Protein homeostasis in LGMDR9 (LGMD2I) – The role of ubiquitin–proteasome and autophagy–lysosomal system

Limb‐girdle muscular dystrophy R9 (LGMDR9) is an autosomal recessive disorder caused by mutations in the fukutin‐related protein gene (FKRP), encoding a glycosyltransferase involved in α‐dystroglycan modification. Muscle atrophy, a significant feature of LGMDR9, occurs by a change in the normal balance between protein synthesis and protein degradation. The ubiquitin–proteasome system (UPS) and autophagy–lysosomal system play a key role in protein degradation in skeletal muscle cells, but their involvement in the pathology of LGMDR9 is still largely unknown. We have aimed at clarifying whether proteolysis through the UPS and the autophagy–lysosomal pathway is dysregulated in LGMDR9 patients.


INTRODUC TI ON
Limb-girdle muscular dystrophy R9 (LGMDR9, also known as LGMD2I) is an autosomal recessive disorder caused by mutations in the gene-encoding fukutin-related protein (FKRP) [1]. FKRP is a ribitol-5-phosphate transferase, which participates in the glycosylation of α-dystroglycan, an important protein link between the extracellular matrix and the cytoskeleton. FKRP deficiency causes insufficient α-dystroglycan glycosylation, resulting in dysfunction of laminin binding and instability of the dystrophin-glycoprotein complex in skeletal muscle [2][3][4]. Genetic variants in the FKRP gene cause a broad spectrum of phenotypes ranging from mild clinical symptoms in LGMDR9 to more severe symptoms seen in congenital muscle dystrophy (MDC1C), Walker-Warburg syndrome (WWS) and muscle-eye-brain disease (MEB), which are also associated with severe brain defects and mental impairment [1,5]. Dilated cardiomyopathy and ventilatory impairment are common clinical features in these patients [6,7].
The FKRP variant c.826C>A (p.Leu276Ile) is the most frequent cause of LGMDR9 in patients from Northern Europe [8,9]. The spectrum of LGMDR9 phenotypes ranges from mild calf and thigh hypertrophy with disease onset in the second or third decade to early onset Duchenne-like disease with loss of ambulation in the teens [10]. Prominent proximal muscle weakness and atrophy are significant clinical signs in LGMDR9 [6,11]. However, the molecular basis underpinning atrophy in LGMDR9 is lacking.
Muscle atrophy occurs by a change in the normal balance between protein synthesis and protein degradation. There are two major proteolytic systems contributing to protein degradation in skeletal muscle -the ubiquitin-proteasome system (UPS) and the autophagy-lysosomal system [12].
UPS degradation is achieved through peptide hydrolysis by the 26S proteasome [13]. The E3 ubiquitin ligases Atrogin-1 (also called MAFbx) and MuRF1 (also known as TRIM63) are important biomolecular markers of the UPS degradation pathway. However, although they have been found to be significantly upregulated in several rodent and human models of atrophy [14,15], recent reports have indicated that Atrogin-1 and/or MuRF1 might not be chronically upregulated in humans during muscle wasting (reviewed in [16,17]).
Macroautophagy (hereafter referred to as autophagy) is a highly conserved homeostatic process for lysosome-mediated degradation of cytoplasmic components, including damaged or dysfunctional organelles and toxic protein aggregates [18]. Autophagy can be organelle selective, as demonstrated with mitochondria (mitophagy) [19].
In autophagy, double-membrane-bound vesicles known as autophagosomes sequester cellular components and organelles and fuse with lysosomes where the content of the vesicles is degraded [20].
A certain basal level of autophagy is required for the maintenance of protein homeostasis and muscle mass. Muscle wasting is associated with increased autophagy with excessive autophagosome formation. Suppression of autophagy, on the other hand, results in accumulation of protein aggregates and abnormal organelles, leading to myofibre degeneration [21][22][23][24].
Dysfunction in the UPS and/or autophagy has been reported in various muscular dystrophies (reviewed in [25,26]). However, it is currently unknown if proteolysis through the E3 ubiquitin ligases, Atrogin-1 and MuRF1, and/or the autophagic lysosomal pathway is dysregulated in LGMDR9 patients. Therefore, the purpose of this study was to determine the expression of key markers of UPS, and furthermore, autophagy and its regulatory pathways in LGMDR9 muscle. Skeletal muscle sections were investigated by immunoblot, TEM and indirect immunofluorescence. Protein expression data were correlated with the age of onset, disease duration at biopsy or the level of atrophy in LGMDR9 patients to reveal potential clinical significance.

Muscle biopsies
Muscle specimens from 12 patients with LGMDR9 (6 males and 6 females, average age at biopsy 25.58 years, range 9-51 years) and from 8 normal controls (4 males and 4 females, average age at biopsy 25.05 years, range 42-65 years) were used in this study. Muscle biopsies were obtained from the vastus lateralis, flash frozen in isopentane, cooled in liquid nitrogen and then stored at −80°C for further analysis. All patients affected by LGMDR9 harboured the c.826C>A/ c.826C>A FKRP genotype and were subjected to investigation in the previous studies [9,27,28]. Information concerning age at onset of disease and disease duration was collected via a questionnaire as explained in [9] (Table S1). Additional information about patients can be found in [28].

Statistical analysis and data presentation
All data are presented as means ±SEM, plotted with GraphPad

UPS degradation in LGMDR9 muscle
The E3 ubiquitin ligases Atrogin-1 and MuRF1 are important biomolecular markers of the UPS degradation pathway and have been shown to be upregulated in multiple physiological settings of atrophy [13]. To investigate if this would also be the case in LGMDR9, protein levels of Atrogin-1 and MuRF1 were estimated by immunoblot analysis in skeletal muscle from patients and normal controls.
Results obtained from LGMDR9 muscle biopsies did not show any significant difference in the expression of these two markers with respect to controls ( Figure 1A,B). Moreover, expression levels of Atrogin-1 and MuRF1 did not correlate with the extent of atrophy, age at onset or duration of symptoms in LGMDR9 patients (data not shown). We also assessed the levels of protein ubiquitination as well as α5 (PSMA5) and β5 (PSMB5) subunits of 20S catalytic core of 26S proteasome. Western blot analysis revealed an increase in total ubiquitin content in LGMDR9 muscle biopsies as compared to controls. This was accompanied by a significantly increased expression of β5 (PSMB5) and a trend towards increased expression of α5/ PSMA5 (p = 0.1). Overall, levels of K48-linked ubiquitination did not differ between LGMDR9 biopsies and controls, however, variability in band patterns and intensities was observed between the two groups ( Fig. S1).

Autophagy in LGMDR9 muscle
A number of studies have indicated abnormal regulation of autophagy in various forms of muscular dystrophy (reviewed in [25]). To investigate if the autophagy-lysosomal degradation pathway is dysregulated in LGMDR9, protein expression levels of autophagy markers Atg7, Atg12, Beclin-1, LC3B and p62 were assessed in muscles from LGMDR9 patients and normal controls.
Immunoblot analysis showed significant increase in Atg7 expression in LGMDR9 patients as compared to normal controls, however, no difference in the levels of Atg12 and Beclin1 was observed. The lipidated form of LC3B, LC3B-II, which is generated during autophagosome formation, was highly increased in LGMDR9 muscle, whereas the level of p62, a protein known to be incorporated into autophagosomes and efficiently degraded during autophagy, was significantly decreased. Altogether, the results indicate an activation of autophagy in LGMDR9 muscle (Figure 2A,B).
With the exception of LC3B-II, which showed an inverse correlation with age at onset, there were no obvious relationships between the level of atrophy, age at onset or duration of symptoms in LGMDR9 and the expression levels of the above-mentioned autophagy markers (data not shown).

LC3B/p62/Lamp2 colocalization in LGMDR9 muscle
Increased levels of LC3B-II in skeletal muscle from patients with LGMDR9 could be caused either by autophagy induction or by failure in autophagosomes degradation. During functional autophagy, autophagosomes fuse with lysosomes to form autolysosomes, followed by degradation of autophagosomal cargo. Therefore, in order to gain insight into autophagosome-lysosome fusion in LGMDR9 muscle, the abundance of the common lysosomal marker LAMP2 and its intracellular localization relative to LC3B were investigated. Furthermore, since p62 is a resident of autophagosomes, it was also of interest to verify its localization relative to LC3B. No differences in the levels of LAMP2 were observed between controls and LGMDR9 samples upon immunoblot analysis ( Figure 3A,B).
Immunofluorescence analysis demonstrated that LC3B-, p62-and LAMP2-positive signals were far more abundant in LGMDR9 muscles as compared to control muscle. LC3B co-localized with both p62 and LAMP2 in LGMDR9 muscle, and to a higher extent than in control muscle ( Figure 3C,D). This suggests that the content of autophagosomes/autolysosomes in LGMDR9 is increased.  Figure 4A-E, Fig. S2A,B). In addition, fibres contained concentric multi-lamellar structures, highly indicative of an ongoing process of autophagy in LGMDR9 muscle (Fig. S2C,D).
Furthermore, morphologically altered mitochondria and dilated sarcoplasmic reticulum (SR) were observed in LGMD2R9 muscles ( Figure 4D-G, Fig. S2E-H). Defective mitochondria can be selectively degraded in the process of mitochondrial autophagy (mitophagy) [29,30]. Therefore, expression of mitophagy markers, proteins that prime damaged mitochondria and allow the recruitment of the autophagosome for mitophagy, such as Bnip3 and the E3 ubiquitin-protein ligase Parkin (PARK2), was assessed. The results showed significantly reduced expression of Bnip3 and Parkin in LGMDR9 muscle as compared to controls. Notably, a shift towards a higher molecular weight and a smeary pattern of Parkin were detected in LGMDR9 samples ( Figure 5).

Akt/mTORC1
Autophagy is regulated by multiple signalling pathways. To elucidate the pathways involved in autophagy enhancement seen in LGMDR9 muscle, the expression and the activation of specific upstream proteins, known to control autophagy, was quantified. Akt/PKB is known to be one of the most potent modulators of autophagy in skeletal muscle, blocking the formation of autophagosomes and the lysosomal degradation of their content [31][32][33]. Activation of Akt leads to the downstream activation of mTORC1, which besides its stimulatory effect on muscle growth and hypertrophy, also leads to inhibition of autophagy [34,35]. In addition, Akt phosphorylates and inactivates GSK3β, a positive regulator of autophagy [36,37]. Therefore, the next step was to determine whether the induction of autophagy observed in LGMDR9 skeletal muscle could be due to diminished Akt signalling.
There were no significant changes in the expression level of total Akt or its activated (phosphorylated) form pAkt (Ser473) in LGMDR9 muscles as compared to those of controls. The level of phosphorylation of mTORC1 (Ser2448) and GSK3β (Ser9), however, was decreased in LGMDR9 muscle as compared to normal controls.

AMPK and TRAF6
Other proteins involved in the regulation of autophagy independent of Akt signalling are AMPK and TRAF6. AMPK is involved in autophagy activation through inhibition of mTORC1 as well as direct phosphorylation of ULK1, an important factor in autophagy activation [38,39]. TRAF6 is required for optimal activation of JNK, AMPK, FoxO3 and NF-kB pathways, crucial regulators of autophagy [40]. Previous studies have suggested that musclespecific depletion of TRAF6 inhibits expression of autophagyrelated molecules and autophagosome formation in atrophying skeletal muscle, and that increase in TRAF6 leads to activation of autophagy [40,41]. Therefore, expression of AMPK and its acti- LGMDR9 and control muscles were observed. This was also the case for TRAF6 protein expression (Figure 7).

ER stress/UPR
It has been shown that ER can be a source of membranes during the formation of autophagic vesicles and experimentally generated ER stress can induce autophagy in mammalian cells [42][43][44][45]. Several canonical unfolded protein response (UPR) pathways are implicated in this process [44,46,47].
In order to assess whether the changes in autophagy seen in LGMDR9 muscle could be due to increased ER stress, expression levels of folding catalysts BiP (also known as Grp78), Grp94, calnexin, PDI, the level of eIF2α, peIF2α (Ser51) and IRE1α, the most conserved ER stress sensor, were measured. With regard to the levels of Bip, Grp94, calnexin and eIF2α, there were no substantial differences in protein expression LGMDR9 and control muscles. However, a significant increase in the expression of PDI and peIF2α (Ser51) was found in LGMDR9 muscle when compared to controls. There was also a decrease in the level of IRE1α in LGMDR9 muscle as compared to controls, although it did not reach statistical significance (p = 0.056) (Figure 8).
Decreased levels of IRE1α could be a sign of chronic ER stress in patients with LGMDR9. Under prolonged ER stress, UPR signalling can trigger cell death. One major event responsible for this switch between pro-survival and pro-apoptotic responses is the induction of protein CHOP (also known as Growth Arrest and DNAdamage-inducible 153, GADD153) (reviewed in [48]). Significantly increased level of CHOP expression was seen in LGMDR9 skeletal muscle when compared to controls ( Figure 8). Moreover, an increase in pro-apoptotic Bax and a decrease in anti-apoptotic Bcl2, both transcriptionally regulated by CHOP, were observed in LGMDR9 samples suggesting an activation of apoptosis in these patients.
However, the level of cleaved Parp, another apoptotic marker, remained unchanged (Fig. S3). Conversely, increased autophagy flux has been suggested to be involved in muscle atrophy and degeneration in MDC1A as well as in
LGMDR9 patients are characterized by proximal muscle atrophy.
However, the cascade of molecular events leading from the causative FKRP mutations to the muscle atrophy is largely unknown.
Here, we have examined the role of the UPS and the autophagy-lysosomal system, the two main proteolytic systems associated with atrophy in muscular dystrophy. All LGMDR9 patients included in this study were c.826C>A homozygotes which allowed us to assess changes in different molecular mechanisms on a homogeneous FKRP genotypic background.

UPS markers Atrogin-1 and MuRF1 are not upregulated in LGMDR9 muscle
The E3 ubiquitin ligases Atrogin-1 and MuRF1 have been reported to be elevated in a wide range of atrophy-inducing conditions, and thus have become recognized as key markers of muscle atrophy.
However, upregulation of these UPS markers seems not be an universal finding in muscular dystrophies [49][50][51][52]. In the presented study, expression of Atrogin-1 and MuRF1 in LGMDR9 muscle biopsies did not differ from the control samples. Moreover, there was no apparent correlation between the levels of Atrogin-1 and muscle biopsies presented in this work. Therefore, our data suggest that the atrophy observed in LGMDR9 is likely not to be caused by upregulation of the E3 ubiquitin ligases MuRF1 and Atrogin-1. However, the involvement of other UPS-associated E3 ligases, such as TRIM32, MUSA1 and NEDD4-1, cannot be ruled out [16,53]. Furthermore, it is important to point out that we possess no information concerning the kinetics of protein translation or decay of these ligases in LGMDR9 muscle. It has been previously reported that, under conditions of disuse, the mRNA levels of Atrogin-1 and MuRF1 rapidly increase for a relatively short period of time, and then return to baseline within 2 weeks, suggesting their potential involvement in the remodelling process [17]. Therefore, induction of these genes at early stages of LGMDR9 cannot be rejected. Although the levels of Atrogin-1 and MuRF1 were not changed in LGMDR9 muscle as compared to normal controls, increased level of total ubiquitin as well as higher content of 20S subunits may suggest enhanced rates of protein degradation.
Alternatively, increased ubiquitin levels may result from decreased rates of protein deubiquitination. These questions remain to be elucidated.

Autophagy markers and TEM indicate enhancement of autophagy in LGMDR9 muscle
To examine if autophagy is activated in skeletal muscle from pa- LGMDR9 muscle, as compared to normal controls.
However, we recognize that our data represent a static measurement of autophagic activity and cannot provide information about the movement of cargo through the autophagy pathway. Therefore, although the data presented here strongly support the idea of autophagy induction, rather than autophagy blockage, autophagy flux experiments in vitro would provide additional and important information in order to clarify whether autophagy is activated or hampered in skeletal muscle of LGMDR9 patients.
No correlation was observed between the expression levels of autophagy markers and the disease duration at biopsy or the degree of atrophy in LGMDR9 patients. This might suggest that autophagy proceeds independently of disease progression in these patients.
However, more studies specifically designed to tackle this issue would be needed to confirm this hypothesis.

Mitophagy in LGMDR9 muscle
Mitophagy is a two-step process characterized by the induction of general autophagy followed by priming of the damaged mitochondria for mitophagic recognition and elimination [30]. Hence, since autophagy signalling is generally upregulated in LGMDR9 muscle, one would expect that this would enhance the induction of mitophagy.
This notion is consistent with the observation of mitochondria enveloped in autophagic vacuoles as seen by TEM ( Figure 4E). However, decreased expression levels of mitophagy markers Bnip3 and Parkin in LGMDR9 muscle indicate deficiency in this degradation process rather than its activation. This would be in line with the observed accumulation of swollen mitochondria and mitochondria without cristae which normally would be removed by mitophagy. On the other hand, this cannot explain the presence of areas lacking mitochondria in muscle fibres of some LGMDR9 patients. Furthermore, importantly, it should be noted that there exists a multitude of adaptors and substrates related to mitophagy, of which many function independently of the PINK1/Parkin pathway [58].
Notably, Parkin in LGMDR9 muscles demonstrated a shift in molecular mass of approximately 10 kDa, which potentially could correspond to the addition of mono-ubiquitin. However, whether such modification of Parkin would affect mitophagy in LGMDR9 muscle remains to be investigated.
Accumulation of abnormal mitochondria is associated with oxidative stress and increased production of reactive oxygen species (ROS). Increase in ROS is known to be involved not only in induction of apoptosis [59] but also in positive regulation of autophagy [60]. This opens for a pathway in which induction of autophagy in LGMDR9 muscle could be stimulated by ROS. Moreover, increased levels of ROS might explain the decreased level of anti-apoptotic Bcl2 and increased pro-apoptotic Bax.

ER stress is induced in LGMDR9 muscle
Dysfunction of ER has been suggested to play a significant role in several muscular dystrophies. Expression of ER stress markers has been shown to be increased in DMD, Myotonic dystrophy and Tibial muscular dystrophy (reviewed in [61]). Since ER stress can induce autophagy to participate in the degradation of unfolded proteins and in the removal of superfluous ER membranes, [44,62] we investigated if markers of UPR were changed in LGMDR9 muscle.
Skeletal muscle from LGMDR9 patients reported herein showed significant changes in some (i.e. PDI, peIF2α, CHOP and IRE1α) but not all UPR markers. While there is a possibility that the expression of these proteins may also be governed through mechanisms independent of the ER stress response, ultrastructural changes in the sarcoplasmic reticulum seen in most of our LGMDR9 samples support the notion of increased ER stress in LGMDR9. Moreover, evidence of enhanced ER stress/UPR in LGMDR9 has previously been reported [63]. These authors found increased mRNA expression of GRP78 and CHOP, and furthermore, rough endoplasmic reticulum proliferation, dilated SR cisternae, triad reduplication as well as high numbers of lipid droplets in skeletal muscle of patients with LGMDR9. Similar features were observed in the present study, although the level of lipid droplets in our collection of LGMDR9 muscle biopsies was not to the extent as that reported in [63]. ER stress/UPR activation was also demonstrated in an FKRP knock-down Zebrafish model [64].
However, these alterations were not detected upon examination muscle samples from dystroglycanopathy patients with mutations in FKRP. The timing at which biopsies were obtained and/or the suitability of muscle tissue for measurement of changes in ER stress/ UPR were suggested by authors as the potential factors causing the differences between the Zebrafish model and human patient samples [64].
The results presented here indicate ER stress in LGMDR9 muscle. However, the question concerning its activation as well as the mechanisms by which ER stress would invoke autophagy in LGMDR9 still remain elusive and require further investigation.

Summary
By means of vastus lateralis biopsies derived from patients with LGMDR9, we have investigated the pathological mechanisms underlying muscle wasting in this disorder. Our results show that activation of atrophy in LGMDR9 appears to be independent of the activation of E3 ubiquitin ligases, Atrogin-1 and MuRF1. Skeletal muscle of LGMDR9 patients has a protein expression profile that is concordant with enhanced autophagy and induction of ER stress. These conclusions are supported by ultrastructural alterations seen on TEM images. However, more experiments are needed to clarify the regulatory mechanisms and to determine if these catabolic pathways would be suitable for therapeutic intervention in LGMDR9.

ACK N OWLED G M ENTS
We thank Kenneth B. Larsen and Randi Olsen from the Advanced Microscopy Facility (AMCF), Department of Medical Biology, UiT The Arctic University of Norway, Tromsø, Norway for their technical assistance. We also thank Peter Franek for assisting the preparation and layout figures.

CO N FLI C T O F I NTE R E S T
No conflicts of interest, financial or otherwise, are declared by the authors.

E TH I C A L A PPROVA L
All patient biopsies were obtained with written informed consent and the study was approved by The Regional Committee for

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.