Genes of the serotonergic system & susceptibility to psychiatric disorders : A gene-based haplotype approach

Abstract: Psychiatric disorders are complex, non-mendelian disorders. Complex disorders are ultimately determined by a number of genetic and environmental factors, and the effect of each factor may be obscured or confounded by others. Although detection and precise characterization of the relationship and interactions between genes involved and/or environmental factors may not be ascertained by any available methods, several strategies of investigation may be used. Candidate gene analysis is one disclosing approach, which was used in this thesis to test for associations between a particular gene variant and a disease. This thesis work aimed to design a simple strategy to construct sets of polymorphic loci limited by gene boundaries, referred to as gene-based haplotypes, to be applied in case-control clinical studies, as opposed to the classical, ancestral haplotype block approach, which is mostly used for genome-wide and large scale population analyses. The gene-based haplotype approach developed here is not influenced by parental transmission data, is cost and time effective, and is more informative than single locus associations. Also, linkage disequilibrium (LD) analysis within the gene and between the markers increases the likelihood to detect concealed causative alleles. This strategy was applied to two candidate susceptibility genes of the serotonergic system implicated in the pathophysiology of several psychiatric disorders: (i) the serotonin transporter (5-HTT) gene, and (ii) the tryptophan hydroxylase 1 (TPH-1) gene. Focusing on these genes, the strategy was adapted to a classical case-control design to seek for association between specific gene-based haplotypes that are common in the population while potentially carrying risk for the disorder. Three different psychiatric disorders were explored: Major Depressive Disorder (MDD), Schizophrenia (SCZ) and Borderline Personality Disorder (BPD). Single locus analyses provided a few positive associations with the disorders, which were mostly weak especially after correction for multiple testing. This was expected, as it is often the case for single-locus association studies. However, when multi-loci sets were constructed using a population genetics algorithm, several haplotypes were identified that displayed strong statistical associations with disease. The significance level obtained for haplotype markers in association studies was substantially higher than that observed for single loci.