Mechanisms of protein degradation in atrophying muscles: What have we learned during the past decade?

Abstract

Skeletal muscle atrophy occurs in diverse conditions, including aging, disuse, cancer cachexia, and chronic disease. It results from an imbalance between protein synthesis and degradation, where excessive proteolysis drives loss of contractile proteins, weakness, and metabolic decline. Recent advances in structural biology, multi-omics approaches, and high-resolution imaging have uncovered how sarcomeric and cytoskeletal components are gradually degraded by ubiquitin ligases, proteasomes, and autophagy. Mechanical loading and mechanotransduction emerge as key regulators of proteostasis, linking tension to anabolic signaling. Transcriptional and epigenetic control through IGF1-Akt-mTOR, TGF-β, inflammatory cytokines, and circadian rhythms, as well as non-coding RNAs and miRNAs, also contribute to wasting. This review summarizes these recent findings and novel therapeutic strategies, such as restoring mitochondrial function and modulating RNA networks and mechanosensitive signaling to preserve muscle mass and function.

Skeletal muscle is critical for overall health as it supports movement and posture and is the major site for glucose disposal. Preservation of muscle mass and strength is essential for longevity and quality of life (1, 2) and influences not only metabolic health but also bone integrity and repair (3). During muscle wasting, there is a significant reduction in muscle mass and strength primarily due to the loss of the contractile myofibrilar proteins, which comprise more than 70% of muscle proteins. Their reduced production and destruction cause frailty, fatigue, and, if prolonged, disability and mortality.

This loss of muscle mass and strength (namely atrophy) is a hallmark of various physiological and pathological conditions, including aging, immobilization, cancer cachexia, and chronic diseases. The breakdown of the contractile apparatus is a highly ordered and tightly regulated process that enables muscle contraction to continue even during rapid wasting, such as during starvation or cancer. Recent advances in molecular biology and super-resolution imaging approaches have substantially advanced our understanding of these proteolytic mechanisms, facilitating the development of rationally designed therapeutics to combat muscle wasting.

Sarcomere architecture and myofibril breakdown

The sarcomere, the fundamental contractile unit of muscle, is comprised of highly ordered, precisely aligned arrays of myosin in thick filaments and actin in thin filaments, delimited by Z-lines (4). Cryo-EM imaging by the Raunser group recently uncovered the molecular architecture of Z-lines, where thin filaments and the desmin intermediate filaments are crosslinked and stabilized by the bona fide Z-line protein α-actinin (4). Degradation of this intricate structure of the myofibrils must be gradual and tightly regulated to ensure that muscles continue to contract even during rapid wasting (e.g. during starvation or cancer). During the past decade, the molecular mechanisms for the gradual loss of proteins from the myofibril became clearer, and distinct ubiquitin ligases play critical roles. In atrophy caused by nerve sectioning (i.e. denervation), the breakdown of sarcomeric proteins follows a defined sequence. Initially, small regulatory proteins that stabilize the thick filaments, like myosin binding protein C (MyBP-C), are targeted for degradation primarily by the ubiquitin ligase MuRF1. This process likely enhances the susceptibility of the more loosely bound myofibrils to the catalytic activity of ubiquitin ligases and the proteasome because at a more delayed phase, myosin heavy chain undergoes a similar fate (5, 6). Degradation of actin and other thin filament components, such as myosin light chains and tropomyosin, is associated with the initial disassembly of the Z-lines and the desmin cytoskeleton, and seems to require another ubiquitin ligase, TRIM32 (7, 8), although MuRF1 was also shown to play a role (9, 10). This loss of desmin filaments requires phosphorylation by GSK3-β, ubiquitylation by the ubiquitin ligase TRIM32, and disassembly by the coordinated actions of the calcium-specific protease calpain-1 and the AAA-ATPase ATAD1 (11, 12, 13, 14, 15). Recent evidence indicates that MyBP-C normally bridges and stabilizes both thick and thin filaments (16), suggesting that the initial loss of MyBP-C may also reduce the stability of thin filaments during myofibril breakdown. Therefore, inhibition of GSK3-β, the critical kinase responsible for desmin phosphorylation could be an attractive approach to combat atrophy by blocking the initial disassembly of the desmin cytoskeleton and the attached myofibrils. In fact, muscle-specific downregulation of GSK3-β reduces atrophy during fasting (11) or spaceflight (17), and pharmacological inhibition of this kinase improves muscle function in a mouse model of Duchenne muscular dystrophy (DMD) (18).

Mechanical forces imposed on muscle by exercise are known to halt this excessive proteolysis and override various types of atrophy (19). The Irving group have significantly advanced our understanding of muscle structure-function relationships by demonstrating that force generation is regulated not only by calcium-dependent thin filament activation but also via mechanosensing by thick filaments. Using high-resolution X-ray diffraction and fluorescence polarization techniques in skeletal muscle fibers, they demonstrated that myosin heads adopt a helical OFF state at rest and transition into an active configuration in response to mechanical strain, supporting a dual-filament regulation model (20, 21, 22). This model transforms the way we think about how muscle contraction is activated, and highlights the role of thick filaments in sensing and responding to load. In addition, the structural dynamics of the sarcomere are tightly coupled to calcium signaling and are altered in disease, including cardiomyopathy (23, 24). These structural insights can explain how muscle adapts to use and disuse, with direct implications for atrophy, where mechanical and molecular disruptions to regulation of thick filament may promote impaired force production and fiber degeneration. Consistently, our group and others (5, 6, 25) have demonstrated that thick filaments are selectively lost early in atrophy, even before thin filament loss. Their selective loss (and the loss of stabilizing filaments like the desmin cytoskeleton, see below) represents an initiating event driving overall proteolysis rather than being a downstream cosnequence of muscle wasting. This implicates defects in mechanotransduction early during atrophy and highlights a major clinical need for therapeutic strategies that preserve sarcomere integrity, particularly for bedridden or ventilated patients where atrophy is rapid, and exercise is not feasible.

Molecular drivers of proteolysis in atrophy

A set of common transcriptional adaptations drives proteolysis and muscle atrophy in diverse catabolic conditions (15, 19). Analysis of aging muscle transcriptomes using real-time PCR revealed increased stochasticity (i.e., greater transcriptomic randomness) with age, which plateaued in very old mice (27–29 months). This entropy analysis uncovered the membrane protein Klotho as a critical regulator of muscle mass. Low levels of Klotho during aging are associated with muscle atrophy and weakness, and its accumulation specifically in the liver by systemic delivery (via tail vein) using AAV vectors improved muscle structure and function in old mice (21–24 months) but failed to do so in the oldest cohort (26). These findings suggest a therapeutic window for Klotho early during aging and before transcriptomic disorganization sets in. In another spatiotemporal transcriptomic analysis coupled with immunofluorescence of innervated, denervated and re-innervated muscles, perturbations of polyamine pathway were found associated with atrophy of glycolytic fibers (27). A consecutive study by the same group demonstrated that polyamine metabolism is also perturbed in amyotrophic lateral sclerosis (ALS), and perturbation of this pathway in muscle is sufficient to recapitulate the pathological manifestations of ALS (28).

A hallmark of muscle atrophy is the induction of genes that promote proteolysis, including ubiquitin proteasome system (UPS) components. Various ubiquitin ligases promote protein ubiquitylation and degradation during atrophy, and their expression is sufficient to trigger wasting, though their induction per se cannot be a direct measure for protein degradation (29). The relative contributions of the numerous ubiquitin ligases that are activated during atrophy likely differ, with some acting primarily to initiate myofibril disassembly, while others function downstream to amplify proteolysis once structural integrity is compromised. Among these enzymes are MuRF1 (TRIM63) and MAFbx/atrogin-1 (FBXO32) (30, 31), the F-box protein FBXL22 (32), and ASB2β (ASB2) (33), which facilitate loss of muscle mass via distinct mechanisms that ultimately activate myofibril breakdown. Disassembly and degradation of myofibrillar proteins often follow breakdown of the desmin cytoskeleton, which is activated by GSK3β-dependent phosphorylation (7, 11, 13, 14, 34, 35, 36, 37, 38). Then, various ubiquitin ligases, including MuRF1 (5, 6, 10, 39), Trim32 (7, 8, 40), and UBR2 (41), and E2-conjugating enzymes like UBE2L3 (9) act on the major myofibrillar components myosin and actin and promote their ubiquitylation. Post-translational modifications like acetylation can protect myosin from degradation because deacetylation of myosin by HDAC4 appears essential for degradation (42), though it remains unclear whether acetylation affects myosin ubiquitylation. Subsequent extraction of the ubiquitylated components from the myofibrils is primarily mediated by the AAA-ATPase p97/VCP and is required to enhance susceptibility of the solubilized myofibrillar proteins to the catalytic core of the proteasome (19, 43, 44). Interestingly, activation of p97/VCP in vivo is enhanced by the skeletal muscle-specific Methyltransferase, METTL21C, which trimethylates p97/VCP on Lys315 residue and consequently facilitates p97/VCP hexamer formation and ATPase activity (45). Ubiquitin-independent degradation by the 20S proteasome has also been shown to target intrinsically disordered and oxidized proteins. In particular, the 20S proteasome subunit PSMA3 can selectively bind unstructured proteins and facilitate their degradation in a ubiquitin-independent manner (46). However, it remains unclear whether this mechanism applies to muscle atrophy or contributes meaningfully to the catabolic process.

In addition to ubiquitylation, protein UFMylation have been recently associated with skeletal muscle atrophy in a mouse model of ALS, contributing to reduced strength (47). Simultaneously, the expression of certain factors increase in muscle during atrophy most likely as a compensatory mechanism to protect muscle from the induced catabolic program. For example, the protein macroautophagy and youth optimizer (MYTHO) is essential for maintenance of normal muscle mass as it functions as a scaffold for autophagosome formation and enhances autophagic flux (48, 49). Thus, its induction in aged atrophying muscle likely represents a protective response; accordingly, MYTHO deficiency in c. elegans has been shown to reduce survival upon exposure to oxidative stress (49).

Proteolysis early in atrophy largely increases through the FOXO3-mediated expression of the atrophy program. This transcription factor has long been recognized as the main transcription factor that induces atrophy-related genes, and its activation is sufficient to cause wasting (50). Recent findings have uncovered additional, coordinated roles for multiple transcription factors including (e.g. PAX4, NRF1/αPAL), especially in the induction of proteasome subunit genes and assembly chaperons (51). These factors stimulate proteasome expression, assembly and activation during the later stages of atrophy, by distinct regulatory proteins, such as ZFAND5/ZNF216 (52). Other transcription factors that have been reported to induce atrophy-related genes and promote wasting are Kruppel-like factor (KLF), which is activated by TGF-β in pancreatic cancer (53) and Forkhead box P1 (FoxP1) (54). In addition, muscle-specific deletion of Kelch-like ECH-associated Protein 1 (Keap1), a negative regulator of the transcription factor NRF2, improves force production but does not block atrophy upon hindlimb unloading (55) (Fig. 1). These observations argue against the existence of a single master regulator of proteasome subunits gene induction in muscle. Instead, proteasome biogenesis appears to be governed by coordinated and temporally distinct transcriptional programs, whose relative importance may vary with the duration and nature of the catabolic stimulus.

Figure 1.

Transcriptional control of muscle homeostasis and atrophy. Under homeostasis, mTOR activates the transcription factors NRF2 and HIF to maintain redox balance and support high intracellular Fe²⁺ levels, thereby preserving muscle mass. By contrast, during atrophy, a distinct set of transcription factors, including PAX4, NRF1-αPAL, and FOXO family members, drive the expression of atrogenes and proteasome subunits, leading to enhanced protein degradation. FOXO-dependent catabolic transcription is further reinforced by FoxP1 and facilitated by the proteasome regulator ZFAND5, while PRMT5 restrains FOXO activity to limit wasting. Low intracellular Fe²⁺ levels also contribute to the atrophy program. These transcriptional regulators orchestrate the shift from muscle homeostasis to muscle atrophy by controlling redox status, iron handling, and proteolytic gene expression.

The rapid loss of muscle mass and strength is often accompanied by a decrease in rates of protein synthesis, especially in humans (56). Studies using SUnSET-based proteomics and phosphoproteomic analyses of immobilized mouse skeletal muscles concluded that this reduction in protein synthesis is mediated by a decrease in translational efficiency (57). In times of scarcity (limited nutrient availability, as in fasting), the reduced rates of protein synthesis result primarily from inhibition of the major growth pathway in muscle, insulin-PI3K-AKT-mTOR signaling. In proliferating cells, this pathway drives cell division, but in mature, non-dividing muscle cells, it stimulates protein synthesis and prevents protein breakdown. Recent studies in yeast models have demonstrated that mTOR1 phosphorylates and inhibits the ribosome preservation factor Stm1 to activate dormant ribosomes. Under nitrogen starvation, ribosomes are selectively degraded, with a small pool of dormant 80S ribosomes preserved via Stm1-dependent mechanism (58). By contrast, nutrient replenishment activates Stm1 phosphorylation by mTOR and causes rapid reactivation of ribosomes (58). To date, it remains unclear if this regulatory mechanism applies to fasting-induced muscle wasting.

Signaling pathways regulating atrophy

Insulin-PI3K-AKT signaling and its regulatory cross-talk

Muscle mass is tightly regulated by the balance between anabolic and catabolic signaling, and the PI3K-AKT-mTOR pathway plays a central role. Activation of this pathway promotes protein synthesis and muscle growth, whereas its inhibition leads to atrophy (59, 60). Multiple components of the UPS can suppress PI3K-AKT-mTOR signaling and inhibiting them could be of value to combat muscle wasting. These include various ubiquitin ligases such as Trim72 (61), Trim32 (40), and CHIP (62)), E2-conjugating enzymes like Ube2H (61), and deubiquitylating enzymes such as USP1 (63) and USP19 (64).

Recent studies uncovered cross-talks between insulin-PI3K-AKT pathway and other signaling cascades, including Wnt, TAK1, PRMT5, and TGF-β (Fig. 2). For example, the extracellular ligand Wnt7a that signals through its receptor Fzd7 activates AKT-mTOR signaling and blocks cachexia in C26-colon tumor-bearing mice (65, 66) and is also required for muscle regeneration (via satellite cells expansion) in mdx mouse model for DMD (67). Whether Wnt7 signaling is inhibited under catabolic states remains an open question. Despite its anabolic role, chronic activation of mTOR can paradoxically enhance proteolysis, promote degeneration of neuromuscular endplate and accelerate atrophy through a negative feedback loop that inhibits AKT activity (68, 69). Accordingly, recent transcriptomic profiling of young and aged muscles established an mTOR-dependent gene atlas demonstrating that fine-tuning of mTOR activity is essential for muscle homeostasis (70). Pharmacological inhibition of mTORC1 by rapamycin significantly alleviates age-related sarcopenia in rodents, whereas sustained mTORC1 activation in muscle fibers is sufficient to induce molecular signatures of sarcopenia (70, 71).

Figure 2.

Signaling pathways regulating muscle atrophy in catabolic states. In diverse catabolic conditions, multiple extracellular cues converge on a limited set of intracellular pathways that inhibit anabolic pathways and induce expression of atrophy-related genes and stimulate protein degradation via the UPS and autophagy. Myostatin and activin A activate ActRII-SMAD2/3 signaling to promote catabolic transcriptional programs, whereas BMP ligands can either drive atrophy through BMPR–SMAD1/5 or support muscle maintenance through MuSK-dependent activation of PI3K-AKT signaling. Wnt7 also enhances PI3K-AKT signaling to preserve muscle mass. By contrast, reduced insulin/IGF-I signaling through InsR/IGFR decreases PI3K-AKT-mTOR activity, releasing inhibition of FOXO transcription factors and enabling the induction of genes that promote proteolysis. Inflammatory cytokines acting via Fn14 further suppress PI3K-AKT-mTOR signaling to promote wasting, while EDA2R activates NF-κB to stimulate the atrophy program. These pathways signal through FOXO, SMAD and NF-κB transcriptional effectors to coordinate the catabolic responses that underlie skeletal muscle atrophy.

The TNF receptor superfamily member fibroblast growth factor-inducible 14 (Fn14) is another regulator of AKT activity (Fig. 2). Mice lacking Fn14 exhibit increased AKT and FOXO3 phosphorylation, reduced UPS gene expression, and attenuated atrophy on denervation (72). Similarly, inducible deletion of the insulin and IGF-1 receptors in muscle causes insulin resistance and atrophy (73, 74), driven by FOXO-dependent reduction in mitochondrial respiration, complex I core subunit expression, and ATP production (75). The resulting impaired glucose uptake further compromises recovery from atrophy, as glucose transport is required for muscle homeostasis and regeneration (76).

The ubiquitin ligase UBR5 (and other members of this family) is another critical regulator of muscle mass. This enzyme promotes recovery following inactivity (77), and its loss reduces protein synthesis and activates proteolysis and atrophy (78). Additional anabolic regulators include the desmosomal component plakoglobin, which stabilizes the insulin receptor and the dystrophin glycoprotein complex (DGC) on the muscle membrane (73), and is reduced early during atrophy, and the methyltransferase PRMT5, which methylates and destabilizes FoxO1 to suppresses autophagy (79). The kinase transforming growth factor-β activated kinase 1 (TAK1) also maintains muscle mass by promoting mitochondrial integrity and redox balance (80).

Dual roles of BMP and TGF-β signaling in muscle atrophy

In addition to insulin signaling, accumulating evidence indicates the importance of bone morphogenetic protein (BMP) signaling for maintenance of normal muscle mass (Fig. 2). BMP activation preserves neuromuscular junction integrity and mitigates cancer-associated atrophy (81). The key kinase in this pathway is MuSK, which helps maintain insulin-mTOR signaling and the mass of preferentially slow-twitch muscle fibers (82). Canonical BMP receptor (BMPR) and Activin Receptor II (ActRII) signaling exert opposing effects on skeletal muscle (1, 81). However, recent studies demonstrate that BMPR signaling can also promote wasting depending on the downstream SMAD effectors. Both ActRII and BMPR ligands can induce atrophy and inhibit human skeletal muscle differentiation via SMAD2/3 or SMAD1/5, respectively (1, 83). A neutralizing antibody against ActRIIA/B blocks both effects, suggesting that some BMPs also act through ActRII in muscle. In mice, overexpression of BMP9 specifically in the liver causes systemic wasting and liver toxicity, while local accumulation of BMP7/9 promotes ectopic bone formation (83). Elevated BMP-SMAD1/5 signaling was also reported in sarcopenic rat muscle, further suggesting that this pathway can contribute to age-related wasting and may thus represent a therapeutic target alongside ActRII blockade.

EDA2R-NIK signaling emerges as novel catabolic pathway promoting cachexia

The Ectodysplasin A2 receptor (EDA2R), a member of the TNF receptor superfamily, activates NIK signaling and is a novel pathway driving cancer cachexia primarily through NF-κB (84) (Fig. 2). EDA2R is induced in muscles of tumor-bearing mice and patients with cachectic cancer, as well as during aging and following denervation, where it promotes the expression of atrophy-related genes (85). Muscle-specific deletion of EDA2R or NIK protects tumor-bearing mice from systemic wasting, identifying EDA2R as a key upstream regulator of NF-κB in muscle during catabolic states. The upstream catabolic signals that induce EDA2R expression during wasting are still unknown. It is also unclear whether the attenuation in muscle atrophy upon EDA2R depletion is due to the activation of anabolic pathways such as PI3K-AKT-mTOR or BMP signaling. In addition, whether EDA2R signaling cooperates with other catabolic pathways to induce wasting, such as myostatin-ActRII or HSP90-STAT3-FOXO1 signaling (86), is another important question for future research.

Circadian inputs promote proteolysis in muscle atrophy

Circadian control adds another layer of regulation. The clock transcription factor BMAL1 maintains muscle mass and function, and its downregulation in obesity (87) leads to glucose intolerance and ultimately atrophy (87, 88). In addition, the clock-controlled ubiquitin ligase FBXL21 targets the sarcomere component TCAP for degradation (89), and loss of TCAP induces mitophagy, dystrophy, and cardiomyopathy (90). Degradation of TCAP by FBXL21 requires phosphorylation by GSK-3β, a kinase activated by the fall in insulin-AKT-FOXO signaling and promotes overall proteolysis by facilitating desmin filament breakdown (discussed above) (11, 12, 13). These various studies and recent findings suggest that diverse signaling inputs, anabolic, metabolic, and circadian, all converge on a limited set of transcriptional effectors that control muscle proteostasis, many of them acting through or cooperating with FOXO transcription factors. Thus, the various catabolic cues may differ less in their ultimate cellualr outcome than in the timing, rate, context, and extent of their activation. Understanding how these transcriptional programs integrate to drive or restrain muscle atrophy will be essential for developing new targeted therapies to preserve muscle mass.

Metabolic and mitochondrial dysfunction in atrophy

Mitochondria are essential for energy production and cellular homeostasis, and their dysfunction is a prominent feature of muscle atrophy. During disuse or denervation, mitochondrial content and oxidative capacity decline, accompanied by alterations in morphology and network dynamics. Using ³¹P magnetic resonance spectroscopy and blood plasma metabolomics, recent evidence suggests that impaired mitochondrial oxidative capacity precedes the loss of muscle mass and strength (91, 92). Degradation of mitochondrial proteins is tightly controlled by matrix proteases such as LONP1, which is induced under catabolic conditions. Loss of LONP and the resulting impaired mitochondrial protein turnover are sufficient to promote mitochondrial dysfunction and reduction in muscle mass and strength (93). Mitochondrial nicotinamide adenine dinucleotide (NAD⁺) levels also decline during atrophy, and NAD⁺ repletion has been shown to suppress proteasomal activity and FOXO signaling (Table 1 and see below).

Table 1.

Mitochondrial regulation and metabolic dysfunction in muscle Atrophy

Regulator	Function	Levels during atrophy	Consequence	Reference
LONP1	Matrix protease, protein quality control	↑ Expression under catabolic stress	Mitochondrial protein degradation, dysfunction	(93)
NDUFA4L2	Complex I subunit (low oxygen adaptation)	Unclear	Ectopic expression reduces ATP/NAD+, and activates apoptosis	(99)
MCUR1	Mitochondrial Ca2+ uptake	↓ Expression	↓ Respiration, energy metabolism	(97)
DRP1	Mitochondrial fission	Unclear	Deficiency leads to fragmented mitochondria and fiber death	(96)
NAD+	Metabolic cofactor	↓ Levels	↑ Proteolysis via FOXO and the proteasome	(98,99,100)
Fe+2	Metabolic cofactor	↓ Levels	↑ oxidative stress	(94)

Similarly, perturbations in iron metabolism and availability contribute to oxidative stress and muscle wasting, particularly in aging and chronic disease. Iron supplementation is sufficient to preserve muscle mass and contractile capacity in tumor-bearing mice and human patients (94). Thus, maintenance of iron homeostasis is essential for muscle integrity, and appears to depend on mTOR-mediated expression of iron-related genes via the transcription factors NRF2 and HIF (95) (Fig. 1).

Several mitochondrial regulators have emerged as key players in atrophy. The mitochondrial pro-fission dynamin-related protein 1 (DRP1) is required for maintenance of normal muscle mass; its muscle-specific deficiency disrupts mitochondria fission, enhances calcium influx into the mitochondria, and promotes muscle fiber death (96). Age-related muscle atrophy is also associated with reduced mitochondrial calcium uptake, partly due to the downregulation of the mitochondrial calcium uniporter regulator 1 (MCUR1), leading to impaired mitochondrial respiration and mitochondrial dysfunction (97). Systemically, these detrimental effects on mitochondria function during aging may result from the fall in natural alkaloid trigonelline levels in blood plasma, and the resulting decrease in NAD ⁺ pool as was demonstrated in muscles from c. elegans, mice and primary myotubes from aged humans (98). Another regulator of NAD⁺ metabolism and muscle mass is the mitochondrial protein NADH dehydrogenase (ubiquinone) one alpha subcomplex, 4-like 2 (NDUFA4L2). Overexpression of NDUFA4L2 in mouse skeletal muscles using adenovirus leads to low oxygen consumption, reduced energy levels (ATP, NAD⁺), and loss of muscle mass and force production, also via the activation of apoptosis (99). Thus, the reduction in the levels of key intramuscular metabolites, including adenine nucleotides and NAD⁺, are hallmarks of mitochondrial dysfunction that may drive muscle wasting (Table 1). Nevertheless, recent evidence in mice suggests that cellular decline in NAD⁺ levels can occur without adverse effects on muscle mass and strength, mitochondrial function, or overall health (100). In parallel, certain metabolites (e.g. phosphorus) and phospholipids accumulate in aged muscles due to impaired metabolism and are associated with sarcopenic decline (101, 102).

Muscle-secreted factors and inter-organ communication

Pro-inflammatory cytokines and systemic inflammation

Skeletal muscle is increasingly recognized as a secretory organ that communicates with distant tissues through myokines, cytokines, metabolites, and extracellular vesicles (EVs) (103) (Table 2). Acting in autocrine, paracrine, and endocrine fashions, these factors coordinate systemic energy balance and organ crosstalk. Their blood levels change markedly during muscle atrophy and influence both local and systemic homeostasis (104).

Table 2.

Muscle as an endocrine organ

Myokine/Factor	Source	Role in Atrophy	Notes
Myostatin/Activin A	Muscle, kidney	Strong inducer of proteolysis by inhibiting PI3K-AKT signaling	Inhibition reverses atrophy
IL-6/TNF-α/IFNγ	Muscle, immune, brain	Promote NF-κB/TGF-β/JAK-STAT signaling	Elevated in sepsis, cancer cachexia
IL-15	Muscle	Promotes autophagy
FGF21	Muscle	Reduces protein synthesis and enhances autophagy via Bnip3	Elevated in fasting, metabolic disorders
EV-associated miRNAs (e.g., miRs-206/1/208)	Muscle EVs	Protective, partly via CRK inhibition	Muscle regeneration, NMJ integrity, reduced fibrosis in sarcopenia
EV-associated miR-102	Tumor-derived EVs	Suppresses protein O-GlcNAcylation in muscle and promotes wasting	Elevated in cancer cachexia
Fibcd1, FNDC1	Muscle	Protective, partly via ERK	Mitigate cancer-induced diaphragm atrophy

Pro-inflammatory cytokines, including tumor necrosis factor-α (TNFα), interleukin-6 (IL-6), IL-1, and interferon-γ (IFNγ), TNF-like weak inducer of apoptosis (TWEAK), are elevated in sepsis, cancer and other catabolic conditions. These cytokines trigger muscle wasting by binding cell surface receptors (e.g., TNFα receptor, FN14), by activating NF-κB and TGF-β signaling (105), or by stimulating the release of other cytokines (19). In parallel, they suppress anabolic signals, exacerbate mitochondrial dysfunction, and inhibit satellite cell activity. Emerging evidence implicates a brain-muscle signaling axis in the pathogenesis of atrophy. Inflammation within the central nervous system can elicit skeletal muscle fatigue. The accumulation of reactive oxygen species in the brain leads to secretion of IL-6 to the circulation, activation of JAK-STAT pathway in muscle, and subsequent mitochondrial dysfunction and motor impairment (106). In addition to systemic inflammation, local inflammation responses within the muscle tissue have been reported in diverse models of wasting (107, 108).

Myostatin and activin signaling

Among secreted catabolic factors, myostatin, a member of the TGF-β family, is a potent negative regulator of muscle mass. It suppresses muscle growth and promotes wasting, partly by inhibiting insulin-PI3K-AKT signaling (109, 110). Myostatin levels increase in various human diseases, including cancer and during aging, and its inhibition consistently mitigates cachexia in preclinical models. Recent single-cell transcriptomics identified a crosstalk between the kidney and muscle in which activin A, a myostatin homolog is secreted from kidney cells in chronic kidney disease and drives systemic muscle wasting (111).

Anabolic and catabolic myokines

Atrophying muscle shows a marked shift in its myokine profile. Anti-inflammatory or anabolic myokines, such as, Fibcd1 (112), FNDC1 (113), and IGF-1 (114) decline, whereas catabolic myokines, such as FGF21 (115), are induced and act in autocrine and endocrine fashions to modulate metabolism and proteostasis. Fibroblast growth factor 21 (FGF21) secretion from muscle increases in response to mitochondrial stress during fasting or metabolic disorders, leading to reduced protein synthesis and enhanced autophagy via activation of Bnip3 (115, 116). Accordingly, mice lacking FGF21 are protected from fasting-induced atrophy, whereas FGF21 accumulation is sufficient to activate autophagy and trigger muscle wasting (115). In the liver, FGF21 expression rises during ketogenic diet feeding in mice to alleviate lipid overload, and pharmaceutical FGF21 analogues are in clinical trials for the treatment of fatty liver disease (116, 117). Another skeletal muscle-secreted cytokine, interleukin 15 (IL-15), has been implicated in the regulation of muscle size, because transgenic mice overexpressing IL-15 exhibit enhanced autophagy and reduced gastronemius muscle mass (118) (Table 2). However, its precise role in muscle atrophy remains unclear.

Recent efforts to identify additional myokines using Drosophila transgenic screening and mammalian models have revealed that the evolutionarily conserved myokine Fibcd1 can mitigate cancer-induced diaphragm atrophy in mice bearing patient-derived melanoma xenografts and LLC carcinomas, and preserve muscle mass via ERK signaling (112). Interestingly, the lysosomal enzyme, Hexosaminidase A (HexA) is secreted by the liver in fatty liver disease and improves glucose tolerance in muscle (119).

Muscle-derived EVs

EVs released from muscle fibers are emerging as critical mediators of intercellular communication (Table 2). They play key roles in muscle growth (120) and regeneration in muscular dystrophies (121, 122) and aging (123). Muscle-derived EVs, are enriched with miRNAs characteristic of striated muscles, miR-1/miR-133a/miR-206, which promote regenerative signaling. Although their mRNAs and protein cargoes remain largely unknown, mounting evidence supports a therapeutic potential for EVs as natural delivery vehicles.

Profiling of EVs cargo in different mouse models for atrophy using transcriptomics, proteomics, and miRNAs analysis has identified miRNAs with protective effects. For example, miR-1 and miR-208a alleviate atrophy and reduce fibrosis in sarcopenic muscle (124, 125). The canonical muscle miRNAs, miR-1/miR-133a/miR-206, also sustain neuromuscular junction (NMJ) integrity and function (126), which is critical for maintenance of muscle size and health (127). These miRNAs inhibit the adaptor protein CRK post-transcriptionally, leading to activation of the small GTPase RAC1 and maintenance of NMJ function (126). Systemic miRNA profiling of mouse muscles atrophying due to dexamethasone treatment identified numerous differentially expressed miRNAs that target key signaling pathways, including PI3K-AKT-FOXO and MAPK (128). Using miRNA screen and luciferase reporter assay, a recent study further identified 33 (out of 42 tested) miRNAs capable of counteracting atrophy of cultured myotubes, many of which are downregulated in aged human muscles (129). Eighteen of these miRNAs can inhibit Atrogin1/MAFbx expression, and among these miRNAs, miR-376c-3p attenuates atrophy and improves muscle strength in old mice (129).

Exercise also modifies the miRNA content of circulating EVs. High-intensity training in mice elevated miR-133a and miR-133b levels in EVs, and treatment of sedentary mice with these vesicles enhances glucose tolerance, insulin sensitivity, and reduces plasma levels of triglycerides (130). Conversely, tumor-derived EVs that carry miR-122 suppresses protein O-GlcNAcylation in muscle and promotes wasting (131). These findings highlight the remodeling of the muscle secretome during atrophy, consequently affecting not only muscle-intrinsic signaling but also systemic physiology. Restoring anabolic myokine output or inhibiting catabolic and inflammatory cues that drive muscle wasting could offer great therapeutic potential.

Non-coding RNA in muscle wasting

Long non-coding RNAs (lncRNAs), widely denoted as at least 200 nucleotide transcripts with no protein encoding potential, have emerged as critical regulators of skeletal muscle myogenesis, regeneration, atrophy and hypertrophy. They exert their functions by regulating gene expression, acting as molecular sponges that protect target mRNA from miRNA-dependent degradation, or serving as scaffolds for nuclear bodies to drive droplet formation via liquid–liquid phase separation (LLPS) (132, 133, 134). An atlas of skeletal muscle-expressed lncRNAs identified fiber-type specific expression, subcellular localization, and association with pathways disrupted during muscle atrophy. For example, the lncRNA Pvt1 promotes atrophy of primarily slow-twitch muscle fibers because its downregulation protects these fibers from denervation-induced wasting (135). lncRNA-MEG3 is another key factor that is preferentially expressed in slow-twitch muscle fibers and is critical for muscle development and homeostasis. Its deficiency causes muscle wasting, mitochondrial dysfunction, and impaired regenerative capacity, whereas its overexpression enhances oxidative fiber composition, endurance, and muscle mass (133). lncRNA-MEG3 accumulation in mdx mice significantly alleviates muscle wasting and fat infiltration. By binding to polycomb repressive complex two subunit (Suz12), lncRNA-MEG3 regulates LLPS of Suz12 to maintain histone h3 lysine 27 tri-methylation (H3K27me3) modifications of key muscle-specific genes, such as Fhl3 and Rnf128 (133). Another lncRNA that is specifically enriched in skeletal muscle is lnc-mg, whose muscle-specific conditional deficiency in transgenic mice causes muscle atrophy and reduces endurance during exercise (132). Conversely, skeletal muscle-specific overexpression of lnc-mg promotes muscle hypertrophy. Other lncRNAs, such as lncRNA-MUMA and SMUL, attenuate muscle wasting by inhibiting atrophy-related miRNAs or by stabilizing anti-catabolic proteins such as SMURF2. The lncRNA SMUL promotes skeletal muscle atrophy by suppressing SMURF2 expression through nonsense-mediated mRNA decay, consequently stimulating TGF-β/SMAD pathway. Notably, SMURF2 reduces muscle atrophy and promotes a switch from fast-to slow-twitch fibers (136), highlighting SMUL as a potential therapeutic target in muscle wasting. Another lncRNA named mechanical unloading-induced muscle atrophy-related lncRNA (lncMUMA) is enriched in muscle and decreases in expression during muscle atrophy-induced by hindlimb suspension. Its accumulation may represent an attractive therapeutic strategy because it enhances differentiation and blocks atrophy by inhibiting miR-762 function (137).

Recent RNA-seq analysis identified 33 known lncRNAs and 18 novel lncRNAs that are commonly expressed in various mouse models for atrophy (138). The expression of lncRNA H19 and its encoded miR-675 are significantly reduced in unloaded soleus and gastrocnemius muscles, probably as a protective mechanism because the overexpression of lncRNA H19 is sufficient to inhibit IGFI-PI3K-AKT signaling and induce Atrogin-1 and atrophy via miR-675 (139). The lncRNA NEAT1 promotes glucocorticoids-induced wasting by increasing FoxO1 expression and subsequent protein degradation (140). Recent findings reveal that certain lncRNAs harbor translated small open reading frames (sORFs) with coding potential. These sORFs can produce functional micropeptides, hence expanding the regulatory landscape of the non-coding transcriptome in muscle homeostasis (141).

Structural and mechanical regulation of muscle size

The structural integrity of skeletal muscle is essential for its function and resistance to atrophy. A central structural module is the DGC, which connects the cytoskeleton to the extracellular matrix and integrates mechanical and metabolic cues, including insulin signaling (142, 143). The association of the insulin receptor with the DGC is endowed by the desmosomal component plakoglobin (γ-catenin) and appears critical for insulin signaling activity and muscle homeostasis. Consistently, reduced DGC stability is a hallmark of various muscle wasting conditions, including type-2 diabetes (73), cancer (144), prolonged bed rest (145), and DMD (73). Moreover, phosphorylation of dystrophin at serine 3059 has been shown to stabilize dystrophin and β-dystroglycan association, attenuate atrophy on denervation (146), alleviate cachexia and prolong survival in colon-26 tumor bearing mice (147), and protect cultured myotubes from inflammation-induced atrophy (148).

In addition to its structural role and function as a signaling hub, the DGC acts also as a mechano-sensing complex. Mechanical unloading, whether due to immobilization, spinal cord injury, bed rest or spaceflight, disrupts DGC signaling and accelerates proteolysis. Not in vein the heart, which is continuously contracting, is relatively protected from atrophy compared with skeletal muscle. In fact, mechanical stimulation can slow disuse atrophy in intensive care unit (ICU) models by suppressing the expression of genes that promote proteolysis, including autophagy genes and the ubiquitin ligases Fbxo31 and SMART (149). These findings suggest that activating mechano-transduction pathways (e.g., by reinforcing DGC stability) could represent promising therapeutic strategies to promote recovery after prolonged inactivity. Mechanotransduction pathways also include genes such as Arrestin domain containing two and 3 (Arrdc2/3), which are downregulated during disuse atrophy and are implicated in membrane trafficking and adaptation to mechanical load (150). Similarly, phosphorylation of DGC components, such as dystrophin at serine 3059, enhances complex stability and mitigates muscle wasting in models of cachexia and denervation (147, 148, 148). Mechanical forces seem coupled to molecular regulation of contractility because thick filaments can sense strain and undergo load-dependent conformational changes in response to mechanical strain (20, 21, 22), which is critical for maintenance of muscle size. However, although physical activity remains one of the most effective countermeasures to prevent muscle atrophy, it is not feasible for bed-ridden patients and the frail elderly. Therefore, mimicking mechanical signaling pharmacologically or through gene therapy may offer an alternative strategy to preserve muscle mass and function in clinical contexts when mobility is limited.

Emerging therapeutic approaches

Recent preclinical studies have identified several promising strategies to counteract muscle wasting, ranging from genetic interventions to pharmacological and tissue engineering approaches. Glucocorticoids, for example, are known to induce atrophy during fasting (19), and also after denervation (151). Chronic daily administration causes marked atrophy and even blunts the protective effect of myostatin inhibition on dystrophic muscles (152). However, recent findings challenge this traditional view and show that intermittent glucocorticoid therapy (weekly dosing of prednisone) can shift muscle metabolism from catabolic to anabolic. In models of DMD (153), obesity (154) or aging (155), this regimen enhanced mitochondrial activity, reduced atrophy, and activated PGC1α (155).

Another promising approach involves bromodomain and extra-terminal domain (BET) protein inhibitors, such as those targeting BRD4. These compounds protect C26-tumor-bearing mice from cachexia by suppressing STAT3 and FOXO3 signaling, improving survival even without reducing tumor size (2). Gene-based therapies have also gained attention. Overexpression of PRDM16, ERRγ, and LMCD1 promote oxidative metabolism and satellite cell activation. In mice, transgenic expressing of PRDM16 (a PR-domain-containing protein) or estrogen-related receptor gamma (ERRγ) reduces fat infiltration and improves muscle function following injury (156, 157), in part by reducing myostatin expression (157). Similarly, transient expression of the LIM and cysteine-rich domains 1 (LMCD1) in mouse gastrocnemius muscle using adenovirus enhances muscle strength, whereas its silencing reduces muscle force without causing atrophy (158).

Progress in tissue engineering has led to the development of pre-innervated muscle organoids that, when implemented into a rat model for Volumetric muscle loss, showed beneficial effects on muscle mass, satellite cell density, and NMJ integrity (159). Mitochondrial transplantation has also shown promise in attenuating fibrosis, fat infiltration and muscle wasting in preclinical Sprague Dawley rat models of rotator cuff injury (160), although the translational potential remains to be fully established.

At the molecular level, single-cell RNA sequencing recently identified musculoskeletal embryonic nuclear protein 1 (Mustn1), which is secreted from smooth muscle cells and affects ECM deposition during muscle disuse and recovery from atrophy (161). In sequential studies, the same group discovered a newly characterized RNA-binding protein, Zfp697, that regulates miRNA expression to promote muscle recovery following injury or disuse (162). Together, these advances highlight emerging discoveries of distinct mechanistic insights that converge toward clinical translation, and the integration of multi-omic technologies with targeted delivery systems provides a powerful foundation for next-generation therapies.

Conclusion

Muscle atrophy arises from a coordinated breakdown of structural components, impaired proteostasis, mitochondrial dysfunction, and systemic inflammation. Over the past decade, our understanding of these detrimental processes advanced remarkably. Studies integrating multi-omics data with in vivo models have identified key molecular drivers of atrophy, ranging from ubiquitin ligases and non-coding RNAs, to myokines and EVs, that may represent new therapeutic targets and biomarkers for therapeutic interventions. These common pathways that are activated in diverse types of atrophy and promote proteolysis are unlikely to contribute equally to muscle atrophy across all physiological and pathological conditions. Instead, muscle wasting appears to arise from a coordinated and context-dependent interplay between structural destabilization, transcriptional reprogramming, metabolic dysfunction, and systemic cues, whose relative impact evolves over time and likely dictates the rate of atrophy. Recognizing this process as a tightly regulated sequence of cellular events may help reconcile apparent discrepancies in the literature and facilitate the development of more selective time-window-specific therapies, in which distinct molecular factors are targeted at different phases of atrophy. Strategies that stabilize structural components, maintain mechano-transduction, suppress proteolytic pathways, and enhance anabolic signaling hold promise for clinical translation to treat muscle wasting. Future progress will likely depend on personalized approaches, and as mechanistic insights facilitate clinical application, a new era of precision medicine in muscle atrophy is within reach.

Conflict of interest

The author declares that she does not have any conflicts of interest with the content of this article.

Reviewed by members of the JBC Editorial Board. Edited by Henrik Dohlman