More than mitochondrial biogenesis: alternative roles of PGC-1α in exercise adaptation

Endurance exercise training is characterised by sustained contractile activity against a low opposing resistance or repeated high-intensity contractions interspersed with brief recovery intervals. A hallmark adaptation to endurance exercise training is skeletal muscle mitochondrial biogenesis, resulting in an enhanced respiratory capacity and subsequent improvements in metabolic control (Holloszy, 1967). Mitochondrial biogenesis requires the coexpression of both nuclear and mitochondrial genomes, and the requisite signal transduction pathways involved have focused heavily on the transcriptional coactivator and posited ‘master regulator’ of mitochondrial biogenesis, the peroxisome proliferator-activated receptor (PPAR)-γ coactivator-1α (PGC-1α). PGC-1α docks on and activates several transcription factors to increase the expression of proteins ultimately incorporated into existing and/or new mitochondria. PGC-1α mRNA abundance increases rapidly after a single bout of endurance exercise, although such changes precede an increase in both total transcriptional protein abundance necessary for mitochondrial biogenesis and functional mitochondria (Baar et al. 2002). Indeed, it takes up to 72 h for an exercise-induced increase in skeletal muscle mitochondrial function (i.e. tighter respiratory control as manifested by a closer matching of ATP production with ATP hydrolysis). In contrast, the time course and cellular regulation of other, more immediate adaptation responses to endurance exercise, such as those regulating skeletal muscle glycogen depletion and lactate accumulation, are not known.

A recent article in The Journal of Physiology by Kim and colleagues (2014) provides novel insight to the cellular mechanisms mediating changes in glycogenolytic and glycolytic capacity following endurance exercise training. In their study, Wistar rats performed either two separate 3 h swimming sessions interspersed with 45 min recovery, or the same exercise session repeated for three consecutive days. To determine the effect of prior exercise on glycogen depletion and lactate accumulation, animals underwent in situ electrical stimulation of the triceps muscle at baseline and 16–20 h following both a single exercise bout and 3 days of training. The contracting triceps muscle was then clamp frozen for subsequent biochemical analysis. The authors demonstrated that skeletal muscle glycogen breakdown and lactate accumulation were inversely proportional to the volume of training, and that these changes paralleled reductions in the total content of several glycogenolytic (glycogen phosphorylase, phosphorylase kinase) and glycolytic (PFK, GAPDH, LDH) enzymes that were downregulated with consecutive training days. Despite an increase in PGC-1α protein 18 h following acute exercise, no differences in mitochondrial function (rates of pyruvate oxidation) were observed 3 days after training. To delineate a potential role for PGC-1α in these ‘glycogen sparing’ adaptive responses, the researchers utilised electroporation to transfect PGC-1α DNA into the animal's triceps muscle and also cultured C₂C₁₂ myoblasts in the presence of a PGC-1α shRNA plasmid to overexpress and disrupt the PGC-1α gene. Overexpression of PGC-1α resulted in ∼50% reductions in glycogen phosphorylase, phosphorylase kinase, and PFK protein levels, whereas PGC-1α knockdown in culture increased the expression of these proteins. Collectively, these findings demonstrate that the rapid reductions in proteins of the glycogenolytic/glycolytic pathway are mediated by PGC-1α within 24 h of a single or consecutive endurance exercise bouts.

The mechanistic basis underlying the early exercise-induced events activated to preserve metabolic control are not clear, but the authors hypothesise that PGC-1α may indirectly upregulate the expression of transcriptional repressors of genes encoding for glycogenolytic and glycolytic enzymes. Indeed, recent insight from Summermatter et al. (2013) showed that muscle-specific PGC-1α transgenic mice produce significantly less blood lactate compared to wild-type littermates following fatiguing endurance exercise. In that study, PGC-1α was also shown to regulate lactate dehydrogenase (LDH) isozyme composition by modulating the activity of several LDH gene transcription factors, which collectively increased the LDH B/A isozyme ratio, in turn favouring pyruvate accumulation over lactate (Summermatter et al. 2013). Previous experiments also show that PGC-1α regulates the expression of PDK4, an enzyme responsible for restraining glucose oxidation, along with genes involved in post-exercise glucose uptake (GLUT1 and GLUT4 transporters) and glycogen storage (hexokinase) (Wende et al. 2007). The results from these studies suggest that in addition to its established role in mitochondrial biogenesis, PGC-1α can reduce the initial, fatigue-inducing effects of unaccustomed endurance exercise via regulation of the glycogenolytic enzymatic machinery to attenuate glycolytic flux, preserve muscle glycogen, and attenuate lactate accumulation. As these changes arise independently of improvements in oxidative phosphorylation (which, independently, would also elicit a glycogen sparing effect), it is possible this metabolic adaptation comprises a PGC-1α-mediated ‘stress response’ rapidly activated to preserve cellular homeostasis in the absence of newly synthesised mitochondrial protein (Fig.1).

Figure 1.

Schematic representation of PGC-1α-mediated stress responses to unfamiliar and chronic contractile activity

Preceding detectable increases in mitochondrial biogenesis, PGC-1α enhances glycogen storage via downregulation of glycogenolytic flux, lactate accumulation and glucose oxidation. In addition, unfamiliar contractile stress triggers protein misfolding in the endoplasmic reticulum (ER), which, as a result of unfolded protein response (UPR) signal transduction, may restrain maximal mitochondrial protein synthesis rates. With repeated training, the UPR declines, providing necessary ATP for mitochondrial protein synthesis. Glu, glucose; GLUT1/4, glucose transporters; HK, hexokinase; G6P, glucose 6-phosphate; PFK, phosphofructokinase; PhK, phosphorylase kinase; GPh, glycogen phosphorylase; LDH A/B, lactate dehydrogenase isozymes; PDK4, pyruvate dehydrogenase kinase 4; ATF6α, activating transcription factor 6α.

As several mitochondrial proteins have long (∼7 days) half-lives, and as already noted, ∼72 h is required before any meaningful exercise-invoked changes in mitochondrial function take place, the authors attribute the rapid PGC-1α stress response to a means of promoting survival when one is faced with an immediate threat (such as escaping predators or a natural disaster) and is thereby forced to undertake unfamiliar contractile activity (Kim et al. 2014). In support of this thesis, a single running bout was sufficient to induce endoplasmic reticulum (ER) stress in untrained skeletal muscle from mice and activate a compensatory adaptive unfolded protein response (UPR), the magnitude of which then subsequently declined with repeated training (Wu et al. 2011). UPR signalling ameliorates the aggregation of misfolded proteins and possible calcium disturbances that promote ER stress following strenuous exercise. Wu and colleagues showed that PGC-1α coactivated a putative UPR signal transducer, ATF6α, to increase the expression of various chaperones necessary for ER luminal protein folding (Wu et al. 2011). However, the UPR concomitantly triggers a temporary global translational downregulation that preferentially enables translation of UPR-related proteins to accommodate restoration of ER homeostasis. Therefore, while the cell attempts to alleviate ER disturbances following contractile stress, it is conceivable that the incorporation of nascent protein into the mitochondria is repressed until a new, augmented steady-state homeostasis is reached. Support for this notion also stems from the fact that similar to protein synthesis, ER protein folding reactions consume ATP; therefore this potentially prioritises available ATP away from mitochondrial protein synthesis to sustain the peak protein folding demands of the UPR in the immediate recovery period following exercise in the untrained state. The conservation of energy substrate (principally glycogen) as directed by PGC-1α may provide a readily available ATP source to facilitate the UPR, although this hypothesis needs to be verified.

In light of animal data showing that PGC-1α mediates multiple adaptation processes with exercise contraction, a question remains: how we can translate these adaptation responses to the wealth of human exercise studies, particularly acute, that report post-exercise increases in PGC-1α mRNA and protein? Before ascribing mitochondrial biogenesis to PGC-1α and its coregulators, a number of factors related to subject training status, type of training habitually performed, the type of exercise intervention under investigation, and nutrient availability (or lack thereof) need to be considered. For example, in well-trained endurance athletes there is a lower ceiling for increasing total mitochondrial mass, thus a given PGC-1α signal in this population may relate to a role in post-exercise substrate partitioning as a result of exogenous carbohydrate ingestion. Alternatively, in the case of a middle-distance athlete, a greater intracellular reliance upon PGC-1α modulation of lactate homeostasis could predominate to counter untoward fatigue in these types of events. Thus, a combination of factors probably dictates the PGC-1α-mediated biochemical response.

In conclusion, Kim and colleagues provide valuable insight to the multiplicity and complexity of the functions of PGC-1α in mammalian metabolism following endurance exercise. It seems prudent to ask how their experimental model might be translated to whole-body integrative physiology, particularly for untrained individuals, given the ‘unphysiological’ exercise stimulus provided in their study. As noted previously, it appears that perhaps the only obligatory response to exercise is the defence of cellular and whole-body homeostasis, and in this regard, PGC-1α is highly versatile, interacting with multiple transcription factors to activate distinct biological programmes that maintain cellular integrity.

Additional information

Competing interests

None declared.