The dependency of autophagy and ubiquitin proteasome system during skeletal muscle atrophy

Abstract

Among the four proteolytic systems in the cell, autophagy and the ubiquitin-proteasome system (UPS) are the main proteolytic events that allow for the removal of cell debris and proteins to maintain cellular homeostasis. Previous studies have revealed that these systems perform their functions independently of each other. However, recent studies indicate the existence of regulatory interactions between these proteolytic systems via ubiquitinated tags and a reciprocal regulation mechanism with several crosstalk points. UPS plays an important role in the elimination of short-lived/soluble misfolded proteins, whereas autophagy eliminates defective organelles and persistent insoluble protein aggregates. Both of these systems seem to act independently; however, disruption of one pathway affects the activity of the other pathway and contributes to different pathological conditions. This review summarizes the recent findings on direct and indirect dependencies of autophagy and UPS and their execution at the molecular level along with the important drug targets in skeletal muscle atrophy.

Introduction

Skeletal muscle atrophy causes a reduction in muscle mass marked by shrinkage of myofibers due to an enhanced rate of protein degradation over synthesis and is the functional cause of diseases/pathologies such as cachexia, sarcopenia, and generalized organ dysfunction (Fanzani et al. 2012). Out of the four known proteolytic systems, namely (i) the ubiquitin-proteasome system (UPS), (ii) the autophagy system, (iii) the calpains dependent pathway, and (iv) the caspase-dependent pathway, it is the autophagy and UPS system that performs the majority of proteolysis and is predominantly responsible for skeletal muscle atrophy. Macroautophagy (herein also termed autophagy) plays a crucial role in the turnover of myofibrillar proteins via the formation of autophagosomes to entrap protein aggregates or damaged organelles followed by their concomitant lysosomal degradation (Mizushima et al. 2011). In skeletal muscles, growth factors/insulin trigger PtdIns mediated activation of Akt activate mTOR (via TSC1/2) to stimulate p70^S6K/, PHAS-1/4E-BP1 thereby restraining autophagy (Bodine et al. 2001b). On the contrary, under ATP deficit conditions, AMPK impedes mTOR and stimulates FoxO3/, TSC1/2, and phosphorylates ULK1, thereby inducing autophagy. Moreover, in the fusion of autophagosome with the lysosome, two key membranous proteins LAMP2 and ESCRT are crucial; hence, decrease in ESCRT and LAMP2 activities is shown to disrupt the fusion of autophagosome and lysosome (however, the exact mechanism by which this occurs is not known) (Nakamura and Yoshimori 2017).

Short and soluble myoproteins are proteolyzed through UPS via the E1, E2, and E3 machinery (Morreale and Walden 2016). From the vast pool of E3 ligases, Atrogin-1/MAFbx and MuRF1 have been reported in skeletal muscle atrophy (Bodine et al. 2001a). Atrogin-1 is known to augment the degradation of key sarcomeric structural proteins like myosin, vimentin, desmin alongside some essential muscle transcription factors like MyoD and eIF3f (Xie et al. 2009). In muscles, Atrogin-1 is known to identify and ubiquitinate a specific motif, LXXXLL in the MyoD polypeptide (a key monogenic transcription factor) thereby promoting its degradation by the UPS machinery (Jogo et al. 2009). Parallel to this, MuRF1 regulates the half-life of myosin heavy chain isoforms, myosin light chains 1 and 2, myosin-binding protein C (Anderson et al. 2018), troponin I, and some other important structural proteins (Conraads et al. 2010).

Besides this, key factors like Akt, FoxO, NFκB, myostatin, Smad2/3, myogenin, JunB, and AMPK are also involved in multiple signaling cascades affecting the skeletal muscle mass (Piccirillo et al. 2014). Under physiological conditions, IGF-1 binds to its receptor to activate Akt, which phosphorylates downstream targets i.e., isoforms of FoxO proteins mainly FoxO1, FoxO3, and FoxO4 converting them into their inactive forms (Shukla 2014). This halts the expression of their target genes markedly MuRF1, Atrogin-1, USP14, Fbxo31, Fbxo30, Fbxo21, USP14, Ube4b, and some proteasomal subunits transcribing genes (Margolis et al. 2018). However, under energy-deficit conditions, increased AMP levels activate AMPK, which triggers FoxO3 via phosphorylating Akt-independent sites of FoxO3. AMPK-activated FoxO3 promotes the expression of Atrogin-1 and MuRF-1 that elevates proteasomal degradation of myofibers (Stefanetti et al. 2014). Glucocorticoids also trigger KLF15 to activate FoxO3. Additionally, KLF15 with REDD1 inhibits mTOR to promote skeletal muscle atrophy (Braun et al. 2013). Hence, glucocorticoids promote the expression of MuRF1 instead of Atrogin-1 via nuclear-translocated factor NF-kB. IL-6 and TNF-α trigger phosphorylation and proteasomal degradation of IkB (inhibitor of NF-kB) to set free janus kinase (JAK) and IkB kinase, respectively, to activate NF-kB (Peterson et al. 2011). Along with IL-6 and TNF-α, another key factor, TWEAK promotes skeletal muscle atrophy via NF-kB signaling. Under stress conditions, TWEAK binds on Fn14 receptor, which allows NF-κB and TRAF6 activation for MuRF1 and Atgs gene expression (Mittal et al. 2010). Moreover, myostatin-like factors also promote myofibrillar atrophy through the Smad2/3 pathway which induces the expression of Atrogin-1/MuRF1 (Lokireddy et al. 2011). Although it is still suspected that Smad2/3 with Smad4 (heterodimer) pathways affect the Akt–mTOR pathway to promote muscle atrophy (Trendelenburg et al. 2009).

E3 ligases like Atrogin-1 or MuRF1 are important for proteolysis and are degraded in an ATP-dependent manner via 19S regulatory particle (19S RP) and β1, β2, and β5 subunits in the core 20S subunit of 26S proteasome (Groll and Huber 2003; Da Fonseca et al. 2012). The 20S proteasome degrades target proteins by generating oligopeptides of length from 3 to 15 amino acid residues, which, in turn, are hydrolyzed to amino acids by oligopeptidases, found within the 26S proteasome (Nickell et al. 2009). Until recently, it was believed that the UPS and autophagy systems worked completely independently of each other without any point of crosstalk during skeletal muscle atrophy. However, recent studies have indicated the crosstalk/regulatory points between these two degradation systems, importantly ubiquitination (Kocaturk and Gozuacik 2018). In what follows, this review will attempt to summarize the major and key crosstalk points between autophagy and UPS.

Crosstalk factors between autophagy and UPS

Transcriptional factors FoxO, p53, and NF-kB

Several transcription factors have been identified to play important role in skeletal muscle dystrophies. FoxO3, a downstream target of Akt increases the expression of ULK1/2, Beclin1, Vps34, Atg12-Atg5, LC3, cathepsin L, etc. to stimulate autophagy, while Atrogin-1 and MuRF1 endorse ubiquitination of various muscular proteins for their subsequent proteasomal degradation (Fig. 1). Importantly, AKT phosphorylates FoxO family proteins that are ubiquitinated by MDM2 and finally degraded via 26S proteasome (Huang et al. 2015). During skeletal muscle atrophy induced by starvation, AMPK activates FoxO3 via phosphorylating at Ser413 and Ser588, thus promotes transcription of the above-described factors mandatory for both UPS and autophagy (Davila et al. 2012). AMPK also phosphorylates ULK1 to stimulate autophagy (Milan et al. 2015). Moreover, during fasting or cachectic conditions, the reticence of Akt pathway stimulates FoxO1/3 which increases the transcription of MAFbx and MuRF1 to promote degradation of myoproteins through UPS (Reed et al. 2012). FoxO1 is also essential for ECM modulation and to activate Akt by increasing RICTOR expression and inhibiting TRB3 expression (Chen et al. 2010). Hence, FoxO1 is an important player to regulate the modulation of adipose tissue and skeletal muscle mass to maintain energy homeostasis. Alike FoxO1 and FoxO3, caloric restriction limits sarcopenia by Akt phosphorylation that inhibits transcription of FoxO4, MAFbx, and MuRF1 (Sanchez et al. 2014). FoxO4 binds with insulin response elements (IREs) to activate transcription of IGFBP1 and downregulates HIF1A expression (Greer et al. 2007). A study reported that the energy-sensing AMPK can phosphorylate FoxO transcription factors at six regulatory sites, but it failed to report the phosphorylation of isotypes (Dogra et al. 2007).

Fig. 1.

FoxO3 protein upregulates various E3 ligases to promote UPS and autophagy promoting genes. FoxO3 itself was promoted via AMPK and inhibited via Akt as well as MDM2. Additionally, MDM2 also blocks the p53 through its ubiquitination and then proteasomal degradation in normal conditions. However, in stress, p14/p19 or ARF blocks the MDM2; now, p53 becomes free to acts as a transcription factor for autophagy promoting genes

Moreover, under physiologicalconditions, p53 levels are controlled by HDM2/MDM2 (E3 ligase)—proteasomal system. However, during stress conditions, pl4/pl9/ARF proteins associate with HDM2/MDM2, to inactivate it and indirectly stabilize p53 (Fig. 1; Table 1). Stabilized p53 activates transcription of downstream targets, i.e., ATG2, ATG4, ATG7, ULKI, BNIP3, and other stress or death-related genes (Crighton et al. 2006). While a fact that needs to be improvised is that during non-genotoxic stress conditions induced by either rapamycin/tunicamycin or nutrient deficit, p53 degradation is mediated by HDM2/MDM2 through proteasome (Kip and Van Veen 2015). The transcription factor p53 is also reported to regulate the activity of NF-κB. NF-κB and p53 inhibit each other to control the relative levels of transcribed genes. Exposure to UV light induces the p53 activation which, in turn, inhibits the NF-κB-induced gene expression. It is due to the competition of NF-κB and p53 to interact with the limited pool of transcriptional co-activator proteins p300 and CREB-binding protein (CBP) (Schneider et al. 2010). Hence, activation of NF-κB is limiting the activity of p53.

Table 1.

Genes or processes correlate UPS to autophagy in skeletal muscles

Symbol	Description	Functions between autophagy and UPS	Reference
CHIP (STUB1)	STIP1 homology and U-box containing protein 1	CHIP has a tetratricopeptide repeat and U-box domain to translocate ubiquitinated targets to UPS and autophagy.	Gómez et al. (2017)
BAG	BCL-2-associated athanogene	Interact with CHIP or Hsp70 to translocate proteins to UPS and autophagy.	Wang and Pessin, (2013)
TRAF6	TNF receptor-associated factor 6	TRAF6 upregulates LC3, Gabarapl1, Beclin1, MuRF, MAFbx to enhance autophagy and proteasomal deprivation of myoproteins.	Paul and Kumar, (2011).
Ubiquitin	Ubiquitin protein	Tag protein for selective autophagy and proteasomal degradation.	Lu et al. (2015), Korolchuk et al. (2010)
p62	SQSTM1	Common adaptor protein for lysosomal and proteasomal substrates, that correlates both systems in a compensatory manner.	Ji and Kwon (2017)
LC3	Microtubule-associated proteins light chain 3	Autophagosome marker having proteasome-dependent degradation.	Mizushima et al. (2011)
Beclin1	Mutual control mechanism	Protein promotes autophagy, however, it self-degraded via proteasomal system after the action of NEDD4 and RNF216 like E3 ligases.	Mizushima et al. (2011)
FoxO	Forkhead box O	Upregulates various E3 ligases to promote UPS and numerous Atg to enhance autophagy.	Reed et al. (2012), Sanchez et al. (2014)
p53	TP53 or tumor protein	In normal conditions, degraded via the proteasome, however during stress, acts as a transcription factor to stimulate autophagy promoting genes.	Crighton et al. (2006), Kip and Van Veen (2015)
NF-kB	Nuclear factor kappa-light-chain-enhancer of activated B cells	During stress, NF-kB acts as a transcription factor to upregulate the MuRF1, promotes UPS, and Beclin1 like atrogenes to endorse SMA via both proteolytic systems.	Bhatnagar et al. (2012)
TWEAK	Tumor necrosis factor-like weak inducer of apoptosis	Correlates UPS with autophagy via NF-kB	Bhatnagar et al. (2012)
ATF4	Activating transcription factor 4	Upregulating autophagy genes when the proteasomal system is inhibited.	Zhu et al. (2010)
Mul1	Mitochondrial ubiquitin ligase	Facilitate the ubiquitination of specific OMM proteins, for their proteasomal deprivation to persuade autophagy/mitophagy.	Yun et al. (2014)
Parkin	Parkin RBR E3 ubiquitin-protein ligase	Facilitate the ubiquitination of specific OMM proteins, for their proteasomal deprivation to persuade autophagy/mitophagy.	Kraft et al. (2010)
PINK1	PTEN induced putative kinase 1	Helps in the action of Parkin to correlate both UPS and autophagy.	Lazarou et al. (2012)
TLR4	Toll-like receptor 4	Activates p38 MAPK for autophagosome establishment as well as upregulation of Atrogin-1/MAFbx and MuRF1 to promote UPS.	Doyle et al. (2011)
TLR3	Toll-like receptor 3	Via TRIF/TBK1 and TRAM adapter proteins activate NF-κB to induce SMA via both systems.	Kawasaki and Kawai (2014)
TLR2	Toll-like receptor 2	TLR4 promotes proteasomal degradation of muscle proteins.	Sin et al. (2019)
TLR9	Toll-like receptors 9	After sensing mtDNA, TLR9 interacts with TLR4 to induce UPS and autophagy.	Delgado and Deretic (2009), Gough (2016)
AMPK	AMP-activated protein kinase	Enhances the expression of FoxO protein to correlate both systems.	Grumati et al. (2011)
KLF15	Krüppel-like factor 15	Promotes the FoxO3 expression to induce SMA via autophagy and UPS.	Braun et al. (2013).
REDD1	Regulated in development and DNA damage responses 1	REDD1 inhibits mTOR to induce autophagy.	Braun et al. (2013).
TNF- α	Tumor necrosis factor-α	Indirectly crosslinks UPS with autophagy through NF-kB activation.	Peterson et al. (2011)
IL-6	Interleukin 6	Indirectly crosslinks UPS with autophagy through NF-kB activation.	Peterson et al. (2011)
IL-1	Interleukin 1	Indirectly crosslinks UPS with autophagy through NF-kB activation.	Peterson et al. (2011)
Rpn1, Rpnl0, Rpn13	Mutual control mechanism	Proteasomal subunits, degraded via p62-mediated autophagy.	Nickell et al. (2009)
Ribophagy	Autophagy of ribosomes.	Stacked mRNA-protein complexes because of stress, degraded via both the proteasomal and lysosomal systems.	Kocaturk and Gozuacik (2018)

The proinflammatory cytokine IL-1, IL-6, TNF-α, TWEAK, and myostatin are the major factors that activate NF-κB and TRAF6 to promote skeletal muscle atrophy. Most of these factors have their renowned receptors but common downstream targets, i.e., NF-κB which is known to control autophagy and apoptosis (Zhu et al. 2017). Shreds of evidence show the crosstalks between NF-κB and autophagy . NF-κB is shown to stimulate the gene expression of proteins involved in autophagosome formation such as LC3, Beclin 1, ATG5, and BAG3-HspB8 complex. On the contrary, NF-κB is also reported to inhibit autophagy by stimulating its repressors, like Bcl-2, A20, phosphatases, and PTEN/mTOR or through suppression of BNIP3, JNK1, p53, and ROS. Hence, there is cross-regulation between autophagy and NF-κB, through IKK and NIK degradation process. That may be due to that the degradation of both IKK and NIK is largely through autophagy and not through UPS. Moreover, NF-kB is also vital for the expression of MuRF1 to promote proteasomal systems (Bhatnagar et al. 2012). There are numerous good reviews of ubiquitination in the NF-κB pathway (Chen and Chen 2013; Sun and Ley 2008); therefore, we summarized only important interlinking pathways. Overall, NF-κB activity is controlled not only by autophagy but also by UPS and p300 (Table 1; Fig. 8).

Fig. 8.

Systematic representation of linking markers between autophagy and ubiquitin-proteosomal system

E3 ligases TRAF6 and CHIP

E3 ligase, tumor necrosis factor receptor-associated factor 6 (TRAF6) plays an indispensable role in interlinking UPS and autophagy. Increased expression as well as activity of TRAF6 in cachectic, sarcopenic, and denervation models have been reported and related to skeletal muscle atrophy (Kumar et al. 2012). The catabolic fragments of degrading proteins stimulate TRAF6 E3 ligase. TRAF6 is required for optimal activation of NF-ĸB, AMPK, and JNK pathways required to upregulate the expression levels of LC3, Gabarapl1, Beclin1, MuRF, MAFbx, key components of autophagy, and UPS. Interestingly, TRAF6 also ubiquitinates Beclin 1, crucial for autophagy. Moreover, TRAF6 intermediates the apoptotic process of cells through the activation of caspase 8 (He et al. 2006). Alongside, TRAF6 mediates ubiquitination of catabolic polypeptides at Lys63 position to make them collective targets for selective autophagy. Evidentially, there is no appropriate synthesis of lysosomal and proteasomal components in muscle-specific TRAF6 knockout mice (Paul et al. 2010). Furthermore, TRAF6 deletion in cachexic or denervated conditions rescues regular distribution of myofibers via blocking autophagosome formation and proteasomal component expression (Paul and Kumar 2011). Additionally, the role of TRAF6 is also reported in toll-like receptor (TLR)-mediated cytokine production, wherein TLRs make a complex with TRAF6 to activate MAPKKK for the expression of cytokine genes (Walsh et al. 2015). However, ablation of TRAF6 can partially inhibit skeletal muscle atrophy (Paul et al. 2010), indicating the role of other E3 ligases. Hence, other partners E3 ligase or helping proteins need to be explored involved in muscle protein degradation to understand the proper molecular mechanism of TRAF6 in skeletal muscle atrophy.

Apart from TRAF6, another E3 ligase namely carboxy terminus of Hsc70 interacting protein (CHIP) along with BAG3 (the chaperon protein) degrades myoproteins and filamin C (a muscle protein found in the Z-line) via proteasomal as well as autophagosomal system. Filamins always experience unfolding and refolding cycles as required for muscle contraction, hence prone to quick degradation during the unfolded state. BAG3 present in the complex of Hsc70, HspB8, and CHIP binds with unfolded filamin C. N-terminal tetratricopeptide repeat domain of CHIP interacts with the molecular chaperones, through the C-terminal domain of Hsc/Hsp70 and Hsp90 to aid in their ubiquitination and proteasomal degradation. Whereas the C-terminal, U-box domain of CHIP acts as an E3 ligase to direct the filamin C toward selective autophagy (Gómez et al. 2017). CHIP ubiquitinates filamin C as well as BAG3 and p62 delivers them to the lysosome. Moreover, BAG1 makes a crosslink in-between the ATPase domain of Hsp70 and proteasome via its ubiquitin-like (UBL) domain, for translocation of Hsp70 to 26S proteasome, hence promotes the proteasomal breakdown of Hsp70 (Fig. 2). On the other hand, BAG3 endorses Hsp70, LC3, and p62 for selective autophagic degradation via recruiting the autophagic machinery to the target components (Wang and Pessin 2013). However, another E3 ligase ASB2β also ubiquitinates filamin B and degrades it through UPS in skeletal muscle at least during myogenesis (Bello et al. 2009). Hence, the programmed transcription of these E3 ligases seems to be directive which remains elusive. Moreover, factors that decide the selective degradation via UPS or autophagy of proteins having dual sites by CHIP remain elusive.

Fig. 2.

E3 ligase, CHIP has tetratricopeptide domain that directs target myoprotein degradation via UPS and U-box domain that directs substrate deprivation by the lysosomal system

Ubiquitination of proteins

E3 ligases like MuRF1, Atrogin-1, USP14, Fbxo21, Fbxo30, Fbxo31, Ube4b, etc. ubiquitinylates the protein at Lys 48, Lys 11, and Lys 29 to decide the translocation of proteins to the proteasome (Lu et al. 2015). The opponent E3 ligases like Parkin, Hul5, CHIP, and E6-AP ubiquitinylate at Lys 63 residue of proteins and direct the target protein/organelle like mitochondria, ER, ribosomes, and liposomes for selective autophagy (Korolchuk et al. 2010). Moreover, ATP deficit hampers the proteasomal degradation of ubiquitinylated proteins; hence, these proteins are translocated to macroautophagy through interactions with p62 or NBR1 (Ma et al. 2019). Though, in the case of skeletal muscle atrophy, with the increased level of E3 ligases, there is enhanced ubiquitination at specific positions of target myoproteins to guides their degradation via UPS or autophagosome-lysosomal pathways (Table 1).

The change in the ratio of E3 ligase isoforms MuRF2B and MuRF2A regulates the sequential activation of proteasomal or autophagic degradation of target components (Pizon et al. 2013). MuRF2B and MuRF2A both are required for autophagy whereas MuRF2B interacts with LC3 to promote autophagy. However, the absence of MuRF2B pushes MuRF2A to activate UPS and shift the proteolytic machinery (Pizon et al. 2013). Overall, p62, MuRF2A, and MuRF2B are key factors determining the switches between autophagy and UPS. It is also reported that monoubiquitinated targets are degraded in lysosomes while polyubiquitinated targets are degraded by the proteasome (Schreiber and Peter 2014). On the contrary, Pax3, an important transcription factor for myogenic differentiation, is reported to be monoubiquitinylated by Taf1 and degraded via UPS (Boutet et al. 2010). Hence, the extent of ubiquitination that decides the degradation path of these proteins and regulating factors during skeletal muscle atrophy needs further exploration.

Cargo protein like p62/SQSTM1/A170

Protein p62 contains three domains namely N terminal PhoxBEM1 (PB1) domain, LC3-interacting region (LIR), and C-terminus ubiquitin-binding domain (UBD). In skeletal muscles, the UBD of p62 interacts with Lys 63 ubiquitinated target proteins, and LIR with LC3b to translocate the targeted proteins into the growing autophagosome. However, the interaction of Lys 48 ubiquitin-conjugated target protein with the N-terminal PB1 domain of p62 promotes the proteasomal degradation of proteins in muscles (Fig. 3). Specifically, p62 has a slightly higher affinity for Lys 63 than the Lys 48 ubiquitinated target components; thus, p62 promotes autophagy in comparison with UPS (Ji and Kwon 2017). It is also reported that PB1 and TB domains of p62 stimulate the NF-kB and TRAF6 signaling, respectively; thus, p62 serves as a bridge between UPS and autophagy (Schwalm et al. 2015).

Fig. 3.

Components of adaptor molecule, p62, which establishes the correlation between proteasome (UPS) and phagophore (autophagy) through PB1 and LIR domains

Certain stress conditions such as starvation lead to proteasomal inhibition due to energy deprivation. Attenuated UPS promotes the phosphorylation of p62 at S405 and S409 through ULK1/Atg1 which enhances the protein translocation to autophagy. Here, phosphorylated p62 competes with the Nrf2 to bind with Keap1 and increase active Nrf2 levels. Interaction between Keap1 and phosphorylated p62 shift ubiquitinated aggregates towards autophagy. On the contrary, reduced autophagy and abundant p62 direct the protein substrates towards proteasomal degradation and promote cell survival (Liu et al. 2016). The competition between p62 and p97 also determines the fate of ubiquitinated target proteins in cells as over-expression of p97/VCP inhibits the binding of p62 with ubiquitinated targets and directs the substrates for proteasomal degradation (Korolchuk et al. 2010). Reduced expression of p97 also hinders FoxO3-induced muscle atrophy, by reducing overall protein turnover (Shi et al. 2011). Hence, p62 is a crosslinking protein, whose levels and interactions with other factors decide the fate of ubiquitinated proteins to be degraded by either autophagy or UPS.

Toll-like receptor coordinately activates UPS and autophagy

Till date, 12 TLRs have been recognized in mouse and 10 in human, with their foremost rolls in the immunological system. But in recent years, it is well reported that TLRs are also crucial in muscular atrophies via affecting or collaborating with the proteolytic systems of the cell. Lipopolysaccharide (LPS)-induced TLR4-mediated activation of p38 MAPK stimulates atrophy in C2C12 myotubes via promoting autophagosome establishment as well as upregulation of Atrogin-1/MAFbx and MuRF1. Furthermore, p38 MAPK inhibition or TLR4 knockout eliminates LPS-induced muscle atrophy. Thus, TLR4 mediates muscle degradation via the activation of both autophagosomal and proteasomal systems (Doyle et al. 2011). Moreover, HSP70 and HSP90 activated TLR4 and TLR2 were reported to upsurge muscular protein catabolism via the proteasomal system (Sin et al. 2019). Additionally, TLR3 and TLR4 are reported to persuade downstream signaling independently from MyD88, via TRIF/TBK1 and TRAM adapter proteins, those phosphorylate to activate NF-κB and induce autophagy (Kawasaki and Kawai 2014). The aforementioned studies showed that NF-kB tends to activate both proteasomal as well as autophagosomal systems; hence, TLR3 and TLR4 can induce autophagy and UPS indirectly in skeletal muscles (Table 1; Fig. 8).

Mitochondrial DNA (mtDNA) released in cytosol because of mitochondrial stress causes the activation of mtDNA sensing TLR9 and cyclic GMP-AMP synthase (cGAS) that promote IRF3-dependent signaling to induce autophagy (Riley and Tait 2020). Moreover, TLR9 is incapable to persuade autophagy independently, but in coordination with TLR4, it stimulates autophagy (Delgado and Deretic 2009; Gough 2016). A study also reported that TLR9 knockout mice experience more growth in muscle fibers due to the reticence of the SIRT1/Smad pathway. However, activated SIRT1/Smad signaling via TLR9 upsurges the muscular degradation in gastrocnemius muscle (Lyu et al. 2019). Thus, there may be some possibilities that through Smad pathway, TLR9 may affect skeletal muscle atrophy via both proteasomal and lysosomal systems, but it lacks the evidence (Fig. 4). Besides all, Duchenne muscular atrophy also shows the upregulation of TLR7 with its adaptor molecule MyD88 in myoblast cells (Henriques-Pons et al. 2014).

Fig. 4.

LPS-induced TLR4 endorses UPS and autophagy via p38MAPK. TLR4 in coordination with TLR9 after induction of mtDNA activates IRF3 to induce autophagy and SIRT1/Smad pathway to activate autophagy and UPS via FoxO. Moreover, Hsp70/90 activates TLR2–TLR4 complex to persuade UPS, while TLR3–TLR4 complex, via TRIF/TRAM, upregulates NF-kB to persuade both UPS and autophagy

Crosslinking of autophagy and UPS in organelle degradation

Mitophagy

Mitophagy is characterized by the partial or fully non-functional mitochondria engulfed in autophagosome and translocated to lysosomes for degradation. Mitochondria being the powerhouses of the cell serve the most crucial role of supplying the myofibers with the huge flux of ATPs required to perform the characteristic functions of skeletal muscles. Hence, mitophagy promotes myofiber degeneration due to increased ROS production in defective mitochondria. ROS activates AMPK that suppresses mTOR and promotes autophagy as well as UPS (Grumati et al. 2011). Importantly, in the case of in vitro and in vivo skeletal muscle atrophy induced by fasting or denervation, FoxO1/3 upregulates mitochondrial E3 ligase named Mul1, whose interactions increase the ubiquitination of outer mitochondrial membrane (OMM) proteins mainly mitofusin-2 for its degradation via the proteasomal system. Mitofusin isoforms Mfn1 and Mfn2 are vital for the proper maintenance of OMM. Ubiquitinated mitofusins and other OMM proteins are the inducers of mitophagy. However, these proteins are more prone to be degraded by the proteasomes and result in the formation of smaller fragments of mitochondrial subunits for autophagic destruction (Yun et al. 2014). The above-discussed mechanism of mitophagy through Mul1 is represented in Fig. 5.

Fig. 5.

In the case of mitophagy, Mul-1 and Parkin1-dependent ubiquitination of specific outer mitochondrial protein (mitofusins and VDAC) decides the degradation of mitochondria by autophagy or UPS

Additionally, Mfn1, Mfn2, membrane translocases (TOM70, TOM40, and TOM20), and voltage-dependent anion channel (VDAC) proteins are the substrates for Parkin (E3-ligase), thus increase mitophagy (Gegg et al. 2010). In normal conditions, PINK1 is translocated to mitochondria for post-translational modifications by proteases and its degradation via cytoplasmic proteasomes (Yamano and Youle 2013). However, during mitochondrial stress, PINK1 instead of being degraded accumulates on the OMM that recruits and activates Parkin onto OMM (Lazarou et al. 2012). Ubiquitination of proteins like Mfn1 and Mfn2 on the OMM via Parkin makes them susceptible to UPS degradation and promotes mitochondrial fission to facilitate engulfment of mitochondrial portions by autophagosomes (Kraft et al. 2010). Hence, mitophagy is dependent on the proteasomal system. Moreover, most ubiquitinated OMM proteins like VDAC and TOM are recognized by the p62 and make them susceptible to mitophagy/autophagy (McLelland et al. 2014). Some evidence showed that suppression of Mul1 partially reduces muscle wasting even after response to various types of muscle atrophy stimuli (Lokireddy et al. 2012).

Ribophagy and pexophagy

Under different stress and cachexia conditions, ribosomes undergo disintegration and subsequent degradation via autophagosome known as ribophagy. Proteins like NUFIP1-ZNHIT3 directly tag ribosomes to LC3, hence promote ribophagy (Wyant et al. 2018). However, in yeast and mammalian cells, under ER stress conditions, primarily ubiquitination of ribosomal subunits and p97/VCP-mediated translocation lead these for proteasomal degradation (An and Harper 2018). Importantly, the clearance of mRNA-protein complexes that are stuck in the translational process due to stress activates both the proteasomal and lysosomal systems (Kocaturk and Gozuacik 2018). However, the exact mechanism behind ribophagy in the case of skeletal muscle atrophy still needs to be explored. Additionally, ribosome biogenesis initiated during denervated condition activates the mTOR and prevents the soleus muscle atrophy (Table 1; Fig. 6), but the exact approaches behind these activation/inhibitions are still contradictory (Ossareh-Nazari et al. 2010).

Fig. 6.

In skeletal muscles, ribosomes are connected with LC3 via NUFIP-1 and ZNHIT3 to induce autophagy, and also, ribosomes are ubiquitinated to be translocated to proteasome via unknown factors. Moreover, stacked mRNA-protein complexes because of stress degraded via both the proteasomal and lysosomal systems

Peroxisome undergoes degradation under stringent conditions through autophagic-mediated machinery termed pexophagy. Peroxisomes express a specific receptor called peroxisomes proliferator activator receptors (PPARs) in different isoforms essentially required for the regulation of lipid metabolism, glucose homeostasis, regulation of inflammatory responses, etc. In skeletal muscles, PPARγ is reported to act as an E3 ubiquitin ligase that physically interacts with the homology domain region of NF-κB and p65 to induce its Lys28 and Lys48-linked ubiquitination and degradation. Moreover, an E3 ubiquitin ligase, neural precursor cell expressed developmentally downregulated protein 4 (NEDD4) was found to interact with PPARγ, which increases PPARγ stability by inhibiting its proteasomal degradation and regulating adipogenesis. Additionally, PPARβ regulates the metabolic function and numerical abundance of peroxisomes via Pex11β in parallel to osteoblast differentiation, showing the importance of PPARs in the maintenance of peroxisomes in skeletal muscles or osteoblasts (Qian et al. 2015). However, a unified pathway that clearly defines the role of peroxisomes and their connecting link with autophagy or other proteolytic systems during skeletal muscle atrophy still needs to be explored.

The degradation of membrane-less organelles

Membrane-less organelles are cytoplasmic assemblies of biological macromolecules and important for several fundamental processes of the cell, including various biochemical reactions as well as RNA and protein transport. Synthesis of membrane-less organelles (MLOs)/non-membrane bound compartments inside the cell, separated from the liquid cytoplasm, occurs through liquid-liquid de-mixing or liquid-liquid phase separation (LLPS) process (Pancsa et al. 2019). Important examples of MLOs are centrosomes (which nucleate microtubules), nucleoli (which make ribosomes inside the nucleus), processing bodies (promyelocytic leukemia nuclear bodies), Cajal bodies (which make spliceosomes), P granules, and stress granules (Decker and Parker 2012). Skeletal muscle differentiation or myoblast to myotubes formation is regulated by the microtubule-organizing center (MTOC) by switching MTOC from a centrosome to a non-centrosomal location. During this cell cycle phase, two RING-family E3 ligases, i.e., anaphase-promoting complex/cyclosome (APC/C) and the Skp1/Cullin/F-box (SCF) proteins, are the key regulatory ligases that promote proteasomal degradation of the centrosome. The centrosome is the complex aggregate of proteins, and the aggregated proteins may bind to the 19S regulatory complex and resist the proteolysis in the 20S. UPS is involved to remove ubiquitin chains from these aggregates and triggers the autophagy for the clearance of these aggregates (Ben-Nissan and Sharon 2014). Hence, these aggregates are thought to be digested through autophagy instead of UPS. Both UPS and autophagy are complementary to each other for degradation of LLPS or MLOs. LLPS is also observed in mTORC1 regulation during cellular stress. LLPS regulates mTORC1 dissolution from stress granules mediated via the active DYRK3 kinase (Wippich et al. 2013). Under acute hyperosmotic stress, it was disclosed that proteasome-containing nuclear foci transient structures also have ubiquitinated proteins, p97, RAD23B (a substrate-shuttling factor for the proteasome), and multiple proteasome-interacting proteins. In mechanistic terms, proteasomal degradation followed by LLPS is triggered by interactions of ubiquitin chains with two ubiquitin-associated domains of RAD23B (Yasuda et al. 2020).

The prevalent view is that ubiquitinated target proteins are aggregated by the p62 scaffold protein, and these aggregates are then engulfed in autophagosome for autophagic degradation. Hence, p62 forms droplets in vivo having liquid-like properties; however, in vitro addition of Lys 63 polyubiquitinated chains to p62 induces its phase separation and results in enhancement of ubiquitin signals in p62 droplets. Mixing of recombinant p62 with p62 knockout cells resulted in polyubiquitination-dependent p62 phase separation. Thus, p62 phase separation depends on the interaction between p62 and polyubiquitinated chain, triggering autophagic cargo concentration, and autophagic degradation (Khaminets et al. 2016). Additionally, in vitro studies show that unconjugated LC3 is recruited into the p62 droplets while it is recruited in p62 bodies during in vivo conditions. Evidently, in vivo, mutant p62 proteins without phase separation process also show impaired p62 body formation and autophagic degradation (Sun et al. 2018). Moreover, cell stress efficiently activates the formation of stress granules in proliferating and differentiated skeletal muscle cells, shown by the presence of distinct TIA-1 and DDX3-containing foci in the cytoplasm (Ravel-Chapuis et al. 2016).

Under various stress conditions like exposure to an environmental toxin and carcinogen, arsenic, stress granules consist of cytoplasmic assemblies of mRNPs are stalled in translation initiation. Besides, arsenite stress also triggers the accumulation of ubiquitinated defected ribosomal proteins in stress granules, which are rapidly degraded by the 26S proteasome during recovery. The proper molecular mechanisms of stress granule synthesis and turnover are not clearly understood. However, AN1-type zinc finger protein 1 (ZFAND1) acts as an evolutionarily conserved regulator of stress granule clearance. ZFAND1 interacts with the p97 and 26S proteasome to recruit them to arsenite-induced stress granules. During recovery from arsenite stress, along with p97 and 26S proteasome, stress granules dissociate in a process requiring the chaperone complex HSPB8-BAG3-HSP70 (Rodriguez-Ortiz et al. 2016). Evidently, in ZFAND1 knockout conditions, arsenite-induced stress granules lack the 26S proteasome and p97, while the defected proteins accumulate in the close locality to stress granules. These stress granules are recruited into LC3 of autophagosomes by p62 for their autophagosomal-lysosomal degradation (Turakhiya et al. 2018). Additionally, homeostasis of stress granules is recently proposed to be a chief pathogenic mechanism in various neuro-muscular degenerative disorders (Ramaswami et al. 2013). From the above-described shreds of evidence/findings, we can conclude that the degradation of MLOs, mainly of stress granules, is done by autophagy as well as UPS.

Mutual dependency of UPS and autophagy

Interdependency of autophagy and UPS

From previous studies, it is known that in skeletal muscles, both the concerned proteolytic systems are interconnected. It is reported that inhibition of one system affects the functioning of the other. Inhibition of proteasome by the inhibitors like bortezomib, carfilzomib, and oprozomib in carcinomas and melanoma cells is reported to increase ATF4 levels that promote the expression of ATG5 and ATG7 to promote autophagy (Zhu et al. 2010). Most proteasome inhibitors stabilize p53 and activate AMPK by reducing GSK- β activity, activating unfolded protein response (UPR), phosphorylating eIF2α, and increasing the conversion of LC3-I (free form) to LC3-II (PE-conjugated form), hence increases autophagy (Zhu et al. 2010). Similarly, inhibition of autophagy activates the proteasomal system. In colon cancer cells, the knockdown of ATG genes by siRNA resulted in the upregulation of proteasomal subunits (Wang et al. 2013). Moreover, inhibition of autophagy in Atg7 null muscles increased the levels of Atrogin-1 and MuRF1 along with polyubiquitinated targets (Masiero and Sandri 2010). On the contrary, knockdown of ATG7 and ATG12 in HeLa cells was reported to inhibit autophagy along with proteasomal activity despite the accumulation of ubiquitinated proteins and p62 via Nrf1-dependent pathways (Sha et al. 2018). These accumulated p62 proteins compete with other adaptor proteins like p97 for ubiquitinated proteins, hence delays the availability of substrate for the proteasome. Consistently, p62 knockdown cells salvaged the levels of UPS substrates in autophagy-deficit cells (Korolchuk et al. 2010).

Importantly, in skeletal muscles, to determine the dependency of autophagy on the muscle-specific E3 ligases, MuRF1 and Atrogin-1 knockout mice were transfected with GFP-LC3, and after 7 days, the mice fasted for 24 h. Here, autophagosome formation was unaffected during the loss of E3 ligase genes (Lindsten et al. 2003). Moreover, to determine whether autophagy is dependent on the activity of the proteasome, the proteasome inhibitor MG262 was used. Here, the initiation of GFP-LC3 labeled autophagosomes by FoxO3 was unaffected (Lindsten et al. 2003). Also, Masiero and Sandri generated inducible Atg7 knockout mice, to block autophagy in skeletal muscles. The Atg7 null muscles disclosed unexpected phenotypes, characterized by the presence of protein aggregates, abnormal mitochondria, sarcoplasmic reticulum distension, oxidative stress, apoptosis, and enhanced skeletal muscle atrophy (Masiero and Sandri 2010). Moreover, during fasting and denervation conditions, inhibition of autophagy does not protect muscles from atrophy but instead promotes greater loss of skeletal muscles. Importantly, proteasome function, studied in vivo by expressing a fluorescent proteasomal substrate, is not compromised in Atg7 null muscles (Masiero and Sandri 2010). Hence, the relation between these two systems might be different in cell types and conditions. Moreover, contradictions need further exploration. Skeletal muscle atrophy being multifactorial dependent is more imperative to find out controlling signaling pathways with their interconnections and modulations during diverse atrophic conditions (Fig. 7).

Fig. 7.

Interdependency of autophagy and UPS. Inhibition of proteasome promotes autophagy, but inhibition of autophagy can promote as well as inhibit the proteasome in condition-dependent manner

Mutual degradation of UPS or autophagy components

It has been perceived in skeletal muscles that numerous autophagic proteins are degraded via the proteasomal system, while the whole proteasome undergoes autophagic degradation (Fig. 8). Studies also revealed that deprivation of 26S proteasomes occurs with the help of lysosomes via TOLLIP and p62-regulated proteaphagy (Lu et al. 2014). In higher animals, conditions similar to amino acid starvation endorses ubiquitination of various 19S proteasome subunits like Rpn1, Rpnl0, Rpn13 to promote their degradation through p62-mediated autophagy (Cohen-Kaplan et al. 2017). On the contrary, Beclin1 is degraded via the proteasome through ubiquitination by E3 ligases like NEDD4 and RNF216 (Platta et al. 2012). It is worth mentioning that in nutrient-rich conditions, Cullin-4 (E3 ligase) endorses the ubiquitination of AMBRA1 for degradation via the proteasomal system. Moreover, LC3 protein is also degraded by the proteasome in a ubiquitin and ATP-independent manner (Gao et al. 2010). These facts noticeably establish that there are crosslinking points in between UPS and various types of autophagy. However, the interactions among the signaling pathways are required to understand the response of UPS and autophagy during skeletal muscle atrophy.

Conclusion

The UPS system has been largely considered as the principal player in the generation of skeletal muscle atrophy. However, without the additional process of autophagy, it is impossible to scavenge the cell and organelle debris after protein breakdown. Recent studies have shown that UPS and macroautophagy have various points of coordination and crosstalk. These crosslinking points have been explored in various disease states and under normal conditions, but so far they have not been critically examined for the situation of skeletal muscle atrophy which is linked with the pathologies of cachexia, sarcopenia, and immobilization. Crosstalk between these two systems is very important in skeletal muscle due to the fact that muscle units are fused together and therefore not able to be digested alone by either autophagy or UPS. Hence, crosstalk between the UPS and autophagy systems is crucial and needs to be understood to allow pharmaceutical intervention into skeletal muscle atrophy. Selective ubiquitination at positions Lys-48 and Lys-63 of target proteins, respectively, act as a signal for degradation of target proteins via the proteasome and autophagy system. In some cases, ubiquitination is specific, but in the majority of cases, it is not known whether this ubiquitination is a random process or specific to the target protein. Hence, the ubiquitination of other proteins important in skeletal muscles needs to be studied. Moreover, both ubiquitination-specific domains, i.e., tetratricopeptide repeat and U-box domains specific for UPS and autophagy, are present in some proteins. However, it remains to understand what factors dictate their degradation route. Other molecules like FoxO1/3 and NF-kB represent an important link between autophagy and the proteasomal system because of their ability to stimulate both autophagy and UPS promoting gene transcription. Such findings indicate the interdependency of both systems. There are cases where UPS is essential to start autophagy and an excellent example is mitophagy where Mul1 or Parkin1 is required to induce ubiquitination of mitofusin-2 followed by its subsequent degradation via the proteasome, thereby initiating mitophagy. Skeletal muscles are very rich in mitochondria. Hence, mitophagy is a major process in skeletal muscle atrophy that requires coordination of UPS and autophagy. Beclin1 and LC3 are important proteins for autophagy, which are first degraded by UPS; indeed, the proteasomal system itself is degraded by autophagy. However, the factors which decide and promote their degradation rate are not well known. Moreover, studies that indicate inhibition of one system affects the activity of another one need to be explored using in vivo models. Apart from the above-discussed crosslinking points, we have also raised the point in this review that TLRs also relate UPS function to autophagy. Here, TLR4 both independently and in conjunction with other TLRs (like TLR2, TLR3 and TLR9) act to correlate both systems to more tightly control skeletal muscle atrophy. Recent studies have shown that both autophagy and UPS degrade membrane-less organelles that are formed by LLPS (also known as biomolecular condensates or liquid droplets) via an unclear molecular mechanism(s), which need further exploration. Most studies regarding crosstalk and their interdependent molecular mechanism are not well reported in skeletal muscle atrophy. It is a well-established fact that skeletal muscle atrophy is one of the major contributors to human mortality and morbidity. Therefore, interventions in skeletal muscle atrophy are of high therapeutic value to retard or block mortality and morbidity. Hence, it is crucial to reveal the interdependency and crosstalk points between UPS and autophagy to better understand the interdependency of signaling pathways and to develop drugs or compounds to treat skeletal muscle atrophy.

Abbreviations

Akt/PKB

protein kinase B

AMBRA1

activating molecule in BECN1-regulated autophagy

AMPK

AMP-activated protein kinase

ARF

ADP ribosylation factor

ATF4

activating transcription factor 4

Atg

autophagy-related genes

ATP

adenosine triphosphate

BAG3

Bcl-2-associated athanogene-3

BNIP3

BCL2/adenovirus E1B 19-kDa protein-interacting protein 3

CHIP

C-terminus HSP-70 interacting protein

DYRK3

dual specificity tyrosine-phosphorylation-regulated kinase 3

eIF2α

eukaryotic initiation factor 2 α

ERK

extracellular signal-regulated kinase

ESCRT

endosomal sorting complexes required for transport

Fbxo21

F-box protein 21

Fn14

fibroblast growth factor-inducible molecule 14

FoxO3

forkhead box O3

GAIP

G-α interacting protein

Hip

hsc70-interacting protein

Hsp

heat shock protein

IkB

inhibitor of NF-kB

IL-6

interleukin 6

IRF3

interferon regulatory factor 3

JAK

Janus kinase

JunB

transcription factor Jun-B

KLF15

Krüppel-like factor 15

LAMP

lysosomal-associated membrane protein

LC3

microtubule-associated protein 1A/1B-light chain

LIR

LC3-interacting region

LPS

lipopolysaccharides

Lys

lysine residue

MAPK

mitogen-activated protein kinase

MDM2

mouse double minute 2 homolog

Mfn

mitofusin

mtDNA

mitochondrial DNA

mTOR

mammalian target of rapamycin complex 1

MuRF1

muscle RING finger protein 1

MyD88

myeloid differentiation primary response 88

NBR1

neighbor of BRCA1 gene 1 protein

NEDD4

neural precursor cell expressed developmentally downregulated protein 4

NF-κB

nuclear factor kappa-light-chain-enhancer of activated B cells

Nrf2

nuclear factor erythroid 2 (NFE2)-related factor 2

p62

sequestosome 1

PB1

N terminal PhoxBEM1

PtdIns

phosphatidylinositol 3 phosphate

RAD23B

UV excision repair protein RAD23 homolog B

RAF

rapidly accelerated fibro sarcoma

REDD1

regulated in development and DNA damage responses 1

Rpn

regulatory particle of non-ATPase

RSK

ribosomal S6 kinase

Smad

small mothers against decapentaplegic

TIA-1

T cell intracellular antigen-1

TLR

toll-like receptor

TNF

tumor necrosis factor

TOLLIP

toll interacting protein

TOM

the outer membrane

TRAF6

tumor necrosis factor receptor associated factor 6

TRAM

transverse rectus abdominis myocutaneous flap

TRIF

TIR-domain-containing adapter-inducing interferon-β

TSC1

tuberous sclerosis protein 1

TWEAK

tumor necrosis factor-like weak inducer of apoptosis

UBD

C-terminus ubiquitin binding domains

Ube4b

ubiquitination factor E4B

ULK

Unc-51 like autophagy activating kinase

UPS

ubiquitin proteasome system

USP14

ubiquitin-specific protease 14

Vps

vacuolar protein sorting

ZFAND1

zinc finger AN1-type containing 1