Classically, circulating TNF-α is reported to be higher in COPD patients than controls and negatively correlated to lean mass [23,24], as well as to muscle strength, including upper muscles [25]. However, this correlation seems to be consistent only in cachectic patients [26]. Moreover, with regards to muscular TNF-α, significantly elevated, decreased, and unaltered levels have been reported in the quadriceps of COPD patients [27,28,29], or even undetected at the protein level [30,31], although the effects of TNF-α could also be mediated by the elevated circulating levels. Mechanistically, TNF-α plays a dual role in muscle mass regulation, possibly via differential signaling according to receptor binding (TNF-R1 or TNF-R2) [32,33]. On the catabolic counterpart, TNF-α promotes the inflammatory response by binding on TNF-R at the macrophage surface and activating the nuclear factor-kappa B (NF-кB) pathway via the inhibitory-kappa B kinase alpha (IKKα) activation, thus supporting the M1-biased (inflammatory) phenotype. TNF-α secreted by myeloid cells (CD68+ macrophages) inhibits myotubes fusion in vivo [34], supporting its catabolic role. In vitro, it can be induced in C2C12 myotubes by serum amyloid A, which is increased during COPD exacerbations [35]. Possibly in a paracrine/autocrine fashion, its secretion activates glycolytic metabolism in C2C12 myotubes in a NF-кB-dependent manner, as well as enhancing the secretion of HIF1α, a key transcription factor induced in hypoxic or inflammatory conditions [36]. In the same study, HIF1α and its downstream target vascular endothelial growth factor (VEGF) were found elevated in quadriceps biopsies of COPD patients and correlated to the level of muscular TNF-α. Finally, a recent study has found that the downregulation of histone deacetylase 2 (HDAC2) by RNA interference in cultured primary myotubes increased TNF-α production and cell apoptosis [37]. HDAC2 belongs to a family of enzymes which classically downregulate inflammatory genes’ expression. The authors also evidenced a decrease of HDAC2 level in the quadriceps of COPD patients, associated with an increase in the NF-кB pathway.

However, some in vivo models seem to counteract in vitro data. Indeed, a mice model of TNF-α-signaling knock out (TNF-R2 KO) exhibited a more severe phenotype of skeletal muscle wasting (including reduced fibers’ CSA and an increase of type IIx myofibers) than wild-type littermates [32] after cigarette smoke (CS) exposure. Other data support the anabolic effects of TNF-α, which acts as a mitogen for satellite cells [38]. In myocytes, the binding of TNF-α to its receptor TNF-R activates the p38-mitogen-activated protein kinase (MAPK) pathway, leading to a silencing of target genes such as Pax7 and Notch1 (via an epigenetic control), promoting satellite cell differentiation into myotubes [39], in contrast to that previously reported [40]. The latter study also showed an increase in mitochondrial content in IKKα-expressing myotubes. Other models tend to corroborate in vitro data, such as a mouse model exposed to CS and exhibiting increased circulating TNF-α levels, which were correlated to detrimental outcomes such as muscle mass, decreased capillarization, and increased catabolism [41]. Moreover, an overexpression of TNF-α in the lung leads to an increased muscle fatigue, decreased muscle mass and mitochondrial content, and increased catabolism (mostly in male mice) [42]. These data might suggest that part of the deleterious effects of inflammation are secondary. Of note, therapies against TNF-α have not proven to be effective for various outcomes such as dyspnea or a number of exacerbations [43], suggesting that this cytokine is not the primary or at least the sole driver of COPD inflammation.

Circulating IL-6 is also classically elevated in COPD patients compared to healthy subjects, both during and outside exacerbations [44] and associated with reduced quadriceps strength [45]. Like TNF-α, IL-6 is considered as a “double-edged sword” with proinflammatory (catabolic) and anti-inflammatory (anabolic) effects (see the review in [46]) according to local immune cells and the cytokines’ microenvironment. Overall, the pilot study of Tsujinaka and colleagues showed that transgenic mice chronically overexpressing IL-6 exhibit a marked muscle atrophy [47]; however, the knockout of IL-6 was not sufficient to prevent sarcopenia in a sepsis mouse model [48]. In muscle from COPD patients, IL-6 has been reported as unchanged at the transcriptional level (at both stable and exacerbation states) compared to that of healthy controls [27,28].

In vitro, IL-6 bears anabolic effects by regulating satellite cells’ function and enhancing glucose metabolism; furthermore, a loss of IL-6 signaling in myoblasts results in a reduced proliferation and migration [49,50].

In muscle from COPD patients, IL-8 has been reported as unchanged at the transcriptional level (at both stable and exacerbation states) compared to that of healthy controls [28]. Surprisingly, despite an important role for IL-8 in COPD pathophysiology and in particular lung inflammation, few data are available regarding its expression in the skeletal muscles of COPD patients.

One study has found an elevation of IL-18, another proinflammatory cytokine, in the plasma and the quadriceps of COPD patients (at the mRNA level) compared to healthy controls [51]. Immunohistochemistry showed that the increase was predominant in type II fibers. Of note, muscle mRNA levels of TNF-α and IL-6 were unchanged in that study. Moreover, IL-18 mRNA levels were not significantly different in the quadriceps of COPD patients compared to that of healthy smokers. Of note, contrary to IL-6, IL-18 was not altered by exercise [51]. It was hypothesized that this increase could participate in skeletal muscle wasting by increasing local apoptosis.

In vitro, IL-15 has an anabolic effect by inducing myosin chain synthesis in myotubes [52], notably in response to TNF-α stimulation [53]. This is particularly interesting given that IL-15 was increased and correlated with TNF-α in a rat model of COPD (induced by CS exposure as well as LPS instillations), in the serum and in both peripheral and respiratory skeletal muscles [54]. In the same study, this was paralleled by an increase in ubiquitin–proteasome markers (as expected), which were positively correlated to IL-15 levels. Furthermore, IL-15 possibly promotes the effector function of memory CD8+ T cells [55]. However, no data have been reported in human COPD muscle samples to the best of our knowledge.

IFN-γ has been reported as unchanged at the protein level in COPD quadriceps compared to controls [27]. However, a recent transcriptional analysis shows a globally downregulated interferon response in COPD quadriceps [56]. In vitro, IFN-γ does not activate catabolic pathways on C2C12 myoblasts [57] but induces the expression of proangiogenic factors such as angiopoietin-2 in primary human myoblasts [58]. Taken together, these data may indicate that IFN-γ downregulation could be part of the pathological mechanisms leading to skeletal muscle wasting, although other observational as well as functional data lack at this stage.

IL-10 exerts a consistent anti-inflammatory, pro-regenerative role (as demonstrated by acute injury models). It bears potent anabolic effects by inhibiting the atrophy signaling induced by TNF-α in myotubes [59]. As a cytokine implicated in muscle regeneration in acute injury models (see below), its potential protective role in muscular dystrophy is not surprising [60]. However, to date, it has not been studied in skeletal muscles of COPD patients.

Other myokines can also play a role in the control of muscle mass, most of which having not been thoroughly investigated in COPD, such as IL-4 (promoting myoblast fusion in vitro) [61], IL-7, which secretion by skeletal muscle gradually decreases with age [20], or irisin, known to decrease oxidant-induced apoptosis in diabetes mellitus, and recently reported to be decreased in COPD serum [62]. Two studies outside the COPD context also point towards a prominent role of the Toll-like receptor (TLR)-4 pathway in muscle wasting, via the activation of the p38-MAPK atrophy pathway [63,64]. However, to date, the expression of IL-17, a cytokine activating TLR4 signaling, has not been investigated in COPD muscle.