Oral Presentation 9th GeneMappers Conference 2012

Dissection of Human QTLs Influencing Serum BDNF in Mexican American Pedigrees (#37)

Marcio A Almeida 1 , Joanne E Curran 1 , David C Glahn 2 , John Blangero 1
  1. Texas Biomedical Research Institute, San Antonio, Texas, United States
  2. Department of Psychiatry, Yale University, New Haven, Connecticut , United States

Brain-derived neurotrophic factor (BDNF) is a polypeptide involved in regulating neuronal survival, differentiation, and outgrowth. Serum levels of the BDNF protein have been associated with a number of psychiatric diseases such as depression and also with neurodegenerative diseases. We measured sBDNF levels in a sample of 860 San Antonio Family Heart Study (SAFHS) individuals who are members of extended Mexican American pedigrees. Quantitative trait linkage analysis utilizing high density SNPs to calculate identity-by-descent probabilities were employed to identify genomic regions harboring causal genes. Using this approach, we localized four potential quantitative trait loci (QTLs). Observed LOD scores were 4.45 (chromosome 22), 3.29 (chromosome 1), 2.41 (chromosome 19) and 2.39 (chromosome 5). We then examined existing deep sequence data for a subset of 491 individuals to search for causal variants/genes in these genomic regions. WGS data (obtained as part of the T2D-GENES Consortium) were available for 491 total individuals and was examined in the QTL regions.  Variants, prioritized by their functional potential, within a QTL-specific 1-LOD support interval were tested for association with BDNF levels using a measured genotype model in a variance component framework. Several of the QTL regions yielded substantial evidence for causal gene identification. For example, the chromosome 22 QTL, we identified a non-synonymous variant (N759D) in the PLXNB2 gene that was significantly associated with BDNF. This variant is predicted to be strongly deleterious. The PLXNB2 gene is involved in axon growth and guidance. Similarly, the chromosome 5 QTL strongly pointed towards involvement of the IRX1 gene with the identification of a highly associated (p = 1×10-6) evolutionarily conserved private novel variant. This gene represents an excellent causal candidate with a known major role in neurodevelopment.  These results empirically demonstrate the potential of deep sequence data to follow up QTL localization studies to identify novel likely causal variants/genes.