Oct 30, 2017 - Oral sessions: 09 Epigenetics; 10 Ethical Issues and Psychiatric Genetic Counselling; 11 Gene ...... The speakers in this session will discuss various approaches to finding ...... Venkata Mattay1, Daniel Weinberger1, Andreas Meyer-Lindenberg1 ...... cokinetic interaction was ranked (greatest to least) UMs >.
ß 2006 Wiley-Liss, Inc.
684
Abstracts
Programme Outline SATURDAY, OCTOBER 28th 2006 09:00–19:00 10:30–12:30 12:30–15:00 15:00–16:00 16:00–17:00 16:00–17:00 17:30–19:00 19:00–20:00
Registration Round Table ‘‘Depression: Just a Cultural Factor?’’ ISPG Board Lunch 2007 Programme Committee Meeting 2008 Programme Committee Meeting Workshop 1. ‘‘The chromosome 8 in Bipolar Disorder’’ Opening Ceremony Plenary Lecture PL1: Genetics, a philosophy for the future
SUNDAY, OCTOBER 29th 2006 08:00–19:00 10:00–19:00 08:30–09:30 09:30–10:00 10:00–12:00 10:00–12:00 12:00–13:30 13:30–15:30 15:30–17:30 17:30–18:30
Registration Exhibitions Plenary Lecture PL 2 Genetics, a philosophy for the future: the molecular biology side ‘‘The Impact of Williams Syndrome Mutations on Neural Organization and Brain Function: A Window into Social Cognition’’ Break Symposium S1 Genetics and Brain Imaging Symposium S2 Do ADHD and Autism share a genetic basis? Lunch and Poster Sessions Oral sessions: 01 Affective Disorders (I); 02 Bipolar Disorder (I); 03 Schizophrenia (I); 04 Substance Abuse Oral sessions: 05 Candidate Genes (I); 06 Childhood Disorders (I); 07 Endophenotypes (I); 08 Pharmacogenetics/Genomics (I) Oral sessions: 09 Epigenetics; 10 Ethical Issues and Psychiatric Genetic Counselling; 11 Gene Expression (I); 12 Neurodegeneration
MONDAY, OCTOBER 30th 08:30–19:00 09:00–19:00 8:30–9:30 9.30–10:00 10:00–12:00 10:00–12:00 12:00–13:30 13:30–15:30 15:30–17:30 16:00–18:00 17:30–18:30
Registration Exhibitions Plenary Lecture PL 3: Genetics, a philosophy for the future: the philosopher side ‘‘What is a gene today?’’ Break Symposium S3 The Genetics of Bipolar Disorder Symposium S4 Implications of psychiatric genetic research: Ethical, legal, social, and psychological aspects Lunch and Poster Sessions Oral sessions : 13 Bipolar Disorder (II); 14 Childhood Disorders (II); 15 Endophenotypes (II); 16 Schizophrenia (II) Symposium S5 The Genetics of Schizophrenia Symposium S6 The Genetics of Depression Oral session: 17 Bioinformatics/population Genetics (I); 18 Eating Disorders ; 19 Pharmacogenetics/Genomics (II)
TUESDAY, OCTOBER 31st 08:30–19:00 09:00–19:00 08:30–09:30 09:30–10:00 10:00–12:00 10:00–12:00 12:00–13:30 13:30–14:30 14:30–15:30 15:30–16:30 16:30–19:00 19:00–20:00
Registration Exhibitions Workshop 2. ‘‘An Update on The knockout mouse project and EUCOMM insertional mutagenesis project’’ Break Symposium S7 Isolated populations: a valid tool to find genes in psychiatric diseases? Symposium S8 Drug and alcohol addiction: genetics and pharmacogenetics Lunch and Poster session Oral session: 20 Bioinformatics/population Genetics (II); 21 Gene Expression (II) Plenary Lecture PL 4 ‘‘The multiple rare-variants model of common disease’’ Workshop 3: Psychiatric Applications of Pharmacogenetics Testing: CYP450 2D6 (CYP2D6) and CYP450 2C19 (CYP2C19) Oral session: 22 Affective Disorders (II); 23 Candidate Genes (II); 24 Childhood Disorders (III); 25 Schizophrenia (III) ISPG Assembly
WEDNESDAY, NOVEMBER 1st 8:30–10:30 8:30–10:30 10:30–11:00 11.00–12:30 12:30–13:00 13:00
Symposium S9 Epigenetics and Chromatine Symposium S10 System Biology Break Oral sessions: 26 Affective Disorders (III); 27 Bioinformatics/population Genetics (III); 28 Candidate Genes (III); 29 Neurobiology/Animal Studies BEST POSTER AWARD Closing remarks
Abstracts PL2 THE IMPACT OF WILLIAMS SYNDROME MUTATIONS ON NEURAL ORGANIZATION AND BRAIN FUNCTION: A WINDOW INTO SOCIAL COGNITION J.R. Korenberger University of California Wlliams Syndrome (WS), due to a hemizygous deletion of chromosome band 7q11.23, features a relatively invariable suite of physical, cognitive and behavioral characteristics. We have studied sociability in WS individuals with typical and atypical deletions, combining fine structure genetic and behavior analyses to suggest genes whose altered expression contributes to altered behavior or neural activity as seen in functional MRI. The significance of this is that these individuals differ from each other by the deletion of only one to two genes. Their social cognition also differs, suggesting that these genes may be involved in mediating a subset of the social responses of WS. For example, we have reported an individual with an atypically small deletion in this region who exhibits some but not all WS characteristics. In particular, the almost invariable hypersociability is absent, while language development and visuospatial function are relatively normal. We conclude and suggest that genes normally deleted in WS but not in this patient are essential for normal social, language and visuospatial development. Thus from our human work, we have identified specific genes whose variation is associated with WS social behavior. By using these to explore gene expression in a non-human primate brain, the Macaque, we have elucidated a novel pathway that links the expression of WS genes to brain systems including the cingulated cortex and other limbic structures which in part constitute the ‘social brain’ of the Macaque. Our results provide dramatic evidence that modulation of these systems may underlie social behaviors in humans, and striking data on which to base the design and testing of essential models both in primates and in humans. That is, taken together, our human and macaque data support the definition of a neural system that may underlie significant variations in human social behavior. Moreover, our results suggest ways in which the pathways defined by these genes may also be responsible for social behavior and its derangement in the normal human population.
PL3 WHAT IS A GENE TODAY? P.E. Griffiths University of Queensland, Australia Historians of genetics agree that multiple conceptions of the gene have coexisted at each stages in the history of genetics and that the resulting partial ambiguity has often contributed to the success of genetics, both because workers in different areas have needed to communicate and to draw on one another’s results despite wrestled with very different scientific challenges, and because empirical findings have often challenged the presuppositions of existing conceptions of the gene. Today, a number of different conceptions of the gene coexist in the biosciences. An ‘instrumental’ gene similar to that of classical genetics retains a critical role in the construction and interpretation of experiments in which the relationship between genotype and phenotype is explored via hybridization between organisms or directly between nucleic acid molecules. It also plays an important theoretical role in the foundations of disciplines such as quantitative genetics and population genetics. A ‘nominal’ gene, defined by the practice of genetic nomenclature, is a critical practical tool and allows communication between bioscientists in a wide range of fields to be grounded in welldefined sequences of nucleotides. This concept, however, does not embody major theoretical insights into genome structure or function. Instead, a ‘post-genomic’ conception of the gene embodies the continuing project of understanding how genome structure supports genome function, but with a deflationary picture of the gene as a structural unit. This final concept of the gene poses a significant challenge to earlier assumptions about the relationship between genome structure and function, and between genotype and phenotype. PL4 THE MULTIPLE-RARE VARIANTS MODEL OF COMMON DISEASE E.S. Gershon University of Chicago The ‘‘multiple-rare-variants common-disease’’ hypothesis is a plausible explanation for recent findings of linkage and association that have not
685
led to consistent functional molecular variants. Testing of this hypothesis requires a conceptual reorientation for clinical genetics research. It is not statistically efficient to screen for rare variants using common haplotypes, composed of SNPs with minor allele frequency (MAF) > _ 0.05, as is now proposed for genome-wide association (GWA). Resequencing will be required to identify less common and rare variants, and direct association testing will be needed with these variants. Resequencing is not feasible for the entire genome, so a candidate gene strategy is mandated. Consider the dopamine and glutamate neurotransmission pathways. These two pathways are central to the most prominent current understandings of the biology and pharmacology of several major mental illnesses. Consideration of the two pathways implies at least 60 gene targets, including metabolic enzymes, transporters, receptors, and intracellular signaling molecules. A plausible research strategy would include resequencing of the known and presumed functional sequences of each gene, including extended sequencing of presumed promoter regions. These would identify new rare and uncommon variants, as well as common variants. Genetic association tests: For genes with an excess of rare variants evident in the resequencing, polymorphisms to be genotyped will include all rare variants. In addition, one could genotype rare but not singleton alleles (where a singleton allele has MAF < 0.001), newly discovered functional variants, and common variants not in LD with HapMap common variants. W1 8q24 BIPOLAR DISORDER SYMPOSIUM Chair: Melvin McInnis, University of Michigan Presenters: P. Zandi1, S. Zoellner2, F. McMahon3, J. Nurnberger4, J. Kelsoe5, S. Cichon6, 1Johns Hopkins University 2 University of Michigan 3 NIMH 4 Indiana University 5 University of California, San Diego 6 University of Bonn The chromosomal region surrounding 8q24 has been identified in linkage studies as highly likely to contain susceptibility genes for BP disorder. Cichon et al. (2001) reported a LOD of 3.62 in 75 German, Israeli and Italian families. McInnis et al. (2003) reported an NPL of 3.13 in 65 N American families of N European descent in an adjacent region of 8q24. A recent analysis (McQueen et al., 2005) of combined data that included the above families and consisted of 1,067 BP families, primarily of Caucasian descent, solidified 8q24 as one of the strongest and most consistent linkage finding to date. We have followed up on the linkage findings by typing 1,461 SNPs in 737 Bipolar families across 15 Mb (for an average density of 1 per every 11 Kb) on 8q24. These families include the waves 1–4 of the NIMH Genetics Initiative families and multiplex BP families ascertained and studied at the Johns Hopkins University. Of the genotyped subjects, 1,815 had either SABP, BPI, or BPII (574 had no genotyped parents, 643 had 1 genotyped parent, and 598 had 2 genotyped parents). SNP genotyping was completed by the Center for Inherited Disease Research (CIDR) using the Illumina platform. Initial analyses have been performed using FBAT. There are 3 gene containing regions with p < 10–3 and include the genes ADCY8, KCNQ3, and SIAT4a. There is a region in the telomeric area of this region that identifies a SNP, rs2978607, with evidence of association at the p < 10–5 level. This SNP is in a gene desert, the nearest genes being a provisional gene called KHDRBS3 (1.5 Mb centromeric) and COL22A1 (1.4 Mb telomeric). These results are to be considered preliminary due to the identification of improbable inheritance vectors in several SNPs and data curation is ongoing at the time of abstract submission. The data have been distributed to investigators within the collaborative of the NIMH Genetics Initiative for Bipolar disorder. At this chromosome 8q24 symposium results from the analyses from several collaborative groups will be presented, which will include different methodologies as well as analysis following stratification according to phenotype schemes. There will be an 8q24 presentation from the Bonn/Mannheim groups by Cichon et al., who have performed LD studies in two peak regions showing highest evidence for linkage of BP disorder and 8q24 in their genome scan of 2001. They genotyped a total of 400 SNPs between the coordinates 120,419,000 bp–124,925,000 bp (region 1) and 128,074,000–134,150,000 (region 2) (NCBI build 35) in a German BPI sample (400 cases, 400 controls) using Illumina’s GoldenGate Assays. Nominally significant SNPs from the German
686
Abstracts
sample are currently replicated in an independent sample of 400 BPI cases and 400 controls. Acknowledgements: Data and biomaterials were collected as part of the National Institute of Mental Health (NIMH) Bipolar Disorder Genetics Initiative. Principal Investigators and Co-Investigators: Indiana University, J Nurnberger, Miller MJ, Bowman ES, Rau L, Ryan Moe, Samavedy N, El-Mallakh R, (University of Louisville), Manji H (Wayne State University), Glitz DA ( Wayne State University), Foroud T, Dick D, Edenberg H; Washington University, St. Louis, Rice J, Reich T, Goate A, Bierut L; Johns Hopkins University, McInnis M, DePaulo R, MacKinnon D, Mondimore F, Potash J, Zandi P; University of Pennsylvania, Berrettini W; University of California at Irvine, Byerley W and Vawter M; University of Iowa, Coryell W and Crowe R; University of Chicago, Gershon E, Badner J, McMahon F, Liu C, Sanders A, Caserta M, Dinwiddie S; University of California at San Diego, Kelsoe J; Rush University, Scheftner W, Kravitz HM; NIMH Intramural Research Program, McMahon F, Kassem L, Detera-Wadleigh S Investigator from the German Bipolar Collaborative: S Cichon1, J Schumacher2, M Weingarten1, T Feulner1, T Becker3, T Schulze4, A Georgi4, A Van Den Bogaert5, W Maier6, M Schwarz7, P Propping2, M Rietschel4, M No¨then1, 1Department of Genomics, Life Brain Center, University of Bonn, Bonn, Germany, 2Institute of Human Genetics, University of Bonn, Bonn, Germany, 3Institute of Medical Biometry, Informatics, and Epidemiology (IMBIE), University of Bonn, Bonn, Germany, 4Central Institute of Mental Health, Mannheim, Germany, 5Department of Molecular Genetics, VIB8, University of Antwerp, Antwerp, Belgium, 6Department of Psychiatry, University of Bonn, Bonn, Germany, 7Nordbaden Center of Psychiatry, Wiesloch, Germany.
W2 AN UPDATE ON THE KNOCKOUT MOUSE PROJECT AND EUCOMM INSERTIONAL MUTAGENESIS PROJECT INTERNATIONAL EFFORT TO DISRUPT AND CREATE SINGLE STRAIN MUTATIONS FOR EVERY GENE IN THE MOUSE GENOME BY INSERTIONAL MUTAGENESIS AND HOMOLOGOUS RECOMBINATION ON A DEFINED GENETIC BACKGROUND J. Pollock1, G. Hicks2, W. Wurst3 1 National Institute on Drug Abuse, Munich, Germany 2 University of Manitoba, Munich, Germany 3 GSF-Institute Developmental Genetics, Munich, Germany Now that the human and mouse genomes have been sequenced, the next challenge is to determine the function of the genes contained in these genomes. Toward this goal, an international meeting was convened at the Banbury Conference Center at the Cold Spring Harbor Laboratory in October 2003 to discuss the development of an international project to inactivate every gene in the mouse genome, with each mutant strain of mice created carrying a single null mutation. The meeting participants recommended a phased approach. In the first phase it was recommended that a comprehensive collection of null, ideally conditional, alleles with a reporter be created in mouse embryonic stem (ES) cells. The meeting attendees recommended that, subsequently, mutant mice be created from this ES cell resource and characterized by a standardized battery of phenotyping tests and gene expression analyses, conducted at several central facilities. Investigator initiated projects would then complete specialized and in depth analysis of the mutants. The Banbury conference led to at least three international initiatives: one led by the Canadians called North American Conditional Mouse Mutagenesis Program (NorCOMM), one led by the European Union called European Conditional Mouse Mutagenesis Program (EUCOMM), and the NIH led Knockout Mouse Project (KOMP). These initiatives will be described in this symposium.
THE NIH KNOCKOUT MOUSE PROJECT (KOMP) J. Pollock National Institute on Drug Abuse The NIH KOMP initiative http://www.nih.gov/science/models/mouse/ knockout/index.html aims to: 1) support various approaches, including transposon mutagenesis, gene trapping, and gene targeting, to make a
resource of null alleles marked with a high-utility reporter in C57BL/6 by 2010; 2) support a repository to house the products of this resource, as well as an additional ‘repatriation’ effort to bring into repositories 1000 of the existing high priority mouse knockouts; 3) develop C57BL/6 ES cells with high germ line transmission that may be used in a high throughput manner for generating this resource; and 4) administer a data coordination center which will make the results of the production effort available to the research community. The projects awarded by KOMP RFAs (Request for Applications) for HG-05-007: The Completion of a Comprehensive Mouse Knockout Resource, DA06-009: Development and Improvement of Inbred ES Cell Lines for Use in Generation of Mouse Mutants, HG-05- 008: A Data Coordination Center for the Knockout Mouse Project (KOMP) will be announced and described. The RFA for Repository and Repatriation is currently being planned.
THE EUROPEAN CONDITIONAL MOUSE MUTAGENESIS PROGRAM (EUCOMM) W. Wurst GSF National Research Center for Environment and Health, Institute of Developmental Genetics, Munich, Germany The decoding of the sequences of the human and mouse genomes has been heralded as an historical milestone. This massive accomplishment has required the concerted action of the international scientific community, who have joined forces, exchanged technical and scientific knowledge and pooled resources. Analysis of the mouse and human genomes has resulted in the identification of approximately 25.000 genes as well as of thousands conserved non-coding regions. Now attention has turned to the next phase of the project, elucidation of gene function in the context of the entire organism in a coordinated large scale fashion. As part of an international consortium EUCOMM will produce, in a systematic high throughput way, conditional mutations throughout the mouse genome. A collection of up to i) 20.000 conditionally mutated genes will be generated in mouse embryonic stem (ES) cells using gene trapping and gene targeting approaches in the next three years; ii) 320 mouse mutants will be produced for genes of medical and biological relevance and iii) 20 transgenic mouse lines expressing cre-recombinase in a time and tissue specific regulated fashion will be generated. This EUCOMM library will enable mouse mutants to be established worldwide in a standardized and cost-effective manner, making mouse mutants available to a much wider biomedical research community than has been possible previously. All material including targeting vectors, mutant ES cells, mouse resources, and Standard Operating Procedures (SOPs) generated by EUCOMM will be displayed to the scientific community via the EUCOMM web site (http:// www.EUCOMM. org) and distributed by the RZPD (http://www. rzpd.de) and European Mouse Mutant Archive (EMMA, http:// www.emmanet.org). These resources will make major contributions to speed up functional genomics by the provision of animal models to the scientific community.
S1.2 IMAGING GENETICS: EXPLORING THE INTERPLAY OF GENES, BRAIN AND BEHAVIOR IN THE PATHOPHYSIOLOGY OF MOOD DISORDERS A.R. Hariri University of Pittsburgh, PA, USA Identifying specific biological pathways that contribute to complex cognitive and emotional behaviors is paramount to our understanding of how individual differences in these behaviors emerge and how such differences may confer vulnerability to neuropsychiatric disease. Recent advances in both molecular genetics and noninvasive functional neuroimaging have begun to provide the tools necessary to explore these and other behaviorally relevant biological mechanisms. Because of its unique ability to assay the structure and function of neural systems, neuroimaging provides a powerful tool to explore the relationship between genes, brain, and behavior. In this talk, I will outline an experimental strategy by which genetic effects on brain function has been explored and highlight the effectiveness of this strategy to delineate serotonergic pathways and mechanisms contributing to the emergence of individual differences in corticolimbic brain function that potentially bias temperament and risk for mood disorders.
Abstracts S1.3 IMAGING GENETICS MODEL FOR THE STUDY OF NEUROPSYCHIATRIC DISEASES S.G. Potkin, F. Macciardi, L. Friedman, J. Turner, J. Fallon, W. Bunney, D. Keator University of California, USA Given the known importance of both genetics and environment in brain function, and the role of neuroimaging in revealing brain dysfunction, the integration of genetics with brain imaging can enhance our fundamental understanding of disease. Genetic studies of disease are often association studies looking for genetic contributions to diagnostic status, which are inevitably small effects and require inordinately powerful sample sizes. The integration of intermediate phenotypes in imaging genetics holds promise for finding more specific and larger genotypic effects. This integrative approach incorporates >500K SNPs and >20K imaging voxels and must address a number of statistical issues. We utilize an ordered analysis incorporating candidate genes and genomewide strategies, a novel extension of a method for identifying significant results, and the use of a number of replication/ validation strategies. Initial imaging phenotypes are defined from the fMRI patterns that distinguish schizophrenic from healthy controls. The effects of genotype and diagnosis on the imaging phenotype is then analyzed. This analysis is not an association of genotype with diagnosis, but a combined analysis of the effects of genotype, diagnosis, and their interaction, on the quantitative trait of the imaging phenotype. Initial attention in the ordered analysis is focused on a priori selected candidate genes, and followed with a genome-wide analysis, with replication in new samples as a key step. This combines the advantages of both the genome-wide scans and the candidate gene approach and addresses the concern of false positives in combining imaging and gene wide scan approaches. S1.4 IDENTIFICATION OF NEURAL RISK MECHANISMS FOR PSYCHIATRIC DISORDERS USING IMAGING GENETICS A. Meyer-Lindenberg NIMH Neuroimaging can play a pivotal role in delineating neural systems related to genetic risk for psychiatric disorders. Understanding neural interactions on the systems level links genetic variation to behavioral and clinical phenotypes and defines new treatment targets. In this talk, we will focus on two circuits: striatal-prefrontal interactions related to risk for schizophrenia and variation in genes involved in dopaminergic neurotransmission (COMT, PRODH2, DARPP-32), and regulatory circuits linking amygdala to prefrontal cortex influenced by genes implicated in risk for anxiety, violence, depression and autism (MAO-A, 5-HTTLPR, AVPR1A). In each case, the importance of using neuroimaging methods to characterize functional and effective connectivity for reaching a mechanistic understanding of the investigated circuits will be stressed. S2.1 AUTISM GENETICS: FROM GENE TO BRAIN TO BEHAVIOR D. Geschwind UCLA School of Medicine Autism is a neurodevelopmental disorder characterised by language difficulties, social deficits and repetitive, stereotyped behaviour. Although autism has a strong genetic component, heterogeneity has limited power to detect autism-related loci. We have used quantitative endophenotypes in the AGRE resource to provide more analytic power than the categorical phenotype alone. The most fruitful have involved male or female sex, language delay and autistic-like social behaviour. We have recently provided the first confirmatory evidence (of genomewide significance) for an autism locus on chromosome 17, and confirmed a language-related quantitative trait locus (QTL) in multiplex autism families that overlaps a major autism locus on chromosome 7q3. We have moved from these loci to comprehensively examine specific regions by high density SNP mapping. On chromosome 7q3, 2758 single nucleotide polymorphisms (SNPs) covering a 15 Mb interval at an average inter-marker distance of 5.6 k b were genotyped in 174 parent/child trios from the Autism Genetic Resource Exchange (AGRE) in Stage 1. Stage 1 associated genes were subsequently tested in an independent sample of 300 AGRE trios in Stage 2 for replication, supporting CNTNAP2 as a n autism susceptibility gene. An association with autism was most pronounced in the subset of families with
687
seizures. In addition, a rare copy number variant was identified in CNTNAP2. Expression of CNTNAP2 was remarkably restricted to frontotemporal-subcortical circuits providing a link between genetic susceptibility and specific brain circuits involved in autism. S2.2 AUTISTIC SYMPTOMS IN ADHD—A REPORT FROM THE INTERNATIONAL MULTI-CENTER ADHD GENE PROJECT (IMAGE) A. Mulligan1, R. Anney1, W. Chen2, L. Butler1, M. Fitzgerald1, M. Gill1, and the IMAGE Consortium3 1 Trinity College Dublin 2 Institute of Psychiatry 3 IMAGE Introduction: Sub-threshold autistic symptoms and their relationship with oppositional symptoms are examined in an ADHD sibling pair sample. Methods: 1047 sibling pairs (1047 ADHD cases diagnosed using PACS and 1356 unselected siblings) had ADHD and oppositional symptoms assessed using Connors’ Rating Scales and autistic symptoms measured by Social Communication Questionnaire (SCQ), excluding autism and IQ < 70. Autistic symptoms were (1) compared in ADHD and sibling groups and (2) correlate d with oppositional symptoms. Results: Mean SCQ scores of ADHD subjects (8.32, 95%CI: 7.96–8.69) differed significantly from sibling scores (4.46, 95%CI: 4.21–4.70), using a regression method based on Huber-White variance estimator robust to equal variance assumption while modelling family cluster. SCQ correlated with oppositional symptoms in ADHD (r ¼ 0.26, p < 0.001) and siblings (r ¼ 0.37, p < 0.001). The total sample was subdivided into 3 groups: ‘No-Disorder’, ‘ADHD-only’ and ‘ADHDþ Conduct Disorder’. Mean SCQ scores were 4.42 (95%CI:4.18–4.66), 7.63 (95%CI:7.22–8.03) and 10.59 (95%CI:9.74–11.44) respectively (F(2,1056) ¼ 211.88, p < 0.001), confirming significant differences of all three groups. Conclusion: ADHD probands have more autistic symptoms than their siblings. Children with ADHD and conduct disorder have significantly more autistic symptoms than those with ADHD alone. We demonstrated clinical overlap between ADHD and autism symptoms and will explore whether genetic overlap is also present. S2.3 THE FIRST GENOMEWIDE ASSOCIATION STUDIES OF SOCIAL AND NONSOCIAL AUTISTIC TRAITS A. Ronald Institute of Psychiatry, King’s College, London Recent twin studies suggest that two of the core behaviours underlying autism spectrum disorders (ASDs) social impairments (sASD), such as unusual eye gaze and difficulty with social interactions, and nonsocial behaviours (nASD), such as sticking to rigid routines and repetitive behaviours are both highly heritable. However, analyses indicate that sASD and nASD, when assessed as traits in the population, are partly genetically distinct. The present research is the first to capitalize on this finding by conducting genomewide association scans for sASD and nASD separately. Both studies screened 500,000 SNPs on the Affymetrix 500K GeneChip microarray using DNA pooled for lowversus high-scoring groups selected from a representative sample of 5,000 children. Low-scoring sASD individuals were compared to highscoring sASD individuals (n ¼ 372 & n ¼ 373, respectively) in the first study, and in the second study, low-scoring nASD individuals were compared to high-scoring nASD individuals (n ¼ 436 & n ¼ 434, respectively). Each of these four groups was split into five independent ‘sub-pools’, with each sub-pool assayed on a separate microarray to capture sampling variance. As predicted from the twin findings, different SNPs were associated with sASD and nASD. Separately for sASD and nASD, SNPs showing the largest significant differences between high and low groups were individually genotyped on 5,086 children. Although each SNP accounts for a small amount of variance, they can be combined in an ‘SNP set’ that can be used as genetic risk indices for sASD and nASD. S2.4 THE INTERSECTION OF AUTISM AND ADHD: A GENETIC EPIDEMIOLOGY PERSPECTIVE R.D. Todd Washington University, School of Medicine Autism and attention deficit/hyperactivity disorder (ADHD) are usually thought of as distinct conditions with very different
688
Abstracts
characteristics. Not only are the two characterized by different domains by psychopathology but autism is usually thought of as a rare, severe disorder while ADHD is thought to be a common, less severe entity. Clinical studies from Scandinavia, genetic linkage and population based studies from around the world have recently emphasized the overlap of autistic and ADHD symptoms in a fraction of children. The goal of this presentation is to review several population-based studies of these two symptom domains and demonstrate their co-heritability of severe degrees of autistic characteristics with specific subtypes of ADHD. In particular, birth-record based twin studies in the state of Missouri where individuals were assessed for both ADHD and autistic characteristics will be reviewed to emphasize the basic underlying structure of the two disorders in the general population and the intersection of severe autistic symptomatology with combined type ADHD as defined using population-based criteria. The impact of such co-occurring versus comorbid symptomatology will be assessed using a variety of measures of school and social competence. The implication of these findings for genetic linkage studies of autism and ADHD, when considered separately, will be reviewed. Moreover, the impact of such characterizations on a variety of neuroimaging, candidate gene and neurophysiological studies will be addressed. S3.1 WHOLE GENOME ASSOCIATION STUDY OF BIPOLAR DISORDER: THE WELLCOME TRUST CASE CONTROL CONSORTIUM (WTCCC) N. Craddock for the Wellcome Trust Case Control Consortium Department of Psychological Medicine, Cardiff University, Cardiff, UK Recent improvements in genotyping technology together with reductions in costs have made it possible to undertake whole genome association (WGA) studies of complex genetic disorders including psychiatric phenotypes. Our group is leading the bipolar disorder component of the Wellcome Trust Case Control Consortium (WTCCC), a large-scale study funded by the Wellcome Trust and coordinated through the Sanger Institute, Cambridge, UK. WTCCC involves analysis of 2,000 UK samples from each of seven diseases (bipolar disorder, type 1 diabetes, type 2 diabetes, coronary heart disease, hypertension, rheumatoid arthritis and inflammatory bowel disease) and a common set of 3,000 UK (1,500 from the 1958 British Birth Cohort; 1500 blood donors). Two thousand cases of each disease and 3,000 controls are being genotyped using the Affymetrix 500K chip. Close attention is being paid to quality control and optimizing genotype scoring procedures. Genotyping will be completed during autumn 2006. Data will be made freely available to bona fide researchers after a period of 6 months. This presentation will describe the rationale, design and progress of the bipolar disorder component of the WTCCC study. S3.3 DESIGN AND ANALYSIS OF TWO-STAGE ASSOCIATION STUDIES: APPLICATION TO BIPOLAR I M. Boehnke, A. Skol, L. Scott, G. Abecasis, J. Li, R.C. Thompson, F. Meng, W. Guan, D. Absher, H. Akil, S. Watson, M. Burmeister, R. M. Myers University of Michigan, School of Public Health Data on human genetic variants from the International HapMap project and the precipitous drop in genotyping costs have made genomewide association studies a practical approach to study the genetics of complex diseases. Such studies require genotyping hundreds of thousands of genetic markers on hundreds or thousands of subjects. I will discuss optimal design and analysis of two-stage association studies in which a subset of the samples is genotyped on all genetic markers in stage 1, and the remaining samples are genotyped on the most interesting markers in stage 2. Consistent with Satagopan et al., we find that two-stage designs can maintain nearly the same power to detect association as the corresponding one-stage design in which all samples are genotyped for all markers. We find that joint analysis of stage 1 and 2 samples is nearly always more powerful than replication-based analysis, despite the need to account for a much larger number of tests. We address the impact on optimal study design of proportion of samples in stage 1, proportion of markers followed up in stage 2, per genotype cost ratio between stages 1 and 2, and etiologic heterogeneity between stages 1 and 2. We also describe practical issues that arise in the design of a two-stage genome wide association study of bipolar disorder and in the analysis of stage 1 data on 476 bipolar I cases and 470 ethnically-matched controls generously provided by NIMH Bipolar Disorder Genetics Initiative investigators and Dr. Pablo Gejman.
S4.1 NORMALITY AND PATHOLOGY IN A BIOLOGICAL AGE N. Rose, M. White BIOS Centre for the Study of Bioscience, Biomedicine, Biotechnology, and Society, London School of Economics and Political Science In this talk I consider some social implications of recent advances in the sciences of life, especially those in psychiatric genetics and behavioural genomics. I focus in particular on the reshaping of ideas about susceptibility to mental disorder associated with the move away from simple genetic determinism in psychiatry and the developing genetic technologies that enable rapid screening for variations in multiple sequences. I argue that the possibilities of presymptomatic genetic screening for psychiatric disorders, coupled with ideas of prevention and precaution, are reframing the opposition of normal and pathological. I suggest that this is likely to lead to the increasing use of early intervention with psychiatric drugs for those who are not suffering from a mental health problem but are considered to be ‘at risk’. I consider this in the context of two other developments: the re-unification of explanations of normal variations of personality and abnormal conditions of mental illness in molecular neuroscience, and the expansion of the scope of psychiatric diagnosis associated with new perceptions of the prevalence of mental disorder. I suggest that, taken together, these developments are reshaping perceptions of personhood, identity and responsibility concerning mental health, and outline some implications. S4.2 GENES, MENTAL ILLNESS AND STIGMA J. Phelan Columbia University, USA One likely consequence of the genetics revolution is an increased tendency to understand human behavior in genetic terms. How might this ‘‘geneticization’’ affect stigma associated with mental illness? Attribution theory predicts a reduction in stigma via reduced blame, anger and punishment and increased sympathy and help. On the other hand, according to ‘‘genetic essentialist’’ thinking, genes are the basis of human identity and strongly deterministic of behavior. If such ideas are common, geneticization should exacerbate stigma by increasing perceptions of differentness, persistence, seriousness and transmissibility, which in turn should increase social distance and reproductive restrictiveness. I present data testing these predictions with a vignette experiment embedded in a nationally representative survey. I found little support for attribution-theory predictions. Consistent with genetic essentialism, genetic attributions increased the perceived seriousness and persistence of the mental illness and the belief that siblings and children would develop the same problem. Genetic attribution did not affect reproductive restrictiveness or social distance from the ill person but did increase social distance from the person’s sibling, particularly regarding intimate forms of contact involving dating, marriage and having children.
S4.3 GENETIC TESTS, COMMERCIALIZATION, AND CONFLICT OF INTEREST N.A. Holtzman The Johns Hopkins University, Baltimore, MD, USA Using a novel device, I will address three problems: (1) Discoveries of genotypes for common, complex diseases, such as schizophrenia and bipolar affective disorder, receive publicity in the mass media before they have been confirmed, generating excessive public expectations. (2) Some gene-discovery is conducted by university-affiliated scientists who receive support from for-profit ventures, raising conflict-of-interest issues. (3) ‘‘Inhouse’’ genetic test development and marketing is not subject (at least in the USA) to regulation to assure safety and effectiveness. From our study of the mass media the following hypothetical press release is emblematic: A team of scientists from universities in Europe and the United States announced the discovery of a major gene for schizophrenia at the Fifteenth International Congress of Genetic Psychiatry, in Cagliari, Sardinia today. They studied the genes from an international collection of schizophrenic patients, their normal parents and siblings. The schizophrenic cases were five times more likely to possess a genetic variant (allele) of a gene than their unaffected relatives. The scientists have named this gene Schizo-12. Dr. Henry Johnson, the molecular geneticist who reported the findings, said that within a year a test to detect the presence of the variant would be on the market and that a drug
Abstracts
689
that would prevent schizophrenia in patients who tested positive could be expected within five years. The work was supported by Psychotropics‘R-Us, an American pharmaceutical company. I conclude that premature publicity and commercialization, without adequate regulation, can impede progress in psychiatric genetics.
investigation. The DISC1-PDE4B pathway brings together learning, memory and mood with the neurodevelopmental hypothesis in schizophrenia while GRIK4 lends credence to the glutamate hypothesis. These genes, discovered though molecular cytogenetics, are providing key insights into the molecular aetiology of schizophrenia and bipolar disorder but can be equally applied to other areas or biological psychiatry.
S4.4 REINFORCING THE BIOLOGICAL: UNDERSTANDING THE BEHAVIOURAL IMPACT OF PREDICTIVE GENETIC TESTING T. Marteau Institute of Psychiatry, King’s College, London
S5.3 DECONSTRUCTING SCHIZOPHRENIA: ONE DISORDER OR SEVERAL? A. Jablensky University of Western Australia
Genetic tests are being developed that can predict with increasing precision the probability of common complex diseases and responsiveness to a range of drugs. Such testing is intended to prevent disease by motivating those found to be at increased risk to engage in risk-reducing actions. These might include increasing levels of physical activity, stopping smoking or taking medication. The effectiveness of such prediction in reducing risk depends critically upon how people will respond to learning of a genetic predisposition to disease. Some have suggested that genetic predictive testing will increase motivation to engage in risk reducing behaviours, by providing personalised, salient risk information. Others have argued that, given the strong belief that genetic risks are immutable, such testing will reduce motivation to engage in actions to reduce the risks. From the few studies conducted to date it would appear that such testing does not reduce perceptions of control over disease. It does, however, seem to alter perceptions of how control is achieved: when risk assessments include a genetic test, drug therapies are seen as more effective means of reducing risk, and behavioural means less effective (Wright et al., 2003; Marteau et al., 2004). Results from general population surveys of responses to predictive genetic testing for psychiatric conditions report a similar pattern (Phelan et al., 2006). These data can be understood within Leventhal’s Self-Regulation Model of Health and Illness which describes the operation of a series of IF-THEN rules governing responses to health risk information. Evidence will be presented to support the thesis that when genetic tests are used to predict a threat, the threat takes on a genetic identity; if a threat has a genetic identity then it activates a representation of reducing it that is biological in nature, representing the operation of a symmetry heuristic concerning the natures of causes and their solutions (Marteau & Weinman, 2006). Of concern is that using genetic predictions could result in the eschewal of what will sometimes be more effective behaviourally based interventions for less effective biologically-based interventions, an effect that may be evident for psychological as well as physical disorders. S5.1 LINKAGE AND ASSOCIATION STUDIES: PAST PRESENT AND FUTURE M. O’Donovan The Henry Wellcome Building for Biomedical Research in Wales The speakers in this session will discuss various approaches to finding genes for schizophrenia. In this presentation, I will briefly review the successes and failures that have thus far emerged from linkage and candidate gene studies and what that might tell us about the task facing gene identification studies now. I will also illustrate with some recent data how gene analyses might be informed by combining simple association and gene-gene interaction data with expression and linkage data from other public resources to provide convergent evidence implicating functional pathways relevant to schizophrenia pathogenesis. S5.2. GENES AND PATHWAYS THROUGH CYTOGENETICS: THE DISC1, PDE4B AND GRIK4 PARADIGMS D.J. Porteous, W. Muir, D. Blackwood, K. Millar, B. Pickard University of Edinburgh The genetic architecture of major mental illness is unknown, but heterogenous. The prevailing strategies of linkage and association may be valid, but have been of limited success in identifying and validating vulnerability genes and aetiological pathways. Molecular cytogenetics has a well established track record of success in other disease domains. Here we describe the strategy as successfully applied to the identification of DISC1, PDE4B, GRIK4 and other biologically plausible candidates that have been subsequently validated genetically by linkage or association studies and experimentally by clinical and laboratory
Phenotypic variability and likely extensive genetic heterogeneity have been confounding the search for the causes of schizophrenia since the inception of the diagnostic category. The inconsistent results of genetic linkage and association studies using the diagnostic category as the sole schizophrenia phenotype suggest that the current broad concept of schizophrenia does not demarcate a homogeneous disease entity. Approaches involving subtyping and stratification by covariates to reduce heterogeneity have been successful in the genetic study of other complex disorders, but rarely applied in schizophrenia research. An overview of various past and present attempts at delineating schizophrenia subtypes based on clinical features, statistically derived measures, putative genetic indicators, and intermediate phenotypes, highlights the utility of multi-domain neurocognitive endophenotypes. S5.4 COMPREHENSIVE LINKAGE DISEQUILIBRIUM MAPPING OF SCHIZOPHRENIA CANDIDATE GENES IN A LARGE EUROPEAN-ANCESTRY SAMPLE P.V. Gejman1, J. Duan1, M. Martinez2, A.R. Sanders1, G. Burrell1, C. Hou1, D. He1, D. Schwartz1, N.G. Buccola3, B.J. Mowry4, R. Freedman5, F. Amin6, D.W. Black7, J.M. Silverman8, W.F. Byerley9, R.R. Crowe7, C.R. Cloninger10, D.F. Levinson11 1 ENH & Northwestern Univ, Evanston, IL 2 INSERM, Toulouse, France 3 LSU Health Sciences Center, New Orleans, LA 4 QCSR and University of Queensland, Brisbane, Australia 5 Univ. of Colorado Health Sciences Center, Denver, CO 6 Atlanta VA Med Ctr & Emory Univ, Atlanta, GA 7 Univ. of Iowa, Iowa City, IA 8 Mt. Sinai School of Medicine, New York, NY 9 UCSF, San Francisco, CA 10 Washington University, St. Louis, MO 11 Stanford Univ, Palo Alto, CA Linkage, association and cytogenetic data have nominated plausible schizophrenia (SZ) candidate genes, but replication studies have produced mixed results, and there is no definitive evidence for a DNA sequence with an identifiable pathogenic mechanism. We have studied 14 SZ candidate genes (AKT1, ARVCF, CHRNA7, COMT, DAOA, DISC1, DRD2, DTNBP1, HTR2A, NRG1, PPP3CC, RGS4, and the STX7-TAAR6 gene cluster) in a European-ancestry sample of 1,673 cases with final diagnoses of SZ or schizoaffective disorder and 2,146 controls (screened by self-report to exclude those with possible psychotic or bipolar disorders). We used SNPlex and Taqman methods to genotype comprehensive maps of each gene (800 SNPs in 2.2 Mb of sequence) consisting of a framework of HapMap SNPs to tag common variation and additional SNPs that had been reported as associated to SZ, or are non-synonymous or within functional elements; and 187 ancestry-informative markers (AIMs) for continental populations, analyzed in relation to self-reported ancestry. We achieved high call rates and low error rates. Analysis of AIMs demonstrated minimal case-control allele frequency differences. We have analyzed 118 SNPs in ARVCF, COMT, DRD2 (Ser311Cys), HTR2A, NRG1 (core SNP haplotype), and TAAR6. We observed nominal p-values < 0.05 for SNPs located in HTR2A, NRG1 and TAAR6. Interpretation of the findings will require completion of analyses of the remaining genes and of empirical gene- and experiment-wise p-values, as well as further analyses of possible population substructure. S6.1 A PHARMACOGENETIC STUDY OF TREATMENT OUTCOME AND ADVERSE EFFECTS IN THE STAR*D COHORT F.J. McMahon Genetic Basis of Mood & Anxiety Disorders, NIMH, Bethesda, MD, USA
690
Abstracts
Important components of the individual variability in treatment outcome and adverse effects among patients treated with antidepressants may have a basis in common genetic variation that can be revealed directly through genetic association methods. We are genotyping DNA samples obtained from a cohort of 1953 patients with major depressive disorder who were treated with antidepressants in the Sequenced Treatment Alternatives to Relieve Depression (STAR*D) study and prospectively assessed. So far we have focused on a selection of 68 candidate genes which we have genotyped with 768 single nucleotide polymorphism markers chosen to detect common genetic variation. The goals are 1) to identify genetic markers of treatment outcome and 2) to identify genetic markers of treatment emergent adverse effects. We have detected significant and reproducible association between response to citalopram treatment and markers in HTR2A, which encodes the serotonin 2A receptor (McMahon et al., 2006), and GRIK4, which encodes a kainate-type glutamate receptor. Preliminary results also implicate BCL2, which is involved in neurogenesis. Taken together with prior neurobiological findings, these genetic association data make a compelling case for key roles of HTR2A and GRIK4 in the mechanism of antidepressant action and suggest a possible role for BCL2 as well. Studies of treatmentemergent adverse events, such as suicidal ideation, insomnia, and anxiety, are currently underway. Genes involved in these events may also modulate risk for bipolar or other disorders. Genetic markers that identify individuals at high risk for treatment failure or particular adverse effects may ultimately help guide treatment decisions. S6.2 GENETIC ASSOCIATION DATA FROM GENDEP, A MULTICENTRE EUROPEAN STUDY K. Aitchison1, P. Huezo-Diaz2, R. Williamson2, F. Hoda2, M. Nash2, M. Rietschel3, C. Schmael3, T. Schulze3, A. Farmer2, J. Hauser4, N. Henigsberg5, W. Maier6, A. Zobel6, E.R. Larsen7, O. Mors7, A. Marusic8, J. Mendlewicz9, D. Souery9, J. Perez10, P. McGuffin2, I. Craig2 1 Institute of Psychiatry 2 Institute of Psychiatry, MRC SGDP Centre, King’s College London 3 Zentralinstitut fu¨r Seelische Gesundheit (CIMH) 4 Akademia Medyczna im Karola Marcinkowskiego w Poznaniu (Academic Dept of Medicine) Hrvatski institut za istrazivanje mozga, Medicinski fakultet Sveucilista u Zagrebu 6 Universita¨tsklinikum Bonn, Anstalt des o¨ffentlichen Rechts, fu¨r den Fachbereich Medizin 7 Afdeling for Psykiatrisk Demografi, Institut for Psykiatrisk Grundforskning 8 Institut za varovanje zdravja Republike Slovenije 9 Universite´ Libre de Bruxelles, Hopital Erasme, De´partement de Psychiatrie 10 Centro San Giovanni di Dio-Fatebenefratelli 5
Background and Aims: Despite the availability of a number of different antidepressant medications, about 30–50% of patients with depression will show inadequate therapeutic response to these drugs. GENDEP (Genome-based therapeutic drugs for depression) is a European multicentre integrated pharmacogenomics study aiming to identify genomic and proteomic correlates of antidepressant response. Methods: In the clinical component of GENDEP, the inclusion criteria are: major depression of at least moderate severity, and white European ethnicity. Subjects are randomly allocated to two antidepressants: escitalopram (proserotenergic) and nortriptyline (pronoradrenergic), with serial clinical ratings and collection of clinical samples. In our first wave of genetic association analysis, we are genotyping SNPs and repeat polymorphisms in 12 candidate genes (SLC6A2, SLC6A4, TPH1, TPH2, HTR1A, HTR2A, BDNF, TRKB, GNB3, ADRA2A, CYP2D6, and CYP2C19). Results: There was an association between response on the MADRS and the serotonin transporter (SLC6A4). (Data on CYP2D6 and CYP2C19 are reported separately.) Conclusions: The 5-HTTLPR finding replicates findings of many other investigators (Smits et al., 2004). Further data analysis is in progress. Acknowledgements: this project is funded by the European Commission, contract number LSHB-CT-2003–503428.
S6.3 A LARGE SCALE DEPRESSION CASE-CONTROL STUDY AND ITS IMPACT ON DRUG DISCOVERY & DEVELOPMENT P. Muglia, F. Tozzi, I. Prokopenko, A. McCarthy, S. Brewster, J. Perry, S. Chissoe, P. St. Jean, A. Bansal, C. Allan, J. Riley, E. Domenici, E. Merlo-Pich, F. Holsboer, L. Middleton, A. Roses GlaxoSmithKline, Research and Development The discovery of new antidepressants has moved from serendipitous findings that 50 years ago produced the first antidepressants into the development of new compounds based on neurobiological hypotheses. More recently human genetics association studies have offered a new and challenging opportunity to generate a new wave of potential therapeutic targets. The disease-associated genes should provide new antidepressant targets with a better probability of delivering efficacious antidepressants. The analysis of the disease-associated genes for biologically plausible and clinically relevant depression subphenotypes offer a unique opportunity to explore whether specific groups of patients may benefit more than others from modulation of the specific targets. GlaxoSmithKline has been investing extensively in genetics and has collected large and clinically thoroughly characterized samples for at least seventeen common diseases (Roses et al., 2005). For depression 1,000 recurrent depression cases and 1,000 age, gender and ethnic matched controls have been collected from three recruiting psychiatric hospitals from Munich area in Germany under the coordination of the Max Plank Institute and the direction of Prof Florian Holsboer. Depression was diagnosed according to DSM-IV using the SCAN interview and assessment included other clinical variables such as personality traits and life events. All cases and controls have been genotyped for 1,827 tractable genes with up to 6,500 SNPs as part of the so-called HiTDIP program at GlaxoSmithKline. Results from the case control analysis of HiTDIP genes in depression and depression symptom dimension subtypes will be presented. S6.4 GENOME-WIDE LINKAGE ANALYSIS USING ADDITIONAL FAMILIES FROM THE DEPRESSION NETWORK STUDY B. Todd Webb1, E. J. van den Oord1, P. McGuffin2, J. Knight2, G. Breen2, S. Brewster3, P.R. Boyd3, N. Craddock4, M. Gill5, A. Korszun6, W. Maier7, L. Middleton3, O. Mors8, M.J. Owen4, J. Perry3, M. Preisig9, T. Reich10, J. Rice10, M. Rietschel11, L. Jones12, P. Sham2, A.E. Farmer2, P. Muglia3 1 VIPBG, Virginia Commonwealth University 2 Institute of Psychiatry, King’s College London 3 GlaxoSmithKline, Research and Development 4 Department of Psychological Medicine, School of Medicine, Cardiff University 5 Department of Psychiatry, Trinity Centre for Health Sciences 6 Barts and The London, Queen Mary’s School of Medicine and Dentistry 7 Department of Psychiatry, University of Bonn 8 Department of Psychiatry, University of Aarhus 9 Department of Adult Psychiatry, University Hospital of Lausanne 10 Department of Psychiatry, Washington University 11 Central Institute of Mental Health 12 Department of Psychiatry, University of Birmingham Genome-wide linkage analysis for recurrent major depressive disorder (MDD) using the first phase of the Depression Network Study (DeNt) consisting of 497 concordant sib pairs was previously reported. A second phase of an additional 325 families has been genotyped and analyzed in an attempt to replicate the results from the first phase. The previously reported suggestive evidence for linkage on chromosome 1p36 does not appear to replicate in the second phase of the genome scan. The DeNt offers ample opportunity to explore alternative dimensional rather than categorical definitions of depressive phenotypes such as neuroticism and symptom severity. These quantitative measures were analyzed in the combined sample. To control the risk of false positives in the analysis of multiple phenotypes, False Discovery Rate methods were applied. Preliminary results appear to yield interesting leads in locating novel loci involved in depression.
S7.1 METHODOLOGICAL ISSUES FOR COMPLEX TRAIT MAPPING IN ISOLATED POPULATIONS C. Bourgain INSERM U535—Universite´ Paris XI—Villejuif, France
Abstracts
691
Random individuals from isolated populations have a non negligible probability of being related, though this probability strongly depends on the population history. Consequently, inbreeding is often present and large pedigrees connecting all living individuals may be available. The existence of inter and intra-individual correlations has important consequences on most mapping methods and deserves specific methodological developments. If virtually all individuals are eligible for linkage studies, the extensive pedigrees where individuals are related through multiple lines of descent are far too complex for standard linkage analysis. Alternatives include inbreeding based linkage or efficient breaking of the genealogy into smaller sub-pedigrees for standard non parametric linkage. Case-control tests are a powerful tool for association studies in the absence of population stratification, provided that the potential bias introduced by the existence of intra and inter-individual correlations is correctly taken into account. The different solutions for performing valid tests of association depend on the quality of the genealogical information available. When the genealogy is entirely known, the correlations may be computed exactly and appropriate tests for both quantitative and qualitative traits are available. In the absence of exhaustive genealogical information, use of random markers throughout the genome as genomic controls is an interesting way to estimate the correlations and perform valid tests of association. The different methodological issues for both linkage and association studies will be discussed and illustrated with studies from the Cilento (South Italy) and the Hutterites (USA).
challenged, and only few samples exist to prove if it really would be more straightforward in such populations to detect disease-related haplotype signatures through association studies. Detailed information of the population history is increasingly understood as the crucial factor in genetic studies of common diseases. I will describe the features of the Finnish population and our efforts to search for disease genes for rare and common phenotypes. Our studies on schizophrenia, autism and cardiovascular diseases exemplify the strategies used to identify disease genes in various Finnish data sets and study samples. Our studies demonstrate how the impact of specific allelic variants of disease genes, identified in families, can be addressed in large population cohorts containing excessive amount of quantitive phenotype information. (The Center of Excellence, Academy of Finland; GenomEUtwin–project of EU, no. QLG2-CT-2002-01254; NIH grants: R01 NS 43559 and 1R01 HL 70150).
S7.2 ASSESSING THE DEGREE OF POPULATION STRATIFICATION IN ASSOCIATION STUDIES: AN EXAMPLE FROM ICELAND A. Helgason deCODE Genetics, Iceland
1022 subjects from 448 families in the Collaborative Study of the Genetics of Alcoholism (COGA) were assessed at two time points separated by an interval of 4–5 years. The age of subjects at Time 2 was 17.4þ3.4 years. At Time 2, 256 subjects recei ved a diagnosis of alcohol dependence or abuse by one or more criterion sets. Several single genes, including ADH4, GABRA2 and CHRM2, have been shown to be associated with alcohol dependence in adults from these families. In the offspring sample, affected status at Time 2 was associated with Age at First Drink as reported at Time 1 (chi-square ¼ 30.8, p < .0001). The ADH4 (Alcohol dehydrogenase 4) risk allele predicts earlier age at first drink (chi-square ¼ 4.0, p < .05) and greater risk f or alcohol problems (Fisher’s p < .005). GABRA2 (the alpha 2 subunit of the GABA-A receptor) genotype is expressed as conduct problems (F ¼ 4.0, p ¼ .02) in younger adolescents. In older adolescents, GABRA2 genotype is associated with alcohol dependence (F ¼ 11.2, p < .0003). CHRM2 haplotype (muscarinic cholinergic receptor 2) is associated with alcohol dependence and depression in adults; in adolescents, the phenotype is more likely to be depression than alcohol dependence. Additional singl e genes, including hTAS2R16 on chromosome 7q and GABRG3 on chromosome 4p, are also being studied. This work is part of a longitudinal study that will sample phenotypic correlates of risk genes at multiple developmental stages.
The impact of population structure on association studies, undertaken to identify genetic variants underlying common human diseases, is an issue of growing interest. Spurious associations of alleles with disease phenotypes may be obtained or true associations overlooked when allele frequencies differ notably among subpopulations that are not represented equally among cases and controls. Recent research indicates that population structure influences even carefully designed studies and can affect the validity of association results. Most study designs address this problem by sampling cases and controls from groups that share the same nationality or self-reported ethnic background, with the implicit assumption that no substructure exists within such groups. We examined population structure in the Icelandic gene pool using extensive genealogical and genetic data. Our results indicate that sampling strategies need to take account of substructure even in a relatively homogenous genetic isolate. This is likely to be even more important in larger populations. S7.3 IDENTIFICATION OF COMPLEX DISEASE GENES IN A GENETIC ISOLATE: EXAMPLE OF FINLAND Leena Peltonen, M.D., Ph.D., Academy Professor Department of Molecular Medicine, National Public Health Institute and Department of Medical Genetics, University of Helsinki, Finland; The Broad Institute, MIT, Boston, MA, USA The Human Genome Project has produced a high number of catalogued sequence variants enabling genome-wide studies of genetic loci behind common disease-related phenotypes. However, multiple uncertainties must be solved before the best possible strategy for the gene hunt can be designed and large-scale genome-wide investigations undertaken. Which phenotypes to include, which study population (families or cases/isolated or outbred) to choose, which type of markers to be employed (multiallelic or SNPs), and how to select the variants to be genotyped? Rapidly increasing information of the structural or functional variability within the genome will greatly affect the interpretation of data. European population isolates like Finland have been very useful for mapping and cloning genes for rare disorders; in such isolates genetic drift leads to an overabundance of disease-alleles for particular disorders, and a high proportion of patients share these alleles, identical by descent. The concept that the isolates are similarly advantageous for genetic studies of common diseases has been
S8.1 SEPARATE DEVELOPMENTAL PATHWAYS FOR SINGLE GENES ASSOCIATED WITH ALCOHOL DEPENDENCE J.I. Nurnberger1, R. Wiegand1, L. Bierut2, K. Bucholz2, T. Foroud1, H. Edenberg1, E.T. Meyer1, A. Katschke1, J. Kramer3, D. Dick2, W. Reich2, S. Kuperman3, V. Hesselbrock4, A. Goate2, B. Porjesz5 1 Indiana University School of Medicine 2 Washington University at St. Louis 3 University of Iowa 4 University of Connecticut 5 State University of New York
S8.2 GENETICS OF NICOTINE DEPENDENCE—GENOME WIDE ASSOCIATION AND CANDIDATE GENE STUDIES L. Bierut Washington University, at St. Louis Smoking is the leading source of preventable death in the United States, and twin studies consistently demonstrate strong genetic contributions to smoking. The NICSNP study uses a hybrid methodologic approach to survey the entire genome while incorporating prior biologic knowledge. The NICSNP sample is a large case/control study of 1050 nicotine dependent and 879 non-dependent smokers. All participants were selected from two community-based studies, the Collaborative Genetics Study of Nicotine Dependence (U.S.) (PI- Laura Bierut) and the Nicotine Addiction Genetics Project (Australia) (PI- Pamela Madden). The genome wide association (GWA) consisted of pooled genotyping of 2.4 million single nucleotide polymorphism (SNPs) followed by individual genotyping of the top 40,000 signals. The second arm involved a comprehensive candidate gene study, where individual genotyping was conducted in over 400 genes chosen for their biological significance by experts in the field of addiction. There are convergent findings in these complementary approaches. Common variants in candidate genes exhibited the highest level of statistical significance and are among the top 25 signals from the GWA study. In addition, the
692
Abstracts
GWA study results identify novel loci not previously associated with the risk for nicotine dependence. These findings are a major advance in the genetics of complex human disease. This study is a collaborative effort involving scientific dialogue between investiga tors from academic centers, the National Institute on Drug Abuse, the National Cancer Institute, and Perlegen Sciences. These will be publicly available upon publication for further analyses by the scientific community. S8.4 THE MU OPIOID RECEPTOR GENE AND ADDICTIONS Wade H. Berrettini Dept. of Psychiatry, U. of Pennsylvania, Philadelphia, PA 19104 In QTL studies of C57BL/6J and DBA/2J mice, the proximal region of murine chromosome 10, containing the mu opioid receptor gene (OPRM1), is known to contain sequences that influence multiple opioid-related phenotypes, such as voluntary oral morphine consumption, morphine analgesia and density of mu receptors. These data are consonant with mu receptor null mutant experiments, showing that the mu opioid receptor is essential for the rewarding and analgesic effects of morphine. We have created reciprocal congenics for the interval of proximal chromosome 10 containing the OPRM1 locus, and phenotypic characterization reveals capture of the chromosome 10 locus for voluntary morphine oral consumption. Several reports of linkage disequilibrium between opioid dependence (OD) and the human OPRM1 locus will be reviewed. At a functional variant in the human OPRM1 gene, A118G, the G allele predicts response to naltrexone among alcohol dependent individuals. These diverse results suggest that the OPRM1 gene plays a multifaceted role in addition genetics and pharmacogenetics.
S9.1 MOLECULAR CLOCK: GENETICS AND EPIGENETICS BASIS OF CIRCADIAN RHYTHMS P. Sassone-Corsi Department of Pharmacology, University of California, 92697 Irvine, California, USA Circadian rhythms (from the Latin circa diem meaning ‘about a day’) of 24 hours constitute a fundamental physiological function observed in almost all organisms, from prokaryotes to humans. In mammals, circadian rhythms are generated in pacemaker cells within the suprachiasmatic nuclei (SCN) of the hypothalamus, and are entrained by environmental cues, principally light. Disruption of circadian rhythms may have profound influence to human health. For example, disruption of circadian rhythmicity has been linked to insomnia, Alzheimer’s disease, Huntington’s disease and depression. It is now well established that circadian clocks operate via feedback autoregulatory loops of clock proteins and genes in SCN neurons. A finding of fundamental importance has been that peripheral tissues also contain independent clocks, whose function is orchestrated by the SCN. To date, nine core circadian clock genes have been identified in mammals. The molecular mechanism of the circadian oscillator is based on transcriptional-translational feedback loops, as genetic analyses in Drosophila and mammals have shown. Two transcriptional activators, CLOCK and BMAL1, regulate gene expression by interacting with promoter elements termed E-box. Target genes of these activators encode several repressor proteins, including PER1, PER2, PER3, CRY1, and CRY2, that function by inhibiting the CLOCK/BMAL1 complex, thereby generating a circadian oscillation of their own transcription. A long-standing unanswered question concerns the signaling pathways that influence circadian rhythms, and thereby various physiological and metabolic functions. In the past three decades, a number of studies have implicated dopamine in the physiological signaling to the circadian clock. We have investigated the role of the D2 receptor, as it is highly expressed in the retina, a tissue that bears a powerful clock which is directly influenced by light. We will show data that implicate the D2R in signaling to the CLOCK-BMAL1 activator complex and the molecular mechanism by which this process is achieved at the transcriptional level. We will also demonstrate how signaling to the molecular clock involves remarkable changes in histone modifications and chromatin remodeling, underscoring the importance of epigenetic events in the regulation of physiological responses.
S9.2 THE EPIGENETIC MODIFICATION OF THE PRADER–WILLI/ANGELMAN SYNDROMES REGIONAL CONTROL CENTER Ruth Shemer Department of Cellular Biochemistry and Human Genetics, The Hebrew University, Hadassah Medical School, Jerusalem, Israel 91120 Genomic imprinting involves the marking of genes during gametogenesis or early embryo development to achieve monoallelic, parent-oforigin-specific expression. The 2Mb domain on chromosome 15 contains a group of imprinted genes that are paternally expressed, and one gene which is expressed exclusively from the maternal allele (UBE3A). Genetic aberrations in this domain result in two clinically distinct neurobehavioral disorders, Prader–Willi (PWS) and Angelman (AS) syndromes. PWS is a result of molecular defects that bring about silencing of the paternally expressed genes, while AS is caused by molecular defects that result in a loss of expression of UBE3A gene on the maternal copy. The monoallelic expression of all imprinted genes in the domain is under the control of an imprinting center. We found that this imprinting center has a parent-of-origin-specific epigenetic modification, which is acquired during gametogenesis and early embryo development. MeCP2 is a nuclear protein that binds specifically to methylated DNA and functions as a general transcription repressor in the context of chromatin remodeling complexes. MeCP2 binds the PWS/AS imprinting center and regulates its epigenetic structure. Mutations in the MeCP2 gene cause Rett syndrome, a neurodevelopmental disorder, which shares clinical features with those of Angelman syndrome. We found that this similarity is due to the fact that MeCP2 deficiency causes epigenetic aberrations of the PWS/AS imprinting center. These epigenetic aberrations lead to the reduction of UBE3A expression in Rett syndrome patients.
S9.3 WHAT ARE IMPRINTED GENES DOING IN THE BRAIN? L.Wilkinson, A. Isles, W Davies Behavioural Genetics Group, Psychology and Psychological Medicine, Cardiff University, UK A major interest of our group is in the brain and behavioural functions of imprinted genes, where expression in the offspring is dependent on whether the allele has passed through the male or female germline. We have recently reviewed the arguments as to why this is a particularly fascinating and important area of work1 but basically it has become increasingly apparent that; (i) many imprinted genes are expressed in mammalian brain, (ii) imprinted genes appear to be important in fundamental aspects of brain development/organisation, (iii) a number of human mental disorders show parent-of-origin inheritance patterns consistent with the involvement of imprinting and (iv) direct manipulation of imprinted loci in mouse models give rise to behavioural effects, both pre-weaning and in the adult animal. In my talk I will consider the challenges stemming from the possibility that imprinted genes may be influencing complex aspects of behaviour; including basic questions as to the likely scope and extent of any imprinted effects, diverse issues related to biological mechanism and perhaps the most difficult challenge of all, how to reconcile such effects in the adult with current evolutionary theories explaining the origin and maintenance of imprinted genes, theories which pit the maternal and paternal genomes against each other in an intragenomic battle of the sexes. 1. Davies W, Isles AR, Wilkinson LS. (2005). Imprinted gene expression in the brain. Neuroscience and Biobehavioral Reviews 29, 421–430.
S9.4 THE TWO FACES OF EPIGENETIC INHERITANCE Andras Paldi Ecole Pratique des Hautes Etudes, GENETHON, Evry, France The term of epigenetic inheritance is used to designate a set of heritable phenomena that can not be explained by the transmission of the DNA. According to a more conservative interpretation, epigenetic inheritance is restricted to the mitotic and/or meiotic transmission of a gene’s functional state. Maintenance of the cellular phenotype after division is an example of mitotic epigenetic inheritance. When transmitted through meiosis, the epigenetic characters do not follow Mendelian inheritance making the classical linkeage analysis inefficient. In all
Abstracts cases however, genetically identical cells or organisms display different heritable phenotypes. One of the best characterized examples of epigenetic inheritance is genomic imprinting where the activity state of the gene depends on its genealogical history, but independent of the DNA sequence itself. The two parental alleles of an imprinted gene are embedded in chromatin regions with different structure on the two homologous chromosomes leading to different replication and expression patterns. Studies on DNA methylation and covalent modifications of chromatin components led to the notion that epigenetic inheritance is based on these molecular mechanisms, usually termed epigenetic as well. The epigenetic mechanisms influence the accessibility of the DNA sequence to the transcription machinery by changing the structure and dynamics of the chromatin. Human pathologies associated to the defects in the imprinting mechanism demonstrate the high stability of the epigenetic states. High stability of epigenetic inheritance is increasingly recognized as an important factor in the development and evolution of multicellular organism. However, less attention is paid to the second face of epigenetic inheritance. Indeed, substantial evidence shows that environmental stimuli can influence the transmission of epigenetic traits. The environmental instability re-introduces a Lamarckian flavour in the understanding of heredity. The role of the interface between the external stimuli and epigenetic inheritance is played by the epigenetic mechanisms of chromatin modification. Little is known about how this interface functions. The interconnected network of epigenetic mechanisms may function in a bistable manner. Small perturbations usually have no effect; they are successfully buffered by the interactions within the network. Large perturbations may push the epigenetic system in a new locally stable equilibrium. I shall discuss the potential role of the cellular metabolism in the epigenetic switch. S10.1 MOLECULAR TOOLS TOWARDS SINGLE CELL AND MOLECULAR ANALYSES IN SYSTEMS BIOLOGY Ulf Landegren, O. Ericsson, M. Gullberg, S. Gustafsdottir, M. Howell, S. Henriksson, J. Jarvius, M. Jarvius, M. Kamali, C. Larsson, K.J. Leuchowius, J. Melin, M. Nilsson, E. Schallmeiner, O. So¨derberg, M. Thorselius, C. Wa¨hlby The Rudbeck Laboratory, Department of Genetics and Pathology, Uppsala University, Uppsala, Sweden In order to go from mere pattern recognition of molecular states to mechanistic and systems-level understanding, it is now imperative to conduct analyses at the single-cell level. It therefore emerges as an important goal to enable investigations at single-molecule levels with quantitative image analysis and parallel read-out formats, generating digital as opposed to analog information. It is therefore critically important to devise extremely specific molecular recognition reactions, coupled with background-free signal amplification. We have established a unique set of molecular tools suitable to interrogate single DNA, RNA and protein molecules, and interactions among these, directly in the complexity of the cell. These probes can provide parallel analyses of sets of molecules, and they reveal the location and interactions among molecules directly in cells and tissues. Padlock probes are linear oligonucleotides that are converted into DNA circles when both ends hybridize to a target DNA sequence. The probes can be added in large numbers to simultaneously interrogate great numbers of target sequences. It is also possible to couple the target-dependent ligation of the probes to a localized amplification reaction via rolling-circle replication, permitting detection of the location of target DNA or RNA sequences. Proximity ligation makes use of specific binding reagents, for example antibodies, coupled to oligonucleotides for proximity-dependent ligation reactions that depend on dual recognition of target molecules. The assay currently permits detection of as little as a few hundred target proteins in solution, and even single protein molecules in situ. By a suitable choice of binding reagents also non-protein molecules can be detected, and the interaction between pairs or trios of molecules can be visualized directly in cells.
O1.1 META-ANALYSIS OF GENETIC STUDIES IN MAJOR DEPRESSIVE DISORDER A. Cecile, J.W. Janssens1, Sandra Lopez Leon2, Angela M. Gonza´lezZuloeta Ladd3, Jurgen Del-Favero4, Stephan Claes4, Ben A. Oostra5
693
1
Dep. of Public Health, Erasmus University, Rotterdam Erasmus MC, Den Haag Genetic-Epidemiology Unit, Dep. of Epide, Rotterdam 4 Dep. of Molecular Genetics, Flanders Int, Antwerp 5 Department of Clinical Genetics, Erasmus, Rotterdam 2 3
Introduction: The genetic basis of major depressive disorder (MDD) has been extensively investigated. However, identification of MDD genes has been hampered by conflicting results from underpowered studies. We reviewed all genetic case-control association studies published to date and performed meta-analyses for polymorphisms that had been investigated in at least three studies. Methods: Electronic databases were searched up to April 2006 for all case-control studies in which cases were diagnosed with MDD according to standard diagnostic criteria (DSM, ICD or RDC), with controls derived from the general population. Pooled odds ratios (ORs) with 95% confidence intervals (CI) were calculated using randomeffects models. Results: The 162 papers that met our search criteria studied 274 polymorphisms in 82 genes. Only 19 polymorphisms (17 genes) had been investigated in at least 3 studies. In addition to our previous metaanalysis on DRD4 (OR, 1.73; CI, 1.29–2.32), we found statisticallysignificant associations with MDD for GNB3 C825T (OR, 1.49; CI, 1.141.95), MTHFR C677T (OR, 1.24; CI, 1.00–1.55), APOE a˚2 (OR, 0.43; CI, 0.18–1.02) and SLC6A4 (OR, 1.12; CI, 1.02–1.21). Conclusion: Our meta-analysis identified five MDD susceptibility genes. Yet we were only able to perform meta-analyses for 21% of the proposed candidate genes. Further research is needed to replicate the association studies of the remaining MDD candidate genes. O1.2 RESULTS OF A WHOLE GENOME ASSOCIATION STUDY FOR MAJOR DEPRESSIVE DISORDER (MDD) Hakan Sakul1, David A. Hinds2, Kelly A. Frazer2, David R. Cox2, Craig Hyde1 1 Pfizer Global R&D, CT 2 Perlegen Sciences, CA Many genetic and environmental factors affect response to treatment with an SSRI or placebo in MDD clinical trials. While many candidate genes have been studied, results are often discordant and difficult to interpret due to small sample size, multiple testing, etc. The primary objective of this study was to identify genes affecting SSRI or placebo response, as well as time-to-response, to improve the design of MDD trials. DNA from 1,024 Caucasian subjects from 8 MDD trials were genotyped using 250,000 SNPs and analyzed using various statistical models. Primary endpoints were Total HAMD, and three subscales: Core Depression, Anxiety and Insomnia. Statistical significance was based on overall false discovery rate (q-value). The following numbers of SNPs were significant and annotated as CNS-relevant or novel based on the literature: Placebo
SSRI
Response variable
CNS
Novel
CNS
Novel
Binary Quantitative Time to response
4 11 1
168 145 153
14 11 8
169 142 129
False discovery rates were also computed for 11 CNS candidate genes and SNP rs10891539, located in the TTC12 gene (11q23.1) within 50 kb of the D2 dopamine receptor (DRD2), was the most significant marker. This SNP is a part of a 3 SNP haplo-block spanning both genes (D’ > 0.9). Most literature reports for DRD2 highlight associations with the bipolar disorder while our data is from unipolar depression. As such, our finding may have a significant impact on MDD trials. Further, the effects of each allele of this SNP suggested that with all studies combined, screening out subjects homozygous for the reference allele results in SSRI treatment significance that is double the observed value without screening. For 2 of the 5 trials where SSRI treatment response was not significant, such screening would result in significant SSRI response. In conclusion, we have identified several genes with significant effects on either SSRI or placebo response in MDD trials. Our analyses also demonstrate that even the large sample size of this study has limited statistical power for replication, and we anticipate that our results will facilitate external replication studies and possible collaborations to join datasets.
694
Abstracts
O1.3 SIGNIFICANT LINKAGE OF MAJOR DEPRESSION ON 15q25-q26 AFTER SNP FINE-MAPPING Douglas Levinson1, Oleg Evgrafov2, James A. Knowles2, James B. Potash3, Myrna M. Weissman2, William A. Scheftner3, R. Raymond, Jr. DePaulo3, Raymond R. Crowe4, Kathleen Murphy-Eberenz5, Diana H. Marta3, Melvin G. McInnis5, Philip Adams2, Madeline Gladis5, Erin B. Miller3, Jo Thomas3, Peter A. Holmans4 1 Stanford University, Palo Alto, CA 2 Department of Psychiatry, Columbia University, New York, NY 3 Department of Psychiatry, Johns Hopkins, Baltimore, MD 3 Department of Psychiatry, Rush University, Chicago, IL 4 Department of Psychiatry and Mental Health, Iowa City, IW 5 Department of Psychiatry, University of Philadelphia, PA 4 Department of Psychological Medicine, Ca, Cardiff The authors studied a dense map of SNP markers on chromosome 15q25-q26 to maximize the informativeness of genetic linkage analyses in a region where they previously reported suggestive evidence for linkage of recurrent, early-onset major depressive disorder (MDDRE). In 631 European-ancestry MDDRE families, 88 SNPs were genotyped and multipoint linkage analyses carried out (ALLEGRO). Markermarker linkage disequilibrium (LD) was minimized (r2 < 0.05); a simulation study using founder haplotypes from these families suggested that linkage scores were not inflated by LD. The dense SNP map increased information content from around 0.7 to over 0.9. The maximum evidence for linkage was ZLR ¼ 4.69 (Kong-Cox LOD ¼ 4.78) at 109.8 cM (92.6 Mb, NCBI Build 34). The exact p-value (0.0000014) was below the genome-wide significance threshold (0.00002). By contrast, in the genome scan using microsatellite markers at 9 cM spacing, the maximum ZLR score for Europeanancestry families was 3.43 (106.53 cM, LOD ¼ 2.56). It was estimated that the linked locus or loci in this region might account for a 20% or less population-wide increase in risk to siblings of cases. This region has produced modestly positive evidence for linkage to depression and related traits in other studies. These results suggest that DNA sequence variations in one or more genes in the 15q25-q26 region can increase susceptibility to major depression, and that efforts are warranted to identify these genes.
O1.4 ASSOCIATION OF THE MUSCARINIC ACETYLCHOLINE RECEPTOR M2 (CHRM2) GENE WITH MAJOR DEPRESSIVE DISORDER Sarah Cohen1, Dasha Gaysina1, Farzana Hoda1, Ania Korszun2, Michael J Owen3, Nick Craddock3, Anne Farmer1, Ian Craig1, Peter McGuffin1 1 Institute of Psychiatry, London 2 Queen Mary School of Medicine and Dentistry, London 3 Cardiff University, Cardiff The substantial genetic contribution to depression predisposition is well-established; however, specific genes have yet to be consistently identified. The CHRM2 gene is located within a linkage region at chromosome 7q31–35, and has been associated with a range of psychiatric disorders. A significant association has been reported with depression in four SNPs in the CHRM2 gene, and a three SNP haplotype. A second group has also published a positive association; however, the SNPs investigated were not identical to those in the first report. We set out to replicate the initial haplotype (rs1824024, rs2061174, rs2350786) using a large case-control sample. The sample, collected at 3 UK Centres, consists of 1204 cases that have had 2 or more episodes of ICD10/DSMIV depression of at least moderate severity assessed by SCAN interview and 849 healthy controls screened by questionnaire and telephone interview for absence of psychiatric disorder. A significant association was observed with the TCT haplotype and major depression (p ¼ 0.002), and a trend with the TCA haplotype and major depression (p ¼ 0.06). Although our results support emerging evidence that the CHRM2 gene may be an important factor in the genetic aetiology of major depressive disorder, they should be approached with caution, as the significant results relate to a rare haplotype and differ somewhat from the results of the previous group. Acknowledgements: Depression case-control collection was supported by a grant from the Medical Research Council (UK); SC is in receipt of an MRC studentship.
O1.5 GENOTYPE ENVIRONMENT INTERACTION FOR RISK OF DEPRESSION Nicholas Martin1, Will Coventry1, Michael James1, Nathan Gillespie2, Lindon Eaves2 1 Queensland Institute of Medical Research, Queensland 2 Virginia Institute of Psychiatric and Be, Richmond In 2003, Caspi and colleagues published a landmark paper in Science reporting an interaction between a polymorphism (5HTTLPR) in the promoter region of the serotonin transporter gene (5HTT renamed SLC6A4) and the number of life events experienced, in predicting risk of depression. This paper has been widely cited and has become the dominant research paradigm in psychiatric genetics, since it appears to reconcile the competing claims of genetic and environmental etiology. There have been numerous replication studies, with highly inconsistent results, some supporting the interaction as reported by Caspi, some finding no interaction, and others reporting interaction in the opposite direction. Furthermore, it has recently become apparent that genotyping of the 5HTTLPR polymorphism is somewhat more complicated than previously thought, casting doubts on the veracity of earlier studies. Most importantly, it has been suggested that the GxE detected may be an artifact of the unsatisfactory psychometric properties of the psychiatric constructs being employed. We are analysing data from our large twin family collection (5,200 individuals) for whom depression diagnoses, risk factors and DNA samples are available, to bring considerable power to the question of GxE interactions for depression. Furthermore, we are investigating whether the inconsistently-replicated interaction of 5HTTLPR (and, independently, any SNP haplotype) with stressful life events on risk of depression is moderated by levels of social support, and whether this could explain the inconsistency. We are also exploring the extent to which IRT modelling will help resolve the effects of GxE from artifacts of measurement. O1.6 5HTTLPR s/s GENOTYPE MODIFIES THE EFFECT OF STRESSFUL LIFE EVENTS INCREASING RISK FOR DEPRESSION: THE PREDICT-GENE STUDY Jorge Cervilla1, Margarita Rivera1, Esther Molina1, Juan de Dios Luna1, Francisco Torres-Gonza´lez1, Juan Bello´n1, Berta Moreno-Ku¨stner1, Michael King2, CORE GROUP PREDICT STUDY3, Blanca Gutie´rrez1 1 University of Granada, Granada 2 Royal Free Hospital, London 3 PREDICT STUDY, EUROPE The PREDICT-Gene study attempts to replicate reported GxE interactions between the s/s genotype at the 5HTTLPR locus and previously-suffered life-events (LEs) increasing risk for depression. ICD-10 depression was ascertained using the CIDI. Exposure to LEs was assessed using the List of Threatening Experiences. Subjects were genotyped to determine 5HTTLPR genotype (s/s, s/l or l/l). Logistic regression was used to test interaction. 641 subjects (183 depressed24.5%- and 458 controls) were included. Among depressed patients, 30.6% had the s/s genotype, compared to 21.4% among controls (OR ¼ 1.62; 95%CI:1.1–2.3; p ¼ 0.01). Having 2þ LEs was also significantly more common amongst the depressed compared to controls (OR ¼ 1.86;IC95% [1.47–2.35] p < 0.0001). Regarding the likelihood ratio for the interaction between the s/s genotype and exposure to 0, 1 or 2 þ LEs, we found a significant result showing LEs by s/s genotype interaction (LR X2 ¼ 6.28, p ¼ 0.043). Thus, after having been exposed to just one LE, s/s subjects had increased risk for depression compared to l/l or l/s subjects (OR ¼ 3.3 95% CI:1.72–6.5; p ¼ 0.0001). Those subjects with l/l or l/s genotypes required a greater degree of exposure to LEs (2þ) for a similar risk level for depression (Test for different probabilities across all levels of exposure: LRX2 ¼ 44.33; p < 0.0001). O1.7 STRESS RELATED NEGATIVE AFFECTIVITY AND GENETICALLY ALTERED 5-HTT FUNCTION: EVIDENCE FOR SYNERGISM IN SHAPING RISK FOR DEPRESSION Nele Jacobs Open University the Netherlands, Heerlen Context: Genetic moderation of the depression-inducing effects of stressful life events (SLEs) has been reported, but findings suggest that
Abstracts genes may not moderate the effects of SLEs per se, but instead moderate the risk of depression associated with the stable tendency to develop negative emotions in response to minor environmental experiences. Objective: To examine whether a functional polymorphism of the serotonin transporter gene (5-HTTLPR) moderates the association between negative affectivity (neuroticism, N) and depression, and to what degree this can explain previous findings involving SLEs. Setting and design: A population-based sample of ethnically homogeneous adult female twins (n ¼ 274) in a longitudinal study design. Main outcome measure: A continuous score of self-reported depressive symptoms. Results: The depressogenic effect of SLEs in the three months prior to interview was significantly greater in women with two short (S) alleles compared to women with one or none. However, this effect disappeared after taking into account the effect of SLEs conditional on N. Similarly, the depressogenic effect of N increased progressively with the number of S alleles, a situation that remained unchanged after taking into account the effect of N conditional on SLEs. Conclusion: Genotype environment interactions in depression may be more productively interpreted by involving mechanisms closer to the psychological experience itself. The probability that stress-related cognitive vulnerabilities for depression result in symptom formation may be moderated by a neurobiological phenotype characterised by altered processing of negative emotions, associated with variation in 5-HTTLP. O1.8 5-HTTLPR GENOTYPE INFLUENCES 5-HT1A RECEPTOR BINDING IN DEPRESSION Peter Eichhammer1, Susan E. Kennedy2, Tiffany M. Love2, Robert A. Koeppe3, Philipp Sand4, Jon-Kar Zubieta2 1 Dep. of Psychiatry, Uni Regensburg, Regensburg, Bavaria 2 MBNI, University of Michigan, Ann Arbor, Michigan 3 Dep. of Nuclear Medicine, University of Ann Arbor, Michigan 4 Dep. of Psychiatry, University of Regens, Regensburg, Bavaria Recent studies have given rise to the assumption that the crosstalk between the human 5-HT transporter and 5-HT1A receptors is genetically determined. In particular, a functional promoter polymorphism of the 5-HT transporter gene (5-HTTLPR) alters HT1A receptor availability in healthy individuals, as assessed by brain positron emission tomography (PET). Since both molecular structures are closely involved in the regulation of emotional behaviour, we attempted to assess the relationship between 5-HTTLPR genotype and 5-HT1A receptor availability in major depressive disorder (MDD). 15 unmedicated patients with MDD and 17 healthy controls were examined using the 5-HT1A-selective ligand [11C]WAY 100635 and underwent genotyping. Binding analyses revealed a significant decrease in 5-HT1A receptor availability in carriers of the short 5-HTTLPR allele relative to other subjects. Effects on receptor BP were most pronounced in temporal/hippocampal regions of the brain (p < .01). Based on these findings, it is tempting to conclude that 5-HTTLPR may influence inhibitory and neurogenic processes in depressive patients, at least in part via regulation of 5-HT1A receptor availability. These results may further contribute to a better understanding of 5-HTT/5-HT1A interplay and its relevance in shaping clinical phenotype and outcome in major depression. O2.1 BIPOLAR DISORDER CANDIDATE GENE CUSTOM SNP PANEL: DESIGN AND PRELIMINARY RESULTS Margit Burmeister1, Laura J. Scott2, Yun Li2, Robert C. Thompson2, Jun Li3, Fan Meng2, Weihua Guan2, Devin Absher3, Marquis P. Vawter4, Prabhakara Choudary5, Hiroaki Tomita4, Simon J. Evans2, William E. Bunney4, Edward G. Jones5, Jack D. Barchas6, Alan F. Schatzberg3, Huda Akil2, Stanley J. Watson2, Richard M. Myers3, Michael Boehnke2 1 University of Michigan, Ann Arbor MI 2 Univ. of Michigan, Ann Arbor, MI 3 Stanford Univ., Stanford, CA 4 UC Irvine, Irvine, CA 5 UC Davis, Davis, CA 6 Weill Medical College of Cornell, New York, NY
695
We designed a Bipolar Disorder candidate gene panel for the 1536 SNP Illumina Golden Gate assay. Inclusion criteria were association with Bipolar Disorder in 1 large study or > 2 smaller studies, reproducible or RT-PCR confirmed difference in expression between bipolar disorder and controls in microarray studies from brain mRNA, strong evidence of biological implication or mouse model with relevant phenotypes. Genes were defined as all exons and introns, plus 10 kb 5’ to the start of transcription and 5 kb 3’ to the last base of the longest transcript. HapMap II SNPs with minor allele frequency > 0.05 were identified for all SNPs in > 120 genes. After eliminating SNPs with an Illumina design score < 0.6, optimal SNPs were chosen to cover all HAPMAP bins and singletons by at least 1 SNP within LD of r2 > 0.8. All non-synonymous SNPs were added regardless of MAF and HapMap. Some large genes with > 100 HapMap SNPs had to be excluded for efficiency, as were genes without HapMap SNPs, usually due to genome map uncertainty. The final list includes 93 genes, with a median of 11 SNPs per gene (average 17; range 1 to 118). We will present preliminary results using this panel in an association study of 476 cases and 470 controls from the NIMH genetic initiative. We thank the NIMH Human Genetics Initiative and the many contributors to that collection for making the samples available, and the Pritzker Neuropsychiatric Disorders Research Fund L.L.C. for funding. O2.2 ASSOCIATION OF GRK3 AND P2RX7 WITH BIPOLAR DISORDER IN A FRENCH CANADIAN POPULATION Nicholas Barden1, John Kelsoe2, Tom Barrett2, Bernard Gagne´3, Michel Labe´3, Mario Harvey3 1 CHUL research Centre, Quebec 2 UCSD, San Diego 3 CRCHUL, Quebec Polymorphisms in P2RX7 on chromosome 12q24.31 predispose to bipolar disorders (BP). While studies interrelate immunity and depression and the role of the purinergic receptor P2X7 in immune responses is well documented, the relationship of this to BP remains unknown. Recent functional studies demonstrated interaction of P2RX7 with another susceptibility gene for BP, GRK3. This gene encodes for the adrenergic receptor tyrosine kinase beta 2 (ADRBK2) and phosphorylates the intracellular domain of P2X7 receptors to recruit beta-arrestins and dynamin and to initiate receptor endocytosis, degradation or redistribution. Genetic analysis in BP triads reported association with markers in the 50 -regulatory region of GRK3. Since mutations in P2RX7 are associated with BP in the SLSJ region of Quebec, we tested whether polymorphisms in GRK3 could also be associated with BP in this population. We performed a SNP-based genetic analysis upstream of GRK3 in a BP case/control sample (cases ¼ 214, controls ¼ 208) and 43 pedigrees (551 ascertained individuals) from the SLSJ region. We performed allelic and genotypic association studies on 4 SNP markers using CLUMP and found one significantly positive (p < 0.05) allelic association. Analysis of SNP haplotypes using PHASE and cocaphase confirmed the association between the GRK3 5’-regulatory region and BP. It is possible that interaction between GRK3 and P2XR7 are important for the development of BP. Evaluation of the role of GKR3 in ATP-mediated P2RX7dependent pathways could lead to the identification of new pathways involved in the pathophysiology of mood disorders. O2.3 IDENTIFICATION OF A BIPOLAR DISORDER SUSCEPTIBILITY LOCUS ON CHROMOSOME 15Q Erica Zoe McAuley1, Ian P. Blair1, Jan M. Fullerton1, Jennifer A. Donald2, Philip B. Mitchell3, Peter R. Schofield4 1 Garvan Institute of Medical Research, NSW 2 Macquarie University, NSW 3 University of New South Wales and Black, NSW 4 Garvan Institute of Medical Research, NSW Bipolar disorder (BP) is a highly heritable, common, severe mood disorder with lifetime prevalence up to 4%. We conducted a 10 cM genome scan on 35 multigenerational pedigrees with 288 genotyped individuals and identified linkage to chromosome 15q25-26. Using a broad disease definition including individuals diagnosed with BPI, schizoaffective disorder mania type, BPII or recurrent unipolar depression, 130 individuals in the cohort were affected. We found a maximum two point LOD score of 3.85 and a maximum multipoint LOD score of 4.76 at D15S130 using 60% age-specific penetrance, the dominant model under broad disease definition. Markers on chromosomes 4q, 6q, and 13q suggested linkage, in addition to a second marker
696
Abstracts
adjacent to D15S130. The region on 15q25-26 was further investigated by typing 8 additional microsatellite markers, yielding significant scores with a maximum two point LOD score of 4.66 and a maximum multipoint LOD score of 5.17 at marker D15S1004. A 95% confidence interval calculated by the Zmax-1 method suggests a support interval spanning 17 cM between markers D15S979 and D15S816 (86.692.8 Mb). Haplotype analysis based on pedigree-specific, identicalby-descent, allele sharing supported the 95% confidence interval estimates. The 15q25-26 locus is supported by linkage findings from studies on recurrent early-onset major depressive disorder, BP with psychotic features and a study of schizophrenic and BP subjects, suggesting that the locus might contain a gene conferring susceptibility to both mood and psychotic disorder.
O2.4 GENOME-WIDE SIGNIFICANT EVIDENCE FOR LINKAGE ON CHROMOSOME 10q IN 10 MULTIPLEX BP FAMILIES Tine Venken1, Maaike Alaerts1, Daniel Souery2, Cornelia van Duijn3, Julien Mendlewicz2, Christine Van Broeckhoven1, Stephan Claes2, Jurgen Del-Favero1 1 University of Antwerp, VIB8, Antwerpen 2 University of Brussels, Brussels 3 Erasmus Medical Center, Rotterdam 2 University Hospital Gasthuisberg, Leuven This study presents a 10 cM density genome-wide scan in multiplex pedigrees with bipolar affective disorder (BP). As part of the Scientific Programme on ‘Molecular Neurobiology of Mental Illness’, funded by the European Science Foundation, 10 of 15 families were selected for inclusion, of which 2 were of Jewish Ashkenazi and 8 of EuropeanCaucasian descent. Genome-wide linkage analysis revealed 2 loci with evidence suggesting linkage located on chromosomes 10q21.3-23.3 (LOD 1.97) and chromosome 19q13.2-13.4 (LOD 2.41). Follow-up genotyping applying a 2 cM density STR map confirmed linkage at both loci. There was less evidence for the chromosome 19q region, but it still suggested linkage (LOD 2.13), whereas evidence for the chromosome 10q region increased to a significant level (LOD 3.28). Most of the evidence for the chromosome 10q region comes from a single, large Ashkenazi Jewish pedigree. Haplotype analysis on this pedigree shows that all affected individuals share the same haplotype between markers D10S210 and D10S1765, comprising a chromosomal region of 30 cM. This region has been found to be a candidate region for BP in several linkage analysis studies and in a large meta-analysis. It has also been implicated in a linkage study on schizophrenia (SZ) in Ashkenazi Jewish families. These results suggest that a mutation in a gene on chromosome 10q21-q23 is present in the Ashkenazi Jewish population, increasing susceptibility both to SZ and BP.
O2.5 HAPLOTYPE ASSOCIATION STUDIES OF THE DISC1 GENE IN FINNISH BIPOLAR FAMILIES Outi Marika Palo, William Hennah, Helena Kilpinen, Kaisa Silander, Pia Soronen, Mervi Antila, Annamari Tuulio-Henriksson, Tuula Kieseppa¨, Timo Partonen, Jouko Lo¨nnqvist, Leena Peltonen, Tiina Paunio National Public Health Institute, Helsinki Bipolar disorder (BPD) and schizophrenia have at least partially convergent etiologies and thus may share some susceptibility loci predisposing individuals to mental disorders. Multiple linkage and association studies have suggested that DISC1 (1q42) is involved in some neurodevelopmental conditions. In the Finnish population, allelic DISC1-haplotypes showed association with schizophrenia, and tentatively with autism and Asperger’s syndrome. Our study sample consisted of 179 BPD families (n ¼ 720), of which 227 individuals were diagnosed with bipolar type I, II, or cyclothymia, and 254 have bipolar disorder with psychotic features. 57 randomly-selected Finnish trios were used as a control sample. We genotyped 13 haplotypetagging SNPs in the cluster of TSNAX, DISC1 and DISC2 genes. A two-SNP haplotype at the 3’ end of the DISC1 yielded significant association (p ¼ 0.00064) with the bipolar disorder spectrum. The previously-described two-SNP risk haplotype HEP3, and a three-SNP haplotype located at the 5’ end, displayed association with bipolar disorder with psychotic features in males (p ¼ 0.00011 and p ¼ 0.00123, respectively). The present study provides evidence for association between BPD with psychotic features and a schizophrenia-associated
haplotype of DISC1. Another two-SNP haplotype association was observed for BPD at the telomeric end of DISC1. These results support the involvement of DISC1 in BPD and suggest that different variants within DISC1 could contribute to variations in the phenotype. We intend to analyse quantitative variables derived from neuropsychological test data, some of which may represent intermediate phenotypes of BPD. O2.6 EVIDENCE FOR INTERACTING SUSCEPTIBILITY GENES ON CHROMOSOME1P AND 4q AS WELL AS ON 2q AND 6q IN 108 FAMILIES WITH BIPOLAR AFFECTIVE DISORDER Rami Abou Jamra1, Robert Fu¨rst2, Kaneva Radka3, Antonia Flaquer2, Thomas G. Schulze3, Guillermo Orozco Diaz4, Fabio Rivas4, Vihra Milanova5, Sven Cichon6, Peter Propping7, Marcella Rietschel3, Thomas F. Wienker2, Markus M. Nothen6, Johannes Schumacher7 1 University of Bonn, Bonn 2 Institute of Medical Biometry, Informati, Bonn 3 Laboratory of Molecular Pathology, Medic, Sofia 3 Central Institute of Mental Health, Mannheim 4 Civil Hospital Carlos Haya, Malaga 5 Department of Psychiatry, Medical University, Sofia 6 Department of Genomics, Life&Brain Cente, Bonn 7 Institute of Human Genetics, University, Bonn Linkage evidence between bipolar affective disorder (BPAD) and loci on chromosomes 1p, 2q, 4q, and 6q was observed in four European samples consisting of 108 families with German, Spanish, Bulgarian, and Roman descent. Here we report on a systematic interaction analysis using all 4 regions under a broad and narrow disease definition. In each region, non-parametric linkage analysis was performed after conditioning on linkage evidence to the other three regions. On chromosome 2q we observed a delta-NPL-Score of 5.93 (increase of the initial NPL finding) under a broad disease definition when conditioning on linkage signals to chromosome 6q. vice versa, we found a delta-NPL-Score of 5.08 on 6q when conditioned on the initial linkage data to chromosome 2q. On chromosome 1p, a delta-NPL-Score of 3.12 was observed using a narrow disease definition and conditioned on linkage findings to ¨ NPL-Score of 2.41 on chromosome 4q was found chromosome 4q, and a A when conditioned on linkage signals to chromosome 1q. Our results point to an interaction between chromosomal loci on 1p and 4q, and between 2p and 6q. Susceptibility genes located in these regions may have interfering regulative effects or interact together within the same biological pathway. O2.7 DOCK9 VARIANTS ARE ASSOCIATED WITH SEVERAL BIPOLAR DISORDER PHENOTYPES AND INTERACT WITH G72 Sevilla D. Detera-Wadleigh1, Chun-yu Liu2, Imer Cardona1, Winston Corona1, Nirmala Akula1, CJM Steele1, Mukta Kundu1, Tom I. Bonner3, Judith A. Badner2, Manjula Maheshwari4, Richard Gibbs4, Elliot S. Gershon2, Francis J. McMahon1 1 Genetic Basis for Mood and Anxiety, Mood and Anxiety Disorders Program, National Institute of Mental Health Intramural Research Program, National Institutes of Health, Bethesda, MD 20892 2 Department of Psychiatry, University of Chicago, Chicago, IL 60637 3 Laboratory of Genetics, National Institute of Mental Health, National Institutes of Health, Bethesda, MD 20892 4 Baylor College of Medicine, Houston, TX 77030 To identify variation(s) underlying the bipolar disorder linkage on 13q32, we have studied DOCK9, a candidate gene detected via a linkage disequilibrium (LD) screen. LD mapping in the NIMH 1 to 4 datasets displayed association with promoter and intron-1 SNPs in the NIMH 1–2 series (9 SNPs, p ¼ 0.02–0.006), supported in NIMH 3 (one SNP, p ¼ 0.039), but not in NIMH 4. To examine whether association may have been masked by phenotypic heterogeneity, we analyzed familial component phenotypes of bipolar disorder. Increased evidence observed in NIMH 1–2, replicated in NIMH 3; e.g., racing thoughts during mania: NIMH 1–2, 16 SNPs (p ¼ 0.037–0.0008) and NIMH 3, 9 SNPs (p ¼ 0.032–0.007). Similar results were seen for psychosis, suicide attempt, and mania at onset. These signals were either absent in NIMH 4 or involved opposite alleles, suggesting epistasis. To test this hypothesis, we performed a family-based likelihood test of transmission of DOCK9 alleles conditional on genotype at the G72
Abstracts marker M23. This test identified overtransmission of the ancestral genotype of M23 and DOCK9 alleles in the entire NIMH dataset (p-0.02). Overexpressed G72 peptide has been shown to activate D-amino acid oxidase, catalyzing the oxidation of D-serine, an NMDA receptor agonist. DOCK9 activates Cdc42, a RhoGTPase. These pathways converge at the level of dendritic growth, suggesting a biological basis for the observed association at both genes and the apparent epistasis between them. O2.8 VARIATION AT GRM3 INFLUENCES SUSCEPTIBILITY TO PSYCHOTIC BIPOLAR DISORDER Elaine Green1, Detelina Grozeva2, Nadine Norton2, Ian Jones2, Lisa Jones3, Mick O’Donovan2, Mike Owen2, George Kirov2, Nick Craddock2 1 University Hospital of Wales, Cardiff 2 Cardiff University, Cardiff 3 University of Birmingham, Birmingham The metabotropic glutamate receptor 3 (GRM3) is a schizophrenia (SZ) candidate gene. Although findings have been inconsistent, support has come from multiple studies. Interestingly, Weinberger’s group reported the major allele (A) of rs6465084 was associated with an increased risk of SZ and alteration of glutamate neurotransmission, prefrontal and hippocampal physiology and cognition. Our previous GRM3 study in UK SZ case-control sample showed no evidence of association. GRM3 has been less studied in bipolar disorder (BP), although one previous study reported an association. GRM3 lies on chromosome 7q21.1-q21.2, the region that received strongest support in our first stage UK-Irish BP sibling-pair genome screen. We examined polymorphisms at GRM3 in our UK DSMIV BPI disorder cases (n ¼ 661) and UK controls (n ¼ 1377). We found significant association at rs6465084 (A allele: cases: 78.5%; controls: 75.0%; odds ratio [OR] ¼ 1.19, 95% confidence interval [CI], 1.02–1.40; p ¼ 0.027). The signal came predominantly from the BPI cases with lifetime psychotic features (n ¼ 417) (A allele: 80%; OR ¼ 1.30, 95% CI, 1.08–1.58; p ¼ 0.0061) and the signal was stronger in the subset of cases in which at least half of all mood episodes had prominent psychotic feature (n ¼ 166) (A allele: 82.5%, OR ¼ 1.54, 95% CI, 1.15–2.08; p ¼ 0.0038). Our findings suggest that GRM3 may play a role in influencing susceptibility to illness with both bipolar and prominent psychotic features. O3.1 GENOMEWIDE SNP LINKAGE SCAN OF SCHIZOPHRENIA IN A LARGE MULTICENTER SAMPLE Douglas Levinson1, Pablo V. Gejman2, Claudine Laurent3, Michael J. Owen4, Ann E. Pulver5, Brien Riley6, Dieter B. Wildenauer7, Kenneth S. Kendler6, Jacques Mallet3, Bryan J. Mowry4, Gerald Nestadt5, Michael O’Donovan4, Alan R. Sanders2, Sibylle G. Schwab7, Nigel Williams4, Margot Albus5, Ste´phanie Bauche´3, Nicola deMarchi4, Dimitris Dikeos4, Jubao Duan2, Maurice Jay3, Virginia K. Lasseter5, F. Bernard Lerer6, Wolfgang Maier4, Deborah A. Nertney4, Ivan Nikolov4, Nadine Norton4, Anthony O’Neill5, George Papadimitriou4, Ricardo Segurado4, Jeremy M. Silverman5, Dermot Walsh6, Hywel Williams4, Peter A. Holmans4 1 Stanford University, Palo Alto, CA 2 Center for Psychiatric Genetics, Northwestern University, Evanston, IL 3 Laboratoire de Ge´ne´tique Mole´culaire de, Paris 4 Department of Psychological Medicine, Ca, Cardiff 5 Department of Psychiatry, Johns Hopkins, Baltimore, MD 6 Virginia Institute for Psychiatric and B, Richmond, VA 7 University of Western Australia, Perth, 4Department of Psychiatry, University of Brisbane 5 State Mental Hospital, Haar 4 Clinica Psichiatrica, Seconda Universita`, Naples 6 Department of Psychiatry, Hadassah-Hebre, Jerusalem 5 Department of Psychiatry, Queens University, Belfast 5 Department of Psychiatry, Mt. Sinai School of Medicine, New York, NY 6 The Health Research Board, Dublin A genomewide linkage scan was carried out in pedigrees from 8 samples, including 971 families with 4540 genotyped subjects, 2120 affected under a narrow diagnostic model (schizophrenia or schizoaffective disorder). Of 5992 non-Y SNP markers (Illumina map), 5903 met
697
QC criteria. Marker-marker linkage disequilibrium was analyzed (Haploview), and maps constructed that maximize marker heterozygosities while constraining marker-marker rsquared to 0.05 or less. Multipoint linkage analyses were carried out (ALLEGRO, exponential model). Analyses of six chromosomes were completed so far for the primary analysis (755 European-ancestry pedigrees, informative by the narrow model, with 3479 genotyped individuals, 1727 affected). Suggestive evidence for linkage was observed on chromosome 8p at 46.8 deCode cM (pointwise p < 0.0005, information content 0.90), consistent with findings in other samples, and it is of interest that the wellknown candidate gene NRG1 lies outside the 1-lod support interval (42.6–50.5 cM), but it is not yet known whether NRG1 and/or other loci account for the linkage evidence observed in these studies. Notably, on the chromosomes analyzed so far, 25–30% of SNPs had to be removed to reduce LD, otherwise very large spurious linkage peaks were observed in regions with multiple high-LD marker pairs. Ongoing are analyses of all chromosomes for the European-ancestry families, of all families, of linkage taking cross-site heterogeneity into account, and of the effect of these and other recent results on Genome Scan Meta-Analysis results for schizophrenia. O3.2 GENETIC ANALYSIS OF THE MYELIN OLIGODENDROCYTE GLYCOPROTEINGENE IN SCHIZOPHRENIA Nagahide Takahashi1, Iritani Shuji1, Aoyama Nagisa1, Saito Shinichi1, Ishihara Ryoko1, Inada Toshiya2, Ozaki Norio1 1 Nagoya University, Nagoya, 2 Teikyo University, Chiba Introduction: Several lines of evidence suggest that abnormal myelination and oligodendrocyte function are related to the development of schizophrenia. The Myelin Oligodendrocyte glycoprotein gene (MOG), located on 6p21, is proposed to play an important role in maintenance and completion of the myelin sheets. Recently, Tkachev et al. showed that the expression of MOG significantly reduced in the postmortem brain of patients with schizophrenia using microarray and Q-PCR analyses. Additionally, it has been reported that MOG is significantly associated with schizophrenia in the Chinese and Canadian populations. Methods: Following evaluation of linkage disequilibrium, singlemarker and haplotype analyses were performed in the polymorphisms located in MOG in 384 subjects with schizophrenia meeting the criteria of DSM-IV-TR and 384 healthy controls. All subjects are Japanese descent. Results: Although haplotype analysis showed weak association of this gene with schizophrenia (p value ¼ 0.033), none of the single nucleotide polymorphisms (SNPs) reached statistical significance after the Bonferroni’s correction. Mutation search detected one SNP which causes non-synonymous amino acid change in the exon 4, however, no association was observed with schizophrenia in this SNP. A microsatellite marker located on the 3’ UTR of MOG yielded significant evidence of association with schizophrenia (p value ¼ 0.00034). Conclusion: The present results suggest that MOG contribute to the etiology of schizophrenia in the Japanese population. Neuropathological and the expression studies using the postmortem brain is now underway to elucidate the exact mechanisms of MOG for the development of schizophrenia in Japanese subjects. O3.3 DTNBP1, MUTED, AND BLOC1S2: MULTIPLE SUSCEPTIBILITY GENES IN THE BLOC-1 COMPLEX IMPLICATES VESICULAR TRAFFICKING DEFECTS IN SCHIZOPHRENIA Richard Straub, Kristin Nicodemus, Michael Egan, Terry Goldberg, Robyn Honea, Joseph Callicott, Daniel Weinberger NIMH, NIH, Bethesda, MD Genes and pathways functionally linked to dysbindin (DTNBP1) have naturally become primary targets of inquiry. Neither the specific cellular functions of dysbindin or the consequences of the reduced dysbindin transcript and protein levels found in postmortem schizophrenic brains is known. What is known is that dysbindin is a component of BLOC-1 (biogenesis of lysosome-related organelles complex), which includes at least 7 other proteins from the genes MUTED, PLDN, SNAPAP, GCN5L1, BLOC1S2, BLOC1S3, and CNO. Defects in some of these proteins affect the biogenesis and/or trafficking of intracellular organelles found in specialized secretory cells such
698
Abstracts
as melanocytes, and platelets. We sequenced in these genes, genotyped, and performed family-based and case-control tests of association and epistasis, using the clinical and intermediate phenotypes from the CBDB/NIMH Sibling Study. We found very few coding SNPs, so risk haplotypes act primarily via aberrant levels and timing of expression. Dysbindin, MUTED and BLOC1S2 were associated with the clinical phenotype. Epistasis between dysbindin and MUTED was evident, and each was in epistasis with COMT, an important determinant of cortical dopamine levels. Functional imaging showed that dysbindin and MUTED influenced activation of prefrontal cortex and hippocampus during working memory tasks and that MUTED affected hippocampal volume. Dysbindin strongly influenced a broad range of cognitive phenotypes, especially IQ / processing speed, but the other genes did not, suggesting that dysbindin may have cellular functions and effects on cognition independent of its role in BLOC-1. O3.4 HIGH-DENSITY OLIGONUCLEOTIDE TILING ARRAY BASED VERY-HIGH RESOLUTION CGH AND ITS APPLICATION TO THE ANALYSIS OF BREAKPOINTS IN CHROMOSOME 22Q AND IN 22Q11 DELETION SYNDROME Alexander Eckehart Urban1, Jan Korbel1, Fabian Grubert1, Rebecca Selzer2, Todd Richmond2, April Hacker3, Roland Green2, Beverly S. Emanuel3, Mark Gerstein1, Sherman M. Weissman1, Michael Snyder1 1 Yale University, New Haven, CT 2 NimbleGen Systems, Madison 3 Children’s Hospital of Philadelphia, Philadelphia 3 University of Pennsylvania, Philadelphia We have developed Very-High Resolution CGH based on high-density oligonucleotide tiling microarrays, which allows predicting breakpoints in the genomic sequence with an accuracy of regularly up to 200 bp while covering the length of an entire chromosome with a single experiment. We applied VHR-CGH to the analysis of several different cases of chromosomal aberrations on Chromosome 22q [Urban et al., PNAS 2006]. 22q11 Deletion Syndrome (22q11DS) results from a common submicroscopic deletion in 22q11 and is characterized by various developmental abnormalities, with 25–30% of the patients developing schizophrenia. 22q11DS thus offers a promising model for the study of the molecular etiology of schizophrenia. But because deletions are large and varied and the phenotypes are complex many challenges remain and advanced technologies are needed. Mskless Array Synthesis (MAS) is used to generate tiling microarrays with 385,000 oligonucleotide probes, with an 85 bp tiling path along the nonrepetitive sequence of Cs22q. The probe length is variable to have a uniform Tm over all probes, for maximum performance. We used these MAS-arrays to map deletion breakpoints in 12 patients with 22q11DS and can demonstrate that separate typical deletions, undistinguishable by conventional methods because they reach into LCR regions, may differ by up to several hundred thousand bp. We also mapped insertion breakpoints in one patient with der22-Syndrome and CatEye-Syndrome, respectively, and two patients with Dup22-Syndrome. VHR-CGH analysis readily allowed determining breakpoints of several aberrations and variations with PCR and DNA sequencing.
O3.5 PROTEOME PROFILING OF THE PREFRONTAL CORTEX (PFC) AND HIPPOCAMPUS OF SCHIZOPHRENIA BRAINS Sinthuja Sivagnanasundaram1, Nicole Caixeiro2, Maryam Nesvaderani2, Irina Dedova2, Stuart Cordwell3, Izuru Matsumoto2 1 NISAD Senior Research Officer, Sydney 2 NISAD, University of Sydney, Sydney 3 University of Sydney, Sydney Background: Structural and functional abnormalities have been identified in a number of brain regions, particularly the hippocampus and PFC (both grey and white matters) of schizophrenic subjects, including volume reduction and cytoarchitectural irregularity. Abnormalities in the activities and/or volume of the PFC and hippocampus (both anterior and posterior regions) seem to be well correlated with a number of cognitive, negative and positive symptoms. The molecular mechanisms underlying the neuronal pathology in both
regions are largely unknown. Hypothesis: We hypothesise that these mechanisms will be elucidated by applying a high- throughput proteomics approach to determine alterations in the protein expression profile between the schizophrenic and control brains. Method: Matched brain tissue samples of schizophrenic and control subjects were provided by the NSW Tissue Resource Centre (Sydney); BA46 grey matter (10 cases and 10 controls), BA46 white matter (10, 10), anterior (7, 7) and posterior (9, 9) hippocampus. Proteins were extracted and analysed by two-dimensional gel electrophoresis (2DGE) in duplicates. Protein profiles were analysed using Nonlinear Phoretix 2D Expression software, and proteins identified using MALDI-TOF mass spectrometry (MS). Results: We examined and compared the expression levels of over 600 and 1000 protein spots varying in intensity for the BA46 and hippocampus samples respectively. Differentially-expressed proteins and their isoforms involved in a number of cellular processes including myelination, microtubule formation and cytoskeleton regulation were identified. Conclusion: 2DGE in combination with MS can be successfully applied to identifying molecular factors and mechanisms that may underlie neuropathology in schizophrenic brains. O3.6 A TWO-STAGE DESIGN USING DATABASE AND NOVEL SNPS FROM 18 DOPAMINERGIC GENES SUGGESTS DOPAMINE TRANSPORTER AND DOPAMINE D3 RECEPTOR POLYMORPHISMS ARE RISK FACTORS FOR SCHIZOPHRENIA Michael Talkowski1, K.V. Chowdari1, Hader Mansour1, Joel Wood1, Brian Shirts1, Jianping Zhang2, Bernie Devlin1, Vishwajit Nimgaonkar1 1 University of Pittsburgh, Pittsburgh, PA 2 Mayo Clinic, MN Dopaminergic dysfunction has long been hypothesised in the etiology of schizophrenia (SZ). To test the dopamine (DA) hypothesis, we evaluated selected DA genes in a two-stage study design, using published SZ linkage information (Lewis et al., 2003) to prioritise our analyses. In phase I, we analysed 101 database SNPs from 18 genes among 150 SZ trios and 280 controls. Phase II involved follow-up in an enlarged case sample (n ¼ 478) and independent control sample (n ¼ 500) using a comprehensive set of tag SNPs for genes localised to linked regions (COMT, DRD2, NR4A2, VMAT1) and functional candidates from phase I meeting a pre-defined threshold for suggestive associations after corrections for multiple testing (p < 0.003 for individual SNP tests; DRD3 and SLC6A3 (DAT)). Tag SNPs (r2 > 0.8 between loci) were derived using HapMap data alone (DRD2, NR4A2, VMAT1, SLC6A3; 148 SNPs/58 tags) or in combination with novel SNPs identified by sequencing functional regions (COMT; 58 SNPs/20 tags) or the entire gene and flanking sequence (DRD3; 69 SNPs/20 tags). We also genotyped 31 genomic control SNPs. Phase II analyses continued to implicate only SLC6A3 and DRD3 (SLC6A3, 6 SNPs associated; DRD3, 5 SNPs). The associations could not be attributed to population substructure and are detectable using different analytic designs. Post-hoc testing at COMT revealed a significant interaction between rs737865 and gender, replicating previous reports. These analyses represent a comprehensive and novel evaluation of positional dopaminergic candidates in schizophrenia genesis. Replicate studies are warranted. O3.7 DYSREGULATION OF RETINOID TRANSPORTERS EXPRESSION IN BODY FLUIDS OF SCHIZOPHRENIA PATIENTS Chunling Wan Shanghai Jiao Tong University, Shanghai This study aims to find the biomarkers or associated proteins of schizophrenia in human body fluids. We applied proteomic technology, combining two-dimensional electrophoresis with Coomassie blue staining and mass spectrometry identifying procedure for clinical screening of disease-influenced proteins in two sets of samples, plasma from 19 schizophrenia patients who were given an 8-week course of antipsychiatric medication and cerebrospinal fluid (CSF) from drugtreated schizophrenic patients (n ¼ 35) and age/sex-matched controls (n ¼ 36). We compared 56 plasma protein spots in schizophrenia patients before and after medical treatment. The expression of transthyretin (TTR) tetramer increased significantly in plasma of
Abstracts schizophrenic patients after two months valid in-hospital antipsychotic treatment (P ¼ 0.002). This result was confirmed by West-blotting of native PAGE with rabbit polyclonal anti-TTR IgG. To establish whether abnormalities in TTR tetramer in plasma could be confirmed, we further investigated CSF protein expression in 35 schizophrenia patients and 36 normal controls. We determined 80 protein spot expression levels, and the expression of TTR tetramer and apolipoprotein E (ApoE) were observed down-regulated by up to 1.68 and 3.62 times in schizophrenia patients’ CSF compared to that of normal controls. This result confirmed our observation of changed TTR tetramer expression in schizophrenia patient plasma. In humans, TTR tetramer and ApoE are both retinoid transporters. The present results suggest insufficient transport of retinol to the brain. Retinoid dysfunction might be involved in the pathology of schizophrenia. O3.8 HAPLOTYPES ENCODING SPEC2, PDZ-GEF2 AND ACSL6 GENES ARE ASSOCIATED WITH SCHIZOPHRENIA Xiangning Chen1, Xu Wang1, Shaon Hossain1, F. Anthony O’Neill2, Walsh Dermot3, Lora Pless4, Kodavali V. Chowdari4, Vishwajit L. Nimgaonkar4, Sibylle G. Schwab5, Dieter B. Wildenauer5, Patrick F. Sullivan6, Edwin Van den Oord1, Kenneth S. Kendler1 1 Virginia Commonwealth University, Richmond, Virginia 2 The Queens University, Belfast, Northern Ireland 3 The Health Research Board, Dublin 4 University of Pittsburgh, Pittsburgh, PA 5 University of Western Australia, Perth, WA 6 University of North Carolina, Chapel Hill, NC Chromosome 5q22-33 is a region where studies have repeatedly found evidence of linkage to schizophrenia. In this report, we took a stepwise approach to systematically map this region in the Irish Study of High Density Schizophrenia Families (ISHDSF) sample and found a 750 kb interval coding for the SPEC2/PDZ-GEF2/ACSL6 genes to be consistently associated with the disease. This finding was replicated in 4 more independent samples. Using sex and genotype conditioned analyses, we found, in all 5 samples, that multiple markers were associated with the disease and the associations derived largely from one of the two sexes. Three risk haplotypes were identified in the 5 samples, and each haplotype was found in at least two samples. Our data showed that the interactions of sex and risk haplotypes played a critical role in this region. From these data, we concluded that haplotypes underlying the SPEC2/PDZ-GEF2/ACSL6 region are associated with schizophrenia. However, due to the extended high LD in this region, we were unable to distinguish whether the association signals came from one or more genes. O4.1 GENOME-WIDE LINKAGE ANALYSIS OF HEROIN DEPENDENCE IN HAN CHINESE: RESULTS FROM WAVE ONE OF A MULTI-STAGE STUDY Ming Tsuang1, Jessica Su2, Shao Zhu3, Ruimin Zhang4, Bo Zhang4, Jixiang Li4, Xiaobo Yuan4, Jianhua Li4, Michael Lyons5, Stephen Faraone2 1 University of California, San Diego, La Jolla, CA 2 SUNY Upstate Medical University, Syracuse, NY 3 Harvard Medical School, Boston, MA 4 Yunnan Institute of Drug Abuse, Kunming 5 Boston University, Boston, MA The contribution of genes to the etiology of heroin dependence is greater than for any other illicit drug. The specific genes mediating this effect remain unknown, despite several candidate gene association studies of the condition. Here we report results of a genome-wide search for heroin dependence susceptibility loci using multipoint linkage analysis. In phase I, we ascertained 207 independent affected sibling pairs from 202 Han Chinese families from Yunnan Province, China (near Asia’s ‘‘Golden Triangle’’). After data-cleaning, 194 fully independent sibling pairs (i.e., with no overlapping individuals) from 192 families were genotyped on 404 short tandem-repeat markers spaced at an average inter-marker distance of 9 cM. Although none of our findings achieved genome-wide significance, we found two regions with nonparametric linkage (NPL) Z-scores greater than 2.0. An NPL Z-score of 2.19 (p ¼ 0.014) was observed at D4S1644, located at 143.3 cM on chromosomal region 4q31.21. The highest NPL Z-score of 2.36 (p ¼ 0.009) was observed at 53.4 cM on chromosomal region 17q11.2 at marker D17S1880. This is among the first published reports of a genome-wide linkage analysis of heroin dependence. Forthcoming
699
results from other groups and from two additional waves of ascertainment (one planned, one currently ongoing) for our own study should be able to support or refute the putative susceptibility loci we have identified, after which positional candidate genes can be evaluated as risk factors for the illness. O4.2 FINE-MAPPING AN 18 Mb ALCOHOL DEPENDENCE SUSCEPTIBILITY LOCUS ON 4q22-q32 USING TAG SNPS G. Kalsi1, P-H Kuo1, J. Alexander1, P.F. Sullivan2, EJCG. van den Oord1, D.G. Patterson3, D. Walsh4, K.S. Kendler1, C.A. Prescott5, B.P. Riley1 1 Virginia Commonwealth University, Richmond, USA 2 University of North Carolina, Chapel Hill, USA 3 Shaftsbury Square Hospital, Belfast, Northern Ireland, UK 4 Health Research Board, Dublin, Ireland 5 University of Southern California, Los Angeles, USA A genome wide linkage scan conducted in the Irish Affected Sib Pair Study on Alcohol Dependence (IASPSAD) produced significant evidence for linkage to number of DSM-IV alcohol dependence (AD) symptoms over a broad region of chromosome 4, from 4q22 to 4q32 (peak multipoint LOD ¼ 4.59, P ¼ 2.1 106, at D4S1611). A one-LOD interval under this peak delineates a 17.6 Mb region defined by the microsatellite markers D4S1572 and D4S427. We fine-mapped an 18.5 Mb region containing 65 genes and expressed sequences (ESTs) in the current build of the human genome (hg18, NCBI Build 36.1). We used Tagger and Phase 2 HapMap data (Jan ‘06 build) to select a minimum set of gene- and EST-based tagSNPs based on pairwise r2 to efficiently extract maximum information with the smallest genotyping burden. This approach identified 523 tag SNPs, of which 460 (88%) were suitable for high-throughput multiplex genotyping in the first stage of a two-stage False Discovery Rate (FDR) control design in a sample of 328 each cases and controls. 20 of the 61 genes contained markers that satisfied the phase one FDR significance cutoff value of p < 0.1. The most significant signal is in a gene involved with drug metabolism, PAPSS1, which produced allelic association of p ¼ 0.006 and 3-marker haplotypic association of p ¼ 0.007. Signifcant signals were also produced by genes influencing oligosaccharide catabolism (MANBA, P ¼ 0.031), transcription (PRDM5, P ¼ 0.034; PITX2, P ¼ 0.043), phosphorylation (ALPK1, P ¼ 0.032), receptor binding (ANK2, P ¼ O.041), protein degradation and liver damage. Stage 2 follow-up studies are underway currently.
O4.3 ALCOHOL DEPENDENCE IS ASSOCIATED WITH THE ZNF699 GENE, A HUMAN LOCUS RELATED TO DROSOPHILA HANGOVER, IN THE IRISH AFFECTED SIB PAIR STUDY OF ALCOHOL DEPENDENCE (IASPSAD) SAMPLE Brien Riley1, Po-Hsiu Kuo1, Gursharan Kalsi1, Vladimir Vladimirov1, Dawn Thiselton1, Jennifer Vittum1, Brandon Wormley1, Michael Grotewiel1, Diana Patterson2, Patrick Sullivan3, Edwin van den Oord1, Dermott Walsh2, Kenneth Kendler1, Carol Prescott2 1 Virginia Commonwealth University, Richmond, VA 2 Shaftesbury Square Hospital, Belfast 3 University of North Carolina, Chapel Hill, NC 2 Health Research Board, Dublin 2 University of Southern California, Los Angeles, CA Because tolerance is an important aspect of alcohol dependence (AD) in humans, recent evidence showing that the Drosophila gene hangover is critically involved in the development of alcohol tolerance in the fly suggests that variation in related human loci might be important in the etiology of alcohol-related disorders. The orthology of hangover in mammals is complex, but a number of human gene products (including ZNF699) with similar levels of amino acid identity (18–26%) and similarity (30–37%), are consistently identified as the best matches with the translated hangover sequence. We tested for association between the dichotomous clinical phenotype of alcohol dependence and seven single nucleotide polymorphisms (SNPs) in ZNF699 in our sample of 565 genetically independent cases and 496 siblings diagnosed with AD, and 609 controls. In analyses of genetically independent cases and controls, four of the seven single markers show strong evidence for association with AD (Fisher’s exact p 0.00003–0.001), and the most significant single marker, rs7254880, tags an associated haplotype with frequency 0.071 in cases compared to
700
Abstracts
0.034 in controls (chi-square 15.563, p < 0.00008, 5000 permutation p < 0.001, OR 2.17); inclusion of affected siblings gives similar results. Expression analyses conducted in independent postmortem brain samples show that expression of ZNF699 mRNA is significantly reduced in the dorsolateral prefrontal cortex of individuals carrying this haplotype compared with other observed haplotype combinations. O4.4 GABRA2: A CANDIDATE GENE FOR ALCOHOL DEPENDENCE? Ulrich Preuss Dept Psychiatry MLU Halle, Halle Several lines of evidence from animal and human studies indicate that the most important inhibitory neurotransmitter, GABA, plays an important role in alcohol dependence including alcohol withdrawal. The GABAA-receptor, an ion channel, is a pentamer which consists of 5 different subunits. One of these subunits, GABA A2 and its Gene, GABRA2, was identified from linkage analysis to be associated with the liability for alcohol dependence. Subsequent studies in several samples were able to confirm this association. The aim of this study is, to replicate this finding and investigate if genetic variants of the GABRA2 receptor increase the risk for alcoholism-associated phenotypes. 421 inpatient alcohol-dependent individuals were enrolled into the study. Characteristics of alcohol dependence and alcohol withdrawal were obtained by the SSAGA (Semi-Structured Assessment on Genetics in Alcoholism). 10 SNPs across GABRA2 were determined using PCR. One genetic variant was detected to significantly differ between alcohol-dependent subjects and controls. Two common haplotypes and their complementary alleles were identified containing this SNP and were present in 90.5% of controls and 93.5% of the alcoholdependent individuals. One of the haplotypes (TCACATTC) was significantly associated with alcohol dependence and characteristics of alcohol withdrawal and severity of alcohol dependence (delirium tremens, withdrawal seizures). These findings support and extend the two previous studies implicating the GABA-A receptor as contributing not only to the genetic risk for alcohol dependence but also increase the liability for severe alcohol withdrawal. O4.5 ADOLESCENT EXPOSURE TO ETHANOL IN A SOCIAL CONTEXT SELECTIVELY INCREASES ALCOHOL PREFERENCE IN VERVET MONKEYS: INDEPENDENT AND INTERACTIVE GENE-ENVIRONMENT EFFECTS Roberta Palmour1, Peter Szatmari2, Amy Beierschmitt3, Frank Ervin1 1 McGill University, Montreal QC 2 McMaster University, Hamilton ON 3 Behavioural Science Foundation, Estridge Estate Early use of beverage ethanol in human populations is well-established as a risk factor for later alcohol abuse and dependence, the extent to which environmental, genetic or interaction effects mediate this relationship is unknown. 96 adolescent vervet monkey offspring of alcohol-preferring and non-preferring parents were configured into groups of 8, balanced with respect to gender and initial ethanol preference. In half of the groups, ethanol as a 10% w/v solution was freely available 4 days per week for a period of 1 year; in the other groups, isocaloric sucrose was available on the same schedule. Ethanol preference was measured at 6, 12 and 24 months, and social behavioural measures were collected throughout the first year of the study. There was a main effect of ethanol exposure in the social context at 6, 12 and 24 months (p < 0.0001, p < 0.005 and p < 0.001, respectively), and main effects of gender at 6 and 12 (p < 0.005, p < 0.01), months. There was a main effect of baseline consumption (a measure of genetic vulnerability) at 6 months only (p < 0.005) and a strong interaction between baseline consumption and exposure at 24 months (p < 0.0001) and a weaker interaction between baseline consumption, exposure and gender at 24 months (p < 0.01). Social adolescent exposure to ethanol had a persistent effect in individuals with high genetic vulnerability and a transient effect in those with low or moderate vulnerability to preference drinking. O4.6 TESTING FOR MEASUREMENT INVARIANCE IN GENETIC ANALYSES OF SMOKING AND NICOTINE DEPENDENCE Hermine Maes, Michael C. Neale, Kenneth S. Kendler Virginia Commonwealth University, Richmond, Virginia
Understanding the contribution of genetic and environmental influences to the various stages of smoking initiation to more severe forms of smoking behaviour (regular smoking, smoking persistence, nicotine dependence) is an important first step in finding specific etiological factors for smoking. Previous behaviour genetic research has focused on using a conditional approach including smoking initiation, regular smoking and a sum score of scales established to measure nicotine dependence (FTQ (Fagerstro¨m & Schneider, 1989), FTND (Heatherton et al., 1991)) to estimate the heritability of nicotine dependence. However, this approach does not allow for possible measurement invariance of the scale by sex and age. To overcome this, we used an item response theory approach to examine the relationship between genetic and environmental risk factors for smoking initiation, regular smoking, and nicotine dependence in twins from the population-based Mid-Atlantic Twin Registry. Smoking behaviour was assessed in female, male and opposite sex 20 to 60 year old adult twin pairs by personal interview. Results indicated significant mean and variance differences by sex, age and their interaction in the latent ‘nicotine dependence’ factor. Allowing for differences in factor loadings in addition to factor mean differences improved the fit of the model. Further model improvement was obtained by allowing factor variance and threshold differences by age and sex. Heritability of the latent liability to nicotine dependence was estimated at about 80% without evidence for shared environmental influences, when allowing for measurement differences by sex and age. O4.7 EVIDENCE OF STRONG FAMILIALITY FOR PSYCHOSTIMULANT ADDICTION Thorgeir E. Thorgeirsson1, Frank Geller1, Andrew Hicks1, Thorlakur Jonsson1, Valgerdur Runarsdottir2, Hogni Oskarsson3, Anna Wiste4, Andres Ingason1, Hreinn Stefansson1, Jeffrey Gulcher1, Augustine Kong1, Thorarinn Tyrfingsson2, Kari Stefansson1 1 Decode genetics, Reykjavik 2 SAA-National Center of Addiction Medicine, Reykjavik 3 Therapeia, Reykjavik 4 Emory University, Atlanta Over 18,000 individuals have been treated for addiction at the National Center of Addiction Medicine in Iceland (SAA), over the last three decades. This population-based treatment cohort provides a unique opportunity for studying the familiality of addiction. An extensive genealogy database covering the entire present day population and stretching back several centuries contains information on distant genealogical relationships, allowing the study of addiction familiality for any genealogical relationships. The relative risk (RR) of substance dependence was determined for up to fifth degree relatives of probands diagnosed with alcohol, cannabis, sedative, amphetamine, opiate and cocaine dependence. For first degree relatives, we observed highly significant RR values for all substances ranging from 2.3 for alcohol to 20.9 for cocaine. The RR values were near 5.0 for cannabis and sedatives and 6.6 for opiates. Although RR values drop monotonically with increasing genealogical distance, significant values are observed well beyond the nuclear family, where the effect of shared environmental factors is expected to be minimal. Interestingly, we found very high RR for amphetamine dependence, 9.4 (1st degree) – 1.21 (5th degree). While we found evidence for a strong familial component for all major substances of abuse, the magnitude of risk in psychostimulant dependence in this treatment cohort is particularly striking. This finding emphasizes the role of genetics in the etiology of addiction, and highlights the importance of substance-specific effects. O4.8 AN INTEGRATED APPROACH TO IDENTIFY NEUROBIOLOGICAL MECHANISMS OF ADDICTIVE BEHAVIOUR Gunter Schumann1, Michael Smolka2, Rainer Spanagel2, Karl Mann2 1 Institute of Psychiatry, London, 2 Central Institute of Mental Health, Mannheim, Germany Addiction disorders are frequent oligogenic disorders with complex inheritance patterns and interplay of genetic and environmental factors contributing to their phenotype. They are heterogenous by nature with common and distinct genes contributing to different phenotypes of substance use disorders. To identify the genetic and neurobiological basis of addiction disorders we took these characteristics into account by identifying candidate genes shown to alter drug taking behaviour in animal models and analyse them for association
Abstracts
701
with analogous phenotypes of substance use disorders in humans. A recent example of this approach, the characterization of the role of a circadian rhythm gene Period 2 in human alcohol drinking behaviour will be presented (Nature Medicine, 2005; 11: 35–42). Neuroimaging permits reduction in phenotypic heterogeneity by measuring specific brain functions implicated in the etiology of addiction disorders and link them to genetic variations and behavioural characteristics relevant to disease processes. We will provide examples demonstrating the validity of this approach by analyzing the association of 5-HTT and COMT genotypes with fMRI-response to negative emotional cues (Nature Neuroscience, 2005;8:20–21, J. Neuroscience, 2005;25:836– 842). The integration of our candidate gene approach using behavioural animal models and neuroimaging studies will be described by providing an outlook of a recently funded EU-integrated gene-neuroimaging project ‘‘IMAGEN’’.
hsa-miR-189. Due to the overlap between GTS and OCD, it has been postulated that, in some cases, OCD and GTS may be alternative manifestations of the same underlying illness. We hypothesised that OCD patients may carry sequence variants in the 30 UTR of the family of SLITRK genes, especially in the regions predicted to be targets for miRNAs. We have performed re-sequencing of the six genes of the SLITRK family in 86 OCD patients. We have identified four previouslydescribed SNPs and eight sequence variants not found in available public databases. We performed case-control analysis between known and new SNPs in 95 control samples to test whether the presence of these variants confers susceptibility to development of OCD. Finally, conservation and miRNA target prediction programmes have been used to define the potential disruption in complementarities between known miRNAs expressed in the central nervous system and nucleotide changes affecting the 30 UTR of SLITRK genes.
O4.9 CHRONIC ETHANOL EXPOSURE IMPAIRS NEURONAL DIFFERENTIATION OF SH-SY5Y CELLS C.V. Wernicke1, J.D. Hellmann 1, U. Finckh2, H. Rommelspacher1 1 Clinical Neurobiology, Dept. of Psychiatry, Charite´—University Medicine Berlin, Germany 2 Laboratoriumsmedizin Dortmund, Dortmund, Germany
O5.2 TRANSCRIPTION FACTOR AP-2 AND MONOAMINERGIC FUNCTIONS IN THE CENTRAL NERVOUS SYSTEM Mattias Damberg, Kent Nilsson, Rickard Sjo¨berg, Lars Oreland, Jerzy Leppert Uppsala University, Va¨stera˚s
Recently, we showed that human neuroblastoma SH-SY5Y cells express dopamine transporter, and both splice variants (D2l, D2s) of dopamine D2 receptors (DRD2). Furthermore, D2s transcription was more enhanced, than D2l during neuronal differentiation of SH-SY5Y cells by retinoic acid (RA). In vivo D2s represents the high-affinity autoreceptor and D2l the low-affinity postsynaptic receptor. In lactotrops, overexpression of D2s but not of D2l leads to mitogenactivated protein kinase (MAPK) induction. The question was how chronic ethanol exposure influences the RA induced neuronal differentiation of SH-SY5Y cells, as well as the expression of D2l/s, and whether the MAPKs (Raf, MEK, ERK) and the Raf kinase inhibitor protein (RKIP) as regulators of cell survival and cell death are involved. Cells were incubated with 10M RA (controls) or RA and ethanol (100 mM) for up to 31 days. Transcription of the mentioned genes was investigated by real-time RT-PCR. RA treated cells showed differentiated state with clusters connected by very long abundant and branched neurites. Ethanol treated cells also displayed a network of neurites, but did not form such clusters. Ethanol diminished the increase of D2s, seen in RA only treated cells (p ¼ 0.004, df ¼ 5), leading to a significant decreased ratio of D2s/D2l compared to the controls (p ¼ 0.005, df ¼ 5). Withdrawal from ethanol for 3h and 24h did not completely restore the diminished D2s/D2l ratio. Also RKIP transcription, which correlated with the state of neuronal differentiation, was systematically reduced in ethanol treated cells, whereas withdrawal for 24h led to an entire compensation in RKIP transcription. Although total ERK transcription levels as well as basal activity were not affected by chronic ethanol exposure, the ERK responsiveness to BDNF remained significantly reduced in chronically ethanol treated cells. Our results show, that SH-SY5Y cells are well suited to study the influence of ethanol on the development/adaptation of dopaminergic features. Whether DRD2 gene polymorphisms influence the gene expression is under investigation.
In the central nervous system, transcription factor AP-2 family a critical regulatory factor in neural gene expression and neuronal development. Several genes in the monoaminergic systems display AP2 binding sites in regulatory regions. In addition, brainstem levels of transcription factor AP-2alpha and AP-2beta are positively correlated to monoamine measures in rat forebrain. Moreover, AP-beta genotype correlates with CSF-HVA and MHPG in females, suggesting a regulatory role of AP-2 in the adult brain as well. Great changes in psychiatric phenotypes due to genetic factors are rarely the result of single gene polymorphisms but are more often due to several genetic mechanisms. Identification of combinations of candidate genes that are all linked to one disease or psychiatric phenotype has recently been discussed. The expression of these candidate genes might be regulated by the same transcription factors, e.g. AP-2. Transcription factor AP2beta genotype has been studied in relation to personality, platelet monoamine oxidase (MAO) activity, binge-eating disorder, CSF-levels of monoamine metabolites, premenstrual dysphoric disorder and schizophrenia. We have studied interactions between AP-2beta genotype and psychosocial variables to predict vulnerability to depression. In females, AP-2beta genotype interacted with both types of housing (p ¼ 0.0018) and separated parents (p ¼ 0.0003) to predict depressive symptoms. On the whole, we suggest that transcription factor AP-2 family is involved in regulating the monoaminergic systems, both pre- and postnatally, and, therefore, might be involved in the pathophysiology of neuropsychiatric disorders.
O5.1 30 UTR-SLITRKs VARIANTS IN miRNAs TARGETS IN PATIENTS WITH OBSESSIVE-COMPULSIVE DISORDER Ester Saus1, Mo`nica Grataco`s2, Maria del Pino Alonso3, Juan R. Gonza´lez2, Jose´ Manuel Mencho´n3, Cinto Segala`s3, Mo`nica Baye´s2, Javier Labad3, Xavier Estivill4 1 Center for Genomic Regulation, Barcelona 2 Center for Genomic Regulation; National, Barcelona 3 OCD Clinic, University Hospital of Bellvitge, Barcelona 4 National Genotyping Center, Barcelona Obsessive-compulsive disorder is characterised by recurrent unwanted thoughts and/or repetitive behaviour, with a prevalence estimate of 2.5% in the Spanish population. Abelson et al. (2005) have recently studied SLITRK1 as a candidate gene for Gilles de la Tourette Syndrome (GTS) on chromosome 13q31.1, and identified two mutations in SLITRK1 in three out of 174 GTS probands. One of these mutations was a base change in the 30 UTR of the gene, and corresponds to a highlyconserved nucleotide within the predicted binding site for a miRNA,
O5.3 AN ANALYSIS OF POLYMORPHISMS IN THE REMAINING PRODH ALLELE, SERUM PROLINE LEVELS AND BRAIN FUNCTION MEASURES IN INDIVIDUALS WITH THE 22q11.2 DELETION SYNDROME Jacob Vorstman1, Alecia Willis2, Monique De Sain-van der Velden3, Bert Dorland3, Frits Beemer4, Rene´ Kahn5, David Valle2, Beverly Emanuel3, Herman van Engeland4, Chantal Kemner4 1 University Medical Centre Utrecht, Utrecht 2 Howard Hughes Medical Institute and Inst, Baltimore, Maryland 3 Metabolic Diseases, University Medical Centre, Utrecht 4 Medical Genetics, University Medical Centre, Utrecht 5 Psychiatry, U.M.C. Utrecht and Rudolf Ma, Utrecht 3 Human Genetics, Children’s Hospital, Philadelphia, PA 4 Child Psychiatry, U.M.C. Utrecht and Rud, Utrecht The PRODH gene at 22q11.2 encodes the catabolic enzyme Proline Dehydrogenase. In a previous study with 22q11DS patients, increased levels of plasma Proline were reported. Interestingly, increased Proline levels may exert a modulatory influence on glutamatergic neurotransmission. Therefore, PRODH, through its key role in Proline metabolism, may be an interesting candidate gene for the etiology of psychotic disorders in the context of 22q11DS. We have previously confirmed significantly-increased Proline levels in a sample of children with the 22q11.2 deletion syndrome. Our current goal is a) to study potential modulatory effects of Proline levels on brain function parameters in our study sample of children with a confirmed 22q11.2 deletion (n60) and
702
Abstracts
b) to examine associations between polymorphisms in the remaining PRODH allele with serum Proline levels. Our results thus far indicate a significant effect of serum Proline level on Smooth Pursuit performance. Furthermore, sequencing of PRODH has been completed for exons 3–15, including the 3’ UTR. We have found 8 synonymous, 6 missense and 23 intronic and UTR variants. Of the 6 missense variants, two are novel variants for our lab; the other four had been previously identified and tested for effect on Proline Dehydrogenaseoxidase activity. We plan to test the effect of the novel variants on Proline Dehydrogenaseoxidase activity. We will discuss our results and their potential implications for our understanding of the etiology of psychopathology in the context of the 22q11.2 deletion syndrome, as well as implications for future research. O5.4 GENETIC VARIATION IN PRODH AND ZDHHC8 ON 22q11 IMPACTS ON BRAIN STRUCTURE AND CORTICAL FUNCTION IN HEALTHY HUMANS Lucas Kempf1, Richard Straub1, Bhaskar Kolachana1, Beth Verchinski1, Joseph Callicott1, Michael Egan2, Venkata Mattay1, Daniel Weinberger1, Andreas Meyer-Lindenberg1 1 NIMH, Bethesda 2 Merck and Co., Philadelphia, PA Introduction: Several genes on 22q11 have been implicated in risk for schizophrenia, including COMT, PRODH, expressed in the striatum and involved in glutamate synthesis, and ZDHHC8, which is thought to be involved in palmitoylation. In murine knock-out models these genes have shown PPI deficits, a reflex regulated through a thalamic-striatalcortical loop. We used a multimodal imaging strategy to investigate the impact of this genetic variation on brain structure and function in a large sample of healthy genotyped humans. Methods: We used voxel-based morphometry and two FMRI tasks known to be sensitive to genetic variation in 122 healthy controls to study the effects of the following SNPs: COMT, rs4680, PRODH, rs372055 and ZDHHC8, rs175174. Activation and thalamic functional connectivity was analysed for both tasks. Results: PRODH had a significant effect on the activation of the dorsal striatum. ZDHHC8 had significant impact on the structure of the thalamus and functional activity of the posterior cingulate, known to have functional and structural connectivity with the thalamus. A significant interaction between genetic variants in COMT and PRODH was seen in caudate activation. Conclusions: These results suggest genes in 22q11 impact on subcortical-cortical circuits implicated in schizophrenia, providing a possible neural substrate for epistatic interactions contributing to schizophrenia risk attributed to this region. O5.5 SUPPORT FOR THE ADULT NEUROGENESIS THEORY OF DEPRESSION: ANXIETY AND DEPRESSION DISORDERS ASSOCIATED WITH PLXNA2 Naomi Wray1, Michael James1, Steven P. Mah2, Matthew Nelson2, Gavin Andrews3, Patrick F. Sullivan4, Grant Montgomery1, Andrew J. Birley1, Andreas Braun2, Nicholas G. Martin1 1 Queensland Institute of Medical Research, Brisbane 2 Sequenom Inc, CA 3 University of New South Wales, NSW 4 University of North Carolina, NC Reduction in adult neurogenesis has been proposed as a mechanism for onset of depression. Sempahorins and their co-receptors, plexins, have been implicated in the development and maintenance of the nervous system and in adult neurogeneis. A recent genome-wide association study for schizophrenia identified a variant of the gene encoding plexin A2 (PLXNA2) to be most consistently associated across discovery and replication samples. Common genetic liabilities exist between schizophrenia, depression, anxiety and measures of personality, but few examples exist of genetic variants common to two or more of these phenotypes. With these factors in mind, a genetic association study was conducted between six SNPs from the PLXNA2 gene and five phenotypic measures: anxiety or depression, anxiety, depression, neuroticism and psychological distress. Our study sample consisted of 1,834 people selected as concordant or discordant sibling pairs with respect to extreme neuroticism scores from a population cohort of 18,742 Australian twin individuals and their siblings. All participants completed detailed Composite International Diagnostic Interview (CIDI) and personality
questionnaires. Association analysis found evidence for association between SNP rs2478813 (and other SNPs correlated with it) and anxiety, depression, neuroticism and psychological distress; the association with anxiety (empirical p ¼ 0.00010) can be declared significant after stringent Bonferroni correction for multiple testing. Our results suggest that PLXNA2 is a likely candidate for causal variation in a broad spectrum of psychiatric phenotypes and provide support for the theory of adult neurogenesis being implicated in these disorders. O5.6 UPDATE ON THE SEARCH FOR SCHIZOPHRENIA LIABILITY VARIATION IN THE DTNBP1 GENE IN THE IRISH STUDY OF HIGH DENSITY SCHIZOPHRENIA FAMILIES (ISHDSF) Brien Riley1, Dawn Thiselton1, Vladimir Vladimirov1, Robert Ribble1, Brandon Wormley1, Greg Frank1, Babak Tabatabai1, Benjamin Neale2, Bradley Todd Webb1, Edwin van den Oord1, Frances Anthony O’Neill2, Dermott Walsh3, Kenneth Kendler1 1 Virginia Commonwealth University, Richmond, VA 2 Institute of Psychiatry, Richmond, VA 2 Queens University, Belfast 3 Health Research Board, Dublin The dystrobrevin binding protein 1 (DTNBP1) gene is widely associated with schizophrenia. We identified a DTNBP1 high risk haplotype (HRH) in the ISHDSF. We sequenced 156 Kb of the genomic locus on genetically independent HRH backgrounds, and identified 281 polymorphisms (1.8 SNPs/Kb); 76 were novel. 253 (90.7%) could be mapped to specific haplotypes, 26 were specific to the HRH, and 21 (80.7%) were novel. The sequence dataset provides several insights to the DTNBP1 association data and the capacity to test specific hypotheses. Published associations cluster on 3 specific haplotypes, and variation specific to these may account for all the associations simultaneously. We identified 22 markers (2 novel) where the common allele is shared by all three associated haplotypes, even though one is evolutionarily distant from the others. Assessment of these markers in two samples showing association with two different haplotypes is underway currently. Screening for evolutionarily conserved regions (ECRs) identifies a number of conserved intronic and upstream regions, in addition to the exons. 26 SNPs map within ECRs and another 14 map within 100 bp of them. 7 of these are specific to the Irish HRH and 4 have alleles shared by the set of associated haplotypes. Finally, the sequence data suggest one of two possibilities intriguing possibilities. Either the rates of SNP evolution are markedly different on specific haplotypes, or the published clade is inverted and the most common haplotype is actually the most recent and expanded possibly by selection to its high frequency. O5.7 BLOC-1 COMPONENTS DYSBINDIN (DTNBP1) AND MUTED CAN MODULATE DOPAMINE D2 RECEPTOR ENDOCYTOSIS Richard Straub, Yoshitasu Sei, Zhen Li, Daniel Weinberger, Yukihiko Iizuka NIMH, NIH, Bethesda, MD Genetic variation in DTNBP1 influences cognition and risk for schizophrenia. Dysbindin is a component of BLOC-1 (biogenesis of lysosomerelated organelles complex 1) which includes MUTED (a dysbindin binding partner) and BLOC1S2, and both genes we have discovered are also susceptibility genes. Dysbindin expression is reduced in schizophrenics, but little is known about BLOC-1 function in neurons. Increased dopamine D2 receptor (DRD2) density has been observed in some brain regions in schizophrenics and the therapeutic effect of antipsychotics may require antagonism of DRD2. To examine whether BLOC-1 affects DRD2 regulation, using flow cytometric analysis of human SH-SY5Y neuroblastoma cells, we studied the effects of transfection of DTNBP1 and MUTED siRNAs on surface DRD2 levels and on dopamine induced internalization. We also studied B lymphoblasts from schizophrenics and controls. DTNBP1 siRNA reduced dysbindin protein, increased surface expression of DRD2 and blocked dopamine-induced internalization. MUTED siRNA produced similar effects that were additive to those of DTNBP1 siRNA. DRD2 was also upregulated by siRNA in B lymphoblasts. DTNBP1 risk haplotypes contribute to the reduction in dysbindin observed in schizophrenia, and reduced dysbindin may compromise BLOC-1 function, increasing surface DRD2. More broadly, receptor dysregulation may be an important mechanism by which genetic variation in DTNBP1 influences risk for schizophrenia. Results of these and additional studies of other dopamine
Abstracts receptors as well as the visualization of receptor internalization and trafficking by confocal microscopy and the effect of risk haplotypes on BLOC-1 function will be presented. O6.1 GENETIC PREDICTORS OF ADHD CONTINUITY — A LONGITUDINAL STUDY Kate Langley1, Tom A. Fowler2, Michael J. Owen2, Michael C. O’Donovan2, Anita Thapar2 1 University Hospital of Wales, Cardiff 2 Cardiff University, Cardiff Attention Deficit Hyperactivity Disorder (ADHD) has been demonstrated to be genetically influenced. A number of susceptibility gene variants are associated with the disorder (notably a DRD4 VNTR, DAT1 VNTR, DRD5 microsatellite marker) with findings replicated in meta-analyses. ADHD is increasingly recognised to continue into adolescence and adulthood, although it is unclear whether risk factors which influence the origins of the disorder also play a role in its continuity. In one of the first reported studies of its type, we follow up children with ADHD into adolescence, 5 years after initial participation in a genetic study, to test to what extent susceptibility gene variants associated with ADHD also influence its continuity. All children underwent detailed assessments including research diagnostic interviews, meeting DSM III-R or IV criteria for ADHD at baseline. Time 1 assessment -mean no. DSM-IV ADHD symptoms ¼ 14.65, sd 2.37. At Time 2 assessment five years later, 79% continued to meet diagnostic criteria—mean no. DSM-IV ADHD symptoms ¼ 12.17, sd 4.79. The DRD4 variant significantly predicted both total number of ADHD symptoms (t ¼ 2.20, p ¼ 0.03) and diagnosis of ADHD (chi square ¼ 4.72, p ¼ 0.03) at time 2. The DAT1 480bp VNTR and DRD5 CA(n) microsatellite repeat did not influence continuity. These findings suggest the DRD4 7 repeat allele is an important predictor of ADHD continuity. The importance of examining genetic risk factors influencing the outcomes as well as origins of psychiatric disorder is discussed. O6.2 AUTISM CANDIDATE GENES IDENTIFIED WITH A GENE INTERACTION NETWORK SOFTWARE-TOOL Wouter Staal1, Lude Franke2, Jacob Vorstman2, Ron Hochstenbach2, Emma van Daalen3, Cisca Wijmenga2, Herman van Engeland3 1 UMC Utrecht, Child & Adolescent Psychiatry, Utrecht 2 Dep Human Genetics, Utrecht 3 Child & Adolescent Psychiatry, Utrecht The etiology of autism is largely determined by genetic factors. With linkage analysis, several loci have been related to autism, although replication of these studies has proven difficult. We have recently published an overview of cytogenetic abnormalities in autism in an attempt to identify alternative regions of interest. Identification of candidate genes in linkage loci and the cytogenetic regions of interest (CROI) would be a logical next step. However, the total number of genes in the linkage loci and the CROI’s is large. A novel approach to the computational prediction of disease genes has recently been developed in our laboratory, allowing the measurement of functionally closely related genes that are located on separate disease loci. This method -Prioritizerworks on the assumption that the number of pathways disturbed in a disease is extremely limited and that genes that are involved in a pathway are closely connected. Consequently, genes that are located in disease loci and work in the same biological pathway are more likely to be candidate genes for a disease than genes that are not related. Prioritizer was used to identify candidate genes in the relevant linkage loci and CROI’s. Previously suggested candidate genes for autism were found, including: MBP, MBD1, MBD2, BCL2, COMT and UBE3A. Several lowcopy repeat genes of the proximal portion of chromosome 15 were also identified. Furthermore, alternative potentially interesting candidates were listed, such as the mitogen-activated protein kinases (MAPK), which are involved in neuronal development and apoptosis. O6.3 GENOME-WIDE SCAN FOR AUTISM SPECTRUM DISORDERS IN AN EXTENDED PEDIGREE FROM FINLAND Leena Helena Kilpinen1, Tero Ylisaukko-oja2, Joni A. Turunen2, Reija Alen3, Raija Vanhala4, Lennart von Wendt4, Teppo Varilo2, Leena Peltonen2 1 National Public Health Institute, Helsinki 2 National Public Health Institute, Helsinki 3 Department of Child Neurology, Central H, Jyva¨skyla¨ 4 Unit of Child Neurology, Hospital for Children, Helsinki
703
As part of the genetic study of autism spectrum disorders (ASDs) in the Finnish population, an extensive genealogical search was conducted to identify common ancestors for a set of ASD families. Based on this search back to the 17th century, eighteen ASD families (33 affected) were merged into one extended pedigree. Hypothesizing that this pedigree could expose ASD gene(s) enriched in this internal subisolate, we performed a genome-wide scan using 1107 STS-markers (intermarker spacing 3.4 cM). A joint analysis of linkage and linkage disequilibrium was performed with Pseudomarker statistics and a traditional multipoint linkage analysis with Simwalk2. Nine loci exceeded the chosen significance level of –log(p) > 2.5: 1q21-23, 15q12, 13q12, 14q11, 8q23, 6p24, 5q32, 5q35 and 6p22. Of these, 1q21-23 (p ¼ 0.00082) is among the best loci, also in previous genomewide scans for autism and Asperger syndrome in Finland. 15q11-13 (p ¼ 0.00081) is a well-recognised site for cytogenetic abnormalities associated with autism. Best multipoint linkage was observed at 19p13 (–log(p) ¼ 3.57). Regional candidate genes were analysed with dense sets of SNPs, using an extended sample with an additional 126 families with ASDs (281 affected). Candidate genes on 1q21-23 include RGS4, NOS1AP, ATP1A4, ATP1A2, KCNJ9 and KCNJ10, and on 15q11-13 the GABAA receptor gene cluster. Best evidence of association was observed with intragenic SNPs of ATP1A2 (p ¼ 0.0004) and GABRB3 (p ¼ 0.001), but additional genotyping of the other loci is currently ongoing.
O6.4 HOW GENERAL ARE GENES FOR GENERAL COGNITIVE ABILITY? SNP-SET ANALYSES Yulia Kovas1, Lee M. Butcher2, Nicole Harlaar2, Robert Plomin2 1 SGDP, Institute of Psychiatry, London 2 SGDP Centre, Institute of Psychiatry, London Introduction: A set of SNPs associated with general cognitive ability (‘g’) in 7-year-olds has been reported at this meeting from a genomewide association scan using 500,000 SNPs (Butcher et al.). We used this ‘g’ SNP set to test the ‘generalist genes’ hypothesis derived from multivariate quantitative genetic analyses which predicts that genes associated with ‘g’ will also be associated with mathematics and reading. Methods: In a sample of 3000 pairs of 7-year-old twins, we used multivariate quantitative genetic analysis to assess genetic correlations between ‘g’, mathematics and reading. We used the ‘g’ SNP set in multiple regression analyses to identify which SNPs in the SNP set are also associated with mathematics and reading. RESULTS: Genetic correlations were substantial between ‘g’ and mathematics (.67), between ‘g’ and reading (.58) and between mathematics and reading (.74). Nonetheless, some significant specificity of genetic effects on these three traits also emerged. Results for the ‘g’ SNP set supported the generalist genes hypothesis in that most of the SNPs in the ‘g’ SNP set were also associated with mathematics and reading. Similar results were found for the entire distribution (ability) and for the low-performing extreme (disability). Conclusions: Multivariate quantitative genetic analyses and molecular genetic analyses using a ‘g’ SNP set support the generalist genes hypothesis, which is likely to have far-reaching implications for diagnosis and treatment as well as prediction and prevention of learning disabilities.
O6.5 AN ASSOCIATION SCAN USING POOLED DNA AND 100K SNPS IDENTIFIES A ‘SNP SET’ FOR EARLY READING Emma Meaburn, Lee M. Butcher, Nicole Harlaar, Charles Curtis, Ian W. Craig, Leonard C. Schalkwyk, Robert Plomin Institute of Psychiatry, London Introduction: About 60% of the variation in both reading disability and ability in the early school years is attributable to genetic factors. Method: We report a multistage association scan of more than 100,000 SNPs, using pooled DNA to identify some of the quantitative trait loci (QTLs) that explain genetic variation in early reading in a large representative UK sample of 7-year-olds. In an initial screening stage, pooled DNA for low (n ¼ 755) versus high (n ¼ 747) reading groups was compared. In order to assess sampling variation, each group was divided into 10 independent pools, and each pool assayed on a 100K
704
Abstracts
SNP microarray set. Significant associations were confirmed by individually genotyping an independent unselected sample of 4,600 children, enabling us to identify SNPs significant in the low versus high screening stage that also show a linear association in the expected direction across the distribution of reading abilities, as predicted by QTL theory. Results: 20 SNPs were significant at p < .01, with the additive effects of these SNPs accounting for 2.5% of the variance in reading scores when combined in a ‘SNP set.’ No QTL effect sizes greater than 5% were detected, even though we had 99% power to detect QTL associations of .5% effect size. Implications: This genome-wide association scan suggests that QTL effect sizes for even highly heritable common disorders and quantitative traits such as early reading disability and ability might be much smaller than previously believed.
O6.6 COMT, BDNF AND CHILDHOOD BEHAVIOR: A PHENOME-WIDE APPROACH Oliver Palfrey-Davis, Robert Plomin Institute of Psychiatry, London The Generalist Genes hypothesis suggests that polymorphisms in candidate genes for cognitive and behavioural disorders are likely to be related to a range of phenotypes, rather than just one specific outcome measure. Therefore, instead of cherry-picking associations from a dataset with multiple measures, we used a systematic approach to examine associations between the childhood behavioural phenome as assessed in the longitudinal Twins Early Development Study (TEDS) and two of the most widely-reported functional polymorphisms: Catechol-O-Methyltransferase (COMT) val158met polymorphism (rs4680) and Brain-Derived Neurotrophic Factor (BDNF) val66met polymorphism (rs6265). These two polymorphisms were genotyped on a subsample of TEDS (about 6700 twins) who had been assessed on a broad range of measures of language, reading, mathematics, cognitive ability, academic achievement and behaviour problems across six ages from 2 to 10 years. Our systematic ‘phenome-wide’ screen has up to 99% power to detect an association that accounts for 0.3% of the variance (p < .01, two-tailed). Despite this statistical power, we find no systematic pattern of significant associations across these diverse measures. We conclude that BDNF and COMT are not linearly associated with the behavioural phenome in childhood.
O6.7 A CLUSTERING APPROACH FOR MAPPING DISEASE PATHWAYS IN ADHD Rosalind Neuman1, Lingwei Sun2, Richard Sherva2, Richard Todd2 1 Washington University Medical School, St. Louis, MO 2 Washington University, St. Louis, MO Multiple genetic and environmental mechanisms contribute to the clinically heterogeneous ADHD phenotype. To increase our knowledge of underlying risk factors for this disorder, ADHD typologies must be reduced to etiologically homogeneous forms that will inform genetic studies and treatment interventions. The approach taken here was to devise a mixture model to differentiate disease etiology in phenotypically similar subjects. Our sample consisted of twins, 7–19 years, ascertained from Missouri birth records. Dopaminergic and nicotinic polymorphisms and non-genetic risk factors for ADHD were used as input to a latent class model to predict membership within latent groups. Standard methods were used to identify, posthoc, group differences relating to prenatal substance use, psychopathological disorders, family psychosocial stressors, and pregnancy, birth, and infancy complications. A set of typologies was defined, distinguished by the proportion of genotypic and covariate risk factors prevalent within each cluster. Eg., 57% of 1 small cluster consisted of population-based combined ADHD twins all homozygous for the DAT440 allele, 85% with 1 copy of the DRD4-7 allele, and 83% with prenatal smoke exposure. Posthoc analyses indicated that, in this cluster, an increased number of problems occurred at delivery of the twins along with a modest rise in maternal blood pressure during the pregnancy. We demonstrate a flexible latent class approach using genetic and non-genetic covariates for refinement of the ADHD subtype. Our results indicate that this approach may be a valuable tool for assessing the genetic and nongenetic risk factors on a trait’s expression.
O6.8 FURTHER EVIDENCE THAT KIAA0319 IS ASSOCIATED WITH DEVELOPMENTAL DYSLEXIA Denise Harold1, Silvia Paracchini2, Thomas Scerri2, Mc¸gan Dennis2, Natalie Cope1, Gary Hill1, Valentina Moskvina1, Janet Walter2, Alex Richardson2, Michael Owen1, John Stein2, Eric Green3, Michael O’Donovan1, Julie Williams1, Anthony Monaco2 1 Cardiff University, Cardiff 2 University of Oxford, Oxford 3 National Human Genome Research Institute, Bethesda, Maryland The DYX2 locus on chromosome 6p22 is the most replicated region of linkage to developmental dyslexia (DD). Two candidate genes within this region have been implicated in the disorder: KIAA0319 and DCDC2. Five polymorphisms within DCDC2 have shown association with DD in a US and a German sample. We have genotyped these variants in two independent UK samples from Cardiff and Oxford. Nominally significant associations were detected between three DCDC2 SNPs and reading traits in the Oxford sample. However, no association with DD was observed for any DCDC2 variant in the Cardiff sample. Having previously found evidence that variation in KIAA0319 confers susceptibility to DD, we also sought to refine this genetic association by genotyping an additional 36 SNPs in the gene. Ten SNPs, predominantly clustered around the first exon, showed the most significant association with DD in one or both UK samples, including rs3212236 (P ¼ 0.00003) and rs761100 (P ¼ 0.0004). The location of this association may indicate the presence of susceptibility variants influencing gene regulation. The KIAA0319 and DCDC2 proteins share functional similarities: both appear to play a role in neuronal migration. We therefore tested for epistatic interactions between variants in the two genes. The most significant interaction observed was between rs793862 in DCDC2 and rs761100 in KIAA0319 (P ¼ 0.007). Thus our data provide further support for KIAA0319 as a DD susceptibility gene in the UK population. The evidence implicating DCDC2 is inconsistent, but it may contribute to DD susceptibility through epistatic interactions with KIAA0319. O7.1 STRUCTURAL AND FUNCTIONAL IMPLICATIONS OF DISC1 GENOTYPE IN SCHIZOPHRENIA Anil Malhotra1, Philip Szeszko1, Colin Hodgkinson2, Todd Lencz1, Pamela DeRosse1, Katherine Burdick1, John Kane1, David Goldman2 1 Zucker Hillside Hospital, Glen Oaks 2 NIAAA, Bethesda, MD Disrupted in Schizophrenia I (DISC1) has been hypothesized to be a susceptibility gene for schizophrenia and schizoaffective disorder, but little is known regarding the mechanisms through which it may confer increased risk. Given that DISC1 plays a role in cerebral cortex development, polymorphisms in this gene may have relevance for neurobiological models of schizophrenia that have implicated cortical deficits in its pathophysiology. Previous studies suggest that the minor allele (phe) of a missense variant in exon 9 of DISC1 (leu607phe) and that variants of a 3-locus haplotype containing the phe allele increase risk for disease. We investigated whether the leu607phe polymorphism was associated with prefrontal grey matter volumes and positive and negative symptoms in patients with schizophrenia or schizoaffective disorder. Among patients and healthy volunteers, phe carriers had significantly less grey matter in the superior frontal gyrus and anterior cingulate gyrus compared to leu/leu homozygotes. Leu607phe genotype accounted for 11.2% of the variance in superior frontal gyrus grey matter volume and 12.6% of the variance in anterior cingulate gyrus grey matter volume. In addition, patients who were phe carriers had significantly greater severity of positive symptoms (hallucinations) compared to patients who were leu/leu homozygotes. Lastly, less left superior frontal gyrus grey matter volume correlated significantly with increased severity of hallucinations among patients. These findings implicate genetic involvement of DISC1 in variation of prefrontal cortical volume and positive symptoms, thus providing a potential mechanism through which DISC1 may confer increased risk for schizophrenia-related disorders. O7.2 CHACTERISATION OF DYSBINDIN POSITIVE SCHIZOPHRENIA USING CLINICAL, COGNITIVE, AND NEUROPHYSIOLOGICAL MEASURES Gary Donohoe1, Derek W. Morris1, Ian H. Robertson1, Sarah Clarke1, Jeanne-Marie Nangle1, Siobhain Schwaiger1, Elena Magno1, Pierfilipo DeSanctis2, Jennifer Montesi2, Hugh Garavan1, Daniel C. Javitt2, Michael Gill1, John Foxe2, Aiden Corvin1
Abstracts 1
Trinity College Dublin, Dublin Nathan Kline Institute, New York
2
Background: Susceptibility genes for schizophrenia have been hypothesized to mediate liability for the disorder at least partly by influencing neurocognitive function. Dysbindin has recently been associated with poorer general cognitive ability and higher negative symptom severity. We investigated the association between a dysbindin risk haplotype previously identified in our schizophrenia sample and measures of cognitive, neurophysiological, and clinical symptoms. Methods: Dysbindin risk haplotype carriers were compared with nonrisk haplotype carriers on a) measures of cognition (working memory, episodic memory and attention) taken from the Weschler and CANTAB batteries b) an EEG high density mapping study of the P1 response during early visual processing, and c) a principal components factor analysis of PANSS symptom scores. Results: We found that risk carrier status explained a significant proportion of variance in performance on a spatial working memory task. We also found that risk haplotype carriers showed significantly attenuated P1 responses in visual cortex relative to non-risk carriers. We finally observed that risk haplotype carriers were less likely to show elevated scores on affective components of the PANSS. Conclusions: In our sample of patients with chronic schizophrenia, Dysbindin risk was associated with deficits at both earlier and later stages of information processing and reduced affective symptoms. Building on recent cognitive and clinical evidence, these data support the hypothesis of dysbindin’s association with a more trait-like and disabling presentation of schizophrenia, which may help further elucidate its biological significance for the disorder. O7.3 BIPOLAR DISORDER WITH PERSECUTORY DELUSIONS: A MORPHOMETRIC SUBGROUP? Heike Tost, Gabriele Ende, Matthias Ruf, Christine Schmael, Caroline Knorr, Monika Deschner, Andrea Schmitt, Christian Vollmert, Fritz A. Henn, Marcella Rietschel Central Institute of Mental Health, Mannheim Morphometric brain research in bipolar disorder (BD) has yielded conflicting results. Recent findings from our group support the idea that BD patients with persecutory delusions (PD) constitute a distinct disorder subgroup overlapping with schizophrenia (SCZ). Based on this finding, we hypothesised that (A) the analysis of BD patients with PD increases the power of brain morphometry in BD, (B) anomalies displayed by this subgroup resemble the morphometric pattern previously described for SCZ. 42 BD type I patients and 42 controls were studied using structural magnetic resonance imaging (MRI) on a 1.5 Tesla Siemens Vision scanner. Analysis of gray matter (GM) volumes was performed with SPM2 using an optimised voxel-based morphometric procedure. While the statistical comparison between the entire BD sample and healthy controls was bland except for a tiny circumscribed area of GM decrease in the left middle temporal gyrus, selective analysis of BD patients with PD revealed a pronounced and widespread fronto-temporal GM decrease in the superior and middle temporal gyrus, dorsolateral and ventrolateral prefrontal cortex and the anterior cingulate gyrus (p < 0.05 corrected). Comparison of psychotic BD patients with PD with patients without PD confirmed a prefrontal GM decrease for the PD subgroup (x ¼ 40, y ¼ 52, z ¼ 14, t ¼ 6.3, p ¼ 0.04 corrected). BD patients with PD constitute a biologically-meaningful subgroup with brain morphometric features similar to SCZ patients. This result further questions the validity of current Kraepelinian concepts of BD and SZ as being distinct and exclusive entities. O7.4 ALLELES OF REELIN GENE SHOW ASSOCIATION WITH WORKING MEMORY IN FINNISH SCHIZOPHRENIA FAMILIES Juho Olavi Peltonen1, Annamari Tuulio-Henriksson2, Anu Loukola3, Jesper Ekelund3, Tiina Paunio3, William Hennah3, Teppo Varilo3, Jaana Suvisaari2, Timo Partonen2, Jouko Lo¨nnqvist2, Leena Peltonen3 1 National Public Health Institute, Helsinki 2 Department of Mental Health and Alcohol, Helsinki 3 Department of Molecular Medicine, Nation, Helsinki Using the large, well-characterized nationwide collection of schizophrenia families we have obtained evidence for multiple genetic loci,
705
one of them being 7q22. The initial linkage evidence emerged from a sibpair analysis of 52 (n ¼ 236) families. We later replicated this finding in an independent sample of 352 nuclear families (n ¼ 1626). RELN gene on 7q22 encodes the reelin protein critical for neuronal development. It regulates neuronal migration by guiding neurons and glial cells to their positions during brain development, and is expressed in brain areas with synapse remodelling. No reports exist for genetic association of RELN with the clinical diagnosis of schizophrenia. However, specific alleles of RELN have been suggested to contribute to the response to antipsychotics. We hypothesized that since RELN takes part in essential brain functions it might associate with cognitive traits affected in schizophrenia. We collected detailed neuropsychological data from 186 schizophrenia families (n ¼ 618) and tested association with five microsatellites in RELN using sex, age, and affection status as covariates. This revealed associations (p ¼ 0.002 to 0.000001) with traits related to working memory, controlled by the frontal brain areas and known to be impaired in schizophrenia. Importantly, carriers of the most associated allele scored lower in the visual working memory (p ¼ 0.0004) and verbal working memory tests (p ¼ 0.000005), suggesting a role of RELN in frontal brain functions and raising the possibility that some allelic variants of this gene might contribute to schizophrenia related cognitive disturbances. O7.5 IDENTIFYING USEFUL ENDOPHENOTYPES FOR SCHIZOPHRENIA: INTEGRATING COGNITIVE PSYCHOLOGY AND TWIN METHODOLOGIES William Kremen1, Kristen Jacobson2, Hong Xian3, Seth Eisen3, Lindon Eaves4, Ming Tsuang1, Michael Lyons2 1 University of California, San Diego, La Jolla, CA 2 University of Chicago, Chicago, IL 3 Washington University, St. Louis, MO 4 Virginia Commonwealth University, Richmond, VA 2 Boston University, Boston, MA The ability to find gene-cognition associations in schizophrenia should be enhanced by refining cognitive endophenotypes. We focus on two factors or improving endophenotypes. First, strong evidence for the heritability of specific putative endophenotypes is not always available. Second, most cognitive measures are highly multi-determined; deficits may reflect very different cognitive processes in different people. The Wisconsin Card Sorting Test (WCST), for example, is widely used in schizophrenia research but several different cognitive abilities may account for performance; moreover, data from our twin sample (345 pairs) indicate that the WCST is not heritable. We suggest that a promising strategy is the integration of experimental cognitive psychology—which is aimed at parsing cognition—and twin/behavioural genetic methodologies. Twin samples provide the most powerful means for determining heritability, but cognitive assessment of twins has primarily involved a ‘broad brushstroke’ approach that is not aimed at parsing cognition. Integrating these approaches, we can delineate specific cognitive component processes and identify the extent of genetic and environmental influences. We present genetic analyses aimed at differentiating storage and executive components of working memory. In a multivariate twin analysis, we have, for example, identified an executive function component of a working memory task that is influenced by genes that are independent of other component processes involved in this task (overall h2 ¼ .51; executive component h2 ¼ .28). By more precisely specifying genetically-influenced component processes, this approach is likely to generate improved cognitive endophenotypes with greater likelihood of manifesting significant associations with candidate genes. O7.6 D2 RECEPTOR BINDING AND DOPAMINERGIC NEUROTRANSMISSION IN 22q11DS Therese van Amelsvoort1, Erik Boot2, Janneke Zinkstok2, Nico Abeling2, Lieuwe de Haan2, Don Linszen2, Jan Booij2 1 Academic Medical Center, Amsterdam 2 Academic Medical Center, Amsterdam Introduction: People with 22q11 deletion syndrome (22q11DS) are at increased risk of developing schizophrenia-like psychosis. The gene for Catechol-O-methyl-transferase (COMT), an enzyme involved in dopamine catabolism, is located at 22q11. Subsequently, adults with 22q11DS are haploinsufficient for the COMT gene. How dopaminergic neurotransmission is affected in 22q11DS is not known. Thus we
706
Abstracts
studied striatal D2 receptor binding and other dopaminergic markers at baseline and following dopamine depletion in adults with 22q11DS. Methods: Twelve non-psychotic, neuroleptic naı¨ve adults with 22q11DS and twelve age- and gender-matched healthy controls underwent 123I-IBZM SPECT at baseline. Two weeks later all subjects received 3 500mg of alpha-methyl-para-tyrosine (AMPT) given over a 4 hour period. Blood samples were drawn at 0, 3, and 6 hours following the first dose of AMPT to determine plasma levels of prolactin (PRL), vanilmandelic acid (VMA), 3-methoxy-4-hydroxyphenylglycol (MHPG) and homovanillic acid (HVA). Five hours after the first AMPT dose a second 123I-IBZM SPECT was made. Results: Following dopamine depletion, a significant increase in PRL levels and decrease in HVA, MHPG, VMA levels were observed in both groups. Baseline and post-AMPT HVA levels were significantly lower in 22q11DS subjects compared to control subjects. Post-AMPT PRL levels were significantly lower in 22q11DS subjects. There were no group differences in D2 receptor binding at baseline or following AMPT. Conclusions: Dopaminergic neurotransmission in 22q11DS is disrupted, and this may contribute to their increased risk for psychosis. O7.7 MAG BUT NOT CNP GENE IS ASSOCIATED WITH TOTAL BRAIN WHITE MATTER IN FIRST EPISODE PSYCHOSIS Aristotle Nicholas Voineskos1, Gwyneth Zai1, Natalie Bulgin1, Sajid Shaikh1, Donna Lang2, William Honer2, James Kennedy1 1 University of Toronto, Toronto 2 University of British Columbia, Vancouver Introduction: White matter abnormalities are increasingly important in the pathophysiology of schizophrenia. Converging evidence from postmortem brain gene expression studies, case control and family studies strongly implicate myelin system genes in schizophrenia. White matter abnormalities are frequently found in neuroimaging studies of psychosis. Myelin associated glycoprotein (MAG) and cyclic nucleotide phosphodiesterase (CNP) are candidate genes of great interest in psychosis, and in white matter changes in particular. Methods: Two SNPs for MAG (rs720309, rs720308) (2) and one SNP for CNP (rs 207106)(5) were genotyped in an ABI 7000 using standard protocols. Subjects (n ¼ 47) with first episode psychosis and controls (n ¼ 24) were scanned once with 1.5T, GE, T1 weighted MRI from which total white matter was calculated. Genotypes were compared with total white matter using ANCOVA (covariates: age, sex, intracranial volume). Results: Both markers for MAG showed significant association with total brain white matter (rs720309: p ¼ 0.016; rs720308: p ¼ 0.002) in the psychosis group, but no association in controls. CNP showed no association with total white matter. Discussion: To our knowledge, this is the first study to associate genetic variation with white matter brain imaging measures. We found a significant association for both markers of the MAG gene in intron 8, with the phenotyope of total brain white matter in first episode psychosis cases. Future work in this promising area should include replication with larger sample sizes, more imaging in regions of interest, Diffusion Tensor Imaging and genotyping measures of the myelin system. O7.8 FYN KINASE GENE AND PERFORMANCE ON WISCONSIN TEST IN MAJOR PSYCHOSES AND CONTROL SUBJECTS Janusz Rybakowski1, Alina Borkowska2, Maria Skibinska3, Joanna Hauser1 1 Department of Adult Psychiatry, Poznan 2 Clinical Neuropsychology Unit, Bydgoszcz 3 Laboratory of Psychiatric Genetics, Poznan Genes related to the glutamatergic system have been implicated in the pathogenesis of major psychoses and in prefrontal cognitive dysfunctions. The Src-family tyrosine kinase Fyn plays a key role in interaction between the brain-derived neurotrophic factor and glutamatergic receptor NMDA, in the prefrontal cortex (PFC). The aim of this study was to estimate the association between polymorphisms of Fyn gene and performance on prefrontal neuropsychological tests in schizophrenic and bipolar patients and in controls. The study included 188 patients with schizophrenia, 181 patients with bipolar mood disorder and 105 controls. All subjects were genotyped for the three polymorph-
isms of the Fyn gene (93 A/G, IVS10þ37T/C, and Ex12þ894T/G). They also performed the Wisconsin Card Sorting Test (WCST), where the number of perseverative errors may be a neuropsychological measure of PCF function. Schizophrenic patients with T/T genotype of IVS10þT/C polymorphism and T/T genotype of Ex12þ894T/G polymorphism obtained significantly fewer perseverative errors on WCST compared to patients with other genotypes. No significant differences in WCST performance were found in schizophrenic patients as to 93 A/G polymorphism. Likewise, no significant differences were found between any Fyn polymorphism and WCST performance in bipolar patients and control subjects. The results obtained point to a possible relationship between some polymorphisms of the Fyn gene, related to the function of the glutamatergic system, and performance on neuropsychological tests measuring PFC activity, in schizophrenic patients but not in bipolar illness or control subjects. O8.1 ANALYSIS OF CANDIDATE GENES FOR PROSPECTIVELY ASSESSED RESPONSE TO LITHIUM IN A COLLECTION OF BIPOLAR DISORDER PATIENTS FROM SARDINIA David A Collier1, Leonardo Tondo2, Janet Munro1, Sara Campos1, Gianfranco Floris2, Gerome Breen1, Robert Kerwin1, Davd Collier1 1 Institute of Psychiatry, London 2 Centro Bini, Department of Psychology, University of Cagliari 2 Assessorato Igiene Sanita` Assistenza Sociale, Cagliari The most frequently used drug for long-term treatment of bipolar disorder is lithium carbonate. It is an effective mood stabiliser for acute mania and prophylaxis and has significant antidepressant properties. Bipolar patients display differential response to lithium therapy ranging from no response in a third, to the virtual elimination of affective symptoms in another third. Some of this may be explained by genetic factors, as lithium responsiveness is a familial trait Our goal was to identify genes which influence variability in response to lithium when patients with bipolar I or bipolar II are treated with the drug. The study is based on DNA and clinical data collected from 300 bipolar I or II patients who have taken lithium for at least one year starting in the 1970’s, in Cagliari, Sardinia. These patients have been prospectively assessed for up to 25 years for response to treatment, and other clinical variables. Candidate genes for lithium response, selected from molecular targets of lithium or aetiological factors in depressive illness, including the serotonin transporter, were analysed. Four categories of Liþ response were used 4: good response (complete remission) 3: partial good response (marked improvement but not episode-free); 2: poor response (minor or modest improvement in frequency of episodes or admissions; significant morbidity); and 1: no response (no improvement or worsening of illness). We found that LL genotype and L alleles of the serotonin transporter are more common in the non-responding group (no or poor response), consistent with previous studies. O8.2 GENE-GENE INTERACTION ASSOCIATED WITH DRUG RESPONSE IN UNIPOLAR MAJOR DEPRESSION DISORDER Alison Motsinger, Kirsten Haman, Maureen Hahn, Angela Steele, Brett English, Hugh Fentress, Regina Myers, Lisa Hazelwood, Michelle Mazei, Elaine Sanders-Bush, Randy Blakely, Marylyn Ritchie, Richard Shelton Vanderbilt University, Nashville, TN Unipolar major depression disorder (MDD) is a complex disorder that affects nearly 10 million people each year in America. Fortunately, the disorder is often treatable. Several antidepressants are currently available. However, not all patients respond, and many factors are assumed to contribute to this. We hypothesise that genetic variants may be associated with treatment response. In the current study, we ascertained genetic data on a group of 50 subjects (mean age 42.8, 39% male) with MDD. Response was assessed using the Hamilton Rating Scale for Depression (HAM-D). Individuals were treated with serotonergic antidepressants, and response was measured at 16 weeks. Of the subjects ascertained, 18 had no response to medication and 32had a significant response or remitted. We genotyped 14 single nucleotide polymorphisms (SNPs) in 8 genes hypothesised to be important in the pharmacogenomics of depression. To detect genetic variations associated with drug response, we used Multifactor Dimensionality Reduction (MDR), a powerful statistical approach used to detect gene-gene interactions with or without main effects. We detected an interesting
Abstracts interaction between SNPs 5HT2A A_1438G and MAOA VNTR long short variant that predicted response with 73.33% (p < 0.055) accuracy. This model is marginally significant and is likely to reach highly significant levels with increased sample size. These findings highlight the importance of looking for gene-gene interactions in the study of complex phenotypes, like drug response. Such associations hold the promise of individualised medicine, where appropriate treatment regimens are tailored to individuals’ genetic profile. O8.3 PHARMACOGENETIC STUDIES IN CATIE, A LARGE RANDOMIZED CLINICAL TRIAL FOR SCHIZOPHRENIA Patrick Sullivan1, Goldstein David2 1 UNC/Genetics, Chapel Hill 2 Duke University, Durham The CATIE trial (N Engl J Med 2005;353:1209–23) examined the treatment responses of 1,640 individuals with schizophrenia to structured, double-blinded, and randomized treatment with perphenazine or olanzapine, quetiapine, risperidone, or ziprasidone for up to 18 months. DNA samples were collected on N ¼ 765 individuals. We have genotyped all subjects to classify predicted phenotype for 10 drug metabolizing enzymes plus > 3,000 SNPs in 118 pharmacokinetic candidate genes (eg, all known primary components of dopaminergic and serotonergic systems). The analysis plan was posted on the web prior to beginning analyses. Major hypotheses include: (i) prediction of treatment-related phenotypes (optimized dose, treatment response, neurocognitive change, and serum drug levels), (ii) prediction of adverse drug reactions (extrapyramidal effects, tardive dyskinesia, and changes in body mass and other components of the metabolic syndrome). All genotyping was completed in late 5/06, analyses are on-going, and will be finalized by the time of presentation. O8.4 PHARMACOGENETICS OF DOPAMINE RELATED GENES AND TREATMENT RESPONSE TO RISPERIDONE Nakao Iwata1, Masashi Ikeda1, Yoshio Yamanouchi1, Reiji Yoshimura2, Tsuyoshi Kitajima1, Yoko Kinoshita1, Jun Nakamura2, Norio Ozaki3 1 Fujita Health University, Toyoake 2 University of Occupational and Environment, Kitakyushu 3 Nagoya University, Nagoya To achieve personalized medicine in antipsychotic medication, a considerable number of pharmacogenetic studies have been performed in recent years. Individual variation in dopaminergic transmission systems has been hypothesized to be involved in the efficacy of antipsychotic treatment in schizophrenia. Aside from five distinct dopamine receptors (D1-D5 receptors), recent investigations indicate that the AKT/GSK3 signaling cascade plays an important role in dopaminergic systems. Here we conducted a pharmacogenetic study to evaluate whether variants in dopamine related genes (dopamine receptor genes (DRD1-DRD5), AKT1 and GSK3beta genes) can predict the improvement with risperidone (RIS) treatment for schizophrenia. One hundred and twenty one neuroleptics-naı¨ve schizophrenia patients were treated with RIS monotherapy for 8 weeks, and clinical symptoms were evaluated by PANSS. Twenty variants in these genes were genotyped, and multiple linear regression analysis was performed. TwoSNPs (TaqIA in DRD2 and AKT1SNP5) were significant predictors for treatment response to RIS. The patients with A1/A1 genotype of TaqIA and C/C genotype of AKTSNP5 showed significantly better improvement than those without these genotypes. We have identified possible predictor SNPs in DRD2 and AKT1 for RIS treatment in Japanese schizophrenia patients, since the contributions of these SNPs were greater than those of baseline PANSS score which is considered to be a reasonable contributor to improvement in PANSS after RIS treatment. Further study will be required for conclusive results to elucidate the relation of DRD2 and AKT1 to schizophrenia for use in personalized medicine. O8.5 POSTPARTUM DEPRESSION AND SEROTONIN TRANSPORTER-LINKED PROMOTER POLYMORPHISM: RESPONSE TO PHARMACOLOGIC TREATMENT Rocı´o Martı´n-Santos1, Luisa Garcı´a2, Mo´nica Grataco´s3, Puri Navarro2, Carlos Ascaso3, Estel Gelabert4, Ana Plaza2, Ricard Navine´s4, Cecilia Garcı´a3, Xavier Estivill4
707
1
Institut Municipal d’Investigacio´ Me`dica, Barcelona; Perinatal Psychiatric Unit and Gender Re, Barcelona Genes and Disease Program, and National, Barcelona 3 Public Health Department, UB, Barcelona 4 Pharmacology Research Unit, Institut Mun, Barcelona 4 Genes and Disease Program, National Geno, Barcelona 2 3
The prevalence of major depression occurring in the postpartum (PMD) is estimated at 4%–6% and results in considerable morbidity for women, their infants and families. PMD describe depression occurring anytime within 12 months following delivery. The period of higher risk appear to occur close to the time of birth between 8 to 24 weeks. A biological component of PMD is suggested by family history and twin studies. A serotonin transporter-linked promoter polymorphism (5-HTTLPR) has been linked to major depression and with selective serotonin reuptake inhibitors antidepressants (SSRI) response. As this relationships has not been studied in PMD, we have recruited a sample of 120 consecutive women diagnosed of major depression and 120 without any psychiatric disorder evaluated after birth at the Perinatal Psychiatric Unit. Diagnose was established through a DSMIV-structural interview. Therapeutical SSRI response was evaluated by: 1) 9 score in the Edinburgh Postnatal Depression Scale, 2) 50% improvement of the Hamilton Rating Scale for Depression, and 3) 3 score in the Severity of Affective disorders, at baseline, 8 and 24 weeks of treatment. Preliminary results in a subsample of 90 patients indicate that the 5-HTTLPR polymorphism is not related to pharmacologic response in PMD. In spite of that, inclusion of other covariates and the analysis of the whole sample should clarify the involvement of a genetic susceptibility to antidepressant response. This ´ -TV3 (PI: Garcı´a) and study has been done in part with MARATO RGPG:GO3/184 (PI: Martı´n-Santos) grants. O8.6 GENES OF THE DOPAMINE AND SEROTONIN SYSTEMS AND RESPONSE TO METHADONE THERAPY Csaba Barta1, Agnes Szilagyi1, Katalin Hoffmann2, Zsolt Demetrovics2, Maria Sasvari-Szekely1 1 Semmelweis University, Budapest 2 Eotvos Lorand University, Budapest Candidate genes coding for proteins involved in the brain’s reward system which comprises the dopaminergic mesocorticolimbic pathway and is under modulation by the serotonin system were studied. The genes that are possibly involved in the development of drug dependence include the dopamine D4 receptor (DRD4), dopamine transporter (DAT), serotonin transporter (SERT), catechol-O-methyltransferase (COMT) and monoamino oxidase A (MAO-A). The aim was to identify genetic effects influencing response to methadone maintenance therapy (MMT) in heroin dependent patients. This pharmacogenetic study reports genotyping of ten polymorphisms of the above genes in heroin dependent subjects who were phenotypically characterized in detail by a battery of psychological questionnaires and clinical data. The analyzed polymorphic sites include the coding region (exon 3 VNTR) and the 50 upstream region (521CT, 615AG, 616CG and 120 bp duplication) of the DRD4 gene, the DAT 3’ VNTR and intron 8 VNTR, the 50 upstream region (5-HTTLPR), and the intron 2 polymorphisms (STin2) of SERT, the 158AG SNP in COMT and the 50 VNTR of the MAO-A gene. Significant genetic effects were observed in case of certain genotypes of the DRD4 gene with relation to good versus poor response to MMT, which might in the future serve as an aid to establish individualized therapy of heroin dependent patients. O8.7 VULNERABILITY, RESILIENCE AND RESPONSE TO PSYCHOTROPIC DRUGS: SHARED GENETIC FACTORS? Hans H. Stassen, C. Scharfetter Psychiatric University Hospital, Zurich The efficacy of psychotropic drugs in the treatment of psychiatric disorders depends essentially on the severity of psychopathology at baseline, with placebo-drug differences hardly ever reaching significance in mild cases. Likewise, family studies of major psychiatric disorders demonstrate that the magnitude of genetic predisposition (vulnerability) is correlated with severity of illness, thus suggesting a close relationship between response to psychotropic drug treatment, severity of illness and genetic predisposition. Based on data from 71 multiplex nuclear families (527 subjects) ascertained through index
708
Abstracts
cases diagnosed with bipolar illness or schizophrenia, we have addressed the question of diagnosis-independent susceptibility to functional psychoses. Subjects were genotyped for 553 polymorphisms of a genome scan along with 48 candidate genes hypothesized to be involved in psychotropic drug response and lying in genomic regions harboring vulnerability loci for functional psychoses. Our analyses revealed both positive and negative genotype-phenotype correlations, and thus the presence of co-existing vulnerability and resilience factors. Most interestingly, there was some overlap with genomic regions identified through our study on psychotropic drug response (257 patients treated with antidepressants or antipsychotics and genotyped for 178 candidate genes), where positive as well as negative correlations likewise suggested the existence of mechanisms that can support or impede recovery. In consequence, our data indicate that shared genetic factors can increase or decrease the likelihood for developing the disorder, as they can support or impede recovery. Detailing the characteristics of these factors will ultimately lead to considerably improved treatments along with more personalized medicines. O8.8 PHARMACOGENETICS OF ANTIDEPRESSANT EFFICACY Alessandro Serretti Bologna University, Bologna Pharmacogenetic studies on mood disorders attempt to identify gene polymorphisms associated with the therapeutic efficacy and side effects of antidepressants. Antidepressant drugs are the first-line treatment for major depression, but a therapeutic response in clinical practice is expected in about two-thirds of patients. The great inter-individual variability in the pharmacological response pattern has been partially ascribed to heritable factors. During recent years, the possible influence of a set of candidate genes as possible genetic predictors of antidepressant response efficacy was investigated and will be reviewed here. The functional polymorphism in the upstream regulatory region of the serotonin transporter gene (5-HTTLPR), the C(1019)G variant in the 5HT1A receptor gene, the A218C gene variant in the tryptophan hydroxylase gene (TPH), the 102TC variant in the 5HT2A receptor gene, the G-protein beta3-subunit (Gbeta3) C825T gene variant, the glucocorticoid receptor-regulating cochaperone (FKBP5) and the Circadian Locomotor Output Cycles Kaput (CLOCK) gene variants were independently associated with short-term SSRIs antidepressant efficacy. Marginal associations were reported for COMT, ADRB1, NET, 5HT6, CRHR1, ACE I/D, Mao-A, BDNF and IL-1beta. A number of other gene variants were not associated with outcome. Although still preliminary, the results obtained on the pharmacogenetics of antidepressants are promising for individualised therapy. O9.1 X-LINKED PARENT-OF-ORIGIN EFFECTS IN BRAIN: IMPLICATIONS FOR SEX-SPECIFIC NEUROBIOLOGY William Davies1, Anthony Isles2, Ji-Yun Song1, Doreen Burrmann1, Paul Burgoyne2, Lawrence Wilkinson2 1 The Babraham Institute, Cambridge 2 Cardiff University, Cardiff 2 National Institute for Medical Research, London Imprinted genes are monoallelically expressed in a parent-of-origin dependent manner and may influence neurobiological phenotypes. Two murine X-linked imprinted genes (Xlr3b and Rhox5) have recently been described, one or both of which may underlie an X-linked parent-of-origin effect on reversal learning. Using microarrays, we compared autosomal gene expression in the brains of adult 39,XpO and 39,XmO mice (inheriting a single X chromosome paternally or maternally respectively) to identify brain processes dependent upon X-linked imprinting. Six genes were found to differ significantly in their expression between the two groups, including reelin and a cluster of genes encoding GABA receptor subunits, and these now constitute downstream candidates by which X-linked imprinted genes may elicit parent-of-origin effects on cognition. The asymmetric inheritance of the X chromosome between males and females means that X-linked imprinted genes may contribute towards sexually dimorphic phenotypes. To assess the extent of this contribution, we use a mouse model in which the effects of sex-linked genes may be dissociated from the effects of gonadal hormones, to examine brain gene expression and cognitive behaviours dependent upon the parental origin of the X. The present work aims to identify molecular and behavioural pathways upon which X-linked imprinting impacts, and which may therefore be sexually dimorphic. Hence, it will
provide insights into normal cognitive sexual dimorphism, into why men and women are differentially vulnerable to certain common mental disorders, and into sex-specific responses to therapeutics. O9.2 NATURE, NURTURE AND NOISE: THE IMPORTANCE OF INTRINSIC DEVELOPMENTAL VARIATION IN BRAIN WIRING Kevin Mitchell Trinity College Dublin, Dublin There is mounting evidence that many psychiatric disorders may have underlying defects in connectivity. A growing number of genes involved in axon guidance or cell migration have been implicated in psychiatric disorders, including Robo-1 (dyslexia), Slitrk6 (Tourette’s), Nrg1 and DISC1 (schizophrenia). Variation in genes controlling brain wiring is thus a major potential contributor to psychiatric disorders. At the same time, twin and adoption studies have clearly shown that the heritability of psychiatric disorders is less than one, which has been interpreted as evidence for environmental effects. This interpretation overlooks another major source of phenotypic variance in brain wiring: intrinsic developmental variation. The processes of axon guidance and cell migration that underlie brain wiring are incredibly complex and inherently noisy, although they are normally heavily buffered and the system is usually remarkably robust. Genetic perturbations can dramatically sensitise the system, however, both to further genetic modifiers and also to the effects of intrinsic noise. This intrinsic variability can be directly observed in comparisons of neurodevelopmental phenotypes between genetically identical organisms or even between repeated structures within a single organism. Wiring events are especially sensitive to threshold effects and random noise may thus cause discrete rather than graded differences in outcome (developmental ‘canalisation’). In summary, we should expect a significant contribution to phenotypic variance from intrinsic variation in the processes controlling brain wiring. Many psychological or psychiatric traits may therefore be largely innate, despite being only partly inherited. O9.3 EPIGENOMIC MICROARRAY PROFILING OF DNA METHYLATION VARIABILITY IN MONOZYGOTIC AND DIZYGOTIC TWINS Zachary Kaminsky1, Sun-Chong Wang1, Sigrid Ziegler1, Irving Gottesman2, Albert Wong1, Allan McRae2, Peter Visscher2, Nick Martin2, Arturas Petronis1 1 Centre for Addiction and Mental Health, Toronto, Ontario 2 University of Minnesota, Minnesota 2 Queensland Institute of Medical Research, Brisbane The DNA sequence-based paradigm offers little insight into the mechanisms of phenotypic discordance of monozygotic (MZ) twins, while epigenetic metastability and the critical role of epigenetic regulation in phenotypic outcomes may account for all or part of this phenomenon. A novel microarray- based technique for DNA methylation profiling was employed to investigate the levels of epigenetic differences and similarities in MZ and dizygotic (DZ) twins. Twin DNA was obtained from 9 sets of MZ and the same number of DZ twins consisting of 5 female and 4 male twin sets aged from 12 to 14 yrs old from the Australian Twin Registry. The unmethylated fraction was enriched using peripheral blood leukocyte DNA and interrogated on the human CpG island microarray containing over 12,000 features. Comparison of twin vs. co-twin hybridizations (biological variance group) and twin vs. self control hybridizations (technical variance group) demonstrated numerous loci exhibiting biological DNA methylation differences within pairs of MZ co-twins. Microarray data from MZ and DZ twin groups was subjected to hierarchical clustering, which identified genomic regions of epigenetic variability consistent across all twins. A within-array across-feature correlation analysis suggested a higher variability in the DZ twins as compared to MZ twins. A larger degree of epigenetic variation in DZ twins could be due to 1) epigenetic variation in the germline, 2) DNA sequence impact on epigenetic profiles, or both. This study will contribute to epigenetic analyses of identical twins discordant for complex disease. O9.4 COORDINATED GENETIC AND EPIGENETIC REGULATION OF COMT, RELN, DRD GENE EXPRESSION IN SCHIZOPHRENIA AND BIPOLAR DISEASE Cassandra Smith, Hamid Abdolmaleky, Sam Thiagalingam Boston University, Boston
Abstracts Genetic and environmental factors early in development are important in schizophrenia and bipolar disease etiology. Genetic studies implicate many genes; yet different genes are mutated in different families. Previously, our research (confirmed by others) showed that methylation of the RELN promoter correlated with schizophrenia. DNA methylation of promoter regulates gene expression epigenetically. Here, the goal was to understand how DNA sequence variation and promoter methylation correlated with gene expression and activity, and disease. RNA and DNA samples (105) from post-mortem brain were studied using conventional PCR and sequencing methods. Hypomethylation of the MB-COMT promoter, particularly in the left frontal lobe, of schizophrenia and bipolar patients vs controls was detected (74% (P ¼ 0.005) and 71% (P ¼ 0.008) versus 40%, respectively) while the RELN and DRD2 promoters were hypermethylated (P ¼ 0.01) in the patient samples. The Val158 (hyperactive) COMT allele was more frequent in patients, and a correlation was found between VAL158 genotype and RELN promoter methylation (e.g. ten VAL158 homozygotes, eighteen VAL158MET homozygotes and, heterozygotes had methylation levels of 100%, 50%, and 76%, respectively (P ¼ 0.01)). Hypomethylation of the RELN and DRD2 promoters correlated with hypomethylated vs hypermethylated MB-COMT promoter in controls vs patients, respectively. An inverse relationship between RELN and MB-COMT methylation vs expression level, and a direct correlation between DRD1 expression and RELN expression level was detected. These results suggest MB-COMT hyperactivity reduces synaptic dopamine reduces DRD1 and RELN expression contributing to the pathogenesis of disease.
O9.5 ENVIRONMENTAL FACTORS IN OBSESSIVECOMPULSIVE SYMPTOMATOLOGY: WHAT CAN WE LEARN FROM MONOZYGOTIC TWINS DISCORDANT ON OC SYMPTOMS? Danielle Caroline Cath1, Daniel S. van Grootheest2, Gonneke Willemsen3, Dorret I. Boomsma3 1 VU University, NL 2 Biological Psychiatry/Psychiatry VU Uni, Amsterdam 3 Biological Psychology, Amsterdam Our twin data suggest that 53–64% of the variation in ObsessiveCompulsive (OC) phenomenology in adults is explained by environmental factors. To date, no twin studies have examined the nature of such factors. The few clinical studies available have suggested diverse individual-specific factors (such as low SES, childhood trauma, exposure to postpartum events in women), as well as common environmental factors (such as dysfunctional parenting), underlie exacerbation of OC behaviour. To explore environmental factors that protect or exacerbate OC symptoms, two types of twin comparisons are particularly informative, i.e. 1) within-twin comparisons of MZ discordant twins to study individual-specific environmental factors, and 2) between-twin comparisons of MZ twins concordant high and concordant low on OC symptoms to study common environmental factors. In a large twin sample in the Netherlands encompassing 2672 adult twin pairs, an abbreviated 12-item version of the Padua Inventory (Sanavio, 1988) was administered in 2002, assessing OC symptomatology, along with a variety of questions on family structure, health, lifestyle, birth complications and life events. Applying a stringent criterion for discordant OC behaviour (one twin scoring > 17, the other < 7 on the PI; scores based on thresholds in clinical samples), we find that 66 MZ twin pairs are discordant for OC symptoms, 29 MZ twins concordant high (both > 17) and 512 MZ pairs concordant low (both < 7) on OC symptoms. Their data provide information on the type of unique and common environmental factors of importance in OC symptomatology.
O10.1 PSYCHIATRIC GENETICS: EMERGING PERSONAL AND FAMILY IMPLICATIONS Katie Featherstone Cardiff Univeristy, Cardiff In this paper, we examine the probable social implications of recent and ongoing developments in the field of psychiatric genetics for individuals and their families affected by schizophrenia and bipolar affective disorder. We argue that the social implications of genetic risk
709
assessments are likely to display some particular characteristics that, while not unique, are particularly urgent in this context. The potential for stigma and shame among family members, pressures for secrecy and non-disclosure among kindred and the medicalisation of moral categories have special salience in psychiatry. We have already identified complex dynamics of information-exchange and the disclosure or withholding of genetic risk information amongst individuals and families affected by a wide range of familial conditions. Families are sites of mutual surveillance and scrutiny and they are also sites for complex patterns of belief and understanding of genetics, inheritance, risk, health and illness. Thus, in this paper we will build upon these findings to discuss potential ways in which these themes are likely to be manifested in the context of recent and ongoing developments in the field of psychiatric genetics. O10.2 THE SEARCH FOR SUSCEPTIBILITY GENES IN A BRAVE NEW WORLD: PSYCHIATRIC GENETICS, ETHICS AND LAW Daniel L. Ambrosini McGill University, Montreal, Quebec The fields of psychiatry and genetics are converging as the new field of psychiatric genetics. As they reposition themselves from the theoretical to the practical domain, international guidelines and a code of practice should be established to ensure that scientific research is conducted ethically. The research methods used by psychiatric geneticists in identifying ‘susceptibility genes’ for mental illness are complex and have intrinsic limitations that are not widely apparent. Psychiatric geneticists have an ethical obligation to be sensitive to psychosocial perceptions, the effects of stigmatisation and past historical abuses in eugenics, sexual sterilisation and genetic discrimination. Competing societal and individual concerns must be reconciled. Psychiatric genetics raises new ethical and legal obligations for clinicians and researchers in the areas of confidentiality, privacy and the commercialisation of genes. Without a fuller understanding of how this research affects social policy considerations, such as the employment context, predictive school testing, adoption and insurance schemes, we risk repeating past ethical mistakes within psychiatry. In order to obtain ethics approval and Agency funding, Research Ethics Board members need greater knowledge and awareness of the science and ethical ramifications governing this field. Psychiatric genetics is also having a modest effect on the philosophy of criminal law, as the search for susceptibility genes threatens to displace traditional notions of criminal responsibility and mens rea, and new classes of defence and scientific evidence are being questioned in the courtroom context. O10.3 PHARMACOGENETICS OF ADDICTIVE BEHAVIOURS: PHILOSOPHICAL QUERIES AND PUBLIC CONCERNS Sarah Wilson1, Elisa Pieri Lancaster University, Lancaster ‘New’ genetic technologies, such as pharmacogenetics, predictive genetic testing and nutrigenomics, are thought to generate new responsibilities at the individual level by promoting personalised approaches to healthcare and ‘healthy’ living. The individualisation promoted through these genetic technologies raises ethical and social concerns associated with the tension between individual and collective rights and responsibilities, particularly vis a vis the role of the State in the provision of healthcare. More specifically, research on pharmacogenetics of behaviours, such as pharmacogenetics of addiction, has the potential to impact notions of personal responsibility and choice, and reconfigure our understandings of moods, behaviours and mental health. Also of relevance are concerns relating to pharmacogenetics and the potential for increased discrimination for ethnic minorities, and how this may exacerbate apparent diagnostic differential already existing in mental health in the UK context. This paper builds on work carried out at the ESRC Centre for Economic and Social Aspects of Genomics (CESAGen) and in collaboration with Nowgen, one of the six genetics knowledge parks funded by the UK Department of Health. It is based on a study of public views and values on personalised medicine in the UK, and it addresses the ways in which notions of individual and governmental responsibility for health and well being are being challenged and transformed through research into pharmacogenetics of addictive behaviours
710
Abstracts
O10.4 PSYCHIATRIC GENETIC TESTING: WHO WOULD WANT IT AND WHY? Mett Marri Laegsgaard1, Ole Mors2 1 Centre for Basic Psychiatric Research, Risskov 2 Centre for Basic Psychiatric Research, Aarhus Psychiatric genetic research might lead to the development of presymptomatic and prenatal tests. Who would want these tests and for what reasons? Prior studies investigating intentions to test included exclusively patients with bipolar disorder, and reported only intentions to test, not factors influencing these intentions. We asked 397 psychiatric patients about their intention to take a psychiatric genetic test, and about their expectations and fears concerning psychiatric genetics. Patients were diagnosed with recurrent depression (n ¼ 228), anxiety (n ¼ 105), bipolar disorder (n ¼ 29) or schizophrenia (n ¼ 35). SCAN 2.1 was used for psychiatric diagnoses. 83% of patients expressed intention to test: 45% would test only if effective treatment/prevention exists, while 38% would test notwithstanding treatment possibilities. More patients with depression than patients with anxiety expressed intention to test (OR 2.3, p ¼ 0.02), and more parents than non-parents expressed that intention (OR 2.6, p ¼ 0.004). Patients are more in favour of testing when expecting to feel better prepared, knowing of the presence of risk genes (OR 9.3, p < 0.0001), and less in favour of testing when fearing that psychiatric genetics leads to an attitude of some lives being worth more than others (OR 0.5, p ¼ 0.013), or fearing that knowing of the presence of risk genes could bring on the disease in question (OR 0.5, p ¼ 0.002). Knowledge of patients’ expectations and fears concerning genetic testing is of great importance in a psychiatric genetic counselling service.
O11.1 EXPRESSION OF THE KIAA0319 GENE FROM A HAPLOTYPE ASSOCIATED WITH DEVELOPMENTAL DYSLEXIA Megan Dennis1, Silvia Paracchini2, Richard Wade-Martins2, Antonio Velayos-Baeza2, Eric D. Green3, Anthony P. Monaco2 1 University of Oxford, Oxford 2 Wellcome Trust Centre for Human Genetics, Oxford 3 Genome Technology Branch, National Human, Bethesda, MD Developmental dyslexia (DD) is a condition that affects an individual’s ability to read and spell in the absence of any profound sensory or neurological impairment, despite adequate intelligence and educational opportunity. Recently, a 77-kb region of human chromosome 6p22 was found to be associated with DD, and a ‘‘risk’’ haplotype spanning this region has been identified. This genomic segment encompasses the first four exons of the KIAA0319 gene, the entire TTRAP gene, and the first exon of the THEM2 gene. Allele-specific expression analysis of various cell lines has shown a relative lower expression of KIAA0319 associated with the risk haplotype, while TTRAP and THEM2 expression from the risk and non-risk haplotypes is the same. We sought to identify the specific functional variant on the risk haplotype affecting KIAA0319 expression by studying bacterial artificial chromosomes (BACs) containing the risk versus non- risk haplotype. Two BACs containing each haplotype were sequenced and all genetic variants catalogued. Variants residing within multi-species conserved sequences (MCSs) were tested for association with DD in a collection of 264 families. We are now focusing our functional analyses on variants falling in or near the KIAA0319 promoter region using luciferase-based reporter plasmids constructed from risk and non-risk BACs. Transfection studies in human cell lines show that the risk haplotype confers lower expression compared to a non-risk haplotype. These results suggest that a DD-associated functional variant may reside in the 4-kb sequence upstream of the KIAA0319 gene.
O11.2 KLINEFELTER SYNDROME AS A MODEL FOR VERBAL COGNITIVE DISORDERS Lynn E. DeLisi1, Philip D. Harvey2, Marquis Vawter3 1 New York University, New York 2 Mt Sinai School of Medicine, NY 3 UC-Irvine, Irvine, CA Background: Klinefelter Syndrome (a chromosomal karyotype with one or more extra X chromosomes) is thought to be due to overexpression of genes escaping inactivation on the X chromosome.
However, this concept has not been tested. We thus performed the following study. Methods: Lymphoblastoid cell lines from 11 males with documented Klinefelter Syndrome and 6 XY control males were examined using a comprehensive genome-wide Affymetrix microarray containing transcripts from the X and Y chromosomes as well as all autosomes. Results: Ninety-three genes were found to be differentially expressed in the XXY males compared with the XY males after correction for multiple testing. These genes included only 4 on the X chromosome, with the remainder located on autosomes. Nine transcripts showed significant correlation of expression with multiple measures of verbal cognition. Expression of one pseudoautosomal gene, GTPBP6 (GTP binding protein 6, putative; alias PGPL) was inversely correlated with verbal IQ (r ¼ 0.86, p < 0.001) and four other measures of verbal ability. A second gene, RECQL (RecQ protein-like (DNA helicase Q1like) also displayed several significant correlations with cognitive measures. Conclusions: Although many genes on the X chromosome escape inactivation on the extra X chromosome in normal females, it is unclear whether the same mechanism applies to extra X chromosomes in Klinefelter men. Further examination of genes that show correlations of expression with verbal cognition, such as GTPBP6 and RECQL, may establish a causal link between these genes and language function.
O11.3 GENE EXPRESSION PROFILING IN WHOLE BLOOD: FINDING BIOMARKERS FOR DEPRESSION Witte J.G. Hoogendijk1, S. De Jong2, J. Van Zanten3, B. Penninx4, R. Van Dyck4, J. Smit5, B. Ylstra6, A.B. Smit6, S. Spijker6 1 VU University Medical Center, Amsterdam 2 VU University (VU), Amsterdam 3 GGZ Buitenamstel (GGZBA), Amsterdam 4 VUmc, Amsterdam 5 GGZBA, Amsterdam 6 VU, Amsterdam Major Depressive Disorder (MDD) is a heritable, highly-prevalent disorder. In order to gain biomarkers for depression and to identify patient subclasses, we analysed the gene expression of whole blood cell challenges. Genome responsiveness in the individual genetic background of patients and controls was probed in 12 unmedicated wellmatched depressed cases and 11 healthy controls, as collected by the Netherlands Study of Depression and Anxiety (NESDA). Whole blood samples were either kept at baseline (directly taken from the heparine collection tube) or challenged with lipo-polysaccharides (LPS; 10 ng/ml blood) for 5 h. LPS-induced, as well as base-line gene expression, was determined by micro array analysis (Agilent whole genome arrays). LPS appeared to be a potent stimulus, because 392 out of the 12,205 genes analysed (i.e. 3.2%), were regulated more than 4-fold in all samples. From these genes, the majority was up-regulated by LPS (288 out of 392). Then, MDD-specific changes in LPS-induced gene expression were examined. We observed 89 genes that were differentially regulated by LPS in MDD cases vs controls (p < 0.005). Gene expression that was found to be specific for the disease-state will be verified using real-time quantitative PCR. Besides aiding in more accurate diagnosis, pharmaco-treatment and follow-up strategies for MDD, analysis of gene expression in whole blood samples might also be of use in identifying individuals at risk for the (re) occurrence of depression.
O11. 4 CO-CHAPERONES OF THE GLUCOCORTICOID RECEPTOR AND DEPRESSION DURING PREGNANCY Elyse Katz, Elisabeth Binder, Todd Deveau, Jeff Newport, Zachary Stowe, Joseph Cubells, Charles Nemeroff Emory University, Decatur, GA Depression during pregnancy is a major women’s health concern. Pregnancy induces dramatic changes in the function of the hypothalamic-pituitary-adrenal axis, leading to elevated maternal cortisol levels. Glucocorticoid receptor (GR)-resistance occurs in 80% of pregnant women, and appears to be a specific mechanism limiting the activity of cortisol during pregnancy. Impairment of this adaptive response to elevated cortisol may contribute to risk for depression during pregnancy. We hypothesise that altered expression of GR cochaperones, which regulate GR signalling, modulates GR sensitivity in
Abstracts pregnancy. We investigated the peripheral-blood monocyte mRNA expression profiles of the GR and 12 of its co-chaperones in 8 women at 6 time points (pre-conception, 12, 24 and 36 weeks gestation and 4 and 12 weeks postpartum) and in 8 depressed and 8 non-depressed women at 24 weeks gestation. During gestation, we observed a significant upregulation of FKBP5 and NCOA1 message, while FKBP4 andSTUB1 mRNA were significantly down-regulated. A significant up-regulation of GR message was also observed. The GR and its co-chaperone transcripts returned to preconception levels by 12 weeks postpartum, except for PPIA, which was elevated at both 4 and 12 weeks postpartum. Levels of FKBP5 and BAG1 mRNA correlated negatively with Beck depression scores. The down-regulation of FKBP5 and BAG1 in women with peripartum depression presumably reduces GR resistance by increasing the biological activity of cortisol, which may contribute to depression during pregnancy. O11.5 A COMT REGULATORY LENGTH VARIANT IS PART OF A SCHIZOPHRENIA RISK HAPLOTYPE Philipp Sand1, Margot Albus2, Dieter Bernd Wildenauer3, Margitta B Borrmann-Hassenbach2, Sibylle Gabriele Schwab3, Peter Eichhammer4 1 University of Regensburg, Regensburg 2 District Hospital, Division of Teaching, Haar 3 Western Australian Institute of Medical, Perth 4 University of Regensburg, Dept. of Psych, Regensburg The COMT gene has long been viewed as a candidate gene in schizophrenia on the grounds of linkage findings, the psychopathology associated with deletion of the gene in velocardiofacial syndrome and the participation of COMT in dopamine metabolism. While many previous studies have produced mixed results with regard to association of the common missense variant Val158Met, COMT gene regulatory variation has received little attention. We performed a combined analysis of 94 nuclear families, plus 221 unrelated control subjects, using four markers spanning the entire gene. Variants under investigation comprised a common 67bp-length polymorphism located in the first intron, which affected gene expression in transient transfection experiments. Out of 13 four-marker haplotypes observed, only one comprising the rare, low-expressing allele, was significantly over-transmitted to affected off-spring (nominal p ¼ 0.03). This haplotype was also the only four-marker risk haplotype to emerge in parallel from the case-control approach (Wald ¼ 2.79, p ¼ 0.005) and replicates a nested three-marker risk haplotype identified in two earlier familybased studies. Unlike the haplotype, single alleles were not associated with the disease in either sample. The present data suggest that the heterogeneity of findings in schizophrenia with regard to association of the Val158Met variant may be due to interactions with the newlyidentified modifier of COMT expression. O12.1 CYTOGENETIC FINDINGS OF CEREBRAL CORTEX CELLS IN ALZHEIMER’S DISEASE Biljana Spremo-Potparevic1, Lada Zivkovic2, Vladan Bajic3, Ninoslav Djelic4 1 Faculty of Pharmacy, Belgrade, Serbia 2 Faculty of Pharmacy, Department of Biology, Belgrade 3 Institute of Biomedical Research Galenika, Belgrade 4 Faculty of Veterinary Medicine, Department of Biology, Belgrade Premature centromere division (PCD) represents loss of control over sequential separation and segregation of chromosome centromeres. It is associated with various diseases. PCD on the X chromosome was first described in ageing women. Its verification was also confirmed in individuals suffering from Alzheimer disease (AD). A vast number of cytogenetic investigations in AD patients were done on chromosomes of peripheral blood lymphocytes and skin fibroblasts. PCD was described as an alteration which can be seen during mitotic metaphase. Recent data has shown that PCD appears earlier than previously presumed. The FISH method applied to the centromere region of the X chromosome in interphase nuclei of peripheral lymphocytes in AD patients, demonstrated that PCD appears after completion of DNA replication, in G2 phase of the cell cycle. In view of the fact that neurons can re-enter into the cell division cycle, present study were done on neurons of the frontal cerebral cortex. Interphase nuclei from neurons of the frontal cerebral cortex were analyzed in 5 sporadic AD patients after autopsy and 5 agematched controls. Our results confirm re-entry in the cell cycle of brain cells of the frontal lob and that PCD in the interphase of the cell cycle
711
shows a pattern of genome instability. Based on the findings of PCD in the nerve cells and the peripheral blood lymphocytes of Alzheimer patients we can conclude that PCD of the X chromosome can be used as a cytogenetic biomarker of Alzheimer patients. O12.2 HETEROZYGOCITY RATE OF HUMAN MINISATELLITE UPS29 AT PARKINSON DISEASE Eugene Patkin1, Darija Shubina1, Natalija Slominska1, Sofia Pchelina2, Aleksandr Schwarzman2, Irina Suchkova1 1 The Institute of Experimental Medicine, Saint-Petersburg 2 St. Petersburg State Medical University, Saint-Petersburg We examined length polymorphism (LP) and heterozygosity (HZ) of GC-rich human minisatellite UPS29 in healthy volunteers and patients with the early-onset form of Parkinson disease (PD). UPS29 consists of 46 bp imperfect repeats within intron 13–14 of gene CENTB5 (1p36.33). Three loci (PARK6, PARK7, PARK9) associated with early-onset disease are in region 1pter-1p36.33. PCR analysis of healthy and PD patients as a whole revealed one main 16-repeat allelic variant (rate 88.9% and 85.0% respectively), and additionally 13 allelic variants consisting of 4.5 to 29 repeats (rate from 0.48% to 5.29% in norm and 1.1% to 5.3% in PD). Comparable groups differed strikingly in HZ rate (13.46% and 31.91% respectively). One allele was constant, 16 repeats in length, whereas the second allele varied, but was always the short one (from 4.5 to 14 repeats). Thus minisatellite UPS29 is low polymorphic and probably non-hypervariable, and the appearance of the short allelic variant of UPS29 is associated with the early-onset form of PD, even in a heterozygous state. The influence of this mechanism is still unclear, and is under investigation in our laboratory. Keeping in mind that UPS29 is localised in the cluster of genes (SCNN1D, KCNAB2, HTR1D, ACOT7,ESPN, TP73) implicated in various neuron functions, it could be anticipated that UPS29 instability might also be associated with other neurological and psychiatric diseases and conditions. Work supported by RFBR 1 04-04-4816. O12.3 PLEIADES PROMOTER PROJECT: GENOMIC RESOURCES ADVANCING THERAPIES FOR BRAIN DISORDERS Elizabeth M. Simpson1, Wyeth W. Wasserman2, Robert A. Holt3, Steven J. Jones4, Daniel Goldowitz5, Stephen J.A. Ward6, Sue Kingsley7 1 Centre for Mol. Medicine & Therapeutics, Vancouver, British Columbia 2 Centre for Mol. Medicine & Therapeutics, Vancouver, British Columbia 3 BC Genome Sciences Center, Vancouver, British Columbia 4 BC Genome Sciences Center, Vancouver, British Columbia 5 UT Research Center of Excellence in Genomics & Bioinformatics, Memphis, Tennessee 6 School of Journalism, University of British Columbia, Vancouver, British Columbia 7 International BioPharma Solutions Ltd., Vancouver, British Columbia Brain disorders represent an enormous, growing social and economic burden and a major unmet treatment need. Therapeutic obstacles include the blood-brain barrier, unwanted side effects, non-dividing cells and high-risk surgical procedures. However, gene therapy, directly or via stem cells, holds great therapeutic promise. The Pleiades Promoter Project (www.cisreg.ca/pleiades) aims to generate 160 fully characterised, human DNA MiniPromoters (less than 4 kb) to drive gene expression in defined brain regions of therapeutic interest for diseases such as Alzheimer’s, Parkinson’s, Huntington’s, Amyotrophic Lateral Sclerosis, Multiple Sclerosis, Spinocerebellar Ataxia, Depression, Autism and Cancer. As part of the Mouse Atlas project (www.mouseatlas.org), seventeen brain regions of therapeutic interest were isolated from mouse by laser capture microscopy, and SAGE (Serial Analysis of Gene Expression) libraries prepared. We are presently using these data and existing genomic expression resources to identify genes with medium- to high-level, region-specific expression. For each gene, four MiniPromoter constructs will be generated by PCR (often stitching together non-contiguous DNA) and cloned into gene-targeting vectors to make 200 new stains of knock-in mice. Detailed histological analyses will be conducted to document the region
712
Abstracts
and cell-type specificity of these MiniPromoters. Thus, we will develop a richly-characterised ‘tool-set’ of resources for both basic and therapeutic research. To this end we are recruiting future collaborators who may be interested in using the MiniPromoter mice, embryonic stem cells, promoter constructs or promoter-design software in their research.
and was localized to cytoplasmic vesicles in Neuro2A cell culture. In brain sections, dynapsin was localized to neurons and neuritic plaques. We suggest that dynapsin represents a novel linker protein between vesicles and the dynein-dynactin complex and is involved in endosomal trafficking. We conclude these results strongly suggest a role for dynapsin in the genesis of AD.
O12.4 GLUTATHIONE S-TRANSFERASE P1 *C VARIANT PREDICTS COGNITIVE AND FUNCTIONAL OUTCOME IN ALZHEIMER DISEASE Gianfranco Spalletta1, Sergio Bernardini2, Fabio Macciardi3, Giuseppina Bonaviri4, Alberto Trequattrini5, Carlo Caltagirone2, Bossu` Paola1 1 IRCCS Fondazione Santa Lucia, Rome 2 Tor Vergata University, Rome 3 Milan University, Milan 4 La Sapienza University, Rome 5 ASL Citta` di Castello, Citta` di Castello
O13.1 MOOD-INCONGRUENT PSYCHOTIC FEATURES IN BIPOLAR DISORDER: FAMILIAL AGGREGATION AND SUGGESTIVE LINKAGE TO 2p11-q14 AND 13q21-33 Fernando Goes1, Peter Z. Zandi2, Kuangyi Miao3, Francis J. McMahon4, Jo Steele4, Virginia L. Willour3, Dean F. MacKinnon3, Francis M. Mondimore3, Barbara Schweizer3, John I. Nurnberger4, John P. Rice5, William Scheftner6, William Coryell7, Wade H. Berrettini8, John R. Kelsoe9, William Byerley10, Dennis L. Murphy4, Elliot S. Gershon5, J. Raymond DePaulo3, Melvin G. McInnis4, James B. Potash3 1 Johns Hopkins, Baltimore, Maryland 2 Johns Hopkins Bloomberg School of Public Health, Baltimore, MD 3 Johns Hopkins School of Medicine, Baltimore, MD 4 National Institute of Mental Health, Bethesda, MD 4 Indiana University School of Medicine, Indianopolis, IN 5 Washington University School of Medicine, St. Louis, MO 6 Rush University Medical Center, Chicago, IL 7 University of Iowa School of Medicine, Iowa City, IA 8 University of Pennsylvania School of Med, Philadelphia, PA 9 University of California, San Diego, San Diego, CA 10 University of California, San Francisco, San Francisco, CA 5 University of Chicago, Chicago, IL 4 University of Michigan Medical School, Ann Arbor, MI
Objective: Oxidative stress has been suggested as a contributor of Alzheimer’s Disease (AD) neurodegeneration, particularly in those patients with late onset (LOAD). We studied glutathione S-transferase (GST) polymorphisms because the genes code for enzymes that are critical in defending against oxidative stress. Methods: In a three-year longitudinal study we analyzed the effect of GST P1-M1-T1 polymorphisms on the clinical course of global cognitive level, as measured by the Mini Mental State Examination (MMSE), behavioural and psychological symptoms, as measured by the Neuropsychiatric Inventory (NPI) and basic activities of daily living, as measured by the Physical Self-Maintenance Scale (PSMS) in 130 LOAD patients who were drug naive and had undergone the first clinical examination for the diagnosis of AD. Results: Carriers of GSTP1 *C allelic variant (n ¼ 21) have a faster decline in global cognitive functions (repeated measures ANOVA: F ¼ 4.001; df ¼ 3,360; p ¼ 0.008) and in basic activities of daily living (repeated measures ANOVA: F ¼ 4.988; df ¼ 3,360; p ¼ 0.002). This faster decline was independent from having or not the ApoE e4 allele. Increased neuropsychiatric symptom severity in carriers of GSTP1 *C allelic variant was not significantly different between GSTP1 *C carriers and noncarriers. Conclusion: These data, together with our previous study that indicated GSTP1 *C allelic variant as a risk factor for developing LOAD, support and strengthen the hypothesis that oxidative damage is a prominent feature both in the pathogenesis and in the clinical progression of LOAD. O12.5 THE NOVEL PROTEIN DYNAPSIN IS ASSOCIATED WITH A SUBGROUP OF LATE-ONSET ALZHEIMER DISEASE Stephan Zuchner1, John R. Gilbert1, Gaofeng Wang1, Carrie Browning-Large1, Paula G. Bronson1, Chunrong Lu1, Sean M. Garvey1, Joelle M. van der Walt1, Mike A. Slifer1, Donald E. Schmechel1, William K. Scott1, Eden R. Martin1, Jonathan L. Haines2, Margaret A. Pericak-Vance1 1 Duke University Medical Center, Durham 2 Vanderbilt University Medical Center, Nashville, TN Alzheimer disease (AD) is a genetically heterogeneous and complex disorder with the majority of cases characterized by an age at onset > 60 years. We identified a chromosomal AD locus on 2q33.3. Ordered subset analysis (OSA) of 437 families optimized linkage in a subset of 31 AD families with at least one family member with early-onset AD (50–60 years) resulting in a multipoint peak LOD score of 3.79. Converging linkage data with an extensive SAGE analysis of post mortem AD hippocampus identified a candidate gene. Genotyping a dense array of 74 SNPs across the linkage region also targeted this gene with significant evidence for association (minimum p ¼ 0.008). The association strengthened in an extended sample of 73 AD pedigrees meeting the OSA criteria (minimum p ¼ 0.003). This gene, which we have designated dynapsin, has not been characterized previously. We demonstrated as many as eight different isoforms and showed that isoform 7 was higher expressed in a subset of AD brains compared to controls. We designed an antibody and showed that dynapsin immunoprecipitated with components of the dynein-dynactin motor complex. Dynapsin contains a conserved Pleckstrin homology domain
Introduction: Mood-incongruent psychosis (MICP)in Bipolar Disorder may constitute a form of the illness with greater severity, possibly representing phenotypic manifestations of susceptibility genes shared with schizophrenia. We attempt to characterise clinical correlates, familial aggregation and genetic linkage of subjects with MICP. Methods: Subjects were drawn from The NIMH Genetics Initiative BP Collaborative sample, consisting of 708 families. Subjects with MICP and mood-congruent psychosis were compared on clinical variables. Familial aggregation was tested using a proband-predictive model and generalised estimating equations. A genome-wide linkage scan incorporating a MICP covariate was performed. Results: MICP was associated with increased rates of hospitalisation and attempted suicide. Having a proband with MICP predicted MICP in relatives with BPI when compared to all other cases (OR ¼ 2.60, p < 0.0005), and when compared to cases with mood-congruent psychosis (OR ¼ 2.14, p ¼ 0.004). The presence of MICP increased evidence for linkage on 13q21-33 (LOD ¼ 2.99; empirical chromosomewide p ¼ 0.012) and 2p11-q14 (LOD ¼ 3.15; empirical chromosomewide p ¼ 0.018). These LODs and their increase from baseline met genomewide suggestive criteria for significance. Conclusions: Mood-incongruent psychosis showed evidence of more severe course, familial aggregation and suggestive linkage to two regions previously implicated in mental illness susceptibility. The 13q21-33 finding supports prior evidence of bipolar/schizophrenia overlap in this region, while the 2p11-q14 finding is the first to suggest that this schizophrenia linkage region might also harbor a bipolar susceptibility gene. O13.2 INCONGRUENT PSYCHOSIS IN BIPOLAR DISORDER: SUGGESTIVE LINKAGE AT 1q32.3 (LOD ¼ 4.15), 7p13 (LOD ¼ 3.32) AND 20q13.31 (LOD ¼ 2.98) Marian L Hamshere1, Thomas Schulze2, Johannes Schumacher3, Aiden Corvin4, Michael J. Owen5, Peter Propping3, Wolfgang Maier4, Guillermo Orozco Diaz5, Fermin Mayoral5, Fabio Rivas5, Ian Jones5, George Kirov5, Michael Gill4, Peter A Holmans5, Markus M No¨then6, Sven Cichon6, Marcella Rietschel2, Nicholas J. Craddock5 1 Cardiff University, Cardiff 2 Central Institute of Mental Health, Univ, Mannheim 3 Institute of Human Genetics, University, Bonn 4 Departments of Genetics and Psychiatry, Dublin 5 Department of Psychological Medicine, Un, Cardiff
Abstracts 4
Department of Psychiatry, University of Bonn Civil Hospital Carlos Haya, Ma´laga Institute of Human Genetics, Bonn
5 6
Recent evidence from molecular genetic linkage and association studies supports the possibility that lifetime occurrence of mood-incongruent psychotic features may identify a genetically-distinct subset of Bipolar disorder cases. We have undertaken a genome-wide linkage analysis on a large bipolar dataset of 381 affected relative pairs (ARPs), conditioned by the presence of incongruent psychosis symptoms. The sample comprises three original bipolar microsatellite genome screen data sets of Caucasian origin, collected from 1) Andalusia, 2) Germany and Italy and 3) UK and Ireland, each genotyped independently. Individuals were considered affected if they were diagnosed with DSMIV BPI or SABP. Incongruent psychosis was defined as the lifetime occurrence of mood-incongruent psychotic features. Pedigree structures were confirmed statistically. The familiality of mood-incongruent psychosis was assessed with an intraclass correlation coefficient (ICC). Genomewide ARP covariate linkage analysis was performed, conditioned on the presence or absence of incongruent psychosis. Genome-wide significance levels were obtained through simulation. Significant familiality of incongruent psychosis was observed (ICC ¼ 0.31, p < 0.001). Covariate linkage analysis provided three regions with genome-wide suggestive evidence for linkage on chromosomes 1q32.3 close to the DISC1 locus (LOD ¼ 4.15, expected 0.12 times per genome scan), 7p13 (LOD ¼ 3.32) and 20q13.31 (LOD ¼ 2.98). Our results provide molecular support for the hypothesis that genes exist influencing susceptibility to specific subsets of Bipolar cases, dependent on the presence of incongruent psychosis. This has implications for understanding the relationship between mood and psychotic disorders. O13.3 Combining clinical and biochemical phenotypes in mapping susceptibility genes for bipolar disorder Martin Alda1, Gustavo Turecki1, Xiujun Sun2, Jun-Feng Wang2, Iris Jaitovich-Groisman1, Catalina Lopez de Lara1, Paul Grof2, Anne Duffy1, Martin Ruzickova2, Tomas Hajek2, Angus Beck3, Guy Rouleau4, L Trevor Young2 1 McGill University, Montreal 2 University of Toronto, Toronto 2 Dalhousie University, Halifax 3 Community Mental Health, Charlottetown 4 Universite de Montreal, Montreal Background: Previous work of several groups suggests that bipolar disorder (BD) is clinically heterogeneous. Here we report results of a full genome scan of BD responsive to lithium. We also investigated a promising biochemical phenotype in a subset of families. Methods: We conducted full genome scan study of BD with 811 DNA markers. The clinical sample consisted of 36 families with 282 genotyped subjects (135 affected with BD or recurrent depression). All probands were responders to lithium monotherapy; the prevalence of non-affective psychiatric disorders (such as schizophrenia, anxiety, or substance abuse) in these families was at the general-population level. In eleven families (44 subjects) we measured pCREB levels in cultured lymphoblasts and their response to forskolin stimulation. Results: We found suggestive evidence of linkage in 3p, 6p, 14q, and 21q regions, but not in 8p, 13q, 22q, or 18q reported by others in samples of BD patients with prominent psychosis, or with co-morbid panic disorder and/or rapid mood switching. Affected as well as unaffected family members showed elevated basal pCREB levels and blunted response to forskolin compared to healthy controls. Both basal and stimulated levels appear heritable (intra-family correlations of 0.7). Analyzing the biochemical phenotype we obtained moderate evidence of linkage in the same regions as when using the clinical phenotype. Conclusions: Clinically defined subtypes of BD appear to be linked to distinct chromosomal regions. pCREB levels may represent an endophenotype that could provide a useful lead in search for BD susceptibility genes. O13.4 SP4 GENE CONVEYS SUSCEPTIBILITY TO BIPOLAR DISORDER AND SCHIZOPHRENIA John Kelsoe1, Xianjin Zhou1, Shengzhen Guo2, Tatyana Shehktman1, Lin He2, Mark Geyer1 1 University of California, San Diego, La Jolla, CA 2 Bio-X Center, Shanghai Jiao Tong University, Shanghai
713
SP4 is a transcription factor in the SP1/SP4 family that regulates the expression of brain genes and plays an important role in brain embryogenesis. We have examined this gene for its possible role in psychiatric illness. We generated a SP4 hypomorphic mouse by knockin of a LacZ-LoxP construct that reduced SP4 expression to 3–5% of wild type. These hypomorphic mice exhibited vacuolization and astrogliosis in the dentate gyrus of the hippocampus, and decreased expression of neurotrophin-3. Behavioural testing revealed decreased exploratory behaviour in open field tests, diminished prepulse inhibition of startle and decreased fear conditioning, which are behavioural models of psychosis. These behavioural deficits were reversed by rescue of the SP4 gene by breeding with Cre mice. SP4 null mice displayed abnormal neuronal proliferation and synaptogenesis in the hippocampus. Haploinsufficiency of the SP4 gene also caused a prepulse inhibition deficit that was reversed by the administration of carbamazepine. SP4 maps to human chromosome 7 near a region implicated in linkage studies of schizophrenia. We examined 9 SNPs spanning the SP4 gene in 278 bipolar triads of Caucasian origin. One SNP in intron 5 of the gene showed significant association to bipolar disorder (p ¼ 0.0008). The same SNP also showed association to schizophrenia in an independent sample of 305 Chinese triads (p ¼ 0.01). Together these data suggest that variation in the SP4 gene may convey a susceptibility to bipolar disorder and schizophrenia through an effect on brain embryogenesis or gene expression. O13.5 DETAILED ANALYSIS OF THE SEROTONIN TRANSPORTER GENE IN THE NORTHERN SWEDISH POPULATION Maaike Alaerts1, Tine Venken1, Sonia De Zutter1, Lien Heyrman1, Karl-Fredrik Norrback2, Rolf Adolfsson2, Jurgen Del-Favero1 1 University of Antwerp, VIB8, Antwerpen, 2 University of Umea˚, Umea˚ Through active reuptake of serotonin into presynaptic neurons, the serotonin transporter (5HTT) plays an important role in regulating serotonin concentrations in the brain. Therefore it has been hypothesized that this transporter is involved in the etiology of bipolar (BP) disorder. Two functional variations in the SLC6A4 gene encoding 5HTT, a 44bp insertion/deletion polymorphism in the promoter region and a VNTR in the second intron, have been extensively studied for association in different populations, with conflicting results. We genotyped and analysed these variations in a Northern Swedish association sample consisting of 274 BP patients and 364 matched controls, but found no association. We subsequently broadened the study, investigating the LD-pattern in and around the gene, based on information provided by the HapMap project. A set of SNPs that fully characterises the LD-structure of SLC6A4, namely 20 htSNPs in 3 LDblocks and 6 SNPs in the recombination hotspot in the5’ regulatory region, were analysed. No single SNP or haplotype was found to be associated with BP. In addition, we explored the gene on copy number polymorphisms in the 274 Swedish BP patients using a multiplex amplicon quantification technique, but no deletions or duplications were detected. We also performed mutation analysis on 5microRNA target sites in the 3’ UTR of SLC6A4 in this patient population, but no variants were found. We can conclude that variation in the SLC6A4 gene or its regulatory regions does not contribute to susceptibility for BP in the Northern Swedish population. O13.6 NO ALLELIC ASSOCIATION OR INTERACTION OF 3 KNOWN FUNCTIONAL POLYMORPHISMS WITH BP DISORDER Ann Van Den Bogaert1, Kristel Sleegers1, Sonia De Zutter1, Lien Heyrman1, Karl-Fredrik Norrback2, Rolf Adolfsson2, Jurgen Del-Favero1 1 University of Antwerp, VIB8, Antwerpen, 2 University of Umea, Department of Clinical Sciences, Umea Most genetic association studies on bipolar (BP) disorder have focused on genes involved in major neurotransmitter systems or brain development. Functional polymorphisms in the serotonin transporter (5-HTTLPR), catechol-O-methyltransferase (Val158Met) and dopamine D3 receptor (Ser9Gly) genes have all been associated with BP disorder. We attempted to investigate whether these functional variants contribute to the genetic etiology of BP disorder in a northern Swedish isolated population. Moreover, we wanted to collect informa-
714
Abstracts
tion about the synergistic contribution of these functional variants. For this purpose, we genotyped the functional polymorphisms in a sample of 182 BPI patients and 364 healthy individuals matched for gender and age. On the whole, our data suggest that these functional variants are not involved in the etiology of BP disorder in the northern Swedish population, nor did gene-gene interaction analysis support a central role of these variants in BP disorder. O13.7 NEUROCOGNITION IN BIPOLAR DISORDER: FROM PHENOTYPE TO GENOTYPE Katherine Burdick1, Birgit Funke2, Nisali Gunawardane1, Todd Lencz1, Philip Szezsko1, Pamela DeRosse1, Raju Kucherlapati2, Anil Malhotra1 1 The Zucker Hillside Hospital, Glen Oaks, NY 2 Harvard Partners, Boston, MA There is a paucity of data on the genetic underpinnings of neurocognitive dysfunction in bipolar disorder (BPD). For example, few studies have assessed cognition in euthymic BPD patients, there are limited data on the co-segregation of neurocognitive deficits in unaffected siblings of patients with BPD, and there are limited genetic association studies utilizing neurocognitive measures in BPD. We have examined a cohort of euthymic BPD patients, their unaffected siblings (past the modal age of onset and free of Axis I diagnoses) and matched healthy controls on a battery of neurocognitive measures. We have also conducted a candidate gene study of 4 SNPs within the catechol-omethyl-transferase (COMT) gene, including the Val158Met polymorphism previously linked with cognition in non-BPD subjects, in a sample of BPD subjects drawn from the Zucker Hillside Hospital DNA data base (1339 probands and 752 healthy controls) in order to investigate the relationship between COMT genetic variation and neurocognitive function in BPD. Our data indicate that: 1) there are significant and widespread neurocognitive impairments in euthymic BPD patients; 2) unaffected siblings of BPD patients also demonstrate impairments on aspects of neurocognition and, 3) prefrontal aspects of verbal memory in patients with BPD may be related to COMT genotype. Although preliminary, these data suggest that neurocognitive impairment may represent a promising endophenotype for future molecular genetic studies of BPD. Further, these data may lead to a better understanding of the molecular pathophysiology of BPD and, perhaps, could suggest new treatment targets in BPD. O14.1 MOLECULAR ANALYSIS OF 8 AUTISM PATIENTS WITH CYTOGENETIC ABNORMALITIES John Vincent1, Christian Marshall2, Abdul Noor1, Christian Windpassinger1, Shin-ichi Horike2, Sanaa Choufani2, Beata Stachowiak1, Jennifer Skaug2, Leon Sloman1, Peter Kroisel2, Erwin Petek2, Wendy Roberts2, Stephen Scherer2 1 Centre for Addiction and Mental Health, Toronto 2 Hospital for Sick Children, Toronto 2 Institute of Human Genetics, Graz In our search for genes that may cause autism, we have analysed cytogenetic breakpoints in a number of patients. These include 4 patients with breakpoints mapping to the 7q31 autism linkage region, as well as 3 patients with translocations t(19;21)(p13.3;q22.1), t(6;14)(q13;q21), t(5;18)(q32;q21), and a patient with an inversion on 4p. Some of the identified breakpoints directly interrupt genes-coding genes, which include RAY1/ST7, GABRG1, dystonin, SH3TC2. Others interrupt non-coding genes, such as CB338058, or may interfere with adjacent genes by position effect. Such genes include WNT2, NPTX2, KCND2, GABRA2, GABRA4 and GABRB1, ADRB2, HTR4. We will discuss our strategies to determine the role (if any) of these genes in autism. We have also performed 500K Affymetrix SNP microarray analysis on these individuals to determine whether other cryptic deletions or duplications may be present. This analysis has revealed a number of interesting candidate regions, including 1p21.3, 3p13-p14.1 and 11q25. These findings will be discussed. O14.2 ASSOCIATION OF A GLUTAMATE TRANSPORTER GENE (SLC1A1) WITH THALAMIC VOLUME IN PEDIATRIC OCD Paul Daniel Arnold Centre for Addiction and Mental Health, Toronto
An association between the glutamate transporter gene SLC1A1 and obsessive-compulsive disorder (OCD) has recently been identified in two independent populations (Arnold et al., Dickel et al., in press 2006). The purpose of this study was to determine if SLC1A1 variants were associated with volumetric differences in brain regions previously implicated in OCD. We studied 25 medication-naı¨ve children and adolescents with OCD, who had been scanned using magnetic resonance imaging (MRI). Genotyping was performed on three polymorphisms (rs301434, rs3087879, rs3056) within a haplotype block of SLC1A1 previously demonstrated to be associated with OCD. For each variant, genotype groups were examined using ANCOVA for differences in total volumes of the orbitofrontal cortex, anterior cingulate, caudate, putamen, globus pallidus and thalamus, after controlling for age and total intracranial volume. The rs3056-GG genotype was significantly associated with increased total thalamic volume (p ¼ .02). This finding was consistent with a previous family-based study conducted by our group in which a haplotype containing the G allele was over-transmitted to OCD probands. These results, which were not corrected for multiple comparisons, provide preliminary evidence that SLC1A1 sequence variation may be associated with increased total thalamic volume. Previous neuroimaging studies conducted on this population have demonstrated increased thalamic volume in pediatric patients with OCD (Gilbert et al., 2000). Taken together, the genetic and neuroimaging data therefore suggest that thalamic volume warrants further study as a potential endophenotype of OCD. Replication in larger samples is needed to confirm these findings. O14.3 PARENTAL CONTRIBUTIONS TO ATTENTION DEFICIT HYPERACTIVITY DISORDER IN CHILDREN: QUALITATIVE AND QUANTITATIVE DIFFERENCES Lisa Goos1, Payam Ezzatian2, Russell Schachar1 1 Hospital for Sick Children, Toronto 2 University of Toronto, Toronto, ON The goal of the present study was to investigate parent-of-origin effects in Attention Deficit Hyperactivity Disorder (ADHD). There are two main sources of parent-of-origin effects in ADHD: differences in the relative quantity of risk factors transmitted by each parent, or qualitative differences in the risks transmitted, such as those carried on the sex chromosomes or regulated by genomic imprinting. 60 children with maternal-only history of ADHD and 131 children with a paternal-only history were compared on three domains for which prior evidence suggested parent-of-origin effects may exist: core symptoms, disruptive behaviour and depression. Dependent variables were derived from previously validated, age-appropriate and standardised parent and teacher interviews and questionnaires. Depression levels were rated using the Child Depression Inventory. Consistent with previous research and the predictions derived from threshold models of ADHD etiology, the maternal history group received higher ratings of behavioural disorder (ADHD, conduct disorders and oppositional symptoms) than the paternal history group from both parents and teachers. Parent-of-origin effects were also observed for depression, with the paternal history group rating themselves significantly more depressed than children in the maternal history group, particularly girls. Heightened paternal transmission relative to maternal is suggestive of genomic imprinting, and the interaction with proband sex indicates the involvement of the sex chromosomes or sexspecific physiological or hormonal factors. Interpretations of these data in terms of environmental and genetic factors, including epigenetic and sex-linked hypotheses, are explored. O14.4 PRESCHOOL LANGUAGE IMPAIRMENT AND READING OUTCOMES Emma Hayiou-Thomas1, Robert Plomin2 1 University of York, York 2 Institute of Psychiatry, London The current study examines the association between preschool speech and language impairment and later reading skills. At 4½ years of age, a subsample of children from the Twins Early Development Study (TEDS) was administered an extensive battery of language measures. Early reading skills (primarily decoding) were assessed at 7 and reading comprehension skills at 10 years of age. The phenotypic association between reading and earlier impairments in both expressive phonology and non-phonological language was moderate: impaired
Abstracts children scored at least half a standard deviation below their peers on reading measures at 7 and 10. Bivariate De-Fries-Fulker extremes analyses were used to partition phenotypic associations into genetic and environmental components. Both genetic and shared environmental factors appear to mediate the relationship between speech and language impairments at 4 1/2 and reading skills at 7: 18% of the phenotypic association between early language impairment and reading was attributable to genetic factors, while 74% was due to shared environmental factors. For early speech problems, 44% of the association was genetic, and 50% shared environmental. A different pattern emerged for reading comprehension skills at 10 years of age: genetic factors accounted for 74% of the phenotypic association between early language impairment and reading comprehension, and for 100% of the link between early speech impairment and reading comprehension. O14.5 SUBTYPES VS. SEVERITY DIFFERENCES IN ADHD IN THE 1985–86 NORTHERN FINNISH BIRTH COHORT Gitta Lubke1, Bengt Muthen2, Irma Moilanen3, Sandra Loo2, James Swanson3, May Yang2, Marjo-Ritta Jarvelin3, Susan Smalley2 1 University of Notre Dame, Notre Dame Indiana 2 UCLA, Los Angeles 3 University of Oulu, Oulu 3 UC Irvine, Irvine Objective: To investigate whether inattention, hyperactivity, and impulsivity behaviour in the NFBC support an interpretation of qualitatively distinct subgroups in ADHD or variants along a continuum of severity. Method: Latent class models, exploratory factor models and factor mixture models are applied to questionnaire data of ADHD behaviour obtained from the Northern Finland Birth Cohort of adolescents. The three types of statistical models correspond to different sets of assumptions regarding the structure of the phenotype of ADHD. Latent class analysis corresponds to an interpretation in terms of qualitatively distinct classes such as unaffected, affected and subtypes; factor analysis corresponds to differences in severity, and factor mixture analysis represents a hybrid model that allows for both distinct subtypes and severity differences within subtypes. Results: A comparison of the fit of the different models to the data shows that models which distinguish between a low scoring majority class (‘unaffecteds’) and a high scoring minority class (‘affecteds’), and which, in addition, allow for two factors (inattentive, hyperactiveimpulsive) with severity differences provide the best fit. Conclusions: Our analysis provides evidence that a high scoring minority group (8.8% of males and 6.8% of females)probably reflects an ‘ADHD group’ in the Northern Finland Birth Cohort, while the majority of the population falls into a low scoring group of unaffecteds. Distinct factors consisting of items of inattention and hyperactivity-impulsivity are evident for both genders in the population, with considerable variability in severity within each class. O14.6 DIFFERENTIAL FAMILY AND PEER ENVIRONMENTAL FACTORS IN ADHD: A DISCORDANT SIBLING PAIR STUDY Cathelijne Buschgens1, Marcel A.G. Van Aken2, Sophie H.N. Swinkels1, Jan K. Buitelaar1 1 Radboud University Nijmegen Medical Cent, Nijmegen 2 Utrecht University, Utrecht Most studies that have explored differences in familial and psychosocial factors between children who have ADHD and those who do not, generally used between-group comparisons. However, earlier behavioural genetic studies imply that salient environmental influences operate within families, making siblings in a family different rather than similar. This study is the first one looking at the significance of differential environmental sibling experiences within ADHD families. Participants were 45 Dutch ADHD probands and their unaffected sibling aged 10–18 years who participated in IMAGE, a large international collaborative clinical study. Paired t-tests were used to compare ADHD children with their unaffected biological sibling on four domains of non-shared environmental influences as measured by the SIDE. Correlations and regression analyses were used to determine association with and the relative contribution of non-shared environmental influences and differences in child psychopathology. The results indicate that ADHD probands and their unaffected siblings report
715
differences in the domains of sibling interaction, parental treatment, peer group characteristics and events specific to the child. In addition, perceived differences are related to differences in the level of the child’s emotional and behavioural adjustment. Both differential sibling interaction and differential peer characteristics contribute considerably to adjustment differences between the siblings. Environmental influences operating within ADHD families are important for understanding the etiology and development of ADHD children and their siblings. Further research is necessary to determine whether our findings hold for larger samples consisting of twins and/or concordant sibling pairs.
O14.7 GENETIC VULNERABILITY TO PSYCHOPATHIC TENDENCIES IN CHILDREN Essi Viding University College London, London Recent evidence suggests that psychopathic personality traits are heritable in both adults and children. In addition to demonstrating simple heritability of a trait, genetically informative study designs are particularly useful for investigating aetiological heterogeneity and can refine the search for developmental pathways to persistent antisocial conduct. This paper will outline results from the Twins Early Development Study (TEDS), indicating that antisocial behaviour in the presence of psychopathic traits is more heritable than antisocial behaviour in the absence of such traits. This finding is present at both seven and nine years of age. In addition, early risk factors associated with antisocial behaviour show a different pattern in children with/ without psychopathic traits. For example, children with psychopathic tendencies show higher levels of conduct problems throughout early childhood and stronger genetic mediation of the longitudinal relationship between early and late conduct problems. These findings suggest that antisocial children with psychopathic traits form an aetiologically distinct subgroup and that molecular genetic research may be particularly fruitful in this group. Molecular genetic investigations will be particularly exciting when combined with neuroimaging approaches, which we are currently undertaking.
O15.1 THE ANDROGEN RECEPTOR IS A QUANTITATIVE TRAIT LOCUS FOR MEMORY FUNCTION IN MALES David A. Collier1, Denes Kovacs2, Xiehe Liu3, Xueli Sun3, Junmei Hu3, Gerome Breen1, Peter Tompa2, Tao Li3 1 Institute of Psychiatry, London 2 Institute of Enzymology, Budapest 3 Sichuan University, Chengdu 3 Institute of Psychiatry/Sichuan University, London The androgen receptor (AR) is a soluble intracellular DNA binding transcription factor. Androgens initiate conformational changes that affect receptor-protein and -DNA interactions (Gao et al., 2005). A polyglutamine tract in the AR gene, which maps to the X-chromosome, is highly polymorphic, and its expansion causes single gene disorders such as SMBA. The AR may also influence memory, particularly in males (Janowsky, 2005). ARs are prominent in the hippocampus, amygdala and prefrontal cortex, and androgen deprivation causes significant loss of synapses in the hippocampus (Leranth et al., 2003; 2004), affects hippocampal dependent learning and memory (Frye et al., 2001; Kritzer et al., 2001) and affects neurotransmission in the prefrontal cortex. In humans, this may result in cognitive impairment, particularly of memory function, in older males. As part of a programme examining memory function as an endophenotype of psychosis, we examined whether the polyglutamine tract in AR is associated with cognitive function in a healthy, non-elderly population of 400 subjects from Sichuan Province, SW China. We used fourteen neuropsychological tests in five neurocognitive domains, Attention and Speed of Information Processing, Memory and Learning, Verbal Function, Visuoconstructive Abilities and Executive Function. The AR repeat was not associated with cognitive function in females but was strongly associated in males, particularly for delayed and immediate visual memory. We conclude that the AR gene may be a quantitative trait locus for memory function in the general population, and we are presently evaluating its effect on memory function in males with schizophrenia.
716
Abstracts
O15.2 NORMAL TRAIT VARIATION IN ADULT VISUAL ATTENTION: MAPPING OF NOVEL CANDIDATE LOCI IN HUMANS Stefan Pulst1, Hilary Coon2, Michael Goldsmith3, Andy Peiffer2, Mark Leppert2 1 Cedars-Sinai Medical Center, Los Angeles, CA 2 University of Utah, Salt Lake City, UT 3 Cedars-Sinai Medical Center, Los Angeles, CA Attention is the cognitive process of selectively focusing on one stimulus while excluding others. Although dysfunctional attention is a key symptom of Attention Deficit/Hyperactivity Disorder (ADHD), attentional mechanisms are altered in a variety of psychiatric diseases such as schizophrenia and mood disorders. Prior studies of attention in children using continuous performance tests (CPTs) have usually revealed lack of heritable components, despite the fact that the clinical ADHD phenotype has a significant genetic component. We have examined a large cohort of Utah adults for attention traits consisting of 594 individuals in 41 three-generation CEPH pedigrees. To determine quantitative attention traits, we phenotyped this adult cohort using a CPT of the X- type and measured reaction time, errors of omission and commission and a composite ADHD score. Formal genetic analysis was performed using the Pedigree Analysis Package (PAP) and SOLAR. We determined significant heritabilities for individual attention traits including reaction time and the composite ADHD score. A genome scan with 537 microsatellite markers identified a multipoint lod score on chromosome 9q (empirical p ¼ 0.0006) for the ADHD score and additional linkage signals for reaction time. To our knowledge, this is the first study showing evidence for significant heritabilities for reaction time and other attention traits in normal adults. Focus on normal adults may avoid the problem of coexisting comorbid conditions and may uncover novel linkage signals contributing to normal and potentially pathological attentional mechanisms. O15.3 DIMENSIONALITY AND HERITABILITY OF OBSESSIVE-COMPULSIVE SYMPTOMS Carol Mathews1, Denise Chavira2, Amin Azzam3, Monica Bagnarello4, Helena Garrido4, Victor Reus1 1 University of California, San Francisco, San Francisco, CA 2 University of California, San Diego, CA 3 University of California, Berkeley, Berkeley, CA 4 National Children’s Hospital, Costa Rica, Costa Rica The division of obsessive-compulsive symptoms (OCS) into specific factors is now widely accepted. However, the utility of these categories for genetic studies remains unclear, as studies examining their heritability have been contradictory. Less attention has been paid to the possibility that clinically significant obsessionality is primarily determined by a small ‘‘core’’ group of OCS that crosses the boundaries between the symptom subgroups. The aim of this study is to determine whether such a core group exists, and to compare its heritability to that of the more traditionally-derived symptom factors. Methods: We examined the properties and heritability of obsessional symptoms in college students, medical students and OCD families. Results: In each of the three samples, we identified a core group of symptoms that comprised a single unique factor and accounted for over 90% of the variation of the four more traditional symptom factors. This core factor was highly correlated with OCD, YBOCS total score and hoarding in our OCD families and had a heritability estimate of.19 when OCD was not included as a covariate and.49 when OCD was included as a covariate. In contrast, the four symptom factors were not found to be heritable. Conclusions: There appears to be an underlying unidimensional component to obsessionality, both in non-clinical and clinical samples. This component, which is heritable, accounts for the majority of the variation in the more traditionally- derived symptom factors, and is composed of OCS that are not specific to any of the symptom subgroups. O15.4 ALTERED CORTICAL THICKNESS IN ADOLESCENTS AND YOUNG ADULTS AT GENETIC RISK FOR SCHIZOPHRENIA Larry Seidman Harvard Medical School, Boston, Massachusetts Introduction: First-degree relatives of persons with schizophrenia carry elevated genetic risk for the illness. An important question is
whether genetic risk for the illness is expressed in altered brain volumes prior to onset of psychosis. We assessed this by using a novel approach to measuring cortical thickness. Methods: Participants were 27 non-psychotic, un-medicated firstdegree relatives of persons with a DSM-IV diagnosis of schizophrenia or schizoaffective disorder, depressed type, and 48 un-medicated controls, recruited from the community and hospitals in metropolitan Boston (ages 13–28). MRI data was acquired on a Siemens 1.5T MR scanner.Data were analysed using morphometric approaches combining the Cardviews and Freesurfer tools. Differences were tested with two-tailed tests. Results: The high-risk participants have significantly thinner cortices than controls in a number of regions typically altered in schizophrenia, even though they are not psychotic.In the right hemisphere, differences are noted in the frontopolar and dorsolateral prefrontal cortex (DLPFC), in the orbitofrontal cortex (OFC), the insula, mainly in its anterior portion, anterior and posterior cingulate, anterior and posterior parahippocampus, temporal pole, temporal occipital, fusiform, precuneal and occipital areas. Similarly, in the left hemisphere, differences are localised in the frontopolar cortex and DLPFC, OFC, anterior insula, anterior and posterior cingulate, paracingulate, posterior parahippocampus and occipital areas. Discussion: These data agree with volumetric findings observed in these cortical areas in schizophrenia. These promising data indicate altered cortical thickness in those at risk for schizophrenia, independent of psychosis.
O15.5 DOES SEXUAL SELECTION LINK HEIGHT AND INTELLIGENCE? Rosalind Arden King’s College London, London Females prefer mates who are taller than themselves. Females also prefer mates who are intelligent. Intelligence and height are reported to be phenotypically associated and, in adult males, to share some genetic covariance. In this study we aim to find out whether the association between height and intelligence varies between the sexes and whether genetic covariance between the traits manifests in childhood or in adolescence. In a community sample of British sevenyear old twins (n > 3000) and in a sample of American adolescent twins (n ¼ 1232, mean age 15), we analysed the phenotypic and genetic correlations between the two traits in the two sexes. We found a small phenotypic association between height and intelligence in the sevenyear olds, but no genetic correlation between the traits in either sex. In the adolescent samples, we found a phenotypic association and a genetic correlation between the traits in males, but in female adolescents the two traits were phenotypically and genetically uncorrelated. These results hint that traces of the influence of sexual selection might be visible in the covariance structure between traits. Further, they demonstrate a way in which behavioural genetics’ methods can investigate questions motivated by evolutionary theory.
O15.6 THE DEVELOPMENTAL PATHWAY OF AN ENDOPHENOTYPE: COGNITIVE TIMING IN ADHD Nanda Rommelse Clinic of Neuropsychology, Free University of Amsterdam The subjective sense of time seems to be affected in children with Attention-Deficit/Hyperactivity Disorder (ADHD) which results, for example, in impatience, lack of planning and underestimation of passed time. Whether this deficit is partly genetically based as is an endophenotype, and therefore can be measured in non-affected siblings of an affected individual, is not yet clear. A second point of interest is whether this deficit is stable during the development or whether children with ADHD seem to out grow this deficit, as would be predicted by the delayed maturation hypothesis. To investigate this, 226 children with ADHD, 188 non-affected siblings and 162 control children between the ages of 5 and 19, were subjected to a time reproduction task in which they had to reproduce time intervals of 4, 8, 12, 16 and 20 seconds. Analyzed were the precision of timing and the tendency to under or over reproduce time durations. Overall, children with ADHD were more imprecise, as were their non-affected siblings, in comparison
Abstracts to controls. Furthermore, the tendency to under reproduce longer durations was more strongly pronounced in ADHD children and slightly more in non-affected siblings. Both findings are indicative of an endophenotype. Most interestingly, was the developmental pathway of these deficits: In ADHD children and non-affected siblings, timing capabilities improved more strongly with age than in controls. Yet, as non-affected siblings seemed to overcome their deficits completely by late adolescence, ADHD children did not.
O15.7 BRIDGING THE GAP BETWEEN GENOTYPE AND PHENOTYPE: ENDOPHENOTYPES IN TWIN STUDIES Kristen C. Jacobson, PhD Department of Psychiatry, University of Chicago With recent advances in molecular genetics and genomics, as well as increased attention to the interactive effects of genes and environments, one might begin to question whether traditional behavioural genetic research has anything more to contribute to our understanding of individual differences in psychiatric traits and behaviours. It is argued that traditional behavioural genetic methodologies of twin, sibling, and adoption studies still have a unique place in research. Specifically, while one of the goals of psychiatric genetic research is to uncover specific genetic variants that are risk factors for psychiatric disorders, it is unlikely that genes code for the disorders and behaviours directly, but rather it is more likely that genes influence disease and behaviour through indirect mechanisms or pathways. To gain a better understanding of these potential genetic mechanisms, twin studies can be used to identify biologically-based traits that may serve as endophenotypes for psychiatric disease. This paper provides empirical evidence from multiple studies of both children and adults that twin studies can be successfully used to identify potential endophenotypes for antisocial and aggressive behaviour. Candidate endophenotypes include characteristics that tap physiological arousal, impulsivity, and socio-emotional processing. Moreover, it is demonstrated that innovative use of multivariate twin models can provide important information about the underlying structure of aggressive and antisocial behaviour. These results suggest that using quantitative, normally distributed endophenotypes as outcome variables in molecular genetic studies may be a worthwhile strategy, and may offer greater power than the use of traditionally-defined psychiatric disorders.
O16.1 VARIANTS IN THE ESTROGEN RECEPTOR ALPHA GENE ITS MRNA ISOFORMS CONTRIBUTE TO SCHIZOPHRENIA Cynthia Weickert, Ana Miranda-Angulo, William Perlman, Sarah Ward, Joel Kleinman, Vakkalanca Radhakrishna, Richard Straub, Daniel Weinberger NIMH, Bethesda, MD Estrogen modifies mood, cognition and psychiatric symptoms; therefore, variation in the ability of the brain to respond to estrogen may impart risk to developing mental illness. We asked if variation in the estrogen receptor alpha (ESR1) gene ormRNA is associated with the diagnosis of schizophrenia. By re-sequencing functional regions of the ESR1 gene, we discovered a novel insertion/deletion (indel) promoter polymorphism. In a case-control genetic association analysis, three distinct haplotypes surrounding this indel were significantly more frequent in schizophrenics than in normal controls (p ¼ 0.001 to 0.009, in African Americans). In a follow-up family-based association analysis, we found that a SNPs in intron 1 (PvuII), partof the ‘‘at-risk haplotype’’, was significantly associated with schizophrenia (p ¼ 0.03 in African American and Caucasians). Furthermore, PvuII genotype significantly influenced brain ESR1 mRNA levels. Next, we asked if the type of ESR1 mRNA transcripts differed in the brains of patients with schizophrenia compared to normals and patients with bipolar illness and depression. We detected decreased frequencies of the wild type and ¨ 7 ESR1 in schizophrenic and depressed increased frequencies of A individuals, respectively, compared to controls. A novel ESR1 transcript coding for a truncated ESR1 protein was detected in one patient with schizophrenia. A SNP in intron 6 was associated with presence of ¨ 7 ESR1 mRNA but this SNP not associated with risk for schizophrenia A
717
or with ESR1 mRNA levels. We provide multiple lines of evidence that variation in ESR1 is associated with schizophrenia. O16.2 ASSOCIATION EVIDENCE FROM NUDEL AND PDE4D SUPPORT THE DISC1-PATHWAY CONCEPT IN THE ETIOLOGY OF SCHIZOPHRENIA Liisa Tomppo1, William Hennah1, Pa¨ivi Lahermo2, Anu Loukola1, Jesper Ekelund1, Timo Partonen1, Aarno Palotie2, Jouko Lo¨nnqvist1, Leena Peltonen1 1 National Public Health Institute, Helsinki 2 Finnish Genome Center, University of Helsinki Disrupted in Schizophrenia 1 (DISC1) is currently one of the best candidate genes for severe mental illness, and has been further associated in reduced grey matter volume, deficiencies in working memory, cognition and the P300 event related potential. We have reported a specific DISC1 haplotype, named HEP3, showing association to schizophrenia in Finnish families. Through the employment of this finding in a genome-wide analysis, by ascertaining the families for the DISC1 risk allele, we identified another locus 16p13. The gene of a DISC1 binding partner, NDE1, is located in this region and found to display association to schizophrenia with a haplotype spanning the gene. Since both DISC1 and NDE1 display association to schizophrenia in the Finnish study sample, we hypothesized that other DISC1 interactors may also have an impact on the etiology of schizophrenia. We selected 10 additional genes encoding the molecules of the DISC pathways and monitored for their association to schizophrenia related traits. We initially genotyped the SNPs in 171 families and followed any suggestive finding in the second independent set of 303 families. In the stage one sample, SNPs in both PDE4D and NUDEL displayed p-values of 0.007, and a SNP in PDE4B displayed a p-value of 0.03. These were replicated in the stage two sample (p-values of 0.03, 0.001 and 0.04, respectively). These and other converging lines of evidence further underline the importance of DISC1-related molecular pathways in the etiology of schizophrenia. O16.3 SNAP-25 IS ASSOCIATED WITH SCHIZOPHRENIA IN 270 IRISH HIGH DENSITY FAMILIES, FOLLOWING GENOME-WIDE SIGNIFICANT LINKAGE TO CHROMOSOME 20p12.3-q13.12 USING LATENT CLASSES OF PSYCHOTIC ILLNESS Ayman Fanous1, Edwin van den Oord2, Zhongming Zhao2, Brandon Wormley2, Richard Amdur1, F. Anthony O’Neill2, Dermot Walsh3, Kenneth Kendler2, Brien Riley2 1 Washington VA Medical Center, Washington, D.C. 2 Virginia Commonwealth University, Richmond, VA 2 Queen’s University, Belfast 3 Health Research Board, Dublin Background: We have previously reported genome-wide significant linkage to chromosome 20p12.3-q13.12 in a genome scan of latent classes of psychotic illness that were based on psychopathology (1). This region had previously resulted in little evidence of linkage using operationalized diagnostic criteria in our sample, but has been linked in other samples. SNAP-25 (20p12.2) is a component of the SNARE complex. Its expression in brain has been reported to be decreased in schizophrenia. Furthermore, its expression may be induced by overexpression of dysbindin. Methods: The Irish Study of High Density Families comprises 270 families and 1408 individuals, among whom 722 were genotyped and had at least one psychotic episode. SNAP-25 was selected for association testing based on its location, as well as its function in neurotransmission. We genotyped eighteen intronic haplotype-tagging SNPs in the gene, which were obtained from HapMap. Family-based association tests were performed in FBAT. Results: Six of 18 SNPs were associated with at least one diagnostic category (P < .01), from narrow to very broad. We estimated the risk of these signals being false positive at merely 10%. Several 2marker haplotypes were over-transmitted. Conclusions: This is the first report of association between SNAP-25 and schizophrenia to our knowledge. Replication in other samples is warranted. Clinically homogeneous phenotypes could yield greater power to detect linkage and identify susceptibility genes. References: (1) Fanous et al., Genome Scan of Latent Classes of Psychotic Illness, ISPG Abstracts, 2005.
718
Abstracts
O16.4 GENE IN COMMON TO SCHIZOPHRENIA AND BIPOLAR DISORDER MAPPED TO 17q21.31-q22 Hreinn Stefansson1, Thorgeir Thorgeirsson1, Einar Gudfinnson1, Valgerdur Steinthorsdottir1, Andres Ingason1, Augustine Kong1, Jeffrey Gulcher1, Thordur Sigmundsson2, Engilbert Sigurdsson2, Hannes Petursson2, Kari Stefansson1 1 deCODE Genetics, Reykjavı´k 2 University Hospital, Reykjavı´k Schizophrenia and bipolar disorder are complex disabling disorders that continue to exact a heavy toll not only upon those afflicted but also on relatives and on our health and welfare systems. Family history of either disorder significantly increases the risk of developing these disorders. The two disorders may share important etiological factors while some aspects of their pathogenesis may differ. Most probably they represent a continuum of several complex biological dimensions that contribute to the phenotype to a variable degree in every subject. Our linkage analysis on integrated schizophrenia and bipolar families, 648 subjects in 239 families, gave a strong linkage signal (LOD ¼ 4.1) on 17q21.31-q22. In a recent study linkage analysis of a completely ascertained sample of familial schizophrenia and bipolar disorder from Palau, Micronesia showed linkage to 17q32.2 (LOD ¼ 2.8). Furthermore, in a large sib pair analysis on schizophrenia samples from UK, Williams and coworkers mapped a schizophrenia gene to 17p11.2-q25.1 (LOD 3.35) and meta-analysis by Lewis and colleagues of both disorders showed evidence for linkage at 17q21.33 – 24.33. There is evidence to show that psychiatric disorders like schizophrenia and bipolar disorder can be caused by large scale rearrangements. Chromosome 17 is among those chromosomes known to represent several segmental duplications and inversion polymorphisms. It therefore presents a special challenge for positional cloning. Rare, large scale rearrangements such as inversions, deletions and duplications may account for the strong linkage signal at the 17q locus. O16.5 COMPARISON OF RESULTS FOR WHOLE GENOME ASSOCIATION ANALYSIS OF SCHIZOPHRENIA USING POOLED DNA SAMPLES AND INDIVIDUALLY GENOTYPED SAMPLES John Verner Raelson, Pascal Croteau, Valeri Perepetchai, Paul Van Eerdewegh, Rene Allard, Helene Fournier, Nathalie Laplante, Micheline Lapalme, Quynh Nguyen-Huu, Nouzha Paquin, Bruno Paquin, Jonathan Segal, Jean-Michel Vidal, Tim Keith, Majid Belouchi Genizon Biosciences, Ville St-Laurent, Quebec We performed whole genome association studies on both pooled DNA and individually genotyped schizophrenia samples from the Quebec founder population. The pooling study used 469 cases and 469 controls divided into 8 case pools and 8 control pools consisting of between 51 and 52 individuals each. Pools were genotyped for approximately 200,000 SNP markers by Perlegen Sciences Inc. Differences in mean allele frequency between cases and controls were determined and exact P values for single markers and for Fisher combined P values for multimarker windows were determined by a permutation of pool status. The maximum significance possible, due to the limited number of combinations of case and controls pools was – log P ¼ 4.4. All markers or multimarker windows significant at—log P > 3.5 were individually genotyped and retested using permutation of individual case-control status. Thirty-eight percent of these remained at—log P > 3.5. The genomewide scan was repeated using individual genotyping for 510 patients and 1235 controls. Individuals used in the pooling study form a subset of individuals used in the individual genotyping study. Each individual was genotyped for a panel of approximately 370,000 markers using the Illumina Infinium II platform. All significantly associated regions are being further fine mapped. Here we examine similarities and differences in results produced by the pooling and individual genotyping studies. These results are part of our program for gene discovery in complex diseases which includes ADHD and Alzheimer’s disease. O16.6 IN SILICO MAPPING OF MOUSE PREPULSE INHIBITION SUGGESTS TSN IS A SCHIZOPHRENIA CANDIDATE GENE Bradley Todd Webb, Joseph L. McClay, Edwin J. van den Oord VIPBG, Virginia Commonwealth University, Richmond, VA
We conducted an in silico genome scan using common inbred mouse strains for prepulse inhibition (PPI), a possible endophenotype for schizophrenia.Multiple highly significant associations (p-value 8.1e07, genome wide FDR < 0.045) were found in a 3 to 6 megabase region on mouse chromosome 1 that is homologous to a region of human chromosome 2q14 implicated by meta-analysis of schizophrenia genome scans. The associated SNPs tagged a single underlying haplotype which included the gene Translin (Tsn), known to alter PPI when knocked out in mice. Microarray data from a subset of inbred lines used to map the PPI association was available. The relationship between Tsn expression levels and associated SNPs was examined. Expression levels from one of the four probe sets for Tsn were highly correlated (p-value 1.67e-05) with SNPs associated with PPI. The 3 probe sets not correlated tagged different sets of transcripts with alternative 30 UTRs. Using WebQTL and data from the BXD RIlines, we were able to confirm the presence of a cis eQTL for Tsn. The relationship between schizophrenia and PPI makes Tsn a good candidate for schizophrenia association studies. In addition, Tsn interacts functionally with translin-associated factor X (TSNAX or TRAX) and multiple associations with schizophrenia and haplotypes covering both DISC1 and TRAX have been reported. We are now working to confirm the relationship between Tsn expression and the PPI haplotype in mice and genotyping SNPs in this gene in human samples. O16.7 ASSOCIATION WITH SCHIZOPHRENIA IN THREE CANDIDATE GENES: GRM-3, AKT-1 AND NRG-1 Sibylle Schwab1, Christie Plummer1, Adam Doyle1, Margot Albus2, Margitta Borrmann-Hassenbach2, Wolfgang Maier3, Dieter Wildenauer1 1 University of Western Australia, Nedlands 2 Bezirkskrankenhaus Haar, Haar 3 University of Bonn, Bonn Background Association of DNA variants with schizophrenia has been reported for the metabotropic glutamate receptor 3 (GRM3) the protein kinase B (AKT-1) and neuregulin 1 (NRG1). Whereas NRG1 is located in a chromosomal region with good support for linkage with schizophrenia, there has been only weak support for linkage to the chromosomal locus of GRM3 with schizophrenia and no support for linkage to the chromosomal region of the AKT1 locus. Methods We genotyped SNPs for all three regions based on previous reports and attempted replication of association with schizophrenia using a family based design. Statistical analysis was performed using the program Famhap. Results AKT1: A P value of 0.0005 was obtained for rs4983559. This SNP is located approximately 17kb upstream of the AKT1 gene and revealed the lowest P value obtained for 9 SNPs analysed in the region. GRM3: Five SNPs have been analysed. SNP rs2237562 was nominally associated with schizophrenia (P ¼ 0.01).NRG1: We genotyped 13 SNPs located in the exon dense 3’ region of the gene, and one SNP and two microsatellite markers previously implicated in susceptibility to schizophrenia by Stefansson et al (AJHG, 2002, 72:83). Statistical analysis revealed association of schizophrenia with rs10503929 (p ¼ .017). Haplotype analysis revealed lowest P value of P ¼ .000004 for a five-marker haplotype and a global P value of P ¼ 0.00035. Conclusions: We conclude that DNA variation in all three candidate genes contribute to susceptibility for schizophrenia. O16.8 A GENOME-WIDE SCAN FOR SCHIZOPHRENIA SUSCEPTIBILITY LOCI IN FAMILIES OF MEXICAN AND CENTRAL AMERICAN ANCESTRY Michael Escamilla1, Alfonso Ontiveros2, Humberto Nicolini3, Henriette Raventos4, Ricardo Mendoza5, Rolando Medina1, Rodrigo Munoz2, Douglas Levinson3, Juan Manuel Peralta4, Albana Dassori1, Laura Almasy4 1 University of Texas Health Science Cente, San Antonio, TX 2 INFOSAME, Monterrey 3 Study Group of Medicine and Families CAR, Mexico City 4 University of Costa Rica, San Jose 5 UCLA, Los Angeles, CA 2 Family Health Centers of San Diego, San Diego, CA 3 Stanford University, Palo Alto, CA 4 Southwest Foundation for Biomedical Rese, San Antonio, TX Schizophrenia is a complex psychiatric disorder, likely to be caused in part by multiple genes. In this study, linkage analyses were performed
Abstracts to identify chromosomal regions most likely to be associated with schizophrenia and psychosis in multiplex families of Mexican and Central American origin. We genotyped 459 individuals from 99 families, containing at least two siblings with schizophrenia or a related diagnosis, using 404 microsatellite markers. Multipoint nonparametric linkage analyses were performed using broad (any psychosis) and narrow (schizophrenia and schizoaffective disorder) models. Under the broad model, three chromosomal regions (1pter-p36, 5q35, and 18p11) exhibited evidence of linkage, with NPL scores greater than 2.7 (equivalent to empirical p values of less than 0.001), with the peak multipoint NPL ¼ 3.42 (empirical p value ¼ 0.00003), meeting genome-wide evidence for significant linkage in the 1pter-p36 region. Under the narrow model, the same three loci showed (nonsignificant) evidence of linkage. These linkage findings (1pter-p36, 18p11 and 5q35) highlight where genes for psychosis and schizophrenia are most likely to be found in persons of Mexican and Central American ancestry, and correspond to recent linkages of schizophrenia or psychosis in other populations which were formed in part from emigrants from Iberia and Western Europe in the 15th and 16th centuries. O17.1 AN EXAMINATION OF POPULATION DIFFERENCES IN THE IMAGE SAMPLE ACROSS 47 AUTOSOMAL GENES IMPLICATED IN PSYCHIATRIC DISORDERS Ben Neale1, Philip Asherson1, Shaun Purcell2, Keeley Brookes1, Kaixin Zhou1, Wai Chen1, Jan Buitelaar2, Richard S. Ebstein3, Michael Gill4, Tobias Banaschewski5, Iris Manor6, Ana Miranda7, Bob Oades8, Herbert Roeyers9, Ari Rothenberger5, Jo Sergeant6, Hans-Christoph Steinhausen7, Edmund Sonuga-Barke8, Pak Sham1, Steve Faraone2 1 Institute of Psychiatry, London 2 Massachusetts General Hospital, Boston, MA 2 Radbound University Nijmegen Medical Cen, Nijmegen 3 Herzog Memorial Hospital, Jerusalem 4 Trinity Centre for Health Sciences, Dublin 5 University of Goettingen 6 Geha MHC, Tel Aviv 7 University of Valencia 8 University Clinic for Child and Adolescents, Essen 9 Ghent University, Ghent 6 University of Amsterdam 7 University of Zurich 8 University of Southampton 2 SUNY Upstate Medical University, Syracuse, New York For association mapping, population differences pose a number of difficulties. At the simplest level, in case-control association analysis, population structure can give rise to false positive association. With the advent of the HapMap, association mapping has become more sophisticated, utilizing linkage disequilibrium(LD) information to tag human variation. Questions still remain about the extent to which population differences give rise to differences in LD structure, and the utility of the CEPH families to tag European populations effectively. The International Multi-site ADHD Genetics project collected 673 families from eight different countries (Belgium, Germany, Holland, Ireland, Israel, Spain, Switzerland, and the United Kingdom) and densely typed single nucleotide polymorphisms (SNPs) on 47 autosomal genes. Gene selection focused on hypothesized biological relevance to ADHD (and by extension many other psychiatric disorders). With an average marker density of approximately 1 SNP per 2.5 kb, this dataset provides an opportunity to explore the nature of population differences within Europe across these gene regions. Approximately a third of the markers show nominally significant differences for a simple test of heterogeneity, when including all populations. Exclusion of the Israeli samples from the test for heterogeneity reduces the proportion to about 8 percent. On average fewer tags are required to characterize the Northern European populations compared to the Swiss, Spanish, and Israeli, which could indicate reduced variation in Northern Europe, or more accurate tag SNP selection driven by the CEPH sample. O17.2 RESOLVING THE EVOLUTIONARY PARADOX OF COMMON, HARMFUL, HERITABLE MENTAL DISORDERS Matthew Keller1, Geoffrey Miller2 1 VIPBG, Richmond, VA 2 U New Mexico, Albuquerque
719
Given that natural selection is so powerful at optimizing complex adaptations, why does it seem unable to eliminate susceptibility alleles that predispose to common, harmful, heritable mental disorders, such as schizophrenia or bipolar disorder? We assess three leading explanations or this apparent paradox from evolutionary genetic theory: ancestral neutrality (susceptibility alleles were not harmful among ancestors), balancing selection (susceptibility alleles sometimes increased fitness), and polygenic mutation-selection balance(mental disorders reflect the inevitable mutational load on the thousands of genes underlying human behaviour). The first two explanations are commonly assumed in psychiatric genetics and Darwinian psychiatry, while mutation-selection has often been discounted. All three models can explain persistent genetic variance in some traits under some conditions, but the first two have serious problems in explaining human mental disorders. Ancestral neutrality fails to explain low mental disorder frequencies and requires implausibly small selection coefficients against mental disorders given the data on the reproductive costs and impairment of mental disorders. Balancing selection (including heterozygote advantage, antagonistic pleiotropy, and frequencydependent selection) tends to favor environmentally contingent adaptations (which would show no heritability) or high-frequency alleles (which psychiatric genetics would have already found). Only polygenic mutation-selection balance seems consistent with the data on mental disorder prevalence rates, fitness costs, the likely rarity of susceptibility alleles, and the increased risks of mental disorders with brain trauma, inbreeding, and paternal age. This evolutionary-genetic framework for mental disorders has important implications for the types of designs likely to find susceptibility alleles. O17.3 EVIDENCE OF INTERACTION BETWEEN DTNBP1 AND IL3 IN SCHIZOPHRENIA Todd Edwards1, Xu Wang2, Brandon Wormly2, Brien Riley2, F. Anthony O’Neill3, Dermot Walsh4, Kenneth S. Kendler2, Marylyn D. Ritchie3, Chen Xiangning2 1 Vanderbilt University, Nashville, TN 2 Department of Psychiatry, Virginia Commonwealth Univ., Richmond, VA 23298 3 The Queens University, Belfast, Northern Ireland 4 The Health Research Board, Dublin 3 Center for Human Genetics Research, Vand, Nashville, TN 37232 Schizophrenia is a psychotic mental disorder with a lifetime prevalence of approximately 1%. Heritability estimates of approximately 80% indicate this disease is largely genetic. Reports of risk loci and an epistatic model in the literature suggest a complex genetic etiology. In this study, we explored gene-gene interactions in population-based (657 cases: 414 controls) and family-based (269 families; 654 cases: 737 controls) datasets of English or Irish ancestry. Sixty markers in 8 genes were genotyped in families and thirteen markers in 3 genes were genotyped in population data. Results from the family dataset had an intermediate disease definition, and the population sample used the DSM-IIIR schizophrenia diagnosis or poor outcome schizoaffective disorder. Multifactor Dimensionality Reduction Pedigree Disequilibrium Test (MDR- PDT) was used to explore epistasis in families. A highly significant 3-locus interactive model was identified. The 3-locus interactive model was between genes IL-3, RGS4 and dystrobrevinbinding protein 1 (DTNBP1) (rs2069803, rs2661319 and rs2619539 respectively), p ¼ 0.006. To verify these findings, we used MDR on the case-control dataset containing the same markers typed in the 3 genes. While we could not replicate the 3-locus interaction, we saw evidence of a joint effect of IL3 and DTNBP1 with different markers (rs31400 in IL3 and rs760761 in DTNBP1), p < 0.001, odds ratio ¼ 1.25, 95% CI [0.95, 1.65]. This is the first report of gene-gene interaction associated with schizophrenia in both population-based and family-based association studies. O17.4 GENETIC INHERITANCE OF SCHIZOPHRENIA AND BIPOLAR DISORDER IN AN ITALIAN POPULATION ISOLATE Chiara Scapoli1, Sabina Rampinelli2, Giovanni Vazza1, Pio Peruzzi2, Gabriella De Sanctis2, Arianna Di Florio2, Cinzia Bertolin1, Maria Luisa Mostacciuolo1 1 Dipartimento di Biologia, Universita` di, Ferrara 2 Dipartimento di Neuroscienze, Universita`, Padova 2 Dipartimento di Salute Mentale, ULSS 14, Chioggia/Padova
720
Abstracts
Family, twin and adoption studies have shown that schizophrenia (SZ) and bipolar disorder (BP) are complex and highly heritable disorders. Heritability has been estimated around 80% in both disorders (Owen, 2005). Both Spectrum Disorders are genetically heterogeneous diseases and findings of linkage on different chromosomes in various studies corroborate this assumption (Kirov, 2005). Genetic isolates provide outstanding opportunities for identification of susceptibility genes for complex diseases. To assess the mode of inheritance of SZ/BP spectra disorders, a complex segregation analysis, has been performed in a total sample of 40 pedigrees (553 individuals; 96 nuclear families) originating from Chioggia population, that can be considered an ‘‘isolated community’’. POINTER indicates that polygenic, multifactorial and single locus models provide a significantly (P < 0.001) better fit to the data than a sporadic model. The general single locus model was not significantly different from additive or dominant hypothesis; the estimate of the susceptibility gene frequency was around 0.009. To test for the presence of a residual polygenic component, the mixed model was also examined: a main additive gene plus a residual polygenic component resulted the better hereditary model for Chioggia sample (p < 0.001), characterized by a high proportion of mixed (SZþBipolar) pedigrees. The concept of "additivity" may conceal a complex pattern: more genes with modifying effects may result in an additive pattern and this indication is in complete agreement with the well known complexity of the SZ/BP disorders, both on a clinical and genetic level ( Riley & Kendler 2006) O17.5 METRIC ISSUES IN THE GENETIC ANALYSIS OF COMPLEX TRAITS Michael Neale, Steve Aggen Virginia Commonwealth University, Richmond, VA Complex traits such as psychiatric disorders and substance abuse phenotypes are typically difficult to measure. Both behaviour and introspective reports about psychological states are usually measured with a series of items which may be binary or ordinal. Analysis of diagnoses derived from such items rely on a number of assumptions about measurement, which are rarely tested. It is shown that incorrect conclusions about the impact of genetic and environmental factors and their interaction may be drawn when the measurement model is flawed. Both increased statistical power and protection from erroneous conclusions may be obtained by multivariate genetic analysis. O18.1 BDNF Val66Met AND 270C > T POLYMORPHISMS AND PERSONALITY DIMENSIONS IN ANOREXIA NERVOSA Filip Rybakowski1, Monika Dmitrzak-Weglarz2, Agnieszka Slopien2, Maria Skibinska2, Andrzej Rajewski2, Joanna Hauser2 1 Poznan Univ. of Med. Sciences, Poznan 2 Univ. of Med. Sci., Poznan Background: Two polymorphic variants of BDNF gene (Val66Met and –270C > T) have been reported as associated with the pathogenesis of eating disorders. Several personality dimensions, such as high harm avoidance and high persistence probably predispose to the onset of anorexia nervosa (AN). These temperamental dimensions show high heritability, which may indicate their mediating role between genotype and AN phenotype. Method: One hundred-forty-eight AN patients (age 17.6 2.5) and 100 control females (age 20.7 1.3) were recruited for the study. Personality dimensions were assessed with Temperament and Character Inventory (TCI). We genotyped two polymorphisms in Brain Derived Neurotrophic Factor (BDNF) gene – Val66Met and –270C > T. Results: No significant associations between analysed polymorphisms and the diagnosis of AN were observed. However, we found that Val66Met and – 270C > T were significantly associated with harm avoidance dimension in AN patients, respectively (Anova F ¼ 3.022; p ¼ 0.05) and (Anova F ¼ 7.927; p ¼ 0.006). The significant association between –270 C > T polymorphism and trait of persistence in both control females (Anova F ¼ 3.711; p ¼ 0.05) and AN patients (Anova F ¼ 4.037; p ¼ 0.05) was also observed. Conclusions: We did not observe the association between analysed polymorphisms in BDNF gene with AN, which may be related to the relatively small sample size. Association of BDNF variants with personality dimensions predisposing to anorexia nervosa may suggest the use of personality assessment in the refinement of phenotype definition in the genetic studies of AN.
O18.2 GENETIC CONTINUITY AND CHANGE FOR BODY MASS INDEX IN MIDDLE CHILDHOOD Claire Haworth1, Lee M Butcher1, Susan Carnell2, Jane Wardle2, Robert Plomin1 1 Institute of Psychiatry, London 2 University College London, London In a large representative sample of twins, we used both quantitative and molecular genetic methodologies to investigate the developmental etiology of body mass index (BMI) from 7 to 10 years of age. BMI data were collected at 7 years (3452 pairs) and at 10 years (5208 pairs). Heritabilities of BMI were.75 at both ages. Longitudinal analyses yielded a genetic correlation of.63, indicating substantial genetic continuity. Nonetheless, genes also contribute to change in that significant new genetic influence emerged at 10 years that was independent of 7 years. These results lead to two predictions for molecular genetic research: (1)many QTLs associated with BMI at 7 will also be associated with BMI at 10, and (2) some QTL associations will be specific to 10. Using DNA from 5000 individuals, we tested these hypotheses for 10 single nucleotide polymorphisms (SNPs) reported to be associated with obesity phenotypes. Although only two SNPs showed significant associations with BMI in the reported direction, they confirmed our predictions in that one SNP (rs1801282) was associated at both ages; the other SNP (rs1800206) was significantly associated only with BMI at 10. The replication of only two of the ten previouslyreported associations, despite our large sample size, is typical of complex traits, especially as many reported studies were underpowered to detect associations of small effect. Nonetheless, our finding of genetic change as well as continuity leaves open the possibility that some of these reported SNP associations with BMI might emerge in adolescence or adulthood. O18.3 ASSOCIATION BETWEEN THE INTERLEUKIN 1 RECEPTOR ANTAGONIST (IL1RN) INTRONIC REPEAT AND DYSREGULATED EATING BEHAVIOUR Richard Ebstein Hebrew University, Jerusalem IL-1 receptor antagonist (IL1RN), present in peripheral and CNS tissues, is a highly selective receptor antagonist and blocks all action of IL1. IL-1Ra can attenuate the suppression of appetite in response to leptin. IL1RN is characterised by length variation in intron 2 due. IL1RN*2 was shown to be associated with reduced levels of IL-1Ra protein. The role of IL1Ra in feeding and appetite prompted us to examine the role of the IL1RN polymorphism in eating behaviour in 737 healthy women, assessed by scores on the EAT-26 and EDI, as well as in a group of 182 women diagnosed with DSM IV Anorexia Nervosa (AN). Family-based associations were tested using PBAT and UNPHASED. Significant association using PBAT was observed between the IL1RN*2 allele, EAT-26 scores (p ¼ 0.0177) and EDI BD (p ¼ 0.0135). The presence of the 2 repeat predicted lower scores on these questionnaires. The 4 repeat was associated with higher EAT scores. Although IL1RN was not in itself associated with DSM IV AN, conditional analysis using UNPHASED between IL1RN and 3 SNPs in the IGF2 gene (previously associated by us with EAT-26 scores1 in healthy subjects) showed significant association with DSM IV AN (LRS ¼ 18.531 DF ¼ 6 p ¼ 0.005). This is the first report linking the IL1RN repeat polymorphism to eating regulation in both healthy females and AN. 1. Bachner-Melman R et al., Am J Psychiatry. Dec 2005;162(12):2256–2262. O18.4 THE ROLE OF 5-HTTLPR POLYMORPHISM IN PREDICTION OF DIAGNOSTIC SHIFT IN EATING DISORDERS Maria Cristina Cavallini1, Massimiliano Grassi2, Stefano Erzegovesi3, Alessandra Bosaia2, Francesca Mapelli2, Laura Bellodi2 1 Universita` Vita-Salute San Raffaele, Italia 2 Universita` Vita-Salute San Raffaele, Milano 3 Universita` Vita-Salute San Raffaele, Milano Eating Disorders (ED) have an unstable course. From 8 to 62% of patients affected with Anorexia (AN) develop bulimic symptoms requiring complex clinical management during their lifetime. We plan to construct a model that predicts the risk of shift for ED patients, including at the first level clinical variables such as temperament and
Abstracts character dimensions as measured on the TCI- R scale, familial variables, onset, genetic factors (role of 5-HTTLPR polymorphism, already involved in BN liability). We recruited 180 patients satisfying AN restricting subtype criteria (DSM IV) at onset, from a sample of 568 ED patients in our Eating Disorder Unit. 35% of these patients, having an average illness duration of 7.8 years, at some time during their lives were diagnosed with BN. Along with linear analyses, we applied a nonlinear approach (Artificial Neural Network, ANN) able to model the relationships among variables without an a priori description of the features of these relationships. Preliminary results, supported by linear and non-linear analysis, did not confirm the involvement of 5-HTTLPR genotypes or alleles in vulnerability to developing binge or purging behaviours in AN patients, while self-directness and harm avoidance and age at onset may play some role in the probability of crossover. At this stage of the analysis, ANN did not show greater advantages than linear techniques. The singling out of predicting factors could point the way to more specific therapeutic intervention.
O18.5 GENOMIC DNA HYPOMETHYLATION IN FEMALES WITH ANOREXIA NERVOSA Helge Frieling1, Andreas Gozner2, Bernd Lenz1, Thomas Hillemacher1, Dominikus Bo¨nsch1, Martina de Zwaan1, Johannes Kornhuber1, Stefan Bleich1 1 University of Erlangen-Nuremberg, Erlangen 2 Unversity of Erlangen-Nuremberg, Erlangen Background: Epigenetic mechanisms such as DNA-methylation or histone modifications are believed to play a crucial role in the pathogenesis of several psychiatric diseases. The aim of the present study was to examine whether alterations of global(genomic) DNA methylation exist in females with eating disorders. Methods: Genomic DNA methylation was assessed in 21 females suffering from anorexia nervosa (AN), 22 suffering from bulimia nervosa (BN) and 30 healthy women as controls. After extraction from EDTA-blood, DNA was digested methylation sensitive (using HpaII and MspI endonucleases). A single nucleotide elongation assay using Cy5-labelled dCTPs was performed to measure genomic DNA methylation. Results: We found a global DNA hypomethylation in patients with AN (47.2% 19.9; ANOVA: F ¼ 6.87; R2 ¼ 0.16; P ¼ 0.002) as compared to BN (60.4% 21.7; Bonferroni: P ¼ 0.039) and controls (65.2% 10.5; Bonferroni: P ¼ 0.001). No significant differences were found between BN and controls. Discussion: Our study reports for the first time alterations in genomic DNA methylation in patients suffering from AN. As already known for psychiatric diseases like alcoholism and schizophrenia, epigenetic mechanisms may play a role in the pathophysiology of this disorder. O19.1 ALTERNATIVE STRATEGIES FOR QUANTITATIVE TRAIT HAPLOTYPE ANALYSIS: APPLICATION TO MDR1 GENE IN TARDIVE DYSKINESIA Vincenzo De Luca1, Clement Zai1, Daniel J Muller1, Jeffrey Liebreman2, Herbert Meltzer3, Gary Remington1, James L. Kennedy1 1 University of Toronto, Toronto 2 University of North Carolina, Chapel Hill 3 Vanerblit Univeristy, Nashville Pharmacogenetics, using a candidate gene association approach, investigates pharmacodynamic and pharmacokinetic gene factors that may influence drug response and side effects. For example, about 20% of patients with schizophrenia receiving antipsychotic treatment develop tardive dyskinesia (TD), an involuntary movement disorder associated with long-term conventional antipsychotic administration. To date, however, there are no published reports correlating certain candidate genes (e.g. drug transporters) with the pathogenesis of TD. In addition to genetically variable receptor proteins, drug transporters are recognized as a key determinant of drug disposition and response, influencing blood and tissue levels of many drugs by controlled transport across compartmental boundaries. Major determinants eliminating drugs and toxics from CNS are ATP-dependent efflux pumps such as p-glycoprotein (e.g. MDR1), which can influence the bioavailability and CNS concentrations, as well as the disposition of drugs. In this report, we applied different analysis methods to detect
721
the effect of MDR1 haplotypes on the AIMS score of 235 schizophrenics with TD. When we compared the global p-value calculated with QTPHASE and EHAP, we found a value of 0.629 and 0.540 respectively. The hypothesis that MDR1 can influence the severity of TD is rejected by both programmes; however, this report suggests that different programmes yield similar but not the same p-values and thus threshold p-values should always be rejected, given the possibility of false positives by using different programmes. O19.2 REPLICATION OF GENETIC MARKERS ASSOCIATEDWITH CLOZAPINE-INDUCED AGRANULOCYTOSIS Maria Athanasiou1, Michael Dettling2, Bradley Dain1, Michael Lachowicz1, Anil Malhotra2, Gerson Stanton3, John Kane2, Ingolf Cascorbi3, Igor Mosyagin3, Carol Reed1 1 Clinical Data, Inc., New Haven, CT 2 Charite´ — Universita¨ts medizin Berlin, Berlin 2 The Zucker Hillside Hospital, New York, NY 3 Case Western Reserve University, Cleveland, OH 3 University clinic Schleswig-Holstein, Kiel Clozapine is a highly efficacious drug for the treatment of schizophrenia, but its use is limited, in part due to the side effect of agranulocytosis. In order to reduce the incidence of clozapine-induced agranulocytosis (CIA), patients are required to submit to a blood monitoring program. If a genetic test for CIA were developed, it is possible that clozapine could become a safer drug and more patients could benefit from its use. We previously conducted a candidate genebased case-control study to discover genes associated with CIA. Cases were patients with ANC < 500 during clozapine treatment, and controls were treated for at least one year without a significant reduction in WBC or ANC. Seventy-four candidate genes were sequenced and a haplotype analysis was conducted using logistic regression. Genetic markers in HLA- DQB1, HLA-C, NTSR1, DRD1 and CSF2RB were significant with a permutation test adjusted p-value < 0.05. We attempted to replicate the findings from 4 of these genes in an independently collected cohort of 49 cases and 78 controls. Cases were patients with an ANC < 500 during clozapine treatment and controls were treated for at least two years without an adverse effect on the WBC or ANC. Permutation tests were performed for markers within genes, and the Benjamini-Hochberg FDR method was applied to the best permutation p-values from the four genes. A genetic marker in one gene replicated. The strength of the findings is sufficient to develop a diagnostic test for CIA risk. O19.3 ANTIPSYCHOTIC TREATMENT INDUCED WEIGHT GAIN IN SCHIZOPHRENIA ASSOCIATED WITH THE DRD2 GENE? Daniel J. Mu¨ller1, Rudi Hwang2, Clement Zai2, Tricia Sicard2, Subi Tharmalingam2, Nicole King2, Jan Volavka3, Pal Czobor3, Jeffrey Lieberman4, Herbert Meltzer5, James Kennedy2 1 University Medicine of Berlin (CCM), Berlin 2 Neurogenetics Section, Centre for Addict, Toronto, ON 3 Nathan S. Kline Institute for Psychiatry, Orangeburg, NY 4 Columbia University, New York, NY 5 Psychiatric Hospital at Vanderbilt, Nash, Nashville, TN Treatment-emergent weight gain varies within the broad class of antipsychotics and is highest for clozapine and olanzapine. However, an individual’s propensity to gain weight depends largely on genetic factors. Candidate gene studies have produced significant findings in the 5-HT2C and ADRA2a receptor genes, as well as in the leptin, GNB3 and SNAP-25 genes. Surprisingly, the dopaminergic system has received little attention, although this system activates reward circuits stimulated by food intake, which is thought to increase following antipsychotic treatment (‘‘food craving’’). Furthermore, the DRD2 is known to play a key role in the efficacy of antipsychotics. We will present new analyses that we conducted spanning the entire DRD2 gene comprising 12 SNPs. We analysed a population of 139 chronic schizophrenia patients derived from two study samples (A ¼ 80, B ¼ 59). The majority were treated with clozapine and were followedup for weight changes for up to 14 weeks. Carriers of the C/C genotypes of the (functional relevant) C957T polymorphism (n ¼ 52) gained on average 5.76 kg as compared to 3.42 kg in carriers of the C/T genotypes (n ¼ 50) and 2.51 kg in T/T carriers (n ¼ 21; F [2,120] ¼ 4.29, p ¼ .01).
722
Abstracts
Haplotype analyses yielded the most significant findings when including the adjacent C939T and Taq1D polymorphisms (Global: LRS ¼ 19.39, p ¼ .003; Individual: chi-square ¼ 15.5, p ¼ .0001). In summary, our preliminary analyses indicate that DRD2 gene variants are associated with antipsychotic-induced weight gain. O19.4 INTERACTION OF ADRA2A GENE AND ADRB3 GENE ON BONE DENSITY IN MALE SCHIZOPHRENIC PATIENTS WITH EXPOSURE TO ANTIPSYCHOTICS Tsuo-Hung Lan Yu-Li Hospital, Department of Health, Hualien Objective: A lower bone density is suspected prevalent among schizophrenia patients while compared to the general population. This study is to explore the gene-gene interaction on bone density in male schizophrenic patients from ADRA2A gene and ADRB3 gene. Method: This is a multi-center, investigator-initiated, naturalistic study project. Here we enrolled 119 male inpatients meeting DSM-IV criteria for schizophrenia or schizoaffective disorder from two psychiatric hospitals in Taiwan. Bone density indicated in BUA was measured for all subjects by using QUS-2 over calcaneal bone. Descriptive data of age, sex, height, body weight, medical or surgical past history, activity level, smoking and the antipsychotics used in the past two years were collected. Four candidate genes including ADRA1A, ADRA2A, ADRB3, HTR2A have been genotyped by PCR or DHPLC. Results: After controlling for gender, age, previous antipsychotics treatment, and BMI, the absolute regression coefficient of BUA of an interaction term between ADRA2A gene and ADRB3 gene for an individual male is 21.0 10.0 (p-value ¼ 0.03). Conclusion: It is suggested that the interaction between ADRA2A gene and ADRB3 gene indeed modified the bone density effect in our male schizophrenic patients significantly. O19.5 EFFECTS OF ANTIDEPRESSANT TREATMENT ON GENE EXPRESSION IN MOUSE CULTURED CELLS Ursula M. D’Souza, Karen Sugden, Carmine M. Pariante, Helen Byers, Malcolm Ward, Leonard C. Schalkwyk, Peter McGuffin, Katherine J. Aitchison Institute of Psychiatry, London Purpose of the study: Antidepressant response varies between individuals and is partly determined by genetic factors. It is likely to be effected through complex molecular mechanisms including variations in transcription and post-translation. We tested the hypothesis that two antidepressants, escitalopram (an SSRI) and nortriptyline (a pronoradrenergic tricyclic antidepressant), exert their effect through alterations in expression levels of a number of neurotransmitter and drug-transport associated genes. These include serotonin transporter (5-HTT), brain-derived neurotrophic factor (BDNF), glucocorticoid receptor (NR3C1), P-glycoprotein 1a and 1b (ABCB1a and ABCB1b), Pglycoprotein associated protein 1 (ABCC1) and FK-506 binding protein 5 (FKBP5). Transcriptomic (microarray) and proteomic analyses are being carried out. Methods: Mouse fibroblast cell line L929 was treated in vitro with the two antidepressants at 1 uM and 10 uM concentrations for 30 minutes, 24 hours and 48 hours. Cells were then harvested and RNA extracted, before cDNA was generated and used in quantitative RT-PCR assays. The gene expression data were normalised to a variety of housekeeping genes. Results: Results of non-parametric tests suggest that 5-HTT, ABCB1a, ABCC1 and FKBP5 are differentially expressed following antidepressant treatment. The effects are observed over varying times and at different drug concentrations. Conclusions: These data suggest that antidepressants might exert some of their effects through changes in expression levels of genes involved in serotonergic neurotransmission (5-HTT), regulation of glucocorticoid receptor function (FKBP5) and drug transport (ABCB1a and ABCC1). O20.1 A POWERFUL MEANS FOR ANALYSES OF GENETIC EPISTASIS IN COMPLEX TRAITS: A GOOD DEMONSTRATION IN THE STUDY OF SCHIZOPHRENIA SUSCEPTIBILITY LOCI INTERACTION YongYong Shi1, Qiang Liu1, XinZhi Zhao1, RuQi Tang1, Wei Tang1, Yun Liu1, Ke Huang1, GuoYin Feng1, Jurg Ott2, Lin He1 1 Shanghai Jiao Tong University, Shanghai 2 Rockefeller University, New York
Schizophrenia is a severe mental disorder affecting 1% of the population, and its complex genetic etiology should be influenced both by principal effects and the epistasis of causal loci. To analyse susceptibility gene interactions in large scale association studies, we developed an analytical method, PDSE, which focuses purely on causal epistasis in complex traits. We applied several models to theoretically ensure PDSE is powerful and feasible in mapping genetic susceptibility epistasis loci on an extremely large scale, even genome-wide. By carrying out PDSE on a screen of 91 SNPs in 10 previously-reported candidate genes, DTNBP1, CAPON, TNFB, DBH, DRD3, DISC1, COMT, PIK4CA, SLIT3 and ZDHHC8, we found that the epistatic effects of two SNPs,rs11755055 and rs9464794, were significantly associated with schizophrenia (p ¼ 5.87e-15), after multiple test correction (the corrected significance threshold is 1.22e-5). These two SNPs lie in intron regions of DTNBP1, implying that functional or structural interactions are involved in schizophrenia etiology in this gene/region. All SNPs are genotyped on 707 patients and 689 controls from the Chinese Han population. We recommend that genome-wide association studies of schizophrenia or other complex diseases should use both traditional chi-square tests and PDSE to respectively deal with principal susceptibility and epistatic effects(in another application, it can analyse 116,204 SNPs in pairs in 6.5 hours on a PC with one 3.6GHz CPU, which is acceptable). O20.2 A GENOMEWIDE ASSOCIATION SCAN FOR GE INTERPLAY IN CHILDHOOD Lee M. Butcher, Emma Meaburn, Ian W. Craig, Leonard C. Schalkwyk, Robert Plomin Institute of Psychiatry, London Introduction: Molecular genetic studies focus on finding main effects of genes then look for interactions with environmental measures. We conducted the first genomewide association scan that focuses on the interplay (correlation as well as interaction) between genes (G) and environment (E). Method: Using a sample of 12,000 7-year-olds assessed on general cognitive ability (‘g’) and ‘household chaos’, we conducted a genomewide association scan with 500,000 SNPs in a 2 2 design: high versus low ‘g’ and high versus low ‘household chaos’. From the extremes of the sample of 12,000 children, about 200 unrelated children were selected for each cell of the 2 2 design. After screening the 500,000 SNPs using pooled DNA allelotyped on microarrays, SNPs were selected for individual genotyping in a representative sample of 5,000 children. Results: SNPs were screened separately for associations (1) with ‘g’, (2) with ‘household chaos’ (GE correlation), and (3) the interaction between ‘g’ and ‘household chaos’ (GE interaction), using a combination of permutation testing and novel quality control metrics. More than 20 different SNPs were significantly associated with each of these three analyses in the sample of 5000 children. Although the effect sizes of the associations were small (.2% on average), the SNPs can be combined into SNP sets and used as genetic risk indices of ‘g’, GE correlation and GE interaction. Conclusions: Genomewide association scans for quantitative traits can be identified not only for main effects of genes but also for GE correlation and GE interaction.
O20.3 A SIMULATION STUDY: NON-RANDOM SNP GENOTYPING ERROR IMPACT ON FAMILY BASED QTL LINKAGE ANALYSIS Kaixin Zhou, Fruhling Rijsdijk, Philip Asherson Institute of Psychiatry, London Linkage study based on a dense SNP marker panel might be more efficient than traditional miscrosatellites (Evans, 2004). One advantage of the SNP-based linkage panel is high genotyping accuracy, 99% or better. The other special characteristic of the SNP genotyping microarray is its non-random distribution of genotyping errors. Although previous simulation studies suggested that 1% genotyping error will dramatically influence QTL linkage output (Abecasis 2001), it is not clear whether this holds true for the new SNP panel, due to the ideal conditions assumed in previous studies and the new features of SNP linkage panels. In this simulation study, we applied the genetic map and allele frequency spectrum of a real SNP-based linkage map, the Illumina linkage Panel IV (ILP-IV), to investigate both random and non-random genotyping error impact on family-based QTL linkage
Abstracts studies under single ascertained or affected sib pair designs. The sensitivity of linkage test methods to genotyping errors will be discussed on this occasion. Our results replicated Abecasis’ result that the effect of 1% error is severe in most cases. The Merlin-regression method is preferred rather than a non-parametric test when population parameters are available for primitive genome scanning. Non-random genotyping errors are only more worrisome when the regression-based method is used under the affected sib pair design. We also demonstrated that a genotyping error rate of 99.9% would be favourable in a dense SNP-based QTL linkage study to assure a reliable result.
O20.4 12q14: A NOVEL LINKAGE REGION IN EXTENDED AUTISM FAMILIES Deqiong Ma1, J. Jaworski1, DA Skaar1, A Konidari1, C Haynes1, R.K. Abramson2, H.H. Wright2, M.L. Cuccaro1, J.R. Gilbert1, J.L. Haines2, M.A. Pericak-Vance1 1 Center for Human Genetics, Durham; NC 2 University of South Carolina School of Medicine, Columbia, SC 2 Center for Human Genetics Research, Nashville, TN Autism is a common neurodevelopmental disorder, with a significant genetic component and substantial locus heterogeneity. No consensus linkage region or associated candidate gene from any proposed region has emerged despite extensive studies. The primary substrate for linkage studies has been small nuclear families. A contrasting and powerful alternative is the use of large, extended pedigrees to identify significant linkage results, as these families contain more potential linkage information than sibpair families. We collected 26 extended autistic families (65 affected, 184 total individuals), with each family having two to four affected individuals comprised of either avuncular or cousin pairs. We then performed a genome-wide linkage analysis using a high-density single-nucleotide-polymorphism (SNP), Affymetrix GeneChip Human Mapping 10K array, and identified a genome-wide significant linkage signal on chromosome 12q14-15. Nine more microsatellites covering the interested region have now been typed on these families and confirm the linkage signal with the peak at D12S1686 [78.75 cM] (heterogeneity LOD [HLOD]: 3.49 [recessive model]). The linkage signal was enhanced in 17 male-affected only families with a 4.56HLOD under recessive model. Using the power of the extended pedigrees, we constructed haplotypes and narrowed the minimum candidate region to a 4 cM segment from 75cM to 79 cM. This region contains 10 known genes, including interesting candidates AVPR1A; SRGAP1; WIF1; GRIP1 and TPH2. Analysis of these genes is underway.
O20.5 SCHIZOPHRENIA IS LINKED TO CHROMOSOME 20p13 IN A LARGE, INBRED, ARAB ISRAELI PEDIGREE Yoav Kohn1, Kyra Kanyas1, Osnat Karni1, Adi Levi1, Omri Teltsh1, Mira Korner2, Adnan Hamdand3, Bernard Lerer1 1 Biological Psychiatry Laboratory, Department of Psychiatry, Hadassah-Hebrew University Medical Center, Jerusalem 2 The Center for Genomic Technologies, Institute of Life Sciences, Hebrew University, Jerusalem 3 Regional Mental Health Center, Taibe, Israel Schizophrenia is a common and severe heritable disorder. Despite extensive research, knowledge of its pathophysiology is limited. Linkage and association studies have repeatedly drawn attention to several chromosomal regions and to genes within them. Conflicting patterns of association and the lack of a clear functional significance of the associated variants limit the interpretation of these results. The use of rare pedigrees where genes with large effect cause the disorder has proven beneficial in studies of other complex disorders. Our objective was to use this advantage by performing a genome wide linkage analysis for schizophrenia in a large, multiplex Arab pedigree from Israel. We genotyped 347 microsatellite markers in 24 pedigree members affected with schizophrenia spectrum disorders and 32 unaffected relatives. Further fine mapping was performed in regions of interest. Linkage analysis with Superlink demonstrated a significant LOD score of 3.37 on chromosome 20p13 under an autosomal dominant mode of inheritance. Further fine mapping narrowed down the linked region to 2–5 cM. A haplotype containing the markers D20S193, D20S889 and D20S116, which spans 1.8 Mb was found to be shared by
723
most affected pedigree members. This interval contains two strong candidates for schizophrenia genes—PANK2 and ATRN. Our study exemplifies the power of rare and unique pedigrees to draw attention to novel and less recognized regions for genetic studies of schizophrenia. O21.1 POLYMORPHIC VNTRS IN THE 5HT TRANSPORTER GENE ALLOW A MECHANISM FOR GENETIC AND ENVIRONMENTAL INTERACTIONS IN THE DEVELOPMENT OF BEHAVIOURAL DISORDERS John Quinn Physiology Laboratory, Liverpool A genetic predisposition to a number of neurological disorders has been attributed to the human serotonin transporter gene (5HTT). In particular, two variable number tandem repeat (VNTR) domains within Intron 2 and a second in the 5’ promoter of 5HTT have been hypothesized as predisposing genetic factors for such conditions. We have demonstrated that the Intron 2 VNTR supports differential gene expression in a tissue-specific and stimulus-inducible fashion and that it is bound and regulated by the transcription factors YB-1 and CTCF. We now demonstrate synergistic action between the 50 Promoter VNTR and that within Intron 2. We propose a synergistic interaction between the two domains to be important for the final transcriptional response of a specific allele in both normal physiology and in conditions which result in behavioural disorders. We will present reporter gene data on the function of these VNTRs, in addition to biochemical analysis of transcription factor binding to the domains. Furthermore, we will correlate modulation of transcription factor binding in vivo with endogenous transporter gene expression in response to lithium. We postulate that this transcriptional mechanism allows for genetics and the environment to interact in the development of behavioural disorders. Our aim is to achieve a better understanding of how polymorphic domains spanning the serotonin transporter gene might modulate invariant regulatory domains to mediate a transcriptional response to challenge. O21.2 GENOME-WIDE CORRELATION OF DNA VARIATION AND RNA EXPRESSION WITHIN HUMAN CORTEX Amanda Myers1, Kristen Rohrer2, Alice Zhao2, Lauren Marlowe2, Mona Kaleem2, Leslie Bryden2, Doris Leung2, John Hardy2 1 NAS/NIA/NIH, Bethesda, 2 NIA/NIH, Bethesda In our studies of neurologic disease we have found that non-coding haplotype variation is necessary to define disease risk. We hypothesized that these variants might effect expression of the risk genes. To examine this, we are correlating DNA SNP variation genome-wide with changes in RNA levels. Since our series of samples (n ¼ 765) contains tissue from multiple regions, our first experiment was to define the regional expression of transcripts within the normal human brain. We have 8 neuropathologically normal samples from which there is tissue from multiple regions. We probed RNA extracted from frontal, temporal, parietal and cerebellar cortex of these 8 samples on the illumina sentrix bead chip. Comparisons of transcript levels were made using the Illumina beadstation software. We found that the correlation between the expression profiles of frontal and temporal cortex was 0.99, between temporal and parietal cortex it was 0.98 and between temporal and cerebellar cortex it was 0.90. There were 35 transcripts that had significantly different (p ¼ 0.05) expression between frontal and temporal cortex, 322 transcripts that were significantly different between parietal and temporal cortex and 5778 transcripts that were significantly different between cerebellum and temporal cortex. Three transcripts were significantly different between all areas of cortex, thus making the total sum of the significantly different transcripts among the cortical regions in our series around 350 transcripts. Currently, we are analyzing the remaining transcripts and looking for changes in expression levels which correlate to DNA variation. O21.3 CIS- AND TRANS-ACTING LOCI INFLUENCE EXPRESSION OF DTNBP1,A SUSCEPTIBILITY GENE FOR SCHIZOPHRENIA Nick Bray, Pete Holmans, Marianne van den Bree, Nigel Williams, Lesley Jones, Nick Craddock, Michael Owen, Michael O’Donovan University Hospital of Wales, Cardiff
724
Abstracts
Susceptibility to complex disease appears to be partly mediated by heritable differences in gene expression. Genetic influences on gene expression can be trans-acting as well as cis-acting, and are therefore not limited to the gene locus itself. Recent studies have demonstrated that expression of a gene can serve as quantitative trait for linkage analysis. These studies have indicated that a large proportion of the heritable variance in gene expression is attributable to trans-acting loci. Our studies are based on the principle that, if genetic variation influences disease susceptibility through cis-acting effects on a gene’s expression, then other genes containing polymorphism with a transacting effect on expression of that gene may also be susceptibility loci. Thus, a disease gene’s expression may serve as an endophenotype for mapping novel loci that modulate disease susceptibility. We have applied this method to the DTNBP1 gene, encoding dysbindin, for which there is strong evidence that association with schizophrenia is mediated by cis-acting effects on DTNBP1 expression. Quantitative linkage analysis was performed on real-time PCR measures of DTNBP1 expression in lymphoblastoid RNA from large CEPH sibships. A linkage peak was observed at the DTNBP1 locus on chromosome 6p22–23, consistent with cis-acting effects on DTNBP1 expression. However, highest LOD scores ( > 4) were observed at other chromosomal loci, suggesting the additional importance of polymorphic trans-acting influences on expression of DTNBP1. O21.4 EXPRESSION OF VARIANT COMT IS DEPENDENT ON POSTTRANSCRIPTIONAL CONTROL Philipp Sand, Gabriele Korte, Martina Domani, Nikola Konstandin, Alexandra Karl, Katrin Wagner, Matthias Dinnbier University of Regensburg, Regensburg The human COMT gene encodes for isoforms of catechol-O-methyltransferase, a widely expressed enzyme involved in the degradation of catecholamines. COMT activity is determined by a common missense substitution at residue 108 (158) which has become a focus of research in schizophrenia and other neuropsychiatric disorders. The authors provide proof that the region harboring the missense substitution, however, is subject to posttranscriptional modification that abrogates the variant’s functionality. This may help to reconcile conflicting association data that have accumulated in schizophrenia, ADHD and bipolar disorder. O21.5 EXPRESSION OF GENES AFFECTING GLUTAMATE RECEPTOR EDITING IS DIFFERENTIALLY MODULATED BY MOOD STABILIZERS Philipp Sand1, Gabriele Korte1, Peter Schreier2, Peter Eichhammer3 1 University of Regensburg, Regensburg 2 University of Wuerzburg, Wuerzburg 3 Molecular and Behavioral Neuroscience In, Ann Arbor Glutamate is the principal excitatory neurotransmitter in the brain and interacts with both metabotropic and ionotropic postsynaptic receptors. Glutamate receptor overstimulation increases intracellular calcium by directly opening ion channels and secondarily affecting neuronal calcium homeostatic mechanisms. The editing of the GluR2 subunit at the Q/R site, however, is known to counteract excitotoxic effects of glutamate and protects vulnerable neurons from ischemic insults. In a pilot investigation, we addressed the effects of three mood stabilizers (lithium chloride, valproic acid and lamotrigine) on the expression of genes that encode for RNA-editing enzymes, and examined editing at the GluR2 Q/R site at different dosages in human HL-60 promyelocitic cells. The expression of genes that enhanced pre-mRNA editing was enhanced in a dose-dependent fashion up to 4-fold following exposure to valproic acid and, to a lesser extent, following exposure to lamotrigine. In contrast, lithium chloride downregulated expression of the genes under study. The present data suggest that distinct effects of mood stabilizers on excitatory transmission may be achieved by an enhancement, or an attenuation, of glutamate receptor subunit editing. O22.1 ASSOCIATION OF THE IL-10 RECEPTOR 1 GENE WITH CHILDHOOD-ONSET MOOD DISORDERS Virginia Misener1, Lissette Gomez1, Karen G. Wigg1, Paola Luca1, Nicole King2, Agnes Vetro3, Eniko Kiss3, Zsuzsa Tamas4, James L. Kennedy2, Maria Kovacs3, Cathy L. Barr4 1 Toronto Western Research Institute, Toronto, Ontario
2
Neurogenetics Section, Centre for Addiction, Toronto, ON Department for Child and Adolescent Psyc, Szeged Vadaskert Hospital, Budapest 3 School of Medicine, University of Pittsburgh, Pittsburgh, PA 4 Toronto Western Research Institute; Brai, Toronto, ON 3 4
Inflammatory cytokines induce a syndrome of ‘‘sickness behaviour’’, resembling symptoms typically seen in depression. This has led to the hypothesis that cytokine activity contributes to depressive disorders. A logical extension of this hypothesis is that some variations in cytokine system genes confer risk for depression. To test this, we are investigating cytokine network genes in childhood-onset mood disorders (COMD). Childhood-onset depression is highly heritable, making this a useful approach for identifying genes underlying depression. To test for associations with COMD, we used the transmission/disequilibrium test (TDT), analysing 386 families with children diagnosed with depression before age 14. Twenty-five polymorphisms in 7 genes were investigated: IL-1alpha (1), IL-1beta (7), IL-6 (4), TNF-alpha (2), IL-10 (2), IL-10 receptor 1 (IL-10R1) (2), IL-1Ra (7). TDT analyses showed no significant evidence for association with these polymorphisms individually (pvalues>0.05).However, TDT analyses of haplotypes showed significant evidence for association with a 2-marker haplotype (rs4252272[A]/ rs2229113 [A]) of IL-10R1 (chi-squared ¼ 5.03; p ¼ 0.02). Each of these variants changes the amino acid (aa) sequence of the IL-10R1 protein (aa138:Ser/Gly and aa330:Gly/Arg, respectively). Moreover, functional evidence for a proinflammatory effect of the 330Arg variant, and the possibility of altered ligand binding due to the Ser138Gly change, have been reported. Thus, the association of these IL-10R1 variants with COMD suggests their involvement in conferring risk for depression, and warrants further investigation. O22.2 A BRAIN-DERIVED NEUROTROPHIC FACTOR HAPLOTYPE IS ASSOCIATED TO ANTIDEPRESSANT TREATMENT RESPONSE IN MOOD DISORDERS Virginia Soria1, Mo`nica Grataco`s2, Juan R Gonza´lez2, Mo`nica Baye´s2, Rafael de Cid2, Jose M. Crespo3, Julio Vallejo3, Mikel Urretavizcaya3, Xavier Estivill2 1 University Hospital of Bellvitge, Barcelona 2 Center for Genomic Regulation, Barcelona 3 University Hospital of Bellvitge, Barcelona To test whether BDNF is a susceptibility locus for mood disorder phenotypes and antidepressant treatment response, we performed a case-control study with eight TagSNPs (including the Val66Met functional variant), covering the entire BDNF region. We have genotyped 342 control subjects and a clinical sample of 374 patients with mood disorders diagnosed according to DSMIV criteria. Single SNP case-control association analysis regarding the mood disorder phenotype identified a significant association with an SNP located in the 5’ upstream region of the gene (p ¼ 0.029), which did not remain significant after Bonferroni correction. We further investigated the association of this nominal positive SNP with the antidepressant treatment response phenotype. Patients were allocated to one of two groups according to the Thase and Rush staging scheme: those with clinical remission after a major depressive episode treated with adequate antidepressant monotherapy were denoted ‘responders’. After testing for five genetic models of inheritance, the additive model was the one that best fit the data. Thus, patients heterozygous for variant alleles have almost 3-fold probabilities of being responders, almost 6-fold if they are homozygous carriers (OR ¼ 2.95; p ¼ 0.0025). We have further identified a haplotype conferring risk to this particular phenotype through a 3 SNPs sliding window approach (OR ¼ 2.71; p ¼ 0.0078). While our results failed to detect a positive association of BDNF with mood disorder diagnosis, they support the implication of BDNF in the therapeutic response to antidepressants. O22.3 GENETIC MARKERS OF SUICIDAL IDEATION EMERGING DURING ANTIDEPRESSANT TREATMENT Gonzalo Laje1, Silvia Buervenich2, Husseini Manji1, A. John Rush2, Alexander Wilson3, Dennis Charney4, McMahon Francis1 1 National Institute of Mental Health, Bethesda 2 Karolinska Institutet, Stockholm 2 University of Texas Southwestern, Dallas 3 National Human Genome Research Institute, Baltimore 4 Mount Sinai School of Medicine, New York
Abstracts Background: Suicidal ideation is an uncommon but potentially dangerous symptom than can emerge during antidepressant treatment, fueling concern about the safety of antidepressant medication. The biological basis of treatment-emergent suicidal ideation (TESI) is unknown. Genetic markers may shed light on the causes of TESI and help identify individuals at high-risk who may benefit from closer monitoring, alternative treatments, and/or specialty care. Methods: DNA samples were collected from a clinically-representative cohort of 1938 outpatients with major depressive disorder who were treated with the selective-serotonin reuptake inhibitor (SSRI), citalopram, under a standard protocol for up to 14 weeks. DNA was genotyped for 768 single nucleotide polymorphisms in 68 candidate genes. Allele and genotype frequencies were compared between the 120 participants who developed TESI and those who did not. Results: Two markers were significantly associated with TESI (marker 239, allele-wise p ¼ 7.84 10–5, OR ¼ 1.94, permutation p ¼ 0.01; marker 157, genotype-wise p ¼ 2.43 10–5, OR ¼ 8.23, permutation p ¼ 0.003). Both markers reside within genes involved in glutamate signaling. Taken together, these markers identified cases of TESI with a sensitivity of 57% and a specificity of 99.8% in this sample. Conclusion: Two genetic markers in the glutamate signaling pathway were associated with TESI during SSRI therapy. These markers may be useful in identifying patients at increased risk of this potentially dangerous adverse event. O22.4 INFLUENCE OF SEXUAL ABUSE ON SUICIDALITY: MODULATION BY POLYMORPHISMS IN THE 5-HTT AND THE TH1 GENE Alain Malafosse1, Isabelle Jaussent2, Fabrice Jollant3, Nader Perroud1, Patrick Baud1, Catherine Buresi1, Franck Bellivier2, Flavie Mathieu2, Marion Leboyer2, Philippe Courtet3 1 University Hospital of Geneva, Cheˆne Bourg, Gene`ve 2 INSERM, Montpellier 3 University Hospital of Montpellier, Montpellier 2 University Hospital of Cre´teil, Cre´teil A stress-diathesis model has been proposed for suicidal behaviour (SB). This suggests that individuals’ sensitivity to stressful events depends on their genetic makeup. Adults reporting childhood physical and sexual abuse experiences have a higher rate of SB and greater impulsivity. Recent meta- analyses yielded statistically significant association between the tryptophan hydroxylase gene (TPH1) and the serotonin transporter gene (5HTT) and SB. The goal of the present study was to determine whether the intron 7 A218C TPH1 and the promoter ins/del 5HTT polymorphisms moderate the influence of childhood trauma on SB. Using the Childhood Trauma Questionnaire we evaluated patients with (N ¼ 392) and without (N ¼ 51) a history of SB. We created a composite index of SB using the number of suicide attempts, the French versions of the Risk Rescue Rating Scale (RRRS), and of the Suicidal Intent Scale (SIS). Analyses show that sexual abuse is related with high scores of suicidality (OR ¼ 1.60, 95%CI [1.04– 2.45]). This association is even more important among individuals carrying two copies of the short allele of the 5-HTT promoter polymorphism (OR ¼ 3.10, 95%CI [1.14-8.46]) but not among those carrying one long allele. Similarly, sexual abuse is associated with high scores of suicidality among intron 7 A218C TPH1 heterozygotes (OR ¼ 2.51, 95%CI [1.32–4.67]) but not among both homozygotes. These results suggest that both TPH1 and 5HTT polymorphisms modulate the effect of sexual abuse on suicidality. O22.5 The MTHFR C677T POLYMORPHISM, MAJOR DEPRESSIVE DISORDER AND PERSONALITY TRAITS Daria Gaysina1, Sarah Cohen2, Farzana Hoda2, Ania Korszun3, Michael Owen4, Nick Craddock4, Ian Craig2, Anne Farmer2, Peter McGuffin2 1 Institute of Biochemistry and Genetics, Ufa 2 MRC, Social, Genetic & Developmental Psychiatry, London 3 Wolfson Institute of Preventative Medicine, London 4 University of Cardiff, Cardiff, Wales Biochemical studies indicate that folate deficiency increases risk of depression. The enzyme 5,10-methylentetrahydrofolate reductase (MTHFR) is responsible for the final step in the conversion of dietary forms of folate to 5-methyltetrahydrofolate. A SNP in the MTHFR gene (C677T) results in the production of a mildly dysfunctional thermo-
725
labile enzyme. The MTHFR 677T/T genotype is associated with significant elevation in the circulating concentrations of the homocystein and a decrease in serum folate concentrations, which may parallel a similar reduction in 5-methyltetrahydrofolate in the CNS and lead to a reduction in monoamine neurotransmitter function and elevated risk of MDD. Some previous studies have suggested an association between the MTHFR gene and MDD. A recent genome-wide linkage analysis for recurrent MDD has interesting evidence of linkage on chromosome 1p36, where the MTHFR gene is located. To test the hypothesis that MTHFR C677T polymorphism may be involved in predisposition to MDD and/or to special phenotypic subtypes based on personality trait measures, we conducted an association study on a sample of 1222 patients with MDD and 833 control subjects. There are no significant differences in genotype or allele frequencies between depressive patients and controls, neither in total samples, nor in females and males separately. No significant differences were obtained in scores of Neuroticism, Extraversion or Psychoticism EPQ scales between patient-carriers of the 677T allele and homozygotes of the 677C allele using GLM analysis. We are currently testing the locus in relation to other intermediate phenotypes of MDD. O22.6 RGS2, THE HUMAN ORTHOLOG OF A MOUSE ANXIETY GENE, IS ASSOCIATED WITH ANXIOUS TEMPERAMENT Jordan Smoller, Jes Fagerness, Shaun Purcell, Lesley Yamaki, Lisa Watras, Joseph Biederman, Pamela Sklar, Jerrold Rosenbaum Massachusetts General Hospital, MA Background: Behavioural inhibition to the unfamiliar (BI) is an anxiety-related temperamental profile that is a heritable risk factor for the development of panic and phobic anxiety disorders. Core behavioural and biological features of BI are closely paralleled by mouse models of unconditioned fear behaviour. The gene encoding regulator of G protein signaling 2 (Rgs2) was recently shown to be a quantitative trait gene influencing BI-like anxiety behaviour in mice. RGS2 accelerates the deactivation of G protein coupled receptor signaling in brain regions associated with anxiety. Objectives and Design: We examined whether variation at the human RGS2 locus is associated with BI in a family-based sample (N ¼ 120 families) of children who underwent laboratory-based behavioural assessments of temperament. Single marker and haplotype-based analyses of 15 single nucleotide polymorphisms (SNPs) encompassing RGS2 were performed using family-based association methods. Results: The 15 marker panel achieved a density of 1 SNP/2kb across a 32 kb region encompassing RGS2. Nine of the 15 markers were nominally associated with BI and three remained significant after Bonferroni correction. A haplotype comprising all 15 markers was strongly associated with BI (permuted p value ¼ 3 105). Conclusion: The mouse anxiety gene RGS2 is associated with shy, inhibited temperament (BI), a risk factor for anxiety disorders. Modulators of RGS2 may provide a novel therapeutic approach to the treatment of anxiety disorders. O22.7 SIGNIFICANT LINKAGE ON CHROMOSOME 14 FOR ANXIETY Christel Middeldorp1, Jouke Jan Hottenga1, P. Eline Slagboom2, Patrick F. Sullivan3, A. Leo Beem1, Gonneke Willemsen1, Dorret I. Boomsma1 1 Vrije Universiteit, Amsterdam 2 Leids Universitair Medisch Centrum, Leiden 3 University of North Carolina, Chapel Hill, NC A genome-wide linkage analysis was performed on anxiety, assessed with the Spielberger State-Trait Anxiety Inventory—trait version (1). These scores are closely associated with diagnoses of anxiety disorders. The sample consisted of 2488 subjects from 717 families, coming from the Netherlands Twin Register (NTR). Part of the sample was selected using an EDAC design. The subjects had participated at least once in a longitudinal survey study. Mean scores on 1 to 5 occasions were analysed. The large sample and quantitative measure makes it a relatively powerful design. Evidence for linkage was found on chromosome 14 on 105cM (LOD 2.1). A second LOD score of 1.4 was found on chromosome 16 on 104 cM. Recently, suggestive linkage at the same location on chromosome 14 was reported for a quantitative score
726
Abstracts
that reflected a common vulnerability for anxiety disorders (2), suggesting congruence with our results. In that study, families were ascertained through probands with panic disorder. Our sample was not based on a clinical sample, possibly indicating that the locus on chromosome 14 influences the risk of developing symptoms of anxiety in general and not only the risk of developing anxiety disorder. 1. Spielberger et al.: STAI Manual for the State-Trait Anxiety Inventory. Palo Alto, CA: Consulting Psychologists Press. 2. Kaabi B et al.: Genome scan for loci predisposing to anxiety disorders using a novel multivariate approach: strong evidence for a chromosome 4 risk locus. Am. J. Hum. Genet. 2006;78:543–553. O22.8 TWIN FAMILY STUDIES OF BORDERLINE PERSONALITY DISORDER IN THE NETHERLANDS, BELGIUM AND AUSTRALIA Dorret Boomsma1, Marijn Distel2, Nick Martin3, Gonneke Willemsen2, Catherine Derom3, Tim Trull4 1 Vrije Universiteit, Amsterdam 2 Vrije Univ, Amsterdam 3 QIMR, Brisbane 3 Katholieke Univ, Leuven 4 University of Missouri, Columbia There is ample evidence from twin and family studies that genetic factors influence most psychiatric diagnoses. Very little genetic research has been done on Borderline Personality Disorder. We carried out a twin-family study in 3 countries. A 24-item questionnaire (PAIBOR) was mailed to twins, sibs, parents and spouse of twins in The Netherlands and Belgium and to twins in Australia. We analysed the total sum scores on the 24 items. Completed questionnaires were returned by 8531 Dutch participants (3956 twins), 1662 Belgian participants (883 twins) and 600 twins from Australia. Total scores differed very little between men and women, or between younger and older subjects. Familial resemblance among twins and sibs for Borderline Personality Disorder indicated a heritability of around 45%, which did not differ as a function of age. A low, but significant, correlation between spouses was observed, associated with duration of marriage. O22.9 GENETIC PROPERTIES OF THE DEPRESSION OF ALZHEIMER DISEASE Michael Slifer, Margaret Pericak-Vance Duke University Medical Center, Durham, NC In 2002 the American Psychiatric Association published diagnostic criteria for depression of Alzheimer disease (DAD), recognising the phenomenological distinctiveness and clinical importance of a disorder affecting half of those suffering from Alzheimer disease (AD). However, little is known regarding the genetic properties of DAD. In this study, three independent datasets of AD patients (from the Collaborative Alzheimer Project and Duke Twins Study of Memory in Aging) are used to demonstrate a familial component to DAD. We observed significantly elevated sibling recurrence risk for DAD versus AD without depression among non-twin full siblings (n ¼ 96; Odds Ratio ¼ 8.3; p < 0.001), as well as monozygotic twins (n ¼ 38; Odds Ratio ¼ 25; p ¼ 0.006). Furthermore, including individuals with depression who had a nonAD etiology for their dementia decreased sibling risk, suggesting that DAD is specific to AD and not simply the consequence of cognitive disorders more generally. Moreover, including individuals with AD but depressive disorders that were unrelated to their AD also decreased sibling risk, illustrating that DAD does not appear to be the product of reducing the threshold for expression of conventional depressive disorders. In the independent dataset of unrelated individuals, we observed decreased allelic association between the apolipoproteinE4 susceptibility allele (ApoE4) and DAD (n ¼ 257; ApoE4 frequency ¼ 0.311) relative to AD without depression (n ¼ 297; ApoE4 frequency ¼ 0.458; p ¼ 0.001). Results concur with the two sibling datasets and may represent genetic heterogeneity between DAD and the parent disorder of AD. O22.10 BRAIN DERIVED NEUROTROPHIC FACTOR AND ITS RECEPTOR NTRK2 SHOW ASSOCIATION WITH OBSESSIVE-COMPULSIVE DISORDER Maria del Pino Alonso1, Mo´nica Grataco´s2, Juan R Gonza´lez2, Jose´ Manuel Mencho´n3, Rafael de Cid4, Cinto Segala`s3, Mo`nica Baye´s2, Javier Labad3, Julio Vallejo4, Xavier Estivill2
1
Hospital Universitario de Bellvitge, Hospitalet de Llobregat, Barce Genes and Disease Program and National G, Barcelona Department of Psychiatry. Hospital de Be, Hospitalet de Llobregat, Barcelona 4 National Genotyping Center (CeGen). Cent, Barcelona 4 Department of Psychiatry. Hospital de B, Hospitalet de Llobregat, Barcelona 2 3
Growth factors promote cell survival and neuronal differentiation and are known to be involved in synaptic efficiency and neuronal plasticity. Brain-derived neurotrophic factor (BDNF) belongs to the neurotrophine family and has been involved in several psychiatric disorders. To test the involvement of BDNF and its receptor (NTRK2) in obsessivecompulsive disorder (OCD), we performed an association study of 54 TagSNPs covering both genes in a sample of 134 patients (OCD Clinic of Bellvitge Hospital) and 342 controls. TagSNPs were selected from the HapMap project (Phase I data freeze, dbSNP b124) considering the genotypes corresponding to the 60 individuals from the CEPH-30-trios of European ancestry. Eight TagSNPs were selected that cover BDNF and 46 for coverage of NTRK2. We performed both single case-control association and haplotype analysis. No significant differences in genotype and allele distributions of BDNF, including the functional Val66Met variant, were detected between OCD patients and controls. A significant association of an intronic SNP in NTRK2 receptor gene and the presence of OCD was identified (p ¼ 0.001). When considering OCD family background, both a single SNP and a composite haplotype were identified for BDNF (p ¼ 0.0003 and p ¼ 0.01). Furthermore, an intronic SNP from the NTRK2 gene and OCD with positive OCD family history emerged (p ¼ 0.0002). Our findings suggest that NTRK2 may constitute a susceptibility gene for OCD, while BDNF involvement in OCD vulnerability appears to be limited to the subgroup of patients with positive family history of OCD.
O23.1 EVIDENCE IMPLICATING BRD1 WITH BRAIN DEVELOPMENT AND SUSCEPTIBILITY TO BOTH SCHIZOPHRENIA AND BIPOLAR AFFECTIVE DISORDER Jacob Severinsen1, Carsten R. Bjarkam2, Stine Kiær-Larsen3, Inger Marie Olsen4, Maria Mærsk Nielsen3, Jenny Blechingberg3, Anders L. Nielsen3, Ida E. Holm4, Leslie Foldager4, Bryan D. Young5, Walter J. Muir6, Douglas H. Blackwood6, Thomas J. Corydon3, Ole Mors4, Anders D. Børglum3 1 Institute of Human Genetics, Aarhus 2 Institute of Anatomy, University of Aarhus 3 Institute of Human Genetics, University of Aarhus 4 Centre for Basic Psychiatric Research, Aarhus 4 Department of Pathology, Aalborg Hosp, Aalborg 5 Cancer Research UK, Department of Med, London 6 Division of Psychiatry, University of Edinburgh Linkage studies suggest that chromosome 22q12–13 may contain one or more shared susceptibility genes for schizophrenia (SZ) and bipolar affective disorder (BPD). In a Faeroese sample we previously reported association between microsatellite markers located at 22q13.31-qtel and both disorders. The present study reports an association analysis across 5 genes (including 14 single nucleotide and two microsatellite polymorphisms) in this interval using a case-control sample of 162 BPD, 103 SZ patients and 200 controls. The bromodomain-containing 1 gene (BRD1), which encodes a putative regulator of transcription showed association with both disorders with minimal p-values of 0.0046 and 0.00001 for single marker and overall haplotype analysis, respectively. A specific BRD1 2-marker ‘‘risk’’ haplotype showed a frequency of 10% in the combined case group versus 1% in controls (p-value 2.8 10–7). Expression analysis of BRD1 mRNA revealed widespread expression in mammalian brain tissue which was substantiated by immunohistochemical detection of BRD1 protein in the nucleus, perikaryal cytosol and proximal dendrites of the neurons in the adult rat, rabbit and human CNS. Quantitative mRNA analysis in developing fetal pig brain revealed spatiotemporal differences with high expression at early embryonic stages, with intense nuclear and cytosolar immunohistochemical staining of the neuroepithelial layer and early neuroblasts, whilst more mature neurons at later embryonic stages had less nuclear staining. The results implicate BRD1 with SZ and BPD susceptibility and provide evidence that suggests a role for BRD1 in neurodevelopment.
Abstracts O23.2 MUTATION, ASSOCIATION, AND EXPRESSION STUDIES OF HUMAN NR2E1: A CANDIDATE FOR IMPULSIVE-AGGRESSIVE BEHAVIOURS, BIPOLAR DISORDER, AND SCHIZOPHRENIA Ravinesh A. Kumar1, Stephen Leach2, Russell Bonaguro1, Alan McLean2, Brett S. Abrahams1, Emil Coccarro2, Sheilagh Hodgins3, Gustavo Turecki4, Walter Muir2, Douglas Blackwood2, Angela R. Brooks-Wilson2, Elizabeth M. Simpson1 1 University of British Columbia, Vancouver 2 Canada’s Michael Smith Genome Sciences C, Vancouver, British Columbia 2 The Royal Edinburgh Hospital, Edinburgh 2 University of Chicago, Chicago, Illinois 3 King’s College London, Institute of Psychiatry, London 4 Douglas Hospital Research Centre, Montreal, Quebec Nr2e1/ mice (fierce) have forebrain hypoplasia and are pathologically violent. We have corrected the fierce phenotype using a human genomic clone of NR2E1, demonstrating functional protein and regulatory equivalency between these orthologs in a mouse. Notably, NR2E1 maps to a putative 6q21 loci for bipolar disorder and schizophrenia. Thus, NR2E1 is a functional and positional candidate for behavioural and psychiatric disorders. To test the hypothesis that some humans with impulsive-aggressive behaviours, bipolar disorder, or schizophrenia may have mutations in NR2E1, we sequenced the complete coding, 50 and 30 untranslated, splice-site, proximal promoter, and evolutionary-conserved noncoding regions in 126 humans with features resembling Nr2e1/mice, including families showing genetic linkage to the NR2E1 locus. We did not detect any coding region changes; however, we identified 11 candidate regulatory mutations that were absent from 160 controls. We used in silico tools to demonstrate that seven candidate mutations were predicted to alter known neural transcription factor binding sites. We also performed a case-control association analyses in 295 –359 patients with bipolar disorder, 309 – 382 schizophrenia, and 270– 402 ethnically-matched controls using exonic, promoter, and UTR markers. We did not detect any significant associations. Finally, we examined NR2E1 expression in normal human adult brain, and detected expression in cortex, occipital lobe, frontal lobe, temporal lobe, and putamen. We conclude that protein-coding changes in NR2E1 do not contribute to impulsive-aggressive behaviours, bipolar disorder, and schizophrenia in the subjects examined here; however regulatory mutations may play a role.
O23.3 GENETIC VARIATION OF SEROTONIN RECEPTORS PREDICTS PSYCHIC ANXIETY IN SCHIZOPHRENIA Monsheel Sodhi1, Lisa Hazelwood2, Chun Li1, David Airey1, Elaine Sanders-Bush1, Herbert Y. Meltzer1 1 Vanderbilt University, Nashville TN, 2 NINDS, Rockville, MD Epidemiological studies indicate a high incidence of anxiety disorders and depression in schizophrenia. Accumulating pharmacological and genetic association data show that the genes encoding 5-HT1A, 5HT2A and 5-HT2C receptors (R) are strong candidates for investigation. Therefore we hypothesized that SNPs in these genes would be associated with traits of anxiety and depression in schizophrenia. We examined 2 samples of Caucasian American schizophrenia patients (n ¼ 81 and n ¼ 185) assessed for severity of psychic anxiety, somatic anxiety, depressed mood and guilt using the Brief Psychiatric Rating Scale (BPRS). While the 5-HT1A (-1019 C/G), 5-HT2A (102 T/C) and 5-HT2C (759 C/T, 697 C/G, 23 cys/ser) SNPs were uninformative, the 5-HT2A 452 His/Tyr change was associated with psychic anxiety but not somatic anxiety in both schizophrenia samples (overall ordinal logistic regression ¼ 10.4, 2df, p ¼ 0.005). Subjects with the Tyr452 allele(Tyr/Tyr and His/Tyr) had lower mean psychic anxiety scores (1.8þ/1.5)compared with His/His subjects (2.6þ/1.5). Previous in vitro studies show that the 5-HT2A 452Tyr variant receptor is more rapidly desensitised than the 5-HT2A 452His variant, so subjects with the Tyr452 allele are predicted to have a less active 5-HT2A receptor with less anxiogenic effects on stimulation. Therefore we conclude that 5-HT2A 452 His/Tyr variation could play an important role in the development of psychic anxiety in schizophrenia.
727
O23.4 THE FIRST ASSOCIATION STUDY BETWEEN G72/G30 AND UNIPOLAR DEPRESSION IN A LARGE SAMPLE OF PATIENTS AND CONTROLS OF GERMAN DESCENT Thomas G. Schulze1, Alexander Georgi1, Frederike Schirmbeck1, Anna Karpushova2, Susanne Hoefels3, Rami Abou Jamra4, Johannes Schumacher4, Wolfgang Maier3, Lars Beckmann4, Peter Propping4, Markus M. No¨then2, Sven Cichon2, Marcella Rietschel1 1 Central Institute of Mental Health, Mannheim 2 Dpt. of Genomics, Life & Brain Center, U, Bonn 3 Dpt. of Psychiatry, University of Bonn 4 Inst. of Human Genetics, University of Bonn 4 German Cancer Research Center, Heidelberg G72/G30 is considered a strong susceptibility gene for both schizophrenia (SZ) and bipolar disorder (BD). We have recently reported association between identical G72/G30-haplotypes and SZ, BD and panic disorder (PD). An association study on major depression has not yet been conducted. We studied a German sample of 500 MD patients and 1030 population-based controls. We were interested in whether our previously-identified risk haplotype of markers M22, M23, and M24 was also associated with MD. To further explore any relationship between G72/G30 and MD, we genotyped 9 additional SNPs highlighted in other studies. The haplotype C-C-T was significantly more frequent in MD patients than in controls (40.5 % vs. 36.1 %; p ¼ 0.027; OR ¼ 1.2). Exploratory analysis on 9 further G72/G30 SNPs of interest yielded significant associations for M12 (p ¼ 0.038) and M14 (p ¼ 0.046), while M15 (p ¼ 0.090) and rs1935062 (p ¼ 0.055) did not reach significance. Permutation analysis adjusting for the 9 SNPs yielded a global p ¼ 0.176. This is largest case-control study on G72/G30 and MD to date. We found an association between MD and the same risk haploptype that we had previously found associated with SZ, BD or PD. We also found indicative associations with other variants. Given that depressive symptoms are present across these diagnostic groups, G72/ G30 might predispose to ‘‘core symptoms’’ prevalent in all four disorders. Further studies will be needed to elucidate these core symptoms. O23.5 TDT AND ASSOCIATION ANALYSIS OF HYPNOTIZABILITY AND RELATED ATTENTIONAL CHARACTERISTICS Maria Sasvari-Szekely Med.Chem.Mol.Biol.Path., Semmelweis Univ, Budapest Introduction: Hypnotizability is a stable, trait-like characteristic with a considerable inherited component, and is related to higher cognitive functions. Recent studies demonstrate importance of an enzyme involved in dopamine metabolism (COMT) in responsiveness to hypnosis. Our previous results showed that the DRD4 VNTR 7-repeat allele was associated with lower ‘Persistence’ scores using the TCI selfreport measure of temperament, and generally slower overall reaction time performance in cognitive tasks. Our present results identify genetic components of hypnotizability and related attentional characteristics. Methods: The Waterloo-Stanford Group C (WSGC) was used to test hypnotic susceptibility and candidate genetic polymorphisms of the dopaminergic and serotoninergic system. Personality and cognitive characteristics were portrayed based on the TCI personality questionnaire and cognitive tests of attentional performance in a sample of 110 healthy Hungarian young adults and their families. Results: Associations found in population-based association analyses were attested by TDT analysis. The role of the 7-repeat allele was reassured with relation to overall cognitive performance. Conclusions: The present results provide further support that dopaminergic polymorphisms contribute to personality traits as well as laboratory measures of attention. Our results suggest that genetic variations in the coding region of the dopamine D4 play an important role in mediating cognitive performance in maintaining focused attention across time. Results are consistent with a role for this gene in disorders such as ADHD that are characterized by poor performance in neuropsychological tests of executive function. O23.6 GENE ENVIRONMENT INTERACTION AND PERSONALITY/BEHAVIOUR—MARKED SEX DIFFERENCES Lars Oreland, J. Hallman, K. W. Nilsson, R. L. Sjoberg, J. Leppert Uppsala University, Uppsala
728
Abstracts
The MAO-A and 5-HTT genes are two key players in the regulation of serotonin turnover within the brain. Both have functional variations in the promoter regions (Sabol et al., 1998; Lesch et al., 1996). AP-2ß is a transcription factor of importance for development and maintenance of e.g. the monoamine brain-stem nuclei. An AP2ß gene polymorphism has been shown to be associated with personality and monoamine turnover (see Damberg, 2005). The results from a stratified sample of 6000 adolescents show that in boys the short MAOALPR allele interacted with psychosocial factors for risk for criminal activity (Nilsson et al., 2005) and destructive behaviour when drunk. In girls a similar result was associated with the long MAO-A allele (Sjoberg et al., 2006). The short 5-HTTLPR allele was protective against depressive symptoms in boys, while in girls this allele interacted with bad family relations to increase the risk for depressive symptoms (Sjoberg et al., 2005). AP-2ß genotype interacted with psychosocial environment for risk for antisocial behaviour in boys. In girls such an interaction was found for depressive symptoms. With regard to personality, using Cloninger’s TCI, both 5HTTLPR and AP2ß genotypes were mainly associated with the Self Transcendence subscale Spiritual Acceptance, in an additive way among boys and interactively among girls. In conclusion, genotype and a wide variety of psychosocial factors interact significantly, to precipitate a variety of behavioural and psychiatric disorders. Distinct sex differences are at hand in this regard. O23.7 THE GENE FOR INTERLEUKIN 6: THE CANDIDATE GENE OF NEUROPSYCHIATRIC ILLNESS? Omar Sery1, Radim Sˇtaif2, Rena´ta Pitelova´2, Jan Lochman2, Vladimı´r Mikesˇ2, Ivana Drtı´lkova´3, Petr Zvolsky´4, William Didden5 1 Institute of Biochemistry, Brno, Czech Republic 2 Institute of Biochemistry, Faculty of Science, Brno, Czech Republic 3 Department of Psychiatry, Faculty of Medicine, Brno, Czech Republic 4 Department of Psychiatry, Ist Medical Fa, Praha, Czech Republic 5 GHC Research Institute, Houston, Texas IL-6 plays the role as a physiological neuromodulator involved in dopaminergic, serotonergic and other neurotransmissions. Genetic variation (-174 G/C polymorphism) in the promoter region of the IL-6 gene influences the expression of IL-6. Our objective was to determine whether 174 G/C polymorphism of IL-6 is associated with alcoholism and ADHD. In our two independent association studies, we investigated the relationship between the above-mentioned polymorphism and alcoholism in 400 alcoholics and 400 controls, and in an ADHD study on 49 hyperactive boys and 40 control boys. We found a statistically significant difference (p < 0.03) in allelic frequencies between alcoholics and controls and a statistically significant difference (p < 0.001) between controls and ADHD subjects using Fisher exact test. The C allele was observed more frequently in alcoholics, as well as in the ADHD group of boys as compared with control subjects. Our association studies provided new data on the relationship between 174 G/C polymorphism of IL-6 gene and alcoholism/ADHD. These findings highlight the need for further investigation of the influence of cytokine genes on the pathogenesis of various psychiatric disorders. In accordance with recent in vitro experiments, we expect to find a relationship between the IL-6 polymorphism studied and the embryonal development of specific human brain structures. O23.8 A FLEXIBLE SNP GENOTYPING SYSTEM TO STUDY ALLELIC DETERMINANTS OF BRAIN FUNCTION Paola Iacopetti, Irene Tarantino, Arianna Cozza, Samuele Santi, Graziana Savarino, Stefano Fogli, Pietro Pietrini University of Pisa, Pisa We developed a system for SNP-genotyping consisting of two ‘‘microchips’’, one investigating 40 SNPs in genes involved in the hepatic metabolism of antidepressant drugs, the other 100 SNPs in genes involved in brain response to antidepressants. Candidate genes were chosen by literature screening. SNPs were selected on HapMap data, using the pairwise algorithm combined with the ‘‘Best N’’ method; a few SNPs, reported as associated with treatment outcome, were added to the set. The main advantages of our experimental design are: a) simultaneous analysis of several genes (22) involved in distinct aspects of antidepressant response; b) a comprehensive search for common variability in candidate genes by HapMap ‘‘tag-SNPs’’. DNA samples are genotyped by ‘‘padlock probes’’, which specifically recognise the two variants of each SNP; padlock probes are amplified with universal
primers, and contain unique tags for hybridisation to known positions on ad hoc designed micro-arrays, where the two alleles are distinguished by distinct fluorophores. Since reactions carried out on multiplex and signal intensities on microarrays are homogeneous, as they are not affected by SNP flanking sequences, tens of SNPs from tens of individuals can be analysed simultaneously on a single slide, with high accuracy and reproducibility. Considering that 100 SNPs from many genes related to neural transmission, plasticity and neurogenesis are included in the second microchip and that each of them can easily be substituted with others if needed, our system represents a powerful tool for investigating genetic influence on normal and pathological brain function. O23.9 HOW MULTIPLE STATISTICAL COMPARISONS CAN PRODUCE FALSE POSITIVE FINDINGS Patrick Sullivan UNC/Genetics, Chapel Hill Psychiatric geneticists who conduct association studies are almost always aware of the problem of multiple comparisons to produce false positive findings. However, the impact may be far more damaging that is generally appreciated. To explore this issue, we used realistic simulations. Using ‘‘HapSample’’, a program written by Dr. Fred Wright and based on HapMap genotypes, we used resampling methods to create 1,000 association samples consisting of 500 cases and 500 controls. Genotypes were generated for 10 COMT functional and htSNPs under a null model. LD patterns were identical to those from HapMap. All association samples were analyzed for a typical suite of analyses - main effect, main effects stratified by sex, haplotype analyses using a sliding window, and five endophenotypes available in cases only. All data were generated under a null model - ie, any positive findings are by definition false positives. Findings: 96.8% of the studies produced significant finding at p < 0.05 with a median of 6 false positives. All positive studies were ‘‘replicated’’ 10 times. If the definition of replication was lax (any SNP, any phenotype/test), all studies ‘‘replicated at least 7 times and 67% replicated 10/10 times. This pattern of findings is uncomfortably similar to that for many candidate gene studies in neuropsychiatry. Ways to avoid false positives and to avoid the propagation of false positives are shown. O23.10 TWIN PAIRS ASCERTAINED FOR PSYCHOTIC DISORDERS REVEAL ASSOCIATION OF AN AKT1 ALLELE WITH VERBAL LEARNING AND MEMORY AND GREY MATTER INTENSITY OF AFFECTED BRAIN REGIONS Olli Pietila¨inen1, Tiina Paunio2, Anu Loukola2, Annamari Tuulio-Henriksson2, Tuula Kieseppa¨2, Pia Soronen2, William Hennah2, Joni A. Turunen2, Juho O. Peltonen2, Outi M. Palo2, Kaisa Silander2, Jouko Lo¨nnqvist2, Jaakko Kaprio2, Tyrone D. Cannon3, Leena Peltonen2 1 National Public Health Institute, Helsinki 2 National Public Health Institute, Depart, Helsinki 3 UCLA, Departments of Psychology and Psychiatry, Los Angeles Several neurocognitive functions are affected in schizophrenia (SCZ) and bipolar disorder (BPD). Cognitive abilities represent assessable, quantitative traits, providing more information for genetic analyses than end state diagnoses and may be more directly associated with the underlying biology. Modern imaging techniques provide tools to measure alterations in volume or in white/gray matter composition of specific brain regions, related to impairments of cognitive functions or diagnostic entities. Twins represent ideal study subjects to address the genetic background of complex neurobehavioral traits since they share the prenatal and infancy environment. We tested variables measuring neurocognitive functions in a Finnish twin sample consisting of 59 SCZ twin pairs (8 concordant, 51 discordant), 20 BPD twin pairs (5 concordant, 15 discordant) and 62 control twin pairs matched for age, sex and demographics. We genotyped 3 SNPs in AKT1 and assessed their association with cognitive functions and detailed 3D brain MRI images (48 SCZ, 47 control twin pairs (190 individuals)). We used linear regression with age, gender, psychosis, zygosity, and the co-twin being affected as covariates. Three neuropsychological variables, related to verbal learning and memory showed association with an AKT allele defined by rs1130214 (p < .006). Importantly, this allele also associated with decreased gray matter density in medial and dorsolateral prefrontal cortex (p < .05), an association that was confirmed through permutation analysis. Our study further highlights the importance of
Abstracts quantitative neurocognitive features as endophenotypes for SCZ and BPD and imply the role of AKT as one of the genetic determinants of them. O24.1 INTERACTION OF SEROTONIN-TRANSPORTER VARIANTS AND ADVERSE CHILDHOOD ENVIRONMENT IN ADHD Christine Freitag1, Wolfgang Retz2, Petra Retz-Junginger2, Marc Schneider2, Denise Wenzler2, Christian Kissling3, Johannes Thome3, Michael Ro¨sler2 1 Saarland University Hospital, Homburg 2 Institute for Forensic Psychology and Psychiatry, Homburg 3 Department for Psychiatry, The School of Medicine, Swansea Although attention-deficit/hyperactivity disorder(ADHD)is highly inheritable, environmental conditions play an important role in its manifestation during childhood development. Here, we report the results of an investigation on the interaction of an adverse childhood environment with a functional polymorphism of the serotonin transporter promoter gene (5-HTTLPR) and its impact on ADHD psychopathology in young adult delinquents. Method: 184 male delinquents were assessed for childhood and current ADHD and adverse childhood environment using standardised instruments. Each subject was genotyped for 5-HTTLPR long (L) and small (S) alleles. Results: Logistic regression analysis revealed independent effects of high childhood environmental adversity and the 5-HTTLPR LLgenotype on childhood ADHD symptoms. In addition, a significant gene for environmental interaction was found, indicating that carriers of at least one 5-HTTLPR short allele are more sensitive to childhood environmental diversity than carriers of the LL-genotype. Discussion: The results confirm prior findings of an association between an ADHD and 5-HTTLPR LL-genotype and an adverse childhood environment and underline the need for further investigation of genes as regards environmental interaction with respect to ADH. O24.2 COMT AND MAOA GENE VARIANTS ARE ASSOCIATED WITH PSYCHOPATHY IN ADOLESCENTS WITH ADHD Tom Alan Fowler, Kate Langley, Darko Turic, Naureen Whittinger, Frances Rice, Marianne van den Bree, Michael J Owen, Michael O’Donovan, Anita Thapar Cardiff University, Cardiff Antisocial behaviour (ASB)in ADHD indexes a more strongly genetic and neurobiological phenotype than ASB alone. Psychopathy is an aspect of ASB considered to define a more homogenous phenotype and has strongly biological origins, especially the ‘affective’ (emotional low responsiveness) component. Prefrontal cortex (PFC) and amygdala function have been implicated. COMT and MAO-A genes encode enzymes that metabolise neurotransmitters, COMT particularly being linked with PFC functioning.Both genes have functional variants found to be associated with ASB. We aimed to test whether the COMT Val158Met polymorphism and MAO-A 30bp VNTR are associated with psychopathy in adolescents with childhood ADHD. Children with ADHD were followed-up 5 years after participation in a genetic study (n ¼ 156, participation rate ¼ 79%). Psychopathic traits were assessed using the Hare PCL:YV. For analysis, MAO-A 30-bp VNTR high versus low activity alleles and COMT Val158Met VAL/VAL versus VAL/MET and MET/MET alleles were considered. Higher ‘affective ‘ psychopathy scores were associated with possession of the COMT VAL/VAL genotype (ß ¼ 0.23, p ¼ 0.03) and low activity MAO-A alleles (ß ¼ 0.20, p ¼ 0.04). MAO-A was also associated with total psychopathy scores (ß ¼ 0.27, p ¼ 0.02). Although these genes have been previously linked with DSM-IV conduct disorder (CD) symptoms, the associations with psychopathy were not entirely explained by associated CD. To our knowledge, this is the first study to report associations of these gene variants with psychopathic traits in ADHD. O24.3 EVIDENCE FOR ASSOCIATION OF THE GENE KIAA0319-LIKE (KIAA0319L) ON CHROMOSOME 1P TO READING DISABILITIES Jillian M Couto1, Lisette Gomez1, Karen Wigg1, Tasha Cate-Carter2, Jennifer Archibald2, Barbara Anderson2, Rosemary Tannock2, Maureen W Lovett2, Tom Humphries2, Cathy L. Barr3
729
1
Toronto Western Research Institute, Toronto, Ontario The Hospital for Sick Children, Toronto, ON Toronto Western Research Institute, Toronto, ON
2 3
Developmental dyslexia or specific reading disabilities (RD) is a learning disability characterised by difficulties with fluent word recognition and by poor spelling and decoding abilities. These result from deficits in the phonological components of language, and can occur despite at least average intelligence and effective classroom instruction. Many studies have found linkage and/or association of RD to several chromosomal regions. A locus on chromosome 1p34–36 has been linked to RD in three independent samples. Within this region we have identified a gene KIAA0319Like (KIAA0319L) which is an interesting candidate based on its position, and its homology to KIAA0319. The latter is a candidate for RD susceptibility on chromosome6p in a number of independent association studies. In this study, association of KIAA0319L was assessed using 5 tagging SNPs in a sample of 263 nuclear families ascertained through a proband with reading difficulties. Significant evidence of association was found for both a single marker (p ¼ 0.009; corrected p ¼ 0.035), and a haplotype (p ¼ 0.017) to RD defined as a categorical trait. This haplotype showed a significant association to the quantitative measure of word reading efficiency and strong trends towards association to spelling and rapid naming of objects and colours. We conclude that KIAA0319L is likely the RD susceptibility gene on chromosome 1p and contributes to multiple reading components. These results also strengthen the case for the homolog KIAA0319 as the susceptibility gene on 6p. A screen for functional DNA changes in the gene is underway. O24.4 INVESTIGATION OF THE GENES ENCODING ARGININE VASOPRESSIN (AVP) AND ITS RECEPTOR (AVPR1B) IN CHILDHOOD-ONSET MOOD DISORDER Emma Dempster1, Irina Burcescu1, Karen Wigg1, Eniko Kiss2, ´ gnes Vetro´2, Jim Kennedy3, Kovacs Marika4, Ildiko´ Baji3, A Cathy Barr1 1 Toronto Western Research Institute, Toronto 2 Department for Child and Adolescent Psyc, Szeged 3 Vadaskert Hospital, Budapest 3 Neurogenetics Section, Centre for Addiction, Toronto 4 University of Pittsburgh School of Medicine, Pittsburgh Disturbances in stress hormones have been identified in depression, in particular in the hyperactivity of the hypothalamic-pituitary-adrenal (HPA)-axis. Arginine vasopressin (AVP) plays a crucial role in modulating the HPA-axis, and does so through a G-protein receptor, AVPR1B. To determine if genetic variation in AVP or AVPR1B could be contributing to vulnerability to depression, we genotyped genetic markers (SNPs) across both genes in a large family-based sample of childhood-onset mood disorder (COMD). Six SNPs were genotyped across AVPR1B, 2 were novel non-synonymous polymorphisms and the further 4 were constituents of a haplotype that has previously been shown to be protective against depression. Three SNPs were chosen to capture the majority of genetic variations in AVP. The sample was comprised of 394 Hungarian family-trios ascertained through affected probands with a diagnosis of COMD. Two of the AVPR1B SNPs showed significant association (Lys65Asn:p ¼ 0.007; SNP5:p ¼ 0.027). Haplotype analysis showed the significant over-transmission of the most frequent 6 marker haplotype (p ¼ 0.00011). However, the AVP SNPs showed no association with COMD either alone or as a haplotype. The results of this study suggest the involvement of AVPR1B, although not AVP, in the pathogenesis of depression. This finding is of interest, as avpr1b mice knockouts have impaired stress-induced catecholamine release, implying that this receptor regulates catecholamine stress response. Thus, genetic variation in AVPR1B may have implications on HPA-axis dysregulation in depression. O24.5 A CASE-CONTROL STUDY OF THE RELATIONSHIP BETWEEN 13 CANDIDATE GENES AND ADHD Radim Sˇtaif1, Omar Sˇery´2, Rena´ta Pitelova´2, Jan Lochman2, Vladimı´r Mikesˇ2, Ivana Drtı´lkova´3, Pavel Theiner3, William Didden4 1 Institute of Biochemistry, Brno, Czech Republic 2 Institute of Biochemistry, Faculty of Science, Brno, Czech Republic 3 Department of Psychiatry, Faculty of Medicine, Brno, Czech Republic 4 GHC Research Institute, Houston, Texas
730
Abstracts
Attention deficit hyperactivity disorder (ADHD) is a multifactorial disorder with a significant genetic contribution. Dopamine genes have been the initial candidates, based on the site of action of psychostimulant drugs. Other genes from neurotransmitters and the immune system have been focused on as well. We have investigated whether there is an association between ADHD and polymorphisms of 13 candidate genes: DAT1, DRD2, COMT, MAOB, ACE, ANG, CCR5, BDNF, MU, IL-2, IL-6, TNF-alpha and M235T. The aim of our casecontrol study was to determine whether ADHD boys (118) have different genotype and allelic frequencies of examined polymorphisms of the above-mentioned genes as compared to controls (153). No significant differences were found between groups of boys in the allelic and genotype frequencies of COMT, MAOB, ACE, ANG, CCR5, BDNF, MU, IL-2, IL-6, TNF-alpha and M235T genes (p < 0.05). For DAT1, statistical analysis using Fisher exact test did not confirm the positive association of the previously-investigated 10-repeat allele or 10/10 genotype of the DAT1 gene (p ¼ 0.531). For DRD2, ADHD probands were more likely than controls to have the A1 allele and genotype A1A1 (p ¼ 0.0013). Our results showed that ADHD in Czech children is associated with TaqI A polymorphism of the DRD2 gene. The presence of the DRD2 A1 allele increases the risk of ADHD.
the clinical phenotype of ADHD. Previous neuropsychological studies have shown inconsistent associations between the 7-r allele and performance on tasks designed to assay sustained attention in children with ADHD. We investigated the performance of children with (n ¼ 61) and without (n ¼ 48) ADHD on the Sustained Attention to Response Task (SART). We grouped all children into either the 7r-present group (possessing at least one copy of the 7r allele) or the 7r-absent group. Using a step-wise linear regression, we examined the variance associated with clinical diagnosis and IQ and then the remaining variance associated with possession of the risk variant (7r). The ADHD group made significantly more Commission and Omission errors than the Control group, but no significant genotype effects were found on these measures. The ADHD group was significantly more variable in response time (RT) than the Control group, in terms of standard deviation of RT, fast- and slow-frequency variability. After controlling for clinical group, participants in the 7r-absent group were significantly more variable than those with at least one copy of the 7r allele, for the SD of RT, the fast- and slow- frequency variability measures. These results, found in both the Random and Fixed SARTs, suggest that the DRD4 VNTR is associated with variability in RT in ADHD and control children.
O24.6 MUTATION-SCREEN AND ASSOCIATION ANALYSIS OF THE GLUCOCORTICOID RECEPTOR GENE (NR3C1) IN CHILDHOOD-ONSET MOOD DISORDERS (COMD) ´ gnes Vetro´3, Jonathan Mill1, Karen Wigg2, Irina Burcescu2, A E. Kiss3, K. Kapornai3, Zsuzsa Tama´s4, Ildi Baji4, Julia Ga´doros4, James Kennedy1, Maria Kovacs2, Cathy Barr2 1 Centre for Addiction & Mental Health, Toronto 2 Toronto Western Research Institute, Toronto, Ontario 3 Department for Child and Adolescent Psyc, Szeged 4 Vadaskert Hospital, Budapest 2 School of Medicine, University of Pittsburgh, Pennsylvania
O24.8 ASSOCIATION OF NEUROTROPHIC TYROSINE KINASE RECEPTOR TYPE 3GENE IN CHILDHOOD-ONSET MOOD DISORDER Cathy Barr1, Yu Feng1, Karen Wigg1, Nicole King2, Agnes Vetro3, Eniko Kiss3, Gabriella Daro´czy3, Krisztina Kapornai3, James Kennedy2, Maria Kovacs3 1 The Toronto Western Hospital, Toronto 2 Centre for Addiction and Mental Health, Toronto 3 Szeged University, Szeged 3 University of Pittsburgh School of Medicine, Pittsburgh
Background: Glucocorticoid-mediated feedback inhibition is impaired in individuals with depression and stress-related psychopathology. Given the moderate heritability of these disorders, it is plausible that polymorphisms in the glucocorticoid receptor (GR) gene act to increase susceptibility. Maternal behaviour in rats epigenetically alters a NGF1-A transcription factor binding-site in the promoter region of the GR gene, providing a mechanism by which environmental cues can regulate GR expression. The analogous region of the human GR gene (NR3C1) has not been studied, but it is possible that polymorphisms in this region may disrupt binding of transcription factors known to regulate GR expression. Methods: We first examined SNPs across the coding-region of NR3C1 in a large sample of 394 Hungarian family-trios ascertained through affected probands with a diagnosis of COMD. Bioinformatic analyses were performed on the NR3C1 promoter-region to identify conserved promoter sequences, and these regions were screened for novel polymorphisms. Results: Initial genotyping provides some evidence that NR3C1 is involved in the aetiology of COMD, and additional markers are being typed. Bioinformatic analysis identified a NGF1-A transcription factor binding-site in a conserved region of the promoter. A number of rare polymorphisms were identified in this region that are promising candidates for a role in depression. Conclusion: A number of novel polymorphisms were identified in the promoter-region of NR3C1. Preliminary studies on the gene putatively support its role in the aetiology of COMD, although further studies are needed before firm conclusions can be drawn. O24.7 ASSOCIATION BETWEEN DRD4 7R VNTR AND RESPONSE VARIABILITY IN CHILDREN WITH AND WITHOUT ADHD Katherine A Johnson1, Ian H. Robertson2, Simon P. Kelly3, Edwina Barry4, Aoife Da´ibhis2, Michael Daly2, David Lambert4, Ziarih Hawi4, Michael Gill4, Mark A. Bellgrove5 1 Trinity College Dublin, Dublin 2 Institute of Neuroscience, Trinity College, Dublin 3 Cognitive Neurophysiology Laboratory, Na, Orangeburg, New York 4 School of Medicine and Health Sciences, Dublin 5 School of Behavioural Science, Melbourne, Dublin Genetics studies have consistently shown an association between the 7repeat (7r) allele of a 48-base pair VNTR in exon 3 of the DRD4 gene and
Major depressive disorder (MDD) is multifactorial disorder with moderate heritability. Family and twin studies show the highest relative risk for MDD in families of probands with early age at onset and/or recurrent episodes compared with the risk in the general population. Chronic antidepressant treatment up-regulates the neurotrophin signaling pathways which is involved in neural plasticity and survival and deficits in neural plasticity have been suggested to underlie the development of depression. A genome scan of families multiply affected with recurrent, early onset MDD by Holmans et al. (2004) revealed significant linkage on chromosome 15q25.3-q26.2 (LOD ¼ 3.73). One candidate gene located in this 15q region, NTRK3 (neurotrophic tyrosine kinase, receptor type 3) was particularly interesting because NTRK3 is a key gene in neural plasticity. We investigated the relationship of this gene to childhood-onset mood disorders (onset before age 14) in a sample of 394 families with 459 affected children recruited from 21 mental health facilities across Hungary. The results of family based association studies of the NTRK3 gene show a significant association with 3 markers in our sample (p < 0.035). Screening of the exons flanking these markers is currently in progress. O24.9 PUTATIVE IDENTIFICATION OF SUSCEPTIBILITY GENES FOR AUTISM ON 15q11–q13: ROLE OF UBE3A AND ATP10A Guia Guffanti Masetti1, MT Bonati2, Alessandra Gessi1, Florinda Cavalleri2, Francesca Cogliati2, Margherita Marchi2, Luisa Strik Lievers3, Ernesto Caffo3, Lidia Larizza1, Fabio Macciardi1, Silvia Russo2 1 Universita` degli Studi di Milano, Segrate 2 Istituto Italiano Auxologico, Milano 3 Universita` di Modena, Modena The etiology of autism is still largely unknown despite our current understanding from family and twin studies that genetics play a substantial role in the etiology of the disorder. Moreover, integrating data from linkage studies and analyses of chromosomal abnormalities allow identification of 15q11–q13 as one of the regions of particular etiopathogenetic interest for autism and autism related disorders. In an effort to find the autism susceptibility genes potentially harbored in this chromosomal region we have screened a set of markers spanning two known imprinted, maternally expressed genes, UBE3A and ATP10A, selected because they are both positional and candidate genes. We replicated evidence of Linkage Disequilibrium at marker
Abstracts 0
D15S122, located at the 5 end of UBE3A and originally reported by Nurmi (2001). In addition, our analyses also show one significant haplotype that includes D15S122 at UBE3A and SNP3 and D15S1534 at ATP10A. These findings are of particular interest considering that the association of D15S122 has not previously been replicated to date and that UBE3A is the gene responsible for the Angelmann Syndrome, that shares neurological and behavioural abnormalities with the autism spectrum disorders. Despite the limited power to detect genes of minor effect with a low density SNPs, our data support a potential role of UBE3A in the complex pathogenic mechanisms of autism. To strengthen our findings, we are currently genotyping a denser set of SNPs across the region and using a larger sample, with the ultimate goal of identifying the specific polymorphism(s) responsible for the association. O25.1 SCHIZOPHRENIA GENES CO-LOCALIZE TO MULTIPLE, INTERSPERSED FRAGILE SITE REGIONS Cassandra Smith1, Hamid Abdolmaleky1, Andrew Bolton1, Giang Nguyen1, Nam Nguyen1, Charles Kaufman2 1 Boston University, Boston 2 Columbia Unviersity, NYC While many genes (and chromosomal regions) are linked to schizophrenia, no single gene or small number of genes is present in the majority of patients. This research tested the possibility that genes linked to schizophrenia occur in of the 119 interspersed unusual genomic regions called fragile sites. Fragile site are prone to mutation, especially sensitive DNA replication inhibitors and associated with genomic remodeling, plasticity and instability. Many genes within fragile sites are regulated epigenetically, and these regions are frequent sites for deletions, translocations, retroviral insertions etc. The sequence of only 15 fragile sites is known, and some, but not all contain simple or complex repeating sequences. The relationship between schizophrenia and fragile sites was explored by comparing the chromosomal addresses of fragile sites and over 200 genes and chromosomal abnormalities linked to schizophrenia. A larger than expected number of genes and chromosomal abnormalities linked to schizophrenia in one or more families co-localize with fragile sites (Chi Square, P ¼ 0.001). Negative controls (all human genes listed in Genbank, or genes tested but not found to be associated with schizophrenia) were not linked to fragile sites. These results link schizophrenia to multiple, interspersed fragile sites and support our hypothesis that schizophrenia is linked to genomic instability. We speculate that DNA changes at fragile sites can accounts for the seemingly disparate biological and environmental factors that are linked to this disease. O25.2 TOWARDS UNDERSTANDING THE SCHIZOPHRENIA CODE: AN EXPANDED CONVERGENT FUNCTIONAL GENOMICS APPROACH Alexander B. Niculescu Indiana University School of Medicine, Indianapolis, IN 46202 Identifying genes for schizophrenia through classical genetic approaches has proven arduous. We present a comprehensive convergent analysis that translationally integrates brain gene expression data from a relevant pharmacogenomic mouse model (involving treatments with a psychomimetic agent—phencyclidine (PCP), and an anti-psychotic—Clozapine), with human genetic linkage data and postmortem brain data, as a Bayesian strategy of cross validating findings. Topping the list of candidate genes, we have three genes involved in GABA neurotransmission (GABRA1, GABBR1 and GAD2), one gene involved in glutamate neurotransmission (GRIA2), one gene involved in neuropeptide signaling (TAC1), two genes involved in synaptic function (SYN2 and KCNJ4), five genes involved in myelin/ glial function CNP, MBP, PLP1, MAL, GFAP), and one gene involved in lipid metabolism (LPL). These data suggest that schizophrenia is primarily a disorder of brain functional and structural connectivity, with GABA neurotransmission playing a prominent role. These findings may explain the EEG gamma band abnormalities detected in schizophrenia. The analysis also revealed other high probability candidates genes (neurotransmitter signaling, other structural proteins, ion channels, signal transduction, regulatory enzymes, neuronal migration/ neurite outgrowth, transcription factors, RNA regulatory and circadian clock genes), pathways and mechanisms of likely importance in pathophysiology. Some of the pathways identified
731
suggest possible avenues for augumentation pharmacotherapy of schizophrenia with other existing agents, such as benzodiazepines, anticonvulsants and lipid lowering agents. Other pathways are new potential targets for drug development. Lastly, a comparison with our earlier work on bipolar disorder illuminates the significant molecular overlap between schizophrenia and bipolar disorder. O25.3 BDNF HAPLOTYPE DIFFERENTIATES SCHIZOAFFECTIVE DISORDER AND MAJOR AFFECTIVE DISORDERS FROM SCHIZOPHRENIA Todd Lencz1, Robert Lipsky2, Pamela DeRosse1, Katherine Burdick1, Judith Jaeger1, John Kane1, David Goldman2, Anil Malhotra1 1 The Zucker Hillside Hospital, Glen Oaks, NY 2 NIAAA, Bethesda, MD Background: Recent population genetic studies suggest a partial overlap in genetic transmission of schizoaffective disorder and mood disorders. Allelic variaton in the gene encoding Brain Derived Neurotropic Factor (BDNF) has been associated with affective disorders in some studies, but schizophrenia studies have been largely negative.It is hypothesized that BDNF allelic variation is associated with the mood component of disorder, and that haplotype frequencies will be similar in patients with schizoaffective disorder and primary mood disorders, as distinct from patients with schizophrenia and healthy volunteers. Methods: We tested for an association between a 5-marker BDNF haplotype and SCID-based, consensus DSM-IV diagnosis, in 373 Caucasian patients and 220 Caucasian healthy comparison subjects. Primary diagnoses included schizophrenia (n ¼ 208), schizoaffective disorder (n ¼ 60), bipolar disorder (n ¼ 76) and major depressive disorder (n ¼ 29). Results: We detected a significant association between BDNF haplotypes and illness manifestation. The common haplotype (containing the valene variant of the Val66Met polymorphism) was overrepresented in patients with schizoaffective disorder and affective disorders compared to healthy volunteers. Moreover, the common haplotype showed significantly greater frequency in schizoaffective patients compared to patients with schizophrenia. Patients with schizophrenia did not significantly differ from healthy volunteers. Conclusions: This is the first candidate gene study to differentiate schizoaffective disorder from schizophrenia. These data suggest that the effect of BDNF genetic variation may be associated with the clinical phenotype of affective dysregulation across several DSM-IV diagnostic categories, representing a step towards the molecular classification of psychiatric illness. O25.4 A SYSTEMATIC HAPMAP-BASED SURVEY OF SCHIZOPHRENIA CANDIDATE GENES IN THE GERMAN POPULATION Sven Cichon1, Thomas Mu¨hleisen1, Axel Hillmer1, Robert Fu¨rst1, Rami Abou Jamra2, Johannes Schumacher2, Per Hoffmann2, Margrieta Alblas1, Susanne Kopmann1, Alexander Georgi2, Thomas G. Schulze2, Peter Propping2, Marcella Rietschel2, Markus M. No¨then1 1 Life & Brain Center, University of Bonn, Bonn 2 Institute of Human Genetics, University, Bonn 2 Central Institute of Mental Health, Mann, Mannheim Recent association studies have identified a number of potential schizophrenia susceptibility genes that currently await replication in independent, large samples of schizophrenia. In our present study, we systematically covered 6 promising genes with haplotype tagging SNPs capturing all haplotypes with a frequency > 1% using HAPMAP phase I and II data, and tested these SNPs for association in a total of 2,230 individuals, comprising a family-based sample of 210 parent-offspring trios with schizophrenia and in an independent sample of 800 schizophrenia patients and 800 controls, all of German origin (all four grandparents and parents German). All diagnoses were made based on DSMIV criteria. Using Illumina’s GoldenGate assays, we genotyped a total of 628 SNPs in these samples, covering the RGS4 genes in chromosomal region 1q23.3 (15 SNPs), the DISC1/DISC2/TSNAX locus on 1q42.1 (138 SNPs), ENTH on 5q33 (30 SNPs), DTNBP1 on 6p22.3 (45 SNPs), NRG1 on 8p21 (384 SNPs) and COMT on 22q11.2 (16 SNPs). We have completed genotyping of all individuals. Genotypes are currently under statistical analysis using TDT and FAMHAP. For associated
732
Abstracts
variants/haplotypes, detailed genotype/phenotype correlations will be performed. O25.5 MIDDLE BRAIN STRUCTURES IN SCHIZOPHRENIA PATIENTS’ FAMILIES: MRI CHARACTERISTICS OF PREDISPOSITION AND NEURODEGENERATION Vera Orlova1, Nina Voscresenskaya2, Eugene Miloserdov3, Leonid Gubsky3, Dmitry Kupriyanov3, Nikolay Anisimov4 1 Mental Health Research Centre, RAMS, Moscow 2 Mental Health Research Centre, RAM, Moscow 3 Centre of Magnetic Resonance Tomography, Moscow 4 Centre of Magnetic Resonance Tomog, Moscow In order to investigate characteristics of middle brain structures in schizophrenia patients and their relatives 157 subjects were examined by MRI: 38 probands, 77 parents and siblings, and 42 controls. Morphometric analysis included assessment of linear (width) and volume parameters of lateral ventricles (LV), III ventricle and their indices as well as linear and angular indices (LAI1) reflecting interconnection of middle brain structures (ventricles, corpus callosum—CC, pons) and being sensible to deviations of liquor camera (on part of material). Genetic mathematical analysis (least square method) was used. The results showed that the majority of studied ventricle parameters has high ( > 50%) and medium (30–50%) heritability coefficients. Correlations of some displacement parameters of CC with LV volume in patients and in relatives were demonstrated (r from 0.5 to 0.9, p from 0.0003 to 0.04). Alteration of LAI was demonstrated in patients and their relatives (at the level of tendency). The established tendency indicates to complex nature of ventriculomegalia in schizophrenia which relates both with predisposing factors (primary hydrocephalia) and atrophic process (presumably). The data support unitary model summarizing mechanisms of aberrant brain development and degenerative process. The work was supported by Russian Foundation for Basic Research and Royal Society grants. 1 Yoram S., Metser U., Beni-Adani L. et al. Morphometric study of the midsagittal MR imaging plan in cases of hydrocephalus and atrophy and in normal brains // AJNR Am. J. Neuroradiol. – 2001. – Vol. 22. – P. 1674–1679. O25.6 ASSOCIATION OF DYSBINDIN WITH OUTCOME MEASURES IN A COHORT OF ITALIAN SCHIZOPHRENIC PATIENTS Sarah Tosato1, Mirella Ruggeri2, Chiara Bonetto3, Maria Elena Bertani2, Giovanna Marrella3, Antonio Lasalvia2, Doriana Cristofalo2, Giuseppe Aprili3, Michele Tansella2, Paola Dazzan5, Marta DiForti5, Robin Murray5, David Collier6 1 DMSP, Section of Psychiatry, Verona, Italy 2 DMSP, Section of Psychiatry, Verona 3 DMSP, Section of Psychiatry, Verona 3 Blood Transfusion Service, Azienda Ospedaliera, Verona 5 Institute of Psychiatry, Division of Psy, London 6 Institute of Psychiatry, SGDP Centre, London The dysbindin (DTNBP1) gene is considered a susceptibility gene for schizophrenia. In the present study, we performed exploratory analysis to determine if DTNBP1 might be associated with schizophrenia in a real clinical population from Verona, Italy. We analysed a prevalence cohort of patients (n ¼ 141) with a diagnosis of schizophrenia and follow-ups at 3 and 6 years. First, we performed a case-control study to test if polymorphisms of DTNBP1 are associated with schizophrenia. Then we tested whether clinical symptoms are cross-sectionally associated with the presence of DTNBP1 variants and if patients carrying these DTNBP1 variants have a more unfavourable longitudinal clinical outcome. The patients’ clinical status was assessed at baseline using BPRS and GAF and then followed-up after 3 and 6 years. We genotyped 80 patients and 106 healthy controls for polymorphisms in DTNBP1. A trend towards statistical significance (p ¼ 0.058) for rs2619538 was found. Using GENECOUNTING software, we found that rs2619538-P1583 (p-value ¼ 0.048), P1320-P1757 (pvalue ¼ 0.034) and rs2619538-P1583-P1578 (p-value ¼ 0.040) haplotypes occurred more often in cases as compared to controls. The analysis of the clinical course of schizophrenia over time showed that the rs2619538-P1583 haplotype was more likely to be transmitted to subjects with more severe and persistent psychopathology for the items ‘‘conceptual disorganisation’’ and ‘‘emotional withdrawal’’. The results are compatible with the view that DTNBP1 is a candidate gene for schizophrenia and is associated with more severe psychopathology.
O25.7 STATISTICAL EPISTASIS BETWEEN NRG1, ERBB4, AKT1 AND CAPON INFLUENCES RISK FOR SCHIZOPHRENIA Kristin Nicodemus1, Bhaskar Kolachana2, Radhakrishna Vakkalanka2, Richard E Straub2, Daniel R. Weinberger2 1 GCAP/CBDB/NIMH/NIH, Johns Hopkins SPH, Bethesda MD 2 GCAP/CBDB/NIMH/NIH, Bethesda MD The protein encoded by NRG1, a candidate gene for schizophrenia, interacts with proteins in other putative candidate genes. We assessed whether variation in NRG1 was associated with schizophrenia using a dense map of NRG1 SNPs and microsatellites. In addition, we utilised exploratory statistical methods to test for interaction between SNPs in NRG1 and its protein or signalling pathway interaction partners. Using a family and a case-control sample, we conducted FBATs, logistic regression and a score-test haplotype analysis to assess association between NRG1 and schizophrenia. To test for interaction between NRG1 and its partners, we used a recently-proposed Markov chain Monte Carlo boolean logic tree-based algorithm and a CART-based algorithm. 3 SNPs in NRG1 were marginally associated with risk of schizophrenia in case-control/family analyses. We observed weak association in family-based analysis for HapICE SNPs NRG221132 (p ¼ 0.075), and for haplotypes containing NRG221533 (p ¼ 0.086), NRG241930 and NRG243177 (p ¼ 0.098). Both epistasis methods created best trees using 3 SNPs in NRG1 and 1 SNP each in AKT1, CAPON and ErbB4. Focusing solely on these SNPs to reduce multiple testing, we were able to confirm pairwise gene-gene interaction (LRT p: NRG1*CAPON p ¼ 0.0014; CAPON*AKT1 p ¼ 0.031; NRG1*ErbB4 p ¼ 0.077) using standard logistic regression. The association between NRG1 and schizophrenia was significantly increased by the use of novel epistasis methods to detect interaction. We are genotyping an independent sample to confirm our data-derived interaction models. O25.8 A COMPLEMENTARY STRATEGY TO APPROACH THE GENETICS OF SCHIZOPHRENIA Dan Rujescu University of Munich, Dept. of Psychiatry, Munich We use complementary strategies to approach the pathobiology and genetics of schizophrenia, including genetic association and family studies, as well as animal and cell-culture models. We first aim to identify genes in a large case-control, family-based study. Over 500 patients, 200 first-degree relatives and 1600 healthy volunteers, entered the study. High-throughput genotyping was done and initial results will be presented. Furthermore, we use endophenotypes as a complementary approach. These comprise, among others, electrophysiological (e.g. P50 suppression and eye movement), neuropsychological (e.g. working memory, attention/vigilance, verbal/visual learning and memory, speed of processing, and problem solving) and neuroimaging endophenotypes. Additionally, we use an NMDAreceptor antagonist animal model which mimics aspects of psychosis to identify candidate genes which can be used in human studies. In our model, chronic, low-dose treatment with MK801 alters the expressionof NMDA receptor subunits in a pattern similar to schizophrenia. On a cellular level, the number of parvalbumin-positive interneurons was selectively decreased, a finding which parallels observations in post mortem brains. On a functional level, recurrent inhibition of pyramidal cells was altered, as postulated from histological findings. Finally, on a behavioural level, these animals showed cognitive deficits, which again parallel findings in schizophrenia. We used a functional genomic approach for the identification of hippocampal candidate genes for psychosis-related traits and identified differentially-expressed genes. In human genetic analyses, some of these novel genes were found to be strongly associated with schizophrenia and/or above-mentioned endophenotypes. 026.1 EXPLAINING SUICIDALITY IN A POPULATION-BASED YOUTH SAMPLE Richard Todd Washington University School of Medicine, Saint Louis, Missouri Increased rates of suicidal behaviour have been reported in youths with depression, bipolar disorder (BPD), and obsessive compulsive disorder (OCD). We examined the interaction of these factors on risk for suicidal behaviour using quantitative measures of mood disturbance and OCD
Abstracts in a population-based Missouri twin sample (N ¼ 1,325). Child Behaviour Checklist (CBCL) profiles for BPD and OCD, and semistructured interview diagnoses of mood disorders, behaviour disorders, and suicidal behaviour were included. Individuals meeting the CBCLBPD and CBCL-OCD profiles overlapped substantially (n ¼ 28). Few subjects had CBCL-JBD alone (n ¼ 5), while CBCL-OCD alone was more prevalent (n ¼ 108). The CBCL-BPD profile was influenced primarily by additivegenetic factors (67%) while the CBCL-OCD profile was moderately heritable (47%). Examination of the DAT1 440 allele in subjects with both phenotypes (CBCL-BPD-OCD) indicated a 4-fold excess of homozygotes (p ¼ 0.04). Risk for suicidal thoughts, plans, or attempts was significantly increased in subjects meeting CBCL-OCD only (OR ¼ 2.91(1.40–6.07)) and was higher in subjects having CBCLBPD-OCD (OR ¼ 4.82(1.39–16.77)). Comorbid DSM-IV conduct disorder, ADHD, and BPD did not alter this association. Comorbid ODD and depression were significantly associated with suicidality, but did not account for the association with CBCL-OCD or CBCL-JBD-OCD. Subjects with CBCL-JBD-OCD and CBCL-OCD are at increased risk for potentially lethal behaviours, not due to the presence of comorbid mood disorders.
O26.2 INFLUENCE OF SEXUAL ABUSE ON SUICIDALITY: MODULATION BY POLYMORPHISMS IN THE 5-HTT AND THE TH1 GENE A. Malafosse2,3,4, I. Jaussent 1,2, F. Jollant 1,2, N. Perroud 3, P. Baud 3, C. Buresi 3, F. Belivier 1,2, F. Mathieu 1,2, M. Leboyer 1,2, Ph. Courtet 1,2 1 Department of Psychological Medicine and Psychiatry, CHU Montpellier, France 2 INSERM E 361, Montpellier, France 3 Department of Psychiatry, University Hospital of Geneva, Switzerland 4 Department of Medical Genetics and Development, School of Medicine, Geneva, Switzerland A stress-diathesis model has been proposed for suicidal behaviour (SB). This suggests that individuals’ sensitivity to stressful events depends on their genetic makeup. Adults reporting childhood physical and sexual abuse experiences have a higher rate of SB and greater impulsivity. Recent meta-analyses yielded statistically significant replication of first reports of association between the tryptophan hydroxylase gene (TPH1) and the serotonin transporter gene (5HTT) and SB. The goal of the present study was to determine whether the intron 7 A218C TPH1 and the promoter ins/del 5HTT polymorphisms moderate the influence of childhood trauma on SB. Using the Childhood Trauma Questionnaire we evaluated patients with (N ¼ 392) and without (N ¼ 51) a history of SB. We created a composite index of SB using the number of suicide attempts, the French versions of the Risk Rescue Rating Scale (RRRS), and of the Suicidal Intent Scale (SIS). Analyses show that sexual abuse is related with high scores of suicidality (OR ¼ 1.60, 95%CI [1.04–2.45]). This association is even more important among individuals carrying two copies of the short allele of the 5-HTT promoter polymorphism (OR ¼ 3.10, 95%CI [1.14– 8.46]) but not among those carrying one long allele. Similarly, sexual abuse is associated with high scores of suicidality among intron 7 A218C TPH1 heterozygotes (OR ¼ 2.51, 95%CI [1.32–4.67]) but not among both homozygotes. These results suggest that both TPH1 and 5HTT polymorphisms modulate the effect of sexual abuse on suicidality.
O26.3 A MOLECULAR GENETIC STUDY OF COMPLETED SUICIDE IN CENTRAL SLOVENIA Tina Zˇagar1, Borut Jerman1, Metka Paragi1, Karen Sugden2, Ian Craig2, Anne Farmer2, Maja Zorko1, Sasˇka Rosˇkar1, Milosˇ Kravanja1, Olga Balazˇic3, Jozˇe Balazˇic3, Andrej Marusˇicˇ1 1 Institute of Public Health of the Republic of Slovenia, Ljubljana, Slovenia 2 Social, Genetic and Developmental Psychiatry Centre, Institute of Psychiatry London, UK 3 University of Ljubljana, Faculty of Medicine, Institute of Forensic Medicine, Ljubljana, Slovenia Background: Suicidal behaviour has been associated with dysfunction of genes responsible for the serotonergic system.
733
Aims: To determine genetic risk factors for completed suicide in a case-control study and investigate correlations with other relevant factors. Method: The DNA analyses included three single nucleotide polymorphisms (SNP) and microsatellite in TPH1. We also analysed an insertion/deletion in the promoter region (HTTLPR) and VNTR2 in the serotonin transporter gene (HTT), one SNP in the receptor HTR2A and another SNP in receptor HT1B, all genes related to the serotonergic neurotransmission. The last investigation was VNTR within gene for MAOA, referring to serotonin catabolism. All analyses were done on 224 suicide victims, who committed suicide using different violent or non-violent methods, on 146 of their matched. Results: The main findings of our study was an association between the TPH1 A779C, MAOA and suicide victims older than 48 years. The latest results of gene-environment interactions will be presented. Conclusion: Our results indicate the influence of other variables, like age and gender as possible covariants of genetic risk factors for the completed suicide in Slovenia.
O26.4 RELATIONSHIP OF SEROTONIN TRANSPORTER GENE POLYMORPHISMS AND HAPLOTYPES TO SUICIDAL BEHAVIOUR AND MRNA TRANSCRIPTION Ph. Courtet1,2, F. Jollant1,2, N. Perroud3, C. Buresi3, A. Malafosse2,3,4 1 Department of Psychological Medicine and Psychiatry, CHU Montpellier, France 2 INSERM E 361, Montpellier, France 3 Department of Psychiatry, University Hospital of Geneva, Switzerland 4 Department of Medical Genetics and Development, School of Medicine, Geneva, Switzerland There is compelling evidence that serotonin system dysfunction is associated with certain behavioural disorders, such as suicidal behaviour and impulsive aggression. We and others have reported that a functional serotonin transporter promoter polymorphism, 5-HTTLPR, alters the risk of violent suicide behaviour (SB). Subsequently, meta-analyses confirmed that the 5-HTTLPR s allele was associated with SB (OR ¼ 1.17 CI: 1.04–1.32, p ¼ 0.009), specifically with violent SB (OR ¼ 3.32 CI: 1.51–7.30, p ¼ .0001). Two other serotonin transporter functional variants have been described, one in the promoter consisting in an A/G SNP, and the second in the intron 2 (STin2 VNTR). These three polymorphisms, as well two others located in intron 6 (rs140700) and in the 30 -UTR region (rs3813034), were genotyped in 1015 suicide attempters, 108 suicide completers and 450 controls, from the University Hospitals of Montpellier and Geneva. In addition, quantification of 5-HTT mRNA was done by Real-Time PCR in both lymphoblastoid cell lines and prefrontal tissues. Allelic associations or linkage disequilibrium (LD) (D0 ) matrices were computed across the seven SNPs using Haploview version 2.04 Software. Allele, genotype and haplotype frequencies were compared using the w2 distribution. Preliminary results suggest an 5-HTT haplotype association with violent SB in the whole sample (w2 ¼ 8.0, p ¼ 0.045), and lower 5-HTT mRNA expression in cells (p ¼ 0.03) and tissues (p ¼ 0.05) with the ssþlGlG genotypes than in those with the lAlA genotypes.
O26.5 ASSOCIATION BETWEEN THE BRAIN DERIVED NEUROTROPHIC FACTOR (BDNF) GENE POLYMORPHISM (VAL66MET), DEPRESSIVE DISORDERS AND SUICIDAL BEHAVIOUR M. Sarchiapone1, L. Iacoviello2, V. Carli1, C. Cuomo1, M. De Gaetano2, M.C. Latella2 1 Department of Health Sciences, University of Molise, Campobasso, Italy 2 Laboratory of Molecular Genetics, Catholic University, Campobasso, Italy Background: BDNF has been studied in the brain of depressed patients and suicide victims and lower levels of BDNF expression were found in comparison with normal controls. The aim of our study was to examine the association between the Val66Met functional polymorphism of the BDNF gene and suicidal behaviour in depressed patients.
734
Abstracts
Method: Study participants comprise 225 patients with a DSM IV diagnosis of Unipolar Major Depression. From every patient a blood sample was collected in order to perform the genetic analysis. Severity of depressive symptoms has been evaluated with the Hamilton Depression Rating Scale (HDRS). Detailed information about suicidality (suicidal ideation, number, intentionality and lethality of attempts, age of first attempt) was also collected. Specific tools have also been used for the evaluation of impulsivity and aggression. Results: No association was found between the BDNF polymorphism and suicidal behaviour when performing a case-control analysis. Patients’ genotype was not related to the presence of specific depressive symptoms nor to the overall severity of depression. When focusing the analysis on the subgroup of suicidal patients, no association was found with suicidal ideation or with number, severity or lethality of suicide attempts. Impulsivity and aggression did not been to be related with a specific BDNF polymorphism. Conclusion: In the studied sample, no association was found between the BDNF genetic functional polymorphism, characteristics of the depressive illness and suicidality. O26.6 CATECHOL-O-METHYLTRANSFERASE POLYMORPHISM (COMT) IN SUICIDE ATTEMPTERS: A POSSIBLE GENDER EFFECT ON ANGER TRAITS Patrick Baud1, Philippe Courtet2, Nader Perroud1, Fabrice Jollant2, Catherine Buresi1 1 University Hospital of Geneva, Cheˆne Bourg, Gene`ve 2 Department of Psychological Medicine, Montpellier Anger-related personality traits are considered risk factors for suicidal behaviour. According to twin studies, they are partially under genetic control and their various clinical expressions have been associated with serotonergic and catecholaminergic activities. A functional polymorphism on the human COMT, which codes for the catecholamines inactivating enzyme catechol-O-methyltransferase, has been shown to influence aggressive and anger-related traits in various clinical populations. The aim of the present study was to investigate the association between anger traits (as characterized by the State-Trait Anger Expression Inventory) and COMT Val158Met polymorphism in suicide attempters (n ¼ 427) and control subjects (n ¼ 185). Results showed that suicide attempters have a higher frequency of the Val/Val genotype than control subjects (29% vs 18.4%; p ¼ 0.021), as well as a higher Val allele frequency (54.6% vs 47.3%; p ¼ 0.019). Moreover, the Val/Val genotype markedly affected scores on two STAXI subscales — Trait Anger and Anger Control — in female suicide attempters, thus suggesting a possible gender effect of the COMT genotype on a stable personality trait (Trait Anger scores: p ¼ 0.002, F 6.26, df 2/298, KW: 0.002; df 2; X2 12.55 ; Anger Control scores: p ¼ 0.003, F ¼ 4.98, df 2/298, KW: 0.002, df 2, X2 12.8). These results are discussed in the light of recently published data on the effect of COMT Val158Met polymorphism on cognitive performances and behavioural tendencies in clinical and non-clinical populations. O26.7 MAO A AND MAO B IN SUICIDAL BEHAVIOR Ina Giegling, Annette M. Hartmann, Hans-Ju¨rgen Mo¨ller, Dan Rujescu Molecular and Clinical Neurobiology, Dept. of Psychiatry, University of Munich, LMU, Germany Risk of suicide-related behavior is supposed to be determined by a complex interplay of sociocultural factors, traumatic life experiences, psychiatric history, personality traits, and genetic vulnerability. Family and twin studies point towards a partial heritability of suicidal behavior and aggression-related traits. We have initiated a large scale case control genetic association study which comprises of 250 suicide attempters and 1600 healthy volunteers and investigated the role of a comprehensive set of candidate genes in this behavior. We will present new data on the MAO A and MAO B genes a candidate genes for these traits. We studied several gene variants of MAO A and MAO B in a sample of 571 suicide attempters, healthy controls and suicide completers. We also analyzed the relation to aggression-related traits, assessed by STAXI and FAF. Our results provide evidence that the MAO A and MOA B may influence aggression-related traits. Given that these represent intermediate phenotypes of suicidal behavior, MAO A and MAO B might also act on suicidal behavior through these traits. The observed associations warrant further replications.
O27.1 DO TAG MARKERS TELL US THE TRUTH? Valentina Moskvina Cardiff University, Cardiff In view of the linkage disequilibrium structure of the genome, the selection of maximally informative SNP markers is a fundamental issue in the design of association studies. Currently used selection methods rely on pairwise marker correlation or informativity measures for subsets of markers. Nevertheless, the selected markers do not provide a completely satisfactory description of the individual remaining markers. In some cases, it is altogether impossible to detect an association between a marker and a disease using markers selected based upon pairwise r2 (Terwilliger and Hiekkalinna, 2006). The number of tag markers can be further reduced by using haplotypic information, but then the results of association analysis are difficult to interpret. The presented work shows examples where it is impossible to find an association of a marker (B) with the disease status (C) using a tag marker (A) which is in r2 ¼ 0.8 with the marker B. Moreover, the necessary sample size NAC to detect association between B and C using A is commonly calculated as NAC ¼ NBC /r2(AB); here the question of how far this rule over-/underestimates the sample size is systematically studied. A new alternative methodology of informative marker selection, allowing for control and optimisation of information content and full marker reconstruction, is also discussed. Reference: Terwilliger, J D and Hiekkalinna, T. (2006) An utter refutation of the ’Fundamental Theorem of the HapMap’. European Journal of Human Genetics: 14(4) p.426–437. O27.2 IDENTIFICATION OF NOVEL GENES AND GENE CLUSTERS RELATED TO HUMAN MEMORY Andreas Papassotiropoulos1, Dietrich A. Stephan2, Matthew J. Huentelman2, Eric M. Reiman2, Dominique J.-F. de Quervain1 1 University of Zurich, Zurich 2 Translational Genomics Research Institute, Phoenix, AZ Background: Experimental work in animals has shown that memory formation depends on a cascade of molecular events. In humans, heritability estimates of 50% suggest that genetic factors have an important impact on this fundamental brain function. Objective: To identify memory-related genes and gene-clusters in humans. Methods: We used a combination of whole-genome association (500K SNP chip), candidate gene approach (gene clustering) and fMRI. Conclusions: Variability of human memory performance is related to variability in genes encoding proteins of a signaling cascade, including NMDA receptor, metabotrobic glutamate receptor, adenylyl cyclase, CAMKII, PKA and PKC. Functional magnetic resonance imaging during memory formation reveals that this genetic profile correlates with activations in memory-related brain regions, including the hippocampus and parahippocampal gyrus. The whole-genome association study revealed the existence of novel genes significantly related to human memory performance and brain activation. Relevant literature: de Quervain and Papassotiropoulos, PNAS 2006; 103 (11): 4270–4. de Quervain et al., Nature Neuroscience 2003; 6(11): 1141–2. O27.3 MAQ: STRAIGHTFORWARD DETECTION AND ANALYSIS OF SPECIFIC COPY NUMBER POLYMORPHISMS Jurgen Del-Favero1, Lien Heyrman2, Boris Harding2, Wim Glassee2, Peter De Rijk2, Dirk Goossens2 1 University of Antwerp, Antwerp 2 University of Antwerp, Antwerp Recent studies show that Copy number polymorphisms or CNPs are far more common in the human genome than expected. Since these variations are not picked up by sequencing based mutation detection schemes, other methods need to be applied. Here we present a multiplex PCR (mPCR) based method for the detection and analysis of CNPs. This method, designated Multiplex Amplicon Quantification (MAQ), consists of the simultaneous PCR amplification of several fluorescently labeled target and reference sequences. The comparison of normalized intensities between the test individual and control individuals results in a dosage quotient indicating the copy number of the target amplicon. The PCR primers of the MAQ assays are designed by proprietary software allowing simultaneous analysis of several
Abstracts genomic regions, genes or even exons of interest. MAQ is highly comparable to Multiplex Ligation-dependent Probe Amplification (MLPA), with the important difference that the experimental setup consists of a single mPCR reaction followed by fluorescent fragment analysis, resulting in a considerable reduction of labor and time. Together with the method, we developed a program, MAQr, to facilitate the analysis of experimental data. We developed and applied a MAQ assay for a set of 3 genes (GLUR7, CACNG2 and AKAP5) previously reported to show DNA copy number variation in psychiatric patients (Wilson et al., 2006). Subsequent MAQ analysis of over 400 DNA samples of psychiatric patients did not reveal the presence of DNA copy variation within these genes. Wilson et al., (2006) Human Molecular Genetics 5(5): 743–749. O27.4 ESTIMATION OF GLOBAL AND LOCAL GENOMIC SHARING USING GENOME-WIDE SNP DATA Shaun Matthew Purcell1, Benjamin N. Neale2, Mark J. Daly3, Pak C. Sham4 1 Center for Human Genetic Research, Boston 2 SGDP Research Centre, King’s College London, London 3 Broad Institute & CHGR, MGH, Boston 4 Genome Research Centre, University of Hong Kong, Hong Kong Recent technological developments have made possible the efficient genotyping of hundreds of thousands of SNPs in individual DNA samples. Genetic data of this quantity and detail will provide information on the ancestry of individuals and on the relationship between pairs of distantly-related individuals. We have developed a simple identity-by-descent method to estimate the global genomic sharing between any two individuals from the same population. This method was used to detect previously unknown relationships among individuals in the HapMap population samples. We are now developing a Hidden Markov Model (HMM) to obtain estimates of local genomic sharing between pairs of individuals, given the genotype data and the estimated global genomic sharing. The methodology will take some account of complications such as genotyping errors and linkage disequilibrium between SNPs, and results in estimated probabilities of sharing none, one or both haplotypes by descent, at any section of the genome. The method is be implemented in the PLINK software (http:// pngu.mgh.harvard.edu/purcell/plink/) and applied to the HapMap Phase 2 data. PLINK is a whole genome association analysis toolset, offering both computationally efficient implementations of existing methodologies (data management, summary statistics, sample matching based on inferred ancestry and a variety of association statistics) as well as novel approaches such as IBD estimation. Finally, we will discuss the potential use of this methodology to gene-mapping studies of complex diseases and quantitative traits. O27.5 INTEGRATION OF IMAGING AND GENETICS: AN ANALYTICAL STRATEGY TO PERFORM GENOME-WIDE STUDIES FOR IMAGING GENETIC PHENOTYPES (IGPs) Fabio Macciardi1, Lee Friedman2, Jessica Turner2, Hal Stern2, Steven Potkin2 1 University of Milano, Segrate 2 University of California, Irvine, Irvine Our goal is to develop novel approaches addressing the interplay of imaging and genetics within the context of a specific disease. Key to our approach is a general linear model (GLM) analysis of a quantitative imaging phenotype. The model can accommodate a range of factors expected to affect the observed imaging activation (e.g., ethnicity, behaviour/performance): in the most basic form, the GLM is: Imaging Phenotype ¼ Overall Mean þ Genotype Effect þ Diagnosis Effectþ Genotype-Diagnosis Interaction Effect To perform a genome-wide analysis with 500,000þ SNPs, we developed a method building on the work of Satagopan et al. (2004) and modifying their approach of using the GLM summary test statistic on imaging phenotypes instead of traditional test statistics comparing allele frequencies across diagnostic groups. Our approach is to assume that a set ‘‘T’’ of the 500,000þ SNPs are loci with true genotype-diagnosis interaction effects on the imaging phenotype. The goal is to identify a subset ‘‘t’’ of these T true disease-related loci using the loci corresponding to the highest ‘‘t’’ teststatistics. The value of t is determined by considerations of power. The power of the statistical strategy is defined as the probability that the t
735
highest test statistics correspond to t out of the T true disease (and imaging) related loci. This power can be calculated by simulation as a function of sample size, T and t, the minor allele frequencies in the patient and control groups and the interaction effect size.
O27.6 EFFICIENT CALCULATION OF EMPIRICAL P-VALUES FOR GENOME WIDE LINKAGE AND ASSOCIATION THROUGH WEIGHTED MIXTURES Sarah Medland, B.T. Webb, Po-Hsiu Kuo, Michael C. Neale VIPBG, Richmond, VA Currently the two most popular methods for obtaining empirical pvalues for genome-wide linkage and association involve performing ‘gene-dropping’ simulations or permutation tests (in which the coefficient of genotypic sharing, is randomized across relative pairs) at each locus. However, as the time required to obtain empirical significance is directly proportional to the time taken to complete the initial analyses, the time required to obtain empirical p-values for multivariate analyses using these methods is prohibitive. Utilizing the distributional dependencies between the coefficient of genotypic sharing and the permutation test we report a new method of efficient permutation. In summary, the distribution of the coefficient of genotypic sharing at locus j is estimated as a weighted mixture of the coefficients of genotypic sharing drawn from a pool of ‘modal’ distributions. This weighting scheme is then applied to the distribution of the permutation tests at the ‘modal’ loci to obtain an empirical pvalue at locus j (which is asymptotically distributed as the permutation test at loci j). This weighted-mixture approach greatly reduces the number of permutation tests performed making it suitable for use in multivariate analyses and other computationally intensive linkage and association analyses. In addition, as the distribution of genotypic sharing is a property of the genotypic data for a given sample and is independent of the phenotypic data, the weighting scheme can be applied to any phenotype (or combination of phenotypes) collected from that sample. O28.1 GENETICS OF SUICIDE-, ANGER- AND AGGRESSION-RELATED BEHAVIOUR Dan Rujescu University of Munich, Dept of Psychiatry, Munich Risk of suicide-related behaviour is supposed to be determined by a complex interplay of socio-cultural factors, traumatic experiences, psychiatric history, personality and genetic vulnerability. This view is supported by adoption and family studies indicating suicidal acts have a genetic contribution that is independent of the heritability of Axis I and II psychopathology. The heritability for serious suicide attempts was estimated to be 55%. Neurobiological studies have shown that serotonergic dysfunction is implicated in suicidal behaviour. These findings stimulated the investigation of variations in serotonergic genes in this context. We have initiated a large-scale case control genetic association study which comprises 250 suicide attempters and 1600 healthy volunteers and has investigated the role of a comprehensive set of serotonergic candidate genes. We will present new data on a comprehensive set of serotonergic candidate genes. Since both aggression-related traits and serotonergic activity are partially heritable and correlate inversely, variations in these genes might then, to some extent, account for variations in aggression-related behaviour. Thus, we also investigated the relationship between genes and anger, as a subtype of aggression-related behaviour. Additionally, we conducted a large-scale gene expression analysis using cDNA-microarrays to identify new candidate genes. We found several genes to be differentially expressed in the orbitofrontal cortex of suicide completers. Crossvalidation experiments using quantitative RT-PCR validated genes so far. These genes were genotyped as well to look for associations with suicide-, anger- and aggression-related behaviour and these results will also be presented. O28.2 PAM AS A CANDIDATE GENE FOR AUTISM Susan Santangelo1, Stephen Haddad2, Tulio Santos3, Jesen Fagerness2, Irma Moilanen3, Marja-Leena Mattila3, Katja Jussila3, Sanna Kuusikko3, Jaakko Ignatius4, Lesley Yamaki2, Priya Moorjani2, David Pauls2, Vijaya Ramesh3 1 Harvard Medical School/Mass General Hosp, Boston, MA
736
Abstracts
2
Psychiat Neurodev Genetics Unit, CHGR, M, Boston, MA Center for Human Genetic Research, Mass Gen, Boston, MA University of Oulu, Dept of Child Psychiatry, Oulu 4 University of Oulu, Laboratory of Genetics, Oulu 3 3
The co-occurrence of autism spectrum disorders (ASD) and Tuberous Sclerosis Complex (TSC) may yield clues to the etiology of autism. We recently identified Pam, (Protein Associated with the Myc oncogene), as an interactor of the tuberin-hamartin protein complex in mammalian CNS. Tuberin and hamartin function together to inhibit mammalian target of rapamycin (mTOR) signaling and cause TSC. The multiple functional domains of Pam, its highest expression in CNS, as well as the role of Pam homologs in synapse development, make it a strong candidate gene for autism. The gene for Pam (aka MYCBP2) maps to chromosome 13q22, and spans 280kb, with 84 exons. In a sample of 156 ASD parent-offspring trios, we genotyped 70 htSNPs (MAF > 5%, r2 > 0.8), 58 of which remained after excluding errors. Affecteds were diagnosed via the Autism Diagnostic Interview and/or Autism Diagnostic Observation Schedule. Using Haploview, we identified a nominally significant (p ¼ 0.006) 5-SNP haplotype in block 1. We then used the sliding window approach in whap to examine all non-overlapping 5-SNP haplotype blocks, which yielded a primary omnibus permutation test p-value of 0.02 for the first block. Individual analyses of the first 5 SNPs returned a global permutation p-value of 0.02 (corrected for multiple testing) and revealed the significance of the fourth SNP (likelihood ratio test ¼ 8.09, permutation p-value ¼ 0.007, on 10,000 permutations). We are currently carrying out a replication study in several hundred AGRE families. O28.3 HIGH-DENSITY SNP ASSOCIATION STUDY OF THE AUTISM SUSCEPTIBILITY LOCI ON CHROMOSOME 7q Nuala Sykes1, J. Lamb2, E. Maestrini3, L. Winchester2, A. Morris2, H. Butler2, E. Bacchelli3, F. Blasi3, G. Barnby2, I. Sousa2, A. J. Bailey3, A. P. Monaco2, IMGSAC3 1 University of Oxford, Oxford 2 Wellcome Trust Centre for Human Genetics, Oxford 3 University of Bologna, Bologna 3 University Department of Psychiatry, Oxford 3 http://www.well.ox.ac.uk/maestrin/iat.h, Oxford Autism is a severe neurodevelopmental disorder that usually occurs due to a complex genetic predisposition. Previously, linkage for autism has been found to chromosome 7q in the IMGSAC genome screen, which has been replicated by several groups and confirmed by metaanalysis. The peak of linkage is 40Mb and contains 200 genes. Attempts at fine mapping have been unsuccessful in decreasing the region of linkage and candidate gene screening has failed to find any etiological variants. A gene based, high-density SNP typing and association study was therefore carried out to identify susceptibility variants. Haplotype-tagging SNPs were selected across the linkage region, using Haploview and the Gabriel algorithm for block definition. Blocks overlapping all known genes and their putative regulatory elements were selected for genotyping. 1536 SNPs were genotyped (including SNPs to assess mis-typing and sample stratification) using the Golden Gate assay in a sample of 576 individuals. The sample was made up of 125 trios from multiplex families, selected based on their allele sharing status, and 192 Caucasian sex-matched controls. Casecontrol and family based association analyses were carried out using a block based approach and a GENEBPM algorithm for posterior probability. A replication study of the results obtained is now being carried out that will cover 339 SNPs from the top 5% of associated blocks and the positive genes, in a new independent sample of 200 trios and 200 controls.
O28.4 ASSOCIATION OF THE GAD1 CANDIDATE GENE WITH AUTISM Madalena Martins1, Catarina Correia2, Anto´nio Currais1, Ana Margarida Coutinho1, Carla Marques2, Assunc¸a˜o Ataı´de2, Teresa Sa˜o Miguel2, Joana Almeida2, Celeste Bento2, Teresa Morgadinho3, Guiomar Oliveira2, Astrid Moura Vicente2 1 Instituto Gulbenkian de Cieˆncia, Oeiras 2 Instituto Gulbenkian de Cieˆncia/Institut, Oeiras/Lisboa 2 Hospital Pedia´trico de Coimbra, Coimbra 3 Faculdade de Medicina da Universidade de, Coimbra
Autism is a neurodevelopmental disorder characterised by deficits in reciprocal communication and social interaction and by repetitive and stereotyped behaviour. An important genetic component has been recognised for autism, and several genome-wide scans for susceptibility genes have been carried out. Among others, a putative susceptibility locus on chromosome 2q31–33 has been identified. In the present study we tested the candidate gene glutamate decarboxylase 1 (GAD1), mapping to this linkage region, for genetic association with autism. The GAD1 gene encodes GAD67, a protein responsible for the conversion of glutamic acid into GABA, which has been found reduced in autistic patients. We have examined the association of autism with five SNPs within the GAD1 gene, two located in the 50 UTR region and three in intronic regions, in a sample of 245 nuclear families and 132 controls. The Transmission Disequilibrium Test (TDT) showed significant association with autism for the markers 3749034 (P ¼ 0.0190), in the 50 UTR region. The Transmit Test showed significant transmission for six haplotypes (0.0061 < P < 0.02), suggesting association with autism. Moreover, given the known interactions between the glutamate, GABA and serotonin neurotransmission systems, we tested the hypothesis of an involvement of GAD1 in the occurrence of hyperserotonemia in autism. No evidence of association was found between serotonin distribution and GAD1 variation. The positive results obtained for TDT of the 50 UTR marker and the significant transmission for six haplotypes suggest a contribution of the GAD1 gene to autism pathogenesis. O28.5 ANALYSIS OF 51 CANDIDATE GENES IN DSM-IV COMBINED TYPE ADHD Philip Asherson1, Keeley Brookes1, Neale Ben1, Kaixin Zhou1, Wai Chen1, Jan Buitelaar2, Richard Ebstein3, Micheal Gill4, Tobias Banaschewski5, Iris Manor6, Ana Miranda7, Bob Oades8, Herbert Roeyers9, Ari Rothenberger5, Jo Sergeant6, Hans-Christoph Steinhausen7, Edmund Sonuga-Barke8, Robert Plomin1, Peter McGuffin1, Pak Sham1, Xu Xiaohui1, Steve Faraone2 1 Institute of Psychiatry, London 2 Radboud University Nijmegen Medical Cent, Nijmegen 3 S. Herzog Memorial Hospital, Jerusalem 4 Trinity Centre for Health Sciences, Dublin 5 University of Goettingen 6 Geha MHC, Petach-Tikva 7 University of Valencia 8 University Clinic for Child and Adolescents, Essen 9 Ghent University, Ghent 6 University of Amsterdam 7 University of Zurich 8 University of Southampton 2 SUNY Update University, New York ADHD is a common neuro-developmental disorder, starting in early childhood and persisting into adulthood, in the majority of cases. Family and twin studies demonstrate the importance of genetic factors. Candidate gene association studies have identified several loci that exert small but significant effects on ADHD (particuarly DRD4 and DRD5). To provide further clarification of reported associations and identify novel associated genes, we examined 928 SNPs spanning 51 candidate genes involved in the regulation of neurotransmitter pathways, particularly dopamine, norepinephrine and serotonin pathways, in addition to circadian rhythm genes. Analysis used TDT in a sample of 776 DSM-IV combined type cases ascertained for the International Multi-centre Genetics Project (IMAGE). We found nominal significance with one or more SNPs in eighteen genes, including key findings in the literature: DRD4 and DAT1. Gene-wide tests, adjusted for the number of SNPs analysed in each gene, identified associations with TPH2, ARRB2, SYP, DAT1, ADRB2, HES1, MAOA and PNMT. A replication set of 383 families will provide further evidence for or against the associations observed. O28.6 EPIGENETIC EFFECTS IN ADHD: PARENT-OF-ORIGIN EFFECTS IN IMAGE SAMPLE Richard Anney1, Ziarah Hawi2, Karen Sheehan2, Keeley Brookes3, Kaixin Zhou3, Tobias Banaschewski4, Jan Buitelaar5, Richard. P Ebstein6, Ana Miranda7, Robert D Oades8, Herbert Roeyers9, Aribert Rothenberger10, Joseph Sergeant11, Edmund Sonuga-Barke12, Hans-Christoph Steinhausen13,
Abstracts 3
4
Phillip Asherson , Stephen Faraone , Michael Gill2, _ The IMAGE Consortium3 1 Trinity College Dublin, Dublin 2 Trinity College Dublin, Dublin 3 Institute of Psychiatry, London 4 University of Gottingen, Gottingen 5 University Medical Center Nijmegen, Utrecht 6 Hebrew University, Jerusalem 7 Universidad de Valencia 8 University of Duisberg, Essen 9 University of Gent 10 Gottingen University, Gottingen 11 Amsterdam Free University, Amsterdam 12 The University of Southampton 13 Zurich University, Zurich 4 State University of New York Attention-deficit/hyperactivity disorder (ADHD [MIM143465]) affects 2%–17% of school-age children. It is typically characterised by inattention, excessive motor activity, impulsivity and distractibility. In a recent manuscript, Hawi and colleagues observed that genetic risk alleles associated with ADHD were predominantly inherited from the paternal chromosome. This parent-of-origin effect (POE) was reexamined within the IMAGE sample. Of the 52 genes examined, 22 show at least a weak association signal (p < 0.1) and were defined as putative risk alleles for the purpose of this study. We examined these putative risk alleles for POE in a sample of 674 families. We do not observe a general trend for paternal over-transmission in the IMAGE sample. For some genes, we observe increased association signal from maternal chromosomes (eg. HES1, HES6 and TPH2), and for others, increased association signal from paternal chromosomes (eg. SLC6A2/ NET and SLC6A3/DAT1). We will discuss the key findings and the implications of this research for the field of psychiatric genetics O29.1 REQUIREMENT FOR TYPE III (CRD) NEUREGULIN1 IN SENSORIMOTOR GATING AND WORKING MEMORY Jasmine Ying-Jiun Chen, Madeleine Johnson, Michael Lieberman, Jay Gingrich, Holly Moore, Rose Goodchild, David Talmage, Lorna Role Columbia University, New York, NY Neuregulin1 (Nrg1) signalling has emerged as a key regulator of neuronal development, migration, myelination and synaptic maintenance. The Nrg1 gene has also been implicated as a susceptibility gene for schizophrenia. To understand its contribution to neural circuits and behaviour, we investigated Type III isoforms of Nrg1, the most abundant, neuron-specific class of Nrg1s in the brain. Expression of Type III Nrg1 is detected in the medial prefrontal cortex as well as in CA3, the ventral hippocampus and subiculum, regions that are associated with attention and working memory. Analysis of 5i transcriptional start sites for Nrg1 mRNA revealed distinct promoter usage for Type I vs. Type III Nrg1 isoforms, suggesting differential regulation of Nrg1 isoform expression at the transcriptional level. Adult mice with mono-allelic deletion of Type III Nrg1 (heterozygotes) exhibit enlarged lateral ventricles, impaired working memory and profound prepulse inhibition (PPI) deficits. These anatomical and behavioural anomalies are reminiscent of endophenotypes associated with schizophrenia. The PPI deficits of heterozygotes are normalised by nicotine treatment, a finding parallel to that on nicotine-induced enhancement of sensory gating in patients with schizophrenia. Finally, heterozygotes have blunted corticosterone response under acute stress, a phenomena linked to overall dysregulation of the hypothalamic-pituitary-adrenal (HPA) axis. Disturbances in the HPA axis have also been reported in schizophrenia. On the whole, our data support Type III Nrg1 as important in neural circuits pertaining to attention and memory and suggest a key role of Nrg1 signalling in regulation of appropriate responses to stress. O29.2 VALIDATION OF THE ROMAN LOW—AVOIDANCE (RLA) RAT LINE AS A MODEL OF GENETICALLY DETERMINED VULNERABILITY TO STRESS AND DEPRESSION Osvaldo Giorgi1, Giovanna Piras2, Maria G. Corda2 1 Dept of Toxicology—University of Cagliari; 2 Dept. of Toxicology—University of Cagliari
737
There is relative paucity of valid animal models to investigate the influence of the genotype on the pathogenesis of depression. The selective breeding of the Roman high-(RHA) and low-avoidance (RLA) rat lines for, respectively, rapid vs poor acquisition of two-way active avoidance in the shuttle-box has generated two phenotypes differing in their responsiveness to aversive and rewarding stimuli. Moreover, the RLA phenotype is reminiscent of several cardinal characteristics of mood disorders. The present study was undertaken to assess the heuristic value of the RLA rat line as a genetic model of depression. We report that RLA rats demonstrate, relative to their RHA counterparts: (1) reduced locomotor activity when exposed to aversive conditions, (2) increased ‘‘anxiety’’ in the conflict test, (3) longer duration of immobility in the forced swim test (FST), and (4) anhedonia-like behaviour when exposed to palatable food. Thus, RLA rats fulfil the criterion of face validity. We also show that the mesolimbic dopaminergic projections involved in the control of motivation and reward are hypofunctional in RLA rats relative to their RHA counterparts. This finding suggests that RLA rats meet the criteria for the dopaminergic hypothesis of anhedonia and support the construct validity of the model. Finally, the longer duration of immobility in the FST of the RLA line is normalised by antidepressant agents like desipramine and fluoxetine. Hence, we propose that the RLA line is a genetic model of depression that fulfils the criteria of face, construct, and predictive validities. O29.3 ANIMAL KNOCKOUT AND HUMAN STUDIES IDENTIFY SEMA6A AND PLXNA2 AS SCHIZOPHRENIA CANDIDATE GENES Derek Morris1, Annette Runker1, Colm M.P. O’Tuathaigh2, William P. Gilks1, Emma H. Allott1, Kerstin Hakansson2, Fumikazu Suto3, Hajime Fujisawa4, Toshiya Manabe5, Michael Gill1, John L. Waddington2, Aiden P. Corvin1, Kevin J. Mitchell1 1 Trinity College Dublin, Dublin 2 Royal College of Surgeons in Ireland, Dublin 3 National Institute of Genetics, Mishima 4 Nagoya University Graduate School of Sci, Nagoya 5 University of Tokyo, Tokyo Genes involved in axon guidance and cell migration are good candidates for the aetiology of psychiatric disorders with underlying connectivity defects. Here, we combine neuroanatomical and behavioural studies in animal knockout models with association analyses in human samples to investigate the role of SEMA6A (5q21–22) and PLXNA2 (1q32.2) in schizophrenia susceptibility. SEMA6A, with its receptor PLXNA2, is involved in axon guidance and cell migration in the developing brain. Both genes map to schizophrenia linkage peaks. In homozygous SEMA6A mutant mice, we observed a spectrum of neuroanatomical phenotypes involving regions implicated in schizophrenia (e.g. defects in axon guidance of thalamocortical projections and the anterior commissure). Overlapping defects were observed in PLXNA2 mutant animals. Behavioural data from SEMA6A knockout mice indicate that disruption of SEMA6A function is associated with changes in both distinct elements of exploratory behaviour within the natural repertoire, as well as abnormalities in general motor function. SEMA6A and PLXNA2 were tested for association in an Irish sample of 373 schizophrenia cases and 812 controls. Co-incidentally, association was reported elsewhere between schizophrenia and PLXNA2 (PMID: 16402134). We did not replicate association with PLXNA2 but detected association with a variant in a conserved region of intron 3 of SEMA6A (rs258015, p ¼ 0.022). These parallel and complementary studies indicate that: (i) SEMA6A plays an important role in wiring of neural systems implicated in schizophrenia; (ii) animal knockouts for SEMA6A exhibit abnormal interaction with their environment; (iii) variation in this gene may contribute to schizophrenia in humans. O29.4 G72/G30 TRANSGENIC MICE AND GENE EXPRESSION PROFILING Chunyu Liu, Cheng Lijun, Hattori Eiji, Nakajima Akira, Tang Yaping, S. Gershon Elliot University of Chicago, Chicago, IL The G72/G30 gene region has multiple reports of association with both bipolar disorder and schizophrenia. We have introduced a human BAC clone containing G72/G30 gene into B6/CBA mice. In the F4 generation of the transgenic mice, real-time PCR revealed stable transmission of BAC DNA between generations. We performed G72/G30 expression
738
Abstracts
study in cDNAs from 16 different human tissues, fetal brain, five brain subregions (amygdala, substantia nigra, thalamus, cerebral cortex, hippocampus), and RNA from human fetal and adult whole brains. Eight new alternative splicing forms of G72 were identified, including two isoforms from testis; two from brain (substantia nigra and amygdala); four from transgenic mice brain cortex tissues. G72 expresses at very low level in human testis and brain. In fourteen other human tissues no G72 expression was detected using multiple methods. G72 expresses at a high level in transgenic mouse brain cortex. The transgenic mice may form a living model to study G72 function and its role in disease etiology. O29.5 A MICROARRAY GENE EXPRESSION STUDY OF THE EFFECTS LITHIUM CARBONATE ON MOUSE BRAIN MRNA IN ORDER TO UNDERSTAND THE NEUROBIOLOGY OF MOOD STABILISATION AND TREATMENT OF BIPOLAR AFFECTIVE DISORDER Andrew McQuillin, Mie Rizig, Hugh Gurling University College London, London Objectives: Lithium is the most widely prescribed and effective moodstabilizing drug used for the treatment of bipolar disorder. In order to understand how lithium produces changes in the brain we studied brain mRNA from ten mice after treatment with lithium and mRNA from ten controls. Methods: We used MAS5.0, Smudgeminer and GC-RMA packages to determine gene expression changes in the mouse brain using Affymetrix MOE430v2 microarrays after two weeks of lithium treatment. Results: After Bonferroni correction 136 genes showed significant changes in expression. Three genes with the most changed mRNA expression were alanine glyoxylate aminotransferase2 like1 (Agxt2l1), c-mer proto-oncogene tyrosine kinase (Mertk) and sulfotransferase family 1A phenol-preferring member1 (Sult1a1). Also amongst the 136 genes that survived Bonferroni correction were the period gene Per2 (2.47 fold increase), the glutamate receptor Grm3 (0.70 fold decrease) and secretograninII (Scg2 1.28 fold increase) as well several myelin related genes and protein phosphatases. Taking a significance value of p < 0.05 identified a further 2940 genes showing possible changed mRNA expression. These mRNAs included several that had already been implicated in response to lithium such as BDNF (1.44 fold increase) beta phosphatidylinositol transfer protein, (Pitpnb 1.26 fold increase) and inositol myo-1-monophosphatase1 (Impa1 1.22 fold increase). Changes in mRNA expression were found in 45 other genes related to phosphatidylinositol metabolism. Gene ontology analysis showed that lithium significantly affected a cluster of processes associated with nucleotide and nucleoside metabolism.
protein node for haloperidol but cAMP response element binding protein (CREB) was a major node for Clozapine. Ng A et al. (2006) pSTIING: Nucleic Acids Res. 2006; 34:D527–34.
O29.7 GABAB RECEPTORS REGULATE SPONTANEOUS PPI DEFICIT IN DBA/2 MICE M. Bortolato, R. Frau, M. Orru`, M. Fa`, F. Genogu, M. Puligheddu, F. Marrosu, G. Mereu University of Cagliari Psychotic patients typically exhibit deficits in informational processing. In particular, one of the major neuropsychological endophenotypes in schizophrenia-spectrum disorders is a deficit in sensorimotor gating. Gating functions can be dependably measured through the evaluation of the prepulse inhibition (PPI) of the acoustic startle reflex (ASR), consisting in the reduction of the ASR that occurs when eliciting stimulus is immediately preceded by a nonstartling prestimulus. In animals, PPI can be disrupted by psychotomimetic drugs and efficiently restored by antipsychotics. Although drug-induced PPI impairments offer a viable animal model of psychotic disorders, they cannot reliably mimic the genetic and neurophysiological impairments that underpin such disturbances. Thus, animals exhibiting spontaneous PPI deficits are highly preferable for genetic studies and as a framework for inbreeding protocols aimed at the segregation of genes critically involved in schizophrenia. Previous findings have documented that DBA/2 mice spontaneously exhibit low PPI, which is enhanced by antipsychotics. Our lab has characterized that baclofen, the GABAB receptor agonist, restores PPI deficits induced by NMDA receptor blockade in rats. In view of these premises, we tested the impact of baclofen in PPI of DBA/2 mice. Baclofen (5 mg/Kg i.p.) consistently increased PPI in a fashion similar to the atypical antipsychotics, suggesting a role of GABAB receptors in the control of information processing. Interestingly, such increases were prevented by the GABAB antagonist SCH50211 (50 mg/Kg i.p.). Our data point to GABAB receptors as an important substrate for sensorimotor gating reduction in DBA/2 mice, and highlight DBA/2 mice as a target for further investigations, aimed at the characterization of the genetic imbalances that underpin gating deficits.
O29.6 A STUDY OF PROTEIN NETWORKS AND MICROARRAY GENE EXPRESSION CHANGES INDUCED BY HALOPERIDOL AND CLOZAPINE IN THE MOUSE BRAIN Hugh Gurling University College London, London
P01 PROTECTIVE EFFECT OF DYSBINDIN GENE (DTNBP1) FOR MOOD DISORDERS Chi-Un Pae*1,2, Alessandro Serretti3, Laura Mandellic3, Diana De Ronchic3, Ashwin A. Patkarb2, Jung-Jin Kima1, Chul Leea1, In-Ho Paika1 1 Department of Psychiatry, Kangnam St.Mary’s Hospital, The Catholic University of Korea College of Medicine 2 Department of Psychiatry and Behavioral Sciences, Duke University Medical Center, Duke University Department of Psychiatry 3 St. Paul’s Hospital, The Catholic university of Korea College of Medicine 3 Institute of Psychiatry, University of Bologna, Bologna, Italy
We carried out a gene expression study of the effects of clozapine and haloperidol. In the mouse brain 1,457 genes were significantly (p < 0.05) up or down regulated by clozapine but only 361 were significantly changed by haloperidol. Using a t test (p < 0.05), 436 genes were differentially expressed by the two drugs. cDNA arrays and in situ hybridisation were then used to confirm changes found with Affymetrix arrays. The cannabis receptor cnr1 was down regulated by clozapine (p ¼ 0.03). Brain derived neurotrophic factor was up regulated by clozapine. The gene most up regulated by clozapine was foxp1. Both drugs down regulated huntingtin interacting protein 1, a neuronal apoptosis gene. Both Clozapine (t test p ¼ 0.009) and haloperidol (t test p ¼ 0.031) down regulated presenilin 1 gene expression. Both drugs up regulated microtubule associated protein 1 Tau. Other genes genetically implicated in schizophrenia such as PCM1 and the serine threonine kinase gene UHMK1 showed decreased expression in response to clozapine. Changes in mRNA were mapped on to known protein interaction networks in the mouse and human (Ng et al 2006). A very different distribution of the nodes of convergent protein and mRNA change were observed for the two drugs. For example Receptor-interacting serine-threonine kinase 3 was a key
Dysbindin gene (DTNBP1) is localized in the synaptic terminals in wide anatomical regions in the brain and it is involved in the modulation of synaptic signal transduction and plasticity. Recent studies suggested that DTNBP1 may be considered as a possible susceptibility gene for schizophrenia and mood disorders. The authors investigated a possible association between DTNBP1 gene variants and mood disorders. Four SNPs within DTNBP1 (rs3213207, rs1011313, rs760761 and rs261952) were genotyped for 151 patients with bipolar disorder I (BD), 188 patients with major depression (MD) and 478 controls. We observed a significant protective association of the haplotype A-C-T-A both in BD (all SNPs - p ¼ 0.00016) and MD (all SNPs - p ¼ 0.0007). Sliding windows analysis revealed a major contribution due to T-A haplotype both in BP (the last 2 SNP - p ¼ 0.00007) and MD (the last 2 SNP p ¼ 0.000026). Single marker and subgroup (e.g., psychotic features, age at onset, family history, etc) analyses showed no significant association. In conclusion, specific DTNBP1 haplotypes, previously associated with schizophrenia, may be also associated with mood disorders. Studies from different ethnicities will be necessary to confirm our findings.
Abstracts P02 COMORBID PSYCHIATRIC DISORDERS IN SUBJECTS WITH FAMILIAL MAJOR DEPRESSIVE DISORDER Maaike Verhagen1, Joost Janzing2, Annemarie Van der Meij2, Maartje Louter2, Barbara Franke2, Jan K. Buitelaar2 1 Radboud University Nijmegen Medical Cent, Nijmegen 2 Department of Psychiatry, Radboud University, Nijmegen Introduction: Major Depressive Disorder (MDD) is often associated with high rates of lifetime psychiatric comorbidity. However, knowledge regarding the prevalence of comorbid disorders in familial MDD is scarce. This study investigates which lifetime comorbid psychiatric disorders are present in adults with lifetime MDD with familial vulnerability to MDD. In addition, gender, severity, onset and course of MDD in relation to comorbidity are studied. Methods: 125 Dutch adults with lifetime MDD, according to DSM-IV, and at least one first-degree relative positive for MDD, were phenotyped for psychiatric disorders using the Composite International Diagnostic Interview. Odds ratios and 95% confidence intervals (CI) were calculated for gender, onset, severity and course of MDD. Results: Psychiatric comorbidity was observed in 87.2% of participants. The most prevalent comorbid disorders were Generalized Anxiety Disorder (49%), Dysthymia (47%) and Specific Phobias (37%). Early age of onset of MDD ( < 26 yrs) was particularly associated with high risk of comorbidity. Compared to women, men had significantly more comorbid Alcohol Use Disorders (OR 4.6 95% CI 1.5–.13.4, P ¼ .003). Panic Disorder was more prevalent in severe MDD, as compared to mild MDD (OR 3.6, 95% CI 1.6–7.9, P ¼ .001). Conclusions: Comorbid disorders are highly prevalent in familial MDD. The presence of particular lifetime comorbid disorders may be dependent on MDD characteristics like age of onset and severity. Clinical heterogeneity in MDD may reflect underlying genetic subtypes of MDD and should be taken into account in genetic studies. P03 ASSORTATIVE MATING FOR OBSESSIVE-COMPULSIVE BEHAVIOUR IN A POPULATION-BASED SAMPLE Daniel Sebastiaan van Grootheest, S.M. van den Berg, J.J. Hottenga, D.C. Cath, D.I. Boomsma Vrije Universiteit, Amsterdam Background: Assortative mating means that mated pairs are more similar in a phenotypic trait than would be expected by chance. If a significant marital correlation exists for liability to psychiatric disorders, this should be taken into account when analysing data from twin and family studies, because the effects may be confused with shared environmental factors. We aimed to test whether there is a significant association for OC behaviour between husbands and wives in a population-based sample. We further evaluated whether marital resemblance for OC behaviour is primary or secondary to assortment for correlated variables. Methods:OC behaviour was measured by 12 items on the Padua Inventory Revised. We assessed correlation betweens twins and their spouses, and between the twins’ fathers and mothers. A model for mate selection addressed whether the correlation between mates for OC behaviour arises from direct assortment or through correlation with other variables for which assortment occurs. Results: Small but significant correlations of between.1 and.2 for OC behaviour were found between parents of the twins and between the twins and their partners, indicating that resemblance between spouses does not increase as a function of duration of marriage. Only a small amount of the observed marital resemblance for OC behaviour could be explained by assortment for correlated variables such as age, religious attendance and education. Conclusions: Little primary assortment exists for OC behaviour. Ignoring this assortment within a twin study causes only a negligible bias. P04 DISTAL AND PROXIMAL CORRELATES OF SUICIDE ATTEMPTS IN SUICIDAL IDEATORS: A POPULATION-BASED STUDY Jelena Brezo1, Joel Paris2, Richard Tremblay3, Frank Vitaro3, Mark Zoccolillo4, Martine He´bert5, Gustavo Turecki1 1 McGill group for Suicide Studies, Montreal, QC 2 Department of Psychiatry, McGill University, Montreal, QC 3 GRIP, University of Montreal, Montreal, QC
739
4
McGill University, Montreal, QC University of Quebec, Montreal, QC
5
Identification of factors that distinguish between ideators who act on their suicidal thoughts from those who do not is an important research goal, given that such thoughts may not only be correlates but also predictors of more serious suicidal acts. We examined putative correlates of suicide attempts in suicidal ideator members of a French-Canadian, school-based cohort. Suicidal thoughts were evaluated in adolescence and early adulthood. Analyses were performed on the total sample of suicidal ideators, persistent ideators and ideating men and women. The most consistent correlates of suicide attempts were psychiatric psychopathology and gender. Externalising disorders were significant contributors in persistent ideators [drug misuse (OR ¼ 3.4; 95% CI 1.3- 9.0)] and in men [disruptive disorders (OR ¼ 5.9; 95% CI 2.2–16.0)]. In women, adult psychiatric comorbidity also had a significant effect (OR ¼ 1.6; 95% CI 1.1–2.1). Childhood sexual abuse was of relevance in both women (OR ¼ 1.2;95%CI 1.1–1.4) and persistent ideators (OR ¼ 1.3; 95% CI 1.04–1.5). Personality traits showed a gender-specific contribution, with affective instability (OR ¼ 1.1; 95% CI 1.01–1.1) and anxiousness (OR ¼ 1.3; 95% CI 1.1– 1.7)contributing in men and disruptive aggression (OR ¼ 1.1; 95% CI 1.03–1.3) in women. The heterogeneity in suicide attempts correlating across subgroups of ideators observed in our population-based study deserves further exploration and follow-up. P05 THE s/s GENOTYPE AT THE 5-HTTLPR POLYMORPHISM INCREASES RISK FOR DEPRESSION IN A PRIMARY CARE SAMPLE Margarita Rivera1, Blanca Gutie´rrez1, Esther Molina1, Francisco Torres1, Juan Bello´n2, Berta Moreno-Ku¨stner1, Jose´ A. Lorente1, Jorge Cervilla1 1 Universidad de Granada, Avd. Madrid, 11 Granada 2 CAP El Palo Ma´laga, Ma´laga Inconclusive results exist on association between the 5-HTTLPR polymorphism at the serotonin transporter gene and depression. This case-control study sets out to further explore the association between the 5- HTTLPR marker and ICD-10 depressive episode using a large primary care sample. 737 Caucasian primary care Spanish attendees, who took part in the PREDICT study (King et al., 2006), were included in the analysis. 262 participants fulfilled ICD-10 criteria for a depressive episode and the remaining 475 were considered controls. Depressive symptoms were elicited using the Composite International Diagnostic Interview. Subjects were genotyped for the 5-HTTLPR polymorphism to establish s/s, s/l or l/l genotypes. Data were gathered on gender, age, education, living arrangement and family history of psychological problems. Data were analysed using multivariate logistic regression to adjust for potential confounders. We found a rough association between the s/s genotype and depression (P ¼ 0.009; OR ¼ 1.541; 95%CI: 1.092–2.176). Such association remained significant after adjusting for gender, age, living alone, educational level and family history of psychological problems (P ¼ 0.032; OR ¼ 1.476; 95%CI: 1.034–2.108). In addition, there was a statistically-significant lineal trend association between increasing number of s alleles and depression (Mantel-Haenszel 2 ¼ 4.42, p ¼ 0.036). In conclusion, our data suggest liability for depression amongst s allele carriers. Acknowledgements: We thank Prof. Michael King and the rest of the PREDICT-study core group for their collaboration. P06 AN ASSOCIATION CANDIDATE-GENE STUDY ON GENETIC VARIATIONS WITHIN HTR2A IN PATIENTS SUFFERING FROM PANIC DISORDER Paul Gerson Unschuld1, Angelika Erhardt2, Marcus Ising2, Susanne Lucae2, Stefan Kloiber2, Martin Kohli2, Elisabeth B. Binder3, Bertram Mu¨ller-Myhsok2, Martin E. Keck3, Florian Holsboer2 1 Max Planck Institut fu¨r Psychiatrie, Mu¨nchen 2 Max Planck Institut fu¨r Psychiatrie, Mu¨nchen 3 Department of Psychiatry and Behavioral, Atlanta 3 Division of Psychiatric Research, Zu¨rich Anxiety disorders and specifically panic disorder (PD) are caused by complex interactions of environmental and genetic factors. We performed an association candidate-gene approach and analysed 15 single nucleotide polymorphisms (SNPs) within the gene coding for the
740
Abstracts
serotonin-receptor 2A (HTR2A) in patients suffering from PD and a control sample. We found that the SNP rs2296972 shows an association between the number of T-alleles and severity of symptoms in PD. By performing tests according to the Fisher product method, an association between HTR2A and the personality trait reward dependence could be shown. The polymorphisms rs3742278, rs2296972 and rs2770292 form a haplotype, which may be associated with higher susceptibility to PD. P07 THREE CIRCADIAN CLOCK GENES CONTRIBUTE TO THE SEASONAL PATTERN IN MOOD DISORDER Jens Treutlein1, Gunter Schumann2, Asude Alpman3, Josef Frank4, Carolina Johansson3, Martin Depner4, Liviu Aron3, Marcella Rietschel4, Stefan Wellek5, Pia Soronen6, Tiina Paunio6, Andreas Koch7, Ping Chen8, Mark Lathrop8, Rolf Adolfsson9, Maj-Liz Persson10, Siegfried Kasper11, Martin Schalling3, Leena Peltonen4, Timo Partonen5 1 Central Inst. of Mental health, Mannheim 2 Section of Addiction Biology, Institute, London 3 Dept of Molecular Medicine, Karolinska Inst., Stockholm 4 Div.Genetic Epidemiology in Psych, Mannheim 5 Dept. Biostatistics, Central Ins, Mannheim 6 Dept. Molecular Medicine, Nation, Helsinki 7 Inst. for Clinical Molecular Biology, Kiel 8 Centre National de Ge´notypage, Evry 9 Division of Psychiatry, Dept Cl, Umea˚ 10 Section of Psychiatry, Neurotec Dept, Stockholm 11 Dept. of General Psychiatry, Medica, Vienna 4 Program in Medical and Population Genetics, Boston 5 Dept. Mental Health and Alcohol, Helsinki Seasonal affective disorder (SAD) is characterized by genetic and environmental components. We hypothesized that three circadian clock genes forming a functional unit, period homolog 2 (Per2), aryl hydrocarbon receptor nuclear translocator-like (Arntl) and neuronal PAS domain protein 2 (Npas2), could affect the phenotype of SAD. We analysed the three genes in a sample of 189 patients and 189 matched controls. The a priori framework for selection of SNPs was guided by assessing potential functionality in silico. Gene-wise logistic regression analysis revealed that SAD is associated with variations in each of the three genes: in a combined genetic analysis, we found additive effects of the three genes. We hypothesize that the circadian clock genes play a role in the pathogenesis of SAD and present a hypothetical mechanism of action for the observed effect: SNP10870 of Per2, which is located in an intronic region, showed association with the SAD phenotype. Per2 is known to be an important regulator of the entire transcriptional network of endogenous clock genes. The stretch of DNA in which SNP10870 is located may constitute an intronic transcriptional enhancer element. The rare risk allele of the SNP may create a binding site for SP1, which is not present in most of the normal population, as indicated by the low amount in the controls. Supportive evidence for the hypothesis of quantitative Per2 action is provided by fibroblast cell lines over-expressing mPer2 mRNA levels, which severely impairs the expression of circadian clock genes (Yamamoto et al., 2005). P08 P2RX7, A GENE CODING FOR A PURINERGIC LIGAND-GATED ION CHANNEL, IS ASSOCIATED WITH MAJOR DEPRESSIVE DISORDER Susanne Lucae1, Daria Salyakina1, Nicholas Barden2, Mario Harvey2, Bernard Gagne´2, Michel Labbe´2, Elisabeth Binder3, Manfred Uhr1, Marcelo Paez-Pereda2, Inge Sillaber2, Marcus Ising1, Tanja Brueckl1, Roselind Lieb1, Florian Holsboer1, Bertram Mu¨ller-Myhsok1 1 Max Planck Institute of Psychiatry, Munich 2 Neuroscience, CHUL Research Centre and U, Quebec 3 Department of Psychiatry and Behavioral Sciences, Atlanta 2 Affectis Pharmaceuticals, Munich The P2RX7 gene is located within a region on chromosome 12q24.31 that has been identified as a susceptibility locus for affective disorders by linkage and association studies. P2RX7 is a purinergic ATP-binding calcium channel expressed in neurons, as well as in microglial cells in various brain regions. We investigated 29 single nucleotide polymorphisms (SNPs) within the P2RX7 gene and adjacent genes in a sample of 1000 German Caucasian patients suffering from recurrent major depressive disorder(MDD). They were contrasted with diagnosed
healthy Caucasian controls from the same population (n ¼ 1029). A non-synonymous coding SNP in the P2RX7 gene (rs2230912), previously found to be associated with bipolar disorder, was significantly associated(p ¼ 0.0019) with MDD. This polymorphism results in an amino acid exchange in the carboxy-terminal cytosolic domain of the P2RX7 channel protein, suggesting that the observed P2RX7 polymorphism might play a causal role in the development of depression. P09 POLYMORPHISMS IN LEPTIN ARE ASSOCIATED WITH RESISTANCE TO ANTIDEPRESSANT TREATMENT AND COGNITIVE IMPAIRMENT IN MAJOR DEPRESSION Stefan Kloiber, Stephan Ripke, Martin Kohli, Simone Reppermund, Thomas Bettecken, Daria Salyakina, Hildegard Pfister, Marcus Ising, Susanne Lucae, Sonja Horstmann, Tatjana Dose, Andreas Menke, Benno Puetz, Bertram Mu¨ller-Myhsok, Josef Zihl, Florian Holsboer Max Planck Institute of Psychiatry, Munich The discovery of the adipose tissue hormone leptin as a signal to the brain regarding the amount of peripheral fat stores was crucial. Furthermore, leptin is involved in several CNS functions like cognition, sleep, behaviour and appetite regulation, and evidence is emerging that leptin plays a role in brain development, e.g. hippocampal formation. To investigate the role of leptin in major depressive disorder (MDE), we genotyped single-nucleotide polymorphisms (SNPs) in the leptin gene in depressed patients. In 338 inpatients with MDE, we assessed weekly psychopathological ratings during hospitalisation. Leptin plasma levels, attention and memory function were analysed on admission and prior to discharge. 18 SNPs in the leptin gene region were genotyped. Quantitative traits were analysed by ANOVA, qualitative traits by Fisher-Exact-Test, both after correction for age, gender and Body-Mass-Index. 5 different inheritance-models were investigated. We found significant associations of several SNPs in the leptin gene with resistance to treatment with antidepressants (p < 0.01) and with the number of treatment attempts (p < 0.01). Various SNPs were associated with impaired attention performance (p < 0.01) and impaired memory function (p < 0.01). After Westfall/Young correction for multiple testing with 1.000 permutations, all results remained significant at a level of 0.05. In addition, we used the Fisher product method using 100.000 permutations (p < 0.01). These findings indicate association of the leptin gene with resistanc to treatment with antidepressants and impaired cognitive function in depressed individuals.
P10 ASSOCIATION OF POLYMORPHISMS IN P2RX7 AND CAMKKB WITH ANXIETY DISORDERS Angelika Erhardt1, Susanne Lucae1, Paul G. Unschuld1, Marcus Ising1, Nikola Kern1, Daria Salyakina1, Roselind Lieb1, Manfred Uhr1, Elisabeth Binder2, Martin E. Keck1, Bertram Mu¨ller-Myhsok1, Florian Holsboer1 1 Max Planck Institute of Psychiatry, Munich 2 Department of Psychiatry and Behavioral Sciences, Atlanta There is considerable evidence that genetic factors play an important rol in the pathophysiology of affective disorders including bipolar disorder, major depressive disorder and anxiety disorders. Long-term follow up studies, as well as drug treatment studies, suggest that these clinica conditions share a number of pathophysiological characteristics, including genetic variables. One possible candidate region is located on chromosome 12q24.31, originating from previous linkage and association studies with bipolar disorder. This region contains two candidate genes for purinergic ligand-gated ion channels, P2RX7 and P2RX4, and one gene coding for calmodulin dependent protein kinase kinase beta (CaMKKb). We investigated the genetic associations between 26 SNPs in the candidate genes P2RX7, P2RX4 and CaMKKb on chromosome 12q24.31 in 179 patients with anxiety disorder and 462 healthy controls. Nominal case-control associations could be detected for the P2RX7 polymorphism rs6489795 (Ser357Thr) and for a SNP in the 50 UTR region of P2RX4, which did not remain significant after correction for multiple testing. We found, however, a prominent association between severity of panic and agoraphobia symptoms and an exonic SNP (rs3817190) in the CaMKKb gene and a trend for association with an exonic SNP in P2RX7 (rs1718119) with severity scores on the panic- and agoraphobia scale. The locus 12q24.31 seems to be an important genetic region for anxiety, bipolar and unipolar
Abstracts disorder, further confirming the hypothesis of a genetic overlap in the group of affective disorders.
P11 CASE-CONTROL STUDY OF SEROTONIN TRANSPORTER POLYMORPHISM INMAJOR DEPRESSIVE DISORDER Krisztina Hejjas1, Eszter Domotor2, Anna Szekely3, Gabriella Balogh2, Andrea Sarosi2, Gabor Faludi2, Maria Sasvari-Szekely3 1 Semmelweis University, Budapest 2 Dept. Psychiatry, Ku´tvo¨lgyi Clinical Cen, Budapest 3 Institute of Psychology, Eo¨tvo¨s Lora´nd U, Budapest 3 Inst. Medical Chem. Mol. Biol. and Patho, Budapest Several candidate gene association studies have shown the importance of the polymorphic alleles of the Serotonin transporter (SLC6A4) as possible risk factors of depressive disorders. Recently, several metaanalyses have been conducted on clinical studies related to both the 44 base-pair insertion/deletion polymorphism of the SLC6A4 promoter (5-HTTLPR) and intronic repeat variants (StIn2) comprising 9, 10 and 12 copies of a 17- base-pair sequence, suggesting the role of the StIn2 rather than the 5-HTTLPR polymorphism as a risk factor in completed suicide and schizophrenia. Here we investigated the two common serotonin transporter gene polymorphisms as possible risk factors for major Depressive Disorders (MDD) in a case-control setup. The genotype frequencies of the 5-HTTLPR and the StIn2 polymorphism did not show significant deviation from the Hardy-Weinberg equilibrium in our Caucasian samples of clinical and control subjects. Association analysis did not show significant differences between the 5HTTLPR genotype frequencies of the control and the clinical population. On the other hand, comparison of StIn2 genotype distribution revealed a statistically-significant difference: the relatively rare 10/10 genotype was twice as frequent among psychiatric patients. These results suggest that the StIn2 polymorphism of the serotonin transporter gene might represent a genetic risk factor for MDD. Our results, together with the outcome of a recent meta-analysis by deLara et al. (2005) demonstrate the specific role of the 10 repeat allele in predisposition to depression and suicide.
P12 POLYMORPHISMS IN THE GLUCOCORTICOID RECEPTOR GENE AND SUSCEPTIBILITY TO MAJOR AFFECTIVE DISORDERS Detelina Grozeva1, Elaine Green2, Sian Caesar3, Sally Hyde3, Christine Fraser2, Ian Craig3, Anne Farmer3, Allan Young3, Nicol Ferrier3, Peter McGuffin3, Lisa Jones3, Ian Jones2, Michael Owen2, George Kirov2, Nick Craddock2 1 Department of Psychological Medicine, Cardiff 2 Department of Psychological Medicine, Ca, Cardiff 3 Department of Psychiatry, University of, Birmingham 3 Institute of Psychiatry, London A major hypothesis that has been suggested for the pathogenesis of Bipolar and Unipolar disorders is the corticosteroid receptor hypothesis. It focuses on the impaired corticosteroid receptor signalling as a primary factor for the developing of the disorders. To understand Bipolar disorder, we have to understand the mechanisms that control the range and stability of emotions, and it is shown that variations in expression of the glucocorticoid receptor (GR) gene can contribute to the tuning of emotional stability. There is evidence that chronic dysregulation of hypothalamus-pituitary-adrenal axis activity could be associated with the onset and course of Depression. The GR gene is a member of the steroid receptor family and is located on chromosome 5– 5q31.3. In the brain, GR is thought to modulate emotional behaviour and cognitive functions. We examined whether four polymorphisms (a SNP altering a BclI RFLP, N363S, rs33388 and rs33389) in the GR gene confer susceptibility to Affective Disorders. These SNPs have been shown to be associated with changes in sensitivity to glucocorticoids. We studied two large well-diagnosed samples of patients with Bipolar Disorder and Major Depression, and matched controls. We did not observe association between developing the diseases and any of the variants. Our results show that none of these polymorphisms exerts a major influence on susceptibility to Bipolar and Unipolar Affective disorders. Currently we are examining the data to study the haplotype pattern, and to test the possible influence of clinical covariates and subtypes.
741
P13 THE SEROTONIN TRANSPORTER GENE AND DEPRESSION IN PSYCHOTIC DISORDERS Michael Escamilla1, Humberto Nicolini2, Betty Camarena2, Javier Contreras1, Alfonso Ontiveros2, Henriette Raventos3, Ricardo Mendoza4, Rodrigo Munoz5, Albana Dassori1, Regina Armas2, Rolando Medina1, Salvador Contreras1 1 University of Texas Health Science Center, San Antonio, TX 2 Study Group of Medicine and Families CAR, Mexico City 2 INFOSAME, Monterrey, NL 3 University of Costa Rica, San Jose 4 UCLA, Los Angeles, CA 5 Family Health Centers of San Diego, San Diego, CA 2 University of California at San Francisco, San Francisco, CA Variation in the serotonin transporter gene (5HTT) has been shown to influence depression in persons who have been exposed to a number of stressful life events. Subjects with a history of psychotic disorders experience a number of unfavourable life events such as unemployment, difficulties with marital relationships, homelessness, medical illness and legal problems. The present study tested the association between the 5-HTTLPR polymorphism and a history of lifetime depression in persons with psychotic disorders. Subjects were originally recruited for a family-based linkage study of schizophrenia and schizoaffective disorder. One subject per family (the first affected subject recruited in each family) was used for the current analysis (N ¼ 129). A best-estimate consensus process was used to assign final diagnoses. We genotyped each subject at the s/l promoter region polymorphic site. Consensus best-estimate diagnoses for the 129 subjects revealed a range of psychotic disorders: Schizophrenia (105), Schizoaffective Disorder (18), other psychotic disorders (6)). All subjects had a history of psychosis. Of the total sample, 77 individuals (60%) had had at least one full episode or syndrome of major depression. We found a significant association between having the ‘‘ss’’ or sl’’ genotype and having a history of depression (X2 ¼ 4.84, p ¼ 0.028). This finding suggests that having an ‘‘ss’’ or ‘‘sl’’ genotype at the 5-HTTLPR promoter polymorphic locus increases the risk of persons with psychotic disorders of developing major depression during the course of their illness.
P14 DEFINING THE PHENOTYPE: DEPRESSIVE EPISODES IN BIPOLAR AND UNIPOLAR DISORDERS Liz Forty1, Lisa Jones2, Ian Jones1, Nick Craddock1 1 Cardiff University, Cardiff 2 University of Birmingham, Birmingham Whether depressive states in bipolar disorder are aetiologically similar to those seen in unipolar disorder is not clear. The aim of the current study is to assess whether there are specific features of depressive episodes that could be used in genetic studies to identify depression that is aetiologically related to mania. We compared depressive episodes in individuals meeting DSMIV criteria for bipolar I disorder (BPI N ¼ 404) and individuals meeting DSMIV criteria for Recurrent unipolar major depression (MDDR N ¼ 601). All individuals were recruited as part of our ongoing molecular genetic studies of affective disorders and were interviewed using the Schedules for Clinical Assessment in Neuropsychiatry, reviewing case notes. We found significant differences between the two groups in several variables: the bipolar group was characterised by more episodes of depression; depressive episodes in the bipolar group were characterised by a shorter duration of the longest episode and increased suicidal behaviour. Typical depressive symptoms, such as initial insomnia and weight loss, were significantly more prevalent in the unipolar group, whereas the bipolar group experienced significantly more hypersomnia and slowed activity. Our findings suggest that, although there are numerous similarities between depressive episodes in unipolar and bipolar disorders, there are also a variety of clinical differences. These may be useful in identifying individuals with depressive episodes who are more likely to have a ‘‘bipolar’’ pathogenesis.
P15 THE NEUROPSYCHIATRIC PHENOTYPE OF DARIER’S DISEASE Katherine Gordon-Smith1, Lisa Jones2, Nick Craddock1 1 Cardiff University, Birmingham 2 University of Birmingham, Birmingham
742
Abstracts
Darier’s Disease is a rare autosomal dominantly-inherited skin disorder in which co-occurrence of neuropsychiatric abnormalities, including affective disorders, suicidal ideation and mild learning disability, has frequently been reported. Families showing intrafamilial co-segregation of Darier’s Disease and major affective illness have been described. The disease is caused by mutations in a single gene, ATP2A2, which maps to 12q23-q24.1 and encodes SERCA2 calcium pump involved in intracellular calcium transport. Because SERCA2 is expressed in the brain as well as skin, it is plausible that ATP2A2 mutations could have neuropsychiatric effects. We have conducted the first systematic study to investigate the neuropsychiatric phenotype in Dariers Disease. A battery of neuropsychiatric tests has been administered to 100 unrelated individuals with Darier’s Disease. This battery contains measures assessing the presence of mood disorders, neurological abnormalities and intellectual functioning. 46 % of the sample had a lifetime DSM-IV diagnosis of affective disorder and 13% had made a suicide attempt. The latter figure is three times higher than that found during a large national survey of psychiatric morbidity among adults in Great Britain, in which 4.4% of respondents reported a suicide attempt during their lives. Our findings support the previously-suggested association between Darier’s Disease and neuropsychiatric abnormalities, including mood disorders and suicidal ideation. Mutation detection is currently being undertaken to allow investigation of genotype-phenotype relationships between the type and locations of mutations in ATP2A2 gene and the neuropsychiatric abnormalities that have been observed. P16 LACK OF ASSOCIATION BETWEEN G1463A (ARG441HIS) TPH2 POLYMORPHISM IN GERIATRIC MAJOR DEPRESSED PATIENTS. Humberto Correa, Maria Aparecida Bicahlo, Guilherme Pimenta, Fernando Neves, Edgar Morais, Luiz Armando De Marco, Marco Aure´lio Romano-Silva UFMG, Belo Horizonte-Minas Gerais
For understanding Bipolar disorder, we have to understand the mechanisms that control the range and stability of emotions and it is shown that variations in expression of the glucocorticoid receptor (GR) gene can contribute to the tuning of emotional stability. There is evidence that chronic dysregulation of hypothalamus-pituitaryadrenal axis activity could be associated with the onset and course of Depression. The GR gene is a member of the steroid receptor family and is located on chromosome 5– 5q31.3. In the brain GR is thought to modulate emotional behaviour and cognitive functions. We examined whether four polymorphisms (a SNP altering a BclI RFLP, N363S, rs33388 and rs33389) in the GR gene confer susceptibility to Affective Disorders. These SNPs have been shown to be associated with changes in the sensitivity to glucocorticoids. We studied two large well-diagnosed samples of patients with Bipolar disorder and Major Depression, and matched controls. We did not observe association between developing the diseases and any of the variants. Our results show that none of these polymorphisms exerts a major influence on susceptibility to Bipolar and Unipolar Affective disorders. Currently we are examining the data to study the haplotype pattern, and to test the possible influence of clinical covariates and subtypes. P18 MALE DEPRESSION AND ADULT ATTACHMENT STYLES: GENETIC AND ENVIRONMENTAL INFLUENCES Carol Franz1, Matthew Pannizon2, Michael J Lyons2, Michael D Grant2, Seth A Eisen3, William S Kremen1 1 University of California San Diego, La Jolla, 2 Boston University, Boston, MA, 3 Washington University, St. Louis, MO
Background: Recently a functional mutation (G1463A) was described and associated with major depression in some patients. This result is however rather controversial since it was not replicated by other studies. A possible reason is the fact that all patients assessed in this first report were 60 years or older. Methods: We investigated exon 11 of the TPH2 gene in samples of patients from a Geriatric Reference Center in Belo Horizonte, Brazil. Our cohort comprised 102 patients (87 females, 15 males, age 62- 94, the mean age at inclusion was 77.3 years) from many ethnic backgrounds, with a diagnosis of late onset major depression according to the DSM-IV and the Geriatric Depression Scale, some of them with severe depression. Results: We did not find the A-allele (His441) in any of the 102 patients nor in the controls. Conclusion: Our data support the notion that this rare TPH2 variant is also not a major determinant of genetic risk for depression in geriatric patients. The high frequency of the minor allele previously found in a geriatric sample might be the result of population stratification. However, considering the failure to replicate their findings in a large independent sample (over 5,000 subjects from different populations) and the frequency of the observation of the His441 variant in the initial sample, including a few homozygotes, possible technical artifacts, other than genotype readings cannot be ruled out.
We examined 350 middle-aged male twin pairs from the Vietnam Era Twin Study of Aging (VETSA) to investigate genetic and environmental influences on depression and attachment style. Anxious attachment style reflects abandonment fears and worries about availability of romantic partners; avoidant attachment reflects tendencies to be emotionally detached or wary. Previous research shows essentially complete genetic overlap between anxiety and depression. The relationship of avoidant attachment and depression has received little attention. However, beyond anxiety, depression includes components of isolation and emotional detachment that might underlie avoidant attachment styles. A trivariate biometrical modeling analysis was performed decomposing the relationship among anxious and avoidant attachment and depression symptoms (CES-D). There was strong evidence for a single latent genetic factor influencing all three measures. Phenotypic correlations among the measures ranged from.36 to.48, but genetic correlations approached unity. Most of the remaining variance was accounted for by non-shared environmental factors specific to each measure. Despite the theoretical assumption that adult attachment styles are learned from childhood family relationships, common environment had negligible influence on attachment measures. However, common environment accounted for about 20% of the variance in depression. It may be genes predisposing to depression also establish susceptibility to emotional difficulties in close relationships. Non-shared environmental influences (e.g., choice of a romantic partner and that partner’s attachment style) may exacerbate or attenuate these difficulties. Both avoidant and anxious attachment styles have similar genetic overlap with depression, but each is relatively independent with regard to environmental influences.
P17 POLYMORPHISMS IN THE GLUCOCORTICOID RECEPTOR GENE AND SUSCEPTIBILITY TO MAJOR AFFECTIVE DISORDERS Detelina Grozeva1, Elaine Green2, Sian Caesar3, Sally Hyde3, Christine Fraser2, Ian Craig3, Anne Farmer3, Allan Young3, Nicol Ferrier3, Peter McGuffin3, Lisa Jones3, Ian Jones2, Michael Owen2, George Kirov2, Nick Craddock2 1 University Hospital of Wales, Cardiff 2 Department of Psychological Medicine, Ca, Cardiff 3 Department of Psychiatry, University of Birmingham 3 Institute of Psychiatry, London
P19 SEROTONIN TRANSPORTER GENE AND SOCIAL SUPPORT MODERATE POST-TRAUMATIC STRESS DISORDER IN HURRICANE EXPOSED ADULTS Karestan Koenen1, Dean Kilpatrick2, Ronald Acierno2, Kenneth Ruggiero2, Sandro Galea3, Heidi Resnick2, John Boyle3, Joel Gelernter4 1 Harvard School of Public Health, Boston, MA 2 Medical University of South Carolina, Charleston, SC 3 Medical University of Michigan, Ann Arbor, MI 3 SRBI, New York City, NY 4 Yale University, New Haven, CT
A major hypothesis that Bipolar and Unipolar hypothesis. It focuses signalling as a primary
In an epidemiologic sample of adults who experienced the 2004 Florida hurricanes, we tested why exposure to natural disasters leads to the development of post-traumatic stress disorder (PTSD) in some individuals and not in others. Random digit dialing was used to recruit
has been suggested for the pathogenesis of disorders is the corticosteroid receptor on the impaired corticosteroid receptor factor for the developing of the disorders.
Abstracts a representative population sample of hurricane affected adults (n ¼ 579 for this analysis). PTSD, prior trauma-exposure, hurricane exposure and social support were assessed via structured telephone interview within 9 months of hurricane exposure. The presence of the short allele of the serotonin transporter gene promoter polymorphism moderated the risk for PTSD associated with high hurricane exposure and low social support. In adults who experienced high levels of hurricane exposure and had few social supports, each s allele was associated with an increased risk of posthurricane PTSD symptoms and diagnosable PTSD (PTSD % s/ s ¼ 17.6%, s/l ¼ 8.8%, comparison ¼ 2.2%,p < .001). Effect of s allele on risk for PTSD symptoms (beta ¼ .25,t ¼ 3.44, p < . 001) and diagnosis (OR ¼ 6.2, 95% CI ¼ 1.3,29.3,p ¼ .02) among adults with high levels of hurricane exposure and low social support persisted after adjusting for sex, age, ancestry, and history of potentially traumatic events. Ancestry was determined by the ancestry-informative marker method (Yang et al., 2005). Similar results were found for post-hurricane depression. These data are the first of which we are aware to suggest that genotype moderates the mental health effects of a recent natural disaster. P20 ASSOCIATION TO MARKERS IN THE 19P13 REGION IN PANIC DISORDER SAMPLES FROM THE GENETICALLY ISOLATED FAROE ISLANDS Hans Atli Dahl1, August Wang2, Thomas Als3, Henrik Ewald3, Ole Mors3, Torben Kruse1 1 Odense University Hospital, Odense 2 Copenhagen University Hospital, Copenhagen 3 Psychiatric Hospital in Aarhus, Aarhus 13 patients, meeting the ICD-10 and DSM-IV criteria for panic disorder with agoraphobia, from a sub-isolate of the Faroe Islands were used in a genome-wide search for candidate loci. In the search for possible shared segments inherited identically by descent, the patient haplotypes were reconstructed from the parental genotypes and compared to haplotypes from 43 healthy individuals from the same region on the Faroe Islands. Although sampled independently, a subsequent genealogical assessment revealed a common ancestor 12 generations back for all 13 patients, indicating possible founder effect. Analyses were performed on either single marker allele frequencies or two locus haplotype frequencies as implemented in the program CLUMP (Shamand Curtis 1995). Significant association below a predefined cut-off p-value of 0.005 was observed in the single marker analysis at marker D19S394 and likewise below a predefined cut-off p-value of 0.01 for a two locus haplotype segment involving markers D19S583-D19S394 on chromosome 19p13. A closer look at the region showed LD extending both centromeric and telomeric to the two-locus haplotype, indicating probable IBD inherited regions. Within the presumable 11 cM. IBD region, two separate regions are observed with a relatively skewed distribution of segments in the case group, producing low p-values in Fisher’s exact test. The sizes of the narrower regions are 2 cM. and 6 cM. respectively, and we suggest that one of these regions harbours a susceptibility gene for panic disorder. P21 5HTTLPR GENOTYPE INFLUENCE ON HIPPOCAMPAL VOLUME AND DEPRESSION IN A FEMALE TWIN SAMPLE Kelly Botteron1, Elizabeth Lobos2, Casey Babb3, Tomoyuki Nishino2, Alexandre Todorov2, Michael Miller3, Andrew Heath3, Vesselin Chorbov3, Richard Todd3 1 Washington University School of Medicine, St Louis, MO 2 Washington University, St Louis, MO 3 Washington University, St Louis, MO 3 Johns Hopkins University, Baltimore, MD Some studies report an interaction of serotonin transporter allele (5HTTP) polymorphisms, trauma history and Major Depressive Disorder (MDD). Additionally, a significant relationship between 5-HTTP alleles and limbic activation and regional brain volume has been reported. The current study reports from an ongoing MRI investigation of early-onset MDD in a sample of young female twins. Employing a classic twin design, monozygotic (MZ) and dizygotic (DZ) twin pairs are enrolled. Affected pairs have at least one twin with a h/o MDD > 4 weeks. MDD diagnosis and trauma data are based on structured interview. Anatomical MRI scans (three T1-weighted MPRAGE) were acquired, coregistered and summed to increase signal
743
to noise ratio. Hippocampal volumes were determined based on published methods (MIM) employing automated high-dimensional fluid deformation mapping. Genotyping was based on established methods (RDT, BAL) to establish S, La or Lg 5HTTP alleles. Recently, Lg alleles have been demonstrated to have low activity similar to S alleles. Allele distribution of sample included 40% S, 52% La and 8% Lg. Volume differences were found between low and high activity alleles: high activity alleles (La) have a 6% right hippocampal reduction (p ¼ 0.0003) in all subjects. Subdividing by MDD had a similar significant right reduction. When including trauma, differences were highly significant: 16.4% reduction associated with trauma and La allele (p ¼ 0.0007). These results lend further support to a potential relationship between 5HTTP and hippocampal volume in MDD subjects which is further modulated by a history of trauma. P22 SNPS IN THE NTRK2 GENE AND GRK3 GENES PREDICT RESPONSE TO LITHIUM IN PATIENTS WITH EUPHORIC AND DYSPHORIC MANIA RESPECTIVELY Susan Leckband1, Rebecca McKinney2, Tatyana Shehktman2, John R. Kelsoe3 1 VA San Diego Healthcare System, San Diego, CA 2 Department of Psychiatry, VA San Diego H, La Jolla, CA 3 Department of Psychiatry, University of, La Jolla, CA Response to lithium shows substantial inter-individual variation that may be genetically influenced. We have previously reported a study of the genetics of lithium response in 92 bipolar patients who were lithium responders and 92 who were non-responders. Lithium response was determined retrospectively using SCID or DIGS interview, detailed life chart and medical records. Consistent with previous literature several clinical traits were associated with poor lithium response including dysphoric mania and rapid cycling. A novel result was poor response in those with co-morbid PTSD. 88 SNPs in 9 genes were examined in these subjects in phase I of the study. In phase II, 48 SNPs in an additional 6 genes and a previous gene were examined and are reported here. Six SNPs in the gene NTRK2 were nominally significant, including the most significant SNP (p ¼ 0.0062). Nominally significant SNPs were also identified in the genes: IMPA1, IMPA2 and Bcl2. Analyses of these data also revealed interactions with the clinical trait of dysphoric mania. A SNP in the NTRK2 gene was associated in an allele dose dependent fashion only in subjects with predominantly euphoric mania. Another SNP in the gene GRK3 was similarly associated but only in those with dysphoric mania. NTRK2 codes for tyrosine kinase receptor Trkb, which is the receptor for the growth factor BDNF that has been implicated in lithium’s mechanism of action. Together these data implicate variation in NTRK2 gene in lithium response and suggest different disease and therapeutic mechanisms in dysphoric mania. P23 INTERACTIONS BETWEEN EARLY TRAUMATIC LIFE EVENTS, AND MAO-AGENOTYPE ASSOCIATED WITH INCREASED PHYSICAL AGGRESSION IN ADULTHOOD Giovanni Frazzetto1, Giorgio Di Lorenzo2, Valeria Carola3, Luca Proietti2, Alberto Siracusano2, Cornelius Gross3, Alfonso Troisi2 1 EMBL, Monterotondo (RM) 2 Department of Neurosciences, School of Medicine, Rome 3 Mouse Biology Unit, European Molecular Biology Lab, Monterotondo (RM) Several recent studies have demonstrated that a functional polymorphism in the monoamine oxidase-A gene (MAO-A LPR) can moderate the association between early adverse life events and increased risk for anti-social behaviour, conduct disorder, and aggression. To examine whether variations in parenting style might also contribute to risk for aggressive and impulsive behaviour in adulthood, and to determine whether such associations could be moderated by the MAO-A LPR, we genotyped a combined population of outpatients and healthy volunteers (N ¼ 197) who had been administered a battery of standardized psychometric tests. By an ANCOVA analysis, our findings confirm a significant positive effect of early traumatic life events (ETLE) on physical aggression as measured by the Aggression Questionnaire (AQ-PA; P ¼ 0.003). Self-reported physical aggression scores were significantly higher in those subjects that had adverse experiences early in development (0–15 years of age). There was no
744
Abstracts
main effect of the low-activity variant of the MAO-A LPR on aggression scores. We found a significant three way gene x environment x sex interaction (P ¼ 0.04) with increased AQ-PA scores for males reporting more than one ETLE and carrying the low MAO-A activity allele. We also examined correlations between parenting style as measured by the Parental Bonding Instrument (PBI) and AQ-PA in persons carrying the low or high functioning MAO-A LPR allele to investigate whether parenting style moderates the effects of ETLE on risk for aggression in genetically susceptible individuals.
P24 CRYPTIC RELATEDNESS AMONG CASES AND CONTROLS IN TWO ISOLATED POPULATIONS Thomas Als1, Hans A. Dahl2, Friedrik P. Wikman3, Tracey J. Flint4, August G. Wang5, Beatriz Marcheco-Teruel6, Torben F. Orntoft3, Anders D. Børglum4, Maria Vang5, Torben A. Kruse2, Ole Mors4 1 Centre For Basic Psychiatric Research, Risskov 2 Department of Clinical Biochemistry and Genetics, Odense 3 Department of Clinical Biochemistry, Ske, Aarhus 4 Centre for Basic Psychiatric Research, Aarhus 5 Dept. of Psychiatry, Amager Hospital, Copenhagen 6 National Center of Medical Genetics, Havana 4 Institute of Human Genetics, University, Aarhus 5 Dept. of Psychiatry, Landssju´krahusid (N), Torshavn Classical case-control analysis might detect differences between cases and controls due to improperly-accounted relatedness among individuals rather than owing to true association between marker and trait. Such cryptic relatedness is an overlooked problem in genetic analyses of isolated populations. We used a sample of cases with bipolar affective disorder, schizophrenia and autism, in addition to 44 randomlyselected controls from the isolated population of the Faroe Islands. We used 660 genetic markers to obtain genetic estimates of kinshipand inbreeding-coefficients for all pairs of individuals, to assess the level of genetic relatedness and inbreeding. Cases were on average no more related to one another than they were to controls. Controls were likewise, on average, no more related to one another than they were to cases. However, although there was no average difference between within-group and between-group levels of relatedness, several pairs of individuals appeared to be significantly related, indicating some degree of cryptic relatedness of both cases and controls. Several individuals also proved to be inbred. Such inter- and intra-individual correlations may lead to spurious associations, but could potentially be corrected by using genetically-estimated kinship-coefficients. Estimation of pairwise kinship-coefficients using a 10K SNP dataset of 25 cases of inbred families and 40 random controls from an isolated sub-region in eastern Cuba will provide another example regarding relatedness and inbreeding. How much impact this pattern of relatedness structure has on association studies is still to be explored.
P25 POLYMORPHISMS OF 5HT2A, 5HTT, MAOA, AND APOE IN INDIGENOUS POPULATION OF MEXICO Blanca Zoila Gonza´lez-Sobrino1, Carlos Cruz2, Marı´a Antonieta Ochoa-Ocan˜a1 1 Universidad Nacional Auto´noma de Me´xico, Mexico City 2 Instituto Nacional de Psiquiatrı´a Ramo´n, Mexico City In this presentation, we report frequencies of allelic variants of 5HT2A, 5HTT-LPR, MAOA (EcoRV, FnuH4I, VNTR) and APOE. They have been utilised as markers in association studies on psychiatric disorders, as there is very little information in this sense regarding the indigenous Mexican population. The great variety and distribution of this population in different parts of the country is due to geographical distances, a series of migrations (before and after the Spanish Conquest)and eventual isolation caused by economic, cultural and historical processes. We studied a sample of 212 unrelated individuals: two different groups of Nahuas from the Sierra Madre Oriental, and Cora/Huichol groups in the Sierra Madre Occidental; they show a heterozygosity value of 0.29 to 0.42. FST statistics, measuring endogamy between populations, is 0.06 in Coras/Huicholes, 0.09 in Nahuas, and 0.13–0.18 in both groups. A third mixed population in Mexico City shows intermediate values. Differences between neighbouring groups are probably a consequence of their strong sense of identity and different life styles.
P26 AN ASSOCIATION STUDY OF SCHIZOPHRENIA WITH COMPLEX POLYMORPHIC REGION IN THE BDNF GENE Yoshimi Iijima1, Takeya Okada2, Ryota Hashimoto3, Tomoko Shizuno2, Hiroaki Hori2, Tetsuo Nakabayashi3, Tadafumi Kato4, Takashi Asada5, Masahiko Tatsumi6, Asako Kosuga6, Kunitoshi Kamijima6, Harada Seiichi3, Arima Kunimasa3, Saitoh Osamu3, Kunugi Hiroshi2 1 National Institute of Neurosciense, NCNP, Tokyo 2 National Institute of Neuroscience, NCNP, Tokyo 3 Department of Psychiatry, Osaka Universi, Osaka 3 Department of Psychiatry, Musashi Hospit, Tokyo 4 Laboratory for Molecular Dynamics of Men, Saitama 5 Dept. of Psychiat., Institute of Clinic, Ibaraki 6 Department of Psychiatry, Showa University, Tokyo Brain-derived neurotrophic factor (BDNF) belongs to the neurotrophic factor family and promotes the development, regeneration, survival and maintenance of neuron function. It has recently been reported that a microsatellite polymorphism located approximately 1.0 kb upstream from the translation initiation site of BDNF is associated with psychiatric disorders. We found that this polymorphism is not a simple dinucleotide repeat, but is highly polymorphic, with a complex structure containing three types of dinucleotide repeats of 50 (CG)del/ 4/5, (CA)9–15 and (GA)2/3. We obtained evidence for association between this polymorphic region (BDNF-linked complex polymorphic region: BDNF-LCPR) and bipolar disorder; furthermore, a luciferase reporter gene assay in rat primary cultured neurons suggests that the risk allele is associated with lower transcription activity, compared to the other alleles. In this study, we performed an association study between the BDNF-LCPR and schizophrenia. Subjects were 390 Japanese patients with schizophrenia (DSM-IV) and 388 controls matched for age and sex. Genotyping of BDNF-LCPR was done by pyrosequencing. We found that homozygosity for the 50 (CG)del allele was more common in patients with schizophrenia than in controls (p ¼ 0.003, odds ratio 4.6, 95% CI: 1.5–13.8). In the haplotype-based analysis, the del-12–3 haplotype was significantly more common in schizophrenics than in controls. These results suggest that the BDNFLCPR is involved in the pathophysiology of schizophrenia, as well as in bipolar disorder.
P27 THE GENETICS OF ADDICTION IN ICELAND Thorlakur Jonsson1, Thorgeir Thorgeirsson1, Valgerdur Runarsdottir2, Andrew Hicks1, Natasa Desnica1, Andres Ingason1, Anna Wiste1, Frank Geller1, Augustine Kong1, Jeffrey Gulcher1, Hogni Oskarsson2, Thorarinn Tyrfingsson2, Kari Stefansson1 1 Decode Genetics, Reykjavik 2 SAA—National Center of Addiction Medic, Reykjavik 2 Therapeia, Reykjavik The genetic contribution to addiction liability is widely recognised although the identification of specific genetic risk factors and genes involved in the molecular basis of addiction remains a major challenge. Here we describe three cohorts being used to study addiction genetics in Iceland. 1) An alcohol and drug addiction cohort from the SAA— National Treatment Center, the major center for the treatment of addiction disorders in Iceland. The accumulated list of admissions from 1977 onward contains DSM diagnoses for over 18,000 individuals. 2) A nicotine addiction cohort with genotype and phenotype information for 40,000 individuals who answered questions on lifetime smoking habits (24,000 smokers). 3) A group of 1,600 individuals diagnosed with substance disorders, mostly nicotine and/or alcohol dependence, in a study of the genetics of anxiety and depression. Phenotypic characterization of participants recruited to the addiction study includes webbased questionnaires on addiction (i.e. Fagerstro¨m Test of Nicotine Dependence, and questions addressing DSM criteria for nicotine dependence) and personality (NEO-FFI), and computerized interviews (CIDI, SSAGA-II). The large collection of phenotypes and genotypes coupled with the complete Icelandic genealogy is being utilized to isolate genetic variants conferring risk of addiction using linkage and association methods. The extensive phenotypic information can also be leveraged in genotype-phenotype correlation studies to study whether at-risk variants confer risk of addiction in general, addiction to specific substances, or addiction co-morbid with other psychiatric or personality disorders.
Abstracts P28 TAMAL (v2)—A PROGRAM FOR RAPID & COMPREHENSIVE SNP SELECTION FOR LARGE GENE LISTS Patrick Sullivan UNC/Genetics, Chapel Hill TAMAL (Technology And Money Are Limiting) is a web site that allows flexible and rapid selection of SNP for genetic association studies. To do this currently requires the difficult, time-consuming and error-prone integration of data from many different sources. In TAMAL, this has already been done. To use TAMAL, the user simply inputs a list of standard gene names and selects options—the population of interest, htSNP selection method, a MAF filter and which predicted ‘‘functional’’SNPs are of interest. Functional SNPs are based on in silico preditions (ie, SNPs in predicted promoters, regulatory regions, transfactor binding sites, miRN target sites and highly conserved regions). TAMAL is available free at http://neoref.ils.unc.edu/tamal/. P29 TPH GENE POLYMORPHISMS IN BIPOLAR PATIENTS WITH AND WITHOUTCOMORBID SOCIAL PHOBIA Nader Perroud1, Patrick Baud2, Catherine Buresi2, Martin Preisig3, Bruno Etain4, Franck Bellivier4, Sophie Favre2, Nadja Reber2, Franc¸ois Ferrero2, Marion Leboyer4, Alain Malafosse2 1 University Hospital of Geneva, Cheˆne-Bougerie 2 Department of Psychiatry, University Hospital, Cheˆne-Bourg 3 University Department of Adult Psychiatry, Prilly-Lausanne 4 AP-HP, Department of Psychiatry, Hospital, Creteil Epidemiological and genetic findings suggest comorbid anxiety disorders may help to define more genetically homogeneous forms of bipolar disorder (BD). Recent family-genetic studies suggest that BD with comorbid anxiety disorder constitutes a distinct clinical entity. We have previously shown that social phobia (SP) was linked to a more severe phenotype of BD with a higher rate of suicide attempts. SP could therefore be a component of this particular genetic subtype of BD and increase the association strength between BD and genes that have already been suspected in that disorder. The serotonin system has been implicated in the etiology of BD. In this study, polymorphisms of tryptophan hydroxylase (TPH) genes 1 and 2 were compared in BD patients with or without SP. Method: Genotype and allele frequencies of TPH1 polymorphisms (rs7310929, rs211104, rs1800532 and a 50 UTR microsatellite) and TPH2 polymorphisms (rs11178997, rs11179000, rs11179001, rs1386496, rs7305115, and rs1487279) were compared in all BD subjects (N ¼ 453), BD subjects with SP (N ¼ 58), without SP (N ¼ 395) and controls (N ¼ 478) using a chi-square test. Results: TPH1 rs1800532 and 50 UTR microsatellite were significantly associated with BD. However, this association was not strengthened by comorbid SP. Haplotype constructions showed that the more common TPH1 haplotpype was significantly associated with BD. However, this association was not influenced by comorbid SP. Conclusion: Our results support previous findings suggesting association between TPH and BD. However, they did not show that SP could represent a BD genetic subtype preferentially associated with these genes. P30 ANALYSIS OF FYN POLYMORPHISMS WITH BIPOLAR DISORDER AND LITHIUM RESPONSE IN BIPOLAR PATIENTS Aleksandra Szczepankiewicz, Aleksandra Suwalska, Maria Skibinska, Monika Dmitrzak-Weglarz, Anna Leszczynska-Rodziewicz, Janusz K. Rybakowski, Joanna Hauser Poznan University of Medical Sciences, Poznan According to the neurodevelopmental hypothesis, Fyn kinase gene (Fyn) has been implicated as a candidate gene for psychiatric disorders. Within the Fyn gene, three polymorphisms (-93A/G in the 570 -flanking region, Ex12þ894T/G in the 30 UTR and IVS10þ37T/C in intron 10)have been described. The aim of this study was to find possible associations of three polymorphisms of Fyn gene with bipolar disorder and links between lithium response and Fyn genotypes in the group of bipolar patients. We analysed 452 bipolar patients and 519 subjects from the control group. Consensus diagnosis was made for each patient, according to DSM-IV and SCID. We also analysed 89 bipolar patients
745
with varying prophylactic lithium response. Genotypes of three analysed polymorphisms were established by PCR-RFLP. We found an association of T/G and T/C polymorphisms in the whole group of bipolar patients (p ¼ 0.001 and p ¼ 0.009, respectively). We also observed an association of these polymorphisms with the early age at onset of bipolar illness (p ¼ 0.0001 and p ¼ 0.009, respectively). No significant differences were observed between Fyn polymorphisms and lithium response. Haplotype analysis revealed that T/G and T/C polymorphisms are in linkage disequilibrium (r2 ¼ 0.84,D’ ¼ 0.925 with 95%CI: 0.89–0.95). The observed association of T/G and T/C polymorphisms with bipolar disorder and the early age at onset may suggest their involvement in modulating the onset age of bipolar disorder. P31 ASSOCIATION BETWEEN NOREPINEPHRINE TRANSPORTER GENE (SLC6A2) POLYMORPHISMS AND BIPOLAR DISORDER B. Etain1, Flavie Mathieu2, Ste´phane Jamain2, Frank Bellivier3, Chantal Henry4, Marion Leboyer3 1 De´partement de psychiatrie, Cre´teil 2 INSERM U513, Cre´teil 3 Hoˆpital Albert Chenevier, Cre´teil 4 Hoˆpital Charles Perrens, Bordeaux Background: Complex dysfunctions in norepinephrine neurotransmission havebeen implicated in the pathophysiology of bipolar affective disorder (BPAD). SLC6A2, which encodes the norepinephrine transporter protein, is a potential physiological and functional candidate gene, although its involvement in the genetic susceptibility of BPAD has yet to be demonstrated. Methods: We genotyped two SLC6A2 polymorphisms (T182C: rs2242446 ; G1287A: rs5569) in 399 BPAD patients and 99 healthy controls. Association was performed using logistic regressive models, in order to adjust for age, sex and age at onset, using the SAS Institute statistical package. Haplotype analysis was carried out using EH software. Results: No association was detected between BPAD and either of the two polymorphisms. Haplotype frequencies were not significantly different between BPAD patients and controls. However, significant associations were found between suicidal behaviour in BPAD patients and T182C polymorphism (Padjust ¼ 0.008), and between the presence of comorbid panic disorder in BPAD patients and the G1287A polymorphism (Padjust ¼ 0.008). Haplotype analysis suggested that bipolar patients carrying a ‘TG’ haplotype had a higher risk of suicidal behaviour (OR ¼ 1.43 [1.1–1.9], p ¼ 0.01) or comorbid panic disorder (OR ¼ 1.44 [1.0–2.1], p ¼ 0.05). Conclusions: This is the first report on independent associations between SLC6A2 gene haplotypes, suicidal behaviour and panic disorder in a large sample of BPAD patients. We suggest that the SLC6A2 gene is not associated with BPAD. However, this gene may modulate bipolar disorder phenotypic expression, especially comorbid conditions. P32 GENOME-WIDE PARAMETRIC LINKAGE ANALYSES OF MULTIPLEX PEDIGREES IN BIPOLAR DISORDER Jessica Ross, William Byerley Langley Porter Psychiatric Institute, San Francisco, CA Bipolar mood disorders are complex psychiatric disorders characterised by recurring episodes of mania and depression, with an estimated prevalence of 1% in most populations. Family, twin, and adoption studies all suggest strong genetic susceptibility for these disorders, but linkage studies have failed to reproduce findings. The NIMH Genetics Initiative for Bipolar Disorder was created in 1991 in an attempt to ascertain sufficient numbers of multiplex families and detailed clinical data to conduct powerful genetic linkage studies. Approximately 3000 individuals have been genotyped in four different ‘‘waves’’ of data collection, from a total of 644 multiplex families with at least two affected first-degree relatives. We report the results of the genome-wide linkage analysis of data from this complete dataset using parametric linkage methods, which have been shown to have greater power than non-parametric analyses in many complex genetic models. We performed analyses using 3 nested bipolar affective disorder phenotypic models with both dominant and recessive inheritance patterns. We found a genome-wide significant two point LOD score of
746
Abstracts
3.3 on chromosome 16p using the BP2 dominant model for inheritance. Multipoint analyses of this region using the same inheritance model identified a LOD score of 4.1. These findings support previouslyreported linkages at this loci in samples from independent populations. Several other LOD scores were > 1.5 and thereby suggestive of linkage. Many of these loci have also been identified in previous genome-wide linkage studies for affective disorder.
P33 ASSOCIATION OF ALLELES AT GSK3b50 WITH BIPOLAR DISORDER AGE OF ONSET Alison Fitches1, Deborah Clark2, Peter R. Joyce3, Martin A. Kennedy4, J. Elisabeth Wells5, Robin J. Olds2 1 University of Otago, Dunedin 2 Department of Pathology, Dunedin School, Dunedin 3 Deparment of Psychological Medicine, Chr, Christchurch 4 Department of Pathology, Christchurch Sc, Christchurch 5 Department of Public Health and General, Christchurch Glycogen synthase kinase 3aˆ (GSK3aˆ) has a well-defined role in the Wnt/aˆ-catenin pathway. We have examined the association of alleles at GSK3aˆ 50C/T single nucleotide polymorphism (SNP)on age of onset and markers of severity for bipolar disorder (BPD). The investigation included 258 individuals with a diagnosis of BPD. Age of onset was defined as the first reliably diagnosed hypo-/manic or depressive episode. The mean age of onset was 21.5 8.8 yr for BPD-I and 19.0 7.6 yr for BPD-II. Suicide attempts were reported by 43.8% of BPD-I and 41.3% of BPD-II subjects. The age of 1st hypo-/manic episode was 20.8 6.6 yr, 26.3 10.5 and 25.0 10.1 for C/C, C/T and T/T genotypes, respectively (F ¼ 6.52, p ¼ 0.002, Welch’s ANOVA). Age of first depressive episode was 18.6 4.5, 22.3 8.4 and 21.7 10.4 yr for C/C, C/T and T/T genotypes, respectively (F ¼ 5.56, p ¼ 0.005). The C/C genotype was associated with an earlier age of first suicide attempt for those with BPD-I (p ¼ 0.02). The data suggest that the age of onset is influenced by the GSK3aˆ 50 SNP. The average age difference between C/T and T/T genotype groups is insignificant, consistent with the T allele exerting a dominant effect.
P34 IDENTIFYING POTENTIAL CANDIDATE GENES IN AN IRISH BIPOLAR DISORDER SAMPLE LINKED TO 14q21-32 Fiona Cassidy1, Chengfeng Zhao2, Jonathon Badger2, Catherine Delaney3, Lisa Mooney3, Seth Dobrin2, Siobhan Roche1, Patrick McKeon3 1 Smurfit Institute of Genetics, Dublin 2 Marshfield Clinic Research Foundation, Wisconsin 3 St. Patrick’s Hospital, Dublin A 10 cM whole genome scan was previously performed in a collection of 60 Irish bipolar affective disorder (BPAD) affected sib pairs (ASPs). The most significant result was on chromosome 14 at 76 cM (68.21 MB, peak marker D14S588, multipoint NPL ¼ 3.28, narrow BPAD type I disease model). Since the region of the chromosome with P values < 0.05 was quite substantial (47 cM to 122 cM), we undertook a fine-mapping analysis of this region. 100 SNP markers (0.6 cM resolution) were analysed in an extended sample of 88 ASPs. Linkage analysis resolved our original linkage peak into 3 separate peaks. The most significant NPL was at rs1885590 (66.84 MB, multipoint NPL ¼ 3.21, narrow disease model), which is located within intron 2 of the MPP5 gene. An additional 80 SNPs are currently being genotyped to further resolve this region to 0.2 cM. Two SNPs were associated with BPAD (p < / ¼ 0.05: rs2885128 at 75.85 MB, and rs2148564 at 93.14 MB). The P values for an additional two SNPs approached significance (p < / ¼ 0.06: rs229646 at 64.26 MB and rs2010649 at 94.87 MB). Data from a separate whole genome expression array study on Stanley Foundation post-mortem brain samples is also being used to identify potential candidate genes within the region of linkage. Here we present the complete results from our linkage, association, and expression analyses, detailing the novel candidate genes under investigation. P35 INVESTIGATION OF SUSCEPTIBILITY LOCI FOR BIPOLAR AFFECTIVE DISORDER ON CHROMOSOME 21 Siobhan Roche1, Fiona Cassidy2, Chengfeng Zhao3, Badger Jonathon3, Lisa Mooney4, Catherine Delaney4, Seth Dobrin3, Patrick McKeon4 1 Smurfit Institute of Genetics, Dublin
2
Trinity College, Dublin Center for Human Genetics, Marshfield Cl, Wisconsin St. Patrick’s Hospital, Dublin
3 4
A genome-wide linkage scan in 60 bipolar affective disorder (BPAD) affected sib-pairs identified a potential susceptibility locus on chromosome 21 at 21q21 (NPL ¼ 1.9, P ¼ 0.029, BPAD type I model). We will present the fine-mapping results of this region. A second region of linkage was obtained at 21q22 (D21S1446), a BPAD susceptibility locus supported by multiple studies (NPL ¼ 1.42, P ¼ 0.08, narrow model). Interestingly, although our findings at 21q22 only approach significance, the peak marker is located 12Kb downstream of S100B, a calcium-binding protein that exhibits neurotrophic and neurodegenerative effects. Serum S100B protein levels are elevated in schizophrenic and bipolar patients and a two marker haplotype is associated with schizophrenia. We investigated whether variants within S100B are also associated with BPAD in a collection of 125 BPAD type I trios. SNPs rs283935 (P ¼ 0.02) and rs378826 (P ¼ 0.03) were associated with BPAD. The P values for two additional SNPs approached significance. The results of the single- and multi-marker association analyses will be presented. Fine-mapping analyses of region 21q22 have reported TRPM2, which is 2Mb upstream of S100B, as a possible BPAD susceptibility gene. The results of association tests of TRPM2 will also be presented. In summary, region 21q22 is a highly-replicated BPAD susceptibility locus. We will present evidence to support S100B as a candidate susceptibility gene within this region and will investigate an additional locus at 21q21. P36 AN ALLELE SHARING METHOD FOR FINE MAPPING DISEASE LINKED LOCI: APPLIED TO BIPOLAR AFFECTIVE DISORDER Andrew Lee University of Edinburgh, Edinburgh A founder mutation contributing to some disease will occur against a distinct haplotype background. Over time, the haplotype common to individuals carrying the mutation will be broken down due to recombination. Individuals carrying the mutation from disparate families will be expected to show some degree of sharing of the mutation depending on the number of generations since a common ancestor. In such families, linkage studies are an effective means of identifying a wide region associated with the resultant illness. However, there is often no obvious candidate gene, so in order to minimise the search region I have developed a method using a dense genotype map across a common linkage region to identify any ancestral haplotype shared as regards the disease-linked mutation. I have developed a simulation model under which we have successfully tested the method, and I have gone on to use this model to develop a set of criteria according to which the method is expected to be most effective. I have also successfully tested the method on a number of families known to carry a founder mutation responsible for cystic fibrosis. Finally, we have applied the method to four families that show linkage to a region connected with bipolar affective disorder. In doing so, we have narrowed down the region of interest from an 8Mb linkage region to a 200kb region of allele sharing. P37 BDNF Val66Met POLYMORPHISM AND HIPPOCAMPAL VOLUME IN ELDERLY BIPOLAR PATIENTS Stevin Zung, Quirino Cordeiro, Fa´bio Duran, Geraldo Busatto, Homero Vallada University of Sao Paulo Medical School, Sao Paulo, Sao Paulo Introduction: The Val66Met polymorphism of the BDNF gene is associated with susceptibility to bipolar disorder and also with variations in hippocampal volume in healthy subjects. The present study investigated whether this polymorphism was associated with hippocampal volume in elderly bipolar patients. Methods: Magnetic resonance imaging data were obtained in 49 bipolar patients aged 60 years or older. Patients were divided into two groups: Val homozygotes (n ¼ 28) and Met homozygotes or heterozygotes (n ¼ 21). The volume of the hippocampus was measured and compared between the groups using voxel-based morphometry (VBM), based on the Statistical Parametric Mapping (SPM) technique. The voxels mapped to hippocampus were circumscribed using the ‘‘small volume correction’’ tool available in the SPM, and differences were
Abstracts reported as significant if surviving family-wise error (FWE) correction for multiple comparisons (p < 0.05). Results: There was no difference in hippocampal volume between elderly bipolar patients with Val/Val genotype and those with Met alleles. Conclusion: The volume of the hippocampus was not associated with the BDNF Val66Met polymorphism in our sample of bipolar patients. Studies with larger samples are required to investigate the possible association between BDNF Val66Met polymorphism and hippocampal volume. P38 EFFECT OF MITOCHONDRIAL DNA 8701/10398 POLYMORPHISMS ON MITOCHONDRIAL MATRIX pH An-a Kazuno1, Kae Munakata1, Satoshi Shimozono1, Nobumasa Kato2, Atsushi Miyawaki1, Tadafumi Kato1 1 RIKEN, Brain Science Institute, Wako, Saitama, 2 University of Tokyo, Bunkyo, Tokyo Recent studies suggest mitochondrial abnormality in bipolar disorder. We previously reported the altered energy metabolism in the brains of patients with bipolar disorder using phosphorus-31 magnetic resonance spectroscopy. The comorbidity of affective disorders in patients with mitochondrial encephalopathy has been reported. Moreover, it has always been stated that mitochondrial DNA (mtDNA) 10398 polymorphism is associated with Parkinson’s disease, Alzheimer’s disease, and bipolar disorder. It is generally accepted that bipolar disorder is accompanied by altered intracellular calcium signalling. The driving force of mitochondrial calcium uptake is the proton gradient in the mitochondrial inner membrane. Based on this evidence, we hypothesized that the mitochondrial dysfunction in bipolar disorder might be partly due to altered mitochondrial matrix H linked with the mtDNA polymorphism associated with bipolar disorder. To examine this hypothesis, we generated a pH indicator for measurements of mitochondrial matrix pH. We measured basal mitochondrial matrix pH in transmitochondrial hybrid cell lines (cybrids) with mtDNA polymorphisms associated with bipolar disorder. We found that the basal mitochondrial matrix pH is significantly lower in cybrids with 8701A/ 10398A mtDNA associated with bipolar disorder than in those with 8701A/10398G.
P39 LARGE-SCALE ASSOCIATION ANALYSIS OF THE CHROMOSOME 4p15–16 CANDIDATE REGION IN BIPOLAR AFFECTIVE DISORDER AND SCHIZOPHRENIA Andrea Christoforou1, Stephanie Le Hellard1, Philippa A. Thomson1, Stewart W. Morris1, Albert Tenesa2, Benjamin S. Pickard1, Naomi R. Wray1, Walter J. Muir1, Douglas H. Blackwood1, David J. Porteous1, Kathryn L. Evans1 1 University of Edinburgh, Edinburgh 2 MRC Human Genetics Unit, Edinburgh Several independent linkage studies have identified chromosome 4p1516 as a putative region of susceptibility for bipolar disorder (BP) and schizophrenia (SCZ). We initially reported a maximum multipoint LOD score of 4.8 in a large Scottish pedigree multiply affected with BP. Haplotype analysis of this family and three other chromosome 4p linked families has revealed two priority linkage regions (B and D), each shared by three of the four families. We performed an association analysis of Regions B and D. SNP genotype data for the CEU trios was downloaded from the International HapMap Project. Linkage disequilibrium maps of the two regions were constructed and haplotype tagging (ht) SNPs selected. A total of 408 htSNPs were then tested for association in 368 BP, 386 SCZ and 458 control individuals. The significance threshold was determined using principal component analysis as implemented in the program SNPSpD. In Region B, one individual SNP and 21 haplotypes were associated at the experimentwide level (p < 0.0002), while two individual SNPs and 15 haplotypes met the region-wide threshold (p < 0.0005). The SNPs and haplotypes overlapped and fell into three distinct clusters. In Region D, 25 haplotypes were associated at the experiment-wide level, while 10 haplotypes met the region-wide threshold (p < 0.0003). No individual markers were significant. These overlapping haplotypes fell into eight clusters. Our study has identified significant regions within the chromosome 4p linkage region, including candidate genes, for further investigation. It warrants replication in an independent sample.
747
P40 ANALYSIS OF SEROTONIN TRANSPORTER VNTR HAPLOTYPES IN BIPOLAR AFFECTIVE DISORDER Sara Campos de Sousa1, Gerome Breen2, Sarah Osborne1, Janet Munro1, Robert Kerwin1, David St Claire2, David Collier1 1 King’s College London, London 2 King’s College, London 2 University of Aberdeen, Aberdeen Bipolar disorder (BPD) is a frequent, severe psychiatric disorder with lifetime prevalence between 1 and 2% in the general population, being equally distributed between sexes. The serotonin transporter5-HTT or SERT is a strong candidate gene for BPD, as serotonin seems to play a key role in the regulation of mood and the treatment of affective symptoms. Consequently, two functional polymorphisms in the gene, a VNTR in the second intron with 3 alleles (Stin2.9, Stin2.10 and Stin2.12) and a 44 base pair insertion/deletion within the promoter region, the 5-HTTLPR with two alleles (long variant, L and short variant, S), have been studied. Both VNTRs are functional, with the S allele of the 5-HTTLPR having lower transcriptional activity (Lesch et al., 1996; Collier et al., 1996) and the intron 2 having differential enhancer activity (Fiskerstrand et al., 1999). A third SNP polymorphism in the gene, close to the 5-HTTLPR, may also confer a functional effect. A recent meta-analysis (Cho et al., 2006) suggested that both of these VNTR polymorphisms have a modest effect on risk of developing BPD. Surprisingly, despite the evidence in favour of these polymorphisms as risk factors for BPD, the haplotype of the two VNTRs has rarely been examined. In the present study, we genotyped 500 bipolar patients and 500 controls from the UK for these two polymorphisms, and analysed it. This allowed us to determine whether the polymorphisms have an independent or haplotypic effect on risk.
P41 SEARCHING SUSCEPTIBILITY LOCI FOR BIPOLAR DISORDERS: A SIB PAIRS STUDY ON CHROMOSOME 12 Cristina Lorenzi1, Adele Pirovano2, Danilo Dotoli2, Fanny Bongiorno2, Elena Marino2, Marco Catalano2, Enrico Smeraldi2 1 University Vita-Salute—S. Raffaele, Milan 2 University Vita-Salute—S.Raffaele, Milan Objective: Many studies have demonstrated the role of genetic factors in determining susceptibility to mood disorders, particularly to bipolar ones. Nevertheless, few candidate genes have been clearly identified; for this reason, detection of the genetic bases for mood disorders remains an important challenge for understanding the outcome and treatment of these diseases. Past linkage studies focused particularly on bipolar disorders, since genetic factors have a far greater etiologic role in manic-depressive disorder than in major depression. Several chromosomal regions, in linkage with bipolar disorders, were identified, among them, some broad areas of chromosome 12. Methods: We performed a genome-wide scan (cM 10) on chromosome 12, in a sample of 160 Italian sib pairs of subjects affected by mood disorders and their healthy sibs. We performed a non-parametric linkage analysis using the QTL Express programme. Results: We detected significant linkage between manic phases and the marker D12S85 at 62 cM from p-ter (Approximate LOD: 4.594) and a suggested linkage between clinical depression and the region between the markers D12S310 and D12S1617, at 38 cM (Approximate LOD: 3.744). Conclusions: Given the complexity of bipolar disorders, we could assume the existence of different regions in the human genome, presumably related to susceptibility for mania or for depressive episodes. Specific genetic factors for manic facets, acting in synergy with other ones controlling depressive manifestations, could result in bipolar disorders. P42 FINE MAPPING OF A NEW BIPOLAR AND UNIPOLAR AFFECTIVE DISORDER LOCUS ON CHROMOSOME 1p36 Radhika Kandaswamy1, Nicholas Bass1, Khalid Choudhury1, Vinay Puri1, Andrew Mcquillin1, Jacob Lawrence1, David Curtis2, Hugh Gurling1 1 University College London, London 2 St Bartholomew’s and Royal London School, London Linkage studies of bipolar disorder have found several positive linkages on chromosome 1p36 near the telomere. In our bipolar genome scan
748
Abstracts
(Curtis et al., 2003), we found a lod score above 3.00 in a single family and an overall lod score of 2.7. One other linkage study found a lod of 3.6 for unipolar affective disorder at 1p36 near the telomere (Zubenko et al., 2003). We have now fine mapped a potential candidate gene near the telomere of 1p by finding positive allelic association between a microsatellite marker and bipolar disorder. We employed a case-control sample of 600 bipolar cases of British ancestry and 450 supernormal controls. Four microsatellite markers spanning a region of 500kb were genotyped. Significant allelic association with bipolar disorder was found for the marker D1S243 using CLUMP analysis. Further fine mapping analysis of the region using multiple SNPs around D1S243 is currently being carried out. Protein kinase C, zeta type (PRKCZ) is very close to D1S243 and it is very highly expressed in parts of the brain thought to be involved in bipolar disorder. These findings have led us to sequence the candidate gene PRKCZ. Curtis, D. et al. (2003) Psychiatr Genet 13, 77–84. Zubenko, et al. (2003) Am J Med Genet B Neuropsychiatr Genet 123, 1–18.
P43 VOTE-COUNTING META-ANALYSIS OF SUSCEPTIBILITY LOCI FOR BIPOLAR DISORDER ON CHROMOSOME 3 Sara Szente Center for Medical Education, Medical University, Vienna Studies converge on evidence for substantial genetic contribution to the susceptibility for bipolar disorder (BP). Combined and meta-analyses identified several chromosomal regions that might harbour susceptibility loci. Several molecular genetic studies have resulted in support for possible linkage with markers on chromosome 3. Findings supportive for 3q29 were obtained from our Austrian BP sample in a genome-wide scan and a follow-up linkage analysis (Bailer et al. 2002; Schosser et al. 2004). The present study was a systematic quantitative review of genome-wide scans on chromosome 3. Genome-wide scans in BP-affected families published up to May 2006 were retrieved via electronic literature search. Sample overlap across studies was accounted for. In our analysis, LOD (log of the odds) and NPL (nonparametric LOD) scores >1 for chromosome 3 were used by the meta-analytical vote-counting approach. Twenty two out of 32 genomewide scans with LOD or NPL scores >1 for chromosome 3 were identified. Ten studies yielded linkage scores >1 for 3p, eight for 3q, and four for both 3p and 3q. Several studies presented more than one peak. This analysis suggests that chromosome 3 deserves further investigation in BP. It is worthy of note that the dopamine 3 receptor gene (DRD3) is located on chromosome 3q13.3 and our group found suggestive genome-wide evidence for linkage with BP on chromosome 3q29 (p ¼ 0.0005). However, linkage signals from different studies do not overlap, although some are located on or close to 3q29.
P44 SEX-SPECIFIC ASSOCIATION BETWEEN BIPOLAR DISORDER IN WOMEN AND GPR50 Pippa Thomson1, Hazel Kinnell1, Maria Paola Piccardi2, Maria Del Zompo2, Douglas Blackwood1, David Porteous1 1 University of Edinburgh, Edinburgh 2 University of Cagliari, Cagliari Studies of mood disorder and obesity have suggested that, while the majority of obese people do not have mood disorders, depression with atypical symptoms in females is significantly associated with being overweight (McElroy et al., 2004). Previously we reported association between females with bipolar disorder and a deletion polymorphism in GPR50, an X-linked orphan G protein–coupled receptor (p ¼ 0.0002, Thomson et al., 2005). GPR50 is expression is restricted to the pituitary and hypothalamus, areas increasingly implicated in mood disorders. The deletion results in the loss of four amino acids from the C-terminal tail, a region known to be important in protein interactions and GPCR signalling. In addition, GPR50 has recently been implicated in studies of obesity, with polymorphisms being linked to increased serum fasting triglyceride levels and decreased HDL-cholesterol levels, supporting a role for GPR50 in lipid metabolism (Bhattacharyya et al., 2006). We are continuing to study this gene, its genetics, pattern of expression and protein interactors. To date we have performed case-control association study within a second independent bipolar disorder population from Sardinia and yeast two hybrid screens to identify proteins interacting with the C-terminal tail.
P45 ASSOCIATION ANALYSIS BETWEEN BIPOLAR DISORDER AND VARIANTS AT THE G72/G30 LOCUS IN A RUSSIAN SAMPLE A. Karpushova1, A. Georgi2, F. Schirmbeck2, A. Tiganov3, G. Panteleyeva4, L. Abramova3, V. Artyuch3, M. Chananashvili3, V. Krasnov4, S. Kapiletti4, S. Mosolov4, A. Chuchalin5, G. Babadjanova4, J. Schumacher5, R. Abou Jamra5, M. No¨then1, S. Cichon1, M. Rietschel2 1 Dpt. of Genomics, Life & Brain, University, Bonn 2 Central Institute of Mental Health, Mannheim 3 Mental Health Res. Ctr., Moscow 4 Russian State Medical University, Moscow 4 Moscow Research Institute of Psychiatry, Moscow 5 Russian State Medical University, Moscow 5 Inst. of Human Genetics, University of Bonn Since the initial report by Chumakov et al. several studies have replicated the association between G72/G30 and schizophrenia (SZ). G72/G30 variants have furthermore consistently been found associated with bipolar disorder (BD). We recently reported association between identical G72/G30-haplotypes and SZ in a German as well as BD in both a German and a Polish sample. The markers M22, M23, and M24 showed the strongest associations. The study by Chumakov et al. found (mainly genotypewise) associations with M23 and M24 in a moderately-sized sample from Russia. In this population, a study on G72/G30 and BD has not yet been performed. We studied 253 Russian BD (DSM-IV) patients and 246 controls. We genotyped markers M22, M23, and M24. We performed allelewise, genotypewise, and haplotypebased tests. Given that in our previous studies, the association with BD was mainly driven by cases with a life-time history of persecutory delusions, we also performed an analysis using the subsample of 35 cases with persecutory delusions. There were no differential distributions of alleles, genotypes, and haplotypes between cases and controls, either in the overall or the sample stratified for a history of persecutory delusions. This is the first case-control analysis studying a potential involvement of G72/G30 in the etiology of BD in the Russian population. Our results do not suggest such involvement. Larger samples, allowing for powerful genotype-phenotype correlation analyses, may be needed to study the role of this gene in the Russian population.
P46 IS THE 5-HTTPRL POLYMORPHISM ASSOCIATED WITH BIPOLAR DISORDER OR WITH SUICIDAL BEHAVIOUR OF BIPOLAR PATIENTS? Humberto Correa, Fernando Neves, Gustavo Silveira, Luiz Armando De Marco, Marco Aure´lio Romano-Silva UFMG, Belo Horizonte-Minas Gerais Background: The serotonin transporter gene (5-HTT) is a candidate for the pathogenesis of bipolar disorder. This gene has a 44bp insertion/ deletion polymorphism within the promoter region (5-HTTLPR) with two allelic forms, the long (L) and the short (S) variants. An association between the low-activity S variant and bipolar disorder has been shown but the replication has not been consistent. On the other hand it has been more consistently shown that S allele is associated with suicidal behaviour. Since suicidal behaviour is a rather frequent event in bipolar disorder, an important question is whether suicidality, instead of bipolarity itself, could be related to S-allele. Methods: We assessed 147 bipolar inpatients and 182 controls. Diagnosis was performed by a psychiatrist using a structured interview (MINI-PLUS), according to DSM-IV. Suicidal behaviour was assessed using a semi-structured instrument and a review of medical records. Genotyping for the 5-HTTLPR was performed using PCR. Results: Bipolar patients and health controls showed comparable genotypic and allelic frequencies. Patients carrying a S-allele made violent suicide attempts more frequently (w2 18.4; p ¼ 0.001) and made more suicide attempts over time (F ¼ 3.6; p ¼ 0.025). Conclusion: We were able to show an association between the S-allele and suicidal behaviour but not with bipolar disorder. Suicidality in bipolar patients can help us to understand why the association studies between the 5-HTTPRL and bipolardisorder are so controversial.
Abstracts P47 LINKAGE ANALYSIS ON A HIGHLY INBRED MULTI-GENERATIONAL PEDIGREE SEGREGATING FOR MAJOR PSYCHIATRIC AND MENDELIAN NEUROLOGICAL DISORDERS Ivan Nikolov1, Stuart Macgregor2, Peter Holmans1, George Kirov1 1 University Hospital of Wales, Cardiff 2 Queensland Institute of Medical Research, Queensland We have collected DNA and genealogy information from 96 individuals from a multi-generational, highly inbred pedigree of Gypsy settlers in a rural region of Bulgaria, totalling more than 300 individuals. They come originally from Turkey and their main occupation has been livestock dealing. The genealogy could be traced back up to six generations and the pedigree has been developing with a reasonable degree of isolation during the past 60 years. Rare Mendelian traits (such as spastic paraplegia) have been found to segregate in this family as well as major psychiatric disorders (Schizophrenia, Affective Disorders). We designed a genome-wide linkage study using Illumina Linkage IV SNP genotyping platform. Linkage scan using SNPs rather than microsatellites has been shown to provide greater mapping resolution. Merlin and Simwalk2 software have been used for linkage analysis. High number of complex inbreeding loops is found on every genealogical level which poses difficulties when designing the parametric linkage analysis. So far, analyses have been done on a nuclear family segregating for spastic paraplegia. Suggestive linkage was found to loci on Chrs. 2, 12 and 15. Linkage analyses on schizophrenia spectrum and bipolar disorders have also been run. P48 ASSOCIATION OF BIPOLAR DISORDER TO CHROMOSOME 3q29: PRELIMINARY RESULTS Alexandra Schosser1, Karoline Fuchs2, Theresa Scharl3, Friedrich Leisch4, Jochen Kindler5, Nilufar Mossaheb5, Ursula Bailer5, Siegfried Kasper5, Werner Sieghart2, Kurt Hornik3, Harald N. Aschauer5 1 University Hospital for Psychiatry, Vienna 2 Center for Brain Research, Vienna 3 Department of Statistics and Probability, Vienna 4 Department of Statistics, University Lud, Munich 5 Department of General Psychiatry, University, Vienna 3 Department of Statistics and Mathematics, Vienna We recently published a genome scan and a follow-up linkage analysis within the chromosome 3q29 region in bipolar affective disorder families, resulting in a NPL score Zall of 4.09 (p ¼ 0.0005) in the first, and a NPL score of 3.56 (p ¼ 0.001) in the latter case. Subsequently, when genotyping 11 additional evenly spaced SNP markers, we again found support for linkage (p ¼ 0.00055). In the current study, we conducted an association analysis within an independent sample (casecontrol design). Our sample consisted of 71 bipolar disorder patients (DSM-IV) and 267 psychiatric healthy controls. Genotyping of SNPs rs225, rs3747672 and rs522174 was performed using PCR-RFLP. Fisher´s exact test was employed to evaluate possible differences between the groups in the frequencies of alleles and genotypes; the standard X2 test was used for Hardy-Weinberg equilibrium. We found a difference in genotype distribution (wt/wt, wt/mut, mut/mut) for SNP rs225 between patients (19.72%, 40.84%, 39.44%) and controls (15.36%, 55.80%, 28.84%; p ¼ 0.074). No such difference was seen for SNPs rs522174 and rs3747672. As for SNP rs225, there was a significant difference in the proportion of homozygous patients and controls (p ¼ 0.032). No such difference was seen for SNPs rs522174 and rs3747672. The current results regarding the association of SNP rs225 with bipolar disorder within an independent sample add further support to our previous linkage findings of bipolar disorder to chromosome 3q29. P49 FAILURE TO CONFIRM ALLELIC ASSOCIATION BETWEEN COMT AND BIPOLAR DISORDER IN A LARGE UK SAMPLE Nick Bass1, Andrew McQuillin1, Jacob Lawrence1, Khalid Choudhury1, Vinay Puri1, Gursharan Kalsi1, David Curtis2, Hugh Gurling1 1 University College London, London 2 St Bartholomew’s and Royal London School, London Robust evidence for a bipolar susceptibility locus on chromosome 22q was found in the first meta-analysis of whole genome linkage scans in
749
bipolar disorder and schizophrenia, (Badner and Gershon, 2002). The Catechol O-methyltransferase (COMT) gene has been considered a strong functional candidate due to reports of low enzyme activity in patients with both depression and bipolar disorder. The COMT gene is located on 22q11. The Val58Met polymorphism has been shown to influence the enzyme’s activity. However, results of association studies between COMT and bipolar disorder have been contradictory. Here we report the findings from a large UK case-control sample (600 cases, 450 controls). No association was found between bipolar disorder and the Val58Met COMT polymorphism. A tetranucleotide repeat, identified from the sequence database, also failed to show association with bipolar disorder. In addition we will present data for rs 2075507, a SNP in the COMT promoter region.
P50 HETEROZYGOUS BETA-THALASSAEMIA IN BIPOLAR PROBANDS AND AFFECTED SIBS Alberto Bocchetta, Stefania Serio Dept Neurosciences University of Cagliari, Cagliari A potential association between heterozygous beta-thalassaemia and bipolar disorder has been suggested by case reports and by surveys in psychiatric populations (1 –4). The objective of this study was to clarify the nature of previous observations by studying affected sib pairs. Bipolar probands with reduced mean corpuscular volume and increased haemoglobin A2 concentration (i.e. beta-thalassaemic heterozygotes) were selected from outpatients followed at the Department of Neurosciences, University of Cagliari. Haematological phenotype (blood count, haemoglobin electrophoresis and/or chromatography) was studied in sibs suffering from major affective disorders, as diagnosed according to Research Diagnostic Criteria. Phenotype concordance was found in 26 of 37 affected sibs (70%; 95% Confidence Interval, 53%-84%). As the mendelian expectance of identity by descent in sibs is 50%, these preliminary phenotype data are consistent with the hypothesis that heterozygous beta-thalassaemia may act as a susceptibility factor in bipolar spectrum disorders in specific populations or that there is linkage disequilibrium between the beta-globin gene and a gene for affective disorders. 1. Joffe RT et al. Bipolar affective disorder and thalassemia minor. Am J Psychiatry 1986;143:933. 2. Harada H et al. Bipolar affective disorder associated with beta- thalassaemia minor. Biol Psychiatry 1995;37:477 –80. 3. Bocchetta A, Del Zompo M. Bipolar affective disorder and heterozygous beta-thalassaemia. Am J Psychiatry 1990;147:1094. 4. Bocchetta A. Heterozygous beta-thalassaemia as a susceptibility factor in mood disorders: excessive prevalence in bipolar patients. Clin Pract Epidemol Ment Health 2005;1:6.
P51 FINE MAPPING ANALYSIS ADDS SUPPORT FOR SEVERAL BIPOLAR DISORDER SUSCEPTIBILITY LOCI IN ROMA FAMILIES Radka Kaneva1, Dora Angelicheva2, Vihra Milanova3, Joachim Hallmayer4, Stuart Macgregor5, Spiridon Alexiev3, Mina Ivanova4, Rositza Vladimirova3, Vessela Stoyanova3, Tinka Milenska4, Assen Jablensky5, Luba Kalaydjieva2 1 Molecular Medicine Centre, Sofia 2 Western Australian Institute for Medical Research, Perth 3 Department of Psychiatry, Medical University, Sofia 4 Department of Psychiatry and Behavioral Sciences, Palo Alto 5 Genetic Epidemiology, Queensland Institute of Medical Research, Brisbane 4 Laboratory of Molecular Pathology, Sofia 4 Regional Mental Health Services, Kjustendil 5 School of Psychiatry and Clinical Neuroscience, Perth Population genetic studies in the population of Rome have characterised the area as a young genetic isolate, with limited genetic diversity, strong founder effect for many mendelian disorders and a complex sub- isolate structure. As suggested by previous studies of isolated populations, the latter could also afford several advantages in the search of genes contributing to complex genetic disorders. Recent linkage analysis of genome-scan and fine mapping data provided best evidence for linkage between bipolar affective disorder (BPAD) and loci on chromosomes 1p36, 2p14-p16, 4q31, and 6q24 in three large families of Roman descent. All additional loci showing suggestive evidence of linkage in the genome scan (8p21, 9p23, 11p13, 13q31, 17q23, and
750
Abstracts
18q12) were subjected to a systematic fine mapping study. To deal with the problem of statistical analysis of the extended pedigrees, including new affected members and branches, both Genehunter and Sim Walk2 were used and the results compared. Higher-density mapping provides additional support for BPAD susceptibility loci on 4q31 and 6q24. Suggestive evidence for linkage is confirmed for 11p13, 13q31, and allows further narrowing down of the linkage regions. Genetic heterogeneity, both intra and inter-familial is observed even within this small set of families. P52 GENOME SCAN FOR SCHIZOPHRENIA/BIPOLAR DISORDER SUPPORTS A SUSCEPTIBILITY LOCUS IN 15q26 Giovanni Vazza1, Cinzia Bertolin2, Eva Scudellaro2, Andrea Vettori2, Francesca Boaretto2, Sabina Rampinelli3, Gabriella De Sanctis3, Giulia Perini3, Pio Peruzzi4, Maria Luisa Mostacciuolo2 1 University of Padova, Padova 2 Dept. of Biology, University of Padova, Padova 3 Dept. of Neurosciences, University of Pa, Chioggia 4 Dept. of Mental Health, ULSS 14 Chioggia, Chioggia Schizophrenia (SZ) and bipolar disorder (BPD) are two psychiatric diseases with a strong genetic component. According to the ‘‘continuum theory’’ suggesting an overlap between these disorders, we performed a 7.5cM genome scan in a sample of 16 multigenerational families with SZ/BPD (62 affected subjects), all originating from Chioggia: a little island in the Venetian lagoon in the north-east of Italy. We identified linkage peaks on four regions (1p, 1q, 4p, 15q) that were subjected to a follow up study with an increased marker density. The strongest linkage was obtained on chromosome 15q26 with a non parametric LOD score (NPL) of 3.05 (nominal P ¼ 0.00197) for marker D15S1014. In the same locus, positive values were observed by parametric analysis with a maximum recessive HLOD of 2.43 (alpha 0.91) and a dominant HLOD of 2.12 (alpha 0.82). Genome-wide simulations were performed to asses thresholds for suggestive (NPL>2.2) and significant (NPL>3.15) linkage. Results suggest that the linkage peak on chromosome 15q meets the statistical criteria for suggestive linkage, corresponding to a genomic P value of 0.07. Interestingly, evidence for linkage in 15q26 has been previously reported by Maziade (2005) who performed a genome scan on SZ/BPD families from Quebec, using a similar ‘common locus’ model. In conclusion, our results represent the first independent replication of a locus in 15q26 and support the hypothesis that gene/s in this region may influence susceptibility to both SZ and BPD. P53 A GENOME-WIDE SCAN FOR BIPOLAR DISORDER IN A LARGE AND COMPLEX CUBAN PEDIGREE CONFIRM PREVIOUSLY REPORTED LINKAGE FINDINGS AND REVEALS A NEW SUSCEPTIBILITY LOCUS AT 5q21.3-22.3 Beatriz Marcheco-Teruel1, T.J. Flint2, F.P. Wikman3, M. Torralbas4, L. Gonza´lez4, L. Blanco4, Q. Tan5, H. Ewald2, T. Orntoft3, T.A. Kruse5, A.D. Børglum6, O. Mors2 1 National Center of Medical Genetics, Havana 2 Centre for Basic Psychiatric Research, A, Aarhus 3 Department of Clinical Biochemistry, Ske, Aarhus 4 University Psychiatric Hospital, Holguin 5 Department of Clinical Biochemistry and, Odense 6 Institute of Human Genetics, University, Aarhus A genome-wide scan of a large and complex pedigree with an apparently dominant bipolar I disorder from the Velasco region in the Holguin province, in the eastern part of Cuba was performed using the early access version of the Genechip Mapping 10K Xba array from AFFYMETRIX. Parametric and non-parametric linkage analyses under dominant and recessive models were performed using genotypes from 28 affected members. Two phenotypic models were included in the analyses: bipolar I disorder and recurrent depressive disorder or bipolar I disorder only. A suggestive parametric LOD score was obtained under the dominant model and the broader phenotype at chromosome 14q11.2–12 (LOD ¼ 2.05). In the same region a nonparametric LOD score (NPL) close to genome-wide significance was also reached (NPL ¼ 7.31, p-value ¼ 0.07). Chromosome 5q21.3–22.3 also showed close to genome-wide significant linkage (NPL ¼ 7.26, p ¼ 0.07) In addition, 2p25.1–25.3, 2p13.3–14, 3p14.2, 6p22.3–24.1,
7p14.1–14.2, 8q12.2–12.3, 10q21.1–21.2, 14q13.1–21.1, 15q15.1– 21.2 and 22q12.3–13.32 showed suggestive linkage. The majority of these potential susceptibility loci overlaps with or is close to previous linkage findings. The locus on 5q may represent a novel susceptibility locus. P54 COMT AND MANIC SYMPTOMS IN SCHIZOPHRENIA Pamela DeRosse1, Birgit Funke2, Katherine Burdick1, Todd Lencz1, Terry Goldberg1, John Kane1, Raju Kucherlapati2, Anil Malhotra1 1 Zucker Hillside Hospital, Glen Oaks 2 Harvard Partners Center for Genetics and, Boston, MA Catechol-O-methyltransferase (COMT) is an enzyme involved in the degradation of the neurotransmitter dopamine (DA). Due to the long held belief that dysfunction within the DA system contributes to the pathogenesis of many psychiatric disorders, the functionality of COMT genetic variation has led to extensive investigation of this gene in psychiatric populations. To date, the strongest association of COMT genetic variation has been to schizophrenia but these findings have been inconsistent. We suggest that inconsistencies in the relation between COMT genetic variation and schizophrenia may be clarified by a careful delineation of the target phenotype. The present study reports a significant association between a COMT haplotype and the severity of manic symptoms in 162 patients with schizophrenia or schizoaffective disorder (SZ). These data suggest that the effect of COMT variation may be associated with comorbid manic symptoms in SZ. P55 FAMILIALITY OF SUBPHENOTYPES IN THE NATIONAL INSTITUTES OF MENTAL HEALTH BIPOLAR SAMPLE Laura Scott, Erika Saunders, Melvin McInnis, Margit Burmeister University of Michigan, Ann Arbor, MI Bipolar-related subphenotypes that cluster within families may help identify subsets of patients that are more genetically homogeneous. Environmental or assessment factors that segregate by family may influence estimates of familiality. We aimed to determine familiality of subphenotypes of bipolar disorder (BP), accounting for effects of age, sex, diagnosis and site/wave of ascertainment. We studied 589 sibships with 1416 siblings affected with bipolar I (BPI), schizoaffective disorder, bipolar type (SAB), bipolar II (BPII) or recurrent unipolar depression (RUDD). Sibships were from families with > 2 BPI cases collected by the NIMH Bipolar Genetics Initiative (NIMHBGI). Rapid cycling showed the strongest evidence for familiality (OR (95%C.I.) ¼ 2.02 (1.43, 2.85), p ¼ 6 10–5) in a model including age, sex, diagnosis, and site/wave of ascertainment. Additional significantly familial traits were co-morbid alcohol abuse/dependence (p ¼ 2 10–4) and co-morbid panic disorder (p ¼ 8 10–3), as well as psychosis, decreased appetite or increased appetite in depression, suicidal thoughts, and rapid mood switching (p ¼ .01-.04). Omission of the effect of site/wave of ascertainment from the model inflated the significance level of the apparent familial association of all subphenotypes from one to four orders of magnitude. We have found evidence of familiality for subphenotypes of BP, some of which had not been identified previously identified as familial. In multicenter samples, familiality may be overestimated if variability in diagnosis of subphenotypes between site/wave of ascertainment is not considered. P56 DOPAMINERGIC AND SEROTONERGIC GENES IN BIPOLAR DISORDER: A CASE CONTROL STUDY IN A SARDINIAN POPULATION Giovanni Severino, Donatella Congiu, Mirko Manchia, Caterina Chillotti, Raffaella Ardau, Maria Del Zompo Unit of Clinical Pharmacology, Department of Neurosciences ‘‘B.B. Brodie’’, University of Cagliari, Italy Objectives: Several lines of evidence point to a role for dopamine and serotonin neurotransmission in mood disorders and in particular in Bipolar Disorder. We conducted a case-control study in order to investigate a possible association between the Bipolar Disorder using the: Ser9Gly variant in the D3 receptor gene (DRD3), 40-bp VNTR in the 30 untranslated region of the dopamine transporter gene (SLC6A3), the T102C polymorphism in 5-HT2A receptor gene (HTR2A), the
Abstracts Cys23Ser substitution in the 5-HT2C receptor gene (HTR2C) and the 5HTTLPR polymorphism in the serotonin transporter gene (SLC6A4 ). Methods: We genotyped a sample of 300 unrelated Sardinian bipolar patients (BP) (110 males, 190 females) according to RDC and 277 healthy Sardinian control anonymous blood donors (133 males, 144 females). BP were recruited at the outpatients unit of the Section of Clinical Pharmacology, Department of Neuroscience, University of Cagliari. Lifetime consensus diagnoses were achieved using the SADSL and available medical records. The study was approved by the local Ethics Committee and informed written consent obtained from all subjects. Polymorphisms were analyzed by PCR. Statistical analysis was performed using Chi-square or Monte Carlo test. Results: BP consisted in 46 bipolar II, 180 bipolar I and 74 schizoaffective disorder bipolar types. We did not found any statistically significant difference in allele frequencies between BP and controls for all genes: DRD3 [p-value ¼ 0.81; OR ¼ 1.04 (95% CI ¼ 0.80- 1.33)]; SLC6A3 [p-value ¼ 0.34]; HTR2A [p-value ¼ 0.08, OR ¼ 1.23 (95% CI ¼ 0.97–1.55)]; HTR2C [p-value ¼ 0.45; OR ¼ 0.86 (95% CI ¼ 0.61–1.21)]; SLC6A4 [p-value ¼ 0.15; OR ¼ 1.19 (95% CI ¼ 0.94–1.50)]. Conclusion: Our preliminary analysis does not show any significant association between Bipolar Disorder and these candidate genes.
P57 ASSOCIATION ANALYSES OF PDLIM5 GENE IN BIPOLAR DISORDER A. Squassina, D. Congiu, F. Manconi, M. Manchia, MP. Piccardi, M.Del Zompo Unit of Clinical Pharmacology, Department of Neurosciences ‘‘B.B. Brodie’’, University of Cagliari, Italy ENH (enigma homolog) is an adaptor protein that selectively binds PKCe to its substrate N-type Ca2þ channels in brain neurons. Since an altered activity of ENH is responsible for a disregulation of the PKC signalling cascade and alterations in PKC activity are supposed to be involved in Bipolar Disorder, ENH may play an important role in the pathophysiology of BD. ENH coding gene (PDLIM5) is located at 4p22, a region for which a linkage with BD has been detected. Previous investigations found altered expression of PDLIM5 in post-mortem brains of bipolar and schizophrenic patients. To date, two studies found association between markers spanning the ENH gene region and mental disorders (Kato et al., 2005; Horiuchi et al., 2006). In this study we investigated the association of SNPs rs1008257, rs2433320 and rs2433322 located within the 50 flanking region and the first intron of PDLIM5 in a sample comprised of 300 BP patients and 277 controls from Sardinia (Italy). All SNPs were in Hardy Weinberg Equilibrium in the bipolar group whereas only SNP rs1008257 was in HWE in controls. Therefore we carried out association analyses only for the first SNP and discarded haplotypes investigation. No association was found for the tested polymorphism when the sample was considered as a whole. However, when the investigation was carried on Bipolar I and Bipolar II disorders separately, an association between SNP 1 and BP II was detected (w2 ¼ 4.68; p ¼ 0.03). This result, if replicated in independent samples, supports a role of PDLIM5 gene in bipolar disorder.
P58 THE EFFECT OF THE DBH G444A POLYMORPHISM ON SPATIAL WORKINGMEMORY Ciara Greene1, Mark Bellgrove2, Ziarah Hawi1, Ian H. Robertson1, Michael Gill1 1 Trinity College Dublin, Dublin 2 2 University of Melbourne, Melbourne Background: The DBH gene codes for the enzyme dopamine betahydroxylase (DbetaH) which catalyses the conversion of dopamine to noradrenaline. A number of polymorphisms at or near the DBH locus have been linked with variation in DbetaH levels in the plasma and cerebrospinal fluid. A G to A substitution polymorphism at position 444 on Exon 2 of the gene has been identified, the A allele of which is associated with lower plasma DbetaH activity than the G allele. Possession of the A allele has previously been shown to have a deleterious effect on spatial working memory. Methods: 75 healthy undergraduates completed a spatial working memory task in which they were required to hold one, two or three spatial locations in working memory.
751
Results: Accuracy on the task across all genotype groups decreased with increasing memory load. When analysed with respect to DBH G444A genotype, it was found that accuracy was superior in the heterozygous (G/A) group than in either homozygous group. This trend was particularly significant at the highest memory load. Conclusions: The observed increase in accuracy with the heterozygous genotype is contrary to previous findings and may be explained by a number of mechanisms. It has been suggested that a balance of prefrontal dopamine and noradrenaline may be necessary for cognitive performance. The intermediate levels of DbetaH activity occasioned by the heterozygous genotype may therefore provide an optimal balance of dopamine and noradrenaline for working memory performance. P59 POSITIVE ASSOCIATION BETWEEN POU1F1 AND MENTAL RETARDATION IN YOUNG FEMALES He Lin1, Sun Yun2, Zhang Fuchang3, Gao Jianjun4, Guo Tingwei4, Gao Xiaocai3, Zhang Kejin3, Shi Yongyong4, Zhang Zijian5, Tang Wei4, Zheng Yonglan6, Li Sheng6, Li Xingwang4, Feng Guoyin7, Sheng Xiaoming8 1 Bio-X Center, Shanghai Jiao Tong Uni 2 Institute for Nutritional Sciences, SIBS, Chinese Academy of Sciences 3 Institute of Population and Health, Northwest University 4 Bio-X Center, Shanghai Jiao Tong University 5 Institute of Population and Health, Northwest University 6 Institute for Nutritional Sciences, SIBS, Chinese Academy of Sciences 7 Shanghai Institute of Mental Health 8 Xin Hua Hospital, Shanghai Jiao Tong University School of Medicine Genetic defects attributable to the genes involved in the hypothalamus-pituitary-thyroid gland (HPT) axis can cause abnormal thyroid hormone function and mental retardation (MR). Pit-1, encoded by the POU1F1 gene, is a pituitary-specific transcript on factor responsible for the expression of several pituitary hormones. One of the symptoms of patients with POU1F1 mutations is hypothyroidism and abnormalities of the nervous system in the period shortly after birth. We performed a case-control association study and a quantitative analysis of IQ to investigate the possible genetic contribution of POU1F1 in the Chinese Han population.Pairwise linkage disequilibrium (LD) analysis showed that rs300996, snp-7057 and rs300977 were in strong LD. There were significant differences in allele, genotype and haplotype frequencies in these three SNPs between cases and controls. When we conducted a breakdown comparison between cases and controls within different gender groups, no positive results in males were found. In females, however, we found significant differences between cases and controls in allele frequency distribution of rs300996 (P ¼ 0.0003), snp-7057 (P ¼ 0.0001) and rs300977 (P ¼ 0.0005) and in the distributions of common haplotypes combined by these SNPs (global P ¼ 0.0050). The P-value was 0.0301 for rs300996 and 0.0397 for the haplotype combination of rs300996-snp-7057-rs300977 in the analysis of the quantitative effects of the alleles and haplotypes on IQ in females. Our data suggest that POU1F1 may affect MR through a genderspecific mechanism. P60 THE COMT Val158Met POLYMORPHISM AND BRAIN MORPHOMETRY IN HEALTHY YOUNG ADULTS Janneke Zinkstok1, Nicole Schmitz2, Therese van Amelsvoort2, Maartje de Win3, Wim van den Brink4, Frank Baas5, Don Linszen2 1 Academic Medical Center Amsterdam, Amsterdam 2 Department of Psychiatry, Academic Medic, Amsterdam 3 Department of Radiology, Academic Medica, Amsterdam 4 Amsterdam Institute of Addiction Research, Amsterdam 5 Neurogenetics Laboratory, Academic Medic, Amsterdam Background: Catechol-O-Methyltransferase (COMT) is the most important mechanism for dopamine degradation in the prefrontal cortex and contains a functional polymorphism (Val158Met) influencing enzyme activity. The low-activity Met allele has been associated with better performance on cognitive tasks relying on the prefrontal cortex. Whether COMT also affects brain structure is still unclear. This study investigated the relationship between the COMT Val158Met polymorphism and brain anatomy in healthy young adults. Methods: In a cross-sectional study, structural MRI data and DNA for COMT genotyping were obtained from 154 healthy young adults. Statistical Parametric Mapping software (SPM2) and optimised voxel-
752
Abstracts
based morphometry were used to determine total and regional gray and white matter density differences between genotype groups, as well as age-related gray and white matter density differences within the genotype groups. Results: We found a significant effect of COMT genotype on agerelated differences in gray and white matter density in females but not in males. In female Val carriers, increased gray matter in the temporal lobe and the cerebellum and increased white matter in the frontal lobes were positively correlated with age; in female Met homozygotes, decreased gray matter density in the parietal lobe and decreased white matter density in the frontal lobes, the parahippocampal gyrus and the corpus callosum were positively correlated with age. Conclusion: These results suggest that the COMT Val158Met polymorphism in females may affect age-related differences in gray and white matter density. P61 INTERACTIONS BETWEEN BDNF Val66Met AND GRIN2B 366C/G AND rs890G/T POLYMORPHISMS Aleksandra Rajewska-Rager1, Maria Skibinska1, Monika Dmitrzak-Weglarz, Pawel Kapelski, Anna Leszczynska-Rodziewicz, Agnieszka Slopien, Joanna Hauser University of Medical Sciences, Poznan Summary: Brain-derived neurotrophic factor (BDNF) is a neurotrophic factor that regulates a variety of neuromodulatory processes during development and adulthood. There is growing evidence for the role of NMDA receptor genes GRIN2B (glutamate receptor, ionotropic, N-methyl-d-aspartate 2B) in the pathophysiology of schizophrenia. The aim of this study was to find possible interactions between Val66Met polymorphism BDNF gene and polymorphisms: 366C/G, and rs890G/T of GRIN2B gene. Methods: We analysed 450 patients with schizophrenia, diagnosed according to DSM-IV and ICD-10 criteria, and 484 healthy control subjects. Patients with schizophrenia where divided into subgroups: male patients (n ¼ 245), female patients (n ¼ 205), patients with early onset (n ¼ 99) and patients over 18 years old (n ¼ 351). Polymorphisms analysed: Val66Met BDNF, 366C/G GRIN2B and rs890G/T GRIN2B were established by PCR-RFLP. Results: We have found interactions between the V/M/T/T combination of polymorphisms: Val66Met BDNF and rs890G/T GRIN2B in the female group (p ¼ 0.0503) and also in the over-18-year-old subgroup (p ¼ 0.02). For Val66Met BDNF and polymorphism 366C/G GRIN2B, we have found an association with the V/M/C/C combination in the over-18 subgroup(p ¼ 0.0389) This data indicated that a combination of specific genotypes may be associated with schizophrenia, and further studies assessing the interactions between BDNF and NMDA gene polymorphisms should be performed on larger groups. P62 N-BACK TEST IN HEALTHY SUBJECTS: RELATION TO CANDIDATE GENES FOR MAJOR PSYCHOSES Alina Borkowska1, Janusz K. Rybakowski2, Marta Tomaszewska3, Monika Wilkosc3, Maria Skibinska4, Monika Dmitrzak-Weˆglarz5, Joanna Hauser4 1 Neuropsychology Unit UMK Collegium Med, Bydgoszcz 2 Department of Adult Psychiatry, Poznan U, Poznan 3 Clinical Neuropsychology Unit, Nicolaus, Bydgoszcz 4 Dept Adult Psychiatry, Laboratory of Psy, Poznan 5 Dept. Adult Psychiatry, Laboratory of P, Poznan Genes connected with the dopaminergic, serotonergic, glutamatergic and neurotrophin systems have been implicated in the pathogenesis of major psychoses and in prefrontal cognitive dysfunctions. This study attempted to establish an association between the polymorphisms of selected candidate genes in these systems and performance on N-back test, measuring prefrontal cortex (PFC) function in healthy subjects, as compared to schizophrenic and bipolar patients. The study included 108 healthy subjects, 121 patients with schizophrenia and 154 patients with bipolar disorder, aged 18–56 years. All subjects performed N-back test (Coppola version), measuring visuospatial working memory and motor performance: the % of correct reactions and reaction time were assessed. Genotyping was done for polymorphisms of dopaminergic (DRD1: 48A/G), serotonergic (5HTTPRL), glutamatergic (FYN: 93A/G, IVS10þ37T/C and Ex12þ894T/G) and neurotrophin (brainderived neurotrophic factor—BDNF: C/T, Val/Met)genes. For the % of
correct reactions on the test, significant associations with the polymorphisms of FYN gene were obtained in healthy subjects and in schizophrenic patients, but not in bipolar patients. For reaction time on the test, a marginal significance (p ¼ 0.06) was found in healthy females for 5HTTPRL and BDNFC/T polymorphism, which was not reflected in patients with major psychoses. The Src-family tyrosine kinase FYN plays a key role in the interaction between BDNF and glutamatergic receptor NMDA in the prefrontal cortex, and the results obtained may underscore a role of the glutamatergic system in PFC activity in healthy subjects and in schizophrenic patients. P63 THE ROLE OF SINGLE NUCLEOTIDE POLYMORPHISMS (SNPs) IN THE MINERALOCORTICOID RECEPTOR (MR) IN STRESS RESPONSE AND PSYCHOPATHOLOGY Nienke van Leeuwen1, Stefan Wust2, O.C. Meijer3, E.R. de Kloet3, F.G. Zitman4, R.H. de Rijk3 1 Medische Farmacologie/LACDR, Leiden 2 Dept. of Psychobiology, University of Trier 3 Medical Pharmacology, University of Leiden 4 Dept. of Psychiatry, LUMC, Leiden Stress causes activation of the hypothalamic-pituitary-adrenal (HPA)axis, resulting in the secretion of corticosteroids, which facilitate behavioural adaptation. Variability in the stress-response in individuals is thought to underlie differences in susceptibility to psychopathology. Central in the regulation of the stress-response are two brain corticoid receptors, the mineralocorticoid receptor (MR) and the glucocorticoid receptor (GR). This project is designed to test whether genetic polymorphisms in the MR-gene underlie variability in stressresponsiveness and if this genetic variability is associated with psychopathology. The MR-gene was screened in 50 anonymous DNA samples from the Dutch population for SNP frequency using direct sequencing. Several haplotypes were established. One SNP is significantly associated with an enhanced cortisol secretion and heart rate response (R.H. DeRijk/S. Wu¨st article submitted) in a group of healthy volunteers (n ¼ 118) subjected to a psychosocial challenge (Trier Social Stress Test). In addition to stress- induced cortisol levels, differences in the (morning) cortisol awakening rise (CAR) are also observed among healthy subjects. Interestingly, a second SNP was found to be associated with this CAR. In vitro functionality assays showed clear changes in the capacity of the two SNPs to activate luciferase by different glucocorticoid responsive elements (GRE’s), when cortisol was used as a ligand. These functional SNPs will now be tested in different groups of patients with mood, anxiety and/or somatoform disorders. Groups are currently being collected and extensively phenotyped. P64 THE STROOP TEST IN HEALTHY SUBJECTS AND CANDIDATE GENES FOR MAJOR PSYCHOSES Marta Tomaszewska1, Alina Borkowska2, Monika Wilkosc2, Maria Skibinska3, Monika Dmitrzak-Weglarz3, Joanna Hauser3 1 Clinical Neuropsychological Unit, Bydgoszcz 2 Clinical Neuropsychology Unit, Nicolaus, Bydgoszcz 3 Laboratory of Psychiatric Genetics, Pozn, Poznan The Stroop Color Word Interference Test is a special method for the assessment of verbal working memory, reflecting an activity of the frontal part of the cingulate cortex. In this study, an association between performance on the Stroop Color-Word Interference Test in healthy subjects and the polymorphism of selected candidate genes for major psychoses were assessed. The study included 108 healthy volunteers (39 males, 69 females), aged 18–60 (mean 32 10. 6)years. Neuropsychological assessment was performed using the Stroop Color-Word Interference Test, Parts A and B. Part A of the test (Reading Color Names in black -RNCb), measures verbal abilities and attention; Part B (Naming Color of Word- different - NCWd), measures verbal working memory and executive functions. Genotyping was done for polymorphism of dopaminergic (D1 receptor, DRD1: 48 A/G; catechol-O-methyltransferase, COMT 108/158 Val/Met) glutamatergic (FYN kinase: 93 A/G, IVS10þ37T/C and Ex12þ894T/G) and neurotrophin (brain-derived neurotrophic factor, BDNF: C/T, Val66Met) genes. Performance on Stroop A was associated with the polymorphism - 48 A/G of DRD1 (better results in heterozygotes), with COMT 108/158 Val/Met polymorphism (better results in val/val subjects). Performance on Stroop B in males was associated with BDNF C/T polymorphism; people with C/C genotype
Abstracts obtained better results as compared to other genotypes. The results obtained may suggest a contribution of selected candidate genes for major psychoses to performance on the Stroop test in healthy subjects. P65 WISCONSIN CARD SORTING TEST IN HEALTHY SUBJECTS AND PSYCHIATRIC CANDIDATE GENES Monika Wilkosc1, Alina Borkowska2, Marta Tomaszewska3, Maria Skibinska4, Aleksandra Szczepankiewicz4, Monika Dmitrzak-Weglarz4, Joanna Hauser5, Janusz Rybakowski6 1 Clinical Neuropsychology Unit CM UMK, Bydgoszcz 2 Clinical Neuropsychology Unit Nicolaus C, Bydgoszcz 3 Clinical Neuropsychology Unit, Nicolaus, Bydgoszcz 4 Laboratory of Psychiatric Genetics, Pozn, Poznan 5 Laboratory of Psychiatric Genetics, Poz, Poznan 6 Department of Adult Psychiatry, Poznan U, Poznan The Wisconsin Card Sorting Test (WCST) is a principal tool used for neuropsychological assessment of prefrontal cortex activity. The aim of this study was assessing association between performance on the WCST, measuring working memory and executive functions, and polymorphisms of selected candidate genes for major psychoses. The study was performed on 108 healthy volunteers aged 18–60 years, with no history of substance abuse or serious illnesses. Neuropsychological assessment was performed using the WCST Heaton Version and evaluating the following domains: perseverative errors (inability to change reaction due to ignorance of relevant stimuli), nonperseverative errors (attentional inability to avoid distraction), number of completed categories (ability to utilise new information and previous experiences), percent of conceptual responses (ability of conceptual thinking) and set to complete 1st category (ability to formulate a logical conception). Genotyping were done for polymorphism of dopaminergic (D1 receptor, DRD1-48A/G, catechol-O-methyltransferase, COMT108/158Val/Met), glutamatergic (FYNkinase:93A/G, IVS10þ37T/C, Ex12þ894T/G) and neurotrophin (brain-derived neurotrophic factor, BDNF:C/T, Val66Met) genes. Males with A/Gpolymorphism of DRD1 gene obtained better results on trials to complete 1st category, compared to G/G polymorphism. Better performance on nonperseverative errors was observed in females with val/val genotype of COMT. The C/T genotype of C-270T BDNF polymorphism was associated with higher percentages of conceptual responses. The results obtained suggest a contribution of dopaminergic and neurotrophin system genes to working memory and executive function efficiency, connected with prefrontalcortex activity, in healthy subjects. P66 TGFB-INDUCED FACTOR (TGIF): CANDIDATE GENE FOR PSYCHOSIS in18p11.3 Ivan Chavarria-Siles1, Consuelo Walss-Bass2, Byung Dae Lee2, Paulina Quezada2, Albana Dassori2, Salvador Contreras2, Rolando Medina2, Mercedes Ramirez2, Armas Regina3, Laura Almasy4, Rodolfo Salazar5, Henriette Raventos5, Michael Escamilla2 1 University Texas Health Science Center, San Antonio, Texas 2 Psychiatric Genetics Research Center, Un, San Antonio, Texas 3 Langley Porter Psychiatric Institute, Uni, San Francisco, California 4 Southwest Foundation for Biomedical Research, San Antonio, Texas 5 Centro de Investigacion en Biologia Celu, San Pedro, San Jose We previously identified a putative gene locus for the phenotype of psychosis in chromosome 18p11.3 (marker D18S63) in 393 subjects from the Central Valley of Costa Rica (CVCR). We replicated the association to a specific allele (allele 11on D18S63) in an independent sample of the CVCR (312 families, 880 subjects, 342 with history of psychosis) using Family Based association Test (FBAT) analysis (Bi-allelic P value ¼ 0.0233). In a combined analysis of all 476 families of the CVCR (1273 subjects, 490 with a history of psychosis), Allele 11 on D18S63 showed evidence of association (FBAT Bi-allelic P value ¼ 0.0005). Next, we completed fine mapping analyses to identify candidate genes within this region using 9 SNPs surrounding D18S63 in the combined sample. The markers studied span four genes: Myomesin-1 (MYOM1), Myosin regulatory light chain (MRLC2), TGFB-induced factor (TGIF) and Disks large-associated protein (DAP1). Using FBAT, none of the additional SNPs showed independent association with the phenotype of psychosis. Linkage disequilibrium (LD) analysis showed that D18S63 lies within the TGIF gene and was in complete LD (D’ ¼ 1.00)with the 3 SNPs within TGIF, and none of the other markers showed strong LD with D18S63.A specific haplotype for
753
markers D18S63- rs47988096-rs3866990 (all within the TGIF gene) showed evidence of association (bi-allelic P value ¼ 0.0017), although the global haplotype test was non-significant. Altogether, our results point to the possible involvement of TGIF in the pathophysiology of psychotic-related disorders. P67 DIFFERENTIAL DOPAMINE RECEPTOR D4 ALLELE ASSOCIATION WITH ADHD DEPENDANT OF PROBAND SEASON OF BIRTH Xiaohui Xu1, Keeley Brookes1, Ben Neale1, Wai Chen1, Jan Buitelaar2, Richard Ebstein3, Tobias Banaschewski4, Edmund Sonuga-Barke5, Iris Manor6, Ana Miranda Ana7, Bob Oades8, Herbert Roeyers9, Ari Rothenberger10, Jo Sergeant11, Hans-Christoph Steinhausen12, Steve Faraone13, Michael Gill14, Anita Thapar15, Philip Asherson1 1 Institute of Psychiatry, London 2 Radboud University Nijmegen Medical Cent, Nijmegen 3 S. Herzog Memorial Hospital, Jerusalem 4 University of Goettingen 5 University of Southampton 6 Geha MHC, Petach-Tikva 7 University of Valencia 8 University Clinic for Child and Adolescents, Essen 9 Ghent University, Ghent 10 University of Go¨ttingen 11 University of Amsterdam 12 University of Zurich 13 SUNY Update University, New York 14 Trinity Centre for Health Sciences, Dublin 15 University of Cardiff Season of birth is an indicator of seasonal environmental risk factors that has been associated with attention deficit hyperactivity disorder (ADHD) in two existing studies. In addition, one study reported interaction between season of birth and genotypes of the dopamine D4 receptor gene. In this study, we investigated the season of birth association with ADHD in four independent samples collected for molecular genetic studies of ADHD and found a small but significant increase of 3.3% in the number of summer births compared to UK census data (nominal P ¼ 0.007). The data showed a small but non-significant excess transmission of the 7-repeat allele in each of the four independent datasets, with the combined dataset just reaching nominal significance for the allele-specific association (nominal p < 0.04, odds ratio ¼ 1.18). We further observed under-transmission of the 2-repeat allele (OR ¼ 0.3, nominal-P ¼ 0.05) and over-transmission of the 7-repeat allele (OR ¼ 1.33, nominal- P ¼ 0.02)in ADHD probands born in the winter months. Overall there was a significant difference in the transmission of the 2-repeat allele between the two seasonal groups (nominal-P ¼ 0.009). Inspection of the combined dataset analysed month-by-month did not however show a clear pattern of seasonal variation in transmission of either allele to ADHD probands, and these findings are therefore more likely to represent random sampling rather than a true effect. P68 ASSOCIATION STUDY BETWEEN GILLES DE LA TOURETTE AND THE ROUNDABOUT3 AND ROUNDABOUT4 GENES Debora Miranda1, Karen Wigg2, Yu Feng2, Paul Sandor2, Cathy Barr3 1 UFMG, Belo Horizonte 2 Toronto Western Research Institute, Toronto 3 Toronto Western Research Institute, Toronto Gilles de la Tourette Syndrome (GTS) is a neuropsychiatric disorder characterised by tics and long known to be familial. Evidence from twin studies indicate that it has a substantial genetic component. Three different genome scan studies found evidence suggesting linkage to chromosome region 11q23-24. In this region we identified two possible susceptibility genes, the roundabout (ROBO)3 and roundabout (ROBO)4 genes. ROBOs are the receptors for SLIT protein. SLITROBO proteins are known to play a role in axonal midline crossing in the spinal cord during neurodevelopment. Mutations in the SLITRK1 gene have recently been identified in three patients with GTS. While the SLITRK1 gene appears to account for only a few cases of GTS, this finding may provide leads to other susceptibility genes. On this basis, we investigated the ROBO3 and ROBO4 genes located on the chromosome 11q24.2 linked region. We examined the inheritance of polymorphisms in ROBO3 and ROBO4 genes in 155 nuclear families
754
Abstracts
identified through a GTS proband with 255 affected children. We used markers that tagged the major haplotypes of these genes to ensure adequate coverage. We found no significant evidence for association using the transmission disequilibrium tests for these two genes. Based on our families, we conclude that these genes are not major genetic factors contributing to susceptibility to GTS; however, other genes in the SLITRK-ROBO pathway may be involved and should be investigated. P69 SEX SPECIFIC ASSOCIATION BETWEEN D2 DOPAMINE RECEPTOR POLYMORPHISM AND PANIC DISORDER Se-won Lim1, Kang-Seob Oh1, Min-Soo Lee2 1 Kangbuk Samsung Hospital, Seoul 2 College of Medicine, Korea University, Seoul Purpose: To examine (1) the association of DRD2 TaqI A polymorphism and panic disorder and (2) gender difference of possible association. Methods: 114 patients with panic disorder (62 males (54.4%), mean age 40.96 10.11 years) and 200 comparison subjects (114 males (57.0%), mean age 35.57 8.81 years) were tested for DRD2 TaqI A polymorphism. We excluded panic patients with comorbid alcohol related disorders, bipolar disorders and any kind of psychotic disorders because there are already some reports on association of these diseases with DRD2 TaqI A polymorphism. Results: There was some difference in the frequency of the genotype in DRD2 polymorphism between patients and controls (w2 ¼ 6.09, df ¼ 2, p ¼ 0.048). A1þ allele (A1A1 and A1A2) frequency analysis also showed significant association (w2 ¼ 4.08, df ¼ 1, p ¼ 0.043). In addition, we observed a stronger and more specific association between panic disorder and the A1þ allele of the DRD2 TaqI A polymorphism for men (w2 ¼ 4.71,df ¼ 1, p ¼ 0.03), but not for women (w2 ¼ 0.45, df ¼ 1, p ¼ 0.50). Conclusions: These results in our Korean sample suggest that the DRD2 TaqI A polymorphism may be associated with panic disorder. The observed association was more significant with A1 allele carrier frequency analysis. Furthermore, we found sex-specific association of DRD2 A1 allele with panic disorder only in males. P70 BOTH NATURE AND NURTURE ENTAIL VIOLENCE: SEROTONERGIC GENES AND ADVERSE CHILDHOOD ENVIRONMENT Andreas Reif1, Christine Freitag2, Michael Ro¨sler3, Marc Schneider3, Andrea Eujen4, Denise Wenzler3, Christian Kissling4, Christian Jacob4, Petra Retz-Junginger3, Johannes Thome4, Klaus-Peter Lesch4, Wolfgang Retz3 1 Dept. of Psychiatry, Univ. of Wu¨rzburg, Wu¨rzburg 2 Department of Child and Adolescent Psych, Homburg 3 Institute for Forensic Psychology and Ps, Homburg 4 Clinical and Molecular Psychobiology, Wu¨rzburg 4 Department for Psychiatry, The School of Swansea Aggressive behaviour is influenced by variation in genes of the serotonergic circuitry and also early life experience. The present study aimed to investigate the contribution of polymorphisms shown to moderate transcription of two genes involved in serotonergic neurotransmission (serotonin transporter, 5HTT, and monoamine oxidase A, MAOA) of the development of violence and to test for gene environment interactions relating to adverse childhood environment. A cohort of 184 adult male volunteers referred for forensic assessment participated. Each individual was assigned either to a violent or non-violent group. Degree of adversity during childhood was significantly associated with violent behaviour. 45% of violent, but only 30% of non-violent individuals carried the low-activity, short MAOA allele (p ¼ 0.040). In the violent group, carriers of low-function short variants of 5HTT were found in 77%, as compared to 59% in the nonviolent group (p ¼ 0.015). Thus, both variants were significantly associated with violent crime. Logistic regression was performed and the best fitting model, with a predictive power of 70%, revealed a significant, independent effect of childhood environment and MAOA genotype. In addition, a significant influence of interaction between childhood environment and 5HTT genotype on violent behaviour was found. None of the models supported a MAOA 5HTT interaction effect, MAOA environment interaction or an independent effect of 5HTT genotype. MAOA genotype is independently associated with
violent crime, while there is a relevant 5HTT environment interaction, with heterozygous individuals showing the greatest vulnerability to childhood adversity. P71 EFFECT OF ADENOSINE A2A RECEPTOR GENE POLYMORPHISMS ON MIGRAINE WITH AND WITHOUT AURA Christa Hohoff1, Martin Marziniak2, Klaus-Peter Lesch3, Ju¨rgen Deckert4, Claudia Sommer5, Rainald Mo¨ssner3 1 Dept. of Psychiatry, University of Mu¨nster 2 Dept. of Neurology, University of Homburg, Homburg and Wu¨rzburg 3 Dept. of Psychiatry, University of Wu¨rzburg 4 Dept. of Psychiatry, University of Mu¨nster 5 Dept. of Neurology, University of Wu¨rzburg The adenosine A2A receptor facilitates effects of calcitonin generelated peptide (CGRP) and vasoactive intestinal polypeptide (VIP), two important neuropeptides in migraine pathophysiology. Furthermore, the A2A receptor is the molecular target of caffeine, which is effective in the treatment of migraine attacks in combination with analgesic drugs. In the present study we therefore investigated the effect of A2A receptor gene polymorphisms on migraine susceptibility. 265 migraine patients (122 with aura and 143 without aura), diagnosed according to the International Headache Society, and 154 migrainefree controls were assessed. They were genotyped for six single nucleotide polymorphisms (SNPs), three of which have been previously described as validated exonic variants. The additional three SNPs cover the intronic and the 50 and 30 flanking regions of the A2A receptor gene and were derived from the Hapmap database. All SNPs were in Hardy-Weinberg-equilibrium and in high linkage disequilibrium. While no significant difference existed at the single marker level, a six-marker haplotype was found to be more frequent in migraine patients with aura (p < 0.01) but not in patients without aura, compared to the control group. This indicates a contribution of A2A receptor gene polymorphisms to the pathogenesis of migraine with aura. Nevertheless, replication studies in independent, preferably family-based samples are warranted. P72 CANDIDATE GENES AND SUICIDE COMPLETION: SEROTONERGIC PATHWAY Alja Videtic1, Pungercic Galina2, Tomazˇ Zupanc3, Zpanic Pajnic Irena3, Balazic Joze3, Tomori Martina3, Komel Radovan2 1 Medical Faculty, Ljubljana 2 Medical Faculty, Institute of Biochemist, Ljubljana 3 Medical Faculty Ljubljana, Institute of Ljubljana Suicide is an important mental and health problem and is among the ten leading causes of death in all ages worldwide. Slovenia ranks among countries with the highest suicide rate (25/100,000 people, reported in 2003). Several molecular genetic studies emphasise the involvement of genetic factors in the etiology of suicidal behaviour. There is considerable evidence that genes coding for proteins involved in the serotonergic system (tryptophan hydroxylase (TPH), serotonin (5-HT) transporter (5-HTT) and serotonin receptors (5-HTR)) are major candidate genes. The study population consisted of 226 suicide victims and 225 controls, all of Slovenian origin. Genes of the serotonergic pathway included in the present study had been previously linked to psychiatric disorders. Variations in seven polymorphisms in four serotonin receptor genes were investigated: 5-HTR1A C(47)T and G(815)A, 5-HTR1B G(861)C and A(161)T, 5-HTR1F C(78)T and C(528)T, and 5-HTR2A C(1420)T. On serotonin transporter gene we analysed insertion/deletion of 44bp in the promoter region and VNTR in the second intron. No statistically-significant differences in genotype and allele frequency distributions between the suicide and control groups were found. Only for polymorphism C(1420)T of 5-HTR2A was a weak association with suicide. The informative nature of our results may contribute to elucidation of candidate genes for suicide.
P73 A FUNCTIONAL POLYMORPHISM IN THE MAOA GENE PROMOTER PREDICTS CENTRAL DOPAMINE FUNCTION AND BMI Francesca Ducci1, Timothy K. Newman2, Sam Funt2, Matti Virkkunen3, David Goldman2
Abstracts 1
NIAAA/NIH, Rockville, MD NIAAA/LNG, Rockville, MD University of Helsinki, Helsinki, Finland
2 3
Monoamine oxidase A deaminates the monoamine neurotransmitters serotonin, dopamine and noradrenalin. The MAOA gene contains a length polymorphism in its promoter region (MAOA-LPR) that affects transcriptional efficiency. Our goals were to test 1) whether MAOALPR contributes to variation in monoamine activity, assessed using levels of cerebrospinal fluid (CSF) monoamine metabolites and 2) whether MAOA-LPR genotype influences alcoholism and/or body mass index (BMI). Male, unrelated alcoholics(N ¼ 278) and controls (N ¼ 227) were collected from a homogeneous Finnish source population. CSF concentration of 5-hydroxyindoleacetic acid, homovanillic acid (HVA) and 3-methoxy-4- hydroxyphenylglycol were available from 208 participants. MAOA-LPR genotypes had a significant effect on HVA concentration (p ¼ 0.01) but explained only 3% of the total variance. Specifically, the genotype conferring high MAOA activity was associated with lower HVA levels in both alcoholics and controls, a finding that persisted after accounting for the potential confounds of alcoholism, BMI, height and smoking. MAOA-LPR genotype predicted BMI (p < 0.005), the high activity genotype being associated with lower BMI. MAOA-LPR genotypes were not associated with alcoholism or related psychiatric phenotypes in this dataset. Our results suggest that MAOA-LPR allelic variation modulates dopamine turnover in the CNS, but does so in a manner contrary to our prior expectation that alleles conferring high activity would predict higher HVA levels. Finally, our work suggests that MAOA may be involved in the regulation of BMI. Independent samples are necessary to confirm this hypothesis. P74 LACK OF ASSOCIATION OF DRD2 AND COMT POLYMORPHISMS WITH BEHAVIOURAL MOTIVATION Katrina Jane Light, Peter R. Joyce, Allison L. Miller, Martin A. Kennedy University of Otago, Christchurch Based on their central roles within the dopamine system, the dopamine D2 receptor (DRD2) and catechol-O-methyltransferase (COMT) have been investigated as candidate genes for behaviour motivation. Single nucleotide polymorphisms in both the COMT and DRD2 genes have been associated with a reduction in enzymatic activity and receptor density, respectively. Based on the apparent functional effect of these two gene variants, attempts have been made to associate these polymorphisms with measures of behavioural motivation. Reuter et al. (2006) have recently reported that interaction of the DRD2 TaqI A and COMT Val polymorphisms influences an individual’s behavioural approach system (assessed based on Gray’s Reinforcement Sensitivity Theory). Based on the conceptual comparability of Gray and Cloninger’s personality models, we attempted to replicate this finding. We administered the Temperament and Character Inventory (TCI) to two independent samples of depressed outpatients (for both samples n ¼ 146). No significant effect of DRD2 or COMT, either independently or epistatically, on novelty-seeking or harm-avoidance scores was observed, within either sample. Age, but not gender, was significantly associated with personality scores, whereby novelty seeking declined with age. This study finds no support for the COMT-DRD2 Equilibrium Model of Positive Emotionality hypothesised by Reuter et al. (2006). Because of differences in the personality measure and/or sample composition, our findings do not exclude the possibility that these genes interact to influence personality. Future analyses will investigate further variation within these genes and alternative measures of behaviour motivation, such as reward dependence. P75 AMISYN, SCAMP5 AND NBEA ARE CANDIDATE GENESFOR AUTISM AND SUGGEST A ROLE FOR NEURON VESICLE TRAFFICKING IN THE PATHOGENESIS OF AUTISM Dries Castermans, Kathleen Freson, Rita De Vos, Wim Van de Ven, Christel Ven Geet, Jean Steyaert, John Creemers, Koen Devriendt Catholic University of Leuven, Leuven Autism is a neurodevelopmental disorder of unknown cause and pathogenesis. We initiated a positional cloning strategy, starting from three unrelated persons with idiopathic, non-familial autism carrying a de novo chromosomal translocation. We identified genes in which the
755
corresponding proteins have a possible function in vesicle trafficking in the brain. Patient 1: Neurobeachin (NBEA) is a neuron-specific Akinase anchoring protein, recruiting PKA to the appropriate subcellular location. Patient 2: SCAMP5 is the brain-specific member of the secretory carrier membrane protein family involved in vesicle budding from trans-Golgi. Patient 3: Amisyn is involved in regulating SNARE complex formation, by binding to syntaxin-1 and SNAP25. We studied the involvement of these genes in the regulated secretory pathway of dense-core secretory granules (DCSG, counterparts of neuronal large dense-core vesicles) by overexpression versus RNAi-mediated knockdown in mouse beta-TC3 neuroendocrine cell line. In a transient secretion-induction assay, we showed a role for NBEA, amisyn and SCAMP5 as negative regulators of neuron vesicle trafficking and/or fusion. Interestingly, for all 3 genes, we observed structural abnormalities of DCSGs in these cell lines. In addition, preliminary in vivo data on the blood platelets of these patients revealed specific abnormalities in the morphology of dense granules as well. As a whole, the identification of genes affected by translocations in patients with autism, in combination with the in vitro and in vivo functional and morphological data, implies that vesicle trafficking in neurons may be involved in the pathogenesis of autism.
P76 GENETICS AND PERSONALITY TRAITS IN PATIENTS WITH SOCIAL ANXIETY DISORDER: A CASE-CONTROL STUDY IN SOUTH AFRICA Lochner Lochner1, Sian Hemmings2, Soraya Seedat1, Craig Kinnear2, Renata Schoeman3, Kristina Annerbrink4, Marie Olssen5, Elias Eriksson4, Johanna Moolman-Smook2, Christer Allgulander3, Dan Stein3 1 MRC Unit on Anxiety and Stress Disorders, Cape Town 2 MRC/US Centre on Molecular and Cellular, Cape Town 3 Department of Psychiatry, University of Cape Town 4 Department of Pharmacology, Go˜teborg Uni, Sweden 5 Department of Pharmacy, Apoteket AB, Sah, Sweden 3 Karolinska Institutet, Sweden, Sweden Although there is evidence that social anxiety disorder (SAD) has a familial basis, there are few studies of potential candidate genes. There is also the possibility that temperamental risk factors for SAD may be genetically transmitted. Our aims were: i.) to compare patients and controls’ w.r.t. personality traits, and ii.) to investigate the role of geneencoding components of serotonergic (5-HT) and dopaminergic (DA) pathways in SAD patients and to compare differences in temperament between carriers of different alleles for selected polymorphisms. 63 patients with SAD and SPIN-scores > 18, and controls (n ¼ 150) were included. The Temperament and Character Inventory measured personality. DNA was genotyped to investigate the role of select candidate genes encoding components in serotonergic and dopaminergic pathways in mediating SAD. To achieve this, the frequency of variants in 5-HT and DA genes was compared between a Caucasian subset of SAD patients (n ¼ 41) and controls (n ¼ 88), using case-control association analyses. The frequency of variants in 5-HT and DA-related genes across temperament was investigated using ANOVA. In Caucasians, there was a difference between patients and controls in distribution of the 5-HT2A T102C polymorphism, with significantly more patients harbouring T-containing genotypes (p ¼ 0.012). After correction for multiple testing, a trend towards significance was observed between the 5-HT2A T102C T-containing genotypes and decreased reward dependence (p ¼ 0.096). The results suggest a possible role for reward dependence and the 5-HT2A T102C polymorphism in SAD.
P77 GENOMIC ALTERATIONS IN A PATIENT WITH SCHIZOPHRENIA, CROHN’s DISEASE AND POLYCYSTIC KIDNEY DISEASE Mette Nyegaard1, Eigil Kjeldsen2, Søren Pedersen3, Ole Mors4, Anders Børglum5 1 Institute of Human Genetics, Aarhus 2 Cancercytogenetics Laboratory, Dept. of Medical Haematology, Aarhus 3 Department of Clinical Genetics, The Barthol, Aarhus 4 Institute for Basic Psychiatric Research, Aarhus 5 Institute of Human Genetics, The Barthol, Denmark
756
Abstracts
A patient diagnosed with schizophrenia, Crohn’s disease and polycystic kidney disease was analysed for quantitative genomic changes using the Agilent 60-mer oligo-based CGH array, which contains over 40,000 probes. Genomic amplifications and/or deletions were found in multiple regions of the genome, including chromosomes 3 and 6. To validate these alterations, quantitative PCR was performed using genomic DNA from the patient and compared to multiple healthy control individuals. In three genomic regions, the alterations originally found in the patient were also found in healthy controls. In one region representing a gene, however, the patient was shown to harbour an approximately 30,000 bp deletion, not found in any of the control individuals. This gene is located on chromosome 3 and it widely expressed in the brain. The gene encodes a protein in the voltage-dependent calcium channel complex and could be a candidate susceptibility gene for schizophrenia. P78 A TREND ASSOCIATION BETWEEN s/s 5-HTTLPR GENOTYPE AND ANXIETY DISORDERS Esther Molina1, Blanca Gutie´rrez1, Margarita Rivera1, Francisco Torres1, Juan Bello´n2, Berta Moreno-Ku¨stner1, Jose´ A. Lorente1, Juan de Dios Luna1, Jorge Cervilla1 1 Universidad de Granada, Avd. Madrid, 11 Granada 2 CAP ‘‘El Palo’’ Ma´laga, Ma´laga Dysfunctions in the serotonergic system seem implicated in both generalised anxiety disorder (GAD) and panic disorder (PD). Furthermore, treatment of such anxiety disorders focuses on inhibiting the serotonin transporter (SERT). To evaluate the association between variation in the SERT gene and anxiety disorders, we genotyped the 5-HTTLPR polymorphism using 605 subjects who attended primary care centres in Spain. Participants were assessed for psychiatric diagnosis using the PRIME-MD scale to establish the presence of both GAD and PD according to ICD-10 criteria. 162 subjects had GAD and/or PD and 443 were controls. Patients who had depression without either of these anxiety disorders were excluded from this analysis. We found a trend for association between s/s genotype and anxiety disorders, although differences did not reach statistical significance (OR ¼ 1.453; 95%C.I. ¼ 0.96–2.19, p ¼ 0.073). Nevertheless, when we reanalysed data using comorbid anxiety and depression as outcomes, we found a significantly high frequency of s/s genotypes amongst comorbid cases as compared to controls (30.77 vs 21.44, p ¼ 0.02). The association remained significant after adjusting for age and sex (O.R ¼ 1.593; 95%C.I: 1.02–2.49, p ¼ 0.041).Our findings indicate that anxiety, particularly if comorbid with depression, is possibly associated with the s/s genotype. However, we need further statistical power to confirm the trend we report here. We thank Prof. Michael King and the PREDICT study core group for their cooperation. P79 A HAPLOTYPE-BASED ANALYSIS OF NF1 GENE IN BIPOLAR DISORDER AND SCHIZOPHRENIA Diego Forero1, Liesbet Herteleer2, Karl-Fredrik Norrback3, Patrick Callaerts2, Rolf Adolfsson3, Jurgen Del-Favero4 1 Department of Molecular Genetics, VIB, Antwerpen 2 Laboratory of Developmental Genetics, VI, Leuven 3 Department of Clinical Sciences, Univers, Umea˚ 4 Applied Molecular Genomics Group, Depart, Antwerp Basic models of psychiatric disorders may be useful for the discovery of additional candidate genes. Preliminary expression data in a Drosophila model for bipolar (BP) disorder, in conjunction with available knowledge about its biological and genetic features, show that neurofibromatosis 1 (NF1; located on 17q11.2, a linkage region for BP) is an interesting functional and positional candidate gene for further exploration in psychiatric genetics. For that purpose, we genotyped 17 haplotype-tagging SNPs selected from the HapMap database on a Sequenom platform in 182 BP patients, 182 schizophrenia (SZ) patients and 364 healthy controls, all of Northern Swedish origin. Subsequent statistical analysis showed, consistent with the HapMap data, a high degree of LD in and around NF1. Indeed, the 17 tagging SNPs (spanning 200 kb) are clustered in only one haplotype block. Using different statistical approaches, we found no significant differences in allele, genotype or haplotype frequencies between BP or SZ patients and controls. Although previous studies demonstrated a key role of NF1 in the crosstalk between intracellular pathways implicated in neural function (CREB and RAS) and in some neuropsychiatric disorders (mutations in neurofibromatosis-related
alterations), the present study did not confirm a major role of NF1 as a susceptibility gene for BP and/or SZ in Northern Sweden. P80 NO ASSOCIATION BETWEEN ADULT ATTENTIONDEFICIT/HYPERACTIVITY DISORDER AND BDNF OR ITS SPECIFIC RECEPTOR NTRK2 IN A SPANISH SAMPLE Marta Ribase´s1, Josep Antoni Ramos2, Rosa Bosch2, Lucas Brunso2, Mariana Nogueira2, Nuria Go`mez2, Ester Garcı´a2, Mo`nica Baye´s3, Bru Cormand4, Miquel Casas2 1 Psychiatric Department, Barcelona 2 Psychiatric Department, Hospital Univers, Barcelona 3 Genes and Diseases Program, Center for Genetics, Barcelona 4 Genetics Department, Facultat de Biologi, Barcelona Genetic studies and animal models suggest that brain-derived neurotrophic factor (BDNF) and its specific receptor Tyrosine kinase B (NTRK2) could play an essential role in Attention-deficit/hyperactivity disorder (ADHD), and that alterations in this neurotrophic system may participate in the etiology of this neuropsychiatric disorder. In this regard, a strong association between ADHD and the Val66 allele of the BDNF gene has recently been reported; the expression of this neurotrophic factor is regulated by different psychostimulants used in the pharmacological treatment of ADHD, and conditional deletion of BDNF in the postnatal brain of knockout mice leads to hyperactivity after exposure to stressors. To evaluate the participation of this neurotrophic system in ADHD, we considered the 270C > T and Val66Met (rs6265) SNPs of the BDNF gene and the 69G > C (rs1187325), IVS15þ40G > A (rs1047896) and IVS19125T > C (rs1078947) polymorphisms of the NTRK2 gene and performed a population-based association study on a sample of 88 ADHD adult Spanish patients and 88 healthy controls. No significant association was observed when allele, genotype or estimated haplotype frequencies of the five polymorphisms were considered. The subdivision of patients according to clinical subtypes of inattentive, hyperactive-impulsive or combined ADHD did not produce any significant association either, which could be explained by the relatively small sample size. We also discarded sex influence orepistatic effects between both genes. These results do not support the participation of BDNF or its receptor NTRK2 in genetic susceptibility to ADHD.
P81 MULTIPLE SUICIDAL BEHAVIOR: STRESSFUL LIFE EVENTS, GENDER AND THE GENETIC POLYMORPHISM OF 5-HT TRANSPORTER AND MAO-A Tullio Giraldi1, Giulia Schillani2, Serena Goljevscek2, Davide Carlino3, Maurizio De Vanna3, Eugenio Aguglia3 1 Dept. of Bomedical Sciences, Trieste 2 Dept. Biomed. Sci., Trieste 3 Univ. Psychiatric Clinic, Trieste Background: Stressful life events, gender and disregulated monoaminergic transmission have been related to suicide. The severity of suicidal behaviour in multiple attempters has thus been examined in relation to life events, gender and polymorphisms of the promoter region of serotonin transporter (SERT) and MAO-A, shown to be relevant in combination with environmental factors. Methods: 39 consecutive multiple suicidal attempters of both sexes, admitted to the Psychiatric Clinic after having committed their last attempt, were interviewed, and the stressful life events reported were collected in a Life Calendar. The patients were genotyped for 5HTTLPR polymorphisms of (SERT) and for VNTR polymorphism of MAO-A. Results: The severity of suicidal behaviour depends significantly on the number of life events experienced in the 6 months before each suicide attempt. The functional activity of SERT caused by 5-HTTLPR polymorphism plays a significant role, alone or in combination with the gender of patients. Contingency tables show that patients with low functional activity of SERT, in combination with a high exposure to life events, have increased severity of suicidal behaviour in female subjects. The VNTR MAO-A polymorphism is not significantly associated with suicidal behaviour of the subjects. Conclusion: These results encourage the study of a larger number of multiple suicide attempters to confirm the present findings and clarify the role played by SERT genetic polymorphism in suicidal behaviour.
Abstracts P82 EXCLUSION OF MAJOR CANDIDATE LOCI IN MULTIPLEX ADHD FAMILIES Jobst Meyer1, Susanne Walitza2, Haukur Palmason3, Christiane Seitz2, Tobias J. Renner2, Jessica Sigmund3, Christian Vogler3, Savira Ekawardhani3, Dirk Moser3, Marcel Romanos2, Andreas Warnke2, Christine M. Freitag2, Klaus Peter Lesch3 1 Graduate School for Psychobiology, Trier 2 Department of Child and Adolescent Psych, Wuerzburg 3 Graduate School of Psychobiology, Trier 3 Department of Psychiatry and Psychotherapy, Wuerzburg Attention deficit/hyperactivity disorder is a childhood syndrome affecting 3% to 9% of school-age children. Epidemiologic studies indicate that boys represent a large majority of this group. The disorder persists into the adult years in a substantial number of patients, up to 70%. Genetic factors contribute significantly to the condition, with heritabilities ranging from 60% to 90%. ADHD shows a complex mode of inheritance. Association was replicated for alleles of the dopaminergic system, such as variants of genes encoding the dopamine transporter (DAT1, SLC6A3) and dopamine receptor 4 (DRD4). Furthermore, markers on chromosomal regions 4q13.2, 5q33.3, 5p1213, 11q2, 16p13 and 17p11 have shown linkage with the disorder in affected sib-pair studies. Here, we demonstrate that polymorphic marker alleles derived from these loci do not co-segregate with disease in six multiplex ADHD families with a total of 145 family members, 61 of these affected with ADHD. This observation may point to a noncomplex mode of inheritance, and the involvement of other loci in these families.
P83 INVESTIGATIONS INTO A CHROMOSOME 4P BIPOLAR DISORDER SUSCEPTIBILITY GENE; ALLELIC EXPRESSION ASSAYS AND AN ASSOCIATION STUDY IN PI4K2B Lorna Houlihan, David J. Porteous, Douglas H. R. Blackwood, Kathryn L. Evans University of Edinburgh, Edinburgh Genetic studies have identified linkage of bipolar disorder (BP) to chromosome 4p15-16. The 4p15-16 region segregates with illness in a large Scottish family (F22) and three smaller linked families. Haplotype analysis has prioritised two sub-regions, each shared by three of these families. One sub-region includes the phosphatidylinositol 4-kinase type-II beta (PI4K2B) gene. PI4K2B is a priority candidate gene because it is a member of the phosphoinositol (PI) pathway, which is targeted by lithium for therapeutic effect in BP. There are three aspects to our investigation of PI4K2B. Primarily, we are using lymphoblastoid cell lines from F22, which provide a unique resource to examine PI4K2B expression. Allelic contribution to PI4K2B gene expression has been assessed in these F22 cell lines by allele-specific quantitative RT-PCR using Taqman, complimented by Peakpicker technology. PI4K2B protein expression has also been quantified in the cell lines. To date, there is no evidence for altered PI4K2B expression in F22 lymphoblastoid cell lines. However, tagging SNPs from the region containing PI4K2B have been used in case-control association study on 400 bp, schizophrenia cases and control individuals. The association study identified a SNP located in the same haplotype block as PI4K2B, 10 kb upstream of the coding region, as a potentially important variant in SCZ (allele p ¼ 0.006, odds ratios: 1.314, 95% CI 1.08–1.59). Thus, PI4K2B merits further investigation to assess its potential role as a candidate gene in psychiatric illness.
P84 NO ASSOCIATION BETWEEN A 30 bp VNTR IN INTRON 8 OF THE DAT1gene AND ADULT ATTENTION DEFICIT HYPERACTIVITY DISORDER (ADHD) IN A GERMAN SAMPLE Daniel Bru¨ggemann1, Esther Sobanski1, Barbara Alm1, Jens Treutlein1, Markus H. Skowronek1, Hannah Schmalzried1, Alexandra Phillipsen2, Gerome Breen3, Schubert Thomas4, Alexander Georgi1, Monika Deschner1, Marcella Rietschel1 1 Central Institute of Mental Health, Mannheim 2 Department of Psychiatry, University of Freiburg 3 SGDP, IOP, London 4 Zentrum Nervenheilkunde Stadtmitte, Mannheim
757
The dopamine transporter gene (DAT1) is a primary candidate for attention deficit hyperactivity disorder (ADHD). Association studies focused on the ten-repeat allele of a 40 base pair 30 VNTR polymorphism in the 30 UTR of DAT1, yielding conflicting results. A 30 bp VNTR located in intron 8 has recently been suggested as a novel functional site. Its 3-repeat allele was found to be associated with ADHD in samples of English and Taiwanese children. We tested the association between this novel VNTR and ADHD using a German sample of 122 adult ADHD patients and 174 community-based controls. We studied adult ADHD patients because in the families of adults with ADHD, very high rates of ADHD have been reported, suggesting that persistent ADHD is a very useful phenotype for molecular genetic studies. Allele frequencies did not differ between cases and controls. The allele frequencies for the most common allele, the 3-repeat-allele, were 91.4% in cases vs. 94.8% in controls; frequencies for the second most common allele (2-repeat) were 33.6% in cases vs. 33.9% in controls. Likewise, no differential allelic distribution was observed, when cases were stratified according to their clinical subtype (inattentive vs. mixed inattentive-hyperactive). Our results do not support an association of the 30 bp VNTR of Intron 8 with ADHD in adults in the German population. P85 A HIGH THROUGHPUT MUTATION SCREEN OF HUMAN ION CHANNEL GENES IN EPISODIC NEUROLOGICAL DISORDERS Ronald Lafreniere1, J.-F. Poulin1, M. Simoneau1, N. Gupta1, F. Lafreniere1, K. Boisvert1, A. Anton1, S. McLaughlan-Edwards1, M. Albanese1, M.K. Charles1, S. Laurent1, G. Ebers2, Z. Cader3, J. Sequeiros4, J.M. Pereira Monteiro5, G. Turecki6, M. Alda6, P. Grof7, S. Chouinard8, B. Brais8, P. Cossette8, G.A. Rouleau8 1 Emerillon Therapeutics Inc., Montreal, Quebec 2 Wellcome Trust Centre for Human Genetics, Oxford 3 Oxford University, Oxford 4 University of Porto, Porto 5 Hospital Santo Anto´nio, Porto 6 McGill University, Montreal, Quebec 7 University of Ottwaw, Ottawa, Ontario 8 Universite de Montreal, Montreal, Quebec Common brain diseases such as epilepsy and migraine are episodic, showing significant heritability. Mutations in several ion channel genes have been identified as the genetic basis of these disorders in some patients. Additional diseases such as bipolar disorder, essential tremor and Tourette syndrome share some aspects of episodic disorders, and may be caused by dysfunction of ion channel genes. In an effort to identify predisposing genes for these disorders, we have initiated a high throughput mutation detection screen of brainexpressed human ion channel genes. Our goal is to screen 150 genes in 368 DNA samples from unrelated patients with these five disorders. Patient samples were selected based on clear family history and diagnosis, and availability of additional family samples to allow segregation analysis. Samples are screened for DNA variants using denaturing HPLC and DNA sequencing. RFLP analysis or allelespecific oligonucleotide hybridisation is then used to determine allele frequencies. To date, we have screened 140 ion channel genes in our patient cohort and have identified over 1200 variants, of which twothirds are novel (not found in public SNP databases). Missense and frameshift variations are being further validated using: (1) association studies comparing allele frequencies in patients and controls and (2) segregation analysis in extended pedigrees. An interesting candidate gene for bipolar disorder has been identified and will be presented. Identification of genetic factors causing these brain disorders will benefit patients and families through better molecular diagnosis and improved therapeutics. P86 MUTATION SCREENING OF THE SGCE GENE IN OBSESSIVE-COMPULSIVE DISORDER AND TOURETTE SYNDROME Hilga Katerberg1, D. Cath2, M.A.J. Tijssen3, E.M.J. Foncke3, Y. van de Leemput4, J.A. den Boer5, P. Heutink6, F. Baas7 1 VUMC, deparment of Medical Genomics, Amsterdam 2 Department of Psychiatry, VU University, Amsterdam 3 Department of Neurology, Academic Medica, Amsterdam 4 Department of Neurogenetics, Academic Medica, Amsterdam 5 Department of Biological Psychiatry, Uni, Groningen
758
Abstracts
6
Department of Human Genetics, Amsterdam Department of Neurogenetics, Amsterdam
7
Mutations in the epsilon sarcoglycan gene (SGCE gene) are associated with obsessive-compulsive behaviour in some families with MyoclonusDystonia (M- D). Therefore, we screened the SGCE gene for mutations in patients with Obsessive-Compulsive Disorder (OCD) and/or Gilles de la Tourette Syndrome (GTS), a genetically related tic-disorder. We screened the coding region and flanking intronic regions of the SGCE gene for mutations in 88 patients with OCD and/or GTS with a positive family history for tics and/or obsessive-compulsive behaviour and 5 patients with non-familial OCD. No sequence variants were found in the coding region of the SGCE gene, and several common polymorphisms were present in the intronic regions of the SGCE gene. However, we identified three new sequence variants in the 30 UTR of the SGCE gene. These variants are predicted to be within putative microRNA binding sites and could have a biological function. Two of these sequence variants were also present in control chromosomes from the general Dutch population. The third variant was not present in the control chromosomes and segregated with GTS but not with OCD. Future studies to investigate the significance of this variant are warranted. P87 DNA COPY-NUMBER ANALYSIS IN INDIVIDUALS WITH PSYCHOSIS AND CONGENITAL ANOMALIES Angelo Ricciardi1, H. Shaker Naeeni2, V.D.M. McAllister3, P. Papili2, P. McGuffin3, M. Di Forti2, C. Pariante3, P. Dazzan2, R.M. Murray2, K.J. Aitchison3 1 Institute of Psychiatry, Maudsley Hosp., London 2 Division of Psychological Medicine and Psychiatry, London 3 COAST Team, Westways Resource Centre, Cr, London 3 MRC SGDP Centre, Institute of Psychiatry, London 3 Clinical Neuropharmacol., Division of Ps, London Background: Some chromosomal abnormalities (such as velocardiofacial syndrome, VCFS) are associated with congenital physical anomalies and an increased risk of psychotic illnesses (Craddock et al., 2005). Mantripragada et al. (2004) reported that gene copy number profiling would be a reliable technique for investigating VCFS and related conditions. Marfan’s syndrome has been associated with various psychiatric disorders in a large Sardinian family study (Mercuro et al., 1997). Method: We report a 21-year-old man, with Marfanoid features, presenting symptoms consistent with schizophrenia, and a 24-year-old woman with transposition of the great vessels, presenting schizoaffective features. Both were karyotyped and one was also investigated for VCFS using in situ hybridisation (with probe TUPLE1). The former was analysed for Marfan’s syndrome (awaiting results). Results: The karyotyping and the in situ test revealed normal results. However, the test performed did not exclude the possibility of other mutations in that region. We have therefore collected samples for DNA, RNA and proteomic analysis in these individuals, and intend to conduct a comprehensive copy number analysis using CNAT (Affymetrix) and expression studies. Conclusions: These two cases are examples of congenital anomalies and schizophrenia spectrum illnesses. The detailed genetic and expression analysis methodologies that we propose to use here will provide valuable insights regarding the genetic basis for physical and psychiatric comorbidity in these cases, and may even lead to the identification of novel candidates for schizophrenia. Acknowledgements: This study was funded by South London & Maudsley NHS Trust. P88 GENETIC ASSOCIATION AND POSTMORTEM IMMUNOREACTIVITY OF RGS2 IN SUICIDE VICTIMS Huxing Cui1, Naoki Nishiguchi1, Elena Ivleva1, Masaya Yanagi1, Masaaki Fukutake1, Hideyuki Nushida1, Yasuhiro Ueno1, Noboru Kitamura1, Osamu Shirakawa2, Kiyoshi Maeda1 1 Kobe University, Kobe City, Hyogo prefecture 2 Kobe General City Hospital, Kobe City, Hyogo prefecture Objective: Regulator of G–protein signalling (RGS) is known to be a molecule negatively regulating intracellular signalling of G proteincoupled receptors, such as serotonin. Recent studies using knock-out mice or mice behavioural quantitative trait locus methods suggested that the RGS2 played a part in anxiety and/or aggressive behaviour. To
test the possibility that RGS2 gene is involved in suicide through alteration in serotonergic neural transmission, or anxiety and/or aggressive personality traits, we carried out an association study of RGS2 gene polymorphisms in suicide victims. Immunoreactivity in postmortem brain samples was further investigated. Material and Method: Genomic DNA from 189 suicide victims and 210 controls was used in a genetic association study. Both the amygdala and prefrontal cortex of postmortem brains were selected for immunoreactivity investigation. Comparison of genotype distribution and allele frequencies was effected using a chi-square test. The significance of immunoreactivity differences was tested by Mann-Whitney U test. Results: Four SNPs were genotyped. Significant differences were found in genotype distributions of the 2 SNPs and allele frequencies of all four SNPs. RGS2 immunoreactivity significantly increased in the postmortem brains of suicide victims. Conclusions: We found the polymorphisms of RGS2 were associated with suicide, and immunoreactivity significantly increased in the postmortem brains of suicide victims as compared with controls. These findings suggest that the RGS2 gene is involved in biological susceptibility to suicide.
P89 AN ASSOCIATION ANALYSIS OF FUNCTIONAL SEROTONIN TRANSPORTER POLYMORPHISMS IN AUTISM SPECTRUM DISORDER Sarah Curran1, John Powell2, Luca Pugliese2, Ben M. Neale2, Declan Murphy2, Patrick F. Bolton2 1 Institute of Psychiatry, London 2 Institute of Psychiatry, London Serotonin (5-HT), in addition to its role as a neurotransmitter, acts as a trophic factor during brain development. Abnormalities in the serotoninergic systems have been reported in autism spectrum disorders (ASD). Platelet hyperserotonemia has been reported in a proportion of ASD patients and their relatives. Patients may have a decrease in platelet 5- HT2 receptor binding, an increase in the density of platelet 5-HT transporters and antibodies against brain 5-HT receptors. Furthermore, functional MRI (fMRI) also suggests there may be abnormal serotonin synthesis rates in ASD children. Acute dietary depletion of the 5-HT precursor tryptophan is associated with a brief exacerbation of maladaptive behaviour in some people with ASD, and 5-HT reuptake inhibitors give significant benefit in treating repetitive behaviour in people with autism. Thus SLC6A4 may be involved in the aetiology and/or phenotypic expression of the ASD syndrome. The polymorphism most intensely studied is the variable number tandem repeat (VNTR) polymorphism in the promoter region of SLC6A4. It has two alleles, S (short, lesser expressing) and L (long, greater expressing) alleles, though activity may be modified by an A/G SNP within the long allele. A second VNTR in intron 2 can also influence expression of the transporter. We have genotyped 190 subjects and their parents with ASD for these polymorphisms. The Transmission Disequilibrium test is being employed.
P90 SEROTONIN PATHWAY GENES: INFLUENCE OF GENDER, GENOTYPE AND BRAIN REGION ON EXPRESSION Karen Sugden1, Clare Wright2, Nadeem Khan3, Ian W. Craig2, Ursula M. D’Souza2 1 SGDP Centre, Institute of Psychiatry, London 2 SGDP Centre, Institute of Psychiatry, London 3 Department of Neuropathology, Brain Bank, London There is strong evidence to suggest that variation in genes within the serotonergic system is associated with neuropsychiatric disorders. Several studies have focused on understanding the functional effects of specific polymorphisms on expression levels. However, the combination of genotype, gender and brain region could have an overall effect on gene expression. In this study, we report expression of five serotonergic genes (5-HTT, TPH1, TPH2, 5-HT2C, and 5-HT2A) in human postmortem brain tissue using ‘TaqMan’ real-time quantitative PCR. In addition, the effect of genotype, brain region and gender on their expression levels was determined. The sample consisted of several brain regions obtained from 15 individuals post-mortem (9 males, 6 females). These brain tissues included the superior frontal gyrus (SFG), superior temporal gyrus (STG), cerebellum, midbrain, striatum thalamus and hippocampus.
Abstracts Results: mRNA from the five genes was detected in all brain regions tested. 5-HT2C, 5-HT2A and TPH2 showed differential relative expression between brain regions, with an overall high expression in the SFG and STG compared to the cerebellum. TPH1 and TPH2 were similarly expressed in all brain regions; in addition, no significant differences in their levels of expression were detected after controlling for brain region and gender. No significant effect of gender or genotype on expression was seen for any of the genes assayed. These data might help to clarify the role of gender, tissue origin and gene sequence variation in the regulation of serotonergic neurotransmission. P91 ASSOCIATION OF THE b-INTEGRIN SUBUNIT GENE WITH AUTISM Catarina Catarina Correia1, Madalena Martins2, Ana Margarida Coutinho1, Carla Marques2, Teresa Miguel3, Assunc¸a˜o Ataı´de3, Joana Almeida2, Luı´s Borges2, Louise Gallagher3, Judith Conroy3, Michael Gill3, Guiomar Oliveira2, Astrid Moura Vicente2 1 Instituto Gulbenkian de Cieˆncia, Oeiras 2 Instituto Gulbenkian de Cieˆncia/Institut, Oeiras/Lisboa 2 Centro de Desenvolvimento da Crianc¸a, Ho, Coimbra 3 Direcc¸a˜o Regional de Educac¸a˜o da Regia˜o, Coimbra 3 Department of Genetics, Smurfit Institut, Dublin Autism is a complex behavioural disorder probably determined by interaction of multiple genes, still largely unidentified. The results from several genome scans have provided convincing evidence for the presence of an autism susceptibility locus onchromosome 2q31-q33. One of the functional candidate genes that maps in this region is the ITGA4 gene, which encodes the Integrin alpha-4 precursor. In this study, we tested eight single nucleotide polymorphisms (SNP) in the ITGA4 gene for association with autism in a sample of 164 autistic nuclear families. We found significant transmission disequilibrium of alleles at the rs155100 marker (P ¼ 0.019) and of specific haplotypes containing this SNP. It is reported that b-integrins play an important role in the recruitment of circulating activated T-cells, monocytes and macrophages to the central nervous system and in the modulation of the behaviour of microglia. In a previous study, we have shown that an autoreactive response to a specific brain protein occurs significantly more frequently in autistic children than in age-matched controls, which may represent the immune system’s neuroprotective response to a previous brain injury occurring during neurodevelopment. The finding that ITGA4 rs1449263 marker is associated with this specific autoreactivity prompts us to suggest that the association found between autism and the ITGA4 gene is due to a possible involvement of a´4 integrin in this process of neuroinflammation. Our study provides new insights into the role of integrins in autism etiology, to which little attention has been paid thus far. P92 ASSOCIATION ANALYSIS OF POLYMORPHISMS IN SEROTONINE AND DOPAMINE RELATED GENES AND PERSONALITY TRAITS IN BULGARIAN SUBJECTS Mina Ivanova1, Vesselin Chorbov2, Christian Kostov3, Vihra Milanova3, Ivo Kremensky2, Radka Kaneva2 1 Laboratory of Molecular Pathology, Sofia 2 Laboratory of Molecular Pathology, Sofia 3 First Psychiatric Clinic, Sofia A group of healthy Bulgarian subjects (n ¼ 144, females ¼ 62%, males ¼ 38%) was assessed by the Cloninger’s Temperament and Character Inventory (TCI) and genotyped for 7 polymorphisms. We focused on polymorphisms in genes of the serotoninergic and dopaminergic neurotransmitter systems. An association was tested between the TCI performance and 1438G/A polymorphism in the HTR2A gene, Cys23Ser in 5-HT2C gene, two 5-HTT polymorphisms (ins/del 5- HTT-PR and the intron 2 VNTR), the 30 UTR VNTR in DAT1 gene, Val158Met in COMT and 30 VNTR in the MAOA gene. It was found that the low activity met158 variant of COMT is more common among people with low scores on reward dependence (F ¼ 2.595, df ¼ 8, p ¼ 0.012), self directness (F ¼ 2.357, df ¼ 8, p ¼ 0.022) and cooperativeness (F ¼ 3.751, df ¼ 8, p ¼ 0.001). Association between homozygosity for the short MAOA allele and low performance for the TCI subscale cooperativeness was found (B ¼ –11.56 SE ¼ 5.75, p ¼ 0.047). Individuals with 10/10 DAT genotype were more likely to have higher persistence scores than the others (B ¼ 24.16, SE ¼ 8.96, p ¼ 0.008). The
759
presence of HTR2A-pr allele 2 was associated with low Self-Transcendence (B ¼ –19.65 SE ¼ 6.66 p ¼ 0.004). The results of our study delineate some polymorphic variants of enzyme and transporter genes with possible contribution to character and temperament in healthy Bulgarian subjects. P93 ASSOCIATION OF THE NRG1 GENE IN SCHIZOPHRENIA AND BIPOLAR DISORDER IN A SECOND SCOTTISH SAMPLE Kathy Evans, Pippa Thomson, Andrea Christoforou, Ben Pickard, David Porteous, Walter Muir, Douglas Blackwood University of Edinburgh, Edinburgh Neuregulin 1 (NRG1) is a strong candidate for involvement in susceptibility to schizophrenia. A haplotype, initially identified as showing association in the Icelandic and Scottish populations, has shown a consistent effect size in multiple European populations. Additionally, NRG1 has been implicated in susceptibility to bipolar disorder. Here we used haplotype-tagging SNPs to identify single markers and haplotypes associated with schizophrenia and bipolar disorder in an independently ascertained Scottish population. Haplotypes in two regions met an experiment-wide significance threshold of P ¼ 0.0016 and were permuted to correct for multiple testing. Region A overlaps with the Icelandic haplotype and shows nominal association with schizophrenia (P ¼ 0.00032), bipolar disorder (P ¼ 0.0011), and the combined case group (P ¼ 0.0017). This region includes the 50 exon of the NRG1 GGF2 isoform and overlaps the EST cluster Hs.97362. However, no haplotype in Region A remains significant after permutation analysis (P > 0.05). Region B contains a haplotype associated with both schizophrenia (P ¼ 0.00014), and the combined case group (P ¼ 0.000062), although it does not meet Nyholt’s threshold in bipolar disorder alone (P ¼ 0.0022). This haplotype remained significant after permutation analysis in both the schizophrenia (P ¼ 0.024) and combined case groups P ¼ 0.016). It spans a 136 kb region that includes the coding sequence of the SMDF isoform and 30 exons of all other known NRG1 isoforms. Our study identifies a new region involved in schizophrenia and bipolar disorder in the Scottish population. P94 ASSOCIATION OF ALLELIC VARIANTS OF THE DOPAMINE RECEPTOR D2Gene WITH ADHD AND LOW PERSISTENCE TRAIT IN FINNISH POPULATION-BASED BIRTH COHORTS Emma Sofia Nyman1, Anu Loukola2, Teppo Varilo2, Matti Joukamaa3, Irma Moilanen4, James McGough5, Marjo-Riitta Ja¨rvelin6, Susan Smalley5, Leena Peltonen2 1 National Public Health Institute, Helsinki 2 Department of Molecular Medicine, National Public Health Inst., Helsinki 3 Social Psychiatry Unit, Tampere School of Public Health, Tampere 4 Department of Pediatrics, University of Oulu 5 Div. of Child and Adolescent Psychiatry, Los Angeles CA 6 Department of Public Health and General Practice, Oulu The dopaminergic system has been implicated in ADHD and personality traits. Multiple studies have addressed dopamine receptor genes, especially the DRD2 Taq1A-RFLP, with inconclusive results. We investigated the role of DRD2 in ADHD and personality traits in two large Finnish population-based birth cohorts. The ADHD study sample originates from the Northern Finland 1986 Birth Cohort (n ¼ 9432) from the two northernmost provinces of Finland, representing a genetically homogeneous population. Its geographic isolation also contributes to the population being more environmentally homogeneous. ADHD cases (n ¼ 179) and matched controls (n ¼ 158) were ascertained based on specially designed questionnaires and subsequent clinical examinations using DSM-IV criteria. The personality trait study sample emerges from the Northern Finland 1966 Birth Cohort (NFBC66, n ¼ 12058). Allelic variants of DRD2 were monitored using 9 evenly distributed SNPs, including the Taq1A-RFLP. We found an association between ADHD in males and three SNPs in DRD2. Odds ratios for the risk-conferring minor alleles were 1.89 (1.03–3.45, p ¼ 0.039), 2.08 (1.14–3.80, p ¼ 0.017) and 1.93 (1.05–3.55, p ¼ 0.033). Interestingly, the same alleles were also associated with low persistence in NFBC66 (adjusted for sex), with p-values of 0.0267, 0.0091 and 0.0149. Our study provides further evidence for the role of DRD2 in
760
Abstracts
ADHD and the persistence personality trait in two large populationbased cohorts, and suggests a shared biology in ADHD and the low persistence trait.
these analyses, we will also exploit quantitative variables derived from neuropsychological test data and circadian phenotype features that can be considered as intermediate phenotype for bipolar disorder.
P95 POSSIBLE ASSOCIATION BETWEEN GENETIC VARIANTS AT THE GRIN1 GENE AND SCHIZOPHRENIA WITH LIFETIME HISTORY OF DEPRESSIVE SYMPTOMS IN A GERMAN SAMPLE Alexander Georgi1, Rami Abou Jamra2, Johannes Schumacher2, Tim Becker3, Christine Schma¨l1, Monika Deschner1, Torsten Paul1, Susanne Ho¨fels2, Maren Wulff3, Markus Schwarz3, Norman Klopp4, Thomas Illig4, Peter Propping5, Sven Cichon6, Markus M. No¨then6, Thomas G. Schulze1, Marcella Rietschel1 1 Central Institute for Mental Health, Mannheim 2 Institute of Human Genetics, University, Bonn 3 IMBIE, University of Bonn, Bonn 2 Department of Psychiatry, University of Bonn 3 Nordbaden Center of Psychiatry, Wiesloch 4 Institute of Epidemiology, GSF, Neuherberg 5 Institute of Human Genetics, University, Bonn 6 Life & Brain Center, University of Bonn, Bonn
P97 ASSOCIATION BETWEEN THE PERSONALITY DIMENSION ‘‘NOVELTY SEEKING’’ AND VARIANTS AT THE G72/G30 LOCUS IN A LARGE POPULATION-BASED SAMPLE FROM GERMANY Jana Strohmaier1, Susanne Hoefels2, Anna Karpushova3, Johannes Schumacher4, Rami Abou Jamra4, Wolfgang Maier2, Peter Propping4, Markus M. Noethen3, Sven Cichon3, Thomas G. Schulze1, Marcella Rietschel1 1 Central Institute for Mental Health, Mannheim 2 Department of Psychiatry, Bonn 3 Life & Brain Center, Bonn 4 Institute of Human Genetics, Bonn
Disturbed glutamatergic neurotransmission has been implicated in the pathogenesis of schizophrenia (SCZ). The NR1 subunit of N-methy-Daspartate receptors (NMDARs) is encoded by the gene GRIN1 on chromosome 9q34.3, a linkage region for SCZ. Controversy surrounds a potential association between SCZ and GRIN1. We present a casecontrol study on GRIN1 in a German sample. In 354 SCZ patients (44% female, 56% male) and 323 population-based controls (58% female, 42% male), we genotyped a microsatellite (position 137303343, build Nov 2002) and the SNPs rs4880213, rs11146020, rs6293 and rs10747050 (called Marker 1 to 5). Previous genetic studies in SCZ suggested that SCZ cases with a history of major depressive episodes (MDE) outside of psychotic episodes delineate a genetically distinct subgroup. Therefore, we also analysed the subgroup of 87 patients with MDE. There was no differential distribution of alleles or haplotypes between the cases and controls. The comparison between the subgroup of SCZ patients with MDE and controls showed several significant associations, e.g. a global p of 0.0113 for the four-marker-haplotype 1-2-3-4. When comparing the MDE subgroup with the remaining SCZ patients, this association still held. However, after Bonferroni-correction for multiple testing owing to the sliding-window approach, our results were lacking in significance. Although our exploratory study revealed no associations holding up to correction, our results suggest that GRIN1 be considered a candidate for future investigations. These should consider stratifying SZC cases on the presence of MDE. P96 FAMILY BASED STUDY OF 13 FUNCTIONAL CANDIDATE GENES AND BIPOLAR DISORDER Pia Soronen, Outi M. Palo, Kaisa Silander, Mervi Antila, Annamari Tuulio-Henriksson, Tuula Kieseppa¨, Leena Peltonen, Timo Partonen, Jouko Lo¨nnqvist, Tiina Paunio National Public Health Institute, Helsinki Bipolar disorder (BPD) is a severe mood disorder with a strong genetic background. We studied 84 single nucleotide polymorphisms (SNPs) of 13 functional or positional candidate genes in a Finnish bipolar family sample. The genotyping was performed by a mass spectrometry based method (MassARRAYTM, Sequenom), and the results were analysed by Pseudomarker using the recessive model. In the analysis of 63 families (350 individuals, 116 affected with BPDI, 36 with SA-BPD, BPD II or MDD), several variants of the glycogen synthase kinase 3 beta gene (GSK3B) showed suggestive evidence of linkage and association (p ¼ 0.01–0.05 from two-point analysis). Moreover, receptors for brain derived neurotrophic factor NTRK2 showed suggestive association (p ¼ 0.01). In females, some evidence for association was obtained for the D-amino acid oxidase activator (DAOA) (p ¼ 0.001). Using the psychosis status, a suggestive association was obtained with SNPs of the genes encoding for adrenergic beta, receptor kinase 2 (ADRBK2; phosphorylates G protein coupled receptors) (p ¼ 0.0008) and brain-specific form of a rate limiting enzyme of serotonin synthesis tryptophan hydroxylase 2 (TPH2) (p ¼ 0.02). GSK3B is highly interesting gene because it regulates circadian rhythm and lithium is a potent inhibitor of GSK3B. We are currently genotyping the selected variants in the complete sample of 161 families (715 individuals). In
Introduction: G72 has previously found to be associated with schizophrenia, bipolar disorder -in particular with persecutory delusions-, and panic disorder. The gene variants we discovered to be associated with these disorders occur at a high frequency in the general population. Genotype-phenotype correlation studies in the general population of disease-associated personality traits could assist in unraveling the mechanisms which lead to disturbed functioning in the above mentioned disorders. Methods: We assessed a population-based sample of 1105 German individuals using the Tridimensional Personality Questionnaire (TPQ). Individuals were genotyped for the G72/G30-markers M23 and M24. Logistic regression analyses were performed, with genotypes as dependent variables and the scales ‘‘harm avoidance’’ (HA), ‘‘novelty seeking’’ (NS), ‘‘reward dependence’’ (RD), and the respective subscales as independent variables. Results: Individuals with genotypes carrying the respective previously identified risk alleles, i.e. ‘‘C’’ (M23) and ‘‘T’’ (M24) had significantly higher scores on NS (p ¼ .031 and p ¼ .013) and its subscale ‘‘exploratory excitability vs. stoic rigidity’’ (p ¼ .048 and p ¼ .046) than non-carriers of the alleles. There were no associations with HA, RD, and other subscales. Discussion: We found an association between G72 disease alleles and exploratory excitability which characterizes individuals who are extremely responding to environmental stimuli. Due to the observation that G72 activates DAAO which acts on the NMDA receptors it may be hypothesized that G72 disease alleles contribute to increased stimuli response (vs stoic rigidity) a feature which is common to the above mentioned disorders. P98 COMT CONTRIBUTES TO GENETIC SUSCEPTIBILITY SHARED BY ANXIETY SPECTRUM PHENOTYPES John Hettema1, Seon-Sook An2, Michael Neale2, Edwin van den Oord2, Kenneth Kendler2, Xiangning Chen2 1 Virginia Commonwealth University, Richmond, Virginia 2 Virginia Commonwealth University, Richmond, VA Catechol-O-methyl transferase (COMT) has been investigated for its possible role in a wide range of psychiatric phenotypes. In particular, several studies support association of this gene with panic disorder and, possibly, other anxiety-related traits. In this study, we examined the COMT gene for association with genetic risk across a range of anxiety spectrum phenotypes. We used multivariate structural equation modeling to select twin pairs that scored at the extremes of a latent genetic risk factor shared by neuroticism, several anxiety disorders, and major depression from the Virginia Adult Twin Study of Psychiatric and Substance Use Disorders. Using one member from each of these pairs, the resulting sample of 589 cases and 539 controls were entered into a 2- stage association study in which genetic markers were screened in stage 1, the positive results of which were tested for replication in stage 2. The functional val158met polymorphism (rs4680) plus nine other SNPs selected to capture the major allelic variation across the COMT locus were analyzed for differences between cases and controls. Two markers, including val158met, showed criterion differences between cases and controls in the stage 1 sample. We are genotyping, in our stage 2 sample, these markers plus two others that make up significant haplotypes. Our preliminary results suggest replication of some of these associations in stage 2. Consistent with prior studies, our findings appear to be female specific. Variations
Abstracts in the COMT gene increase susceptibility across a range of anxiety disorders, major depression, and neuroticism.
P99 TRICHOTILLOMANIA IS ASSOCIATED WITH MUTATIONS IN SLITRK1 Allison Ashley-Koch1, Michael Cuccaro2, Khanh Nhat Tran-Viet2, Heidi Cope2, Ranga Krishnan2, Margaret Pericak-Vance2, Harry Wright3, Stephan Zuchner2 1 Duke University, Durham, NC 2 Duke University Medical Center, Durham, NC 3 University of South Carolina, Columbia, SC Trichotillomania (TTM) is a chronic behavioural disorder characterized by the irresistible urge to pull out one’s hair, resulting in noticeable hair loss. Clinically, TTM is classified as an impulse-control disorder and believed to belong to the obsessive-compulsive (OCD) spectrum. Tourette’s Syndrome (TS), also belonging to the OCD spectrum, was recently found to be caused by mutations in the slit and trk like 1 (SLITRK1) gene. Abelson and colleagues (2005) identified a frame-shift mutation and a 30 -UTR micro-RNA binding site variation in three independent TS families. The mother of one proband carried the SLITRK1 mutation and was diagnosed with TTM. Therefore, we hypothesized that sequence variation in SLITRK1 may also contribute to TTM not associated with TS. We identified two novel nonsynonymous sequence changes that occurred in two out of 44 independent TTM subjects of European descent. Both changes occurred within a short distance between the second LRRCT domain and the transmembrane domain of SLITRK1 and co-segregate with psychiatric conditions in their respective families. We did not detect the altered genotypes in a sample of 2192 non-TTM Caucasian controls. Our results confirm that rare variations in SLITRK1 result in disorders of the OCD spectrum, and challenge the prevailing belief that psychiatric conditions are caused only by common genetic variants. Furthermore, these data provide the prospect of genetic testing for OCD spectrum disorders which could have substantial consequences for the patients and their families. P100 A CASE-CONTROL ASSOCIATION STUDY ON THE NPAS3 GENE Benjamin Pickard, Andrea Christoforou, Philippa Thomson, Kathryn Evans, Stewart Morris, David Porteous, Douglas Blackwood, Walter Muir University of Edinburgh, Edinburgh The NPAS3 gene encodes a transcription factor expressed in the brain. Two lines of evidence suggest it is a good candidate gene for psychiatric illness. Firstly, we have documented a chromosomal abnormality in a mother and daughter with schizophrenia and learning disability that directly disrupts this gene. Secondly, a mouse Npas3 knockout displays a behavioural phenotype characteristic of mouse models of schizophrenia. In addition, these mice have a deficit in adult hippocampal neurogenesis that is relevant to the morphological changes documented in post mortem studies of schizophrenic brains. We have now gone on to assess the role that NPAS3 variation may have in altering risk of psychiatric illness in the wider population. We have carried out a casecontrol association study involving 68 tagging SNPs across the gene, genotyped in 368 bipolar disorder cases and 386 schizophrenia cases comprised of patients from hospitals in southeast and south central Scotland, and 458 controls of matching geographical distribution. Two significant markers, rs729217 and rs173383, were identified associated with an increased risk of bipolar disorder. Further SNPs are currently being typed to expand the haplotype definition in these regions in the light of recent HapMap data. The results from these studies will be presented and discussed in the context of future resequencing strategies. P101 HIGH THROUGHPUT STRATEGY FOR THE IDENTIFICATION AND CHARACTERIZATION OF GENES INVOLVED IN SCHIZOPHRENIA AND AUTISM Julie Gauthier1, Pierre Drapeau2, Jacques Michaud3, Christian Neri4, Marie-Pierre Dube3, Patrick Cossette3, Michel Bouvier3, Savatore Carbonetto4, Eric Fombonne4,
3
3
761
4
Beatrice Godard , Sylvie Perrault , Pejmun Haghighi , Ridha Joober4, Guy A. Rouleau3 1 CHUM Research Centre, Montreal 2 Universite´ de Montre´al, Montreal, QC 3 Universite de Montreal, Montreal, QC 4 INSERM, Paris 4 McGill University, Montreal, QC Schizophrenia (SCZ) and autism (AUT) are common, devastating and poorly-treated mental disorders. Converging evidence suggests that genetically-disrupted synaptogenesis and synaptic plasticity during development may underlie the pathogenesis of such brain disorders. We therefore hypothesize that a significant fraction of SCZ and AUT cases are a result of frequent novel mutations in many different genes involved in synapse formation and function. Our strategy is to identify the underlying synaptic genetic factors predisposing to SCZ and AUT by using a new, two-step strategy: first the direct re-sequencing of 1000 genes coding for proteins which regulate an entire brain-specific ‘‘machine’’, the synapse, in a large number of patients, and secondly, evaluating the biological and functional effects of disease-related variants of these genes in several animal models. This will allow us to identify the most promising mutations for future functional genomic analyses, including transgenic mouse models. In contrast to the all-toofew mutations discovered to date, we expect to identify a significant number of genes directly causing or increasing susceptibility to SCZ and AUT. P102 ASSOCIATION BETWEEN COMT, HTR1B, AND SUICIDE Brett Kia-Keating1, Stephen J. Glatt2, Ming T. Tsuang2 1 University of California, San Diego, San Diego, CA 2 University of California, San Diego, CA Several studies have suggested that suicidal behaviour is at least partially determined by genetic factors, supporting a search for related genes. Two genes that have been highlighted in the literature as being involved are HTR1B, which codes for serotonin, and COMT, which is related to the inactivation of catecholamines. In order to assess the combined evidence for the numerous studies that have been published relating suicidal behaviour to these two genes, two different random effects meta-analyses were conducted utilizing seven studies related to HTR1B and six studies related to COMT. For HTR1B, the combined evidence from 1,051 case and 1,243 control subjects/participants, suggested that there was no significant association between the HTR1B G861C polymorphism and suicidal behaviour. For COMT, based on the combined data from 624 cases and 935 control subjects, there was evidence of a significant association between the COMT 158Met polymorphism and suicidal behaviour. Although the results for COMT were not influenced by publication bias, the significance of the combined results was not robust and was related to the gender of the case and control subjects. The results for COMT support past literature that has found a relationship between suicidal behaviour and COMT, and has also found that the relationship differs for males and females. Given that in the United States a higher proportion of females attempt suicide, and the higher proportion of males successfully commit suicide, it is possible that COMT is related to the lethality of suicide attempts. P103 CYP2D6 AND CYP2C19 ASSOCIATION DATA FROM GENDEP, A MULTICENTRE EUROPEAN STUDY Patricia Huezo-Diaz1, R. Williamson1, M.W. Nash1, S.C. Sim2, S. Alhberg2, F. Hoda1, M. Rietschel2, C. Schmal2, A. Farmer1, T. Schulze2, J. Hauser3, N. Henigsberg4, W. Maier5, A. Zobel5, E.R. Larsen6, O. Mors6, J. Mendelewicz7, A. Marusic8, J. Perez9, L. Schalkwyk1, P. McGuffin1, I. Craig1, M. Ingelman-Sundberg2, K.J. Aitchison1 1 Institute of Psychiatry, London 2 Karolinska Institute, Stockholm 2 CIMH, Mannheim 3 Academic Department of Medicine, Poznan 4 Croatian Institute for Brain Research, Zagreb 5 University of Bonn, Bonn 6 Department of Psychiatry, Aarhus 7 Free University Hospital, Brussels 8 Institute of Public Health, Ljubljana 9 Ordine Ospedaliero di San Giovanni di Di, Brescia
762
Abstracts
Background: GENDEP is a European multicentre pharmacogenomics study focussed on the prediction of therapeutic response to antidepressants and adverse effects using two antidepressants (escitalopram and nortriptyline). CYP2C19 and CYP2D6 are two enzymes responsible for the metabolism of many antidepressants. There is strong evidence for the role of CYP2C19 in the metabolism of escitalopram, and for CYP2D6 in the metabolism of nortriptyline, thus making the genes for these enzymes eminently suitable candidates for study within GENDEP. Both these genes code for enzymes with varying degree of activity within the population. Each gene has many variants that in combination allow for the prediction of the likely enzymatic activity of CYP2D6 and CYP2C19 (poor metabolisers up to ultra-rapid metabolisers). Hypothesis: The response and/or tolerance of patients to the study medications will be associated with CYP2D6 and CYP2C19 enzyme activity as predicted by genotype. Methods: DNA has been collected from 371 Caucasian subjects ( > 18 years). Subjects were randomly allocated to escitalopram and nortriptyline treatment with serial clinical ratings being performed. First wave genetic association analysis included genotyping relevant SNPs on both the CYP2D6 and CYP2C19 genes. Genotyping was conducted using the AmpliChip CYP450 Test and ABI SNPlex technology. Results: There is evidence of association between CYP2C19 metabolic enzyme activity and escitalopram treatment in this sample. Preliminary analysis showed an association between CYP2D6 metabolic enzyme activity drop-out status. Conclusion: Our study supports the hypothesis that CYP2C19 and CYP2D6 enzyme activity may be related to response to the antidepressants in our study.
P104 NEUREGULIN POLYMORPHISM ASSOCIATES WITH DENSITY OF THE VESICULAR GLUTAMATE TRANSPORTER 1 IN HUMAN STRIATUM AND HIPPOCAMPUS Gavin Reynolds1, Pritsana Piyabhan1, Emma Dempster2, David Collier2 1 Queen’s University Belfast, Belfast 2 The Institute of Psychiatry, London There is substantial evidence for glutamatergic dysfunction in schizophrenia. This includes disturbances of the cortico-striatal pathway and hippocampal innervation, reflected by the deficits of the vesicular glutamate transporter 1 (VgluT1) we have identified in striatal and hippocampal regions. Several of the genes implicated in schizophrenia, including neuregulin and dysbindin, may influence function of glutamatergic synapses. We tested whether the neuregulin polymorphism providing the greatest contribution to risk in an Icelandic cohort might be associated with this indicator of glutamatergic dysfunction in human post-mortem tissue. Using tissue from the Stanley Neuropathology Consortium, we have determined density of VgluT1 in striatal and hippocampal regions by immunohistochemical staining and investigated the association of VgluT1 density with the SNP8NRG221533 neuregulin polymorphism. Significant association with neuregulin genotype was found for VgluT1 density in putamen and in the dentate gyrus of the hippocampus in which subjects homozygous for the high schizophrenia risk C allele demonstrated lower densities of this glutamatergic marker. Reanalysis of CC homozygotes vs. the remaining sample generally increased significances, indicating a further significant association with VgluT1 in the nucleus accumbens. These findings demonstrate that genotype of a genetic risk factor for schizophrenia is associated with the density of a marker of cortico-striatalglutamatergic innervation and hippocampal glutamatergic terminals in which there are deficits in the disease. This suggests that these glutamatergic deficits may, in part, be genetically determined, either from early development or by the at-risk genotype imparting an increased vulnerability to subsequent damage of glutamatergic terminals. P105 PATHOLOGICAL GAMBLING: ANALYSIS OF DOPAMINE GENES IN DISCORDANT SIB-PAIRS Daniela Sabbatini Lobo1, Homero Vallada2, Joanne Knight3, Silvia Martins4, Hermano Tavares2, Valentim Gentil2, James Kennedy3
1
Centre for Addiction and Mental Health, Toronto, ON Department and Institute of Psychiatry U, Sao Paulo, SP Social and Genetics Developmental Psychiatry, London 4 Department of Mental Health, Johns Hopkins University, Baltimore, MD 3 Neurogenetics Section at the Centre for, Toronto, ON 2 3
Molecular genetic studies have associated polymorphisms in dopamine receptor genes to pathological gambling (PG). Using a family-based association study design we investigated the relationship of functional polymorphisms in dopamine receptor genes (DRD1 800T/C, DRD2 TaqAI, DRD3 ser9gly; DRD4 exon III VNTR, and DRD5 (CA)n) with PG. Method: We evaluated 140 patients (mean age SD ¼ 40 8.9 years-old; 70 males, 70 females) who fulfilled DSM-IV criteria for PG and 140 non-gambler siblings (mean age SD ¼ 40 9.8 years-old; 70 males, 70 females). Pathological gamblers were excluded if they met lifetime diagnostic criteria for bipolar disorder. Lifetime history of psychosis or neurological diseases were considered as exclusion criteria for both groups. One functional polymorphism on each of the dopamine receptor genes was chosen for analysis to reduce multiple testing. Pedigree disequilibrium tests were performed and 2-tailed significance level of p < 0.05 was adopted. Results: An association of PG and allele 1 of DRD1 - 800T/C was found (p ¼ 0.03). None of the other polymorphisms was associated to PG. Conclusion: These findings could be regarded as further evidence of DRD1 involvement in PG. PG has been considered as a behavioural addiction. Investigation on the functional aspects of the D1 receptor has focused on its involvement in cue-induced craving, which plays an important role in the treatment of PG. We are currently genotyping other dopamine receptor genes polymorphisms to further confirm these findings. P106 PRION PROTEIN HAPLOTYPES IN SUBJECTS WITH ALZHEIMER’S DISEASE AND FRONTOTEMPORAL LOBAR DEGENERATION Bernd Ibach1, Klaus Schmidtke2, Christopher Pratt1, Tatjana Nonenmacher1, Go¨ran Hajak1 1 University of Regensburg, Regensburg 2 University of Freiburg, Freiburg The cellular prion protein (PRNP) is a cell surface glycoprotein mostly expressed in neurons and has been implicated in the susceptibility to familial spongiform encephalopathies. Whether the prion protein confers vulnerability to other neurodegenerative disorders is a matter of ongoing debate. The present study addresses three PRNP gene variants in a case-control investigation of Alzheimer’s disease and dementias with frontotemporal lobar degeneration. Methods: 327 subjects were investigated, including 64 subjects with a diagnosis of probable Alzheimer’s disease (NINCDS-ADRDA), 156 subjects with a consensus diagnosis of dementia associated with frontotemporal lobar degeneration (Neary et al., 1998) (frontotemporal dementia, semantic dementia and primary progressive aphasia) and 107 healthy, age-matched controls. We tested for an association of neurodegenerative diseases with a PRNP missense variant (M129V), a silent exonic variant (rs8124241), and an intronic variant (rs13045348). Results: All genotype data complied with the Hardy-Weinberg equilibrium. No single polymorphism under study, however, gave genotype or allele distributions neither in Alzheimer’s disease, nor for dementias with frontotemporal lobar degeneration, that differed significantly from the respective distributions in healthy controls (p ¼ .09). Conclusions: At the present marker density, this is the first investigation of the PRNP gene in both Alzheimer’s disease and in dementias with frontotemporal lobar degeneration. Our data argue against a major role of the gene in these neurodegenerative diseases but do not entirely rule out minor effects. Future studies will need to control for possible risk modifiers, e.g. age at onset, or APOE status. P107 A FIRST EPISODE PSYCHOSIS CASE-CONTROL GENETIC ASSOCIATION STUDY Katherine Aitchison1, P. Chow2, M. Di Forti3, L .Curtis2, M.J. Arranz3, R.W. Williamson2, D. Gayzina2, S. Cohen2,
Abstracts 2
2
3
3
P. Huezo-Diaz , F. Hoda , A. Butt , C. La Cascia , A. Miorelli3, V. Mondelli3, S. Navari3, J. Eyeson3, D. Clifford4, R. Gafoor3, C. Morgan3, P. Papili3, A. Ricciardi3, H. Shaker2, F. Saardzadeh-Sardahaee2, V. Mantua2, P. Fearon3, P. Jones4, I. Craig2, C. Pariante3, P. Dazzan3, J. Powell4, D.A. Collier2, P. McGuffin2, R.M. Murray3 1 Institute of Psychiatry, London 2 Institute of Psychiatry, MRC SGDP Centre, London 3 Institute of Psychiatry, Division of Psychological Medicine, London 4 South London and Maudsley NHS Trust, London 4 University of Cambridge, Cambridge 4 Institute of Psychiatry, Department of Neuroscience, London Background: GAP (genes and psychosis) is a case-control study of first episode psychosis conducted in London and Cambridge, which aims to identify genes conferring susceptibility to psychosis, and associated phenotypes including cognitive dysfunction and cerebral morphology. Methods: First episode psychosis cases have been recruited in South London and Maudsley NHS Trust and in Cambridge. A variety of demographic and clinical data have been collected. In a subset of these, neurocognitive assessments and MRIs have been performed. Samples have been taken for DNA, and in a subset for RNA and proteomic analysis. Genetic association analysis is being undertaken using a candidate gene approach. The genes chosen for the first wave of analysis include the current most promising candidates for susceptibility to psychosis (NRG1, dysbindin, DISC1, G72, etc) as well as candidates for susceptibility to cannabis misuse (COMT), cognitive dysfunction, dysregulation of brain morphology or susceptibility to bipolar disorder (e.g. LIS1), and candidates in the dopamine and serotonin neurotransmitter systems. Results: DNA has been collected from 302 patients to date. Of these, 72% are male, and the mean age is 25 years; 187 are Caucasian; 115 are of black origin; and the rest are of other or mixed ethnicity. Genotyping is being undertaken in this sample and in matched controls. Conclusions: Data is being reported separately for a number of phenotypes for which there is already some data. This presentation will report the overall genetic association results. Acknowledgements: Funded by South London and Maudsley NHS Trustees.
P108 COMT GENE POLYMORPHISM (VAL158MET) IS ASSOCIATED WITH VERBAL WORKING MEMORY IN A HEALTHY POPULATION Mari Aguilera1, Neus Barrantes-Vidal2, Ba´rbara Arias3, Jorge Moya4, ´ ngeles Ruizpe´rez4, Helena Villa4, Ignacio Iba´n˜ez4, Maria A Genero´s Ortet4, Lourdes Fan˜ana´s3 1 University of Barcelona, Barcelona 2 Dep. de Psicologia Clı´nica i de la Salut, Barcelona 3 Unitat d’Antropologia, Departament Biolo, Barcelona 4 Dept. de Psicologia Ba`sica, Clı´nica i Ps, Castello´n Catechol-O-Methyltransferase (COMT) is the main enzyme of DA in the prefrontal cortex (Tunbridge et al., 2006). High activity Val allele (Val158Met polymorphism on COMT gene –cr 22q11-) results in lower levels of available DA and, coherently, has been related with inefficient prefrontal-dependent executive tasks, mostly to WCST performance (Egan et al., 2001; Rosa et al., 2004). However, few studies have analysed whether COMT genotypes also relate to working memory (WM) (Bruder et al., 2005). This study analysed whether VAL carriers performed worse on verbal working memory (VWM). The sample consisted of 351 Spanish healthy subjects [(mean age ¼ 24 (5.9), range 18–55; education ¼ 13.9 (2.4); 166 men]. VWM was assessed by WAISIII Letter and Number subtest. The Val158Met genotypes were determined using the Taqman 50 exonuclease assay. No statistical differences were detected when comparing genotype frequencies by age, sex, education or IQ. An ANOVA comparing the genotype frequencies yielded a trend for significance being the Val/Val individuals who score lower in VWM (F ¼ 2.54, p ¼ 0.08). In fact, when grouping by genotype (Val/Val vs Met/–), we found that Val/Val homozygous subjects have poorer performance on VWM compared to Met carriers (F ¼ 4.99; p ¼ 0.026). This finding is consistent with previous results showing the influence of COMT genotype on the prefrontally mediated functions such as WM. No sociodemographic or general cognitive ability differences can account for this result. PND, Ministerio de Sanidad-Consumo: 05I317/Fundacio´ SENY.
763
P109 VARIANT BRAIN-DERIVED NEUROTROPHIC FACTOR IN CHRONIC TINNITUS Philipp Sand, Berthold Langguth, Tobias Kleinjung, Birgit Fischer, Go¨ran Hajak, Christopher Pratt, Sandra Pfluegl, Helene Niebling, Peter Eichhammer University of Regensburg, Regensburg Objectives: Neurotrophic factors play key roles in the developing auditory pathway including the sensory epithelium of the inner ear, and structures involved in the central nervous processing of auditory stimuli. In the present investigation, we explored a possible implication of variant BDNF in the susceptibility to chronic tinnitus. Methods: 222 subjects complaining of chronic tinnitus were recruited from a tinnitus clinic and underwent detailed neuro-otological examinations including otoscopy, stapedius reflexes, middle ear pressure measurements, pure tone audiometry, tinnitus pitch and loudness matches. Subjects were genotyped for a biallelic BDNF missense variant (Val66Met). Prevalence of the substitution was compared to the prevalence in an ethnically homogenous group of healthy controls (N ¼ 317). Results: Carriers of the Met variant were significantly less likely to develop chronic tinnitus with comorbid hearing impairment (p ¼ .02, OR ¼ 1.62, 95% CI ¼ 1.1–2.5). When no assumptions of dominance were made for the minor allele, the Met allele still conferred protection against tinnitus with hearing impairment (p ¼ .05, OR ¼ 1.42, 95% CI ¼ 1.0–2.0). Conclusions: The present study suggests a role of variant BDNF in modulating the genetic susceptibility to chronic tinnitus with hearing impairment. Possible implications of this finding include a differential response to the pharmacological treatment of tinnitus, and specifically, to the neurotrophic effects of antidepressants. P110 CONTRIBUTION OF SERT AND BDNF GENES ON NEUROTICISM IN HEALTHY GENERAL POPULATION Mari Aguilera1, Barbara Arias2, Neus Barrantes-Vidal3, Helena Villa4, Jorge Moya4, Iban˜ez Ignacio4, M. Angeles Ruizperez4, Genero´s Ortet4, Lourdes Fan˜ana´s2 1 University of Barcelona, Barcelona 2 Un. Antropologia, Dp Bio Animal, Fac Bio, Barcelona 3 Dep de Psicologia Clı´nica i de la Salut, Barcelona 4 Dep de Psicologia Ba`sica, Clı´nica i Psic, Castello´n Serotonergic neurotransmission is a major modulator of emotional behaviour. In this sense, recent evidence shows the involvement of SERT and BDNF genes in personality dimensions such as Neuroticism (N) or Harm Avoidance (HA) (Ebstein, 2006 for review). The aim of our study was to analyse the role of these genes in relationship to personality dimensions measured in healthy general population. Sample: 335 healthy Spanish subjects [152 men; mean age ¼ 23.5 (5.3), range 18– 55]. Personality dimensions were assessed with the Big Five Questionnaire (BFQ) and Temperament and Character Inventory (TCI). The 5-HTTLPR polymorphism was determined using standard PCR amplification and agarose gel and Val66Met genotypes using the Taqman 50 exonuclease assay. Variability at the BDNF gene was not associated to personality dimensions. However, we found statistical differences in the distribution of N scores (BFQ) according to SERT genotypes in the group of males: L/L homozygotes have higher scores on N than S carriers (F ¼ 4.8; p ¼ 0.009). No significant association was found for HA (TCI). These results are consistent with previous studies carried out by Brummett et al. (2003) or Flory et al. (1999). Our data suggest a contribution of SERT genetic variability in N but not in HA in men. A close relationship between N and anxiety-depression traits (with a clear serotonergic basis) has been suggested, that may explain the discordant results found for both dimensions in relation to SERT gene. Acknowledgements: Projects SEJ2005-09307, SAF 2005-07852C02-01 P111 INVESTIGATION OF THE 5-HTTLPR IN CHRONIC TINNITUS Philipp Sand, Berthold Langguth, Tobias Kleinjung, Birgit Fischer, Petra Stoertebecker, Sandra Pfluegl, Helene Niebling, Go¨ran Hajak, Peter Eichhammer University of Regensburg, Regensburg Serotonin plays a key role in the processing of acoustic information in the auditory pathway, and is known to affect tone-evoked responses in a
764
Abstracts
frequency-dependent fashion. The present study addressed serotonin reuptake, as determined by a common length polymorphism in the serotonin transporter gene (5-HTT), with regard to tinnitus susceptibility. Tinnitus is a highly prevalent condition marked by the perception of sound in the absence of external stimuli. 186 subjects who complained of primary tinnitus lasting for a minimum of 6 months, were recruited from a tinnitus clinic and were included in an association study of the 5-HTT promoter length polymorphic region (5-HTTLPR). Genotype and allele distributions were compared to data obtained from a control group of 256 healthy individuals. When contingency tables were calculated, carriers of the short, low-expressing 5-HTTLPR allele did not differ between the two populations under investigation (p > 0.86, OR ¼ 0.97, 95% CI ¼ 0.64–1.46). Contrary to our conjecture, the 5-HTT variant thus does not per se appear to confer an increased risk of tinnitus. It cannot be excluded, however, that the variant under investigation is involved only in the etiology of specific forms of chronic tinnitus, e.g. in subjects suffering from comorbid affective disorders. P112 HTR2C, HTR1A, NOS-I AND –III GENE VARIANTS IN SUICIDE ATTEMPTERS AND COMPLETERS Laura Mandelli2, Dan Rujescu1, Ina Giegling1, Barbara Schneider3, Annette M. Hartmann1, Axel Schnabel4, Konrad Maurer3, Hans-Ju¨rgen Mo¨ller1, Diana De Ronchi2, Alessandro Serretti2 1 Department of Psychiatry, Ludwig Maximilians University, Munich, Germany 2 Institute of Psychiatry, University of Bologna, Bologna, Italy 3 Department of Psychiatry, Psychosomatics, and Psychotherapy, Johann Wolfgang Goethe-University, Frankfurt/Main, Germany 4 Institute of Forensic Medicine, Johann Wolfgang Goethe-University, Frankfurt/Main, Germany. Nitric oxide (NO) is a neurotransmitter which modulates depressive and aggressive behaviours. NOS exists in three forms, termed NOS-I, NOS-II and NOS-III. The serotonin 2C (HTR2C) and 1A (HTR1A) receptors have been also involved in suicide-related behaviours. We studied gene variants in 259 suicide patients and 312 German healthy subjects. The analyses in relation to suicidal behaviours did not reveal any significant association for HTR2C and HTR1A. On the other hand, a NOS-I risk haplotype C-G-G (rs2682826 / rs1353939 / rs693534) was associated with suicidal behaviour (p ¼ 0.01). Sliding windows analysis showed similar results for the haplotype G-G (rs1353939 / rs693534) (p ¼ 0.01). Additionally, the G-allele of rs693534 was associated with suicidal behaviour (p ¼ 0.005). Interestingly, the same haplotypes (CG-G and G-G) as well as the rs693534 G allele were also associated with increased aggression as measured by the Questionnaire for Measuring Factors of Aggression (FAF). Regarding NOS-III, a protective haplotype C-T-A (rs2070744 / rs1799983 / rs891512) was observed (p ¼ 0.01) with a pronounced effect against suicide completion (p ¼ 0.005). Sliding window analysis showed the same effect of haplotype T-A (rs1799983 / rs891512) (p ¼ 0.01) which was again protective against completed suicide (p ¼ 0.006). Single marker analysis showed the same protective effect of the rs891512 A-allele (p ¼ 0.009) again especially against suicide completion (p ¼ 0.007). Additionally, a second haplotype (T-T-G) was associated with increased aggression (p ¼ 0.0002; sliding haplotype T-G, p ¼ 0.002). In conclusion, our study suggests a possible involvement of NOS-I and NOS-III gene variants in suicidal behaviour and related intermediate phenotypes.
P113 5-HT2A RS643627 AND RS594242 POLYMORPHISMS AND PERSONALITY TRAITS Raffaella Calati2, Dan Rujescu1, Ina Giegling1, Laura Mandelli2, Barbara Schneider3, Annette M. Hartmann1, Axel Schnabel4, Konrad Maurer3, Hans-Ju¨rgen Mo¨ller1, Diana De Ronchi2, Alessandro Serretti2 1 Department of Psychiatry, Ludwig Maximilians University, Munich, Germany 2 Institute of Psychiatry, University of Bologna, Bologna, Italy 3 Department of Psychiatry, Psychosomatics, and Psychotherapy, Johann Wolfgang Goethe-University, Frankfurt/Main, Germany 4 Institute of Forensic Medicine, Johann Wolfgang GoetheUniversity, Frankfurt/Main, Germany
We analyzed the rs643627 and the rs594242 polymorphisms of the serotonin 2A receptor gene (5-HT2A) and personality traits, measured by the Temperament and Character Inventory (TCI), in a patient sample of 167 suicide attempters and suicide completers (34.7% males, age: 39.9 13.9) and in a control sample of 312 healthy subjects (44.2% males, age: 45.1 14.9). Single genotypes were not associated with personality dimensions. However, controlling for sex, age and educational level, the rs643627 marker showed a marginal effect on Novelty Seeking (NS) (p ¼ 0.042). Subjects with the A/A genotype showed higher NS scores, particularly in males. Moreover, controlling for suicide attempt, the rs594242 genotype had a significant effect on Self Directedness (SD) (p ¼ 0.038) and Cooperativeness (C) (p ¼ 0.01). The G/G genotype was associated with lower SD scores in controls (F ¼ 6.1, df ¼ 2,286, p ¼ 0.002) and higher C scores in suicide attempters (F ¼ 3.06, df ¼ 2,219, p ¼ 0.049). A similar association on SD, although in the opposite direction, was previously reported, using the A-1438G variant in a sample of 146 healthy adults (Ham et al., 2004). Analyzing haplotypes, in the whole sample we did not find a positive association. However, including suicide attempt as covariant, the effect of the haplotypes on NS scores was close to significance, with the G-G haplotype showing lower scores on NS. Moreover, covaring for sex, age and education, the G-G haplotype effect remained. Summarizing, 5-HT2A polymorphisms could marginally modulate some personality traits. Future research is required. Ham, B. J., et al., 2004. Neurosci Lett 354, 2–5. P114 ASSOCIATION STUDY BETWEEN DR2, 5HTT, AND ATTENTION DEFICIT AND HYPERACTIVITY DISORDER Daniel Moreno De Luca Universidad Auto´noma de Bucaramanga, Bucaramanga, Santander Attention Deficit and Hyperactivity Disorder (ADHD) is the most common neuro-behavioral alteration in childhood, with an incidence of 8 to 18%. Due to its unknown aetiology, it has been considered as a complex condition, and monozygotic twin studies have shown a strong genetic influence on its development. Through triad tests, this study aims to investigate the association between polymorphisms in the dopamine receptor (DRD2) and the serotonin transporter (5HTT) genes, and ADHD. Methods: One-hundred patients from Bogota´ - Colombia, previously evaluated by neuropediatrics and/or child psychiatry and with living parents who gave their consent, were included in the study. Patients with genetic syndromes, neurological diseases and other psychiatric conditions were excluded. DSM-IV criteria were used for the psychiatric evaluation and detection of comorbilities, and the 5HTT and DRD2 polymorphisms were identified by means of VNTRs and RFLPs respectively. Results: No significant differences were found, among the 70 triads in which a blood sample could be extracted, between the presence of the patient’s polymorphic alleles and the polymorphic alleles from his parents. Conclusions: Results lead us to believe that there is no association between the presence of the above-mentioned polymorphisms and ADHD. Nevertheless, larger studies need to be performed, since the biological plausibility suggests that the serotonin and dopamine pathways might have an additive or multiplicative effect on the risk of ADHD. P115 GENES AND NEUROPSYCHOLOGICAL TEST PERFORMANCE: ARE THE TWO RELATED IN ADULT ADHD? Barbara Franke Radboud University Nijmegen Med. Centre, Nijmegen Introduction: Attention deficit/hyperactivity disorder, ADHD, is a neuropsychiatric disorder with high heritability, characterised by hyperactivity, inattention and impulsivity. It is found in 3–5% of school-aged children; in 30–60% of cases the disorder persists into adulthood. Up to now research into the genetics of ADHD has been focused mainly on the disease in children. The objective of the current study was to establish a first connection between several genetic polymorphisms and performance on neuropsychological tests in adults with ADHD. Method: We grouped 45 adults with ADHD (after extensive clinical assessment for ADHD and comorbidity) according to ADHD risk
Abstracts genotype or the presence of an ADHD risk allele, respectively, at four polymorphisms (DRD4 48 bp repeat, DRD4 120 bp duplicated repeat, SLC6A3 40 bp VNTR and COMT Val158Met). We then compared the test performance of the different groups on a battery of neuropsychological tests (tests of executive functioning and others). Results: The COMT Val158Met polymorphism was found to be related to differences in IQ and reaction time (with higher IQ scores and better reaction time for the low expressed COMT allele), both of the DRD4 polymorphisms (48 bp repeat and 120 bp duplication) showed an association with verbal memory skills (with reduced availability or function of the receptor associated with better performance) and the 40 bp VNTR in SLC6A3 was linked to differences in inhibition (with the supposedly higher expressed genotype performing best). Conclusions: Our findings tentatively point towards possible endophenotypes for adult ADHD. P116 DYSLEXIA SUSCEPTIBILITY LOCUS ON CHROMOSOME 1P CONFIRMED IN DUTCH SIB PAIR COLLECTION Barbara Franke1, C.G.F. de Kovel2, F.A. Hol2, J.J.P. Lebrec3, B. Maassen4, H. Brunner2, G. Padberg3, J. Platko4, D. Pauls4 1 Radboud University Nijmegen Med. Centre, Nijmegen 2 Department of Human Genetics, Radboud Un, Nijmegen 3 Department of Medical Statistics and Bio, Leiden 4 Department of Medical Psychology, Radbou, Nijmegen 3 Department of Neurology, Radboud Univers, Nijmegen 4 Psychiatric & Neurodev. Genetics Unit, M, Boston Introduction: Developmental dyslexia (DD) is defined as a specific and significant impairment in reading ability, not explained by deficits in intelligence, learning opportunity, motivation or sensory acuity. DD is one of the most frequent disorders in childhood, representing a major educational and social problem. It is well established that DD is a heritable trait, but only a few susceptibility genes have been reported, and none has been replicated. In this study, we attempted to confirm genetic linkage to developmental dyslexia and reading-related quantitative traits of loci that have been shown in previous studies to be associated with dyslexia. Methods: Families with two or more dyslexic children were ascertained by advertisements in magazines and newspapers. All parents and children were tested with a battery of tests, and families were included if at least two children were identified as dyslexic. Analyses were performed on 108 nuclear families. Results: The categorical trait (DD yes/no) showed the strongest linkage to 1p36 (NPL-LOD ¼ 2.0). LOD-scores for quantitative traits: word-reading, non-word reading and rapid naming were highly correlated and peaked near the same location as the categorical trait, as well as on chromosome 2. Non-word repetition showed little phenotypic correlation with dyslexia or with the other quantitative traits, and this trait showed linkage peaks on 11p and 15q. Conclusion: Our results tentatively confirm the dyslexia locus on chromosome 1 and identify a region for further study in the Dutch sib pair cohort. P117 REPLICATION OF AN ASSOCIATION BETWEEN ADHD AND A SPECIFIC DAT1 HAPLOTYPE IN THE IMAGE SAMPLE Keeley Brookes1, Ben Neal2, Xiaohiu Xu2, Jan Buitelaar3, Richard Ebstein4, Michael Gill5, Wai Chen2, Tobias Banaschewski3, Edmund Sonuga-Barke4, Iris Manor5, Ana Miranda6, Bob Oades7, Herbert Roeyers8, Ari Rothenberger3, Jo Sergeant4, Hans-Christoph Steinhausen5, Steve Faraone6, Philip Asherson2 1 SGDP Center, London 2 Institute of Psychiatry, London 3 Radboud University Nijmegen Medical Cent, Nijmegen 4 S. Herzog Memorial Hospital, Jerusalem 5 Trinity Centre for Health Sciences, Dublin 3 University of Go¨ttingen, Goettingen 4 University of Southampton, Southampton 5 Geha MHC, Petach-Tikva 6 University of Valencia, Valencia 7 University Clinic for Child and Adolescent, Essen 8 Ghent University, Ghent 4 University of Amsterdam 5 University of Zurich 6 SUNY Update University, New York
765
The behavioural condition Attention Deficit Hyperactivity Disorder (ADHD) has been shown to be associated with the 10-repeat allele of a polymorphism located in the 30 UTR of the dopamine transporter gene DAT1. Our previous study found that in two independent samples, this allele forms a specific two-marker haplotype with the 3-repeat allele of another VNTR polymorphism located within intron 8 of the gene that is associated with ADHD. Here we report an initial replication of this finding, and additional evidence for association spanning the 50 regulatory region of the gene has been observed in a sample of 776 DSM-IV combined type cases ascertained for the International Multicentre Genetics (IMAGE) project, contributing to our understanding of the potential role of this gene in risk for ADHD. P118 WHY DO READING DIFFICULTIES CO-OCCUR WITH OTHER LEARNING DISABILITIES? SNP-SET ANALYSES Nicole Harlaar, Emma Meaburn, Yulia Kovas, Robert Plomin Institute of Psychiatry, London Introduction: At this meeting, a set of 20 single-nucleotide polymorphisms (SNPs) associated with early reading disability and ability has been reported from a genome-wide association scan (Meaburn et al.). We use this reading SNP set to test the theory of ‘generalist genes’, which predicts that many of the genes associated with reading will also be associated with other learning abilities. Methods: A large UK sample of twins was assessed at age 7 on telephone measures of reading and cognitive abilities, and year-long teacher assessments of reading and mathematics achievement. Quantitative genetic analyses were used to examine genetic and environmental influences on the covariation between reading, mathematics and general cognitive ability (‘g’). Multiple regression analyses were conducted to assess the extent to which mathematics and ‘g’ can be predicted by the 20 SNPs and a composite SNP set associated with reading at age 7. Results: As predicted by the ‘generalistgenes’ theory, our quantitative genetic analyses indicated that more than half of the genetic influences on reading disability overlapped with genetic influences on mathematics and ‘g’. There were also significant independent genetic influences on reading. Similarly, more than half of the 20 SNPs in the reading SNP set were also associated with mathematics and ‘g’. Logistic regression analyses yielded similar results for low mathematics performance and low ‘g’ scores. Conclusions: Our reading SNP set confirms results from quantitative genetic analyses indicating that many of the genes that affect reading ability and disability also affect mathematics and ‘g’. P119 IMPACT OF SMOKING DURING PREGNANCY ON ATTENTIONAL CONTROL IN ADHD Marieke Altink Radboud University Medical Centre, Nijmegen Objective: To explore the effect of maternal smoking during pregnancy, whether or not in interaction with family-genetic vulnerability, on attentional control. Method: The study consists of 50 concordant and discordant ADHD sib-pairs of Dutch descent and 61 healthy control sib-pairs, aged 6–18 years. Performance on a visual continuous performance task was assessed for all children. Fluctuations in tempo (SMST) on the task reflect attentional control. Maternal smoking during pregnancy was specified as: ‘‘yes’’ versus ‘‘no’’. Results: Twenty-seven percent of children with ADHD had a history of maternal smoking during pregnancy, compared with 13% of the healthy controls. This indicates more than a two-fold increase in risk for a diagnosis of ADHD in those individuals whose mothers smoked during pregnancy (OR 2.9, 95% CI 1.43-5.87 p < 0.05). Significant differences were found for attentional control in the ADHD sib-pair group between children whose mothers smoked during pregnancy and those who did not (mean SMST ¼ 3.55, SD ¼ 2.09, mean SMST ¼ 2.68, SD ¼ 1.96, p ¼ 0.028). The regression model for SMST of siblings included an effect of age, of proband performance on SMST, of tobacco use during pregnancy of the sibling and an interaction between SMST of proband and prenatal smoking (p ¼ 0.039). Conclusion: Mean scores and the correlation between the performance of the proband on SMST and the sibling differ when the mother uses tobacco during pregnancy. This reflects gene-environment interaction.
766
Abstracts
P120 PRKCB1 IS NOT ASSOCIATED WITH AUTISM IN THE IRISH POPULATION Mao-Sheng Yang, Judith Conroy, Ziarih Hawi, Louise Gallagher, Michael Gill Trinity College, Dublin Introduction: The protein kinase C-beta 1 (PRKCB1) gene encodes both PRKCB1-1 and PRKCB1-2 isoforms and is highly expressed in the granule cells of the brain. The isoforms of the protein kinase C family play a critical role in the regulation of synaptic transmission. PRKCB12 is highly expressed in the synaptic endings of the parallel fibres of granule cells. A decreased number of granule cells has been reported in brain tissue from autistic patients. The gene encoding PRKCB1 is mapped to chromosome 16p11.2. This region was suggested to harbour possible susceptibility genes for autism. Methods: We examined six single nucleotide polymorphisms (SNPs) (rs3785387, rs196002, rs1873423, rs198182, rs182068, rs169143) in 171 Irish trio families with Autism. Transmission disequilibrium tests (TDT) for SNP markers and haplotypes were carried out using the TDTPHASE and PDTPHASE from the UNPHASED version 2.404 programmes. Results and conclusion: TDT analysis showed no evidence of association with autism in both SNPs and halpotypes of the 6 SNPs. Our data do not support the suggestion that this locus PRKCB1 contributed to the development of autism, at least in the Irish population. P121 NO ASSOCIATION BETWEEN TPH2 GENE POLYMORPHISMS AND ADHD IN A UK SAMPLE Sheehan Karen1, Ziarih Hawi1, Michael Gill1, Lindsey Kent2 1 Trinity College, Dublin, 2 University of Cambridge, Cambridge Tryptophan Hydroxylase 2 (TPH2) is rate-limiting enzyme in the biosynthesis of serotonin, which is exclusively expressed in the brain. Recent molecular studies have reported significant association between markers mapped to TPH2 and psychiatric conditions, including ADHD. We have examined four single nucleotide polymorphisms (SNPs), two of which (rs1843809 rs1386493) were reported to associate with ADHD in an Irish ADHD sample. Transmission disequilibrium analysis revealed no significant association between any of these markers and ADHD. Dividing by the sex of the transmitting parent has also failed to replicate the previously- reported paternal over-transmission of the associated alleles to ADHD probands. A larger sample size will be required to clarify whether or not TPH2 alleles are associated with ADHD. P122 FOUR GENES ANALYSED FOR ASSOCIATION WITH AUTISM AND DBH LEVELS Lynne Elizabeth Cochrane1, Judith Conroy1, George Anderson2, Louise Gallagher1, Michael Gill1 1 Trinity College, Dublin 2 Yale, Connecticut Autism is a developmental disorder typically characterised by dysfunctions in communication, socialisation and/or imagination. Although its aetiology is unknown, it has a strong genetic component. Previous studies have reported decreased levels of the enzyme DBH in autistic individuals. Markers in the DBH gene itself are found to be responsible for approximately 55% of the variance in DBH expression in control populations. In this study, we examined genes that may influence expression of the DBH gene. Phox2a and Phox2b are transcription factors known to act on DBH transcriptional activity. In addition, PNMT and aldehyde reductase are genes downstream of DBH in the dopamine pathway. TDT analysis conducted on individual markers in Phox2a and Phox2b showed no association with autism. However, a two-marker haplotype in Phox2b showed a trend towards association (p < 0.09). QTLSNP analysis showed nominal association of a marker rs7689304 in Phox2b with DBH levels in a parental sample, using a dominant model (p < 0.05). In addition, TDT analysis showed no significant variations in transmission of alleles from parents to affected children in 3 SNPs at PNMT. Analysis of 8 SNP markers in aldehyde reductase showed significant association between autism and rs2088102 (p < 0.04), and a trend towards association with rs622789 (p < 0.09). QTLSNP analysis of aldehyde reductase showed one marker (rs12049503) associated with DBH levels in parents, in both the
codominant and recessive models (p < 0.06 and p < 0.03 respectively). These findings need to be replicated in other populations. P123 AN INVESTIGATION OF THE NEUROTROPHIC FACTOR GENES GDNF, NGFAND NT3 IN SUSCEPTIBILITY TO ADHD Lindsey Kent1, Zahoor Syed2, Frank Dudbridge3 1 University of Cambridge, Cambridge 2 Addenbrookes Hospital, Cambridge 3 MRC Biostatistics Unit, Cambridge Attention Deficit Hyperactivity Disorder (ADHD) is a common, highly heritable, neurodevelopmental disorder with onset in early childhood. Genes involved in neuronal development and growth are, thus, important etiological candidates, and neurotrophic factors have been hypothesized to play a role in the pathogenesis of ADHD. Glialderived neurotrophic factor (GDNF), nerve growth factor (NGF(Beta subunit)) and neurotrophic factor 3 (NT3) are members of the neurotrophin family and are involved in the survival, differentiation and maintenance of neuronal cells. We have examined 9 coding and intronic SNP’s across GDNF, NGF and NT3 in a family-based association sample of 120 DSM IV ADHD probands and their biological parents, as well as a case control analysis with 120 sexmatched controls. Significant overtransmission of the C allele of a non-synonymous C/T snp (rs6330) in NGF which codes an alanine/ valine change, was found in the family-based sample (Chi sq ¼ 3.69, p ¼ 0.05), but not confirmed in the case control sample. Although this snp is located in the 50 pro-NGF sequence and not the mature NGF protein, it may affect intracellular processing and secretion of NGF. No other significant results were found in either the family-based or case control samples. P124 ASSOCIATION ANALYSIS OF SLC25A12 AND EN2 IN THE FINNISH FAMILIES WITH AUTISM-SPECTRUM DISORDERS Joni Turunen1, Tero Ylisaukko-oja1, Helena Kilpinen1, Karola Rehnstro¨m1, Elli Kempas1, Raija Vanhala2, Taina Nieminen-von Wendt2, Lennart von Wendt2, Leena Peltonen1 1 National Public Health Institute, Helsinki 2 Hospital for Children and Adolescents, Helsinki Recently, two candidate genes for autism have been identified. First, Ramoz et al. reported a significant association between autism and the mitochondrial aspartate/glutamate carrier gene (SLC25A12) on chromosome 2q. This finding has since been replicated; however, negative reports also exist. Second, the homeodomain transcription factor gene ENGRAILED 2 (EN2) on chromosome 7q has been associated with autism spectrum disorders (ASD) in multiple independent samples. The purpose of this study was to investigate the previously associated SNPs in two separate Finnish study samples with ASDs. The samples included 100 families with strictly defined autism and 29 extended families with Asperger syndrome. SNPs rs2056202 and rs2292813 in SLC25A12 and SNPs rs1861972 and rs1861973 in EN2 were genotyped using either fluorogenic 50 nuclease allelic discrimination chemistry (TaqMan1) with an ABI Prism1 7900 Sequence Detection System or direct sequencing. The association analyses were performed using the family based association program FBAT. Two SNPs in EN2 yielded no evidence of association, but rs2292813 in SLC25A12 was associated significantly with autism (p ¼ 0.0018). Also other family-based analysis methods supported this finding. Analysis of the Asperger syndrome sample provided no evidence for association at any of the SNPs tested. In conclusion, our study supports a potential role for SLC25A12, but not for EN2, in the etiology of autism-spectrum disorders in the Finnish population. P125 ADHD IN CHILDREN WITH LEARNING DISABILITIES Eleni Zarafoniti, Alexandra Michopoulou, Barbara Belesioti, Vasiliki Goula, Panagiota Mihalopoulou, Dimitrios Giannios, Katerina Pantazopoulou Penteli’s Children Hospital, Penteli Objectives: Parents consulting with their children regarding learning difficulties at the Psychology Department in Penteli’s Children’s Hospital often complain that their children also present hyperactivity
Abstracts and/or difficulties in concentration. The aim of the present study is to examine the co-morbidity between learning difficulties and ADHD. Methods: The sample consisted of 50 children, aged 6 to 14. Children were categorised into three different groups: a) children with slight learning disabilities b) children with dyslexia and c) children with learning and intellectual disabilities. Parents were asked to complete the ADHD-IV scale (DuPaul, Power, Anastopoulos and Reid, 1998). In order to compare the three groups, we used the x2 test. Results: It was found that co-morbidity between learning difficulties and ADHD does exist, but not to the extent that parents report. The type of learning difficulties did not seem to affect the results. Conclusions: A considerable number of children with learning difficulties were found to present ADHD. However, it seems that parents usually tend to overestimate their children’s hyperactivity and/or difficulties in concentration. P126 GENETIC INFLUENCES ON MECHANICALLY ASSESSED ACTIVITY LEVEL IN CHILDREN Alexis Caroline Wood1, Kim Saudino2, Hannah Rogers1, Philip Asherson1, Jonna Kuntsi1 1 Institute of Psychiatry, London 2 Boston University, Boston Activity level is part of the symptom domain of impulsivityhyperactivity within Attention Deficit Hyperactivity Disorder. Yet it is poorly understood, partly due to limitations on parent and teacher ratings that are typically used as measurements. This project aims to improve understanding of the genetic and environmental influences within the etiology of overactivity, by obtaining objective measurements of activity level using actigraphs, in addition to standard behavioural rating scales. Our aim in this first wave of analysis was to investigate the heritability of objectively-measured activity. A general population sample of 400 twin pairs, ages 7–9, wore actigraphs on the waist and dominant leg throughout a 2.5 hour test session, including a 20-min break. This gave minute-by-minute readings of the number and intensity of movements. Using comparisons between monozygotic and dizygotic twins, twin model fitting was conducted. This suggested a heritability of 53% on the lab-based session, with no role for common environment. For an actigraph score taken during the break, heritability was 35%, with common environment accounting for 37% of the variability. Bivariate model fitting showed a genetic correlation of 1.0, indicating that the same genes underlie activity level in the two situations, but have different strengths of influence on the phenotype. Actigraph data show a phenotype with a moderate degree of genetic influence. Further analyses will investigate actigraph data’s contribution to a phenotype that is maximally sensitive to the underlying genotype and how this may inform molecular genetic studies. P127 A POSITIVE HAPLOTYPE ASSOCIATION OF THE CHROMOSOME 15 GAMMA-AMINO BUTYRIC ACID BETA 3 SUBUNIT GENE (GABRB3) WITH AUTISM SPECTRUM DISORDER John Powell, Sarah Curran, Ben Neal, Katharina Dworzynski, Patrick Bolton Institute of Psychiatry, Kings College, London Autism spectrum disorder (ASD) comprises the related diagnostic categories: autism, atypical autism, Asperger’s syndrome and other pervasive developmental disorders (ICD-10). Several genes are believed to contribute to the underlying genetic risk for the development of ASD. However, etiologic heterogeneity may impede the discovery of ASD-susceptibility genes. Chromosome 15q11-q13 (also known as the Prader-Willi/Angelman Syndrome critical region) has been identified as a strong candidate region on the basis of autistic symptoms in patients with chromosomal abnormalities in that region and numerous suggestive linkage and association findings. We have tested 148 simplex U.K. families, characterised by both ADI-R and ADOS to have an ASD, for association with the GABRB3 gene on chromosome 15q11-13. Single nucleotide polymorphisms (SNPs), spaced on average by 13 kb, were genotyped across the gene that had not been previously parsed in autism. Global significance for all marker p-values was obtained with region-wide empirical significance. Short range linkage disequilibrium was observed, but there is evidence for large amounts of recombination in the gene. A high LD haplotype of 5 SNPs was tested, including the two most significant markers, and
767
was significant at p when haplotypes of >1% were included. The marker responsible for the association in this region is a mere 12 kB from a significant marker already reported by McCauley et al., 2004. P128 A BIRD IN HAND IS WORTH TWO IN THE BUSH? THE ABILITY TO DELAY GRATIFICATION IN CHILDHOOD Krisztina Lakatos Inst Psychology, Hung. Acad. Sci, Budapest Delaying access to rewards has been associated with social influences, such as maternal behaviour, but also with the functioning of the reward pathway, attention and behavioural problems. In the current study, 86 children participated in the Mischel delayed gratification paradigm at 7.5 years of age, and were genotyped independently. The latency to wait for the reward was weakly influenced by polymorphisms of the serotonin transporter (STin2) and the D4 dopamine receptor (DRD4) genes. Children homozygous for the high transcription activity (12x) allele of the STin2 showed the longest waiting times (F(2,74) ¼ 3.339, p ¼ 0.041, eta2 ¼ 0.083). Absence of the 7-repeat allele of DRD4 increased the chance that children would stop waiting before time was over (2 ¼ 4.036, df ¼ 1, p ¼ 0.045, phi ¼ 0.225). The significant statistical interaction between these polymorphisms indicated that the simultaneous presence of the 10/12 genotype and the absence of the 7x allele of the DRD4 gene decreased the time children were able to delay rewards (F(2,74) ¼ 5.189, p ¼ 0.008, eta2 ¼ 0.123). Animal studies indicate the role of the serotonergic system in controlling impulsive choices, whereas DRD4 receptor gene polymorphisms are associated with ADHD. Children diagnosed with ADHD were shown to have better neuropsychological functioning in the presence of the 7-repeat allele. The present results also indicate involvement of the serotonergic and dopaminergic systems in the ability regulate behaviour in order to delay rewards. P129 POLYMORPHISMS IN DRD2 ARE ASSOCIATED WITH AD/HD Michael Cuccaro1, Scott Collins1, Arthur Anastopoulos2, Ave Lachiewicz1, Erin Kane2, David FitzGerald1, Melanie Kail1, Linda Exelbierd1, John Gilbert1, Allison Ashley-Koch1 1 Duke University Medical Center, Durham, NC 2 University of North Carolina Greensboro, Greensboro, NC We are conducting a candidate gene analysis of ADHD families collected primarily in North Carolina. Eligible families contain at least onechild between 5 and 12 years of age who meets DSMIV criteria for ADHD. All available parents and siblings are collected and sampled. One aim has been the evaluation of genes in the dopaminergic system as this system has been implicated in previous candidate gene analyses. We have genotyped 9 haplotype tagging SNPs in the DRD2 gene in 66 families containing a total of 227 sampled and phenotyped individuals. These SNPs were analyzed for association with the ADHD phenotype using several measures of genetic association: the PDT (Martin et al., 2000), the genotype-PDT (Martin et al., 2003a) and the association in the presence of linkage (APL) test (Martin et al., 2003b). The most significant association was observed with marker rs4274224 (PDT p ¼ 0.008; geno-PDT p ¼ 0.03; APL p ¼ 0.02). Markers rs2075654 and rs1079596 were also associated (APLp ¼ 0.03 and 0.01, respectively). These associations remained when our analysis was restricted to Caucasian families (n ¼ 57). Few studies have found evidence for an association with DRD2 and ADHD. However, many studies have only evaluated the Taq 1A polymorphism which has primarily been associated with risk for alcoholism. Our data suggest that genetic variation in DRD2 may indeed be associated with risk for ADHD and warrant further examination to determine if these preliminary results are replicable or possibly a population-specific association. P130 A GENE CENTRIC ASSOCIATION STUDY OF 1500 SNPs IN THE CHROMOSOME 2q AUTISM SUSCEPTIBILITY LOCUS Elena Bacchelli1, Francesca Blasi2, Simona Carone2, Claudio Toma2, Janine Lamb3, Nuala Sykes3, Gabrielle Barnby3, Andrew Morris3, Laura Winchester4, Helen Butler4, Anthony J. Bailey5, Anthony P. Monaco4, Elena Maestrini5, Consortium IMGSAC6 1 University of Bologna, Bologna 2 University of Bologna, Bologna
768
Abstracts
3
2
4
3
Wellcome Trust Centre for Human Genetics, Oxford Wellcome Trust Centre for Human Genetic, Oxford 5 Park Hospital for Children, Oxford 5 University of Bologna, Bologna 6 http://www.well.ox.ac.uk/maestrin/iat.h, Oxford Autism is a lifelong neurodevelopmental disorder that shows strong evidence for a complex genetic etiology. The IMGSAC performed a whole genome screen for linkage and identified the strongest result on chromosome 2q, which has been independently replicated. The linkage peak is very broad, comprising 40 Mb and containing almost 200 genes. We have conducted mutation screening on 19 genes among the best candidates, but no evidence was found that any of these genes strongly contributes to autism susceptibility. Therefore, we have decided to undertake a systematic gene-based high-density SNP genotyping and association study. Utilising data from the HapMap project we evaluated the patterns of linkage disequilibrium (LD) and the distribution of haplotype blocks across the 40 Mb region. We selected the haplotype-tagging SNPs (htSNPs) from blocks located within all known genes and their putative regulatory elements, resulting in approximately 1500 htSNPs genotyped in 576 samples (126 autistic trios from multiplex families and 188 sex-matched controls), using the GoldenGate1 assay. 1497 out of 1536 SNP assays (97.5%) were successful and genotyping efficiency was 99.79%. Familybased and case-control analysis allowed to identify a number of haplotypes within genes showing strong evidence of association with autism. Replication of these results is now in progress and it will consist in genotyping 400 SNPs (htSNPs from the top 5% of associated blocks and additional SNPs from genes showing the most significant evidence for association) in a new independent sample of 200 trios and 200 controls. P131 MUTATION SCREENING OF THE PTEN GENE IN PATIENTS WITH AUTISM AND MACROCEPHALY Catalina Betancur1, Guiquing Cai2, Pauline Chaste3, Gudrun Nygren4, Juliet Goldsmith2, Jennifer Reicher2, Henrik Anckarsater4, Maria Rastam4, Marion Leboyer3, Christopher Gillberg4, Alain Verloes5, Joseph D. Buxbaum2 1 INSERM U513, Creteil 2 Mount Sinai School of Medicine, New York 3 INSERM U513, Universite´ Paris XII, Cre´teil 4 Department of Child and Adolescent Psych, Goteborg 5 Clinical Genetics Unit, Hoˆpital Robert D, Paris The tumor suppressor gene PTEN, on 10q23, is mutated in Cowden disease, Bannayan-Riley-Ruvalcaba syndrome, Lhermitte-Duclos disease, and Proteus syndrome, collectively referred to as PTENhamartoma tumor syndromes. Individuals with germline PTEN mutations are prone to tumors, sometimes associated with neurological features such as macrocephaly, mental retardation and seizures. Furthermore, PTEN mutations have been described in a few patients with autism and macrocephaly. In this study, we screened PTEN for mutations and deletions in 88 patients with autism spectrum disorders and macrocephaly (head circumference > þ2 SD) recruited by the Paris Autism Research International Sibpair (PARIS) study or by Mount Sinai School of Medicine/Autism Genetics Resource Exchange. Mutation analysis was performed by direct sequencing of all exons, flanking regions and promoter. Dosage analysis of PTEN was carried out using multiplex ligation-dependent probe amplification. No exonic or whole gene deletions were observed. We identified a previously undescribed missense mutation in exon 8 (D326N) in a 5-year-old boy with autism, mild mental retardation and marked macrocephaly (þ10 SD). Segregation analysis showed that the mutation was de novo; study of 15 polymorphic markers excluded non-paternity. Physical examination was negative except for macrocephaly, operated bilateral hexadactyly of the feet and 3 cafe´-au-lait spots. These findings indicate that molecular screening of PTEN is indicated in patients with autism and marked macrocephaly. P132 COGNITIVELY INTACT CHILDREN CARRYING THE APOLIPOPROTEIN E EPSILON (E) 4 ALLELE HAVE A THINNER MESIAL TEMPORAL CORTEX Philip Shaw1, Jason Lerch2, Kristen Taylor3, Jeffrey Seal3, Alan Evans2, Judith Rapoport3, Jay Giedd3 1 NIMH, Bethesda
Montreal Neurological Institute, Montreal Child Psychiatry Branch, NIMH, Bethesda Child Psychiatry Branch, Bethesda
3
The e4 polymorphism of the apolipoprotein gene is associated with late onset Alzheimer’s Disease (AD) and has effects on brain function in asymptomatic young adults. We hypothesized that the allele would also be associated with neuroanatomic changes in children particularly in the entorhinal cortex and parahippocampal gyrus, where the earliest pathological changes of AD are found. We acquired neuroanatomic magnetic resonance images on 258 healthy children and adolescents, 64% were scanned at least twice. A measure of cortical thickness was extracted and children genotyped (65 e4 heterozygotes; 7 e4 homozygotes and 176 non-carriers). Using mixed model regression, we examined the effects of e4 carriers status on cortical thickness, and whether these effects were static or evolved throughout adolescence. For our entorhinal regions of interest we adopted a p < 0.05; throughout the remainder of the cortex we used the false discovery rate procedure (q ¼ 0.05) to adjust for multiple comparisons. e4 allele carriers had a significantly thinner cortex in the entorhinal cortex, and to a lesser extent throughout the remaining parahippocampal gyrus, but not in the neocortex. There was no evidence of any significant difference between e4 carriers and others in the trajectories of cortical development. There were no also no effects of genotype on Weschler IQ scores. Our findings are compatible with the concept that the e4 allele imbues carriers with a cortical phenotype in the mesial temporal lobe, which is present from early development and may contribute to a propensity to develop AD.
P133 THE GENOMIC COMPARISON BETWEEN MONOZYGOTIC TWINS DISCORDANT FOR AUTISTIC DISORDER Aoi Hashida1, Akira Imamura1, Hideji Yamashita2, Naohiro Kurotaki1, Yuji Okazaki2, Hiroki Ozawa1 1 Nagasaki University, Nagasaki 2 Kyushu Tokai University, Aso 2 Tokyo Metropolitan institute of Medical, Tokyo Recently several researchers reported that monozygotic (Mz) twins discordant for some phenotypes were caused by genomic differences. To compare the genomes of Mz twins discordant for autism, we performed the fluorescence representational genomic profiling (FRGP) method and SNPs analysis. The subjects were a pair of female Mz twins discordant for DSM-IV Autistic Disorder. We performed FRGP, in which the whole genome was scanned using two-dimensional gel electrophoresis after digestion with a methylation-sensitive restriction enzyme and fluorescent labeling. Also, we analyzed SNPs using Affymetrix GeneChip Human Mapping 100K Set (50K Xba Array, 50K Hind Array). After comparing patterns with approximately 600 spots as restriction landmarks, we detected at least four spots that differed in intensity between the twins. Two of the four target spots were successfully cloned. In SNPs detection, a hundred and one SNPs had different signal intensity among all 116204 SNPs. The FRGP method and SNPs analysis enabled us to detect genomic differences between Mz twins discordant for autism. We expect that such differences might reflect the etiology of autism. P134 PROTON MAGNETIC RESONANCE SPECTROSCOPIC INVESTIGATION IN CHILDREN WITH ATTENTION-DEFICIT/ HYPERACTIVITY DISORDER Leila Ben Amor, Sophie Chantal, Karine Morasse, Mario Harvey, Nicholas Barden Laval University, Le´vis (Quebec) Attention Deficit Hyperactivity Disorder (ADHD) is a neuro-developmental disorder with a high genetic component that causes high levels of activity, impulsivity and/or inattentiveness and significantly impairs social behaviours. Accumulated evidence evokes the alteration of dopaminergic and glutaminergic transmission, which in turn modifies brain metabolite levels. The fronto-striatal circuit seems implicated most consistently. We used proton magnetic resonance Spectroscopy (1H-MRS) to investigate neurochemical changes between children with ADHD (n ¼ 13) and healthy controls (n ¼ 6) in the prefrontal cortex, striatum and cerebellum. We examined markers of
Abstracts
769
brain activity: N-acetylaspartate (NAA), choline (Cho), myo-Inositol (mI), glutamate (Glu), and glutamate-glutamine (Glx) ratios relative to creatine (Cr) concentration. Children with ADHD had significantly lower Glu/Cr and Glx/Cr ratios in the right prefrontal cortex, compared to controls (1.36 vs. 1.57, p ¼ 0.04 and 2.11 vs. 2.47,p ¼ 0.03). Moreover, children with ADHD demonstrated a significant reduction of NAA/Cr ratios in the left cerebellum (1.06 vs.1.43 p ¼ 0.001), but no significant differences with any analyzed metabolites were found in striatal regions. Conclusion: Our results confirm neurometabolic changes in prefrontal cortex in children with ADHD compared to healthy controls. In contrast to previous studies, we report decreased levels of GLx/Cr in the right prefrontal cortex. This may indicate an imbalance in right/left glutamate metabolites rather than a regional modulation. Our results also raise the possible implication of cerebellar dysfunction in the ADHD pathophysiology.
memory, and executive dysfunctions in areas such as planning, working memory, and motor organisation. These cognitive deficits were associated with structural brain anomalies, primarily in the medial frontal, temporal and cerebellar grey matter. Conclusions: This complex cognitive phenotype is probably the effect of differential genetically determined brain structural development.
P135 SOCIAL DYSFUNCTIONING IN ADHD: A REVIEW Judith Simone Nijmeijer1, P.J. Hoekstra1, C.A. Hartman1, J.K. Buitelaar2, R.B. Minderaa1 1 University of Groningen, Groningen 2 University of Nijmegen, Nijmegen
Introduction: Cognitive function is closely correlated with a linear decrease in dopamine from early adulthood through to death during normal aging (1), thus it is possible that SLC6A3 may play a role in cognitive functioning in non-demented individuals. Here we have characterised and investigated the role of a novel intronic 63bp-VNTR with cognitive abilities. Methods: Samples were available from Caucasian non-demented volunteers (n ¼ 701, age > 50 years) comprising 243 males and 558 females, who undertook a series of tests that measured fluid intelligence, processing speed, vocabulary ability and immediate/ delayed memory. Volunteers were genotyped for a novel intronic SLC6A3 63-base-pair VNTR. The different alleles were characterised by sequencing and then assayed by the use of luciferase constructs in eukaryotic cell lines. Results: We identified three alleles comprised of 6, 7 and 8 repeats (allele frequency 0.1%, 81.2% and 18.7% respectively). We found that the longest variant (8 repeats) was associated with improved memory abilities using both the recessive (p ¼ 0.013) and additive models (p ¼ 0.040). The basal level of expression differed for each allele and it was selectively modified by CTCF transcription factor. Conclusion: These data suggest a functional role of this novel VNTR polymorphism. A synergic effect with of other functional variants may also exist and thus further studies are warranted. Reference: 1. Badcock, J Psychiatr Res2005; 39:11–19.
Attention deficit hyperactivity disorder (ADHD), characterized by symptoms of inattention, impulsivity, and hyperactivity, affects 3–5% of children worldwide. The profound influence of ADHD on social functioning may be its most disabling aspect. ADHD children tend to be rejected by peers, have few mutual friends, and show conflicts with adults, mainly resulting from their disruptive, intrusive, and aggressive behaviours. These behaviours are an intrinsic part of ADHD, but also reflect the high co-occurrence rate of oppositional defiant and conduct disorder with ADHD. Although common in ADHD, problems in social functioning vary in severity between individual children, and their influence on long-term outcome varies likewise. An intriguing question is whether findings of severe levels of social dysfunctioning in a subgroup of ADHD children may be related to autistic type deficits. The link to autism is interesting, as some children with ADHD show phenotypical similarities to children with autism spectrum disorders. Furthermore, genetic linkage findings have partially suggested the same genetic regions of interest in both ADHD and autism, offering an attractive explanation for the overlap in symptoms. In this review, the literature on the aforementioned themes is summarized. Furthermore, plans for future research on the subject of social functioning in ADHD are presented. These include the precise phenotyping of social disability in ADHD children, and relating social phenotypic data to genetic linkage findings on areas found in autism and to geneenvironment interaction effects, based on data collected in the International Multi-Centre ADHD Gene Project (IMAGE). P136 COGNITION AND BRAIN IN CHILDREN WITH 22q11.2 DELETION SYNDROME Linda E. Campbell1, Angela F. Stevens2, Rayna Azuma2, Robin Morris2, Declan G.M. Murphy2, Kieran C. Murphy3 1 Centre for Mental Health Studies, Newcastle 2 King’s College London, Institute of Psychiatry, London 3 Royal College of Surgeons in Ireland, Dublin Background: Velo-cardio-facial syndrome or 22q11.2 deletion syndrome (22qDS) is the most common known microdeletion syndrome characterised by a high frequency of congenital heart disease, characteristic dysmorphology, structural brain anomalies, learning disabilities and psychiatric disorder. Method: In this study, our aim was twofold: 1) to characterise memory, executive function and attentional abilities of children with 22qDS (N ¼ 47) compared with age, gender and parental socioeconomic status matched sibling controls (N ¼ 26); and to 2) relate cognitive function to structural brain anatomy. We used standardised neuropsychological batteries, the Children’s Memory Scale (CMS), the Cambridge Automated Neuropsychological test battery (CANTAB) and the Maudsley attention and response suppression task battery (MARS). In addition, topographically unrestricted voxel-based morphometry as well as regions of interest (ROI) analysis was utilised to investigate structural neural correlates of cognitive impairments. Results: Overall, the 22qDS group had reduced memory scores, with specific deficits in visual memory and in particular facial recognition
P137 FUNCTIONALITY AND ASSOCIATION STUDY OF A NOVEL VARIABLE NUMBER OF TANDEM REPEAT (VNTR) POLYMORPHISM OF DOPAMINE TRANSPORTER (SLC6A3) WITH GENERAL COGNITIVE ABILITIES Fabio Miyajima1, Kate Haddley2, Jill Bubb2, J.F. William Deakin1, Neil Pendleton1, Mike Horan1, William Ollier1, Antony Payton1, John Quinn2 1 University of Manchester, Manchester 2 University of Liverpool, Liverpool
P138 A t(1;16) TRANSLOCATION DISRUPTS A MYOSIN GENE THAT OVERLAPS NUDE ON CHROMOSOME 16 p M. Pat Malloy1, Ben S. Pickard2, David J. Porteous2, Douglas H.R. Blackwood1 1 Division of Psychiatry, University of Ed, Edinburgh 2 Medical Genetics, University of Edinburgh, Edinburgh The search for genes that raise the susceptibility to schizophrenia has benefited from a positional approach to candidacy based on cloning disrupted coding sequences from chromosomal re-arrangements in individuals with psychosis. An individual has been found where DSMIV chronic schizophrenia co-exists with a balanced t(1;16) translocation. We have cloned the breakpoint on 16p and found that it directly disrupts a myosin related gene MYH11, which itself overlaps the immediately adjacent gene NUDE. NUDE is a DISC1 interactor through the LIS1 system. DISC1 is one of the best genes for altering risk to psychosis with replicated linkage, association and mutation data to support its candidacy. We hypothesise that in the individual under study there is haploinsufficiency of the myosin related protein, and also a possible alteration of NUDE expression which may relate to the illness of the individual. P139 AN UNUSUAL UNBALANCED DICENTRIC TRANSLOCATION t(18;21) RESULTS IN 18pSYNDROME—CLINICAL AND CYTOGENETIC FEATURES Tanya Thiagarajah1, M. Pat Malloy2, David J. Porteous3, Douglas H.R. Blackwood2, Ben S. Pickard3 1 LD Directorate, NHS Lothian, Edinburgh 2 Division of Psychiatry, University of Edinburgh 3 Medical Genetics, University of Edinburgh, Edinburgh The first chromosomal deletion detected in the human species was that of the short arm of chromosome 18. It is, in fact, one of the more frequent
770
Abstracts
autosomal deletions occurring in isolation or as the unbalanced consequence of a translocation. We describe a female with mild to moderate learning disability (US Mental Retardation) and mild physical dysmorphisms who was discovered to have a karyotype 45 XX t(18q;21q) resulting in deletion of one whole short arm of chromosome 18 (in addition, there was deletion of 21p, which carries little genic information). Although the individual had previouslydescribed phenotypic associations of 18p- including ptosis and depressed nasal bridge and autoimmune disorders (including juvenile onset diabetes mellitus and hypothyoidism), there were no cardiac abnormalities or dystonias. The translocation was unusual in that two centromeres were preserved on the derivative chromosome.
P140 FIRST CASE OF MOSAIC TRISOMY OF CHROMOSOME 1 IDENTIFIED IN THE SCHIZOPHRENIC BRAIN Yuri Yurov1, Ivan Iourov1, Svetlana Vorsanova2, Thomas Liehr3, Irina Demidova2, Viktor Vostrikov1, Natalia Uranova1 1 Mental Health Research Center, Moscow 2 Institute of Pediatrics and Children Surgery, Moscow 3 Institute of Human Genetics and Anthropology, Jena Genomic instability manifested as mosaic numerical chromosome imbalances (aneuploidy) in the brain are supposed to be related to the pathogenesis of schizophrenia (Yurov et al., 2001; Iourov et al., 2006). Here, we report a molecular cytogenetic study of the postmortem brain of a female with an intravital diagnosis of schizophrenia. Interphase multiprobe fluorescence in situ hybridisation (mFISH) has indicated the presence of an additional chromosome 1 in about 5% of cells in the cerebral cortex. The rate of trisomy involving other chromosomes was 0.3–0.4% in the cerebral cortex of the patient, identical to that of unaffected counterparts, including the trisomy of chromosome 1. Therefore, we have suggested the index case to represent low-level chromosomal mosaicism for trisomy of chromosome 1 in the brain. To confirm the assumption of mosaic trisomy 1 and exclude a derivative chromosome 1 (unbalanced translocation or a supernumerary marker chromosome), we have applied interphase chromosome-specific multicolour banding (MCB), with the probe set for chromosome 1. MCB analysis has shown that this case is a true mosaic trisomy of chromosome 1. The present case supports the hypothesis that genetic alterations associated with schizophrenia could be manifested as lowlevel chromosomal mosaicism due to genome (chromosome) instability in the brain. The work was supported by INTAS 03-51-406.
P141 INCREASED RATE OF LOW-LEVEL CHROMOSOMAL MOSAICISM IN AUTISM: AN INTERPHASE FISH SURVEY Svetlana Vorsanova1, Ivan Iourov2, Victoria Voinova-Ulas1, Natalia Gorbachevskaya2, Alexei Kolotii1, Alphia Beresheva1, Irina Demidova1, Viktor Ktavetz1, Viktor Monachov2 1 Institute of Pediatrics and Children Surgery, Moscow 2 Mental Health Research Center, Moscow Numerous cytogenetic and molecular genetic studies of individuals with autistic spectrum disorders are targeted at uncovering specific genes that may be causative for this disease. However, mosaic genetic alterations in autism remain to be studied. Through monitoring chromosome variations in a cohort of 30 individuals with autism and 30 matched controls by interphase multiprobe FISH, we have detected that 11 cases of autism (about 30%) are characterised by low-level mosaic aneuploidy involving sex chromosomes (8 cases) and autosomes (3 cases), while only one individual in the control group has low-level mosaicism due to an additional chromosome X in 3% of cells. Since numerical chromosomal changes represent genome variations that affect the whole gene set of a chromosome, it should be recognised that low-level chromosomal mosaicism could significantly affect the normal development and function of the brain without specific visible clinical and phenotypic appearance. It is difficult to reveal mosaic chromosomal alterations by standard cytogenetic and moleculargenetic studies without applying molecular cytogenetics. Our data suggest that chromosomal mosaicism due to developmental genomic instability (as hypothesised by Iourov et al., 2006) should not be excluded when attempting to determine the genetic basis of autism. This work was supported by INTAS 03-51-4060 and RGRF (Russian Federation).
P142 THE SCHIZOPHRENIA BRAIN DEMONSTRATES DISCREPANT SOMATIC CHROMOSOME PAIRING Ivan Iourov1, Svetlana Vorsanova2, Thomas Liehr3 1 Institute of Pediatrics and Children Surgery, Moscow 2 Institute of Human Genetics and Anthropology, Jena 3 Mental Health Research Center, Moscow The investigation of chromosome organisation, a little-explored epigenetic phenomenon in neuronal cells of the human brain, can bring new insight into understanding the pathogenesis of major psychiatric diseases. The present study was targeted at investigation of the positioning of interphase chromosomes (i.e. non-random chromosome association or somatic pairing) in neuronal and glial cells of the postmortem schizophrenia brain, respecting the same epigenetic processes in unaffected counterparts. Molecular cytogenetic studies of chromosomes 1, 9, 15, 16, 17, 18 and X, through the application of interphase multiprobe FISH with chromosome enumeration probes and multicolour chromosome-specific banding, has provided evidence that the schizophrenic brain differs in the rate of somatic pairing by both heterochromatic and euchromatic chromosome regions. The frequency of somatically-paired heterochromatic regions was 1.5–3 times higher for chromosomes 1, 9, 16 and 18, and for euchromatic regions of chromosomes 1 and 18 in the schizophrenic cerebral cortex, while pairing of chromosomes 15 and 17 has occurred more frequently in the non-diseased brain. Since nuclear organisation defines proper functioning of a cell, the present findings propose for the first time alterations of chromosome organisation in the brain as a possible epigenetic mechanism involved in the pathogenesis of schizophrenia. Work supported by INTAS 4. P143 SEGMENTAL UNIPARENTAL ISODISOMY ON 5q32-qter IN A PATIENT WITH CHILDHOOD-ONSET SCHIZOPHRENIA Jeffrey Seal1, Anjene´ Addington2, Michele Gornick2, Nitin Gogtay2, Philip Shaw2, Deanna Greenstein2, Peter Gochman2, Marge Coffey2, Wavrant-De Vrie`ze Fabienne3, Janet Brooks3, Zugen Chen4, Barry Merriman4, Stanley Nelson4, John Hardy3, Judith Rapoport2 1 National Institute of Mental Health-CHP, Bethesda, MD 2 Child Psychiatry Branch, NIMH, NIH, Bethesda, MD 3 Laboratory of Neurogenetics, NIA, NIH, Bethesda, MD 4 Department of Human Genetics, UCLA Schoo, Los Angeles, CA Schizophrenia is a severe mental disorder affecting approximately 1% of the world’s population. Though the etiology of schizophrenia is complex and multifactorial, with estimated heritabilities as high as 80%, genetic factors are the most compelling. Childhood onset schizophrenia (COS), defined as onset of schizophrenia before age 13, is a rare and malignant form of the illness which may have more salient genetic influence. Here we describe the first known case of paternal segmental uniparental isodisomy on 5q32-qter in a patient with COS, which adds to the previously known high rates of chromosomal abnormalities reported in this sample. Uniparental isodisomy (iUPD) is a rare genetic condition in which the offspring receives two chromosomal homologues from one parent. Segmental UPD is defined as UPD on a portion of a chromosome with biparental inheritance seen in the rest of the homologous pair. Complications due to this abnormality may arise from malfunctioning imprinted genes or homozygosity of recessive disease-causing mutations. This aberration became apparent during whole genomic screening of a COS cohort and is of particular interest because 5q has been implicated in schizophrenia by several genome-wide linkage studies and positive gene associations. This patient report, therefore, presents more evidence that some schizophrenia susceptibility gene, or genes, may be found on distal 5q. P144 TagSNP GENOTYPING OF NEUROTROPHIN’S FAMILY GENES IN EATING DISORDERS PATIENTS Josep Maria Mercader1, Mo`nica Grataco`s1, Rafael de Cid1, Anna Carreras1, Anna Puig1, Juan Ramon Gonza´lez1, Mo`nica Baye´s1, Fernando Ferna´ndez Arranda2, Xavier Estivill1 1 Center for Genomic Regulation, Barcelona 2 Psychiatric Service, Ciutat Sanitaria Be, Barcelona Increasing evidence of the participation of neurotrophins in the pathophysiology of Eating Disorders (ED) such as Anorexia Nervosa
Abstracts (AN) and Bulimia Nervosa (BN) has recently been suggested. In addition, both animal models and association studies propose BDNF and its high-affinity receptor, NTRK2, as key regulators of eating behaviour. To shed light on the potential involvement of other neurotrophins as susceptibility genes for ED, we performed a genetic case-control study for the following candidate genes: BDNF, NTRK2, NTF3, NTRK3, CNTF and CNTFR. The clinical samples included 249 eating disorders (113 AN, 136 BN) and 150 unrelated control subjects. From the HapMap project dataset, we utilised genotypes from public release 16 corresponding to CEU trios. Only SNPs having a unique mapping location on the NCBI b34 assembly and a minor allele frequency (MAF) higher than 10%, were considered for further analysis. A total of 138 Tag SNPs encompassing all candidate genes was selected using the HapMap-LDSelect-Processor. All SNPs were genotyped using SNPlex, a technology developed by Applied Biosystems, which has a current plexing level of 45–48 SNP loci in a single reaction. We performed univariate association studies using logistic regression. For each SNP computed, the odds ratios (adjusted by gender) were calculated assuming five different modes of inheritance: codominant, dominant, recessive, overdominant and additive. Supported by the Psychiatry Genetics Network (G03/184), SAF200501005, PI041257 and the National Centre of Genotyping-CEGEN funded by Genoma Espan˜a and Danone Institute.
P145 GHRELIN GENE POLYMORPHISMS Arg51Gln AND Leu72Met IN EATING DISORDERS: AN ASSOCIATION STUDY Jochen Kindler1, Ursula Bailer2, Martina De Zwaan3, Karoline Fuchs4, Friedrich Leisch5, Alexandra Strnad2, Claudia Lennkh-Wolfsberg2, Nadja El-Giamal2, Kurt Hornik3, Werner Sieghart4, Siegfried Kasper2, Harald Aschauer2 1 Medical University of Vienna, Vienna 2 Department of General Psychiatry, University of Vienna 3 Department of Psychosomatic Medicine and Psychotherapy, Erlangen 4 Center for Brain Research, Vienna 5 Department of Statistics, University of Munich 3 Department Statistics and Mathematics, University of Vienna Eating disorders (ED) are of unknown etiology. DSM-IV recognizes several subgroups (Anorexia nervosa, Bulimia nervosa, Binge eating disorders) which are considered to share common vulnerability. Ghrelin is a gastrointestinal peptide hormone that stimulates growth hormone secretion, food intake and body weight gain. The Ghrelin gene is located at 3p25-26. The present study examines two polymorphisms (G346A ¼ Arg51Gln, C408A ¼ Leu72Met) of the Ghrelin gene for association with ED. A total of 114 (111 females, 3 males; mean age: 32.35) unrelated Austrian Caucasians with a DSMIV current diagnosis of Anorexia Nervosa (n ¼ 23), Bulimia Nervosa (n ¼ 53) or Binge Eating Disorder (n ¼ 38) were identified as probands. In addition, 164 (164 females, 0 males; mean age: 32.19) unrelated healthy controls were examined. Statistical significance was set at a p < 0.05. Allele-specific multiplex PCRs were used for detection of G346A and C408A in genomic DNA. No significant difference was seen in the allele frequency of ED (G346A: 3.0%, C408A: 8.3%) and healthy controls (G346A: 2.1%, C408A: 5.5%) by Fisher’s exact test. Similarly, there was no significant difference in the genotype distribution of G346A and C408A between ED and healthy controls. Our results indicate that the Arg51Gln and Leu72Met polymorphisms of the Ghrelin gene are not associated with ED but the statistical power of our study is limited. Our findings are in line with those of Monteleone et al. (2006). Further studies should investigate subgroups of eating disorders separately.
P146 ANOREXIA NERVOSA, PERFECTIONISM AND COMMON GENETIC POLYMORPHISMS: D4 RECEPTOR (DRD4), INSULIN-LIKE GROWTH FACTOR 2 (IGF2) AND ARGININE VASOPRASSIN 1a RECEPTOR (AVPR1a) Rachel Bachner-Melman1, Ada H. Zohar2, Ilana Kremer3, Inga Gritsenko4, Elad Lerer4, Richard P. Ebstein5 1 Hebrew University of Jerusalem, Jerusalem 2 Ruppin Academic Center, Emek Hefer 3 HaEmek Hospital, Afula 4 Herzog Hospital, Jerusalem 5 Hebrew University of Jerusalem, Herzog H, Jerusalem
771
We tested association between the dopamine D4 receptor (DRD4) and anorexia nervosa (AN) in a group of 169 AN trios and 418 non-clinical families (542 daughters). We used UNPHASED for 5 DRD4 polymorphic loci: 3 promoter region SNPs (-C521T, -C616G & –A809G), the 120 bp promoter region tandem duplication and the exon III repeat. We also tested these 5 loci for association with the Child and Adolescent Perfectionism Scale (CAPS) in both the clinical and non-clinical groups. Single locus analysis showed significant association (p ¼ 0.009) between the ‘C’ –C521T allele and AN. Multi-loci (haplotype) analysis also showed significant association, particularly a 4 locus haplotype (exon III-120 bp repeat- C521T-A809G; global p value ¼ 0.0008). A single haplotype was most significant (exon III 4 repeat, 120 bp single repeat, ‘C’ –521 SNP & ‘A’ –809 SNP; p ¼ 0.0017). 3 of these 4 loci have been shown to increase DRD4 transcription. Association was also observed between DRD4 and CAPS scores for both the AN and the nonclinical subjects. Two additional genes, the insulin-like growth factor 2 (IGF2) and the arginine vasopressin 1a receptor (AVPR1a) that we have previously shown to be associated with dysfunctional eating behaviour are also associated with CAPS scores (IGF2 ApaI SNP: p ¼ 0.001; AVPR1A RS3 microsatellite: p ¼ 0.01). Results suggest that one pathway to eating pathology is mediated by the impact of variations in DNA sequences on perfectionism. P147 A NOVEL MULTIVARIATE METHOD FOR THE ANALYSIS OF PET DATA AND BEHAVIORAL TRAITS Tiffany Greenwood, Walter Kaye, Nicholas Schork University of California, San Diego, La Jolla, CA Novel statistical methods are needed to model the complex manner in which neuronal circuits interact with behaviour and genetics. We describe a method for testing the relationship between variation in a similarity matrix and ancillary information collected on a sample of individuals that obviates the need for cluster analysis by testing more global hypotheses about the patterns of similarity in the matrix. This method is also the perfect companion for heat map and tree-based representations of high-dimensional data organized by some feature or grouping factor meant to reveal the relationship between the variables used for their construction. We have applied this method to a sample of 16 control women and 17 subjects recovered from bulimia nervosa for whom the following data were collected: 1) 5-HT1A receptor binding in frontal–limbic regions of the brain (the prefrontal cortex, orbital frontal cortex, lateral orbital frontal, subgenual cingulate, and amygdala) as assessed by positron emission tomography (PET); 2) Behavioural measures including the Temperament and Character Inventory (TCI) and the Barratt Impulsivity Scale (BIS); and 3) LS/SS variants of the human 5-HTT gene (SLC6A4). We have found that selftranscendence from the TCI and interpersonal behaviour from the BIS explain significant portions of matrices constructed from 5-HT1A receptor density in the limbic system in normal controls (34.4%, p ¼ 0.016) and recovered bulimics (35.7%, p ¼ 0.003), respectively. Variants of the 5-HTT gene were not informative for either 5-HT1A receptor binding or behaviour. P148 GENETIC ASSOCIATIONS WITH REACTION TIME VARIABILITY IN AD/HD Scott Kollins1, Jeff Epstein2, Simon Tonev3, Arthur Anastopoulos4, Ave Lachiewicz1, David Fitzgerald1, Erin Kane4, Melanie Kail1, Michael Cuccaro1, Linda Exelbierd1, John Gilbert1, Allison Ashley-Koch1 1 Duke University, Durham, NC 2 Cincinatti Children’s Hospital Medical Center, Cincinatti 3 Duke Clinical Research Institute, Durham, NC 4 University of North Carolina, Greensboro, Greensboro, NC Reaction time (RT) and RT variability on neuropsychological tests have been proposed as endophenotypes for AD/HD since AD/HD children are usually slower and more variable in their RTs than nondiagnosed children. We examined genetic associations with four measures from the Conners’ continuous performance task (CPT): commission errors, mu, sigma and tau. The latter three parameters are derived from an exGaussian distributional model and represent the mean (mu) and standard deviation (sigma) of the normal portion of the RT distribution, while tau is a descriptor of the exponential portion of the distribution, representing any positive skew in the RT distribution. CPT data were collected for 53 nuclear families with at least one child aged 5 to 12 years who met DSM-IV criteria for AD/HD. We genotyped a series of
772
Abstracts
candidate genes involved in neurotransmitter pathways using a haplotype tagging approach. Data analysis using the QTDT suggests that commission errors and mu are associated with the same genetic markers in BDNF (p < 0.05 for rs1048221, rs7124442 and rs7934165). Also, tau is associated with genetic variation in DBH (p < 0.05 rs2073833, hcv3274654, rs1108580, hcv2253949 and rs1541333). The association of tau with DBH is particularly exciting as hcv2253949 is a coding SNP. These data suggest that measures of CPT performance, especially those that capture the non-normal distribution of RT, are associated with genetic variation and that such endophenotypes maybe useful for better understanding the biological mechanisms that contribute to the pathophysiology of AD/HD. P149 SEARCHING FOR DOPAMINE-RELATED ENDOPHENOTYPES IN A CHILD PSYCHIATRIC SAMPLE Zsofia Nemoda1, Zsanett Tarnok2, Judit Gervai3, Eva Kereszturi1, Julia Gadoros2, Maria Sasvari-Szekely1 1 Semmelweis University, Budapest 2 Vadaskert Child and Adolescent Psychiatr, Budapest 3 Hungarian Academy of Sciences, Budapest Attention Deficit Hyperactivity Disorder (ADHD), Tourette-syndrome (TS) and Obsessive Compulsive Disorder (OCD) are often present as comorbid conditions, therefore presuming a common neurobiological and genetic origin. Family and twin studies have shown considerable heritability in these child psychiatric disorders. Since dysfunction of the basal ganglia and the prefrontal cortex (thalamico-cortical loops) is hypothesised in these neurodevelopmental disorders, most of the candidate genes belong to the dopaminergic systems. Recently, arbitrary diagnostic categories are being replaced by dimensional intermediate phenotypes in genetic association analyses. Neurocognitive tests assessing prefrontal functions are major endophenotype candidates. In this study, we performed a comparative genetic analysis in order to specify certain dopaminergic system-related endophenotypes in the background of ADHD, TS, and OCD (total n ¼ 165). The following neurocognitive tests were administered: Stroop, Wisconsin Card Sorting, Trail-making A, B, Semantic Fluency, and Digit Span forward and backward. The investigated polymorphisms included the dopamine D4 receptor (DRD4) gene 48 bp VNTR, the dopamine transporter (DAT) gene 40 bp VNTR in the 30 region and the catechol-Omethyltransferase (COMT) gene Val158Met polymorphism. In our clinical child population we did not observe any important effects of the COMT and DAT polymorphisms. The 7-repeat allele (7-present genotype category) was associated with Trail-making in children with ADHD diagnosis. 7 patients performed better on the Trail-B test. P150 NEUREGULIN-1 RISK GENOTYPE AND EYE MOVEMENTS IN SCHIZOPHRENIA Hannes Petursson1, Magnus Haraldsson1, Ulrich Ettinger2, Brynja Magnusdo´ttir1, Engilbert Sigurdsson1, Thordur Sigmundsson1 1 Landspitali University Hospital, Reykjavik 2 Institute of Psychiatry, London Aim: To investigate the potential effects of the 5 SNP at-risk neuregulin-1 (NRG-1) haplotype on antisaccade (AS) and smooth pursuit (SP) eye movements in schizophrenia patients and healthy controls. Background: Recent studies provide evidence for an association between variations in the NRG-1 gene on chromosome 8p and the risk of schizophrenia. The phenotypic heterogeneity of schizophrenia complicates linkage and candidate gene studies. By studying endophenotypes, it may be possible to circumvent this problem. Previous studies have demonstrated impaired AS and SP eye movements in patients with schizophrenia. Findings: That first-degree relatives of schizophrenia patients have significant SP and AS impairments indicates that the latter may be linked to schizophrenia susceptibility genes. Methods: Analyses included 72 medicated patients with schizophrenia and 63 healthy controls. Eye movements were measured using infrared oculography. Results: For AS and SP data, the main group effect (patients vs. controls) was significant, but the main effect of genotype and group by genotype interaction was not significant. Conclusions: Preliminary results do not show a significant effect of the 5SNP at risk core haplotype of the NRG-1 gene on SP and AS eye
movements. However, the power of the sample to identify small- to medium-sized effects is limited, and the risk of type II error must be kept in mind. Therefore, it is probably premature to claim that findings either support or do not support the assertion that patient and control carriers perform worse than non-carriers. P151 CHRNA7 IS A COMMON SUSCEPTIBILITY GENE FOR COGNITIVE DEFICITS IN SCHIZOPHRENIA AND BIPOLAR DISORDER Ana Barabash1, Ine´s Ancı´n2, Blanca Va´zquez-Alvarez2, Jose Luis Santos3, Eva Sa´nchez-Morla3, Natalia Alvarez2, David Jime´nez2, Magdalena Marin˜o2, Maria Jose´ Bescos2, Isabel Argudo2, Sergio Torrijos2, Raquel Rodriguez2, Juan Jose´ Lo´pez-Ibor2, Jose´ Antonio Cabranes2 1 Hospital Clinico San Carlos, Madrid 2 Hospital Clı´nico San Carlos, Madrid 3 Hospital Virgen de la Luz, Cuena Biochemical and genetic studies suggest that the nicotinic cholinergic system is involved in the pathophysiology of schizophrenia (SCZ) and bipolar disorder (BD), particularly in cognitive and sensory abnormalities. We have examined the association of the 86C/T promoter variant of the CHRNA7 gene with SCZ and BD and with the presence of: deficit inhibition of P50 in evoked potentials, deficit in sustained attention in the Continuous Performance Test, verbal memory impairment in the California Verbal Learning Test (CVLT) in BD and SCZ or history of psychotic symptoms in BD. We have studied 73 BD, 88 SCZ and 67 controls. The distribution of genotypes considering the presence or absence of the variant allele T was significantly different between patients and controls (X2 ¼ 6.085;p ¼ 0.031). Risk associated with carrying at least one allele T was 2.9 (1.05–7.98) for BD and 3.45 (1.22–9.73) for SCZ. Moreover, we found a significant association between genotypes with the variant allele and worse performance on the following items of CVLT: learning (p < 0.047) and short delay cued recall (p < 0.015) in controls and in SCZ (p < 0.009 and p < 0.040); short delay recall (p < 0.043) in SCZ and impaired perseveration (p < 0.010) in BD. We found no association with sensory gating, sustained attention or history of psychotic symptoms. We concluded that CHRNA7 may be a common susceptibility gene for a shared cognitive endophenotype in BD and SCZ. P152 NEUREGULIN-1 IS ASSOCIATED WITH SLOWER COGNITIVE PROCESSING IN SCHIZOPHRENIA Elvira Bramon1, Emma Dempster2, Sophia Frangou1, Muriel Walshe1, Colm McDonald1, Pak Sham1, David Collier1, Robin Murray1 1 Institute of Psychiatry, London 2 Toronto Western Research Institute, Toronto Background: Neuregulin-1 (NRG1) is one of the best-supported susceptibility genes for schizophrenia (Stefansson et al., 2002; 2003; Li et al., 2006). Investigating the effects of NRG1 on disease components or endophenotypes will help us to understand the role of this gene in psychosis (Gottesman and Gould, 2003; Harrison and Weinberger, 2005). Objective: To investigate the association between Neuregulin-1 and the P300 wave, a neurophysiological endophenotype of schizophrenia. Methods: Participants were 62 patients with schizophrenia or schizoaffective disorder, 94 of their unaffected relatives and 33 unrelated controls. The P300 wave was extracted from the electroencephalogram whilst subjects conducted a two-tone discrimination task. The effect of three markers from the core NRG-1 at-risk haplotype (Stefansson et al., 2002), including single nucleotide polymorphism SNP8NRG221533 and two microsatellites (478B14-848 and 420M91395), on P300 amplitude and latency was examined using multilevel modelling. Results: There was no overall association between NRG-1 and P300 amplitude. The two microsatellites had no influence on P300 latency. However, the SNP marker SNP8NRG221533 had a significant dosedependent influence on P300latency, with a higher number of C alleles associated with a greater latency delay [Coef. ¼ 32.4 ms; 95%CI: 13.2 to 51.6 ms; p < 0.01]. Conclusions: The P300 latency is a psychosis endophenotype reflecting the speed of neural transmission. We hypothesise that variation in NRG1 conveys risk for schizophrenia by disrupting neural connectivity and leading to a slower speed of cognitive processing.
Abstracts P153 NEUROPSYCHOLOGICAL BATTERY ASSESSMENT IN CUBAN SCHIZOPHRENIC PATIENTS AND THEIR FIRST DEGREE RELATIVES. A FAMILY STUDY Rau´l mendoza Quin˜ones Centro de Neurociencias de Cuba, Ciudad de la Habana There is increasing evidence that the use of endophenotypes or vulnerability markers may enhance the discriminative capacity of linkage and association studies. Cognitive dysfunctions are potential endophenotypes of schizophrenia. Despite the limitations of the neuropsychological battery assessments available at present, previous studies have shown that many neuropsychological functions are impaired in schizophrenic patients and their first-degree relatives, suggesting a genetic substrate for the illness. The purpose of this study was to examine the presence of familial association in a Cuban sample of schizophrenic patients and their relatives in a family study. Methods: 60 patients with a DSM-IV diagnosis of schizophrenia, 97 first-degree non-affected relatives and 72 normal comparison subjects were included in the study. All subjects were administered a comprehensive neuropsychological test battery to explore executive functioning, working memory, selective attention and perception. ANOVAglobal comparisons were used to find differences betweens the groups. Results: As expected, schizophrenic patients showed more impairment on neuropsychological tests than first-degree relatives and controls. Data has also shown that patients and their relatives performed worse than control subjects on negative emotion recognition tasks and executive functions as assessed by the Wisconsin Card Sorting Test (WCST). These data strongly suggest that executive function emotion recognition impairments indicate a genetic liability for schizophrenia. P154 A TWIN STUDY OF GENETIC AND ENVIRONMENTAL INFLUENCES ON SENSORY GATING Andrey Anokhin1, Andrey Vedeniapin2, Andrew Heath2, Oleg Korzyukov3, Nashaat Boutros3 1 Washington Un. School of Medicine, St. Louis 2 Washington University School of Medicine, St. Louis, MO 3 Wayne State University School of Medicin, Detroit, MI Sensory gating abnormalities indicated by poor suppression of P50 auditory-evoked potential has been implicated in the biological bases of schizophrenia and some other psychiatric disorders and proposed as a candidate endophenotype for genetic studies. In addition, it has been shown that gating deficits in schizophrenics extend to event-related brain potential (ERP) components associated with early attentive processing stages (the N1/P2 complex). However, evidence for heritability of these sensory gating indices in the general population is very limited. This study examined heritability of P50, N1, and P2 amplitudes and the corresponding gating indices using a dual-click auditory paradigm with ERP recording. A community-based sample of twins included 54 monozygotic and 55 dizygotic pairs. Main outcome measures were amplitudes and gating indices of P50, N1 and P2 ERP components. Genetic model-fitting analysis showed high heritability (58–76%) of peak amplitudes of P50, N1, and P2 waves. Genetic influences on P50 gating (S2/S1) were modest (36%), while heritability of N1 and P2 gating was high and significant (57 and 54%, respectively). The alternative gating measure (S1-S2 difference) showed significant heritability (41–70%) for all three ERP components. Relatively weak genetic influences on P50 gating can be related to its poor test-retest reliability, demonstrated in previous studies. These results suggest that gating measures derived from the N1/P2 wave complex may be more suitable endophenotypes for population-based genetic studies of the sensory gating function and its impairments in psychopathology. P155 DATA MINING APPROACH TO IDENTIFY ENDOPHENOTYPES OF COMPLEX TRAITS WITH ILLUSTRATIVE APPLICATION TO SCHIZOPHRENIA Cathy S.J. Fann1, Amy C.L. Hsu1, Chee-Jen Chang2, Hai Kwo Hwu3 1 Academia Sinica, Taipei 2 Chang-Kung University, Taipei 3 Taiwan University, Taipei A complex trait results from a combination of genetic and environmental effects and their interactions. In general, genetic effects are
773
mild to moderate, so that it is difficult to detect disease susceptibility genes among thousands of markers, many of which may yield a weak positive signal. Endophenotypes are thought to be surrogates through which genes influencing the corresponding traits may be localised and characterised. This study proposes to identify endophenotypes of complex traits using data mining tools such as Classification and Regression Trees, which clarify the heterogeneity among populations or differences in the relationships between predictors and a response in different parts of the measurement space. The power and specificity of the proposed method were evaluated by simulation studies. We used data from the Schizophrenia family study to illustrate our approach. We found that data mining may provide an effective method to identify endophenotypes of complex traits while adjusting for genetic factors, environmental factors and variables that do not correlate with the disease. P156 POLYMORPHISM OF THE HUMAN PBR GENE IN RELATION TO STRESS RESPONSES Yurie Nakamoto1, Kazuhiko Nakamura2, Go Mugishima3, Mitsuko Sato1, Masako Miwa1, Mitsunobu Yoshii1 1 Tokyo Institute of Psychiatry, Tokyo 2 Hamamatsu Univ Sch Med, Hamamatsu 3 Fukuoka Pref Univ, Fukuoka Peripheral-type benzodiazepine receptors (PBR) are involved in the production of various steroids, including stress hormones. It has been shown that PBR are increased after acute stress and decreased under chronic stressful conditions. In our previous studies, expression of platelet PBR was associated with anxiety traits in normal human subjects (Nakamura et al., Psychopharmacology 162: 301–303, 2002), and probably related to polymorphism of the PBR gene (Nakamoto et al., Soc Neurosci Abstr 29: 641.3, 2003). In the present study, we have further analysed these relationships in 162 normal human subjects. Single nucleotide substitution 485G > A (Arg162His) was identified in the coding region of the human PBR gene, showing G/G (61.1%), G/A (32.1%) and A/A (6.8%). In male subjects, the density of platelet PBR showed a significantly different variance among genotypes (G/G > G/ A > A/A), presumably representing the order of stress sensitivity. In animal studies using Slc Wistar rats, chronic social isolation (for 12 weeks after weaning) caused a marked increase in the densities of platelet PBR, which was observed exclusively in males (p < 0.05). The results suggest that males are more sensitive than females to environmentally-induced stress, as far as PBR responses are concerned. The G allele in 485G > A polymorphism of the human PBR gene may be a risk factor for stress, as in the case of panic disorders (Nakamura et al., Am J Med Genet B Neuropsychiatr Genet 141(3) 222–226, 2006). P157 AN IMAGING GENOMICS APPROACH TO ENDOPHENOTYPE ANALYSIS OF PSYCHOSIS: GENETIC ANALYSIS OF VERBAL FLUENCY, EMOTIONAL FACE RECOGNITION AND FRACTIONAL ANISOTROPY OF WHITE MATTER Diana Prata1, Andrea Mechelli2, Marta DiForti2, David Dempster2, Cindy Fu2, Marco Picchioni2, Sukhi Shergill2, James Woolley2, Sridevi Kalidindi2, Colm MacDonald2, Arsime Demjaha2, Gerome Breen2, Robin Murray2, Phillip McGuire2, David Collier2 1 Institute of Psychiatry of London, London 2 Institute of Psychiatry, London A potential advantage of the endophenotype approach to genetic analysis of complex disorders is that smaller samples can be used to assess the functional impact of genetic variation on the brain, since genes may have a more direct relationship with endophenotypes than with categorical disease diagnosis. Furthermore, it may provide valuable information on the metabolic consequences of genetic variation. Functional or structural neuroimaging data, such as, respectively, the physiological responses during specific information processing assessed by fMRI or the white matter integrity in DTI, can be used as endophenotypes. This study combined fMRI and DTI with molecular genetics to investigate the aetiology of psychosis. The functional endophenotypes used were verbal fluency (VF) and emotional face recognition tasks (EFR), and the structural measurement used was fractional anisotropy (FA) of white matter. We genotyped a sample of 79 controls and 85 patients (and 28 at-risk)
774
Abstracts
scanned for VF, 53 controls and 44 patients (and 17 at-risk) scanned for EFR and 129 controls and 105 patients with psychosis (and 17 at-risk) who underwent DTI. The patient group included subjects with DSMIV schizophrenia or bipolar disorder, with psychotic symptoms that were diagnosed and scanned at the Institute of Psychiatry of London during the past five years. We examined the correlation of the observed endophenotypic differences within these disorders with the genetic variation in putative schizophrenia susceptibility genes such as Dysbindin 1, COMT and 5-HTT, which will contribute to the understanding of the brain metabolic pathways putatively relevant to Psychosis. P158 GENETIC FACTORS PREDICTING EXECUTIVE FUNCTIONING IN HEALTHY SUBJECTS AND OBSESSIVE COMPULSIVE SPECTRUM DISORDER PATIENTS Maria Cristina Cavallini1, Paolo Cavedini2, Claudia Zorzi2, Monica Piccinni2, Antonia Stazi3, Laura Bellodi2 1 Universita` Vita-Salute San Raffaele, Italia 2 Universita` Vita-Salute San Raffaele, Milano 3 Registro Italiano Gemelli- Istituto Suer, Roma Specific patterns of altered executive neurofunctioning have been described for Obsessive Compulsive (OCD) spectrum disorders. Our aims are to characterise genetic contributions to these patterns in the general population and to evaluate whether the altered neurofunctional profile of patients, as measured by neuropsychological assessment, might rely on a specific genetic liability related to a neurobiological basis. In additional to clinical diagnoses, subjects will be characterised on the basis of a neurofunctional profile that could help in identifying the genetic mechanisms determining specific patterns of modification of pathological behaviour in OCD patients, in relationship to drug or behaviour therapies. Considering the neurofunctional profile as an endophenotype, genetic information will be used to achieve better knowledge of the functioning and dysfunctioning of brain circuits involved in the patho-physiology of the disorders, and therefore to better understand related mental processes. Four hundred pairs of twins (Italian Twin Registry) from the general population underwent neuropsychological testing as a measure of neurocognitive performance of the executive functions (i.e. decision-making, problem solving, self-shifting ability) in the general population. The same tests will be performed on 300 OCD probands and their first degree relatives. Co-transmission of impaired neurocognitive performance and disorders will be assessed. We anticipate the definition of genetic contribution to neurofunctional pathways and then identification of a common neurofunctional alteration for OC spectrum disorder, which could be directly related to a specific genetic liability.
the serotonergic receptors polymorphisms studied in our Brazilian sample. P160 VOXEL-BASED MORPHOMETRIC ANALYSIS OF GRAY MATTER IN FIRST-EPISODE SCHIZOPHRENIA AND THEIR UNAFFECTED RELATIVES: A 3T MR INVESTIGATION Wei Deng1, Xiao-hong Ma2, Su Lui3, Li-jun Jiang2, He-han Tang3, David Collier4, Qi-yong Gong3, Tao Li4 1 Psychiatric laboratory, Chengdu, Sichuan Province 2 Psychiatric lab & Department of Psychiatry, Chengdu, Sichuan Province 3 HMRRC, Department of Radiology, West Chi, Chengdu, Sichuan Province 4 Division of Psychological Medicine, Inst, London The aim of the study is to explore the alterations of Voxel-based morphometry (VBM) in schizophrenia and their unaffected relatives and to reveal potential endophenotypes by assessing brain morphometry. 13 first-episode, treatment-naive schizophrenic patients (6 males and 7 females), 20 first-degree non-psychotic relatives of the patients (9 males and 11 females) and 33 age-, gender- and handednessmatched healthy controls (13 matched with patients and 20 matched with relatives) were recruited. High resolution T1-weighted images were acquired using a 3T MR imaging system (GE), with an 8 channel phase array head coil. Images were analyzed by SPM2. Optimized VBM protocol, including creation of customized templates, segmentation, normalization, and spatially smoothing were adopted. In comparison with the controls, GMV of patients showed significant reduction in widespread cortical regions, including right amygdaloid nucleus, superior, middle and inferior frontal gyri, inferior temporal gyrus, parietal lobe, occipital lobe and cerebellum, left middle frontal gyrus, superior and inferior temporal gyri, middle cingulate gyrus, inferior insular lobe and parietal lobe. When comparing relatives with control groups, we observed reduced GMV localized to the bilateral prefrontal lobes, temporal lobes and right anterior cingulate gyrus in the male relatives. However there were no statistical difference between female relatives and controls. In conclusion, there were cortical reduction in cingulate gyrus, frontal lobe and temporal lobe of patients with schizophrenia and their unaffected male relatives, which could be partly due to a shared genetic liability. Funded by NSFC, Wellcome Trust, NARSAD.
P159 SEROTONERGIC POLYMORPHISMS AND NEUROENDOCRINE RESPONSE TO INTRAVENOUS CITALOPRAM IN OCD Fabio Corregiari, Quirino Cordeiro, Stevin Zung, Marcio Bernik, Homero Vallada University of Sa˜o Paulo, Sa˜o Paulo
P161 IS CEREBRAL ASYMMETRY ASSOCIATED WITH SEROTONIN TRANSPORTER IN PSYCHOSIS? Pietro Papili1, Katherine J. Aitchison2, Kevin Morgan3, Paul Fearon3, Craig Morgan3, Julia Lappin3, Marta Di Forti3, Monica Aas3, Alexander Butt3, Joanna Eyeson3, Caterina La Cascia3, Ana Miorelli3, Valeria Mondelli3, Serena Navari3, Carmine M. Pariante3, Logos Curtis4, Collier David2, Robin M. Murray3, Paola Dazzan3 1 Istituto di Psichiatria ‘‘P.Ottonello’’, Bologna 2 MRC SGDP Centre and Clinical Neuropharm, London 3 Division of Psychological Medicine, Inst, London 4 MRC SGDP Centre, Institute of Psychiatry, London
Introduction: The acute blockade of the serotonin transporter with intravenous (IV) infusion of serotonin reuptake inhibitors provokes a rapid increase of the serotonin levels in the central nervous system, and a stimulation of the thalamic receptors with the release of various hormones, such as prolactin. This kind of neuroendocrine test has revealed a blunted hormonal reaction in patients with obsessive-compulsive disorder (OCD). Thus serotonergic receptors genes may influence the neuroendocrine test in OCD patients. Methods: A sample of 56 DSM-IV patients and 30 matched healthy controls were studied in a test of neuroendocrine stimulation with infusion of IV citalopram and the measurement of plasmatic prolactin level. Four serotonergic receptors polymorphisms were analysed and compared with prolactin release between the groups of patients and controls: T102C polymorphism of the 5-HT2A gene; C516T polymorphism of the 5-HT2A gene; Cys23Ser polymorphism of the 5-HT2C gene; andG861C polymorphism of the 5-HT1D beta gene. Results: The group of OCD patients presented a blunted prolactin response. However there was no association between the hormonal reaction and the polymorphisms investigated. Conclusions: The patients with OCD present a blunted prolactin reaction related to the IV citalopram test, which is not associated with
Background: Reduced structural cerebral asymmetries have been reported in schizophrenic patients and in their relatives. A candidate for the regulation of cerebral asymmetry is the serotoninergic system. There is evidence that serotonin has influences on brain development in mammals (Esaki, 2005) and that serotonin transporter function is an early step in left-right patterning (Fukumoto, 2005). Furthermore a metanalysis suggests an association between schizophrenia and the STin2.12 allele of the intron2 VNTR of the serotonin transporter (Fan, 2005). We therefore hypothesised that in psychotic patients reduced cerebral asymmetry is related to the presence of STin2.12 allele. Methods: We evaluated 92 patients (38% female) at their first episode of psychosis (affective psychosis 28%, schizophrenia/schizophreniform 47%, other psychoses 25%) and 84 healthy controls (42% females). Volumetric measurements of prefrontal, premotor, sensorimotor, occipitoparietal and temporal regions were obtained from 1.5T MRI images. In a subsample of 54 patients and 36 controls DNA was extracted by standard procedures and the serotonin transporter intron2 VNTR was genotyped by fluorescent technology. Results: Patients showed a significantly reduced asymmetry of the temporal region compared to healthy controls (p ¼ 0.03). We are
Abstracts currently exploring whether the STin2.12 allele is associated with this altered pattern of cerebral asymmetry. Conclusions: Temporal asymmetry is reduced in psychotic patients. It is possible that risk genes for psychosis are associated with genes regulating cerebral asymmetries. Acknowledgements: This study is funded by the UK Medical Research Council, Stanley Medical Research Institute and South London and Maudsley NHS Trust. P162 NEUROCOGNITIVE DEFICITS IN FIRST-EPISODE SCHIZOPHRENIC PATIENTS AND THEIR FIRST-DEGREE RELATIVES Xiaohong Ma1, Qiang Wang2, P.C. Sham3, Xiehe Liu2, S. Rabe-Hesketh3, Xueli Sun2, Huaqing Meng3, Wei Chen3, EYH Chen4, Wei Deng2, R.C.K. Chan4, R.M. Murray5, D.A. Collier3, Tao Li3 1 Psychiatric Laboratory, Chengdu 2 Psychiatric Lab and Department of Psychiatry, Chengdu, Sichuan 3 Division of Psychological Medicine, SGDP, London 3 Department of Biostatistics and Computin, London 3 Department of Psychiatry, The First Affi, Chongqing 4 Department of Psychiatry, University of Hong Kong 5 Division of Psychological Medicine, IOP, KCL, London Some neuropsychological abilities, particularly those affecting memory, attention and executive function, are impaired amongst both schizophrenic patients and their unaffected relatives, implying that these deficits are at least partly genetic in origin. The objective of this study was to determine whether selected neurocognitive abilities are impaired in first-onset schizophrenic patients and their relatives compared to controls. We examined attention and speed of information processing, memory and learning, verbal function, visuoconstructive abilities and executive function in 207 first-episode schizophrenic patients (163 of whom were drug naive), 322 of their first-degree relatives and 133 unrelated normal controls. The data were subjected to multilevel modelling to compare neurocognitive performance between schizophrenic probands, relatives and controls while taking into account potential correlations among members of the same family; age, gender and years of education were included as covariates. Of the three groups, schizophrenic patients performed poorest on all neuropsychological tests, suggestive of a broad range of neurocognitive deficits. Their first-degree relatives showed a narrower pattern of poor performance on Digit Symbol, Digit Span, Trail Making, Verbal Fluency Test, Tower of Hanoi and WCST-M tests. Our findings show that selected neurocognitive deficits, especially attention and executive function, are impaired in the families of schizophrenic patients. These patterns of neurocognitive deficits may represent 8endophenotypes denoting varying degrees of vulnerability to schizophrenia. This work was supported by funding from the Chinese National Natural Science Foundation (TL), the Wellcome Trust (TL, DAC, PS), NARSAD (TL), and the Schizophrenia Research fund (TL). P163 TEMPORAL AND HEMISPHERE SPECIFICITY OF COMT POLYMORPHISM AND PAIRED-CLICK GATING Jose M. Canive1, Brett L. Lu2, Ashely K. Smith3, J. Christopher Edgar2, Aaron P. Jones3, Carolyn Albers3, Stephen F. Lewis2, Mingxiong Huang3, Michael Escamilla2, Gregory A. Miller3 1 New Mexico VA & University of New Mexico, Albuquerque, NM 2 Department of Psychiatry, University of Albuquerque, NM 3 Biomedical Research Institute of New Mexico, Albuquerque, NM 3 Department of Radiology, San Diego VA, San Diego, CA 3 Department of Psychology and Beckman Ins, Urbana, IL An endophenotype of schizophrenia, a gating deficit in a paired-click paradigm is shown as an abnormally high ratio of evoked testing (S2) amplitude to conditioning (S1) amplitude, measured at Cz around 50 ms post-stimulus (P50) using electroencephalography (EEG). We recently detected an effect of the Catechol-O-Methyltransferase (COMT) Val158Met polymorphism on P50 gating, with the Val homozygotes, associated with the most prefrontal (PFC) synaptic dopamine degradation, exhibiting poorest gating. To examine the temporal and hemisphere specificity of the COMT/gating relationship, paired-click gating was localized to left and right primary auditory cortex (the main generators of the P50 response at Cz) using magnetoencephalography (MEG) source modeling, at both 50 ms (M50) and 100 ms (M100). Paired-click task and COMT genotyping
775
were performed on controls (n ¼ 25) and patients with schizophrenia (n ¼ 41) to derive M50 and M100 gating in bilateral primary auditory cortex. Consistent with past reports, the schizophrenia group exhibited poorer left M50 and bilateral M100 gating. A Diagnosis X Genotype ANOVA detected an effect of COMT on left M100 gating (p ¼ 0.04, partial 2 ¼ 0.12), with the Val/Val homozygotes having the worst gating. Exploratory analysis suggested that the effect of COMT on left M100 gating is greater in the control group via its direct effect on the S2 response. Findings suggest a top-down, PFC dopamine-mediated effect on left-hemisphere gating, lending further support to a lateralized frontal-temporal functional disconnection in information filtering as a hallmark of schizophrenia. P164 POLYMORPHISMS IN THE 5-HTT GENE AND PITUITARY VOLUME IN FIRST EPISODE PSYCHOSIS Serena Navari1, Carmine M. Pariante1, Katherine J. Aitchison2, David Collier3, Peter McGuffin3, Logos Curtis3, Marta DiForti1, Monica Aas1, Joanna Eyeson1, Ana Miorelli1, Valeria Mondelli1, Pietro Papili1, Alex Butt1, Craig Morgan1, Paul Fearon1, Julia Lappin1, Kevin Morgan1, Robin M. Murray1, Paola Dazzan1 1 Institute of Psychiatry, KCL, London 2 MRC SGDP Centre, KCL; Institute of Psychiatry, London 3 MRC SGDP Centre, KCL, London Background: Subjects at their first psychotic episode show hyperactivity of the hypothalamic-pituitary-adrenal axis (HPA) during the acute phase of their illness, which may reflect an activation of the hormonal stress response. This activation of the HPA axis may induce an enlargement of the pituitary gland. Stress vulnerability has been related to the presence of polymorphisms relevant to serotonergic neurotransmission; in particular, the short allele of serotonin transporter promoter linked polymorphic region (5-HTTLPR) seems to influence stress reactivity in depressed patients. It remains unclear whether, in psychotic patients, the enlargement of the pituitary is associated with a genetic liability to stress reactivity. We evaluated the relation between pituitary volume, as an index of the HPA-axis activation, and polymorphisms in the 5-HTTLPR. Methods: 78 patients at their first episode of psychosis (mean age: 27.8 8.6 years; 36.2% female) and 78 age- and gender-matched controls. Pituitary volume was measured using 1.5 mm, coronal, 1.5 T, high resolution MRI images. DNA was extracted from peripheral blood by standard procedures and the 5-HTTLPR including a SNP within this region was genotyped by techniques developed within our laboratories. Results: Patients had a significantly larger pituitary volume than controls (22%, p < 0.001). This enlargement was independent from pharmacological treatment. We are currently exploring whether, in our sample, differences in pituitary volume were mediated by polymorphisms in the 5-HTTLPR. Acknowledgments: UK Medical Research Council; Stanley Medical Research Institute; South London & Maudsley NHS Trust. P165 ASSOCIATION OF MEDIAL FRONTAL HYPOACTIVATION WITH PSYCHIATRIC FAMILY HISTORY IN PATIENTS WITH PANIC DISORDER: A NIRS STUDY Yuji Okazaki1, Yukika Nishimura2, Ken Inoue3, Hisashi Tanii2, Hisanobu Kaiya3, Masato Fukuda4 1 Metropolitan Matsuzawa Hospital, Setagaya-ku 2 Mie University Graduate School of Medicine, Tsu, Mie 3 MIe University Graduate School of Medicine, Tsu, Mie 3 Warakukai Clinical Research Center of Panic Disorder, Nagoya & Tokyo 4 Gumma University Graduate School of Medicine, Maebashi, Gumma Near-infrared spectroscopy (NIRS) is an optical imaging technique that allows non-invasive in vivo measurements of changes in the concentration of oxygenated ([oxy-Hb]) and deoxygenated ([deoxy-Hb]) hemoglobin that reflect activation of particular brain regions. One hundred and nine patients with panic disorder and 96 normal controls were measured during a word fluency task by a multi-channel NIRS machine (HITACHI ETG 4000) placing 33 probes on the frontal region. Patients with panic disorder showed significantly less activation in all channels compared with normal controls without lateralization effect. Furthermore, patients with family history of panic disorder (N ¼ 9) in the first degree relatives demonstrated very significant hypoactivation in medial frontal region compared with the rest of the patients
776
Abstracts
(N ¼ 100). Patients with family history of panic disorder, depression and psychiatric treatments (N ¼ 26) showed similar results of medial frontal hypoactivation as compared with the rest of the patients (N ¼ 83). There was no significant difference between patients with and without panic disorder/family history concerning age, sex ratio, age of onset, performance of the word fluency task, scores of PDSS-J and SDS. Medial frontal dysfunction in patients with panic disorder seems to be associated with family history of panic disorder/related psychiatric disorders.
P166 GENETICS OF SCHIZOPHRENIA: SEGREGATION ANALYSIS OF ENDOPHENOTYPES Maartje Aukes, Behrooz, Z. Alizadeh, Margriet, M. Sitskoorn, Jean-Paul Selten, Richard, J. Sinke, Rene´, S. Kahn UMC Utrecht, Utrecht Endophenotyping is a promising strategy for genetic research on complex psychiatric disorders as endophenotypes improve power of a study while they may reduce genetic architecture. In the present study, we investigated heritability and genetic transmission patterns of several endophenotypes including verbal memory, set shifting and fine motor functioning in Dutch families with schizophrenia. Thirty highrisk families including 138 subjects (15 probands and 123 of their relatives) were administered the CVLT delayed recall task, the Trailmaking task and the Purdue pegboard task. A segregation analysis was performed on the age and sex adjusted levels using the SEGREG module of S.A.G.E. We found significant evidence of oligogenic inheritance of the three studied endophenotypes. We found no evidence for a Mendelian mode of transmission. The segregation analysis showed a homogeneous general model of transmission best fitted the verbal memory data and a heterogeneous general model best fitted the fine motor functioning data. When compared with a model with no genetic effect, the Akaike’s Information Criteria were significantly lower for verbal memory (525.28 versus 531.72; p ¼ 0.04) and fine motor functioning (698.88 versus 700.50; p ¼ 0.05). For set shifting the environmental model could not be rejected. Our findings suggest an underlying genetic transmission for verbal memory and fine motor functioning. The parameters estimated can augment further genetic study relevant to identify genetic factors involved in these traits in these families. Further analysis on heritability and familial risk estimation for the endophenotypes will be presented. P167 BIPOLAR DISORDER IN SARDINIAN POPULATION: AGE AT ONSET AS AN ENDOPHENOTYPE? Mirko Manchia, Simona F. Lampus, Stefania Margiani, Caterina Chillotti, Raffaella Ardau, Giovanni Severino, Maria Del Zompo Unit of Clinical Pharmacology, Department of Neurosciences ‘‘B.B. Brodie’’, University of Cagliari, Italy The discovery of the genetic factors implicated in the predisposition to Bipolar Disorder (BD) may greatly profit from genetic studies in isolated populations, as Sardinian population, with lesser genetic heterogeneity and probably fewer environmental differences than outbred populations. The identification of endophenotypes, such as a subgroup with early age at onset, may facilitate the finding of genes involved in the etiology of BD. Assesing the presence of an early onset group of bipolar subjects in a Sardinian population of at least four generations of Sardinian ancestry with the objective to design a linkage study with candidate genes. 300 probands were recruited using the FH-RDC. Pypsi, a software planned for the recording of clinical data, enabled as the management of the acquired informations. Age of onset was defined as the first reliably diagnosed hypo-/manic or depressive episode. Admixture analysis was used to decompose the age-at –onset distribution into a mixture of theoretical normal distributions. We clinically characterized all probands by sex, diagnosis (BPI, BPII, SAM), age-at-onset and suicide attempts. Referring to 180 BPI probands the age-at-onset consisted of a mixture of four normal distributions with means of 18.13 (SD ¼ 0.56), 25.06 (SD ¼ 3.36), 42.13 (SD ¼ 4.80) and 59.94 (SD ¼ 2.05) with a p-value ¼ 0.0106. Cutoff points at ages 19 and 37 were derived from this analysis and used to define ageat-onset subgroups.
Our study demonstrates the presence of a group of bipolar subjects with an early age at onset. Early age at onset may possibly be an endophenotype that can be used in linkage studies with candidate genes.
P168 EPIGENETIC INFLUENCES ON CANDIDATE GENES FOR MENTAL DISORDERS Selma Stuffrein-Roberts, Mike Eccles, Vicky Cameron, Peter Joyce, Martin Kennedy University of Otago, Christchurch Psychiatric illnesses are caused by environmental circumstances and genetic predisposition, yet dissecting out the heritable component has proven difficult. Genetic association studies, which work under the assumption of biallelic gene expression, have largely overlooked epigenetic mechanisms (e.g. DNA methylation) that can influence the level and/or pattern of gene expression. Evidence is accumulating that epigenetics plays a crucial role in various human illnesses and contributes to the observed phenotypic diversity. We are investigating mental disorders from an epigenetic perspective by screening a variety of psychiatric candidate genes for their allelic expression pattern. Our primary hypothesis is that some candidate genes will show monoallelic or differential gene expression. A secondary hypothesis is that these expression patterns will occur in a tissue-specific fashion. To test these hypotheses, we obtained hybrid offspring from a cross between two inbred mouse strains. These offspring potentially possess different alleles for most genes. Sequence differences between the alleles were identified and a single base extension assay developed. This assay allows the distinction of mRNA expressed by two different alleles and was used to analyse expression patterns of several genes, including Drd2 and Tph2 in cerebellum and midbrain tissue sections. No differential, monoallelic or tissue-specific gene expression has been found; however, a third brain region (olfactory) is currently being investigated, along with more genes. Any epigenetic effects on the expression of psychiatric candidate genes would help explain some of the genetic complexity of mental disorders and would significantly influence the interpretation of association studies.
P169 WHOLE-GENOME AMPLIFICATION (WGA) OF SODIUM BISULFITE-TREATED DNA ALLOWS ACCURATE METHYLATION PROFILING ON LIMITED DNA RESOURCES. Jonathan Mill, Simin Yazdanpanah, Eva Gueckel, Arturas Petronis Centre for Addiction & Mental Health, Toronto Introduction: Increasing evidence suggests an important role for epigenetic processes such as DNA methylation in the aetiology of psychiatric disorders. DNA methylation is usually assessed by modifying DNA with sodium bisulfite, a process that converts unmethylated cytosines to uracil. Methylated cytosines can be subsequently detected using techniques such as bisulfite genomic sequencing or single-nucleotide primer-extension. A major problem with bisulfite-based methylation analyses is that large quantities of starting genomic DNA material are needed, a luxury not usually afforded in epigenetic studies of psychiatric disorders. Methods: Bisulfite-treated DNA was amplified using two WGA methods: optimised primer-extension preamplification (PEP) using degenerate primers, and the commercially-available REPLI-g kit from Qiagen. Following bisulfite-PCR, methylation profiles were assessed using three approaches: direct bisulfite sequencing, the sequencing of cloned PCR products and single-nucleotide primer-extension; they were then compared to those obtained from bisulfite-treated DNA not subjected to WGA. Results: All loci tested could be amplified using bisulfite-PCR, and the methylation-profiles obtained from bisulfite-DNA amplified with WGA were comparable to those obtained from unamplified bisulfite-DNA. We estimate that WGA results in at least a 200-fold increase in usable bisulfite-treated DNA. Conclusions: WGA provides a useful tool for investigating DNA methylation in cases where DNA availability is low. Bisulfite-treated DNA amplified by WGA can be used for a range of downstream quantitative methylation profiling techniques.
Abstracts P170 EXPLORATION FOR THE EPIGENETIC VARIATION IN SAMPLES DERIVED FROM PSYCHIATRIC PATIENTS Kazuya Iwamoto1, Miki Bundo1, Junko Ueda1, Yoko Nakano1, Mizuho Ishiwata1, Chihiro Kakiuchi1, Takahiro Tsujita2, Ichiro Kusumi3, Yuji Okazaki4, Tadafumi Kato1 1 RIKEN Brain Science Institute, Wako 2 Department of Psychiatry, Nagasaki University, Nagasaki 3 Department of Psychiatry, Hokkaido University, Hokkaido 4 Department of Psychiatry, Mie University, Mie Alterations in the epigenetic status such as DNA methylation and histone modifications are suggested to be involved in pathogenesis of complex psychiatric disorders. We have been searching for the epigenetic differences in samples derived from psychiatric patients including bipolar disorder and schizophrenia by the following approaches: 1) candidate gene approach, where epigenetic status of genes that are well-replicated in the genetic or expression studies are examined. 2) Expression-based approach by screening of the genes whose expressions are up-regulated by the treatment of 5-azadeoxycitidine or tricostatin A to the lymphoblastoid cells, and 3) whole-genome approach, where DNA methylation pattern are directly mapped to the human genome using tiling arrays. Although, so far, we did not obtain significant epigenetic differences in samples derived from psychiatric patients, we identified altered DNA methylation status of SOX10 that are associated with altered expression in brains of patients with schizophrenia by the candidate gene approach, and identified differences in DNA methylation patterns of some genes such as GATA3 and CHST5 between monozygotic twins discordant for bipolar disorder. Such alterations together with candidate genes derived from ongoing tiling array experiment may be important for the regulation of long-lasting gene expression changes in psychiatric patients. P171 EPIGENETIC REGULATION OF THE HERP GENE IN PATIENTS WITH ALCOHOL DEPENDENCE Stefan Bleich, Bernd Lenz, Sonja Beutler, Helge Frieling, Johannes Kornhuber, Dominikus Bo¨nsch University of Erlangen-Nuremberg, Erlangen Background: This study was conducted to investigate whether DNA methylation pattern within the HERP (homocysteine-induced endoplasmic reticulum protein) promoter region and expression of HERP mRNA are altered in patients with alcohol dependence. Methods: Expression of HERP mRNA levels was measured by quantitative PCR in the blood of 66 male alcoholic patients and 55 non-drinking healthy controls. DNA methylation status of the HERP promoter was measured by quantitative PCR after methylationsensitive enzymatic digestion (using HpaII and MspI endonucleases). Results: HERP promoter DNA methylation was significantly increased in patients with alcohol dependence when compared with healthy controls (88.0%, SD 11.8 vs. 80.4%, SD 14.5; t test, t ¼ 2.45, p ¼ 0.02), which was significantly associated with their elevated homocysteine levels (multiple linear regression, p ¼ 0.007). Furthermore, we found a significantly lower HERP mRNA expression in patients with alcohol dependence (t test, delta CT ¼ 7.61; SD 1.87, p < 0.001) when compared with healthy controls (delta CT ¼ 6.04; SD 2.41). The lowered HERP mRNA expression in alcoholic patients was best explained by the hypermethylation of the HERP gene promoter (regression analysis, p ¼ 0.004). Conclusions: This study reports for the first time a lowered HERP mRNA expression that is explained by a specific gene promoter hypermethylation in alcoholic patients. As HERP has been considered to play an essential role within the intracellular defence system, these findings may be useful in the understanding and treatment of different disease conditions associated with alcohol dependence. P172 EPIGENETIC MECHANISMS IN NEURODEVELOPMENTAL DISORDERS Patrick Jean Rogue Universite´ Louis Pasteur—Me´decine, Strasbourg Epigenetic processes are involved in human diseases. Thus, Rett syndrome (RS) is a common neurodevelopmental disorder generally attributable to mutations in the X-linked MeCP2 gene, whose product binds specifically to methylated DNA and functions as a transcriptional
777
repressor (by acting as a bridge between two global epigenetic modifications, DNA methylation and histone methylation) and as a regulator of alternative splicing. Epigenetic factors could be important in other neurodevelopmental disorders including Infantile Autism (AI) and Angelman syndrome (AS), which share several characteristics in common with RS. Multiple genetic factors have been described in the etiology of these disorders, including GABRB3 (AI) and the ubiquitin ligase UBE3A (AS). MeCP2 has also been implicated although mutations in the coding region of MECP2 are not frequently found. However the regulation of MeCP2 expression during normal brain development is complex, suggesting that other types of mutations may affect normal MeCP2 expression. Furthermore it has recently been shown (Samaco et al. 2004) that MeCP2 deficiency results in significant reduction of UBE3A and GABRB3 expressions in mouse brain without apparent alterations in expression, suggesting UBE3A and GABRB3 defects observed in autism-spectrum disorders are consequences of abnormal MECP2 function. Both UBE3A and GABRB3 genes are encoded within 15q11–q13, indicating a common epigenetic defect at the level of the regulation of gene expression within this region in the three phenotypically similar but genetically distinct autism-spectrum disorders. We will discuss how altered epigenetic mechanisms may be involved in neurodevelopmental disorders with special focus on MeCP2. P173 ATTITUDES TOWARDS PSYCHIATRIC GENETICS Magdalena Kaczmarkiewicz-Fass1, Aleksandra Rajewska-Rager2, Anna Leszczynska-Rodziewi2, Joanna Hauser2 1 Department of Adults Psychiatry, Poznan 2 Department of Adults Psychiatry, Poznan Objectives: According to the dynamic evolution of molecular genetics, it is probable that in the near future some genes related to the etiology of many mental disorders will be singled out. Therefore, it will be possible to carry out predictive genetic tests. The purpose of the study undertaken was to analyse patients’ opinions on problems connected with the possibility of carrying out predictive genetic testing in psychiatry. Methods: Participants in this study were 173 patients with bipolar disorder. The questionnaire: ‘‘Ethical problems related to genetic examinations’’ included questions about genetics and psychiatry. Results: Patients have a positive attitude towards psychiatric genetic research (67%). Opinions of patients referring to the possibility of obligatory genetic tests were positive in 35% of cases, and they often agreed to render test results to members of the family (42.7%) and the family doctor (53.75%). The majority of patients (77%) would not allow genetic tests results to be given to their insurance company or healthinsurance agent. More than 65% would agree with prenatal genetic testing of their children. Conclusions: The majority of patients with a diagnosis of bipolar disorder were in favour of genetic research and predictive testing in psychiatry. References: Illes F, Rietz C, Rudinger G, Matschinger H, Angermeyer MC, Rietschel M: Der GenEthik-Fragebogen. Ein Instrument zur Erfassung der Einstellung zu psychiatrischen-genetischer Untersuchung und Forschung, Bonn/Berlin. Pace. ISBN, 2002,3,9301181. 30–34. P174 GENETIC COUNSELING IN INDIA Elizabeth Gettig1, Triptish Bhatia2, Vishwajit L. Nimgaonkar1, Smita N. Deshpande2 1 University of Pittsburgh, Pittsburgh, PA 2 Ram Manohar Lohia Hospital, Delhi Genetic counseling is well established in North America and has expanded to at least a dozen other countries globally (http:// igce.med.sc.edu/programsIntro.htm). While many other nations are still in the era of infectious diseases, their global share of genetic diseases is also increasing (WHO, 1999). In the following review of research in genetic testing and counseling in India, we focus on evidence for requirements of genetic counseling in India. While genetic tests are available to urban Indian populations but few facilities exist for rural India. Genetic counseling is however difficult to come by in both urban and rural India. Different surveys have found that genetic counseling is welcomed. People come forward for counselling wherever services are available but are still are apprehensive about termination of pregnancies and prenatal diagnoses. The diseases for which
778
Abstracts
counseling is available though only in cities in India are thalassemia, sickle cell anemia and chromosomal abnormalities. Though some psychiatric genetic programs are running successfully in India, genetic counseling in this field is still in its infancy. Projects on genetics of schizophrenia were initiated at RML-New Delhi in 1996 in collaboration with Western Psychiatric Institute and Clinic, Pittsburgh, USA. In 2004, this work was extended to training qualified personnel in genetic epidemiology, counseling and ethics. Funded by the Fogarty International Research Institute (1 D 43 TW006167), NIH, USA, we aim to collect genetic risk data for schizophrenia, bipolar disorder and major depressive disorder and initiate genetic counselling. P175 IMPACT OF ETHICS COMMITTEES ON THE STUDIES OF GENETIC RISK FACTORS IN MENTAL DISEASES Carta Paolo, De Lisa Roberto, Serio Stefania, Ligas Franca, Chillotti Caterina, Del Zompo Maria Section of Clinical Pharmacology, Department of Neurosciences ‘‘B.B. Brodie’’, University of Cagliari, Italy All clinical trials should first be approved by an Ethics Committee (EC) before being put into practice. The Ethics Committee is an independent institution made up of healthcare professionals and other figures, aimed at guaranteeing the rights of recruited trial subjects and expressing an ethical and scientific opinion on the overall trial. In order to suitably evaluate scientific, methodological and ethical aspects of the trial, in accordance with legislative regulations decreed by the Italian Ministry of Health on 18/03/98, ECs are obliged to include several professional figures, one of the most relevant being the Clinical Pharmacologist. The Section of Clinical Pharmacology of the Department of Neuroscience ‘‘B.B. Brodie’’, University of Cagliari, among many other activities acts as Scientific Secretariat for numerous local ECs. The EC supervises drug-applied research and is called upon to assess protocols of clinical/therapeutic research as well as genetic research. The latter consists in the identification of the basis of pathologies and the identifying of new molecular targets with the aim of implementing individual therapies and genetic tests. Genetic trials do not grant any immediate benefit to trial subjects but aim rather to provide information for future use; such trials therefore do not concur with the customary experimental concept, representing an important contribution to the progress and advancement of the entire scientific community. The fundamental objective is to find an ‘‘agreement’’ between Population (the Rights of enrolled subjects) and Scientific Community (Progress of Knowledge): in this scenario EC plays a fundamental role. In order to facilitate the drawing up of genetic research protocols, the Scientific Secretariat has published a set of Guidelines in line with reports published in national and international literature. From 2002 to 2006 the Section of Clinical Pharmacology, in the role of Scientific Secretariat of local ECs, has evaluated numerous protocols of genetic research, totalling 6% of all trials assessed, among which several genetic trials and psychiatric genetic protocols. The application of Guidelines has stimulated the debate and assessment of studies of genetics and pharmacogenetics, thereby enhancing approval of trials in full respect of human rights. P176 MICROARRAY ANALYSIS IN THE PREFRONTAL CORTEX OF SCHIZOPHRENIA Masaya Yanagi1, Takeshi Hashimoto2, Noboru Kitamura3, Masaaki Fukutake2, Kentarou Mouri2, Naoki Nishiguchi2, Osamu Shirakawa2, Kiyoshi Maeda2 1 Kobe University, Kobe Hyougo-ken 2 Kobe University, Kbe, Hyougo-ken 3 Kobe City General Hospital, Kobe, Hyoogo Introduction: The etiology and genetic risk factors for schizophrenia are complex. Genome-wide gene expression analysis using DNA microarrays is the potential tool for finding unanticipated susceptibility genes for complex diseases. In this study, we carried out a microarray, HU133A chip (Affymetrix), in the postmortem brains of schizophrenia patients. Method: Total RNA was extracted from frozen blocks of orbital prefrontal cortex using ISOGEN (Nippon Gene). The purity of total RNA was evaluated by the OD260/OD280 ratio, and its integrity was evaluated by denaturing agarose gel electrophoresis. Microarray analysis was performed according to the manufacturer’s protocol
(Affymetrix). This study was approved by the Ethics Committee of the Kobe University School of Medicine, and informed consent was obtained from all subjects and their relatives. Results and Discussion: Gene expression data generated by the microarray were normalised by dividing each expression value by the median gene expression value. Genes of interest were genes whose expression was under half in the patients as compared to controls, and whose genetic loci were on 8p or 13q or 22q, reported to have the strongest evidence for susceptibility loci for schizophrenia in metaanalysis. Of 22 284-probe sets examined, 6 genes satisfied all criteria. These genes contained two intracellular signalling molecules, two transcription factors, one cell adhesion molecule and one nucleocytoplasmicshuttling protein. Further examination of these genes might clarify the pathophysiology of schizophrenia. Badner JA, Gershon ES. Mol Psychiatry. 2002;7(4):405–411. P177 CHANGES OF SEROTONIN TRANSPORTER mRNA EXPRESSION IN LYMPHOCYTES OF MAJOR DEPRESSIVE PATIENTS TREATED WITH DULOXETINE Blanca Gutie´rrez, Esther Molina, Jose´ A. Lorente, Margarita Rivera, Jorge Cervilla Universidad de Granada, Granada Lymphocytes from human peripheral blood exhibit a series of markers of neurotransmitters, such as specific receptors and transporters. Some authors have suggested that the expression of such molecules in peripheral blood lymphocytes could reflect brain status. The aim of the present study is to determine if the administration of duloxetine, a SSNRI, can vary mRNA levels of serotonin transporter and how these changes are related to improvement in depression symptoms. Analyses are being carried out on patients with DSM-IV-TR major depression. All individuals undergoing duloxetine treatment are prospectively rated for response according to the Montgomery-Asberg Depression Scale (MADRS) and Global Clinical Impression (ICG) scale. A checklist for secondary effects is also used during follow-up of these patients. Blood samples are being collected in TempusTM blood RNA tubes (Applied Biosystems) containing RNA stabilising reagents that ensure the stabilisation of transcript profiles. Samples are obtained before and after 3 and 6 weeks of treatment with duloxetine. We have started expression analysis using real-time technology for relative quantification of RNA levels. Preliminary results based on 5 patients show a significant reduction in serotonin transporter mRNA levels in their lymphocytes before treatment and an increase after it. We are now enlarging our sample in order to have further statistical power that can confirm the trend we report here and the relation of those changes to relief of depressive symptoms. Work supported by MEC(SAF2004-01310). P178 PLASMA OXYTOCIN AND VASOPRESSIN ARE HIGHLY HERITABLE IN A PEDIGREE OF VERVET MONKEYS Julia Bailey1, Karen Bales2, Matthew Jorgensen3, Lynn Fairbanks3 1 UCLA, Los Angeles, CA 2 Dept. of Psychology, California National, Davis, CA 3 Center for Primate Neuroethology, Dept of Psychiatry, UCLA, Los Angeles, CA Introduction: Oxytocin and vasopressin are nine-amino acid peptides produced in the hypothalamus and secreted by the posterior pituitary. Both oxytocin and vasopressin are crucial in the formation of normal social bonds, and have been implicated in conditions in which social bonding is impaired (ex. autism spectrum disorder). Methods: Study subjects were 259 captive, socially-living vervets (Chlorocebus aethiops) from a large multigenerational pedigree at the UCLA/VA Vervet Research Colony. Blood samples were collected under anesthesia during annual health exams, and kept on ice until spun and frozen at 808C. Oxytocin and vasopressin were measured by commercial enzyme immunoassay (Assay Designs, Ann Arbor, MI) validated for the vervet monkey. Heritability and bi-variate analyses were assessed using variance component modelLing as implemented in SOLAR. Results: Both oxytocin (h2 ¼ 0.61 0.11, p < 0.0001) and vasopressin (h2 ¼ 0.29 0.12, p ¼ 0.002) are heritable in this sample. Neither variable was significantly influenced by the effects of age or sex. The genetic correlation between the two variables is highly significant (RHOg ¼ 1, p < 0.00001), indicating that these two traits are controlled by pleiotrophic or linked genes.
Abstracts Conclusions: Plasma Oxytocin and vasopressin are highly heritable in this non-human primate species, and genes encoding for their expression appear to be linked, as is seen in Homo sapiens. This may have implications for the heritability of social behaviour such as mothering and heterosexual social bonds, as well as disorders such as depression and autism spectrum disorders. P179 SUPPORT FOR RETICULON 4 (RTN4) AS A POTENTIAL SUSCEPTIBILITY GENE FOR SCHIZOPHRENIA Tim Pierce, Nick Bray, Nigel Williams, Dobril Ivanov, Michael Owen, Michael O’Donovan University Hospital of Wales, Cardiff Over-expression of Reticulon 4 (RTN4) has been reported in post mortem brains of schizophrenic patients. Two independent studies have reported association with schizophrenia with polymorphisms (77878 AAC-Insertion and 77394 VNTR) in the 30 UTR of RTN4, although association was restricted to the female subgroup in one study. There have also been four negative replication studies published. We tested these two polymorphisms for association with schizophrenia in our Cardiff case (n ¼ 709) control (n ¼ 710) sample. Individual genotyping revealed an association between the 77394 VNTR and schizophrenia at the genotypic (p ¼ 0.02) and allelic level (p ¼ 0.02). Consistent with one previous study, association was limited to the female subgroup (p ¼ 0.004). No association was detected for the AAC-Ins polymorphism, although the previously-associated insertion allele was over-represented in cases, as was the putative risk haplotype (AAC-Ins, shorter VNTR). We extended our study by sequencing the RTN4 coding sequence and genotyping all of the polymorphisms in a pooled association sample. This revealed no evidence for association with schizophrenia. Additionally, we sought to determine whether altered expression of RTN4 could be attributable to cis-acting sequence variation. We measured relative expression of brain mRNA from 35 informative individuals using marker (rs6545468), which revealed highly significant expression differences in some individuals (max differences 40%), confirming the presence of cis-acting variants influencing RTN4 expression. Our results provide modest support for the hypothesis that genetic variation within RTN4 contributes to susceptibility to schizophrenia. P180 ANALYSIS OF PDLIM5 (LIM) EXPRESSION IN THE MOUSE BRAIN Yasue Horiuchi1, Shinya Iwasaki2, Aoi Syu2, Minori Koga3, Hiroki Ishiguro3, Tadao Arinami3 1 University of Tsukuba, CREST, Tsukuba, Ibaraki 2 University of Tsukuba, Tsukuba, Ibaraki 3 University of Tsukuba, CREST, Tsukuba, Ibaraki The PDLIM5 (PDZ and LIM domain 5) gene contains an LIM domain and a PDZ domain and regulates intracellular calcium levels by linking calcium channels and protein kinase C. The expression of the PDLIM5 is up-regulated in the postmortem brains of patients with schizophrenia and those with bipolar disorders. Previous studies showed that the SNPs in PDLIM5 are associated with schizophrenia and with gene expression. In this study, we investigate the influences of methamphetamine (METH), MK-801 and haloperidol on Pdlim5 in mouse brain. Mice were administrated saline (control group), METH (2 mg/kg/day), MK-801(0. 5 mg/kg/day), haloperidol (0.5 mg/kg/day) for 1 (acute group), 12 or 28 (chronic group) days. Animals were sacrificed 3 h after the final injection and brain regions were removed. The levels of mRNA expression of Pdlim5 were measured by a real-time RT-PCR. Pdlim5 mRNA expressions significantly increased in the prefrontal cortex (PRC) after 12-day METH treatment (p < 0.001) and decreased after 28-day haloperidol treatment (p < 0.05). MK-801 treatment did not produce differences in the expression level of Pdlim5. These data suggest that PDLIM5 is involved in the dopaminergic signalling cascade. P181 LINKING PER GENE EXPRESSION AND ALCOHOL MISUSE Sylvane Desrivieres1, Li Dong2, Jens Treutlein3, Gunter Schumann2 1 Institute of Psychiatry, London 2 Institute of Psychiatry, King’s College, London 3 Central Institute of Mental Health, Uni, Mannheim
779
We have recently linked disturbances of the circadian clock to alcoholdrinking behaviour. Mice in which the circadian clock gene Period 2 (Per2) is mutated show increased alcohol intake, while in humans, there is an association between alcoholic individuals and genetic variations in the PER2 gene (Spanagel et al., Nat Med 2005 Jan;11(1):35–42). We have now extended these studies and investigated the role of another circadian clock gene, PER1, in patterns of human alcohol drinking and its potential contribution to alcohol dependence. We performed a search for genetic variations in the PER1 gene and identified 15 single nucleotide polymorphisms (SNPs), finding a significant association between one of these SNPs and alcohol drinking behaviour in adolescents. Interestingly, for PER1 and for PER2, the SNPs showing significant association with high versus low alcohol intake are located in the 50 regulatory region and enhancer-like structure, suggesting that differential expression of these genes might contribute to the observed phenotype. To test this hypothesis, we have generated genotype-specific reporter-gene constructs and investigated the effects of individual SNPs on the transcriptional activation of PER genes in a human neuroblastoma cell line. This analysis provides important information as to the molecular mechanisms linking alcohol misuse and the circadian clock. P182 SNPs AND GENE EXPRESSION ANALYSIS IN MONOZYGOTIC TWINS DISCORDANT FOR SCHIZOPHRENIA Shinji Ono1, Akira Imamura2, Naohiro Kurotaki2, Nishihara Koji3, Aoi Hashida4, Yuji Okazaki5, Hiroki Ozawa2 1 Kamigoto Hospital, Minamimatsuura 2 Nagasaki University, Nagasaki 3 Nagasaki Medical Center, Nagasaki 4 Joya Hospital, Nagasaki 5 Matsuzawa Hospital, Tokyo We studied SNPs and the gene expression profiles in monozygotic twins discordant for schizophrenia and compared between twins, as a trial to identify susceptible gene for schizophrenia. The subjects were a pair of monozygotic male twins (age: 51), one had DSM-IV schizophrenia, undifferentiated type, and the other had a job, was married, and was well-adapted to social life. This study was approved by the Ethics Committee of Nagasaki University. The samples were genomic DNA and RNA from the subjects. The SNPs were detected using Affymetrix GeneChip Human Mapping 100K Set (50K Xba Array, 50K Hind Array) and the gene expressions were analyzed by Affymetrix Human Genome U133 Plus 2.0 Array. In SNPs detection, forty six SNPs had different signal intensity among all 116204 SNPs. Also, we measured the expression level of 12857 genes. As a result, we detected 4 SNPs that caused possible different gene expression. In addition, we found there were 21SNPs that might influence amino acid change or splicing within some genes. We investigate the associations between these SNPs and schizophrenia. P183 GENOTYPE-PHENOTYPE CORRELATIONS OF THE WILSON DISEASE IN BASHKORTOSTAN Aliya Magzhanova1, Elza Khusnutdinova1, Rim Magzhanov2, Aleksandra Karunas1 1 Ufa Center Russian Academy Science, Ufa 2 Bashkir State Medical University, Ufa Wilson’s disease is an autosomal recessive disorder of hepatic copper metabolism caused by mutation in the gene encoding a coppertransporting P-type ATPase (13q14.3–q21.1) and leading to heavy hepatic and neurological disorders. The purpose of this research was to analyse the correlation between clinical features (neurological, neuropsychological and liver disorders) and types of mutations. We observed 34 patients from 28 families from Bashkortostan, using clinical examination, biochemical analyses of blood and instrumental methods (hepatic scanning, hepatography). Neuropsychological investigation was performed using HADS, Spilberger’s scale and Beck’s scale. We had carried mutation analysis in 28 families with WD from Bashkortostan. Using SSCP analysis followed by sequencing of 10,14,15,16,17,18,19 exons were identified 4 mutations in 60% of chromosomes: H1069Q—46.4%, E1064K—3.6%, 3402delC—5.6%, 3559þ1G > T—5.6%. Investigation of genotype-phenotype correlations determined that manifestation of homozygous for the H1069Q mutation occurs at a later age than the heterozygous for this mutation
780
Abstracts
or undetermined type of mutation. Homozygous for the H1069Q mutation have more frequently late trembling form of disease and more rarely have liver disorders. P184 POLYMORPHIC CIS-ACTING INFLUENCES ON DISC1 EXPRESSION IN THE HUMAN BRAIN Jesse Hayesmoor, Nick Bray, Michael Owen, Michael O’Donovan University Hospital of Wales, Cardiff The DISC1 gene was originally identified in a large Scottish family in which a balanced chromosomal translocation t(1;11)(q42;q14) cosegregated with several psychiatric diagnoses, including schizophrenia, bipolar disorder and major depression. Since then, several studies have reported association between DISC1 haplotypes and both schizophrenia and bipolar disorder in various populations. Expression of DISC1 in the brain is developmentally regulated with highest levels in the limbic system, particularly the hippocampus. The protein is thought to play an important role in neuronal cell migration and neurite extension, since, when gene expression is downregulated, impairments in these processes are observed. Evidence from cell lines derived from t(1;11) carriers shows that there is a 50% reduction in the normal level of DISC1, which lends support to the view that reduced expression may have pathogenic effects. The present study investigated the hypothesis that DISC1 expression in the human brain is subject to polymorphic cisacting influences. Highly quantitative measures of relative allelic expression were applied to postmortem brain tissue from 65 Caucasian individuals. Reproducible distortion of allelic expression > 20% was observed in three individuals, with maximal disortion of 49%. These results indicate the presence of polymorphic cis-acting influences on DISC1 expression in the human brain, which may be relevant to the aetiology of mental illness. P185 THE DIFFERENCES OF EEG COHERENCE BETWEEN SCHIZOPHRENIA AND BIPOLAR DISORDER Yu-Sang Lee1, Seong-Deok Park1, Jong Won Park1, Yong Kyu Kim1, Kyung Sue Hong2, Jae Seung Jang3 1 Yongin Mental Hospital, Yongin-City 2 Department of Psychiatry, Sungkyunkwan U, Seoul 3 Clinical Research Institute, Seoul Natio, Seoul Objectives: EEG coherence could imply connectivity between two different areas of the brain, known to be important in the pathophysiology of bipolar I disorder (BPDI) and schizophrenia. The authors investigated EEG coherence in patients with BPDI and schizophrenia to examine the connectivity of neural circuits. Methods: EEGs were recorded in 15 schizophrenia and 14 bipolar disorder patients and 14 age-matched normal control subjects, by 16 electrodes with linked-ear reference. Spectral parameters and coherence were calculated for the alpha bandwidth (813Hz) by a multichannel autoregressive model, using 20 artifact-free 2-second epochs, and differences among the three groups were compared by two different statistical methods: F-test and Kruskal-Wallis test. Furthermore, when there were significant differences among the three groups, Scheffe’s multiple comparison tests were provided and JonckheereTerpstra tests for ordered alternatives were given. Results: In the intra-hemispheric comparison, left frontal coherence increased from controls to BPDI to schizophrenia. In the interhemispheric comparison, (1) inter-prefrontal coherence in BPDI was significantly higher than in normal controls, and (2) inter-prefrontal coherence in schizophrenia was significantly lower than in controls. Conclusion: These results suggest that (1) both schizophrenia and BPDI are diseases having abnormal neural circuit connectivity in both the frontal and prefrontal lobes, and (2) abnormality is more severe in schizophrenia than in BPDI. Furthermore, the data indicate that common pathogenetic processes may take place in both schizophrenia and BPDI. P186 META-ANALYSIS REVEALS NO ASSOCIATION OF THE Val66Met POLYMORPHISM OF BDNF WITH EITHER SCHIZOPHRENIA OR BIPOLAR DISORDER Tetsufumi Kanazawa Osaka Medical College, UCSD, Osaka Background: A long-term controversy exists on whether or not major psychotic disorders can be divided into two groups, for example
schizophrenia and bipolar disorder. Many genes and polymorphisms have been studied for a role in both disorders, including the Val66Met (also known as rs 6265) variant of BDNF (brain-derived neurotrophic factor). Many case-control association studies have been performed to see if BDNF could serve as a useful clinical diagnostic biomarker for schizophrenia or bipolar disorder, but results have been equivocal. Objectives: To determine, by meta-analysis, if the Val66Met polymorphism of BDNF influences risk for schizophrenia, bipolar disorder, or both. Methods: We searched Pubmed, Medline and PsycInfo using keywords including Val66Met, Rs6265, BDNF, schizophrenia and bipolar disorder. A total of 13 studies for schizophrenia and 11 studies for bipolar disorder were combined by random-effects meta-analysis. Main results: The pooled results from the schizophrenia sample (2955 patients, 4035 controls) and the bipolar disorder sample (3143 patients, 6347 controls) indicated a lack of significance with either of the two psychoses, with pooled odds ratios of 1.00 (p ¼ 0.944) and 0.95 (p ¼ 0.161), respectively. Conclusion: Although there are some limitations on the study, our results indicate there is a lack of association between the Val66Met polymorphism and either of the two psychoses. A larger sample size and evaluation of more single nucleotide polymorphisms are needed to obtain clearer and more conclusive findings regarding the relationship between the BDNF gene and psychosis. P187 ASSOCIATION OF MITOCHONDRIAL COMPLEX I SUBUNIT GENE NDUFV2 AT 18P11 WITH SCHIZOPHRENIA Shinsuke Washizuka1, Mizue Kametani2, Mamoru Tochigi3, Tadashi Umekage4, Kazuhisa Kohda3, Tsukasa Sasaki4, Tadafumi Kato2 1 Shinshu University, Matsumoto 2 Brain Science Institute, RIKEN, Wako 3 Faculty of Medicine, University of Tokyo, Tokyo 4 Health Administration Center, University, Tokyo Bipolar disorder and schizophrenia are two representative major mental disorders sharing a common genetic background. Several loci such as 18p11, 13q32 and 22q11-3 were commonly linked with these diseases. Since mitochondrial dysfunction has been suggested in both of these disorders, NDUFV2 at 18p11, encoding a subunit of the complex I, NADH ubiquinone oxidoreductase, is a candidate gene for both bipolar disorder and schizophrenia. We previously reported that single nucleotide polymorphisms (SNPs) in the upstream region of NDUFV2 were associated with bipolar disorder in Japanese. The association of haplotype consisting of two SNPs, 3542G > A and 602G > A, with bipolar disorder was also seen both in Japanese and the National Institute of Mental Health Pedigrees trios. Furthermore, we found that 602G > A altered promoter activity. 3542G > A was predicted to reach the binding site of heat shock transcription factor 1. In this study, two polymorphisms, 3542G > A and 602G > A, were investigated in 229 schizophrenic patients as compared with controls. The Ethics Committees of the Brain Science Institute and participating institutes approved this study. Individual genotypes were not associated with schizophrenia. However, the haplotype consisting of these two SNPs was significantly associated with schizophrenia. This finding indicated that the haplotype in the promoter region of NDUFV2 may be a common genetic risk factor for bipolar disorder and schizophrenia. P188 BIPOLAR AFFECTIVE DISORDER DIMENSION SCALES (BADDS)—EXTENDED RATING SCALES Elen Elizabeth Russell Cardiff University, Heath Park Psychiatric research continues to be dominated by the use of categorical operational diagnostic systems. However, the limitations of this approach have become clear as molecular genetic studies provide increasing support for overlap across the traditional Bipolar DisorderSchizophrenia divide. This highlights the importance of alternative tools which pick up on variation both within and across diagnostic boundaries. We have previously described the Bipolar Affective Disorder Dimensional Scales (BADDS) that provide lifetime ratings of mania, depression, psychosis and psychotic mood congruence on 0– 100 scales. Here we describe 11 lifetime clinical scales that supplement BADDS and are relevant for patients with affective and/or psychotic symptoms. An iterative process was undertaken in which scale definitions were agreed upon and piloted across sets of data. Modifica-
Abstracts tions were made to improve reliability and ease of utility. The scales measure the following: predominance of manic and mixed episodes within affective illness, relationship between affective and psychotic features, fluctuation of mood within acute phases of illness, instability of clinical state, periodicity of acute phases of illness, most dysphoric manic episodes, extent to which illness is prototypically psychotic or affective, predominance of chronic defect state, disorganised behaviour and catatonic features. These scales can be used alongside and independently of traditional diagnostic systems to provide a richer description of lifetime psychopathology. It is hoped that when used in conjunction with genetic data, they will enable the identification of biologically-validated phenotypic subgroups which will facilitate understanding of mechanisms behind different manifestations of psychotic and affective illness.
P189 DETECTION AND CHARACTERISATION OF THE BIPOLAR SUSCEPTIBILITY GENE SLYNAR IN THE MOUSE Ana Catarina P. Pereira1, Andrew McQuillin1, Ole Mors2, Inger M. Olsen2, Vinay Puri1, Khalid Choudhury1, Nicholas J. Bass1, Hugh M. Gurling1 1 University College London, London 2 Aarhus University Hospital, Aarhus Four linkage studies of bipolar or unipolar disorder have found lod scores greater than 3.0 between markers in the chromosome 12q24 region and affection status. Fine mapping carried out in Denmark and the UK has localised a bipolar susceptibility locus within a region of 278 kb. The most promising candidate gene is Slynar (Kalsi et al., 2006). Slynar has no known function, but both the Human and Macaque brain have been shown to possess Slynar mRNA transcripts. The mouse shows conservation with the human Slynar in exons 2, 3 and 4. However, no mRNA had been identified in the mouse. In order to determine if Slynar is transcribed in the mouse and to assess the suitability of making a mouse model for bipolar disorder, we made mouse specific primers for the most conserved region covering Slynar exon 3 and exon 4. mRNA was extracted from mouse brain; RTPCR screening with mouse-specific primers detected a Slynar transcript in both brain and testes. Sequence analysis of clones from the RTPCR product confirmed that the transcripts were derived from Slynar. This suggests that Slynar is expressed in the mouse brain. Riboprobes prepared from one of these clones were used to carry out Northern blots on mRNA from a variety of mouse tissues. Kalsi et al.: Allelic and haplotypic association with bipolar disorder on chromosome 12q24 identifies the Slynar gene (AY070435) and related brain expressed sequences as a candidate gene for genetic susceptibility to affective disorders. Am J Psychiatry 2006 (in press). P190 IDENTIFICATION OF GENETIC DETERMINANTS OF SOCIAL MEMORY USING CHROMOSOME SUBSTITUTION MOUSE STRAINS Hilgo Bruining1, Lourens Nonkes2, Hugo Oppelaar2, Herman van Engeland3, Hanna Swaab4, Martien Kas2 1 University Medical Centre Utrecht, Utrecht 2 Department of Pharmacology and Anatomy of Utrecht 3 Department of Child and Adolescent Psych, Utrecht 4 Department of Education and Child Studie, Utrecht Objective: To discover genetic pathways underlying social recognition and formation of social memory in mice using chromosome substitution (CS) strains. Method: To identify genetic loci responsible for social recognition and memory, a set of recently-generated CS-strains was screened in a social discrimination paradigm. In each of the CS-strains, a single chromosome from the donor strain (A/J) has been substituted in the genetic background of the host strain (C57BL/6J), providing an efficient approach to identify QTLs affecting developmental, physiological and behavioural processes. In this social discrimination paradigm, individual CS-mice were subsequently exposed to one and two intruder mice (one familiar and an unknown intruder) with two inter-exposure intervals (IEI) of 5 minutes and 24 hours, respectively. Social behaviour was scored using ObserverTM software (Noldus Information Technology, Wageningen). Results: Our preliminary data showed that some CS-strains, in contrast to the C57BL/6J control line, did not discriminate between the
781
familiar and unknown intruders after short- (5 min) and long-term (24 hours) IEI. Thus far, 3 CS-strains showed only short-term social discrimination and one CS-strain discriminated efficiently after 24 hours but not after 5 min. Conclusion: Our data suggest that short- and long-term social discrimination is controlled by different genetic pathways that can be further identified using combined genetic and genomic approaches. Understanding the molecular mechanisms underlying these dissected processes may be relevant for neuropsychiatric syndromes with impaired social cognitive performance, including Klinefelter Syndrome, autism and schizophrenia.
P191 INVOLVEMENT OF NR2B TYROSINEPHOSPHORYLATION IN EMOTIONAL RESPONSES Mina Delawary, Takanobu Nakazawa, Yuji Kiyama, Toshiya Manabe, Tadashi Yamamoto Medical Science, University of Tokyo, Minato-ku, Tokyo The NR2B protein is a critical structural and functional subunit of the N-methyl-D-aspartate (NMDA) glutamate receptor. NMDA receptor is involved in most aspects of higher brain function such as cognition, memory and learning. Dysfunction of NMDARs has been implicated in the etiology of schizophrenia. It is known that NR2B subunits are tyrosine-phosphorylated by various stimuli such as taste learning, ethanol treatment and lithium treatment. To investigate the biological role of NR2B tyrosine-phosphorylation, we have generated mice with a knockin mutation at the major phosphorylation site on NR2B, Tyr-1472 to phenylalanine (YF mice). YF mice show defects related to amygdala functions. In the elevated plus-maze test, time spent in openarms was reduced in YF mice as compared to that in wild-type mice. A similar phenotype was seen in corticotropin-releasing factor (CRF) overexpressing mice. This phenotype of YF mice was cancelled by the administration of CRF receptor antagonist. As expected, in YF mice, the expression level of CRF in the amygdala increased compared with that in wild-type mice. In the slice of amygdala from wild-type mice, NMDA application induced de-phosphorylation of Tyr-1472 and upregulation of the CRF mRNA level. Given that CRF is critically important in emotional responses in the amygdala, these data strongly suggest that phosphorylation of NR2B is involved in controlling emotional responses by regulating amygdaloid CRF content.
P192 INVOLVEMENT OF CANNABINOID CB2 RECEPTOR IN ALCOHOL PREFERENCE IN MICE AND ALCOHOLISM IN HUMANS Hiroki Ishiguro1, Shinya Iwasaki1, Lindsey Teasenfitz2, Susumu Higuchi3, Yasue Horiuchi4, Toshikazu Saito5, Emmaneul Onaivi2, Tadao Arinami4 1 Dept. of Med. Genet. Univ. of Tsukuba, Tsukuba, Ibaraki 2 William Paterson University, Wayne, NJ 3 Kurihama Alcoholism Center, Yokosuka, Kanagawa 4 University of Tsukuba, CREST/JST, Tsukuba 5 Sapporo Univeisity, Sapporo Cannabinoid type 2 receptor (CB2), mainly expressed in peripheral immune tissue, appears to play a role in alcohol abuse/dependence. In C57/BJ6 mice, expression of the Cnr2 gene was examined in brain regions with different tolerance/reinforcement levels for alcoholdrinking. Mice that developed more alcohol preference showed reduced CB2/Cnr2 gene expression, whereas mice with little reinforcement showed no changes in Cnr2 expression in the ventral midbrain. The amount of high-concentration alcohol (16–32%) was correlated with Cnr2 expression in that brain region. To evaluate the role of CB2 in alcohol-related behaviour combined with environmental factors, ethanol consumption was evaluated in mice treated daily with CB2 agonist JWH015 or antagonist AM630, in conjunction with chronic mild stress. Alcohol preference was enhanced in mice treated with JWH015 when subjected to chronic stress, whereas AM630 prevented development of alcohol preference. An association between the Q63R polymorphism of the human CNR2 gene and alcoholism in a Japanese population was found [p ¼ 0.007; Odds ratio 1.25, 95% confidence interval, (1.06–1.47)]. CB2 receptor gene, together with environmental factors, is associated with the physiologic effects of alcohol, and CB2 antagonists may have potential as therapies for alcoholism.
782
Abstracts
P193 INVERSE CORRELATION BETWEEN GAMMA BANDS AND SPIKING DISCHARGES IN DBA2J MICE SUGGEST A GABAB ROLE IN BEHAVIOURAL PERFORMANCE Francesco Marrosu1, Roberto Frau2, M. Orru`3, Monica Puligheddu4, Marco Bortolato3, Mauro Fa3, Luigi Barberini4, Giampaolo Mereu3 1 Universita` di Cagliari, Cagliari 2 Dipartimento Neuroscienze, Universita` di Cagliari 3 Dipartimento di Neuroscienze, Cagliari; Dipartimento di Scienze Cardiovascolari, Cagliari Purpose: DBA2J mice present decreased behavioural performances, cortical EEG alterations and characteristic spiking pattern (SSP), a landmark of rodent absence epilepsy. It has been suggested that GABAB-agonists worsened, while GABAB antagonists improved SSP. As recent studies suggested gamma EEG rhythms (21–50Hz) might be involved in optimally integrating sensory-motor performances, we carried out a study aimed at correlating EEG frequency with SSP after GABAB-receptor agonist (baclofen) followed by the antagonist SCH50911. Methods: Mice were EEG tested with the GABAB agonist L-baclofen followed by GABAB antagonist SCH 50911. Digital signals were processed at baseline, 10 and 20 min after L-baclofen; following SCH50911 antagonism, the EEG was also calculated at 10 and 20 min from drug administration. Power spectrum analysis included delta, theta alpha, beta and gamma (21–50 Hz) bands. Results: SSP and power spectrum of beta activity were increased by up to 80% following administration of 5 mg/Kg L-baclofen, while gamma decreased to the same extent. Following 50 mg/kg SCH 5911, SSP and beta power frequencies were reduced ( > 80%), while gamma increased (correlation 0.92; p < 0.001). Conclusions: Recent reports show that GABAB antagonists improve ‘‘cognitive’’ performances (e.g. maze solutions) in pharmacological rodent models of absence; it seems feasible to consider gamma frequency increase after SCH5091 in DBA2J a possible neurophysiological correlate of the above. However, as to whether the proportionally-inverse relation between SSP reduction and gamma increase can affect possible behaviour improvement in DBA2J mice remains to be established. P194 CHARACTERIZATION OF TRANSGENIC MICE EXPRESSING A TRUNCATED DISRUPTED-IN-SCHIZOPHRENIA-1 Sanbing Shen, David St Clair University of Aberdeen, Aberdeen Schizophrenia is a severe mental illness, which is diagnosed by a combination of positive and negative symptoms and impaired cognitive function. There are no objective tests, nor is there a convincing animal model. One of the most exciting findings in the genetics of schizophrenia is the discovery of the Disrupted-in-Schizophrenia-1 (Disc1) gene, which is truncated from intron 8 in a large Scottish family that develop schizophrenia. DISC1 encodes a coiled-coil protein that forms a large developmentally regulated, cytoskeleton-associated and multifunctional complex including Nudel, Lis1, Dynein, Fez1, 4-3-3e and PDE4B. Depletion of the endogenous Disc1 or expression of the truncated Disc1 impairs neurite outgrowth in vitro and cerebral cortical development in vivo. To examine how the truncated Disc1 is involved in the pathogenesis of schizophrenia, we have generated transgenic mice expressing the truncated Disc1 tagged with an EGFP to the exon 8, using a 148 kb mouse bacterial artificial chromosome. The transgenic heterozygotes were viable and appeared to express 2 copies of the truncated Disc1 at a level comparable to 2 copies of the endogenous Disc1, closely mimicking the genetic ratio in the Scottish family. Preliminary morphometric evidence showed that transgenic foetuses displayed trends of dilated lateral ventricles, and reduced cerebral cortex and hippocampus. However, they showed no gross motor abnormalities. Both hetero- and homozygotic transgenic mice will be fully analysed for neuroanatomical and behavioural changes, and the findings will be discussed in relation to human schizophrenic pathology and cognitive function. P195 FINE MAPPING OF TWO QTLs ON MOUSE CHROMOSOME 1 FOR SEDATIVE-HYPNOTIC DEPENDENCE AND WITHDRAWAL Kari Buck, Laura Kozell, Deaunne Denmark, Nikki Walter, John Belknap, John Crabbe Oregon Health & Science Univ. & VAMC, Portland, Oregon
Physiological dependence and associated withdrawal episodes can constitute a powerful motivational force that perpetuates drug use and abuse. Using robust behavioral models of drug physiological dependence in mice and fine mapping using a series of novel interval-specific congenic strains (ISCSs), we have identified two linked QTLs within a 3.4 Mb interval of chromosome 1 syntenic with human 1q21–q23. Together, these QTLs contribute equally for 26% and 25% of the genetic variance in alcohol and pentobarbital withdrawal severity in mapping populations derived from the DBA/2J (D2) and C57BL/6J (B6) progenitor strains, respectively. The more proximal 1.1 Mb QTL interval contains 39 genes and the distal 2.3 Mb QTL interval contains 59 genes. Additional ISCSs are being developed and tested for severity of withdrawal from ethanol and pentobarbital (and other sedativehypnotics) to further narrow these two QTL intervals. Comparison of D2 vs. B6 and ISCS vs. B6 microarray results detected evidence for genotype-dependent expression in the brain for 11 genes in the more proximal QTL interval and 10 genes in the distal QTL interval. Using confirmative QrtPCR, bioinformatics, and coding sequence SNP analyses, we are performing a systematic, nonbiased evaluation of the candidates for these two QTLs to identify the most promising genes to influence sedative-hypnotic dependence and withdrawal. Supported by NIH grants DA05228, AA011114, AA10760 and a VA Merit grant.
P196 THE FLINDERS RAT MODEL OF DEPRESSION: GENE-ENVIRONMENT INTERACTION AND ANTIDEPRESSANT EFFECTS BY PROTEOMIC ANALYSIS Chiara Piubelli1, Lucia Carboni2, Miriam Vighini2, Susanne Gruber3, Weronica Andersson3, Aram El Khoury3, Aleksander Mathe’3, Enrico Domenici2 1 GLAXOSMITHKLINE, Verona 2 GlaxoSmithKline—Psychiatry CEDD, Verona 3 Karolinska Institutet, Stockholm The Flinders Sensitive Line (FSL) has been proposed as a rat model for depression, due to its ability to mimic several features of the human disease. With the aim of investigating gene-environment interaction, we have examined the impact of an early life stress, such as maternal separation, and the effect of antidepressant treatment on the hippocampus of FSL and their controls, the Flinders Resistant Line (FRL). 2D-gel proteomics was selected as an unbiased approach for wide-scale analysis of hippocampal protein expression. This project is part of the GENDEP European Integrated Project, combining clinical pharmacogenomic/pharmacogenetic studies on depressed patients with preclinical investigations on animal models. FSL and FRL rats were subjected to maternal separation and both stressed and control animals were split into groups receiving chronic escitalopram treatment or vehicle. Hippocampal protein expression was evaluated by 2Dgel electrophoresis and subsequent image analysis. Comparisons among groups were performed by univariate and multivariate statistical analysis to detect changes induced by the stress exposure and the pharmacological treatment, associated with a different genetic background. This study shows that different groups of proteins are modulated by stress in hippocampus of the FRL and FSL rats as well as by the antidepressant treatment in the two gene-environment backgrounds. Identified proteins are involved mainly in cellular primary metabolism and biosynthesis, in protein processing, cytoskeleton organization, and in the regulation of neurogenesis and CNS development, suggesting new molecular correlates of vulnerability to depression, and of response to antidepressants. P197 THE PSYCHOGENETICALLY SELECTED ROMAN LOW- AND HIGH-AVOIDANCE RATS DIFFER IN THE BEHAVIORAL RESPONSES IN THE FORCED SWIMTEST Giovanna Piras1, Maria G. Corda2, Osvaldo Giorgi2 1 Dept of Toxicol—University of Cagliari, CAGLIARI 2 Department of Toxicology—University of, CAGLIARI The selection of Roman low-(RLA) and high-avoidance (RHA) rats to assess, respectively, poor vs rapid acquisition of active avoidance in a shuttle-box led to the segregation of other phenotypic traits that influence the higher emotionality of RLA rats as compared with RHA rats. Thus, stressors induce a more pronounced activation of the hypothalamo-pituitary- adrenocortical axis in RLA than in RHA rats.
Abstracts The Roman lines also differ in the functional properties of serotoninand dopamine-mediated neurotransmissions, which play key roles in the expression of anxiety- and depression-related behaviors. Our study was designed to compare the behavioral profiles of the Roman lines in the forced swim test (FST) in naı¨ve rats and in animals treated with the antidepressants desipramine (DMI) and fluoxetine (FLUO). RLA rats exhibited significantly longer duration of immobility than RHA rats in both, the pre-test (episodes/15 min: RLA, 79.8 3.7; RHA, 56.2 2.6, p < 0.05) and the control test session (episodes/5 min: RLA, 25.6 1.4; RHA, 13.0 1.8, p < 0.05). The administration of DMI (5, 10, and 20 mg/ kg x 3, SC) or FLUO (5, 10, and 20 mg/kg x 3, SC) before the test session induced a more robust reduction of the duration of immobility in RLA rats than in their RHA counterparts. These findings provide experimental support to the face and predictive validities of the RLA rat line as a genetic model of depression. P198 STUDY OF SGAs’ METABOLIC SIDE EFFECTS IN A MURINE MODEL WITH AND WITHOUT DIETARY RESTRICTION Cristina Mas1, Fe`lix Junyent2, Juani Utrera2, Ba´rbara Arias3, Carme Auladell2, Lourdes Fan˜ana´s3 1 University of Barcelona, Barcelona 2 Dep. Biologia Celular, Facultat de Biolog, Barcelona 3 U.Antropologia, Dep. Biologia Animal, Fa, Barcelona Background: Chronic treatment with certain second generation antipsychotics (SGA) has been associated with undesirable metabolic effects such as weight gain, lipid dysregulation, hyperglycaemia and diabetes. The aim of this study was to investigate whether metabolic side effects could be mediated through mechanisms independent of intake and weight gain. Methods: Two cohorts of mice (23 with ad libitum access to food (A) and 15 undergoing dietary restriction (B)) were assigned to three different groups of treatment (olanzapine (OL), quetiapine (QUE) and saline (C)). Mice were injected daily and controlled for their weight and intake for 21 days. Blood levels of glucose, insulin and triglycerides were obtained at the end of the experiment. Results: In cohort A, mice treated with OL increased their intake in comparison with the control group, however these differences were not significant. Interestingly, in cohort B, mice treated with OL showed a tendency towards weight gain in comparison with the control group (average weight gain: 0.13g and 0.03g, respectively). Moreover, both glucose and insulin levels were higher in OL group respect to the control group (t ¼ 2.30; P ¼ 0.05 and t ¼ 3.23; P ¼ 0.01; respectively). Triglycerides showed no clear differences between groups. The metabolic profile associated with QUE was similar to control group. Conclusions: Despite the limited sample size of the cohorts, our results suggest that olanzapine could increase weight and disturb metabolic parameters such as glycaemia and insulinemia through mechanisms independent of intake. Acknowledgements: Project SAF2005-07852-C02-01 P199 BEHAVIORAL STUDIES IN SYNAPSIN III KNOCKOUT MICE: RELEVANCE TO SCHIZOPHRENIA Hung-Teh Kao1, Ramona Rodriguiz2, John Gilbert IV2, Jian Feng3, Paul Greengard4, Barbara Porton1, William Wetsel2 1 NYU School of Medicine, Orangeburg, New York 2 Duke University School of Medicine, Durham, North Carolina 3 SUNY Buffalo, Buffalo, New York 4 Rockefeller University, New York, New York The synapsin III gene has been implicated in schizophrenia by previous postmortem and genetic studies. The role of this gene in neural function is also consistent with the neurodevelopmental hypothesis of this disorder. Here, we report that postmortem levels of synapsin III are significantly decreased in the dorsolateral prefrontal cortex of individuals with schizophrenia, a finding that is consistent with a previous report demonstrating decreased levels of synapsin III in the hippocampi of individuals with schizophrenia. Reductions of synapsin III protein in two brain regions thought to play a role in schizophrenia provide a strong rationale for investigating the behavioral phenotype of mice lacking the synapsin III gene. A battery of behavioral tests were performed. Synapsin III knockout mice are hyperactive in the open field and they display abnormal responses to amphetamine and MK-801. These mice are also deficient in prepulse inhibition and this deficit was
783
restored by clozapine or raclopride, but not by haloperidol. In addition, synapsin III knockout mice exhibit abnormalities in latent inhibition, contextual and cued fear conditioning, and expression of fear potentiated startle. As the synapsin III KO mice do not harbor obvious anatomical deficits or neurological disorders, they may represent a unique neurodevelopmental model for dissecting the molecular pathways that contribute to certain behavioral abnormalities associated with schizophrenia. P200 GENERATION AND CHARACTERIZATION OF TEMPORAL TISSUE-SPECIFIC INDUCIBLE CB1 KNOCKOUT MICE M. Collu, M.F Marongiu, S. Porcu, M. Melis, W. Fratta, P. Fadda, M.S. Ristaldi Dept. of Neuroscience, University of Cagliari and CNR Neurogenetics and Neuropharmacology Institute, Cagliari, Italy Cannabis abuse has been related to several psychiatric disorders such as dependence, anxiety, depression, cognitive impairment, and psychosis, and the endocannabinoid systems seems to be involved in the pathophysiology of different mental disorders. Several approaches have been employed to determine the biological function of CB1 receptors. Recently the new technology of gene targeting in mice has permitted to study the function and physiological role of the cannabinoid system using knockout (KO) mice in which the gene encoding the CB1 receptor has been deleted. Such mutant animals (CB1 KO mice) are a powerful model to study the receptor function using physiological, biochemical, morphological and behavioural approaches. However, as for all conventional KOs, some compensations for the loss of a gene product are obviously present. To overcome this limitation we have used a gene expression system that may be induced any time in mice from the embryonic to the adult life. The tetracyclineregulated expression system, the most utilized and best developed approach to inducible gene regulation, allows turning ‘‘off’’ or ‘‘on’’ the expression of a transgene through the assumption of tetracycline or its derivative doxicyclin. We have generated an inducible temporal and tissue-specific KO CB1 mice in which the ability to modulate the expression of CB1 gene in specific brain areas with the CaMKII (Ca2þ/ Calmodulin-Dependent Protein Kinase II) promoter represents a powerful tool for studying the physiological and pathological role of the cannabinoid system.
P201 TUMOUR NECROSIS FACTOR-ALPHA -863C/A POLYMORPHISM IS ASSOCIATED WITH ALZHEIMER’S DISEASE Atsushi Tsutsumi, Masaki Nishiguchi, Hiroki Kikuyama, Akira Hyoukyou, Jun Koh, Hiroshi Yoneda Osaka Medical College, Takatsuki, Osaka Background: In Alzheimer’s disease there is increasing evidence of neurotoxicity which is mediated by CNS inflammatory and apoptotic processes. TNF-alpha is one of the proinflammatory cytokine. TNFalpha and glutamate synergistically stimulate toxic activation of their respective signaling pathways in neurons as a contributing mechanism of cell death in Alzheimer’s disease. It is reported that 308G/G polymorphism of TNF-alpha gene has a correlation to Alzheimer’s disease in Caucasian, but in Japanese the manifestation of 308G/G polymorphism is only around 1%, and the association to Alzheimer’s disease is denied. A high-risk TNF-alpha haplotype (1031C, 863C, 857C) for Alzheimer’s disease was identified in Chinese. These three polymorphisms were emerged in 0.17–0.30, and have a high frequency in Japanese. We investigated whether these three polymorphisms (1031T/C, 863A/C, 857C/T) might be responsible for susceptibility to Alzheimer’s disease in Japanese. Methods: A total of 69 Japanese patients with a DSM-IV diagnosis of Alzheimer’s disease and 96 healthy controls were recruited. Genomic DNA was isolated from whole blood according to standard procedure. The genotyping was performed by the fluorescence resonance energy transfer method using the LightCycler System. Results: Frequency of the 863C allele is significantly increased in patients with Alzheimer’s disease as compared to controls. Conclusion: These results showed that the 863C allele of TNF-alpha gene is a genetic risk factor in Japanese patients with Alzheimer’s disease.
784
Abstracts
P202 ROLE OF APOE, ACT AND CHRNA7 GENES IN THE CONVERSION FROM MILD COGNITIVE IMPAIRMENT TO DEMENTIA Jose Antonio Cabranes1, Carmen Y de Ugarte1, Blanca M Va´zquez-Alvarez1, Ine´s Ancı´n2, Alberto Marcos2, Pedro Gil2, Raquel Rodriguez2, Marta Encinas2, Natalia Alvarez2, Juan Jose´ Lo´pez-Ibor2, Ana Barabash2 1 Hospital Clinico San Carlos, Madrid 2 Hospital Clı´nico San Carlos, Madrid The combination of tests on genes involved in the pathogenesis of Alzheimer Disease (AD) may be informative of progression from Mild Cognitive Impairment (MCI) to dementia. The apolipoprotein E (APOE) genotype, the Ala/Thr polymorphism at the signal peptide of the alpha1-antichymotripsin (ACT) gene and the 86 C/T polymorphism at the promoter region of nicotinic receptor alpha 7 gene (CHRNA7) were studied in 89 patients with MCI. After a median follow-up of 49.6 months, 27 subjects remained with MCI and 62 subjects developed AD. 40 converted to AD before 17 months. A similar control group (n ¼ 90) with MMSE of 28 was included. Possession of at least one APOE4 allele was associated with risk of MCI (OR:6,04; 95% CI:2,76€C 13, 23; p < 0.001) but did not have an effect on the cumulative hazard of developing AD. ACT or CHRNA7 genotypes were not associated with MCI but oppositely modified the predictive cumulative hazard of progressing to AD. Presence of the T allele of CHRNA7 was associated with a 50% reduction in the probability to progress to AD (HR ¼ 0,5 95% CI: 0,2€C 1,5; p ¼ 0,1), while presence of allele T of ACT was associated with higher risk of faster progression to AD (HR ¼ 2,03; 95% CI: 1–4,6; p ¼ 0,06). We concluded that APOE is a risk gene for MCI and that ACT and CHRNA7 are modifier genes of conversion to AD.
P203 HOMER1 PROMOTER ANALYSIS IN PARKINSON’S DISEASE: ASSOCIATION STUDY WITH PSYCHOTIC SYMPTOMS A. de Bartolomeis1, G. Annesi2, G. Nicoletti2, P. Pugliese2, J.L. Kennedy3, P. Barone1, A. Quattrone2 1 Universita` di Napoli ‘Federico II’, Napoli 2 CNR, Cosenza 3 University of Toronto, Toronto Homer1 gene expression has been linked to abnormal movements in animal models. Genetic variability within different candidate genes has been implicated in the clinical severity of sporadic Parkinson’s (PD) in many populations. The continuing neurodegeneration of PD and the prolonged use of levodopa are associated with motor complications such as dyskinesia and psychotic side-effects, including hallucinations and paranoid delusions. Approximately 25% of patients with idiopathic Parkinson’s disease experience hallucinations. We are now investigating the polymorphisms located in the 50 flanking region of the HOMER1 gene within a sample of 261 sporadic Parkinson’s from Southern Italy, using a three SNPs genotype and haplotype combination (rs4704559, rs1094289, rs470560). Our study implicates the effect of allele A of rs4704559 marker in susceptibility to psychotic symptoms in PD (chisq ¼ 4.35, 1df, p ¼ 0.036) and more specifically defines an OR (95% CI) of 9.09 (1.18–69.69) for genotype AA, which is overrepresented in patients with hallucinations. Even though our results are preliminary, this HOMER1 gene variant may represent a biomarker for longitudinal evaluation of PD patients. P204 SEROTONIN TRANSPORTER AND GLICOGEN SYNTHASE KINASE GENES IN ALZHEIMER’S DISEASE Elena Marino1, Cristina Lorenzi2, Paola Artioli2, Adele Pirovano2, Marco Catalano2, Enrico Smeraldi2 1 University Vita-Salute—S. Raffaele, MILAN 2 University Vita-Salute—S. Raffaele, MILAN Literature has shown that Serotonin Transporter (SERT), Tryptophan Hydroxilase (TPH) and Glycogen Synthase Kinase-3 (GSK3) genes play a partially-unknown role in dementia. Recent studies have identified a significant age-related decline in brainstem and SERT density. These findings support the role of SERT as a candidate for neurodegenerative disease. In Alzheimer’s disease (AD) patients, increased GSK-3b protein levels are found in the brain, and GSK was associated with paired helical filaments. This study attempted to evaluate genetic
assessment on 163 subjects (dements n ¼ 85 versus healthy subjects n ¼ 78), focusing on SERTPR polymorphism, on TPH SNP (218A/C) and on two GSK-3b SNPs (50C/T, 1727A/T). The state of dementia was evaluated by two tests: Mini Mental State Exam (MMSE) and Milan Overall Dementia Assessment. Subjects underwent three MMSEs, at three different times. In our sample, chi-square analysis confirmed that females are more susceptible to dementia (p ¼ 0,04) than to AD. Moreover, we demonstrated that subjects affected by AD showed significantly worse cognitive impairment, compared to other demented subjects, during observation time (MMSE1 p ¼ 0.0051; MMSE2 p ¼ 4.79 105; MMSE3 p ¼ 5.19 105). Genetic analysis showed an association trend (p ¼ 0.06) between GSK-1727 polymorphism and dementia; genotype AA seems to play a protective role against dementia in healthy males. This preliminary result, although only slightly significant, prompts us to perform further studies on the influence of genetic patterns on dementia.
P205 PLASMA HOMOCYSTEINE AND THE RISK OF PROGRESSION IN MILD COGNITIVE IMPAIRMENT Tomasz Gabryelewicz1, Maria Styczynska1, Wojciech Androsiuk2, Maria Barcikowska1 1 Polish Academy of Sciences, Warsaw 2 Dep. of Psychiatry, Medical University of Warsaw Objective: To establish the rate of progression to dementia in MCI. To investigate the risk of progression for different subtypes, and to find the potential predictors of progression. Methods: MCI (n ¼ 105) individuals enrolled in longitudinal study at an Alzheimer’s Day Clinic in Warsaw received annual clinical and psychometric examinations for up to mean 3 years. The diagnosis of MCI according to Mayo Clinic Petersen’s Criteria was made by the panel of specialists. 42 subjects were classified as amnestic MCI and 63 as multiple-domain MCI with memory impairment. Results: After 3 years of follow-up, 23 of 105 subjects with MCI were diagnosed with dementia. 40 showed cognitive decline not dementia, 34 were stable, while 8 displayed cognitive improvement. The risk of conversion to dementia was higher among the subjects with multipledomain MCI. Moderately elevated plasma tHcy levels (14.3 mol/l; SD 4.2) were observed in 60.95% of MCI sample compared with normal ranges. The multiple-domain group had statistically higher mean plasma tHcy levels than amnestic group (15.2 vs 12.8), and subjects who converted to dementia had statistically significant higher baseline mean plasma tHcy levels than non-converters (16.53 vs 14.36). Conclusions: We conclude that the risk of conversion to dementia was higher among the subjects with multiple-domain MCI than amnestic MCI, and the occurrence of elevated plasma tHcy levels may constitute a predictor for those who are more likely to progress to dementia.
P206 APOLIPOPROTEIN E AND DEMENTIA IN THE CUBAN POPULATION Beatriz Marcheco-Teruel1, J.J. Llibre2, L.E. Fuentes1, Y. Herna´ndez1, L.C. Marı´n1, M. Go´mez1, C. Ferri2, M. Prince2 1 National Center of Medical Genetics, Havana 2 Carlos J. Finlay University Hospital, Hi, Havana 2 Section of Epidemiology, Institute of Ps, London In Cuba, Alzheimer’s disease is the most frequent cause of dementia followed by vascular dementia. Apolipoprotein E-epsilon 4 isoform has been associated with increased risk for developing Alzheimer’s disease in different populations. The ApoE genotype was determined in 1909 individuals 65 years and older. All of them were clinically evaluated: 1718 were normal and 191 had a diagnosis of dementia. ApoE genotyping was performed by polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) and the lipid profile was assessed. The ApoE allele frequencies were 6.4%, 85.3% and 8.3% for ApoE2, ApoE3 and ApoE4 respectively. The presence of ApoE-epsilon 4 allele was significantly associated with increased risk of dementia (Chisquare ¼ 13.41, p ¼ 0.0003, OR ¼ 1.93) and high levels of cholesterol (Chi-square ¼ 7.58, p ¼ 0.0059, OR ¼ 1.6). The ApoE-epsilon 4/4 genotype was more frequent in individuals of African descent compared to Caucasian descent (70% VS 30%). The origin and admixture of the Cuban population is a valuable resource to study genetic risk factors and ethnic variability in the dementias.
Abstracts P207 INTERLEUKIN-1 ALPHA GENE IN SPORADIC ALZHEIMER DISEASE. ASSOCIATION WITH GENDER AND DEPRESSIVE SYMPTOMS Paolo Olgiati1, Antonis Politis2, Alessandro Serretti1, Diego Albani3, Gianluigi Forloni3, Petros Malitas2, Christina Piperi4, Evangelia Stamouli2, Antonis Mailis2, Anna Rita Ratti1, Monica Morri1, Mandola Ujkaj1, Francesca Prato3, Sara Batelli3, Diana De Ronchi1, Anastasios Kalofoutis4 1 Institute of Psychiatry, University of Bologna, Bologna, Italy 2 Department of Psychiatry, Eginition Hospital, University of Athens Medical School, Athens, Greece 3 Neuroscience Department. Istituto di ricerche Farmacologiche ‘‘Mario Negri’’, Milan, Italy 4 Laboratory of Biological Chemistry, University of Athens Medical School, Athens, Greece Since inflammatory processes have been implied in the pathogenesis of Alzheimer Disease (AD), a number of studies have addressed the effects of interleukin-1 (IL-1) gene polymorphisms on the risk for developing AD. IL-1alpha is coded by IL1A gene. The IL1A has a common polymorphism in the 50 regulatory region (a C to T transition at position 889 relative to the start site of transcription—rs1800587) with possible functional effects. IL1A T/T genotype has been associated in a recessive manner to Alzheimer Disease (Rainero et al., 2004), to liver cirrhosis in patients with chronic hepatitis C (Bahr et al., 2003) and to schizophrenia (Katila et al., 1999). The same variant has been associated with depressive symptomatology and antidepressant outcome (Yu et al., 2003). The present study compared IL-1 alpha rs1800587 variants between 75 patients with sporadic AD and 60 nonpsychiatric controls. Forty-seven patients and forty-seven controls were from Athens, Greece, while twenty-eight patients and thirteen controls from Faenza, Italy. No stratification effect was observed for the two ethnic groups (Mantel-Haenszel test p ¼ 0.50). No association between IL1A gene and AD was found in the overall sample. However a higher frequency of C/T heterozygotes vs C/C homozygotes (p ¼ 0.04) and a trend towards greater prevalence of T allele carriers (p ¼ 0.09) were reported among females with AD compared to controls. Regardless of gender T/T homozygosity was found to be coupled with lower depression (assessed by the Cornell Scale for Depression in Dementia) in AD (p ¼ 0.03) after covarying for age and illness duration. The present findings, although preliminary, are in line with previous reports and are cues to understand the role of inflammatory response in the pathophysiology of sporadic AD. This might be prominent in females and in the onset of mood disturbance.
P208 MIGRAINE AND TUMOR NECROSIS FACTOR GENE POLYMORPHISM: AN ASSOCIATION STUDY IN A SARDINIAN SAMPLE Asuni C, Stochino M.E, Cherchi A, Manconi M., Congiu D, Del Zompo M. Centro per lo Studio e la Terapia delle Cefalee Primarie ‘‘Franco Tocco’’, Sezione di Farmacologia Clinica, Dipartimento di Neuroscienze ‘‘B.B. Brodie’’—Universita` di Cagliari—ASL8, Cagliari. Background: Migraine is a complex genetic disease, in which susceptibility genes and environmental factors contribute to development of this illness. The physiopathology of migraine is still unknown but the ‘‘sterile inflammation’’ or ‘‘neurogenic inflammation’’ seems to have a key role. The Tumor Necrosis Factors (TNF) (TNF- e TNF-) are cytokines implicated in inflammatory reactions and endothelial function. Several studies suggest that TNF may be involved in migraine. The TNF- and TNF- genes are located on chromosome 6p21.3 in the human leukocyte antigene (HLA) region. Objective: To assess the possibility of an association between TNF genes polymorphisms and migraine disease, a case-control study was performed in a Sardinian sample. Methods: We evaluated 299 patients affected by migraine without aura (I.H.S. criteria 2004) and 278 migraine-free controls. The polymorphisms G308A in the promoter region of the TNF gene, and G252A in the first intron of TNF- polymorphisms were determined by NcoI restriction fragment length polymorphism analysis. Results. We found a statistically significant difference in allele (p ¼ 0.018; O. R. ¼ 1.46 95% CI: 1.066 to 2.023) and genotype (Trend w2 ¼ 5.46 df ¼ 1 p ¼ 0.019) frequencies of TNF-, between cases and controls. Allele and
785
genotype frequencies of TNF G308A polymorphism did not differ significantly between the two groups. Conclusion. These data suggest that subjects with TNFB2 allele have a low risk of developing migraine without aura and that the polymorphism of the TNF- gene is in linkage disequilibrium with other migraine responsible genes in the HLA region. P209 ALZHEIMER’S DISEASE AND GENES INVOLVED IN NEURONAL CHOLINERGIC TRANSMISSION: A CASE-CONTROL ASSOCIATION STUDY IN A SARDINIAN SAMPLE F. Manconi1, D.Congiu1, A. Squassina1, R.M. Mereu2, G. Obino2, D. Viale2, P.F. Putzu2, C. Chillotti1, M.P. Piccardi1 1 Unit of Clinical Pharmacology, Department of Neurosciences ‘‘B. B. Brodie’’, University of Cagliari 2 Local Health Agency 8, Cagliari, Italy Background: Alzheimer’s disease (AD) is characterized by an extensive loss of cholinergic neurons, and their cortical projections, from the basal forebrain area. The resulting reduction in cholinergic activity is associated with decreased levels of the neurotransmitter acetylcholine (ACh), decreased activity of acetylcholinesterase (AChE), choline acetyltransferase (ChAT), and increased butyrylcholinesterase (BChE) activity. In the present study, we investigated the possible involvement of AChE, ChAT and BChE polymorphisms in the pathogenesis of AD. Methods: AD patients (n ¼ 158), exclusively of Sardinian ancestry (parents, grandparents and great-grandparents), were recruited from the Unit of Clinical Pharmacology, Department of Neurosciences, University of Cagliari, and Division of Geriatrics Local Health Agency 8. Patients were diagnosed according to DSM-IV, and National Institute of Neurologic and Communicative Disorders and Stroke–AD and Related Disorders Association (NINCDS-ADRDA) criteria for possible or probable AD. Cognitive screening was performed by means of Mini-Mental State Examination (MMSE). Healthy controls (n ¼ 118), of Sardinian ancestry, were recruited from among healthy relatives of patients and from religious and sport associations. All patients and control subjects gave informed consent for participation in the study. SNP analysis was performed by PCR/RFLP or an ABI Prism 7900HT instrument using Assays-On-Demand reagents from Applied Biosystems, Inc. (Foster City, CA) Results: No significant differences in allele, genotype or haplotype frequencies were observed between patients with AD and controls. Conclusion: These results suggest that the AChE, ChAT and BChE polymorphisms do not constitute a major genetic risk factor for susceptibility to Alzheimer’s disease in a Sardinian population. Acknowledgements: this project is funded by the Ministero della Salute- Progetti di Ricerca Finalizzata
P210 ASSOCIATION OF HTR2C WITH METABOLIC SYNDROME IN PATIENTS USING ANTIPSYCHOTICS Barbara Franke1, H. Mulder2, A. Aart van der–Beek van der3, J. Arends4, F.W. Wilmink4, H. Scheffer5, A.C.G. Egberts6 1 Radboud University Nijmegen Med. Centre, Nijmegen 2 Department of Clinical Pharmacy, Wilhelm, Assen 3 Department of Clinical Pharmacy, Martini, Groningen 4 Psychiatric Hospital GGZ Drenthe, Assen 5 Department of Human Genetics, Radboud Un, Nijmegen 6 Department of Pharmacoepidemiology and P, Utrecht Introduction: The use of antipsychotic drugs is associated with severe weight gain, and therefore with an increased risk of cardiovascular morbidity. Association studies on weight gain and obesity with SNPs within the promoter region of the X-chromosomal HTR2C gene encoding the serotonin 2C receptor have yielded conflicting results. Methods: Patients were identified from a schizophrenia diseasemanagement programme. In this programme, patients’ blood pressure, triglycerides, HDL-cholesterol, weight and waist circumference are measured regularly during follow-up. Primary endpoint was the prevalence of the metabolic syndrome as classified by the NCEP: ATP III. Primary determinants were polymorphisms in HTR2C (promoter: c.1-142948(GT)n,rs3813928:G > A (997 G/A), rs3813929:C > T (759 C/T), rs518147:G > C (697 G/C); intron 5:rs1414334:C > G). The association of HTR2C variants and components of the metabolic
786
Abstracts
syndrome was evaluated with logistic regression and expressed as adjusted odds ratios (OR, 95% CI). Results: In total, 112 patients with chronic psychiatric disorders were included. They mainly ( > 80%) used atypical antipsychotics. Carriership of variant alleles of the polymorphisms HTR2C:c.1-142948(GT)n and rs1414334:C > Gwas associated with components of the metabolic syndrome (OR 2.50 (1.11–5.66) and OR 3.76 (1.25–11.29), respectively). Haplotype analysis showed an OR of 4.69 (1.34–16.45) for the combined presence of these variants. Conclusion: HTR2C polymorphism is associated with an increased risk of the metabolic syndrome in patients taking antipsychotics.
P211 COMPARISON OF WHOLE-GENOME AMPLIFICATION METHODS USING THREE GENOTYPING PLATFORMS CJC Curtis, E Meaburn, IW Craig, R Plomin MRC SGDP Centre, London Introduction: In order to identify quantitative trait loci (QTLs) of small effect size, large samples and large numbers of genetic markers are required. Consequently, the availability of large quantities of goodquality DNA is critical unless permanent cell lines are available, which are expensive for large samples. During the past decade, several techniques have been developed to achieve in vitro amplification of genomic DNA. The goal of whole-genome amplification (WGA) is to boost DNA stocks and end the need for re-sampling by providing high yields of unbiased amplification products of the entire DNA template from small amounts of DNA. Method: We compared four WGA methods for quantity, quality and accuracy of DNA product produced. We also investigated different DNA sample types by using four samples obtained from buccal swabs and one sample from blood. These samples were amplified using the WGA methods and interrogated using three diverse genotyping platforms: an in-house 18- marker simple sequence repeat (SSR) multiplex system, a 48-marker multiplex system (SNPlexTM), and the Affymetrix GeneChip1 Human Mapping 100K SNP Array. DNA from the four WGA methods was compared to the original non-amplified DNA for the five individuals. Results: Although the four WGA methods differed in relation to performance, cost and ease of use, they all yielded high-quality amplified DNA. Conclusion: WGA is a viable technique to amplify DNA for high density SNP analysis. P212 THE HIGH ACTIVITY ALLELE OF CATECHOL-O-METHYLTRANSFERASE (COMT) GENE IS ASSOCIATED WITH RITALIN NON-RESPONSE IN ADHD CHILDREN Eva Kereszturi1, Zsofia Nemoda1, Zsanett Tarnok2, Julia Gadoros2, Judit Gervai3, Ma´ria Sasvari-Szekely1 1 Semmelweis University, Budapest 2 Adaskert Child and Adolescent Psychiatry, Budapest 3 Hungarian Academy of Sciences, Budapest Introduction: Stimulant medications, such as methylphenidate (Ritalin), are the most commonly used, effective treatment for Attention Deficit Hyperactivity Disorder (ADHD). Ritalin acts primarily by inhibiting the dopamine transporter (DAT1), a protein responsible for the reuptake of dopamine from the synapse into presynaptic terminals. The catechol-O-methyltransferase (COMT) has also been found to influence drug response. However, individual differences between patients’ response to methylphenidate are known. Methods: ADHD was diagnosed according to DSM-IV; the repeat polymorphisms of the DAT1 30 -untranslated region (30 -UTR) variable number tandem repeats (VNTR) and COMT Val158Met polymorphism were analysed by PCR-based methods. A case-control association study was carried out to compute association between the phenotype and candidate polymorphisms. Results: Ritalin treatment was effective for 94 patients (72.9%). Allele and genotype frequencies of COMT variants were significantly different comparing responders and non-responders. The Val allele was observed more frequently in responders (P ¼ 0.007). Genotype distribution of this polymorphism was also significantly different (P ¼ 0.029). The Val/Val genotype possessing high activity was represented twice as frequently in the Ritalin-responder group. On
the other hand, no association was found between the DAT1 repeat polymorphism and ADHD. Conclusion: A significant association was found between the Val/Val genotype of the COMT polymorphism and ADHD. This allele variant was previously described as resulting in higher enzyme activity of the protein. These findings could contribute to clarifying the molecular pathology of ADHD. P213 USING CYP2D6 GENOTYPE TO UNDERSTAND PAROXETINE-PIMOZIDE PHARMACOKINETICS IN MAN Elizabeth Foot1, Linda Briley2, Linda McCarthy2, Rachel Moate2, Brian Reck2, Keith Muir2, Carole Stapleton2, Charles Cox2, David Dow2, Nalini Mehta2, Ian Latham2, Neil Goodgame2, Marika Vousden2, Gianluca Nucci2 1 GlaxoSmith Kline, Greenford, Middlesex 2 GlaxoSmithKline, Greenford, Middlesex Introduction: Paroxetine, a selective serotonin reuptake inhibitor (SSRI) for treatment of unipolar depression and anxiety disorders, is a known inhibitor of CYP2D6 with little or no effect on CYP3A4 in vitro. Pimozide, a dopamine receptor antagonist for management of Tourette’s Disorder and which can cause QT prolongation, is considered metabolized primarily by CYP3A4, although CYP2D6 and CYP1A2 may contribute. A paroxetine/pimozide interaction study was initiated when Sertraline, another SSRI, was found to increase pimozide exposure by 40%, leading to contraindications for co-administration. Methods: An open label, single sequence clinical study to investigate the safety and pharmacokinetics (PK) of oral pimozide when coadministered with repeat oral dosing of 60mg paroxetine in healthy volunteers. Pharmacogenetic blood samples were obtained. Results: In a post-hoc analysis, 33 subjects were genotyped for CYP2D6 polymorphisms known to alter CYP2D6 enzymatic function. Based on CYP2D6 genotype, subjects’ predicted metaboliser status were: UM (1), EM (25) IM (4), PM (2) and unknown (1). The two subjects with highest pimozide AUC were CYP2D6 PMs, suggesting an important role for CYP2D6 in Pimozide metabolism. Pimozide exposure was increased in the combination group (AUC increased 2.5-fold), suggesting pimozide metabolism was inhibited following repeated dosing with paroxetine. The magnitude of the pharmacokinetic interaction was ranked (greatest to least) UMs > EMs > IMs > PMs. These exploratory results suggested CYP2D6 metaboliser status impacts the metabolism of pimozide and that the PK interaction between pimozide and paroxetine can be explained using the known CYP2D6 inhibitory properties of paroxetine. P214 ASSOCIATION BETWEEN 5-HT2A RECEPTOR GENE T102C AND 1438G/APOLYMORPHISM AND CLOZAPINE TREATMENT RESPONSE Hasan Herken1, M. Emin Erdal2, Mesut Cetin3, Omer Barlas2, Cengiz Basoglu3, Ozcan Uzun4, Salih Selek5, Tuba Gokdogan2 1 Pamukkale University, Denizli 2 Medical Biology and Genetics Departme, Mersin 3 Department of Psychiatry, GATA Haydar, Istanbul 4 Department of Psychiatry, GATA Etlik, Istanbul 5 Department of Psychiatry, Gaziantep U, Gaziantep Aim: 5-HT2A receptor gene is a polymorphic gene which probably predisposes to schizophrenia and whose polymorphism is related to therapeutic response. In this study we aimed to explore the relationship between the response to clozapine and receptor gene T102C and 1438 G/ A polymorphism. Method: The study was conducted with 103 schizophrenia patients having had 2 months of clozapine treatment in three clinic outpatient units. Method: Psychotic Rating Scale (BPRS) was applied to each patient before and after 2 months of treatment. Results: 5-HT2A receptor 1438 G/A variant G/G genotyped patients had significantly higher hospital admission rates than G/A A/A genotyped patients (P ¼ 0.001, df ¼ 89, F ¼ 7.41). Similarly, T102C variant C/C genotyped patients had higher hospital admission rates than T/C ve T/T genotyped patients (P ¼ 0.001, df ¼ 89, F ¼ 10.8). 5-HT2A receptor 1438 G/A variant G/G genotyped patients also responded less to clozapine treatment than A/A G/A genotyped patients and vice versa (2 ¼ 7.2, df ¼ 2, P ¼ 0.027). 1438G/A polymorphism G/G genotyped patients, who responded less and were more treatment
Abstracts resistant, had significantly more T102C polymorphism C/C genotype, which was associated with treatment resistance ((2 ¼ 118.1, df ¼ 4, P ¼ 0.0001). Conclusion: Our findings point out that there may be an association between 5-HT2A gene polymorphism and antipsychotic treatment. P215 POLYMORPHISMS OF THE D2 AND D3 RECEPTOR GENES AND CLINICAL RESPONSE TO OLANZAPINE Luiza Olajossy-Hilkesberger, Beata Godlewska, Marcin Olajossy, Halina Marmurowska-Micha3owska Department of Psychiatry, Lublin Objective: Dopamine D2 and D3 receptors are strongly implicated in the mode of action of antipsychotic drugs. The aim of the study was to investigate whether polymorphisms of the D2 receptor gene (141C Ins/Del) and D3 receptor gene (Ser9Gly) influence olanzapine response in patients with schizophrenia. Methods: We studied 99 Caucasian subjects meeting DSM-IV criteria for schizophrenia, undergoing treatment with olanzapine (daily dose— 20mg). Psychopathology was measured before and after 6 weeks of treatment with PANSS scale. Clinical improvement was quantified as change in PANSS total scores and subscores as shown by the percentage improvement below baseline score. Results: A significant difference in response to treatment in terms of PANSS positive and negative symptom subscores improvement was seen between males and females (greater improvement in females in both subscores: p < 0.05; p < 0.01). The Ser9Gly polymorphism of D3 receptor gene was significantly associated with the percentage improvement in PANSS negative symptom subscore (better response in Ser/Gly heterozygotes - p < 0.05). The analysed polymorphism of D2 receptor gene was not significantly associated with clinical improvement determined by changes in PANSS scale. Conclusions: These results suggest that variations in the D3 receptor gene may influence individual and particularly negative symptom response to olanzapine. P216 THE PROTEOMICS OF ANTIDEPRESSANT DRUG RESPONSE Patrick Christopher McHugh, Geraldine R Rogers, Peter R Joyce, Martin A Kennedy University of Otago, Christchurch Genetic variation may affect individual responses to antidepressant drug treatment. As a means of identifying molecules that may be important in antidepressant response, we explored the molecular effects of the selective serotonin reuptake inhibitor paroxetine on cultured neural cells. For this purpose, we developed a mouse embryonic stem (ES) cell-derived neural culture model and determined the presence of several cell types, including serotonergic and catecholaminergic neurons. We exposed the cultured neurons to one micromolar paroxetine for two weeks and applied proteomic analysis to identify proteins with altered expression levels. We identified eleven proteins that were differentially expressed between cultured neural cells exposed or not exposed to paroxetine. Of these eleven proteins, two molecular chaperones (PDIA3 and HSPA9A) and two intermediate filament proteins (GFAP and VIM) were observed to have altered expression, and these may be of relevance to clinical responses to antidepressant treatment. It is possible that inherited variation in the corresponding genes of the proteins identified could contribute to interindividual variability in antidepressant response. We propose to examine polymorphic variants of the human orthologues, and test these by association studies in appropriate patient cohorts for relevance to antidepressant drug response. In summary, using a novel model cell culture system, we looked at the effects of paroxetine on protein expression by employing a proteomic approach. We identified a number of proteins that may have relevance to antidepressant function and may provide insights into the neurobiology of depression. P217 5-HT2C GENE VARIANT INFLUENCES OLANZAPINE-INDUCED WEIGHT GAIN IN MEN BUT NOT IN WOMEN Beata Godlewska1, Luiza Olajossy-Hilkesberger2, Magdalena Ciwoniuk1, Jerzy Landowski1
787
1
Medical University of Gdansk, Gdansk Medical University of Lublin, Lublin
2
Introduction: Weight gain is a major limitation associated with treatment using certain antipsychotic drugs. There is substantial evidence that individual variability in drug-induced weight gain may relate to genetic factors, of which the 759C/T polymorphism in the 5-HT2C receptor is best, although not totally consistently, established. There is also a 697G/C polymorphism in partial linkage disequilibrium with 759 which has not previously been investigated in this context. We have undertaken an analysis of these polymorphisms in a well-characterised sample of patients receiving olanzapine. Methods: Patients (53 males, 53 females) were monitored over 6 weeks following initial olanzapine administration, and weight and body-mass index (BMI) recorded. DNA from blood was genotyped for the 759 and 697 5-HT2C polymorphisms and association with change in weight and BMI assessed. Results: While no significant overall association of genotype with measures of weight gain was observed, an association within the males, but not females, was identified in which the-759T allele predicted less percentage weight gain (p ¼ 0.03) and less increase in BMI (p ¼ 0.027). This remained after removing 11 obese subjects (BMI > 30), when significance increased. Association with 697G/C polymorphism was not significant in either group. Conclusion: These data demonstrate that, in this Polish sample, olanzapine-induced weight gain is predicted in part by a functional 5-HT2C promoter polymorphism in male, but not female, subjects. It would appear that the 697G/C polymorphism has less effect on druginduced weight gain. P218 ASSOCIATION STUDY OF THERAPEUTIC RESPONSE WITH SERT GENE POLYMORPHISM IN FEMALE SCHIZOPHRENIC PATIENTS Martina Rojnic1, Nada Bozina2, Vesna Medved3, Nikolina Jovanovic3 1 Department of Psychiatry, University Hos, Zagreb 2 Clinical Institute of Laboratory Diagnos, Zagreb 3 University Hospital Center Zagreb, Zagreb Scientific findings suggest that the serotonin system may be implicated in treatment response to second generation antipsyhotics (SDAs) in psychotic disorders. Since serotonin actions on its receptors are limited by serotonin transporter (SERT), we hypothesised that serotonin transporter gene variants might influence treatment response. The functional polymorphism of the serotonin (5-HT) transporter linked promoter region (SERTPR) consists of the presence or absence of 44 base-pair segments producing both a long (L)variant, which enhances basal activity of the SERT gene, and a short (S) variant. We investigated the relationship between L/S promoter (SERTPR) polymorphic variants of serotonin transporter with treatment response in 105 female schizophrenic patients, treated with olanzapine for up to 3 months. To assess and evaluate improvement in clinical psychotic symptoms and therapeutic response to olanzapine, all patients were rated using the Positive and Negative Syndrome Scale (PANSS). Overall, the presence of L SERTPR variant was associated with significantly better treatment response measured with total PANSS and general PANSS subscales. These findings identify genetic factors associated with treatment response to olanzapine in female schizophrenic patients. P219 GENETIC POLYMORPHISMS OF SEROTONIN TRANSPORTER, MONOAMINE OXIDASE AND RESPONSE TO SSRI ANTIDEPRESSANT DRUGS IN TERMINALCANCER PATIENTS Tullio Giraldi1, Giulia Schillani2, Marianna Capozzo3, Roberta Vecchi4, Maria Anna Conte4 1 Dept. of Bomedical Sciences, Trieste 2 Dept. Bioned. Sci., Trieste 3 Dept. Biomed. Sci., Trieste 4 Hospice Pineta DC, Trieste The serotonin transporter (SERT) and monoamine oxidases (MAO) play a crucial role in monoaminergic transmission, contributing to mood regulation, and are the molecular target of antidepressant drugs. 5HTTLPR and VNTR polymorphisms have been identified in the promoter
788
Abstracts
region of SERT and MAO-A genes, and were shown to interact with stressful life events, leading to the development of mental disorders. The aim of this study is to examine the role of these polymorphisms in the development of psychological suffering in terminal cancer patients and in their response to treatment with SSRI antidepressants. 53 terminal patients with various tumours (30 females, 23 males, average age ¼ 71, range ¼ 48–95) admitted to the Hospice, were considered. The patients were assessed psychometrically using the Hospital Anxiety and Depression Scale (HADS) and Mini-Mental Adjustment to Cancer (Mini-MAC) scales, at admission and after 14 days of treatment with sertraline or citalopram: patients were also genotyped for 5-HTTLPR and MAO-A VNTR. The results obtained indicate that these polymorphisms are not associated with differences in the score on the psychometric scales used before drug treatment. On the contrary, sertraline reduced the HADS depression score in all patients, whereas Citalopram was effective only in patients with the highly functional SERT genotype. Moreover, sertraline significantly increased MAC fighting spirit in the patients with highly functional SERT genotype. The MAO-A genotype and the other concomitant drug treatment did not cause any significant effects on the psychological dimensions considered.
extracted from patients’ whole blood and sequenced in order to detect the different kinds of alleles wihtin the 5-HTTLPR. The rare 14-B s variants proved to be non-responsive to antidepressant treatment, and significantly less responsive than the most common 14-A s variants. Moreover, the rare 16-F l variants were associated with a worse response (like 14-A s) than the common 16-A l variants, while the 16-D were shown to be the most responsive to antidepressant treatment. The screening of gene-regulation online database showed that ‘‘long’’ alleles and ‘‘short’’ alleles have different consensus sequences for transcriptional factor binding. The interaction with different transcriptional factors could explain the different regulation of the SERT gene.
P220 VAMP2 GENE POLYMORPHISMS AND FLUVOXAMINE RESPONSE IN JAPANESE DEPRESSIVE PATIENTS Shinichi Saito1, Nagahide Takahashi1, Ryoko Ishihara1, Masashi Ikeda2, Tatsuyo Suzuki2, Tsuyoshi Kitajima2, Yoshio Yamanouchi2, Nakao Iwata2, Mitsuhiko Yamada3, Keizo Yoshida1, Toshiya Inada2, Norio Ozaki1 1 Dept of Psychiatry, Nagoya Univ, Nagoya, Aichi 2 Dept of Psychiatry, Fujita Health Univ, Toyoake, Aichi 3 Dept of Psychogeriatrics, National Cente, Kodaira, Tokyo 2 Dept of Psychiatry, Teikyo Univ, Ichihar, Ichihara, Chiba
The gene encoding for serotonin transporter (SERT) is polymorphic and is considered to be a strong candidate for a pharmacogenetic marker of antidepressant response. A 44-base-pair insertion/deletion polymorphism in the promoter region (SERTPR) seems to modulate the gene’s transcription. The frequency distribution of short (S) and long (L) allele differ significantly between different ethnic groups. In this study, we investigated the frequency distribution of S and L variants of SERTPR in a healthy Croatian population and patients with major depressive disorder (MDD). We also investigated the correlation of SERTPR polymorphism with response to selective serotonine reuptake inhibitor paroxetine in 114 patients with MDD (diagnosed by Structured Clinical Interview for DSM-IV disorders). SERT genotyping was performed by the PCR (polymerase-chain reaction) method. Assessments of depression symptom severity were rated on the Hamilton Depression Scale at baseline and at 1,2,3 and 6 weeks after initiation of paroxetine treatment (20–40 mg/per day). Subjects’ response patterns were categorised as responders and non-responders. Responders were defined as those subjects with a decrease in the HAMD 17-item scale 50% from baseline to week 6. Frequency distribution of L and S alleles and genotypes of SERTPR were not statistically different in healthy participants and MDD patients. Concerning SERTPR polymorphism and drug response, the L allele was associated with better response to paroxetine treatment (p ¼ 0.032).
Vesicle-associated membrane protein 2 (VAMP2) is a key component of the synaptic vesicle docking/fusion machinery, and its mRNA reportedly increases in the frontal cortex of rats following chronic antidepressant treatment. VAMP2 is therefore thought to be involved in the mechanism of action of antidepressants and may alter their efficacy. The purpose of this study was to investigate whether the VAMP2 gene is associated with clinical responses to a specific antidepressant, fluvoxamine. A total of 106 patients with major depressive disorder were given fluvoxamine (50–200 mg/day) for 8 weeks and assessed for severity of depression using the Semi-Structured Interview Guide of the Hamilton Depressive Scale (SIGH-D) (17 items) at 0 and 8 weeks. We defined a clinical response as more than a 50% reduction in baseline SIGH-D within 8 weeks, and defined clinical remission as less than a SIGH-D score of 7 at 8 weeks. Following evaluation of the linkage disequilibrium (LD) of VAMP2, single nucleotide polymorphisms (SNPs), which cover 90% of haplotypes in each LD block, were selected as haplotype tag SNPs (htSNPs). Single-marker and haplotype analyses of these SNPs were performed between responder and nonresponder, remission and nonremission. Genotying was performed by PCR-RFLP. VAMP2 gene was composed of one LD block, and 2 htSNPs were selected. Analysis of htSNPs, as well as haplotype analysis, did not reveal any significant associations. Our results suggest that the VAMP2 gene is unlikely to play a major role in efficacy of fluvoxamine.
P221 POSSIBLE PHARMACOGENETIC IMPLICATIONS OF RARE ALLELES IN SEROTONIN TRANSPORTER GENE Adele Pirovano1, Cristina Lorenzi2, Paola Artioli2, Elena Marino2, Marco Catalano2, Enrico Smeraldi2 1 University Vita-Salute—S. Raffaele, MILANO 2 University Vita-Salute—S. Raffaele, MILANO Nakamura et al. examined the polymorphic region of the promoter region of serotonin transporter gene (5-HTTLPR) in detail: the alleles, previously reported as *s and *l, are divided into four and six kinds of allelic variants, respectively. Long alleles (l*) consist of 16 imperfect repeats and could be of six different kinds (16A, 16B, 16C, 16D, 16E, 16F). In the same way, the short alleles (s*) consist of 14 imperfect repeats and could be of four different kinds (14A, 14B, 14C, 14D). In this study, we investigated antidepressant response to fluvoxamine in subjects carrying different *l and *s variants according to Nakamura’s nomenclature. 273 patients were included in this study and evaluated using the 21-item Hamilton Rating Scale for Depression. DNA was
P222 SEROTONIN TRANSPORTER POLYMORPHISM AND PAROXETINE DRUG RESPOND IN MAJOR DEPRESSIVE DISORDER Nada Bozina1, Alma Mihaljevic- Peles2, Marina Sagud3, Miro Jakovljevic2, Jadranka Sertic3 1 Clinical Hospital Center Zagreb, Zagreb 2 University Psychiatric Clinic Rebro Uni, Zagreb 3 University Psychiatric Clinic Rebro Univ, Zagreb 3 Clinical Institute of Laboratory Diagno, Zagreb
P223 EFFECTS OF SEROTONERGIC POLYMORPHISMS IN OCD TREATMENT WITH SEROTONIN REUPTAKE INHIBITORS Quirino Cordeiro, Fabio Corregiari, Stevin Zung, Marcio Bernik, Homero Vallada University of Sa˜o Paulo, Sa˜o Paulo Introduction: Serotonergic drugs play an important role in the pharmacologic treatment of obsessive-compulsive disorder (OCD). However 30–40% of OCD patients display an inadequate response to serotonin reuptake inhibitors (SRIs). Methods: 60 DSM-IVOCD patients were investigated. Poor responders (n ¼ 32) were taken as those patients with at least three trials with SRIs, one of which clomipramine, in maximal or maximal tolerated doses, for at least eight weeks, who presented a reduction of no more than 25% in Y-BOCS. Serotonergic polymorphisms were analysed and compared to clinical response in these patients. Results: 5-HT2A T102C polymorphism (alleles: (2 ¼ 0.08, 1d.f., p ¼ 0.77; genotypes: (2 ¼ 0.19, 2d.f., p ¼ 0.90); 5-HT2A C516T polymorphism (alleles: (2 ¼ 7.20, 1d.f., p ¼ 0.007; genotypes: (2 ¼ 8.10, 2d.f., p ¼ 0.01); 5- HT2C Cys23Ser polymorphism (alleles: (2 ¼ 0.25, 1d.f., p ¼ 0.61; genotypes: (2 ¼ 2.14, 2d.f., p ¼ 0.34); 5-HT1D beta G861C polymorphism (alleles: (2 ¼ 1.25, 1d.f., p ¼ 0.26; genotypes: (2 ¼ 6.56, 2d.f., p ¼ 0.03); 5- HTT ‘‘short/long’’ promoter region polymorphism (alleles: 2 ¼ 0.36, 1d.f., p ¼ 0.55; genotypes: (2 ¼ 0.49, 2d.f., p ¼ 0.78). Conclusions: These results suggest that the polymorphisms C516T of the 5-HT2A gene and the G861C of the 5-HT1D beta gene are associated with response to treatment with SRIs in OCD patients.
Abstracts P224 A PHARMACOGENETIC STUDY OF OLANZAPINE FOR URGE TO SMOKE AND NICOTINE WITHDRAWAL John McGeary1, Damaris Rohsenow2, Jennifer Tidey1, Robert Miranda1, Robert Swift2, Kent Hutchison3, Alan Sirota2, Peter Monti2 1 Brown University, Providence 2 VA, Providence, RI 3 CU-Boulder, Boulder, CO Olanzapine (OLAN), an atypical antipsychotic medication with mixed 5-HT2/DA antagonist properties, should affect biobehavioral mechanisms linked to smoking relapse. After 10 hrs smoking deprivation, OLAN was predicted to dose dependently decrease urge to smoke and withdrawal prior to and after smoking cue exposure especially for those with the S allele of the 5HTTLPR or 7-repeat variant of the DRD4 VNTR. Methods. The 24 community smokers participated in a doubleblind placebo-controlled within-subjects randomized trial comparing: (1) placebo, (2) 2.5 mg OLAN, and (3) 5.0 mg OLAN, one dose per session a week apart, with dose order counterbalanced. Urge to smoke, tobacco withdrawal symptom severity, were assessed within cue reactivity. Results. OLAN reduced withdrawal severity before (2.5 mg) and during (both doses) cue exposure, and decreased positive-affect urge to smoke before and during cue exposure (2.5 mg only). Blood pressure (MAP) change during cue exposure was lower in the 5 mg condition than in the 2.5 mg condition but did not differ from placebo. Participants homozygous for the short allele of the DRD4 dopamine receptor allele were more sensitive to OLAN effects on withdrawal symptoms and MAP. Conclusions. The results indicate a beneficial effect of 2.5 mg OLAN on tobacco withdrawal and urge to smoke; the beneficial effects on withdrawal are moderated by the DRD4 gene. Combined 5HT/DA antagonists should be considered for future development of pharmacotherapies for smoking cessation. P225 NO ASSOCIATION BETWEEN THE hOGG1 GENE POLYMORPHISM (Ser326Cys) AND TARDIVE DYSKINESIA Shinichi Sakata1, Takahiro Shinkai2, Hiroko Hori1, Kensuke Utsunomiya2, Yui Naoe2, Yuko Fukunaka1, Kazuko Shimizu1, Chima Matsumoto2, Osamu Ohmori3, Jun Nakamura1 1 Univ of Occupatl and Environmentl Health, Kitakyuushu 2 Univ of Occupatl & Environmetl Health, S, Kitakyuushu 2 Univ of Occupatl & Environmetl Health, K, Kitakyuushu 2 Univ of Occupatl & Environmetl Health, N, Kitakyuushu 3 Univ of Occupatl & Environmetl Health, W, Kitakyuushu Oxidative stress such as free radical-mediated neuronal dysfunction may be involved in the pathophysiology of tardive dyskinesia (TD). The human 8-oxoguanine DNA glycosylase (hOGG1) plays an important role in the repair of damaged DNA. The aim of this study is to examine whether a functional polymorphism, a serine (Ser) to cysteine (Cys) substitution at codon 326 (Ser326Cys) of the hOGG1 gene (3p26.2), is associated with susceptibility to TD in a Japanese sample of schizophrenia patients. It is reported that the potential capacity of hOGG1 with the 326Ser to repair the damaged DNA is 7 times higher than that with the 326Cys, suggesting that 326Cys allele may give an individual more susceptibility to the formation of 8-hydroxyguanine in DNA. Our sample includes 200 patients with schizophrenia (DSM-IV). We evaluated TD using the Abnormal Involuntary Movement Scale score. Informed consent was a premise for participation, and this study was approved by the Ethics Committee of the University of Occupational and Environmental Health. Genotyping was assessed by the TaqMan allele-specific assay method using ABI PRISM1 7000 (Applied Biosystems). Differences in allele and genotype distribution between cases with and without TD were evaluated using the chi2 test. No significant association between the hOGG1 Ser326Cys polymorphism and TD was found (genotype: chi2 ¼ 0.125, d.f. ¼ 2, p ¼ 0.939; allele frequency: chi2 ¼ 0.006, d.f. ¼ 1, p ¼ 0.938). Our results suggest that the hOGG1 Ser326Cys polymorphism may not confer susceptibility to TD in our sample. P226 5-HT2C RECEPTOR AND LEPTIN POLYMORPHISMS, OBESITY AND METABOLIC SYNDROME IN SCHIZOPHRENIA Olga Yevtushenko, Stephen Cooper, Ryan O’Neill, Jeniffer Doherty, Gavin Reynolds Queens University Belfast, Belfast
789
Excessive weight gain, glucose dysregulation, hyperlipidaemia and metabolic syndrome are the major side effects of antipsychotic treatment. Recent pharmacogenetic studies associate several candidate genes with weight gain. To evaluate the relative contribution of the functional polymorphisms in the leptin and 5HT2C receptor genes to metabolic changes in patients with schizophrenia or schizoaffective disorder we are conducting a cross-sectional study. Measurements taken included blood pressure, random blood glucose, cholesterol, high density triglycerides levels; waist circumference, body mass index. Interim analysis has been undertaken on 130 subjects; metabolic syndrome (IDF criteria) was determined in 46 cases. All samples were genotyped for the 5-HT2C receptor 759C/T and leptin 2548A/G polymorphisms. An association between leptin polymorphism and measures of obesity (BMI, waist circumference) was found in males, G allele being associated with greater obesity. No significant effects of the 5-HT2C polymorphism were found. Combined genotype analysis of 5HT2C 759C/T and leptin polymorphisms showed highly significant interaction on these measures (p < 0.001). In patients with 759 C or CC genotype, the leptin genotype effect on obesity reached p < 0.01. Leptin-2548G allele carriers exhibited significantly higher rate of metabolic syndrome in males, but not in females. The effect of leptin genotype on frequency of metabolic syndrome within the 759C/CC group was very strong (p ¼ 0.001). These findings present further evidence of association between 5-HT2C receptor 759C/T, and leptin 2548A/G polymorphisms in determining both obesity and metabolic syndrome in patients with schizophrenia. P227 POSITIVE ASSOCIATION OF CYP2D6 AND COMT VARIANTS WITH ECSTASY-INDUCED CHANGES IN WATER HOMEOSTATIC MEASURES Eva Maria Tsapakis1, Kim Wolff1, Patricia Huezo-Diaz1, Adam Winstock1, Robert Kerwin1, Mary Forsling2, Katherine Aitchison1 1 Institute of Psychiatry, London 2 GKT School of Medicine, London In a Caucasian UK sample of regular clubbers, ingestion of ecstasy (‘E’) was associated with the induction of inappropriate antidiuretic [or vasopressin (AVP)] secretion (SIADH). ‘E’ toxicity may be determined by poor metabolism, with CYP2D6 being responsible for demethylenation, and cathechol O-methyl transferase (COMT) for O-methylation. Both CYP2D6 and COMT are highly polymorphic genes. We hypothesized that genetically determined poor metabolism of ‘E’ may be a predisposing factor to ‘E’-associated SIADHs, and hence a risk factor for potentially fatal cerebral oedema. Blood was obtained from 30 (17 male) experienced clubbers (mean 6.6 years clubbing). Pre- and post-clubbing measurements were performed to compare various biochemical indices. We found a positive association between ‘E’ use by CYP2D6 genotype and decrease in plasma osmolality (p < 0.001) and a significant association between ‘E’ use by COMT rs4680 genotype and decrease in plasma sodium (P ¼ 0.030). CYP2D6 poor metaboliser status by ‘E’ use was significantly associated with decreased plasma osmolality (P ¼ 0.003) and plasma Naþ (P ¼ 0.012), whereas COMT rs4680 low activity by ‘E’ use was significantly associated with decreased plasma Naþ (P ¼ 0.019) and increased plasma AVP (P ¼ 0.034). Our findings indicate that CYP2D6 and COMT metaboliser status could be used to predict individuals who might be particularly at risk of SIADH on ‘E’ consumption, and hence potentially lethal cerebral oedema. P228 MDMA: STRESS REGULATION, CORTISOL, 5-HTTLPR AND COMT POLYMORPHISMS Eva Maria Tsapakis1, Kim Wolff1, Carmine Pariante1, Adam Winstock1, David Holt2, Robert Kerwin1, Katherine Aitchison1 1 Institute of Psychiatry, London 2 St. George’s University of London, London This study investigates the influence of genotype on the metabolism of 3,4-methylenedioxy methamphetamine (MDMA, ‘ecstasy’) and its influence on stress regulation and cortisol secretion. Fifty-one subjects (33 men, mean age 25 years) self-nominating experienced clubbers (mean 6.6 years clubbing) were tested ‘in the field’. Pre- and postclubbing measurements were performed to compare: stress regulation; cortisol concentration and drug use in those who had and had not consumed MDMA. Of the 60.8% (31/51) subjects whose urine tested positive for a psychoactive substance, 21 were positive for MDMA.
790
Abstracts
Plasma cortisol concentrations increased in all subjects, with changes being significantly greater in the MDMA-positive group (670.5 101.6 compared to 411.5 65 mmol/l, p < 0.004). Grouping the Met/Met low functionality COMT genotype versus the high (Val/Val) and intermediate (Val/Met) functionality genotypes, we found a positive association with low COMT functionality and MDMA-induced significant increases in plasma cortisol post-clubbing (p ¼ 0.016). We failed, however, to show an association with 5-HTTLPR genotype and MDMA-induced cortisol changes (p ¼ 0.398). The observed association between MDMA use, cortisol level and COMT rs4680 genotype suggest that consumption of MDMA may have consequences relating to dysfunction of the HPA axis and stress regulation, as well as to the genetically determined metabolic status of the user.
P229 A POSSIBLE INTERACTION OF OX1R AND DRD2 GENES ON SUSCEPTIBILITY TO POLYDIPSIA IN SCHIZOPHRENIA Yuko Fukunaka1, Takahiro Shinkai2, Hiroko Hori3, Yui Naoe4, Shinichi Sakata3, Kensuke Utsunomiya4, Kazuko Shimizu3, Chima Matsumoto4, Osamu Ohmori5, Jun Nakamura3 1 Univ of Occupatl & Environmentl Health, Kitakyushu 2 Univ of Occupatl & Environmetl Health, S, Kitakyushu 3 Univ of Occupatl & Environmetl Health, Kitakyushu 4 Univ of Occupatl & Environmetl Health, K, Kitakyushu 4 Univ of Occupatl & Environmetl Health, N, Kitakyushu 5 Univ of Occupatl & Environmetl Health, W, Kitakyushu The underlying pathophysiology of primary polydipsia in schizophrenia is poorly understood. Our previous study, however, suggested that this condition may involve a genetic component (Shinkai et al., 2003). We previously reported a significant association between functional dopamine D2 receptor gene (DRD2) polymorphisms and polydipsia in schizophrenia (Matsumoto et al., 2004). Recently, important interactions between dopaminergic neurotransmission and the orexin system have been discovered (Fadel et al., 2002). In the present study, we examined the association between orexin-1 receptor gene (OX1R) and DRD2 polymorphisms and polydipsia in a Japanese schizophrenia sample. Our sample included 149 patients with schizophrenia (DSMIV) (65 with polydipsia and 84 without polydipsia). Informed consent was obtained prior to the subjects’ participation and this study was approved by the Ethics Committee of the University of Occupational and Environmental Health. Genotyping was assessed by the PCRRFLP and the TaqMan allele-specific assay method using ABI PRISM1 7000 (Applied Biosystems). Differences in genotype distribution between cases and controls were evaluated using the chi2 test. We found a trend for additive effects of OX1Rand DRD2 on susceptibility to polydipsia (genotype: 2 ¼ 10.983, df ¼ 5, p ¼ 0.052). Our data suggest that the polymorphisms in OX1R and DRD2 are involved individually or in combination in the predisposition to polydipsia in schizophrenia.
P230 TARDIVE DYSKINESIA PHARMACOGENETICS IN AFRO-CARIBBEAN PATIENTS ON NEUROLEPTIC TREATMENT Richard Bruggeman1, Bob Wilffert2, Asmar F.Y. Al Hadithy3, Vishnu Jangbahadoer Sing4, Glenn Matroos5, Hans W. Hoek4, Jim van Os5, Jacobus R.B.J. Brouwers3, P.N. van Harten4 1 University Medical Centre Groningen, Groningen 2 Zorggroep Noorder-breedte and De Tjonge, Leeuwarden 3 Groningen Institute for Drug Exploration, Groningen 4 Parnassia Psychiatric Institute, The Hague 5 Dr. D.R. Capriles Clinic, Curac¸ao 5 South Limburg Mental Health Research and, Maastricht 4 Symfora Group Psychiatric Centre, Amersfoort Background: Up to 75% of patients chronically exposed to neuroleptics may develop motor side-effects. Tardive dyskinesia (TD), a severe and potentially irreversible motor side-effect, accounts for about 30%. Age, gender and polymorphism of D3, 5HT2A and 5-HT2C receptors are described to affect the risk for TD. Aim: to investigate the association between the polymorphisms Ser9Gly (D3), C102T (5-HT2A) and Cys23Ser (5-HT2C) with TD in an Afro-Caribbean population treated with neuroleptics. Patients: 109 subjects from van Harten et al. study (Schizophr. Res.1996;19:195).
Methods: Relationships between the genotype and orofaciolingual dyskinesia (TDof; AIMS items 1–4) and limb-truncal dyskinesia (TDaa; AIMS items 5–7) were analyzed with regression analysis. Only statistically significant results (p < 0.05) are mentioned. Results: After correction for age, higher TDof-values were found with Ser9/Gly9 than with Gly9Gly genotypes (males 1.6; females 3.3 points higher). In females, TDof-values were 5.2 points higher with 102TT than with 102CC. For TDaa in males a slightly higher TDaa-value was found for Ser9/Gly9 than for Gly9/Gly9 (0.5), whereas in females Ser9/ Gly9 results in lower (1.1) and Ser9/Ser9 in higher (1.7) TDaa-values. In females, Cys23/Ser23 and 102TT genotypes were associated with higher TDaa-values (1.0 (p < 0.058) and 4.8, respectively). Conclusion: In our population, the associations between TD-parameters and genotypes are gender-dependent and differ strongly between TDof and TDaa. P231 IDENTIFICATION OF DIAGNOSTIC BIOMARKERS FOR DEPRESSION BY MULTI ANALYTE PROFILING OF PERIPHERAL SAMPLES FROM CASE/CONTROL COLLECTIONS Enrico Domenici1, Inga Prokopenko2, David Wille3, Federica Tozzi2, Alun McCarthy2, Barbara Weber2, Lefkos Middleton2, Pierandrea Muglia2 1 GSK Psychiatry CEED, Verona, Italy 2 GSK—Translational Medicine & Genetics, Verona 3 GSK—Discovery Statistics, Harlow The definition, classification and diagnosis of psychiatric diseases still relies mostly on clinical evaluation of subjective symptoms and the assessment of response to psychiatric medications lacks objective, reliable and accurate measures. Currently there are no validated biological markers reflecting a particular trait or state of the illness, and the search for peripheral biomarkers has generated a number of hypotheses requiring replication and further qualification to demonstrate their clinical relevance. In order to identify peripheral biological markers for depression, we have carried out proteomic profiling of samples from depressed patients and controls selected from genetic collections available within GSK. Patients were diagnosed according to DSM criteria using structured interviews and a number of clinical variables and demographic information were available for the analysis. Plasma samples from 250 depressed patients and 250 controls were submitted to multi analyte profiling allowing the assessment of up to 76 proteins, including a series of cytokines, chemokines and neurotrophins previously suggested to be involved in the pathophysiology of depression. Samples from schizophrenic patients were included in order to validate the specificity of any finding. Univariate data analysis has shown more significant p-values than expected by chance and highlighted known ad novel proteins as potential depression markers. Multivariate analysis was also carried out to identify the most informative combination of analytes enabling to differentiate disease state and subtypes from controls. The results of this study will be presented and discussed for their potential value as objective measures of drug response in longitudinal drug efficacy trials. P232 NO ASSOCIATION BETWEEN THE DOPAMINE D3 RECEPTOR (DRD3) GENE POLYMORPHISM (Ser9Gly) AND TARDIVE DYSKINESIA Kensuke Utsunomiya1, Takahiro Shinkai2, Shinichi Sakata1, Hiroko Hori1, Chima Matsumoto2, Yui Naoe1, Yuko Fukunaka1, Kazuko Shimizu1, Osamu Ohmori2, Jun Nakamura1 1 Univ of Occupatl & Enviromentl Health, Kitakyusyu 2 Univ of Occupatl & Enviromentl Health,S, Kitakyusyu 2 Nissan Technical Center Health Promotion, Atsugi 2 Univ of Occupatl & Enviromentl Health, Kitakyusyu The relation between dopamine D3 receptor (DRD3) and tardive dyskinesia (TD) is biologically plausible because DRD3 is selectively expressed in a brain region implicated in locomotor function (Suzuki et al., 1998). A recent combined analysis also suggests positive association between the DRD3 polymorphism in schizophrenia and TD (Lerer et al., 2002). The aim of this study is to replicate whether a functional polymorphism, a serine (Ser) to cysteine (Gly) substitution at codon 9 (Ser9Gly) of the DRD3 gene, is associated with susceptibility to TD in a Japanese sample of schizophrenia patients. Our sample includes 196 Japanese schizophrenia patients (DSM-IV). We evaluated
Abstracts TD using the Abnormal Involuntary Movement Scale (AIMS) score. Informed consent was a premise for participation, and this study was approved by the Ethics Committee of the University of Occupational and Environmental Health. Genotyping was assessed by the TaqMan allele-specific assay method using ABI PRISM1 7000(Applied Biosystems). Differences in allele and genotype distribution between cases with and without TD were evaluated using the chi2 test. Difference in AIMS score among the three genotypic groups was evaluated using the Kruskal-Wallis test. No significant association between the DRD3 Ser9Gly polymorphism and TD was found (genotype: chi2 ¼ 0.17, d.f. ¼ 2, p ¼ 0.92; allele frequency: chi2 ¼ 0.00, d.f. ¼ 1, p ¼ 1.00). No significant difference in AIMS score among the three genotypic groups was found (p ¼ 0.66). Our results suggest that the DRD3 Ser9Gly polymorphism may not confer susceptibility to TD.
P233 HPA AXIS GENES AND CLINICAL OUTCOME AND RESPONSE TO SSRIs IN MAJOR DEPRESSION Barbara Arias1, Sergi Papiol2, Cristo´bal Gasto´3, Rosa Catala´n4, Blanca Gutie´rrez2, Lourdes Fan˜ana´s2 1 Universitat de Barcelona, Barcelona (Spain) 2 Un Antropologia (Dp Biologia Animal). Un, Barcelona (Spain) 3 CSM of Hospital Clı´nic de Barcelona., Barcelona (Spain) 4 CSM of Hospital Clı´nic de Barcelona y ID, Barcelona (Spain) Dysregulation of the hypothalamic-pituitary-adrenal (HPA) axis has been reported in depression (Holsboer, 2000). Normalization of HPA axis has been suggested to play a role in the mechanisms of action of antidepressants (Nemeroff, 2002). We analysed the influence of genetic variants in CRHR1, CRHR2, CRH-BP and FKBP5 genes on the vulnerability for depression and the response to antidepressant treatment. Sample: 159 depressive outpatients and 96 healthy controls (Spanish origin). Patients were assessed for clinical features such as seasonality or age of onset. Episode was treated with citalopram and followed along 12 weeks. Symptom severity was evaluated on inclusion and then monthly along the follow-up using a 21-item HDRS. SNPs were assayed using Applied Biosystems SNaP-Shot and TaqMan technology. Results: rs110402, in CRHR1 gene, was associated with an increased risk of presenting a seasonal pattern and an early age of onset of the first depressive episode. Allele G carriers of rs2270007 of CRHR2 gene, showed a worse response to citalopram in time (Genotype effect F ¼ 7.45, P ¼ 0.007) and an increased risk of 2.93(95%IC [1.24–6.90]) for non responding at 4th week (2 ¼ 7.59, df ¼ 1, P ¼ 0.006). Variability at genes encoding proteins with a pivotal role in HPA axis regulation seems to influence i) the phenotypic expression of a differential clinical picture of depression and ii) early clinical response to SSRI antidepressants. Acknowledgements: Spanish Ministerio de Ciencia-Tecnologia (SAF 2005- 07852-C02-01)
P234 NO ASSOCIATION BETWEEN A FUNCTIONAL POLYMORPHISM IN THE HUMAN MULTIDRUG-RESISTANCE GENE (MDR1) AND TARDIVE DYSKINESIA Takahiro Shinkai1, Hiroko Hori1, Kensuke Utsunomiya2, Shinichi Sakata1, Yui Naoe2, Yuko Fukunaka1, Kazuko Shimizu1, Chima Matsumoto2, Rudi Hwang3, Osamu Ohmori4, Jun Nakamura1 1 Univ of Occupatl & Environmetl Health, Kitakyushu 2 Univ of Occupatl & Environmetl Health, K, Kitakyushu 2 Univ of Occupatl & Environmetl Health, N, Kitakyushu 3 Neurogenetics Section, Centre for Addict, Toronto 4 Univ of Occupatl & Environmetl Health, W, Kitakyushu P-glycoprotein, which is coded by MDR1 gene, in the brain capillary endothelial cell limits the entry of many drugs including antipsychotics into the brain. The aim of our study is to examine whether a functional polymorphism, a C to T substitution at position 3435 in exon 26 of the MDR1 gene, is associated with susceptibility to tardive dyskinesia (TD) in a Japanese sample of schizophrenia patients. It has been reported that individuals homozygous for this polymorphism had significantly lower duodenal MDR1 expression levels and function of MDR1 (PNAS, 97: 3473–3478, 2000). Furthermore, the brain entry of risperidone and 9- hydroxyrisperidone has been shown to be greatly limited by P-
791
glycoprotein (Int J Neuropsychopharmacol, 7: 415–419, 2004). Our sample includes 183 patients with schizophrenia (DSM-IV). This study was approved by the Ethics Committee of the University of Occupational and Environmental Health, and all subjects for this study were recruited with fully informed written consent. We evaluated TD using the Abnormal Involuntary Movement Scale score. Genotyping was assessed by the TaqMan allele-specific assay method using ABI PRISM1 7000 (Applied Biosystems). Differences in genotype distribution and in allele frequency between cases with and without TD were evaluated using the 2 test. No significant association between the MDR1 C3435T polymorphism and TD was found (genotype: 2 ¼ 1.26, d.f. ¼ 2, p ¼ 0.53; allele: 2 ¼ 0.69, d.f. ¼ 1, p ¼ 0.41). Our results suggest that MDR1 C3435T polymorphism may not confer susceptibility to TD. P235 POLYMORPHISMS IN DOPAMINE RECEPTOR DRD1 AND DRD2 GENES AND PSYCHOPATHOLOGICAL AND EXTRAPYRAMIDAL SYMPTOMS IN PATIENTS ON LONG-TERM ANTIPSYCHOTIC TREATMENT Vita Dolzˇan1, Blanka Kores Plesnicˇar2, Alessandro Serretti3, Laura Mandelli3, Diana De Ronchi3, Paolo Olgiati3, Bojan Zalar4, Jure Koprivsˇek2, Katja Breskvar1 1 Institute of Biochemistry, Faculty of Medicine, Ljubljana, Slovenia 2 Dept. of Psychiatry, Teaching Hospital Maribor, Maribor, Slovenia 3 Institute of Psychiatry, University of Bologna, Bologna, Italy 4 University Psychiatry Clinic, Ljubljana, Slovenia DRD1 and DRD2 receptor gene variants have been associated with clinical aspects of schizophrenia, however only specific features were analyzed in different samples. To assess the complex interaction between genetic and clinical factors, we studied the possible crossinteractions between DRD1 and DRD2 dopamine receptor gene polymorphisms, symptomatology of schizophrenia and schizoaffective disorders and the occurrence of treatment induced side effects taking into consideration possible clinical confounding variables. One hundred and thirty one outpatients in stable remission meeting the DSMIV criteria for schizophrenia spectrum disorders and receiving long-term maintenance therapy with haloperidol, fluphenazine, zuclopethixole or risperidone were genotyped for DRD1 A-48G, DRD2 Ins141CDel and DRD2 Ser311Cys polymorphisms. Psychopathological symptoms were assessed with the Positive and Negative Syndrome Scale for Schizophrenia (PANSS). Extrapyramidal side effects were assessed with the Simpson-Angus Extrapyramidal Side Effects Scale (EPS), the Barnes Akathisia Scale (BARS) and the Abnormal Involuntary Movement Scale (AIMS). Drug dosage was included as covariant because it was associated with the severity of symptomatology, akathisia and parkinsonism. Our results suggest a possible association between DRD2 311 Ser/Cys genotype and tardive dyskinesia, though this finding is due to only 3 subjects. No major association was observed for the other polymorphisms and with the other clinical variables considered. Our study evidenced that DRD2 Ser/Cys rare variants may influence tardive dyskinesia independently from the other clinical variables. This effect was observed in a small number of subjects and it requires caution in interpreting it. P236 DYSBINDIN ASSOCIATED WITH SSRI ANTIDEPRESSANT EFFICACY Chi-Un Pae1,2, Alessandro Serretti3, Laura Mandelli3, Diana De Ronchi3, Ashwin A. Patkar2, Jung-Jin Kim1, Chang-Uk Lee1, Soo-Jung Lee1, Chul Lee1, In-Ho Paik1 1 Department of Psychiatry, Kangnam St. Mary’s Hospital, The Catholic University of Korea College of Medicine 2 Department of Psychiatry and Behavioural Sciences, Duke University; 3 Institute of Psychiatry, University of Bologna, Bologna, Italy In the search for further genes influencing antidepressant response we focused on the dysbindin gene (dystrobrevin-binding-protein 1, DTNBP1). One hundred and four Korean inpatients affected by Major depressive disorder (MDD) were treated with various antidepressants at standard therapeutic daily doses and rated with the 10-items ˚ sberg Depression rating scale (MADRS) at baseline and Montgomery-A discharge. Five DTNBP1 variants (rs3213207 A/G, rs1011313 C/T,
792
Abstracts
rs2005976 G/A, rs760761 C/T and rs261952 A/C) were analysed for all subjects. Rs2005976 was found to be significantly associated with final MADRS scores, with the rarest A allele associated with higher final scores (p ¼ 0.00055), rs760761 showed also a significant association (p ¼ 0.0058) and rs2619522 showed a positive trend (p ¼ 0.025). Markers were not significantly associated with CGI scores. Five marker haplotypes were mildly associated with MADRS final scores. When considering the block composed by the 3 SNPs individually associated with response (rs2005976, rs760761 and rs2619522) results were more marked (p ¼ 0.0096), with the G-C-A haplotype associated with a positive outcome. Despite limitations due to the sample size and the mild antidepressant response, we observed a significant association between DTNBP1 variants and antidepressant response. P237 NO ASSOCIATION BETWEEN THE METABOTROPIC GLUTAMATE RECEPTOR TYPE 3 GENE (GRM3) AND SCHIZOPHRENIA IN A JAPANESE POPULATION Mamoru Tochigi1, Motomu Suga2, Jun Ohashi2, Takeshi Otowa2, Hidenori Yamasue2, Kiyoto Kasai2, Tadafumi Kato3, Yuji Okazaki4, Nobumasa Kato2, Tsukasa Sasaki2 1 MDMD, Brain Science Institute, RIKEN, Saitama 2 University of Tokyo, Tokyo 3 Brain Science Institute, RIKEN, Saitama 4 Mie University, Mie Several lines of evidence have suggested the metrotropic glutamate receptor 3 (GRM3) gene is a possible susceptibility gene for schizophrenia. To our knowledge, six studies have investigated the genetic association between GRM3 and schizophrenia; however, the results have been quite controversial. In the present study, we investigated the association between the GRM3 gene and schizophrenia in 402 Japanese by analysing 10 single nucleotide polymorphisms (SNPs), including all SNPs which had shown significant results in previous studies. No significant difference was observed in allelic frequencies or genotypic distribution of the 10 SNPs between controls and patients. Likewise, the permutation test showed no significant differences inestimated haplotype frequencies between controls and patients. Thus, the present study provides no positive evidence of association between the GRM3 gene and schizophrenia in the Japanese population. However, in two SNPs (rs6465084 and rs1468412), significant deviations from the Hardy-Weinberg equilibrium were observed in patients (p ¼ 0.00084 and 0.022, respectively, uncorrected). This encourages further studies using larger sample sets to draw more definitive conclusions. P238 LACK OF ASSOCIATION BETWEEN IL4 GENE PROMOTER POLYMORPHISM AND SHIZOPHRENIA Masaki Nishiguchi, Tsutsumi Atsushi, Houkyou Akira, Kikuyama Hiroki, Koh Jun, Yoneda Hiroshi Osaka Medical College, Takatsuki, Osaka Background: An immune process, characterized by relative predominance of the T helper-2 (Th2) system, may be involved in the pathophysiology of schizophrenia. Functional polymorphisms in the interleukin-4 (IL-4) genes appear to be principal candidates for genetic research. IL-4 plays a major role as a Th2-type response mediator. It induces T-helper cells to differentiate into Th2 cells while suppressing development of Th1 cells. The IL-4 coding gene is located within the cytokine gene clusteron chromosome 5q31.1,iwhich is a susceptible region for schizophrenia, identified by linkage analyses. A high-risk IL-4 590CC genotype for schizophrenia was identified in a Caucasian population. We investigated whether the IL-4 590C/T single nucleotide polymorphism (SNP) might be responsible for susceptibility of Japanese patients to schizophrenia. Method: A total of 173 Japanese patients with a DSM-IV diagnosis of schizophrenia and 123 healthy controls were recruited. Genomic DNA was isolated from whole blood according to standard procedure. Genotyping was performed by the fluorescence resonance energy transfer method using the LightCycler System. Results: No significant differences in allele or genotype frequencies were observed between patients with schizophrenia and controls. Conclusion: These results suggest that the IL-4 590C/T SNP is not a major genetic risk factor for susceptibility to schizophrenia in a Japanese population.
P239 INTERACTIONS BETWEEN BDNF AND DOPAMINE RECEPTOR: 1, 2, 3, 4POLYMORPHISMS IN SCHIZOPHRENIA Maria Skibinska, Monika Dmitrzak-Weglarz, Piotr Czerski, Pawel Kapelski, Agnieszka Slopien, Joanna Hauser University of Medical Sciences, Poznan Brain-derived neurotrophic factor (BDNF) plays a critical role in neuronal proliferation, survival and differentiation of the dopaminergic system. BDNF stimulates dopamine activity and turnover. By controlling the expression of specific genes, BDNFmay be an important factor in neurodevelopmental psychiatric diseases. The study was performed on 342 patients with schizophrenia (193 males, 149 females), diagnosed according to DSM-IV and ICD-10 criteria, and 355 controls (138 males, 214 females). Polymorphisms: Val66Met BDNF, 48A/G DRD1, 141C Ins/Del DRD2, Ser9Gly DRD3 and 521C/T DRD4 were analysed by PCR-RFLP. Fisher’s exact test was used to test interaction between Val66Met and dopamine receptor polymorphisms. Interactions were found interactions between the Val/Met/Ins/Ins combination of BDNF and DRD2 polymorphisms in the entire group (p ¼ 0.018) and Val/Val/Ins/Ins in the females (p ¼ 0.0326). For BDNF and DRD3 polymorphisms, we have found an association with the Val/Met/Ser/Ser combination in the whole group (p ¼ 0.039) and in the male group (p ¼ 0.044). In the female group, we have found an association (p ¼ 0.015) with the Val/Met/A/G combination of BDNF and DRD1 polymorphisms. The results suggest a protective role of Val/Met/Ins/ Ins, Val/Met/Ser/Ser and Val/Met/A/G genotype combinations against schizophrenia. Val/Val/Ins/Ins in the female group is associated with the illness. Our results suggest an epistatic effect of the BDNF and dopamine receptor genes in schizophrenia. P240 NTRK2 A3848G & T278870C SNP POLYMORPHISMS IN SCHIZOPHRENIA Monika Dmitrzak-Weglarz1, Maria Skibinska2, Pawel Kapelski2, Agnieszka Slopien2, Anna Leszczynska-Rodziewicz2, Janusz Rybakowski2, Joanna Hauser2 1 Poznan Univeristy of Medical Sciences, Poznan 2 Poznan University of Medical Sciences, Poznan Brain-derived neurotrophic factor (BDNF)is a key regulator of the survival, differentiation, growth and plasticity of neurons in the central nervous system. Several investigations confirm a role of BDNF in complex psychiatric illnesses. As the biological effect of BDNF is mediated by neurotrophic tyrosine kinase receptor type 2 (NTRK2), it seems to be good candidate gene in association studies on schizophrenia. In detail, recent evidence shows abnormalities in levels of TrkB in the hippocampus of schizophrenic patients, as well as decreased mRNA levels of TrkB after exposure to haloperidol. In the present work, we studied the possible association of two SNPs polymorphism, localised in the 2nd and 16th introns, with schizophrenia using a Polish sample. The RFLP method was used to identify the A3848G and T278870C polymorphisms of NTRK2 gene. Genotypes and alleles were estimated in patients with schizophrenia (n ¼ 487) and healthy controls (n ¼ 450). Genotype distribution and allele frequency for both analysed polymorphisms in patients with schizophrenia did not differ significantly from controls. We also found no differences in genotype and allele distribution in the male and female subgroups. Schizophrenic patients were also analyse with regard to age at onset (18 > years n ¼ 99 and 18 < years n ¼ 351). There was no significant difference between patients and controls for either A3848G or T278870C polymorphisms. Analysed separately, the A3848G and T278870C polymorphisms of NTRK2 gene showed no association with schizophrenia in the Polish population, but possible involvement may be estimated after analysis of interaction between these polymorphs. P241 LACK OF ASSOCIATION BETWEEN IL-2 GENE PROMOTER POLYMORPHISMAND SCHIZOPHRENIA Akira Houkyou, Masaki Nishiguchi, Atsushi Tsutsumi,Hiroki Kikuyama, Jun Koh, Hiroshi Yoneda 1 Osaka Medical College, takatsuki city osaka, Department of Neuropsychiatry, Osaka Medical College, Osaka, Japan Background: The alteration of T helper 1 (TH1) and TH2 responses and related cell-mediated immunity has been supposed to be associated with the immunological pathogenesis in the development of schizophrenia. Interleukin-2 (IL-2) acts as a growth factor for Tcells. A significantly decreased production of IL-2 by peripheral lymphocytes is
Abstracts one of the best-replicated immunological findings in schizophrenia. The IL-2 coding gene is located within the cytokine gene cluster on chromosome 4q,a chromosomal region again identified by linkage analyses as containing a susceptibility gene for schizophrenia. A highrisk IL-2 330TT genotype for schizophrenia was identified in a Caucasian population.We investigated whether IL-2 330T/G single nucleotide polymorphism (SNP) might be responsible for susceptibility in Japanese patients with schizophrenia. Methods: A total of 173 Japanese patients with a DSM-IV diagnosis of schizophrenia and 123 healthy controls were recruited. Genomic DNA was isolated from whole blood according to standard procedure. Genotyping wasperformed by the fluorescence resonance energy transfer method using the LightCycler System. Results: No significant differences in allele or genotype frequencies were observed between patients with schizophrenia and controls. Conclusion: These results suggest that the IL-2 330T/G SNP is not major genetic risk factor for susceptibility to schizophrenia in a Japanese population. P242 INTERACTIONS BETWEEN BDNF AND 3848A/G AND 278870 T/C POLYMORPHISMS OF NTRK2 IN SCHIZOPHRENIA Anna Leszczynska-Rodziewicz, Maria Skibinska, Monika Dmitrzak Weglarz, Pawel Kapelski, Agnieszka Slopien, Joanna Hauser University of Medical Sciences, Poznan Brain-derived neurotrophic factor (BDNF) plays a crucial role in the survival, proliferation and differentiation of neurons. BDNF acts via receptor TrkB, and since it controls the expression of specific genes, it may be an important factor in theethiology of neuropsychiatric diseases. The study was performed on 342 patients with schizophrenia (193 males, 149 females), diagnosed according to DSM-IV and ICD-10 criteria, and 355 control subjects (138 males, 214 females). Polymorphisms: Val66Met BDNF, 3848 A/G, 278870 T/C of TrkB genes were analysed by PCR-RFLP. Fisher’s exact test was used to test interaction between Val66Met and TrkB receptor polymorphisms. We found interactions between the Val/Met/G/G combination of BDNF and TrkB polymorphisms in the whole group (p ¼ 0.0504) and Val/Met/G/G in the group with disease onset after 18 years of age (p ¼ 0.0287). These results suggest a protective role of Val/Met/G/G genotype combinations in schizophrenia. P243 SEARCH OF SUSCEPTIBILITY SNP FOR SCHIZOPHRENIA IN THE PAEL-R GENE Keiichiro Mori1, Mutsuo Harano2, Naohisa Uchimura2 1 Kurume-University, Kurume, Fukuoka 2 Kurume University, Kurume, Fukuoka Introduction: Pael-R (Parkin-associated endothelin receptor-likeReceptor) is a substrate of Parkin, a gene product responsible for autosomal recessive juvenile Parkinsonism (AR-JP). It is reported that Pael-R affects dopamine levels in the brain,and the Pael-R gene is located on 7q31 in human chromosomes, one of the chromosomal loci contributing to vulnerability to bipolar disorder. Therefore, we hypothesised that the Pael-R gene is a candidate susceptibility gene for schizophrenia, and we searched the SNP of the Pael-R gene. Methods: The Pael-R gene was screened for mutations using genomic DNA samples from patients by direct sequencing. This study was approved by the Ethics Committee of Kurume University. After a complete description of the study was provided, written informed consent was obtained from all subjects before study entry. The subjects were Japanese patients with schizophrenia who had been diagnosed according to DSM-IV criteria. Results and Conclusion: We identified two SNPs which change Phe to Leu in the first exon and Glu to Lys in the second exon. We are now performing association analysis between the disease group and healthy control group regarding these two SNPs. Our report at the Congress will include those analysis results. P244 FAMILY-BASED ASSOCIATION STUDY OF THE DOPAMINE D1, D3, D4 RECEPTOR GENES AND SCHIZOPHRENIA Pawel Kapelski, Piotr Czerski, Maria Skibin˜ska, Monika Dmitrzak, Aleksandra Szepankiewicz, Janusz Rybakowski, Joanna Hauser University of Medical Sciences, Poznan
793
The dopamine system is a preferred object of biological research in schizophrenia. The dopamine D1, D3 and D4 receptor (DRD1, DRD3 and DRD4) genes are one of the main objects of this research. The present study was conducted on an ethnically-homogenous Polish population. The group consisted of 118 patients with DSM-IV schizophrenia and their healthy parents, a total of 236 individuals. Weinvestigated the polymorphisms: A-48G of DRD1 gene, Ser9Gly of DRD3 gene and 521C/T of DRD4 gene in trios using the transmission disequilibrium test. No significant differences were observed between transmitted and non- transmitted alleles for Ser9Gly DRD3 gene polymorphism (p ¼ 0,3827) and a promoter polymorphism 521C/T of DRD4 gene (p ¼ 0,3213). An excess of transmission of the A allele of the A-48G of DRD1 gene polymorphism was detected (p ¼ 0,0612). Our study found an excess of transmission of the A allele of the A-48G of DRD1 gene polymorphism, indicating the possibile involvement of this polymorphism in the pathogenesis of schizophrenia. P245 ASSOCIATION BETWEEN POLYMORPHISM IN THE VESICULAR MONOAMINE TRANSPORTER 1 GENE (VMAT1/SLC18A1) ON CHROMOSOME 8P AND SCHIZOPHRENIA Falk W. Lohoff, Thomas N. Ferraro, Wade H. Berrettini University of Pennsylvania, Philadelphia, PA Linkage studies have suggested a susceptibility locus for schizophrenia (SZ) exists on chromosome 8p21-22. The vesicular monoamine transporter 1 gene (VMAT1/slc18a1) maps to this SZ susceptibility locus. Vesicular monoamine transporters are involved in the presynaptic vesicularpackaging of monoamine neurotransmitters which have been postulated to play a role in the etiology of SZ. Variations in the VMAT1 gene might affect transporter function and/or expression and might be involved in the etiology of SZ. Genotypes of 62 SZ patients and 188 control subjects were obtained for four missense SNPs (Thr4Pro, Thr98Ser, Thr136Ile, Val392Leu) and 2 non- coding SNPs (rs988713, rs2279709). All cases and controls were of European descent. The frequency of the minor allele of the Thr4Pro polymorphism was significantly increased in SZ patients when compared to controls (p ¼ 0.0140; df ¼ 1; OR: 1.69). Assuming a recessive mode of inheritance, the frequency of homozygote 4Pro carriers was significantly increased in the SZ patients (24%) when compared to controls (8%) (p ¼ 0.0006, df ¼ 1; OR: 3.74). Haplotype analysis showed nominal significance for an individual risk haplotype (p ¼ 0.013). Expression analysis confirmed VMAT1 in post mortem human brain at the mRNA level. Results suggest that variations in the VMAT1 gene may confer susceptibility to SZ in patients of European descent. Further studies are necessary to confirm this effect and to elucidate the role of VMAT1 in the genetic origins of SZ. P246 ASSOCIATION STUDY OF CSF2RB WITH SCHIZOPHRENIA IN THE IRISH STUDY OF HIGH-DENSITY SCHIZOPHRENIA FAMILIES Qi Chen1, Xiangning Chen1, Xu Wang1, Anthony O’Neill2, Dermot Walsh3, Kenneth S. Kendler1 1 Virginia Commonwealth University, Virginia 2 The Queens University, Belfast 3 The Health Research Board, Dublin Background: CSF2RB (Colony Stimulating Factor 2 Receptor Beta is a subunit of the functional receptor for interleukin 3 (IL-3), colony stimulating factor 2 and interleukin 5, and is responsible for the initiation of signal transduction triggered by ligand binding. In our previous study, we showed evidence that IL-3 gene is associated with schizophrenia. CSF2RB is located on 22q13.1, a region implicated in harbouring risk genes for schizophrenia in linkage studies. These findings suggestthat CSF2RB may be a candidate gene for schizophrenia. Methods: We studied 9 SNPs (rs2072707, rs2284031, rs17811365, rs909486, rs2075936, rs11705394, rs7285064, rs131840, rs131842) in the CSF2RB gene in the Irish study of high density schizophrenia families (ISHDSF) to test whether it is associated with schizophrenia. Results: Using the pedigree disequilibrium test, we found positive association (p ¼ 0.024) for SNP rs11705394. Another marker, rs7285064, showed a trend (p ¼ 0.075). Interestingly, when we analysed by gender, these two SNPs showed gender-specific associations (for males, p ¼ 0.0066 and 0.011, for females, p ¼ 0.49 and 0.37
794
Abstracts
respectively). In haplotype analyses, we also found male-specific associations. Conclusions: Our results suggest that CSF2RB may be associated with schizophrenia, and the associations observed are gender-specific. P247 HOMOZYGOSITY MAPPING IN A CONSANGUINEOUS FAMILY WITH SCHIZOPHRENIA CO-MORBID WITH EPILEPSY Helen Miranda Knight1, Alan Maclean1, Ben Pickard1, Walter Muir1, Douglas Blackwood1, Mohammad Ayab2 1 University of Edinburgh, Edinburgh 2 St Lukes Hospital, Middlesborough Homozygosity mapping with consanguineous families is a powerful method of localising autosomal recessive genes, as parental consanguinity provides a high degree of assurance that all affected children will be homozygous- by-descent (HBD) over the genomic region that includes the disease gene (s). The kindred investigated comes from Pakistan, a region where consanguineous marriages are relatively common. The parents have no detectable clinical disorders; however, five out of the six children have a rare presentation of schizophrenia comorbid with epilepsy and sensori-neural hearing impairment. This clustering of phenotypes in several offspring of a first cousin marriage is strongly suggestive of a rare, Mendelian recessive disorder. An initial genome-wide scan, performed using microsatellite markers, identified 6 markers showing homozygosity by descent (HBD) pattern located on chromosome 22q13. HBD in this region was confirmed through a second SNP genome-wide scan, mapping the putative disease locus to a 12Mb region on 22q12.3-13.3. Moreover, SNP genotype analysis also indicated a second HBD interval on chromosome 2p24.3. The putative disease loci on chromosome 2p24.3 overlaps with a deafness locus, DFNB47, linked to autosomal recessive nonsyndromic hearing impairment mapped in two Pakistani kindreds. In contrast, chromosome 22q12 shows linkage for epilepsy, and positive findings for affective psychosis and schizophrenia have been reported to cluster in a region of 4-5cM on 22q13.1. The sequencing of candidate genes in these newlyidentified intervals may lead to the discovery of causative mutations associated with this rare phenotype. P248 SYSTEMATIC MUTATION SCREENING STUDY FOR DISC1 AND NRG1 IN SCHIZOPHRENIA Chih-Min Liu1, Hai-Gwo Hwu1, Yuli Liu2, Ueng-Cheng Yang3, Pei-Chun Hsu3, Cathy S-J Fann4, Wei-Chih Yang4 1 National Taiwan University Hospital, Taipei 2 National Health Research Institute, Taipei 3 Institute of Bioinformatics, National Ya, Taipei 4 Institute of Biomedical Science, Academi, Taipei Disruptions in Schizophrenia 1 (DISC1) and Neuregulin 1 (NRG1) have been reported to be associated with schizophrenia in literature and in our previous association studies. The aim of this study is to search for risk polymorphisms of the two genesfor schizophrenia, using the systematic mutation screening method. We have established standardised bioinformatic processes by using several tools to define the exon regions, promoter regions and highly conserved regions of the two genes. We have sequenced about 21 kb and 19 kb of the above regions of DISC1 and NRG1, respectively. Samples for direct sequencing are 100 schizophrenic patients and 50 normal controls. We have identified 93 genetic variations in DISC1, of which 63 have not beenreported in NCBI dbSNP125, and 51 genetic variations in NRG1, of which 36 are novel. We will do genotyping work on 37 and 14 genetic variations in DISC1 and NRG1 respectively in large case-control samples. P249 SCHIZOPHRENIA AND MBL DEFICIENCY: PROPER HANDLING OF MBL DETECTION LIMIT? Leslie Foldager, Thomas Damm Als Centre for Basic Psychiatric Research, Risskov Increased risk for schizophrenia has been associated with a history of autoimmune diseases (Am. J. Psychiatry 163:521–528). The most common hereditary defect in the human immune system is mannanbinding lectin (MBL) deficiency, predisposing to infection and autoimmunity in individuals with immature or compromised adaptive immunity (Mol. Immunology 38:133–149). This leads to a hypothesis of
association between schizophrenia and hereditary immune defects. The MBL serum concentration level defining deficiency is an open question, and various levels have been used; < 10, < 50, < 100, < 500 ng/ml (personal communication, JC Jensenius). A problem arises, as 10–15% below a 10 ng/ml detection limit is common. We conducted a simulation study to evaluate methods dealing with this problem when examining association between MBL in serum and SNPs in the MBL encoding gene (MBL2). Samples of 100 cases and 300 controls with genotypes for three MBL2 promoter region SNPs and three SNPs within MBL2 exon1 were generated using HapSim (Bioinformatics 21:4309–4311), assuming week association between cases and promoter region SNPs. MBL serum concentrations were generated with a larger proportion of MBL deficient cases than controls. MBL deficiency was defined arbitrarily as < 100 ng/ml, and a 10 ng/ml detection limit was assumed. Methods and software developed for tobit regression or survival analysis could potentially be used if haplotype frequencies were known. But, as this is often not the case, application of haplotype association methods (AJHG 70:425–434, Human Heredity 53:79–91) was evaluated. P250 GLOBAL TRANSCRIPTOME ANALYSIS OF THE PREFRONTAL CORTEX USING A RODENT PCP MODEL OF SCHIZOPHRENIA Brian Morris1, H Ozeki2, Catherine Winchester3 1 Yoshitomi Research Institute of Neurosci, Glasgow 2 YRING, Glasgow 3 YRING & University of Strathclyde, Glasgow The cognitive deficits of schizophrenia (impaired working memory, attentional and executive function deficits) are considered a core feature of the disease resulting from dysfunction in the prefrontal cortex in particular. Despite concerted efforts into the cause of schizophrenia, the genetic components of this complex disease remain elusive. We performed a global transcriptome screen to identify schizophrenia-associated genes differentially expressed in the prefrontal cortex, utilisinga phencyclidine (PCP) rat model of schizophrenia (Cochran et al., 2003, Neuropsychopharmacology. 28 (2):265– 275) and rat oligonucleotide GeneChips from Affymetrix. This PCP model of schizophrenia developed in our laboratories produces a pattern ofmetabolic hypofunction, neurochemical changes and behavioural deficits in the rodent brain that closely mirror those observed in the brains of schizophrenic patients (Cochran et al., 2003 and Morris et al., 2005, Current Opinions in Pharmacology 5:101–106). 327 differentially-expressed transcripts were identified. Many of the associated genes map to key schizophrenia loci, including the neurodevelopmental gene DPYSL2 that maps to 8p21-22 and has been previously implicated as a susceptibility gene for schizophrenia, and the AMPA receptor subunit gene GRIA1 (GluR1) that maps to SCZD1 schizophrenia susceptibility locus 1) on 5q33. We present the microarray identification of schizophrenia-associated genes and real-time PCR data on differentially-expressed genes mapping to key schizophrenia loci. We conclude that microarray analysis of a validated animal model of schizophrenia can lead to the identification of genes involved in the cognitive aspects of the disease. P251 SCREENINGS OF PATIENT-SPECIFIC ALU INSERTIONS SITES OF GENES ASSOCIATED WITH SCHIZOPHRENIA Mikako Ueno1, Akihisa Akahane2, Takanori HATA2, Kenichi Matsumura2, Hiroko Seki2, Shinichiro Nanko2 1 Dep. of Psych., Teikyo Univ. Sch. of Med, Tokyo 2 Department of Psychiatry, Teikyo Univers, Tokyo Human mobile elements are DNA segments that can be inserted into or deleted from the human genome. They can alter the construction and expression of neighbouring genes, ultimately leading to human diversity. Both mobile elements and DNA methylation have recently been reported as the cause of human diseases. We speculated that the behaviour of mobile elements in genes under low or aberrant methylation may be one of the factors responsible for mental disorders. Based on this hypothesis, we attempted to detect risk genes by screening insertion sites of mobile elements in patients with schizophrenia. To discover the flanking regions of the mobile element, retroposon Alu insertion in genomic DNA of the 11 trios (a patient with schizophrenia and his/herparents) was amplified by Alu-specific PCR.
Abstracts Southern hybridization of PCR products using candidate genes for schizophrenia was performed to detect de novo or patient-specific insertion sites. Screenings using Alu-specific PCR andsouthern hybridisation identified several DNA fragments in patients with schizophrenia, which were not recognised in their parents. These patient-specific fragments with hybridisation signals may contain the risk genes for schizophrenia. These results suggest that the behaviour of mobile elements may contribute to susceptibility to schizophrenia and that Alu insertion sites are useful in detecting risk genes associated with schizophrenia. This research was done with the approval of the Ethical Committee for Genetic Research, Teikyo University School of Medicine. P252 ASSOCIATION ANALYSIS BETWEEN SNPS IN DISC1 AND SCHIZOPHRENIA IN THE POLISH POPULATION Piotr M. Czerski1, Anna Karpushova2, Monika Wilkosc3, Pawel Kapelski4, Beata Godlewska5, Marta Tomaszewska3, Anna Leszczynska-Rodziewi4, Thomas G. Schulze5, Alexander Georgi5, Joanna Hauser4, Jerzy Landowski5, Alina Borkowska3, Markus M. No¨then2, Sven Cichon2 1 Poznan University of Medical Sciences, Poznan 2 Department of Genomics, Life & Brain Cen, Bonn 3 Clinical Neuropsychology Unit, Nicolaus, Bydgoszcz 4 Department of Psychiatry, Poznan Univers, Poznan 5 Department of Psychiatry, Medical Univer, Gdansk 5 Division of Genetic Epidemiology in Psyc, Mannheim Objective: Linkage between schizophrenia and chromosome 1q42.2 markers has been reported independently in different populations worldwide. This region contains the gene Disrupted In Schizophrenia 1 (DISC1,) found to be disrupted by a balanced translocation and to cosegregate with schizophrenia in a large Scottish pedigree. Association studies investigating DISC1 single nucleotide polymorphisms (SNPs) in a number of samples were promising but not yet compelling. In the present study, we aimed at investigating the possible contribution of DISC1 variants to the pathogenesis of schizophrenia in the Polish population. Methods: We investigated five SNPs within DISC1: rs1030711, rs1538977, rs2492367, rs2812393 and rs821616. MALDI-TOF mass spectrometry-based SNP genotyping was performed by Sequenom’s iPLEX technology. 150 DSM-IV patients diagnosed with schizophrenia and 150 controls were genotyped. All individuals participating in the study originated from the Polish population and were recruitedin three clinical centres (Poznan, Bydgoszcz, Gdansk). Informed consent from participants was obtained. Cocaphase v2.401 software was used for statistical analysis. Results: None of five analysed DISC1 SNPs was associated with schizophrenia when single marker analysis was performed. Consistently, sliding-window haplotype analysis gave no significant results. Discussion: Our data provide no support for an involvement of the five analysed DISC1 SNPs in predisposition to schizophrenia in the Polish population. Due to the limited sample size and incomplete SNP coverage, we cannot exclude small gene effects or an effect of DISC1 variants not captured by the SNPs selected for this study. P253 PSYCHOPATHOLOGY, PERSONALITY TRAITS, AND SOCIAL DEVELOPMENT OF ADOLESCENT AND YOUNG ADULT FIRST-DEGREE RELATIVES OF SCHIZOPHRENIA PATIENTS Stephen Glatt1, William Stone2, Stephen Faraone3, Larry Seidman2, Ming Tsuang1 1 University of California, San Diego, La Jolla, CA 2 Harvard Medical School, Boston, MA 3 SUNY Upstate Medical University, Syracuse, NY Evaluation of individuals at high genetic risk for schizophrenia is a powerful method for identifying precursors of the illness. To identify the aspects of personality, psychopathology and social development that differentiate high-risk and controlindividuals, adolescent and young-adult first-degree relatives of schizophrenia and schizoaffective disorder patients and control subjects were compared on 36 measures at baseline of a longitudinal study. Measures differentiating high-risk and control subjects were related to four genetic loading indices. Highrisk subjects older than 17 years showed more physical anhedonia, less
795
positive involvement with peers and more problems with peers, siblings and the opposite sex. Older high-risk individuals were also less cooperative, less self-directed and less reward-dependent. Problems with peers and the opposite sex, as well as reward dependence, were related linearly to genetic loading. Alterations in personality traits and social development are present in high-risk individuals, and may be markers for genetic liability toward the illness. P254 LINKAGE OF SCHIZOPHRENIA WITH CHROMOSOME 1q32 IN KOREAN MULTIPLEX FAMILIES Kyung Sue Hong1, Jong Won Kim1, Yu-Sang Lee2, Dong Yeon Park1, Eun-Young Cho1, Hyun Ok Jeun2, Dongsoo Lee1, Kyoo-Seob Ha2, Yong Lee Jang3 1 Sungkyunkwan University, Seoul 2 Yong-In Mental Hospital, Yong-In 2 Samsung Seoul Hospital, Seoul 2 Seoul National University, Seoul 3 Seoul Medical Center, Seoul Chromosome 1q contains a few loci that have been identified as having genetic linkage with schizophrenia in several independent studies. However, markers showing peak linkage signal are dispersed over a large chromosomal region. In addition, inconsistent findings have been generated from different populations or different subgroups in the same population. The purpose of the current study is to determine whether those loci are linked to schizophrenia in the Korean population. We investigated46 Korean multiplex schizophrenia families, initially using eleven microsatellite markers spanning around 91 cM region of 1p2242. In a non-parametric linkage analysis, D1S249 located on 1q32.1 showed statistical evidence suggestive of linkage. During second-stage analysis to narrow down the region, four additional nearby markers were genotyped. In the single point analysis, we found another statistically-significant linkage signal at D1S2891. The highest NPL score of 2.67 (P ¼ 0.0039)was obtained in the multi-point analysis. This study provides additional evidence for linkage of chromosome 1q32 with schizophrenia. P255 SERT GENE POLYMORPHISM AND THERAPEUTIC RESPONSE IN CROATIAN FEMALES WITH SCHIZOPHRENIA Nikolina Jovanovic1, Martina Rojnic Kuzman2, Vesna Medved2, Nada Bozina3 1 University Hospital Zagreb, Zagreb, Croatia 2 University Hospital Center, Department o, Zagreb 3 University Hospital Center, Clinical Ins, Zagreb Recent scientific findings suggest that functional polymorphisms of the serotonin transporter (SERT)gene may be involved in the therapeutic response of psychotic disorders to the second generation of antipsychotics. Although intron parts of a gene are not transcribed, both common repeats of the variable-number-tandem-repeat (VNTR), the short (s) and long (l)alleles, are purported to increase its transcription. The aim of the study was to determine association between the serotonin transporter intron2 VNTR region (SERTin2) and treatment response. Ninety-five olanzapine-treated female patients with schizophrenia were included in the study. Assessment and evaluation of improvement in clinical psychotic symptoms and therapeutic response to antipsychotic medication included the Positive and Negative Symptom Scale (PANSS).Genomic DNA was isolated from a whole blood sample, and SERT transporter intron polymorphisms were identified by PCR-RFLP (polymerase chain reaction-restriction fragment length polymorphisms). We found no association between SERTin2 genotype and treatment response, measured with total PANSS, positive, negative and general psychopathology subscales. Although it will be important to extend the present study to a larger sample, our initial results suggest that l/s intron2 (SERTin2) gene polymorphism is not likely to have predictive value as a pharmacogenetic marker of treatment response to olanzapine in Croatian females with schizophrenia. P256 GENETIC HETEROGENEITY AT THE CHROMOSOME 15q14 CANDIDATE SCHIZOPHRENIA LOCUS Savira Ekawardhani1, Dirk Moser2, Ulrike Schuelter2, Christian Vogler2, Haukur Palmason2, Klaus-Peter Lesch3, Jobst Meyer2 1 Dept. Neuro-behavioral Genetics, Trier 2 Division of Neuro-Behavioral Genetics, G, Trier 3 Clinic for Psychiatry and Psychotherapy,, Wuerzburg
796
Abstracts
Our previous results confirmed the presence of a susceptibility locus for psychoses on chromosome 15q14–15q15.1. Non-parametric multipoint linkage analyses comprising 12 multiplex families with periodic catatonia (SCZD10, OMIM #605419) identified a region with a maximum peak of Zall ¼ 5.39 (p ¼ 0.001). Seven of twelve families showed linkage, indicating genetic heterogeneity. Non-parametric multipoint analyses comprising only these chromosome 15-linked families resulted in a narrower susceptibility region, with a peak at marker D15S118 (NPLall ¼ 2.45), and a second peak at marker D15S994 (NPLall ¼ 1.99). Statistical power was reduced as expected when using fewer families, but the shape of peaks improved and appears more helpful in selecting candidate genes for sequencing. The highest peak at marker D15S118 also supported simultaneous findings from our group in one family (F11) for a functional rare variant in SLC12A6, a gene located close to this marker, which co-segregates with thesyndrome. In more detail, our study would show that a rare G variant leads to methylation of the adjacent cytosine in the SLC12A6 promoter. Furthermore, decreased G variant promoter activity could be proven by luciferase assay, especially under stimulated conditions. Conclusions: A functional rare G variant located near the maximum NPL score at SLC12A6 was found, but only in one family. Therefore, we conclude that there should be a second susceptibility gene for catatonic schizophrenia in our families, with linkage at the 15q14.15.1 locus. P257 GENE X CANNABIS INTERACTION: CASE-ONLY DESIGN ANALYSIS IN A FIRST-EPISODE PSYCHOSIS SAMPLE Marta Di Forti, Caterina LaCascia, Alexander Butt, Ana Miorelli, Valeria Mondelli, Joanna Eyeson, Pietro Papili, Serena Navari, Monica Aas, Damien Clifford, Angelo Ricciardi, Raffael Gafoor, Paola Dazzan, Carmine Pariante, Kathy Aitchison, Craig Morgan, Ashalom Caspi, Temi Moffitt, Bernard Freeman, Peter McGuffin, David Collier, John Powell, Robin Murray Institute of Psychiatry, London Cannabis consumption is associated with approximate doubling of the risk of subsequent schizophrenia. There is evidence that a functional val158met polymorphism in the catechol-O-methyltransferase gene (COMT) and use of cannabis interact to increase risk. This gene, with environment (G x E) interaction, is not accounted for by a correlation between cannabis use and the COMT genotype (Caspi, 2005, Henquet, 2005). We predicted a significant difference in the frequency of the COMT Val158 allele between cases consuming and not consuming cannabis, consistent with those homozygous for the Val158 allele being at highest risk of developing schizophrenia when cannabis users. We collected demographic, clinical and cannabis-use data and DNA in a sample of 223 first-episode psychosis patients, recruited as part of the Genetic And Psychosis (GAP) study. The COMT val158met polymorphism was genotyped by standard procedures and a case-only design (Xin Liu et al., 2004) conducted. Results: 123 (55.2%) smoked cannabis; 63.4% of them began their use before age 16, and mean length of cannabis use was 9.5 years. There was no significant difference in cannabis consumption between ethnic groups. As in case-only studies of gene-environment interaction, we calculated an odds ratio(OR) from cross-classification of cannabis use and COMT genotype in patients. Logistic regression was used to check for the effect of potential confounders. A case-series interaction OR > 1 in this study indicates that the effect of cannabis use on later psychosis is moderated by the COMT genotype. P258 EDUCATIONAL ACHIEVEMENT, CANNABIS USE, AND GENOTYPE IN PREDICTING PSYCHOSIS Alexander Butt, Marta Marta, Caterina La Cascia, Ana Miorelli, Valeria Mondelli, Pietro Papili, Damien Clifford, Serena Navari, Monica Aas, Raffael Gafoor, Joanna Eyeson, Carmine Pariante, Paola Dazzan, Craig Morgan, Kathy Aitchison, Peter McGuffin, Bernard Freeman, David Collier, John Powell, Robin Murray Institute of Psychiatry, London Schizophrenia involves both genetic and environmental factors, which by different mechanisms and at different times, disrupt neurodevelopment. Cannabis appears to interact with the catechol-O-methyltransferase (COMT) val158met genotype to increase risk (Caspi, 2005;
Henquet, 2005). In adolescents at high risk of developing psychosis, impaired cognitive ability appears to be an endophenotype marker of genetic vulnerability (Erlenmeyer-Kimling, 2001). There is evidence that genes like dysbindin (DNTBP1), neuregulin 1(NRG1) and disrupted in schizophrenia 1 (DISC1) confer illness susceptibility through their role in neurodevelopment (Kirov, 2005). We expect significant difference in genotype for COMT, dysbindin, DISC1 and neuregulin1 polymorphisms between case subjects using cannabis with educational qualifications (EQ) compared to case subjects not using cannabis without EQ. We collected socio-demographic, clinical, cannabis use data and genotype data in a sample of 223 first-episode psychosis patients. Information on EQ was used to measure cognitive function. Results: 123 (55.2%) smoked cannabis and 63.4% of them began their use before age 16. The mean age at first cannabis use was 16.4 (SD 4.4) years. We had information on EQ in 103 subjects; 62 (60.2%) achieved school qualifications (age 16 and above) and counter intuitively 54 (87%) of them had used cannabis. Patients were divided by the presence or absence of cannabis use and by EQ, and compared for geneticpolymorphisms. Schizophrenia is an aetiologically-complex disorder arising from the interaction of a range of environmental and genetic factors acting at various stages of life. P259 SCHIZOPHRENIA IN VELO-CARDIO-FACIAL SYNDROME Stephen Monks, Kieran Murphy Royal College of Surgeons in Ireland, Dublin Velo-cardio-facial syndrome (VCFS) is usually caused by a deletion of chromosome 22q11. It has been shown that people with VCFS have an extraordinary risk of developing schizophrenia (25%). The schizophrenia phenotype has differed between studies, for example with varying mean age of onset. 50 adults with 22q11 deletions identified by Fluorescent in-situ hybridisation (FISH) were assessed.Intelligence was measured using the Wechsler Abbreviated Scale of IQ (WASI; Wechsler, 1999). Standardised clinical assessment was performed using the Schedules for Clinical Assessment in Neuropsychiatry (Wing et al., 1990). The Oxford-Liverpool Inventory of feelings and emotions was administered to measure the Schizotypy trait. Positive and negative symptoms of schizophrenia were measured with SAPS and SANS (Schedule for Assessment of Positive/Negative symptoms, Andreasson). 29.5% of participants had a psychotic disorder, with 13% fulfilling DSM- IV criteria for schizophrenia. In addition,10.5% had major depression without psychotic features and 5% had an anxiety disorder. Amongst the schizophrenic group, mean IQ was 72, with a mean age of onset of 20.6 years. The mean SAPS and SANS score was 34 and 19 respectively. There was no difference in mean IQ between schizophrenic patients and non- schizophrenic patients with VCFS ( t ¼ 1.5, p ¼ 0.144). These results show a high rate of mental illness, including schizophrenia, in individuals with velo-cardio-facial syndrome. The schizophrenic patients showed more positive symptoms and fewer negative symptoms with an age of onset similar to schizophrenia in the general population. P260 ASSOCIATION MAPPING OF THE GLIAL CELL LINE-DERIVED NEUROTROPHIC FACTOR (GDNF) GENE IN SCHIZOPHRENIA Hywel Williams1, Nadine Norton2, Tim Peirce2, Sarah Dwyer2, Nigel Williams2, Michael O’donovan2, Michael Owen2 1 University Hospital of Wales, Cardiff 2 Cardiff University, Cardiff The glial cell line-derived neurotrophic factor (GDNF) gene is located within a region of chromosome 5 (5p14.1–q13.3) that has been highlighted as a potential schizophrenia susceptibility loci by a number of genome scans. GDNF is neurotrophic and apotential differentiation factor for dopaminergic systems. Due to the hypothesized role of dopaminergic systems in the aetiology of schizophrenia and the observation that neuroleptic medications mediate their action via dopamine neurotransmitters,GDNF is a potential schizophrenia candidate gene. To test this hypothesis, we performed LD-based association mapping across GDNF. Using HAPMAP version2 as a resource and LDSelect software, we chose 9 SNPs that spanned a genomic region of 40kb, fully encompassing the coding sequence for GDNF. SNPs were genotyped in a sample of 673 schizophrenic patients and 716 matched controls, all Caucasian origin and all collected from
Abstracts the UK or Eire. Of the 9 SNPs genotyped, 2 showed significant genotypic association at p < 0.05 level (rs2973050; P-Value ¼ 0.007 and rs2910702; P-Value ¼ 0.039). We then performed 2 and 3 marker haplotype analysis using cocaphase software and detected two, 3 marker haplotypes with simulated p values < 0.05. Given the number of SNPs tested and the level of significance achieved, none of the markers, either individually or combined in a 2 or 3 marker haplotype, were significant when corrected for multiple testing. We conclude that our sample provides no support for the hypothesis that GDNF is a susceptibility gene for schizophrenia. P261 LACK OF SUPPORT FOR ASSOCIATION BETWEEN AKT1 AND SCHIZOPHRENIA IN A UK CASE CONTROL SAMPLE Nadine Norton, Hywel Williams, Sarah Dwyer, Liam Carroll, Tim Peirce, Valentina Moskvina, Ricardo Segurado, Nigel Williams, Michael Owen, Michael O’Donovan Cardiff University, Cardiff AKT1 (V-akt murine thyoma viral oncogene homolog 1) is involved in intracellular signalling pathways postulated as being of aetiological importance in schizophrenia. Markers in the AKT1 gene have also recently been associated with schizophrenia in two samples of European origin and in Japanese and Iranian samples. Aiming to replicate these findings, we examined ten SNPs spanning AKT1 in a UK case-control sample (schizophrenia cases n ¼ 673, controls n ¼ 716). These included all SNPs previously reported to be associated in European and Japanese samples, alone or in haplotypes, as well as additional markers defined by the Haploview Tagger programme (pairwise tagging, minimum r2 ¼ 0.8, minor allele frequency ¼ 0.02). We found no association with single markers (min p ¼ 0.17). We found weak evidence for association (p ¼ 0.016) with a three marker haplotype that had previously been reported as significant in a Japanese sample, but this did not survive correction for the number of specific prior hypotheses specified at the outset of this investigation. Our study provides no support for the hypothesis that AKT1 is a susceptibility gene for schizophrenia. Examination of our own data and those of other groups leads us to conclude that overall, the evidence for association of AKT1 is not robust. P262 NO ASSOCIATION BETWEEN GENETIC VARIANTS AT THE ASCT1 GENE AND SCHIZOPHRENIA OR BIPOLAR DISORDER IN A GERMAN SAMPLE Markus Hans Skowronek1, Alexander Georgi1, Rami Abou Jamra2, Johannes Schumacher2, Tim Becker3, Christine Schmael1, Torsten Paul1, Monika Deschner1, Susanne Ho¨fels2, Maren Wulff3, Markus Schwarz3, Norman Klopp4, Thomas Illig4, Peter Propping2, Sven Cichon3, Markus M No¨then3, Thomas G Schulze1, Marcella Rietschel1 1 Central Institute of Mental Health, Mannheim 2 Institute of Human Genetics, University, Bonn, 3 IMBIE, University of Bonn, Bonn 2 Department of Psychiatry, University of, Bonn 3 Nordbaden Center of Psychiatry, Wiesloch 4 Institute of Epidemiology, GSF, Neuherberg 3 Department of Genomics in Life & Brain C, Bonn Altered glutamatergic neurotransmission is considered a potential etiological factor in schizophrenia (SCZ) and affective disorders. The gene ASCT1 (SLC1A4), a member of the glutamate transporter superfamily, is located on 2p13-14, a region showinglinkage to both SCZ and bipolar disorder (BD). ASCT1 can thus be considered a candidate gene for both disorders. In a German sample, we tested for association between ASCT1 and both SCZ and BD (according to DSMIV). Samples were as follows: for SCZ, 330; for BD, 306 patients; 319 population-based controls (CON). Genotypes for SNPs rs2075209, rs1064512, rs3732062, rs759458 and STR (STR-SLC1A4-11) were in HWE. Power to detect an effect at genotype relative risks of 1.3–1.5 was moderate to good (45–90%). Allele and haplotype frequencies did not differ between cases and controls. Recent findings from our group on the associations between BDNF and SCZ and between G72/G30 and BD suggest that SCZ patients with a history of major depressive episodes (MDE) outside psychotic episodes and BD cases with a history of persecutory delusions constitute genetically distinct subgroups of these disorders. Thus, we hypothesized that restricting case definition to
797
those 95 SCZ individuals with MDE and to those 107 BD patients with a history of persecutory delusions might clarify the relationship between SCZ, BP and ASCT1. However, these stratification approaches did not yield any significant association either. Our results do not support an association of the ASCT1 gene with BD or SCZ in the German population.
P263 ASSOCIATION STUDY OF APOE GENOTYPES IN A SAMPLE OF LATINO SCHIZOPHRENIA SIB PAIR Alfonso Tovilla1, Rogelio Apiquia´n2, Ana Fresan3, Michael Escamilla4, R. Mendoza5, A. Jerez5, R. Mun˜oz4, Henrrietta Raventos4, Jose Can˜ive4, Alfonso Ontiveros4, R. Median4, Beatriz Camarena3, Humberto Nicolini1 1 Universidad Auto´noma de la Ciudad de Me´x, Me´xico D. F. 2 Grupo de Estudios Medicos Caracci, Me´xico, D.F. 3 Instituto Nacional de Psiquiatrı´a Ramo´n, Me´xico, D.F. 4 The Genetics of Schizophrenia in Latino, San Antonio, TX 5 The genetics of schizophrenia in latino, San Antonio, TX Introduction: The role of the APOE gene in schizophrenia is not completely understood. One line of thought suggests that APOE e4 may confer a risk factor for the development of schizophrenia and an earlier age of onset. In this work, we evaluate the participation of APOE in schizophrenia in a sample of affected siblings. Methods: We analysed the genotype frequency and alleles of APOE, in a sample of 60 families of Mexican origin with two sibs with schizophrenia. We genotyped 179 individuals: 128 were affected and 52 not affected. Diagnoses were based on a structured interview with all subjects (DIGS-Spanish version). Results: There were no significant differences observed between groups, in the allelic (w2 ¼ 4.202, df ¼ 6, p ¼ 0.6493) neither for genotype (w2 ¼ 7.827, df ¼ , p ¼ 0.5517) frequencies. When dividing the sample for gender, we observed an increase frequency of the Apo-"2 allele in the group of non-affected male siblings (w2 ¼ 8.010, df ¼ 2, p ¼ 0.0182), as well as for genotype(w2 ¼ 8.41, df ¼ 2, p ¼ 0.0149). We did not observe a decrease in the age of onset of the illness in the probands bearing allele "4. Conclusions: Our results did not show an association between ApoE" 4 and schizophrenia; however, we observed that allele "2 could be associated as a protection allele against developing schizophrenia in males.
P264 ALPORT SYNDROME AND SCHIZOPHRENIA CO-EXIST IN MULTIPLE MENBERS OF A SINGLE SPANISH FAMILY Miguel Ruiz Veguilla1, J. Cervilla2, E. Molina2, P. Torres3, J.C. Alvarez2, B Gutierrez2 1 Hospital Neurotraumatologico de Jae´n, Jae´n 2 Dept. of Legal Medicine and Psychiatry a, Granada 3 Hospital Neurotraumatolo´gico, Jaen We report a family with multiple members (3) affected by both Alport syndrome and schizophrenia. Alport disease (which is characterised by chronic nephritis and hearing loss) has affected subjects in this family for at least 5 generations. In a significant proportion of patients with Alport syndrome, disease seems to map to the long arm of X chromosome. However, in this family, the illness is clearly transmitted following a mitochondrial pattern of inheritance, as individuals in the maternal line present the illness and no male with the disease transmits it to his children. Interestingly, some clinical, genetic and neuroimaging studies have implicated mitochondrial dysfunction in the pathophysiology of schizophrenia. Mitochondrial inheritance has also been described in many forms of familial hearing loss. Within this family, schizophrenia and Alport syndrome co-existed in three out of four sibs. Clinical features were very similar in all three cases, presenting marked cognitive impairment and low IQ. We found no obstetric complications or dermatoglyphic abnormalities, suggesting no neurodevelopmental involvement. The onset of schizophrenia correlated with time at which hearing loss diagnosis was made and to hearing loss severity. The father, suffering from paranoid schizophrenia, was the only evidence of previous history of any psychiatric disorder. Blood samples for genetic analyses were collected from all available subjects (10). Karyotypes were completely normal for all of them. A mutational screening in mitochondrial DNA is now being
798
Abstracts
performed in order to identify the mutation involved in the transmission of Alport syndrome and explore its relationship to schizophrenia. P265 LOCUS-WIDE ASSOCIATION STUDY OF GENES IN THE 5q33.1 REGION WITH SCHIZOPHRENIA Yasuyuki Fukumaki1, Hiroki Goto1, Hiroki Shibata1, Masafumi Takaji1, Hideaki Ninomiya2, Nobutada Tashiro1, Ngamwong Jarusuraisin2 1 Kyushu University, Fukuoka 2 Dazaifu Hospital, Fukuoka 2 Chiang Mai University, Chiang Mai We have reported that 11 chromosomal regions showed nominal significance in a genome-wide scan for linkage (JSSLG 2003). Out of these regions, we selected the 5q33.1 one, where suggested linkage with schizophrenia has been reported in multiple studies, for further analyses. We conducted locus-wide association studies on all annotated genes in the region (43 genes in about 4.9 Mb). We genotyped 100 Japanese cases and 100 controls for 131 single nucleotide polymorphism (SNP) markers distributed at intervals of less than 30 kb within each gene region. We observed significant associations of schizophrenia with 7 SNPs in 6 genes by ‘‘single-point’’ analyses and with 22 pairs of SNPs in 10 genes by haplotype analyses. All nominal significance was subsequently examined by genotyping the second set of Japanese samples (100 cases and 120 controls). We replicated one ‘gsingle-point’h association of schizophrenia with a SNP in SP329 (p ¼ 0.0434 in the first set and p ¼ 0.0140 in the second set). We also observed a significant haplotype association in two SNPs in PDE6A in both sample sets (p ¼ 0.0031 in the first set and p ¼ 0.0409 in the second set). We conclude that SP329 and PDE6A are responsible for genetic susceptibility in Japanese schizophrenia. However, neither of these significant associations was replicated in the additional typing of Thai samples (202 cases and 201 controls), suggesting the ethnic heterogeneity of these genes in schizophrenia susceptibility. P266 GENETIC EVIDENCE FOR INVOLVEMENT OF NRG1/ERBB SIGNALLING INOLIGODENDROCYTES IN SCHIZOPHRENIA Lyudmila Georgieva1, Jennifer Wilkinson1, Nadine Norton1, Takeshi Sakurai2, James Young2, Christopher Plescia2, Kenneth L Davis2, Vahram Haroutunian3, Joseph D Buxbaum2, Nick Craddock1, George Kirov1, Michael J Owen1, Michael C O’Donovan1 1 Cardiff University, Cardiff, 2 Mount Sinai School of Medicine, New York 3 Mount Sinai School of Medicine/MIRECC, B, New York Neuregulin (NRG1) is a schizophrenia susceptibility gene involved in multiple aspects of nervous system function, including neuronal and glial cell development, synapse formation, oligodendrocyte survival and myelination. Recent studies including our own suggest the NRG1 receptor erbB4 also contributes to schizophrenia susceptibility, suggesting that signalling pathways involving NRG1/erbB4 might be affected in schizophrenia. In order to identify downstream signalling molecules of erbB4, wecarried out a yeast two-hybrid screen for proteins interacting with the erbB4 cytoplasmic domain. We identified the MAGI family of scaffolding proteins. MAGIs have been previously shown to interact with the receptor phosphotyrosine phosphatase, RPTPbeta, a molecule which like NRG1 signalling, has also been implicated in oligodendrocyte development. Given their functions, we postulated that erbB4 and RPTPbeta might form a molecular complex that regulates neuregulin signalling. To examine whether such a complex might further contribute to schizophrenia susceptibility, we examined 28 SNPs spanning RPTPbeta for evidence for association with schizophrenia in a large UK case-control sample. Multiple associated markers were observed. Despitethe large number of markers tested, the evidence for association remained significant at a gene-wide level. Moreover, markers in RPTPbeta showed evidence for epistatic effects on disease risk in combination with markers in a gene for which we have previously reported strong evidence for association with schizophrenia, that is a key regulator of oligodendrocyte development (OLIG2). These results suggest that alterations in NRGerbB4- RPTPbeta signalling influence risk of schizophrenia and do so through influencing oligodendrocyte development.
P267 STRESS GENETICS AND SUICIDAL BEHAVIOUR August G. Wang1, David Woldbye2, Pernille Kofoed2, Henrik B. Rasmussen3, Thomas Werge4 1 Dep. of Psychiatry, Amager CUH, Copenhagen S 2 Copenhagen University, Copenhagen 3 Research Dep. Sct. Hans Hospital, Roskilde 4 Research Dep. Sct. hans Hospital, Copenhagen Objective: Evidence suggests an important role for genetic factors in suicidal behaviour, independent of the genetics of the possible psychiatric illness in question. Within the stress-diathesis model, neuropeptide Y has been suggested to play a role. Method: The idea is to search for this association in one psychiatric illness. All participants have been subjected to stress by the illness itself. The object in this preliminary study is to identfy possible associations between variations in the NPY gene and history of suicidal behaviour in patients with schizophrenia. We investigated the occurrence of NPY polymorphisms and the history of suicidal behaviour among about 300 schizophrenic patients. NPY allele, genotype and haplotype frequencies were compared in persons who had no history of attempted suicide, using non-violent methods or violent methods. Results: Results thus far have shown no significant difference in variations in the NPY gene with a history of suicidal behaviour. P268 A NEURONAL PAS GENE, NPAS3, IS A CANDIDATE FOR SCHIZOPHRENIA Georgina Macintyre1, Tyler Alford1, Lan Xiong2, Guy Rouleau2, Philip Tibbo3, Diane Cox4 1 University of Alberta, Edmonton 2 CHUM Research Center, Montreal, Quebec 3 Psychiatry, University of Alberta, Edmonton, Alberta 4 Medical Genetics, University of Alberta, Edmonton, Alberta NPAS3, a gene on human chromosome 14, was previously identified as a susceptibility gene for schizophrenia in our laboratory (Kamnasaran et al., Med Genet. 2003,40:325). An affected proband with schizophrenia and severe learning disability and heraffected mother are carriers of a translocation between chromosomes 9 and 14 that disrupts the NPAS3 gene. NPAS3 belongs to the basic helix-loop-helix Per-ArntSim (bHLH-PAS) transcription factor family, involved in diverse biological processes: regulation of cell differentiation and circadian rhythms, and the development and function of the nervous system. bHLH-PAS proteins have both DNA and protein binding domains. To test the hypothesis that NPAS3 is associated with schizophrenia, we looked for mutations in NPAS3 in patients with schizophrenia (DSMIII/DSMIV criteria) who had an affected parent. DNA samples from 47 patients and 54 ethnically matched, healthy controls were analysed. The twelve individual exons of NPAS3 were sequenced to identify potential disease-associated mutations in fourteen patients and two controls. Synonymous and non-synonymous changes found only in the patient group were investigated in the remaining samples. Two potentially pathogenic changes were associated with disease. These could be disease-causing, as the first may disrupt an important functional domain (p ¼ 0.01). The second mutation, present only in the patient group, may disrupt RNA splicing. Alternatively, the changes may be in linkage disequilibrium with a pathogenic change elsewhere in the gene. These changes could affect NPAS3 function and contribute to schizophrenia in some of our patients. P269 IS SEROTONIN INVOLVED IN WEIGHT GAIN AND METABOLIC SYNDROMEIN PSYCHOSIS? Valeria Mondelli1, Paola Dazzan2, Marta Di Forti2, Monica Aas2, Alex Butt2, Joanna Eyeson2, Ana Miorelli2, Serena Navari2, Pietro Papili2, Katherine J. Aitichison3, Logo Curtis3, Peter McGuffin3, David Collier3, Robin M. Murray2, Carmine M. Pariante2 1 Institute of Psychiatry, London 2 Psychological Medicine, Institute of Psy, London 3 MRC SGDP Centre, Institute of Psychiatry, London Background: Serotonin is known to regulate appetite and eating behaviour. The long allele of serotonin transporter promoter linked to polymorphic region 5-HTTLPR seems to contribute to the susceptibility to binge eating disorder. 5HT2A receptor (HTR2A) gene has been suggested to be associated with disinhibition in eating behaviour. With the introduction of atypical antipsychotics, new concerns have emerged about weight gain and metabolic syndrome in psychosis.
Abstracts Our aim is to clarify the effectof the 5-HTTLPR and HTR2A polymorphisms on weight and metabolic parameters in a sample of psychotic patients. Methods: We measured weight and height in 94 psychotic patients recruited as part of the large Genetic And Psychosis (GAP) study, carriedout in South London. In 30 of these patients, we also collected blood samples to assess metabolic parameters (HbA1c and lipid profile). DNA was extracted from patients’ peripheral blood. The 5-HTTLPR and HTR2A, including SNPs within these regions, will be genotyped by techniques developed in the SGDP laboratories. Results: In our sample, mean weight was 78 19 kg (mean SD) and mean body mass index was 26 6 kg/m2. 61 patients were receiving atypical antipsychotics, 7 typical antipsychotics,1 both atypical and typical, 11 were antipsychotic-free and in 14 patients medication was not specified. This will be the first study examining the effect of 5HTTLPR and HTR2A polymorphisms on weight and metabolic parameters in psychotic patients. Acknowledgements: This study was funded by South London & Maudsley NHS Trust. P270 CAN GENES PREDICT VOLUMETRIC BRAIN DIFFERENCES IN SCHIZOPHRENIA? Ana Miorelli1, Marta DiForti2, Valeria Mondelli3, Joanna Eyeson2, Kevin Morgan2, Julia Lappin2, Craig Morgan2, Alexander Butt2, Damian Cliford2, Pietro Papili2, Serena Navari2, Sara Campos do Sousa3, Bernard Freeman3, Monica Aas3, Paul Faeron2, Carmine Periante3, David Collier3, Robin Murray2, Paola Dazzan2 1 Institute of Psychiatry, London 2 Division of Psychological Medicine, Inst, London 3 Clinical Neuropharmacology and MRC SGDP, London Background: Structural brain volume changes are reported in psychosis. Twin studies indicate that brain volume is highly heritable; furthermore there is genetic correlation with schizophrenia (Rijsdijk, 2005). This indicates that schizophrenia susceptibility genes may also affect brain volume. Thus, genes such as Disrupted in Schizophrenia 1 appear to contribute to genetic risk for schizophrenia through disruptive structural changes(Cannon, 2005). Microcephalin and abnormal spindle-like microcephaly associated gene are associated with cerebral cortex size in humans(Evans, 2003). These genes are not studied in populations with psychosis. Hypotheses: Reduced brain volume in schizophrenia is related to genetic variation in susceptibility genes and genes that control brain size. Method: We recruited 99 first episode psychotic patients and 94 controls. We evaluated brain structure using MRI, and collected DNA. The images were acquired with a GESigna 1.5-T system and processed with Sbamm voxel-based semi-automated method. DNA was extracted from patient’s blood and SNPs genotyped using the ABI taqman system. We analysed SNPs in genes related to microcephaly and schizophrenia risk for association with brain volume. Results: Patients showed significantly smaller whole brain volume than controls (p ¼ 0.013), grey and white matter volumes being significantly smaller (p ¼ 0.01 and p ¼ 0.006 respectively). Conclusion: There are volumetric brain reductions in psychosis. It is possible that risk genes for psychosis are associated with genes regulating brain volume. Acknowledgments: Study funded by UK Medical Research Council, Stanley Medical Research Institute and South London and Maudsley NHS Trust. P271 MRI PARAMETERS OF FRONTAL LOBES IN SCHIZOPHRENIA PATIENT’S FAMILIES: ASSOCIATIONS WITH SCHIZOPHRENIA ENDOPHENOTYPE Nataly Efanova1, Vera Orlova1, Nina Voscresenskaya1, Leonid Gubsky2, Dmitry Kupriyanov2, Nikolay Anisimov2, Tatiana Solokhina1, Rostislav Nikolaev2 1 Mental Health Research Centre, RAMS, Moscow 2 Centre of Magnetic Resonance Tomography, Moscow 2 Psychiatric Hospital, Moscow Existing data suggest the involvement of frontal lobes in the pathogenesis of schizophrenia. The aim of this study was to analyse MRI parameters of frontal lobe volumes and their correlations with
799
liability to schizophrenia. The families of 28 schizophrenia patients were examined: 28 probands, 24 theirparents and 26 healthy controls, a total of 152 subjects. 3 mm Coronal T1-weighted 3D magnetic resonance images were acquired on a 0.5 Tesla magnet Tomikon S50, Bruker (Germany). The volumes of frontal lobes were calculated as total parameter. The frontal lobe characteristics were assessed by biserial coefficients of correlations with liability to schizophrenia (BCCLS) on the basis of standard statistical data in the groups of patients and controls as well as schizophrenia rate population. Nonsignificant reduction of the total frontal lobes volume was found in patients and their relations as compared with controls. BCCLS for total frontal lobe volumes was 0.2 0.01, indicating a significant slight relation with liability to the disease. Results hypothesise that frontal lobes peculiarities may be associated with schizophrenia endophenotype. The work was supported by Russian Foundation for Basic Research and Royal Society grants.
P272 MEDIAL TEMPORAL BRAIN STRUCTURES IN SCHIZOPHRENIA PATIENTS FAMILIES: MRI PECULIARITIES, CORRELATIONS WITH PSYCHOLOGICALCHARACTERISTICS OF NEGATIVE DISORDERS Natalia Purvenis1, T.D. Savina2, V.A. Orlova2, N.I. Voskresenskaja2, L.V. Gubskij3, D. Kupriyanov3, N.V. Anisimov3 1 Russian State Medical University, Moscow 2 Mental Health Research Centre, RAMS, Moscow 3 Moscow State University Magnetic Resonan, Moscow Existing data suggesting involvement of medial temporal brain structures in mechanisms of schizophrenia development. Aim: Estimation of amygdala, hippocampus, lateral ventricles (LV) and their temporal horns (TH)volumes in schizophrenic patients and relatives;and their correlation with psychological characteristics of emotional and cognitive deficit in patients. Methods: 85 subjects were examined by MRI: 35 schizophrenic patients, 27 unaffected relatives and 23 controls. 13 psychological characteristics (Polyakov‘s school methods) which previously demonstrated interconnection with negative symptoms of schizophrenia were assessed. Results: The data showed reduction of left amygdala, widening of both LV and left TH (significant), right amygdala and both hippocampus (tendency) in patients and relatives. In patients degree of amygdala reduction correlated with degree of disturbance precision of identifying emotions (PIE) using nonverbal expression (r ¼ 0.56, p < 0.05), degree of both hippocampus reduction with degree of alteration of attention switching, attention under conditions of continuous concentration, thinking, productivity of voluntary retention by reproduction data(PVRRD), mental activity, rigidity (r 0.5 – 0.7, p < 0.01 – 0.1). The volume of both LV negatively correlated with PVRRD, PIE with introduction verbal channel and tendency to actualize unusual object properties (TAUOP) and positively with frequency of using nonemotional interpretation (r 0.4 – 0.5, p < 0.05 – 0.1). The volume of left TH was positively correlated with TAUOP (r ¼ 0.6, p < 0.05). Conclusion: The data confirm the role of medial temporal structures both in predisposition to schizophrenia and disease process.
P273 G308A TUMOR NECROSIS FACTOR ALPHA FUNCTIONAL POLYMORPHISM AND SCHIZOPHRENIA RISK: META-ANALYSIS PLUS ASSOCIATION STUDY Massimo Gennarelli1, Luisella Bocchio-Chiavetto2, Paolo Valsecchi3, Catia Scassellati2, Patrizio Pasqualetti2, Cristian Bonvicini2, Paola Corsini3, Giuseppe Rossi4, Bruno Mario Cesana5, Sergio Barlati5, Emilio Sacchetti6 1 IRCCS Centro San Giovanni di Dio FbF, Brescia 2 Genetic Unit Center I.R.C.C.S. ‘‘San Giov’’, Brescia 3 Department of Mental Health, Spedali Civ, Brescia 4 Psychiatric Rehabilitation Center I.R.C., Brescia 5 Department of Biomedical Sciences and Bi, Brescia 6 University Psychiatry Unit, University a, Brescia Research on the G308A functional polymorphism in the tumor necrosis factor alpha (TNFa) gene as a susceptibility factor for
800
Abstracts
schizophrenia has provided contrasting results in different populations. Therefore we conducted a meta-analysis of the published casecontrol association studies and a replication study in a large sample. The meta-analysis showed that only the AA genotype was weakly associated with schizophrenia susceptibility (Z ¼ 2.16, p ¼ 0.03, OR ¼ 1.76, 95% CI 1.05–2.92) although in presence of high degree of heterogeneity among the studies, potentially as consequence of relevant confounders. The replication case-control association study (323 DSM-IV-TR schizophrenia patients and 346 controls) showed that the 308A allele conferred an increased susceptibility for schizophrenia only in males (OR ¼ 1.73, 95% CI 1.07–2.79, p ¼ 0.025), and the interaction became more specific when only patients of the paranoid subtype were compared to the controls (relative risk ratio ¼ 3.09, 95%CI 1.28–7.47, p ¼ 0.012). The presence of 308A allele was also associated with a later age at onset of schizophrenia (F1,291 ¼ 7.094, p ¼ 0.008). Our results confirm that the TNFa 308A allele has a weak effect on schizophrenia susceptibility. A number of factors, such as ethnicity, gender, schizophrenia subtypes and age at onset of the disorder, influence the strength of the association.
cortex of schizophrenic patients relative to controls. To verify the involvement of GRIN3A, the human gene coding for the NR3A subunit, in the susceptibility to schizophrenia, we searched for polymorphisms of GRIN3A gene and performed a case-control association study. A cohort of 230 unrelated patients diagnosed as having schizophrenia according to DSM-IV, were included in the study and compared with 256 normal controls. The case-control sample was genotyped for rs1337676 and rs1983812 SNPs, located in the 50 UTR and in 30 UTR of the gene, respectively. Pearson’s exact test did not reveal any difference in allele and genotype frequencies between patients and controls for rs1337676 (alleles: chi squared ¼ 0.89, p ¼ 0.37; genotypes: chi squared ¼ 1.84, p ¼ 0.40) and rs1983812 (chi squared ¼ 0.0441, p ¼ 0.85; chi squared ¼ 1.67, p ¼ 0.43), thus allowing to accept the null hypothesis of no association. This is the first study investigating the association between GRIN3A variations and schizophrenia. Given the relatively high relative risk (OR > 2) that the present analysis could detect, further studies are needed to evaluate a minor role for the GRIN3A gene in the development of schizophrenia.
P274 ASSOCIATION STUDY BETWEEN SCHIZOPHRENIA AND THE AMPA GENES GRIA1, GRIA2 AND GRIA4 Chiara Magri1, Rita Gardella2, Laura Imperadori2, Damiano Podavini2, Iatropoulos Paraskevas2, Stefano Davide Barlati3, Paolo Valsecchi3, Cristian Bonvicini4, Massimo Gennarelli4, Emilio Sacchetti3, Sergio Barlati2 1 University of Brescia, Brescia 2 Department of Biomedical Sciences and Bi, Brescia 3 Department of Mental Health, Brescia Spe, Brescia 4 Genetic Unit, I.R.C.C.S. ‘‘San Giovanni d, Brescia
P276 ASSOCIATION OF COMT GENE WITH OUTCOME MEASURES IN A COHORT OF ITALIAN SCHIZOPHRENIC PATIENTS Mirella Ruggeri1, Sarah Tosato2, Chiara Bonetto2, Mariaelena Bertani2, Giovanna Marrella2, Antonio Lasalvia2, Doriana Cristofalo2, Lorenza Lazzarotto2, Michele Tansella2, Marzia De Gironcoli3, Paola Dazzan4, Marta Di Forti4, Robin Murray4, David Collier5 1 University of Verona, Verona 2 Section of Psychiatry, University of Verona 3 Blood Transfusion Service, Azienda Osped, Verona 4 Institute of Psychiatry, Division of Psy, London 5 Institute of Psychiatry, Sgdp Centre, ki, london
We report the results obtained from association analyses of the GRIA1, GRIA2 and GRIA4 genes, which encode for three subunits of the AMPA glutamate receptors. All these genes can be considered good candidates for susceptibility to schizophrenia, due to the finding of an altered expression in different brain areas of schizophrenia patients. We analysed 8, 5 and 3 polymorphisms of the GRIA1, GRIA2 and GRIA4 genes, respectively, in an Italian case-control sample. For GRIA1 and GRIA2, SNPs were selected on the basis of LD structure, whereas for GRIA4 only those previously reported in association with the disease were typed. The ‘‘single locus’’ data evidenced association with two SNPs (rs707176 and rs2963944) and a microsatellite (rs10631988) in the GRIA1 gene. This association was further confirmed by the haplotype analysis, which showed an increased frequency of the haplotype C9CC (rs1422884-rs10631988-rs707176-rs2963944) in schizophrenia patients (p ¼ 0.009). When the whole patient group was considered, no positive results emerged for GRIA2 and GRIA4. However, a logistic regression analysis on GRIA4 highlighted an increased frequency of the rs641574 AA genotype in the non-paranoid subtype group in respect to the control group (p ¼ 0.022). Our data indicate that GRIA1 is involved in the pathogenesis of schizophrenia, whereas GRIA2 does not seem to contribute to the disease. Concerning GRIA4, we did not replicate the positive association reported in the Japanese population. However, our analysis suggests that GRIA4 might be a possible risk factor, at least in non-paranoid schizophrenia subtypes. P275 THE NMDA RECEPTOR REGULATORY SUBUNIT NR3A (GENE GRIN3A) AS SUSCEPTIBILITY FACTOR FOR SCHIZOPHRENIA: PRELIMINARY RESULTS FROM A CASE-CONTROL STUDY Rossella Gulli1, Bianca Masnata1, Marco Mollica2, Stefano Bignotti3, Cristian Bonvicini4, Massimo Gennarelli4, Emilia Bellone1, Paola Mandich1 1 Dept. of Neuroscience, Ophthalmology and, Genova 2 Dept. of Mental Health, ASL3, Genova 3 Psychiatric Rehabilitation Unit, IRCCS S, Brescia 4 Genetics Unit, IRCCS San Giovanni di Dio, Brescia Growing evidence suggests that NMDA receptor (NMDAR) dysfunction may be involved in the pathophysiology of schizophrenia. Functional NMDARs are composed by NR1, NR2 (NR2A-D) and NR3 subunits (NR3A-B). The NR3A subunit forms, with NR1 and NR2A or NR2B, NMDARs with decreased activity. Expression studies demonstrated that NR3A is overexpressed in prefrontal
Recent studies have identified the gene coding for the catechol-Omethyl transferase (COMT) enzyme as a candidate gene for schizophrenia. A functional polymorphism at codon 158, encoded by the SNP rs4680 results in the substitution of methionine (Met) for Valine (Val) which leads to a thermolabile protein. The COMT Val158Met alleles influence the availability of dopamine in the central nervous system, especially in the frontal cortex, and have been associated with cognitive function and interaction with cannabis use on the development of psychosis. Other polymorphisms in the gene, particularly the SNPs rs737865 and rs165599, have been associated with the risk to develop schizophrenia. To evaluate the relevance of this candidate gene to schizophrenia in a clinical population, we analysed a prevalence cohort of schizophrenic patients (n ¼ 141) in charge at the South Verona Community Mental Health Service then followed up at 3 and 6 years. First, we performed a case-control study to test if the rs4680, rs737865 and rs165599 polymorphisms of COMT are associated with schizophrenia. Then we tested if clinical symptoms are cross-sectionally associated with the presence of COMT polymorphisms. Finally, using the longitudinal clinical data, we tested if COMT polymorphisms are associated with a less favourable clinical outcome. The patients’ clinical status was assessed at baseline using BPRS and GAF and then re-assessed at follow-up. Eighty patients from this cohort and 106 healthy controls were genotyped for rs4680, rs737865 and rs165599. P278 CROSS-POPULATION VALIDATION OF SCHIZOPHRENIA-ASSOCIATION AND FUNCTIONAL ANALYSIS ON mRNA EXPRESSION AND ELECTROPHYSIOLOGICAL PROPERTIES OF TWO ISOFORMS OF GABRB2 Wing Sze Lo1, Cunyou Zhao2, ZhiWen Xu2, Jianhuan Chen2, Zhiliang Yu2, Frank W. Pun2, Ka-Lok Tong2, Siu-Kin Ng2, Shui-Ying Tsang2, Mutsuo Harano3, Micha Gawlik4, Naohisa Uchimura3, Gerald Stober4, Hong Xue2 1 HKUST, n/a 2 Hong Kong University of Science and Tech, n/a 3 Kurume University School of Medicine, Fukuoka 4 University of Wuerzburg, n/a SNPs and haplotypes in GABAA receptor b2 gene (GABRB2) were reported in association with schizophrenia initially in Chinese
Abstracts population, and later with Chinese, German and Portuguese trios, although disagreement exists. The present study aimed to validate the association with Japanese (SZ/CN 304/207) and German (SZ/CN 301/ 190) populations, and to examine the mRNA expression and electrophysiological profiles of two isoforms of GABRB2. Seventeen SNPs within a 1839 bp were analyzed for disease-association with Bonferroni-correction, permutation and random resampling. Association was demonstrated in both Japanese (P ¼ 0.0002–0.0191) and German (P ¼ 0.0033–0.0410), especially for haplotypes containing rs1816072 and rs1816071. Among the Leonhard’s clinical subtypes, the strongest association was exhibited by systematic schizophrenia, the most severe schizophrenia subtype. Real-time quantitative PCR performed with mRNA extracted from postmortem brain of US Caucasian schizophrenics (Male/Female 23/8) and controls (Male/Female 24/7), showed reductions in schizophrenics the expression of long by 21.7% and short by 13.4% isoforms. Significant decrease of long/short ratio was observed in patient (P ¼ 0.023–0.009). Genotypical effects on expression were established for rs1876071 and rs1816072 in male. Electrophysiological profiles of recombinant GABAA receptors containing different isoforms were significantly different. Cross-population validation of disease association along with the genotype-dependent alteration in expression and differential electrophysiological properties of the alternativelyspliced isoforms suggest that potential functional consequences of DNA polymorphisms in GABRB2 may play role in the pathogenesis of schizophrenia.
P279 DOES COMT POLYMORPHISM INFLUENCE COGNITIVE FUNCTION IN FIRST EPISODE PSYCHOSIS? Monica Aas1, Joanna Eyeson2, Valeria Mondelli2, Alex Butt2, Marta Di Forti2, Ana Miorelli2, Serena Navari2, Pietro Papili2, Kevin Morgan2, Julanta Zanelli2, Craig Morgan2, Paul Fearon2, Katherine J. Aitchison3, Peter McGuffin3, David Collier3, John Powell2, Robin Murray2, Paola Dazzan2, Carmine M. Pariante2 1 Institute of psychiatry,KCL, London 2 Psychological Medicine, Institute of Psy, London 3 MRC SGDP Centre, Institute of Psychiatry, London Background: The metabolism of dopamine in the prefrontal cortex is catalysed predominantly by Catechol-O-methylstransferanse (COMT). The coding region of the COMT gene includes a single nucleotide polymorphism leading to Val to Met substitution at amino acid codon 158. The Val allele polymorphism is associated with an increased dopamine catabolism. COMT genotype (Val/Val) is associated with cognitive reduction in the executive functioning and working memory in patients with schizophrenia. Previous studies on cognitive function and COMT polymorphism have focused on chronic psychotics, this will be one of the first studies assessing patients with first-episode psychosis. Method: We recruited 49 patients (67% males) with first-episode psychosis as part of the Genetic and Psychosis (GAP) study and the Aetiology and Ethnicity of Schizophrenia and Other Psychosis (AESOP) study carried out in South London. All patients underwent neuropsychological assessment including measures of executive function by Trails A and B, Letter Number Span and Verbal memory. We collected blood samples from all the patients for DNA. COMT polymorphism will be genotyped using standard methodologies. Results: In our sample the mean age was 27 7, 6 yrs(mean SD). The Premorbid IQ of the sample was 96 17 SD. Conclusion: This study will generate valuable data examining the association between Val allele polymorphism and reduction in cognitive function in first episode psychosis. Acknowledgement: The GAP study was funded by the South London & Maudsley NHS Trust. The AESOP study was funded by the Medical Research Council.
P280 CORRELATIONS BETWEEN MRI PARAMETERS OF SUBCORTICO-FRONTAL BRAIN STRUCTURES AND PSYCHOLOGICAL CHARACTERISTICS OF EMOTIONALITY IN SCHIZOPHRENIA PATIENTS FAMILIES Tatiana Savina1, N. Efanova1, V. Orlova1, N. Voscresenskaya1, L. Gubsky2, D. Kupriyanov2, N. Anisimov2, E. Fedorova1 1 Mental Health Research Centre, RAMS, Moscow 2 Moscow State University Magnetic Resonan, Moscow
801
Estimation of relationship between morphological and neuropsychological traits of schizophrenia endophenotype can be perspective for identification of pathological genes carriers. In order to study correlations of subcortico-frontal brain peculiarities and some emotionality characteristics 22 patients with schizophrenia and 20 their unaffected relatives were examined. 3 mm n˜oronal T1-weighted 3D magnetic resonance images were acquired on a 0.5 Tesla magnet Tomikon S50, Bruker (Germany). Volumes of caudate, nucleus lentiformis and frontal lobes (total) were calculated. Psychological methods of Polyakov‘s school were used. Evaluation of precision of identifying emotions using non-verbal expression was done. 2 estimates were analyzed: frequency of using non-emotional interpretation and precision of identifying emotions in indeterminate situation with information deficit. Results demonstrated correlations of booth caudate and booth nucleus entiformis volumes with values of frequency of using non-emotional interpretation (r from 0.3 to 0.4, p < 0.05) in patients and total frontal lobes volume with value of disturbance of emotions identifying precision in indeterminate situation (r ¼ 0.3, p < 0.05) in their relatives. The data confirm the role of subcortico-frontal brain regions peculiarities in forming of emotionality characteristics in schizophrenia patientsfamilies. They also indicate to differences in mechanisms of emotionality deficit symptoms expression in patients with schizophrenia and of emotionality diathesis in their unaffected relatives. P281 ALLELE SHARING ANALYSIS OF MAO-A GENE IN MEXICAN SCHIZOPHRENIC PATIENTS. Beatriz Camarena1, Alejandro Aguilar2, Rogelio Apiquian3, Ana Fresan2, Rolando Medina2, Henriette Raventos3, J. Can˜ive3, R. Mun˜oz3, R Mendoza4, A. Jerez3, A. Ontiveros3, Michael Escamilla3, Humberto Nicolini3 1 Instituto Nacional de Psiquiatrı´a Ramo´n, Me´xico D. F. 2 Instituto Nacional de Psiquiatria Ramon, Mexico city 3 Mexico city Autonomous University, Mexico city 3 The Genetics of Schizophrenia in Latino, San Antonio, TX 4 The Genetics os Schizophrenia in Latino, San Antonio, TX Schizophrenia is a complex disorder with a genetic contribution. Several studies report gender differences in the age of onset, severity and course of illness. Moreover, a high frequency of X chromosome anomalies among patients with psychosis has been reported. In addition, linkage and association studies have suggested that a locus predisposing to schizophrenia may reside within Xp11. The MAO-A gene are located at this region. Therefore, we decided to analyzed EcoRV and uVNTR polymorphisms in affected sib pairs with schizophrenia using allele- sharing identical-by-state (IBS) analysis. We genotyped 60 affected sib pairs with DSM-IV diagnosis of schizophrenia. Subjects were originally recruited for a family based linkage study ofschizophrenia and schizoaffective disorder. A best estimate consensus process was used to assign final diagnoses. The uVNTR and EcoRV analysis were performed using the primers and conditions described by Sabol et al. (1998), and Hotamisligil and Breakefield (1991), respectively. Analysis by gender of uVNTR polymorphism showed in affected female-female sib pairs a higher number of sib pairs sharing 2 alleles IBS (Chi square ¼ 6.9, df ¼ 2, p ¼ 0.031). Also, EcoRV analysis showed in the female- femalesib pairs a higher number of sib pairs sharing 2 alleles IBS (Chi square ¼ 11.5, df ¼ 2, p ¼ 0.0038). The preliminary findings confirmed the hypothesis that a gene in Xp11 region might confer vulnerability to schizophrenia and support the evidence of a gender effect of MAO-A gene in susceptibility to develop this disorder. P282 THE DYSTROBREVIN BINDING PROTEIN 1 (DTNBP1) GENE IS ASSOCIATED WITH SCHIZOPHRENIA IN THE IRISH CASE-CONTROL STUDY OF SCHIZOPHRENIA (ICCSS) SAMPLE Po-Hsiu Kuo1, Wormley Wormley1, F. Anthony O’Neill2, Dermot Walsh3, Kendler Kenneth S.1, Brien Riley1 1 VIPBG, Richmond, VA, 2 Queens University, Belfast 3 Health Research Board, Dublin DTNBP1 is widely associated with schizophrenia, in ten of fifteen independent samples. Three specific haplotypes are repeatedly associated. We recently collected the large ICCSS sample, currently 1029 cases and 627 controls. Our first goal was to replicate the DTNBP1
802
Abstracts
association. We genotyped 10 SNPs, 9 common to most studies of DTNBP1 and one defining an associated haplotype in another Irish case-control sample. The data were consistent with previous studies from this population in showing reduced LD between rs2005976 (p1757) and rs3213207 (p1635). Of these 10 markers, the strongest association was seen with the common allele of rs760761 (p1320; p ¼ 1.79e-05). Marginal association was observed with the major alleles of rs1474605(p1792; p ¼ 0.032) and rs3213207 (p ¼ 0.041), and with the minor allele of rs2619538 (SNPA; p ¼ 0.045). Haplotype analyses showed association (p ¼ 0.0002) with the haplotype defined by the rare allele of rs2619538 and the common alleles of rs1474605, rs1018381 (p1578), rs2619522 (p1763), rs760761 and rs2005976. These results are in keeping with previous reports from German and Chinese samples of association with the most common observed haplotype. We observe two major differences compared to theresults in the most comparable Irish case-control sample. The rare allele of rs2619538 (SNPA) is present on our associated haplotype. Also, rs2619539 (p1655) makes no contribution to the observed association either as a single marker or as partof a haplotype. These data add further complexity to the world-wide results for DTNBP1 in schizophrenia. P283 FAMILIAL VS SPORADIC PSYCHOSIS: THE ESPIGAS STUDY Antı´a Bran˜as1, Miguel Ruiz-Veguilla2, Marı´a Luisa Barrigo´n1, Manuel Anguita1, Estı´baliz Gordo1, M Ferrı´n2, Blanca Gutie´rrez3, Jorge Cervilla3 1 Hospital Clı´nico San Cecilio de Granada, Granata 2 Hospital Neurotraumatolo´gico de Jae´n, Jae´n 3 Instituto de Neurociencias, Granada The ‘ESPIGAS’ study set out to investigate whether neurodevelopmental markers are associated with early onset psychosis or with a family history of psychosis. To date we have included 76 consecutively identified patients with any functional psychotic disorder who attended community services in Granada. Patients were thoroughly assessed by fully trained psychiatrists using a psychopathological scale (PANSS), a standardized neurological examination (NES), a premorbid social-adjustment scale(Cannon-Spoor), a measure of obstetric complications and milestones (Lewis-Murray), family history of mental disorder, and socio-demographical data. Patients were assigned to the ‘familial psychosis’ group should they have a family history of functional psychosis among at least one first or second degree relative. The remaining patients were assigned to the ‘‘sporadic psychosis’’ group. Mean age is 26.2 years (20% younger than 18), 67% are men and 65% (n ¼ 49) had a diagnosis of schizophrenia. 31% (n ¼ 24) patients were included in the ‘family psychosis’ group. Patients with schizophrenia have a higher proportion of family history of psychosis compared with patients suffering from other psychoses (40% vs. 11,5%, respectively). Univariate analysis showed more language and sphincter control delay among the ‘familiar psychosis’ group than among ‘sporadic psychosis’( 42% vs. 16%; p < 0.05). Neurologic soft signs (complex motor acts subscale) were also found to be significantly more prominent among familial cases (Mean 4 vs. 6; p < 0.05). Only milestone acquisition delay remained significant after adjusting for gender, age and obstetric complications (OR ¼ 4,1 p < 0.05).
P284 SYSTEMATIC SNP POLYMORPHISM SCREENING FOR SCHIZOPHRENIA CANDIDATE GENE OF CALCIUM CHANNEL, VOLTAGE-DEPENDENT, GAMMA SUBUNIT (CACNG2) Hai-Gwo Hwu1, Chih-Min Liu2, Cathy Shen-Jang Fann3, Shih-Feng Tsai4, Ueng-Cheng Yang5, Pei-Chun Hsu5, Yu-Li Liu6 1 National Taiwan Univesity, Taipei 2 National Taiwan University, Taipei 3 Insititute of Biomedical Sciences, Acade, Taipei 4 Division of Molecular and Genomic Medici, Taipei 5 Institute of Bioinformatics, National Ya, Taipei 6 Division of Mental Health and Substance, Taipei The calcium channel, voltage-dependent, gamma subunit (CACNG2) gene located at 22q12 is the gamma subunit of neuronal voltageactivated L-type calcium channels which may stabilize the calcium channel in an inactive state. The CACNG2 protein is similar to the
mouse stargazing protein, which is implicated in AMPA (alpha-amino3-hydroxy-5-methyl-4-isoxazolepropionic acid) receptor trafficking. AMPA receptor mediates fast excitatory synaptic transmission in brain and underlies aspects of synaptic plasticity. To search for novel SNPs existed in CACNG2 gene, we have established bioinformatics models by using several tools to define genetic exon regions and to determine the promoter regions on genomic sequences. Primers were designed for each of the exon and the promoter region. A total of 5 exons and 8 promoter regions for CACNG2 were sequenced on 50 normal controls, 50 simplex (one probands in a schizophrenia family) individuals, and 50 multiplex (at least two probands in a schizophrenia family) individuals which represent more genetic loadings in these cases. We found 48 SNPs within these screened regions, and 33 of these SNP have not been reported by the National Center for Biotechnology Information (NCBI). The differences of the minor allele frequencies among the controls, the simplex and the multiplex groups varied from 0.02% to 18%. In summary, we have found novel SNPs on CACNG2 existed in the ethnic group of Taiwan.
P285 ASSOCIATION STUDY OF THE T102C 5-HT2A POLYMORPHISM IN SCHIZOPHRENIA: DIAGNOSIS, PSYCHOPATHOLOGY ANS SUICIDAL BEHAVIOR Humberto Correa1, Rodrigo Nicolato2, Antoˆnio Teixeira1, Luiz Armando De Marco1, Marco Aure´lio Romano-Silva1 1 UFMG, Belo Horizonte-Minas Gerais 2 UNIFENAS, Belo Horizonte Background: The 5-HT2A gene polymorphism (102T/C) has been previously associated with schizophrenia diagnosis although it may play a major role in suicidal behavior. The goal of this study was to examine the association between the 5-HT2A gene polymorphism (102T/C) and suicide behavior in schizophrenic inpatients. Methods: We studied 214 subjects. Patients (n ¼ 129) met diagnostic criteria for schizophrenia according to a structured clinical interview (MINI-PLUS). Patients underwent a semi- structured interview to assess suicide attempt history and its characteristics. In addition, at least one close relative of the patient was interviewed to assess proband and family suicidal behavior. Healthy controls were students and hospital staff members (n ¼ 85), all free of psychiatric and medical illness.Genotypic frequencies were obtained after DNA extraction, the region of 5-HT2A/T102C containing the polymorphic site amplified by the polymerase chain reaction and digested with the restriction enzyme HpaII. Results: We found no association between suicidal attempt history and suicide attempt characteristics with genotypic or allele frequencies. Moreover, suicide behavior was not associated with demographic or psychopathological characteristics. Conclusion: These results suggest that the 5HT2A gene polymorphism (102T/C) is not involved in the genetic susceptibility to suicidal behavior but further studies with a large sample is needed.
P286 POSTSYNAPTIC DENSITY-95 GENE MUTATIONS AND SCHIZOPHRENIA Chia-Hsiang Chen1, Min-Chih Cheng2, Chen-Hao Li3, Hsiao-Mei Kiao2, Shih-Fen Chen2, Ho-Min Tsai2 1 Tzu-Chi General Hospital, Hualien 2 Institute of Human Genetics, Tzu-Chi Uni, Hualien 3 Institute of Human Genetics, Tzu-CHi Uni, Hualien Synaptic glutamate dysfunction is suggested to be associated with the pathophysiology of schizophrenia. Postsynaptic density-95 (PSD-95) protein is a postsynaptic scaffold protein that plays an important role for the structural and functional integrity of NMDA receptor, one of the glutamate receptor families. Several postmortem study have demonstrated decreased expression of PSD-95 in the frontal cortex, thalamus and hippocampus in patients with schizophrenia and mood disorders, suggesting PSD-95 may be involved in the pathogenesis of certain neuropsychiatric disorders. Hence, PSD-95 is considered as a candidate gene for schizophrenia in this study. To test this hypothesis, we systematically searched for mutations in the promoter region and all the exons of PSD-95 in a sample of Chinese patients from Taiwan using PCR-based autosequencing. Although no variants in the protein-coding sequences of PSD-95 gene were identified, we identified one single
Abstracts nucleotide polymorphism (SNP) at the core promoter region (g.1336C > T) and two SNPs at the 50 untranslated region of PSD-95 (g.456C > T and g.209C > T). Reporter assay showed that the T allele of the g.1336C > T SNP has 1.5 fold activity of T allele, suggesting this is a functional SNP. Further case-control association study between 323 patients and 334 control subjects, however, did not find association of this SNP with schizophrenia. Nevertheless, further association studies of this SNP with other neuropsychiatric disorders are warranted, because PSD-95 may also be involved in other neuropsychiatric diagnoses.
P287 A CRITICAL EVALUATION OF HIGH RESOLUTION DNA MELTING ANALYSIS USING A LIGHTSCANNERTM Sarah Dwyer1, Nigel M. Williams2, Nadine Norton2, Michael J. Owen2, Michael C. O’Donovan2 1 University Hospital of Wales, Cardiff 2 Cardiff University, Cardiff Current candidate gene re-sequencing methods are typically prohibitive, both in terms of labour and cost, for studies of large genes in large samples. As a consequence, candidate gene re-sequencing studies typically yield relatively little power to detect rare variants and are instead largely limited to the direct analysis of susceptibility alleles that are common. The LightScannerTM (Idaho Technologies) is a mutation detection instrument that is capable of performing high resolution DNA melting analysis to identify DNA variants in PCR products by detecting changes in the shape of the melting curve. We have critically assessed the sensitivity of the LightScannerTM by performing a blind analysis of 14 amplimeres in which we had previously characterised 23 SNPs in a sample of 14 unrelated individuals using a Wave1 DHPLC system and DNA sequencing. The amplimeres varied in size (75–557bp), %GC content (42–71%) and the number of SNPs (1 to 4). This study revealed that, depending on the analysis method used, there can be significant variation in the sensitivity of the LightScannerTM to detect DNA variants. However, using an optimal set of analysis conditions we obtained 100% concordance with the known sequence composition of all amplimeres. Our data suggests that the LightScannerTM is a highly sensitive mutation detection system, which because of its low cost (17p/sample) and high speed (>1500 samples/hour) could potentially allow more comprehensive re-sequencing analysis of candidate genes. P288 NO ASSOCIATION BETWEEN THE BRAIN-DERIVED NEUROTROPHIC FACTOR (BDNF) GENE POLYMORPHISM (Val66Met) AND TARDIVE DYSKINESIA Yui Naoe1, Takahiro Shinkai2, Hiroko Hori1, Shinichi Sakata1, Yuko Fukunaka1, Kensuke Utsunomiya2, Kazuko Shimizu1, Chima Matsumoto2, Osamu Ohmori3, Jun Nakamura1 1 Univ of Occupatl & Environmetl Health, Kitakyushu 2 Univ of Occupatl & Environmetl Health, S, Kitakyushu 2 Univ of Occupatl & Environmetl Health, K, Kitakyushu 2 Univ of Occupatl & Environmetl Health, N, Kitakyushu 3 Univ of Occupatl & Environmetl Health, W, Kitakyushu Brain-derived neurotrophic factor (BDNF) plays an important role in the development and maintenance of adult neurons and is an important regulator of synaptic plasticity in the human brain. A recent study showed that plasma BDNF levels were lower in tardive dyskinesia (TD) than non-TD patients and normal controls in a Chinese sample of schizophrenia patients (Schizophr Res 74, 263–270, 2005). TD patients who were heterozygous for the BDNF Val66Met genotypes had higher abnormal involuntary movement scale (AIMS) orofacial scores (NeuroMolecular Med 5, 243–251, 2004).In the present study, we examined the association between the BDNF Val66Met polymorphism and TD in a Japanese sample. Our sample included 208 patients with schizophrenia (DSM-{W). We evaluated TD using the AIMS score. Informed consent was a premise for participation, and this study was approved by the Ethics Committee of the University of Occupational and Environmental Health. Genotyping was assessed by the TaqMan allele-specific assay method using ABI PRISM1 7000 (Applied Biosystems). Differences in allele and genotype distribution between cases with and without TD were evaluated using the chi2 test. No significant association between the BDNF Val66Met polymorphism and TD was found (genotype: 2 ¼ 1.60, d.f. ¼ 2, p ¼ 0.449; allele frequency:
803
¼ 1.21, d.f. ¼ 1, p ¼ 0.27). Our results suggest that the BDNF Val66Met polymorphism may not confer susceptibility to TD in our sample. 2
P289 AN EXAMINATION OF MUTED AS A SCHIZOPHRENIA SUSCEPTIBILITY GENE Amy Gerrish1, Nigel Williams1, Hywel Williams1, Michael Owen1, Michael O’Donovan1 1 Cardiff University Hospital of Wales, Cardiff Keywords BLOC-1, MUTED, Schizophrenia Strong evidence implicates DTNBP1 as a susceptibility gene for chizophrenia, however its pathogenic role remains unclear. The identification of functionally related proteins that themselves confer usceptibility to schizophrenia would establish which of the dysbindin related pathways are involved in the pathogenesis of schizophrenia. Dysbindin is known to be a component of the biogenesis of lysosome related ganelles complex 1 (BLOC-1) whereby it interacts directly with, llidin, muted, cappuccino, SNAP-associated protein and blos-2. Evidence for association between MUTED (6p25.1-p24.3) and schizophrenia has been previously reported either with single markers or through an epistatic interaction with DTNBP1. We have performed mutation screening of all MUTED exons and 1kb of sequence immediately 50 and 30 of the gene. All identified polymorphisms were genotyped through the 30 trios of the HapMap CEU sample and combined with all SNPs of HapMap phase II that span the MUTED locus (total ¼ 90). A total of 19 Tag SNPs, which included 6 SNPs previously reported to be associated with schizophrenia, were genotyped though an association sample of 673 schizophrenic cases and 716 controls. We obtained no significant evidence for allelic association in any of the SNPs tested. In addition, logistic regression analysis using SNPs at MUTED and SNPs previously genotyped at the DTNBP1 locus in the same sample gave no significant evidence of an epistatic interaction. Our data provides no support that MUTED is a susceptibility gene for schizophrenia either as a single locus or through epistatic interaction with dysbindin. P290 GENETIC EVIDENCE FOR INVOLVEMENT OF NRG1/Erbb SIGNALLING IN OLIGODENDROCYTES IN SCHIZOPHRENIA Lyudmila Georgieva1, Jennifer Wilkinson2, Nadine Norton2, Takeshi Sakurai3, James Young3, Christopher Plescia3, Kenneth L. Davis3, Vahram Haroutunian4, Joseph D. Buxbaum3, Nick Craddock2, George Kirov2, Michael J. Owen2, Michael C. O’Donovan2 1 Universisty Hospital of Wales, Cardiff, 2 Cardiff University, Cardiff 3 Department of Psychiatry, Mount Sinai Sc, New York 4 Mount Sinai School of Medicine/MIRECC, B, New York Neuregulin (NRG1) is a schizophrenia susceptibility gene involved in multiple aspects of nervous system function, including neuronal and glial cell development, synapse formation, oligodendrocyte survival, and myelination. Recent studies including our own suggest the NRG1 receptor erbB4 also contributes to schizophrenia susceptibility suggesting that signalling pathways involving NRG1 and erbB4 might be affected in schizophrenia. In order to identify downstream signalling molecules of erbB4, we carried out a yeast two-hybrid screen for proteins interacting with the erbB4 cytoplasmic domain. We identified the MAGI family of scaffolding proteins. MAGIs have been previously shown to interact with the receptor phosphotyrosine phosphatase,RPTPbeta, a molecule which like NRG1 signalling, has also been implicated in oligodendrocyte development. Given their functions, we postulated that erbB4 and RPTPbeta might form a molecular complex that regulates neuregulin signalling. To examine whether such a complex might further contribute to schizophrenia susceptibility, we examined 28 SNPs spanning RPTPbeta for evidence for association with schizophrenia in a large UK case-control sample. Multiple associated markers were observed. Despite the large number of markers tested, the evidence for association remained significant at a gene wide level. Moreover, markers in RPTPbeta showed evidence for epistatic effects on disease risk in combination with markers in a gene for which we have previously reported strong evidence for association with schizophrenia that is a key regulator of oligodendrocyte development (OLIG2). These results suggest that alterations in NRG-erbB4-
804
Abstracts
RPTPbeta signalling influence risk of schizophrenia and do so through influencing oligodendrocyte development. P291 INTERACTION BETWEEN CHILDHOOD ADVERSITY AND 5-HTTLPR POLYMORPHISM IN PSYCHOSIS Joanna Eyeson1, Helen Fisher2, Craig Morgan2, Paola Dazzan2, Marta Di Forti2, Ana Miorelli2, Valeria Mondelli2, Alex Butt2, Monica Aas2, Serena Navari2, Pietro Papilli2, Logos Curtis3, Katherine J. Aitchison3, Carmine Pariante2, Peter McGuffin3, David Collier3, Paul Fearon2, Robin M. Murray3 1 Institute of Psychiatry, London 2 KCL, Institute of Psychiatry, London 3 MRC SGDP centre, KCL, Institute of Psych, London 3 KCL, Institute of Psychiatry, London Background: Adverse experiences in childhood are associated with increased risk of developing psychotic symptoms in adulthood. However, psychosis can develop in the absence of childhood adversity and exposure to such experiences does not always lead to psychosis. Individuals’ genotype may affect susceptibility to psychosis following exposure to adverse childhood experiences. The short allele of a functional polymorphism (5-HTTLPR) located within the promoter region of the gene encoding the serotonin transporter protein has been shown to interact with childhood adversity in the development of psychopathology. Many atypical antipsychotics have a high affinity for serotonin as well as dopamine receptors. Hence, genetic variants in the serotonin system, including the 5-HTTLPR may moderate children’s sensitivity to adversity and subsequently increase vulnerability to later psychosis. Method: Data were collected on 100 first-presentation psychosis cases within South London. Information relating to adverse childhood experiences (multiple family arrangements, institutionalisation, separation from or loss of a parent, physical and sexual abuse) occurring before age 17 was obtained using the Childhood Experiences of Care and Abuse Questionnaire (CECA.Q). Blood samples obtained from each patient were genotyped using fluorescent genotyping technology. Results: Analysis is expected to reveal that having at least one short allele of 5-HTTLPR is more common amongst those patients who have been exposed to adverse childhood experiences, thereby suggesting a possible gene-environment interaction. Acknowledgements: The GAP study is funded by the South London and Maudsley NHS Trust. The AESOP study is funded by the Medical Research Council, UK
P292 ASSOCIATION STUDY BETWEEN THE NMDA RECEPTOR SUBUNIT 2B GENE(GRIN2B) AND SCHIZOPHRENIA Hiroko Hori1, Takahiro Shinkai2, Kensuke Utsunomiya3, Shinichi Sakata4, Yui Naoe3, Yuko Fukunaka4, Kazuko Shimizu4, Chima Matsumoto5, Osamu Ohmori6, Jun Nakamura4 1 Univ of Occupatl & Environmetl Health, Kitakyushu 2 Univ of Occupatl & Environmetl Health, S, Kitakyushu 3 Univ of Occupatl & Environmetl Health, K, Kitakyushu 4 Univ of Occupatl & Environmetl Health, Kitakyushu 5 Univ of Occupatl & Environmetl Health, N, Kitakyushu 6 Univ of Occupatl & Environmetl Health, W, Kitakyushu N-methyl-D-separate (NMDA) receptor dysfunction may be involved in the pathophysiology of schizophrenia. The human gene coding for the 2B subunit of the NMDA receptor (GRIN2B) is considered a candidate gene for schizophrenia based on its selectiveexpression in the brain. To evaluate the hypothesis that GRIN2B acts as a major gene in determining susceptibility to schizophrenia, we performed a casecontrol association study. Our sample consists of 228 Japanese patients with schizophrenia and 188 Japanese normal control subjects. Informed consent was obtained from all subjects prior to participation and this study was approved by the Ethics Committee of the University of Occupational and Environmental Health. Although preliminary data for 5072T/G gave negative results (genotypic distribution; 2 ¼ 0.64, d.f. ¼ 2, p ¼ 0.73, allele frequency; 2 ¼ 0.00, d.f. ¼ 1, p ¼ 1.00), individual genotyping is currently being performed for T-200G as well as 366C/G. Details of association analyses for single markers as well as marker haplotypes will be presented.
P293 A GENETIC ASSOCIATION STUDY IMPLICATES THE CHROMOSOME 11Q23.3 GENE FXYD6 IN SUSCEPTIBILITY TO SCHIZOPHRENIA Khalid Choudhury1, Andrew McQuillin1, Vinay Puri1, Srinivasa Thirumalai2, Jacob Lawrence1, Robert Krasucki1, Susmita R. Datta1, Nicholas J. Bass1, Jonathan Pimm1, David Curtis2, Hugh M.D. Gurling1 1 University College London, London 2 West Berkshire NHS Trust, Reading 2 Queen Mary College, London Previous linkage analyses of multiply affected schizophrenic families have confirmed the involvement of the chromosome 11q23 region in the aetiology of schizophrenia with lod scores of 3.4 and 3.1 (Maziad et al., 1995; Gurling et al., 2001). We attempted to fine map and identify a schizophrenia susceptibility gene within the 11q23 region previously implicated by linkage analyses. Fine mapping was carried out by detecting linkage disequilibrium between genetic markers and schizophrenia in a case- control sample from the United Kingdom. Microsatellite and single nucleotide polymorphism markers were genotyped in 492 unrelated schizophrenia subjects and 443 supernormal controls. Five markers localised within or near the FXYD6 gene exhibited empirically significant allelic associations with schizophrenia (D11S1998, p ¼ 0.024; TTTC20.2, p ¼ 0.047; rs1815774, p ¼ 0.043; rs4938445, p ¼ 0.010; rs4938446, p ¼ 0.026). A test of haplotypic association with markers rs869789 and rs1815774 was significant at p ¼ 0.0001 with an empirical test of significance. FXYD6 may be involved in the molecular pathology of schizophrenia with base pair changes affecting the function, structure or expression of this gene being likely to increase genetic susceptibility to schizophrenia. Sequencing of cases with at risk haplotypes showed some potential aetiological base pair changes in FXYD6. Further results, including additional genotype SNP data and replication of positive SNP markers in a different case-control sample will be presented. Gurling et al (2001) Am J Hum Genet 68:661–73 Maziade et al (1995) Am J Med Genet 60: 522–528. P294 MicroRNA IN HUMAN BRAIN AND SCHIZOPHRENIA: TESTING VARIATION NEAR miR-130b Olga Burmistrova1, Andrey Goltsov2, Evgeny Rogaev3 1
Umass Medical School, Worcester, MA Mental Health Research Center, Moscow Brudnick Neuropsychiatry Research Instit, Worcester, MA
2 3
Introduction: The functional role of non-coding RNA in brain is unknown. We proposed to test the role of small regulatory RNA, specifically miRNAs, in the development of schizophrenia. Methods: We searched for DNA variations in miR-130b precursor gene and studied the frequency of the SNP in patients with diagnosis of paranoid schizophrenia—F20.0x according to ICD-10. The influence of this SNP on expression of miR-130b was studied. Results: MiR-130b was selected as miRNA with putative target in 30 region of methyl-CpG-binding protein 2 (MECP2) gene transcript. The gene for miR-130b is located within the region consistently reported as a susceptibility locus for schizophrenia (22q11). We made expression profile for more than 150 miRNAs in brain tissues of 15 schizophrenic and 15 control individuals. Single nucleotide variation was found in the 50 upstream region. Frequency of SNP was examined in 114 schizophrenic and 105 age- and gender matched control individuals of Russian origin. Expression of miR-130b in correlation with this SNP was studied. Conclusions: No statistically significant differences in the distribution of alleles (63 C/G) or genotypes (63 CC or GG) were found between the control and schizophrenia groups. The other polymorphisms have to be identified for more comprehensive haplotype analysis of miR-130b or other mirRNAs complementary to MECP2 or other neurogenes in schizophrenia. The data demonstrated no correlation between SNP and expression of miR-130b. References: This work was supported by Stanley Research Foundation, RFBR. P295 GENETIC VARIATION IN HT2A AND SUICIDAL IDEATION IN A SAMPLEOF 270 IRISH HIGH-DENSITY SCHIZOPHRENIA FAMILIES Ayman Fanous1, Xiangning Chen2, Xu Wang2, Richard Amdur1, F. Anthony O’Neill2, Dermot Walsh3, Kenneth Kendler2
Abstracts 1
Washington VA Medical Center, Washington, D.C. Virginia Commonwealth University, Richmond, VA Queen’s University, Belfast 3 Health Research Board, Dublin 2 2
Background: Genetic variation in the serotonin 2A receptor (HT2A) has been associated with both schizophrenia and suicidal behavior. Methods: Our sample comprised 270 Irish high-density schizophrenia families. Subjects who had at least one episode of psychosis were rated on the Operation Criteria Checklist for Psychotic Illness (OPCRIT). Lifetime history of suicidal ideation was determined from medical records and psychiatric interviews and scored in the OPCRIT. Twelve SNPs were selected for study. Ten of these were htSNPs derived from HapMap data, along with His452Tyr and T102C. We tested for association with psychotic illness as a whole, as well as stratified by the presence of suicidal ideation, using FBAT and PDTPHASE. Singlemarker as well as haplotype-based tests using a sliding window approach were performed. Results: We observed several 2, 3, and 4 marker haplotypes near the 30 end of the gene that were over-transmitted to psychotic subjects (.02. 04). His452Tyr was included in these haplotypes but was not itself significant. We also observed modest over-transmission of a 2-marker haplotype that included T102C (.04.08), which was also not itself significant in single-marker analyses. Therewas no significant association in the subgroup of the sample with suicidal ideation. Conclusions: Thse results provide modest support for HT2A as a susceptibility gene for psychotic illness, but not for suicidal ideation within psychotic illness. They also suggest that single-marker analysis of HT2A, as has been previously performed, may be inadequate to detect association with schizophrenia. P296 ASSOCIATION ANALYSIS OF NEUREGULIN 1 GENE VARIANTS WITH SCHIZOPHRENIA IN A SPANISH FAMILY SAMPLE Araceli Rosa1, Michelle Gardner2, Manuel J. Cuesta3, Vı´ctor Peralta3, Mar Fatjo´-Vilas4, M. Eula`lia Navarro5, Salvador Miret6, David Comas2, Lourdes Fan˜ana´s4 1 Universitat Pompeu Fabra, Barcelona 2 Unitat Biologia Evolutiva. Departament d, Barcelona 3 Psychiatric Unit. Virgen del Camino Hosp, Pamplona 4 Unitat d’Antropologia. Departament Biolo, Barcelona 5` Area d’Adolescents. Benito Menni, Comple, Sant Boi Llobregat. Barcelona 6 Centre de Salut Mental de Lleida. Servei, Lleida Neuregulin-1 is one of the most exciting candidate genes for schizophrenia since its first association with the disorder in an Icelandic population, where a core at-risk haplotype consisting of 5 SNPs and 2 microsatellite markers was identified (HapICE). Since then, many studies have analyzed allele and haplotype frequencies in different populations yielding varying results. Interestingly, some authors have replicated the association, although with different alleles or haplotypes, whereas others have failed to replicate. These contradictory results might be attributed to population differences. We investigated the association of NRG1 and schizophrenia in a sample of 150 Spanish nuclear families.We tested 8 SNPs along 1,2Mb across NRG1 covering the regions that have been associated to schizophrenia. These SNPs included: 2 SNPs from HapICE, 1 SNP within the HapICE region and 5 SNPs spaced along the gene. The TDT analysis for each marker did not show a significant over- transmission of alleles from the parents to the affected offspring (P>0.05 for all the tests). TRANSMIT was used to estimate haplotype frequencies (constructed from the 3 SNPs contained within the HapICE region) and test for excess of transmission. The analysis did not show a trend for transmission distortion for these haplotypes (p ¼ 0.4). Our study does not seem to detect association between schizophrenia and the investigated NRG1 markers in a Spanish population. Acknowledgement: Fundacio´ La Marato´ de TV3 (014430/31) and Direccio´n General de Investigacio´n, Ministerio de Educacio´n y Ciencia (BFU2004- 04208/BMC). P297 THE MOLECULAR MECHANISM OF INSULIN RESISTANCE GENERATED BY SECOND GENERATION (ATYPICAL) ANTIPSYCHOTICS ´ va Palik1, Ka´roly Cseh2 Gabor Faludi1, Dezso˜ Birkas-Kovacs1, E 1 Semmelweis University, Budapest 2 Ka´rolyi S. Hospital, Budapest
805
Our aim was to investigate fatty tissue derived factors involved in insulin resistance in patients treated with SGA (clozapine, olanzapine, risperidone, quetiapine). The patient group had been taking SGA for at least one year, (41 women and 19 men,15 for each studied drug). Control groups were 82 healthy people of the same age and gender distribution, and 63 non-psychiatric patients with similar age, gender, BMI and similar percentage of carbohydrate abnormality as obese/diabetic referencegroup. After 1 year, 80% of patients became overweight/obese and 35% presented impaired glucose tolerance or diabetes. Fasting ghrelin, resistin, TNF-alpha, insulin, proinsulin, HOMA A, HOMA Bfn were significantly higher and adiponectin levels significantly lower in patients than in healthy controls. In the obese diabetic control group ghrelin levels were significantly lower than in the healthy controls. A significant positive correlation was found between resistin, TNF-alpha, insulin, proinsulin and HOMA A, and a significant negative one between the above parameters and ghrelin and adiponectin levels both in SGA-treated patients and in obese diabetic controls. In people taking SGA the orexigenic effect of the high fasting ghrelin level contributes to weight gain. The increased production of resistin and TNF-alpha in the fat tissue in obesity, and also the decreased production of adiponectin leads to increased insulin resistance. Negative feedback can be observed between adipocytokines and ghrelin production. SGA drugs enhance ghrelin production despite the suppressive effect of adipocytokines. All four SGA drugs were equally obesitogenic and diabetogenic. P298 DYSREGULATION OF COORDINATE EXPRESSION OF CELL SURVIVAL MOLECULES IN THE PI3K/AKT PATHWAY IN DORSOLATERAL PREFRONTAL CORTEX IN SCHIZOPHRENIA AND BIPOLAR DISORDER. Vladimir Vladimirov, Dawn, L. Thiselton, Po-Hsui Kuo, Michael Neale, Kenneth, S. Kendler, Brien, P. Riley VCU, Richmond AKT1 regulates divergent cellular processes including apoptosis, proliferation, differentiation and metabolism. AKT1 was implicated as a liability locus for schizophrenia by haplotypic association and evidence for decreased protein level and phosphorylation in schizophrenic cases. These changes might affect regulation of apoptotic pathways under AKT1 control. We studied mRNA expression profiles of 23 genes involved in AKT1 signaling in dorsolateral prefrontal cortex from schizophrenia and bipolar cases and controls. Gene expression was analyzed by quantitative real-time PCR and normalized against GAPDH and TBP. We evaluated expression differences between diagnostic groups by Kruskal-Wallis one- way ANOVA. After correction for multiple testing, BDNF was underexpressed in both schizophrenic and bipolar cases compared to controls (p ¼ 0.005). GLUT1 was overexpressed (p ¼ 0.023) and GLUT3 underexpressed (p ¼ 0.05) in the bipolar cases only. These differences were not due to effects of potential confounder variables. Maximum likelihood estimated correlations of co-expression of these genes within groups revealed striking differences between controls and cases. In controls, PTEN, p53, IGFBP1, p21, cyclinD1, CREB, BAX, BAD, MDM2, HiF1a´ and GLUT3 all showed highly positive correlation. In schizophrenic cases many correlations were negative within this gene group. In bipolar cases no correlation of expression was observed. Log-likelihood ratio tests show that the variance-covariance structure and correlated expression of these genes was significantly different between groups (schizophrenic cases vs controls: p ¼ 7 10–20, bipolar cases vs controls: p ¼ 3.4 10– 11, bipolar vs schizophrenic cases: p ¼ 3.5 10–17). P299 MOLECULAR CHARACTERIZATION OF DELETION BREAKPOINTS IN ADULTS WITH 22Q11 DELETION SYNDROME WITH AND WITHOUT SCHIZOPHRENIA Andrea Stachon1, Stephen Meyn1, Jeremy Squire2, Laura Moldovan1, Eva Chow2, Anne Bassett2, Rosanna Weksberg1 1 The Hospital for Sick Children, Toronto, ON 2 Ontario Cancer Institute, Toronto, ON 2 Centre for Addiction and Mental Health, Toronto, ON 22q11 deletion syndrome (22qDS) is a genetic disorder that occurs in approximately 1:4,000 live births. One-third of the individuals with this condition develop schizophrenia (SZ) or related psychotic disorder. We used a FISH and real-time qPCR combined approach to fine map 22q11.2 deletions. The sample was divided into two groups: Group 1
806
Abstracts
was comprised of those individuals who met DSM criteria for SZ (n ¼ 22; 10 women and 12 men; mean age, 35þ7.9 years), and Group 2 those who had no history of psychosis (n ¼ 22; 14 women and 8 men; mean age, 27þ8.36 years). The qPCR data were consistent with the TUPLE1 locus deletion status for all 42 patients positive for this FISH probe. Two individuals negative for TUPLE1 FISH probe had atypical deletions detected in the assay, and later confirmed by RP11-138C22 FISH probe. The majority of 22qDS patients shared a common 3Mb hemizygous deletion of 22q11.2 consistent with the literature. The telomeric breakpoint that was placed in several genomic segments previously was localized within a ~352kb genomic segment between D22S936 and HIC2 for the majority of deleted individuals in our study. PRODH deletions occurred with similar frequency in both groups. PRODH was deleted in most but not allpsychotic individuals. Other mechanisms not evaluated could also be involved in the development of SZ, including mutations in the remaining copy of the deleted region, genomic imprinting, modifying loci located elsewhere in the genome, and environmental influences. P300 THE T(1;11) TRANSLOCATION AND DISC1 HAPLOINSUFFICIENCY Jennifer E. Chubb, J.Kirsty Millar, David J. Porteous School of Molecular and Clinical Medicine, University of Edinburgh, Medical Genetics Section, Molecular Medicine Centre, Western General Hospital, Crewe Rd, Edinburgh, EH4 2XU Using cytogenetics our group has identified a balanced translocation between chromosomes 1 and 11 that segregates with major psychiatric illness in a large Scottish family. The translocation directly disrupts two novel genes on chromosome 1, Disrupted-In-Schizophrenia 1 & 2. Genetic studies implicate DISC1 in psychiatric illness and functional studies suggest that DISC1 is important for neurite outgrowth, neuronal migration, cAMP signalling and mitochondrial function, all of which are crucial for brain development and function. Despite this, the mechanistic link between the t(1;11) and disease remains elusive. One hypothesis proposes that an abnormal truncated DISC1 protein, produced as a result of the translocation, is the direct cause of illness. Extensive Real-Time PCR analysis of cell lines derived from t(1;11) family members has shown that although the translocation results in production of an abnormal DISC1 transcript from the derived 1 chromosome, cell lines carrying the translocation have significantly reduced levels of DISC1 in comparison to family members possessing a normal karyotype. In concordance with this 1D and 2D polyacrylamide gel electrophoresis fails to detect a truncated DISC1 isoform in translocation carriers. This has been confirmed using a panel of Nterminal and C-terminal DISC1 antibodies. On the basis of this evidence, the most likely disease mechanism operating in this family is one of haploinsufficiency for DISC1. To test this hypothesis we have established RNA interference for DISC1 in cultured mammalian cell lines. We are currently using this system to characterise cellular abnormalities which result from reduced levels of DISC1. P 301 GENOME-WIDE SEARCH FOR LINKAGE INTERACTIONS IN SCHIZOPHRENIA Jianxin Shi1, Pablo V. Gejman2, Douglas F. Levinson1 1 Departments of Statistics (JS) and Psychiatry (DFL) Stanford University, Stanford, CA (USA) 2 Center for Psychiatric Genetics, Northwestern University and Evanston Northwestern Healthcare, Evanston, IL (USA) In common complex disorders, epistatic interactions among susceptibility genes are likely, but each locus could have an effect too small to detect by linkage analysis. Therefore, it may be useful to search for interactions among loci. Here we do this with a score statistic, using genome scan data for 280 affected sibling pairs (ASPs) from 263 European-ancestry families (Molecular Genetics of Schizophrenia study, Suarez et al., AJHG 2006;78:315–333). For each ASP, IBD was computed (GENEHUNTER 2.1) at each autosomal marker location (N ¼ 385). We tested each pair of marker positions in the genome except pairs on the same chromosome. Z was computed for each marker pair, by taking the sum (across all ASPs) of [IBD(marker1)1][ IBD(marker2)1], divided by the S.D. of the numerator. Distinct interactions were separated by at least 36 cM (4 times the average distance) on at least one chromosome. For each observed Z, the expected
number of Zs of that size per genome was determined by simulation of 1,000 replicates assuming no linkage. The threshold for genomewide significance (once in 20 scans) was Z ¼ 4.5, and for suggestive significance (once per scan) was Z ¼ 3.83. Four interactions reached suggestive significance (which occurred in 7% of simulated replicates): Marker1 (location)
Marker2 (location)
Z
meta-analysis_bin
D2S405 (44 cM) D5S1457 (59 cM) D4S1647 (92 cM) D10S2327 (97 cM)
D10S2325 (29 cM) D19S245 (49 cM) D11S912 (122 cM) D12S1294 (72 cM),
4.13 3.97 3.94 3.84
10.1 11.5
As a secondary analysis, these results are exploratory, but they demonstrate the feasibility of this approach.
P302 ASSOCIATION ANALYSIS BETWEEN nACHR GENES AND SCHIZOPHRENIA PATIENTS WITH OR WITHOUT SMOKING Caixia li1, Xiaohong Ma2, Qiang Wang2, Xueli Sun2, Wei Deng2, Huaqing Meng3, RM Murray4, DA Collier5, Tao Li5 1 Psychiatric Laboratory, Chengdu, Sichuan 2 Psychiatric Laboratory and Department of, Chengdu 610041, P R China 3 Department of Psychiatry, The First Affi, Chongqing 4 Division of Psychological Medicine,Insti, London SE5 8AF 5 Division of Psychological Medicine and S, London SE5 8AF The prevalence of cigarette smoking among schizophrenia patients is significantly higher than in the general population, but the reasons for this are unknown. In order to explore the roles of nACHR genes in schizophrenia and their relatives to smoking status, we selected one SNP, rs1355920 and one microsatellite D15S1360 in CHRNA7, SNP rs1044397, a novel microsatellite marker in each of CHRNA2 and CHRNA4 respectively. Individual genotyping of these five markers were performed on 164 patients with schizophrenia and 330 of their first-degree relatives. We found significant allelic transmission distortion of SNPs rs1355920 from CHRNA7 (P ¼ 0.008). Paired-marker haplotype analysis of rs1355920 and D15S1360 from 7Nachr showed nominally significant association between haplotype rs1355920A-msallele-115bp and chizophrenia (P ¼ 0.03). We also found a weak association between a novel microsatellite marker in 4nACHR and schizophrenia (P ¼ 0.047). There were no significant association between either SNPs rs1044397 in 4nACHR gene or the novel microsatellite marker in nACHR gene and schizophrenia. Using status of smoking in schizophrenic patients as phenotype, we found the novel microsatellite marker in 4nACHR gene was association with status of smoking in schizophrenic patients (P ¼ 0.026). In conclusion, our study suggestion CHRNA7 and CHRNA4 genes may play a role in the pathogenesis of schizophrenia and smoking behaviour in schizophrenics. Thiswork was partly supported by funding from the Chinese National Natural Science Foundation (TL), the Wellcome Trust (TL, DAC, PS), NARSAD (TL), and the Schizophrenia Research fund (TL). P303 ASSOCIATION STUDY BETWEEN THE NrCAM GENE AND PATIENTS WITH METHAMPHETAMINE USE DISORDERS Yuko Okahisa1, Hiroshi Ujike2, Yuji Tanaka1, Kyohei Otani1, Yukitaka Morita1, Makiko Kishimoto1, Akiko Morio1, Toshiya Inada2, Mutsuo Harano2, Tokutaro Komiyama2, Toru Hori2, Mitsuhiko Yamada2, Yoshimoto Sekine2, Nakao Iwata2, Masaomi Iyo2, Ichiro Sora2, Norio Ozaki2, Shigetoshi Kuroda1 1 Okayama University, Okayama 2 Okayama University, Japanese Genetics In, Okayama 2 JGIDA, Ichihara It has been suggested that individual genetic factors are involved in susceptibility to methamphetamine dependence and manifestation of drug-induced psychosis. Ishiguro et al. have recently reported that variants of the neuronal cell adhesion molecule (NrCAM) gene are likely to change expression and alter substance abuse vulnerability in human addiction and animal models of drug reward. They showed that the specific 30 haplotypes and 50 haplotypes of NrCAM gene were associated with substance abuse. The 30 haplotype block includes
Abstracts 0
rs439587, rs401433 and rs6958498, and the 5 haplotype block includes rs3763462, rs1859769 and rs2284284. Therefore, we investigated 6 SNPs in methamphetamine use disorders. We examined 230 subjects who were methamphetamine users and 248 controls by case-control association study. All subjects were Japanese. There were no significant differences in either polymorphism of 30 or 50 region of the NrCAM gene between controls and patients with methamphetamine use disorders. We also examined genetic effects of the NrCAM gene on clinical features, with respect to age at first use, latency of psychosis, spontaneous relapse of psychotic symptoms and multi-substance abuse.We found a significant association of rs2284284 polymorphism (P ¼ 0.034) with latency from first consumption to onset of methamphetamine-induced psychosis, and a significant association of rs1859769 polymorphism (P ¼ 0.03) with multi-substance abuse. The present findings suggested that the NrCAM gene did not play a major role in susceptibility to methamphetamine dependence, but affected some clinical aspects of methamphetamine use disorders. P304 THE FRIZZLED 3 (FZD3) GENE IS ASSOCIATED WITH METHAMPHETAMINE PSYCHOSIS Makiko Kishimoto1, Hiroshi Ujike2, Yuji Tanaka1, Kyohei Otani1, Yukitaka Morita1, Akiko Morio1, Yuko Okahisa1, Tatsuya Kotaka1, Mutsuo Harano2, Toshiya Inada3, Mitsuhiko Yamada4, Tokutaro Komiyama4, Toru Hori4, Yoshimoto Sekine5, Nakao Iwata6, Ichiro Sora7, Masaomi Iyo8, Norio Ozaki9, Shigetoshi Kuroda1 1 Okayama University, Okayama city, Okayama 2 Okayama University, JGIDA (Japanese Gene, Okayama city, Okayama 2 Kurume University, JGIDA(Japanese Geneti, Kurume city, Fukuoka 3 Teikyo University Ichikawa Hospital, JGI, Ichikawa city, Chiba 4 National Center of Neurology and Psychia, Kodaira city, Tokyo 5 Hamamatsu University, JGIDA(Japanese Gen, Hamamatsu, Shizuoka 6 Fujita Health University, JGIDA(Japanese, Houmei city, Aichi 7 Tohoku University, JGIDA(Japanese Geneti, Sendai, Miyagi 8 Chiba University, JGIDA(Japanese Genetic, Chiba city, Chiba 9 Nagoya University, JGIDA(Japanese Geneti, Nagoya city, Aichi FZD3 is a receptor required for the Wnt-signalling pathway that has been implicated in CNS development. We found a significant association between schizophrenia and the FZD3 gene in 2003. Subsequent studies have followed, producing inconsistent results. Since clinical features of methamphetamine psychosis are similar to those of schizophrenia, we analysed the FZD3 gene in methamphetamine psychosis patients to investigate the role of the FZD3 gene in psychotic disorders. Subjects were 188 patients with methamphetamine psychosis and 240 healthy controls. Genotyping the 5 SNPs (SNP1: rs3757888, SNP2: rs960914, SNP3: rs2241802, SNP4: rs2323019, SNP5: rs880481) of the FZD3 gene was performed by PCR-RFLP assays. Although association analyses on a single SNP showed no significant association with methamphetamine psychosis for any SNP examined, haplotype analyses on SNP2-3-4-5 revealed quite a strong asscoation of FZD3 gene with methamphetamine psychosis (p < 0.000001). Estimated haplotype frequency showed that the T-G-A-G haplotype of SNP2-34-5 was less in patients than in controls (P ¼ 4.74 109), implying a negative riskfactor. In our previous study on schizophrenia (Katsu et al., 2003), the T allele of SNP2 and T-G haplotype of SNP2-3 indicated were risk for schizophrenia. Zhang et al.(2004) also reported that A-T-G haplotype of SNP4-rs352203-SNP5 indicated risk for schizophrenia in the Chinese population. These findings indicate that the FZD3 gene may affect susceptibility to two distinct psychoses, endogenous and substance-induced psychosis, although in different ways.
P305 CIRCADIAN RHYTHM GENE PER1 IS ASSOCIATED WITH ALCOHOL DRINKING BEHAVIOR Li Dong Institute of Psychiatry, London Alcohol addiction is a complex neurological disorder caused by both environmental and genetic components. The circadian genes Per1 and Per2 have recently been linked to alcohol sensitivity and alcohol drinking behaviour. We have performed an association analysis between Per1 and alcohol drinking behaviour in humans and identified
807
15 single nucleotide polymorphisms (SNPs) in Per1. We have found an association between one of these SNPs (a C > T substitution) with alcohol intake in alcohol-dependent adult individuals ( < / > ¼ 300g/d). Furthermore, we found that this SNP associates with harmful drinking in adolescents. This association is dependent on high psychosocial stress. To investigate the function of this polymorphism on Per1, we generated genotype-specific Epstein-Barr virus (EBV)-transformed lymphoid cell lines (LCL). These cell lines were established with lymphocytes isolated from peripheral blood of individuals carrying a specific allele of this SNP. We compared the Per1 expression profile in these cell lines after synchronisation with 50% horse serum. Interestingly, we found that Per1 mRNA expression was higher in cell lines carrying the CC genotype (1.7folds) than in cell lines with the TT genotype (p < 0.0001), suggesting that this SNP influences expression of Per1. As a whole, our results suggest that stress-dependent alcohol drinking behaviour is associated with a SNP in Per1. The effect of this SNP on Per1 gene expression may reflect the potential molecularbiological mechanism of this association. P306 BIPHASIC EFFECTS OF ETHANOL IN RETENTION MEMORY PERFORMANCES Francesco Marrosu1, Federico Santoni2, Monica Puligheddu3, Luigi Barberini3, Antonella Tuveri3, Fabrizio Genugu3, Zvani Luigi Rossetti4 1 Universita` di Cagliari, cagliari 2 CRS4, Pula- CA 3 Dipartimento di scienze Neurologiche e Cardiologiche, Cagliari 4 Dipartimento di Neuroscienze, Cagliari Purpose: Cognitive and memory impairment after ethanol intake are commonly observed.Recent studies have challenged at least one part of such common knowledge, given that it has been shown in rats that low doses of ethanol increase spatial working memory (WM) and acetylcholine (Ach) release in the prefrontal cortex of rats, while high doses disrupted WM and decreases Ach release. In the present, preliminary experiments, we tested in healthy volunteers whether a possible equivalent of the WM of animals, represented by memory retention performance (MRP), varies following ethanol intake. Methods: 4 healthy volunteers, homogeneous for sex, age and education, underwent the following experimental double blind design. In three distinct experimental sessions, subjects received 30 cc of orange juice randomly containing 0.5 ml/Kg, 1mg/kg alcohol or no alcohol, the edge of all cups being stained with some drops of alcohol. The task consisted of memorising 4 series of 10 common words; eachsubject was allowed 10 sec for memorisation and 1 min for reporting. The process was randomly repeated in each experimental condition and recorded with a 64 channel digital EEG. R Results: our results show that the intermediate dose of alcohol induced in all subjects a significant increase in MRP retention (p > 0.01). Conclusions: Low alcohol increases MRP, as compared with higher or no doses. Although these preliminary results lack statistical power, they nonetheless encourage larger studies in order to compare them with the experimentally-established biphasic alcohol profile. P307 ALCOHOL ENDOPHENOTYPES AND THE val158met POLYMORPHISM OF THE COMT GENE Chinatsu McGeary1, John McGeary2, Peter Monti2, Damaris Rohsenow2, Jennifer Tidey3, Robert Swift2, Robert Miranda3 1 Brown University, Providence, RI 2 VA, Providence, RI 3 Brown Univ., Providence, RI Background: The val158met polymorphism of the catechol-O-methyl transferase (COMT) gene is ambiguously associated with alcohol dependence, mu opiate receptor response and executive function differences. This study examines associations of this polymorphism with alcohol dependence and alcohol-related endophenotypes in a laboratory trial of naltrexone on alcohol cue reactivity. Method: Data from 94 participants in a larger study of medication effects were used to examine the hypotheses. Thenon-treatment seeking male and female heavy drinkers (64% alcohol dependent) were genotyped for the val158met polymorphism of the COMT gene (val/val [n ¼ 28], val/met [n ¼ 50], met/met [n ¼ 16]). 10 days after randomisation to NTX (50 mg) or placebo, participants completed an alcohol cue reactivity assessment.
808
Abstracts
Results: The val158met polymorphism was neither associated with alcohol dependence nor alcohol-related endophenotypes of cue reactivity and alcohol consumption in the past 90 days. COMT status did not moderate naltrexone’s effects on alcohol cue reactivity. The lack of COMT effects was not likely to be due to inadequate statistical power, since all effect sizes were small. Discussion: These data suggest that the val158met polymorphism is unrelated to alcohol dependence or the craving endophenotype but do not rule out the possibility of an additional mechanism such as stress relief accounting for published positive associations with alcohol dependence. The data further suggest that unlike the OPRM1 Asn40Asp polymorphism, the val158met polymorphism is not a moderator of naltrexone’s effects on alcohol cue reactivity.
P308 POLYMORPHISMS OF COMT GENE AND COCAINE ABUSE/DEPENDENCE IN A BRAZILIAN POPULATION STUDY Homero Vallada1, Nadia Cunha1, Quirino Cordeiro1, Guindalini Camila2, David Collier3, Ronaldo Laranjeira4, Gerome Breen5 1 University of Sao Paulo Medical School, Sao Paulo 2 Institue of Psychiatry, Kings’ College L, London 3 Institute of Psychiatry, Kings’ College, London 4 UNIAD, Federal University of Sao Paulo, Sao Paulo 5 Institute of Psychiatry, Kings College L, London Introduction: Genetic factors play a role in vulnerability to cocaine dependence. The reinforcing properties of cocaine are related to the dopaminergic system. In this system, dopamine is the main neurotransmitter, and the cathecol-O-methyltransferase (COMT) is an important enzyme for its metabolism. Some studies have suggested that the polymorphism Val 108/158 Met (rs 4680) of the COMT may be associated with a higher risk for substance abuse and addiction. However, other studies did not replicate this finding. Data from other independent samples may clarify the putative association. Methods: Three COMT SNPs [Val 108/158 Met (rs 4680), rs737865, rs165599] were investigated in an independent, ethnically-diverse Brazilian population comprising 691 patients with cocaine dependence, compared with 839 community-based controls. Results: No statistically-significant differences were observed for individual SNPs or haplotypes between case and control groups. Conclusion: COMT polymorphisms are not a genetic risk factor for cocaine abuse/dependence in our sample.
P309 THE ROLE OF GENETIC AND ENVIRONMENTAL FACTORS IN THE DEVELOPMENT OF HEROIN ABUSE Agnes Szilagyi1, Csaba Barta1, Anna Szekely2, Zsolt Demetrovics2, Maria Sasvari-Szekely1 1 Semmelweis University, Budapest 2 Eotvos Lorand University, Budapest Our previous results showed that childhood ADHD left untreated could be a predisposing factor for developing substance use disorder (SUD). This effect is more pronounced in the presence of particular genotype forms of DRD4. In this study we increased the number of heroin dependent subjects and extended their phenotypic characterization with psychiatric questionnaires including ADHD RS and TCI. Five genes related to the dopamine and serotonin systems were analyzed to evaluate the role of interactions between genetic and environmental factors in developing SUD. The case population of 182 heroin abusers participated in a methadone response study. The age- and sex-matched control group consisted of 384 healthy individuals. All subjects were genotyped for 11 polymorphisms (dopamine receptor D4: 120 bp duplication, 521C/T SNP, 616C/G SNP, 615A/G SNP, 48 bp VNTR; serotonin transporter: 5-HTTLPR, STin2; dopamine transporter: 30 UTR VNTR, intron8 VNTR; catechol-O-methyltransferase: G158A SNP; monoamino-oxidase A: 50 VNTR). Chi2 test and T probe were used in statistical analyses. In agreement with previous findings, a higher portion of SUD patients reported symptoms of ADHD than controls. Our preliminary data show that the dopamine receptor D4 615G carriers achieved higher hyperactivity and attention-deficit scores than the ones with AA genotype. As the contribution of individual factors in psychiatric conditions is usually very small and
may be modified by other effects,the combined analysis of genetic and environmental factors and their possible interactions should be preferred to verify their role in developing substance dependence.
P310 THE GABRA2 GENE IS ASSOCIATED WITH THE ACUTE SUBJECTIVE EFFECTS OF ALCOHOL, AND POST-MORTEM mRNA EXPRESSION Heather Haughey1, Patrick Finan1, Rex Villanueva2, Francesca Filbey3, Kent Hutchison3 1 University of Colorado at Boulder, Boulder 2 University of Colorado at Boulder, Boulder 3 University of Colorado at Boulder, Boulder GABAA receptors are expressed in several limbic brain regions implicated in drug reward and dependence. Recently, studies ave demonstrated that several single nucleotide polymorphisms (SNPs) within the GABRA2 gene region form a ‘risk’ haplotype for alcohol dependence (Edenberg et al. 2004; Covault et al. 2004). The present study examined which SNPs within the GABRA2 gene influencethe acute subjective effects of alcohol in an alcohol challenge study. 77 subjects participated in a two session alcohol challenge paradigm. Thesessions were randomized, participant-blind and separated by one week.During the alcohol-administration session, participants consumed 3 high-alcoholic beers whereas during the placebo session, participants consumed 3 non-alcoholic beers. Results indicate that the GABRA2 risk-haplotype demonstrated both a significant interaction for trial x GABRA2 (p < 0.031), and a significant main effect on the mood effects of happiness (p < 0.040) and vigor (p < 0.048), as measured by the Profile of Mood States. This riskhaplotype was also found to have a significant main effect (p < 0.04) on the hedonic value of alcohol. In addition, a main effect of the riskhaplotype by mRNA level in post-mortem prefrontal cortex wasalso observed. In summary, this study support previous findings that the GABAergic system plays a role in brain mechanisms of reward and suggests that polymorphisms within GABRA2 may be useful at predicting future risk for alcoholism.
P311 ASSOCIATION ANALYSIS OF THE MU-OPIOID RECEPTOR GENE TO ALCOHOL AND NICOTINE DEPENDENCE IN ICELAND Anna Wiste1, Andres Ingason2, Andrew A. Hicks2, Thorlakur Jonsson2, Frank Geller2, Natasa Desnica2, Ho¨gni Oskarsson3, Valgerdur Runarsdottir4, Augustine Kong2, Jeffrey R. Gulcher2, Thorgeir E. Thorgeirsson2, Thorarinn Tyrfingsson4, Kari Stefansson2 1 Decode Genetics, Reykjavik 2 Decode genetics, Reykjavik 3 Therapeia, Reykjavik 4 SAA, National Center of Addiction Medic, Reykjavik Mu-opioid receptors play an important role in addiction pathways involving substances including opiates, alcohol, and nicotine. An understanding of variation in the mu opioid receptor gene (OPRM1) and its relation to addiction may contribute to the development of addiction therapies. Several studies have reported association of OPRM1 to alcohol and drug dependence, but studies with nicotine dependence are yet to be published. We used a tagging SNP approach to test association of OPRM1 to alcohol and nicotine dependence (AD, ND). Three independent patient cohorts were used: (1) AD (n ¼ 713); (2) heavy smoking (n ¼ 558); (3) AD (n ¼ 560) and/or ND (n ¼ 1397). Population controls (n ¼ 796) were not screened for addiction. Twelve SNPs were selected, including rs 1799971, the A118G polymorphism tested in most reports. No singlepoint associations surviving correction for multiple testing were found in either phenotype. For rs1799971, there was no significant difference (p > 0.05) in the minor allele frequency between controls (9.5%) and any of the patient groups: AD—(1) 8.9%, (3) 8.8%; Smoking—(2) 8.3%, (3) 8.1%). Tag SNPs were chosen to maximize haplotype coverage in the 2 LD blocks. The 6 SNPs in the 50 block form 8 haplotypes with frequency greater than 1%, covering 98% of the haplotype diversity. Haplotype analysis within this block showed suggestive associations, particularly with the combined phenotype of alcohol and nicotine dependence. However, these initial findings require replication in other cohorts to confirm that the observed moderate risks are real.
Abstracts P312 CANDIDATE GENES, PATHWAYS AND MECHANISMS FOR ALCOHOLISM ANDRELATED DISORDERS: AN EXPANDED CONVERGENT FUNCTIONAL GENOMICS APPROACH Alexander B. Niculescu Indiana University School of Medicine, Indianapolis, IN 46202 We describe a comprehensive translational approach for identifying candidate genes for alcoholism. We use cross-matching of animal model brain gene expression data with human genetic linkage data and postmortem data, an approach termed Convergent Functional Genomics. A meta-analysis of three animal model paradigms, based on inbred alcohol preferring (iP) and alcohol non-preferring (iNP) rats, and their response to treatments with alcohol, was used. A comprehensive analysis of microarray gene expression data from five key brain regions (frontal cortex, amygdala, caudate-putamen, nucleus accumbens, and hippocampus) was carried out. The Bayesian-like integration of multiple independent lines of evidence, each by itself lacking sufficient discriminatory power, led to the identification of high probability candidate genes, pathways and mechanisms for alcoholism. These data reveal that alcohol has pleiotropic effects on multiple systems, which may explain the diverse neuropsychiatric and medical pathology in alcoholism. Some of the pathways identified suggest avenues for pharmacotherapy of alcoholism with existing agents, such as angiotensin converting enzyme (ACE) inhibitors. Experiments we carried out in alcohol preferring (iP) rats with an ACE inhibitor show a marked modulation of alcohol intake. Other pathways are new potential targets for drug development. The emergent overall picture is that physical and physiological robustness may permit alcohol preferring individuals to withstand the aversive effects of alcohol. In conjunction with a higher reactivity to its rewarding effects, they may able to ingest enough of this non-specific drug for a strong hedonic and addictive effect to occur.
P314 INTERLEUKIN-1 ALPHA AND BETA, TNF AND HTTLPR GENE VARIANTS ARE NOT ASSOCIATED TO ALCOHOL TOXICITY AND TO DETOXIFICATION OUTCOME Alessandro Serretti1, Ioannis Liappas2, Laura Mandelli1, Diego Albani3, Gianluigi Forloni3, Petros Malitas2, Christina Piperi4, Aikaterini Zisaki4, Elias O. Tzavellas2, Zeta Papadopoulou-Daifoti5, Francesca Prato3, Sara Batelli3, Marzia Pesaresi3, Anastasios Kalofoutis4 1 Institute of Psychiatry, University of Bologna, Bologna, Italy 2 Department of Psychiatry, Eginition Hospital, University of Athens Medical School, Athens, Greece 3 Neuroscience Department. Istituto di ricerche Farmacologiche ‘‘Mario Negri’’, Milan, Italy 4 Laboratory of Biological Chemistry, University of Athens Medical School, Athens, Greece 5 Department of Experimental Pharmacology, University of Athens Medical School, Athens, Greece Genetic factors may influence liability to treatment outcome and medical complications in alcoholism. In the present study we investigated the IL1A rs1800587, IL1B rs3087258, TNFa rs1799724 and the HTTLPR variants in a sample of 64 alcohol dependents and 47 relatives with a set of clinical parameters and outcome measures. Alcohol dependents had a less favorable clinical profile compared to relatives (higher cholesterol, triglycerides, glucose, glutamic oxaloacetic transaminase, glutamic pyruvic transaminase, gamma-glutamyl transpeptidase). After detoxification, all clinical indexes improved and hepatic enzyme levels were similar in alcohol dependents and relatives, except for the GGT that remained significantly higher in alcohol dependents. Alcoholic depressive and anxiety scores were significantly reduced after detoxification. IL-1A, IL-1B, TNF and HTTLPR variants were not associated with any clinical index or change after detoxification. In our sample IL1A, IL1B, TNF and HTTLPR are not liability factors for alcohol toxicity or detoxification outcome.
809
P315 MAJOR DEPRESSIVE DISORDER: GENETIC ASSOCIATION ANALYSES USING SUBPHENOTYPES Inga Prokopenko1, Federica Tozzi1, Elena Vicentini1,5, Alice Vanzo1, Shyama Brewster2, Julia Perry1, Stephanie Chissoe3, Pamela St. Jean3, Aruna Bansal3, Ruchi Upmanyu3, Laura Budde3, Claire Allan4, John Riley3, Enrico Domenici5, Emilio Merlo-Pich5, Alun McCarthy1, Florian Holsboer6, Allen Roses7, Lefkos Middleton1, Wade Berrettini8, James Kennedy9, Pierandrea Muglia1 1 Translational Medicine & Genetics GlaxoSmithKline R&D, Verona (Italy) & Greenford (UK) 2 Pharmacogenetics Sampling, Greenford (UK) 3 Discovery and Pipeline Genetics GlaxoSmithKline, Harlow and Stevenage (UK) & RTP (USA) 4 GR Therapeutic Area Team, Stevenage (UK) 5 Psychiatry CEDD, Verona (Italy) 6 Max-Planck-Institut fu¨r Psychiatrie, Munich (Germany) 7 Genetics Research, RTP (USA) 8 Center for Neurobiology and Behavior, University of Pennsylvania School of Medicine, Philadelphia (USA) 9 Centre for Addiction and Mental Health, Department of Psychiatry, University of Toronto (Canada) Introduction: Depression is a heterogeneous disorder and may be dissected into more homogeneous and heritable sub-phenotypes, each with partly distinct pathogenic mechanisms. A sample of 1,000 cases with recurrent Major Depression and 1,000 gender-, ethnicity- and agematched controls has been collected in collaboration with Academic centres, as part of a large genetic initiative at GSK with the goal of identifying viable therapeutic targets through human genetics studies (Roses et al 2005). Methods: Six thousand five hundred SNPs were genotyped and up to 1827 genes analysed. Each participant had provided information on ethnicity, family medical history, and smoking and drinking habits. Detailed clinical and psychological assessment was conducted on cases, and included DSM-IV diagnosis (based on SCAN interview) and personality traits (Eysenck Personality Questionnaire, EPQ). The associations for a number of genes were stratified for a number of clinically relevant and biologically meaningful depression sub-phenotypes that were sufficiently represented in the sample to provide adequate statistical power. Both dichotomous and continuous traits were tested (e.g. gender; depression family history; anxiety; melancholia; BMI; neuroticism; age of onset). Genotyped SNPs in selected genes were tested for association between controls and the subset of depression patients with the phenotype of interest. Results: Results on some of the analyses performed will be presented, together with algorithms of the subphenotype definition. Conclusions: Results showed that phenotype refinement and definition of sub-phenotypes with higher genetic contribution can help in increasing confidence on disease-associated genes. Furthermore that analysis of large genotyping data can corroborate and substantiate the existence of depression sub-phenotypes.
P316 FURTHER INVESTIGATION OF PSYCHOTIC SYMPTOM DIMENSIONS AS PHENOTYPES FOR GENETIC RESEARCH Alastair G. Cardno Academic Unit of Psychiatry, Leeds Background: Clinical variation in psychoses can be defined, based on factor analysis, in terms of three main symptom dimensions - positive, negative and disorganised, each of which may have subdivisions. Studies of sib-pairs and twins concordant for psychoses show withinpaircorrelations consistent with a familial, possibly genetic, contribution to their aetiology. However, symptom dimensions may be defined in a variety of ways, and it is not clear how familial aggregation varies with varying definitions of symptom dimensions. Aims: To investigate the degree of familial aggregation for a range of definitions of psychotic symptom dimensions in pairs of monozygotic twins concordant for psychosis. Methods: The sample was derived from the Maudsley twin psychosis series,in London (n ¼ 44 concordant MZ pairs). Symptom dimensions were defined using various combinations of lifetime-ever psychotic symptoms rated on the OPCRIT checklist. Within-pair correlations were analysed to determine the degree of familial aggregation.
810
Abstracts
Results: A range of definitions of the disorganised dimension showed notable and similar familial aggregation. Some definitions of the negative dimension showed familial aggregation, but the value of combining symptoms on an ordinal scale was unclear. Definitions of positive symptom dimensions tended to be relatively modest, with a trend for the paranoid sub-dimension to show strongest familial aggregation. Conclusion: The disorganised dimension shows most consistent familial aggregation across different definitions. However, the potential value of each symptom dimension forgenetic research warrants further investigation in independent samples. P317 FURTHER EVIDENCE OF MAO-A GENE VARIANTS ASSOCIATED WITH BIPOLAR DISORDER Alessandro Serretti2, Daniel J. Mu¨ller1,3, Tricia Sicard3, Subi Tharmalingam3, Nicole King3, Laura Mandelli2, James L. Kennedy3 1 Department of Psychiatry, Charite´ University Medicine Berlin, Campus Charite´ Mitte, Berlin, Germany
2
Institute of Psychiatry, University of Bologna, Bologna, Italy Neurogenetics Section, Centre for Addiction and Mental Health, Department of Psychiatry, University of Toronto, Toronto, ON, Canada 3
The aim of this study was to investigate MAOA gene variants in bipolar disorder by using a family-based association approach. The sample included 331 nuclear families from Western and Central Canada with at least one offspring affected with bipolar disorder comprising a total of 1044 individuals. All subjects were genotyped for MAOA–941T > G and uVNTR gene variants using PCR techniques. Haplotype TDT was statistically significant (LRS ¼ 12.17; df ¼ 3; p ¼ 0.0068; permutation global significance ¼ 0.00098), with the T4 haplotype significantly associated with bipolar disorder (OR ¼ 1.63, 95% CI ¼ 1.11–2.37). Single marker analysis evidenced a borderline association for MAOA–941T > G (p ¼ 0.04) but not for the uVNTR. No different results were detected when analyzing female subjects separately. In conclusion, our family-based association study gives mild but further support of the involvement of MAOA variants in bipolar disorder.
Author Index Aas Monica, P161, 774; P164, 775; P257, 796; P258, 796; P269, 798; P270, 799; P279, 801; P291, 804 Abdolmaleky Hamid, O9.4, 708; O25.1, 731 Abecasis G., S3.3, 688 Abeling Nico, O7.6, 705 Abrahams Brett S., O23.2, 727 Abramova L., P45, 748 Abramson R.K., O20.4, 723 Absher D., S3.3, 688 Absher Devin, O2.1, 695 Acierno Ronald, P19, 742 Adams Philip, O1.3, 694 Addington Anjene´, P143, 770 Adolfsson Rolf, O13.5, 713; O13.6, 713; P07, 740; P79, 756 Aggen Steve, O17.5, 720 Aguglia Eugenio, P81, 756 Aguilar Alejandro, P281, 801 Aguilera Mari, P108, 763; P110, 763 Airey David, O23.3, 727 Aitchison K., S6.2, 690 Aitchison K.J., P87, 758; P103, 761 Aitchison Katherine, P107, 762; P227, 789; P228, 789 Aitchison Katherine J., O19.5, 722; P161, 774; P164, 775; P279, 801; P291, 804 Aitchison Kathy, P257, 796; P258, 796 Aitichison Katherine J., P269, 798 Akahane Akihisa, P251, 794 Aken Marcel A.G. Van, O14.6, 715 Akil H., S3.3, 688 Akil Huda, O2.1, 695 Akira Houkyou, P238, 792 Akira Nakajima, O29.4, 737 Akula Nirmala, O2.7, 696 Alaerts Maaike, O2.4, 696; O13.5, 713 Albanese M., P85, 757 Albani Diego, P207, 785; P314, 809 Albers Carolyn, P163, 775 Alblas Margrieta, O25.4, 731 Albus Margot, O11.5, 711; O16.7, 718; O3.1, 697 Alda M., P85, 757 Alda Martin, O13.3, 713 Alen Reija, O6.3, 703 Alexander J., O4.2, 699 Alexiev Spiridon, P51, 749 Alford Tyler, P268, 798 Alhberg S., P103, 761 Alizadeh Z., P166, 776 Allan C., S6.3, 690 Allan Claire, P315, 809 Allard Rene, O16.5, 718 Allgulander Christer, P76, 755 Allott Emma H., O29.3, 737 Alm Barbara, P84, 757 Almasy Laura, O16.8, 718; P66, 753 Almeida Joana, O28.4, 736; P91, 759 Alonso Maria del Pino, O5.1, 701; O22.10, 726 Alpman Asude, P07, 740 Als Thomas Damm, P249, 794 Als Thomas, P20, 743; P24, 744 Altink Marieke, P119, 765 Alvarez J.C., P264, 797 Alvarez Natalia, P151, 772; P202, 784 Ambrosini Daniel L., O10.2, 709 Amdur Richard, O16.3, 717 Amdur Richard, P295, 804 Amelsvoort Therese van, O7.6, 705; P60, 751 Amin F., S5.4, 689 Amor Leila Ben, P134, 768 An Seon-Sook, P98, 760
Ana Ana Miranda, P67, 753 Anastopoulos Arthur, P129, 767; P148, 771 Ancı´n Ine´s, P151, 772; P202, 784 Anckarsater Henrik, P131, 768 and the IMAGE Consortium, S2.2, 687 Anderson Barbara, O24.3, 729 Anderson George, P122, 766 Andersson Weronica, P196, 782 Andrews Gavin, O5.5, 702 Androsiuk Wojciech, P205, 784 Angelicheva Dora, P51, 749 Anguita Manuel, P283, 802 Anisimov N., P280, 801 Anisimov N.V., P272, 799 Anisimov Nikolay, O25.5, 732; P271, 799 Annerbrink Kristina, P76, 755 Annesi G., P203, 784 Anney R., S2.2, 687 Anney Richard, O28.6, 736 Anokhin Andrey, P154, 773 Antila Mervi, O2.5, 696; P96, 760 Anton A., P85, 757 Apiquian Rogelio, P263, 797; P281, 801 Aprili Giuseppe, O25.6, 732 Archibald Jennifer, O24.3, 729 Ardau Raffaella, P56, 750; P167, 776 Arden Rosalind, O15.5, 716 Arends J., P210, 785 Argudo Isabel, P151, 772 Arias Ba´rbara, P108, 763; P110, 763; P198, 783 Arias Barbara, P233, 791 Arinami Tadao, P180, 779; P192, 781 Armas Regina, P13, 741 Arnold Paul Daniel, O14.2, 714 Aron Liviu, P07, 740 Arranda Fernando Ferna´ndez, P144, 770 Arranz M.J., P107, 762 Artioli Paola, P204, 784; P221, 788 Artyuch V., P45, 748 Asada Takashi, P26, 744 Ascaso Carlos, O8.5, 707 Aschauer Harald, P48, 749; P145, 771 Asherson Philip, O17.1, 719; O20.3, 722; O28.5, 736; O28.6, 736; P67, 753; P117, 765; P126, 767 Ashley-Koch Allison, P99, 761; P129, 767; P148, 771 Asuni C., P208, 785 Ataı´de Assunc¸a˜o, O28.4, 736; P91, 759 Athanasiou Maria, O19.2, 721 Atsushi Tsutsumi, P238, 792 Aukes Maartje, P166, 776 Auladell Carme, P198, 783 Ayab Mohammad, P247, 794 Azuma Rayna, P136, 769 Azzam Amin, O15.3, 716 Baas F., P86, 757 Baas Frank, P60, 751 Babadjanova G., P45, 748 Babb Casey, P21, 743 Bacchelli E., O28.3, 736 Bacchelli Elena, P130, 767 Bachner-Melman Rachel, P146, 771 Badger Jonathon, P34, 746 Badner Judith A., O2.7, 696 Bagnarello Monica, O15.3, 716 Bailer Ursula, P48, 749; P145, 771 Bailey A.J., O28.3, 736 Bailey Anthony J., P130, 767 Bailey Julia, P178, 778 Baji Ildiko´, O24.4, 729; O24.6, 730
In this index, the numbers in the italics are the pages in which the abstracts appear.
ß 2006 Wiley-Liss, Inc.
Bajic Vladan, O12.1, 711 Balazˇic Jozˇe, O26.3, 733 Balazˇic Olga, O26.3, 733 Bales Karen, P178, 778 Balogh Gabriella, P11, 741 Banaschewski Tobias, O17.1, 719; O28.5, 736; O28.6, 736; P67, 753; P117, 765 Bansal A., S6.3, 690 Bansal Aruna, P315, 809 Barabash Ana, P151, 772; P202, 784 Barberini Luigi, P193, 782; P306, 807 Barchas Jack D., O2.1, 695 Barcikowska Maria, P205, 784 Barden Nicholas, O2.2, 695; P08, 740; P134, 768 Barlas Omer, P214, 786 Barlati Sergio, P273, 799; P274, 800 Barlati Stefano Davide, P274, 800 Barnby G., O28.3, 736 Barnby Gabrielle, P130, 767 Barone P., P203, 784 Barr Cathy, O24.4, 729; O24.6, 730; O24.8, 730; P68, 753 Barr Cathy L., O22.1, 724; O24.3, 729 Barrantes-Vidal Neus, P108, 763; P110, 763 Barrett Tom, O2.2, 695 Barrigo´n Marı´a Luisa, P283, 802 Barry Edwina, O24.7, 730 Barta Csaba, O8.6, 707; P309, 808 Bartolomeis A. de, P203, 784 Basoglu Cengiz, P214, 786 Bass Nicholas J., P189, 781; P293, 804 Bass Nicholas, P42, 747 Bass Nick, P49, 749 Bassett Anne, P299, 805 Batelli Sara, P207, 785; P314, 809 Bauche´ Ste´phanie, O3.1, 697 Baud Patrick, O22.4, 725; O26.2, 733; O26.6, 734; P29, 745 Baye´s Mo`nica, O5.1, 701; O22.2, 724; O22.10, 726; P80, 756; P144, 770 Beck Angus, O13.3, 713 Becker Tim, P262, 797; P95, 760 Beckmann Lars, O23.4, 727 Beem A. Leo, O22.7, 725 Beemer Frits, O5.3, 701 Behrooz, P166, 776 Beierschmitt Amy, O4.5, 700 Belesioti Barbara, P125, 766 Belivier F., O26.2, 733 Belknap John, P195, 782 Bello´n Juan, O1.6, 694; P05, 739; P78, 756 Bellgrove Mark, O24.7, 730; P58, 751 Bellivier Franck, O22.4, 725; P29, 745; P31, 745 Bellodi Laura, O18.4, 720; P158, 774 Bellone Emilia, P275, 800 Belouchi Majid, O16.5, 718 Ben Neale, O28.5, 736 Bento Celeste, O28.4, 736 Beresheva Alphia, P141, 770 Berg S.M. van den, P03, 739 Bernardini Sergio, O12.4, 712 Bernik Marcio, P159, 774; P223, 788 Berrettini Wade, P315, 809 Berrettini Wade H., O13.1, 712; S8.4, 692; P245, 793 Bertani Maria Elena, O25.6, 732 Bertani Mariaelena, P276, 800 Bertolin Cinzia, O17.4, 719; P52, 750 Bescos Maria Jose´, P151, 772 Betancur Catalina, P131, 768 Bettecken Thomas, P09, 740
812
Author Index
Beutler Sonja, P171, 777 Bhatia Triptish, P174, 777 Bicahlo Maria Aparecida, P16, 742 Biederman Joseph, O22.6, 725 Bierut L., S8.1, 691; S8.2, 691 Bignotti Stefano, P275, 800 Binder Elisabeth, O11.4, 710; P08, 740; P10, 740 Binder Elisabeth B., P06, 739 Birkas-Kovacs Dezso˜, P297, 805 Birley Andrew J., O5.5, 702 Bjarkam Carsten R., O23.1, 726 Black D.W., S5.4, 689 Blackwood D., S5.2., 689 Blackwood Douglas, O23.2, 727; P44, 748; P93, 759; P100, 761; P247, 794 Blackwood Douglas H., O23.1, 726; P39, 747 Blackwood Douglas H.R., P83, 757; P138, 769; P139, 769 Blair Ian P., O2.3, 695 Blakely Randy, O8.2, 706 Blanco L., P53, 750 Blasi F., O28.3, 736 Blasi Francesca, P130, 767 Blechingberg Jenny, O23.1, 726 Bleich Stefan, O18.5, 721; P171, 777 Boaretto Francesca, P52, 750 Bocchetta Alberto, P50, 749 Bocchio-Chiavetto Luisella, P273, 799 Boehnke M., S3.3, 688 Boehnke Michael, O2.1, 695 Boer J.A. den, P86, 757 Bogaert Ann Van Den, O13.6, 713 Boisvert K., P85, 757 Bolton Andrew, O25.1, 731 Bolton Patrick, P127, 767 Bolton Patrick F., P89, 758 Bonaguro Russell, O23.2, 727 Bonati MT, O24.9, 730 Bonaviri Giuseppina, O12.4, 712 Bonetto Chiara, O25.6, 732; P276, 800 Bongiorno Fanny, P41, 747 Bonner Tom I., O2.7, 696 Bo¨nsch Dominikus, O18.5, 721; P171, 777 Bonvicini Cristian, P273, 799; P274, 800; P275, 800 Booij Jan, O7.6, 705 Boomsma D.I., P03, 739 Boomsma Dorret, O22.8, 726 Boomsma Dorret I., O9.5, 709; O22.7, 725 Boot Erik, O7.6, 705 Borges Luı´s, P91, 759 Børglum A.D., P53, 750 Børglum Anders, P77, 755 Børglum Anders D., O23.1, 726; P24, 744 Borkowska Alina, O7.8, 706; P62, 752; P64, 752; P65, 753; P252, 795 Borrmann-Hassenbach Margitta B, O11.5, 711; O16.7, 718 Bortolato M., O29.7, 738 Bortolato Marco, P193, 782 Bosaia Alessandra, O18.4, 720 Bosch Rosa, P80, 756 Botteron Kelly, P21, 743 Bourgain C., S7.1, 690 Boutros Nashaat, P154, 773 Bouvier Michel, P101, 761 Boyd P.R., S6.4, 690 Boyle John, P19, 742 Bozina Nada, P218, 787; P222, 788; P255, 795 Bru¨ggemann Daniel, P84, 757 Bran˜as Antı´a, P283, 802 Brais B., P85, 757 Bramon Elvira, P152, 772 Braun Andreas, O5.5, 702 Bray Nick, O21.3, 723; 779; P184, 780
Bree Marianne van den, O21.3, 723; O24.2, 729 Breen G., S6.4, 690 Breen Gerome, O8.1, 706; O15.1, 715; P40, 747; P84, 757; P157, 773; P308, 808 Breskva Katjar, P235, 791 Brewster S., S6.3, 690; S6.4, 690 Brewster Shyama, P315, 809 Brezo Jelena, P04, 739 Brien, P298, 805 Briley Linda, P213, 786 Brink Wim van den, P60, 751 Broeckhoven Christine Van, O2.4, 696 Bronson Paula G., O12.5, 712 Brookes Keeley, O17.1, 719; O28.5, 736; O28.6, 736; P67, 753; P117, 765 Brooks Janet, P143, 770 Brooks-Wilson Angela R., O23.2, 727 Brouwers Jacobus R.B.J., P230, 790 Browning-Large Carrie, O12.5, 712 Brueckl Tanja, P08, 740 Bruggeman Richard, P230, 790 Bruining Hilgo, P190, 781 Brunner H., P116, 765 Brunso Lucas, P80, 756 Bryden Leslie, O21.2, 723 Bubb Jill, P137, 769 Buccola N.G., S5.4, 689 Bucholz K., S8.1, 691 Buck Kari, P195, 782 Budde Laura, P315, 809 Buervenich Silvia, O22.3, 724 Buitelaar J.K., P135, 769 Buitelaar Jan, O17.1, 719; O28.5, 736; O28.6, 736; P67, 753; P117, 765 Buitelaar Jan K., O14.6, 715; P02, 739 Bulgin Natalie, O7.7, 706 Bundo Miki, P170, 777 Bunney W., S1.3, 687 Bunney William E., O2.1, 695 Buran Fa´bio, P37, 746 Burcescu Irina, O24.4, 729; O24.6, 730 Burdick Katherine, O7.1, 704; O13.7, 714; O25.3, 731; P54, 750 Buresi C., O26.2, 733; O26.4, 733 Buresi Catherine, O22.4, 725; O26.6, 734; P29, 745 Burgoyne Paul, O9.1, 708 Burmeister M., S3.3, 688 Burmeister Margit, O2.1, 695; P55, 750 Burmistrova Olga, P294, 804 Burrell G., S5.4, 689 Burrmann Doreen, O9.1, 708 Busatto Geraldo, P37, 746 Buschgens Cathelijne, O14.6, 715 Butcher Lee M., O6.4, 703; O6.5, 703; O18.2, 720; O20.2, 722 Butler H., O28.3, 736 Butler Helen, P130, 767 Butler L., S2.2, 687 Butt A., P107, 762 Butt Alex, P269, 798; P279, 801; P291, 804 Butt Alexander, P161, 774; P257, 796; P258, 796; P270, 799 Buxbaum Joseph D., P266, 798; P131, 768; P290, 803 Byerley W.F., S5.4, 689 Byerley William, O13.1, 712; P32, 745 Byers Helen, O19.5, 722 Cabranes Jose´ Antonio, P151, 772; P202, 784 Cader Z., P85, 757 Caesar Sian, P12, 741; P17, 742 Caffo Ernesto, O24.9, 730 Cai Guiquing, P131, 768 Caixeiro Nicole, O3.5, 698
Calati Raffaella, P113, 764 Callaerts Patrick, P79, 756 Callicott Joseph, O3.3, 697; O5.4, 702 Caltagirone Carlo, O12.4, 712 Camarena Beatriz, P263, 797; P281, 801 Camarena Betty, P13, 741 Cameron Vicky, P168, 776 Camila Guindalini, P308, 808 Campbell Linda E., P136, 769 Campos Sara, O8.1, 706 Can˜ive J., P263, 797; P281, 801 Canive Jose M., P163, 775 Cannon Tyrone D., O23.10, 728 Capozzo Marianna, P219, 787 Carbonetto Savatore, P101, 761 Carboni Lucia, P196, 782 Cardno Alastair G., P316, 809 Cardona Imer, O2.7, 696 Carli V., O26.5, 733 Carlino Davide, P81, 756 Carnell Susan, O18.2, 720 Carola Valeria, P23, 743 Carone Simona, P130, 767 Carreras Anna, P144, 770 Carroll Liam, P261, 797 Casas Miquel, P80, 756 Cascia C. La, P107, 762 Cascia Caterina La, P161, 774; P258, 796 Cascorbi Ingolf, O19.2, 721 Caspi Ashalom, P257, 796 Cassidy Fiona, P34, 746; P35, 746 Castermans Dries, P75, 755 Catala´n Rosa, P233, 791 Catalano Marco, P41, 747; P204, 784; P221, 788 Cate-Carter Tasha, O24.3, 729 Caterina Chillotti, P175, 778 Cath D., P86, 757 Cath D.C., P03, 739 Cath Danielle Caroline, O9.5, 709 Cavalleri Florinda, O24.9, 730 Cavallini Maria Cristina, O18.4, 720; P158, 774 Cavedini Paolo, P158, 774 Cecile A., O1.1, 693 Cervilla J., P264, 797 Cervilla Jorge, O1.6, 694; P05, 739; P78, 756; P177, 778; P283, 802 Cesana Bruno Mario, P273, 799 Cetin Mesut, P214, 786 Chan R.C.K., P162, 775 Chananashvili M., P45, 748 Chang Chee-Jen, P155, 773 Chantal Sophie, P134, 768 Charles M.K., P85, 757 Charney Dennis, O22.3, 724 Chaste Pauline, P131, 768 Chavarria-Siles Ivan, P66, 753 Chavira Denise, O15.3, 716 Chen Chia-Hsiang, P286, 802 Chen E.Y.H., P162, 775 Chen Jasmine Ying-Jiun, O29.1, 737 Chen Jianhuan, P278, 800 Chen Ping, P07, 740 Chen Qi, P246, 793 Chen Shih-Fen, P286, 802 Chen W., S2.2, 687 Chen Wai, O17.1, 719; O28.5, 736; P67, 753; P117, 765 Chen Wei, P162, 775 Chen Xiangning, O3.8, 699; P98, 760; P246, 793; P295, 804 Chen Zugen, P143, 770 Cheng Min-Chih, P286, 802 Cherchi A., P208, 785 Chillotti C., P209, 785 Chillotti Caterina, P56, 750; P167, 776
Author Index Chissoe S., S6.3, 690 Chissoe Stephanie, P315, 809 Cho Eun-Young, P254, 795 Chorbov Vesselin, P21, 743; P92, 759 Choudary Prabhakara, O2.1, 695 Choudhury Khalid, P42, 747; P49, 749; P189, 781; P293, 804 Choufani Sanaa, O14.1, 714 Chouinard S., P85, 757 Chow Eva, P299, 805 Chow P., P107, 762 Chowdari K.V., O3.6, 698 Chowdari Kodavali V., O3.8, 699 Christoforou Andrea, P39, 747; P93, 759; P100, 761 Chubb Jennifer E., P300, 806 Chuchalin A., P45, 748 Cichon S., P45, 748; W1, 685 Cichon Sven, O2.6, 696; O13.2, 712; O23.4, 727; O25.4, 731; P95, 760; P97, 760; P252, 795; P262, 797 Cid Rafael de, O22.10, 726; O22.2, 724; P144, 770 Ciwoniuk Magdalena, P217, 787 Claes Stephan, O1.1, 693; O2.4, 696 Clair David St, P194, 782 Claire David St, P40, 747 Clark Deborah, P33, 746 Clarke Sarah, O7.2, 704 Clifford D., P107, 762 Clifford Damien, P257, 796; P258, 796 Cliford Damian, P270, 799 Cloninger C.R., S5.4, 689 Coccarro Emil, O23.2, 727 Cochrane Lynne Elizabeth, P122, 766 Coffey Marge, P143, 770 Cogliati Francesca, O24.9, 730 Cohen S., P107, 762 Cohen Sarah, O1.4, 694; O22.5, 725 Collier D.A., P107, 762; P162, 775; P302, 806 Collier Davd, O8.1, 706; P269, 798 Collier David, O25.6, 732; P40, 747; P104, 762; P152, 772; P157, 773; P160, 774; P164, 775; P257, 796; P258, 796; P270, 799; P276, 800; P279, 801; P291, 804; P308, 808 Collier David A., O8.1, 706; O15.1, 715; P276, 800 Collins Scott, P129, 767 Collu M., P200, 783 Comas David, P296, 805 Congiu D., P57, 751; P208, 785; P209, 785 Congiu Donatella, P56, 750 Conroy Judith, P120, 766; P122, 766; P91, 759 Consortium IMGSAC, P130, 767 Conte Maria Anna, P219, 787 Contreras Javier, P13, 741 Contreras Salvador, P13, 741; P66, 753 Coon Hilary, O15.2, 716 Cooper Stephen, P226, 789 Cope Heidi, P99, 761 Cope Natalie, O6.8, 704 Corda Maria G., O29.2, 737; P197, 782 Cordeiro Quirino, P37, 746; P159, 774; P223, 788; P308, 808 Cordwell Stuart, O3.5, 698 Cormand Bru, P80, 756 Corona Winston, O2.7, 696 Correa Humberto, P16, 742; P46, 748; P285, 802 Corregiari Fabio, P159, 774; P223, 788 Correia Catarina Catarina, P91, 759 Correia Catarina, O28.4, 736 Corsini Paola, P273, 799 Corvin Aiden P., O29.3, 737 Corvin Aiden, O13.2, 712; O7.2, 704 Corydon Thomas J., O23.1, 726
Coryell William, O13.1, 712 Cossette P., P85, 757 Cossette Patrick, P101, 761 Courtet Ph., O26.2, 733; O26.4, 733 Courtet Philippe, O22.4, 725; O26.6, 734 Coutinho Ana Margarida, O28.4, 736; P91, 759 Couto Jillian M, O24.3, 729 Coventry Will, O1.5, 694 Cox Charles, P213, 786 Cox David R., O1.2, 693 Cox Diane, P268, 798 Cozza Arianna, O23.8, 728 Crabbe John, P195, 782 Craddock N., S3.1, 688; S6.4, 690 Craddock Nicholas J., O13.2, 712 Craddock Nick, O1.4, 694; O2.8, 697; O21.3, 723; O22.5, 725; P12, 741; P14, 741; P15, 741; P17, 742; P266, 798; P290, 803 Craig I., P103, 761; P107, 762; S6.2, 690 Craig Ian, O1.4, 694; O22.5, 725; O26.3, 733; P12, 741; P17, 742 Craig Ian W., O6.5, 703; O20.2, 722; P90, 758 Craig I.W., P211, 786 Creemers John, P75, 755 Crespo Jose M., O22.2, 724 Cristina Cavallini Maria, P158, 774 Cristofalo Doriana, O25.6, 732; P276, 800 Croteau Pascal, O16.5, 718 Crowe R.R., S5.4, 689 Crowe Raymond R., O1.3, 694 Cruz Carlos, P25, 744 Cseh Ka´roly, P297, 805 Cuccaro M.L., O20.4, 723 Cuccaro Michael, P99, 761; P129, 767; P148, 771 Cuesta Manuel J., P296, 805 Cui Huxing, P88, 758 Cunha Nadia, P308, 808 Cuomo C., O26.5, 733 Currais Anto´nio, O28.4, 736 Curran Sarah, P89, 758; P127, 767 Curtis Charles, O6.5, 703 Curtis C.J.C., P211, 786 Curtis David, P42, 747; P49, 749; P293, 804 Curtis L., P107, 762 Curtis Logo, P269, 798 Curtis Logos, P161, 774; P164, 775; P291, 804 Czerski Piotr, P239, 792; P244, 793 Czerski Piotr M., P252, 795 Czobor Pal, O19.3, 721 Daalen Emma van, O6.2, 703 Dahl Hans A., P24, 744 Dahl Hans Atli, P20, 743 Da´ibhis Aoife, O24.7, 730 Dain Bradley, O19.2, 721 Daly Mark J., O27.4, 735 Daly Michael, O24.7, 730 Damberg Mattias, O5.2, 701 Daro´czy Gabriella, O24.8, 730 Dassori Albana, O16.8, 718; P13, 741; P66, 753 Datta Susmita R., P293, 804 David Collier, P161, 774 David Goldstein, O8.3, 707 Davies W., S9.3, 692 Davies William, O9.1, 708 Davis Kenneth L., P266, 798; P290, 803 Dawn, P298, 805 Dazzan P., P107, 762; P87, 758 Dazzan Paola, O25.6, 732; P161, 774; P257, 796; P258, 796; P269, 798; P270, 799; P276, 800; P279, 801; P291, 804 De Ronchi Diana, P236, 791 Deakin J.F. William, P137, 769
813
Deckert Ju¨rgen, P71, 754 Dedova Irina, O3.5, 698 Del-Favero Jurgen, O1.1, 693; O2.4, 696; O13.5, 713; O13.6, 713; O27.3, 734; P79, 756 Delaney Catherine, P34, 746; P35, 746 Delawary Mina, P191, 781 DeLisi Lynn E., O11.2, 710 Del Zompo M., P208, 785 deMarchi Nicola, O3.1, 697 Demetrovics Zsolt, O8.6, 707; P309, 808 Demidova Irina, P140, 770; P141, 770 Demjaha Arsime, P157, 773 Dempster David, P157, 773 Dempster Emma, O24.4, 729; P104, 762; P152, 772 Deng Wei, P160, 774; P162, 775; P302, 806 Denmark Deaunne, P195, 782 Dennis Mc¸gan, O6.8, 704; O11.1, 710 DePaulo J. Raymond, O13.1, 712 DePaulo Jr., O1.3, 694 Depner Martin, P07, 740 der–Beek van der A. Aart van, P210, 785 Dermot Walsh, O3.8, 699 Derom Catherine, O22.8, 726 DeRosse Pamela, O7.1, 704; O13.7, 714; O25.3, 731; P54, 750 DeSanctis Pierfilipo, O7.2, 704 Deschner Monika, O7.3, 705; P84, 757; P95, 760; P262, 797 Desnica Natasa, P27, 744; P311, 808 Desrivieres Sylvane, P181, 779 Detera-Wadleigh Sevilla D., O2.7, 696 Dettling Michael, O19.2, 721 Deveau Todd, O11.4, 710 Devlin Bernie, O3.6, 698 Devriendt Koen, P75, 755 Diaz Guillermo Orozco, O2.6, 696; O13.2, 712 Dick D., S8.1, 691 Didden William, O23.7, 728; O24.5, 729 DiForti Marta, O25.6, 732; P157, 773; P164, 775; P270, 799; P276, 800 Dikeos Dimitris, O3.1, 697 Dinnbier Matthias, O21.4, 724 Distel Marijn, O22.8, 726 Djelic Ninoslav, O12.1, 711 Dmitrzak Monika, P244, 793 Dmitrzak-Weˆglarz Monika, P62, 752 Dmitrzak-Weglarz Monika, O18.1, 720; P30, 745; P61, 752; P64, 752; P65, 753; P239, 792; P240, 792 Dobrin Seth, P34, 746; P35, 746 Doherty Jeniffer, P226, 789 Dolzˇan Vita, P235, 791 Domani, Martina, O21.4, 724 Domenici E., S6.3, 690 Domenici Enrico, P196, 782; P231, 790; P315, 809 Domotor Eszter, P11, 741 Donald Jennifer A., O2.3, 695 Dong Li, P181, 779; P305, 807 Donohoe Gary, O7.2, 704 Dorland Bert, O5.3, 701 Dose Tatjana, P09, 740 Dotoli Danilo, P41, 747 Dow David, P213, 786 Doyle Adam, O16.7, 718 Drapeau Pierre, P101, 761 Drtı´lkova´ Ivana, O23.7, 728; O24.5, 729 D’Souza Ursula M., O19.5, 722; P90, 758 Duan J., S5.4, 689 Duan Jubao, O3.1, 697 Dube Marie-Pierre, P101, 761 Ducci Francesca, P73, 754 Dudbridge Frank, P123, 766 Duffy Anne, O13.3, 713 Duijn Cornelia van, O2.4, 696
814
Author Index
Duran Fa´bio, P37, 746 Dworzynski Katharina, P127, 767 Dwyer Sarah, P260, 796; P261, 797; P287, 803 Dyck R. Van, O11.3, 710 Eaves Lindon, O1.5, 694; O7.5, 705 Ebers G., P85, 757 Ebstein Richard P., O28.6, 736; P146, 771 Ebstein Richard, O18.3, 720; O28.5, 736; P67, 753; P117, 765 Ebstein Richard S., O17.1, 719 Eccles Mike, P168, 776 Edenberg H., S8.1, 691 Edgar J. Christopher, P163, 775 Edwards Todd, O17.3, 719 Eerdewegh Paul Van, O16.5, 718 Efanova N., P280, 801 Efanova Nataly, P271, 799 Egan Michael, O3.3, 697; O5.4, 702 Egberts A.C.G., P210, 785 Eichhammer Peter, O1.8, 695; O11.5, 711; O21.5, 724; P109, 763; P111, 763 Eiji Hattori, O29.4, 737 Eisen Seth A., P18, 742 Eisen Seth, O7.5, 705 Ekawardhani Savira, P82, 757; P256, 795 Ekelund Jesper, O7.4, 705; O16.2, 717 El-Giamal Nadja, P145, 771 Elliot S. Gershon, O29.4, 737 Emanuel Beverly, O5.3, 701 Emanuel Beverly S., O3.4, 698 Encinas Marta, P202, 784 Ende Gabriele, O7.3, 705 Engeland Herman van, O5.3, 701; O6.2, 703; P190, 781 English Brett, O8.2, 706 Epstein Jeff, P148, 771 Erdal M. Emin, P214, 786 Erhardt Angelika, P06, 739; P10, 740 Ericsson O., S10.1, 693 Eriksson Elias, P76, 755 Ervin Frank, O4.5, 700 Erzegovesi Stefano, O18.4, 720 Escamilla Michael, O16.8, 718; P13, 741; P163, 775; P263, 797; P281, 801; P66, 753 Estivill Xavier, O5.1, 701; O8.5, 707; O22.2, 724; O22.10, 726; P144, 770 Etain B., P31, 745 Etain Bruno, P29, 745 Ettinger Ulrich, P150, 772 Eujen Andrea, P70, 754 Evans Alan, P132, 768 Evans Kathryn, P100, 761 Evans Kathryn L., P39, 747; P83, 757 Evans Kathy, P93, 759 Evans Simon J., O2.1, 695 Evgrafov Oleg, O1.3, 694 Ewald H., P53, 750 Ewald Henrik, P20, 743 Exelbierd Linda, P129, 767; P148, 771 Eyeson J., P107, 762 Eyeson Joanna, P161, 774; P164, 775; P257, 796; P258, 796; P269, 798; P270, 799; P279, 801; P291, 804 Ezzatian Payam, O14.3, 714 Fa` M., O29.7, 738 Fa Mauro, P193, 782 Fan˜ana´s Lourdes, P108, 763; P110, 763; P198, 783; P233, 791; P296, 805 Fabienne Wavrant-De Vrie`ze, P143, 770 Fadda P., P200, 783 Faeron Paul, P270, 799 Fagerness Jes, O22.6, 725 Fagerness Jesen, O28.2, 735
Fairbanks Lynn, P178, 778 Fallon J., S1.3, 687 Faludi Gabor, P11, 741; P297, 805 Fann Cathy S-J, P248, 794; P155, 773; P284, 802 Fanous Ayman, O16.3, 717; P295, 804 Faraone Stephen, O28.6, 736; O4.1, 699; P253, 795 Faraone Steve, O17.1, 719; O28.5, 736; P67, 753; P117, 765 Farmer A., P103, 761; S6.2, 690 Farmer A.E., S6.4, 690 Farmer Anne, O1.4, 694; O22.5, 725; O26.3, 733; P12, 741; P17, 742 Fatjo´-Vilas Mar, P296, 805 Favre Sophie, P29, 745 Fearon P., P107, 762 Fearon Paul, P161, 774; P270, 799; P279, 801; P291, 804 Featherstone Katie, O10.1, 709 Fedorova E., P280, 801 Feng GuoYin, O20.1, 722 Feng Jian, P199, 783 Feng Yu, O24.8, 730; P68, 753 Fentress Hugh, O8.2, 706 Ferrı´n M, P283, 802 Ferraro Thomas N., P245, 793 Ferrero Franc¸ois, P29, 745 Ferri C., P206, 784 Ferrier Nicol, P12, 741; P17, 742 Filbey Francesca, P310, 808 Finan Patrick, P310, 808 Finckh U., O4.9, 701 Fischer Birgit, P109, 763; P111, 763 Fisher Helen, P291, 804 Fitches Alison, P33, 746 FitzGerald David, P129, 767 Fitzgerald David, P148, 771 Fitzgerald M., S2.2, 687 Flaquer Antonia, O2.6, 696 Flint T.J., P53, 750 Flint Tracey J., P24, 744 Florio Arianna Di, O17.4, 719 Floris Gianfranco, O8.1, 706 Fogli Stefano, O23.8, 728 Foldager Leslie, O23.1, 726; P249, 794 Fombonne Eric, P101, 761 Foncke E.M.J., P86, 757 Foot Elizabeth, P213, 786 Forero Diego, P79, 756 Forloni Gianluigi, P207, 785; P314, 809 Foroud T., S8.1, 691 Forsling Mary, P227, 789 Forti M. Di, P87, 758; P107, 762; P161, 774; P257, 796; P269, 798; P276, 800; P279, 801; P291, 804 Forty Liz, P14, 741 Fournier Helene, O16.5, 718 Fowler Tom A., O6.1, 703 Fowler Tom Alan, O24.2, 729 Foxe John, O7.2, 704 Franca Ligas, P175, 778 Francis McMahon, O22.3, 724 Frangou Sophia, P152, 772 Frank Greg, O5.6, 702 Frank Josef, P07, 740 Franke Barbara, P02, 739; P115, 764; P116, 765; P210, 785 Franke Lude, O6.2, 703 Franz Carol, P18, 742 Fraser Christine, P12, 741; P17, 742 Fratta W., P200, 783 Frau R., O29.7, 738 Frau Roberto, P193, 782 Frazer Kelly A., O1.2, 693 Frazzetto Giovanni, P23, 743 Freedman R., S5.4, 689
Freeman Bernard, P257, 796; P258, 796; P270, 799 Freitag Christine M., P82, 757 Freitag Christine, O24.1, 729; P70, 754 Fresan Ana, P263, 797; P281, 801 Freson Kathleen, P75, 755 Friedman L., S1.3, 687 Friedman Lee, O27.5, 735 Frieling Helge, O18.5, 721; P171, 777 Fu Cindy, P157, 773 Fuchang Zhang, P59, 751 Fuchs Karoline, P145, 771; P48, 749 Fuentes L.E., P206, 784 Fujisawa Hajime, O29.3, 737 Fukuda Masato, P165, 775 Fukumaki Yasuyuki, P265, 798 Fukunaka Yuko, P88, 803; P225, 789; P229, 790; P232, 790; P234, 791; P292, 804 Fukutake Masaaki, P176, 778; P88, 758 Fullerton Jan M., O2.3, 695 Funke Birgit, O13.7, 714; P54, 750 Funt Sam, P73, 754 Fu¨rst Robert, O2.6, 696; O25.4, 731 Gabryelewicz Tomasz, P205, 784 Gadoros Julia, O24.6, 730; P149, 772; P212, 786 Gaetano M. De, O26.5, 733 Gafoor R., P107, 762 Gafoor Raffael, P257, 796; P258, 796 Gagne´ Bernard, O2.2, 695; P08, 740 Galea Sandro, P19, 742 Galina Pungercic, P72, 754 Gallagher Louise, P91, 759; P120, 766; P122, 766 Garavan Hugh, O7.2, 704 Garcı´a Cecilia, O8.5, 707 Garcı´a Ester, P80, 756 Garcı´a Luisa, O8.5, 707 Gardella Rita, P274, 800 Gardner Michelle, P296, 805 Garrido Helena, O15.3, 716 Garvey Sean M., O12.5, 712 Gasto´ Cristo´bal, P233, 791 Gauthier Julie, P101, 761 Gawlik Micha, P278, 800 Gaysina Daria, O22.5, 725 Gaysina Dasha, O1.4, 694 Gayzina D., P107, 762 Geet Christel Ven, P75, 755 Gejman P.V., S5.4, 689 Gejman Pablo V., O3.1, 697; P301, 806 Gelabert Estel, O8.5, 707 Gelernter Joel, P19, 742 Geller Frank, O4.7, 700; P27, 744; P311, 808 Gennarelli Massimo, P273, 799; P274, 800; P275, 800 Genogu F., O29.7, 738 Gentil Valentim, P105, 762 Genugu Fabrizio, P306, 807 Georgi A., P45, 748 Georgi Alexander, O23.4, 727; O25.4, 731; P84, 757; P95, 760; P252, 795; P262, 797 Georgieva Lyudmila, P266, 798; P290, 803 Gerrish Amy, P289, 803 Gershon E.S., PL4, 685 Gershon Elliot S., O2.7, 696; O13.1, 712 Gerstein Mark, O3.4, 698 Gervai Judit, P149, 772; P212, 786 Geschwind D., S2.1, 687 Gessi Alessandra, O24.9, 730 Gettig Elizabeth, P174, 777 Geyer Mark, O13.4, 713 Giannios Dimitrios, P125, 766 Gibbs Richard, O2.7, 696
Author Index Giedd Jay, P132, 768 Giegling Ina, O26.7, 734; P112, 764; P113, 764 Gil Pedro, P202, 784 Gilbert J.R., O20.4, 723 Gilbert John R., O12.5, 712 Gilbert John, P129, 767; P148, 771 Gilks William P., O29.3, 737 Gill M., S2.2, 687; S6.4, 690 Gill Michael, O7.2, 704; O13.2, 712; O17.1, 719; O24.7, 730; O28.5, 736; O28.6, 736; O29.3, 737; P58, 751; P67, 753; P91, 759; P117, 765; P120, 766; P121, 766; P122, 766 Gillberg Christopher, P131, 768 Gillespie Nathan, O1.5, 694 Gingrich Jay, O29.1, 737 Giorgi Osvaldo, O29.2, 737; P197, 782 Giraldi Tullio, P219, 787; P81, 756 Gironcoli Marzia De, P276, 800 Gladis Madeline, O1.3, 694 Glassee Wim, O27.3, 734 Glatt Stephen J., P102, 761 Glatt Stephen, P253, 795 Goate A., S8.1, 691 Gochman Peter, P143, 770 Godard Beatrice, P101, 761 Godlewska Beata, P215, 787; P217, 787; P252, 795 Goes Fernando, O13.1, 712 Gogtay Nitin, P143, 770 Gokdogan Tuba, P214, 786 Goldberg Terry, O3.3, 697; P54, 750 Goldman David, O7.1, 704; O25.3, 731; P73, 754 Goldowitz Daniel, O12.3, 711 Goldsmith Juliet, P131, 768 Goldsmith Michael, O15.2, 716 Goljevscek Serena, P81, 756 Goltsov Andrey, P294, 804 Gomez Lisette, O24.3, 729 Gomez Lissette, O22.1, 724 Go´mez M., P206, 784 Go`mez Nuria, P80, 756 Gong Qi-yong, P160, 774 Gonza´lez Juan R., O5.1, 701; O22.2, 724; O22.10, 726 Gonza´lez Juan Ramon, P144, 770 Gonza´lez L., P53, 750 Gonza´lez-Sobrino Blanca Zoila, P25, 744 Gonza´lez-Zuloeta Ladd Angela M., O1.1, 693 Goodchild Rose, O29.1, 737 Goodgame Neil, P213, 786 Goos Lisa, O14.3, 714 Goossens Dirk, O27.3, 734 Gorbachevskaya Natalia, P141, 770 Gordo Estı´baliz, P283, 802 Gordon-Smith Katherine, P15, 741 Gornick Michele, P143, 770 Goto Hiroki, P265, 798 Gottesman Irving, O9.3, 708 Goula Vasiliki, P125, 766 Gozner Andreas, O18.5, 721 Grant Michael D., P18, 742 Grassi Massimiliano, O18.4, 720 Grataco´s Mo´nica, O5.1, 701; O8.5, 707; O22.2, 724; O22.10, 726; P144, 770 Green Elaine, O2.8, 697; P12, 741; P17, 742 Green Eric D., O11.1, 710 Green Eric, O6.8, 704 Green Roland, O3.4, 698 Greene Ciara, P58, 751 Greengard Paul, P199, 783 Greenstein Deanna, P143, 770 Greenwood Tiffany, P147, 771 Griffiths P.E., PL3, 685 Gritsenko Inga, P146, 771 Grof P., P85, 757
Grof Paul, O13.3, 713 Grootheest Daniel S. van, O9.5, 709 Grootheest Daniel Sebastiaan van, P03, 739 Gross Cornelius, P23, 743 Grotewiel Michael, O4.3, 699 Grozeva Detelina, O2.8, 697; P12, 741; P17, 742 Gruber Susanne, P196, 782 Grubert Fabian, O3.4, 698 Guan W., S3.3, 688 Guan Weihua, O2.1, 695 Gubskij L.V., P272, 799 Gubsky L., P280, 801 Gubsky Leonid, O25.5, 732; P271, 799 Gudfinnson Einar, O16.4, 718 Gueckel Eva, P169, 776 Gulcher Jeffrey R., P311, 808 Gulcher Jeffrey, O4.7, 700; O16.4, 718; P27, 744 Gullberg M., S10.1, 693 Gulli Rossella, P275, 800 Gunawardane Nisali, O13.7, 714 Guo Shengzhen, O13.4, 713 Guoyin Feng, P59, 751 Gupta N., P85, 757 Gurling Hugh, O29.5, 738; O29.6, 738; P42, 747; P49, 749 Gurling Hugh M., P189, 781 Gurling Hugh M.D., P293, 804 Gustafsdottir S., S10.1, 693 Gutierrez B., P264, 797 Gutie´rrez Blanca, O1.6, 694; P05, 739; P177, 778; P233, 791; P283, 802; P78, 756 Ha Kyoo-Seob, P254, 795 Haan Lieuwe de, O7.6, 705 Hacker April, O3.4, 698 Haddad Stephen, O28.2, 735 Haddley Kate, P137, 769 Hadithy Asmar F.Y. Al, P230, 790 Haghighi Pejmun, P101, 761 Hahn Maureen, O8.2, 706 Haines J.L., O20.4, 723 Haines Johnathan L., O12.5, 712 Hajak Go¨ran, P106, 762; P109, 763; P111, 763 Hajek Tomas, O13.3, 713 Hakansson Kerstin, O29.3, 737 Hallman J., O23.6, 727 Hallmayer Joachim, P51, 749 Haman Kirsten, O8.2, 706 Hamdand Adnan, O20.5, 723 Hamshere Marian L, O13.2, 712 Haraldsson Magnus, P150, 772 Harano Mutsuo, P243, 793; P278, 800; P303, 806; P304, 807 Harding Boris, O27.3, 734 Hardy John, O21.2, 723; P143, 770 Hariri A.R., S1.2, 686 Harlaar Nicole, O6.4, 703; O6.5, 703; P118, 765 Harold Denise, O6.8, 704 Haroutunian Vahram, P266, 798; P290, 803 Harten P.N. van, P230, 790 Hartman C.A., P135, 769 Hartmann Annette M., O26.7, 734; P112, 764; P113, 764 Harvey Mario, O2.2, 695; P08, 740; P134, 768 Harvey Philip D., O11.2, 710 Hashida Aoi, P133, 768; P182, 779 Hashimoto Ryota, P26, 744 Hashimoto Takeshi, P176, 778 HATA Takanori, P251, 794 Haughey Heather, P310, 808 Hauser J., P103, 761; S6.2, 690 Hauser Joanna, O7.8, 706; O18.1, 720; P30, 745; P61, 752; P62, 752; P64, 752; P65,
815
753; P173, 777; P239, 792; P240, 792; P242, 793; P244, 793; P252, 795 Hawi Ziarah, O28.6, 736; P58, 751 Hawi Ziarih, O24.7, 730; P120, 766; P121, 766 Haworth Claire, O18.2, 720 Hayesmoor Jesse, P184, 780 Hayiou-Thomas Emma, O14.4, 714 Haynes C., O20.4, 723 Hazelwood Lisa, O8.2, 706; O23.3, 727 He D., S5.4, 689 He Lin, O13.4, 713; O20.1, 722 Heath Andrew, P21, 743; P154, 773 He´bert Martine, P04, 739 Hejjas Krisztina, P11, 741 Helgason A., S7.2, 691 Hellard Stephanie Le, P39, 747 Hellmann J.D., O4.9, 701 Hemmings Sian, P76, 755 Henigsberg N., P103, 761; S6.2, 690 Henn Fritz A., O7.3, 705 Hennah William, O2.5, 696; O7.4, 705; O16.2, 717; O23.10, 728 Henriksson S., S10.1, 693 Henry Chantal, P31, 745 Herken Hasan, P214, 786 Herna´ndez Y., P206, 784 Herteleer Liesbet, P79, 756 Hesselbrock V., S8.1, 691 Hettema John, P98, 760 Heutink P., P86, 757 Heyrman Lien, O13.5, 713; O13.6, 713; O27.3, 734 Hicks Andrew, O4.7, 700; P27, 744 Hicks Andrew A., P311, 808 Hicks G., W2, 686 Higuchi Susumu, P192, 781 Hill Gary, O6.8, 704 Hillemacher Thomas, O18.5, 721 Hillmer Axel, O25.4, 731 Hinds David A., O1.2, 693 Hiroki Kikuyama, P238, 792 Hiroshi Kunugi, P26, 744 Hiroshi Yoneda, P238, 792 Hochstenbach Ron, O6.2, 703 Hoda F., P103, 761; P107, 762; S6.2, 690 Hoda Farzana, O1.4, 694; O22.5, 725 Hodgins Sheilagh, O23.2, 727 Hodgkinson Colin, O7.1, 704 Hoefels Susanne, O23.4, 727; P97, 760 Hoek Hans W., P230, 790 Hoekstra P.J., P135, 769 Ho¨fels Susanne, P95, 760; P262, 797 Hoffmann Katalin, O8.6, 707 Hoffmann Per, O25.4, 731 Hohoff Christa, P71, 754 Hol F.A., P116, 765 Holm Ida E., O23.1, 726 Holmans Pete, O21.3, 723 Holmans Peter, P47, 749 Holmans Peter A., O1.3, 694; O3.1, 697; O13.2, 712 Holsboer F., S6.3, 690 Holsboer Florian, P06, 739; P08, 740; P09, 740; P10, 740; P315, 809 Holt David, P228, 789 Holt Robert A., O12.3, 711 Holtzman N.A., S4.3, 688 Honea Robyn, O3.3, 697 Honer William, O7.7, 706 Hong Kyung Sue, P185, 780 Hoogendijk Witte J.G., O11.3, 710 Horan Mike, P137, 769 Hori Hiroaki, P26, 744 Hori Hiroko, P225, 789; P229, 790; P232, 790; P234, 791; P288, 803; P292, 804 Hori Toru, P303, 806; P304, 807 Horike Shin-ichi, O14.1, 714
816
Author Index
Horiuchi Yasue, P180, 779; P192, 781 Hornik Kurt, P48, 749; P145, 771 Horstmann Sonja, P09, 740 Hossain Shaon, O3.8, 699 Hottenga J.J., P03, 739 Hottenga Jouke Jan, O22.7, 725 Hou C., S5.4, 689 Houkyou Akira, P241, 792 Houlihan Lorna, P83, 757 Howell M., S10.1, 693 Hsu Amy C.L., P155, 773 Hsu Pei-Chun, P248, 794; P284, 802 Hu Junmei, O15.1, 715 Huang Ke, O20.1, 722 Huang Mingxiong, P163, 775 Huentelman Matthew J., O27.2, 734 Huezo-Diaz P., P107, 762; S6.2, 690 Huezo-Diaz Patricia, P103, 761; P227, 789 Humphries Tom, O24.3, 729 Hutchison Kent, P224, 789; P310, 808 Hwang Rudi, O19.3, 721; P234, 791 Hwu Hai Kwo, P155, 773 Hwu Hai-Gwo, P248, 794; P284, 802 Hyde Craig, O1.2, 693 Hyde Sally, P12, 741; P17, 742 Hyoukyou Akira, P201, 783 Iacopetti Paola, O23.8, 728 Iacoviello L., O26.5, 733 Ian Craig W., O20.2, 722 Iba´n˜ez Ignacio, P108, 763 Ibach Bernd, P106, 762 Ignacio Iban˜ez, P110, 763 Ignatius Jaakko, O28.2, 735 Iijima Yoshimi, P26, 744 Iizuka Yukihiko, O5.7, 702 Ikeda Masashi, O8.4, 707; P220, 788 Illig Thomas, P95, 760; P262, 797 Imamura Akira, P133, 768; P182, 779 IMGSAC, O28.3, 736 Imperadori Laura, P274, 800 Inada Toshiya, P220, 788; P303, 806; P304, 807 Ingason Andres, O4.7, 700; O16.4, 718; P27, 744; P311, 808 Ingelman-Sundberg M., P103, 761 Inoue Ken, P165, 775 Iourov Ivan, P140, 770; P141, 770; P142, 770 Irena Zpanic Pajnic, P72, 754 Ishiguro Hiroki, P180, 779; P192, 781 Ishihara Ryoko, P220, 788 Ishiwata Mizuho, P170, 777 Ising Marcus, P06, 739; P08, 740; P09, 740; P10, 740 Isles A., S9.3, 692 Isles Anthony, O9.1, 708 IV John Gilbert, P199, 783 Ivanov Dobril, P179, 779 Ivanova Mina, P51, 749; P92, 759 Ivleva Elena, P88, 758 Iwamoto Kazuya, P170, 777 Iwasaki Shinya, P180, 779; P192, 781 Iwata Nakao, O8.4, 707; P220, 788; P303, 806; P304, 807 Iyo Masaomi, P303, 806; P304, 807 Jablensky A., S5.3, 689 Jablensky Assen, P51, 749 Jacob Christian, P70, 754 Jacobs Nele, O1.7, 694 Jacobson Kristen C., O15.7, 717 Jacobson Kristen, O7.5, 705 Jaeger Judith, O25.3, 731 Jaitovich-Groisman Iris, O13.3, 713 Jakovljevic Miro, P222, 788 Jamain Ste´phane, P31, 745 James Michael, O1.5, 694; O5.5, 702
Jamra R. Abou, P45, 748; O2.6, 696; O23.4, 727; O25.4, 731; P95, 760; P97, 760; P262, 797 Jang Jae Seung, P185, 780 Janssens J.W., O1.1, 693 Janzing Joost, P02, 739 Jarusuraisin Ngamwong, P265, 798 Ja¨rvelin Marjo-Riitta, P94, 759 Jarvelin Marjo-Ritta, O14.5, 715 Jarvius J., S10.1, 693 Jarvius M., S10.1, 693 Jaussent I., O22.4, 725; O26.2, 733 Javitt Daniel C., O7.2, 704 Jaworski J., O20.4, 723 Jay Maurice, O3.1, 697 Jean P. St., S6.3, 690; P315, 809 Jerez A., P263, 797; P281, 801 Jerman Borut, O26.3, 733 Jiang Li-jun, P160, 774 Jianjun Gao, P59, 751 Jime´nez David, P151, 772 Johansson Carolina, P07, 740 Johnson Katherine A, O24.7, 730 Johnson Madeleine, O29.1, 737 Jollant F., O26.2, 733; O26.4, 733 Jollant Fabrice, O22.4, 725; O26.6, 734 Jonathon Badger, P35, 746 Jones Aaron P., P163, 775 Jones Edward G., O2.1, 695 Jones Ian, O2.8, 697; O13.2, 712; P12, 741; P14, 741; P17, 742 Jones L., S6.4, 690 Jones Lesley, O21.3, 723 Jones Lisa, O2.8, 697; P12, 741; P14, 741; P15, 741; P17, 742 Jones P., P107, 762 Jones Steven J., O12.3, 711 Jong S. De, O11.3, 710 Jonsson Thorlakur, O4.7, 700; P27, 744; P311, 808 Joober Ridha, P101, 761 Jorgensen Matthew, P178, 778 Joukamaa Matti, P94, 759 Jovanovic Nikolina, P218, 787; P255, 795 Joyce Peter R., P33, 746; P74, 755; P216, 787 Joyce Peter, P168, 776 Joze Balazic, P72, 754 Jun Koh, P238, 792 Junyent Fe`lix, P198, 783 Jussila Katja, O28.2, 735 Kaczmarkiewicz-Fass Magdalena, P173, 777 Kahn Rene´, O5.3, 701 Kahn S., P166, 776 Kail Melanie, P129, 767; P148, 771 Kaiya Hisanobu, P165, 775 Kakiuchi Chihiro, P170, 777 Kalaydjieva Luba, P51, 749 Kaleem Mona, O21.2, 723 Kalidindi Sridevi, P157, 773 Kalofouti Anastasioss, P207, 785; P314, 809 Kalsi G., O4.2, 699 Kalsi Gursharan, O4.3, 699; P49, 749 Kamali M., S10.1, 693 Kametani Mizue, P187, 780 Kamijima Kunitoshi, P26, 744 Kaminsky Zachary, O9.3, 708 Kanazawa Tetsufumi, P186, 780 Kandaswamy Radhika, P42, 747 Kane Erin, P129, 767; P148, 771 Kane John, O7.1, 704; O19.2, 721; O25.3, 731; P54, 750 Kaneva Radka, P51, 749; P92, 759 Kanyas Kyra, O20.5, 723 Kao Hung-Teh, P199, 783
Kapelski Pawel, P61, 752; P239, 792; P240, 792; P242, 793; P244, 793; P252, 795 Kapiletti S., P45, 748 Kapornai Krisztina, O24.6, 730; O24.8, 730 Kaprio Jaakko, O23.10, 728 Karen Sheehan, P121, 766 Karl Alexandra, O21.4, 724 Karni Osnat, O20.5, 723 Karpushova A., P45, 748 Karpushova Anna, O23.4, 727; P97, 760; P252, 795 Karunas Aleksandra, P183, 779 Kas Martien, P190, 781 Kasai Kiyoto, P237, 791 Kasper Siegfried, P07, 740; P48, 749; P145, 771 Katerberg Hilga, P86, 757 Kato Nobumasa, P237, 791; P38, 747 Kato Tadafumi, P26, 744; P38, 747; P170, 777; P187, 780; P237, 791 Katschke A., S8.1, 691 Katz Elyse, O11.4, 710 Kaufman Charles, O25.1, 731 Kaye Walter, P147, 771 Kazuno An-a, P38, 747 Keator D., S1.3, 687 Keck Martin E., P06, 739; P10, 740 Keith Tim, O16.5, 718 Kejin Zhang, P59, 751 Keller Matthew, O17.2, 719 Kelly Simon P., O24.7, 730 Kelsoe J., W1, 685 Kelsoe John, O2.2, 695; O13.4, 713 Kelsoe John R., O13.1, 712; P22, 743 Kemner Chantal, O5.3, 701 Kempas Elli, P124, 766 Kempf Lucas, O5.4, 702 Kendler K.S., O4.2, 699 Kendler Kenneth, O4.3, 699; O5.6, 702; O16.3, 717; P98, 760; P295, 804 Kendler Kenneth S., O3.1, 697; O3.8, 699; O4.6, 700; O17.3, 719; P246, 793; P282, 801 Kendler S., P298, 805 Kennedy James, O7.7, 706; O19.3, 721; O24.6, 730; O24.8, 730; P105, 762; P315, 809 Kennedy James L., O19.1, 721; O22.1, 724; P203, 784; P317, 810 Kennedy Jim, O24.4, 729 Kennedy Martin, P168, 776 Kennedy Martin A., P33, 746; P74, 755; P216, 787 Kennedy Susan E., O1.8, 695 Kenneth, P298, 805 Kent Lindsey, P121, 766; P123, 766 Kereszturi Eva, P149, 772; P212, 786 Kern Nikola, P10, 740 Kerwin Robert, O8.1, 706; P40, 747; P227, 789; P228, 789 Khan Nadeem, P90, 758 Khoury Aram El, P196, 782 Khusnutdinova Elza, P183, 779 Kiær-Larsen Stine, O23.1, 726 Kia-Keating Brett, P102, 761 Kiao Hsiao-Mei, P286, 802 Kieseppa¨ Tuula, O2.5, 696; O23.10, 728; P96, 760 Kikuyama Hiroki, P201, 783; P241, 792 Kilpatrick Dean, P19, 742 Kilpinen Helena, O2.5, 696; P124, 766 Kilpinen Leena Helena, O6.3, 703 Kim Jung-Jin, P236, 791 Kim Yong Kyu, P185, 780 Kima Jung-Jin, P01, 738 Kindler Jochen, P48, 749; P145, 771
Author Index King Nicole, O19.3, 721; O22.1, 724; O24.8, 730; P317, 810 King, Michael, O1.6, 694 Kingsley Sue, O12.3, 711 Kinnear Craig, P76, 755 Kinnell Hazel, P44, 748 Kinoshita Yoko, O8.4, 707 Kirov George, O2.8, 697; O13.2, 712; P12, 741; P17, 742; P47, 749; P266, 798; P290, 803 Kishimoto Makiko, P303, 806; P304, 807 Kiss Eniko, O22.1, 724; O24.4, 729; O24.8, 730 Kissling Christian, O24.1, 729; P70, 754 Kitajima Tsuyoshi, O8.4, 707; P220, 788 Kitamura Noboru, P88, 758; P176, 778 Kiyama Yuji, P191, 781 Kjeldsen Eigil, P77, 755 Kleinjung Tobias, P109, 763; P111, 763 Kleinman Joel, O16.1, 717 Kloet E.R. de, P63, 752 Kloiber Stefan, P06, 739; P09, 740 Klopp Norman, P95, 760; P262, 797 Knight Helen Miranda, P247, 794 Knight J., S6.4, 690 Knight Joanne, P105, 762 Knorr Caroline, O7.3, 705 Knowles James A., O1.3, 694 Koch Andreas, P07, 740 Koenen Karestan, P19, 742 Koeppe Robert A., O1.8, 695 Kofoed Pernille, P267, 798 Koga Minori, P180, 779 Koh Jun, P201, 783; P241, 792 Kohda Kazuhisa, P187, 780 Kohli Martin, P06, 739; P09, 740 Kohn Yoav, O20.5, 723 Koji Nishihara, P182, 779 Kolachana Bhaskar, O5.4, 702; O25.7, 732 Kollins Scott, P148, 771 Kolotii Alexei, P141, 770 Komiyama Tokutaro, P303, 806; P304, 807 Kong Augustine, O4.7, 700; O16.4, 718; P27, 744; P311, 808 Konidari A., O20.4, 723 Konstantin Nikola, O21.4, 724 Kopmann Susanne, O25.4, 731 Koprivsˇek Jure, P235, 791 Korbel Jan, O3.4, 698 Korenberger J.R., PL2, 685 Korner Mira, O20.5, 723 Kornhuber Johannes, O18.5, 721; P171, 777 Korszun A., S6.4, 690 Korszun Ania, O1.4, 694; O22.5, 725 Korte Gabriele, O21.4, 724; O21.5, 724 Korzyukov Oleg, P154, 773 Kostov Christian, P92, 759 Kosuga Asako, P26, 744 Kotaka Tatsuya, P304, 807 Kovacs Denes, O15.1, 715 Kovacs Maria, O22.1, 724; O24.6, 730; O24.8, 730 Kovas Yulia, O6.4, 703; P118, 765 Kovel C.G.F. de, P116, 765 Kozell Laura, P195, 782 Kramer J., S8.1, 691 Krasnov V., P45, 748 Krasucki Robert, P293, 804 Kravanja Milosˇ, O26.3, 733 Kremen William, O7.5, 705 Kremen William S., P18, 742 Kremensky Ivo, P92, 759 Kremer Ilana, P146, 771 Krishnan Ranga, P99, 761 Kroisel Peter, O14.1, 714 Kruse T.A., P53, 750 Kruse Torben, P20, 743 Kruse Torben A., P24, 744
Ktavetz Viktor, P141, 770 Kucherlapati Raju, O13.7, 714; P54, 750 Kumar Ravinesh A., O23.2, 727 Kundu Mukta, O2.7, 696 Kunimasa Arima, P26, 744 Kuntsi Jonna, P126, 767 Kuo P.-H., O4.2, 699 Kuo Po-Hsiu, O4.3, 699; O27.6, 735; P282, 801; P298, 805 Kuperman S., S8.1, 691 Kupriyanov D., P272, 799; P280, 801 Kupriyanov Dmitry, O25.5, 732; P271, 799 Kuroda Shigetoshi, P303, 806; P304, 807 Kurotaki Naohiro, P133, 768; P182, 779 Kusumi Ichiro, P170, 777 Kuusikko Sanna, O28.2, 735 Kuzman Martina Rojnic, P255, 795 Labad Javier, O22.10, 726; O5.1, 701 Labbe´ Michel, P08, 740 Labe´ Michel, O2.2, 695 LaCascia Caterina, P257, 796 Lachiewicz Ave, P129, 767; P148, 771 Lachowicz Michael, O19.2, 721 Laegsgaard Mett Marri, O10.4, 710 Lafreniere F., P85, 757 Lafreniere Ronald, P85, 757 Lahermo Pa¨ivi, O16.2, 717 Laje Gonzalo, O22.3, 724 Lakatos Krisztina, P128, 767 Lamb J., O28.3, 736 Lamb Janine, P130, 767 Lambert David, O24.7, 730 Lampus Simona F., P167, 776 Lan Tsuo-Hung, O19.4, 722 Landegren Ulf, S10.1, 693 Landowski Jerzy, P217, 787; P252, 795 Lang Donna, O7.7, 706 Langguth Berthold, P109, 763; P111, 763 Langley Kate, O24.2, 729; O6.1, 703 Lapalme Micheline, O16.5, 718 Laplante Nathalie, O16.5, 718 Lappin Julia, P161, 774; P270, 799 Lara Catalina Lopez de, O13.3, 713 Laranjeira Ronaldo, P308, 808 Larizza Lidia, O24.9, 730 Larsen E.R., P103, 761; S6.2, 690 Larsson C., S10.1, 693 Lasalvia Antonio, O25.6, 732; P276, 800 Lasseter Virginia K., O3.1, 697 Latella M.C., O26.5, 733 Latham Ian, P213, 786 Lathrop Mark, P07, 740 Laurent Claudine, O3.1, 697 Laurent S., P85, 757 Lawrence Jacob, P42, 747; P49, 749; P293, 804 Lazzarotto Lorenza, P276, 800 Leach Stephen, O23.2, 727 Leboyer Marion, O22.4, 725; O26.2, 733; P29, 745; P31, 745; P131, 768 Lebrec J.J.P., P116, 765 Leckband Susan, P22, 743 Lee Andrew, P36, 746 Lee Byung Dae, P66, 753 Lee Chang-Uk, P236, 791 Lee Chul, P236, 791 Lee Dongsoo, P254, 795 Lee Jang Yong, P254, 795 Lee Min-Soo, P69, 754 Lee Soo-Jung, P236, 791 Lee Yu-Sang, P185, 780; P254, 795 Leea Chul, P01, 738 Leemput Y. van de, P86, 757 Leeuwen Nienke van, P63, 752 Leisch Friedrich, P48, 749; P145, 771 Lencz Todd, O7.1, 704; O13.7, 714; O25.3, 731; P54, 750
817
Lennkh-Wolfsberg Claudia, P145, 771 Lenz Bernd, O18.5, 721; P171, 777 Leon Sandra Lopez, O1.1, 693 Leppert J., O23.6, 727 Leppert Jerzy, O5.2, 701 Leppert Mark, O15.2, 716 Lerch Jason, P132, 768 Lerer Bernard, O20.5, 723 Lerer Elad, P146, 771 Lerer F. Bernard, O3.1, 697 Lesch Klaus Peter, P70, 754; P71, 754; P82, 757; P256, 795 Leszczynska-Rodziewi Anna, P173, 777; P252, 795 Leszczynska-Rodziewicz Anna, P30, 745; P61, 752; P240, 792; P242, 793 Leuchowius K.J., S10.1, 693 Leung Doris, O21.2, 723 Levi Adi, O20.5, 723 Levinson Douglas F., P301, 806 Levinson D.F., S5.4, 689 Levinson Douglas, O1.3, 694; O3.1, 697; O16.8, 718 Lewis Stephen F., P163, 775 li Caixia, P302, 806 Li Chen-Hao, P286, 802 Li Chun, O23.3, 727 Li J., S3.3, 688 Li Jianhua, O4.1, 699 Li Jixiang, O4.1, 699 Li Jun, O2.1, 695 Li Tao, O15.1, 715; P160, 774; P162, 775; P302, 806 Li Yun, O2.1, 695 Li Zhen, O5.7, 702 Liappas Ioannis, P314, 809 Lieb Roselind, P08, 740; P10, 740 Lieberman Michael, O29.1, 737 Liebreman Jeffrey, O19.1, 721; O19.3, 721 Liehr Thomas, P140, 770; P142, 770 Lievers Luisa Strik, O24.9, 730 Light Katrina Jane, P74, 755 Lijun Cheng, O29.4, 737 Lim Se-won, P69, 754 Lin He, P59, 751 Linszen Don, O7.6, 705; P60, 751 Lipsky Robert, O25.3, 731 Liu Chih-Min, P248, 794; P284, 802 Liu Chun-yu, O2.7, 696 Liu Chunyu, O29.4, 737 Liu Qiang, O20.1, 722 Liu Xiehe, O15.1, 715; P162, 775 Liu Yu-Li, P284, 802 Liu Yuli, P248, 794 Liu Yun, O20.1, 722 Llibre J.J., P206, 784 Lo Wing Sze, P278, 800 Lobo Daniela Sabbatini, P105, 762 Lobos Elizabeth, P21, 743 Lochman Jan, O23.7, 728; O24.5, 729 Lochner Lochner, P76, 755 Lohoff Falk W., P245, 793 Lo¨nnqvist Jouko, O2.5, 696; O7.4, 705; O16.2, 717; O23.10, 728; P96, 760 Loo Sandra, O14.5, 715 Lo´pez-Ibor Juan Jose´, P151, 772; P202, 784 Lorente Jose´ A., P05, 739; P78, 756; P177, 778 Lorenzi Cristina, P41, 747; P204, 784; P221, 788 Lorenzo Giorgio Di, P23, 743 Loukola Anu, O7.4, 705; O16.2, 717; O23.10, 728; P94, 759 Louter Maartje, P02, 739 Love Tiffany M., O1.8, 695 Lovett Maureen W., O24.3, 729 Lu Brett L., P163, 775 Lu Chunrong, O12.5, 712
818
Author Index
Lubke Gitta, O14.5, 715 Luca Daniel Moreno De, P114, 764 Luca Paola, O22.1, 724 Luca Vincenzo De, O19.1, 721 Lucae Susanne, P06, 739; P08, 740; P09, 740; P10, 740 Lui Su, P160, 774 Luna Juan de Dios, O1.6, 694; P78, 756 Lyons Michael, O4.1, 699; O7.5, 705 Lyons Michael J., P18, 742 Ma Deqiong, O20.4, 723 Ma Xiao-hong, P160, 774 Ma Xiaohong, P162, 775; P302, 806 Maassen B., P116, 765 Macciardi F., S1.3, 687 Macciardi Fabio, O12.4, 712; O24.9, 730; O27.5, 735 MacDonald Colm, P157, 773 Macgregor Stuart, P47, 749; P51, 749 Macintyre Georgina, P268, 798 MacKinnon Dean F., O13.1, 712 Maclean Alan, P247, 794 Maeda Kiyoshi, P88, 758; P176, 778 Maes Hermine, O4.6, 700 Maestrini E., O28.3, 736 Maestrini Elena, P130, 767 Magno Elena, O7.2, 704 Magnusdo´ttir Brynja, P150, 772 Magri Chiara, P274, 800 Magzhanov Rim, P183, 779 Magzhanova Aliya, P183, 779 Mah Steven P., O5.5, 702 Maheshwari Manjula, O2.7, 696 Maier W., P103, 761; S6.2, 690; S6.4, 690 Maier Wolfgang, O13.2, 712; O16.7, 718; O23.4, 727; O3.1, 697; P97, 760 Mailis Antonis, P207, 785 Malafosse A., O26.2, 733; O26.4, 733 Malafosse Alain, O22.4, 725; P29, 745 Malhotra Anil, O7.1, 704; O13.7, 714; O19.2, 721; O25.3, 731; P54, 750 Malitas Petros, P207, 785; P314, 809 Mallet Jacques, O3.1, 697 Malloy M. Pat, P138, 769; P139, 769 Manabe Toshiya, O29.3, 737; P191, 781 Manchia M., P57, 751 Manchia Mirko, P167, 776; P56, 750 Manconi M., P208, 785 Manconi F., P209, 785; P57, 751 Mandelli Laura, P01, 738; P112, 764; P113, 764; P235, 791; P236, 791; P314, 809; P317, 810 Mandich Paola, P275, 800 Manji Husseini, O22.3, 724 Mann Karl, O4.8, 700 Manor Iris, O17.1, 719; O28.5, 736; P67, 753; P117, 765 Mansour Hader, O3.6, 698 Mantua V., P107, 762 Mapelli Francesca, O18.4, 720 Marı´n L.C., P206, 784 Marcheco-Teruel Beatriz, P24, 744; P53, 750; P206, 784 Marchi Margherita, O24.9, 730 Marco Luiz Armando De, P16, 742; P46, 748; P285, 802 Marcos Alberto, P202, 784 Margiani Stefania, P167, 776 Margriet, P166, 776 Marin˜o Magdalena, P151, 772 Maria Del Zompo, P175, 778 Marika Kovacs, O24.4, 729 Marino Elena, P41, 747; P204, 784; P221, 788 Marlowe Lauren, O21.2, 723 Marmurowska-Micha Powska Halina, P215, 787
Marongiu M.F, P200, 783 Marques Carla, O28.4, 736; P91, 759 Marrella Giovanna, O25.6, 732; P276, 800 Marrosu F., O29.7, 738 Marrosu Francesco, P193, 782; P306, 807 Marshall Christian, O14.1, 714 Martı´n-Santos Rocı´o, O8.5, 707 Marta Diana H., O1.3, 694 Marta Marta, P258, 796 Marteau T., S4.4, 689 Martin Eden R., O12.5, 712 Martin Kennedy A., P74, 755 Martin Nicholas G., O5.5, 702 Martin Nicholas, O1.5, 694 Martin Nick, O9.3, 708; O22.8, 726 Martina Tomori, P72, 754 Martinez M., S5.4, 689 Martins Madalena, O28.4, 736; P91, 759 Martins Silvia, P105, 762 Marusˇicˇ Andrej, O26.3, 733 Marusic A., P103, 761; S6.2, 690 Marziniak Martin, P71, 754 Mas Cristina, P198, 783 Masetti Guia Guffanti, O24.9, 730 Masnata Bianca, P275, 800 Mathe’ Aleksander, P196, 782 Mathews Carol, O15.3, 716 Mathieu Flavie, O22.4, 725; O26.2, 733; P31, 745 Matroos Glenn, P230, 790 Matsumoto Chima, P225, 789; P229, 790; P232, 790; P234, 791; P288, 803; P292, 804 Matsumoto Izuru, O3.5, 698 Matsumura Kenichi, P251, 794 Mattay Venkata, O5.4, 702 Mattila Marja-Leena, O28.2, 735 Maurer Konrad, P112, 764; P113, 764 Mayoral Fermin, O13.2, 712 Mazei Michelle, O8.2, 706 McAllister V.D.M., P87, 758 McAuley Erica Zoe, O2.3, 695 McCarthy A., S6.3, 690 McCarthy Alun, P231, 790; P315, 809 McCarthy Linda, P213, 786 McClay Joseph L., O16.6, 718 McDonald Colm, P152, 772 McGeary Chinatsu, P307, 807 McGeary John, P224, 789; P307, 807 McGough James, P94, 759 McGuffin P., P103, 761; P107, 762; P87, 758; S6.2, 690; S6.4, 690 McGuffin Peter, O1.4, 694; O19.5, 722; O22.5, 725; O28.5, 736; P12, 741; P17, 742; P164, 775; P257, 796; P269, 798; P279, 801; P291, 804 McGuire Phillip, P157, 773 McHugh Patrick Christopher, P216, 787 McInnis Melvin G., O1.3, 694; O13.1, 712 McInnis Melvin, P55, 750 McKeon Patrick, P34, 746; P35, 746 McKinney Rebecca, P22, 743 McLaughlan-Edwards S., P85, 757 McLean Alan, O23.2, 727 McMahon F., W1, 685 McMahon F.J., S6.1, 689 McMahon Francis J., O2.7, 696; O13.1, 712 McQuillin Andrew, O29.5, 738; P42, 747; P49, 749; P189, 781; P293, 804 McRae Allan, O9.3, 708 Meaburn E., P211, 786 Meaburn Emma, O6.5, 703; O20.2, 722; P118, 765 Mechelli Andrea, P157, 773 Median R., P263, 797 Medina Rolando, O16.8, 718; P13, 741; P66, 753; P281, 801
Medland Sarah, O27.6, 735 Medved Vesna, P218, 787; P255, 795 Mehta Nalini, P213, 786 Meij Annemarie Van der, P02, 739 Meijer O.C., P63, 752 Melin J., S10.1, 693 Melis M., P200, 783 Meltzer Herbert, O19.1, 721; O19.3, 721 Meltzer Herbert Y., O23.3, 727 Mencho´n Jose´ Manuel, O5.1, 701; O22.10, 726 Mendelewicz J., P103, 761 Mendlewicz J., S6.2, 690 Mendlewicz Julien, O2.4, 696 Mendoza R, P281, 801; P263, 797 Mendoza Ricardo, O16.8, 718; P13, 741 Meng F., S3.3, 688 Meng Fan, O2.1, 695 Meng Huaqing, P162, 775; P302, 806 Menke Andreas, P09, 740 Mercader Josep Maria, P144, 770 Mereu G., O29.7, 738 Mereu Giampaolo, P193, 782 Mereu R.M., P209, 785 Merlo-Pich E., S6.3, 690 Merlo-Pich Emilio, P315, 809 Merriman Barry, P143, 770 Meyer E.T., S8.1, 691 Meyer Jobst, P82, 757; P256, 795 Meyer-Lindenberg A., S1.4, 687 Meyer-Lindenberg Andreas, O5.4, 702 Meyn Stephen, P299, 805 Miao Kuangyi, O13.1, 712 Michaud Jacques, P101, 761 Michopoulou Alexandra, P125, 766 Middeldorp Christel, O22.7, 725 Middleton L., S6.3, 690; S6.4, 690 Middleton Lefkos, P231, 790; P315, 809 Miguel Teresa, P91, 759 Miguel Teresa Sa˜o, O28.4, 736 Mihalopoulou Panagiota, P125, 766 Mikesˇ Vladimı´r, O23.7, 728; O24.5, 729 Milanova Vihra, O2.6, 696; P51, 749; P92, 759 Milenska Tinka, P51, 749 Mill Jonathan, O24.6, 730; P169, 776 Millar J.Kirsty, P300, 806 Millar K., S5.2., 689 Miller Allison L., P74, 755 Miller Erin B., O1.3, 694 Miller Geoffrey, O17.2, 719 Miller Gregory A., P163, 775 Miller Michael, P21, 743 Miloserdov Eugene, O25.5, 732 Minderaa R.B., P135, 769 Miorelli A., P107, 762 Miorelli Ana, P161, 774; P164, 775; P257, 796; P258, 796; P269, 798; P270, 799; P279, 801; P291, 804 Miranda Ana, O17.1, 719; O28.5, 736; O28.6, 736; P117, 765 Miranda Debora, P68, 753 Miranda Robert, P224, 789; P307, 807 Miranda-Angulo Ana, O16.1, 717 Miret Salvador, P296, 805 Misener Virginia, O22.1, 724 Mitchell Kevin, O9.2, 708 Mitchell Kevin J., O29.3, 737 Mitchell Philip B., O2.3, 695 Miwa Masako, P156, 773 Miyajima Fabio, P137, 769 Miyawaki Atsushi, P38, 747 Moate Rachel, P213, 786 Moffitt Temi, P257, 796 Moilanen Irma, O14.5, 715; O28.2, 735; P94, 759 Moldovan Laura, P299, 805
Author Index Molina E., P264, 797 Molina Esther, O1.6, 694; P05, 739; P177, 778; P78, 756 Mollica Marco, P275, 800 Mo¨ller Hans-Ju¨rgen, O26.7, 734; P112, 764; P113, 764 Monachov Viktor, P141, 770 Monaco A.P., O28.3, 736 Monaco Anthony P., O11.1, 710; P130, 767 Monaco Anthony, O6.8, 704 Mondelli V., P107, 762 Mondelli Valeria, P161, 774; P164, 775; P257, 796; P258, 796; P269, 798; P270, 799; P279, 801; P291, 804 Mondimore Francis M., O13.1, 712 Monks Stephen, P259, 796 Monteiro J.M. Pereira, P85, 757 Montesi Jennifer, O7.2, 704 Montgomery Grant, O5.5, 702 Monti Peter, P224, 789; P307, 807 Moolman-Smook Johanna, P76, 755 Mooney Lisa, P34, 746; P35, 746 Moore Holly, O29.1, 737 Moorjani Priya, O28.2, 735 Morais Edgar, P16, 742 Morasse Karine, P134, 768 Moreno-Ku¨stner Berta, O1.6, 694; P05, 739; P78, 756 Morgadinho Teresa, O28.4, 736 Morgan C., P107, 762 Morgan Craig, P161, 774; P257, 796; P258, 796; P270, 799; P279, 801; P291, 804 Morgan Kevin, P161, 774; P270, 799; P279, 801 Mori Keiichiro, P243, 793 Morio Akiko, P303, 806; P304, 807 Morita Yukitaka, P303, 806; P304, 807 Morri Monica, P207, 785 Morris A., O28.3, 736 Morris Andrew, P130, 767 Morris Brian, P250, 794 Morris Derek W., O7.2, 704 Morris Derek, O29.3, 737 Morris Robin, P136, 769 Morris Stewart, P100, 761 Morris Stewart W., P39, 747 Mors O., P53, 750; P103, 761; S6.2, 690; S6.4, 690 Mors Ole, O10.4, 710; O23.1, 726; P20, 743; P24, 744; P77, 755; P189, 781 Moser Dirk, P82, 757; P256, 795 Mo¨ssner Rainald, P71, 754 Moskvina Valentina, O6.8, 704; O27.1, 734; P261, 797 Mosolov S., P45, 748 Mossaheb Nilufar, P48, 749 Mostacciuolo Maria Luisa, O17.4, 719; P52, 750 Mosyagin Igor, O19.2, 721 Motsinger Alison, O8.2, 706 Mouri Kentarou, P176, 778 Mowry B.J., S5.4, 689 Mowry Bryan J., O3.1, 697 Moya Jorge, P108, 763; P110, 763 Mun˜oz R., P263, 797; P281, 801 Mugishima Go, P156, 773 Muglia P., S6.3, 690; S6.4, 690 Muglia Pierandrea, P231, 790; P315, 809 Mu¨hleisen Thomas, O25.4, 731 Muir Keith, P213, 786 Muir W., S5.2., 689 Muir Walter J., O23.1, 726; P39, 747 Muir Walter, O23.2, 727; P93, 759; P100, 761; P247, 794 Mulder H., P210, 785 Mu¨ller Daniel J., O19.3, 721; P317, 810
Muller Daniel J., O19.1, 721 Mu¨ller-Myhsok Bertram, P06, 739; P08, 740; P09, 740; P10, 740 Mulligan A., S2.2, 687 Munakata Kae, P38, 747 Munoz Rodrigo, O16.8, 718; P13, 741 Munro Janet, O8.1, 706; P40, 747 Murphy Declan, P89, 758 Murphy Declan G.M., P136, 769 Murphy Dennis L., O13.1, 712 Murphy Kieran C., P136, 769 Murphy Kieran, P259, 796 Murphy-Eberenz Kathleen, O1.3, 694 Murray R.M., P87, 758; P107, 762; P162, 775; P302, 806 Murray Robin, O25.6, 732; P152, 772; P157, 773; P257, 796; P269, 798; P270, 799; P276, 800; P279, 801 Murray Robin M., P161, 774; P291, 804 Muthen Bengt, O14.5, 715 Myers Amanda, O21.2, 723 Myers R.M., S3.3, 688 Myers Regina, O8.2, 706 Myers Richard M., O2.1, 695 Naeeni H. Shaker, P87, 758 Nagisa Aoyama, O3.2, 697 Nakabayashi Tetsuo, P26, 744 Nakamoto Yurie, P156, 773 Nakamura Jun, O8.4, 707; P229, 790; P232, 790; P234, 791; P288, 803; P292, 804 Nakamura Kazuhiko, P156, 773 Nakano Yoko, P170, 777 Nakazawa Takanobu, P191, 781 Nangle Jeanne-Marie, O7.2, 704 Nanko Shinichiro, P251, 794 Naoe Yui, P225, 789; P229, 790; P232, 790; P234, 791; P288, 803; P292, 804 Nash M., S6.2, 690 Nash M.W., P103, 761 Navari S., P107, 762 Navari Serena, P161, 774; P164, 775; P257, 796; P258, 796; P269, 798; P270, 799; P279, 801; P291, 804 Navarro M. Eula`lia, P296, 805 Navarro Puri, O8.5, 707 Navine´s Ricard, O8.5, 707 Neal Ben, P117, 765; P127, 767 Neale Ben, O17.1, 719; P67, 753 Neale Ben M., P89, 758 Neale Benjamin, O5.6, 702 Neale Benjamin N., O27.4, 735 Neale Michael, O17.5, 720; P98, 760; P298, 805 Neale Michael C., O4.6, 700; O27.6, 735 Nelson Matthew, O5.5, 702 Nelson Stanley, P143, 770 Nemoda Zsofia, P149, 772; P212, 786 Neri Christian, P101, 761 Nertney Deborah A., O3.1, 697 Nestadt Gerald, O3.1, 697 Nesvaderani Maryam, O3.5, 698 Neuman Rosalind, O6.7, 704 Neves Fernando, P16, 742; P46, 748 Newman Timothy K., P73, 754 Newport, Jeff, O11.4, 710 Ng Siu-Kin, P278, 800 Nguyen Giang, O25.1, 731 Nguyen Nam, O25.1, 731 Nguyen-Huu Quynh, O16.5, 718 Nicodemus Kristin, O3.3, 697; O25.7, 732 Nicolato Rodrigo, P285, 802 Nicoletti G., P203, 784 Nicolini Humberto, O16.8, 718; P13, 741; P263, 797; P281, 801
819
Niculescu Alexander B., O25.2, 731; P312, 809 Niebling Helene, P109, 763; P111, 763 Nielsen Anders L., O23.1, 726 Nielsen Maria Mærsk, O23.1, 726 Nijmeijer Judith Simone, P135, 769 Nikolaev Rostislav, P271, 799 Nikolov Ivan, O3.1, 697; P47, 749 Nilsson K.W., O23.6, 727 Nilsson Kent, O5.2, 701 Nilsson M., S10.1, 693 Nimgaonkar Vishwajit L., O3.8, 699; P174, 777 Nimgaonkar Vishwajit, O3.6, 698 Ninomiya Hideaki, P265, 798 Nishiguchi Masaki, P201, 783; P238, 792; P241, 792 Nishiguchi Naoki, P88, 758; P176, 778 Nishimura Yukika, P165, 775 Nishino Tomoyuki, P21, 743 Noethen Markus M., P97, 760 Nogueira Mariana, P80, 756 Nonenmacher Tatjana, P106, 762 Nonkes Lourens, P190, 781 Noor Abdul, O14.1, 714 Norio Ozaki, O3.2, 697 Norrback Karl-Fredrik, O13.5, 713; O13.6, 713; P79, 756 No¨then M., P45, 748 No¨then Markus M., O2.6, 696; O13.2, 712; O23.4, 727; O25.4, 731; P95, 760; P252, 795; P262, 797 Norton Nadine, O2.8, 697; O3.1, 697; P260, 796; P261, 797; P266, 798; P287, 803; P290, 803 Nucci Gianluca, P213, 786 Nurnberger J., W1, 685 Nurnberger J.I., S8.1, 691 Nurnberger John I., O13.1, 712 Nushida Hideyuki, P88, 758 Nyegaard Mette, P77, 755 Nygren Gudrun, P131, 768 Nyman Emma Sofia, P94, 759 Oades Bob, O17.1, 719; O28.5, 736; P67, 753; P117, 765 Oades Robert D, O28.6, 736 Obino G., P209, 785 Ochoa-Ocan˜a Marı´a Antonieta, P25, 744 O’Donovan M., S5.1, 689 O’Donovan Michael, O21.3, 723; O24.2, 729; O3.1, 697; O6.8, 704; P179, 779; P184, 780; P260, 796; P261, 797; P289, 803 O’Donovan Michael C., O6.1, 703; P266, 798; P287, 803; P290, 803 O’Donovan Mick, O2.8, 697 Oh Kang-Seob, P69, 754 Ohashi Jun, P237, 791 Ohmori Osamu, P225, 789; P229, 790; P232, 790; P234, 791; P288, 803; P292, 804 Ok Jeun Hyun, P254, 795 Okada Takeya, P26, 744 Okahisa Yuko, P303, 806; P304, 807 Okazaki Yuji, P133, 768; P165, 775; P170, 777; P182, 779; P237, 791 Olajossy Marcin, P215, 787 Olajossy-Hilkesberger Luiza, P215, 787; P217, 787 Olds Robin J., P33, 746 Olgiati Paolo, P207, 785; P235, 791 Oliveira Guiomar, O28.4, 736; P91, 759 Ollier William, P137, 769 Olsen Inger M., P189, 781 Olsen Inger Marie, O23.1, 726 Olssen Marie, P76, 755
820
Author Index
Onaivi Emmaneul, P192, 781 O’Neill Anthony, O3.1, 697; P246, 793 O’Neill F. Anthony, O16.3, 717; O17.3, 719; O3.8, 699; P282, 801; P295, 804 O’Neill Frances Anthony, O5.6, 702 O’Neill Ryan, P226, 789 Ono Shinji, P182, 779 Ontiveros A., P281, 801 Ontiveros Alfonso, O16.8, 718; P13, 741; P263, 797 Oostra Ben A., O1.1, 693 Oppelaar Hugo, P190, 781 Oreland Lars, O23.6, 727; O5.2, 701 Orlova V., P280, 801 Orlova V.A., P272, 799 Orlova Vera, O25.5, 732; P271, 799 Orntoft T., P53, 750 Orntoft Torben F., P24, 744 Orru` M., O29.7, 738; P193, 782 Ortet Genero´s, P108, 763; P110, 763 Os Jim van, P230, 790 Osamu Saitoh, P26, 744 Osborne Sarah, P40, 747 Oskarsson Ho¨gni, O4.7, 700; P27, 744; P311, 808 Otani Kyohei, P303, 806; P304, 807 Otowa Takeshi, P237, 791 Ott Jurg, O20.1, 722 O’Tuathaigh Colm M.P., O29.3, 737 Owen M.J., S6.4, 690 Owen Michael, O6.8, 704; O21.3, 723; O22.5, 725; P12, 741; P17, 742; P179, 779; P184, 780; P260, 796; P261, 797; P289, 803 Owen Michael J., O1.4, 694; O3.1, 697; O6.1, 703; O13.2, 712; O24.2, 729; P266, 798; P287, 803; P290, 803 Owen Mike, O2.8, 697 Ozaki Norio, O8.4, 707; P220, 788; P303, 806; P304, 807 Ozawa Hiroki, P133, 768; P182, 779 Ozeki H, P250, 794 Padberg G., P116, 765 Pae Chi-Un, P01, 738; P236, 791 Paez-Pereda Marcelo, P08, 740 Paik In-Ho, P236, 791 Paika In-Ho, P01, 738 Paldi Andras, S9.4, 692 Palfrey-Davis Oliver, O6.6, 704 ´ va, P297, 805 Palik E Palmason Haukur, P82, 757; P256, 795 Palmour Roberta, O4.5, 700 Palo Outi M., O23.10, 728; P96, 760 Palo Outi Marika, O2.5, 696 Palotie Aarno, O16.2, 717 Pannizon Matthew, P18, 742 Pantazopoulou Katerina, P125, 766 Panteleyeva G., P45, 748 Paola Bossu`, O12.4, 712 Paolo Carta, P175, 778 Papadimitriou George, O3.1, 697 Papadopoulou-Daifoti Zeta, P314, 809 Papassotiropoulos Andreas, O27.2, 734 Papili P., P87, 758; P107, 762 Papili Pietro, P161, 774; P257, 796; P258, 798; P269, 798; P270, 799; P279, 801; P291, 804 Papiol Sergi, P233, 791 Paquin Bruno, O16.5, 718 Paquin Nouzha, O16.5, 718 Paracchini Silvia, O6.8, 704; O11.1, 710 Paragi Metka, O26.3, 733 Paraskevas Iatropoulos, P274, 800 Pariante C., P107, 762; P87, 758 Pariante Carmine, P228, 789; P257, 796; P258, 796; P291, 804
Pariante Carmine M., O19.5, 722; P161, 774; P164, 775; P269, 798; P279, 801 Paris Joel, P04, 739 Park Jong Won, P185, 780 Park Seong-Deok, P185, 780 Patkar Ashwin A., P236, 791 Partonen Timo, O2.5, 696; O7.4, 705; O16.2, 717; P07, 740; P96, 760 Pasqualetti Patrizio, P273, 799 Patkarb Ashwin A., P01, 738 Patkin Eugene, O12.2, 711 Patterson D.G., O4.2, 699 Patterson Diana, O4.3, 699 Paul Torsten, P95, 760; P262, 797 Pauls D., P116, 765 Pauls David, O28.2, 735 Paunio Tiina, O2.5, 696; O7.4, 705; O23.10, 728; P07, 740; P96, 760 Payton Antony, P137, 769 Pchelina Sofia, O12.2, 711 Pedersen Søren, P77, 755 Peiffer Andy, O15.2, 716 Peirce Tim, P260, 796; P261, 797 Peles Alma Mihaljevic-, P222, 788 Peltonen Juho O., O23.10, 728 Peltonen Juho Olavi, O7.4, 705 Peltonen Leena, O2.5, 696; O6.3, 703; O7.4, 705; O16.2, 717; O23.10, 728; P07, 740; P94, 759; P96, 760; P124, 766; S7.3, 691 Pendleton Neil, P137, 769 Penninx B., O11.3, 710 Peralta Juan Manuel, O16.8, 718 Peralta Vı´ctor, P296, 805 Pereira Ana Catarina P., P189, 781 Perepetchai Valeri, O16.5, 718 Perez J., P103, 761; S6.2, 690 Periante Carmine, P270, 799 Perica Margaretk-Vance, O22.9, 726 Pericak-Vance M.A., O20.4, 723 Pericak-Vance Margaret, P99, 761 Pericak-Vance Margaret A., O12.5, 712 Perini Giulia, P52, 750 Perlman William, O16.1, 717 Perrault Sylvie, P101, 761 Perroud N., O26.2, 733; O26.4, 733 Perroud Nader, O22.4, 725; O26.6, 734; P29, 745 Perry J., S6.3, 690; S6.4, 690 Perry Julia, P315, 809 Persson Maj-Liz, P07, 740 Peruzzi Pio, O17.4, 719; P52, 750 Pesaresi Marzia, P314, 809 Petek Erwin, O14.1, 714 Petronis Arturas, O9.3, 708; P169, 776 Petursson Hannes, O16.4, 718; P150, 772 Pfister Hildegard, P09, 740 Pfluegl Sandra, P109, 763; P111, 763 Phelan J., S4.2, 688 Phillipsen Alexandra, P84, 757 Piccardi M.P., P209, 785 Piccardi Maria Paola, P44, 748 Piccardi M.P., P57, 751 Picchioni Marco, P157, 773 Piccinni Monica, P158, 774 Pickard B., S5.2., 689 Pickard Ben, P93, 759; P247, 794 Pickard Ben S., P138, 769; P139, 769 Pickard Benjamin, P100, 761 Pickard Benjamin S., P39, 747 Pierce Tim, P179, 779 Pieri Elisa, O10.3, 709 Pietila¨inen Olli, O23.10, 728 Pietrini Pietro, O23.8, 728 Pimenta Guilherme, P16, 742 Pimm Jonathan, P293, 804 Piperi Christina, P207, 785; P314, 809
Piras Giovanna, O29.2, 737; P197, 782 Pirovano Adele, P41, 747; P204, 784; P221, 788 Pitelova´ Rena´ta, O23.7, 728; O24.5, 729 Piubelli Chiara, P196, 782 Piyabhan Pritsana, P104, 762 Platko J., P116, 765 Plaza Ana, O8.5, 707 Plescia Christopher, P266, 798; P290, 803 Plesnicˇar Blanka Kores, P235, 791 Pless Lora, O3.8, 699 Plomin R., P211, 786 Plomin Robert, O6.4, 703; O6.5, 703; O6.6, 704; O14.4, 714; O18.2, 720; O20.2, 722; O28.5, 736; P118, 765 Plummer Christie, O16.7, 718 Podavini Damiano, P274, 800 Politis Antonis, P207, 785 Pollock J., 686; W2, 686 Porcu S., P200, 783 Porjesz B., S8.1, 691 Porteous D.J., S5.2., 689 Porteous David J., P39, 747; P44, 748; P83, 757; P93, 759; P100, 761; P138, 769; P139, 769; P300, 806 Porton Barbara, P199, 783 Potash James B., O1.3, 694; O13.1, 712 Potkin S.G., S1.3, 687 Potkin Steven, O27.5, 735 Poulin J.-F., P85, 757 Powell J., P107, 762 Powell John, P89, 758; P127, 767; P257, 796; P258, 796; P279, 801 Prata Diana, P157, 773 Prato Francesca, P207, 785; P314, 809 Pratt Christopher, P106, 762; P109, 763 Preisig M., S6.4, 690 Preisig Martin, P29, 745 Prescott C.A., O4.2, 699 Prescott Carol, O4.3, 699 Preuss Ulrich, O4.4, 700 Prince M., P206, 784 Proietti Luca, P23, 743 Prokopenko I., S6.3, 690 Prokopenko Inga, P231, 790; P315, 809 Propping Peter, O2.6, 696; O13.2, 712; O23.4, 727; O25.4, 731; P95, 760; P97, 760; P262, 797 Puetz Benno, P09, 740 Pugliese Luca, P89, 758 Pugliese P., P203, 784 Puig Anna, P144, 770 Puligheddu M., O29.7, 738 Puligheddu Monica, P193, 782; P306, 807 Pulst Stefan, O15.2, 716 Pulver Ann E., O3.1, 697 Pun Frank W., P278, 800 Purcell Shaun, O17.1, 719; O22.6, 725 Purcell Shaun Matthew, O27.4, 735 Puri Vinay, P42, 747; P49, 749; P189, 781; P293, 804 Purvenis Natalia, P272, 799 Putzu P.F., P209, 785 Quattrone A., P203, 784 Quervain Dominique J.-F. de, O27.2, 734 Quezada Paulina, P66, 753 Quin˜ones Rau´l mendoza, P153, 773 Quinn John, O21.1, 723; P137, 769 Rabe-Hesketh S., P162, 775 Radhakrishna Vakkalanca, O16.1, 717 Radka Kaneva, O2.6, 696 Radovan Komel, P72, 754 Raelson John Verner, O16.5, 718 Rajewska-Rager Aleksandra, P61, 752; P173, 777
Author Index Rajewski Andrzej, O18.1, 720 Ramesh Vijaya, O28.2, 735 Ramirez Mercedes, P66, 753 Ramos Josep Antoni, P80, 756 Rampinelli Sabina, O17.4, 719; P52, 750 Rapoport Judith, P132, 768; P143, 770 Rasmussen Henrik B., P267, 798 Rastam Maria, P131, 768 Ratti Anna Rita, P207, 785 Raventos Henriette, O16.8, 718; P13, 741; P66, 753; P281, 801 Raventos Henrrietta, P263, 797 Raymond R., O1.3, 694 Reber Nadja, P29, 745 Reck Brian, P213, 786 Reed Carol, O19.2, 721 Regina Armas, P66, 753 Rehnstro¨m Karola, P124, 766 Reich T., S6.4, 690 Reich W., S8.1, 691 Reicher Jennifer, P131, 768 Reif Andreas, P70, 754 Reiman Eric M., O27.2, 734 Remington Gary, O19.1, 721 Rene´, P166, 776 Renner Tobias J., P82, 757 Reppermund Simone, P09, 740 Resnick Heidi, P19, 742 Retz Wolfgang, O24.1, 729; P70, 754 Retz-Junginger Petra, O24.1, 729; P70, 754 Reus Victor, O15.3, 716 Reynolds Gavin, P104, 762; P226, 789 Ribase´s Marta, P80, 756 Ribble Robert, O5.6, 702 Ricciardi A., P107, 762 Ricciardi Angelo, P87, 758; P257, 796 Rice Frances, O24.2, 729 Rice J., S6.4, 690 Rice John P., O13.1, 712 Richard, P166, 776 Richardson Alex, O6.8, 704 Richmond Todd, O3.4, 698 Rietschel M., P45, 748; P103, 761; S6.2, 690; S6.4, 690 Rietschel Marcella, O2.6, 696; O7.3, 705; O13.2, 712; O23.4, 727; O25.4, 731; P07, 740; P84, 757; P95, 760; P97, 760; P262, 797 Rijk Peter De, O27.3, 734 Rijk R.H. de, P63, 752 Rijsdijk Fruhling, O20.3, 722 Riley B.P., O4.2, 699 Riley Brien, O3.1, 697; O4.3, 699; O5.6, 702; O16.3, 717; O17.3, 719; P282, 801 Riley J., S6.3, 690 Riley John, P315, 809 Riley P., P298, 805 Ripke Stephan, P09, 740 Ristaldi M.S., P200, 783 Ritchie Marylyn D., O17.3, 719 Ritchie Marylyn, O8.2, 706 Rivas Fabio, O2.6, 696; O13.2, 712 Rivera Margarita, O1.6, 694; P05, 739; P78, 756; P177, 778 Rizig Mie, O29.5, 738 Roberto De Lisa, P175, 778 Roberts Wendy, O14.1, 714 Robertson Ian H., O7.2, 704; O24.7, 730; P58, 751 Roche Siobhan, P34, 746; P35, 746 Rodriguez Raquel, P151, 772; P202, 784 Rodriguiz Ramona, P199, 783 Roeyers Herbert, O17.1, 719; O28.5, 736; O28.6, 736; P67, 753; P117, 765 Rogaev Evgeny, P294, 804 Rogers Geraldine R, P216, 787 Rogers Hannah, P126, 767
Rogue Patrick Jean, P172, 777 Rohrer Kristen, O21.2, 723 Rohsenow Damaris, P224, 789; P307, 807 Rojnic Martina, P218, 787 Role Lorna, O29.1, 737 Romano-Silva Marco Aure´lio, P16, 742; P46, 748; P285, 802 Romanos Marcel, P82, 757 Rommelse Nanda, O15.6, 716 Rommelspacher H., O4.9, 701 Ronald A., S2.3, 687 Ronchi Diana De, P112, 764; P113, 764; P207, 785; P235, 791 Ronchic Diana De, P01, 738 Rosa Araceli, P296, 805 Rose N., S4.1, 688 Rosenbaum Jerrold, O22.6, 725 Roses A., S6.3, 690 Roses Allen, P315, 809 Rosˇkar Sasˇka, O26.3, 733 Ro¨sler Michael, O24.1, 729; P70, 754 Ross Jessica, P32, 745 Rossetti Zvani Luigi, P306, 807 Rossi Giuseppe, P273, 799 Rothenberger Ari, O17.1, 719; O28.5, 736; P67, 753; P117, 765 Rothenberger Aribert, O28.6, 736 Rouleau G.A., P85, 757 Rouleau Guy, O13.3, 713; P268, 798 Rouleau Guy A., P101, 761 Ruf Matthias, O7.3, 705 Ruggeri Mirella, O25.6, 732; P276, 800 Ruggiero Kenneth, P19, 742 Ruiz-Veguilla Miguel, P283, 802 ´ ngeles, P108, 763 Ruizpe´rez Maria A Ruizperez M. Angeles, P110, 763 Rujescu Dan, O25.8, 732; O26.7, 734; O28.1, 735; P112, 764; P113, 764 Runarsdottir Valgerdur, O4.7, 700; P27, 744; P311, 808 Runker Annette, O29.3, 737 Rush A. John, O22.3, 724 Russell Elen Elizabeth, P188, 780 Russo Silvia, O24.9, 730 Ruzickova Martin, O13.3, 713 Rybakowski Filip, O18.1, 720 Rybakowski Janusz, O7.8, 706; P65, 753; P240, 792; P244, 793 Rybakowski Janusz K., P30, 745; P62, 752 Ryoko Ishihara, O3.2, 697 Saardzadeh-Sardahaee F., P107, 762 Sacchetti Emilio, P273, 799; P274, 800 Sagud Marina, P222, 788 Saito Shinichi, P220, 788 Saito Toshikazu, P192, 781 Sakata Shinichi, P225, 789; P229, 790; P232, 790; P234, 791; P288, 803; P292, 804 Sakul Hakan, O1.2, 693 Sakurai Takeshi, P266, 798; P290, 803 Salazar Rodolfo, P66, 753 Salyakina Daria, P08, 740; P09, 740; P10, 740 Sa´nchez-Morla Eva, P151, 772 Sanctis Gabriella De, O17.4, 719; P52, 750 Sand Philipp, O1.8, 695; O11.5, 711; O21.4, 724; O21.5, 724; P109, 763; P111, 763 Sanders A.R., S5.4, 689 Sanders Alan R., O3.1, 697 Sanders-Bush Elaine, O23.3, 727; O8.2, 706 Sandor Paul, P68, 753 Santangelo Susan, O28.2, 735 Santi Samuele, O23.8, 728 Santoni Federico, P306, 807 Santos Jose Luis, P151, 772 Santos Tulio, O28.2, 735
821
Sarchiapone M., O26.5, 733 Sarosi Andrea, P11, 741 Sasaki Tsukasa, P187, 780; P237, 791 Sassone-Corsi P., S9.1, 692 Sasvari-Szekely Ma´ria, P212, 786 Sasvari-Szekely Maria, O23.5, 727; O8.6, 707; P11, 741; P149, 772; P309, 808 Sato Mitsuko, P156, 773 Saudino Kim, P126, 767 Saunders Erika, P55, 750 Saus Ester, O5.1, 701 Savarino Graziana, O23.8, 728 Savina T.D., P272, 799 Savina Tatiana, P280, 801 Scapoli Chiara, O17.4, 719 Scassellati Catia, P273, 799 Scerri Thomas, O6.8, 704 Schachar Russell, O14.3, 714 Schalkwyk L., P103, 761 Schalkwyk Leonard C., O6.5, 703; O19.5, 722; O20.2, 722 Schalling Martin, P07, 740 Schallmeiner E., S10.1, 693 Scharfetter C., O8.7, 707 Scharl Theresa, P48, 749 Schatzberg Alan F., O2.1, 695 Scheffer H., P210, 785 Scheftner William, O13.1, 712 Scheftner William A., O1.3, 694 Scherer Stephen, O14.1, 714 Schillani Giulia, P81, 756; P219, 787 Schirmbeck F., P45, 748 Schirmbeck Frederike, O23.4, 727 Schma¨l Christine, P95, 760 Schmael C., S6.2, 690 Schmael Christine, O7.3, 705; P262, 797 Schmal C., P103, 761 Schmalzried Hannah, P84, 757 Schmechel Donald E., O12.5, 712 Schmidtke Klaus, P106, 762 Schmitt Andrea, O7.3, 705 Schmitz Nicole, P60, 751 Schnabel Axel, P112, 764; P113, 764 Schneider Barbara, P112, 764; P113, 764 Schneider Marc, O24.1, 729; P70, 754 Schoeman Renata, P76, 755 Schofield Peter R., O2.3, 695 Schork Nicholas, P147, 771 Schosser Alexandra, P48, 749 Schreier Peter, O21.5, 724 Schuelter Ulrike, P256, 795 Schulze T., P103, 761; S6.2, 690 Schulze Thomas, O13.2, 712 Schulze Thomas G., O2.6, 696; O23.4, 727; O25.4, 731; P95, 760; P97, 760; P252, 795; P262, 797 Schumacher J., P45, 748 Schumacher Johannes, O2.6, 696; O13.2, 712; O23.4, 727; O25.4, 731; P95, 760; P97, 760; P262, 797 Schumann Gunter, O4.8, 700; P07, 740; P181, 779 Schwab Sibylle, O16.7, 718 Schwab Sibylle G., O3.1, 697; O3.8, 699 Schwab Sibylle Gabriele, O11.5, 711 Schwaiger Siobhain, O7.2, 704 Schwartz D., S5.4, 689 Schwarz Markus, P95, 760; P262, 797 Schwarzman Aleksandr, O12.2, 711 Schweizer Barbara, O13.1, 712 Scott L., S3.3, 688 Scott Laura, P55, 750 Scott Laura J., O2.1, 695 Scott William K., O12.5, 712 Scudellaro Eva, P52, 750 Seal Jeffrey, P132, 768; P143, 770 Seedat Soraya, P76, 755
822
Author Index
Segal Jonathan, O16.5, 718 Segala`s Cinto, O5.1, 701; O22.10, 726 Segurado Ricardo, O3.1, 697; P261, 797 Sei Yoshitasu, O5.7, 702 Seidman Larry, O15.4, 716; P253, 795 Seiichi Harada, P26, 744 Seitz Christiane, P82, 757 Seki Hiroko, P251, 794 Sekine Yoshimoto, P303, 806; P304, 807 Selek Salih, P214, 786 Selten Jean-Paul, P166, 776 Selzer Rebecca, O3.4, 698 Sequeiros J., P85, 757 Sergeant Jo, O17.1, 719; O28.5, 736; P67, 753; P117, 765 Sergeant Joseph, O28.6, 736 Serio Stefania, P50, 749 Serretti Alessandro, O8.8, 708; P01, 738; P207, 785; P112, 764; P113, 764; P235, 791; P236, 791; P314, 809; P317, 810 Sertic Jadranka, P222, 788 Sˇery´ Omar, O24.5, 729 Sery Omar, O23.7, 728 Severino Giovanni, P167, 776; P56, 750 Severinsen Jacob, O23.1, 726 Shaikh Sajid, O7.7, 706 Shaker H., P107, 762 Sham P., S6.4, 690 Sham P.C., P162, 775 Sham Pak, O17.1, 719; O28.5, 736; P152, 772 Sham Pak C., O27.4, 735 Shaw Philip, P132, 768; P143, 770 Sheehan Karen, O28.6, 736 Shehktman Tatyana, O13.4, 713; P22, 743 Shelton Richard, O8.2, 706 Shemer Ruth, S9.2, 692 Shen Sanbing, P194, 782 Sheng Li, P59, 751 Shergill Sukhi, P157, 773 Sherva Richard, O6.7, 704 Shi Jianxin, P301, 806 Shi YongYong, O20.1, 722 Shibata Hiroki, P265, 798 Shimizu Kazuko, P225, 789; P229, 790; P232, 790; P234, 791; P288, 803; P292, 804 Shimozono Satoshi, P38, 747 Shinichi Saito, O3.2, 697 Shinkai Takahiro, P225, 789; P229, 790; P232, 790; P234, 791; P288, 803; P292, 804 Shirakawa Osamu, P88, 758; P176, 778 Shirts Brian, O3.6, 698 Shizuno Tomoko, P26, 744 Shubina Darija, O12.2, 711 Shuji Iritani, O3.2, 697 Sicard Tricia, O19.3, 721; P317, 810 Sieghart Werner, P48, 749; P145, 771 Sigmund Jessica, P82, 757 Sigmundsson Thordur, O16.4, 718; P150, 772 Sigurdsson Engilbert, O16.4, 718; P150, 772 Silander Kaisa, O2.5, 696; O23.10, 728; P96, 760 Sillaber Inge, P08, 740 Silveira Gustavo, P46, 748 Silverman J.M., S5.4, 689 Silverman Jeremy M., O3.1, 697 Sim S.C., P103, 761 Simoneau M., P85, 757 Simpson Elizabeth M., O12.3, 711; O23.2, 727 Sing Vishnu Jangbahadoer, P230, 790 Sinke J., P166, 776 Siracusano Alberto, P23, 743 Sirota Alan, P224, 789 Sitskoorn M., P166, 776 Sivagnanasundaram Sinthuja, O3.5, 698
Sjo¨berg Rickard, O5.2, 701 Sjoberg R.L., O23.6, 727 Skaar D.A., O20.4, 723 Skaug Jennifer, O14.1, 714 Skibin˜ska Maria, O7.8, 706; O18.1, 720; P30, 745; P61, 752; P62, 752; P64, 752; P65, 753; P239, 792; P240, 792; P242, 793; P244, 793 Sklar Pamela, O22.6, 725 Skol A., S3.3, 688 Skowronek Markus H., P84, 757 Skowronek Markus Hans, P262, 797 Slagboom P. Eline, O22.7, 725 Sleegers Kristel, O13.6, 713 Slifer Michael, O22.9, 726 Slifer Mike A., O12.5, 712 Sloman Leon, O14.1, 714 Slominska Natalija, O12.2, 711 Slopien Agnieszka, O18.1, 720; P61, 752; P239, 792; P240, 792; P242, 793 Smalley Susan, O14.5, 715; P94, 759 Smeraldi Enrico, P41, 747; P204, 784; P221, 788 Smit A.B., O11.3, 710 Smit J., O11.3, 710 Smith Ashely K., P163, 775 Smith Cassandra, O9.4, 708; O25.1, 731 Smolka Michael, O4.8, 700 Smoller Jordan, O22.6, 725 Snyder Michael, O3.4, 698 Sobanski Esther, P84, 757 So¨derberg O., S10.1, 693 Sodhi Monsheel, O23.3, 727 Solokhina Tatiana, P271, 799 Sommer Claudia, P71, 754 Song Ji-Yun, O9.1, 708 Sonuga-Barke Edmund, O17.1, 719; O28.5, 736; O28.6, 736; P67, 753; P117, 765 Sora Ichiro, P303, 806; P304, 807 Soria Virginia, O22.2, 724 Soronen Pia, O2.5, 696; O23.10, 728; P07, 740; P96, 760 Souery D., S6.2, 690 Souery Daniel, O2.4, 696 Sousa I., O28.3, 736 Sousa Sara Campos de, P40, 747 Sousa Sara Campos do, P270, 799 Spalletta Gianfranco, O12.4, 712 Spanagel Rainer, O4.8, 700 Spijker S., O11.3, 710 Spremo-Potparevic Biljana, O12.1, 711 Squassina A., P57, 751; P209, 785 Squire Jeremy, P299, 805 St. Jean P., P63, 690 St. Jean Pamela, P315, 809 Staal Wouter, O6.2, 703 Stachon Andrea, P299, 805 Stachowiak Beata, O14.1, 714 Sˇtaif Radim, O23.7, 728; O24.5, 729 Stamouli Evangelia, P207, 785 Stanton Gerson, O19.2, 721 Stapleton Carole, P213, 786 Stassen Hans H., O8.7, 707 Stazi Antonia, P158, 774 Steele Angela, O8.2, 706 Steele C.J.M., O2.7, 696 Steele Jo, O13.1, 712 Stefania Serio, P175, 778 Stefansson Hreinn, O4.7, 700; O16.4, 718 Stefansson Kari, O4.7, 700; O16.4, 718; P27, 744; P311, 808 Stein Dan, P76, 755 Stein John, O6.8, 704 Steinhausen Hans-Christoph, O17.1, 719; O28.5, 736; O28.6, 736; P67, 753; P117, 765
Steinthorsdottir Valgerdur, O16.4, 718 Stephan Dietrich A., O27.2, 734 Stern Hal, O27.5, 735 Stevens Angela F., P136, 769 Steyaert Jean, P75, 755 Stober Gerald, P278, 800 Stoertebecker Petra, P111, 763 Stochino M.E., P208, 785 Stone William, P253, 795 Stoyanova Vessela, P51, 749 Straub Richard, O3.3, 697; O5.4, 702; O5.7, 702; O16.1, 717 Straub Richard E., O25.7, 732 Strnad Alexandra, P145, 771 Strohmaier Jana, P97, 760 Stuffrein-Roberts Selma, P168, 776 Styczynska Maria, P205, 784 Su Jessica, O4.1, 699 Suchkova Irina, O12.2, 711 Sue Hong Kyung, P254, 795 Suga Motomu, P237, 791 Sugden Karen, O19.5, 722; O26.3, 733; P90, 758 Sullivan P.F., O4.2, 699 Sullivan Patrick, O4.3, 699; O8.3, 707; O23.9, 728; P28, 745 Sullivan Patrick F., O3.8, 699; O5.5, 702; O22.7, 725 Sun Lingwei, O6.7, 704 Sun Xiujun, O13.3, 713 Sun Xueli, O15.1, 715; P162, 775; P302, 806 Suto Fumikazu, O29.3, 737 Suvisaari Jaana, O7.4, 705 Suwalska Aleksandra, P30, 745 Suzuki Tatsuyo, P220, 788 Swaab Hanna, P190, 781 Swanson James, O14.5, 715 Swift Robert, P224, 789; P307, 807 Swinkels Sophie H.N., O14.6, 715 Syed Zahoor, P123, 766 Sykes Nuala, O28.3, 736; P130, 767 Syu Aoi, P180, 779 Szatmari Peter, O4.5, 700 Szczepankiewicz Aleksandra, P30, 745; P65, 753 Szekely Anna, P11, 741; P309, 808 Szente Sara, P43, 748 Szepankiewicz Aleksandra, P244, 793 Szeszko Philip, O7.1, 704; O13.7, 714 Szilagyi Agnes, O8.6, 707; P309, 808 Tabatabai Babak, O5.6, 702 Takahashi Nagahide, O3.2, 697; P220, 788 Takaji Masafumi, P265, 798 Talkowski Michael, O3.6, 698 Talmage David, O29.1, 737 Tamas Zsuzsa, O22.1, 724; O24.6, 730 Tan Q., P53, 750 Tanaka Yuji, P303, 806; P304, 807 Tang He-han, P160, 774 Tang RuQi, O20.1, 722 Tang Wei, O20.1, 722 Tanii Hisashi, P165, 775 Tannock Rosemary, O24.3, 729 Tansella Michele, O25.6, 732; P276, 800 Tarantino Irene, O23.8, 728 Tarnok Zsanett, P149, 772; P212, 786 Tashiro Nobutada, P265, 798 Tatsumi Masahiko, P26, 744 Tavares Hermano, P105, 762 Taylor Kristen, P132, 768 Teasenfitz Lindsey, P192, 781 Teixeira Antoˆnio, P285, 802 Teltsh Omri, O20.5, 723 Tenesa Albert, P39, 747 Thapar Anita, O6.1, 703; O24.2, 729; P67, 753 Tharmalingam Subi, O19.3, 721; P317, 810
Author Index The IMAGE Consortium, O28.6, 736 The Wellcome Trust Case Control Consortium, S3.1, 688 Theiner Pavel, O24.5, 729 Thiagalingam Sam, O9.4, 708 Thiagarajah Tanya, P139, 769 Thirumalai Srinivasa, P293, 804 Thiselton Dawn, O4.3, 699; O5.6, 702 Thiselton L., P298, 805 Thomas Jo, O1.3, 694 Thomas Schubert, P84, 757 Thome Johannes, O24.1, 729; P70, 754 Thompson R.C., S3.3, 688 Thompson Robert C., O2.1, 695 Thomson Philippa, P100, 761 Thomson Philippa A., P39, 747 Thomson Pippa, P44, 748; P93, 759 Thorgeirsson Thorgeir, O16.4, 718; P27, 744 Thorgeirsson Thorgeir E., O4.7, 700; P311, 808 Thorselius M., S10.1, 693 Tibbo Philip, P268, 798 Tidey Jennifer, P224, 789; P307, 807 Tiganov A., P45, 748 Tijssen M.A.J., P86, 757 Tingwei Guo, P59, 751 Tochigi Mamoru, P187, 780; P237, 791 Todd R.D., S2.4, 687 Todd Richard, O6.7, 704; O26.1, 732; P21, 743 Todd Webb B., S6.4, 690 Todorov Alexandre, P21, 743 Toma Claudio, P130, 767 Tomaszewska Marta, P62, 752; P64, 752; P65, 753; P252, 795 Tomita Hiroaki, O2.1, 695 Tompa Peter, O15.1, 715 Tomppo Liisa, O16.2, 717 Tondo Leonardo, O8.1, 706 Tonev Simon, P148, 771 Tong Ka-Lok, P278, 800 Torralbas M., P53, 750 Torres Francisco, P05, 739; P78, 756 Torres P., P264, 797 Torres-Gonza´lez Francisco, O1.6, 694 Torrijos Sergio, P151, 772 Tosato Sarah, O25.6, 732; P276, 800 Toshiya Inada, O3.2, 697 Tost Heike, O7.3, 705 Tovilla Alfonso, P263, 797 Tozzi F., S6.3, 690 Tozzi Federica, P231, 790; P315, 809 Tran-Viet Khanh Nhat, P99, 761 Tremblay Richard, P04, 739 Trequattrini Alberto, O12.4, 712 Treutlein Jens, P07, 740; P84, 757; P181, 779 Troisi Alfonso, P23, 743 Trull Tim, O22.8, 726 Tsai Ho-Min, P286, 802 Tsai Shih-Feng, P284, 802 Tsang Shui-Ying, P278, 800 Tsapakis Eva Maria, P227, 789; P228, 789 Tsuang Ming T., P102, 761 Tsuang Ming, O4.1, 699; O7.5, 705; P253, 795 Tsujita Takahiro, P170, 777 Tsutsumi Atsushi, P201, 783; P241, 792 Turecki G., P85, 757 Turecki Gustavo, O13.3, 713; O23.2, 727; P04, 739 Turic Darko, O24.2, 729 Turner J., S1.3, 687 Turner Jessica, O27.5, 735 Turunen Joni, P124, 766 Turunen Joni A., O6.3, 703; O23.10, 728 Tuulio-Henriksson Annamari, O2.5, 696; O7.4, 705; O23.10, 728; P96, 760 Tuveri Antonella, P306, 807
Tyrfingsson Thorarinn, O4.7, 700; P27, 744; P311, 808 Tzavellas Elias O., P314, 809 Uchimura Naohisa, P243, 793; P278, 800 Ueda Junko, P170, 777 Ueno Mikako, P251, 794 Ueno Yasuhiro, P88, 758 Ugarte Carmen Y de, P202, 784 Uhr Manfred, P08, 740; P10, 740 Ujike Hiroshi, P303, 806; P304, 807 Ujkaj Mandola, P207, 785 Umekage Tadashi, P187, 780 University of Pisa, Pisa, O23.8, 728 Unschuld Paul G., P10, 740 Unschuld Paul Gerson, P06, 739 Upmanyu Ruchi, P315, 809 Uranova Natalia, P140, 770 Urban Alexander Eckehart, O3.4, 698 Urretavizcaya Mikel, O22.2, 724 Utrera Juani, P198, 783 Utsunomiya Kensuke, P225, 789; P229, 790; P232, 790; P234, 791; P288, 803; P292, 804 Uzun Ozcan, P214, 786 van den Oord Edwin J., O16.6, 718 van den Oord Edwin, O3.8, 699; O4.3, 699; O5.6, 702; O16.3, 717; P98, 760 van den Oord E.J.C.G., O4.2, 699 Vakkalanka Radhakrishna, O25.7, 732 Vallada Homero, P37, 746; P105, 762; P159, 774; P223, 788; P308, 808 Valle David, O5.3, 701 Vallejo Julio, O22.2, 724; O22.10, 726 Valsecchi Paolo, P273, 799; P274, 800 van den Oord E.J., S6.4, 690 Van der Walt Joelle M., O12.5, 712 Vang Maria, P24, 744 Vanhala Raija, O6.3, 703; P124, 766 Vanna Maurizio De, P81, 756 Vanzo Alice, P315, 809 Varilo Teppo, O6.3, 703; O7.4, 705; P94, 759 Vawter Marquis, O11.2, 710 Vawter Marquis P., O2.1, 695 Va´zquez-Alvarez Blanca, P151, 772 Va´zquez-Alvarez Blanca M., P202, 784 Vazza Giovanni, O17.4, 719; P52, 750 Vecchi Roberta, P219, 787 Vedeniapin Andrey, P154, 773 Veguilla Miguel Ruiz, P264, 797 Velayos-Baeza Antonio, O11.1, 710 Velden Monique De Sain-van der, O5.3, 701 Ven Wim Van de, P75, 755 Venken Tine, O2.4, 696; O13.5, 713 Verchinski Beth, O5.4, 702 Verhagen Maaike, P02, 739 Verloes Alain, P131, 768 ´ gnes, O22.1, 724; O24.4, 729; Vetro´ A O24.6, 730; O24.8, 730 Vettori Andrea, P52, 750 Viale D., P209, 785 Vicente Astrid Moura, O28.4, 736; P91, 759 Vicentini Elena, P315, 809 Vidal Jean-Michel, O16.5, 718 Videtic Alja, P72, 754 Viding Essi, O14.7, 715 Vighini Miriam, P196, 782 Villa Helena, P108, 763; P110, 763 Villanueva Rex, P310, 808 Vincent John, O14.1, 714 Virkkunen Matti, P73, 754 Visscher Peter, O9.3, 708 Vitaro Frank, P04, 739 Vittum Jennifer, O4.3, 699 Vladimirov Vladimir, O4.3, 699; O5.6, 702; P298, 805
823
Vladimirova Rositza, P51, 749 Vogler Christian, P82, 757; P256, 795 Voineskos Aristotle Nicholas, O7.7, 706 Voinova-Ulas Victoria, P141, 770 Volavka Jan, O19.3, 721 Vollmert Christian, O7.3, 705 Vorsanova Svetlana, P140, 770; P141, 770; P142, 770 Vorstman Jacob, O5.3, 701; O6.2, 703 Vos Rita De, P75, 755 Voscresenskaya N., P280, 801 Voscresenskaya Nina, O25.5, 732; P271, 799 Voskresenskaja N.I., P272, 799 Vostrikov Viktor, P140, 770 Vousden Marika, P213, 786 Waddington John L., O29.3, 737 Wade-Martins Richard, O11.1, 710 Wagner Katrin, O21.4, 724 Wa¨hlby C., S10.1, 693 Walitza Susanne, P82, 757 Walsh D., O4.2, 699 Walsh Dermot, O3.1, 697; O16.3, 717; O17.3, 719; P246, 793; P282, 801; P295, 804 Walsh Dermott, O4.3, 699; O5.6, 702 Walshe Muriel, P152, 772 Walss-Bass Consuelo, P66, 753 Walter Janet, O6.8, 704 Walter Nikki, P195, 782 Wan Chunling, O3.7, 698 Wang August, P20, 743 Wang August G., P24, 744; P267, 798 Wang Gaofeng, O12.5, 712 Wang Jun-Feng, O13.3, 713 Wang Qiang, P162, 775; P302, 806 Wang Sun-Chong, O9.3, 708 Wang Xu, O3.8, 699; O17.3, 719; P246, 793; P295, 804 Ward Malcolm, O19.5, 722 Ward Sarah, O16.1, 717 Ward Stephen J.A., O12.3, 711 Wardle Jane, O18.2, 720 Warnke Andreas, P82, 757 Washizuka Shinsuke, P187, 780 Wasserman Wyeth W., O12.3, 711 Watras Lisa, O22.6, 725 Watson S., S3.3, 688 Watson Stanley J., O2.1, 695 Webb B.T., O27.6, 735 Webb Bradley Todd, O5.6, 702; O16.6, 718 Weber Barbara, P231, 790 Weglarz Monika Dmitrzak, P242, 793 Wei Tang, P59, 751 Weickert Cynthia, O16.1, 717 Weinberger Daniel, O3.3, 697; O5.4, 702; O5.7, 702; O16.1, 717 Weinberger Daniel R., O25.7, 732 Weissman Myrna M., O1.3, 694 Weissman Sherman M., O3.4, 698 Weksberg Rosanna, P299, 805 Wellek Stefan, P07, 740 Wells J. Elisabeth, P33, 746 Wendt Lennart von, O6.3, 703; P124, 766 Wendt Taina Neiminen-von, P124, 766 Wenzler Denise, O24.1, 729; P70, 754 Werge Thomas, P267, 798 Wernicke C.V., O4.9, 701 Wetsel William, P199, 783 White M., S4.1, 688 Whittinger Naureen, O24.2, 729 Wiegand R., S8.1, 691 Wienker Thomas F., O2.6, 696 Wigg Karen, O24.3, 729; O24.4, 729; O24.6, 730; O24.8, 730; P68, 753 Wigg Karen G., O22.1, 724
824
Author Index
Wijmenga Cisca, O6.2, 703 Wikman F.P., P53, 750 Wikman Friedrik P., P24, 744 Wildenauer Dieter, O16.7, 718 Wildenauer Dieter B., O3.1, 697; O3.8, 699 Wildenauer Dieter Bernd, O11.5, 711 Wilffert Bob, P230, 790 Wilkinson Jennifer, P266, 798; P290, 803 Wilkinson L., S9.3, 692 Wilkinson Lawrence, O9.1, 708 Wilkosc Monika, P62, 752; P64, 752; P65, 753; P252, 795 Wille David, P231, 790 Willemsen Gonneke, O9.5, 709; O22.7, 725; O22.8, 726 Williams Hywel, O3.1, 697; P260, 796; P261, 797; P289, 803 Williams Julie, O6.8, 704 Williams Nigel M., P287, 803 Williams Nigel, O3.1, 697; O21.3, 723; P179, 779; P260, 796; P261, 797; P289, 803 Williamson R., P103, 761; S6.2, 690 Williamson R.W., P107, 762 Willis Alecia, O5.3, 701 Willour Virginia L., O13.1, 712 Wilmink F.W., P210, 785 Wilson Alexander, O22.3, 724 Wilson Sarah, O10.3, 709 Win Maartje de, P60, 751 Winchester Catherine, P250, 794 Winchester L., O28.3, 736 Winchester Laura, P130, 767 Windpassinger Christian, O14.1, 714 Winstock Adam, P227, 789; P228, 789 Wiste Anna, O4.7, 700; P27, 744; P311, 808 Woldbye David, P267, 798 Wolff Kim, P227, 789; P228, 789 Won Kim Jong, P254, 795 Wong Albert, O9.3, 708 Wood Alexis Caroline, P126, 767 Wood Joel, O3.6, 698 Woolley James, P157, 773 Wormley Brandon, O4.3, 699; O5.6, 702; O16.3, 717 Wormley Wormley, P282, 801 Wormly Brandon, O17.3, 719 Wray Naomi, O5.5, 702
Wray Naomi R., P39, 747 Wright Clare, P90, 758 Wright H.H., O20.4, 723 Wright Harry, P99, 761 Wulff Maren, P95, 760; P262, 797 Wurst W., 686; W2, 686 Wust Stefan, P63, 752 Xian Hong, O7.5, 705 Xiangning Chen, O17.3, 719 Xiaocai Gao, P59, 751 Xiaohui Xu, O28.5, 736 Xiaoming Sheng, P59, 751 Xingwang Li, P59, 751 Xiong Lan, P268, 798 Xu Xiaohiu, P67, 753; P117, 765 Xu ZhiWen, P278, 800 Xue Hong, P278, 800 Yamada Mitsuhiko, P220, 788; P303, 806; P304, 807 Yamaki Lesley, O22.6, 725; O28.2, 735 Yamamoto Tadashi, P191, 781 Yamanouchi Yoshio, O8.4, 707; P220, 788 Yamashita Hideji, P133, 768 Yamasue Hidenori, P237, 791 Yanagi Masaya, P88, 758; P176, 778 Yang Mao-Sheng, P120, 766 Yang May, O14.5, 715 Yang Ueng-Cheng, P248, 794; P284, 802 Yang Wei-Chih, P248, 794 Yaping Tang, O29.4, 737 Yazdanpanah Simin, P169, 776 Yeon Park Dong, P254, 795 Yevtushenko Olga, P226, 789 Ylisaukko-oja Tero, O6.3, 703; P124, 766 Ylstra B., O11.3, 710 Yoneda Hiroshi, P201, 783; P241, 792 Yonglan Zheng, P59, 751 Yongyong Shi, P59, 751 Yoshida Keizo, P220, 788 Yoshii Mitsunobu, P156, 773 Yoshimura Reiji, O8.4, 707 Young Allan, P12, 741; P17, 742 Young Bryan D., O23.1, 726 Young James, P266, 798; P290, 803 Young L. Trevor, O13.3, 713 Yu Zhiliang, P278, 800
Yuan Xiaobo, O4.1, 699 Yun Sun, P59, 751 Yurov Yuri, P140, 770 Zˇagar Tina, O26.3, 733 Zai Clement, O19.1, 721; O19.3, 721 Zai Gwyneth, O7.7, 706 Zalar Bojan, P235, 791 Zandi P., W1, 685 Zandi Peter Z., O13.1, 712 Zanelli Julanta, P279, 801 Zanten J. Van, O11.3, 710 Zarafoniti Eleni, P125, 766 Zhang Bo, O4.1, 699 Zhang Jianping, O3.6, 698 Zhang Ruimin, O4.1, 699 Zhao Alice, O21.2, 723 Zhao Chengfeng, P34, 746; P35, 746 Zhao Cunyou, P278, 800 Zhao XinZhi, O20.1, 722 Zhao Zhongming, O16.3, 717 Zhou Kaixin, O17.1, 719; O20.3, 722; O28.5, 736; O28.6, 736 Zhou Xianjin, O13.4, 713 Zhu Shao, O4.1, 699 Ziegler Sigrid, O9.3, 708 Zihl Josef, P09, 740 Zijian Zhang, P59, 751 Zinkstok Janneke, O7.6, 705; P60, 751 Zisaki Aikaterini, P314, 809 Zitman F.G., P63, 752 Zivkovic Lada, O12.1, 711 Zobel A., P103, 761; S6.2, 690 Zoccolillo Mark, P04, 739 Zoellner S., W1, 685 Zohar Ada H., P146, 771 Zompo M. Del, P57, 751 Zompo Maria Del, P44, 748; P56, 750; P167, 776 Zorko Maja, O26.3, 733 Zorzi Claudia, P158, 774 Zubieta Jon-Kar, O1.8, 695 Zuchner Stephan, O12.5, 712; P99, 761 Zung Stevin, P37, 746; P159, 774; P223, 788 Zupanc Tomazˇ, P72, 754 Zutter Sonia De, O13.5, 713; O13.6, 713 Zvolsky´ Petr, O23.7, 728 Zwaan Martina de, O18.5, 721; P145, 771
