’Donovan MC. 2005. Schizophrenia: Genes at last? Trends Genet 21:518–525. Purcell S, Cherny SS, Sham PC. 2003. Genetic Power Calculator: Design of linkage and association genetic mapping studies of complex traits. Bioinformatics 19:149–150. Purcell S, Neale B, Todd-Brown K, Thomas L, Ferreira MAR, Bender D, Maller J, Sklar P, de Bakker PIW, Daly MJ, Sham PC. 2007. PLINK: A tool set for whole-genome association and population-based linkage analyses. Am J Hum Genet 81:559–575. Riley B, Kendler KS. 2006. Molecular genetic studies of schizophrenia. Eur J Hum Genet 14:669–680. Schwab SG, Albus M, Hallmayer J, Honig S, Borrmann M, Lichtermann D, Ebstein RP, Ackenheil M, Lerer B, Risch N, Maier W, Wildenauer DB. 1995. Evaluation of a susceptibility gene for schizophrenia on chromosome 6p by multipoint affected sib-pair linkage analysis. Nat Genet 11:325–327. Schwab SG, Knapp M, Mondabon S, Hallmayer J, Borrmann-Hassenbach M, Albus M, Lerer B, Rietschel M, Trixler M, Maier W, Wildenauer DB. 2003. Support for association of schizophrenia with genetic variation in the 6p22.3 gene, dysbindin, in sib-pair families with linkage and in an additional sample of triad families. Am J Hum Genet 72:185–190. Straub RE, MacLean CJ, O’Neill FA, Burke J, Murphy B, Duke F, Shinkwin R, Webb BT, Zhang J, Walsh D, Kendler KS. 1995. A potential vulnerability locus for schizophrenia on chromosome 6p24-22: Evidence of genetic heterogeneity. Nat Genet 11:287–293. Straub RE, Jiang Y, MacLean CJ, Ma Y, Webb BT, Myakishev MV, HarrisKerr C, Wormley B, Sadek H, Kadambi B, Cesare AJ, Gibberman A, Wang X, O’Neill FA, Walsh D, Kendler KS. 2002. Genetic variation in the 6p22.3 gene DTNBP1, the human ortholog of the mouse dysbindin gene, is associated with schizophrenia. Am J Hum Genet 71:337–348. Talbot K, Cho DS, Ong WY, Benson MA, Han LY, Kazi HA, Kamins J, Hahn CG, Blake DJ, Arnold SE. 2006. Dysbindin-1 is a synaptic and microtubular protein that binds brain snapin. Hum Mol Genet 15: 3041–3054. Tosato S, Ruggeri M, Bonetto C, Bertani M, Marrella G, Lasalvia A, Cristofalo D, Aprili G, Tansella M, Dazzan P, Diforti M, Murray RM, Collier DA. 2007. Association study of dysbindin gene with clinical and outcome measures in a representative cohort of Italian schizophrenic patients. Am J Med Genet Part B 144B:647–659. Turunen JA, Peltonen JO, Pietilainen OP, Hennah W, Loukola A, Paunio T, Silander K, Ekelund J, Varilo T, Partonen T, Lönnqvist J, Peltonen L. 1166 Peters et al. 2007. The role of DTNBP1, NRG1, and AKT1 in the genetics of schizophrenia in Finland. Schizophr Res 91:27–36. Van den Bogaert A, Schumacher J, Schulze TG, Otte AC, Ohlraun S, Kovalenko S, Becker T, Freudenberg J, Jonsson EG, Mattila-Evenden M, Sedvall GC, Czerski PM, Kapelski P, Hauser J, Maier W, Rietschel M, Propping P, Nöthen MM, Cichon S. 2003. The DTNBP1 (dysbindin) gene contributes to schizophrenia, depending on family history of the disease. Am J Hum Genet 73:1438–1443. Van den Oord EJCG, Sullivan PF, Jiang Y, Walsh D, O’Neil FA, Kendler KS, Riley BP. 2003. Identification of a high-risk haplotype for the dystrobrevin binding protein 1 (DTNBP1) gene in the Irish study of high-density schizophrenia families. Mol Psychiatry 8:499–510. Vilella E, Costas J, Sanjuan J, Guitart M, De Diego Y, Carracedo A, Martorell L, Valero J, Labad A, De Frutos R, Nájera C, Moltó MD, Toirac I, Guillamat R, Brunet A, Vallès V, Pérez L, Leon M, de Fonseca FR, Phillips C, Torres M. 2007. Association of schizophrenia with DTNBP1 but not with DAO, DAOA, NRG1 and RGS4 nor their genetic interaction. J Psychiatr Res 42:278–288. Wang S, Sun CE, Walczak CA, Ziegle JS, Kipps BR, Goldin LR, Diehl SR. 1995. Evidence for a susceptibility locus for schizophrenia on chromosome 6pter-p22. Nat Genet 10:41–46. Weickert CS, Straub RE, McClintock BW, Matsumoto M, Hashimoto R, Hyde TM, Herman MM, Weinberger DR, Kleinman JE. 2004. Human dysbindin (DTNBP1) gene expression in normal brain and in schizophrenic prefrontal cortex and midbrain. Arch Gen Psychiatry 61:544– 555. Williams NM, Preece A, Morris DW, Spurlock G, Bray NJ, Stephens M, Norton N, Williams H, Clement M, Dwyer S, Curran C, Wilkinson J, Moskvina V, Waddington JL, Gill M, Corvin AP, Zammit S, Kirov G, Owen MJ, O’Donovan MC. 2004. Identification in 2 independent samples of a novel schizophrenia risk haplotype in the dystrobrevin binding protein gene (DTNBP1). Arch Gen Psychiatry 61:336–344. Wing JK, Babor T, Brugha T, Burke J, Cooper JE, Giel R, Jablensky A, Regier D, Sartorius N. 1990. SCAN: Schedules for clinical assessment in neuropsychiatry. Arch Gen Psychiatry 47:589–593. Woodbury MA, Clive J, Garson A Jr. 1978. Mathematical typology: A grade of membership technique for obtaining disease definition. Comput Biomed Res 11:277–298. Zinkstok JR, de Wilde O, Tanck MW, Baas F, Linszen DH. 2007. Association between the DTNBP1 gene and intelligence: A case-control study in young patients with schizophrenia and related disorders and unaffected siblings. Behav Brain Funct 3:19.