Genetic studies in mood disorders

Sammanfattning: Mood disorders, including bipolar disorder (BPD) and major depressive disorder (MDD), are highly complex psychiatric disorders. Decades of genetic studies have generated a large number of putative genetic susceptibility variants. However, with exception of CACNA1C, SYNE1 and ANK3 in BPD no robust association has as yet been identified. In this thesis my aim was to find predisposing genetic risk factors for mood disorders. In paper I, my hypotheses were that the FKBP5 gene is a risk gene for MDD, contributes to severity and is involved in treatment response to an antidepressant, Citalopram. We tested for association of three markers using the STAR'D cohort (Level 1). Rs1360780 was significantly associated with MDD. Rs4713916 was significantly associated with remission following treatment with Citalopram when two study populations were analyzed together. We determined that there is a stratification issue and no correlation can be made for treatment response or severity of MDD. In paper II, my hypothesis was that candidate genes for MDD are acting synergistically to contribute to risk for MDD. We applied three different algorithms to evaluate SNP-SNP interactions. Although none of the interactions survived corrections for multiple comparisons, our results contribute valuable information to future genetic studies in MDD. The logistic regression methods identified large interaction effects. None of the top interactions explain a large proportion of MDD in the general population. Among the top interactions, none of the three algorithms identify identical pairs of markers for risk of MDD. Moreover, none of the top ranked interactions has previously been implicated to act synergistically in MDD-susceptibility. We also noted that markers selected for predicted interaction effects were not among the top interactions. In paper III, my hypothesis was that rare and highly penetrant large structural genomic variations (CNVs) increase the risk for BPD. We searched for CNVs across diagnostic boundaries and included individuals with BPD, schizophrenia (SZ) or schizoaffective disorder (SA). To increase the possibility that the CNV should be highly penetrant we searched for CNVs in affected individuals in BP-pedigrees and identified CNVs in the MAGI1 gene in two families and showed that it was more frequent in individuals with BPD, SZ and SA than controls. In paper IV, my hypothesis was that inherited CNVs contribute with risk to BPD irrespective of their frequency. We developed an algorithm that combines linkage-data with the CNV content within and across families. We identified one significant region with a CNV that maps to 19q13, and stretches over the PSG genes. The PSG proteins has been shown to activate TGF-?. Moreover, two CNV SNPs are reported as likely eQTL’s for regulation of NF?B. Thus, this CNV involves several putative molecular targets in BPD-etiology. In conclusion, the work conducted in this thesis has contributed to our knowledge of the etiology of mood disorders. For BPD we found two new susceptibility loci. In the analysis of MDD we increased the knowledge of the genetic interacting landscape in 63 candidate genes. We also showed that FKBP5 is associated with risk for MDD.