Skeletal muscle AMPK activation: mounting evidence against a role in substrate utilization during acute exercise

Since the seminal findings that pharmacological activation of AMP-activated protein kinase (AMPK) increases skeletal muscle glucose uptake and fatty acid oxidation (FAO), AMPK has gained considerable interest as a regulator of fuel utilization during exercise. This notion arises as the intensity-dependent increase in cellular metabolites (AMPfree, ADP) from ATP hydrolysis and the accumulation of intracellular calcium allosterically regulate AMPK to promote its phosphorylation by upstream kinases (LKB1, CaMKK). Pharmacological activation of AMPK has been shown to induce a myriad of cellular events, including glucose transporter type 4 (GLUT4) translocation to the plasma membrane to facilitate glucose uptake, as well as a co-ordination of several mechanisms up-regulating FAO, including fatty acid transport protein accumulation on the plasma membrane and reduced malonyl-CoA (MCoA) production from acetyl-CoA carboxylase (ACC). However, unlike direct pharmacological activation of AMPK, exercise is a robust physiological stimulus that simultaneously up-regulates numerous cellular processes. As a result, the necessity of AMPK for substrate utilization during exercise has been challenged, particularly as FAO and glucose uptake during exercise or contraction are not altered in mice with muscle-specific AMPKα deletion (Hingst et al. 2020). One way to examine this relationship in humans is by comparing individuals of differing training status, and therefore differing levels of AMPfree during acute exercise. This approach has delineated that a ∼10-fold increase in AMPK activity during exercise at 65% in untrained individuals is completely abolished after 10 days of exercise training (McConell et al. 2005) despite substantial glucose uptake and higher fat oxidation. However, this acute exercise bout was performed at the same absolute intensity, representing a lower relative intensity post-training. To date, the extent of AMPK activation at the same relative intensities between trained and untrained individuals following a prolonged bout of moderate intensity exercise remains unknown. To address these gaps using a cross-sectional study design, a recent paper in The Journal of Physiology by McConell et al. (2020) examined skeletal muscle AMPK activation at 30 and 120 min of a single exercise bout performed at 65% in seven endurance-trained (ET) and 11 untrained (UT) males. As hypothesized, McConell et al. (2020) reported that 120 min of acute exercise increased skeletal muscle AMPKα activity, AMPKα Thr172 phosphorylation,and downstream ACCβ Ser221 phosphorylation in UT individuals, although these metabolic events did not occur at either time-point in ET individuals. While markers of energetic stress (muscle lactate, Cr, AMPfree) were higher in UT compared to ET individuals throughout exercise, respiratory exchange ratios (RER) and net glycogen utilization were similar between groups (despite higher glycogen content in ET individuals). Given the absence of AMPK activation in ET individuals, these data would suggest that AMPK is not required for substrate utilization during exercise. The findings by McConell et al. (2020) have important implications for understanding the role of AMPK during acute exercise, the influence of training on skeletal muscle metabolism during exercise and the potential implications of AMPK activation for post-exercise adaptations. ACC is thought to regulate fatty acid oxidation in skeletal muscle by catalysing the formation of MCoA, a potent inhibitor of carnitine palmitoyl-transferase-1 (CPT-1) and therefore mitochondrial fatty acid transport. AMPK has been implicated in the inactivation of ACC by serine phosphorylation, subsequently reducing MCoA production and removing the inhibition on CPT-1 to facilitate FAO. If AMPK had a dominant effect in the regulation of FAO during exercise, a direct relationship between AMPK activity, ACCβ phosphorylation and FAO would be expected. However, despite a lack of AMPK activation and ACCβ Ser221 phosphorylation in ET individuals during exercise (McConell et al. 2020), RER, a surrogate marker of whole-body fat/carbohydrate utilization, remained comparable between groups. A further dissociation between AMPK activity and FAO is evident as training-induced increases in FAO during exercise performed at the same absolute intensity occur despite attenuated skeletal muscle AMPK activation (McConell et al. 2005). Furthermore, while ACC phosphorylation is blunted in mice with muscle-specific AMPKα deletion, skeletal muscle palmitate oxidation is not decreased during contraction compared to wild-type controls (Hingst et al. 2020). In support of this evidence against a role of the AMPK-ACC-MCoA axis in regulating FAO during exercise, it is well accepted that MCoA content is not reduced during exercise in human skeletal muscle, despite an increase in FAO (Petrick & Holloway, 2019), suggesting mechanisms independent of MCoA content regulate FAO. In this respect, the sensitivity of CPT-1 to MCoA inhibition is acutely decreased during exercise, and the exercise-mediated increase in palmitoyl-CoA availability can further override the inhibitory effects of MCoA (Petrick & Holloway, 2019). Taken together, this would suggest that regulation of CPT-1 enzymatic properties has a more prominent role in increasing FAO during exercise than AMPK activation. These dissociations between AMPK and substrate utilization during exercise are also evident with respect to glucose/glycogen metabolism as net glycogen utilization was comparable between ET and UT individuals despite differences in AMPK activation (McConell et al. 2020). In addition, other findings indicate that manipulating muscle glycogen content (low-carbohydrate diet) does not regulate patterns of AMPK activation during exercise (McConell et al. 2005) and glucose uptake is not reduced in mice with AMPKα deletion (Hingst et al. 2020). As a classical response associated with exercise training, McConell et al. (2020) report a reduction in metabolites indicative of energetic stress in ET individuals. Specifically, skeletal muscle AMPfree increased 17-fold in UT and only 5-fold in ET individuals at 120 min of exercise (McConell et al. 2020). It is interesting that, despite some increase in AMPfree during exercise in ET individuals, there was no influence on AMPK activity. This may suggest a threshold of AMPfree accumulation is required to activate AMPK. While studies between individuals of differing training status that are matched to relative intensity represent a more equivalent metabolic demand than those performed at an identical absolute intensity, the caveat of differing levels of AMPfree between groups remains. Although previous work by McConell et al. (2005) indicates the ratio between AMPK activity/AMPfree is comparable following training, it is not known whether AMPK would be activated in trained individuals when AMPfree reached the same levels observed in untrained individuals. This could be examined by performing exercise at various incremental intensities in trained and untrained individuals, particularly to fatigue when metabolic stress is maximal regardless of training status. As noted by McConell et al. (2020), this could act as an AMPfree ‘titration’ to determine whether the sensitivity of AMPK to AMPfree is altered following training. In addition to AMPfree, several other cellular signals (Ca2+, ROS, LKB1 and CaMKII) have been shown to influence AMPK activation and phosphorylation, and it is possible that training-induced changes in these pathways could contribute to the attenuated AMPK activation following training. Although evidence is accumulating to suggest AMPK may not have a dominant role in regulating substrate selection during exercise, AMPK is nonetheless an important signalling event required for cellular adaptations following exercise. AMPK is known to increase transcription of numerous genes, ultimately increasing mitochondrial content and up-regulating various proteins mediating glucose and fat utilization. Similar to AMPK activation, transcriptional responses to exercise are also influenced by training status. For example, in a time-course analysis, peroxisome proliferator-activated receptor gamma coactivator 1-alpha (PGC1α) mRNA was increased ∼10-fold after a single exercise bout; however, the magnitude was attenuated with successive bouts (Perry et al. 2010). These trends are also evident with respect to protein markers of mitochondrial content (citrate synthase, PGC1α protein), suggesting a pattern of diminishing returns following exercise even when bouts were performed at the same relative intensity (increasing absolute intensity) (Perry et al. 2010). As AMPK is a potent activator of mRNA transcription, the abolished acute AMPK activation in trained individuals (McConell et al. 2020) could be a response mediating the blunted mitochondrial biogenic responses to a single bout of exercise. This attenuation can clearly occur over a short time period, with respect to both AMPK activation (10 days of training; McConell et al. 2005) and mRNA or protein responses (seven sessions of training; Perry et al. 2010). In summary, the study by McConell et al. (2020) has established that AMPK activation does not occur in trained individuals during prolonged moderate exercise matched for the same relative intensity; findings which contribute to the mounting evidence challenging the necessity of AMPK in regulating substrate utilization during exercise. In contrast, given the progressively blunted transcriptional responses to successive bouts of exercise (Perry et al. 2010), these findings support a role of AMPK in mediating post-exercise adaptations. It remains unknown whether AMPK activation could occur to a similar extent when AMPfree levels are identical between trained and untrained individuals, and whether AMP-independent cellular events influence the varying degrees of AMPK activation observed by McConell et al. (2020). While future research in these areas could contribute to current understanding of events mediating the attenuated AMPK activation in trained individuals; the results reported by McConell et al. (2020) provide evidence that, regardless of training status, AMPK is not a key regulator of substrate utilization during exercise. No competing interests declared. SMF, GJD and HLP were responsible for the conception or design of the work. SMF, GJD and HLP were responsible for drafting the work or revising it critically for important intellectual content. SMF, GJD and HLP were responsible for final approval of the version submitted for publication. SMF, GJD and HLP agree to be accountable for all aspects of the work. SMF (Canadian Graduate Scholarship – Master's) and HLP (Canadian Graduate Scholarship – Doctoral) hold scholarships from the Natural Sciences and Engineering Research Council of Canada. We thank Dr Graham Holloway for editing the version of the manuscript submitted for publication.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.