P2–003: Flux and metabolite pathway mapping in the aging and Alzheimer's disease hippocampus

We published two studies on biosimulations for cholesterol and metabolic pathways in the hippocampus for aging and progressive Alzheimer's disease (AD). The objective of this study was to demonstrate the utility of flux and metabolite pathway mapping for visualizing the integrated systems networks. We simulated the hippocampus by linking genome-wide transcription profiles from aging, age matched controls and incipient, moderate, and severe cases of AD (from NCBI GEO GSE 11882 and 1297) with corresponding reactions in a biosimulation model, i.e., to derive parameters for a deterministic kinetic model, Transcriptome-To-Metabolome™ (TTM™) Biosimulation. The model simulates 16 core metabolic pathways, using COPASI®. This model has 4 compartments, 220 species (metabolites), 295 reactions, and 428 parameters derived from expression levels for 238 genes. Time series simulations were run using ordinary differential equations to generate level values for metabolites and flux values for reactions. Pathway maps were generated with NodeXL to include changes in metabolite levels, reaction fluxes, and gene expression levels for the reactions. NodeXL is freeware that integrates into Microsoft Excel. Nodes were species and edges were reactions from the TTM™ model. The nodes represented as spheres where both size and color represented percent differences in metabolite levels were easily distinguished across study groups. Edges were directional and displayed as arrows where size represented percent difference in flux and color the difference in expression level across study groups. Colors ranged from green for decreases to brown for zero and orange to red for increases. Pathways had to be mapped separately and some split in order to allow visual analyses. Figure 1 shows the cholesterol pathway comparing the aged with youngest group; note how easily one can assess the extent of decrease in all metabolites and fluxes. Also note how even in cases of increased gene expression (red arrows) flux was low because the substrates were low. Export of maps into image files allowed assembly of Powerpoint slides for film-strip inspections of pathway changes. Legend: Spheres are reactants (metabolites) and arrows are reactions. Both color and size of spheres reflect percent difference between groups; Only size of arrows reflects magnitude of percent difference between groups for flux in the direction of the arrow. Label on the arrows is the gene(s) responsible for the reaction and the color indicates percent difference in gene expression level. Green are lowest (negative) percent differences and orange to red are increasing positivity; brown color is no change. The highest increase was 25 hydroxycholesterol at 18%, bottom center. Two spheres are labeled (top left & bottom center) with percent change metabolite level values; cholesterol (bottom center) sphere size and color = −85% for scalar information. Flux changes range from −20 to −100%. The key pathways that make up the complete cholesterol biosynthetic pathway are labeled. ∗-mevalonate & squalence; ∗∗-FFMAS, ∗∗∗-TMAS.