Background: Skeletal muscle mass is determined by the balance between protein synthesis and degradation. Mammalian target of rapamycin complex 1 (mTORC1) is a master regulator of protein translation and has been implicated in the control of muscle mass. Inactivation of mTORC1 by skeletal muscle-specific deletion of its obligatory component raptor results in smaller muscles and a lethal dystrophy. Moreover, raptor-deficient muscles are less oxidative through changes in the expression PGC-1α, a critical determinant of mitochondrial biogenesis. These results suggest that activation of mTORC1 might be beneficial to skeletal muscle by providing resistance to muscle atrophy and increasing oxidative function. Here, we tested this hypothesis by deletion of the mTORC1 inhibitor tuberous sclerosis complex (TSC) in muscle fibers. Method: Skeletal muscles of mice with an acute or a permanent deletion of raptor or TSC1 were examined using histological, biochemical and molecular biological methods. Response of the muscles to changes in mechanical load and nerve input was investigated by ablation of synergistic muscles or by denervation . Results: Genetic deletion or knockdown of raptor, causing inactivation of mTORC1, was sufficient to prevent muscle growth and enhance muscle atrophy. Conversely, short-term activation of mTORC1 by knockdown of TSC induced muscle fiber hypertrophy and atrophy-resistance upon denervation, in both fast tibialis anterior (TA) and slow soleus muscles. Surprisingly, however, sustained activation of mTORC1 by genetic deletion of Tsc1 caused muscle atrophy in all but soleus muscles. In contrast, oxidative capacity was increased in all muscles examined. Consistently, TSC1-deficient soleus muscle was atrophy-resistant whereas TA underwent normal atrophy upon denervation. Moreover, upon overloading, plantaris muscle did not display enhanced hypertrophy compared to controls. Biochemical analysis indicated that the atrophy response of muscles was based on the suppressed phosphorylation of PKB/Akt via feedback inhibition by mTORC1 and subsequent increased expression of the E3 ubiquitin ligases MuRF1 and atrogin-1/MAFbx. In contrast, expression of both E3 ligases was not increased in soleus muscle suggesting the presence of compensatory mechanisms in this muscle. Conclusions: Our study shows that the mTORC1- and the PKB/Akt-FoxO pathways are tightly interconnected and differentially regulated depending on the muscle type. These results indicate that long-term activation of the mTORC1 signaling axis is not a therapeutic option to promote muscle growth because of its strong feedback induction of the E3 ubiquitin ligases involved in protein degradation. Keywords: Skeletal muscle, Hypertrophy, Atrophy, Mammalian target of rapamycin complex 1 (mTORC1), Raptor, Tuberous sclerosis complex (TSC), PKB/Akt, FoxO, MuRF1, Atrogin-1/MAFbx * Correspondence: email@example.com Equal contributors Biozentrum, University of Basel, Basel CH-4056, Switzerland Full list of author information is available at the end of the article © 2013 Bentzinger et al.; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Bentzinger et al. Skeletal Muscle 2013, 3:6 Page 2 of 16 http://www.skeletalmusclejournal.com/content/3/1/6 Background rapamycin prevents IGF1-induced growth of myotubes Skeletal muscle is the largest organ, accounting for 30 to , inhibits compensatory hypertrophy in rat skeletal 40% of the total body weight. Muscle tissue is highly muscle  and blocks the growth-stimulating activity of plastic and adapts its size to physical demand. For clenbuterol . Moreover, transgenic overexpression of example, increase in load causes hypertrophy whereas TSC1 causes muscle atrophy in mice , while acute unloading causes atrophy. Importantly, muscle atrophy overexpression of Rheb induces muscle hypertrophy and subsequent wasting are also hallmarks of pathology . Finally, mice deficient for S6K1 show a reduction in muscular dystrophies or in cachexia, the latter being a of muscle fiber size and a blunted response to IGF1 . secondary consequence of a primary disease (for ex- In agreement with these findings, we recently showed ample, AIDS, cancer or sepsis). Several lines of evidence that mice with a skeletal muscle-specific knockout for indicate that muscle mass is controlled by the balance raptor (called RAmKO for raptor muscle knockout) have between protein synthesis and protein degradation [1,2]. a reduced muscle mass and suffer from a progressive In skeletal muscle, protein synthesis can be induced by dystrophy, which causes their death at the age of four to IGF1 (insulin-like growth factor-1), which in turn, acti- six months . Muscles of RAmKO mice also have a vates PI3K (phosphatidylinositol 3-kinase) and PKB decreased oxidative capacity, which can be restored by (protein kinase B; also called Akt). Activated PKB/Akt transgenic expression of PGC-1α . In addition, inhibits the protein complex TSC1-TSC2 (tuberous RAmKO mice show sustained activation of PKB/Akt sclerosis complex), which inactivates the small GTPase because of relieved feedback inhibition onto IRS1 (insulin protein Rheb (Ras homolog enriched in brain). Rheb receptor substrate-1) by the diminished activation of activates mammalian target of rapamycin complex 1 S6K . (mTORC1), which causes an increase in protein transla- Here we investigated the contribution of mTORC1 to tion by phosphorylating its two best characterized tar- muscle atrophy and hypertrophy by targeting rptor (the gets S6K (p70S6 kinase) and 4EBP (eIF-4E-binding gene encoding raptor) or Tsc1 (encoding TSC1) specific- protein). This IGF1-PI3K-PKB/Akt-mTOR signaling ally in mouse skeletal muscle. We show that deletion of pathway controls protein synthesis and cell size in sev- rptor prevents muscle hypertrophy and enhances muscle eral tissues [3,4]. atrophy. Surprisingly, sustained activation of mTORC1 Activation of PKB/Akt also negatively regulates pro- by the genetic deletion of Tsc1 does not induce hyper- tein degradation by phosphorylating the FoxO (Forkhead trophy but rather causes atrophy in all but soleus box O) transcription factors. Protein degradation is muscles. While the TSC1-deficient, hypertrophic soleus mainly carried out by enzymes of the ubiquitin- muscle is also resistant to denervation-induced atrophy, proteasomal and autophagosomal-lysosomal pathways tibialis anterior (TA) muscle atrophies like controls. Bio- [5,6]. Dephosphorylated FoxOs in the nuclei promote chemical characterization shows that regulation of the the expression of the two E3 ubiquitin ligases atrogin-1/ two E3 ligases atrogin-1/MAFbx and MuRF1 differs MAFbx and MuRF1 [7,8]. FoxOs have also been de- between TA and soleus muscles. Furthermore, we dem- scribed to drive expression of autophagy-related genes onstrate that all muscles show an increase in their oxida- [6,9]. The function of active PKB/Akt to simultaneously tive capacity upon mTORC1 activation. In summary, we stimulate protein synthesis and inhibit protein degra- demonstrate that the oxidative capacity in all skeletal dation may explain the profound hypertrophic effect of muscles is controlled by mTORC1, whereas the effect of constitutively active PKB/Akt [10,11]. sustained activation of mTORC1 on muscle size differs mTOR belongs to the PI3/PI4-kinase family; it is between muscles. Hence, our studies decipher a mech- highly conserved from yeast to human and assembles anism of biological robustness that balances the two into two structurally and functionally distinct multi- major metabolic pathways involved in the control of protein complexes, called mTORC1 and mTORC2 skeletal muscle mass. [12,13]. An essential component of mTORC1 is the pro- tein raptor (regulatory-associated protein of mTOR), Methods whereas rictor (rapamycin-insensitive companion of Mice mTOR) is an essential subunit of mTORC2 [3,4]. Most Mice were maintained in a conventional facility with a functions of mTORC1 are acutely inhibited by the fixed dark–light cycle. Studies were carried out according immunosuppressant rapamycin, whereas mTORC2 is to criteria outlined for the care and use of laboratory ani- only repressed by long-term application of rapamycin mals and with approval of the Swiss authorities. RAmKO . In skeletal muscle, the function of mTORC2 seems mice were generated and genotyped as described before to not be essential because mice deficient for rictor have . Floxed Tsc1 mice  were obtained from The no overt phenotype [15,16]. In contrast, mTORC1 par- Jackson Laboratory (Bar Harbor, Maine, USA) and mated ticipates in the control of muscle size. For example, with mice expressing Cre recombinase under the human Bentzinger et al. Skeletal Muscle 2013, 3:6 Page 3 of 16 http://www.skeletalmusclejournal.com/content/3/1/6 skeletal actin (HSA) promoter . Genotyping for the injected with 10 to 30 μl of a mixture containing the conditional Tsc1 allele was performed as described . respective shRNA plasmid and a plasmid coding for TSC-RAmKO mice were generated by intercrossing mice NLS-GFP (2 mg/ml of each construct). The fascia and carrying floxed rptor and Tsc1 alleles. Mice homozygous the skin were sutured and the electroporation was for both floxed alleles were mated with double heterozy- performed using an ECM 830 electroporation system gotes, which also carried the HSA-Cre transgene. Except (BTX Instruments Division, Harvard Apparatus Inc., for overloading experiments and Western blot analysis, Holliston, MA, USA). Eight pulses lasting 20 ms with only male TSCmKO mice were used. Both genders were the frequency of 1 Hz and the voltage set to 180 V/ used in RAmKO and TSC-RAmKO mice. All procedures cm were applied. Mice were analyzed four to six were performed in accordance with the Swiss regulations weeks after electroporation. for animal experimentation and they were approved by the veterinary commission of the Canton Basel-Stadt. Denervation, nerve crush and overloading Mice were anesthetized with ketamine (111 mg/kg) and Rapamycin treatment of mice xylazine (22 mg/kg) by intra-peritoneal injection and sur- Rapamycin treatment began three days before the mice gery was performed under aseptic conditions. For denerv- were challenged with functional overload (FO) or elec- ation, a segment (approximately 5 mm) of the sciatic troporation and continued until mice were sacrificed. nerve at the mid-thigh level was excised . To induce Rapamycin (LC Laboratories, Woburn, MA, USA), dis- muscle re-growth, the nerve was crushed with No 5 solved in saline containing 2% carboxymethylcellulose Dupont forceps (Fine Science Tools GmbH, Heidelberg, (Sigma-Aldrich, St. Louis, MO, USA), was delivered Germany) for 10 seconds at mid-thigh . To induce once daily by i.p. (intraperitoneal) injection at a dose of muscle hypertrophy, a functional overload of plantaris 1.5 mg/kg . muscle was introduced by surgical removal of soleus and gastrocnemius muscles . Surgery was performed on shRNA constructs one leg only. The plantaris muscle of the contralateral leg The methods to construct plasmids encoding shRNA served as control. and the sequences of the Cd4 control shRNA and the NLS-GFP construct have been described elsewhere . The murine 19 nucleotide target sequences correspond to: GTT GAT GCG TAA CCT TCT G (Tsc2), GAT GGA Antibodies CAC TGA TGT TGT G (Tsc1) and GAA TTT TGC TGA The antibodies used were from the following sources: TTT GGA A (rptor). rabbit polyclonal antibodies directed to 4E-BP1 (Phas-I) from Zymed (Life Technologies); those recognizing Tissue culture, transfections and shRNA efficiency Phospho-4E-BP1 (Ser65), PKB/Akt, mTOR, S6 Riboso- Adenoviruses encoding shRNA against Tsc2 and Cd4 mal Protein or Phospho-S6 Ribosomal Protein (Ser235/ were created by cloning the respective shRNA sequence 236) were all from Cell Signaling Technology Inc. (Dan- and H1 promoter from pSuper into pAd-DEST (Life vers, MA, USA); those against FoxO1a were from Technologies Europe B.V., Zug, Switzerland). To test the Abcam plc. (Cambridge, UK); those against TSC1 were efficiency of the Tsc2 shRNA, C2C12 myoblasts, cultured from Bethyl Laboratories (Montgomery, TX, USA). under standard conditions, were transfected with the Rabbit monoclonal antibodies directed against Phospho- Tsc2 and Cd4 shRNA viruses. The efficiency of the rptor Akt (Ser473), IRS-1, FoxO3a (75D8) and phospho- shRNA was tested by co-transfection with an expression FoxO1(Thr24)/FoxO3a(Thr32) (#9466) were from Cell plasmid encoding HA-tagged raptor into COS7 cells using Signaling Technology Inc. Mouse monoclonal antibodies Lipofectamine 2000 (Life Technologies). For PGC1β to α-actinin were purchased from Sigma and antibodies overexpression and knockdown, myoblasts were permitted against HA from Covance Inc. (Geneva, Switzerland). to fuse into multinucleated myotubes for 48 hr and cells Rat monoclonal antibodies directed to the Laminin B2 were infected with adenovirus preparations for an add- Chain (MAB1914) were from Chemicon and sold by itional 48 hr. Adenoviruses (Ad-GFP, Ad-PGC1β,Ad- Millipore AG (Zug, Switzerland). The TSC2 antibodies scrambled or Ad-siPGC1β)were kindlyprovided byBM used were described elsewhere . Mouse monoclonal Spiegelman (Harvard University, Boston, MA, USA). antibodies against myosin heavy chain: slow (A4.840), IIa/IIx (A4.74) and IIb (BF-F3) were purchased from Electroporation of muscle The Developmental Studies Hybridoma Bank (University Plasmids encoding shRNA constructs were electroporated of Iowa, Iowa City, Iowa, USA). Antibodies to puro- into muscle fibers as described before . Briefly, soleus mycin  were a kind gift of Dr. Philippe Pierre (CIML or TA muscle of anesthetized mice was exposed and Parc Scientifique de Luminy, Marseille, France). Bentzinger et al. Skeletal Muscle 2013, 3:6 Page 4 of 16 http://www.skeletalmusclejournal.com/content/3/1/6 Histology and immunohistochemistry Real-time PCR Muscles frozen in liquid nitrogen-cooled isopentane Total RNA was isolated (SV Total RNA Isolation Sys- were cut into 12 μm cross-sections. Cross-sections were tem, Promega AG, Dübendorf, Switzerland) from soleus fixed with 2% paraformaldehyde (PFA) and perme- muscles. RNA concentrations were adjusted between abilized with 1% Triton/PBS for 5 minutes, washed with samples and reverse transcription was carried out using 100 mM glycine/PBS for 15 minutes, blocked with 1% a mixture of oligodT and random hexamer primers BSA/PBS for 30 minutes, and incubated with the pri- (iScript cDNA Synthesis Kit, Bio-Rad Laboratories AG). mary antibody overnight at 4°C. Samples were subse- Sybr Green, real-time PCR analysis (Power SYBR Green quently washed three times for 10 minutes each with 1% Master Mix, Life Technologies) was performed using the BSA/PBS and stained with the appropriate fluorescence la- ABI Prism 7000 Sequence Detector. Expression levels beled secondary antibodies for 1 hr at room temperature. for each gene of interest were normalized to the expres- After washing with PBS, samples were mounted with sion of the housekeeping protein β-actin. The following Citifluor (Citifluor Ltd. London, UK). General histology primers were used: β-actin sense primer: 5' CAG CTT on cross-sections was performed using hematoxylin CTT TGC AGC TCC TT, antisense primer: 5' GCA GCG and eosin (H&E; Merck, Zug, Switzerland). NADH-TR ATATCG TCATCC A; atrogin-1/MAFbx sense primer: (Nicotinamide adenine dinucleotide hydrogen-tetrazolium 5' CTC TGT ACC ATG CCG TTC CT, antisense primer: reductase) staining was done as described . Methods 5' GGC TGC TGA ACA GAT TCT CC; MuRF-1 sense of SDH and COX staining were described elsewhere . primer: 5 ACC TGC TGG TGG AAA ACA TC, antisense Samples were dehydrated and mounted with DePeX primer: 5 AGG AGC AAG TAG GCA CCT CA; Pgc1α mounting medium (Gurr, BDH, VWR International sense primer: 5’ TGA TGT GAA TGA CTT GGA TAC GmbH, Dietikon, Switzerland). AGA CA, antisense primer: 5’ GCT CAT TGT TGT ACT GGT TGG ATA TG; Pgc1β sense primer: 5' GGC AGG TTC AAC CCC GA, antisense primer: 5' CTT GCT AAC In vivo protein synthesis ATC ACA GAG GAT ATC TTG. Quantification of mito- Protein synthesis was measured using the surface sens- chondrial DNA copy numbers was done as described . ing of translation (SUnSET) method  by i.p. injection of 0.040 μmol/g puromycin dissolved in 100 μlofPBS. Quantifications and statistics Mice were sacrificed 30 minutes later and muscles were For muscle fiber size quantification, muscle cross- snap-frozen in liquid nitrogen. Muscles were lysed as sections were stained either for laminin-γ1 or fluores- described below and proteins were separated on 8 to cence labeled wheat-germ agglutinin. Images were 16% SDS-PAGE (Bio-Rad Laboratories AG, Cressier, acquired using a Leica DM5000B fluorescence micro- Switzerland). After transfer to polyvinyl difluoride mem- scope with 10x objective, a digital camera (F-View; Soft branes and blocking of free binding sites with 5% milk Imaging System, Olympus Soft Imaging Solutions powder in Tris-buffered saline with 0.1% Tween 20 GmbH, Münster, Germany), and analySIS software (Soft (TBST), the mouse IgG2a monoclonal anti-puromycin Imaging System). Images of the entire soleus or tibialis antibody (clone 12D10; 1:5,000) was incubated for 1 hr anterior (TA) muscles were aligned with Adobe Pho- at room temperature. After incubation with the appropri- toShop (Adobe Systems Incorporated, San Jose, CA, ate HRP-coupled secondary antibody, blots were de- USA). The minimum distance of parallel tangents at op- veloped using enhanced chemiluminescence reagent. posing particle borders (minimal feret’s diameter) and Coomassie Blue staining was used to verify equal loading. cross-section area (CSA) were measured with analySIS software as described . Data are expressed as mean ± SEM. For statistical comparison of two conditions, the Tissue homogenization, SDS-PAGE and Western blot Student’s t- test was used. The level of significance is Muscles frozen in liquid nitrogen were powdered on dry indicated as follows: *** P <0.001, ** P <0.01, * P <0.05. ice and lysed in cold RIPA buffer supplemented with 1% Triton-X, 10% glycerol, protease inhibitor cocktail tab- Results lets (Roche Diagnostics AG, Rotkreuz, Switzerland), and Acute changes in mTORC1 activity affect muscle fiber size phosphatase inhibitor cocktail I and II (Sigma). Cell To evaluate the potential of mTORC1 in regulating lysates were incubated on ice for 2 hr, sonicated two muscle fiber size, we first tested the effect of mTORC1 times for 15 s and centrifuged at 13,600 g for 30 minutes inhibition or activation in normal weight-bearing mus- at 4°C. Cleared lysates were then used to determine total cles and in acute models of muscle hypertrophy and protein levels (BCA Protein Assay, Pierce, Rockford, IL, atrophy. To this end, we electroporated plasmids encod- USA). After dilution with sample buffer, equal protein ing an shRNA directed against rptor (to inactivate amounts were loaded onto SDS gels. mTORC1) or Tsc2 (to activate mTORC1) into muscle Bentzinger et al. Skeletal Muscle 2013, 3:6 Page 5 of 16 http://www.skeletalmusclejournal.com/content/3/1/6 fibers of mouse soleus muscle using the methods (Additional file 1: Figure S1A) or by infecting myoblasts described . As a negative control, shRNA constructs with adenovirus expressing the corresponding shRNA directed against Cd4 were used. To label targeted muscle construct (Additional file 1: Figure S1B). Four to six weeks fibers, a plasmid coding for nuclear-localized GFP (NLS- after electroporation, transfected muscle fibers were iden- GFP) was co-electroporated with all shRNA constructs. tified by their expression of NLS-GFP in myonuclei Before electroporation into muscle, each shRNA construct (Figure 1) and the size of GFP-positive fibers was com- was tested in tissue culture using either COS cells co- pared with that of neighboring, non-transfected fibers. transfected with the corresponding expression plasmid Knockdown of raptor resulted in a small but significant Figure 1 Acute perturbation of mTORC1 affects muscle fiber size. Soleus muscle was electroporated with plasmids encoding shRNA directed to transcripts encoding CD4 (Cd4), raptor (Rptor) or TSC2 (Tsc2). Plasmids encoding NLS-GFP were co-electroporated to label transfected fibers. After four to six weeks, muscle fiber size was determined by staining mid-belly cross-sections with Alexa-594-labeled wheat germ agglutinin (red). Transfected muscle fibers were identified by the expression of nuclear-localized GFP (green; white asterisks). The experimental paradigms used were innervated muscle (A, B), reinnervated muscle after nerve crush (C, D) and denervated muscle (E, F). Quantifications (B, D, F) of cross- sectional area (CSA) of muscle fibers in each paradigm are given relative to CSA of neighboring, GFP-negative, non-electroporated fibers. Electroporation of plasmids encoding shRNA to Cd4 served as control. Scale bars (A, C, E)=50 μm. Bars (B, D, F) represent mean ± SEM (N ≥3 mice and N ≥200 fibers were measured in each). In case of innervated muscles treated with shRNA to Tsc2 and with rapamycin (Tsc2 + Rapa) and denervated muscles electroporated with shRNA to Cd4, data represent mean ± SD (N = 2). P-values are ***P <0.001; **P <0.01; *P <0.05. Unless otherwise indicated, significance was determined compared to the control (shRNA to Cd4). Bentzinger et al. Skeletal Muscle 2013, 3:6 Page 6 of 16 http://www.skeletalmusclejournal.com/content/3/1/6 decrease in muscle fiber size, whereas knockdown of induced atrophy, followed by fiber re-innervation and TSC2 resulted in a significant increase (Figure 1A, B). re-growth to normal size [27,33,34]. Such “hypertrophy Consistent with the notion that TSC1/2 acts via mTORC1, on recovery” (HOR) was significantly less in muscle fi- rapamycin fully prevented the muscle hypertrophy ob- bers expressing shRNA to rptor and significantly higher served in TSC2 knockdown fibers (Figure 1A, B). As in fibers expressing shRNA to Tsc2 (Figure 1C, D). To expected, electroporation of shRNA constructs targeting test whether shRNA-targeting acted on the initial atro- Tsc1 resulted in a hypertrophy response very similar to the phy or on re-growth, we also examined electroporated Tsc2 knockdown (Additional file 1: Figure S1C, D). muscle fibers in a pure denervation-induced atrophy To test the role of mTORC1 in muscle plasticity, we paradigm. No difference between non-electroporated crushed the sciatic nerve unilaterally immediately after and electroporated fibers was detected in Cd4 controls electroporation, which causes a transient denervation- (Figure 1E, F). In contrast, muscle fibers expressing Figure 2 Conditional inactivation of TSC1 in skeletal muscle. (A) Western blot analysis of soleus muscle from 90-day-old control (ctrl) and TSCmKO mice using antibodies directed against the proteins indicated. α-actinin is used as a loading control. (B) Weight of soleus (Sol), gastrocnemius (GC), plantaris (PL), tibialis anterior (TA), extensor digitorum longus (EDL) and triceps (Tri) muscles of TSCmKO and littermate control (ctrl) mice. Weight is expressed as a percentage of the weight of the same muscle in control mice after normalization to the total body weight (N = 8 to 12 mice for each genotype). Data are mean ± SEM; ***P <0.001; **P <0.01; *P <0.05; Student’s t-test. (C) H&E staining of cross-sections from TA and soleus muscles of control and TSCmKO mice. Scale bar = 50 μm. (D, E)Fiber size distribution in soleus (D) and TA (E) muscles of 90-day-old TSCmKO and control mice (N = 4). More details of fiber size analysis are shown in Additional file 1: Figure S3 and in Additional file 1, Table S1. *P <0.05. Bentzinger et al. Skeletal Muscle 2013, 3:6 Page 7 of 16 http://www.skeletalmusclejournal.com/content/3/1/6 shRNA to Tsc2 were much bigger than non-electroporated phosphorylation of mTOR at the mTORC1-selective site fibers and, like in innervated muscle, the effect of TSC2 Serine 2448 and of the mTORC1 targets S6 and 4EBP knockdown was abrogated by rapamycin (Figure 1E, F). (Figure 2A; Table 1 for quantification). These data are Similar results were obtained by electroporating tibialis similar to those obtained in other tissues where Tsc1 or anterior (TA) muscle (Additional file 1, Figure S1E, F). Tsc2 were conditionally ablated [23,35-37]. These results thus show that acute alteration of mTORC1 Consistent with the activation of the mTORC1 targets activity affects the response of both, the slow oxidative and the role of mTORC1 in the control of protein trans- soleus and fast glycolytic TA muscles to growth-stimulating lation, protein synthesis in EDL muscle of TSCmKO was and atrophy-inducing conditions. increased (Additional file 1: Figure S2A). However, TSCmKO mice gained less weight than their control lit- Constitutive deletion of Tsc1 in skeletal muscle fibers termates. Starting from the age of five weeks, male affects muscles differentially TSCmKO mice were significantly lighter (Additional file 1: To examine whether sustained activation of mTORC1 Figure S2B), whereas the weight difference in females did would lead to the same effects observed in our electro- not reach significance (Additional file 1: Figure S2C). At poration paradigm, mice carrying floxed alleles for Tsc1 least part of this weight difference was due to alteration in  were crossed with mice that express Cre recombin- muscle mass as all but soleus muscles were significantly ase under the control of the muscle fiber-specific human lighter than in control mice (Figure 2B). Thus, despite skeletal actin (HSA) promoter . Mice lacking TSC1 increased protein synthesis, all but soleus muscles are in skeletal muscle (herein called TSCmKO, for TSC lighter in TSCmKO mice than in control mice. muscle knockout) were born at the expected Mendelian To investigate the reason for these muscle-specific dif- ratio and, at birth, could not be visually distinguished ferences in weight, we focused on soleus and TA muscles from their littermate controls. Muscle extracts from in three-month-old mice. Hematoxylin & eosin (H&E) TSCmKO mice were largely devoid of TSC1 (Figure 2A). staining did not reveal any major alterations in either of Moreover, they showed the expected increase in the muscles (Figure 2C). The difference in the muscle Table 1 Quantification of Western blot analysis TSCmKO Ctrl Ratio Number of replicates S2448 p-mTOR 19 ± 2*** 10 ± 1 1.9 4 mTOR 31 ± 6 28 ± 5 1.1 4 S65 P-4E-BP1 23 ± 5*** 12 ± 3 1.9 4 4E-BP1 34 ± 6 39 ± 6 0.9 4 S235/S236 P-S6 53 ± 1*** 13 ± 6 4 4 S6 48 ± 11 42 ± 17 1.1 4 IRS-1 6 ± 2*** 24 ± 1 0.3 4 S473 P-PKB/ Akt 4 ± 2*** 22 ± 3 0.2 4 PKB/ Akt 28 ± 7 22 ± 6 1.2 4 FoxO1 23 ± 5 18 ± 3 1.3 4 FoxO3a 16 ± 7 17 ± 5 0.9 4 T24 T23 P-FoxO1 /3a 2 ± 7**/ 8 ± 3* 11 ± 2/ 20 ± 5 0.2/ 0.4 4 P-PKB/ AktS (Den. TA) n.d. n.d. n.d. 3 S235/S236 P-S6 (Den. TA) 47 ± 12*** 5 ± 2 9.4 3 S473 P-PKB/ Akt (Den. Sol) n.d. n.d. n.d. 3 S235/S236 P-S6 (Den. Sol) 41 ± 10** 3 ± 5 13.7 3 RAmKO Ctrl ratio number of replicates S473 P-PKB/ Akt (Den. TA) 12 ± 2 2 ± 1 6 3 S235/S236 P-S6 (Den. TA) n.d. n.d. n.d. 3 S473 P-PKB/ Akt (Den. Sol) 19 ± 3 3 ± 2 6.3 3 S235/S236 P-S6 (Den. Sol) n.d. n.d. n.d. 3 Proteins were extracted from soleus (Sol) and tibialis anterior (TA) muscles of 90-day-old TSCmKO, RAmKO or control (Ctrl) littermates. “Den.” denotes: denervated. “n.d” denotes: not detectable. The amount of total protein loaded onto the SDS-PAGE was adjusted and Western blots were additionally normalized to α-actinin levels. Numbers given represent average gray values ± SEM after subtraction of the background. “Ratio” represents the average gray value obtained from a knockout animal divided by the gray values from the control littermates. “Number of replicates” represents the number of knockout animals analyzed. The number of Ctrl littermates was always the same or higher than the values given. P-values were determined by Student’s t-test; * P <0.05, ** P <0.01, ***P <0.001. Bentzinger et al. Skeletal Muscle 2013, 3:6 Page 8 of 16 http://www.skeletalmusclejournal.com/content/3/1/6 weight was matched by changes in the muscle fiber size in the phosphorylation of PKB/Akt . As shown in in soleus and TA muscle (Figure 2D, E). Detailed analysis Figure 3A, IRS1 levels were low in soleus muscle of of fiber types showed that both type I and type IIa fibers TSCmKO mice compared to control (Figure 3A; Table 1). were larger in soleus muscle (Additional file 1: Figure S3; In addition, phosphorylation of PKB/Akt and of FoxO1/3 Additional file 1: Table S1). In TA muscle, the glycolytic was substantially decreased in TSCmKO mice compared type IIb fibers were significantly smaller whereas the oxi- to controls (Figure 3A; Table 1). The same alterations in dative type IIa/x fibers were not affected (Additional file 1: expression levels and phosphorylation of the examined Figure S3, Additional file 1: Table S1). In summary, these proteins were detected in TA muscle of TSCmKO mice data show that the response to the activation of mTORC1 (data not shown). Consistent with the low phosphoryl- differs between muscles and fiber types. ation levels of FoxO1a and FoxO3a, transcript levels of We have previously shown that deletion of rptor not atrogin-1/MAFbx or MuRF-1 were much higher in TA only affects the immediate downstream targets of muscle of TSCmKO than in control mice (Figure 3B). Sur- mTORC1, S6K and 4EBP, but also causes a strong increase prisingly, in soleus muscle, transcript levels of atrogin-1/ Figure 3 mTORC1 activation affects the PKB/Akt and PGC1 pathways. (A) Western blot analysis of soleus muscles from 90-day-old control (ctrl) and TSCmKO mice using antibodies directed against the proteins indicated. α-actinin is used as loading control. (B, C) Relative mRNA expression of atrogin-1/MAFbx (Atr-1) and MuRF1 in TA and soleus muscles of TSCmKO and control mice. All values were normalized to the expression of β-actin and control muscles were set to 100% (TA: N ≥4 mice; Sol: N ≥5 mice). (D, E) Relative mRNA expression of Pgc1α and Pgc1β is shown in TA (D) and soleus (E) muscles of TSCmKO and control mice. All values are normalized to expression of β-actin. Relative expression in muscles from control littermates were set to 100%. TA: N ≥4; Sol: N ≥5. Note that levels of Pgc1β but not Pgc1α are up-regulated in TSCmKO mice. (F) Relative mRNA levels of Pgc1α in differentiated C2C12 cells that were infected with adenoviral vectors encoding GFP (ad-GFP), PGC1β (ad-PGC1β), shRNA to a scrambled sequence (ad-siScr) or shRNA to Pgc1β (ad-siPGC1β). Values are normalized to each control (ad-GFP and ad-siScr) and were set to 100% (N = 9). Note that expression of Pgc1α inversely correlates with PGC1β levels. Quantitative data (B-F) represent mean ± SEM. P-values are ***P <0.001; **P <0.01; *P <0.05; Student’s t-test. (G) NADH-TR staining of TA and soleus muscles of 90-day -old control and TSCmKO mice. Both muscles of TSCmKO are more oxidative. Scale bar = 50 μm. Bentzinger et al. Skeletal Muscle 2013, 3:6 Page 9 of 16 http://www.skeletalmusclejournal.com/content/3/1/6 MAFbx and MuRF1 did not differ from controls (Figure 3C) size in experimental paradigms of HOR and denervation- despite the low levels of phosphorylation of PKB/Akt. induced atrophy (Figure 1), we next tested muscle plasti- Thesedataargue that thedifferentialexpressionof the two city in RAmKO and TSCmKO mice. We first used the E3 ligases might be responsible for the selective hyper- synergist ablation/mechanical overload model, in which trophy in soleus muscle. gastrocnemius and soleus muscles including their tendons are surgically removed, a procedure that results in the Sustained activation of mTORC1 increases the oxidative functional overloading (FO) of the remaining plantaris capacity in all muscles muscle [44-46]. Seven or 28 days after surgery, mice were Additional factors that are regulated by mTORC1 euthanized and the plantaris muscle of the overloaded leg [16,22,38] and have been implicated in the control of was compared with plantaris from the contralateral, muscle size are the transcriptional coactivators PGC1α sham-operated leg. In control mice, FO increased muscle and PGC1β [39,40]. Moreover, PGC1α and PGC1β are weight after 7 days to 140% and to more than 200% after major regulators of mitochondrial biogenesis . To 28 days (Figure 4A). Muscle weight also increased in test whether deletion of Tsc1 would also affect the PGC1 RAmKO mice, although the increase was significantly re- pathway and the oxidative capacity of skeletal muscle, duced compared to control animals after 28 days of FO we next compared expression of Pgc1α and Pgc1β in TA (Figure 4A). However, and in contrast to control mice and soleus muscles of TSCmKO mice with littermate (Figures 4B and S5A), individual muscle fibers did not in- controls. Contrary to the expectation, transcript levels of crease in size in RAmKO mice after 7 days (Additional file Pgc1α were decreased in mutant muscles compared to 1: Figure S5B) or after 28 days of FO (Figure 4C). H&E controls (Figure 3D, E). The down-regulation of Pgc1α staining of the plantaris after 28 days of FO did not reveal was more pronounced in soleus muscle, which expresses differences between contralateral and overloaded RAmKO the highest level of PGC1α in wild-type mice . In muscles (Figure 4D). In contrast to RAmKO mice, contrast, mRNA levels of Pgc1β were increased to about TSCmKO muscle responded to FO like control muscle 150% in all examined muscles of TSCmKO mice (Figure 4E-G). (Figure 3D, E). In support of a direct regulation of Pgc1β There is evidence that FO also causes some damage and transcripts by mTORC1, Pgc1β expression was dimin- muscle regeneration and that satellite and other cells out- ished in RAmKO mice (soleus muscle in RAmKO mice: side the muscle’s basal lamina contribute to the weight 73 ± 4.6%; control mice: 100 ± 10.3%; mean ± SEM; increase [47,48]. As HSA-Cre is not expressed in non- N ≥5; P <0.05). Hence, unlike expression of the E3 ubi- muscle cells and satellite cells , we treated control quitin ligases atrogin-1/MAFbx and MuRF1, expression mice with rapamycin during FO to eliminate mTORC1 of Pgc1α and Pgc1β did not differ between TA and soleus function in all cells. This treatment abolished both the muscles in TSCmKO mice. Overexpression and knock- increase in weight and the shift in fiber size distribution down experiments of PGC1β in C2C12 myotubes indi- (Additional file 1: Figure S5C), suggesting that mTORC1 cate that expression of Pgc1α is tightly regulated by expressed in non-muscle cells or in satellite cells might PGC1β (Figure 3F). Such counter-regulation between contribute to the increased weight of plantaris muscles in PGC1α and PGC1β has also been reported in other tis- RAmKO mice after FO. sues . Thus, the increased levels of Pgc1β transcripts As FO induces a relative increase in the number of oxi- in the TSCmKO mice likely suppress expression of dative fibers , we also stained the overloaded plantaris Pgc1α. Interestingly, TSCmKO mice showed an increase from control and mutant mice by NADH-TR. As shown in their capacity for oxidative phosphorylation in TA in Figure 4H, plantaris muscles remained largely non- and soleus muscles as shown by stainings for NADH-TR oxidative in RAmKO mice, whereas in the overloaded (Figure 3G), succinate dehydrogenase (SDH; Additional plantaris of TSCmKO mice even the large myofibers file 1: Figure S4A, B) and cytochrome oxidase (COX; remained highly oxidative (Figure 4I). Additional file 1: Figure S4A, B). This increase was ac- companied by a slight, although not significant, increase Soleus and TA muscles of TSCmKO mice respond in the number of mitochondria as determined by qPCR differently to denervation-induced atrophy of mitochondrial DNA (Additional file 1: Figure S4C). To determine whether mTORC1 activation is sufficient to Taken together, these data suggest that PGC1β is re- prevent atrophy, we next submitted TSCmKO muscle to sponsible for the increased oxidative properties of skel- denervation by cutting the sciatic nerve unilaterally and etal muscle of TSCmKO mice. compared the muscles of the denervated and the contra- lateral (non-denervated) leg six days later. TA and soleus mTORC1 is required for muscle fiber hypertrophy muscles of control mice lost 7% and 14% of their weight, Because acute perturbation of mTORC1 function by respectively (Figure 5A). Importantly, the weight loss in knockdown experiments showed a strong effect on muscle both muscles was significantly higher in RAmKO mice Bentzinger et al. Skeletal Muscle 2013, 3:6 Page 10 of 16 http://www.skeletalmusclejournal.com/content/3/1/6 Figure 4 Growth of muscle upon functional overloading. (A) Plantaris muscles of control (ctrl) and RAmKO mice were functionally overloaded (FO) by ablation of the soleus and gastrocnemius muscles. Muscle weight of plantaris was measured after 7 or 28 days of FO and is expressed as the percentage of the weight of the contralateral, non-overloaded muscle (7 days FO: N ≥5 mice; 28 days FO: N ≥7 mice). (B, C) Fiber size distribution of the contralateral (dashed line) and FO (closed line) plantaris muscle of control (B) and RAmKO (C) mice after 28 days of FO (N = 7). (D) H&E staining of overloaded and contralateral plantaris muscles after FO for 28 days in control and RAmKO mice. (E) Muscle weight after 28 days of FO in control and TSCmKO mice (N = 5). (F) Fiber size distribution of non-overloaded, contralateral (dashed line) and over-loaded plantaris muscles (solid line) after 28 days of FO in TSCmKO mice (N = 5). (G) H&E staining of overloaded and contralateral plantaris muscles after 28 days of FO from TSCmKO mice. (H, I) NADH-TR staining of plantaris muscles after 28 days FO in mice with the indicated genotype. Scale bars (D, G, H, I)=50 μm. Individual data points and bars of quantitative data represent mean ± SEM. P-values are ***P <0.001; **P <0.01; *P <0.05; Student’s t-test. (Figure 5A). In TSCmKO mice, the response to dener- to innervated muscle from control mice. These results vation differed between TA and soleus muscles. Whereas suggest that TA and soleus muscles differ in the response loss of weight in the TA was the same in TSCmKO and to mTORC1 activation under atrophy conditions and they control mice, soleus muscles of TSCmKO mice were suggest that the atrophy observed in the TSCmKO mice largely spared (Figure 5A). H&E staining of the denervated requires adaptive, long-term processes that are not muscles and contralateral muscles did not reveal major induced by acute perturbation of mTORC1 signaling (see structural changes in mutant mice (Figure 5B, C). In soleus Figure 1). In both TSCmKO and control mice, the TA muscles, the substantial weight loss upon denervation in muscle showed a loss of oxidative capacity upon dener- control and RAmKO mice was mirrored by a shift in fiber vation (Figure 5G) whereas the soleus muscle of TSCmKO size distribution. The leftward shift was seen in control mice remained oxidative (Figure 5H). mice (Figure 5D) and was even more pronounced in RAmKO mice (Figure 5E). In TSCmKO mice, muscle Feedback control of PKB/Akt is active during muscle fiber size distribution also shifted slightly toward smaller atrophy size when compared to the hypertrophic, contralateral in- The difference in the atrophy response between TA and nervated soleus muscles (Figure 5F), but remained similar soleus muscles indicated that the underlying signaling Bentzinger et al. Skeletal Muscle 2013, 3:6 Page 11 of 16 http://www.skeletalmusclejournal.com/content/3/1/6 Figure 5 Muscle atrophy induced by denervation. (A) Loss (Δ) of muscle weight in tibialis anterior (TA) and soleus (Sol) muscles after six days of denervation using mice of the indicated genotype. Data are expressed as percentage of weight loss compared to the non-denervated contralateral muscle in thesamemouse.N ≥4 mice for RAmKO and control littermates (ctrl); N ≥5 mice for TSCmKO and control littermates. (B, C) H&E staining of soleus muscle after six days of denervation in mice of the indicated genotype. (D-F) Fiber size distribution in soleus muscle after six days of denervation (solid line) and in the contralateral, non-denervated muscle (dashed line) of mice with the indicated genotype. Note that the most frequent fiber sizein the denervated TSCmKO muscle is the same as that of innervated control muscle (blue arrowheads). N ≥4 for RAmKO and control littermates; N = 5 for TSCmKO and control littermates. (G, H) NADH-TR staining of TA and soleus muscles after six days of denervation in control and TSCmKO mice. Scale bars (B,C, G,H) = 50 μm. Quantification represent mean ± SEM. P-values are ***P <0.001; **P <0.01; *P <0.05using theStudent’s t-test. mechanisms might also differ in the two muscles. To The effect on the expression of the two E3 ubiquitin examine this, we analyzed the changes in expression of ligases was particularly striking in soleus muscles where the E3 ligases atrogin-1/MAFbx and MURF1, and the their expression did not differ from innervated control coactivators Pgc1α and Pgc1β in response to denervation. muscles (Figure 6C). In TSCmKO mice, phosphorylated Denervation has been reported to activate mTORC1, PKB/Akt was too low to be detected in denervated mus- most likely due to the increase in free amino acids . cles (Table 1) but phosphorylation of S6 remained high However, in RAmKO mice phosphorylation of S6K, S6 (Figure 6D). Although phosphorylation of PKB/Akt was and 4EBP remained low six days after denervation low in both TA and soleus muscles, transcript levels of (Table 1 and data not shown) whereas phosphorylation atrogin-1/MAFbx and MuRF-1 were increased in TA but at Serine 473 of PKB/Akt remained high in RAmKO were significantly lower in soleus compared to the dener- mice (Figure 6A). In parallel to the activation state of vated muscles from control mice (Figure 6E, F, Table 1). PKB/Akt, denervation increased transcript levels of The expression of the mTORC1 target PGC1α is also atrogin-1/MAFbx and MuRF-1 in TA and soleus muscles controlled by denervation . In innervated soleus of control mice but not of RAmKO mice (Figure 6B, C). muscle of RAmKO mice, Pgc1α mRNA levels are less Bentzinger et al. Skeletal Muscle 2013, 3:6 Page 12 of 16 http://www.skeletalmusclejournal.com/content/3/1/6 Figure 6 Changes in mTORC1-dependent signaling in denervated muscles. (A, D) Western blot analysis of tibialis anterior (TA) and soleus (Sol) muscles after six days of denervation using antibodies against PKB/Akt phosphorylated at Serine 473 (A) and S6 phosphorylated at Serines 235/236 (D). α- actinin was used as a loading control. (B, C, E, F) Relative mRNA levels of atrogin-1/MAFbx (Atr-1) and MuRF1 as determined by qPCR in TA and soleus (Sol) muscles after six days of denervation. Note that expression of both E3 ligases is blunted in RAmKO mice (B, C), while this response is exaggerated in TA (E) but not in soleus muscles (F)ofTSCmKO mice. (G – J) Relative mRNA levels of Pgc1α and Pgc1β in RAmKO (G, H)and TSCmKO (I, J) mice after six days of denervation. All values are normalized to the expression levels of the transcript measured in innervated muscle of control littermates (set to 100%).N ≥4 mice for TA and N ≥5mice for soleus of each genotype. Values represent mean ± SEM. P-values are ***P <0.001; **P <0.01; *P <0.05; Student’s t-test. than 40%  and Pgc1β mRNA levels are approxi- levels of the two E3 ubiquitin ligases in this particular mately 70% of control muscle. In denervated TA and so- muscle. In contrast, the relative levels of Pgc1α and leus muscles of control mice, expression of Pgc1α and Pgc1β did not differ between TA and soleus muscles Pgc1β was lower than in innervated muscle (Figure 6G, H). upon denervation and are thus unlikely contributors to Similarly, denervation lowered the levels of both transcrip- the differential response. tional co-activators in RAmKO mice although the signifi- cant difference to control mice was lost (Figure 6G, H). Genetic inactivation of mTORC1 reverses the phenotype In contrast, expression of Pgc1α and Pgc1β was very of TSCmKO mice different in TSCmKO mice. While Pgc1α mRNA levels While the inhibitory function of TSC1/2 onto mTORC1 were decreased upon denervation both in TA and soleus is well established, there is evidence that this protein muscles, Pgc1β was significantly increased in both complex can also regulate mTORC2 [50,51]. To test muscles (Figure 6I, J). Taken together, our results show whether any of the effects observed in TSCmKO mice that atrophy is accelerated in RAmKO mice despite low would be maintained in RAmKO mice, we generated levels of atrogin-1/MAFbx and MuRF1. Conversely, the double knockout mice (termed TSC-RAmKO). First, we sparing of soleus muscles from denervation-induced examined phosphorylation of known mTORC1 and atrophy in TSCmKO mice could be based on the low mTORC2 substrates. As shown in Figure 7A, the Bentzinger et al. Skeletal Muscle 2013, 3:6 Page 13 of 16 http://www.skeletalmusclejournal.com/content/3/1/6 Figure 7 TSC1-raptor double knockouts resemble RAmKO mice. (A) Western blot analysis of soleus muscles of TSCmKO, TSC-RAmKO and control (ctrl) mice using antibodies directed against the proteins indicated. An equal amount of protein was loaded in each lane. Loading control was α-actinin. (B) Muscle weight of the tibialis anterior (TA) and soleus (Sol) muscles of TSC-RAmKO and control mice. Muscle weight was first normalized to the body weight and is expressed as percentage of the weight of the same muscle from control mice (N = 3 mice for each genotype). (C)Relative mRNA expression of Pgc1α and Pgc1β in soleus muscle of TSC-RAmKO and ctrl mice. Values obtained in control mice were set to 100% (N = 3 mice). Bars in B and C represent mean ± SEM. P-values are ***P <0.001; **P <0.01; *P <0.05. (D) NADH-TR staining of soleus muscle from TSC-RAmKO and control mice. Scale bar = 100 μm. (E) Schematic drawing of the major signaling pathways regulated by mTORC1 and their influence on protein synthesis and degradation. mTORC1 substrate S6K and S6 were not phosphory- those observed in RAmKO mice and affected all examined lated in TSC-RAmKO mice and phosphorylation of muscles, the effect of mTORC1 activation on muscle size PKB/Akt at Serine 473 was increased compared to con- was unexpected as all muscles except soleus muscles were trol mice. In addition, similar to RAmKO mice, the PKB/ slightly but significantly smaller. Thus, our work highlights Akt target FoxO3a was hyperphosphorylated. The weight the existence of several feed-forward or auto-inhibitory of all muscles including TA and soleus was lower in TSC- loops that allow fine-tuning of the signaling networks in- RAmKO mice than in controls (Figure 7B). Moreover, volved in the control of muscle mass (Figure 7E). transcript levels of both Pgc1α and Pgc1β were lower in Based on the current concepts, mTORC1 activation soleus muscle (Figure 7C) and its oxidative capacity was should result in an increase in muscle mass and muscle decreased (Figure 7D). Finally, the TSC-RAmKO mice de- fiber size. This view is based on the findings that activa- veloped the same pathology as the RAmKO mice and they tion of the mTORC1 upstream components PKB/Akt or eventually died at the age of four to six months (data not IGF-1 receptor causes an increase in muscle mass shown). Thus, all the hallmarks of RAmKO mice are [5,10,11,52,53] and that this increase is rapamycin- present in the double mutants, indicating that TSC acts sensitive [11,53]. Moreover, overexpression of Rheb in mainly via mTORC1 in skeletal muscle. single muscle fibers by electroporation leads to hyper- trophy of the transfected fibers  and whole body Discussion knockout of the mTORC1 target S6K1 results in smaller Here we describe the phenotype of mice in which muscle fibers . Consistent with these experiments, mTORC1 is constitutively active in skeletal muscle acute knockdown of TSC1/2 by shRNA resulted in (TSCmKO) and compare it to mice with inactivated slightly bigger muscle fibers in soleus or TA muscles, mTORC1 signaling (RAmKO). While the oxidative confirming that transient activation of the mTORC1 changes in TSCmKO mice were largely the opposite of pathway is sufficient to induce muscle fiber growth. Bentzinger et al. Skeletal Muscle 2013, 3:6 Page 14 of 16 http://www.skeletalmusclejournal.com/content/3/1/6 However, under conditions of prolonged activation of enhance fiber hypertrophy upon re-innervation. Simi- mTORC1 in TSCmKO mice, all muscles examined, with larly, TSCmKO mice display atrophy resistance to de- the exception of soleus, were smaller than in control nervation in soleus muscle, which shows only moderate mice. As mTORC1 targets are activated and protein syn- expression of the E3 ubiquitin ligases MuRF1 and thesis in EDL muscle of TSCmKO mice is increased, the atrogin-1/MAFbx. By contrast, long-term activation of atrophy induced by chronic mTORC1 activation is likely mTORC1 did not protect TA muscle from atrophy and related to the feedback inhibition of activated S6K onto did not exacerbate the hypertrophy response to IRS1, which in turn, decreases activation of PKB/Akt. overloading of plantaris muscle. These results indicate This tight feedback control of S6K on IRS1-PKB/Akt that the increased protein synthesis by mTORC1 was also observed in mice deficient for raptor or mTOR hyperactivation is not sufficient to maintain muscle mass in some tissues including skeletal and heart muscle in cases where the FoxO-MuRF1-atrogin-1/MAFbx axis [16,54-56] but not in others . Similarly, deletion of is active due to the absence of PKB/Akt signaling. Im- TSC1 strongly decreases activation of PKB/Akt in cul- portantly, both transient and long-term inactivation of tured mouse embryonic fibroblasts , whereas it does mTORC1 increased denervation-induced atrophy and not at all affect PKB/Akt phosphorylation in some tis- prevented muscle growth associated with re-innervation sues [58,59]. These data indicate that the feedback con- or overloading, indicating that increased protein synthe- trol of S6K depends on the cellular context and our data sis is required even when the catabolic proteasomal ac- now show that this feedback is particularly strong in tivity is reduced. Thus, our results provide genetic skeletal muscle. evidence that muscle growth requires mTORC1. Consistent with decreased inhibition of FoxO tran- In our previous work, we demonstrated that raptor- scription factors by PKB/Akt, TA muscle from TSCmKO deficient skeletal muscles show a strongly decreased oxi- mice express high levels of MuRF1 and atrogin-1/ dative capacity due to changes in mitochondrial function MAFbx, involved in protein degradation through the . This loss of oxidative capacity correlated with a proteasome [7,8]. Hence, the atrophy observed in mus- substantial decrease in the transcript levels of Pgc1α, cles of the TSCmKO mice is likely caused by the preva- consistent with the direct regulation of Pgc1α expression lence of the FoxO pathway over mTORC1 activation. by mTOR , and could be restored by transgenic ex- This differs from the muscle hypertrophy observed using pression of PGC1α . Contrary to the expectations the transient, partial activation of mTORC1 with shRNA and the effect of mTORC1 activation in embryonic fi- electroporation. Thus, the atrophy response caused by broblasts , all examined muscles of TSCmKO mice the sustained, saturated mTORC1 activation by genetic showed a decreased expression of Pgc1α but increased Tsc1 deletion may unveil a long-term adaptation of the levels of Pgc1β. Thus, the increase in the oxidative cap- FoxO pathway. Consistently, transient overexpression of acity in TSCmKO mice may be mediated by PGC1β. In- Rheb does not seem to affect PKB/Akt phosphorylation deed, PGC1β has also been shown to be sufficient to , further supporting the idea that muscle atrophy in increase oxidative capacity in skeletal muscle despite the TSCmKO mice is related to the indirect PKB/Akt- concomitant reduction in PGC1α expression . More- dependent activation of FoxO pathways. over, depletion of both PGC1α and PGC1β results in Importantly, contrasting with the atrophic phenotype of much more severe loss of oxidative capacity than deple- most muscles, sustained activation of mTORC1 leads to tion of either protein alone . The reason for the unex- increased mass of soleus muscle in TSCmKO mice. Al- pected down-regulation of Pgc1α transcripts in TSCmKO though PKB/Akt was similarly inhibited in soleus and TA mice might be the counter-regulation of PGC1α and muscles, expression of MuRF1 and atrogin-1/MAFbx was PGC1β. We show here that overexpression of PGC1β in not increased in soleus muscle, indicating that an add- C2C12 myotubes results in a strong suppression of the en- itional regulatory mechanism suppresses their expression, dogenous Pgc1α expression and, conversely, Pgc1β knock- thereby overruling the regulation by PKB/Akt. This differ- down leads to increased expression of Pgc1α transcripts. ential regulation of MuRF1 and atrogin-1/MAFbx expres- These data indicate that the total amount of both PGC1 sion did not seem to be mediated by PGC1α, previously co-activators is tightly controlled in skeletal muscle. identified as a negative regulator of FoxO , because there was no significant difference in PGC1α/β expression Conclusions between soleus and TA muscles from TSCmKO mice. Our study provides new functional insights into the mo- With different atrophy and hypertrophy paradigms, we lecular mechanism of muscle atrophy and hypertrophy. also demonstrate that mTORC1 plays a critical and The data demonstrate that mTORC1 modulation down- complex role in muscle plasticity. Using shRNA electro- stream of PKB/Akt is subject to biological robustness. A poration, we show that transient activation of mTORC1 fine-tuned feedback loop controlled by the anabolic is sufficient to limit denervation-induced atrophy and to mTORC1 pathway mediates crosstalk to E3 ubiquitin Bentzinger et al. Skeletal Muscle 2013, 3:6 Page 15 of 16 http://www.skeletalmusclejournal.com/content/3/1/6 ligase system that increases protein degradation and thus 5. Bodine SC, Stitt TN, Gonzalez M, Kline WO, Stover GL, Bauerlein R, Zlotchenko E, Scrimgeour A, Lawrence JC, Glass DJ, Yancopoulos GD: Akt/ compensates for imbalance. However, this feedback sys- mTOR pathway is a crucial regulator of skeletal muscle hypertrophy and tem fails to fully re-establish muscle homeostasis, lead- can prevent muscle atrophy in vivo. Nat Cell Biol 2001, 3:1014–1019. ing to prevalence of either an anabolic or a catabolic net 6. Zhao J, Brault JJ, Schild A, Cao P, Sandri M, Schiaffino S, Lecker SH, Goldberg AL: FoxO3 coordinately activates protein degradation by the autophagic/ response. Our observations emphasize that muscle growth lysosomal and proteasomal pathways in atrophying muscle cells. requires both activated PKB/Akt and mTORC1 in parallel, Cell Metab 2007, 6:472–483. and they provide a new rationale for the development of 7. Stitt TN, Drujan D, Clarke BA, Panaro F, Timofeyva Y, Kline WO, Gonzalez M, Yancopoulos GD, Glass DJ: The IGF-1/PI3K/Akt pathway prevents pharmacologic agents that target this system. expression of muscle atrophy-induced ubiquitin ligases by inhibiting FOXO transcription factors. Mol Cell 2004, 14:395–403. Additional file 8. Sandri M, Sandri C, Gilbert A, Skurk C, Calabria E, Picard A, Walsh K, Schiaffino S, Lecker SH, Goldberg AL: Foxo transcription factors induce the atrophy-related ubiquitin ligase atrogin-1 and cause skeletal muscle Additional file 1: Contains supplemental figures S1 to S5 and atrophy. Cell 2004, 117:399–412. supplemental Table S1. See text and additional file 1 for more details. 9. Mammucari C, Milan G, Romanello V, Masiero E, Rudolf R, Del Piccolo P, Burden SJ, Di Lisi R, Sandri C, Zhao J, Goldberg AL, Schiaffino S, Sandri M: FoxO3 Abbreviations controls autophagy in skeletal muscle in vivo. Cell Metab 2007, 6:458–471. 4EBP: eIF-4E-binding protein; BSA: Bovine serum albumin; CSA: Cross-section 10. Lai KM, Gonzalez M, Poueymirou WT, Kline WO, Na E, Zlotchenko E, Stitt TN, area; FoxO: Forkhead box O; FO: Functional overloading; H&E: Hematoxylin & Economides AN, Yancopoulos GD, Glass DJ: Conditional activation of akt eosin; HOR: Hypertrophy on recovery; HSA: Human skeletal actin; in adult skeletal muscle induces rapid hypertrophy. Mol Cell Biol 2004, IGF1: Insulin-like growth factor-1; i.p: Intraperitoneal; mTORC1: Mammalian 24:9295–9304. target of rapamycin complex 1; NADH-TR: Nicotinamide adenine 11. Izumiya Y, Hopkins T, Morris C, Sato K, Zeng L, Viereck J, Hamilton JA, Ouchi dinucleotide hydrogen-tetrazolium reductase; NLS-GFP: Nuclear-localized N, LeBrasseur NK, Walsh K: Fast/Glycolytic muscle fiber growth reduces fat GFP; PBS: Phosphate-buffered saline; PFA: Paraformaldehyde; mass and improves metabolic parameters in obese mice. Cell Metab PI3K: Phosphatidylinositol 3-kinase; PKB: Protein kinase B (also called Akt); 2008, 7:159–172. PGC-1: Peroxisome proliferator-activated receptor gamma coactivator 1; 12. Jacinto E, Loewith R, Schmidt A, Lin S, Ruegg MA, Hall A, Hall MN: Raptor: Regulatory-associated protein of mTOR; RAmKO: Raptor muscle Mammalian TOR complex 2 controls the actin cytoskeleton and is knockout; Rheb: Ras homolog enriched in brain; Rictor: Rapamycin-insensitive rapamycin insensitive. Nat Cell Biol 2004, 6:1122–1128. companion of mTOR; S6K: p70/S6 kinase; SUnSET: Surface sensing of 13. Sarbassov DD, Ali SM, Kim DH, Guertin DA, Latek RR, Erdjument-Bromage H, translation; TA: Tibialis anterior; TBST: Tris-buffered saline with 0.1% Tween 20; Tempst P, Sabatini DM: Rictor, a novel binding partner of mTOR, defines a TSC: Tuberous sclerosis complex; TSCmKO: TSC muscle knockout. rapamycin-insensitive and raptor-independent pathway that regulates the cytoskeleton. Curr Biol 2004, 14:1296–1302. Competing interests 14. Lamming DW, Ye L, Katajisto P, Goncalves MD, Saitoh M, Stevens DM, Davis The authors declare they have no competing interests. JG, Salmon AB, Richardson A, Ahima RS, Guertin DA, Sabatini DM, Baur JA: Rapamycin-induced insulin resistance is mediated by mTORC2 loss and Authors’ contributions uncoupled from longevity. Science 2012, 335:1638–1643. CFB, SL and MAR conceived and designed the study. CFB and SL performed most of the experiments and analyzed the data. KR, PC, MG and SS 15. Kumar A, Harris TE, Keller SR, Choi KM, Magnuson MA, Lawrence JC Jr: conducted some of the experiments and CH, LAT and MNH provided Muscle-specific deletion of rictor impairs insulin-stimulated glucose scientific input. CFB, SL, PC and MAR wrote the manuscript. All authors read transport and enhances basal glycogen synthase activity. Mol Cell Biol and approved the final manuscript. 2008, 28:61–70. 16. Bentzinger CF, Romanino K, Cloëtta D, Lin S, Mascarenhas JB, Oliveri F, Xia J, Acknowledgements Casanova E, Costa CF, Brink M, Zorzato F, Hall MN, Rüegg MA: Skeletal We thank Drs. Xian Chu Kong and Céline Costa for their help with the muscle-specific ablation of raptor, but not of rictor, causes metabolic shRNA constructs. We thank Dr. Philippe Pierre (CIML Parc Scientifique de changes and results in muscle dystrophy. Cell Metab 2008, 8:411–424. Luminy, Marseille, France) for providing us with the anti-puromycin antibody. 17. Rommel C, Bodine SC, Clarke BA, Rossman R, Nunez L, Stitt TN, Yancopoulos GD, This work was supported by the Cantons of Basel-Stadt and Baselland, grants Glass DJ: Mediation of IGF-1-induced skeletal myotube hypertrophy by PI(3) from the Swiss National Science Foundation, the Swiss Foundation for K/Akt/mTOR and PI(3)K/Akt/GSK3 pathways. Nat Cell Biol 2001, 3:1009–1013. Research on Muscle Disease, Swiss Life and the Association Française contres 18. Kline WO, Panaro FJ, Yang H, Bodine SC: Rapamycin inhibits the growth and les Myopathies (AFM). muscle-sparing effects of clenbuterol. JAppl Physiol 2007, 102:740–747. 19. Wan M, Wu X, Guan KL, Han M, Zhuang Y, Xu T: Muscle atrophy in transgenic Author details mice expressing a human TSC1 transgene. FEBS Lett 2006, 580:5621–5627. 1 2 Biozentrum, University of Basel, Basel CH-4056, Switzerland. Neuromuscular 20. Goodman CA, Miu MH, Frey JW, Mabrey DM, Lincoln HC, Ge Y, Chen J, Research Center, Department of Biomedicine, University of Basel, Basel Hornberger TA: A phosphatidylinositol 3-kinase/protein kinase B-independent CH-4056, Switzerland. INRA, UMR866, Université Montpellier 1, Université activation of mammalian target of rapamycin signaling is sufficient to induce Montpellier 2, Montpellier, France. skeletal muscle hypertrophy. MolBiolCell 2010, 21:3258–3268. 21. Ohanna M, Sobering AK, Lapointe T, Lorenzo L, Praud C, Petroulakis E, Received: 3 October 2012 Accepted: 15 February 2013 Sonenberg N, Kelly PA, Sotiropoulos A, Pende M: Atrophy of S6K1(−/−) Published: 6 March 2013 skeletal muscle cells reveals distinct mTOR effectors for cell cycle and size control. Nat Cell Biol 2005, 7:286–294. References 22. Romanino K, Mazelin L, Albert V, Conjard-Duplany A, Lin S, Bentzinger CF, 1. Sandri M: Signaling in muscle atrophy and hypertrophy. Physiology Handschin C, Puigserver P, Zorzato F, Schaeffer L, Gangloff YG, Rüegg MA: (Bethesda) 2008, 23:160–170. Myopathy caused by mammalian target of rapamycin complex 1 2. Ruegg MA, Glass DJ: Molecular mechanisms and treatment options for (mTORC1) inactivation is not reversed by restoring mitochondrial muscle wasting diseases. Annu Rev Pharmacol Toxicol 2011, 51:373–395. function. Proc Natl Acad Sci U S A 2011, 108:20808–20813. 3. Wullschleger S, Loewith R, Hall MN: TOR signaling in growth and 23. Kwiatkowski DJ, Zhang H, Bandura JL, Heiberger KM, Glogauer M, metabolism. Cell 2006, 124:471–484. el-Hashemite N, Onda H: A mouse model of TSC1 reveals sex-dependent 4. Laplante M, Sabatini DM: mTOR signaling in growth control and disease. lethality from liver hemangiomas, and up-regulation of p70S6 kinase Cell 2012, 149:274–293. activity in Tsc1 null cells. Hum Mol Genet 2002, 11:525–534. Bentzinger et al. Skeletal Muscle 2013, 3:6 Page 16 of 16 http://www.skeletalmusclejournal.com/content/3/1/6 24. Schwander M, Leu M, Stumm M, Dorchies OM, Ruegg UT, Schittny J, Muller 46. Dunn SE, Michel RN: Coordinated expression of myosin heavy chain U: Beta1 integrins regulate myoblast fusion and sarcomere assembly. isoforms and metabolic enzymes within overloaded rat muscle fibers. Dev Cell 2003, 4:673–685. Am J Physiol 1997, 273:C371–C383. 25. Kong XC, Barzaghi P, Ruegg MA: Inhibition of synapse assembly in 47. Tamaki T, Uchiyama Y, Okada Y, Tono K, Nitta M, Hoshi A, Akatsuka A: mammalian muscle in vivo by RNA interference. EMBO Rep 2004, 5:183–188. Multiple stimulations for muscle-nerve-blood vessel unit in 26. Shavlakadze T, White JD, Davies M, Hoh JF, Grounds MD: Insulin-like compensatory hypertrophied skeletal muscle of rat surgical ablation growth factor I slows the rate of denervation induced skeletal muscle model. Histochem Cell Biol 2009, 132:59–70. atrophy. Neuromuscul Disord 2005, 15:139–146. 48. Tedesco FS, Dellavalle A, Diaz-Manera J, Messina G, Cossu G: Repairing 27. Stockholm D, Herasse M, Marchand S, Praud C, Roudaut C, Richard I, Sebille A, skeletal muscle: regenerative potential of skeletal muscle stem cells. Beckmann JS: Calpain 3 mRNA expression in mice after denervation and J Clin Invest 2010, 120:11–19. during muscle regeneration. Am J Physiol Cell Physiol 2001, 280:C1561–C1569. 49. Quy PN, Kuma A, Pierre P, Mizushima N: Proteasome-dependent activation of mammalian target of rapamycin complex 1 (mTORC1) is essential for 28. Dunn SE, Burns JL, Michel RN: Calcineurin is required for skeletal muscle autophagy suppression and muscle remodeling following denervation. hypertrophy. J Biol Chem 1999, 274:21908–21912. J Biol Chem 2013, 288:1125–1134. 29. van Slegtenhorst M, Nellist M, Nagelkerken B, Cheadle J, Snell R, van den 50. Yang Q, Inoki K, Kim E, Guan KL: TSC1/TSC2 and Rheb have different effects Ouweland A, Reuser A, Sampson J, Halley D, van der Sluijs P: Interaction on TORC1 and TORC2 activity. Proc Natl Acad Sci U S A 2006, 103:6811–6816. between hamartin and tuberin, the TSC1 and TSC2 gene products. 51. Huang J, Dibble CC, Matsuzaki M, Manning BD: The TSC1-TSC2 complex is Hum Mol Genet 1998, 7:1053–1057. required for proper activation of mTOR complex 2. Mol Cell Biol 2008, 30. Schmidt EK, Clavarino G, Ceppi M, Pierre P: SUnSET, a nonradioactive 28:4104–4115. method to monitor protein synthesis. Nat Methods 2009, 6:275–277. 52. Musaro A, McCullagh K, Paul A, Houghton L, Dobrowolny G, Molinaro M, 31. Dunant P, Larochelle N, Thirion C, Stucka R, Ursu D, Petrof BJ, Wolf E, Barton ER, Sweeney HL, Rosenthal N: Localized Igf-1 transgene expression Lochmuller H: Expression of dystrophin driven by the 1.35-kb MCK sustains hypertrophy and regeneration in senescent skeletal muscle. Nat promoter ameliorates muscular dystrophy in fast, but not in slow Genet 2001, 27:195–200. muscles of transgenic mdx mice. Mol Ther 2003, 8:80–89. 53. Pallafacchina G, Calabria E, Serrano AL, Kalhovde JM, Schiaffino S: A protein 32. Briguet A, Courdier-Fruh I, Foster M, Meier T, Magyar JP: Histological kinase B-dependent and rapamycin-sensitive pathway controls skeletal parameters for the quantitative assessment of muscular dystrophy in the muscle growth but not fiber type specification. Proc Natl Acad Sci U S A mdx-mouse. Neuromuscul Disord 2004, 14:675–682. 2002, 99:9213–9218. 33. Tonge DA: Physiological characteristics of re-innervation of skeletal 54. Polak P, Cybulski N, Feige JN, Auwerx J, Ruegg MA, Hall MN: Adipose- muscle in the mouse. J Physiol 1974, 241:141–153. specific knockout of raptor results in lean mice with enhanced 34. Liu M, Zhang D, Shao C, Liu J, Ding F, Gu X: Expression pattern of mitochondrial respiration. Cell Metab 2008, 8:399–410. myostatin in gastrocnemius muscle of rats after sciatic nerve crush 55. Shende P, Plaisance I, Morandi C, Pellieux C, Berthonneche C, Zorzato F, injury. Muscle Nerve 2007, 35:649–656. Krishnan J, Lerch R, Hall MN, Rüegg MA, Pedrazzini T, Brink M: Cardiac 35. Meikle L, Talos DM, Onda H, Pollizzi K, Rotenberg A, Sahin M, Jensen FE, raptor ablation impairs adaptive hypertrophy, alters metabolic gene Kwiatkowski DJ: A mouse model of tuberous sclerosis: neuronal loss of expression, and causes heart failure in mice. Circulation 2011, Tsc1 causes dysplastic and ectopic neurons, reduced myelination, 123:1073–1082. seizure activity, and limited survival. J Neurosci 2007, 27:5546–5558. 56. Risson V, Mazelin L, Roceri M, Sanchez H, Moncollin V, Corneloup C, Richard- 36. Mori H, Inoki K, Munzberg H, Opland D, Faouzi M, Villanueva EC, Ikenoue T, Bulteau H, Vignaud A, Baas D, Defour A, Freyssenet D, Tanti JF, Le-Marchand- Kwiatkowski D, MacDougald OA, Myers MG Jr, Guan KL: Critical role for Brustel Y, Ferrier B, Conjard-Duplany A, Romanino K, Bauché S, Hantaï D, hypothalamic mTOR activity in energy balance. Cell Metab 2009, 9:362–374. Mueller M, Kozma SC, Thomas G, Rüegg MA, Ferry A, Pende M, Bigard X, 37. Zeng LH, Rensing NR, Zhang B, Gutmann DH, Gambello MJ, Wong M: Tsc2 Koulmann N, Schaeffer L, Gangloff YG: Muscle inactivation of mTOR causes gene inactivation causes a more severe epilepsy phenotype than Tsc1 metabolic and dystrophin defects leading to severe myopathy. inactivation in a mouse model of tuberous sclerosis complex. Hum Mol JCell Biol 2009, 187:859–874. Genet 2011, 20:445–454. 57. Gödel M, Hartleben B, Herbach N, Liu S, Zschiedrich S, Lu S, Debreczeni-Mór 38. Cunningham JT, Rodgers JT, Arlow DH, Vazquez F, Mootha VK, Puigserver P: A, Lindenmeyer MT, Rastaldi MP, Hartleben G, Wiech T, Fornoni A, Nelson mTOR controls mitochondrial oxidative function through a YY1-PGC RG, Kretzler M, Wanke R, Pavenstädt H, Kerjaschki D, Cohen CD, Hall MN, -1alpha transcriptional complex. Nature 2007, 450:736–740. Rüegg MA, Inoki K, Walz G, Huber TB: Role of mTOR in podocyte function 39. Sandri M, Lin J, Handschin C, Yang W, Arany ZP, Lecker SH, Goldberg AL, and diabetic nephropathy in humans and mice. J Clin Invest 2011, Spiegelman BM: PGC-1alpha protects skeletal muscle from atrophy by 121:2197–2209. suppressing FoxO3 action and atrophy-specific gene transcription. Proc 58. Mori H, Inoki K, Opland D, Munzberg H, Villanueva EC, Faouzi M, Ikenoue T, Natl Acad Sci U S A 2006, 103:16260–16265. Kwiatkowski DJ, Macdougald OA, Myers MG Jr, Guan KL: Critical roles for 40. Brault JJ, Jespersen JG, Goldberg AL: Peroxisome proliferator-activated the TSC-mTOR pathway in beta-cell function. Am J Physiol Endocrinol receptor gamma coactivator 1alpha or 1beta overexpression inhibits Metab 2009, 297:E1013–E1022. muscle protein degradation, induction of ubiquitin ligases, and disuse 59. Inoki K, Mori H, Wang J, Suzuki T, Hong S, Yoshida S, Blattner SM, Ikenoue T, atrophy. J Biol Chem 2010, 285:19460–19471. Rüegg MA, Hall MN, Kwiatkowski DJ, Rastaldi MP, Huber TB, Kretzler M, 41. Puigserver P, Spiegelman BM: Peroxisome proliferator-activated receptor- Holzman LB, Wiggins RC, Guan KL: mTORC1 activation in podocytes is a gamma coactivator 1 alpha (PGC-1 alpha): transcriptional coactivator critical step in the development of diabetic nephropathy in mice. and metabolic regulator. Endocr Rev 2003, 24:78–90. J Clin Invest 2011, 121:2181–2196. 42. Lin J, Wu H, Tarr PT, Zhang CY, Wu Z, Boss O, Michael LF, Puigserver P, 60. Arany Z, Lebrasseur N, Morris C, Smith E, Yang W, Ma Y, Chin S, Spiegelman Isotani E, Olson EN, Lowell BB, Bassel-Duby R, Spiegelman BM: BM: The transcriptional coactivator PGC-1beta drives the formation of Transcriptional co-activator PGC-1 alpha drives the formation of slow- oxidative type IIX fibers in skeletal muscle. Cell Metab 2007, 5:35–46. twitch muscle fibres. Nature 2002, 418:797–801. 61. Zechner C, Lai L, Zechner JF, Geng T, Yan Z, Rumsey JW, Collia D, Chen Z, 43. Lelliott CJ, Medina-Gomez G, Petrovic N, Kis A, Feldmann HM, Bjursell M, Wozniak DF, Leone TC, Kelly DP: Total skeletal muscle PGC-1 deficiency Parker N, Curtis K, Campbell M, Hu P, Zhang D, Litwin SE, Zaha VG, Fountain uncouples mitochondrial derangements from fiber type determination KT, Boudina S, Jimenez-Linan M, Blount M, Lopez M, Meirhaeghe A, and insulin sensitivity. Cell Metab 2010, 12:633–642. Bohlooly-Y M, Storlien L, Strömstedt M, Snaith M, Oresic M, Abel ED, Cannon B, Vidal-Puig A: Ablation of PGC-1beta results in defective doi:10.1186/2044-5040-3-6 mitochondrial activity, thermogenesis, hepatic function, and cardiac Cite this article as: Bentzinger et al.: Differential response of skeletal performance. PLoS Biol 2006, 4:e369. muscles to mTORC1 signaling during atrophy and hypertrophy. Skeletal 44. Goldberg AL: Work-induced growth of skeletal muscle in normal and Muscle 2013 3:6. hypophysectomized rats. Am J Physiol 1967, 213:1193–1198. 45. Baldwin KM, Valdez V, Herrick RE, MacIntosh AM, Roy RR: Biochemical properties of overloaded fast-twitch skeletal muscle. JApplPhysiol 1982, 52:467–472.
Skeletal Muscle – Springer Journals
Published: Mar 6, 2013