Interpreting the phosphocreatine time constant in aerobically exercising skeletal muscle.

TO THE EDITOR: Francescato et al. (1) contrast the lack of interindividual correlation between the phosphocreatine (PCr) time constant ( PCr) and the ADP concentration ([ADP]) change from rest to steady state ( [ADP]S) during moderate aerobic exercise with the fact that “. . .ADP concentration is commonly thought to be one of the main feedback signals controlling mitochondrial respiration” (1), a view supported by observations that initial postexercise PCr resynthesis rate [ mitochondrial ATP synthesis rate (JP) at end exercise (5)] has a roughly hyperbolic (Michaelis-Menten) relationship to [ADP] (2, 3, 7). The general idea is that JP/JP,MAX is some function f(X), where JP,MAX is “mitochondrial capacity” and a feedback signal X responds via the creatine kinase equilibrium to mismatch between JP and ATP demand (JD) (2, 3). Ignoring the possible role of inorganic phosphate (3, 7), assume that X [ADP], which is proportional to [TCr]/[PCr], where [TCr] is total creatine concentration and [PCr] is PCr concentration. Exponential PCr kinetics in aerobic exercise imply a hyperbolic JP-[ADP] relationship with Km 50 M (2, 3), such that PCr d[PCr]/dJP [TCr]/JP,MAX. Accordingly, PCr varies directly with [TCr] (4) and inversely with mitochondrial content (6). At steady state, JD,S JP,S, so [ADP]S f (JD,S/ JP,MAX); to predict [ADP]S we also need resting [ADP], [ADP]R f (JD,R/JP,MAX), assuming naively (2, 3, 7) that this model applies over the full range (3). Consider three sources of interindividual variation: if only JP,MAX varied, then PCr and [ADP]S would correlate [which they do not (1)]; variation in JD,S would only affect [ADP]S, and variation in [TCr] would only affect PCr, although [ADP]S calculated assuming “normal” [TCr] would correlate negatively with PCr. Analogously, in the linear model, the change in [PCr] from rest to steady state ( [PCr]S) PCrJD,S (5) so variation in JD,S would affect only [PCr], consistent with the first-order system property that response kinetics are independent of perturbation (1); however, variation in JP,MAX would make [PCr]S and PCr covary [which they do not (1)] without obliging rejection of the linear model, e.g., in favor of feedforward mechanisms (1). Francescato et al. (1) also conclude, from the correlation between PCr and [PCr]R, that PCr concentration is “one of the main controllers of oxidative phosphorylation” (1). But consider three sources of variation in [PCr]R: if only JP,MAX varied, adjustments in [ADP]R would yield a negative PCr-[PCr]R correlation, opposite to that observed (1); variation in JD,R (separate from variation in JD,S) would affect [ADP]R and thus [PCr]R, not PCr; but if [TCr] varied, [PCr]R would vary proportionately to maintain [ADP]R, giving a positive PCr[PCr]R correlation with dln PCr/dln[PCr]R 1, close to that observed (1). In summary, without information on variation in ATP turnover and [TCr], the noncorrelation of PCr with [ADP]S and [PCr]S (1) (and as pHS pHR, each noncorrelation implies the other) has no definite implications for ADP-feedback or -feedforward mechanisms (1). Possible causes of PCr-[PCr]R correlation (1) include interindividual variation in [TCr].