Computational modeling indicates anaerobic glycogenolytic ATP synthesis contributes little to quadriceps energy balance during exhaustive bicycling exercise

It has been proposed that anaerobic glycogenolysis may contribute as much as 20% to cellular ATP synthesis during contractile work on basis of indirect evidence from 31P NMR spectroscopic measurements of proton balance in human limb muscles (1). Here, this hypothesis was tested more rigorously using a systems biology approach. We constructed a computational model of energy metabolism in muscle composed of slow-twitch oxidative, fast-twitch oxidative and fast-twitch glycolytic fibers incorporating mitochondrial and glycogenolytic ATP synthetic pathways operating in parallel. Next, we measured the in vivo dynamics of phosphocreatine (PCr), Pi, hexose monophosphates (HMP) and H+ concentrations in quadriceps muscle of human subjects performing in-magnet bicycling exercise using gated 31P NMR spectroscopy (2). The measured PCr dynamics were then compared to model simulations for different magnitudes of glycogenolytic ATP synthetic flux. In contrast to previous indirect analyses (1), we found that anaerobic ATP synthesis contributions of only 10% or less sufficed to explain energy balance during exercise; this in spite of massive activation of glycogenolysis as evidenced by large accumulations of HMP of up to 10 mM directly following exercise (2).