Activation of autophagy by globular adiponectin is required for muscle differentiation

(1)

Activation of autophagy by globular adiponectin is required for

muscle differentiation

Tania Gamberi

a

, Alessandra Modesti

a

, Francesca Magherini

a

, Donna M. D'Souza

b

,

Thomas Hawke

b

, Tania Fiaschi

a,

⁎

a

Dipartimento di Scienze Biomediche, Sperimentali e Cliniche“Mario Serio”, Università degli Studi di Firenze, viale Morgagni 50, 50134 Firenze, Italia

b

Department of Pathology & Molecular Medicine, McMaster University, Hamilton, ON, Canada

a b s t r a c t

a r t i c l e i n f o

Article history:

Received 8 September 2015

Received in revised form 21 January 2016 Accepted 25 January 2016

Available online 26 January 2016

Regulated autophagy is a critical component for a healthy skeletal muscle mass, such that dysregulation of the autophagic processes correlates with severe myopathies. Thus, defining the biological molecules involved in the autophagic processes within skeletal muscle is of great importance. Here we demonstrate that globular adiponectin (gAd) activates autophagy in skeletal muscle myoblasts via an AMPK-dependent mechanism. Acti-vation of autophagy through gAd promotes myoblast survival and apoptosis inhibition during serum starActi-vation and the gAd-activated autophagy orchestrates the myogenic properties of the hormone. Consistent with this con-clusion, inhibition of gAd-activated autophagy by both a pharmacological (chloroquine) or siRNA approach greatly inhibited muscle differentiation, as demonstrated by reductions in myosin heavy chain expression and myotube formation. Further support for the role of adiponectin in autophagy comes from the skeletal muscles of adiponectin KO mice which display decreased LC3 II expression and a myopathic phenotype (heterogeneous fiber sizes, numerous central nuclei). Overall, these findings demonstrate that gAd activates autophagy in myo-blasts and that gAd-activated autophagy drives the myogenic properties of this hormone.

Keywords: Globular adiponectin Myogenesis Autophagy

1. Introduction

Autophagy is important for tissue homeostasis and for the clearance of damaged organelles, misfolded proteins, and pathogens[21,24,27]. Well-regulated autophagy is fundamental for tissue health such that dysregulated autophagy is observed in several human pathologies, including muscle diseases[3,32]. In skeletal muscle, autophagy is sig-ni_{ﬁcantly enhanced in response to pro-atrophic stimuli such as} dener-vation [25], disuse [42], hypoxia [1], and aging [6]. Importantly, autophagy reactivation has been demonstrated to greatly ameliorate the phenotype of diseased muscles[14,15], making it essential, from a therapeutic perspective, to improve the knowledge of the biological molecules which can promote autophagy activation in skeletal muscle. Relative to other hormones, adiponectin is one of the more abundant circulating hormones and has been demonstrated to regulate glucose and lipid homeostasis, as well as having atherogenic and anti-inﬂammatory properties[38]. Globular adiponectin (gAd), originated from the full-length (fAd) form by proteolytic cleavage[35], carries

out crucial biological effects on different tissues including liver, skeletal muscle, and endothelium. Alongside its well-deﬁned metabolic effects in skeletal muscle[39,40], gAd acts on different muscle cell populations promoting muscle differentiation. In particular, gAd promotes myogenesis in myoblasts[10], satellite cells[11], and induces migration and muscle differentiation in the non-resident muscle precursors, mesoangioblasts[13].

Recently, an association between adiponectin and autophagy in other cell populations has been reported[9,17,22,28,31], and given the fundamental importance of autophagy in muscle health[4,6], we hy-pothesized that gAd would be a key regulator of autophagy in skeletal muscle and it's myoblast population. We report that gAd activates au-tophagy on C2C12 myoblasts through an AMPK-dependent pathway. Autophagy activation by gAd inhibits apoptosis in serum-deprived (starved) conditions, thereby enhancing the differentiation of myo-blasts toward myotubes. Interestingly, we demonstrate that autophagy is a critical regulator of myoblast differentiation and that gAd, exoge-nously administered or through paracrine-signaling from other myo-blasts, is critical for autophagy regulation. Theseﬁndings highlight that gAd plays an integral role in autophagy activation in skeletal mus-cle and it's progenitor cells. Furthermore, these_{ﬁndings support the} in-vestigation of adiponectin as a new tool for the amelioration of many inherited muscle diseases which exhibit an inhibited autophagy.

Abbreviations: acrp30, Adiponectin; gAd, Globular adiponectin; MYH, Muscular myosin heavy chain.

⁎ Corresponding author.

E-mail address:tania.ﬁaschi@uniﬁ.it(T. Fiaschi).

http://dx.doi.org/10.1016/j.bbamcr.2016.01.016

Contents lists available atScienceDirect

Biochimica et Biophysica Acta

(2)

2. Materials and methods 2.1. Materials

Unless speciﬁed, all reagents were obtained from Sigma. PVDF membrane was from Millipore; anti-AMPK (sc-25792), anti-phospho-Thr172-AMPK (sc-33524), anti-acrp30 (sc-26497), anti-Myosin Heavy Chain (MYH) (sc-20641), bcl-2 (sc-492), bad (sc-943) anti-bodies, acrp30 siRNA (sc-45891), AdipoR1 siRNA (sc-60124) were from Santa Cruz; anti-LC3, anti-beclin-1, and Alexa 488ﬂuorescent secondary antibodies were from Pierce. gAd was from Alexis. TACS Annexin V-FITC Kit was from Trevigen. JC-1 probe was from Molecular Probe.

2.2. Cell cultures

C2C12 myoblasts were cultured in growing medium composed of Dulbecco's modiﬁed Eagle's medium supplemented with 10% fetal bo-vine serum in 5% CO2humidiﬁed atmosphere. For differentiation,

sub-conﬂuent C2C12 were shifted from growing medium to differentiating medium composed of Dulbecco modiﬁed Eagle's medium containing 2% Horse Serum (HOS) or gAd (1μg/ml) for 4 days.

2.3. Differentiation and fusion indexes

The differentiation index is the percentage of MYH positive cells above the total nuclei. The fusion index is the average number of nuclei present in MYH positive cells containing at least three nuclei above the total nuclei.

2.4. Apoptosis detection

Cells were washed once with PBS buffer, detached from culture plate and then collected by centrifugation at 1000 rpm for 5 min. After a wash with PBS buffer, cells were centrifuged again and then annexin V and propidium iodide positive cells were labeled following manufacturer's instructions. The analysis of early and late apoptotic cells was performed byﬂow cytometry.

2.5. MTT assay

0.5 mg/ml of 3-(4,5-Dimethyl-2-thiazolyl)-2,5-diphenyl-2 H-tetrazolium bromide (MTT) (Sigma–Aldrich, St. Louis, MO, USA) was added to the cells. After 1 h at 37 °C, the cells were extensively washed with PBS and then 1 ml of DMSO was added to the culture. The absor-bance was measured at 570 nm wavelength.

2.6. Immunoblotting

Cells were lysed for 20 min on ice in 500μl of complete radio-immunoprecipitation assay (RIPA) buffer. Lysates were clariﬁed by centrifugation, and total protein contents were obtained using Bradford assay. 20μg of total proteins for each sample were separated by SDS-PAGE and transferred onto PVDF membranes. PVDF membranes were incubated in 2% milk, probed with primary antibodies, and incubated with secondary antibodies conjugated with horseradish peroxidase.

2.7. Immunoﬂuorescence

6 × 104C2C12 cells were grown on glass coverslips. Cells were then washed with phosphate-buffered saline and ﬁxed in 3% para-formaldehyde for 20 min at 4 °C. Fixed cells were permeabilized with three washes with TBST (50 mM Tris–HCl, pH 7.4, 150 mM NaCl, 0.1% Triton X-100), and then blocked with 5.5% HOS in TBST for 1 h at room temperature. Cells were then incubated with speciﬁc primary

antibodies, diluted 1:100 in TBS (50 mM Tris–HCl, pH 7.4, 150 mM NaCl), overnight at 4 °C. Cells were then washed once with TBST and once with TBST with 0.1% BSA and then incubated with secondary anti-bodies (diluted 1:100) for 1 h at room temperature in TBST with 3% BSA. After extensive washes in TBST, cells were mounted with glycerol plastine. The analysis of mitochondrial potential was performed by treating cells with JC-1 probe (5μM final) for 15 min at 37 °C and imme-diately observed. Thefluorescence in all experiments was analyzed using a confocalfluorescence microscope (Leica TCS SP5).

2.8. Adiponectin silencing

Speciﬁc silencing of adiponectin was performed by transfection of C2C12 myoblasts with small-interfering-RNA (siRNA) using TransIT Transfection Reagent according to the manufacturer's instruction. Non-targeting siRNA (scramble) was used as a negative control. 2.9. Adiponectin KO mice

Ad-KO mice were generated as previously described[20]. Adult Ad-KO and WT mice (6 months of age) were housed in standard cages and had access to standard diet and water ad libitum. The animal room was maintained at a constant temperature of 21 °C, a constant humidity of 50%, and a 12:12-h light–dark cycle. All experiments were approved by the Animal Care Committee at York University and were conducted in accordance with guidelines set forth by the Canadian Council for An-imal Care.

2.10. Hematoxylin and eosin (H&E) staining

Cross-sections (8μm thick) of gastrocnemious and soleus muscle were cut on a cryostat and mounted on glass slides. Sections underwent the H&E staining used for analyzing morphological features for both Ad-KO and WT mice. More than 100ﬁbers were analyzed within each mus-cle section.

2.11. Statistical analysis

Data are presented as mean ± S.D. from at least three in-dependent experiments. Analysis of densitometry was performed using ImageJ. Statistical analysis of the data was performed by Student's t-test. P-values_{b0.05 were considered statistically signiﬁcant.} 3. Results

3.1. Activation of autophagy by gAd in myoblasts is an adapting response leading to cell survival

Autophagy is a survival mechanism in response to starvation and various stress conditions. A stress condition to which myoblasts must undergo is the depletion of growth factors, a situation required for the activation of the myogenic program. Our aim was to assay if gAd acti-vates autophagy in myoblasts and deﬁne the involvement of gAd-activated autophagy in the non-metabolic effects of the hormone in myoblasts. Firstly, we observed that treatment of myoblasts with gAd for 24 h promoted autophagy as shown by the increase of the autopha-gic markers LC3 II and beclin-1 (Fig. 1A). In the same conditions, we ob-served that the intracellular adiponectin level is unaffected by 24 h gAd stimulation while an increased phosphorylation of Thr172 in AMPK was detected following gAd treatment (Fig. 1A). Once generated, LC3 II be-comes a component of autophagosomes appearing as dots within the cells. gAd increased the number of autophagosomes in myoblasts, as shown by confocal analysis (Fig. 1B), while the use of the AMPK inhibi-tor, Compound C, robustly decreased the number of autophagosomes within the cells (Fig. 1B), thus showing that activation of autophagy by gAd is AMPK-dependent. To accurately examine the autophagic

(3)

ﬂux, we treated myoblasts with the lysosomal inhibitor baﬁlomycin (BAF) A1which prevents the formation of the autophagolysosomes.

BAF markedly increased LC3 II level in myoblasts treated with gAd as shown by confocal analysis (Fig. 1B). To test if gAd-activated autophagy has pro-survival effects in myoblasts, we performed MTT assay and quantiﬁed the viability of the cells maintained in starving medium, with or without supplemented gAd, in the presence of BAF or Com-pound C.Fig. 1C shows that gAd induced a 1.4-fold increase in cell via-bility and that the increase was greatly affected by the treatment of cells with the autophagy inhibitor BAF and the AMPK inhibitor Com-pound C. Silencing of acrp30 or the receptor AdipoR1 leads to decreased cell viability which was restored when gAd was added to acrp30-silenced myoblasts (Suppl. Fig.1). Theseﬁndings demonstrate that gAd activates autophagy in myoblasts, and in turn, this increased au-tophagic activity leads to cell survival.

3.2. In myoblasts, gAd-activated autophagy is involved in gAd anti-apoptotic effect

To evaluate if gAd-activated autophagy is involved in apoptosis inhi-bition in myoblasts, autophagicflux was blocked by BAF and apoptosis was analyzed by propidium iodide and Annexin V staining. As shown inFig. 2A, gAd induced a ~40% decrease in apoptotic cells in comparison to control while BAF treatment resulted in no difference in apoptotic cell number between groups (Fig. 2A). The anti-apoptotic effect by gAd is due to intrinsic pathway activation since gAd stimulation leads to a marked decrease of bad and a large increase in Bcl-2 (Fig. 2B). The anal-ysis of the mitochondria potential with JC-1 probe shows that the treat-ment of myoblasts with gAd induces a change from greenfluorescence (typical of apoptotic cells) to redfluorescence (typical of viable cells) (Fig. 2C), thus confirming that gAd protects myoblasts from apoptosis acting on intrinsic pathway.

3.3. Activation of autophagy by gAd in myoblasts plays a key role for differentiation

In muscle precursor cells, gAd induces different effects ranging from metalloproteinase production, migration, apoptosis inhibition, and myogenesis[11,13]. Migration of myoblasts is greatly activated by gAd, as shown by wound healing assay (Suppl. Fig. 2). Surprisingly, myoblast motility was greatly impaired when autophagy was inhibited by chloroquine, even in the presence of gAd (Suppl. Fig. 2), indicating that the enhanced myoblast migration is the result of gAd-mediated autophagy. We observed the autophagic marker LC3 II and beclin-1 markedly increased during muscle differentiation, thus sug-gesting that myogenesis and autophagy are associated processes (Fig. 3A). In agreement with our previous results[10], we observed an increased adiponectin expression during muscle differentiation (Fig. 3A). Since gAd promotes myogenesis in muscle precursor cells

[10,11,13], we then investigated the involvement of gAd-activated au-tophagy in the pro-differentiating function of the hormone. Chronic treatment of myoblasts with gAd visibly enhanced myotube formation (Fig. 3B), and increased measures of differentiation and fusion indexes

Fig. 1. gAd activates autophagy in myoblasts inducing cell survival. A) C2C12 myoblasts were serum-depleted overnight and then gAd (1μg/ml) was added for 24 h. Cell lysates were used for the detection of LC3 II, beclin-1, acrp30, and the phosphorylated Thr 172 of AMPK. Bar graph shows the expression level of LC3 II, beclin-1, and acrp30 obtained using actin for normalization and the phosphorylation level of AMPK reported as the ratio between the phosphorylated and total AMPK. a.u.: arbitrary unit. B) Detection of intracellular LC3 II by confocal microscopy. Cells were seeded on coverslips, serum-depleted overnight and then stimulated with gAd (1μg/ml) for 24 h. Compound C (100μM) or BAF A1 (100 nM) was added to the cells together with gAd. Cells were ﬁxed and treated with anti-LC3 antibodies and then with Alexa 488 secondary antibodies (green). Propidium iodide was used for nuclei visualization (red). Bar graph shows the mean of LC3 puncta per cell (100 cells/experiment examined). C) C2C12 myoblasts were serum-depleted overnight and then gAd (1μg/ml) was added for 24 h. MTT assay was performed as reported in“Material and methods” section. Values are reported as fold increase vs the corresponding control (+gAd vs−gAd; + gAd and BAF vs +BAF; + gAd and Compound C vs Compound C). ⁎pb 0.001; ⁎⁎p b 0.01.

Fig. 2. gAd protects myoblasts from apoptotic death through autophagy activation. A) C2C12 myoblasts were serum-depleted overnight and then gAd (1μg/ml) was added for 24 h. Percent of apoptotic cells reported in the bar graph considering 100% control cells (st or st +BAF). BAF A1 (100 nM) were added to the cells together with gAd. The Y axis shows the percentage of annexin V positive (+annexin) and propidium iodide positive (+ PI) cells. B) Analysis of bad and bcl-2 level by immunoblot. Expression of both proteins are normalized using actin immunoblot and reported in the bar graph as arbitrary unit (a.u.). ⁎pb 0.001. C) Analysis of mitochondrial potential using JC-1 probe. Cells were seeded on coverslips, serum-depleted overnight and then gAd (1μg/ml) was added for 24 h. JC-1 probe (5μM) was added to the cells for 15 min at 37 °C and then immediately observed using confocal microscope. Shown experiments are representative of four with similar results.

(4)

(Fig. 3C) as well as the expression of MYH (Fig. 3C). Inhibition of au-tophagy by chloroquine or by BAF markedly decreased myogenesis in gAd-treated myoblasts in comparison to untreated cells, as shown by confocal analysis and MYH expression (Fig. 3B and C). In agreement, the inhibition of autophagy by chloroquine or by BAF in gAd-treated myoblasts induced a decrease of both differentiation and fusion indexes, although a decrease of fusion index was observed in untreated myo-blasts in the presence of chloroquine and BAF (Fig. 3C). Overall, these ﬁndings show that activated autophagy by gAd is involved in differenti-ation of myoblasts.

3.4. Silencing of endogenous acrp30 inhibits autophagy and muscle differ-entiation and enhances apoptosis

Different tissues, ranging from adipose tissue to skeletal muscle, pro-duce adiponectin[7]. In particular, myotubes enhance the expression of adiponectin in inﬂamed and oxidative conditions, thus leading to an au-tocrine/paracrine loop of adiponectin secretion[8,10]. To further con-ﬁrm the involvement of skeletal muscle produced adiponectin in autophagy activation, we silenced endogenous adiponectin in myo-blasts via siRNA. After starving the cells in serum-deprived medium for 24 h, LC3 II and beclin-1 expression levels were analyzed. The trans-fection of acrp30 siRNA into myoblasts successfully decreased endoge-nous adiponectin expression (Fig. 4A). Interestingly, the reduction in

adiponectin expression resulted in similar relative decreases in autoph-agic marker expression (i.e. LC3 II and beclin-1 levels) and in Thr 172 phosphorylation of AMPK (Fig. 4A). In agreement with the protective role on myoblasts, acrp30 silencing induced an increased expression of bad concomitant with a strong decrease of Bcl-2 level, thus supporting the hypothesis that acrp30 silencing promotes the induction of apoptotic death in myoblasts (Fig. 4B). At the same time, promotion of adiponectin-silenced myoblasts toward myotubes resulted in a virtu-ally complete absence of MYH expression and myotube formation or (Fig. 4C and D). In agreement with these observations, quantiﬁcations of both fusion and differentiation indexes were dramatically diminished (Fig. 4E). Altogether, theseﬁndings reveal that skeletal muscle derived adiponectin is involved in autophagy activation and apoptosis inhibition in myoblasts and that endogenously produced hormone induces au-tophagic processes which underlie the pro-myogenic properties of the hormone.

3.5. The administration of gAd to acrp30-silenced myoblasts restores autophagy and differentiation and inhibits apoptosis

After observing that the autocrine production of acrp30 affects au-tophagy, apoptosis, and differentiation in myoblasts, we tested whether this effect was reversible by adding gAd to acrp30-silenced myoblasts. We observed that 24 h of gAd stimulation induced autophagy activation

Fig. 3. Inhibition of autophagy decreases differentiation of myoblasts. A) Subconﬂuent myoblasts was shifted from growing medium to differentiating medium for the period indicated. Amount of beclin-1, LC3, MYH, and acrp30 was analyzed by immunoblot. The expression level of each protein was calculated as the ratio with actin expression and reported in the bar graph as arbitrary unit (a.u.). ⁎pb 0.001 vs point 0;§

pb 0.01 vs point 0;#

pb 0.001 vs point 0. B) Confocal analysis of the differentiation of myoblasts in the presence of chloroquine or BAF. Cells were seeded on coverslips and the growing medium was replaced with differentiation medium for 4 days. Where indicated gAd (1μg/ml), chloroquine (10 μM ﬁnal) or BAF (100 nM) were added to the medium and maintained along the experiment. After 4 days of differentiation, cells were treated with anti-MYH antibodies and then with Alexa-488 secondary antibodies. Green color shows MYH expression while red color shows propidium iodide that labels the nuclei. C) Western blot analysis of MYH expression in cells treated as described in B). Bar graph shows the level of MYH expression calculated using actin immunoblot for normalization and reported as arbitrary unit (a.u.). ⁎pb 0.001; ⁎⁎p b 0.0001. D) Differentiation and fusion indexes calculated after 4 days of differentiation of myoblasts in the presence of gAd, chloroquine, and BAF. The indexes were calculated as reported in “Materials and methods” section. ⁎p b 0.001; ⁎⁎p b 0.0001; ⁎⁎⁎p b 0.01. Shown experiments were repeated at least three times with the same results.

(5)

in acrp30-silenced myoblasts as shown by beclin-1 and LC3 II levels (Fig. 5A). AdipoR1-silenced myoblasts showed a decreased level of both beclin-1 and LC3 II in comparison to control (scramble) highlight-ing that endogenous acrp30 is unable to activate autophagy due to the lack of its receptor. In agreement, gAd stimulation failed to activate au-tophagy in AdipoR1-silenced myoblasts (Fig.5A). The pro-apoptotic protein bad was markedly increased in acrp30-silenced and AdipoR1-silenced myoblasts, and as expected, gAd stimulation led to decreased bad only in acrp30-silenced myoblasts (Fig. 5A). The inhibition of apo-ptotic pathway in acrp30-silenced myoblasts following gAd addition was conﬁrmed by analyzing the mitochondrial membrane potential (Fig. 5B). In agreement, differentiation of myoblasts was restored in acrp30-silenced myoblasts following gAd administration (Fig. 5C). Overall, these_{ﬁndings demonstrate that exogenous gAd reactivates} au-tophagy and differentiation and blocks apoptosis following a blockade of autocrine production.

3.6. Soleus and gastrocnemious muscles from adiponectin KO mice display decreased autophagy and a myopathic phenotype

To verify, in vivo, the involvement of adiponectin in the activation of autophagy within skeletal muscle, we analyzed LC3 II and beclin-1 ex-pression in adiponectin KO mice. LC3 II and beclin-1 levels were mark-edly decreased in both soleus and gastrocnemious muscles in KO mice in comparison to wild type (Fig. 6A), thus showing that adiponectin de-pletion greatly affects skeletal muscle autophagy in vivo. Given the im-portance of muscle satellite cells and their progeny to the growth and maintenance of skeletal muscle[5], we hypothesized that impaired au-tophagy in myoblasts would impact their myogenic capacity and this would manifest as a myopathy within the adiponectin KO mice. Histo-logical analysis of wild-type and adiponectin KO muscles revealed very heterogeneous myofiber sizes, necrotic fibers, and a significant in-crease in the number of centrally located nuclei in gastrocnemious

Fig. 4. Silencing of endogenous acrp30 or AdipoR1 affects autophagy, apoptosis, and differentiation of myoblasts. A) Sub-conﬂuent myoblasts were silenced for acrp30 expression and differentiating medium was added to the cells together with transfection solution. After 4 days anti-LC3, anti-beclin-1, anti-p-Thr-AMPK, and anti-acrp30, immunoblots were performed. B) Myoblasts were silenced for acrp30 and free-serum medium was added together with transfection solution. Anti-bad and anti-bcl-2 immunoblots were performed 72 h after transfection. C) MYH and acrp30 immunoblot 4 days after acrp30 silencing during muscle differentiation. The experiment was performed as described in A). D) Confocal analysis of the acrp30-silenced myoblasts during differentiationfor. Cells were seeded on coverslips and treated as described in panel A). After 4 days of differentiation, cells were treated with anti-MYH antibodies and then with Alexa-488 secondary antibodies. Green color shows MYH expression and red color corresponds to propidium iodide for nuclei staining. E) Differentiation and fusion indexes calculated after 4 days of differentiation of acrp30-silenced myoblasts. The indexes were calculated as reported in“Materials and methods” section. In A), B), and C) panels, actin immunoblot was used for normalization, except for p-Thr-AMPK which was normalized using total AMPK level. The bar graphs show LC3 II, beclin-1, p-Thr-AMPK, and acrp30 (in panel A); bad and bcl-2 (panel B); and MYH and acrp30 (in panel C) expression levels reported as arbitrary unit (a.u.). These experiments are representative of at four different ones with similar results. ⁎pb 0.001; ⁎⁎p b 0.0001.

(6)

muscles of KO mice compared to wild-type (Fig. 6B and C). These ﬁnd-ings provide support for a strong correlation between adiponectin, au-tophagy, and skeletal muscle health in vivo.

4. Discussion

In this paper, we describe the involvement of gAd in the activation of autophagy in skeletal muscle myoblasts and the necessity for this pro-cess in adult myogenesis, and ultimately, in skeletal muscle health. We observed that i) gAd promoted autophagy in myoblasts in an AMPK-dependent fashion; ii) activation of autophagy by gAd played a key role for myoblast survival and apoptosis inhibition; iii) pharmaco-logical inhibition of gAd-induced autophagy impaired survival and dif-ferentiation of myoblasts; iv) silencing of endogenous adiponectin in myoblasts reduces autophagy and cell survival and abrogates myoblast differentiation; andﬁnally, v) skeletal muscles of adiponectin KO mice display decreased autophagy and characteristics of a generalized myop-athy. Thus, we report theﬁrst study illustrating a deep correlation be-tween autophagy and muscle health regulated by gAd.

Recently, it was reported that administration of full-length adiponectin to mice fed with high-fat diet promoted autophagy activa-tion and improved the metabolic profile of these mice (including glu-cose tolerance and insulin sensitivity)[23]. Ourfindings expand on this work by demonstrating that autophagy in myoblasts can be elicited by both endogenous adiponectin and exogenous gAd, which is the more active form of the hormone in skeletal muscle[39]. This is in agreement with the main expression of AdipoR1 in muscle cells, the isoform of adiponectin receptor to which gAd has the greatest affinity[39]. Fur-thermore, we demonstrate that gAd–induced autophagy is associated with adiponectin's myogenic roles[10], thus linking autophagy activa-tion to the non-metabolic effect of gAd in muscle cells. Indeed, it is now clear that adiponectin plays additional roles besides modulating metabolism, among which is the participation in the regenerative pro-cess in many tissues[12]. The involvement of adiponectin in activation of autophagy has been reported in hepatic cells[29], colorectal cancer cells[17], and in mice where adiponectin protects liver from toxicity caused by acetaminophen overdose through an autophagic-dependent mechanism[22].

Fig. 5. gAd stimulation in acrp30-silenced myoblasts re-establishes autophagy, differentiation and inhibition of apoptosis. A) Acrp30 and AdipoR1 was silenced in myoblasts. Where indicated, 48 h from transfection gAd (1μg/ml) was added to the cells for additional 24 h and then immunoblots were performed. B) Acrp30-silenced myoblasts were used for the analysis of mitochondrial potential using JC-1 probe. Cells were seeded on coverslips, acrp30-silenced, and 48 h from transfection gAd (1μg/ml) was added for 24 h. JC-1 probe (5 μM) was added to the cells for 15 min at 37 °C and then immediately observed using confocal microscope. C) Acrp30 was silenced in myoblasts and differentiating medium was added to the cells at the moment of transfection. After 4 days, anti-MYH and anti-acrp30 immunoblots were performed. The bar graphs in panels A and C show protein expression level calculated using actin for normalization and reported as arbitrary unit (a.u.). ⁎pb 0.001 (siRNA acrp30 + gAd vs siRNA acrp30 − gAd); ⁎⁎p b 0.001 (siRNA acrp30 + gAd vs siRNA acrp30 − gAd); ⁎⁎⁎p b 0.01 (siRNA acrp30 + gAd vs siRNA acrp30 − gAd);#

pb 0.001 (siRNA AdipoR1 − gAd vs scramble; siRNA AdipoR1 + gAd vs scramble);§

pb 0.001 (siRNA acrp30 − gAd vs scramble); and pb 0.001 (siRNA AdipoR1 + gAd vs siRNA acrp30 + gAd);##

pb 0.001 (siRNA AdipoR1 + gAd vs siRNA acrp30 + gAd).§§

pb 0.001 (siRNA acrp30 − gAd vs scramble). Shown experiments are representative of four with similar results.

(7)

The involvement of autophagy in maintaining the health of tissues has previously been reported. For example, the muscle-specific deletion of the autophagy gene, Atg7, induces significant muscle atrophy and an age-dependent decrease in force[25]. Furthermore, a key role of au-tophagy during differentiation has been reported for many tissues, such as myofibroblast differentiation[2]and the cardiac differentiation

[19,41]. With respect to skeletal muscle, autophagy increases during myoblast differentiation and activated autophagy protects differentiat-ing myoblasts from apoptosis[26]. In agreement with the latter, our presentﬁndings demonstrate that activation of autophagy by gAd inhibits apoptosis and promotes survival in myoblasts during serum-starvation conditions. Autophagy and apoptosis have functional relationships that could lead to different or convergentﬁnal effects de-pending on cellular setting. Under some circumstances, autophagy con-stitutes a stress adaptation that avoids cell death by suppressing apoptosis, whereas in other cellular conditions, it constitutes an alterna-tive cell-death pathway[24]. In hepatocytes, gAd uses activation of au-tophagy to protect cells from apoptosis induced by ethanol[29]. Here, we report that inhibition of apoptosis through gAd-induced activation of the autophagic pathway in myoblasts is essential for myogenesis. This is in agreement with the fact that myogenesis requires the activa-tion of a transcripactiva-tional program induced by removal of growth factors and, at this stage, inhibition of cell death is essential to obtain optimal muscle differentiation. Nutrient deprivation induces activation of au-tophagy that is crucial for cell survival[24]. We show that gAd treat-ment in a low nutrient condition provokes a further enhancetreat-ment of autophagy activation, thus leading to a great increase of viable myoblasts.

Some papers underline the key role of Bcl-2 and beclin-1 in the re-ciprocal regulation of apoptosis and autophagy. Autophagy activation induced by serum deprivation was accompanied by an up-regulation of Bcl-2 protein levels, thus suggesting that Bcl-2 may participate in balancing the beneﬁcial and detrimental effects of autophagy on cell survival[37]. Moreover, in diabetic cardiomyocytes pharmacological

activation of AMPK stimulates JNK1-Bcl-2 signaling pathway and in-duces the dissociation of Bcl-2–beclin-1 complex. This event normalizes autophagy and decreases apoptosis[18]. Bcl-2–beclin-1 association is regulated by ROCK1 that, phosphorylating Thr119 of beclin-1, leads to disassembly of the beclin-1–Bcl-2 complex[16]. How gAd can balance autophagy activation and apoptosis inhibition, thus leading to myoblast survival and muscle differentiation, needs further investigation. Howev-er, the involvement of gAd in the regulation of Bcl-2–beclin-1 associa-tion may be hypothesized.

Autophagy activation by gAd occurs through AMPK activation. The downstream cascade by activated AMPK leading to autophagy was previously demonstrated in muscle both in vitro[32]and in vivo

[30]. Theseﬁndings highlight AMPK as an important player in autopha-gy activation in muscle. In this regard, our results propose gAd as a physiological molecule able to activate autophagy in vivo through AMPK activation.

Ourfindings demonstrate that the silencing of endogenous acrp30 or AdipoR1 leads to autophagic marker decrease and to bad protein in-crease, suggesting that adiponectin autocrine production is important for autophagy activation and apoptosis inhibition in myoblasts. More-over, the silencing of AdipoR1 (which is the isoform receptor mainly expressed in muscle cells and for which gAd has greater affinity[39]) suggests that myoblasts probably produce gAd in this condition. In this regard, gAd is not considered the circulating form of the hormone, since acrp30 circulates in the plasma as low-, medium-, and high-mo-lecular-weight aggregates of full-length form[34]. Even if it is reported that gAd is produced locally in inflamed microenvironment[8,10], gAd production in basal/not inflamed condition could not be excluded to date.

The involvement of adiponectin in muscle health is supported by our previous results showing that the ex vivo treatment with gAd protects mesoangioblasts from in vivo apoptosis, leading to increased engraft-ment of the cells in the muscles of dystrophic mice[13]. For theﬁrst time, we report a strong association in vivo between adiponectin,

Fig. 6. Analysis of adiponectin KO muscles. A) Immunoblot analysis of LC3 II, beclin-1, and acrp30 expression level in soleus and gastrocnemious. Actin immunoblot is used for normalization. The bar graphs show the ratio between the level of LC3 II, beclin-1, or acrp30 and actin reported as arbitrary units (a.u.). B) H&E staining of muscle cross-sections from wild-type and adiponectin KO mice. White arrows indicate the central nucleatedﬁbers. C) Percentage of central nuclei in WT and adiponectin KO mice.⁎p b 0.001.

(8)

autophagy, and muscle morphology based on the observations of adiponectin KO mice.

Furthermore, myoblast motility is a key process during growth or following damage to allow myoblasts to migrate to sites where they are needed. In accordance with our previousﬁndings showing that gAd elicits a motile program in satellite cells [11] and in mesoangioblasts[13], we report that gAd elicits a motile program in myoblasts through a mechanism dependent on autophagy activation.

Protein degradation in skeletal muscle is controlled by the two major proteolytic systems, the ubiquitin–proteasome and the autopha-gy–lysosome systems. As far as autophagy is concerned, this process must beﬁnely regulated since dysregulated autophagy negatively af-fects muscle health. Indeed, excessive activation of autophagy correlates with muscle wasting[36,42]while inhibition of autophagy causes my-opathies like Pompe and Danon diseases and worsens Duchenne and Ulrich congenital muscular dystrophies[33]and Bethlem myopathy

[14]. Reactivation of autophagy ameliorates myopathic conditions, highlighting the importance of autophagy for overall skeletal muscle health[14,15].

Taken together, ourﬁndings demonstrate that both exogenous addi-tion of gAd and endogenously produced hormone activates autophagy in myoblasts and this activation, mediated through AMPK, orchestrates downstream effects such as muscle differentiation, thus revealing a clear hierarchy between these processes. Based on the key role of au-tophagy in muscle health and since auau-tophagy dysregulation greatly correlates with severe myopathies, the knowledge of the biological mol-ecules able to activate autophagy can represent a new tool to ameliorate myopathic conditions and gAd could be considered a good candidate for this aim.

Supplementary data to this article can be found online athttp://dx. doi.org/10.1016/j.bbamcr.2016.01.016.

Conﬂict of interest

Authors declare no conﬂict of interest. Transparency document

TheTransparency documentassociated with this article can be found in the online version.

Acknowledgments

This work was supported by Ente Cassa di Risparmio di Firenze (Italy) and the Italian Ministry of University and Research (MIUR). References

[1] G. Bellot, R. Garcia-Medina, P. Gounon, J. Chiche, D. Roux, J. Pouyssegur, N.M. Mazure, Hypoxia-induced autophagy is mediated through hypoxia-inducible factor induction of BNIP3 and BNIP3L via their BH3 domains, Mol. Cell. Biol. 29 (2009) 2570–2581.

[2] M. Bernard, M. Dieude, B. Yang, K. Hamelin, K. Underwood, M.J. Hebert, Autophagy fosters myoﬁbroblast differentiation through MTORC2 activation and downstream upregulation of CTGF, Autophagy 10 (2014) 2193–2207.

[3] P. Bernardi, P. Bonaldo, Mitochondrial dysfunction and defective autophagy in the pathogenesis of collagen VI muscular dystrophies, Cold Spring Harb. Perspect. Biol. 5 (2013) a011387.

[4] P. Bonaldo, M. Sandri, Cellular and molecular mechanisms of muscle atrophy, Dis. Model. Mech. 6 (2013) 25–39.

[5] A.S. Brack, T.A. Rando, Tissue-speciﬁc stem cells: lessons from the skeletal muscle satellite cell, Cell Stem Cell 10 (2012) 504–514.

[6] A.L. Bujak, J.D. Crane, J.S. Lally, R.J. Ford, S.J. Kang, I.A. Rebalka, A.E. Green, B.E. Kemp, T.J. Hawke, J.D. Schertzer, G.R. Steinberg, AMPK activation of muscle autophagy pre-vents fasting-induced hypoglycemia and myopathy during aging, Cell Metab. 21 (2015) 883–890.

[7] P. Chiarugi, T. Fiaschi, Adiponectin in health and diseases: from metabolic syndrome to tissue regeneration, Expert Opin. Ther. Targets 14 (2010) 193–206.

[8] A.M. Delaigle, J.C. Jonas, I.B. Bauche, O. Cornu, S.M. Brichard, Induction of adiponectin in skeletal muscle by inﬂammatory cytokines: in vivo and in vitro studies, Endocri-nology 145 (2004) 5589–5597.

[9] E.E. Essick, R.M. Wilson, D.R. Pimentel, M. Shimano, S. Baid, N. Ouchi, F. Sam, Adiponectin modulates oxidative stress-induced autophagy in cardiomyocytes, PLoS One 8 (2013), e68697.

[10] T. Fiaschi, D. Cirelli, G. Comito, S. Gelmini, G. Ramponi, M. Serio, P. Chiarugi, Globular adiponectin induces differentiation and fusion of skeletal muscle cells, Cell Res. 19 (2009) 584–597.

[11] T. Fiaschi, E. Giannoni, M.L. Taddei, P. Chiarugi, Globular adiponectin activates motil-ity and regenerative traits of muscle satellite cells, PLoS One 7 (2012), e34782.

[12] T. Fiaschi, F. Magherini, T. Gamberi, P.A. Modesti, A. Modesti, Adiponectin as a tissue regenerating hormone: more than a metabolic function, Cell. Mol. Life Sci. 71 (2014) 1917–1925.

[13]T. Fiaschi, F.S. Tedesco, E. Giannoni, J. Diaz-Manera, M. Parri, G. Cossu, P. Chiarugi, Globular adiponectin as a complete mesoangioblast regulator: role in proliferation, survival, motility, and skeletal muscle differentiation, Mol. Biol. Cell 21 (2010) 848–859.

[14] P. Grumati, L. Coletto, P. Sabatelli, M. Cescon, A. Angelin, E. Bertaggia, B. Blaauw, A. Urciuolo, T. Tiepolo, L. Merlini, N.M. Maraldi, P. Bernardi, M. Sandri, P. Bonaldo, Au-tophagy is defective in collagen VI muscular dystrophies, and its reactivation rescues myoﬁber degeneration, Nat. Med. 16 (2010) 1313–1320.

[15] P. Grumati, L. Coletto, M. Sandri, P. Bonaldo, Autophagy induction rescues muscular dystrophy, Autophagy 7 (2011) 426–428.

[16] A.U. Gurkar, K. Chu, L. Raj, R. Bouley, S.H. Lee, Y.B. Kim, S.E. Dunn, A. Mandinova, S.W. Lee, Identiﬁcation of ROCK1 kinase as a critical regulator of Beclin1-mediated au-tophagy during metabolic stress, Nat. Commun. 4 (2013) 2189.

[17]B.S. Habeeb, J. Kitayama, H. Nagawa, Adiponectin supports cell survival in glucose deprivation through enhancement of autophagic response in colorectal cancer cells, Cancer Sci. 102 (2011) 999–1006.

[18] C. He, H. Zhu, H. Li, M.H. Zou, Z. Xie, Dissociation of Bcl-2-Beclin1 complex by acti-vated AMPK enhances cardiac autophagy and protects against cardiomyocyte apo-ptosis in diabetes, Diabetes 62 (2013) 1270–1281.

[19] Z. Jia, J. Wang, W. Wang, Y. Tian, W. XiangWei, P. Chen, K. Ma, C. Zhou, Autophagy eliminates cytoplasmic beta-catenin and NICD to promote the cardiac differentia-tion of P19CL6 cells, Cell. Signal. 26 (2014) 2299–2305.

[20] M.P. Krause, Y. Liu, V. Vu, L. Chan, A. Xu, M.C. Riddell, G. Sweeney, T.J. Hawke, Adiponectin is expressed by skeletal muscleﬁbers and inﬂuences muscle pheno-type and function, Am. J. Physiol. Cell Physiol. 295 (2008) C203–C212.

[21]B. Levine, G. Kroemer, Autophagy in the pathogenesis of disease, Cell 132 (2008) 27–42.

[22] Z. Lin, F. Wu, S. Lin, X. Pan, L. Jin, T. Lu, L. Shi, Y. Wang, A. Xu, X. Li, Adiponectin pro-tects against acetaminophen-induced mitochondrial dysfunction and acute liver in-jury by promoting autophagy in mice, J. Hepatol. 61 (2014) 825–831.

[23] Y. Liu, R. Palanivel, E. Rai, M. Park, T.V. Gabor, M.P. Scheid, A. Xu, G. Sweeney, Adiponectin stimulates autophagy and reduces oxidative stress to enhance insulin sensitivity during high-fat diet feeding in mice, Diabetes 64 (2015) 36–48.

[24] M.C. Maiuri, E. Zalckvar, A. Kimchi, G. Kroemer, Self-eating and self-killing: crosstalk between autophagy and apoptosis, Nat. Rev. Mol. Cell Biol. 8 (2007) 741–752.

[25] E. Masiero, L. Agatea, C. Mammucari, B. Blaauw, E. Loro, M. Komatsu, D. Metzger, C. Reggiani, S. Schiafﬁno, M. Sandri, Autophagy is required to maintain muscle mass, Cell Metab. 10 (2009) 507–515.

[26] E.M. McMillan, J. Quadrilatero, Autophagy is required and protects against apoptosis during myoblast differentiation, Biochem. J. 462 (2014) 267–277.

[27] N. Mizushima, B. Levine, A.M. Cuervo, D.J. Klionsky, Autophagyﬁghts disease through cellular self-digestion, Nature 451 (2008) 1069–1075.

[28]S. Nepal, M.J. Kim, A. Subedi, E.S. Lee, C.S. Yong, J.A. Kim, W. Kang, M.K. Kwak, D.S. Arya, P.H. Park, Globular adiponectin inhibits ethanol-induced apoptosis in HepG2 cells through heme oxygenase-1 induction, Biochem. Pharmacol. 84 (2012) 974–983.

[29] S. Nepal, P.H. Park, Activation of autophagy by globular adiponectin attenuates ethanol-induced apoptosis in HepG2 cells: involvement of AMPK/FoxO3A axis, Biochim. Biophys. Acta 1833 (2013) 2111–2125.

[30] M. Pauly, F. Daussin, Y. Burelle, T. Li, R. Godin, J. Fauconnier, C. Koechlin-Ramonatxo, G. Hugon, A. Lacampagne, M. Coisy-Quivy, F. Liang, S. Hussain, S. Matecki, B.J. Petrof, AMPK activation stimulates autophagy and ameliorates muscular dystrophy in the mdx mouse diaphragm, Am. J. Pathol. 181 (2012) 583–592.

[31] G.M. Qi, L.X. Jia, Y.L. Li, H.H. Li, J. Du, Adiponectin suppresses angiotensin II-induced inﬂammation and cardiac ﬁbrosis through activation of macrophage autophagy, En-docrinology 155 (2014) 2254–2265.

[32] M. Sandri, L. Coletto, P. Grumati, P. Bonaldo, Misregulation of autophagy and protein degradation systems in myopathies and muscular dystrophies, J. Cell Sci. 126 (2013) 5325–5333.

[33] P. Spitali, P. Grumati, M. Hiller, M. Chrisam, A. Aartsma-Rus, P. Bonaldo, Autophagy is impaired in the tibialis anterior of dystrophin null mice, PLoS Curr. 5 (2013).

[34] H. Waki, T. Yamauchi, J. Kamon, Y. Ito, S. Uchida, S. Kita, K. Hara, Y. Hada, F. Vasseur, P. Froguel, S. Kimura, R. Nagai, T. Kadowaki, Impaired multimerization of human adiponectin mutants associated with diabetes. Molecular structure and multimer formation of adiponectin, J. Biol. Chem. 278 (2003) 40352–40363.

[35]H. Waki, T. Yamauchi, J. Kamon, S. Kita, Y. Ito, Y. Hada, S. Uchida, A. Tsuchida, S. Takekawa, T. Kadowaki, Generation of globular fragment of adiponectin by leuko-cyte elastase secreted by monocytic cell line THP-1, Endocrinology 146 (2005) 790–796.

[36] X. Wang, C. Blagden, J. Fan, S.J. Nowak, I. Taniuchi, D.R. Littman, S.J. Burden, Runx1 prevents wasting, myoﬁbrillar disorganization, and autophagy of skeletal muscle, Genes Dev. 19 (2005) 1715–1722.

[37] H.D. Xu, D. Wu, J.H. Gu, J.B. Ge, J.C. Wu, R. Han, Z.Q. Liang, Z.H. Qin, The pro-survival role of autophagy depends on Bcl-2 under nutrition stress conditions, PLoS One 8 (2013), e63232.

(9)

[38] T. Yamauchi, T. Kadowaki, Adiponectin receptor as a key player in healthy longevity and obesity-related diseases, Cell Metab. 17 (2013) 185–196.

[39] T. Yamauchi, J. Kamon, Y. Ito, A. Tsuchida, T. Yokomizo, S. Kita, T. Sugiyama, M. Miyagishi, K. Hara, M. Tsunoda, K. Murakami, T. Ohteki, S. Uchida, S. Takekawa, H. Waki, N.H. Tsuno, Y. Shibata, Y. Terauchi, P. Froguel, K. Tobe, S. Koyasu, K. Taira, T. Kitamura, T. Shimizu, R. Nagai, T. Kadowaki, Cloning of adiponectin receptors that mediate antidiabetic metabolic effects, Nature 423 (2003) 762–769.

[40]T. Yamauchi, J. Kamon, Y. Minokoshi, Y. Ito, H. Waki, S. Uchida, S. Yamashita, M. Noda, S. Kita, K. Ueki, K. Eto, Y. Akanuma, P. Froguel, F. Foufelle, P. Ferre, D.

Carling, S. Kimura, R. Nagai, B.B. Kahn, T. Kadowaki, Adiponectin stimulates glucose utilization and fatty-acid oxidation by activating AMP-activated protein kinase, Nat. Med. 8 (2002) 1288–1295.

[41] J. Zhang, J. Liu, L. Liu, W.L. McKeehan, F. Wang, Theﬁbroblast growth factor signaling axis controls cardiac stem cell differentiation through regulating autophagy, Autophagy 8 (2012) 690–691.

[42] J. Zhao, J.J. Brault, A. Schild, P. Cao, M. Sandri, S. Schiafﬁno, S.H. Lecker, A.L. Goldberg, FoxO3 coordinately activates protein degradation by the autophagic/lysosomal and proteasomal pathways in atrophying muscle cells, Cell Metab. 6 (2007) 472–483.