Peroxisome Proliferator-activated Receptor β/δ Induces Myogenesis by Modulating Myostatin Activity*

Background: PPARβ/δ has been implicated in muscle regeneration; however the signaling mechanism(s) is unclear. Results: Activation of PPARβ/δ-promoted Gasp-1 expression blocked myostatin activity and enhanced myogenesis. Conclusion: Activation of PPARβ/δ led to inhibition of myostatin activity and thus increased myogenesis. Significance: PPARβ/δ agonists are novel myostatin antagonists that have potential benefits toward improving postnatal muscle growth and repair. Classically, peroxisome proliferator-activated receptor β/δ (PPARβ/δ) function was thought to be restricted to enhancing adipocyte differentiation and development of adipose-like cells from other lineages. However, recent studies have revealed a critical role for PPARβ/δ during skeletal muscle growth and regeneration. Although PPARβ/δ has been implicated in regulating myogenesis, little is presently known about the role and, for that matter, the mechanism(s) of action of PPARβ/δ in regulating postnatal myogenesis. Here we report for the first time, using a PPARβ/δ-specific ligand (L165041) and the PPARβ/δ-null mouse model, that PPARβ/δ enhances postnatal myogenesis through increasing both myoblast proliferation and differentiation. In addition, we have identified Gasp-1 (growth and differentiation factor-associated serum protein-1) as a novel downstream target of PPARβ/δ in skeletal muscle. In agreement, reduced Gasp-1 expression was detected in PPARβ/δ-null mice muscle tissue. We further report that a functional PPAR-responsive element within the 1.5-kb proximal Gasp-1 promoter region is critical for PPARβ/δ regulation of Gasp-1. Gasp-1 has been reported to bind to and inhibit the activity of myostatin; consistent with this, we found that enhanced secretion of Gasp-1, increased Gasp-1 myostatin interaction and significantly reduced myostatin activity upon L165041-mediated activation of PPARβ/δ. Moreover, we analyzed the ability of hGASP-1 to regulate myogenesis independently of PPARβ/δ activation. The results revealed that hGASP-1 protein treatment enhances myoblast proliferation and differentiation, whereas silencing of hGASP-1 results in defective myogenesis. Taken together these data revealed that PPARβ/δ is a positive regulator of skeletal muscle myogenesis, which functions through negatively modulating myostatin activity via a mechanism involving Gasp-1.

In the late 1960s, work performed by De Duve et al. (1) led to the identification of a series of compounds that promote peroxisome proliferation. These compounds were subsequently grouped into a family known as peroxisome proliferators. Peroxisome proliferators were shown to elicit biological function through binding to ligand-inducible nuclear hormone receptors, of which the first receptor, cloned from mouse liver, was termed peroxisome proliferator-activated receptor (PPAR) 2 (2). Upon ligand binding, PPARs become activated and bind to their target genes by forming heterodimeric complexes with retinoid-X receptors (RXR) (3,4). The activated PPAR-RXR complex then binds to consensus peroxisome proliferator-responsive elements (PPREs) (consisting of a direct repeat sequence, AGGTCA, separated by a single nucleotide) within target gene promoter regions to regulate gene expression (5). Three structurally identical nuclear hormone receptor isoforms (PPAR␣, PPAR␤/␦, and PPAR␥), which are encoded by separate genes (6), have been identified thus far. PPAR␣, expressed predominantly in the liver, heart, brown adipose tissue, kidney, and intestine, is primarily responsible for regulating body energy homeostasis (7). PPAR␥ is expressed in white and brown adipose tissue, intestinal epithelial cells, and immune cells and is essentially involved in adipocyte differentiation and lipid storage in white adipose tissue (8,9). In addition, PPAR␥ has also been shown to be involved in enhancing body insulin sensitivity (10 -12). The third isoform, PPAR␤/␦, is highly expressed in skin, skeletal muscle, adipose tissue, inflammatory cells, and cardiomyocytes. PPAR␤/␦ has been reported to be involved in energy homeostasis (13), lipid metabolism (14), and developmental regulation (15,16). Furthermore, in vitro and in vivo studies using PPAR␤/␦-specific ago-nists, tissue-specific PPAR␤/␦ knockdown, or PPAR␤/␦ overexpressing mouse models have confirmed an array of functions for PPAR␤/␦ in adipose tissue, skin, and muscle, as well as in response to cancer and inflammation. Using overexpressing transgenic mice and agonists, PPAR␤/␦ has been shown to influence skeletal muscle metabolism. Specifically, Luquet et al. (17) have demonstrated that muscle-specific overexpression of PPAR␤/␦ results in hyperplasia of muscle fibers with increased oxidative capability. Similarly, constitutive overexpression of VP6-PPAR␤/␦ in muscles results in a skeletal muscle fiber type switch from glycolytic to slow oxidative. As a result, increased fatty acid oxidation, reduced fat accumulation in adipose tissue, and a lean phenotype is reported in mice (18). In addition, pharmacological activation of PPAR␤/␦ by GW0742 increases angiogenesis, enhances oxidative myofiber number, and improves myonuclear accretion in vivo (19), all features observed in response to exercise (20 -22). Furthermore, pharmacological activation of PPAR␤/␦ may act as a potential therapeutic in preventing the dramatic muscle wasting observed during muscular dystrophy (23). Specifically, activation of PPAR␤/␦ results in increased utrophin A transcript levels in mdx mice (a model of Duchene muscular dystrophy), which is a protein that can functionally compensate for the loss of dystrophin in mdx mice and as such helps to maintain the sarcolemmal integrity of degenerating muscle fibers. A very recent study reports that postnatal activation of PPAR␤/␦ results in a similar effect on muscle metabolism to that observed following inhibition of myostatin (24), a TGF-␤ superfamily member and potent negative regulator of myogenesis (25,26). Postnatal activation of PPAR␤/␦ by GW501516 and neutralization of myostatin activity via PF-879 antibody in ob/ob mice results in reduced fat mass, improved glucose tolerance, and reduced muscle triglyceride and free fatty acid levels (24); this study clearly demonstrates that there is some degree of similarity between PPAR␤/␦ activation and myostatin inhibition, at least during postnatal growth. Given the benefits associated with PPAR␤/␦ activation and skeletal muscle growth, we attempted to delineate the mechanism(s) through which PPAR␤/␦ regulates muscle growth. We report here for the first time that activation of PPAR␤/␦, through the addition of L165041, enhances myogenesis in C2C12 myoblasts via an increase in both myoblast proliferation and differentiation. Consistent with this, loss of PPAR␤/␦ results in reduced proliferation of primary myoblasts and defective differentiation. Microarray analysis revealed the Gasp-1 (growth and differentiation factor-associated serum protein-1) gene as a potential target of PPAR␤/␦. Subsequent expression analysis confirmed up-regulation of Gasp-1 following activation of PPAR␤/␦ and also revealed enhanced association of Gasp-1 with myostatin in response to PPAR␤/␦ activation. Importantly, Gasp-1 has been shown previously to be a potent antagonist of myostatin (27,28), which is a well characterized potent negative regulator of myoblast proliferation and differentiation (29,30) as well as muscle stem cell (satellite cell) activation and self-renewal (31). Therefore, we propose that PPAR␤/␦ positively regulates myogenesis through a mechanism that results in Gasp-1-mediated inhibition of myostatin activity.

EXPERIMENTAL PROCEDURES
Animals-PPAR␤/␦-null mice (mixed genetic background of Sv129/C56BL/6) were kind gifts from Prof. Walter Wahli (University of Lausanne, Lausanne, Switzerland). PPAR␤/␦-null mice were maintained at 20°C with a 12-h light-dark cycle. mdx mice were obtained from the Animal Resources Centre, Canning Vale, Western Australia, Australia. wild type mice (C57BL/6) were purchased from the Center for Animal Resources, National University of Singapore (NUS-CARE), Singapore. All animal procedures were reviewed and approved by the Institute Animal Ethics Committee, Singapore.
Quantitative Real-time PCR (qPCR)-Total RNA from cells and tissue was isolated using TRIzol reagent (Invitrogen). Synthesis of cDNA, qPCR, and subsequent data analysis were performed as described previously (38). The gene-specific primers used in this manuscript are listed in supplemental Table S1.
Microarray Analysis-C2C12 myoblasts were cultured in differentiation medium for 72 h followed by a further 1, 2, 4, 6, 8, 12, or 24 h with or without 10 M PPAR␤/␦ agonist (Sigma-Aldrich). Total RNA was isolated using TRIzol reagent (Invitrogen). RNA was then column-purified using the RNeasy Midi Kit (Qiagen, Valencia, CA) following the manufacturer's guidelines, ethanol-precipitated overnight at Ϫ20°C, and resuspended in RNase-free water. RNA purity was assessed using the Agilent RNA 6000 Nano Kit and 2100 Bioanalyzer (Agilent, Santa Clara, CA). Microarray analysis was performed by Genomax Technologies, Singapore, as per their standardized techniques, using a one-color system and the Agilent SurePrint G3 mouse gene expression array (mouse 44,000 gene array). Following co-hybridization, spots were scanned numerous times and signal intensities were determined using the Agilent feature extraction software. GeneSpring GX 10 software (Silicon Genetics, Redwood City, CA) was then used to combine the data from multiple scans, normalization, and background correction. Differentially expressed genes were identified and considered significant with a -fold change threshold of 1.5 (p Յ 0.05 as determined by ANOVA) with the assistance of GeneSpring GX 10 software.
Gasp-1 Promoter Analysis and Cloning-The entire list of mouse known gene promoter sequences (from University of California, Santa Cruz) was extracted from the evolutionary conserved regions database. The available 1.5-kb proximal Gasp-1 promoter sequence was obtained and subjected to in silico analysis for the identification of conserved transcription factor binding sites using the rVista 2.0 online tool. PCR primers were designed with restriction enzymes sites compatible with both the pGEM-T Easy cloning vector (Promega, Madison, WI) and pGL3-basic luciferase vector (Promega). The proximal 1.5-kb Gasp-1 promoter region was amplified from genomic DNA isolated from wild type mice with the following PCR primers: forward, 5Ј-GCT AGC TGC CGT CTG CAG TGG-3Ј; and reverse, 5Ј-AAG CTT CCG ACT TTA GGC TGT AC-3Ј. A 1-kb truncated promoter fragment was also amplified using the following primer pair: forward, 5Ј-GCT AGC TTC  CAG GGA CAG AA-3Ј; and reverse, 5Ј-AAG CTT CCG ACT  TTA GGC TGT AC-3Ј. Positive sequence-verified clones were selected and subcloned into the pGL3-basic luciferase vector system for gene reporter studies. In addition, the QuikChange site-directed mutagenesis kit (Stratagene, La Jolla, CA) was used in the construction of a mutated Gasp-1 promoter reporter construct, where the DR-1 site was mutated from AGGCCTTTAACCC to TCCCCTTTAACCC. Mutations of the DR-1 site at Ϫ480 to Ϫ482 were introduced using the following oligonucleotides: forward, 5Ј-CTC CAG Gtc cCC TTT AAC CCC TTC CA G-3Ј; and reverse, 5Ј-CTG GAA GGG GTT AAA GGg gaC CTG GAG-3Ј. The mutated Gasp-1 promoter reporter construct was further verified by sequencing to ensure that the mutation was present prior to experimentation.
Transient Transfection and Luciferase Assay-Human myoblasts (36C15Q) were plated at a density of 10,000 cells/cm 2 in 24-well plates. Following an overnight attachment period, human myoblasts were transfected with the 1.5-kb Gasp-1 promoter-luciferase construct (pGL3-Gasp-1), the 1-kb Gasp-1 promoter-luciferase deletion construct (pGL3-Gasp-1 del), or the mutated Gasp-1 promoter reporter construct (mut-pGL3-Gasp-1) together with the control Renilla luciferase vector pRL-CMV and empty vector control (pGL3-basic) using Lipofectamine 2000 (LF2000, Invitrogen) as per the manufacturer's guidelines. Cells were then incubated with the transfection mix in proliferation medium (DMEM, 10% FBS, and 1% penicillin/ streptomycin) at 37°C, 5% CO 2 overnight, after which the medium was replaced with fresh proliferation medium containing either DMSO (control) or 10 M L165041, 30 M GW1929, or 10 M Wy14643 for a further 24 h. Luciferase assays were performed using the Dual-Luciferase assay system as per the manufacturer's protocol (Promega). Relative luciferase activity in each of the extracted protein samples was measured in triplicate using the Fluoroskan Ascent microplate fluorometer and luminometer (catalog No. 5210460, Thermo Fisher Scientific).
Chromatin Immunoprecipitation (ChIP) Assay-C2C12 myoblasts were transfected with pGL3-Gasp-1 promoter and incubated for 48 h. Following incubation, the myoblasts were treated without (DMSO) or with 10 M L165041, 30 M GW1929, or 10 M Wy14643 for a further 24 h. Following agonist treatment, myoblasts were washed twice with PBS and fixed in PBS containing 1% formaldehyde for 10 min at room temperature. The formaldehyde fixation was stopped by adding glycine (0.125 M final concentration), after which the cells were centrifuged at 2000 rpm for 5 min, washed once with ice-cold PBS, and resuspended in lysis buffer (5 mM PIPES, pH 8.0, 85 mM KCl, 0.5% Nonidet P-40, and Complete protease inhibitor mixture). To isolate crude nuclear extracts, lysates were then centrifuged at 2000 rpm for 5 min, washed once with ice-cold PBS, resuspended in high salt lysis buffer (1ϫ PBS, 1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% SDS, and Complete protease inhibitor mixture), and sonicated. After sonication, the extracts were centrifuged at 10,000 rpm for 15 min at 4°C. The protein concentration of the lysates was determined by Bradford assay (Bio-Rad). For immunoprecipitation, 500 g of protein lysates was used. Nuclear extracts were precleared initially through incubation with 50 l of protein A-agarose (Invitrogen) for 30 min at 4°C. The nuclear extracts were then centrifuged at 12,000 rpm for 5 min at 4°C and incubated overnight with 5 g of anti-PPAR␤/␦, anti-PPAR␣, anti-PPAR␥, or anti-IgG antibody at 4°C. Following overnight incubation, 50 l of protein A-agarose was added for 2 h at 4°C followed by centrifugation to pellet the immunoprecipitated complexes. Pellets were washed twice with 1 ml of high salt lysis buffer followed by four washes with wash buffer (100 mM Tris, pH 8.0, 500 mM LiCl, 1% Nonidet P-40, and 1% deoxycholate). Pellets were then resuspended in 400 l of elution buffer (1% SDS and 0.1 M NaHCO 3 ) and incubated for 2 h at 67°C with occasional mixing to reverse formaldehyde cross-linking. Beads were subsequently removed by centrifugation at 12,000 rpm for 10 s, and the supernatant was further incubated at 67°C overnight. Samples were then centrifuged for 3 min at 10,000 rpm, and phenol/ chloroform/isoamyl alcohol (25:24:1) was added to the supernatants, after which the samples were vortexed and centrifuged for 3 min at 14,000 rpm with the aqueous phase collected. DNA was subsequently purified and concentrated using the QIAquick PCR purification kit (Qiagen). The following sets of primers were used for PCR: Gasp Antibody Neutralization-CM from myoblasts exposed to either 10 M L165041 or DMSO was collected as described above. Human myoblasts were plated at a density of 1000 cells/ well (96-well) in proliferation medium and allowed to attach for 18 h, after which the proliferation medium was replaced with CM with or without L165041 and DMSO supplemented with 1 g/ml mouse monoclonal anti-hGASP-1 antibody or an equal volume of PBS. Cells were then fixed at regular 24-h intervals with 10% formaldehyde, 0.9% NaCl fixative prior to assessment of proliferation as described previously (29,30).
Lentiviral Production and Infection-The pCMV-dR8.2 dvpr, hGASP-1 shRNA, or empty pLKO.1 and pCMV-VSVG vectors were transfected into 293T cells using the calcium phosphate precipitation technique (Invitrogen). Briefly, 1 million cells/ml were seeded in 6-well plates and, after an overnight attachment period, were transfected with 5 g of the plasmids in a 2:2:1 ratio (pCMV-dR8.2 dvpr:pCMV-VSVG:hGASP-1 shRNA/pLKO.1). After 16 h of transfection, the medium was replaced with fresh proliferation medium, and the cells were incubated for a further 60 h. After 60 h of incubation, the supernatant was collected as a source of viral particles. The viral particles were then tested for infection efficiency by adding 10 -100 l of virus together with 8 g/ml hexadimethrine bromide (Sigma-Aldrich) to human myoblast cultures. After 8 h of infection, the medium was replaced and the human myoblasts were allowed to differentiate for a further 72 h, after which total RNA and total protein was harvested for analysis of hGASP-1 expression levels.
Smad3 Reporter Assay-The activity of Smad3 was assessed using a Smad3 reporter assay. Briefly, C2C12 myoblasts were transfected with a Smad binding element reporter construct (SBE-4x-Luc (Addgene plasmid 16495), which contains four repetitive Smad3 binding elements linked to the luciferase reporter gene. C2C12 myoblasts were also transfected with the Renilla luciferase vector (pRL-CMV) as an internal control. Cells were transfected via electroporation at 110 volts and 500 ohms resistance using the Gene Pulser MXcell elctroporation system C165-2670 (Bio-Rad), after which the cells were replated and grown for a further 24 h. SBE-4x-Luc-transfected cells were then plated at a density of 7500 cells/well prior to the addition of CM collected from L165041 or control (DMSO)treated cells. SBE-4x-Luc reporter-transfected cells were treated with 2 g/ml sActRIIB and dialysis buffer as control; CM-derived from PPAR␤/␦-null and wild type primary cultures was used as the source to study increased myostatin activity. To assess the effect of rhGASP-1 protein on SBE-4x-Luc reporter activity, SBE-4x-Luc-expressing cells were treated with increasing concentrations (0.5, 1, and 2 g/ml) of rhGASP-1 or BSA for 24 h. Cells were also treated for 24 h with either CM isolated from control Chinese hamster ovary (CHO) cells or myostatin (Mstn) protein CM (1:2 or 1:4 dilution) collected from CHO cells that overexpress and secrete Mstn into the medium (41), to act as a positive control. Lastly, SBE-4x-Luc was also co-transfected with either control shRNA or hGASP-1 shRNA into human myoblasts to assess the effect of GASP-1 knockdown on SBE-4x-Luc reporter activity. Cells were treated with a 1:2 dilution of the Mstn CM for 24 h to act as a positive control. Following 24 h with the respective treatments, all cells were lysed and subjected to luciferase assay using the GlowMax luminometer (Promega) as per the manufacturer's protocol.
Statistics-The data from myoblast proliferation analysis are presented here as means Ϯ S.E. of eight replicates, and an average was taken from three independent experiments. Total myotubes were counted in 12 random images/coverslip, and the mean myotube number Ϯ S.E. from three coverslips/treatment was calculated from three individual experiments. The mononucleated and multinucleated nuclei number was calculated in 20 random images/coverslip, and the mean percentage fusion index Ϯ S.E. from three coverslips/treatment was calculated. Individual myotube area was assessed for all myotubes present in 12 random images taken from three coverslips/treatment. All variations were compared using one-way ANOVA, and values of p Յ 0.05 were deemed significant.

Activation of PPAR␤/␦ via L165041 Agonist Treatment
Enhances Myogenesis-Treatment of C2C12 myoblasts or murine primary myoblasts with L165041 (10 M), a subtypeselective, high affinity ligand for PPAR␤/␦, resulted in a significant increase in myoblast numbers when compared with control-treated cells (DMSO) ( Fig. 1A and supplemental Fig. S1B). The L165041-mediated increase in C2C12 myoblast proliferation was observed as early as 12 h after the addition of L165041 and was maintained up to 96 h (Fig. 1A). Treatment of C2C12 or murine primary myoblasts with L165041 during differentiation also resulted in an observable increase in myotube formation ( Fig. 1B and supplemental Fig. S1C), with an ϳ55 and 52% increase in the myotube number detected at 48 and 72 h, respectively, following the addition of L165041, as compared with control-treated cells (Fig. 1C). Although we observed an increased myotube number, we found no appreciable change in either the myotube fusion index or the myotube area between cells treated with L165041 and control-treated cells (Fig. 1, D and E). However, we did observe an overall increase in the percentage of myotubes, with the average myotube area at 10,000 -250,000 m 2 during differentiation (Fig. 1E). Next we analyzed the expression of critical myogenic regulatory factors involved in the normal progression of myogenic differentiation. Subse- quent qPCR and Western blot analysis of differentiating C2C12 cells revealed increased mRNA expression (Fig. 1, F and G) and protein levels (supplemental Fig. S1A) of both MyoD and myogenin in L165041-treated cells. Furthermore, MyHC protein levels were elevated significantly in differentiating C2C12 myoblasts at 48 and 72 h of differentiation following treatment with L165041 (supplemental Fig. S1A). These data are consistent with the enhanced differentiation and increased myotube number observed following treatment with L165041.

Primary Myoblasts Derived from PPAR␤/␦-null Mice Have Reduced Proliferation and Defective Myogenic Differentiation-
Because L165041-mediated activation of PPAR␤/␦ resulted in enhanced proliferation and differentiation of C2C12 and pri-mary myoblast cultures in vitro, we next studied the myogenic potential of primary myoblast cultures derived from PPAR␤/␦null mice. Consistent with the results obtained following treatment with L165041, the absence of PPAR␤/␦ resulted in reduced myoblast proliferation ( Fig. 2A) as well as reduced myogenic differentiation (Fig. 2, B and C). Specifically, loss of PPAR␤/␦ resulted in the formation of fewer myotubes, with a visible reduction in myotube size and branching, when compared with wild type controls (Fig. 2, B and C). Subsequent quantification revealed a decreased myotube area, with a ϳ55% decrease in the number of large myotubes (4,000 -22,000 m 2 ) in cultures derived from PPAR␤/␦-null mice when compared with wild type controls (Fig. 2D). Furthermore, we found a reduced myotube fusion index during early differentiation (48 h) in the primary cultures derived from PPAR␤/␦-null mice when compared with wild type controls (Fig. 2E), suggesting that the reduced myotube number observed in the absence of PPAR␤/␦ may result from impaired myoblast fusion during early myogenic differentiation.
Identification of Novel Downstream Targets of PPAR␤/␦ in Skeletal Muscle-Next we sought to determine the molecular mechanism(s) through which L165041-mediated activation of PPAR␤/␦ enhances skeletal muscle myogenesis. To this end we performed microarray analysis on RNA collected from L165041-treated and control-treated (DMSO) C2C12 myotubes across a differentiation time course. The results of the microarray are summarized in a heat map (Fig. 3A, left panel); genes that were significantly (p Ͻ 0.05) up-regulated (Table 1) or down-regulated ( Table 2) by more than 1.5-fold were selected. Importantly, genes that had been identified previously as targets of PPAR␤/␦ in muscle, such as Abca1, Abcg1, Angplt4, Adfp, Pdk4, Ucp3, Cpt1b, and Ppargc1a (42)(43)(44)(45), were similarly up-regulated following the addition of L165041 (Table  1). From the list of significantly up-regulated genes, we selected 24 genes (supplemental Table S1) to validate using qPCR, the results of which are summarized in a heat map (Fig. 3A, right panel). Microarray analysis and subsequent confirmation through qPCR revealed the Gasp-1 gene as a novel PPAR␤/␦ target in muscle (Fig. 3A, right panel). Gasp-1 is a secreted protein, which has been shown to interact directly with both the mature and Latency Associated Peptide forms of myostatin resulting in inhibition of myostatin signaling (27,28). Interestingly, loss of myostatin function, much like what we observed following L165041-mediated activation of PPAR␤/␦, results in enhanced myoblast proliferation and differentiation (25,29,30,46) therefore, we propose that PPAR␤/␦ activation enhances myogenesis through a mechanism that involves Gasp-1-mediated inhibition of myostatin activity.
PPAR␤/␦ Regulates Gasp-1 Expression-To further confirm PPAR␤/␦ regulation of Gasp-1 expression, we treated C2C12 myoblasts and differentiating myotubes with L165041 and monitored Gasp-1 expression. Subsequent qPCR results revealed a significant increase in Gasp-1 expression in C2C12 myoblasts following 8-, 12-, and 24-h treatment with L165041 (Fig. 3B). Similarly, an 8-and ϳ12-fold induction of Gasp-1 expression was observed following L165041 treatment at 72 and 96 h of differentiation, respectively (Fig. 3C). Elevated Gasp-1 protein levels were detected at all differentiating time points (24 -96 h) following treatment with L165041 (Fig. 3E). As Gasp-1 is a secreted protein, we next addressed whether PPAR␤/␦ activation increases Gasp-1 protein secretion. Human myoblasts were treated with L165041 for a period of 24 h, after which CM was collected and subjected to Western blot analysis. Consistent with increased Gasp-1 expression, we found that L165041-mediated PPAR␤/␦ activation resulted in enhanced hGASP-1 protein secretion in vitro (Fig. 3D). It is noteworthy to mention that although L165041 treatment resulted in an increase in Gasp-1 expression, the addition of either PPAR␥ agonist GW1929 or PPAR␣ agonist Wy14643 failed to significantly alter Gasp-1 expression in both myoblasts and myotubes when compared with control-treated cultures (supplemental Fig. S2, A and C), suggesting that PPAR␤/␦, but not PPAR␥ or PPAR␣, induces Gasp-1 mRNA expression in skeletal muscle. However, in the same samples we do see a significant increase in the expression of the PPAR␥ target gene adiponectin and the PPAR␣ target gene FABP3 (supplemental Fig. S2, B and D), suggesting that PPAR␥ and PPAR␣ are activated in our system in response to treatment with GW1929 and Wy14643, respectively. Further evidence for PPAR␤/␦ regulation of Gasp-1 is observed in PPAR␤/␦-null mice. Significantly reduced Gasp-1 expression was detected in skeletal muscle tissues isolated from PPAR␤/␦-null mice (Fig. 4A), and moreover, significantly reduced Gasp-1 expression was observed in both slow twitch muscles (soleus) and fast twitch muscles (extensor digitorum longus (EDL)) (Fig. 4A). Furthermore, a significant reduction in Gasp-1 expression was also observed in differentiating primary myoblast cultures derived from PPAR␤/␦-null mice (Fig. 4B). Microarray and subsequent expression analysis collectively confirm that PPAR␤/␦ positively regulates Gasp-1 gene expression; as such, we suggest that Gasp-1 represents a novel muscle-specific downstream target of PPAR␤/␦. Previously published work has revealed that PPAR␤/␦ expression is greater in atrophying muscle tissue isolated from mdx mice, a mouse model of Duchenne muscular dystrophy (23). Based on this observation, we next wanted to ascertain whether increased endogenous PPAR␤/␦ expression, as seen in mdx mice, would induce Gasp-1 expression in vivo. In agreement with the L165041 agonist studies described herein above, we observed an increase in both PPAR␤/␦ and Gasp-1 expression in EDL muscle isolated from mdx mice (Fig. 4, C and D).
A Consensus PPAR Binding Motif (DR-1) Mediates Up-regulation of the Gasp-1 Promoter in Response to PPAR␤/␦ Agonist Treatment-To further investigate the mechanism of Gasp-1 transactivation by PPAR␤/␦, we performed in silico analysis of the 1.5-kb upstream sequence of the Gasp-1 gene promoter. Subsequent sequence analysis identified a putative PPRE, specifically a DR-1 motif (Fig. 4E) within the proximal 1.5-kb region of the Gasp-1 promoter, which has high sequence homology between mouse and human (Fig. 4F). Importantly, the DR-1 sequence we indentified in the Gasp-1 promoter is consistent with a consensus DR-1 sequence that has been predicted previously to be specific for PPAR␤/␦ (47). In studying the role of the DR-1 motif in PPAR␤/␦-mediated activation of Gasp-1, C2C12 myoblasts were transfected with either a prox- imal 1.5-kb Gasp-1 promoter (pGL3-Gasp-1), a truncated 1-kb Gasp-1 promoter (lacking the DR-1 motif) reporter construct (pGL3-Gasp-1 del), or a mutant Gasp-1 promoter reporter construct where the DR-1 site was mutated (mut-pGL3-Gasp-1) and subjected to treatment with L165041. Treatment with L165041 resulted in a ϳ8.5-fold increase in luciferase activity in cells transfected with the proximal 1.5-kb Gasp-1 promoter construct when compared with untreated controls (Fig. 4, G and H); however, no significant increase in luciferase activity was observed in cells transfected with either the truncated or mutated Gasp-1 promoter constructs following L165041 treatment (Fig. 4, G and H, respectively). In addition to analyzing PPAR␤/␦ activation of Gasp-1, we further assessed whether PPAR␥ or PPAR␣ plays a role in regulating Gasp-1 promoter activity. The addition of either the PPAR␥ agonist GW1929 or the PPAR␣ agonist Wy14643 did not significantly increase luciferase activity of the proximal 1.5-kb Gasp-1 promoter construct when compared with untreated controls (Fig. 4, G and  H). Moreover, treatment with the GW1929 or Wy14643 agonist did not significantly alter luciferase activity in cells transfected with the truncated or mutated Gasp-1 promoter constructs (Fig. 4, G and H, respectively). Therefore, these data further confirm that PPAR␤/␦, but not PPAR␥ and PPAR␣, regulates Gasp-1 expression and that the identified DR-1 site in the Gasp-1 promoter is critical in PPAR␤/␦-mediated activation of Gasp-1 expression. To confirm that PPAR␤/␦ binds to the DR-1 site in the Gasp-1 promoter, we performed ChIP analysis. C2C12 myoblasts were transfected with the pGL3-Gasp-1 promoter reporter construct and treated with the PPAR␤/␦ (L165041), PPAR␥ (GW1929), or PPAR␣ (Wy14643) agonist. After treatment cells were collected and subjected to chromatin immunoprecipitation, after which DNA was isolated and puri-

List of genes down-regulated in the microarray data with L165041 agonist treatment when compared with control-treated cells at all time points
Gene accession numbers, symbols, full gene names, and -fold change are given.  fied. As seen in Fig. 4I, we observed binding of PPAR␤/␦ to the DR-1 site specific to the Gasp-1 promoter, which was further enhanced upon treatment with L165041 (Fig. 4I). Importantly, no binding of PPAR␤/␦ to the control ␤-actin promoter was observed (Fig. 4I). In contrast to the above results, we observed no interaction between PPAR␥ or PPAR␣ and the DR-1 site found in the Gasp-1 promoter (supplemental Fig. S2, F and G, respectively). Similar to PPAR␤/␦, no binding of PPAR␥ or PPAR␣ to the control ␤-actin promoter was observed (supplemental Fig. S2, F and G, respectively). These data further confirm that PPAR␤/␦, but not PPAR␥ or PPAR␣, specifically binds to the DR-1 site located in the Gasp-1 promoter region.

Activation of PPAR␤/␦ Enhances Myogenesis through Modulating Myostatin
Activity-Myostatin is a secreted growth factor that acts as potent negative regulator of skeletal muscle growth through targeting and inhibiting both myoblast proliferation and differentiation (29,30). As mentioned earlier, Gasp-1 is a secreted protein that has been demonstrated previously to bind and inhibit myostatin activity (27,28). Therefore, if PPAR␤/␦-mediated induction of Gasp-1 is associated with inactivation of myostatin, we reasoned that treatment with L165041 would increase the levels of secreted Gasp-1, inhibit myostatin activity, and increase myoblast proliferation. In agreement, we found elevated levels of Gasp-1 in CM isolated from L165041-treated cells (Fig. 5A) as well as enhanced interaction between Gasp-1 and myostatin following treatment with L165041, as measured through co-immunoprecipitation analysis (Fig. 5A). Importantly, treatment of myoblasts with CM obtained from L165041-treated cells also resulted in a significant increase in myoblast proliferation when compared with control-treated cells (Fig. 5B). Furthermore, antibody-mediated blockade of hGASP-1 through the addition of a specific anti-hGASP-1 antibody prevented the increased proliferation observed following the addition of L165041 and, in fact, resulted in reduced myoblast proliferation in both agonisttreated and control-treated cells (Fig. 5C). As described above, the absence of PPAR␤/␦ resulted in reduced expression of Gasp-1 in both skeletal muscle tissues and primary myoblast cultures (Fig. 4, A and B). Therefore, we hypothesized that loss of PPAR␤/␦ would lead to reduced levels of secreted Gasp-1, resulting in higher levels of active myostatin. In support of this, we found a significant reduction in the proliferation rate of myoblasts treated with CM obtained from PPAR␤/␦-null primary myoblast cultures (Fig. 5D). Moreover, antagonism of myostatin with sActRIIB partially rescued the reduced proliferation rate observed in PPAR␤/␦-null mice primary myoblasts (Fig. 5E). Taken together these data support the notion that activation of PPAR␤/␦ modulates myostatin activity via induction of Gasp-1 protein. To further confirm that PPAR␤/␦ activation regulates myostatin activity, we measured the effect of PPAR␤/␦ activation on myostatin signaling. To assess the level of myostatin activity, we employed the SBE-4x-Luc Smad3-luciferase reporter system used previously to monitor TGF-␤ function (48). C2C12 myoblasts were transfected with SBE-4x-Luc and treated with CM collected either from cells treated with or without L165041 or from primary myoblast cultures isolated from wild type and PPAR␤/␦-null mice. As seen in Fig.  5F, the addition of CM from L165041-treated cells resulted in a significant 24% reduction in SBE-4x-Luc activity. However, the addition of CM from PPAR␤/␦-null primary myoblast cultures significantly increased SBE-4x-Luc activity by 35%. As a positive control we also treated SBE-4x-Luc-transfected myoblasts with sActRIIB myostatin antagonist, and as expected we observed a significant (42%) reduction in SBE-4x-Luc activity upon addition of sActRIIB, consistent with inhibition of myostatin function (Fig. 5F). A recent study has reported that fenofibrate-mediated activation of PPAR␣ results in decreased Mstn mRNA expression (49). Therefore, we next analyzed the consequence of PPAR␤/␦ activation or deletion on the expression of Mstn. However, unlike the decreased Mstn expression observed following activation of PPAR␣, we found no significant change in Mstn mRNA expression upon L165041 treatment (Fig. 5G). Similarly, myostatin protein levels remained unchanged in gastrocnemius muscle tissue collected from PPAR␤/␦-null mice (Fig. 5H).
Exogenous rhGASP-1 Promotes Myoblast Proliferation and Enhances Myotube Formation-Given that PPAR␤/␦-mediated induction of Gasp-1 is associated with enhanced myogenesis and inhibition of myostatin activity, we next assessed whether the addition of rhGASP-1 to human myoblast cultures influences myoblast proliferation and/or differentiation. As shown in Fig. 6A, treatment with rhGASP-1 resulted in a dosedependent increase in myoblast proliferation, with a significant increase in myoblast number detected following treatment with 1, 2, and 5 g of rhGASP-1 protein when compared with respective control-treated cells (0.01% BSA). Treatment with rhGASP-1 also resulted in enhanced myogenic differentiation (Fig. 6B), with a 29.7% increase in myotube number observed in rhGASP-1-treated cells when compared with the control following 72 h of differentiation (Fig. 6C). In addition, we also found an increase in the myotube area, consistent with myotube hypertrophy (Fig. 6D), following the addition of rhGASP-1. In fact, a 19.2% increase in the number of large (3,000 -33,000 m 2 ) myotubes was detected following treatment with rhGASP-1 (Fig. 6D). Clearly, these data suggest that the addition of exogenous rhGASP-1 promotes myogenesis and induces myotube hypertrophy. To ascertain, whether rhGASP-1 protein treatment results in reduced myostatin activity, we next assessed SBE-4x-Luc reporter activity in the presence of rhGASP-1 protein. As expected, the addition of Mstn protein to SBE-4x-Luc-transfected C2C12 myoblasts resulted in a maximal ϳ5.4-fold increase in SBE-4x-Luc reporter activity (Fig. 6E), which is consistent with enhanced myostatin activity. However, in contrast, treatment of SBE-4x-Luc-transfected C2C12 myoblasts with increasing concentrations of rhGASP-1 protein (0.5, 1, and 2 g/ml) resulted in a dose-dependent decrease in SBE-4x-Luc reporter activity, with 2 g/ml treatment resulting in a ϳ7-fold decrease in SBE-4x-Luc reporter activity when compared with the untreated control (Fig. 6E). Taken together, these data indicate that the addition of exogenous rhGASP-1 protein can significantly interfere with myostatin signaling, which is consistent with the increased myoblast proliferation and differentiation observed in response to rhGASP-1 treatment.
shRNA-mediated Knockdown of hGASP-1 Negatively Regulates Myogenesis-To further confirm the role of hGASP-1 in myogenesis, we next generated cell lines stably overexpressing a lentivirus-based shRNA designed to specifically target and repress hGASP-1 expression. Subsequent analysis revealed a significant reduction in hGASP-1 expression both at the mRNA (ϳ80%) and protein (ϳ75%) level ( Fig. 7A and supplemental  Fig. S2E). Next we assessed myoblast proliferation and differ- B, assessment of C2C12 myoblast proliferation, from 0 to 72 h following treatment with CM collected from L165041 and control-treated (DMSO) myoblasts as monitored by methylene blue staining. C, assessment of human myoblast proliferation at 72 h following treatment with DMSO (Control) or L165041 in the absence (PBS) or presence (ϩAb) of 1 g/ml anti-hGASP-1 antibody as monitored by methylene blue assay. D, assessment of C2C12 myoblast proliferation at 48 h following treatment with CM collected from PPAR␤/␦-null and wild type primary myoblast cultures as monitored by methylene blue assay. E, assessment of PPAR␤/␦-null and wild type mice primary myoblast proliferation at 24 and 48 h following treatment with dialysis buffer (DB) or sActRIIB (2 g/ml) protein as monitored by methylene blue assay. F, assessment of SBE-4x-Luc reporter activity in C2C12 myoblasts treated with CM collected from L165041 or control-treated (DMSO) C2C12 myoblasts; in primary myoblast cultures derived from PPAR␤/␦-null and wild type mice; and in C2C12 cells treated with dialysis buffer or with sActRIIB protein (2 g/ml). All SBE-4x-Luc reporter-transfected cultures were grown for 24 h under proliferating conditions prior to collection. The corresponding graph represents the -fold change in luciferase activity normalized to Renilla luciferase. Each bar represents the mean Ϯ S.E. of triplicate samples from two independent experiments. G, qPCR analysis of Mstn mRNA expression in L165041 and control-treated (DMSO) C2C12 myoblasts. The graph represents -fold change normalized to GAPDH expression. Data are mean Ϯ S. E. (n ϭ 3). H, Western blot analysis of Mstn protein expression in gastrocnemius muscle isolated from PPAR␤/␦-null and wild type mice. The corresponding graph (right) shows the optical density values of Mstn protein expression in PPAR␤/␦-null and wild type mice. ␣-Tubulin expression was analyzed to ensure equal loading of samples. *, p Ͻ 0.05; **, p Ͻ 0.01; ***, p Ͻ 0.001. entiation in the hGASP-1 knockdown cells. As seen in Fig. 7B, lack of hGASP-1 resulted in reduced myoblast proliferation with a significant 28.1 and 24.6% reduction in myoblast number detected at 72 and 96 of proliferation, respectively (Fig. 7B). Furthermore, knockdown of hGASP-1 resulted in an observable reduction in myotube formation (Fig. 7C) with a significant 57.3% reduction in myotube number at 96 h of differentiation when compared with control shRNA-transfected cells (Fig.  7D). Taken together, these data suggest that hGASP-1 plays an important role during the normal progression of myogenesis, specifically through regulation of both myoblast proliferation and differentiation. To analyze the effect of GASP-1 knock-down on myostatin activity, we analyzed SBE-4x-Luc reporter activity in hGASP-1 shRNA-transfected C2C12 myoblasts. In agreement with the above results, we observed a ϳ4.4-fold increase in SBE-4x-Luc reporter activity in response to Mstn protein treatment (Fig. 7E). Furthermore, and consistent with enhanced myostatin activity, we detected a ϳ2.8-fold induction in SBE-4x-Luc reporter activity in hGASP-1 shRNA-transfected myoblasts. These data further confirm that GASP-1 is a potent inhibitor of myostatin activity.

DISCUSSION
Pharmacological activation of the muscle-specific PPAR␤/␦ isoform promotes muscle development, myonuclear accretion,  and satellite cell proliferation and restores sarcolemmal integrity in dystrophic mice models (17,19,23,50,51), strongly supporting a role for PPAR␤/␦ in regulating postnatal muscle growth and development. However, no study has yet clearly revealed the molecular mechanism(s) through which PPAR␤/␦ regulates skeletal muscle growth. Using a selective PPAR␤/␦ ligand (L165041) and the PPAR␤/␦-null mouse model, we show here for the first time that PPAR␤/␦ positively regulates postnatal myogenesis through a mechanism involving transcriptional activation of Gasp-1 and reduced activity of the Gasp-1 downstream target myostatin.
Microarray analysis, with subsequent verification by qPCR, revealed a pattern of gene expression changes similar to that observed previously upon ligand-mediated activation of PPAR␤/␦ (44). Specifically, the addition of L165041 resulted in increased expression of genes involved in lipid transport and storage (Abca1, Abcg1, and Adfp), glucose and fatty acid oxidation (Pdk4 and Cpt1b), energy uncoupling, mitochondrial biogenesis (Ucp3 and Ppargc1a), and angiogenesis (Angplt4). One of the novel and significantly up-regulated genes identified following L165041 treatment was Gasp-1 (WFIKKN2), which is a secreted protein that has been reported previously to function as a specific antagonist of myostatin. Subsequent qPCR and Western blot analysis confirmed up-regulation of Gasp-1 expression in both myoblast and myotube cultures following activation of PPAR␤/␦. In addition, the activation of PPAR␤/␦ also resulted in enhanced levels of secreted and thus potentially active Gasp-1 protein into conditioned medium in vitro. Importantly, the up-regulation of Gasp-1 appeared to be specific for PPAR␤/␦, as activation of PPAR␥ and PPAR␣ failed to alter Gasp-1 levels in both myoblasts and myotubes (supplemental Fig. S2, A and C). Further evidence supporting PPAR␤/␦ regulation of Gasp-1 was observed in mdx mice, which exhibited an elevated endogenous PPAR␤/␦ level in muscle tissue together with increased Gasp-1 mRNA expression. Moreover, promoter-reporter analysis revealed that activation of PPAR␤/␦, but not PPAR␣ or PPAR␥, enhanced Gasp-1-promoter reporter activity and that the identified PPRE (DR-1) within the Gasp-1 proximal promoter region was critical for PPAR␤/␦-mediated regulation. Consistent with this, ChIP analysis confirmed interaction of PPAR␤/␦, but not PPAR␣ or PPAR␥, with the DR-1 site of the Gasp-1 gene promoter, which was further enhanced upon L165041 treatment. In agreement with the results described above, we observed significantly reduced Gasp-1 mRNA expression in differentiating primary myoblast cultures as well as fast and slow muscle tissues isolated from PPAR␤/␦-null mice. Taken together these data confirm Gasp-1 as a downstream target of PPAR␤/␦, further supporting the conclusion that PPAR␤/␦ regulates Gasp-1 expression at the transcriptional level during post natal muscle growth.
Previously published studies have revealed that Gasp-1 family proteins are able to bind to and block the function of myostatin (27,28). Similarly, in the current report we have presented several lines of evidence that support Gasp-1 regulation of myostatin in response to PPAR␤/␦ activation. In addition to increased Gasp-1 secretion (as mentioned above), immunoprecipitation studies revealed that there is more interaction between Gasp-1 and myostatin, despite there being no change in Mstn mRNA expression, upon L165041-mediated activation of PPAR␤/␦. Furthermore, and consistent with reduced myostatin activity, the addition of CM from L165041-treated cells resulted in enhanced myoblast proliferation, which was reversed upon the addition of exogenous anti-hGASP-1 antibody. We also observed decreased SBE-4x-Luc reporter activity in response to treatment with L165041 CM, similar to that observed following sActRIIB-mediated blockade of myostatin. Moreover, we observed impaired myoblast proliferation and increased SBE-4x-Luc reporter activity following treatment with PPAR␤/␦-null myoblast CM, which is consistent with both the reduced Gasp-1 expression observed in PPAR␤/␦-null mice and enhanced myostatin activity. We further suggest that enhanced myostatin activity in PPAR␤/␦-null mice is not due to altered myostatin expression, as we observed no change in Mstn mRNA between wild type and PPAR␤/␦-null mice. Taken together these data support the notion that PPAR␤/␦ is able to post-transcriptionally regulate myostatin activity via a mechanism involving Gasp-1. In agreement with Gasp-1 blockade of myostatin function, we also observed enhanced myoblast proliferation, increases in myotube number/size, and a dose-dependent decrease in SBE-4x-Luc reporter activity in response to treatment with rhGASP-1 protein. Furthermore, shRNAmediated knockdown of hGASP-1 resulted in reduced myoblast proliferation, defective myogenic differentiation, and increased SBE-4x-Luc reporter activity.
In the present report we have described for the first time that PPAR␤/␦ is able to regulate both myoblast proliferation and differentiation, which we suggest is through modulation of myostatin activity. These data, together with previously published reports (19,23,51), strongly support a role for PPAR␤/␦ in positively regulating postnatal skeletal muscle growth. However, in contrast to the results presented here, previously published data from Dressel et al. (44) describe that GW501516mediated activation of PPAR␤/␦ does not affect myogenic differentiation of C2C12 myoblasts. However, it is noteworthy to mention that Dressel et al. (44) treated C2C12 myoblasts with GW501516 only after 96 h of differentiation, whereas here we activated PPAR␤/␦ with L165041 treatment immediately upon initiation of differentiation. Therefore, we propose that timely activation of PPAR␤/␦ during the early initiation stages of myogenic differentiation, rather than after terminal differentiation, may be required to promote enhanced differentiation. Moreover, Dressel et al. (44) neither assessed myoblast proliferation nor studied myogenesis using the PPAR␤/␦-null mouse model we have described here. In agreement with the results presented in the current report, a recent study by Angione et al. (51) reports that treatment of primary myoblasts with GW501516 stimulates myoblast proliferation as assessed through measuring the proliferating cell marker Ki67. Furthermore, a study by Han et al. (52) implicates ligand-mediated activation of PPAR␤/␦ in skeletal muscle regeneration. Specifically, treatment of C2C12 myoblasts with CM collected from GW501516-treated endothelial progenitor cells resulted in increased myoblast proliferation as well as enhanced C2C12 myoblast survival during serum starvation (52). In addition, systemic administration of GW501516 to a mouse hind limb ischemia model resulted in increased regenerating muscle fibers, with characteristic centrally formed nuclei (52). It is interesting to surmise that the increased proliferation observed following GW501516 treatment might also be due to the regulation of circulating growth factors such as myostatin; however, further work will need to be performed to confirm this. Taken together these data further confirm a role for PPAR␤/␦ in postnatal skeletal muscle growth and repair. It is important to mention that treatment of C2C12 myoblasts with L165041 resulted in increased myoblast and myotube number without effecting the myotube fusion index or size. However in contrast, the absence of PPAR␤/␦ resulted in reduced myoblast proliferation and differentiation together with impaired myotube fusion and reduced myotube size. Thus, we propose that the L165041mediated increase in myotube formation may be due to the enhanced myoblast number observed as opposed to enhanced myoblast fusion. However the reduced fusion index and myo-tube size observed in PPAR␤/␦-null mice is consistent with the increased myostatin activity in these mice, and in fact, enhanced myostatin signaling has been shown to promote myotubular atrophy and cachexia-like muscle wasting in vitro and in vivo (41,53,54).
We propose that upon stimulation with either exogenous (L165041) or endogenous PPAR␤/␦ ligands (present following exercise or during muscle wasting), PPAR␤/␦ becomes activated (Fig. 8). Once activated PPAR␤/␦ regulates target gene expression, including Gasp-1, via interaction with the functional PPRE (DR-1) located within the Gasp-1 proximal promoter region. Whether or not PPAR␤/␦ regulates Gasp-1 promoter activity in an RXR-dependent or -independent manner remains unclear, and as such, further work will need to be done in verification. Nonetheless, up-regulation of Gasp-1 gene expression results in enhanced secretion of Gasp-1 protein, which is then able to bind to, and regulate, the activity of Gasp- FIGURE 8. PPAR␤/␦ activation modulates myostatin activity via Gasp-1 during postnatal myogenesis. Both exogenous (L165041) and endogenous skeletal muscle ligands (such as released during exercise and muscle wasting) signal to activate PPAR␤/␦. Activated PPAR␤/␦ heterodimerizes with the co-activator RXR and binds to the PPRE (DR-1) in the Gasp-1 gene to facilitate up-regulation of Gasp-1 expression. Gasp-1 undergoes posttranscriptional modification to add secretory signals prior to being secreted into circulation. The secreted Gasp-1 can then interact with Mstn and block further signaling of myostatin through its receptors, which we propose results in the enhanced skeletal muscle myogenesis observed following activation of PPAR␤/␦. 1-interacting proteins such as myostatin. Subsequent Gasp-1 interaction with myostatin blocks myostatin downstream signaling, resulting in increased postnatal muscle growth and development (Fig. 8). In conclusion, these data suggest that PPAR␤/␦ agonists, such as L165041, may not only have therapeutic potential in muscle metabolism but may also be a novel class of therapeutics that have utility in regulating muscle growth and repair.   Analysis of PPARγ (F) and PPARα (G) interaction with the β-actin promoter, in the absence (-) or presence (+) of GW1929 or Wy14643 respectively, was also performed as a negative control (Lower panels). The relative amounts of both the Gasp-1 and β-actin promoters in the input were also assessed and are indicated. Both isotype-specific IgG and no antibody (No Ab) controls are shown. p<0.05 (*) and p<0.01 (**). I n p u t N o A b I g G a n t i -P P A R α