DESCRIPTION (provided by applicant): Large variations exist in hippocampal neuroanatomy in mice. Specifically, the sizes of the intra-and infrapyramidal mossy fiber (IIPMF) terminal fields differ about 4-5 fold between different inbred strains. Earlier experiments from our laboratory have shown that these variations are hereditary to a large extent, with a heritability of 50% or larger. These variations in neuronal circuitry are predictive of the performance of mice in a spatial learning task in a radial maze: animals with larger IIPMF projections commit fewer errors. Radial maze performance is heritable, too, with heritabilities of 25-30%. The genetic correlation between the size of the IIPMF and radial maze learning is around 0.90, indicating that about 80% of the genetic variation is shared. In this application, we propose to test radial maze learning and measure IIPMF sizes in mice from the BXD Recombinant Inbred Strains (RIS). Up until recently, 34 BXD Rl strains were available, but 45 additional strains have newly been developed, thereby greatly enhancing the power and precision with which Quantitative Trait Loci (QTL) can be mapped using this recombinant inbred strain set. Under Aim 1, we propose to localize QTL regulating radial maze learning. Under Aim 2, we intend to localize QTL influencing IIPMF sizes. Under Aim 3, we will identify pleiotropically acting QTL that influence learning abilities and hippocampal structure simultaneously. Finally, we will use publicly available gene expression data to perform a transcriptome analysis using the Internet.based tool WebQTL (www.webqtl.org). The significance of this work lies in the insight it will provide into the causes of normal cognitive variation. Given the fact that over 99% of all genes are conserved between humans and mice, this will advance our understanding of the causes of pathological variation, such as mental retardation and dementia.