Muscle [phosphocreatine] dynamics during exercise: implication for understanding the regulation of muscle oxidative metabolism

In contracting skeletal muscle, phosphocreatine (PCr) is one of the reactants linking mitochondrial respiration with the rate of muscle ATP usage through feedback activation. In contrast to what happens in heart in which [PCr] is stable, in skeletal muscle 31P-MR spectroscopy shows marked changes in [PCr] in response to skeletal muscle activity. The observation of stable [PCr] in working heart over a wide range of contraction levels leads several authors to suggest that there exists a perfectly balanced direct parallel activation of energy supply and demand in heart, but much less intense direct parallel activation, or even no parallel activation, in skeletal muscle (Korzeniewski, 2007), resulting in a weaker homeostasis of the intermediates in the latter tissue during transition from low to high metabolic rate. Interestingly, the intensity of the parallel activation in skeletal muscle may depend on: (i) experimental conditions (muscle stimulated neurally versus electrically); (ii) the context of perfusion (perfused ex vivo where the influence of some factors, e.g. hormones, are absent versus physiologically perfused by the blood stream in vivo); and (iii) the typology of muscle fibres involved in contraction (oxidative versus glycolytic). Keeping in mind points (i) and (ii), therefore only studies carried out in vivo, have the capacity to evaluate the integrated complexity of the regulation of oxidative phosphorylation in contracting skeletal muscle. In a recent report published in The Journal of Physiology, Jones and colleagues (Jones et al. 2008) developed an approach to study [PCr] dynamics during the early phase of moderate and heavy exercise in humans in vivo. The measured [PCr] was taken to represent the net balance between the rate of ATP usage and the rate of ATP synthesis. The authors studied the influence of baseline metabolic rate on muscle [PCr] kinetics as assessed by using 31P-MR spectra acquired every 6 s from the exercising quadriceps muscle. The subjects completed a series of 6 min steady-state exercise bouts at an intensity corresponding to 40% (moderate intensity) and 80% (heavy intensity) of the highest work-rate attained in an initial incremental exercise (28 ± 4 W). Intensities were chosen to involve dominantly oxidative fibres (40%) then an increasing amount of glycolytic fibres (80%). [PCr] kinetics was modelled by exponential functions characteristic of three transitions from: (i) rest-to-moderate (40%); (ii) rest-to-heavy (80%) and (iii) moderate-to-heavy (40%-to-80%) intensity exercise. The influence of baseline metabolic rate on [PCr] kinetics was quantified by the time constant τ describing the initial ‘fundamental’ decline in muscle [PCr] and by the functional ‘Gain’ of the [PCr] response, i.e. the change in [PCr] per unit change in work-rate. Jones et al. obtained different values for τ over the range from 25 s to about 100 s and different values for Gain from 0.56 mm W−1 to about 1.5 mm W−1 during the three studied transitions. Clearly, the time constant τ progressively increased from rest-to-moderate exercise (25 s), rest-to-heavy exercise (±50 s) and moderate-to-heavy exercise (±100 s), thus indicating a progressive slowing of the [PCr] response. As well, the functional Gain of the [PCr] response increased from 0.6 mm W−1, to about 1 mm W−1 and 1.5 mm W−1 as a function of the exercise intensity. Jones et al. argued that if the energetic system they defined – wherein [PCr] is representative of the intermediates implicated in the metabolic feedback control of oxidative phosphorylation – expresses dynamic linearity, no effect on the response τ or Gain would be expected as a result of manipulations of the baseline work-rate. Based on the variations in both τ and Gain clearly indicated by their experimental work, they concluded that the oxidative phosphorylation is not exclusively under linear first-order control by [PCr]. This conclusion is in good agreement with the concept of direct parallel activation of energy supply and demand in intact contracting muscle studied in vivo (Korzeniewski, 2007), wherein control mediators other than that represented by [PCr] are possibly operative, and can support different strengths of activation of PCr supply and PCr demand. Based on variations in τ and Gain, different strengths of parallel activation should be suspected in the comparison of low-to-moderate versus moderate-to-high transitions. Jones and colleagues (Jones et al. 2008) attribute the slowing in [PCr] response to the shift in motor unit recruitment. Probably, during low-to-moderate intensity exercise, such as that performed at 40% of peak work-rate in their study, the majority of the required muscle power is produced by the type I muscle fibres, whereas during higher intensity exercise, type II fibres significantly contribute to power production. Interestingly, the strength of the parallel activation in intact skeletal muscle might be fibre-type dependent (Korzeniewski, 2007), as suggested by [PCr] kinetics obtained by Kushmerick and colleagues (Kushmerick et al. 1992) in ex vivo arterially perfused cat biceps brachii (type II) and soleus (type I) muscles interrogated by 31P-RM spectroscopy. They showed that there is an overshoot of [PCr] in the isolated cat soleus (type I) during the early phase of recovery following exercise, probably indicating that PCr supply is not only under feedback activation, but rather that other effector molecules/signals directly activate oxidative phosphorylation. This direct activation is not turned off instantly after the termination of sustained contractions, but decays (possibly in an exponential way (Korzeniewski, 2007)) as a result of, e.g. protein dephosphorylation or calcium exchanges. It is of note that the overshoot was not observed in biceps brachii, possibly indicating a much lower (even absence of) parallel activation in this fast-twitch muscle. Kushmerick and colleagues (Kushmerick et al. 1992) also obtained nice results under steady-state conditions of sustained contractions, that provide a similar suggestion: when the steady metabolic rate was experimentally increased there was a deeper decrease in the steady PCr level in the biceps brachii than in the soleus, suggesting a more balanced (so-called higher) parallel activation in oxidative muscle fibres. Based partly on these considerations, B. Korzeniewski embodied many enzyme kinetics into a dynamic computer model of oxidative phosphorylation and ATP usage in skeletal muscle that integrates a broad range of experimental data (Korzeniewski, 2007). He proposed recently that only the (more or less) parallel activation of PCr (ATP) supply and demand could explain the whole set of experimental results dealing with muscle energetics (flux and intermediates dynamics) in the literature. With this approach, oxidative fibres are expected to be under well-balanced parallel activation, whereas in glycolytic fibres, supply is less activated than demand. Thus, the change in [PCr] dynamics found by Jones and colleagues in vivo (Jones et al. 2008) might result from the progressively lower activation of PCr supply when compared to PCr demand when type II fibres contribute significantly to power production. Both the slower [PCr] response and the higher Gain are well explained by this hypothesis. It is possible to quantify the regulation by external effectors of muscle energy metabolism, defined as a two-module supply–demand system where modules are connected by common intermediates represented by [PCr], using tools of metabolic control analysis (Brand, 1997). This approach differentiates the regulation of supply and demand by external effectors (regulation analysis) with the sensitivity of supply and demand modules towards energetic intermediates (control analysis). Thus, one should determine the sensitivity (i.e. local kinetic response quantified by the elasticity coefficient) of each module in the absence of changes in external effectors. This is generally achieved in experimental conditions where the system is at steady state: it means [intermediate] and the flux through the system (here contraction as representative of ATP turnover rate) are effectively constant. Elasticities are obtained around that steady state. The selective modulation of one module induced [PCr] changes and the response of the untouched module is measured (the new steady flux through the system). Figure 1 indicates typical simultaneous changes in [PCr] (the intermediate) and contraction (the flux) that can be obtained in electrically stimulated gastrocnemius muscles of anaesthetized rats in vivo. Energy-supply inhibition by hypoxia inspired fraction of oxygen (FIO2= 9%) made it possible to quantify demand sensitivity by calculating the ‘elasticity coefficient’, i.e. the ratio of change in flux to change in [PCr] (Diolez et al. 2007). Assuming that a small decrease in electrical stimulation selectively affects energy demand by myofibrilar ATPase, the supply elasticity can be quantified as well. Elasticity coefficients are correctly determined only if changes in control mediators other than that represented by [PCr] during the selective inhibition of one module are either absent or have negligible additional effects on the other module. As in vivo it cannot be totally excluded that external effector molecules/factors slightly change during the modulation of intact muscle by electrical stimulation (e.g. variations in Ca2+ concentrations/pulses slightly affecting mitochondria), it can be concluded that such an approach is actually semi-quantitative; it would yet indubitably sharpen our understanding of muscle physiology. Successive 31P MR spectra (acquired every 2 min) and levels of contraction recorded during the modulation around the studied steady state (reference) In this example, respective differences in [PCr] (intermediate) and contraction (flux) experimentally induced by lowering FIO2 to 9% (low supply) and lowering electrical stimulation (low demand) made it possible to calculate elasticity coefficients. In conclusion, the study by Jones and colleagues (Jones et al. 2008) adds to our understanding of in vivo regulation of muscle energy turnover by external effectors and energetic intermediates as represented by [PCr], as did other investigations comparing pulmonary oxygen uptake and muscle [PCr] kinetics in response to step changes in muscle power output in humans in vivo (Rossiter et al. 2002). There are experimental arguments pointing to high parallel activation in oxidative muscle but low parallel activation in glycolytic muscle (Kushmerick et al. 1992). Therefore, in the study by Jones and colleagues, for a given increase in energy demand during exercise transitions, a more or less high direct simultaneous activation by external effectors of energy supply would significantly affect [PCr] kinetics. Keeping in mind the low parallel activation in type II fibres, then it is not surprising that slower kinetics (greater τ) and greater Gain of the [PCr] response (it means marked PCr depletion) were observed during transitions where the contribution of glycolytic fibres to power production is significant. The experiment by Jones and colleagues, inspired by previous experimental development by Rossiter and colleagues (e.g. Rossiter et al. 2002) may provide a nice additional contribution to the concept of fibres-dependent parallel activation in skeletal muscle (Korzeniewski, 2007). Probably, the definition of the modular supply demand system for identifying intermediates and effectors responsible for modules activation (Brand, 1997) will be helpful for further investigations of muscle energetics.