(31)P magnetization transfer measurements of P(i)→ATP flux in exercising human muscle

Fundamental criticisms have been made over the use of (31)P magnetic resonance spectroscopy (MRS) magnetization transfer estimates of inorganic phosphate (P(i))→ATP flux (V(Pi-ATP)) in human resting skeletal muscle for assessing mitochondrial function. Although the discrepancy in the magnitude of V(Pi-ATP) is now acknowledged, little is known about its metabolic determinants. Here we use a novel protocol to measure V(Pi-ATP) in human exercising muscle for the first time. Steady-state V(Pi-ATP) was measured at rest and over a range of exercise intensities and compared with suprabasal oxidative ATP synthesis rates estimated from the initial rates of postexercise phosphocreatine resynthesis (V(ATP)). We define a surplus P(i)→ATP flux as the difference between V(Pi-ATP) and V(ATP). The coupled reactions catalyzed by the glycolytic enzymes GAPDH and phosphoglycerate kinase (PGK) have been shown to catalyze measurable exchange between ATP and P(i) in some systems and have been suggested to be responsible for this surplus flux. Surplus V(Pi-ATP) did not change between rest and exercise, even though the concentrations of P(i) and ADP, which are substrates for GAPDH and PGK, respectively, increased as expected. However, involvement of these enzymes is suggested by correlations between absolute and surplus P(i)→ATP flux, both at rest and during exercise, and the intensity of the phosphomonoester peak in the (31)P NMR spectrum. This peak includes contributions from sugar phosphates in the glycolytic pathway, and changes in its intensity may indicate changes in downstream glycolytic intermediates, including 3-phosphoglycerate, which has been shown to influence the exchange between ATP and P(i) catalyzed by GAPDH and PGK.