Isotopomer modeling and analysis in metabolic systems

Various methods are developed to analyze metabolic fluxes in isolated and intact metabolic systems using metabolically active molecules labeled with stable isotopes. Models that predict the incorporation of isotope from these molecules into various metabolic intermediates and products within the biological system are developed. These models allow the calculation of relative fluxes associated with different biochemical pathways from measurements of the incorporation of isotope into these various intermediates. Relative fluxes of the citric acid cycle and gluconeogenesis were estimated to distinguish important metabolic pathways. When isotope-labeled substrate is depleted, it is demonstrated that current analysis techniques cannot distinguish between metabolically distinct situations. Mathematical models are developed to (i) correct molecular mass number distributions for contributions from natural stable isotope abundance, (ii) determine molecular positions of labeled atoms from molecular mass-number distributions, and (iii) analyze relative fluxes of metabolic pathways. The new technique of accounting for natural stable isotope contributions avoids errors from less rigorous approaches. This technique is implemented in a self-contained software that allows efficient correction of large data sets.;A more comprehensive model of the citric acid cycle is developed which includes several processes shown to be significant from measured isotope labeling patterns of a number of hepatic metabolites. Parameter estimation techniques allow the determination of all relative fluxes within the model based on isotope incorporation information of five metabolites. These studies are the first to quantify the extent of isocitrate dehydrogenase reversibility in isolated metabolic systems. A reduced form of this model allows the calculation of metabolic flux ratios pyruvate carboxylase to pyruvate dehydrogenase and to the citric acid cycle from positional labeling data of molecules excreted in the urine after administration of labeled substrates. Model simulations of non-uniform substrate depletion in the liver leads to a self consistent model of gluconeogenesis and explains why the assumption of single labeled precursor pools can lead to erroneous conclusions about mechanisms.