Metabolic, mitochondrial and behavioral correlations with transcriptional profiles from the CA1 and DG hippocampal regions of young and aged rhesus macaque. Increasing evidence indicates that obesity correlates with impaired cognitive performance during normal aging and is a major risk factor for Alzheimer’s disease. However, little is known regarding how peripheral metabolic variables affect cellular pathways in brain regions important for memory. Brain inflammation, mitochondrial dysregulation, and altered transcriptional regulation have all been found to occur with aging, and recent microarray analyses in rodent models have linked these processes and others to age-related memory impairment. However, whether these brain changes are also associated with metabolic variables is not known. Aging monkeys exhibit several metabolic changes similar to those seen in humans. Here, we tested peripheral-brain relationships in six young (7.0 +/- 0.3 years) and six aged (23.5 +/- 0.7 years) female rhesus monkeys. Animal cognition was gauged with a variable delay task; blood constituents were assessed with a serum chemistry panel emphasizing markers of metabolic dysfunction; mitochondrial function was measured from the hippocampus of one hemisphere; and the CA1 and dentate gyrus regions of the other hippocampus were dissected out for gene expression microarray analysis. Aged animals showed reduced performance on the behavioral task, exhibited aspects of metabolic dysregulation including increased insulin, triglyceride, and chylomicron levels (consolidated into a peripheral metabolic index), and showed a significant age-related reduction in State III oxidation, a measure of mitochondrial function. Microarray analyses revealed hundreds of genes that correlated with the peripheral metabolic index. However, DAVID statistical pathway analyses showed that upregulated inflammatory genes in CA1 and downregulated transcriptional regulation genes in dentate gyrus and CA1 were particularly overrepresented among genes correlated with the peripheral index. Thus, the association of metabolic variables with specific neuropathological processes in different regions of the hippocampal formation may have implications for mechanisms through which peripheral metabolism alters the risk for Alzheimer’s disease.
Keywords: Rhesus hippocampal aging
Overall design: Six young (7.0 +/- 0.3 years) and six aged (23.5 +/- 0.7 years) female rhesus monkeys (Macaca Mulatta) were behaviorally assessed on the delayed match to sample task. Briefly, on each trial the subject observed the experimenter place a treat in either the left or right covered food well, the food wells were screened, and after a variable (zero- no delay, short- 10 seconds, long- 30 seconds) delay, the screens were removed. The subject initially selecting the food well containing the treat was rated as a correct response. Each animal received thirty trials (ten of each delay, in random order) per session, and four sessions over the course of one week. Animals were anesthetized with ketamine, venous blood was drawn for peripheral metabolic assays. Metabolic assays included standard blood chemistry and lipoprotein profiles. Animals were fatally dosed with pentobarbital and brains were removed. Hippocampi were dissected: one hippocampus was used for mitochondrial assays and proteomics, and the other was sub-dissected into CA1 and DG sub-regions for micorarrays. Procedures for mitochondrial assays and proteomics are as published previously (Sullivan et al., 2007). Hippocampal sub-region mRNA extraction, labeling and microarray (Affymetrix Rhesus) hybridization proceeded according to standard protocols (Affymetrix; Blalock et al., 2003, 2004; Rowe et al., 2007). Signal intensity and presence call data were derived with the MAS5 (Affymetrix) algorithm. Data were transferred to Excel, Bioconductor, and Multiexperiment Viewer for further analysis. Statistical tests included paired t-tests across region, and unpaired t-tests across age within each region. Post hoc correlation analyses with metabolic, mitochondrial, and behavioral data were used to refine aging related genes and biological pathways. Lists of regulated genes were analyzed for functional overrepresentation using DAVID, Onto-Express, and GenMapp.
Less...