Genetic factors linking metabolic state, behaviour and adult neurogenesis (DFG-GA CR)
Final Report Abstract
We have previously shown that adult hippocampal neurogenesis is differentially regulated in the spontaneously hypertensive rat (SHR), an animal model of human hypertension and metabolic syndrome, compared to its genetic control. In this study, we applied system genetics tools to indentify genes and pathways that covary with DG neurogenesis and metabolism. We took advantage of natural genomic variation in the genetic reference population of 30 HXB/BXH rat recombinant inbred (RI) strains derived from SHR/Ola and Brown Norway BN-Lx rats. HXB/BXH strains were used to study metabolic diseases for over 20 years; therefore a large collection of metabolic measures and gene expression in the liver, muscle, peritoneal fat, adrenal gland, aorta and ventricle tissues are available for regression analyses in the public databases. Employing Ki67 immunochemistry and BrdU labelling of newborn cells, we provided a quantitative description of two milestones of hippocampal neurogenesis: precursor cell proliferation and survival of newborn cells in the DG of HXB/BXH rats. Both traits followed normal distribution, consistent with polygenic regulation of adult neurogenesis. We also measured gene expression in the hippocampus in all RI and parental strains using microarrays. Finally, we identified a set of 15 physiological phenotypes significantly correlating to neurogenesis, which could be grouped into several functional categories: catecholamine synthesis in the adrenal gland, stress response, glucose and lipid metabolism. Two complementary approaches were used to identify regulatory genes and cell biological pathways underlying these correlations. First, we performed quantitative trait loci (QTL) mapping for neurogenesis and correlating traits, as well as for gene expression data sets. Common QTL for serum glucose levels, net neurogenesis and Telo-2 interacting protein 2 (Tti2), whose expression showed negative correlation with these traits, was detected on chromosome 16. Tti2 participates in the assembly and stabilisation of functional complexes containing phosphatidylinositol 3-kinase-related protein kinases (PIKKs). Our preliminary data suggests differential regulation of mTOR pathway in the parental strains. In the second approach, applying weighted gene coexpression and gene ontology enrichment analyses to gene expression data sets, we identified modules of coexpressed genes, which covaried with physiological traits. The modules were enriched in genes assigned to mitochondrion, particularly respiratory chain, ER and protein folding, and inflammation gene ontology terms. Importantly, hub genes in these clusters were contained within enriched GO terms. Thus, using unbiased systems biology tools and incorporating publicly available data, we identified cell biological processes and genes as candidate links between neurogenesis and metabolism in the rodent model of metabolic syndrome. Natural genomic variation, which by and large results in subtle changes of gene expression rather than loss- or gain of function, is a valid model of variance in the human population. Also, monitoring covariance of the whole transcriptome simultaneously enables the detection of functional gene assemblies related to a phenotype of interest. As a next step, the role of the quantitative trait and hub genes in the regulation of correlating phenotypes needs to be experimentally validated.
