Modeling human muscle metabolism: using constraint-based modeling to investigate nutrition supplements, insulin resistance, and type 2 diabetes

Human muscle metabolism, the biochemical reactions which lead storage and usage of energy, is complex, but important in understanding human health and disease. Optimal muscle metabolism can help maintain a healthy organism by adequately storing and utilizing energy for subsequent use in contraction and recovery and adaption from contraction and exercise. Dysregulated muscle metabolism can lead to insulin resistance and obesity among other health problems. Flux balance analysis (FBA) and constraint-based modeling have successfully elucidated important aspects of metabolism in single-celled organisms. However, limited work has been done with multicellular organisms. The foci of this dissertation are (1) to show how novel applications of this technique can aid in the investigation of human nutrition and (2) to elucidate metabolic phenotypes associated with the insulin resistance (IR) characteristics of Type 2 Diabetes (T2D). First, for human nutrition a novel steady-state constraint-based model of skeletal muscle tissue was constructed to investigate the effect of amino acid supplementation on protein synthesis. Several in silico supplementation strategies implemented showed that amino acid supplementation could increase muscle contractile protein synthesis, which is consistent with published data on protein synthesis in a post-resistance exercise state. These results suggest that increasing bioavailability of methionine, arginine, and the branched-chain amino acids can increase the flux of contractile protein synthesis. Thus, this dissertation introduces the prospect of using systems biology as a framework to investigate how supplementation and nutrition can affect human metabolism and physiology. Second, given the complexity of metabolism, the cause(s) of the altered muscle metabolism in IR remain(s) unknown. Attempting to elucidate this complexity, the constraint-based modeling framework was expanded upon to develop the first in silico analysis of muscle metabolism under varying nutrient conditions and during transitions from fasted to fed states. Systematic perturbations of the metabolic network identified reactions which mimic IR phenotypes: reduced ATP/creatine phosphate synthesis, reduced TCA cycle flux, and reduced metabolic flexibility. Reduced flux through a single reaction is not sufficient to recapitulate the IR phenotypes, but knockdowns in pyruvate dehydrogenase in combination with either reduced lipid uptake or lipid/amino acid oxidative metabolism do so. These combinations also decrease complete lipid oxidation and glycogen storage. Thus, the computational model also provides a novel tool to identify candidate enzymes contributing to dysregulated metabolism in IR