Projects

Project S01 will be responsible for the recruitment and support the basic characterization of the central cohort of 1050 subjects for projects A01-A04. S01 will coordinate the acquisition and quality assurance of a basic neurocognitive and neuroimaging battery (anatomical MRI, resting state MRI, fMRI during inhibitory control), and the biobanking of blood, as well as genome-wide genetic and epigenetic analyses. S01 will provide 15 subprojects of the CRC/TRR 265 with pre-processed psychometric-, MRI- and genetic data to be included in the outcome variable analyses.

WP6 RESEARCH PROJECT Integrative analyses of genetic, epigenetic, and environmental vulnerability factors for affective disorders Prof. Dr. Marcella Rietschel1, Dr. Stephanie Witt1, Dr. Axel Krug2, 1University of Heidelberg, Central Institute of Mental Health (ZI) Mannheim, 2University of Marburg, Department of Psychiatry Summary The aim of this workpackage is to identify how genetic, epigenetic, and environmental factors impact on the etiology of affective disorders. This will be achieved by correlating these factors with each other, with categorial diagnoses, and with endophenotypes. The latter will be assessed in magnetic resonance imaging (MRI) and immunological studies performed in WP1 and WP3. To ensure stateof- the-art coverage of all variants known to be of importance to psychiatric disorders, subjects will be genotyped with the PsychChip, a custom 20,000 probe multiplex array that is specific for psychiatricdisorders. For the statistical analyses, a convergent approach will be applied in collaboration with WP7. This will comprise multimarker-, polygenic score based-, and pathway analytical methods, and will take into account environmental factors. The results of our analyses will provide data concerning diagnosis-specific factors, as well as factors which are of relevance across diagnostic boundaries. Background and own previous work: The heritability of bipolar disorder (BD) has been reported to be as high as 80%, while that of major depressive disorder (MDD) ranges between 40 and 70%. However, the scientific community is calling the conceptualisation of these two disorders as separate entities increasingly into question 317. Formal- and molecular genetic studies also indicate a substiantial overlap in their etiology. The number of identified vulnerability genes for affective disorder is increasing, and our group has contributed substantially to these findings 20,22,318. Identification of the first vulnerability genes has identified various neurobiological mechanisms in the etiology of affective disorders 11,319, and the use of more refined phenotypes is likely to generate further insights. For example, a growing number of studies, including studies performed by our group, are demonstrating the important impact of environmental factors on affective disorders. In a GxE study, we found that BDNF variation moderated the impact of early psychosocial adversity on plasma brain-derived neurotrophic factor and depressive symptoms 38. Given that affective disorders are heterogeneous clinical conditions, it is unlikely that single genetic variants are associated with categorial disease. A more plausible hypothesis is that such variants are associated with sub-dimensions. Therefore, an important prerequisite for understanding why particular individuals will develop the disease whereas others will not, is to elucidate the influence of genetic variation at the level of clinical subdimensions and/or endophenotypes, e.g. MRI or immunological data. With respect to clinical subimensions, we recently found that a genome-wide significant association between BD and a variant in NCAN 22 was due to an association with an overactivity dimension that is also present in MDD 11. Interestingly, NCAN deficient mice displayed a similar behavioral phenotype, rendering them a novel animal model for pharmacological testing of this clinically relevant subdimension. With respect to endophenotypes, the imaging genetic approach has proven to be most successful in terms of delineating the possible mechanisms of action of vulnerability genes 174,280,320,321. This approach will moreover help to elucidate how known environmental risk factors act on the level of individuals 115. Besides genetic factors, epigenetic modifications – which do not alter the genetic code, but regulate gene expression – modulate the susceptibility to psychiatric disorders 50,53,322. Such modifications include DNA methylation. In a recent study, we demonstrated that maternal prenatal stress led to differential methylation in a large number of genes in the newborn. In a collaborative conver- 24 gent approach (with Moshe Syf, Univ. Montreal), which included use of primate animal models, we detected a novel and reliable vulnerability gene for stress-related disorders 323. Most of the genetic studies of affective disorders performed to date have analysed single markers. Using multi-marker approaches, such as pathway/gene set based analysis or polygenic score-based analysis, information contained in many small association signals throughout the genome which is too weak to be detected by single marker tests can be cumulated. This is also a powerful approach for testing shared genetic liability across categorial diagnoses 324,325, as well as across symptom dimensions or sub-groups defined according to more refined phenotypes. Work Program: Genes of interests will be analysed throughout the course of the FOR, once a substantial number of the study subjects have been characterized with the PsychChip. The two genes of interest (NCAN, BDNF) will also be analysed in the animal model (WP2), and were selected because of the following criteria: confirmed association with affective disorders; feasible for use in animal studies (WP2); expressed in the synapse to be a gene of interest for WP4; and of relevance to the immunological hypotheses addressed in WP3. Genotyping and methylation studies will commence at the end of year 1 with n=500 subjects and will be ongoing. Analyses will start as soon as the first findings are available. Data will be replicated in the full cohort, once recruitment is complete. In year 1, we will also conduct feasibility studies to determine which sub-phenotypes are useful and whether the proposed data mining technique is feasible. 1. Genotyping of subjects recruited in WP1 for NCAN and BDNF. 2. Methylation analysis of NCAN and BDNF (pyro-sequencing). 3. Genotyping with the PsychChip (Illumina custom array), which is being developed by the Psychiatric Genomic Consortium. The PsychChip contains 20,000 SNPs and CNVs of relevance to psychiatric disease. SNP selection focuses on top hits from GWAS and meta-analyses, exome SNPs, SNPs from the MHC region, and CNVs implicated in psychiatric disease. 4. Polygenic analyses to quantify the impact of aggregated genetic variation on sub- and endophenotypes in collaboration with WP7. Using an integrative approach, all results from the genotyping and methylation analyses will be combined. Multimarker-, polygenic, and pathway analyses will then be performed to identify as yet undetected genotype-, GxE, and phenotype relationships. One of the available data mining techniques is used in retail market research. Our group recently adapted this method by developing the new software tool RUDI (RUle Discoverer) in order to identify frequent and characteristic genotype patterns showing strong association to phenotype clusters (http://rudi-genetics.net). Outlook: In the second funding period, we will perform genome-wide association studies (GWAS), genome-wide methylation studies in extreme groups, e.g. therapy outcome (good vs. bad responders), and high/low risk groups and transition to disease. We will also take into account environmental factors, such as stress (assessed via questionnaires and biological parameters, i.e. hair cortisol levels). Project specific references Miró X, Meier S, Dreisow ML, Frank J, Strohmaier J, Breuer R, Schmäl C, Albayram Ö, Pardo-Olmedilla MT, Mühleisen TW, Degenhardt FA, Mattheisen M, Reinhard I, Bilkei-Gorzo A, Cichon S, Seidenbecher C, Rietschel M, Nöthen MM, Zimmer A. Studies in humans and mice implicate neurocan in the etiology of mania. Am J Psychiatry. 2012 Sep;169(9):982-90. Rietschel M, et al. Genome-wide association-, replication-, and neuroimaging study implicates HOMER1 in the etiology of major depression. Biol Psychiatry. 2010 Sep 15;68(6):578-85. Krug A, Krach S, Jansen A, Nieratschker V, Witt SH, Shah NJ, Nöthen MM, Rietschel M, Kircher T. The Effect of Neurogranin on Neural Correlates of Episodic Memory Encoding and Retrieval. Schizophr Bull, in press. Schizophrenia Psychiatric Genome-Wide Association Study (GWAS) Consortium. Genome-wide association study identifies five new schizophrenia loci. Nat Genet. 2011 Sep 18;43(10):969-76. doi: 10.1038/ng.940. Psychiatric GWAS Consortium Bipolar Disorder Working Group. Large-scale genome-wide association analysis of bipolar disorder identifies a new susceptibility locus near ODZ4. Nat Genet. 2011 Sep 18;43(10):977-83.