A breath of fresh air for mitochondria in exercise physiology

Mitochondrial affinity for oxygen was recently demonstrated to increase after sprint training and turned out to be closely associated with training-related improvement in pulmonary peak oxygen uptake (VO2peak) during both arm cranking and leg cycling.1 Although logically the final step of the oxygen cascade should be modulated by mitochondrial oxygen affinity with notable impact on VO2peak, experimental evidence for such modulation is scarce. The report of Larsen and colleagues considerably strengthens this conception but also reveals the necessity of integrative considerations and current experimental limitations. The overall research work of these authors is largely contributing to the contemporary understanding of regulatory mitochondrial function in the oxygen transport system. Their studies implicate mitochondrial oxygen affinity, among others, in the control of the basal metabolic rate (BMR) and locomotion efficiency,2 muscle mass-related oxygen extraction3 and training-induced improvements in aerobic capacity.1 Mitochondrial oxygen affinity can be assessed from different fresh mitochondrial preparations, including permeabilized muscle fibres or isolated mitochondria. It is determined through assessment of the oxygen pressure at the mitochondrial level (pO2mito) at which mitochondrial respiration amounts to 50% of the maximal rate of mitochondrial respiration at saturating oxygen concentrations (p50mito). Lowering of the p50mito corresponds to an equivalent increase in the mitochondrial oxygen affinity and vice versa. Owing to the fact that muscle oxygen consumption (by mitochondria) must be the main determinant of (pulmonary) VO2peak, the relationship between VO2peak and the p50mito becomes apparent from the equation: VO2mito = (Vmax × pO2mito)/(pO2mito + p50mito); (Vmax is the oxygen saturated maximal mitochondrial respiration and VO2mito is mitochondrial respiration depending on oxygen availability).1 It may be helpful to briefly consider here the oxygen transport system. The demand of oxygen by the working muscles, for example, leg muscles when cycling, increases tremendously (about 10- to 20-fold, depending on the individual fitness) from rest to maximal work. Consequently, a large amount of oxygenated blood has to be transported to muscles (oxygen delivery, DO2) where oxygen is used by mitochondria for energy production by means of oxidative phosphorylation. A precise interplay between pulmonary ventilation, DO2 (cardiac output x oxygen content of the arterial blood), and the extraction of oxygen by the muscle tissue is a prerequisite for properly matching oxygen need and demand. The peripheral oxygen extraction is primarily driven by the oxygen partial pressure (pO2) gradient and the diffusion distance. The p50mito may optimize the pO2 gradient and thus increase VO2peak. According to the concept of symmorphosis, all steps of the oxygen cascade should be designed proportionally to VO2peak.4 This, however, does not seem to apply well to humans, as several studies suggest a prominent excess capacity of the muscle mitochondrial respiratory rate over DO2 by the blood circulation. Some authors argue that the mitochondrial respiratory capacity exceeds VO2peak to a larger extent when half or more of the body muscle mass (eg leg cycling) is involved during exercise as compared to involvement of small muscle groups (eg knee extension exercise or arm cranking).4 Consequently, the individual steps of the oxygen cascade will contribute differently to VO2peak depending on the muscle mass used,3 which also holds true for the p50mito as displayed by the findings of Larsen and colleagues.1 The short-term sprint training (arm cranking and leg cycling) applied by the authors is a well-chosen training modality to stimulate structural and functional mitochondrial adaptations. After seven sprint training sessions (arm cranking and leg cycling) the p50mito decreased in the m. triceps brachii by 44% and by 35% in the m. vastus lateralis but VO2peak increased on average by 12% during arm cranking and by 6% during leg cycling. The variation of the p50mito changes reported by the authors explained about 40% of the accompanied VO2peak changes indicating the contribution of other alterations along the oxygen cascade, which would be expected according to a coordinated adaptation of all contributing steps.5 At this point, methodological limitations in the assessment of mitochondrial oxidative capacity need to be addressed. Conclusions related to VO2peak may, for example, differ depending on how mitochondrial capacities are derived, for example, ex-vivo preparations like isolated mitochondria or permeabilized tissues are expected to yield different outcomes and ex-vivo respirometry approaches often involve saturating concentrations of substrates for the oxidative phosphorylation machinery, potentially masking insufficient supplies in-vivo. Ex-vivo procedures per se may potentially contribute to overestimation of the mitochondrial oxidative capacity. Confirmation for this assumption comes from a recent experiment where mitochondrial oxidative capacity was predicted using quantitative enzyme histochemistry from skeletal muscle succinate dehydrogenase activity, demonstrating substantially lower overcapacity as previously reported from isolated mitochondria or permeabilized muscle cells.6 Without weakening the relevance of the training-related changes in mitochondrial oxygen affinity reported by Larsen and colleagues,1 these methodological considerations display the difficulty in the interpretation of such effects in the context of VO2peak limitations on the level of the organism. During maximal exercise in real physiological conditions many factors may have a modulatory impact on VO2peak, especially important when compared from pre- to post-training, for example, vasoconstriction in working muscles evoked by severe pulmonary hyperventilation, changes in the oxygen affinity of haemoglobin, myoglobin function, shift in muscle fibre types, etc. Aside from potentially limiting substrates or changes in the ion balance, other factors related to mitochondrial physiology may play significant roles in mitochondrial oxygen extraction in-vivo, for example, supercomplex formation, or mitochondrial membrane potential. Another notable observation by Larsen and colleagues2 demonstrates a close positive association between the mass-related BMR and mitochondrial oxygen affinity derived from isolated mitochondria in humans. Individual differences in mitochondrial COX activity and excess relative to other mitochondrial proteins were suggested as the underlying mechanism. Furthermore, Larsen et al2 observed a similar relationship between mitochondrial oxygen affinity and (pulmonary) VO2 at submaximal cycling as shown for BMR, indicating higher mechanical efficiency (less VO2 for a given workload) at lower mitochondrial oxygen affinity. The authors found confirmation for that in the observation of increased efficiency after supplementation with inorganic nitrate, which inhibits COX activity after conversion into nitric oxide (NO) and increases p50mito. Based on that and the recently demonstrated association between elevation in VO2peak and reduction in p50mito, the authors point to an intriguing physiological explanation for the incompatibility between high mechanical efficiency and high aerobic power.1, 3 An inverse relationship between mechanical efficiency and VO2peak was actually shown in elite cyclers, which was attributed to variations in the percentages of efficient type I and less efficient type II fibres of the main working muscles.7 Another affirmation for the inverse relation between VO2peak and locomotion efficiency, as well as for the involvement of mitochondrial coupling efficiency in this relation, comes from a recent training study based on pre-post evaluation of exercise capacity and mitochondrial function (permeabilized muscle fibres).8 The hypothesis of mitochondrial involvement in mechanical efficiency and aerobic power certainly merits future detailed investigation. It would be expected that it is related to several other aspects of mitochondrial physiology as well. There is, for example, an intriguingly reminiscent molecular mechanism related to mitochondrial dynamics9; conditions of metabolic oversupply trigger mitochondrial fission (related to low mitochondrial oxygen affinity and decreased bioenergetic as well as mechanical efficiency?), while undersupply induces fusion (associated with higher mitochondrial oxygen affinity, bioenergetic and mechanical efficiency?). Better understanding of such potential interrelations will have implications for molecular to organismal energy homeostasis and economy, performance, fatigue, etc, in both physiology and pathology. Particularly attractive from an exercise physiology perspective, sprint-training effectuated increased oxygen affinity will result in increased BMR and VO2peak but unchanged or even worse locomotion efficiency. Incidentally, in contrast to endurance exercise, resistance training may represent an appropriate measure to improve mechanical efficiency because increased strength and muscle size would reduce the dependence on more inefficient type II fibres. In general, the work of Larsen and colleagues highlights the importance of the p50mito as a valuable supplementary tool to ex-vivo respirometry. The p50mito yields additional information on mitochondrial oxygen affinity in an experimental paradigm commonly working with saturating oxygen levels. It thus allows to take into account potential effects of oxygen limitations, not only relevant in exercise physiology but also in various diseases involving mitochondrial dysfunction. I am very grateful to Dr. Johannes Burtscher (Université de Lausanne, Switzerland) for critical proof reading and helpful commenting on this Editorial and to Prof. Erich Gnaiger (Medical University Innsbruck, Austria) for fruitful discussions on this topic. No conflict of interest.