Skeletal muscle dysfunction in COPD: miRNAs, myokines and exercise

Abstract

COPD is a multifactorial and heterogeneous disorder, a leading cause of morbidity and mortality worldwide. Not only does its progression compromise lung function, but it is also associated to systemic complications, including skeletal muscle dysfunction. Skeletal muscle dysfunction affects up to 35% of individuals diagnosed with COPD and is marked by muscle atrophy and altered fibre composition, thus resulting in reduced strength, endurance and physical capacity with an increased mortality risk. Multiple factors, including physical inactivity, oxidative stress, chronic inflammation, mitochondrial dysfunction and impaired autophagy, contribute to the development of skeletal muscle dysfunction. Pulmonary rehabilitation, including exercise training, is a key nonpharmacological intervention that mitigates muscle dysfunction by enhancing protein synthesis and promoting beneficial systemic adaptations. These adaptations are mediated by molecular signals such as myokines and microRNAs (miRNAs), regulating inter-organ communication and gene expression relevant to muscle metabolism and homeostasis. Myokines act as messengers between skeletal muscle and other organs, while miRNAs play pivotal roles in muscle remodelling and exercise adaptation. Therefore, the modulation of specific miRNAs may be a promising therapeutic avenue for addressing skeletal muscle dysfunction in COPD. This review explores the interplay between myokines, miRNAs and skeletal muscle dysfunction in COPD and highlights the potential of miRNAs as biomarkers and therapeutic targets in pulmonary rehabilitation.

Shareable abstract

In COPD, skeletal muscle dysfunction is linked to impaired response to physical exercise. In this setting, specific myokines and miRNAs have been identified, and may constitute novel biomarkers and pharmacological targets. https://bit.ly/49ETpFw

Introduction

COPD is a complex, multifactorial and heterogeneous airway disease and one of the most prevalent chronic diseases in the world. It is a leading cause of morbidity and mortality worldwide with a severe impact on health-related quality of life (HRQoL). It poses a substantial economic burden on healthcare systems, and it is predicted to become the third leading cause of death globally by 2030 [1]. Exposure to cigarette smoke and pollutants, in association with genetic predisposition and developmental factors, contribute to the disease aetiology [2]. Chronic cough, sputum production, dyspnoea and reduced exercise tolerance are the main clinical manifestations of COPD, while its main functional hallmark is a non-completely reversible airflow obstruction at spirometry [3].

In addition to its well-known pulmonary manifestations, COPD also exhibits several systemic manifestations. Among these, skeletal muscle dysfunction is particularly significant, as it originates in the early stages of the disease, diminishes physical activity capacities, and has been linked to an increased mortality [4]. Skeletal muscle dysfunction affects up to 35% of the COPD population and is characterised by skeletal muscle atrophy, a shift in muscle fibre phenotype, and reduced muscle strength and endurance [5]. A number of interacting factors, including physical inactivity/disuse, oxidative stress, low-grade chronic systemic inflammation, mitochondrial dysfunction, and autophagy, are involved in the development of skeletal muscle dysfunction in COPD [6].

Pulmonary rehabilitation programmes including exercise training are nonpharmacological interventions that can target skeletal muscle dysfunction by promoting muscle protein synthesis and reducing protein degradation. Exercise training is the core of any pulmonary rehabilitation programme and promotes adaptive responses not only in the muscles, but in various other organs and tissues through inter-tissue cross-talk mediated by specific signalling molecules [7].

Exercise adaptation response involves the modification of gene expression profile followed by changes in the level of molecules associated with skeletal muscle activity and energy metabolism (such as myokines) as well as those involved in their transcriptional regulation, like microRNAs (miRNAs). Myokines, secreted by the exercising muscles, promote cross-talk between muscle cells and other organs like liver, adipose tissue, bone, gut, the vascular bed, and brain. This interaction regulates metabolism, maintains muscle mass homeostasis and promotes overall health [8, 9].

miRNAs are small, noncoding RNA molecules involved in post-transcriptional regulation of gene expression [10]. These molecules play a key role in relevant cell processes including muscle cell proliferation, differentiation, skeletal muscle and vascular remodelling [11]. Additionally, some miRNAs have been linked to aerobic capacity and exercise training response in both healthy and diseased populations [12].

In this context, the possibility to modulate the expression of various miRNAs which regulate the transcription of gene-encoding molecules involved in muscle growth, repair and metabolism, might represent a promising therapeutic approach to mitigate skeletal muscle dysfunction [13].

In this narrative review article, we aim to provide a summary of the current knowledge on the relationship between myokines, miRNAs and skeletal muscle dysfunction in COPD individuals and how physical activity may improve skeletal muscle function. In addition, we highlight the potential of miRNAs as biomarkers or even therapeutic targets in COPD individuals undergoing pulmonary rehabilitation.

Insights on skeletal muscle homeostasis

Peripheral skeletal muscle is a highly specialised and plastic tissue with crucial functions in movement and metabolism. Skeletal muscle plasticity refers to its ability to change its phenotype in response to external stimuli involving its use and workload [14]. Several factors including physical activity, disuse, ageing or diseases have the potential to affect the phenotype and the metabolic properties of skeletal muscle fibres.

Under physiological conditions, the rates of protein synthesis and degradation are balanced processes, and both occur continuously and concomitantly [15]. Skeletal muscle hypertrophy occurs when the rate of protein synthesis exceeds protein breakdown; in this process, the mammalian/mechanistic target of rapamycin (mTOR) pathway is the key regulator of muscle protein synthesis, under the stimulus of insulin-like growth factor (IGF)-1 [16] (figure 1), which enhances protein synthesis and inhibits protein degradation. In contrast, protein degradation is mainly regulated via the ubiquitin-proteasome system [17] and the autophagy-lysosome pathway [18] (figure 1). Additionally, the small ubiquitin-like modifier (SUMO) pathway regulates various aspects of muscle cell functioning including myogenic differentiation, sarcomere assembly, muscle contraction and metabolism [19], while satellite cells representing the resident stem cells in the muscles ensure the high regenerative capacity of skeletal muscle [20]. Once activated, they proliferate into myoblasts and migrate to the injured site where, through the expression of myogenic factors such as myoblast determination protein (MyoD) and myogenic factor 5 (Myf5), enter the cell cycle [21] and differentiate into myotubes which are either incorporated into damaged fibres to repair them or fused with each other to form new myofibres [22] in a process known as myonuclear accretion.

FIGURE 1.

Schematic representation of the main anabolic and catabolic pathways in the skeletal muscles. The figure shows only the most important steps of both metabolic pathways. AKT: protein kinase B; IGF: insulin-like growth factor; JAK: janus kinase; STAT: signal transducer and activator of transcription; mTOR: mammalian target of rapamycin; PI3K: phosphatidylinositol-3 kinase; p70S6K: 70 kDa ribosomal protein S6 kinase; ActRIIB: activin type IIB receptor; atrogin-1: muscle atrophy F-box; MuRF1: muscle-specific ring finger 1; FOXO: forkhead box O; PGC: peroxisome proliferator-activated receptor γ coactivator; PTEN: phosphatase and TENsin homologue; IRS: insulin receptor substrate; HSP: heat shock protein; Smad: small mothers against decapentaplegic.

Physical exercise represents a powerful stimulus to maintain muscle homeostasis through the stimulation of protein synthesis and the improvement of mitochondrial function. However, the effect of exercise on skeletal muscle depends on its characteristics. Endurance exercise, intended as a moderate-to-high intensity physical activity sustained for prolonged periods, stimulates mitochondrial biogenesis by inducing its master regulator peroxisome proliferator-activated receptor-γ coactivator-1α (PGC1α) and, by blocking NF-κB and the transcription factor Forkhead Box Protein O (FoxO)-3, antagonises both proteasomal and lysosomal protein degradation [23]. Endurance exercise training induces adaptation responses in the cardiovascular and musculoskeletal systems improving exercise capacity. Resistance exercise improves strength and muscle mass by enhancing muscle metabolic capacity [24] and by stimulating myofibrillar protein synthesis, through the activation of the phosphoinositide-3-kinase/activated kinase B (PI3K/AKT) pathway and the overproduction of IGF-1. Conversely, through AKT-mediated phosphorylation, resistance exercise inhibits FoxO-1 and -3, thus preventing the transcription of muscle specific ubiquitin ligases genes [25]. Acute exercise refers to a single session of physical activity, while chronic exercise involves repeated sessions over an extended period, typically defined as training lasting ≥8 weeks; both promote the release of myokines capable of influencing the activity and energy metabolism of skeletal muscles, but also of molecules involved in the regulation of their transcription, such as miRNAs [26].

The interplay of miRNAs and myokines in muscle homeostasis

Some miRNAs and myokines are included in feedback loops, where myokines influence miRNA expression, and miRNAs in turn regulate myokine activity, as shown in table 1. This dynamic interplay is crucial for muscle homeostasis, acute response and chronic adaptation to exercise, and response to injury.

TABLE 1.

The main myokines and microRNAs (miRNAs) involved in muscle homeostasis and their interactions

	Actions and targets	Interactions with other myokines/miRNA	References
IL-6	Secreted by skeletal muscle during exercise; anabolic (low levels → satellite cell activation) or catabolic (chronic/high levels → atrophy); activates JAK/STAT3 → ↑ TNF-α, IL-1β; promotes catabolism via NF-κB pathway	Synergises with TNF-α and other pro-inflammatory cytokines to promote muscle catabolism; reduced by irisin through NF-κB inhibition and TLR4 signalling	[27–31]
Irisin	Produced from FNDC5 under PGC-1α control; promotes insulin sensitivity, mitochondrial function, oxidative stress reduction; induces browning of adipose tissue; enhances muscle repair/regeneration; regulates autophagy; inhibits NF-κB, reduces IL-6, TNF-α, IL-1β	Negatively regulated by myostatin (via Fndc5 and PGC-1α inhibition); influences miR-758 and miR-668; regulated by miR-696	[4, 31–37]
Myostatin	Autocrine inhibitor of muscle growth; binds ActRIIB → activates SMAD → ↑ MuRF1, atrogin-1, autophagy genes; suppresses Akt/mTOR; promotes ferroptosis; antagonised by follistatin	Negatively regulates irisin (via PGC-1α/Fndc5); represses miR-486 and miR-29; regulates expression of miR-1, miR-133a/b, miR-206; targeted by miR-27a/b, miR-208a/b, miR-499	[4, 33, 38–45]
IGF-1	Produced mainly in liver (endocrine) and in muscle (autocrine/paracrine); activates PI3K/Akt/mTOR via IRS-1/2; promotes hypertrophy, regeneration, and satellite cell activation; reduced IGF-1 in COPD → ↓ protein synthesis	miR-1 reduces IGF-1; IGF-1 downregulates miR-1 (via FOXO3a); IGF-1 upregulates miR-133 via myogenin, but is negatively regulated by miR-133; miR-486 both promotes and counteracts IGF-1 signalling	[4, 16, 46–49]
IL-15	Growth factor in skeletal muscle; promotes protein synthesis; stimulates myocytes to accumulate contractile proteins; supports muscle growth; affects fibro-adipogenic progenitors	Acts synergistically with IGF-1 to induce myosin heavy chain synthesis	[50–53]
miR-1	Promotes differentiation via HDAC4 inhibition; involved in glucocorticoid-induced atrophy (targets HSP70); reduces IGF-1	Downregulated by IGF-1 (via FOXO3a); induced when myostatin is absent	[38, 46, 54, 55]
miR-133 (a/b)	Promotes proliferation via SRF inhibition	Upregulated by IGF-1/myogenin; negatively regulates IGF-1R and PI3K/Akt	[38, 46, 56, 57]
miR-206	Regulates myogenesis (targets Pola1, promotes MyoD); linked to inflammation in COPD muscle wasting	Induced when myostatin is absent	[38, 54, 58–60]
miR-486	Promotes growth via PTEN inhibition → activates PI3K/Akt; targets FOXO1a; dual role on IGF-1/p85α	Repressed by myostatin; interacts with IGF-1 signalling (both positive and negative regulation)	[38, 47, 61]
miR-499	Regulates proliferation/differentiation by targeting TGF-βR1	Regulates myostatin; targeted by myostatin repression	[39, 62, 63]
miR-29	Negatively regulates IGF-1 and Akt pathway → ↓ protein synthesis	Repressed by myostatin	[39, 59, 64]
miR-23a	Inhibits atrogin-1 and MuRF1 → protective against catabolism		[59, 65]
miR-27a/b	Regulates myostatin	Targets myostatin	[4, 39]
miR-208a/b	Regulates myostatin	Targets myostatin	[39]
miR-696	Regulates irisin expression	Targets irisin	[4]
miR-758, miR-668	Affected by irisin in adipose tissue	Regulated by irisin	[32]

IL: interleukin; IGF: insulin-like growth factor; JAK: janus kinase; STAT: signal transducer and activator of transcription; TNF: tumour necrosis factor; TLR: Toll-like receptor; PGC: peroxisome proliferator-activated receptor γ coactivator; ActRIIB: activin type IIB receptor; SMAD: small mothers against decapentaplegic; MuRF1: muscle-specific ring finger 1; Akt: protein kinase B; mTOR: mechanistic target of rapamycin; PI3K: posphatidylinositol-3 kinase; IRS: insulin receptor substrate; FOXO: forkhead box O; HDAC: histone deacetylase; HSP: heat shock protein; SRF: serum response factor; MyoD: myoblast determination protein; PTEN: phosphatase and TENsin homologue.

In brief, miR-1 reduces IGF-1 activity, but is itself suppressed by IGF-1 through inhibition of the transcription factor forkhead box O (FOXO)3a. During muscle formation, IGF-1 also increases miR-133, which then inhibits both the IGF-1 receptor and the PI3K/AKT pathway [46]. miR-486 promotes PI3K/AKT signalling by blocking its negative regulators PTEN and FOXO1a, but it may also inhibit positive regulators such as IGF-1 and p85α, creating a complex feedback loop [47].

In mice, the loss of myostatin greatly increases miR-1, miR-133a/b and miR-206, all essential for muscle growth and differentiation [38]. Myostatin can suppress protein synthesis by inhibiting miR-486 and miR-29, while its own expression is regulated by miR-27a/b, miR-208a/b and miR-499 [32]. Transcriptomic studies show that irisin exposure alters two key leptin pathway regulators, miR-758 and miR-668, in adipocytes and white adipose tissue [39], and myostatin also reduces irisin expression by inhibiting Fndc5 and its upstream regulator PGC-1α in skeletal muscle [33].

The impact of physical activity on miRNAs and myokines

The expression of miRNAs and myokines is affected by physical activity, as simplified in figure 2, although with different effects depending on factors such as exercise intensity and duration.

FIGURE 2.

Simplified schematic representation of the effects of physical exercise on the expression of the main microRNAs (miRNAs) and myokines and their interrelationship. Depending on the type, duration and intensity of physical exercise, the expression of some miRNAs can increase or decrease. The diagram illustrates how physical exercise influences the expression of myokines and miRNAs and how, through their interaction, it can stimulate adaptive responses involving skeletal muscles, the cardiovascular system, energy machinery and adipose tissue. IGF: insulin-like growth factor; IL: interleukin.

Exercise strongly influences miRNA expression, but profiles vary depending on type, intensity and duration of activity [12]. Once released into the bloodstream, miRNAs act on multiple tissues to meet exercise-induced metabolic demands [66] and can also serve as biomarkers of acute and chronic training adaptations [67]. Key miRNAs, including miR-1, -133a and -133b, are involved in exercise-induced muscle adaptation, linking to gains in strength and endurance [68, 69]. Resistance training upregulates miR-23a and miR-27a, which activate the AKT/FOXO1 pathway to limit muscle loss [70], and by suppressing miR-1 activate PI3K/AKT/mTOR signalling, promoting muscle protein synthesis and improved muscle function [71]. Several studies highlight miRNA expression differences between high and low responders to resistance exercise. For example, miR-378, -29a, -26a and -451 discriminate low versus high responders [72, 73], with 26 miRNAs differing at baseline and 23 after acute or chronic training in vastus lateralis biopsies [74]. Endurance exercise increases miRNAs promoting fibre-type shifts, mitochondrial biogenesis and oxidative metabolism. In mice, miR-494 decreases after aerobic exercise, with increased mitochondrial transcription factor A and FOXO3 expression [75]. In humans, the muscle-specific miRNAs (myomiRNAs) miR-1 and miR-133a rise after acute endurance exercise [76], paralleling myogenic factors MyoD, myogenin and myogenic regulatory factor (MRF)4 [77]. Circulating miRNAs show variable responses: miR-1, -133a and -133b rise after exhaustive exercise, while miR-206 increases only after high-intensity training [78]. After low-volume sprint training, miR-1, -133a and -133b fall, with miR-206 and miR-499 unchanged [79]; similar decreases were seen for miR-1, -133a and -486 after maximal aerobic testing [80]. In contrast, miR-1, -133a, -133b and -206 decreased after 12-week endurance training despite acute exercise-induced rises [76, 81]. In elderly men, Nair et al. [82] found endurance-trained and sedentary subjects showed different exosomal miRNA responses post-exercise: trained men had IGF-1 pathway activation, while sedentary men showed inhibition. Exercise also drives myokine secretion, mediating muscle–organ and intramuscle communication [7, 83]. Myokines regulate glucose uptake, lipid metabolism and energy balance, preventing metabolic disorders. Interleukin (IL)-6 levels rise up to 100-fold with exercise intensity [84–86], promoting glucose metabolism and suppressing tumour necrosis factor (TNF)-α through p38 mitogen-activated protein kinase activation [50]. Irisin levels increase after acute and chronic exercise in both animals and humans [34, 87], with resistance training boosting irisin more than endurance or combined training [88]. IL-15 mRNA and plasma levels rise after resistance exercise [89, 90]. Finally, myostatin decreases after both acute and chronic exercise in healthy and insulin-resistant individuals, while follistatin rises post-exercise and remains elevated during recovery [9].

Skeletal muscle dysfunction in COPD

Experimental animal models of chronic cigarette smoke exposure suggest that the primary alterations contributing to COPD-associated skeletal muscle dysfunction include reduced type I muscle fibre size and ratio, downregulated mitochondrial activity, and elevated oxidative stress levels [91, 92].

The most consistently reported changes in COPD include loss of quadriceps mass and strength [93] which occur regardless of gender [94] or the extent of airway obstruction; in addition, quadriceps muscle loss gets rapidly worse during exacerbations [95].

Key structural changes in skeletal muscles involve a shift from fatigue-resistant oxidative fibres to force-generating glycolytic fibres, accompanied by a reduction in muscle cross-sectional area [96]. However, these complex molecular alterations go beyond simple fibre-type shifts, as pointed out in a recent work by Chiles et al. [97]. By integrating muscle histology with transcriptomic profiling and using sex-stratified definitions of abnormal myofibre proportions, the authors demonstrated that approximately half of individuals with COPD exhibited reduced type I and/or elevated type IIx/IIax fibres, a phenotype associated with lower muscle strength, impaired exercise capacity, and worse expiratory flow limitation. Transcriptomic analysis identified 29 differentially expressed genes, including NEB, TPM1 and TPM2, alongside co-expression modules enriched for pathways involved in contraction, protein degradation, insulin signalling, and muscle atrophy. Importantly, these changes were not restricted to a single fibre type, but reflected a broader transcriptional remodelling of skeletal muscle. Finally, using a sex-stratified clustering approach, COPD participants with abnormal myofibre proportions demonstrated significantly lower median handgrip strength (26.1 versus 34.0 kg, p=0.022), 6-min walk distance (6MWD) (300 versus 353 m, p=0.039), and forced expiratory volume in 1 s (FEV₁)/forced vital capacity ratio (0.42 versus 0.48, p=0.041) compared with those with normal myofibre proportions. Abnormal myofibre composition, as defined by this clustering strategy, corresponded to <18% type I fibres and/or >22% type IIx/IIax fibres in males, or <36% type I fibres and/or >12% type IIx/IIax fibres in females.

Optimal muscle function is intrinsically linked to optimal mitochondrial activity. Mitochondrial function is central to muscle performance, and in COPD, mitochondrial density and biogenesis are diminished, while reactive oxygen species (ROS) production and apoptosis are increased [98]. These changes impair the muscle's metabolic capacity, promoting atrophy, reduced strength and endurance, and increased fatigability [6].

Beyond mitochondrial dysfunction, several other mechanisms are implicated in COPD-related skeletal muscle dysfunction, often acting synergistically, including genetic and epigenetic modifications, dysregulated myokine secretion, systemic inflammation, oxidative stress, impaired satellite cell function, physical inactivity, and metabolic disturbances [8]. A comprehensive description of the molecular mechanisms underlying skeletal muscle dysfunction in COPD is beyond the scope of the present review, and can be found in several excellent reviews [8, 91, 99]. In the next sections we focus on the relationship between skeletal muscle dysfunction and miRNAs and myokines in COPD.

miRNA expression in COPD

miRNAs play a pivotal role in regulating genes involved in skeletal muscle development and are critically involved in the modulation of skeletal muscle dysfunction in COPD [100], with some miRNAs (such as miR-150-5p) holding the potential to discriminate between COPD individuals and healthy controls [101]. Particularly in the advanced phases of disease, muscle repair and regeneration pathways are impaired and related to altered expression of myomiRNAs both in the circulation and in muscle tissue [13].

A pivotal study by Donaldson et al. [58] identified high circulating levels of miR-1, miR-499, miR-133 and miR-206 among stable COPD patients in comparison to healthy controls, probably reflecting increased muscle catabolism and turnover, as skeletal muscle-specific miRNAs tend to decrease as muscle atrophy worsens in COPD [58]. For instance, miRNA-499 is related to NF-κB p50 and its increased expression in quadriceps muscle has been suggested to be involved in regulating muscle inflammation in COPD [58].

Such findings were aligned to those of a study by Puig-Vilanova et al. [102], who investigated whether epigenetic mechanisms contribute to respiratory muscle dysfunction in COPD by analysing diaphragms from 18 patients with mild-to-severe COPD and 10 sedentary controls undergoing thoracotomy. Clinical assessments included spirometry, exercise testing, transdiaphragmatic pressure and quadriceps strength, while molecular analyses quantified DNA methylation, histone acetyltransferases, histone deacetylases (HDACs), myogenic transcription factors, SUMO ligases, and muscle-enriched miRNAs (miR-1, miR-133, miR-206, among others). Compared with controls, COPD patients showed reduced exercise capacity, diffusion capacity, and diaphragm strength, alongside downregulation of miR-1, miR-133 and miR-206, and upregulation of HDAC4 and MEF2C proteins, whereas DNA methylation, global protein acetylation, SUMOylation, and muscle fibre type/size remained unchanged. These findings suggest that altered microRNA and HDAC4–MEF2C signalling may represent adaptive responses to chronic inspiratory loading rather than overt structural atrophy. However, the cross-sectional design, small sample size limited to thoracotomy patients, and absence of longitudinal follow-up constrain causal inferences and generalisability.

In highly sedentary people with and without muscle weakness compared to controls, the expression of miR-1, miR-206 and miR-27a in vastus lateralis muscle is increased, whereas HDAC4, which inhibits MyHC I gene expression [103], and IGF-1, are decreased [104]. Similarly, it has been reported that the expression of miR-1, miR-206, miR-486 and miR-29b in the vastus lateralis muscle biopsies from non-cachectic COPD individuals compared to healthy controls was significantly upregulated, while no differences between groups were seen in the expression of miR-133a, miR 27a or miR-181a. All the aforementioned miRNAs were significantly lower expressed in cachectic than non-cachectic COPD individuals [105] suggesting that the expression level of miRNAs can vary in relation to disease severity or to a different timing of evaluation in disease course.

However, the findings on miRNAs expression in COPD reported herein are partly in contrast to the results of the study by Lewis et al. [106], in which the authors used quadriceps biopsies from COPD patients in order to investigate skeletal muscle dysfunction-associated alterations. In this study, COPD-related skeletal muscle dysfunction was associated with the downregulation of the myocardin-related transcription factor–serum response factor axis as well as the reduced expression of muscle-specific miRNAs, particularly miR-1 compared to healthy controls. Finally, the COPD group presented with increased levels of miR-1 targets, such as HDAC4 and IGF-1 signalling activity. Importantly, miR-1 expression correlated positively with lung function, exercise capacity, and type I muscle fibre proportion, implicating its reduction in fibre-type shifts and muscle atrophy. The reasons for these contrasting results can be related to differences in the characteristics of the study populations, biological methodologies, smoking history or to different timing of miRNA secretion, related, possibly, to a difference in physical activity or in disease severity.

Myokine expression in COPD

IL-6 is a pleiotropic myokine with a role in muscle differentiation and regeneration [107]. However, in pathological contexts like COPD, its elevated levels (both systemically and within muscle tissue) contribute to muscle atrophy by activating various signalling molecules [108, 109]. IL-6 levels in COPD correlate with disease severity and inversely with muscle strength and exercise capacity [110, 111]. IL-6 also promotes mitochondrial fission and reduces mitochondrial quality in vitro [112], while chronic IL-6 exposure increases muscle fatigability and disrupts mitochondrial function through glycoprotein (gp)130 signalling [113], increasing ROS and oxygen consumption via the JAK/STAT pathway [114].

Irisin, myostatin and IGF-1 may also play a role in COPD-related skeletal muscle dysfunction, although evidence is still sparse and mostly limited to small observational studies.

Dysregulated expression of both irisin and myostatin has been reported in a recent study in a mouse model of COPD with impaired skeletal muscle mass [115], as well as subsequent studies with human participants. Particularly, Kneppers et al. [116] found a significant increase in myostatin expression in the quadriceps muscle of COPD individuals. Cuttitta et al. [117] explored the relationships between body composition, circulating adipocytokines (leptin, adiponectin, haptoglobin) and irisin with functional outcomes in 25 mild-to-severe COPD patients and 26 matched healthy controls. Participants underwent spirometry, 6-min walk test (6MWT), bioelectrical impedance analysis for body composition, and quality-of-life assessments (Short-Form 36-item Health Survey, COPD Assessment Test, modified Medical Research Council, Beck Depression Inventory II, State-Trait Anxiety Inventory-Y), while serum biomarkers were measured using ELISA. COPD patients exhibited significantly greater fat mass, waist circumference and abdominal adiposity than controls, alongside elevated leptin, total adiponectin and haptoglobin levels, whereas irisin and intestinal fatty acid-binding protein (I-FABP) levels were unchanged. Importantly, leptin and haptoglobin correlated inversely with 6MWD and FEV₁, respectively, while irisin correlated negatively with adiponectin and haptoglobin only in COPD patients, suggesting that pro-inflammatory adipokines may counteract beneficial myokine effects on exercise tolerance and lung function. These findings support a complex interplay between adiposity, systemic inflammation, muscle-derived myokines and functional decline in COPD, despite several limitations stemming from the small sample size, the cross-sectional study design and the lack of mechanistic analyses. Another cross-sectional study [118] investigated associations between plasma myokines, inflammatory markers, sarcopenia and functional and respiratory performance in older adults with and without COPD. 86 participants aged ≥60 years (43 COPD, 43 non-COPD) underwent spirometry, dual-energy X-ray absorptiometry for body composition and sarcopenia diagnosis, functional assessments including the Short Physical Performance Battery (SPPB) and 6MWT, and respiratory muscle strength testing via maximal inspiratory (MIP) and expiratory (MEP) pressures. COPD patients exhibited significantly lower brain-derived neurotrophic factor (BDNF), irisin, and soluble TNF receptor (sTNFR)2 levels, alongside reduced 6MWD, SPPB score, MIP and MEP compared to controls. Multivariate regression revealed that COPD-related skeletal muscle dysfunction, lower BDNF and irisin levels, and higher sTNFR1/sTNFR2 levels independently predicted impaired functional and respiratory performance, explaining 18–41% of the variance across outcomes. Overall, findings implicate skeletal muscle dysfunction, reduced myokines and systemic inflammation as contributors to functional decline in older COPD patients, supporting biomarker-driven strategies for early intervention.

Exercise in COPD: effects on myokines and miRNAs

Exercise training is strongly recommended for individuals with COPD by all major international guidelines and is considered the cornerstone of pulmonary rehabilitation programmes. It has been shown to effectively reverse or at least decelerate the loss of skeletal muscle mass and strength [91], despite the considerable heterogeneity in intervention protocols and outcome measures [119]. In addition to exercise type, duration and frequency, the response to training in individuals with COPD is influenced by various factors such as age, sex, disease severity, and the presence of comorbidities.

Therefore, each pulmonary rehabilitation programme should be tailored according to the individual's capabilities. Nonetheless, endurance training and muscle strength training have been proven to be the most suitable strategies to improve HRQoL, reduce inflammation, deflate the patients, and improve physical performance at the 6MWT [120], especially when empowerment-based [121]. Nonetheless, other training strategies have been investigated, and the high-intensity interval training (HIIT) modality has emerged as quite promising. In this context, Hartmann et al. [122] conducted a nonrandomised controlled pilot study (24 patients: 12 cases and 12 controls) and investigated whether 12 weeks of supervised HIIT could improve alveolar–capillary reserve, lung tissue mass, and pulmonary perfusion distribution in patients with mild-to-severe COPD compared with age- and sex-matched healthy controls. The authors observed a severity-dependent reduction in alveolar–capillary reserve in COPD, mainly reflecting reduced pulmonary capillary blood volume rather than changes in alveolar–capillary membrane conductance. The 12-week HIIT programme significantly increased exercise capacity (peak oxygen uptake, peak workload) in both COPD patients and controls, but did not improve alveolar–capillary reserve, lung tissue mass, or perfusion distribution, suggesting that functional gains after HIIT are probably due to extrapulmonary adaptations (e.g. cardiovascular, muscular) rather than structural pulmonary changes.

Despite the well-established benefits of pulmonary rehabilitation in COPD, the extent of the response and its underlying molecular mechanisms remain heterogeneous. In this context, a study [123] that employed hierarchical cluster analysis to investigate skeletal muscle molecular responses to pulmonary rehabilitation identified two major patient clusters despite high variability in the results. Cluster 1 demonstrated a relative decrease in catabolic signalling, a relative increase in anabolic signalling, enhanced expression of oxidative metabolism mediators, and significant improvement in physical function. In contrast, cluster 2 exhibited less pronounced molecular and functional responses to pulmonary rehabilitation. Differential expression of specific molecular markers associated with myogenesis was observed between the two clusters, indicating distinct phases of muscle remodelling. These differences may reflect variations in the timing of initiation and/or the rate of progression of remodelling processes during pulmonary rehabilitation [123].

Comparable findings, though in a different experimental context, were reported by Maniscalco et al. [124], who investigated changes in the metabolic profile of individuals with COPD during pulmonary rehabilitation using nuclear magnetic resonance spectroscopy on exhaled breath condensate. Their study revealed that the metabolic phenotype in COPD evolves throughout pulmonary rehabilitation, highlighting a strong relationship between clinical and molecular parameters. These findings suggest that pulmonary rehabilitation does not follow a uniform, linear progression across patients; rather, it involves an induction phase that varies in duration and pattern for each individual.

Aerobic exercise has been shown to increase irisin expression, thereby improving skeletal muscle atrophy, enhancing resistance to skeletal muscle cell apoptosis, and boosting cellular antioxidant capacity [125]. Long-term, combined resistance and endurance training of moderate-to-high intensity leads to increased levels of IGF-1 and irisin in the skeletal muscles of individuals with COPD, promoting muscle hypertrophy and mitigating muscle atrophy [4]. Exercise also downregulates myostatin mRNA and its downstream muscle atrophy-related factors, MuRF1 and atrogin-1. Resistance training reduces myostatin levels in the quadriceps and increases the myogenin/MyoD ratio, contributing to enhanced muscle synthesis and the restoration of protein balance through increased synthesis and reduced degradation [126].

In individuals with severe COPD, skeletal muscle and serum irisin levels significantly increase after 8 weeks of exercise; however, no change in irisin levels is observed following acute exercise sessions [127]. Conversely, 3–6 months of high-intensity cycling training improved exercise capacity, but did not alter serum irisin levels [128].

Furthermore, 10 weeks of pulmonary rehabilitation resulted in increased IGF-1 mRNA levels in both cachectic and non-cachectic COPD patients. Notably, the rise in IGF-1 protein levels and the decline in myostatin were observed exclusively in non-cachectic individuals [129].

Hypoxaemia can impair muscle adaptation to exercise in COPD, as shown by Costes et al. [130]. While both normoxaemic and hypoxaemic individuals with COPD exhibited significant improvements in exercise capacity after 8 weeks of training, IGF-1 downstream signalling pathways (Akt/mTOR and Akt/GSK-3β) were only mildly activated in normoxaemic patients and significantly downregulated in hypoxaemic ones. Additionally, increases in citrate synthase and lactate dehydrogenase activities (markers of mitochondrial function), along with improvements in muscle fibre cross-sectional area and capillary-to-fibre ratio were observed only in normoxaemic individuals.

The effects of exercise on IL-6 expression appear to be inconsistent. While study reports with healthy subjects report a significant decrease in IL-6 mRNA after endurance training [131], a study conducted in COPD patients found no significant changes in plasma or vastus lateralis IL-6 expression following training [132]. This discrepancy may stem from different training modalities (endurance versus high intensity) and on the sample characteristics, underlying a potential impaired mechanisms occurring in COPD patients.

Some of the beneficial effects of exercise observed in both healthy and diseased individuals are mediated, at least in part, by miRNAs. These molecules play a key role in the adaptive responses of muscle tissue to exercise by regulating myokine gene expression, thereby enhancing mitochondrial function, promoting protein synthesis, and supporting myocyte proliferation [100, 133].

Nonetheless, the connection between exercise-induced alterations in miRNA expression and the adaptive response in COPD remains poorly understood. It has been hypothesised that individuals who respond well to exercise may exhibit a miRNA profile that promotes IGF-1 mRNA expression and activates the Akt/mTOR pathway.

In COPD individuals, exercise inhibits miR-696 [4], in line with previous evidence collected with murine models of muscle atrophy [134, 135]. Additionally, miR-144-3p, which targets the NF-κB pathway, is downregulated in COPD patients following 12 weeks of aerobic training, suggesting that exercise may influence COPD progression through the regulation of miR-144-3p expression [133].

Across the reviewed studies, important inconsistencies emerge in both miRNA and myokine data. These contrasting results probably reflect differences between intramuscular expression, linked to regulatory activity, and plasma levels, potentially indicating passive release during muscle wasting, compounded by cross-sectional designs and variable normalisation strategies. Other discrepancies may arise from differences in body composition, physical activity, systemic inflammation and comorbidity profiles across cohorts, as well as assay variability and small sample sizes limiting statistical power. Finally, most studies have a spatially limited dimension, taking into account only patients from a relatively small geographical area, thus reducing the possibility of strong comparisons. Collectively, these inconsistencies underscore the need for longitudinal, multi-omics studies integrating tissue and circulating biomarkers with standardised methodologies to disentangle whether these signals reflect causal mechanisms, adaptive responses, or epiphenomena of COPD-related muscle dysfunction.

Omic sciences to investigate skeletal muscle dysfunction in COPD

Recent contributions to the field have adopted novel omics modalities to investigate miRNAs and myokine expression, both in pre-clinical and clinical settings.

D’Amato et al. [136] studied a murine model of smoking-induced COPD in order to investigate the potential action of carnosine administration. In the process, they performed a proteomics analysis, which showed how smoke exposure was able to significantly alter the expression of 692 proteins, all related to specific pathways of inflammatory response, fibrosis, oxidative stress and calcium metabolism.

A multicentre case–control study [137] compared systemic proteomic profiles of COPD patients with frequent exacerbations, infrequent exacerbations, and those experiencing an acute exacerbation with healthy controls using two complementary approaches: unbiased label-free liquid chromatography–tandem mass spectrometry (LC–MS/MS) and targeted immune-based multiplex assays. Plasma samples from 20 participants with frequent exacerbations, 20 with infrequent exacerbations, 10 experiencing acute exacerbations, and 20 healthy controls were analysed for differentially abundant proteins, followed by pathway enrichment and protein–protein interaction network analyses. Compared with healthy controls, frequently exacerbating patients displayed extensive dysregulation of inflammatory and immune-related proteins, overlapping partially with profiles seen during acute exacerbation, while complement and coagulation cascades were uniquely altered during acute exacerbation. Notably, reductions in immunoglobulin subtypes and complement pathway activators suggested compromised adaptive immunity and impaired host defense in frequent exacerbation and acute exacerbation groups.

A large-scale meta-analysis [138] integrated genome-wide DNA methylation and transcriptomic data from three independent cohorts (Gene SMART, FUSION, GSE38291) to investigate sex differences in human skeletal muscle epigenetics and gene expression. Using Illumina methylation arrays, RNA-sequencing data, and quantitative PCR validation, the study identified 56 813 differentially methylated positions across the autosomes, 94% of which were hypomethylated in males compared to females, with 10 240 differentially methylated regions annotated to 8420 unique genes. Integration with GTEx and FUSION transcriptomes revealed that 326 genes exhibited sex differences in both methylation and expression, enriched for pathways involved in muscle contraction, substrate metabolism, and chromatin regulation. Notably, hormone receptor related transcription factor binding sites (e.g. androgen, oestrogen, glucocorticoid receptors) were overrepresented among sex-biased loci, while muscle fibre type proportions explained ∼16% of methylation differences, suggesting partial mediation by tissue composition rather than circulating hormone levels. Overall, the study provides the most comprehensive map to date of sex-specific skeletal muscle epigenetics, linking methylation, gene expression, and muscle phenotypes, with implications for sex-informed therapeutic strategies in metabolic and muscle-related disorders.

Finally, Pillon et al. [139], through a FAIR (findable, accessible, interoperable, and reusable) approach, standardised and annotated publicly available exercise and inactivity transcriptomes datasets, enabling transparent cross-study comparisons and user-friendly interrogation of gene expression patterns via www.metamex.eu. This approach overcomes the siloed nature of single-cohort studies, increasing the overall statistical power by overcoming heterogeneous protocols, small sample sizes, and inconsistent reporting. The authors could therefore identify the nuclear receptor 4 A3 (NR4A3) miRNA as one of the most exercise- and inactivity-response gene, with a key role in mitochondrial metabolism.

The main studies investigating the effects of pulmonary rehabilitation programmes on miRNAs, myokines and omics variations in COPD patients are summarised in table 2.

TABLE 2.

Summary of the main studies investigating the effects of pulmonary rehabilitation programmes on microRNAs (miRNAs), myokines and omics variations in COPD patients

First author, year [reference]	Country	Study design	Number of participants	Main findings	Limitations
Boeselt, 2017 [128]	Germany (Marburg and network)	Controlled clinical trial; 6-month individualised high-intensity training versus control	49 enrolled (31 training; 18 control); 37 per-protocol	Training ↑ 6MWT, ↑ rectus femoris area, ↑ QoL; myostatin/irisin/resistin/α-Klotho unchanged overall	Nonrandom allocation; biomarker null findings may reflect timing/assays; attrition from intention-to-treat
Costes, 2015 [130]	France and the Netherlands	Pre–post 2-month exercise training with biopsies; hypoxaemic versus normoxaemic COPD	23 COPD (15 normoxaemic; 8 hypoxaemic)	Training benefits capacity in both, but mitochondrial enzymes, fibre CSA, capillarity, and AKT/mTOR signalling improved only in normoxaemic; hypoxia blunted IGF-1→ AKT pathway	Small hypoxaemic subgroup; short duration; no healthy comparator; mixed modalities of training not fully standardised
Enríquez-Rodríguez, 2024 [137]	Spain	Case–control proteomics	70 (20 frequent exacerbator; 20 infrequent exacerbator; 10 acute exacerbation; 20 healthy control)	Systemic immune/proteomic dysregulation in frequent exacerbators and acute exacerbation; complement/coagulation changes specific to acute exacerbation; suggest impaired humoral defence in frequent exacerbators/acute exacerbation	Cross-sectional; limited power for pathway stratification; no longitudinal sampling across states within individuals
Ijiri, 2015 [127]	Japan	Case–control; serum irisin versus activity; acute and 8-week training substudy	99 (72 COPD; 27 controls)	COPD had lower serum irisin; irisin correlated with physical activity; no acute rise after single exercise bout; ↑ after 8-week training	Observational with training subgroup; biomarker variability; no mechanistic muscle measures; generalisability beyond single centre
Kneppers, 2019 [123]	The Netherlands and Slovenia	Prospective PR cohort with vastus lateralis biopsies pre–post PR; cluster analysis of molecular responses	51 COPD (4-week in-patient PR)	Identified distinct molecular clusters to PR: one with ↓ catabolic/↑ anabolic and oxidative signalling and greater functional gains; another with blunted molecular/functional response	Single-arm PR (no non-PR COPD control), 4-week duration, heterogeneity of responses; omics limited to selected markers
Landen, 2021 [138]		Multicohort meta-analysis integrating muscle methylome + transcriptome	369 methylomes; multicohort RNA-seq	Profound sex differences: 56 813 DMPs; 8420 genes in DMRs; 326 genes with concordant methylation–expression; ∼16% methylation differences explained by fibre-type composition	Cross-sectional; bulk tissue (no cell-type resolution); not COPD-specific but provides context for sex-aware analyses
Maniscalco, 2022 [124]	Italy	Longitudinal metabolomics of EBC during PR (0–2–5 weeks)	50 COPD (35 PR; 15 waiting-list controls)	Metabolic phenotype evolves through PR; methanol decreased in parallel with less dyspnoea/fatigue and ↑ 6MWD; supports molecular tracking of PR response	Modest sample; short follow-up; EBC metabolomics is exploratory and may be sensitive to pre-analytical variability
Pillon, 2020 [139]		FAIR MetaMEx meta-analysis with integration of 66 muscle transcriptome datasets (exercise/inactivity) + in vitro validation	>1100 participants across datasets	Clear segregation of acute versus training versus inactivity programmes; NR4A3 highlighted as top exercise/inactivity-responsive regulator with functional validation in myotubes; public www.metamex.eu resource	Reliant on heterogeneous public datasets; under-representation of women/older adults; not COPD-specific, but mechanistically informative
Vogiatzis, 2007 [132]	Greece	Prospective PR cohort	15 COPD patients	After 10 weeks of high-intensity exercise training 3× per week, peak work rate and muscle fibre CSA increased; mRNA and protein levels of IGF-I, MGF, and MyoD rose significantly; no significant change in local muscle or plasma TNF-α or IL-6 expression	Small COPD sample; no non-COPD control group; limited ability to detect changes in inflammatory mediators; short duration; gene/protein focus limited to selected pathways
Vogiatzis, 2010 [129]	Greece	PR intervention; compared cachectic versus non-cachectic COPD (muscle biopsy)	10 cachectic COPD versus 19 non-cachectic COPD	PR improved work-rate and 6MWD in both groups; fibre CSA ↑ in both, but less in cachectic; ↓ type IIb and ↑ capillary/fibre in both → partial preservation of remodelling in cachexia	Sample size; no non-COPD controls; cachexia definition variability

6MWT: 6-min walk test; QoL: quality of life; CSA: cross-sectional area; AKT: protein kinase B; mTOR: mechanistic target of rapamycin; IGF: insulin-like growth factor; PR: pulmonary rehabilitation; RNA-seq: RNA-sequencing; DMP: differentially methylated position; DMR: differentially methylated region; EBC: exhaled breath condensate; 6MWD: 6-min walk distance; FAIR: findable, accessible, interoperable and reusable; NR4A3: nuclear receptor subfamily 4 group A member 3; MGF: mechanogrowth factor; MyoD: myoblast determination protein; TNF: tumour necrosis factor; IL: interleukin.

Future perspectives and conclusion

Regular physical activity and exercise are well-established lifestyle interventions with substantial benefits for overall health, disease prevention, and management. In both healthy and diseased individuals, physical activity induces adaptive responses by modulating specific molecular pathways across multiple organ systems. These adaptations are mediated through the coordinated actions of myokines and miRNAs.

In COPD, the response to exercise varies widely depending on individual characteristics and disease severity. Therefore, integrating these factors (along with specific patterns of myokine and miRNA expression) is essential for designing exercise-based interventions that address the underlying mechanisms of chronic inflammation, muscle dysfunction, and metabolic disturbances. This understanding might have relevant implications in clinical practice, where skeletal muscle dysfunction might be overlooked; team works with expert therapists, nutritionists, and psychologists, a core characteristic of multidisciplinary pulmonary rehabilitation programmes, might help with the early identification of skeletal muscle dysfunction and its prompt management.

Understanding how different exercise modalities, as well as their intensity, duration and frequency, influence myokine and miRNA expression provides valuable insight for developing optimally personalised exercise programmes. Monitoring changes in circulating miRNAs and/or myokines before and after training not only advances our understanding of the molecular mechanisms and physiological adaptations triggered by various exercise approaches, but also enables clinicians to fine-tune exercise prescriptions to meet the specific needs of each patient.

Nevertheless, accomplishing this goal will require further comprehensive research to elucidate the complex interplay between myokines and miRNAs. Such investigations will be critical in developing tailored exercise protocols for individuals with COPD, grounded in their unique molecular profiles. Furthermore, deepening our understanding of miRNA–myokine interactions may open new avenues for therapeutic strategies aimed at improving muscle function and overall health in COPD.

Given that miRNA expression dynamically responds to various physiological stimuli, including exercise, miRNAs represent promising diagnostic and prognostic biomarkers across a range of pathophysiological conditions. Future research should focus on identifying the intrinsic muscular differences that influence the kinetics of molecular responses to exercise training. In particular, further investigations are needed to understand the muscle-specific factors that govern the temporal dynamics of the molecular response to PR.