Mechanisms regulating skeletal muscle growth and atrophy



S. Schiaffino; M. Sandri, Venetian Institute of Molecular Medicine, Via Orus 2, 35129 Padova, Italy

Fax: +39 49 7923 250

Tel: +39 49 7923 232; +39 49 7923 258




Skeletal muscle mass increases during postnatal development through a process of hypertrophy, i.e. enlargement of individual muscle fibers, and a similar process may be induced in adult skeletal muscle in response to contractile activity, such as strength exercise, and specific hormones, such as androgens and β-adrenergic agonists. Muscle hypertrophy occurs when the overall rates of protein synthesis exceed the rates of protein degradation. Two major signaling pathways control protein synthesis, the IGF1–Akt–mTOR pathway, acting as a positive regulator, and the myostatin–Smad2/3 pathway, acting as a negative regulator, and additional pathways have recently been identified. Proliferation and fusion of satellite cells, leading to an increase in the number of myonuclei, may also contribute to muscle growth during early but not late stages of postnatal development and in some forms of muscle hypertrophy in the adult. Muscle atrophy occurs when protein degradation rates exceed protein synthesis, and may be induced in adult skeletal muscle in a variety of conditions, including starvation, denervation, cancer cachexia, heart failure and aging. Two major protein degradation pathways, the proteasomal and the autophagic–lysosomal pathways, are activated during muscle atrophy and variably contribute to the loss of muscle mass. These pathways involve a variety of atrophy-related genes or atrogenes, which are controlled by specific transcription factors, such as FoxO3, which is negatively regulated by Akt, and NF-κB, which is activated by inflammatory cytokines.


eukaryotic translation initiation factor 4E-binding protein 1


activin receptor 2


activin receptor-like kinase 4/5


AMP-activated protein kinase


BCL2/adenovirus E1B 19 kDa protein-interacting protein 3


F-box only protein


fibroblast growth factor-inducible 14


Forkhead box O


histone deacetylase


insulin-like growth factor 1


IκB kinase β




Krüppel-like factor 15


muscle atrophy F-box


mammalian target of rapamycin


mTOR complex 1/2


muscle RING finger 1


nuclear factor κ light-chain enhancer of activated B cells


neuronal nitric oxide synthase


peroxisome proliferator-activated receptor


PPAR-γ co-activator-1α




phosphatase and tensin homolog-induced putative kinase 1


regulated in development and DNA damage responses 1


serum response factor


transforming growth factorβ


tumor necrosis factor α


TNF receptor-associated factor


tripartite motif-containing protein 32


TNF-like weak inducer of apoptosis


vacuolar protein sorting 34


Yin Yang 1


Skeletal muscle mass and muscle fiber size vary according to physiological and pathological conditions. An increase in muscle mass and fiber size, i.e. muscle growth or hypertrophy, occurs during development and in response to mechanical overload (incapacitation or ablation of synergistic muscles, strength training, reloading after unloading) or anabolic hormonal stimulation (testosterone or β2-adrenergic agonists). A decrease in muscle mass and fiber size, i.e. muscle atrophy, results from aging, starvation, cancer, diabetes, bed rest, loss of neural input (denervation, motor neuron disease) or catabolic hormonal stimulation (corticosteroids). The regulation of muscle mass and fiber size essentially reflects protein turnover, i.e. the balance between protein synthesis and degradation within the muscle fibers. However, skeletal muscle fibers are multi-nucleated structures, thus protein turnover may also be affected by cell or nuclear turnover, i.e. addition of new myonuclei, due to fusion of satellite cells, or loss of myonuclei, due to nuclear apoptosis. Before separately considering conditions of muscle growth and muscle atrophy, it is useful to make some general points. First, skeletal muscles and muscle fiber types vary, often drastically, in their response to the same stimulus. This point is evident in many experimental models used in muscle research, without considering extreme cases such as the differential response to testosterone in sexually dimorphic, androgen-sensitive muscles. For example, denervation in the rat diaphragm muscle causes atrophy of type 2X and 2B fibers, no change in type 2A fibers and slight hypertrophy of type 1 fibers [1]. Similar changes are found in other fast rat muscles (S. Schiaffino and S. Ciciliot, unpublished data); however, the type 1 fibers of the slow soleus show marked atrophy after denervation, thus the same fiber type may undergo opposite changes in different muscles. With regard to nutrient deprivation, slow muscles, such as the soleus, are less sensitive to starvation compared to fast muscles [2]. This response is presumably related to the different sensitivity of fast and slow muscles to corticosteroids [3]. Even within the same muscle, for example the rat diaphragm, corticosteroid treatment causes atrophy of type 2B and 2X fibers but not type 2A and 1 fibers [4]. This differential response probably reflects the fact that contractile activity, which is greater in the continuously active type 1 and 2A fibers, opposes the atrophic process, possibly via the transcriptional co-activator PPAR-γ co-activator-1α (PGC-1α) [5].

Second, changes in protein turnover leading to muscle hypertrophy or atrophy do not always proceed according to the simplistic equations suggested by the ‘balance’ analogy, i.e. muscle hypertrophy results from increased protein synthesis and decreased protein degradation, while muscle atrophy results from decreased protein synthesis and increased protein degradation. Goldberg's analyses of muscle growth in hypophysectomized rats showed that, during hypertrophy of the soleus muscle induced by tenotomy of the gastrocnemius, there is decreased protein catabolism as well as increased synthesis of new proteins, while during hypertrophy of the soleus induced by growth hormone, there is increased protein synthesis without any change in protein degradation rates [6]. Starvation causes decreased protein synthesis and increased protein degradation in both fast and slow rat muscles [2]. However, muscle denervation is accompanied by increased protein degradation and increased rather than decreased protein synthesis [7, 8]. One must also consider that fast and slow muscles differ in their protein turnover rates, with slow muscles showing higher rates of both protein synthesis and degradation [2].

Third, changes in protein turnover and cell/nuclear turnover in skeletal muscle do not always proceed in parallel, thus myonuclear domains may vary. The myonuclear domain size, defined as the cytoplasmic volume per myonucleus, varies among fiber types, being larger in type 2B and 2X fibers compared to type 2A and 1 fibers, and, in contrast to initial reports, is not constant in various conditions [8a]. For example, during muscle atrophy caused by denervation or corticosteroids, there is no loss of myonuclei; therefore the myonuclear domain decreases in proportion to the decrease in cross-sectional area [1, 4, 9, 10]. Conversely, myonuclear domains may increase during postnatal muscle growth and in different hypertrophy models (see below).

Muscle growth

In this section, we focus on muscle growth processes that take place after birth, including muscle growth during postnatal development and the process of muscle hypertrophy induced in adult muscle by functional overload. We do not deal specifically with muscle growth during regeneration, which has been discussed previously [11].

Major signaling pathways controlling muscle growth

Two major signaling pathways control skeletal muscle growth: the insulin-like growth factor 1– phosphoinositide-3-kinase–Akt/protein kinase B–mammalian target of rapamycin (IGF1–PI3K–Akt/PKB–mTOR) pathway acts as a positive regulator of muscle growth, and the myostatin–Smad3 pathway acts as a negative regulator (Fig. 1A). The role of the IGF1 pathway has been supported by a variety of gain- and loss-of-function genetic approaches [12]. For example, muscle-specific inactivation of the IGF1 receptor impairs muscle growth due to reduced muscle fiber number and size [13]. Conversely, muscle-specific over-expression of IGF1 causes muscle hypertrophy [14]. In vivo transfection studies in adult mouse and rat muscles have helped to elucidate the pathways downstream of the IGF1 receptor. IGF1 is known to activate both the mitogen-activated protein kinase/extracellular signal-regulated kinase (MAPK/ERK) and the PI3K–Akt pathways. However, only a Ras mutant that selectively activates the PI3K–Akt pathway was able to induce hypertrophy of transfected fibers, whereas a Ras mutant acting specifically on the ERK pathway did not [15]. Accordingly, constitutively active Akt results in a striking hypertrophy of transfected muscle fibers [16, 17], with a similar effect being seen using inducible muscle-specific transgenic models [18-20].

Figure 1.

Signaling modules responsible for skeletal muscle growth during development, regeneration and overload-induced hypertrophy in the adult. We postulate that all these modules converge to a final common pathway centered on mTOR and its effectors that control protein synthesis. (A) Major signaling pathways. IGF1 stimulates mTOR activity and muscle growth via PI3K–Akt. Follistatin induces muscle growth by inhibiting myostatin and activin A. The two pathways cross-talk by direct interaction between Smad3 and Akt. In addition, transcriptional regulation by Smad3/Smad4 heterodimers may repress mTOR and protein synthesis through mechanisms that have not yet been defined. The arrow connecting mTOR with a myonucleus indicates transcriptional roles of mTOR. (b) Additional pathways controlling mTOR activity and protein synthesis. The SRF (serum response factor), PA (phosphatidic acid) and nNOS (neuronal nitric oxide synthase) pathways may be activated by mechanical overload. The dotted arrow connecting newly fused myonuclei (new mn) to the mTOR pathway indicates the postulated increase in protein synthesis and myotube/myofiber growth associated with myoblast/satellite cell fusion. SC, satellite cell; SSC, satellite stem cell.

Akt stimulates protein synthesis by activating mTOR and its downstream effectors. The kinase mTOR interacts with several proteins to form two complexes: mTOR complex 1 (mTORC1) containing raptor and mTOR complex 2 (mTORC2) containing rictor. Here, we focus on mTOR signaling as it relates to skeletal muscle growth, as a detailed general discussion of mTOR function and regulation has been published previously [21]. It should be stressed that mTOR responds to multiple upstream signals in addition to Akt, including amino acids, and it controls several cellular processes in addition to protein synthesis, including autophagy. The crucial role of mTOR in mediating muscle growth is supported by genetic and pharmacological evidence. Muscle-specific mTOR knockout causes reduced postnatal growth, due to the reduced size of fast but not slow muscle fibers, and severe myopathy [22]. A similar phenotype is found in mice lacking raptor in skeletal muscle, whereas those lacking rictor have a normal phenotype, supporting a major role of mTORC1 in mediating the effect of mTOR on protein synthesis [23]. Rapamycin, a specific mTOR inhibitor, acts especially on mTORC1, although mTORC2 is also affected during chronic treatment [21]. Rapamycin inhibits muscle growth during postnatal development [24], muscle regeneration [17] (Fig. 2) and compensatory muscle hypertrophy induced by synergist elimination [16]. Muscle growth during reloading of unloaded muscles is only partially inhibited by rapamycin [24]. Muscle fiber hypertrophy induced by transfection of adult muscles with a constitutively active Akt construct is also blunted by rapamycin [16, 17].

Figure 2.

Muscle growth in regenerating skeletal muscle is dependent on mTOR activity. Rapamycin, a specific mTOR inhibitor, inhibits growth of regenerating muscle. The section was stained with an antibody specific for embryonic myosin heavy chain [177]. Modified from [17]. Scale bar = 50 μm.

Two major effectors of mTORC1 that promote protein synthesis are eukaryotic translation initiation factor 4E-binding protein 1 (4E-BP1) and S6 kinase 1. Muscle growth is apparently unaffected by 4E-BP1 knockout [25]. In contrast, deletion of S6 kinase 1 causes muscle atrophy and partially prevents the response to constitutively active Akt [26]. However, the control of protein synthesis by mTOR is still incompletely characterized [21]. The growth-promoting effect of mTORC1 is repressed by AMP-activated protein kinase (AMPK), and hypertrophy of soleus muscle has been described in AMPK-deficient mice [26].

Another aspect of mTOR function that is incompletely understood is the role of mTOR in transcriptional regulation. In both yeast and mammalian cells, TOR/mTOR controls cell growth by coordinately regulating the synthesis of ribosomes and tRNAs and activating transcription by all three nuclear RNA polymerases (I, II and III). However, ribosomal RNA accumulation induced in overloaded muscles after synergist ablation is only partially inhibited by rapamycin [27]. On the other hand, blockade of mTOR by rapamycin in cultured myotubes is sufficient to block most IGF1-induced changes in transcription [28]. What are the downstream effectors of mTOR that regulate transcription? The transcription factor Yin Yang 1 (YY1) physically interacts with mTORC1 and mediates mTOR-dependent regulation of mitochondrial gene expression via a YY1–PGC-1α complex [29]. More recently, active mTORC1 was found to induce YY1 phosphorylation, resulting in displacement of the polycomb repressor complex, thereby activating transcription of many genes of the insulin/IGF1–Akt pathway, including Igf1, Irs1, Irs2, Akt1 and Akt2. Conversely, mTORC1 inactivation induced by rapamycin results in YY1 dephosphorylation and recruitment of the polycomb repressor complex to the promoter of these genes, with a consequent block of transcription [30]. Muscle-specific inactivation of the Yy1 gene leads to up-regulation of genes of the insulin/IGF1–Akt pathway [30]; however, the effect of this knockout on muscle growth has not been described.

The second major signaling pathway that controls skeletal muscle growth involves myostatin, a member of the transforming growth factor β (TGFβ) superfamily. Myostatin is produced by skeletal muscle and acts as a negative regulator of muscle growth, as shown by the finding that myostatin mutations in various mammalian species cause muscle hypertrophy [31]. Purified myostatin inhibits protein synthesis and reduces myotube size when added to differentiated myotubes in culture [32]. Furthermore, muscle atrophy is induced in mice by systemically administered myostatin [31]. Muscle hypertrophy may also be induced by inhibitory extracellular binding proteins, such as follistatin, whose effect is even greater than the lack of myostatin, because it binds to other TGFβ superfamily members, such as activin A, that act as negative regulators of muscle growth like myostatin does (Fig. 1A). Myostatin and activin A interact and activate a heterodimeric receptor complex with serine–threonine kinase activity, comprising a type II receptor, activin receptor 2 (ACVR2 and ACVR2B), and a type I receptor, activin receptor-like kinase 4 and 5 (ALK4 and ALK5). A soluble form of ACVR2B acts as a myostatin/activin A inhibitor that is capable of inducing muscle hypertrophy in adult mice. Myostatin/activin A signaling in myofibers is mediated by phosphorylation and nuclear translocation of Smad2 or Smad3 transcription factors, and formation of heterodimers with Smad4.

Although the transcriptional targets of the Smad2/Smad4 and Smad3/Smad4 complexes that mediate the inhibitory effect on growth are not known, it is possible that myostatin/activin A signaling interferes with the Akt–mTOR pathway [33, 34]. For example, muscle hypertrophy induced by transfection of dominant-negative ACVR2B is partially prevented by mTOR-specific siRNAs or by rapamycin [33]. Likewise, follistatin-induced muscle hypertrophy is blunted by blocking the IGF1–Akt–mTOR pathway at the level of the IGF1 receptor, via a dominant-negative IGF1 receptor, at the level of Akt, via dominant-negative Akt, or at the level of mTOR, via rapamycin. However, follistatin-induced muscle hypertrophy was unchanged in S6 kinase 1/2 knockout mice [35]. Another recent study showed that virus-mediated over-expression of a form of follistatin that remains localized within the injected muscle stimulates Akt phosphorylation, mTOR signaling and protein synthesis, leading to a striking muscle hypertrophy that is inhibited by rapamycin [36]. Inhibition of Smad3 activity by follistatin is critical for activation of Akt–mTOR signaling, as constitutively active Smad3 was found to suppress follistatin-induced muscle growth and mTOR activation. It is also possible that a direct interaction between Smad3 and Akt, as demonstrated in other cell systems [37, 38], may be involved in cross-talk between the myostatin/activin A and IGF1 pathways in skeletal muscle.

Additional signaling pathways controlling muscle growth

As shown in Fig. 1B, other important signaling pathways are also known to control skeletal muscle growth. The transcription factor SRF (serum response factor) is required for muscle growth during development, as shown by muscle-specific SRF knockout [39, 40]. Using an inducible, muscle-specific knockout model, SRF was also found to be required for muscle hypertrophy induced by synergist elimination [41]. In this model, the effect of SRF is apparently mediated via release of interleukins 4 and 6 (IL-4 and IL-6), which act in a paracrine manner to induce satellite cell proliferation and fusion [41] (see below). Akt phosphorylation is apparently unchanged in control and SRF mutant muscles at various time points after synergist ablation. However, a previous study reported that SRF is able to activate the Akt pathway via a muscle-enriched microRNA, miR-486, that targets the phosphatase and tensin homolog PTEN, which negatively affects PI3K–Akt signaling [42]. In addition, as activation of mTOR and its downstream effectors were not examined in the previous study [41], it is possible that mTOR activity and protein synthesis are increased by SRF in muscles undergoing overload-induced hypertrophy, either via a transcriptional mechanism mediated by SRF itself or as a secondary consequence of incorporation of new myonuclei resulting from satellite cell activation and fusion. Finally, it should be stressed that SRF is known to control the transcription of several cytoskeletal and sarcomeric protein genes, including those for α-actin, by binding to CArG box regulatory elements.

The Akt–mTOR pathway is also a point of convergence for additional signaling pathways that are known to promote muscle growth. This appears to be the case for androgens and β2-adrenergic agents, which both have well-known anabolic effects on skeletal muscles. Androgens potently stimulate muscle growth: testosterone loss in male mice decreases muscle Igf1 mRNA, Akt phosphorylation and the rate of myofibrillar protein synthesis; these changes are all reversed by nandrolone treatment [43]. Muscle hypertrophy induced by β2-adrenergic agents, such as clenbuterol or formoterol, is accompanied by a significant increase in Akt phosphorylation [44] and is completely blocked by rapamycin [45]. Another signaling pathway that controls muscle growth involves Wnt7a, an extracellular protein that acts both on satellite stem cells, increasing their numbers, and on myofibers by activating the PI3K–Akt pathway via its receptor Fzd7 [46, 47]. A pathway linked to mTOR activation involves neuronal nitric oxide synthase (nNOS). When activated in myofibers by functional overload, nNOS generates nitric oxide (NO) and causes peroxynitrite-dependent activation of the cation channel Trpv1, located in the sarcoplasmic reticulum. The resulting increase in intracellular Ca2+ induced by Trpv1-mediated Ca2+ release triggers activation of mTOR [48]. Furthermore, mTOR is also activated by mechanical stimulation via an Akt-independent pathway involving phosphatidic acid and activated phospholipase D [49].

A novel form of PGC-1α (PGC-1α4), which results from alternative promoter usage and splicing of the primary transcript, is involved in muscle growth, as shown by the finding that mice with skeletal muscle-specific transgenic expression of PGC-1α4 show increased muscle mass and strength [50]. PGC-1α4, which is expressed at significant levels in skeletal muscle, is a shorter, truncated form of the previously described PGC-1α [50a], now referred to as PGC-1α1, which is involved in mitochondrial biogenesis and not in muscle growth. In cultured muscle cells, PGC-1α4 was found to induce IGF1 and repress myostatin, thus promoting myotube hypertrophy, which was blocked by an IGF1 receptor inhibitor. Myotube growth induced by treatment with clenbuterol was also blunted by PGC-1α4 knockdown.

Satellite cell fusion and increase of myonuclei during muscle growth

The issue of whether satellite cell proliferation and fusion contributes to muscle growth has been the subject of debate [51]. There is no doubt that myoblast fusion is essential for muscle growth during early stages of muscle differentiation. For example, myotube growth in culture is impaired when myoblast fusion is inhibited, either during formation of the nascent myotube or during the transition from nascent to mature myotube [52]. IL-6 and IL-4 released by the myotubes act on myoblasts, promoting their proliferation and fusion, respectively [53, 54]. Muscle cells lacking IL-4 or the IL-4α receptor subunit form smaller myotubes with fewer myonuclei [53]. Muscle growth during early postnatal development (from P0 to approximately P21 in mice and rats) is also accompanied by, and presumably dependent on, a continuous increase in the number of myonuclei resulting from satellite cell fusion [55] (an approximately fivefold increase from P3 to P21 in mouse extensor digitorum longus muscle [56]). Muscle regeneration recapitulates many aspects of embryonic and neonatal myogenesis, with satellite cells acting as a major myogenic stem cell, and undergoing active proliferation and fusion during formation of new myofibers [11]. A distinct feature of muscle regeneration, which is missing in normal muscle development, is the central role of inflammation and of various macrophage populations in the muscle growth process [57].

On the other hand, muscle hypertrophy at late postnatal stages takes place without a significant contribution of satellite cell fusion. For example, the approximately twofold increase in myofiber cross-sectional area from P21 to P56 in mouse extensor digitorum longus muscle occurs with a negligible change in myonuclear number [56]. In adult skeletal muscle, clenbuterol-induced hypertrophy does not involve satellite cell fusion [58], although satellite cell activation is induced by androgens [59]. Similarly, myonuclear number is not increased during muscle growth upon reloading of unloaded muscles [9, 60]. Satellite cell proliferation and fusion were not detected during muscle hypertrophy induced by a muscle-specific, inducible and constitutively active Akt1 transgene [18], or by over-expression of JunB [61]. The contribution of satellite cells to muscle hypertrophy induced by blockade of the myostatin/activin A pathway is controversial. In two studies, satellite cell activation was not detected after injection of vectors encoding the myostatin propeptide, which binds non-covalently to myostatin and inhibits its activity [62], or in hypertrophic muscles expressing dominant-negative ACVR2B [33]. In contrast, another study reported an increase in BrdU-positive myonuclei and increased numbers of satellite cells when using a soluble ACVR2B to block this pathway [63]. A recent detailed study concluded that satellite cells play little or no role in myostatin/activin A signaling in vivo, based on the finding that satellite cell and myonuclear number were unchanged in hypertrophic muscles after injection of soluble ACVR2B, and that muscle hypertrophy induced by over-expressing follistatin also occurs in mice lacking syndecan4 or Pax7, which have compromised satellite cell function or number, respectively [64].

A widely used model of muscle hypertrophy in adult mice or rats is compensatory hypertrophy induced by ablation of synergist muscles: for example hypertrophy of the plantaris or soleus muscle after removal of the gastrocnemius, or hypertrophy of the extensor digitorum longus after removal of the tibialis anterior muscle. This model of acute functional overload causes immediate satellite cell proliferation and fusion [65, 66], with a consequent increase in the number of myonuclei [54, 67]. A similarly dramatic increase in mechanical load may be induced in human skeletal muscle by high-intensity eccentric contractions, which also cause proliferation of satellite cells [68] and, when repeated, are known to induce muscle hypertrophy. Depending on the intensity of the exercise and the trained/untrained state of the host, eccentric contractions may cause a spectrum of responses, ranging from severe muscle damage and local inflammation to mild muscle damage (myofibrillar disruptions) without inflammation, to remodeling of the extracellular matrix without obvious damage to the muscle fibers. Similar changes have been reported in rat or mouse muscles after elimination of synergists, and it is possible that the satellite cell activation seen in each of these conditions may reflect a response that is similar to that observed in muscle regeneration. A role of satellite cells in IGF1-induced muscle hypertrophy was suggested by the inhibitory effect of gamma radiation [69]; however, exposure to radiation may affect protein synthesis within the myofibers, thus complicating interpretation of this experiment [70].

IL-6 produced by myofibers and satellite cells is increased in overloaded muscles, suggesting a role for IL-6 and downstream signal transducer and activator of transcription 3 (STAT3) signaling in satellite cell proliferation, myonuclear accretion and hypertrophy [54]. IL-4 was found to control myoblast fusion, thus promoting the transition from nascent to mature myotubes in cultured muscle cells and during postnatal growth [53]. A recent study indicated that satellite cell proliferation and fusion during overload-dependent hypertrophy are induced by IL-6 and IL-4, respectively, and that both cytokines are released by overloaded myofibers under the control of SRF [41]. IL-4 expression is decreased by muscle-specific SRF knockout, which causes reduced postnatal muscle growth and reduced hypertrophy after synergist ablation [40], with both defects being rescued by IL-4 but not IL-6 over-expression [41]. However, another recent study reported that recruitment of satellite cells is not required for muscle hypertrophy, because hypertrophy was unchanged in mice in which more than 90% of satellite cells were ablated by diphtheria toxin A using an inducible Pax7–diphtheria toxin A transgene [71]. How may these opposite conclusions about the mechanism of muscle hypertrophy induced by functional overload be reconciled?

It is clear that skeletal muscle has the capacity to activate two modes of hypertrophy, either with or without satellite cell involvement, as clearly shown by the two stages of muscle growth during postnatal development. These two modes of hypertrophy are also observed at later stages, and may be activated in the adult in response to various stimuli. It is likely that satellite cells are activated and contribute to hypertrophy when an acute stimulus is involved, such as after elimination of synergistic muscles, or after strong exercise with eccentric contractions, i.e. under conditions when some form of muscle damage occurs. In contrast, more gradual exercise, or reloading of unloaded muscles, does not trigger satellite cell activation and fusion. It may be envisaged that, when one of the two available modes of response is artificially impaired, as is the case in gene or cell ablation models, the muscle will use the remaining available mode. Thus, in the absence of satellite cells, a growth response is still induced by functional overload, but via increased protein synthesis alone. Loss-of-function approaches, involving either genes or cells, should thus be interpreted with caution, because compensatory adaptations may occur, not only when the perturbation is produced during development, as in traditional knockout models, but also when it is produced in the adult using inducible systems.

Muscle atrophy

Muscle atrophy involves the shrinkage of myofibers due to a net loss of proteins, organelles and cytoplasm. Acute muscle atrophy, as occurs in many pathological conditions, is due to hyperactivation of the cell's main degradation pathways, including the ubiquitin–proteasome system and the autophagy–lysosome pathway. Recent studies have highlighted a complex scenario whereby these catabolic pathways modulate one another at different levels, and are also coupled at various points to biosynthetic pathways. The result is a coordinated balance between protein degradation and synthesis that reflects the physiological state of the muscle fiber. As muscle accounts for such a large proportion of total body mass, particularly total body protein, this local balance has a significant effect on general protein homeostasis.

The ubiquitin–proteasome and autophagy–lysosome machinery are activated in atrophying muscles

Activation of the cell's proteolytic systems is transcriptionally regulated, and a subset of genes that are commonly up- or down-regulated has been identified in atrophying skeletal muscle, regardless of the catabolic condition [72-75]. These common genes are thought to regulate the loss of muscle components, and were thus designated atrophy-related genes or ‘atrogenes’ [75-77]. Among the up-regulated atrophy-related genes are transcripts belonging to the ubiquitin–proteasome and autophagy–lysosome systems. The up-regulation of several ubiquitin–proteasome and autophagy-related genes is normally blocked by Akt through negative regulation of Forkhead box O (FoxO) transcription factors [77-79].

In muscle, the ubiquitin–proteasome system is required to remove sarcomeric proteins in response to changes in muscle activity. The rate-limiting step of the ubiquitination process, which affects subsequent proteasome-dependent degradation, is catalysed by the E3 enzyme, which is a ubiquitin ligase. Among the known E3s, only a few are both muscle-specific and up-regulated during muscle loss. The first to be identified were atrogin-1/MAFbx (muscle atrophy F-box) and muscle RING finger 1 (MuRF1). Mice lacking atrogin-1/MAFbx and MuRF1 are resistant to muscle atrophy induced by denervation [72]. Moreover, knockdown of atrogin-1 prevents muscle loss during fasting [80], whereas MuRF1 knockout mice (but not atrogin-1 knockout mice) are also resistant to dexamethasone-induced muscle atrophy [81]. So far, very few muscle proteins have been identified as substrates for atrogin-1, and those that have been identified appear to be involved in growth-related processes or survival pathways. For example, atrogin-1 promotes degradation of MyoD, a key muscle transcription factor, and of eukaryotic translation initiation factor 3 subunit F (eIF3-f), an important activator of protein synthesis [82, 83]. In the heart, atrogin-1 ubiquitinates and reduces the levels of calcineurin A, an important factor triggering cardiac hypertrophy in response to pressure overload [84]. Interestingly, immunoprecipitation experiments in C2C12 myoblasts and myotubes have found that atrogin-1 interacts with sarcomeric proteins, including myosins, desmin and vimentin, as well as transcription factors, components of the translational machinery, enzymes involved in glycolysis and gluconeogenesis, and mitochondrial proteins [85]. Whether atrogin-1 ubiquitinates these proteins has yet to be proven. Conversely, MuRF1 was reported to interact with and control the half-life of many important muscle structural proteins, including troponin I [86], myosin heavy chains [87, 88], actin [89], myosin binding protein C and myosin light chains 1 and 2 [90]. Presumably, additional E3s that have not yet been identified are also activated during atrophy to promote the clearance of soluble cellular proteins and to limit/regulate anabolic processes. A recent paper reported that Trim32 (tripartite motif-containing protein 32) is a crucial E3 ligase for the degradation of thin filaments (actin, tropomyosin and troponins), α-actinin and desmin [91]. However, Trim32 knockout mice are not protected from atrophy, but instead show impaired recovery of muscle mass after atrophy [92]. Another E3 ubiquitin ligase that has been found to play a critical role in atrophy is TRAF6 (TNF receptor-associated factor) [93], which mediates the conjugation of Lys63-linked polyubiquitin chains to target proteins. Lys48-linked polyubiquitin chains are a signal for proteasome-dependent degradation, but Lys63-linked polyubiquitin chains play other roles, such as regulating autophagy-dependent cargo recognition by interacting with the scaffold protein p62 (also known as SQSTM1) [94-96]. Muscle-specific TRAF6 knockout mice have a decreased amount of polyubiquitinated proteins, almost no Lys63-polyubiquitinated proteins in starved muscles [97], and are resistant to muscle loss induced by denervation, cancer or starvation [93, 97, 98]. The mechanism of this protection involves both direct and indirect effects of TRAF6 on protein breakdown. In fact, TRAF6-mediated ubiquitination is required for the optimal activation of c-Jun N-terminal kinase, AMPK, FoxO3 and NF-κB [97]. All of these factors are crucial regulators of atrogin-1 and MuRF1 expression and of several autophagy-related genes. Inhibition of TRAF6 reduces the induction of atrogin-1 and MuRF1, thereby preserving muscle mass under catabolic conditions.

Specific ubiquitin ligases may be involved in different models of muscle wasting and at different stages of the atrophy process. For instance, the HECT domain ubiquitin ligase Nedd4-1 has been reported to be up-regulated mainly during muscle disuse. Indeed, deletion of the Nedd4-1 gene specifically in skeletal muscle results in partial protection from muscle atrophy in denervated type II fibers. However, Nedd4-1 knockout mice have smaller muscles, suggesting that this E3 may play additional roles during myogenesis or in the control of protein synthesis [99].

Mul1 is a mitochondrial ubiquitin ligase that plays an important role in mitochondrial network remodeling. Mul1 is up-regulated by the FoxO family of transcription factors under catabolic conditions, such as fasting or denervation, and causes mitochondrial fragmentation and removal via autophagy (mitophagy) [100]. Importantly, knocking down Mul1 spares muscle mass during fasting. Mul1 ubiquitinates the mitochondrial pro-fusion protein mitofusin 2, causing its degradation via the proteasome system. The exact mechanism that triggers Mul1-dependent mitochondrial dysfunction and mitophagy is unclear, but it has been reported that mitofusin degradation is permissive for mitochondrial fission and mitophagy [101].

Carboxy terminus of Hsc70 interacting protein (CHIP) is another ubiquitin ligase, which regulates ubiquitination and lysosomal-dependent degradation of filamin C, a muscle protein found in the Z-line [102]. Filamins undergo unfolding and refolding cycles during muscle contraction, and are therefore prone to irreversible damage [102]. Alterations to filamin structure trigger binding of the co-chaperone BAG3, which is a complex comprising the chaperones Hsc70 and HspB8, as well as the ubiquitin ligase CHIP. CHIP ubiquitinates BAG3 and filamin, which are recognized and delivered to the autophagy system by p62 [102]. Interestingly, filamin B half life is controlled, at least during myogenesis, by another ubiquitin ligase, ASB2β, which is mainly expressed in muscle cells. In this case, ubiquitination of filamin B by ASB2β leads to proteasome-dependent degradation [103].

In skeletal muscle, E3 ligases also have important regulatory functions in signaling pathways. For example, it was recently found that the ubiquitin ligase Fbxo40 (F-box only protein) regulates anabolic signals [104]. Fbxo40 ubiquitinates and affects the degradation of insulin receptor substrate 1, a downstream effector of insulin receptor-mediated signaling. Inhibition of Fbxo40 by RNAi induces hypertrophy in myotubes, and Fbxo40 knockout mice display bigger muscle fibers [104].

Although some E3 ligases involved in muscle protein ubiquitination and breakdown have been identified, very little is known about how ubiquitinated proteins are recognized and delivered to the proteasome. ZNF216 has been identified as an important player in the recognition and delivery of ubiquitinated proteins to the proteasome during muscle atrophy. Interestingly, ZNF216 is up-regulated by FoxO transcription factors in atrophying muscles, and ZNF216-deficient mice are partially resistant to muscle loss during denervation. The absence of ZNF216 in muscle leads to accumulation of polyubiquitinated proteins [105].

Another important system for extraction and degradation of ubiquitinated proteins from larger structures is the p97/valosin containing protein (VCP) ATPase complex. p97/VCP is induced during denervation, and over-expression of a dominant-negative p97/VCP reduces overall proteolysis by the proteasome and lysosome pathways, and blocks the accelerated protein breakdown induced by FoxO3. Interestingly, p97 and its co-factors, Ufd1 and p47, have been found to be associated with specific myofibrillar proteins, suggesting a role for p97 in extracting ubiquitinated proteins from myofibrils [106].

Although a great body of research has focused on the ubiquitination process, little is known about the role of deubiquitination and its contribution to muscle atrophy. The largest class of deubiquitinating enzymes are ubiquitin-specific proteases. So far, only two ubiquitin-specific proteases (USP14 and USP19) have been found to be up-regulated in atrophying muscles [73, 107]. Knockdown of USP19 in myotubes results in decreased protein degradation and reverts dexamethasone-induced loss of myosin heavy chain [108].

Macroautophagy, hereafter referred to as autophagy, is the other proteolytic system that is activated in catabolic conditions and that is under FoxO regulation [109]. The various types of autophagy, including their regulation and involvement in muscle homeostasis, have been reviewed recently [110]. Briefly, autophagy is a highly conserved homeostatic mechanism that is used for the degradation and recycling, through the lysosomal machinery, of bulk cytoplasm, long-lived proteins and organelles [111]. Although autophagy was initially considered a non-selective degradation pathway, the presence of more selective forms of autophagy is becoming increasingly evident. Indeed, autophagy may trigger the selective removal of specific organelles, such as mitochondria, via mitophagy. In mammals, parkin, PINK1 (phosphatase and tensin homolog-induced putative kinase 1), and BCL2/adenovirus E1B 19 kDa protein-interacting protein 3 (Bnip3 and Bnip3L) have been shown to regulate mitophagy, and inactivation of the genes encoding these proteins leads to mitochondrial abnormalities [112, 113]. PINK1 is normally absent in healthy mitochondria because it is constitutively degraded by mitochondrial proteases. However, once mitochondria are damaged, PINK1 is no longer degraded and accumulates. PINK1 induces parkin recruitment to mitochondria, promoting mitophagy through ubiquitination of outer mitochondrial membrane proteins that are recognized by p62, which then brings autophagic vesicles to ubiquitinated mitochondrial proteins [114, 115]. Bnip3 and Bnip3L are BH3-only proteins that are localized at the outer membrane of the mitochondria after cellular stress, and reportedly bind directly to LC3 (MAP1LC3A microtubule-associated protein 1 light chain 3 alpha), thereby recruiting the autophagosome to damaged mitochondria [116, 117]. In atrophying muscle, the mitochondrial network is dramatically remodeled following fasting or denervation, and autophagy via Bnip3 contributes to mitochondrial remodeling [101, 118-120]. Expression of the fission machinery is sufficient to cause muscle wasting in mice, whereas inhibition of mitochondrial fission prevents muscle loss during denervation, indicating that disruption of the mitochondrial network is a crucial amplificatory loop of the muscle atrophy program [101, 118]. Conversely, impairment of basal mitophagy is deleterious to muscle homeostasis, and leads to the accumulation of damaged and dysfunctional mitochondria [121]. Accordingly, the phenotype of mice with muscle-specific inactivation of various genes coding for autophagy-related proteins, such as Atg7, Atg5 or nutrient-deprivation autophagy factor-1 (NAF-1), a Bcl-2-associated autophagy regulator, results in atrophy, weakness and various myopathic features [122-124]. In addition, altered regulation of autophagy-related genes leads to muscle dysfunction. Histone deacetylases 1 and 2 (HDACs) were found to regulate muscle autophagy by controlling the expression of autophagy genes. Muscle-specific ablation of both HDAC1 and HDAC2 results in partial perinatal lethality, and the HDAC1/2 knockout mice that do survive develop a progressive myopathy characterized by impaired autophagy [125, 126].

IGF1–AKT–FoxO signaling

Several studies have shown that the IGF1 and/or insulin signaling suppress protein breakdown while promoting muscle growth [127-129]. Additional data supporting the role of the IGF1 pathway in regulating muscle atrophy have been obtained from studies of Akt. Electroporation of constitutively active Akt in adult myofibers completely blocks muscle atrophy induced by denervation [16]. Akt transgenic mice display muscle hypertrophy and protection from denervation-induced atrophy [19, 20, 130], showing that the Akt pathway promotes muscle growth and simultaneously blocks protein degradation [20, 33]. In particular, Akt regulates both the ubiquitin–proteasome system and the autophagy–lysosome pathway, and this action is mediated by FoxO transcription factors. The FoxO family members that are important for skeletal muscle include three isoforms: FoxO1, FoxO3 and FoxO4. Akt phosphorylates all FoxOs, promoting their export from the nucleus to the cytoplasm. As predicted, the reduced activity of the Akt pathway observed in various models of muscle atrophy leads to decreased levels of phosphorylated FoxO in the cytoplasm and a marked increase in nuclear FoxO [131] (Fig. 3). The translocation and transcriptional activity of FoxO members is sufficient to promote atrogin-1 and MuRF1 expression, and muscle atrophy. Studies utilizing FoxO3 over-expression in adult muscle or muscle-specific FoxO1 transgenic mice showed markedly reduced muscle mass and fiber atrophy [77, 132, 133]. In contrast, FoxO knockdown by RNAi blocks the up-regulation of atrogin-1 expression during atrophy and prevents muscle loss [77, 134].

Figure 3.

Protein degradation regulates protein synthesis. In the presence of growth factors, the PI3K–Akt/protein kinase B pathway sequesters FoxO1/3/4 transcription factors in the cytoplasm. In the absence of growth factors, Akt is inactive, and therefore, FoxOs are translocated into the nucleus and induce the transcription of target genes that regulate the ubiquitin–proteasome and autophagy–lysosome systems. mTOR senses the amino acids derived from the proteasome, or, when localized on lysosomes, the amino acid flux derived from lysosomal protein breakdown, and is therefore activated.

Cross-talk between protein breakdown and protein synthesis is not limited to Akt, but also involves FoxO. Activation of FoxO in Drosophila muscle up-regulates 4E-BP1 [135] and represses mTOR via sestrin [136]. Consistently, in mammals, FoxO3 reduces total protein synthesis in adult muscle [137]. Thus, when Akt is active, protein breakdown is suppressed, and when FoxO is induced, protein synthesis is further suppressed. This is not trivial, as FoxO activity is regulated by several post-translational modifications, including phosphorylation, acetylation and mono- and polyubiquitination [138]. Adding an additional level of complexity, the regulatory consequences of these changes appear to be specific for individual FoxO members. For example, recent evidence suggests that acetylation negatively regulates FoxO3 activity, but has no effect on FoxO1 [139]. Mutants of FoxO3 that mimic the effect of acetylation have cytosolic localization and a reduced capacity to induce transcription of the gene encoding atrogin-1, and cause muscle atrophy [140]. Most of these regulatory mechanisms are Akt-independent, and may play a role in muscle atrophy induced by oxidative or energy stress.

Other studies have revealed a connection between AMPK and FoxO3. AMPK phosphorylates several Akt-independent sites on FoxO3, thereby stimulating its transcriptional activity [141, 142]. Indeed, treatment of muscle cultures with 5-aminoimidazole-4-carboxamide riboside (AICAR), an activator of AMPK, increases protein breakdown and atrogin-1 expression via the FoxO family [143]. It has recently been shown that FoxO3 is activated via AMPK in myofibers to induce expression of atrogin-1 and MuRF1 under conditions of energy stress [101, 144]. Activation of AMPK also leads to induction of some autophagy-related genes encoding proteins such as LC3 and Bnip3.

Increased oxidative stress occurs during denervation and hindlimb suspension. During these disuse conditions, nNOS moves from the sarcolemma, where it is bound to the dystrophin–glycoprotein complex, to the cytosol. Free cytosolic nNOS induces oxidative stress and enhances FoxO3-mediated transcription of atrogin-1 and MuRF1, thereby causing muscle loss [145]. Interestingly, the NF-κB pathway is not involved in nNOS-mediated muscle atrophy [145]. Similarly, when dihydropyridine receptor (DHPR) is reduced in adult muscle by RNAi, muscle atrophy is triggered via nNOS relocalization and FoxO3 activation [146]. However, in this latter setting, the genes up-regulated by FoxO3 are those encoding the autophagy regulators LC3, vacuolar protein sorting 34 (VPS34) and Bnip3 as well as the lysosomal enzyme cathepsin L. In humans, the diaphragm of patients that are mechanically ventilated undergoes rapid atrophy caused by activation of proteolytic systems, including autophagy, through Akt inhibition and FoxO1 induction [147]. Interestingly, oxidative stress is increased and therefore contributes to FoxO activation in this example of disuse-mediated atrophy.

FoxO activity is also modulated by direct or indirect actions of co-factors and by interaction with other transcription factors. FoxOs have been found to interact with PGC-1α, a critical co-factor for mitochondrial biogenesis [148, 149]. Maintaining high levels of PGC-1α under catabolic conditions (either in transgenic mice or by transfecting adult myofibers) spares muscle mass during denervation, fasting, heart failure, aging and sarcopenia – similar to the effect observed for expression of constitutively active FoxO3 [5, 150, 151]. Similar beneficial effects were recently obtained by over-expression of PGC-1β, a homolog of PGC-1α [152]. The positive action on muscle mass of these co-factors is due to inhibition of autophagy–lysosome and ubiquitin–proteasome degradation. PGC-1α and PGC-1β reduce protein breakdown by inhibiting the transcriptional activity of FoxO3 and NF-κB, but do not affect protein synthesis. Thus, these co-factors prevent the excessive activation of proteolytic systems by inhibiting the action of the pro-atrophy transcription factors without perturbing the translational machinery.

We recently reported that the transcription factor JunB blocks atrophy and promotes hypertrophy in adult muscles [61]. Indeed, JunB blocks myofiber atrophy of denervated tibialis anterior muscles and cultured myotubes induced by FoxO3 over-expression, dexamethasone treatment or starvation. Under these conditions, JunB prevents activation of atrogin-1 and partially prevents activation of MuRF1, thereby reducing the increase in overall protein degradation induced by activated FoxO3. Further analysis revealed that JunB does not inhibit FoxO3-mediated activation of the autophagy–lysosome system, but only ubiquitin–proteasome degradation, by inhibiting atrogin-1 and MuRF1 induction under catabolic conditions. In fact, JunB directly binds FoxO3, thereby preventing its recruitment to the promoters of key atrogenes. Moreover, JunB over-expression is sufficient to induce dramatic hypertrophy of myotubes and adult muscle. These hypertrophic changes depend on increased protein synthesis, without affecting the basal rate of protein degradation. The growth-promoting effects mediated by JunB in muscle resemble the effects of inhibiting the TGFβ pathway [33, 34]. Indeed, JunB over-expression markedly suppresses myostatin expression in transfected myotubes and decreases the phosphorylation of Smad3, the transcription factor downstream of the myostatin–TGFβ signaling pathway [61].

Inflammatory cytokines and NF-κB signaling

NF-κB transcription factors are expressed in skeletal muscle and are activated by inflammatory cytokines, particularly tumor necrosis factor α (TNFα). Indeed, inflammation is a potent trigger of muscle wasting and cachexia [153]. NF-κB is maintained in the inactive state by binding of a family of inhibitory proteins called IκB. The increase in the TNFα level induces activation of an IκB kinase (IKKβ) complex that phosphorylates IκB, resulting in its ubiquitination and proteasomal degradation. This leads to nuclear translocation of NF-κB and activation of NF-κB-mediated gene transcription [153].

Transgenic mice that over-express IKKβ specifically in muscle show severe muscle wasting that is mediated, at least in part, by the ubiquitin ligase MuRF1, but not by atrogin-1 [154]. In contrast, muscle-specific inhibition of NF-κB by transgenic expression of a constitutively active IκB mutant does not induce an overt phenotype, but denervation atrophy is substantially reduced [155]. Mice deficient for the p105/p50 subunit of NF-κB are partially resistant to muscle atrophy induced by hindlimb unloading [156]. However, one of the effects of TNFα and pro-inflammatory cytokines is to induce insulin resistance and suppression of the IGF1–Akt pathway [157-159]. Therefore, Akt phosphorylation should always be monitored when NF-κB signaling is altered, as Akt inhibition may substantially contribute to muscle wasting. Indeed IKKβ conditional knockout mice are resistant to muscle atrophy but show activation of Akt [160]. The significance of decreased muscle atrophy following IKKβ ablation and the degree to which this effect is Akt-dependent remains unclear. Nevertheless, these findings highlight the relevance of the cross-talk between the two pathways, and future studies are required to elucidate the respective contributions of the IKKβ–NF-κB and Akt–FoxO pathways to muscle atrophy.

A recent study revealed an unexpected connection between TNFα signaling and myogenin on MuRF1 and atrogin-1 expression: TNFα treatment causes up-regulation of myogenin, MuRF1 and atrogin-1. Interestingly, a G protein-coupled receptor blocks TNFα-mediated myogenin up-regulation by activating Gαi2 [161] and expression of muscle-specific ubiquitin ligases. However, the precise mechanisms of TNFα-mediated myogenin regulation, the interplay with Gαi2 and the implications for muscle wasting are still far from fully understood.

TNF-like weak inducer of apoptosis (TWEAK) is a member of the TNF superfamily that was recently found to induce muscle atrophy [158, 162]. TWEAK acts on responsive cells by binding to fibroblast growth factor-inducible 14 (Fn14), a small cell-surface receptor. Fn14 is up-regulated in denervated muscle, allowing NF-κB activation and consequently MuRF1 (but not atrogin-1) expression [162]. TWEAK knockout mice display reduced atrophy after denervation, as well as reduced NF-κB activation and MuRF1 expression. However, Fn14 does not increase under all conditions of muscle atrophy; for instance, it is not induced by dexamethasone treatment. Another important player in NF-κB signaling is the ubiquitin ligase TRAF6, which is required for Fn14 up-regulation during fasting [97]. As noted earlier, TRAF6 is also required for activation of FoxO3 and AMPK in starved muscles and for induction of the ubiquitin–proteasome and autophagy–lysosome systems [97].

The pro-inflammatory cytokines TNFα, IL-6 and IL-1 also activate the Janus kinase/signal transducer and activator of transcription (JAK/STAT) pathway. Interestingly, sepsis and cancer induce STAT3 phosphorylation in muscles, and STAT3 inhibition spares muscle mass in tumor-bearing mice [163]. Moreover, over-expression of Stat3 is sufficient to induce muscle atrophy and to up-regulate atrogin-1. However, another recent study identified an unexpected role of Stat3 in autophagy regulation. Stat3 has been reported to block VPS34 expression, resulting in alteration of assembly of the Vps34–Beclin1–Vps15–Atg14 complex, and therefore autophagy inhibition and muscle degeneration [164]. This Stat3-dependent regulation of autophagy occurs downstream of Fyn tyrosine kinase.

Other signaling pathways

Myostatin inhibition and its role in muscle growth has been described above; however, the mechanism of myostatin activation and its role and capacity to trigger muscle atrophy remain unclear. Myostatin activation has been reported to induce massive [165], mild or no atrophy at all [166, 167]. However, in muscle cell cultures, myostatin was reported to up-regulate essential atrophy-related ubiquitin ligases. This regulation was found to be FoxO-dependent and NF-κB-independent [168]. Importantly, myostatin expression is controlled by FoxO1, supporting the concept that the myostatin pathway synergizes with Akt–FoxO signaling [169]. A recent study showed that inhibition of myostatin by soluble ACVR2B prevents and fully reverses skeletal muscle loss and atrophy of the heart in tumor-bearing animals [63]. Such treatment dramatically prolongs the survival of these animals, suggesting potential therapeutic efficacy in patients with cancer cachexia. Reports attempting to dissect the downstream signaling have shown that Smad2 and Smad3 are the principle transcription factors that mediate myostatin's effects on muscle mass [33, 34, 36, 100]. However, as mentioned above, specific transcriptional targets of Smad2 and Smad3 are still unknown, and mechanisms of Smad-dependent atrophy remain to be established.

Glucocorticoid levels are increased in many pathological conditions associated with muscle loss. Glucocorticoid treatment induces atrogin-1 and MuRF1 expression and muscle wasting in cell culture and in vivo [16, 77, 88, 127, 170]. In contrast, adrenalectomy or treatment with a glucocorticoid receptor antagonist attenuate muscle loss in some diseases [170, 171]. The mechanisms of glucocorticoid-mediated muscle atrophy were recently unraveled. Once in the nucleus, the glucocorticoid receptor activates expression of two target genes, encoding REDD1 (regulated in development and DNA damage responses 1) and KLF15 (Krüppel-like factor 15) [172]. REDD1 inhibits mTOR activity by sequestering 14-3-3 and increasing TSC1/2 activity. Inhibition of mTOR is permissive for activation of an atrophy program via KLF15. Indeed, mTOR activation attenuates glucocorticoid-induced muscle atrophy. KLF15 is a transcription factor that is involved in several metabolic processes in skeletal muscle, for instance up-regulation of branched-chain aminotransferase 2. KLF15 participates in muscle catabolism via transcriptional regulation of FoxO1, atrogin-1 and MuRF1. Moreover, KLF15 negatively affects mTOR through up-regulation of branched-chain aminotransferase 2, which in turn induces branched-chain amino acid degradation. Interestingly, FoxO1 and glucocorticoid receptor cooperate to up-regulate MuRF1 expression [173].

Proteolysis-dependent regulation of protein synthesis

Synthesis and degradation of proteins are two processes that are intimately connected. Indeed, most of the above-mentioned pathways concomitantly regulate both synthesis and degradation, such that when protein synthesis is induced, degradation is suppressed and vice versa. However, this control appears to be a compensatory mechanism to limit energy expenditure for the production of novel proteins under catabolic conditions. As mentioned above, in denervated muscles, net protein synthesis is increased rather than decreased compared to innervated muscles [8]. This is because a proportion of the amino acids released from protein breakdown stimulate protein synthesis via mTOR, and, if this mechanism is blocked, muscle loss is exacerbated [8]. The direct action of amino acids on translation plays an important role in the rewiring of protein synthesis during catabolic conditions, changing the metabolism and expression of sarcomeric proteins in order to optimize muscle homeostasis and performance. An important example of amino acid-dependent regulation of gene transcription during a catabolic state has recently been described [174] for lysosomal-dependent protein degradation. Nutrients, especially free amino acids, are sensed by the mTOR kinase, which then inhibits autophagy by blocking formation of the Atg1/unc-51-like kinase 1 complex, an important regulatory step for autophagy initiation. The mTORC1 complex is therefore at the center of a variety of cellular process such as protein synthesis, autophagy, aging, mitochondrial function and energy production. These various actions of mTORC1 are exploited by its localization/recruitment to various cellular compartments. For instance, the Rag GTPase complex, which senses lysosomal amino acids, promotes localization of mTORC1 to the lysosomal surface. Accumulation of amino acids within the lysosomal lumen generates an activating signal that is transmitted to the Rag GTPases via vacuolar H+-adenosine triphosphatase ATPase (v-ATPase), recruiting mTORC1 to the lysosomes. This mTOR localization initiates amino acid signaling and protein synthesis [174] (Fig. 3). Concomitantly, mTOR also inhibits transcription factor EB, a master regulator of lysosome biogenesis [175]. Activation of mTORC1 induces phosphorylation and localization of transcription factor EB at the lysosomal membrane, thus inhibiting its transcriptional activity [176]. These data indicate that the content/activity of the lysosome directly regulates lysosome biogenesis via an mTOR–transcription factor EB axis. The implication of this signaling as it relates specifically to muscle homeostasis has yet to be investigated.

Conclusions and perspectives

Our understanding of the mechanisms that control muscle growth and atrophy has greatly advanced during the last ten years. Major milestones in this progress have been identification of the transduction pathways that mediate myostatin and IGF1 signals, in particular the crucial role of Akt and its main downstream effectors, mTOR, which controls protein synthesis, and FoxO3, which controls protein degradation via the proteasomal and autophagic/lysosomal systems. In addition, other pathways, such as SRF, have emerged as potential players in regulation of muscle fiber size, and the contribution of satellite cells has been the object of intensive investigation. It is clear that there is no common mechanism that applies to all models of muscle growth or wasting, and an important objective for future studies will be to define the pathways that are operative in the various situations. For example, autophagy is highly up-regulated in skeletal muscles during starvation, but its induction after denervation, when the ubiquitin–proteasome pathway is strongly activated, is much less important. Satellite cell activation may likewise vary in different models of muscle hypertrophy, and its contribution to the increase in muscle fiber size remains to be established.

Dissection of the pathways that control muscle mass and function will provide useful indications for the development of drugs that are able to boost muscle growth and prevent muscle wasting, a target that is now being actively pursued in both academic and biotechnology/pharmaceutical research, and may have great therapeutic importance for treatment of neuromuscular diseases, systemic disorders, muscle disuse and aging.


Original work reported here is supported by the EC FP7 Project MYOAGE (grant number 223576 to S.S. and M.S.) and the European Research Council (grant number 282310-MyoPHAGY to M.S.).