Key considerations for investigating and interpreting autophagy in skeletal muscle

ABSTRACT

Skeletal muscle plays a crucial role in generating force to facilitate movement. Skeletal muscle is a heterogenous tissue composed of diverse fibers with distinct contractile and metabolic profiles. The intricate classification of skeletal muscle fibers exists on a continuum ranging from type I (slow-twitch, oxidative) to type II (fast-twitch, glycolytic). The heterogenous distribution and characteristics of fibers within and between skeletal muscles profoundly influences cellular signaling; however, this has not been broadly discussed as it relates to macroautophagy/autophagy. The growing interest in skeletal muscle autophagy research underscores the necessity of comprehending the interplay between autophagic responses among skeletal muscles and fibers with different contractile properties, metabolic profiles, and other related signaling processes. We recommend approaching the interpretation of autophagy findings with careful consideration for two key reasons: 1) the distinct behaviors and responses of different skeletal muscles or fibers to various perturbations, and 2) the potential impact of alterations in skeletal muscle fiber type or metabolic profile on observed autophagic outcomes. This review provides an overview of the autophagic profile and response in skeletal muscles/fibers of different types and metabolic profiles. Further, this review discusses autophagic findings in various conditions and diseases that may differentially affect skeletal muscle. Finally, we provide key points of consideration to better enable researchers to fine-tune the design and interpretation of skeletal muscle autophagy experiments.

Abbreviation: AKT1: AKT serine/threonine kinase 1; AMPK: AMP-activated protein kinase; ATG: autophagy related; ATG4: autophagy related 4 cysteine peptidase; ATG5: autophagy related 5; ATG7: autophagy related 7; ATG12: autophagy related 12; BECN1: beclin 1; BNIP3: BCL2 interacting protein 3; CKD: chronic kidney disease; COPD: chronic obstructive pulmonary disease; CS: citrate synthase; DIA: diaphragm; EDL: extensor digitorum longus; FOXO3/FOXO3A: forkhead box O3; GAS; gastrocnemius; GP: gastrocnemius-plantaris complex; MAP1LC3/LC3: microtubule associated protein 1 light chain 3; MAPK: mitogen-activated protein kinase; MYH: myosin heavy chain; PINK1: PTEN induced kinase 1; PLANT: plantaris; PRKN: parkin RBR E3 ubiquitin protein ligase; QUAD: quadriceps; RA: rectus abdominis; RG: red gastrocnemius; RQ: red quadriceps; SOL: soleus; SQSTM1: sequestosome 1; TA: tibialis anterior; WG: white gastrocnemius; WQ: white quadriceps; WVL: white vastus lateralis; VL: vastus lateralis; ULK1: unc-51 like autophagy activating kinase 1.

Background

Skeletal muscle is responsible for generating force and facilitating movement through the activation of heterogeneous cells known as muscle fibers. These fibers exist on a continuum and are defined by the rate at which they generate force and their metabolic characteristics, amongst other properties. The diversity of fibers within skeletal muscle offers several advantages as they allow for variability and flexibility in the production and maintenance of force generation (i.e., speed of contraction, endurance, fatigability). One pivotal aspect of this classification revolves around the expression of specific adult myosin heavy chain (MYH) isoforms, which is often termed “fiber type” with fibers commonly referred to as type I (MYH7), type IIA (MYH2), type IIX (MYH1) and type IIB (MYH4) fibers. Another major classification defines fibers based on metabolic phenotype, which relates to its capacity for energy production and substrate utilization. This classification system describes fibers as being “oxidative” or “glycolytic.” Generally in humans, type I fibers contract at a slower rate and have greater oxidative potential (i.e., slow-twitch/oxidative) compared to type IIA and type IIX fibers which contract at a faster rate, and have greater glycolytic potential (i.e., fast-twitch/glycolytic) [1]. Importantly, type IIB fibers are not present in humans but are in rodents. Type I fibers exhibit a more intricate and densely packed mitochondrial network compared to type II fibers in mice [2]. Investigations into mitochondrial function have revealed higher respiration rates in permeabilized fibers from type I muscles compared to type II muscles in mice [3]. Studies have found comparative cristae density in human type I and type II fibers [4], suggesting similar structural characteristics; however, the elevated mitochondrial respiration in type I fibers may stem from their overall greater mitochondrial content. Importantly, this “matching” of contractile and metabolic characteristics does not necessarily hold true in all mammals. For example, type IIA fibers have both a faster rate of contraction and higher oxidative potential than type I fibers in rodents (see below for further discussion). Thus, we have explicitly noted the species when we discuss findings, as this is an important consideration. Within this paper we will use the terms slow/oxidative and fast/glycolytic to describe skeletal muscles with these “broad characteristics” solely to simplify our classification approach and discussion. However, it is important to note that these are “relative terms” as these characteristics exist on a continuum and can be altered/adapt to various perturbations. Moreover, we stress that these are only two broad characteristics within skeletal muscles that may be differentially expressed or impact “signaling”; thus, additional consideration beyond and even within these broad characteristics is warranted.

Skeletal muscle exhibits remarkable plasticity, capable of adapting to various stimuli by undergoing phenotypic changes. This adaptive capacity allows skeletal muscles to modify their structural and functional characteristics in response to different physiological demands. For instance, exercise training can promote the change in MYH composition (i.e., fiber type) [5,6], and mitochondrial content/function [7–10]. Recruitment patterns also play a pivotal role in shaping skeletal muscle adaptation. The selective activation of skeletal muscle fibers during different activities influences fuel usage and triggers specific adaptive responses [11–13]. Moreover, varied recruitment patterns during exercise can stimulate diverse energy pathways, prompting the skeletal muscle to adapt by optimizing its metabolic profile to efficiently meet the demands of the given task [14].

The divergent contractile and metabolic profiles in skeletal muscle fibers, as well as the dramatic plasticity of this tissue to various conditions may signify distinct activation of cellular signaling pathways, particularly ones that are highly affected by the energy state of the fiber, such as autophagy. Selective forms of autophagy, such as aggrephagy, reticulophagy, and ribophagy, have been implicated in various conditions like exercise and disease [15–18]. However, it remains unclear whether these selective forms of autophagy differ between skeletal muscle/fiber types. Alternatively, selective processes like mitophagy may be dramatically affected by the energy state of the fiber, given basal mitochondrial content and network differences, as well as the dramatic adaptation of this organelle in skeletal muscle to various perturbations. Thus, it is imperative to acknowledge and consider these characteristics to accurately design and interpret autophagy research. For instance, autophagy protein/gene expression and/or flux may change because of alterations in the fiber type composition and/or metabolic profile of the skeletal muscle. As a result, these changes in autophagic signaling or flux may be attributed to altered contractile phenotype, metabolic profiles, or other properties. Here we provide an overview of the autophagic profile and response in skeletal muscles/fibers of different types and metabolic profiles. We also discuss how autophagic findings may be affected in specific conditions/diseases due to skeletal muscle plasticity. Finally, we provide key points of considerations to studying skeletal muscle autophagy to better advance our understanding of this field.

Basal autophagy factor expression and flux in skeletal muscles of different MYH composition and metabolic profiles

Examination of the literature demonstrates a compelling relationship between the fiber type composition, metabolic phenotype of fibers, and the autophagic process (Figure 1; Table 1). As an initial cautionary note, we would like to emphasize that although the MAP1LC3/LC3 (microtubule associated protein 1 light chain 3) ratio (i.e., LC3-II:I) has been used in this literature, it has several limitations and should not be interpretated as a marker of autophagic flux [24]. Under otherwise healthy states, there is a noticeable distinction in autophagy factor expression and flux between different skeletal muscles. Most studies are conducted in rodents due to the ease of studying multiple skeletal muscles compared to humans. In human studies, the majority of work has focused on the vastus lateralis (VL), potentially limiting broad applicability [25–27]. Nonetheless, there seems to be discernible differences in select autophagy proteins/genes between fiber types. For instance, single fiber proteomic analysis found a higher abundance of LC3B in type IIX followed by type IIA and type I human muscle fibers [19]. In rodents, basal levels of ATG (autophagy related) proteins, such as ATG7, ATG12 and BECN1 (beclin 1) are greater in skeletal muscles consisting of fibers of a slower contractile type and that are more oxidative in nature (i.e., slow/oxidative), such as the soleus (SOL) [20,21,23]. Others demonstrate greater levels of select proteins, including ULK1 (unc-51 like autophagy activating kinase 1) and ATG5, in skeletal muscles that are comparably faster and more glycolytic (i.e., fast/glycolytic) such as the tibialis anterior (TA) and gastrocnemius-plantaris complex (GP) [22]. Interestingly, while downstream cargo and recognition proteins such as LC3 are generally higher under basal conditions in mouse slow/oxidative compared to fast/glycolytic skeletal muscle, SQSTM1 is more ambiguous [20–23]. Generally, these findings suggest that there are higher basal levels of several autophagic factors in slow/oxidative muscle; however, flux experiments paint a different picture. Injection of mice with colchicine, a compound that prevents autophagosome-lysosome fusion and degradation of cargo [28], has revealed greater basal autophagic flux in primarily fast/glycolytic muscles, including the TA, GP and EDL (extensor digitorum longus) compared to slow/oxidative muscles like the SOL and diaphragm (DIA) [22,23]. Similarly, fasting experiments have shown rapid and greater elevations in LC3-II in plantaris (PLANT), EDL, and TA compared to the SOL [22,29]. Interestingly, a negative linear correlation between LC3B-II flux and CS (citrate synthase) activity is observed under basal and fasted conditions [22]. Specifically, relatively fast/glycolytic muscles (e.g., TA and GP) display lower CS activity and greater LC3B-II flux, and vice versa compared to relatively slow/oxidative muscles (e.g., SOL and DIA) [22]; suggesting that LC3B-II flux may be better related to metabolic properties than MYH composition. It is important to note that this relationship only takes into consideration one autophagy marker (i.e., LC3B-II flux), and thus may not fully represent the entire pathway. Although there seem to be skeletal muscle/fiber differences in basal autophagic protein/gene levels, the existence and exact nature of this relationship is likely dependent upon the specific skeletal muscles and autophagic markers examined. Furthermore, autophagic protein/gene expression patterns and responses may not be easily explained by a single property/characteristic across all skeletal muscles and fibers; thus, factors such as MYH expression, metabolic characteristics, recruitment patterns, organelle content, among other features require consideration. Ultimately, understanding this intricate interplay provides a pathway to unravel the nuanced regulatory mechanisms that underlie the role of autophagy in skeletal muscle.

Figure 1.

Slow/oxidative muscles have slower contraction speeds, greater mitochondrial content, and higher oxidative potential compared to fast/glycolytic muscles along with greater expression of many autophagy-related factors. In contrast, fast/glycolytic muscles exhibit greater autophagic flux and sensitivity to metabolic stressors (e.g., fasting and exercise shown).

Table 1.

Basal differences in autophagy protein/gene expression and flux between skeletal muscles of different contractile and metabolic properties.

Study	Species	Marker	Expression in Muscle at Baseline
Murgia et al., 2021 [19]	Human	Total LC3B	type IIX and IIA>type I (VL)
McClung et al., 2010 [20]	Mouse	ATG7, BECN1, ATG12, LC3-I:II*	SOL>WVL
		BNIP3	WVL>SOL
Lira et al., 2013 [21]	Mouse	LC3-II:I, LC3-II, ATG7, BECN1, BNIP3	SOL>PLANT>WVL
		SQSTM1	WVL>PLANT>SOL
Mofarrahi et al., 2013 [22]	Mouse	Ulk1, Atg5, Atg12, Autophagic flux, ULK1, ATG5, ATG12–ATG5	TA and GP>SOL and DIA
		Lc3b, Bnip3, Prkn, Becn1, Pik3c3, BECN1, PIK3C3, LC3B-II, BNIP3	SOL and DIA>TA and GP
		PRKN	DIA>SOL, TA, and GP
		Sqstm1, Atg7, Atg4b, SQSTM1, ATG7	SOL=DIA=TA=GP
Paré et al., 2017 [23]	Mouse	Autophagic Flux	EDL>SOL
		SQSTM1, LC3-I, LC3-II	SOL>EDL
		SQSTM1, LC3-I, LC3-II	RQ>WQ
		ATG7, BECN1, LAMP2	RQ=WQ

*note use of unconventional ratio format within cited study.

This may be of considerable importance for the study of organelle-specific autophagy such as the selective clearance of mitochondria, termed mitophagy. In the process of mitophagy, key proteins associated with mitophagy localize to the mitochondria and mark them for degradation. This is relevant because some reports show differential mitophagy-related protein levels in different skeletal muscles [21,22]. For example, slow/oxidative rodent muscles, like the SOL and/or DIA, display higher BNIP3 (BCL2 interacting protein 3) and PRKN (parkin RBR E3 ubiquitin protein ligase) compared to fast/glycolytic TA, GP, PLANT and white vastus lateralis (WVL) muscles [21,22]. However, another report found that BNIP3 is greater in mouse WVL compared to SOL muscle [20]. Not only do these findings demonstrate disparities between the skeletal muscles examined, but it also suggests that levels are likely influenced by factors such as MYH expression and/or mitochondrial content. Given this complexity, particular care needs to be taken when investigating mitophagy markers in skeletal muscle as they may be dependent upon mitochondrial content, and further complicated by the plasticity of this organelle in skeletal muscle to various physiological and pathophysiological states (discussed further below).

Divergent effects of autophagy deficiency in skeletal muscles of different MYH composition and metabolic profile

Tissue-specific autophagy deficiency influences skeletal muscle size and function. The general theme in the limited research is that fast/glycolytic skeletal muscle is more susceptible to impairments caused by autophagy deficiency compared to slow/oxidative muscle [23,30–32]. An early example of this is the constitutive skeletal muscle-specific atg5-deficient (catg5^−/−) mouse model that displays impaired lipidated LC3-II in both red quadriceps (RQ) and white quadriceps (WQ) [23]; however, other models of autophagy deficiency have found accumulation of SQSTM1 and ubiquitin in white gastrocnemius (WG) but not the SOL [31]. While the precise significance of autophagic variances between fast/glycolytic and slow/oxidative skeletal muscle remains uncertain, they may distinctly affect regenerative and remodeling processes, as well as skeletal muscle function. This is demonstrated by findings indicating a greater number of centralized nuclei in the EDL compared to SOL [23], along with functional impairments that were more pronounced and occurred earlier in EDL compared to SOL in autophagy-deficient mice [23]. Another group also observed a loss of mass across all fibers of the TA and gastrocnemius (GAS) in catg7^−/− mice [32,33]. Thus, the current literature suggests that fast/glycolytic muscle may be more sensitive to autophagy-deficiency caused by gene ablation; ultimately resulting in greater skeletal muscle damage and atrophy. This observation aligns with the understanding that rodent fast/glycolytic muscles exhibit a higher autophagic flux when compared to slow/oxidative muscles. It suggests that these fast/glycolytic muscles may depend on a heightened level of autophagic flux to sustain overall function. Consequently, the deficiency of autophagy genes in rodent skeletal muscle results in more pronounced impairments in fast/glycolytic muscles, underscoring the critical role of autophagy in maintaining their function and structural integrity. However, given the differential phenotypic, biochemical, biophysical, and morphological variability and plasticity between skeletal muscles and fibers, there is likely a differential autophagic response depending on the stressors (i.e., metabolic, injury, disease, etc.), thus requiring further consideration.

Metabolic stress-induced autophagy activation in different skeletal muscles

Metabolic stress in skeletal muscle can be induced in several ways; with two common methods being dietary restriction (i.e., fasting) and exercise. Fasting and exercise perturb numerous biochemical and metabolic signaling cascades that could influence the activation of autophagy. Earlier autophagy work has demonstrated that fasting and exercise predominantly work through the activation of AMP-activated protein kinase (AMPK) [34–36]; with lack of functional AMPK resulting in a depressed activation of autophagy in fasted and exercise conditions [34–36].

Fast/glycolytic skeletal muscle exhibits greater sensitivity to fasting-induced autophagy induction

Fasting-mediated activation of autophagy in skeletal muscle has been demonstrated for over two decades [29]. Initial work found that starvation leads to GFP-LC3 puncta accumulation between myofibrils and in the perinuclear regions of the mouse GAS after 24 h. Interestingly, a population of fibers within the same skeletal muscle shows fewer puncta, suggesting a differential response due to fiber type [29]. Moreover, the fast/glycolytic EDL accumulates GFP-LC3 puncta more rapidly (i.e., within 24 h) and to a greater extent compared to the more slow/oxidative SOL (i.e., moderate increase after 48 h) [29]. As a result, it was hypothesized that the difference in fasting-induced autophagy activation may be differentially regulated based on the fiber phenotype [29]. Rapid induction of autophagy signaling and autophagosome formation in fast/glycolytic muscles like the PLANT and TA have been identified as early as 6 h after food deprivation, but to a lesser extent in the slow/oxidative SOL muscle of mice and rats [22,37]. Furthermore, a greater upregulation of Lc3b, Atg7, and Sqstm1 is observed in mouse TA and GP compared to the SOL and DIA in response to food deprivation [22]. Although there is considerable remodeling in skeletal muscle fibers in response to short-term fasting; there is no evidence that suggests short-term fasting can significantly alter MYH composition. Overall, these data suggest that the starvation-induced autophagic response varies across skeletal muscles/fibers, with fast/glycolytic muscles/fibers generally demonstrating a more robust response. This suggests that a skeletal muscles/fibers metabolic properties related to inherent substrate type, utilization, and availability (or sensitivity to changes in these factors) are likely important with respect to the autophagy response during fasting.

Interestingly, human skeletal muscle data contradict the data from rodent studies, which may further suggest species-specific differences. For instance, following overnight (~10 h) fasting in humans, there is higher LC3B-II and lower LC3B-I in isolated type I fibers compared to isolated type II fibers from the VL [38]. Further analyses reveal no differences in SQSTM1, ULK1, ATG5 or ATG12 between isolated human type I and II fibers under fasted conditions [38]. It is important to note that measurement of LC3B in this instance does not fully reflect autophagic flux; thus, only providing a limited understanding of autophagic signaling in human skeletal muscle.

Skeletal muscle autophagic responses to acute and chronic exercise

The metabolic stress-induced autophagy activation by exercise exhibits some interesting responses in different skeletal muscles. Early evidence found elevated LC3-II and autophagosome formation after 1 h of treadmill running in mice [39]. This effect is more pronounced in the TA and VL (primarily type II glycolytic) but not the DIA, which consists of a high proportion of type II oxidative fibers [39]. In support of this, elevated skeletal muscle GFP-LC3 puncta is evident at 80 min of treadmill running in mice with the magnitude of increase being greater in the fast/glycolytic EDL and TA compared to the slow/oxidative SOL [40]. Further, the mouse SOL exhibits elevated phosphorylation of ULK1 S555 (a site proposed to initiate autophagy), LC3-I, and LC3-II, along with lower SQSTM1 at 120 min of exercise [41]. Conversely, mouse TA and GAS show consistent increases (2–5 fold) in LC3-II and concurrent decreases in SQSTM1 following 80–120 min of aerobic exercise [40,42,43]. Flux experiments using the autophagosome-lysosome fusion inhibitor colchicine found elevated autophagic flux in the TA immediately after exercise and following 90 min of recovery from exercise; however, the nature of this response is unclear in other skeletal muscles, particularly slow/oxidative muscles [44]. To complicate matters further, exercise-related responses need to be considered in the context of the skeletal muscle/fiber recruitment pattern, and factors such as type, duration, and intensity of the exercise performed.

Beyond the realm of acute exercise, it is challenging to conclude the effects of long-term training on skeletal muscle autophagy. Nonetheless, long-term training studies provide a valuable starting point for future studies. Four weeks of voluntary wheel running in mice results in autophagic adaptations in both fast/glycolytic (PLANT) and slow/oxidative (SOL) muscles, showcasing greater LC3-II, total LC3, and BECN1 [21]. Additionally, four weeks of voluntary running was accompanied by greater BNIP3 and reduced SQSTM1 in PLANT, whereas SOL exhibited greater ATG7 and SQSTM1 [21]. In contrast, three months of voluntary wheel running does not have a significant impact upon LC3-II or LC3-I in mouse DIA or TA [39]. Five months of voluntary wheel running does not alter the upstream signaling kinases, AMPK, FOXO3 (forkhead box O3), MAPK (mitogen-activated protein kinase), and AKT1 (AKT serine/threonine kinase 1) or autophagy-related proteins (ATG5 ATG7, BECN1, and SQSTM1) in rat PLANT [45]. However, long-term exercise results in increased LC3 puncta, total LC3, LC3-II, and BNIP3 in rat PLANT [45]. It is interesting to note that these changes in LC3 content correlate with the increase in the proportion of type IIA fibers (i.e., fast oxidative) observed with exercise training [45].

While acute exercise triggers some autophagic responses in human skeletal muscle in a manner analogous to rodents, some divergent and unique responses have been observed [27,46]. For instance, phosphorylation of AMPK T172 (active AMPK) only increased immediately following high intensity exercise, while ULK1 S317 phosphorylation increased immediately following low and high intensity exercise in the VL of humans [46]. Interestingly, this is accompanied by a reduction in LC3-II:I but not SQSTM1 immediately following low and high intensity exercise [46]. Similarly, 1 h of moderate intensity exercise elevates phosphorylation of AMPK T172 and ULK1 S555, as well as SQSTM1 in human VL; however, LC3B-II decreases after exercise and remains depressed [27]. These differences in human studies may stem from various factors such as training status, duration, intensity, recruitment patterns, location/depth of biopsy, among other factors. To date, all exercise studies investigating autophagy in human skeletal muscle have taken biopsies from the VL and have not examined other skeletal muscles. Thus, additional studies are warranted to better understand the autophagic response in exercised human skeletal muscle in general, but also in relation to skeletal muscle fiber type.

Alterations in skeletal muscle phenotype and autophagy in disease

Skeletal muscle phenotype is highly adaptable and can change in various conditions or disease states, leading to alterations in MYH composition, mitochondrial content, etc (Figure 2A). These changes are observed in conditions such as skeletal muscle injury, numerous myopathies, and select non-muscle specific diseases. Therefore, it is possible that observed changes in autophagy/mitophagy factor expression or flux are at least in part a consequence of a shift in skeletal muscle MYH composition or general muscle phenotype (e.g., organelle content), and not an altered autophagic signaling pattern due to the condition/disease per se. The complexity of MYH composition changes become evident in different scenarios. For instance, nerve injury in mice, such as denervation, initially results in the loss of all fibers and leads to a shift toward a type IIA phenotype in the long term [47]. Some groups have found significantly elevated autophagy proteins/genes and flux in denervated mouse skeletal muscles, including TA, EDL, and GAS [48,49]. Similarly, catastrophic skeletal muscle injury results in a dramatic loss of type IIA fibers and an increase in IIX and IIB fibers following 5 days of recovery [50]. While the alterations in autophagy flux in regenerating fibers is unclear, studies show impaired regeneration in whole body autophagy deficiency (i.e., atg16l1^-/-), but little to no differences in inducible muscle-specific models (i.e., iatg7^-/-) [50,51].

Figure 2.

Shifts in skeletal muscle phenotype and autophagy. (A) Illustration of skeletal muscle cross-section consisting of mixed fiber composition. Changes in MYH composition within the skeletal muscle, favoring a shift toward a fast/glycolytic or slow/oxidative phenotype, may be linked to alterations in autophagy and mitophagy. (B) Illustration of a skeletal muscle cross-section with associated changes in mitochondrial content, mitochondrial function, and metabolic activity. These changes, in turn, may influence autophagy and mitophagy signaling. Additionally, variations in mitochondrial morphology and network dynamics, such as fusion or fission (conceptually shown), could impact parameters like mitochondrial function, thereby influencing autophagy and mitophagy processes within the skeletal muscle.

In aging and diabetes, specific fibers are preferentially affected. For instance, aged humans and rodent models show a preferential reduction in type II fiber size and a shift toward a slower contractile phenotype [52–60]. Interestingly, there is elevated SQSTM1 and LC3-II in aged rat GAS, while increased autophagic flux in aged mouse TA and SOL [61,62]. Aging is accompanied by altered metabolic function, including reduced ATP synthesis rates, oxidative enzyme activity/content, total mitochondrial mass, and mitochondrial respiration [63–66], which may have an influence on autophagy and mitophagy signaling in aged skeletal muscle. In contrast to aging, diabetic patients experience a loss of type I fibers in the VL [67,68]; however, autophagic levels vary depending on the type of diabetes. For example, autophagy is directly related to insulin sensitivity [69], and as a result, insulin resistance (i.e., type 2 diabetes) is accompanied by suppressed SQSTM1, LC3-II, and ATG5 in VL of humans [70]. Furthermore, studies investigating a mouse model of insulin insufficiency (i.e., type 1 diabetes) find an elevated LC3-II:I ratio in the quadriceps (QUAD) muscle [71].

Diseases originating from non-skeletal muscle tissues, such as chronic obstructive pulmonary disease (COPD), chronic kidney disease (CKD), and hypertension also demonstrate a change in fiber type composition [72–79]. For instance, COPD patients have a higher proportion of type II fibers in the VL, which coincides with the accumulation of autophagosomes and elevated LC3-II:I [72]. Others have also found elevated autophagy and mitophagy protein levels (BECN1, ATG7, LC3-II, BNIP3, and PRKN) in the VL of COPD patients [76]. Similarly, CKD patients display greater LC3, SQSTM1, BNIP3, and PRKN in the VL [77]. This is consistent in a mouse model of CKD, which displays elevated Bnip3, Becn1, Atg12, BNIP3, LC3-II:I, BECN1, and ATG12 as well as reduced SQSTM1 in GAS; a response accompanied by a reduced proportion of type I and IIA fibers in the EDL [74,79]. In addition, hypertensive rats display several skeletal muscle phenotypic alterations. When comparing hypertensive to normotensive rats, autophagic protein and mRNA expression differences have been observed [75,78]; however, these differences vary across skeletal muscles and markers measured. For example, while no differences are observed in SQSTM1, BNIP3 is lower in hypertensive compared to normotensive rats in the WG, red gastrocnemius (RG), and SOL [75,78]. Further, ATG7 is elevated in WG and RG but not SOL, whereas LC3-I is only elevated in WG in hypertensive rats [75,78].

Changes in skeletal muscle fiber type are observed as a secondary consequence of other complex diseases, including cancer, Alzheimer disease, and amyotrophic lateral sclerosis. In patients and rodent models of cancer cachexia there is a documented loss in the size of type II fibers [80,81]. This alteration is closely associated with mitochondrial dysfunction [82] and an upregulation of autophagy molecules, such as LC3-II and SQSTM1, while mitophagy molecules BNIP3, BNIP3L, and PINK1 (PTEN induced kinase 1) show no changes, and PRKN was depressed in the rectus abdominis (RA) muscle from cancer patients [83]. Similarly, in rodent models of Alzheimer disease, reduced type IIB fiber size is observed in GAS, although it is unclear if there is a change in overall skeletal muscle fiber composition [84]. Loss of fiber size and strength is also accompanied by impaired state 3 mitochondrial respiration [85]; however, it remains uncertain whether autophagy signaling or flux is affected in these models. Furthermore, rodent models of amyotrophic lateral sclerosis show a decrease in type IIB MYH content, fiber number, and a shift towards a slower contractile phenotype [86,87]. Once again, this is accompanied by reduced BECN1 and autophagosome accumulation [88].

Myopathies are a classification of diseases that are typically genetic and result in muscle loss, weakness, and are associated with changes in various skeletal muscle characteristics. Duchenne muscular dystrophy is an X-linked myopathy that results in the degradation of all fibers with type II fibers being the first to be degraded [89–91]. GAS from a mouse model of Duchenne muscular dystrophy have reduced LC3-II, Becn1, Atg13, Bnip3, Fundc1, Pink1, and Ulk1 [92]. Myotonic dystrophy type 1 (DM1) displays a loss of type I fiber size and thus a higher proportional area of type II fibers [93], which is accompanied by increased LC3 in drosophila skeletal muscle [94]; however, a reduction in LC3-II, LC3-II:I, and SQSTM1 is observed in the VL of DM1 patients [95]. Pompe’s disease leads to a proportionate loss of type II fibers with disease progression and is accompanied by incomplete autophagic flux [31,96,97]. Furthermore, mouse models of Pompe’s disease exhibit augmented LC3-I and LC3-II in the SOL and WG [31]. Interestingly, the accumulation of autophagosomes and ubiquitinated proteins occurred in WG but not SOL [31,98]. Thus, these data demonstrate the importance of considering multiple skeletal muscles in the same disease state. Moreover, it is necessary to consider how changes in various skeletal muscle properties due to the disease (and potential differential changes across muscles) influence the autophagic profile observed. Although we have primarily focused on changes to MYH composition, it should be noted that several diseases/conditions display alterations in mitochondrial and metabolic function [99–101] as well as other characteristics; therefore, potentially impacting autophagy and likely more specific autophagic signaling pathways such as mitophagy (Figure 2B).

Key factors to consider in designing, measuring, and analyzing autophagy in skeletal muscle

Skeletal muscle diversity

Animal and human skeletal muscle is considerably different. First, humans have three MYH isoforms, while rodents like rats and mice have four [1,19,102–104]. Additionally, skeletal muscle fiber composition differs between small mammals like mice, which primarily have type IIX and type IIB fibers, while larger mammals like humans have predominantly type I and type IIA fibers [1,19,102–104]. Furthermore, there is variation in metabolic phenotype, with human skeletal muscles generally exhibiting lower oxidative potential than rodents [1,19,102–104]. Moreover, even for a specific skeletal muscle, there are notable differences in properties/characteristics across species. For instance, type I fibers constitute approximately 80%, 97%, and 30% of the total fibers within the SOL of humans, rats, and mice, respectively [1,105]. Similarly, while the most oxidative fibers in human skeletal muscle are type I fibers, in mice the most oxidative are type IIA fibers [1]. Additionally, within a species, significant variations exist in other parameters/characteristics across skeletal muscles, such as metabolic properties, even among fibers of the same MYH composition [1]. Moreover, individual and sex differences exist with respect to fiber composition and contractile characteristics [106–109]; and thus, it is crucial to exercise caution when studying and drawing generalizations regarding autophagic responses across various species and even between different skeletal muscles of the same species. Understanding this becomes especially crucial when investigating selective autophagy, such as mitophagy, as mitophagy-related proteins may be associated with mitochondrial content and function [21,22]. Furthermore, the separation of “red” and “white” muscle portions can help control for phenotypic and autophagic differences, particularly in large rodent skeletal muscles like QUAD and GAS. This approach is more challenging to implement in humans; however, single fiber approaches are possible [19,110,111].

Changes in skeletal muscle phenotype

As discussed earlier, autophagic gene/protein expression and flux in skeletal muscle is related to skeletal muscle phenotype, including MYH composition and metabolic profile. Further, numerous conditions and disease states display changes in skeletal muscle phenotype and are accompanied by altered autophagic gene/protein expression and flux. For example, exercise training can result in significant skeletal muscle adaptations, including alterations in MYH composition and metabolic profile [112,113]. From the earlier discussion, it remains uncertain to what extent the observed autophagic responses are directly attributed to autophagic-specific signaling adaptations, or to changes in the skeletal muscle phenotype. Moreover, different perturbations or disease states may result in differential phenotypic adaptations or result in adaptations in specific skeletal muscle properties. Regardless of the cause, it is crucial to recognize that the “autophagic phenotype” may now differ, and determining whether this change is a cause, or a consequence of the altered skeletal muscle phenotype is essential. It is crucial to emphasize that our intention is not to diminish the significance of the observed autophagic responses. Instead, we suggest that these responses might indicate a complex reaction, not solely attributable to condition-specific autophagic signaling. For example, phenotypic changes associated with exercise may affect highly adaptive organelles like the mitochondria, and thus may be accompanied by alterations in mitochondrial remodeling and thus influencing measures of mitophagy [114]. Finally, studies should consider conducting time-dependent experiments, as “response times” may differ across skeletal muscle/fiber types. Further, time-dependent studies could potentially allow for determining the skeletal muscle phenotype-autophagy temporal relationship during adaptation.

Skeletal muscle activation

Another crucial aspect to accurately interpret autophagy in skeletal muscle is the activation of fibers, particularly in the context of exercise and disuse. The diverse modes, durations, and intensities of exercise engage distinct skeletal muscle groups, fibers, and motor units, leading to varying metabolic responses that do not equally rely on anaerobic glycolysis or aerobic respiration. These differences in metabolic responses contribute to distinct activation of signaling cascades, influenced by parameters like ATP concentrations, calcium levels, ROS generation, and numerous other biochemical perturbations. For instance, running on a treadmill at 50% vs. 90% of VO₂ max can result in differential fuel utilization, as well as various other signaling cascades and thus may differentially influence autophagic or mitophagic activation. To complicate matters further, not all skeletal muscles and fibers will receive the “same activity stimulus” to a bout of exercise because of inherent muscle/fiber activation, recruitment, and fatigue patterns. Similarly, limb disuse may not uniformly affect all limb muscles/fibers due to differences in use and recruitment patterns. Considering that exercise and disuse induce significant skeletal muscle adaptation, these complexities necessitate additional muscle metabolic and phenotypic analysis to fully understand the autophagic responses in skeletal muscle.

Conclusion

In summary, the diverse nature of skeletal muscle, characterized by distinct fiber types and metabolic profiles, intricately influences cellular signaling pathways such as autophagy. As presented here, skeletal muscle autophagic responses and interpretation may be complicated by its interplay with MYH composition, metabolic profiles, and other skeletal muscle characteristics. Although there are undoubtedly additional factors to be considered, recognizing the relationships between skeletal muscle MYH composition, metabolic phenotype, and autophagy is essential for advancing both basic knowledge and downstream application. Ideally, researchers should consider providing a comprehensive representation of skeletal muscle phenotype, utilizing diverse criteria like MYH composition, oxidative/glycolytic potential, and mitochondrial content. Additionally, future studies should delve into autophagy at the fiber level and incorporate temporal experiments to elucidate the intricacies associated with changes in autophagy compared to the overall skeletal muscle phenotype. This nuanced understanding holds the key to targeted therapeutic interventions in skeletal muscle disorders, shaping the future of research in this field.

Funding Statement

The work was supported by the Natural Sciences and Engineering Research Council of Canada.

Disclosure statement

No potential conflict of interest was reported by the author(s).