Impaired myogenesis in estrogen-related receptor γ (ERRγ)-deficient skeletal myocytes due to oxidative stress

Abstract

Specialized contractile function and increased mitochondrial number and oxidative capacity are hallmark features of myocyte differentiation. The estrogen-related receptors (ERRs) can regulate mitochondrial biogenesis or mitochondrial enzyme expression in skeletal muscle, suggesting that ERRs may have a role in promoting myogenesis. Therefore, we characterized myogenic programs in primary myocytes isolated from wild-type (M-ERRγWT) and muscle-specific ERRγ^−/− (M-ERRγ^−/−) mice. Myotube maturation and number were decreased throughout differentiation in M-ERRγ^−/− primary myocytes, resulting in myotubes with reduced mitochondrial content and sarcomere assembly. Compared with M-ERRγWT myocytes at the same differentiation stage, the glucose oxidation rate was reduced by 30% in M-ERRγ^−/− myotubes, while medium-chain fatty acid oxidation was increased by 34% in M-ERRγ^−/− myoblasts and 36% in M-ERRγ^−/− myotubes. Concomitant with increased reliance on mitochondrial β-oxidation, H₂O₂ production was significantly increased by 40% in M-ERRγ^−/− myoblasts and 70% in M-ERRγ^−/− myotubes compared to M-ERRγWT myocytes. ROS activation of FoxO and NF-κB and their downstream targets, atrogin-1 and MuRF1, was observed in M-ERRγ^−/− myocytes. The antioxidant N-acetyl cysteine rescued myotube formation and atrophy gene induction in M-ERRγ^−/− myocytes. These results suggest that loss of ERRγ causes metabolic defects and oxidative stress that impair myotube formation through activation of skeletal muscle atrophy pathways.—Murray, J., Auwerx, J., Huss, J. M. Impaired myogenesis in estrogen-related receptor γ (ERRγ)-deficient skeletal myocytes due to oxidative stress.

The estrogen-related receptors (ERRs) are members of the nuclear receptor superfamily of transcription factors, consisting of ERRα, ERRβ, and ERRγ. While no natural ligand has been identified for these receptors, the activity of ERRs is modulated through interactions with coregulatory proteins, including peroxisome proliferator-activated receptor γ coactivator 1α (PGC-1α) and PGC-1β (1–4). Genomic studies combining chromatin immunoprecipitation (ChIP) and gene expression microarray in heart revealed that ERRα and ERRγ regulate genes involved in substrate uptake, the TCA cycle, fatty acid oxidation (FAO), and oxidative phosphorylation (5). Phenotypic characterization of ERRγ whole-body knockout mice demonstrated cardiac electrophysiological conduction defects in embryo as well as reduced expression and activity of TCA and electron transport chain (ETC) complex I enzymes (6). Overexpression of either constitutively active (ERRγ-VP16) or wild-type ERRγ in skeletal muscle increases mitochondrial enzyme activity, enhances exercise capacity, increases expression of genes involved in fat metabolism, and promotes a switch toward more oxidative fiber types and muscle vascularization (7–8).

An integral part of skeletal muscle differentiation is a dramatic increase in mitochondrial number and oxidative capacity. Myotube formation can be almost completely inhibited by blocking mitochondrial metabolism or proliferation (9, 10). Because the ERRs can regulate mitochondrial biogenesis and activity in adult skeletal muscle, they may also have a role in promoting myogenesis. We previously demonstrated that ERRα regulates myocyte differentiation in part through modulation of ERK-MAP kinase signaling during early myogenesis (11). In these studies, we observed that ERRγ expression, which is nearly undetectable in myoblasts (MBs), is highly up-regulated during differentiation. Collectively, these observations support a potential role for the ERRγ isoform in regulating myogenesis, an area of ERRγ function that has not been completely explored.

Reactive oxygen species (ROS) production occurs in resting and contracting muscle as a byproduct of mitochondrial respiration. A major source of ROS in myocytes is the mitochondrial ETC, primarily producing superoxide from complexes I and III that is subsequently converted to H₂O₂ by superoxide dismutase 1 (SOD1) and SOD2 (12, 13). Moderate increases in ROS production, such as during exercise, can promote adaptive induction of antioxidant defense enzymes to respond to subsequent increases in ROS. However, chronically elevated ROS can damage proteins, lipids, and DNA (14). Oxidative stress inhibits myocyte differentiation by reducing expression of myogenic regulatory factors (MRFs) and sarcomeric proteins and promotes muscle atrophy through up-regulation of muscle-specific ubiquitin ligases that mediate protein breakdown. Specifically, treatment with H₂O₂ or depletion of the endogenous radical scavenger glutathione blocks expression of the MRFs, myogenin or MyoD, and the sarcomeric proteins, myosin heavy chain (MHC) and troponin I (TnI), and inhibits myotube (MT) formation (15, 16). Conversely, treatment with ROS scavengers such as N-acetyl cysteine (NAC) and phenyl-N-tert-butylnitrone, rescues H₂O₂ effects and promotes myocyte differentiation (15, 17). Furthermore, elevated ROS levels induce muscle atrophy by up-regulating components of the ubiquitin proteasome system. H₂O₂ treatment of C2C12 myocytes up-regulates the muscle-specific ubiquitin ligases, atrogin-1/MAFbx, and muscle RING-finger protein 1 (MuRF1) (16), which targets MHC for proteasomal degradation (18). Therefore, intracellular ROS levels are an important factor regulating myogenesis and MT growth.

The transcription factor NF-κB is activated by ROS, which can lead to skeletal muscle atrophy. The NF-κB family consists of 5 members: RelA (p65), RelB, c-rel, p50, and p52 (19). Cytoplasmic NF-κB is bound to the inhibitor protein, IκB. In response to stimuli, such as ROS or inflammatory cytokines, the IκB kinases (IKKs) phosphorylate IκBα, leading to IκBα degradation and, thereby, allowing nuclear translocation of NF-κB and transcriptional activation of its target genes. Phosphorylation of the p65 subunit augments NF-κB activity by increasing its stability and nuclear localization or by enhancing binding to other cofactors (20). Oxidative stress activates NF-κB in C2C12 myocytes (21–22), which can, in turn, be inhibited by NAC treatment (23). In response to physiological levels of ROS, NF-κB directly activates antioxidant genes, such as SOD2, SOD1, catalase, and glutathione peroxidase 1 (Gpx1) (24). However, sustained NF-κB activation by oxidative stress is associated with various types of muscle atrophy. NF-κB regulates the ubiquitin proteasome pathway through transcriptional activation of MuRF1, which targets MHC for degradation during atrophy in C2C12 myocytes. Previous studies have shown that activation of NF-κB by constitutively active IKKβ is sufficient to increase MuRF1 expression and cause muscle wasting (25). Muscle-specific IKK deletion prevents MuRF1 induction and protein degradation and maintains muscle fiber type, size, and strength during denervation-induced atrophy (26). Finally, MuRF1-knockout mice maintain MHC expression and are protected from muscle mass and fiber area loss under atrophy conditions (27, 28).

The FoxO family of transcription factors, consisting of FoxO1, FoxO3a, FoxO4, and FoxO6, also mediates the response to ROS and regulates pathological muscle atrophy. FoxO activity is regulated by post-translational modifications, such as phosphorylation, acetylation, and ubiquitination (29). FoxO phosphorylation by Akt inhibits FoxO activity by promoting nuclear export (30, 31). However, oxidative stress drives cytoplasmic-to-nuclear translocation of FoxO, thereby stimulating transcription of the antioxidant enzyme genes, SOD2, and catalase (32–34). FoxO protein levels are also up-regulated during different types of muscle atrophy (35–36). FoxO1 overexpression in skeletal muscle leads to decreased muscle mass and reduced type I fiber gene expression (37), and constitutively active FoxO1 inhibits differentiation in C2C12 myocytes (38). These effects are driven by induction of atrogin-1/MAFbx (39). FoxO3a transcriptionally activates the atrogin-1 gene during disuse muscle atrophy (39, 40). Interestingly, maximal Foxo3a induction of antioxidant genes requires PGC-1α, whereas Foxo3a activation of atrogin-1 is suppressed by PGC-1α (41). In addition, expression of dominant-negative FoxO3a in soleus prevents muscle atrophy under disuse conditions (42). Therefore, the FoxO transcription factors are important mediators of oxidative stress-induced skeletal muscle atrophy.

The current studies reveal an important role for ERRγ in regulating MT formation using a primary myocyte differentiation model. We found that muscle-specific ERRγ^−/− (M-ERRγ^−/−) myocytes form immature MTs with reduced mitochondrial content and altered distribution, increased rates of medium-chain FAO and decreased rates of glucose oxidation. In addition, M-ERRγ^−/− myocytes produce elevated levels of H₂O₂, leading to activation of FoxO and NF-κB and downstream atrophy pathways. The NF-κB inhibitor 6-amino-4-(4-phenoxyphenylethyl-amino)quinazoline (QNZ) reverses MuRF1 induction in M-ERRγ^−/− myocytes and modestly promotes MT formation. Finally, treatment with the antioxidant NAC rescues MT formation and reduces atrogin-1 and MuRF1 expression levels in M-ERRγ^−/− myocytes.

MATERIALS AND METHODS

Animals

Mice carrying a floxed ERRγ allele (exon 2) were obtained from the Institut Clinique de la Souris/Mouse Clinical Institute (MCI; Strasbourg, France) and crossed with MCK-Cre mice, in which Cre recombinase expression is driven by the muscle creatine kinase promoter (Jackson Laboratory, Bar Harbor, ME, USA). Deletion of ERRγ was confirmed by PCR using the following primers: 87, 5′-CCCTTATGCTGATTACCTTCTTGTA-3′; 88, 5′-CAACAATGTAGACACAAAGACATGG-3′; 610, 5′-GTTTTAAAGGCCCTTGGTGATCTCGC-3′; and 612, 5′-CTGCAACCCTTGGACTGCCAGAAC-3′. The Cre transgene was detected by PCR using Cre-specific primers: 5′-GTGAAACAGCATTGCTGTCACTT-3′ and 5′-TAAGTCTGAACCCGGTCTGC-3′; and internal control primers: 5′-CAAATGTTGCTTGTCTGGTG-3′ and 5′-GTCAGTCGAGTGCACAGTTT-3′.

Cell culture

Cell culture reagents were purchased from Mediatech (Manassas, VA, USA) unless otherwise stated. Primary myocytes were isolated from M-ERRγ wild-type (M-ERRγWT) and M-ERRγ^−/− mice, as described previously (11). Growth medium contained 40% DMEM, 40% Ham's F10, 20% FBS (Atlanta Biologicals, Lawrenceville, GA, USA), 100 U/ml penicillin, 100 μg/ml streptomycin, and 2.5 ng/ml bFGF (Promega, Madison, WI, USA). Cells were seeded into collagen-coated plates at densities resulting in ∼70% confluence by 48 h. To induce differentiation, myocytes were switched into DMEM containing 5% horse serum (Atlanta Biologicals). Bright-field images were taken on a Leica DMIL microscope (Leica Microsystems, Buffalo Grove, IL, USA). For NAC treatment experiments, MBs were plated directly into medium containing 0.5 mM or 1 mM NAC (Sigma-Aldrich, St. Louis, MO, USA). To measure the effect of NAC treatment on proliferation, MBs plated in triplicate were counted using a hemocytometer and trypan blue exclusion on 3 consecutive days. Doubling times of untreated M-ERRγWT vs. M-ERRγ^−/− MBs were 24.1 and 27.6 h, respectively.

Quantitative real-time PCR

RNA was isolated from primary myocytes using TRIzol (Invitrogen, Carlsbad, CA, USA). RNA (1 μg) was reverse transcribed using the iScript cDNA synthesis kit (Bio-Rad, Hercules, CA, USA). PCR was performed in 15-μl reactions containing 1× SYBR Green reagent and 0.1 μM gene-specific primers using the iQ5 real-time PCR system (Bio-Rad). Experimental transcript levels were normalized to 36B4 ribosomal RNA analyzed in separate reactions. Primers are listed in Table 1 and, where indicated, were derived from PrimerBank (Harvard PrimerBank; http://pga.mgh.harvard.edu/primerbank/index.html) or previously published (11, 43, 44).

Table 1.

Sequences of primers used for real-time PCR

Gene	Primers
Atrogin-1 (158517905b2)	F: 5′-GAGTGGCATCGCCCAAAAGA-3′
	R: 5′-TCTGGAGAAGTTCCCGTATAAGT-3′
Catalase (6753272a1)	F: 5′-AGCGACCAGATGAAGCAGTG-3′
	R: 5′-TCCGCTCTCTGTCAAAGTGTG-3′
CoxI (43)	F: 5′-ACCATCATTTCTCCTTCTCCTA-3′
	R: 5′-TAGATTTCCGGCTAGAGGTG-3′
Cox6c	F: 5′-TTGTGGCCCTGGGAGTTG-3′
	R: 5′-TCTGCATACGCCTTCTTTCTTG-3′
FoxO1	F: 5′-ACTTTGATAATGTGTTGCCC-3′
	R: 5′-CTGCTGTCAGACAATCTGA-3′
FoxO3a	F: 5′-AACGGCTCACTTTGTCCCA-3′
	R: 5′-TTGATGATCCACCAAGAGC-3′
GLUT1	F: 5′-TGCAGTTCGGCTATAACACTG-3′
	R: 5′-GCGGTGGTTCCATGTTTGAT-3′
Gpx1(145275166b3)	F: 5′-CTACACCGAGATGAACGATCTG-3′
	R: 5′-TCCGAACTGATTGCACGGG-3′
Gpx3 (15011841a1)	F: 5′-CCTTTTAAGCAGTATGCAGGCA-3′
	R: 5′-CAAGCCAAATGGCCCAAGTT-3′
IκBα (226052095b2)	F: 5′-GGCAATCATCCACGAAGAGAA-3′
	R: 5′-GTATTTCCTCGAAAGTCTCGGAG-3′
IL-6 (13624310b2)	F: 5′-TCTATACCACTTCACAAGTCGGA-3′
	R: 5′-GAATTGCCATTGCACAACTCTTT-3′
MuRF-1 (44)	F: 5′-TGCCTGGAGATGTTTACCAAGC-3′
	R: 5′-AAACGACCTCCAGACATGGACA-3′
Ndufa8	F: 5′-GCCAACTCTGGAAGAGCTGAA-3′
	R: 5′-GTGATGGGCGGCAGCTT-3
SDHA	F: 5′-CCATGGTCACTAGGGCTGGTT-3′
	R: 5′-CACGACACCCTTCTGTGATGA-3′
SOD2 (31980762a3)	F: 5′-ACAAACCTGAGCCCTAAGGGT-3′
	R: 5′-GAACCTTGGACTCCCACAGAC-3′
TFIID	F: 5′-TACCTGTGGCATTGATAGCATCC-3′
	R: 5′-AAAGCAATCTGAACTTCTCTGC-3′
UCP2 (188035853b2)	F: 5′-CAGCGCCAGATGAGCTTTG-3′
	R: 5′-GGAAGCGGACCTTTACCACA-3′
UCP3	F: 5′-TGCTGAGATGGTGACCTACGA-3′
	R: 5′-CCAAAGGCAGAGACAAAGTGA-3′

PrimerBank ID numbers are indicated in parenthesis. Primer sequences for other genes were previously published (11). F, forward; R, reverse.

Immunofluorescence

Myocytes plated in 4-well glass-bottom 35-mm dishes (Greiner Bio-One, Monroe, NC, USA) were fixed in 4% formaldehyde for 15 min and blocked for 30 min in 1% BSA, 0.3% Triton, and 3% normal goat serum in PBS. Cells were incubated with primary antibodies to α-actinin (Sigma-Aldrich), TnI (Santa Cruz Biotechnology, Santa Cruz, CA, USA), or cytochrome c (BD Biosciences, Franklin Lakes, NJ, USA) at 4°C overnight, followed by Alexa Fluor 488 (Invitrogen) secondary antibody and Hoechst 33258 (Sigma-Aldrich) nuclear stain. Myocytes were analyzed using the Olympus IX81 inverted microscope.

Histology

Gasctrocnemius muscle from M-ERRγWT or M-ERRγ^−/− mice was fixed in formalin and paraffin embedded. Sections were stained with primary antibody to ERRγ (ab77690; Abcam, Cambridge, MA, USA) and HRP-conjugated secondary antibody (Jackson ImmunoResearch Laboratories, West Grove, PA, USA) followed by detection with SigmaFast DAB (3,3′-diaminobenzidine tetrahydrochloride) with metal enhancer and counterstained with hematoxylin (Sigma-Aldrich).

BODIPY-fatty acid uptake assay

BODIPY-FL-labeled lauric acid or palmitic acid were obtained from Invitrogen. BODIPY stock solution (2.4 mM in DMSO) was diluted to 1 μM in PBS. To conjugate fatty acids to BSA, an equimolar amount of BSA (1 μM) in PBS was added, and solutions were incubated at 37°C for 1 h with rocking. Cells were incubated with BODIPY-fatty acid for 30 min at 37°C, fixed in 4% formaldehyde for 15 min, counterstained with Hoechst 33258 (Sigma-Aldrich), and imaged on the Olympus IX81 inverted microscope.

Glucose uptake assay

Myocytes were incubated in serum-free, low-glucose (1 mg/ml) DMEM at 37°C for 30 min. Subsequently, 20 μM of glucose analog 2-[N-(7-nitrobenz-2-oxa-1,3-diazol-4-yl)amino]-2-deoxyglucose (2-NBDG; Invitrogen) in PBS was added for 20 min at 37°C. Medium was transferred to a black, clear-bottom 96-well plate (Corning, Corning, NY, USA), and fluorescence was measured on a Tecan Infinite M1000 plate reader (λ_ex 465 nm/λ_em 540 nm; Tecan Group, San Jose, CA, USA).

Metabolic assays

Oxidation rates of glucose or fatty acids were determined as described previously (11). For glucose oxidation, cells were incubated for 6 h in serum-free, low-glucose DMEM, then 1 μCi/ml [U-¹⁴C-]-d-glucose (MP Biomedicals, Solon, OH, USA) and 100 μM palmitate were added. Flasks were sealed and incubated for 1 h at 37°C. Following medium acidification with TCA, ¹⁴CO₂ was collected on KOH-saturated filter paper suspended inside the vial and measured by scintillation counting. For FAO, cells were incubated with 125 μM [³H-9,10]-palmitic acid (Perkin Elmer NEN, Waltham, MA, USA) and 1 mM carnitine for 2 h at 37°C. ³H₂O was purified over Dowex 1 × 2-400 resin (Sigma-Aldrich) and quantitated by scintillation counting.

Mitochondrial isolation

Hind limb muscles from M-ERRγWT or M-ERRγ^−/− mice were minced on ice in buffer A (220 mM mannitol, 100 mM sucrose, 100 mM Tris-HCl, and 1 mM EGTA, at pH 7.4) and transferred to centrifuge tubes using 5 ml HM (250 mM sucrose, 100 mM Tris-HCl, and 2 mM EGTA, at pH 7.4). Polytron homogenization and centrifugation were performed as published previously (45). The mitochondrial pellet was resuspended in buffer B (220 mM mannitol, 70 mM sucrose, 20 mM Tris-HCl, and 20 μM EDTA, at pH 7.4). Protein concentration was determined by Micro BCA Protein Assay (Thermo Fisher Scientific, Rockford, IL, USA).

Hydrogen peroxide detection assay

Myocytes were incubated with 50 μM Amplex Red (Invitrogen) and 0.1 U/ml horseradish peroxidase (HRP) for 30 min at 37°C. The Amplex Red solution was then transferred to a 96-well black plate. Mitochondria (185 μg) isolated from M-ERRγWT or M-ERRγ^−/− skeletal muscle were incubated with Amplex Red/HRP in respiration buffer (125 mM KCl, 20 mM HEPES, 3 mM magnesium acetate, 0.4 mM EGTA, 5 mM KH₂PO₄, and 2 mg/ml BSA, at pH 7.1) with 5 mM succinate ± 1 mM ADP. Fluorescence was measured in a Tecan M200 Pro plate reader (λ_ex 540 nm/λ_em 590 nm). H₂O₂ concentrations were calculated using a H₂O₂ standard curve (0.125–8 μM range) and normalized to total protein.

Western blot analysis

Total protein was collected from primary myocytes using RIPA/HBSS as described previously (11), and protein concentration was determined by BCA Assay. Proteins were resolved by SDS-PAGE followed by electroblotting onto nitrocellulose membrane. Membranes were blocked in 5% nonfat milk and hybridized with primary antibodies overnight at 4°C in Tris-buffered saline (TBS) containing 0.1% Tween, 1% BSA unless otherwise indicated. All antibodies were purchased from Cell Signaling Technology (Danvers, MA, USA) unless otherwise specified. Membranes were incubated at RT with phospho-FoxO3a in TBS with 0.1% Tween, 2% BSA for 3 h or with FoxO3a in TBS with 0.1% Tween, 1% BSA for 2 h. The TnI antibody was described above. Antibodies against β-tubulin (developed by Dr. Michael Klymkowsky) and MHC (MF20, developed by Dr. Donald Fischman) were obtained from the Developmental Studies Hybridoma Bank developed under the auspices of the U.S. National Institute of Child Health and Human Development and maintained by the University of Iowa, Department of Biology (Iowa City, IA, USA). Band intensities were quantitated using ImageJ 1.40g software (U.S. National Institutes of Health, Bethesda, MD, USA).

Statistical analysis

All data are presented as means ± se or means ± sd. Significant differences between mean values were determined by unpaired Student's t test. A value of P ≤ 0.05 was considered significant and is denoted by asterisks.

RESULTS

While the ERRs have been shown to regulate mitochondrial respiratory capacity and biogenesis, an integral part of skeletal muscle differentiation, the role of ERRγ in myocyte differentiation has not been studied. Therefore, we investigated the requirement for ERRγ in myogenesis using primary myocytes isolated from mice homozygous for the floxed ERRγ allele (M-ERRγWT) and muscle-specific ERRγ^−/− (M-ERRγ^−/−) mice. Generation of muscle-specific ERRγ-deficient mice was necessary since whole-body ERRγ knockouts exhibit high perinatal mortality (6). Mice containing LoxP sites flanking exon 2 of the Esrrg gene were crossed with mice expressing Cre recombinase driven by the muscle creatine kinase (Mck) gene promoter (46), and disruption of the Esrrg locus in muscle was confirmed by PCR analysis of genomic DNA from skeletal muscle (Supplemental Fig. S1A, B). Transcript analysis in skeletal muscle and liver confirmed muscle-specific deletion of ERRγ in M-ERRγ^−/− tissues (Supplemental Fig. S1C). In addition, detection of ERRγ protein by immunohistochemistry showed strong nuclear staining in M-ERRγWT gastrocnemius muscle, while no ERRγ staining was observed in M-ERRγ^−/− gastrocnemius (Supplemental Fig. S1D). Primary myocytes were isolated from the leg muscles of 5-d-old M-ERRγWT and M-ERRγ^−/− mice. M-ERRγ^−/− MBs exhibit slower growth than M-ERRγWT MBs. When induced to differentiate, M-ERRγ^−/− myocytes formed fewer and shorter MTs relative to M-ERRγWT myocytes at MT day 1 (MTd1) and continued to lag in size and number throughout differentiation (Fig. 1A).

Figure 1.

Differentiation is impaired in M-ERRγ^−/− primary myocytes. A) Bright-field images of M-ERRγWT and M-ERRγ^−/− primary myocytes were taken at the indicated days at ×100. Cells were analyzed in triplicate in several independent trials, and representative images are shown. B) Immunofluorescence analysis was performed for the structural targets actinin and TnI at MTd4 using primary antibodies against the indicated proteins and detected with Alexa Fluor 488 secondary antibody. Representative images from ≥50 cells analyzed are shown at ×600. C) Real-time PCR was carried out for myogenin and MyoD and normalized to 36B4 mRNA levels. Triplicate samples were analyzed, and results are expressed as means ± sd. D) Expression of the sarcomeric components TnI slow, skeletal actin, and MHC slow was analyzed by real-time PCR and normalized to 36B4. Samples were analyzed in triplicate, and data are expressed as means ± sd. *P < 0.05. E) Western blot analysis for MHC and TnI expression was performed in M-ERRγWT and M-ERRγ^−/− myocytes at the indicated differentiation days. β-Tubulin was included as a loading control. Scale bars = 100 μm (A); 10 μm (B).

To assess sarcomere formation, we performed immunofluorescence analysis of actinin and TnI at MTd4 (Fig. 1B). Although M-ERRγWT myocytes exhibited the characteristic dentate pattern of actinin expression within assembled sarcomeres, M-ERRγ^−/− MTs showed a diffuse actinin organization pattern that was more perinuclear, consistent with the immature appearance of M-ERRγ^−/− MTs. M-ERRγ^−/− MTs also showed reduced fluorescence staining of TnI, suggesting reduced expression compared to M-ERRγWT MTs. In agreement, Western blot analysis demonstrated that protein levels of MHC and TnI were reduced in M-ERRγ^−/− myocytes throughout differentiation (Fig. 1E), consistent with the impaired fusion of MBs and slower growth in late-stage M-ERRγ^−/− MTs. We then investigated the transcript expression of differentiation markers, including MRFs and skeletal muscle-specific contractile proteins. MyoD or myogenin transcripts showed a normal expression pattern in M-ERRγ^−/− myocytes throughout differentiation (Fig. 1C). TnI transcript was significantly down-regulated only in M-ERRγ^−/− myocytes at MTd2, while skeletal actin was significantly up-regulated in M-ERRγ^−/− myocytes at MTd1 relative to M-ERRγWT myocytes (Fig. 1D). These results suggest that the myogenic program driven by MRFs is not disrupted in M-ERRγ^−/− myocytes.

Mitochondrial number and distribution are altered in M-ERRγ^−/− myocytes

ERRγ regulates mitochondrial biogenesis factors and enzymes involved in oxidative phosphorylation. Therefore, we investigated mitochondrial distribution and number in M-ERRγ^−/− myocytes relative to M-ERRγWT myocytes. Mitochondria in M-ERRγ^−/− myocytes at MTd4 exhibited a perinuclear distribution, characteristic of immature MTs, in contrast to M-ERRγWT MTs, in which mitochondria were positioned along the myofibril (Fig. 2A). Mitochondrial DNA, a quantitative measure of mitochondrial number, was significantly reduced in M-ERRγ^−/− MTd3 myocytes, while no difference was observed in earlier differentiation stages (Fig. 2B). We then investigated expression of genes encoding mitochondrial ETC protein subunits. Despite being previously implicated as direct ERRγ targets (5), no significant dysregulation was observed for Ndufa8, Sdha, Cycs, or Cox6c in M-ERRγ^−/− myocytes (Fig. 2C). Concurrently, ERRα and PGC-1α transcripts were actually up-regulated in MTd1 or MTd1 and MTd2 myocytes, respectively. Notably, this likely compensatory increase was not sufficient to rescue MT formation in the ERRγ-deficient cells. Furthermore, no significant differences in expression of Tfam or Gabpa, important regulators of mitochondrial biogenesis and function, were found between M-ERRγWT and M-ERRγ^−/− myocytes (data not shown). These results suggest that reduced mitochondrial number and altered distribution may contribute to respiratory dysfunction and impaired M-ERRγ^−/− MT formation. However, these reductions appear to be independent of global ERRγ-dependent regulation of mitochondrial enzyme gene programs.

Figure 2.

M-ERRγ^−/− primary myocytes exhibit mitochondrial defects. A) Immunofluorescence analysis for cytochrome c was assessed in M-ERRγWT and M-ERRγ^−/− myocytes at MTd4. Detection of primary antibodies was carried out using Alexa Fluor 568-conjugated secondary antibodies. At least 50 cells were analyzed, and representative images are shown at ×600. Scale bars = 10 μm. B) To assess mitochondrial number, DNA was isolated from M-ERRγWT and M-ERRγ^−/− myocytes at the indicated days. PCR was carried out for the mitochondria-encoded gene CoxI, and the nuclear encoded gene TFIID, and mitochondrial genome copy number was assessed from the ratio of CoxI to TFIID. Samples were analyzed in triplicate, and results are expressed as means ± se. *P < 0.05. C) Expression of ERRα, PGC-1α, and components of the ETC Ndufa8, Sdha, Cyt c, and Cox6c were analyzed by real-time PCR and normalized to 36B4 expression. Samples were analyzed in triplicate, and data are expressed as means ± sd. *P < 0.05.

Metabolic derangement in M-ERRγ^−/− myocytes

On the basis of the reduced mitochondrial density in M-ERRγ^−/− myocytes, we assessed rates of glucose oxidation and FAO in M-ERRγWT and M-ERRγ^−/− myocytes. In M-ERRγWT myocytes, we observed a 62% increase in glucose oxidation rates from MBs to MTd2 (Fig. 3A), while the rate only rose 19% in M-ERRγ^−/− myocytes. This resulted in 30% lower glucose oxidation rates in M-ERRγ^−/− compared to M-ERRγWT MTd2 myocytes. Interestingly, we observed an increased FAO rate in M-ERRγ^−/− MBs (+34%) and MTd2 (+36%) relative to M-ERRγWT myocytes using ³H-lauric acid, a medium-chain fatty acid (MCFA), as the substrate. However, when using ³H-palmitic acid, a long-chain fatty acid (LCFA), oxidation was reduced in M-ERRγ^−/− MTd2 (−47%) relative to M-ERRγWT myocytes. These results indicate that M-ERRγ^−/− myocytes selectively metabolize MCFAs instead of LCFAs or glucose for energy generation.

Figure 3.

Oxidative metabolism is impaired in M-ERRγ^−/− MTs. A) Glucose oxidation rates were determined by ¹⁴CO₂ release after 1 h incubation with ¹⁴C-glucose. ³H-lauric acid (C12) and ³H-palmitic acid (C16) oxidation rates in M-ERRγWT and M-ERRγ^−/− myocytes were determined by ³H₂O release after a 2-h incubation with the indicated fatty acid. Experiments were performed in triplicate, and results are expressed as means ± se. *P < 0.05. B) Glucose uptake was determined by incubating M-ERRγWT and M-ERRγ^−/− MTd2 with 20 μM 2-NBDG for 20 min at 37°C. Fluorescence was measured on a Tecan Infinite M1000 plate reader (λ_ex 465 nm/λ_em 540 nm). Experiments were performed in triplicate, and results are expressed as means ± se. *P < 0.05. C) Uptake of BODIPY-labeled lauric acid or palmitic acid in M-ERRγWT and M-ERRγ^−/− MTd3 myocytes was assessed by fluorescence microscopy at ×600. At least 50 cells were analyzed from duplicate samples, and representative images are shown. Scale bars = 10 μm. D) Expression of metabolic transcripts in M-ERRγWT and M-ERRγ^−/− myocytes was analyzed by real-time PCR and normalized to 36B4 mRNA levels. Samples were analyzed in triplicate, and data are expressed as means ± se. *P < 0.05.

To determine whether these rate changes were due to altered substrate uptake, we assessed glucose uptake in M-ERRγWT and M-ERRγ^−/− myocytes using the fluorescent glucose analog 2-NBDG (Fig. 3B). Compared to M-ERRγWT, glucose uptake was significantly impaired (−60%) in M-ERRγ^−/− MTd2 myocytes, consistent with the shift from glucose oxidation toward FAO in M-ERRγ^−/− myocytes. Since M-ERRγ^−/− myocytes preferentially metabolize MCFAs, we compared lipid uptake in M-ERRγWT and M-ERRγ^−/− myocytes using BODIPY-labeled palmitic acid or lauric acid by fluorescence microscopy (Fig. 3C). While the uptake of BODIPY-palmitic acid was equivalent, M-ERRγ^−/− MTs accumulated more BODIPY-lauric acid than M-ERRγWT MTs, consistent with the observed difference in FAO.

Given the altered metabolic profile in M-ERRγ^−/− myocytes, we assessed the expression of genes involved in uptake and metabolism of glucose and fatty acids (Fig. 3D). Expression of the fatty acid transporter CD36 was down-regulated in M-ERRγ^−/− myocytes, consistent with reduced utilization of LCFAs. However, the β-oxidation enzyme gene, MCAD, was unchanged in M-ERRγ^−/− myocytes compared to M-ERRγWT (data not shown), while mtCK2 expression was up-regulated in M-ERRγ^−/− myocytes throughout differentiation. Muscle glucose transporters were regulated in an opposing pattern. While GLUT4 transcript was up-regulated, GLUT1 expression was down-regulated in M-ERRγ^−/− myocytes, consistent with their reduced basal glucose uptake. Finally, PDK4 expression was higher in M-ERRγ^−/− myocytes throughout differentiation, correlating with the oxidative substrate switch from glucose oxidation toward FAO in these cells.

Perturbed substrate metabolism is associated with increased ROS in M-ERRγ^−/− myocytes. The metabolic shift from glucose toward fatty acids in the context of reduced mitochondrial number in M-ERRγ^−/− MTs could potentially increase generation of ROS. Therefore, we measured H₂O₂ production in M-ERRγWT and M-ERRγ^−/− myocytes using Amplex Red (10-acetyl-3,7-dihydroxyphenoxazine). H₂O₂ production was increased in M-ERRγ^−/− MBs (+40%) and MTd2 (+70%) relative to M-ERRγWT myocytes (Fig. 4A, top panel). To determine whether ROS production was also elevated in mitochondria of adult M-ERRγ^−/− skeletal muscle, we analyzed H₂O₂ production in isolated mitochondria using Amplex Red (Fig. 4A, bottom panel). H₂O₂ levels were lower in M-ERRγ^−/− mitochondria prior to substrate addition. However, ROS production was significantly elevated in M-ERRγ^−/− mitochondria relative to M-ERRγWT in the presence of the respiratory substrate succinate or with succinate+ADP. These findings reinforce our results in M-ERRγ^−/− myocytes and support that oxidative stress originates from mitochondrial defects in ERRγ-deficient myocytes. In response to increased oxidative stress, antioxidant gene programs were up-regulated in M-ERRγ^−/− myocytes (Fig. 4B). SOD2, UCP2 and UCP3 showed a progressive increase in M-ERRγ^−/− myocytes despite impaired MT growth, while Gpx3 and catalase were more consistently elevated throughout differentiation. In contrast, Gpx1 was not induced in M-ERRγ^−/− myocytes relative to M-ERRγWT myocytes. Collectively, these findings indicate that oxidative stress is elevated in M-ERRγ^−/− myocytes and may contribute to impaired MT formation.

Figure 4.

ROS production is elevated in M-ERRγ^−/− primary myocytes. A) Production of hydrogen peroxide (H₂O₂) was assessed using Amplex Red. Primary myocytes (top) or isolated mitochondria (bottom) from M-ERRγWT and M-ERRγ^−/− myocytes were incubated with 50 μM Amplex Red and 0.1 U/ml horseradish peroxidase for 30 min at 37°C. Isolated mitochondria were incubated with no respiratory substrate (no subs), 5 mM succinate (Succ), or 5 mM succinate plus 1 mM ADP (succ+ADP). Fluorescence was measured on a Tecan Infinite M200 Pro plate reader (λ_ex 540 nm/λ_em 590 nm). Experiments were performed in triplicate, and results are expressed as means ± se. *P < 0.05. B) Expression of the antioxidant SOD2, catalase, Gpx1, Gpx3, and mitochondrial uncoupling proteins 2 and 3 (UCP2 and UCP3) was analyzed by real-time PCR in M-ERRγWT and M-ERRγ^−/− primary myocytes during differentiation. Results were normalized to 36B4 mRNA levels. Experiments were performed in triplicate, and results are expressed as means ± se. *P < 0.05.

Activation of atrophy pathways in M-ERRγ^−/− myocytes

ROS can impair myocyte differentiation and growth through multiple pathways involving up-regulation of muscle-specific atrogenes. We investigated the activation of NF-κB and atrophy pathways in M-ERRγ^−/− myocytes. NF-κB phosphorylation was increased in M-ERRγ^−/− MBs and MTs relative to M-ERRγWT by Western blot analysis (Fig. 5A). In addition, immunofluorescence analysis showed increased phospho-NF-κB nuclear localization in M-ERRγ^−/− MBs compared to M-ERRγWT (Fig. 5B). Along with elevated ROS levels, these results suggested that NF-κB may be up-regulating muscle atrophy gene pathways in M-ERRγ^−/− myocytes. Real-time PCR analysis revealed that expression of the NF-κB target MuRF1 was up-regulated in M-ERRγ^−/− MBs and MTd1 compared with M-ERRγWT myocytes, indicating that NF-κB may promote atrophy at early stages of myogenesis in M-ERRγ^−/− myocytes (Fig. 5C). Consistently, transcript levels of additional NF-κB targets, IL-6 and IκBα, were up-regulated throughout differentiation in M-ERRγ^−/− myocytes. NF-κB can also influence myogenesis through transcriptional regulation of MyoD, YY1, and cyclin D1 (47–49). However, no significant change in MyoD expression was observed in M-ERRγ^−/− myocytes (Fig. 1C). While YY1 was slightly elevated in M-ERRγ^−/− myocytes (data not shown), transcript levels of YY1 target genes, TnI and MHC (48), were unaltered in M-ERRγ^−/− myocytes. Thus, the impairment in MT formation in M-ERRγ^−/− myocytes is not due to inhibition of myogenic pathways but is mediated by the activation of atrogenes.

Figure 5.

NF-κB is activated in M-ERRγ^−/− myocytes. A) Top panels: phosphorylation status of NF-κB was analyzed in M-ERRγWT and M-ERRγ^−/− myocytes by Western blot. β-Tubulin was included as a loading control. Bottom panel: quantitation of P-NF-κB vs. NF-κB levels from ≥2 independent lysates was performed using ImageJ software. B) Immunofluorescence analysis indicates that P-NF-κB is localized in the nucleus of M-ERRγ^−/− MBs. At least 50 cells were analyzed, and representative images are shown at ×600. Scale bars = 10 μm. C) Real-time PCR analysis of NF-κB target genes MuRF1, interleukin-6 (IL-6), and IκBα was performed and normalized to 36B4 expression levels. Experiments were performed in triplicate, and results are expressed as means ± se. *P < 0.05.

To determine whether NF-κB mediates muscle atrophy in M-ERRγ^−/− myocytes, we treated M-ERRγWT and M-ERRγ^−/− myocytes with the NF-κB inhibitor QNZ. QNZ treatment promoted MT formation in M-ERRγ^−/− myocytes, resulting in wider MTs relative to vehicle-treated M-ERRγ^−/− myocytes (Fig. 6A). Consistent with the partial rescue of MT formation, MuRF1 expression returned to wild-type levels with QNZ treatment in M-ERRγ^−/− MTd1 and MTd2 myocytes (Fig. 6B). Similarly, NF-κB inhibition with QNZ also blunted SOD2 induction (−32%) in M-ERRγ^−/− MTd1 myocytes, while IκBα activation was reversed by QNZ treatment in M-ERRγ^−/− myocytes at MTd1 and MTd2. QNZ treatment did not affect elevated IL-6 expression in M-ERRγ^−/− myocytes, although it modestly reduced IL-6 in M-ERRγWT MTd2 myocytes. These results indicate that other signaling pathways may regulate IL-6 expression in M-ERRγ^−/− myocytes. Overall, these results indicate that NF-κB is an important early mediator of muscle atrophy in M-ERRγ^−/− myocytes.

Figure 6.

NF-κB inhibitor QNZ partially rescues MT formation and reduces NF-κB target gene expression in M-ERRγ^−/− primary myocytes. A) Bright-field images of M-ERRγWT and M-ERRγ^−/− MTd2 myocytes treated with the indicated concentrations of QNZ (Santa Cruz Biotechnology) when switched into differentiation medium. At least 50 cells from duplicate samples were analyzed, and representative images are shown at ×100. Scale bars = 100 μm. B) Expression of MuRF1, IκBα, IL-6, and SOD2 in M-ERRγWT and M-ERRγ^−/− primary myocytes treated with the indicated concentrations of QNZ during differentiation was analyzed by real-time PCR and normalized to 36B4 mRNA levels. Experiments were performed in triplicate, and results are expressed as means ± se. *P < 0.05.

Activation of FoxO pathways in M-ERRγ^−/− myocytes

Increased ROS levels can also activate the FoxO transcription factors and promote muscle atrophy. Therefore, we assessed the expression and phosphorylation status of FoxO in M-ERRγ^−/− myocytes. Transcript levels of FoxO1 were elevated throughout differentiation, while FoxO3a expression was significantly up-regulated only at MTd3 in M-ERRγ^−/− relative to M-ERRγWT myocytes (Fig. 7A). Western blot analysis revealed that while FoxO1 phosphorylation was enhanced in M-ERRγ^−/− MBs, phosphorylation levels dropped in M-ERRγ^−/− MTd1 and MTd2 myocytes, indicating FoxO1 activation at these stages (Fig. 7B). Likewise, FoxO3a phosphorylation was also decreased in M-ERRγ^−/− MBs and MTd2 myocytes, consistent with activation of FoxO3a in M-ERRγ^−/− myocytes. We then assessed expression of the FoxO target gene, atrogin-1, which was up-regulated ∼2-fold in M-ERRγ^−/− during differentiation relative to M-ERRγWT myocytes (Fig. 7A). The observed increase in FoxO transcript and activity indicates that the FoxO factors may promote muscle atrophy in M-ERRγ^−/− myocytes.

Figure 7.

FoxO1 and FoxO3a are activated in M-ERRγ^−/− mice. A) Expression of FoxO1, FoxO3a, and their target gene atrogin-1 in M-ERRγWT and M-ERRγ^−/− myocytes during differentiation was analyzed by real-time PCR and normalized to 36B4 expression levels. Experiments were performed in triplicate, and results are expressed as means ± se. *P < 0.05. B) Top panels: Western blot analysis of phospho-FoxO1, FoxO1, phospho-FoxO3a, and FoxO3a in M-ERRγWT and M-ERRγ^−/− myocytes during differentiation. Arrow indicates position of P-FoxO1 band. β-Tubulin was included as a loading control. Bottom panel: quantitation of phospho-FoxO1 vs. FoxO1 and phospho-FoxO3a vs. FoxO3a from ≥2 independent lysates was performed using ImageJ software.

Several upstream signal transduction pathways are modulated by ROS, including the ERK MAPK, p38 MAPK, JNK MAPK, AMPK, and Akt pathways (refs. 50–53 and Supplemental Fig. S2). We compared the activation status of these pathways in M-ERRγWT and M-ERRγ^−/− myocytes. However, we did not find significantly altered phosphorylation patterns of these factors that would contribute to muscle atrophy pathway activation or inhibition of MT formation in M-ERRγ^−/− myocytes.

Treatment of M-ERRγ^−/− with the free radical scavenger NAC rescues MT formation and down-regulates atrogene programs

To determine whether elevated ROS levels inhibit M-ERRγ^−/− MT formation through atrophy pathway activation, we treated myocytes with the antioxidant NAC. NAC treatment increased MB proliferation in both M-ERRγWT (+23±5.9% at MTd1 and +39±6.4% at MTd2) and M-ERRγ^−/− (+56±5.0% at MTd1 and +48±9.0% at MTd2) myocytes, as determined by cell counting. While NAC modestly enhanced M-ERRγWT MT formation, antioxidant treatment completely rescued the phenotype of M-ERRγ^−/− myocytes, such that by MTd2, M-ERRγ^−/− MTs were equivalent to NAC-treated M-ERRγWT MTs (Fig. 8A). Immunofluorescence analysis of cytochrome c demonstrated that NAC also rescued the perinuclear mitochondrial distribution in M-ERRγ^−/− MTs, promoting appropriate localization along MTs similar to M-ERRγWT (Fig. 8B). In addition, NAC treatment reduced elevated atrogin-1 expression levels in M-ERRγ^−/− MBs and MTd2 myocytes to M-ERRγWT levels (Fig. 8C). MuRF1 up-regulation in M-ERRγ^−/− MBs was also significantly reduced by NAC treatment. Although NAC treatment had no effect on MuRF1 expression in MTd2 M-ERRγ^−/− or M-ERRγWT myocytes, this is consistent with NF-κB influencing only early MT formation. Collectively, these results indicate that elevated ROS levels in M-ERRγ^−/− myocytes contribute to up-regulation of atrophy pathways, leading to inhibition of MT formation.

Figure 8.

NAC treatment rescues MT formation and reduces atrogene expression in M-ERRγ^−/− myocytes. A) Bright-field images of M-ERRγWT and M-ERRγ^−/− MTd2 myocytes treated with the indicated concentrations of NAC at the time of plating. Cells were analyzed in triplicate in 3 independent trials, and representative images are shown at ×100. B) Immunofluorescence analysis for cytochrome c in M-ERRγWT and M-ERRγ^−/− MTd3 myocytes treated with 1 mM NAC when plated. At least 50 cells were analyzed, and representative images are shown at ×600. C) Real-time PCR analysis of atrogin-1 and MuRF1 expression in M-ERRγWT and M-ERRγ^−/− myocytes treated with 1 mM NAC when plated. Results were normalized to 36B4 mRNA expression levels. Experiments were performed in triplicate, and results are expressed as means ± se. Scale bars = 100 μm (A); 10 μm (B). *P < 0.05.

DISCUSSION

In this study, we demonstrate that ERRγ is essential for differentiation to proceed in skeletal muscle cells. Loss of ERRγ severely impairs MT formation and results in altered mitochondrial distribution. In addition, M-ERRγ^−/− myocytes exhibit a metabolic switch from oxidation of glucose to MCFAs, associated with increased ROS and up-regulation of muscle atrophy pathways. We demonstrate that elevated ROS are the significant factor leading to impaired differentiation since MT formation, mitochondrial distribution, and atrophy pathway activation are rescued with antioxidant treatment.

M-ERRγ^−/− myocytes exhibit impaired MT formation and immature sarcomere structure. In our previous characterization of differentiation in ERRα^−/− myocytes (11), we found that sarcomere assembly was also impaired, while expression of the structural components, TnI, and MHC was actually induced at both the protein and RNA levels. In contrast, M-ERRγ^−/− myocytes showed no changes in sarcomeric transcript expression but reduced protein levels of TnI and MHC. In ERRα^−/− myocytes, delayed differentiation was due to perturbations in ERK-MAP kinase signaling due to MAP kinase phosphatase-1 dysregulation (11). However, no aberrant ERK activation was observed in M-ERRγ^−/− myocytes (Supplemental Fig. S2A), indicating that disruption of MT formation occurs through distinct mechanisms in ERRα^−/− and ERRγ^−/− myocytes.

Reduced glucose oxidation was accompanied by a parallel decrease in 2-NBDG uptake in M-ERRγ^−/− myocytes. Concurrently, the FAO profile reflected a dramatic shift from LCFA to MCFA oxidation associated with increased BODIPY-lauric acid uptake into M-ERRγ^−/− MTs. Interestingly, CD36 expression was down-regulated in M-ERRγ^−/− myocytes, which may impair LCFA oxidation. MCFAs do not require active transport for uptake into cells, but efficient uptake of LCFAs requires fatty acid transport proteins, such as CD36, in skeletal muscle (54). Recent studies suggest that CD36 protein, in addition to the plasma membrane, may also localize to the mitochondrial membrane in skeletal muscle and is required for LCFA oxidation (55). The mechanisms altering energy substrate utilization and gene expression are potentially complex in this system. Previous studies using in vivo ERRγ gain- and loss-of-function models showed changes in mitochondrial ETC/Oxphos gene expression and oxidative capacity, while changes in substrate preference were not characterized (6–8). Genomic studies did not identify the Cd36 or Slc24a1 (GLUT1) genes, which were down-regulated in M-ERRγ^−/− myocytes, as ERRγ targets; however, Slc24a4 (GLUT4) and Pdk4 are candidate ERR target genes (5, 6, 56). As a result, future studies will explore the involvement of these genes in altered skeletal muscle substrate uptake and oxidation, the mechanism of their dysregulation, and assess basal and insulin-stimulated glucose uptake in M-ERRγ^−/− mice.

The increased reliance of M-ERRγ^−/− myocytes on MCFA as a substrate for energy generation may lead to elevated mitochondrial ROS production, since MCFAs are easily utilized but generate fewer ATPs per molecule than LCFAs. Increased fatty acid flux through the ETC would cause more ROS production in M-ERRγ^−/− myocytes, especially in the context of mitochondrial defects or reduced mitochondrial number. Our data do not support an inherent deficiency in mitochondrial biogenesis in M-ERRγ^−/− myocytes. The mtDNA content is equivalent in M-ERRγWT and M-ERRγ^−/− MBs and only significantly reduced at MTd3, likely as a result of impaired MT formation. ROS generation was also elevated in adult M-ERRγ^−/− skeletal muscle, which show no changes in mtDNA levels or citrate synthase activity, an indicator of mitochondrial capacity (data not shown). Our findings are in agreement with data from whole-body ERRγ-null mice in which defects in cardiac mitochondrial function were not accompanied by reduced mtDNA content (6). These data support the notion that increased ROS production is a consequence of inherent metabolic defects, resulting from ERRγ deletion, which is currently being characterized in adult M-ERRγ^−/− skeletal muscle.

Induction of antioxidant genes, while consistent with ROS-activated signaling, was not sufficient to reduce ROS levels in M-ERRγ^−/− MTd2 myocytes. UCPs may be activated by ROS or byproducts of lipid peroxidation, leading to uncoupling and reduction in ROS generation (12). In addition, UCPs may regulate fatty acid metabolism by exporting fatty acid anions from the mitochondrial matrix, allowing increased FAO rates and reducing mitochondrial membrane potential. Whether UCP2 and UCP3 mediate the metabolic shift toward FAO in M-ERRγ^−/− myocytes is currently being investigated in M-ERRγ^−/− mice. Gpx1 and Gpx3 were differentially regulated in the M-ERRγ^−/− myocytes. Gpx1 is a target of NF-E2-related factor 2, which regulates ROS detoxification enzymes and mitochondrial biogenesis through nuclear respiratory factors (NRFs; refs. 57, 58). Because expression of NRF-2/Gabp2 and Gpx1 was unaltered in M-ERRγ^−/− myocytes during differentiation, our data do not support a role for the NF-E2 pathway in the response to oxidative stress in these cells. Although Gpx3 was previously shown to improve glucose uptake in skeletal muscle by reducing ROS levels (59), Gpx3 induction was either not sufficient to decrease ROS in M-ERRγ^−/− myocytes or may have an alternate function in this system.

Notably, NAC treatment of M-ERRγ^−/− myocytes restored MT formation and mitochondrial distribution, indicating the significant influence of elevated ROS on differentiation. Oxidant signaling plays a complex role in regulating cellular proliferation and differentiation. ROS levels rise during normal differentiation in C2C12 myocytes, correlating with increased mitochondrial biogenesis (60), and low levels of ROS can produce adaptive responses in antioxidant gene expression (14). In agreement, a recent study (61) suggests that mitochondrial ROS production stimulates differentiation in H9c2 rat cardiomyocytes and that treatment with mitochondrial-targeted antioxidants inhibits MT formation. Therefore, basal ROS levels may be an important stimulus in the normal progression of myogenesis, while sustained, elevated ROS can damage proteins, lipids, and DNA, and inhibit myogenesis in C2C12 myocytes (15, 62). M-ERRγ^−/− myocytes have significantly increased ROS levels starting from MBs, which are detrimental to MT formation. Thus, the early and sustained production of ROS in M-ERRγ^−/− myocytes overrrides normal ROS signaling, since MT formation is rescued by antioxidant treatment from the MB stage.

We demonstrate that elevated ROS levels in M-ERRγ^−/− myocytes lead to sustained activation of atrogenes downstream of the NF-κB and FoxO transcription factors. NF-κB was activated early in myogenesis, correlating with MuRF1 up-regulation at MBs and MTd1 in M-ERRγ^−/− myocytes. The NF-κB inhibitor QNZ resulted in significant down-regulation of MuRF1 and IκBα at both MTd1 and MTd2 in M-ERRγ^−/− myocytes, indicating that NF-κB is inhibited and correlating with the modest improvement in MT formation. In contrast, IL-6 expression remained elevated with QNZ treatment. IL-6 expression can be regulated by various signaling pathways, including Ca²⁺/nuclear factor of activated T cells (NFAT), STAT3, and p38 MAPK pathways, in addition to NF-κB in skeletal muscle. However, we did not see changes in STAT3 (data not shown), or p38 (Supplemental Fig. S2) pathways that would promote IL-6 induction. Our data suggest that IL-6 regulation was independent of ROS-mediated effects on M-ERRγ^−/− MT formation, since NAC treatment rescued the phenotype but not IL-6 induction (data not shown). The potential contribution of IL-6 up-regulation to impaired MT formation or metabolic changes in M-ERRγ^−/− myocytes is the subject of ongoing studies. Expression of an additional NF-κB target gene, SOD2, was only reduced in MTd1, indicating that increased NF-κB activation is inhibitory at early myogenic stages. Consistently, QNZ treatment only partially rescued MT formation, suggesting that additional ROS-responsive pathways influencing myocyte differentiation may be activated.

We found sustained activation of FoxO1 and FoxO3a during differentiation in M-ERRγ^−/− myocytes. Although there may be a genomic component to ERRγ regulation of FoxOs, the mechanisms are likely complex. FoxO1 is up-regulated in ERRγ^−/− hearts (6), consistent with our findings, while overexpression of the ERR coactivator, PGC-1α, inhibits FoxO3a occupation of the atrogin-1 promoter (40). Our current data indicate that ROS is the major factor influencing FoxO activation because multiple ROS-responsive pathways were activated in M-ERRγ^−/− myocytes. FoxO activity is also regulated post-translationally by Akt phosphorylation, which inhibits FoxO activity (30–31). However, in M-ERRγ^−/− myocytes FoxO1 and FoxO3a, phosphorylation decreased during differentiation. Because no changes in Akt pathway activation were observed in M-ERRγ^−/− myocytes (Supplemental Fig. S2B), our data support that the elevated ROS signaling accounts for FoxO activation. The sustained up-regulation of atrogin-1 throughout differentiation in M-ERRγ^−/− myocytes indicates that FoxO signaling may be more influential at later differentiation stages in maintaining atrophy pathway activation.

In summary, our data demonstrate that increased ROS are the major causative factor in impaired M-ERRγ^−/− MT formation via activation of atrophy gene programs. Thus, ERRγ may also have a role in muscle atrophy caused by mitochondrial dysfunction during aging or in metabolic conditions, such as diabetes, an area currently being studied in M-ERRγ^−/− mice. Overall, these results indicate that the ERRγ pathway may have a protective role in age-related or disuse atrophy.