Skeletal muscle mass is regulated by the balance between MPS and MPB. Nutrients (i.e., AA) and nutrient derived hormones (i.e., insulin) play a crucial role in regulating the balance between MPS and MPB. In ambulated humans, intake of dietary protein stimulates MPS due to the essential amino acids (EAA) components of proteins (Atherton and Smith, 2012). These anabolic responses are dose-dependent and saturable; at a maximal stimulus, rates of MPS increase ~200–300% for a period of ~2 h following ~20 g protein (Cuthbertson et al., 2005; Atherton and Smith, 2012). In contrast, insulin released following intake of dietary protein and/or CHO, is neither necessary nor sufficient to stimulate MPS (Greenhaff et al., 2008). Reflecting this, the anti-catabolic effects of insulin upon MPB was not recapitulated via AA infusions when insulin concentrations were clamped at 5 μU.ml⁻¹ (post absorptive; Greenhaff et al., 2008). Instead, insulin concentrations of just 15 IU/ml (3 × post absorptive) are sufficient to maximally suppress MPB (Wilkes et al., 2009). This anti-catabolic effect of insulin acting on MPB was confirmed in a recent systematic review and meta-analysis of 44 human studies, which concluded insulin did not significantly affect MPS but has a crucial role in reducing MPB (Abdulla et al., 2016). Therefore, while EAA's stimulate MPS, insulin suppresses MPB (and stimulates muscle glucose uptake). On the basis, EAA and insulin are so vital in maintaining muscle metabolic homeostasis, failure of these mechanisms inevitably leads to skeletal muscle atrophy and IR.

Disuse of human skeletal muscle alters muscle metabolism dynamics (Mallinson and Murton, 2013). For instance, early work by Gibson et al. showed that young men exposed to ULLS exhibited ~30% slower rates of fasted state MPS compared to the contralateral non-immobilized limb (Gibson et al., 1987). Subsequent studies confirmed reductions in MPS; for instance, ~50% reductions in MPS following 2-weeks of bed rest (Ferrando et al., 1996; Paddon-Jones et al., 2004) and ULLS with braces/casting (Glover et al., 2008). In further agreement with this, Kortebein et al. reported ~30% reductions in post-absorptive MPS during a 24 h period in older adults after 10 days of bed rest (Kortebein et al., 2007). Crucially, blunting of MPS in response to muscle unloading is not restricted to fasted periods. Glover et al. reported that ULLS led to ~27% reductions in postprandial rates of MPS at both low and high doses of AA infusions (Glover et al., 2008). Similarly, Drummond et al. reported that 7 days of bed rest blunted fed state MPS following EAA ingestion (Drummond et al., 2012). Similarly, 14 days of ULLS led to a ~30% reduction in MPS after ingestion of 20 g dietary protein (Wall et al., 2013). On this basis, available evidence strongly supports the notion that skeletal muscle atrophy in humans during a period of disuse is driven by blunting of both postabsorptive and postprandial MPS (Rennie et al., 2010; Wall et al., 2013).

The mTOR (mechanistic target of rapamycin) pathway is the major signal transduction network “hub” involved in the regulation of mRNA translation. Cell and rodent based research suggests this system senses important stimuli responsible for the regulation of MPS, i.e., (1) insulin and insulin-like growth factor-1 (IGF-1) through IRS (insulin receptor substrate) and PI3K (Phosphatidylinositol-3) pathways (Wackerhage and Ratkevicius, 2008); (2) AA through leucyl-tRNA/Rag-mTORc1 pathways (Wackerhage and Ratkevicius, 2008), (3) Energy stress through AMPK-eukaryotic elongation factor (AMP activated protein kinase/ eEF2) pathway (Wackerhage and Ratkevicius, 2008), and (4) mechanical stress, e.g., through mechano-sensory pathways (Wackerhage and Ratkevicius, 2008). However, the impact of these factors in the regulation of MPS in human remains unclear. For example, under ambulated conditions, one of the mTOR upstream effector, the class III PI3K hVps34 (human vacuolar protein sorting-34) was shown to be inhibited in amino acid starved state (basal condition) and increased activity of this protein appears to be concomitant with increased S6K1, suggesting it to be key player in mTOR signaling pathway (Dickinson and Rasmussen, 2011). Furthermore, intake of dietary protein typically leads to increase in phosphorylation status of class III PI3K hVps34 (Dickinson and Rasmussen, 2011), mTOR and its downstream substrates regulating mRNA translation, such as p70^S6K (Drummond et al., 2012). However, in response to disuse, blunted phosphorylation of mTOR and p70^S6K was shown after 2-weeks of immobilization while, no significant increases in phosphorylation of mTOR and p70^S6K were noted after immobilization (Wall et al., 2013). In contrast, others have reported no decreases in Akt/mTOR signaling despite reductions in MPS (de Boer et al., 2007; Wackerhage and Ratkevicius, 2008). It is worthwhile noting that peak stimulation of signaling proteins (mTOR and p70^S6K) occurs 1–2 h after protein intake while many studies have muscle biopsies taken 3–6 h after protein intake when the response would be attenuated or perhaps absent (Glover et al., 2010; West et al., 2011; Churchward-Venne et al., 2012a; Drummond et al., 2012). Further studies are needed with frequent biopsy sampling to fully determine a role for deactivation in mTORC1-related signaling networks (or indeed other putative mechanisms) regulating depressions in post-absorptive and post-prandial MPS.

In contrast to the recognized deficits in MPS during immobilization, the role of MPB in disuse atrophy is less clear. This is partly confounded by the technical and clinical challenges in measuring in vivo rates of MPB (e.g., arterio-venous balance methods to measure rate of appearance i.e., breakdown and pulse chase stable isotope approaches to measure fractional breakdown rates; Kumar et al., 2009; Atherton and Smith, 2012; Atherton et al., 2016). In the only study to our knowledge to measure this, Symons et al. demonstrated no increase in MPB in his study of healthy young volunteers exposed to 21 days of microgravity setting using a bed rest model (Symons et al., 2009). Although not directly quantifying MPB, Wall et al. reported that muscle free tracer enrichment over 4-h post prandial period was >3 fold higher after immobilization (Wall et al., 2013). Since MPS has consistently been shown to be reduced with immobilization, a likely explanation is that MPB is actually reduced (rather than increased) and hence the less unlabeled phenylalanine efflux was diluting the muscle free labeled tracer (L-(ring-²H₅) phenylalanine pool (Wall et al., 2013). Alternatively, an accumulation of free tracer could simply be explained by established “anabolic resistance,” i.e., where a failure of AA incorporation into the muscle lead to its accumulation (Glover et al., 2008; Phillips et al., 2009; Rennie, 2009; Wall et al., 2013). Nevertheless, regardless of the driving force behind muscle atrophy (i.e., disuse, aging, cancer, organ failure), blunted postabsorptive and postprandial MPS (Figure 1; anabolic resistance) seem to be the major drivers of disuse atrophy—rather than increases in MPB. Nonetheless, more work is needed across the time-course of unloading to verify this.

In terms of the molecular regulation of MPB, the ubiquitin proteasome system (UPS; Lecker et al., 2006) supplemented by lysosomal and calcium activated calpain (ATP–independent) and caspase dependent cleavage of actinomyosin complexes (Glover et al., 2010) are the major catabolic pathways in muscle. The identification of the “atrogenes” as genes that are uniformly upregulated irrespective of the atrophy stimulus (e.g., denervation, disuse, thermal injury) has received much attention as key drivers of atrophy programming (Bodine et al., 2001a; Jones et al., 2004; Milan et al., 2015). This led to members of the Forkhead Box (Fox) O family (Fox1, 3, and 4) being identified as downstream targets of Akt pathway (Figure 2) and as the main transcription factors regulating MAFbx/atrogin-1expression (Sandri et al., 2004). In terms of disuse atrophy, mRNA expression of two E3 ubiquitin ligases was initially found to be crucial in immobilization, unloading and denervation induced muscle atrophy (Bodine et al., 2001b). These genes, MuRF-1 (Muscle Ring Finger-1), and MAFbx/atrogin-1 (Muscle Atrophy F-box), are expressed in skeletal muscle at low levels but rapidly induced in response to unloading (Bodine et al., 2001b). In humans, after 5 days (Dirks et al., 2014) and 2 weeks (Jones et al., 2004) of immobilization, MAFbx and MuRF1 mRNA were reported to be elevated. Nonetheless, while their expression is thought to be regulated by transcription factors such as FOXO1, FOXO3a, (Sandri et al., 2004) and NFκβ (p50 and Bcl-3)(Wu et al., 2011), no increase in mRNA expression in FOXO's were noted after 4 or 14 days of immobilization (Suetta et al., 2012). De Boer et al reported the expression of MuRF-1 but not MAFbx mRNA was increased after 10 days of ULLS (Jones et al., 2004; de Boer et al., 2007), while both had decreased by 21 days (de Boer et al., 2007). Furthermore, increases in UPP components (particularly UBE2E) were up-regulated 48 h following instigation of ULLS (Urso et al., 2006). In contrast, a recent study by Brocca et al. found that muscle atrophy following ULLS found no change in mRNA expression of ubiquitin-proteasome and autophagy systems (Brocca et al., 2015). Some work has been done in relation to autophagy (and calpain-signaling) in relation to human disuse (Jones et al., 2004). Autophagy is responsible for removing unfolded, damaged and dysfunctional proteins and organelles via the formation of autophagosomes for degradation by lysosomes (Sandri, 2010). Interestingly, up-regulation of autophagy markers such as Beclin-1 suggested increased autophagosome formation and hence a higher activity of the macro-autophagy by 24 days of bed rest; nonetheless, other autophagy markers such as P62, LC3II/I ratio, and cathepsin-L were not up-regulated (Brocca et al., 2012). While details of the pathways discussed above are outside the scope of this review, the readers are referred to reviews by Bonaldo and Sandri (2013) and Sandri (2010). What is clear however is that without more clinical studies with time-course acquisition, including dynamic measures of MPB in tandem with molecular markers spanning different proteolytic systems, no firm conclusions can be made surrounding the mixed results regarding whether existing molecular data suggest MPB is increased, decreased or unchanged in response to disuse in humans.

In addition to the anabolic and catabolic pathways described above, emerging evidence indicates that disturbed redox signaling may also be an important regulator of MPS and MPB in muscle disuse atrophy (Powers et al., 2012; Zuo and Pannell, 2015). Oxidative injury has been shown to occur in muscle fibers during periods of disuse in locomotor skeletal muscles (Min et al., 2011) and in non-load bearing muscle such as the diaphragm during prolonged mechanical ventilation (Kavazis et al., 2009). After 35 days of bed rest, vastus lateralis muscle showed ~18% muscle fiber atrophy and increased protein carbonylation (Dalla Libera et al., 2009). Furthermore, an inverse linear relationship was observed between normalized levels of protein oxidation and muscle fiber CSA (Dalla Libera et al., 2009). An analysis of gene expression showed up-regulation of pathways involved in the oxidative stress response (increase in mRNA for stress response gene heme oxygenase-1) following 48 h of unilateral lower leg suspension (ULLS; Reich et al., 2010). Conversely, in a limb immobilization human model with ~5.7% muscle and 11.8% muscle fiber loss after 14 days of immobilization, no increase in lipid peroxidation and protein oxidation in vastus lateralis was observed (Glover et al., 2010). Although information on oxidative stress and potential mechanisms explaining proteolysis in disuse human muscle is still sparse (Pellegrino et al., 2011), these findings support the extensive evidence available from animal studies that oxidative stress inhibits MPS (Powers et al., 2011a) and increases muscle MPB (via increased gene expression of key proteins involved in the proteolytic pathways such as autophagy, calpains and proteosomes; activation of both calpain and caspase-3 and possibly by modification of myofibrillar proteins which enhances their susceptibility to proteolytic processing; Powers et al., 2011a). Interaction between ROS and insulin signaling pathway has also been described, i.e., ROS may regulate Insulin growth factor-1 (IGF-1) signaling either positively or negatively depending on the amount of ROS produced (Papaconstantinou, 2010). Low levels of endogenous ROS due to their reversible oxidative inhibition of protein tyrosine phosphatases induces phosphorylation of tyrosine residue on the insulin receptor and its substrates triggering IGF-1 signaling (Bashan et al., 2009). In contrast, the IGF-1 signaling pathway is inhibited by higher levels of ROS and recent evidence suggests ROS down regulates the IGF-1 cascade and induces insulin resistance (Bashan et al., 2009; Figure 2). For detailed discussion of the signaling pathways linking ROS and muscle atrophy, the interested reader is referred to recent reviews on oxidative stress and disuse muscle atrophy (Pellegrino et al., 2011; Powers et al., 2011b, 2014; Zuo and Pannell, 2015).

Insulin-mediated glucose uptake is also blunted with muscle disuse (Mikines et al., 1991; Biensø et al., 2012); that is, unloaded muscle becomes IR. This IR can be observed at a whole-body level following bed-rest, but is most apparent at the muscle level across the physiological range of insulin concentrations under clamp conditions (Mikines et al., 1991). Recently, a 1 week bed-rest study in young males by Dirks et al. revealed reduced muscle mass (~1.4 kg lean tissue and ~3% quadriceps CSA) and whole-body insulin sensitivity (~29%)(Dirks et al., 2016). Thus, disuse lowers MPS, induces anabolic resistance to nutrients and impairs insulin-mediated muscle glucose uptake—even in healthy adults (Fink et al., 1983).

The role of IR in driving muscle atrophy however is poorly defined. Evidence from large cross-sectional and longitudinal studies reports accelerated loss of muscle mass in individuals with persistent IR (i.e., people with Type 2 Diabetes), perhaps pointing to mechanistic links. For instance, declines in muscle mass were inversely related to duration of diabetes or HbA1c (Park et al., 2007, 2009; Kalyani et al., 2013) and attenuated with insulin sensitizers (Kuo et al., 2009). Human muscle tissue accounts for 80% of glucose uptake after food ingestion and insulin resistance (HOMA-IR) is associated with reduced quadriceps muscle strength (Kalyani et al., 2013; Leenders et al., 2013), power (Kalyani et al., 2013) and muscle mass (Leenders et al., 2013) in humans. Approximately a 50% more rapid decline in knee extensor strength has been observed in older patients with type 2 diabetes compared with patients without diabetes over a 3 year period, suggesting that decreased muscle strength may be accelerated in type 2 diabetes (Park et al., 2007). In a further study, Volpato et al. reported differences in walking speed, muscle strength, power and muscle quality between individuals with and without diabetes were independent of co-existing peripheral motor neuropathy or peripheral vascular disease, suggesting a direct effect of diabetes per se on muscle performance (Volpato et al., 2012). These findings are important because in catabolic conditions such as diabetes, atrophy in combination with reduced activity decrease quality of life and increase mortality (Zinna and Yarasheski, 2003). Yet despite clear evidence linking accelerated muscle loss in diabetes compared to non-diabetes, studies investigating the direct effect of immobilization on muscle protein turnover in patients with diabetes compared to those without diabetes are scant. Furthermore, clear distinction between Type 1 and Type 2 diabetes needs to be made when investigating patients with diabetes. This is because Type 1 diabetes is a condition with severe depletion of energy stores and reduced mitochondrial function resulting in accelerated muscle protein loss (Hebert and Nair, 2010), which can be reversed by insulin replacement (Workeneh and Bajaj, 2013). In contrast, muscle loss, whilst accelerated in type 2 diabetes, is unaffected by insulin treatment (Workeneh and Bajaj, 2013), possibly due to IR. Hence, skeletal muscle mass loss whilst common, appears to occur less predictably and to varying degree in Type 2 diabetes compared with Type 1 diabetes (Workeneh and Bajaj, 2013). Collectively, these data are consistent with the notion that diabetes causes muscle mass loss possibly due to mechanisms driving muscle IR, however there is lack of data regarding the effects of immobilization or disuse on muscle mass in individuals with diabetes.

The mechanistic regulation of muscle IR in driving muscle atrophy in the setting of “simple disuse” remains vague (Atherton et al., 2016). Early human studies by Shulman et al. showed that, under steady state plasma concentration of both glucose and insulin mimicking postprandial conditions, the mean rate of muscle glycogen synthesis accounted for most of the whole body glucose uptake and virtually all of non-oxidative glucose metabolism in both healthy and diabetic subjects (Shulman, 2000), with defects in muscle glycogen synthesis playing a major role in causing insulin resistance in type 2 diabetes (Shulman, 2000). This may be explained by defects in the insulin receptor substrate (IRS)-1/phosphatidylinositol (PI) 3-kinase pathway, leading to reduced glucose uptake and utilization in insulin target tissues (Draznin, 2006). Free fatty acids induce muscle IR by inhibiting glucose transport/phosphorylation and reductions in both the rate of muscle glycogen synthesis and glucose oxidation (Roden et al., 1996). Additionally, many other mechanisms have been postulated to explain free fatty acid-induced muscle IR, including the Randle cycle, oxidative stress, inflammation and mitochondrial dysfunction (Martins et al., 2012). Full details regarding above mechanisms escape the scope of this article and readers are referred to a review by Martins et al. (2012). With regard to disuse induced muscle atrophy, following (7 days) bed-rest healthy volunteers showed reduced glucose infusion rate and leg glucose extraction (after bed rest) along with reduced muscle GLUT4, hexokinase II, protein kinase B/Akt1, and Akt2 protein levels, and a tendency for reduced 3-hydroxyacyl-CoA dehydrogenase activity (Biensø et al., 2012). Further in the same study, the ability of insulin to phosphorylate Akt and activate glycogen synthase was reduced post bed-rest (Biensø et al., 2012); but whether this observation is causative or a consequence of immobilization is not clear. However, a substantial decline in glucose uptake within 24 h of immobilization would argue against a causative effect (Atherton et al., 2016). Recently, Vigelso et al. showed an inverse association between the increase in muscle pyruvate dehydrogenase complex (PDC) activation and leg lactate release during contraction after 2 weeks unilateral lower limb immobilization, suggesting PDC as a potential key regulator of immobilization induced muscle IR (Vigelsø et al., 2016). Overall the above data suggests that muscle disuse results in development of whole body and muscle specific IR, fuelling the argument that lack of muscle contraction per se may be the main physiological driver for this dysregulation, however a mechanistic explanation for this still remains unclear (Atherton et al., 2016).