Determinants of Neurotransmitters in Cerebrospinal Fluid and Plasma : from Seasonality to Quantitative Genetics

Most psychiatric conditions are complex genetic as the largest proportion of genetic variance is likely to derive from many genetic variants of small effect. Nonetheless, given the intricacies of the human brain and the heterogeneous nature of psychiatric disease entities, dissecting the genetic mechanisms underlying psychiatric conditions has proven convoluted. Intermediate phenotypes - quantifiable continuous traits pertinent to (disease) phenotypes- may be more directly linked to underlying genetic mechanisms than the abovementioned heterogeneous disorders. Such knowledge in turn may deepen the understanding of biological determinants underlying psychiatric illness. The studies described in this thesis have therefore aimed to uncover associations between human genomic variation and intermediate phenotypes measured in cerebrospinal fluid (CSF) and plasma that have relevance to psychiatric conditions, i.e. concentrations of neurotransmitters, coagonists and a neurotrophin. Cerebrospinal fluid was chosen as it is the body fluid compartment most proximate to the central nervous system; blood plasma is relatively straightforward to obtain and metabolite measurements in plasma have proven highly reliable. Quantitation of monoamine metabolite (MM), γ-Aminobutyric acid (GABA), and the N-methyl-D-aspartate receptor (NMDAR) coagonists was conducted using High-Performance Liquid Chromatography (HPLC) in subjects from the general population (with numbers ranging from >300 to 479 depending on the constituent under investigation). For BDNF quantitation, an enzyme-linked immunosorbent assay (ELISA) for a sample of 61 healthy subjects was chosen. Whole-genome data were generated using the Illumina Human OmniExpress Beadchip and imputation, resulting in 5,767,231 SNPs single-nucleotide polymorphisms (SNPs). We first assessed the possible influence of candidate covariates on monoamine metabolite levels and found a peak in serotonin (5-HT) turnover in spring. Seasonality turned out to explain over 5% of the variance in CSF 5-HIAA. The s-allele of the 5-HT transporter polymorphic region (5-HTTLPR) polymorphism dose-dependently determined the degree to which 5-HT turnover varies across seasons. Increased 5-HT seasonality in turn was correlated with depressive symptomatology. We then found that an intronic variant in ERBB4 is associated with CSF GABA levels in a dose-dependent manner. BDNF Val66Met homozygosity was shown not to exert substantial effects on plasma BDNF levels. By genome-wide analyses of N-methyl-D-aspartate receptor (NDMAR) coagonist concentrations, quantitative trait loci at genes encoding a transporter protein (SLC6A20) and enzymes (PRODH and DAO) were identified. Biologically tenable associations in both plasma and CSF were thus detected. A locus at SSTR – encoding a neuropeptide relevant to CNS signal transduction- was associated with CSF monoamine metabolite (MM) levels. This locus in turn was found to associate with mRNA expression of PDE9A, a gene previously implicated in depression and antidepressant response. Finally, although causality and directionality are unknown, concentrations of D-amino acids in plasma and CSF differed between smoking and non-smoking individuals. In conclusion, the studies presented in this doctoral thesis demonstrate that CSF is a powerful and informative target for seasonality and genetic research. Similarly to most psychiatric conditions, CSF constituents are influenced by genetic (polymorphisms) and environmental (seasonal) factors. Ultimately, the highlighted metabolic, cell signaling and transporter pathways may provide novel targets for psychopharmacological treatment modalities