7. Conclusion
Mitochondrial dysfunction is associated with skeletal muscle insulin resistance and studies on human subjects with congenital insulin signalling defects demonstrate unequivocally that mitochondrial defects can result from ab initio insulin resistance. Evidence for a possible causal role of mitochondrial dysfunction in development of insulin resistance is less direct. Proposed mechanisms by which imbalanced bioenergetics can produce deleterious lipids and ROS are intuitively attractive, but support for the conceptually similar models remains circumstantial. The reported dampening effects of DAG, ceramide and hydrogen peroxide on insulin signalling are compelling, but links between mitochondrial oxidative capacity, lipid intermediates and insulin resistance are far from universal, and it is also not conclusively clear if harmful ROS indeed originate in mitochondria. Improved measurement of (i) lipid composition and subcellular location, (ii) real-time oxidative phosphorylation activity, (iii) native rates by which different ROS sources generate superoxide, hydrogen peroxide and, indirectly, hydroxyl radicals, and indeed of (iv) insulin sensitivity itself, which may deteriorate without changes in canonical signalling pathways, should enlighten the causality debate. As it stands, it seems plausible that changes in mitochondrial function that occur relatively late during development of muscle insulin resistance will exacerbate pathology. It is equally conceivable that mitochondrial insufficiencies or deficiencies coincide with harmful effects of nutrients and cytokines on other cellular targets, and that the onset of insulin resistance is triggered by multifarious independent defects. A challenge of this causal scenario would be to unravel the possible interplay between the initial effects, and to establish their relative importance. If mitochondria are indeed not involved during early disease pathology, then it remains to be demonstrated how insulin resistance causes the reported functional mitochondrial defects. Addressing this issue, mitochondrial function will be best evaluated in context of cellular bioenergetic control, as loss of insulin sensitivity likely remodels ATP-consuming processes, and mitochondria will indeed adapt to such remodeling of energy demand. It may thus be fairly obvious to conclude that mitochondrial involvement in obesity-related insulin resistance of skeletal muscle is ‘a case of imbalanced bioenergetics’, but it is currently far less trivial to judge which side of the balance is tipped to upset the peace.