Mitophagy in maintaining skeletal muscle mitochondrial proteostasis and metabolic health with ageing

Abstract

Skeletal muscle is important for overall functionality and health. Ageing is associated with an accumulation of damage to mitochondrial DNA and proteins. In particular, damage to mitochondrial proteins in skeletal muscle, which is a loss of mitochondrial proteostasis, contributes to tissue dysfunction and negatively impacts systemic health. Therefore, understanding the mechanisms underlying the regulation of mitochondrial proteostasis and how those mechanisms change with age is important for the development of interventions to promote healthy ageing. Herein, we examine how impairment in the selective degradation of damaged/dysfunctional mitochondria through mitophagy may play a central role in the loss of mitochondrial proteostasis in skeletal muscle ageing, as well as its broader implications for systemic health. Further, we explore how stimulating mitophagy through exercise may promote healthy ageing.

Abbreviations

AMPK

5ʹAMP‐dependent kinase

ATG's

autophagy‐related proteins

BCL2

B‐cell lymphoma 2

BNIP3

BH3‐only family protein

LAMP2

lysosome‐associated membrane protein‐2

LC3

microtubule‐associated protein 1A/B‐light chain 3

mtPTP

mitochondrial permeability transition pore

PINK1

PTEN‐induced putative kinase 1

ULK1

unc‐51 like autophagy activating kinase 1

Introduction

Skeletal muscle is a highly dynamic organ that is responsible for generating movement required for the activities of daily living, peak physical performance, and metabolic health. Skeletal muscle makes up 40% of the human body and is one of the most metabolically active tissues (Janssen et al. 2000); thus declines in skeletal muscle function have dire consequences. Indeed, progressive decline in skeletal muscle function with age is associated with metabolic dysfunction, chronic disease susceptibility, loss of mobility, and an overall increase in mortality rate (Rantanen et al. 2000; Metter et al. 2002; Nair, 2005; Cruz‐Jentoft et al. 2010). Loss of mitochondrial protein homeostasis (i.e. proteostasis) has been suggested to play a causative role in the age‐related decline in skeletal muscle function (Marcinek et al. 2005; Chabi et al. 2008; Joseph et al. 2012; Kruse et al. 2016). Since mitochondria generate the majority of ATP used to fuel skeletal muscle contractile and metabolic activities, understanding how mitochondrial proteostasis is altered in ageing could lead to the identification of targets for effective interventions to slow ageing and improve age‐related health outcomes.

Pointing fingers: mitochondrial biogenesis or mitophagy?

One hypothesis for the progressive decline in organ function with age is an accumulation in damaged, dysfunctional mitochondria (Stadtman, 1992; Kowald, 1996; Kowald & Kirkwood, 2000; Baraibar et al. 2013). Skeletal muscle mitochondria from both old animals and humans show evidence of an accumulation of protein damage (Lass et al. 1998; Choksi et al. 2008; Wohlgemuth et al. 2010; Beltran et al. 2015) and release of reactive oxygen species (Chabi et al. 2008; Picard et al. 2011), which further contribute to mitochondrial damage and skeletal muscle dysfunction (Picard et al. 2011; Gouspillou et al. 2014). The accumulation of damaged mitochondrial proteins suggests that some facet of protein turnover is compromised as mitochondria are not created de novo, but rather rely upon removal of damaged components (mitophagy) and incorporation of newly synthesized proteins (biogenesis) (Ryan & Hoogenraad, 2007). Data from long‐term stable isotope labelling in mice have revealed that synthesis of mitochondrial proteins does not decline with age in multiple tissues, including skeletal muscle (Miller et al. 2012; Karunadharma et al. 2015), but rather may even increase (Miller et al. 2012). Thus, loss of mitochondrial proteostasis during ageing may not be due to insufficient synthesis of new proteins.

Despite the maintenance of mitochondrial protein synthesis, errors in mRNA translation and/or misfolding of nascent proteins may occur with ageing in skeletal muscle (Orgel, 1963; Stadtman, 1988; Haar et al. 2017), leading to a loss in mitochondrial proteostasis and function. Mitochondrial unfolding protein response (mtUPR) is a stress response pathway that increases the expression of mitochondrial proteases and chaperones to alleviate mitochondrial stress caused by mistranslated and/or misfolded proteins, as well as promoting biogenesis to maintain proteostasis and mitochondrial function (Schulz & Haynes, 2015). However, current understanding of the regulation of mtUPR and its effectors is largely limited to studies in lower organisms (e.g. C. elegans and Drosophila m.) (Nargund et al. 2012, 2015; Baqri et al. 2014; Owusu‐Ansah et al. 2014; Morrow et al. 2016). While these studies illustrate a role for the mtUPR in promoting healthspan and lifespan by promoting mitochondrial proteostasis, mtUPR in mammalian systems appears to be more complex and may not function in the same manner as observed in lower invertebrate model systems (Seiferling et al. 2016). Intriguingly, two recent reports demonstrate that mtUPR promotes the propagation of damaged mtDNA genomes in C. elegans (Gitschlag et al. 2016; Lin et al. 2016). An accumulation of damaged mtDNA in skeletal muscle has long been suggested to play a role in ageing and age‐related disease (Elson et al. 2001; Bua et al. 2006; Stewart & Chinnery, 2015), though whether it is causal to ageing remains debatable. These data may hint at a pleiotropic role of mtUPR where its continued activation with advancing age may hinder healthy ageing by preferential replication of damaged mtDNA. However, it is important to stress that the physiological response of mtUPR in ageing mammalian skeletal muscle is unknown. Development of relevant, muscle‐specific models are needed to address these outstanding questions.

Maintained mitochondrial biogenesis in aged skeletal muscle suggests that impaired breakdown of damaged/dysfunctional mitochondria through mitophagy may contribute to the age‐associated disruption in mitochondrial proteostasis in skeletal muscle. There are data supporting a loss of macroautophagy in skeletal muscle associated with ageing (Wohlgemuth et al. 2010; Sakuma et al. 2015; Potes et al. 2017), but a paucity of direct evidence for reduced mitophagy in skeletal muscle in old animals or humans, although mitophagy has been directly demonstrated to decline in other post‐mitotic tissues (e.g. brain) with age (Sun et al. 2015). As indirect evidence, several studies have reported increased mitochondrial content in aged skeletal muscle, which is often accompanied by reduced oxygen consumption and increased reactive oxygen species production (Barazzoni, 2000; Conley et al. 2000; Tonkonogi et al. 2003; Chabi et al. 2008; Picard et al. 2011). An accumulation of dysfunctional mitochondria in aged skeletal muscle is consistent with the notion that mitophagy is impaired with ageing. Furthermore, ageing skeletal muscle mitochondria become increasingly sensitive to mitochondrial permeability transition pore (mtPTP) opening, increasing the propensity for a loss in membrane potential and release of pro‐apoptotic factors into the cytosolic space (Chabi et al. 2008; Gouspillou et al. 2014). A loss of mitochondrial membrane potential is a potent stimulus for mitophagy (Twig et al. 2008; Gouspillou et al. 2014). The propagation of mitochondria that are prone to lose membrane potential more easily suggests that either mitophagy is impaired or that capacity for effective removal is exceeded in old skeletal muscle. Mitophagy has been shown, in vitro, to be responsible for removing regions of the mitochondrial reticulum that harbour damaged mtDNA (Suen et al. 2010), which, as mentioned above, accumulates with ageing in skeletal muscle (Elson et al. 2001; Bua et al. 2006). Taken as a whole, these data present a strong case for the observed loss of mitochondrial proteostasis with ageing being due to insufficient degradation of damaged/dysfunctional mitochondria through mitophagy (Fig. 1).

Figure 1. Proposed role for mitophagy as a main contributor to loss of mitochondrial proteostasis with age.

 

Mechanisms of skeletal muscle autophagy/mitophagy and importance in healthy ageing

The plausible role of mitophagy in the maintenance of skeletal muscle mitochondrial proteostasis with ageing makes understanding the mechanisms by which mitophagy is regulated critical for the development of interventions to promote healthy ageing. Many of the same regulatory factors that have been described for macro‐autophagy are also involved in mitophagy (reviewed in Kroemer et al. 2010). Briefly, mitophagy is initiated through activation of the master regulator unc‐51 like autophagy activating kinase 1 (ULK1) (Egan et al. 2011). ULK1 recruits and joins in complex with other autophagy‐related proteins (ATG's), namely the PI3K III nucleation complex (BECLIN1, BCL2, AMBRA1, VPS15, and VPS34) and PI3P binding complex (ATG2A/B, ATG9A, and WIPI2) to regulate the formulation of the membrane leading to the creation of the phagophore. The combined efforts of the ATG12 and microtubule‐associated protein 1A/B‐light chain 3 (LC3) ubiquitin‐like conjugation systems facilitate the maturation and elongation of the autophagosome, which engulfs targeted mitochondria selected for degradation. Once the targeted mitochondria have been engulfed within the autophagosome, fusion with the lysosome occurs to form the autolysosome, and degradation of proteins to amino acids occurs. It is important to note, however, that many of the canonical factors in autophagic degradation that have just been briefly described have not been replicated in skeletal muscle systems (Fig. 2).

Figure 2. Current model for autophagy/mitophagy.

Proteins that have been studied in skeletal muscle are in green. Proteins that have been described to be negatively impacted in the context of ageing (i.e. expression or function) are indicated with an asterisk (^*).

While we have presented indirect evidence consistent with insufficient or dysfunctional mitophagy being responsible for the loss in mitochondrial proteostasis in skeletal muscle ageing, consistent data regarding changes of autophagy/mitophagy markers are also emerging in older animal models and humans (Fig. 2). In rodents, for example, multiple markers of insufficient or declining capacity for removal of damaged proteins through mitophagy in skeletal muscle with ageing are evident. Skeletal muscle from old rats and mice have reduced Atg3, Atg5, Atg12, and lysosome‐associated membrane protein‐2 (Lamp2), as well as a reduced Lc3II/I ratio, a common, crude assessment for autophagy flux (Wohlgemuth et al. 2010; Russ et al. 2012; Zhou et al. 2017). Specific to mitophagy, old rats have reduced PTEN‐induced putative kinase 1 (Pink1) expression in skeletal muscle (Zhou et al. 2017). PINK1 is constantly transported to the mitochondria and, under basal conditions, degraded by mitochondrial proteases (Jin et al. 2010). When mitochondria become damaged and lose membrane potential, PINK1 accumulates on the outer mitochondrial membrane where it recruits the E3 ubiquitin ligase, PARKIN, resulting in poly‐ubiquitination of mitochondrial proteins and their subsequent degradation through mitophagy (Jin et al. 2010). In human skeletal muscle, relative to mitochondrial content, PARKIN expression is reduced in older subjects (Gouspillou et al. 2014). Whether this reduced PARKIN expression contributes to reduced mitophagy in skeletal muscle ageing is not known. It is also unclear whether the recruitment of PARKIN to the mitochondria under stressed conditions is impaired with ageing.

Autophagy in mammals has been linked to ageing due to its regulation by pathways known to regulate lifespan and healthspan (e.g. mTOR, IGF‐1, AMPK, etc.) (Rubinsztein et al. 2011). Both deletion and up‐regulation of multiple autophagy genes in lower organisms are well documented to shorten or extend, respectively, lifespan and healthspan (reviewed in Mizushima & Levine, 2010; Rubinsztein et al. 2011). However, reports of muscle‐specific deletion of autophagy genes are not yet as common. The most studied thus far is muscle‐specific deletion of the Atg7 gene, which encodes for an essential component of both Atg12 and Lc3 ubiquitin‐like conjugation systems for the expansion of the autophagosome membrane (reviewed in Xiong, 2015). Deletion of the Atg7 gene in skeletal muscle causes an increase in protein carbonylation and accumulation of enlarged, swollen mitochondria in skeletal muscle (Masiero et al. 2009) and significantly shortens lifespan (Carnio et al. 2014). Similar findings were observed in adult mice with tamoxifen‐inducible deletion of the Atg7 gene in skeletal muscle, which showed an increase in protein carbonylation, mitochondrial ROS production, and sensitivity of mitochondria to depolarize in response to exercise (Lo Verso et al. 2014). Muscle‐specific deletion of another autophagy gene in the Atg12 conjugation system, Atg5, also leads to impaired clearance of mitochondria and exaggerated muscle protein damage (Masiero et al. 2009). Increase in mitochondrial content and abnormal mitochondrial morphology are also observed in muscle‐specific deletion of the upstream and PI3K nucleation complex member, Vsp15 (Nemazanyy et al. 2013). Thus, loss of skeletal muscle autophagy results in a mitochondrial phenotype highly analogous to what occurs with advancing age. However, the disparity between the importance of skeletal muscle mitochondrial proteostasis for healthy ageing and our understanding of mitophagy regulation over the course of the lifespan warrants further investigation.

The importance of skeletal muscle autophagy/mitophagy for metabolic health

Defects in mitochondrial metabolism and function exacerbate declines in tissue/organ function. Thus, declines in mitochondrial proteostasis in skeletal muscle due to insufficient or impaired mitophagy may play an important systemic role in whole body metabolic homeostasis. Mutations of three phosphorylating sites of the autophagy protein B‐cell lymphoma 2 (Bcl2), thus inhibiting the release of Beclin1 from Bcl2–Beclin1 complex for autophagy activation upon stress, reduces sensitivity to insulin and impairs exercise‐induced Glut4 translocation to the plasma membrane (He et al. 2012). Inhibition of autophagy activation has also been demonstrated to block exercise training‐mediated improvements in endurance capacity and glucose tolerance, as well as protection from a high fat diet (He et al. 2012). More specifically, mitophagy in skeletal muscle appears to be required for the metabolic benefits of exercise training. Using tamoxifen‐inducible, skeletal muscle‐specific ULK1 knockout mice, which do not initiate mitophagy in response to exercise, we have observed a lack of improvement in glucose tolerance following exercise training (Laker et al. 2017). When considered in the context of the loss in skeletal muscle mitochondrial proteostasis with ageing, it is tempting to speculate that impaired mitophagy in skeletal muscle plays a significant role in the prevalence of insulin resistance with ageing (Petersen et al. 2003, 2015).

While the capacity for autophagy/mitophagy has systemic, metabolic consequences, it is unclear whether crosstalk exists between autophagy/mitophagy effectors and insulin signalling or whether the metabolic phenotypes observed are indicative of an overall adaptive metabolic response in a given genetically modified model. Recent in vitro data demonstrate that treatment of L6 myoblasts with autophagy inhibitors bafilomycin or chloroquine inhibited insulin‐stimulated glucose uptake as well as insulin‐mediated Irs1 and Akt phosphorylation (Liu et al. 2015). Furthermore, these data were recapitulated when cells were induced to overexpress a dominant‐negative form of Atg5 (Liu et al. 2015), lending some credence to the possibility of a crosstalk signalling mechanism between the two pathways. In contrast, deletion of Atg7 in skeletal muscle paradoxically protects against diet‐induced obesity and insulin resistance through increased Fgf21 secretion, which promotes browning of white adipose tissue and lipid utilization (Kim et al. 2012), indicative of an adaptive phenotype. Thus, different autophagic/mitophagic regulators, whose importance in maintaining proteostasis is well established, have divergent effects on systemic metabolic homeostasis. Untangling the web of interactions between mitophagy and metabolism and discovering potential crosstalk signalling mechanisms and how ageing impacts those relationships will be a fruitful area of research moving forward.

Stimulating skeletal muscle mitophagy to maintain mitochondrial proteostasis

The apparent decline of autophagy/mitophagy in ageing warrants development of interventions to stimulate it in skeletal muscle to negate some of the ageing consequences. Exercise is a potent and reproducible means to promote skeletal muscle mitochondrial proteostasis (Drake et al. 2015). In rodents, life‐long exercise slows age‐related decline in mobility and performance, and maintains lean body mass and metabolic rate (McMullan et al. 2016) as well as mitochondrial function (Young et al. 1983). While exercise exerts many systemic health benefits against ageing through multiple mechanisms, recent data suggest that mitophagy may be a key step in the improvement of mitochondrial proteostasis in skeletal muscle by exercise (He et al. 2012; Lira et al. 2013; Laker et al. 2014, 2017; Pagano et al. 2014). Exercise training increases basal protein levels of Lc3 and p62 as well as several other autophagy‐related genes in skeletal muscle, indicative of an increase in autophagic capacity (Lira et al. 2013). Importantly, expression of the BH3‐only family protein Bnip3 (or the analogous, Nix), which plays a role in recognizing damaged mitochondria by facilitating docking of autophagosomes through binding to Lc3 (Novak et al. 2010), increases with both acute exercise and exercise training in mouse skeletal muscle (Jamart et al. 2013; Lira et al. 2013). An increase in Bnip3 may suggest improved mitophagy capacity in skeletal muscle as a result of exercise training (Lira et al. 2013). Furthermore, life‐long exercise training (combined with an 8% calorically restricted diet) ameliorated age‐associated increases in mitochondrial protein damage in skeletal muscle and restored protein expression of Atg7, Atg9, and Lamp2 in old rats (Wohlgemuth et al. 2010). While it becomes more apparent from these studies that exercise may exert a positive impact on skeletal muscle mitochondrial proteostasis through mitophagy, the mechanism(s) through which exercise regulates mitophagy in skeletal muscle are not well understood.

Mitophagy was first shown to be regulated through the cellular energy sensor, 5ʹAMP‐dependent kinase (Ampk)‐dependent phosphorylation of ULK1 at Ser555 in primary hepatocytes (Egan et al. 2011). Activation of AMPK has long been used as a marker for intensive exercise in both animal and human studies (Winder & Hardie, 1996; Vavvas et al. 1997; Chen et al. 2000). AMPK is activated in response to an elevated AMP level due to increased ATP hydrolysis (Xiao et al. 2007), functioning to antagonize cellular processes that consume energy (e.g. protein synthesis) and promote those that produce energy (e.g. fatty acid oxidation). Acute exercise, in both mice and humans, has been shown to be associated with an increase of ULK1 phosphorylation at Ser555 in skeletal muscle (Pagano et al. 2014; Møller et al. 2015). Recently, we obtained data in genetically engineered mice to show that Ulk1 phosphorylation at Ser555 and subsequent activation of mitophagy in response to acute exercise are indeed dependent upon Ampk activation in skeletal muscle (Laker et al. 2017). These findings suggest the importance of AMPK–ULK1 regulatory axis in the control of mitophagy in skeletal muscle, but its relationship to ageing is not clear. No data on ULK1 activity with age exist, but ULK1 expression has been demonstrated to be reduced in age‐related pathology (Caramés et al. 2010). Furthermore, the sensitivity of upstream Ampk to exercise in skeletal muscle declines with ageing (Reznick et al. 2007). Therefore, it is reasonable to postulate that activation of ULK1 in skeletal muscle by exercise may decline with ageing. Whether life‐long exercise or interventions with other AMPK agonists (e.g. metformin) can maintain the capacity to activate mitophagy during ageing so to maintain mitochondrial proteostasis in skeletal muscle warrants rigorous investigation.

Conclusion

The loss of mitochondrial proteostasis in skeletal muscle has detrimental, systemic consequences for overall health in ageing. Current data suggest that dysfunctional and/or insufficient mitophagy is a primary cause of an accumulation of poorly functional mitochondria in old skeletal muscle. Impairments in mitophagy may have implications for age‐associated, systemic declines in health through either undiscovered signalling crosstalk or as a consequence of maladaptive metabolic responses. Since exercise is a potent means of promotion of mitochondrial proteostasis in skeletal muscle, in part, through upregulation of mitophagy, exercise and/or exercise‐mimetics may be useful tools for us to uncover how skeletal muscle is regulated in the context of ageing and translate those findings for the promotion of healthy ageing in humans. Undoubtedly, understanding how mitophagy is regulated and developing effective ways to augment it are important areas for future development.

Additional information

Competing interests

None declared.

Author contributions

J.C.D and Y.Z. wrote and edited the manuscript. Both authors have approved the final version of the manuscript and agree to be accountable for all aspects of the work. All persons designated as authors qualify for authorship, and all those who qualify for authorship are listed.

Funding

J.C.D. is supported by a ADA postdoctoral fellowship (1‐16‐PDF‐030). Z.Y. is supported by grants from the National Institute of health (R01‐AR050429).

Biographies

Joshua C. Drake obtained his PhD in ‘Human Bioenergetics’ from Colorado State University. He studied mechanisms of proteostatic maintenance in mouse models of slowed ageing as a member of the Translational Research in Aging and Chronic Disease Laboratory under the direction of Drs Karyn L. Hamilton and Benjamin F. Miller. He now works in the laboratory of Dr Zhen Yan studying exercise‐mediated regulation of mitophagy in skeletal muscle, supported by an ADA post‐doctoral fellowship.

Zhen Yan is a Professor of Medicine, Pharmacology, Molecular Physiology and Biological Physics, and resident member of the Robert M. Berne Cardiovascular Research Centre at the University of Virginia School of Medicine. Trained as a physician scientist, he has developed a rigorous research programme using the state‐of‐the‐art molecular and imaging technologies in a variety of animal models with a focus on mitochondrial quality control by exercise and the systemic health impact.