29 Retinoid-related orphan g regulates several that control metabolism in cells: links to modulation of reactive oxygen species production

Suryaprakash Raichur1, Patrick Lau1, Bart Staels2,3 and George E O Muscat1 1Institute for Molecular Bioscience, University of Queensland St Lucia, 4072, Queensland, Australia

2Departement d’Atherosclerose, Institut Pasteur de Lille, Inserm, U545, Lille F-59019, France

3Faculte de Pharmacie et Faculte de Medecine, Universite de Lille 2, Lille F-59019, France

(Requests for offprints should be addressed to G E O Muscat; Email: [email protected])

Abstract

Retinoid-related orphan receptor g (RORg) is an orphan nuclear (NR) that is preferentially expressed in skeletal muscle and several other tissues, including pancreas, thymus, prostate, liver and testis. Surprisingly, the specific role of RORg in skeletal muscle, a peripheral tissue, has not been examined. Muscle is one of the most energy demanding tissues which accounts for w40% of the total body mass and energy expenditure, O75% of glucose disposal and relies heavily on b-oxidation of fatty acids. We hypothesize that RORg regulates metabolism in this major mass lean tissue. This hypothesis was examined by gain and loss of function studies in an in vitro mouseskeletalmusclecellculturemodel.We show that RORg mRNA and protein are dramatically induced during skeletal differentiation. We utilize stable ectopic over-expression of VP16-RORg (gain of function), native RORg and RORgDH12 (loss of function) vectors to modulate RORg mRNA expression and function. Ectopic VP16 (herpes simplex virus transcriptional activator)-RORg and native RORg expression increases RORa mRNA expression. Candidate-driven expression profiling of lines that ectopically express the native and variant forms of RORg suggested that this orphan NR has a function in regulating the expression of genes that control lipid homeostasis (fatty acid-binding protein 4, CD36 (fatty acid translocase), lipoprotein lipase and uncoupling protein 3), carbohydrate metabolism (GLUT5 (fructose transporter), adiponectin receptor 2 and interleukin 15 (IL-15)) and muscle mass (including myostatin and IL-15). Surprisingly, the investigation revealed a function for RORg in the pathway that regulates production of reactive oxygen species. Journal of Molecular Endocrinology (2007) 39, 29–44

Introduction other tissues, including pancreas, thymus, prostate, liver and testis of human (Hirose et al. 1994, Ortiz et al. 1995). Nuclear hormone receptors (NRs) have been demon- RORgt isoform lacks 20 amino acids at amino terminus strated to regulate metabolism in an organ-specific and is restricted to thymocytes (He et al. 1998). manner. The importance of NRs in the context of The NR1F subgroup is closely related to the NR1D promoting and maintaining human health is under- subgroup, including Rev-erba, Rvr/Rev-erbb and the scored by the therapeutic utility of drugs that target Drosophila orphan receptor, E75A, particularly in the dysfunctional hormone signalling in the context of DNA-binding domain and the putative ligand-binding human disease. NRs function as ligand-dependent domain. ROR, Rev-erba and RVR bind as monomers to DNA-binding proteins which translate physiological/ an asymmetric (A/T) 6 RGGTCA motif. ROR functions nutritional/metabolic signals into regulation. The as a constitutive transactivator of , NR superfamily includes the ‘orphan NRs’, which have whereas Rev-erba and RVR do not activate transcrip- no recognized ligands in the ‘orthodox sense’. tion, rather they mediate transcriptional repression, The orphan NR1F subgroup, retinoic and can repress RORa-mediated transactivation from acid receptor-related orphan receptor (ROR/RZR) this motif (Harding & Lazar 1993, Bonnelye et al. 1994, includes three ROR/RZR genes; RORa encodes four Dumas et al. 1994, Forman et al. 1994, Giguere et al. RORa isoforms and is predominantly expressed in blood, 1994, Retnakaran et al. 1994, Adelmant et al. 1996). brain, skeletal muscle and fat cells (Becker-Andre et al. In vivo and in vitro (cell culture) genetic studies have 1993, Giguere et al. 1994). RORb/RZRb is expressed implicated RORa in the regulation of lipid homeostasis. specifically in the brain (Carlberg et al. 1994), and RORg First, NR1F1 (RORa)-deficient mice have a dyslipidaemic encodes two isoforms RORg1 and RORgt. RORg1is phenotype, hypo-a-lipoproteinaemia, muscular atrophy preferentially expressed in skeletal muscle and several and heightened inflammatory responses which lead to

Journal of Molecular Endocrinology (2007) 39, 29–44 DOI: 10.1677/jme.1.00010 0952-5041/07/039–029 q 2007 Society for Endocrinology Printed in Great Britain Online version via http://www.endocrinology-journals.org

Downloaded from Bioscientifica.com at 10/01/2021 04:14:37PM via free access 30 S RAICHUR and others . RORg and skeletal muscle metabolism

atherosclerosis (Vu-Dac et al. 1997, Mamontova et al. 1998, differently. African green monkey kidney COS-1 cells Raspe et al. 1999, Jetten & Ueda 2002, Jetten 2004). were grown in DMEM supplemented with 10% heat- Secondly, ectopic over-expression of RORaDDE in a inactivated fetal calf serum. muscle cell culture model (Lau et al. 2004) leads to the modulation of genes involved in lipid absorption and b-oxidation. RORa (NR1F1) and Rev-erba (NR1D1) RNA extraction and cDNA synthesis opposingly regulate the expression of apolipoprotein Total RNA was extracted from C2C12 cells using TRI- CIII, a component of high density lipoprotein (HDL) and Reagent (Sigma–Aldrich) according to manufacturer’s very low density lipoprotein (VLDL) that regulates protocol. Total RNA was then treated with 2 U Turbo triglyceride (TG) levels and lipoprotein lipase (LPL) DNase 1 (Ambion, Austin, TX, USA) at 37 8C for 30 min activity (Vu-Dac et al. 1997, He et al. 1998, Mamontova et al. followed by purification of the RNA through an RNeasy 1998, Raspe et al. 1999). K K purification column system (Qiagen). RNA was electro- g / Characterization of ROR mice, identified several phoresed to determine the integrity of the preparation. g important immunological functions for ROR ,inthe SuperScript III was used to synthesize cDNA from 3 mg control of lymph node and Peyer’s patch development, K K total RNA using random hexamers according to thymopoiesis and T cell homeostasis. RORg / mice manufacturer’s instructions (Invitrogen). The cDNA display increased apoptosis in the cortex of the thymus, K K was then diluted to 300 ml in nuclease-free water. and RORg / thymocytes placed in culture rapidly undergo apoptosis (Kurebayashi et al. 2000). RORg is highly expressed in skeletal muscle Protein extraction (Hirose et al. 1994). Muscle accounts for w40% of the total body mass and energy expenditure, and O Total soluble protein was extracted from C2C12 cells by 75% of glucose disposal (Baron et al. 1988). This lean the addition of lysis buffer (10 mM Tris (pH 8.0), 150 mM tissue relies heavily on b-oxidation of fatty acids to NaCl, 1% Triton X-100 and 5 mM EDTA) containing supply the extreme energy demands of this organ. protease ‘cocktail’ inhibitors (Rosch). Lysates were However, the fundamental role of RORg in skeletal passed through a 26-gauge needle and centrifuged at muscle lipid and energy homeostasis has not been 10 000 g for 20 min. The supernatant was collected and addressed. We hypothesize that RORg regulates total protein concentration was determined by the metabolism in this major mass lean tissue. This bicinchoninic acid (BCA) as outlined by manufacturer’s hypothesis was tested by ectopic stable over-expression instructions (Pierce Biotechnology Inc., Rockford, IL, of RORg ‘gain and loss’ of function vectors in an USA). in vitro mouse skeletal muscle cell culture model. These studies demonstrated that RORg has a role in controlling the expression of genes that regulate Transient transfections muscle and fat mass (including myostatin and Each well of a 24-well plate of COS-1 cells (w60% interleukin 15 (IL-15)), and lipid homeostasis (fatty confluence) was transfected with a total of w0.6–1.0 mg acid-binding protein 4 (FABP4), CD36 and LPL). of DNA per well using the liposome-mediated transfec- Unexpectedly, the study demonstrated a role for tion procedure as described previously (Lau et al. 1999). RORg in the regulation of reactive oxygen species Briefly, cells were transfected using an N-[1-(2,3-dio- (ROS) production. leoyloxy)propyl]-N,N,N-trimethylammonium methyl- sulphate and metafectene (Biontex Laboratories GmbH, Munich, Germany) liposome mixture in 1! HBS (HEPES-buffered saline; 42 mM HEPES, 275 mM Materials and methods NaCl, 10 mM KCl, 0.4mM Na2HPO4 and 11 mM dextrose (pH 7.1)). The DNA/N-[1-(2,3-dioleoyloxy)- Cell culture propyl]-N,N,N-trimethylammonium methylsulphate/ Mouse myogenic C2C12 cells were cultured in growth metafectene mixture was added to the cells in 0.5– medium called Dulbecco’s modified Eagle’s medium 0.6 ml DMEM supplemented with 10% fetal calf serum. (DMEM) supplemented with 10% heat-inactivated The culture medium was changed 16–24 h later and the Serum Supreme (Cambrex Bio Science, Mt. Waverly, cells were subsequently harvested for the assay of Victoria, Australia) in 6% CO2. Confluent myoblasts were luciferase activity 48 h after the transfection period. differentiated into post-mitotic multinucleated myotubes Fold activation is expressed relative to activity obtained by 5 days (MT5) of serum withdrawal (i.e. cultured in after co-transfection of the promoter–reporter and DMEM supplemented with 2% horse serum). Cells were pSG5/pNL-VP16 vectors only, arbitrarily set at 1. The harvested at the indicated time points, usually 24–120 h mean fold activation values and S.D. were derived from a (1–5 days) after mitogen withdrawal, unless indicated minimum of three independent triplicate experiments.

Journal of Molecular Endocrinology (2007) 39, 29–44 www.endocrinology-journals.org

Downloaded from Bioscientifica.com at 10/01/2021 04:14:37PM via free access RORg and skeletal muscle metabolism . S RAICHUR and others 31

C2C12 stable transfection reactions with either Syber Green or Taqman Tech- nologies (Applied Biosystems). Primers (200 nM) used Myogenic C2C12 cells, cultured in growth medium, for the amplification of target gene sequences have been were co-transfected with pSG5-RORgDH12, pSG5- described in detail (Lau et al. 2004, Ramakrishnan et al. RORg and pNL-VP16-RORg (and pCMVNEO at a m 20:1 ratio of expression vector/NEO vector) by the 2005). PCR was performed with 5 l cDNA and 45 cycles 8 8 liposome-mediated procedure in triplicates. The cells of 95 C for 15 s and 60 C for 1 min. The relative level of were then grown for another 24 h to allow cell recovery expression or fold change and associated errors were and neomycin resistance expression before G418 calculated using the guidelines described by Bookout & selection. After 10- to 14-day selection with 700 mg/ml Mangelsdorf (2003) on the Nuclear Receptor Signaling G418 (Promega) in culture medium, the three Atlas website (NURSA; www.nursa.org/index.cfm)in independent polyclonal pools of stable transfectants accord with the accepted qPCR standards for the National were cultured and maintained on 300 mg/ml G418 Institutes of Health supported NURSA research. medium. Western blot analysis Primers Total soluble protein from the wild-type C2C12 Primers used for qPCR analysis of the mRNA populations myotubes stable pSG5-RORgDH12, pSG5-RORg and have been described in detail (Lau et al. 2004), with the pNL-VP16-RORg C2C12 myotubes were resolved on a exception of primers designed for the detection of 10% SDS–PAGE gel and transferred to a nitrocellulose endogenous RORg using SYBR green (endo RORg membrane. The membranes were blocked overnight or forward, CCCGCCACTCTATAAGGAACTCT and endo for 1 h in 5% skimmed milk in TBS–Tween 20 followed RORg reverse, AGGGCTGAAGGA AATAGAAAGTTGT). by an overnight incubation with either RORg (1:5000, Santa Cruz-28559; Santa Cruz Biotechnology Inc., Santa Cruz, CA, 95060, USA) or RORa (1:2000, Santa Cruz Plasmids 28612), and GAPDH (1:10 000; R&D Systems, Minnea- pSG5-RORa1, mPCP-2x4-tk-LUC(RORaRE) containing polis, MN, USA) antibodies. Following 4!15-min four copies of mouse PCP-2 (GTTATAGTAAC- washes, the membrane was incubated with anti-rabbit TGGGTCAGGGGACT) and pSG5 have been described horseradish peroxidase (HRP) (1:2000) for 1 h. previously (Lau et al. 2004). Mouse RORg1cDNAswere Immunoreactive signals were detected using enhanced amplified from C2C12 total RNA using Pfu Turbo chemiluminescence Super Signal West Pico Substrate (Stratagene) as per manufacturer’s instructions and (Pierce) and visualized by autoradiography on an subsequently cloned into pBluescript. After DNA sequen- XOMAT film developer (Kodak). cing, this insert was cloned into eukaryotic expression vectors pSG5 and pNL-VP16. Subsequently, a truncated version of RORg was amplified (coding 1–494 aa) and Measurement of ROS cloned into pSG5 vectors and used as dominant negative Ten-thousand stable C2C12 cells/well were seeded in a form of RORg.7!(RORgRE)-tk-LUC containing seven 96-well view plate (Perkin–Elmer) and grown in normal copies of RORgRE (GGTAAGTAGGTCAT) was cloned cultured medium (DMEM) with 10% serum. Confluent into pTKLuc as described by Medvedev et al. (1996).The myoblasts were differentiated into post-mitotic multi- human LPL promoter (K1980 bp) in pGL2 reporter vector was kindly provided by Dr Bart Staels (Schoonjans nucleated myotubes by 5 days of serum withdrawal. m et al. 1996). The human Rev-erba promoter (K1733 bp) Myotubes were loaded with 15 MH2DCFDA (ROS in pGL2 reporter vector was kindly provided by Dr probe, Invitrogen) in phenol red-free media for 30 min Vincent Laudet (Adelmant et al. 1996). The mouse at 37 8C and fluorescence was measured. To induce ROS myogenin promoter (K1565 bp) in pCAT reporter was production, after loading ROS probe, we treated the cells kindly provided by Dr E N Olson (Edmondson et al. 1992) with 50 mMH2O2 for about 30 min. The results obtained and, subsequently, it was sub-cloned into pGL2 reporter are from three independent polyclonal pools, assayed in vector in our laboratory. triplicate.

Quantitative real-time PCR (qPCR) Statistical analysis qPCR was performed on an ABI Prism 7500 sequence Statistical analysis was performed on the average of (detection system (Applied Biosystems, Foster city, CA, three independent assays using Student’s t-test or one- USA) in triplicate on three independent RNA prep- way ANOVA, followed by either Tukey’s or Dunnett’s arations. Target cDNA levels were analysed in 25 ml multiple comparison test. www.endocrinology-journals.org Journal of Molecular Endocrinology (2007) 39, 29–44

Downloaded from Bioscientifica.com at 10/01/2021 04:14:37PM via free access 32 S RAICHUR and others . RORg and skeletal muscle metabolism

Results amino acid residues and is conserved in all members of the ROR family. These aromatic amino acid residues are a The RORg mRNA and protein are expressed in a crucial component for recruiting co-factors containing differentiation-dependent manner during skeletal LXXLL motifs (Xie et al. 2005). To measure the RORg- muscle differentiation mediated transactivation on reporter gene, we con- structed the heterologous reporter vector containing g We initially investigated the expression of ROR mRNA sevencopiesofRORgE (GGTAAGTAGGTCAT) in (and protein) relative to GAPDH in the mouse C2C12 pTKLuc (Medvedev et al. 1996). in vitro skeletal muscle cell culture model. Proliferating Initially, we demonstrated that the constitutively active myoblasts can be induced to biochemically and RORg vector (VP16-RORg) significantly increased trans- morphologically differentiate into post-mitotic multi- activation of the synthetic reporter in COS-1 cells relative nucleated myotubes by mitogen withdrawal over several to the native RORg-mediated transactivation (Fig. 2B). days. This transition from a non-muscle to a contractile Subsequently, we demonstrated that the loss of function phenotype is associated with the activation and vector suppressed native RORg-mediated transactivation repression of a structurally diverse group of genes that of a heterologous reporter in COS-1 cells (Fig. 2C). are responsible for the contractile and energy demands Further, the loss of function vector operated in a on this tissue. We utilized qPCR and western blot dominant negative manner and repressed the RORg- analysis to show that RORg mRNA and protein mediated transactivation of the synthetic reporter in a expression (relative to GAPDH) increased during dose-dependent manner (Fig. 2D). skeletal muscle cell differentiation (Fig. 1A). The differentiation-dependent RORg (mRNA and protein) expression was concomitant with the expression Stable ectopic native RORg, VP16-RORg and of several (muscle specific) contractile and metabolic RORgDH12 expression modulates endogenous RORg markers. For example, qPCR demonstrated concomitant mRNA expression induction of myogenin mRNA (a gene that encodes the Further, native RORg,VP16-RORg and pSG5- hierarchical basic helix loop regulator), slow (type I) and RORgDH12 expression vectors were transfected into fast (type II) isoforms of the contractileprotein troponin I skeletal muscle cells, and polyclonal pools of cells (to (Fig. 1B–D). Similarly, differentiation-dependent avoid clonal bias) were isolated after 10–14 days of G418 expression of the genes involved in metabolism (and selection. Subsequently, wild-type C2C12 cells, regulation) of muscle mass was observed. For instance, C2:RORgDH12, native C2:RORg and C2:VP16-RORg fatty acid translocase (CD36), FABP4, myostatin and cell lines were harvested after 5 days of serum withdrawal uncoupling protein 3 (UCP3; Fig. 1E–H). (nZ3). Initially, we measured total RORg mRNA expression in these stably transfected cell lines relative VP16-RORg and RORgDH12 operate as gain and loss to RORg mRNA expression in wild-type C2C12 cells. The of function vectors primers utilized measured total (i.e. both endogenous and ectopic) RORg mRNA expression. This demon- To investigate the metabolic role of RORg in skeletal strated that the stably transfected cell line expressed w20- muscle cells by gain and loss of function strategies, we fold more RORg mRNA. In contrast, western blot analysis proceeded to design vectors (for ectopic expression) that indicated that RORg protein levels did not correlate with enhanced and/or attenuated RORg expression function. the relative total mRNA levels (Fig. 3A). In this context, In order to construct the constitutively activated RORg studies have demonstrated that differential expression of (gain of function) vector, we cloned the VP16 transactiva- mRNAs encoding contractile proteins in cardiac muscle is tion domain (411–456 aa) to the N-terminus of RORg not reflected in protein levels (dos Remedios et al. 1996, (Fig. 2A). We utilized VP16-activated receptor because it Coumans et al. 1997). It should be noted that the antibody still remains unknown whether a natural agonist/ligand utilized cannot discriminate between endogenous exists for the RORg nuclear orphan receptor. Hypothe- (native) and exogenous hybrid RORg proteins, and tically, this fusion would produce a hyperactive receptor. therefore reflects total protein levels. This approach has been used to mimic ligand/agonist- Furthermore, we noticed that although the levels of activated nuclear receptor (Schreiber et al. 2004, Wang total RORg mRNA expression were similar (and not et al. 2004). In addition, to perturb RORg function and to significantly different) in the three stable cell lines, disrupt RORg-mediated gene expression, we constructed endogenous levels of RORg transcripts were signi- the RORgDH12 (dominant negative) expression vector. ficantly reduced in the cell line expressing the This construct encodes amino acids 1–494 and conse- dominant negative form of RORg. Correspondingly, quentlylacks helix12(H12) ofthe ligandbinding domain the lines that express native RORg and constitutively (LBD) that encodes the AF2 transactivation domain active RORg, displayed increased levels of endogenous (Fig. 2A). The AF2 domain contains the LYKELF aromatic RORg transcripts (Fig. 3B, albeit less than twofold).

Journal of Molecular Endocrinology (2007) 39, 29–44 www.endocrinology-journals.org

Downloaded from Bioscientifica.com at 10/01/2021 04:14:37PM via free access RORg and skeletal muscle metabolism . S RAICHUR and others 33

A E CD36 2·5 RORγ 800 ) ) –4 2·0 –4 700 600 1·5 500 1·0 400 300 mRNA/GAPDH mRNA/GAPDH 0·5 Normalized (x10 Normalized (×10 200 0·0 100 RORγ Western blot 0 GAPDH PMB CMB MT1 MT2 MT3 MT4 MT5 PMB CMB MT1 MT2 MT3 MT4 MT5

F FABP4 B Myogenin 0·05

20 ) –4 ) 0·04 –4 15 0·03

10 0·02

5 mRNA/GAPDH Normalized (x10 mRNA/GAPDH 0·01 Normalized (×10 0 0·00 PMB CMB MT1 MT2 MT3 MT4 MT5 PMB CMB MT1 MT2 MT3 MT4 MT5

C TNNI 1 G Myostatin 1200 0·3 ) ) –4 1000 –4 800 0·2 600 400 0·1 mRNA/GAPDH mRNA/GAPDH Normalized (x10 Normalized (×10 200 0 0·0 PMB CMB MT1 MT2 MT3 MT4 MT5 PMB CMB MT1 MT2 MT3 MT4 MT5

D TNNI 2 H UCP3 45 9 40 ) 8 ) –4 35 –4 7 30 6 25 5 20 4 15 3 mRNA/GAPDH

10 mRNA/GAPDH Normalized (×10 2 5 Normalized (×10 1 0 0 PMB CMB MT1 MT2 MT3 MT4 MT5 PMB CMB MT1 MT2 MT3 MT4 MT5 Figure 1 Expression of RORg mRNA and protein during skeletal muscle differentiation in C2C12 myoblasts and differentiated myotubes. Total RNA and soluble protein from proliferating myoblasts (PMBs), confluent myoblasts (CMBs) and MT1–MT5 (days 1–5) of differentiation were analysed by qPCR and western blot analysis. (A) qPCR analysis of RORg mRNA and western blot analysis of RORg protein, GAPDH was used as a loading control in western blot. (B–D) Expression of markers indicative of the acquisition of the muscle specific (myogenin), contractile (TNNI1 and TNNI2) and (E–H) some metabolic genes CD36, FABP4, myostatin and UCP3 during differentiation. Data are expressed as the mean of triplicate samplesGS.D. www.endocrinology-journals.org Journal of Molecular Endocrinology (2007) 39, 29–44

Downloaded from Bioscientifica.com at 10/01/2021 04:14:37PM via free access 34 S RAICHUR and others . RORg and skeletal muscle metabolism

B P<0·001

A P<0·001 250

516aa 1 aa ** 200 RORγ : N A/B DBD Hinge region LBD AF2 C

1aa 516aa 150 VP16-RORγ : N VP16 A/B DBD Hinge region LBD AF2 C 411aa 456 aa 100 P<0·01 1 aa 494aa Relative fold activity 50 RORγ∆H12 : N A/B DBD Hinge region LBD C *

0 7x(RORγE )-tkLuc + + + + pSG5 + – – – pSG5-RORγ – + – – pNLVP16 – – + – pNLVP16-RORγ – – – +

C P<0·001 D P<0·001 8 35 *** ** 7

30 γ 6 *** 25 5 20 4 15 3 *** 10 ** 2 Relative fold activity Fold repression of ROR 5 1

0 0 0·0 0·5 1·0 1·5 2·0 7x(RORγE)-tkLuc + + + + P<0·0001 pSG5 + – + – Ratio of RORγ∆H12/RORγ pSG5-RORγ∆H12 – + – + pSG5-RORγ – – + + Figure 2 (A) Schematic representation of construction of native RORg (pSG5-RORg), gain offunction RORg (VP16-RORg), and loss offunction (pSG5-RORgDH12). (B) Gain offunction VP16-RORg significantly activates RORg-mediated transactivation of heterologous reporter containing RORg-responsive elements. COS-1 cells were co-transfected with 0.5 mg synthetic reporter, 1.0 mg of each of pSG5, pSG5-RORg,pNLVP16 and pNLVP16-RORg. Significance was calculated using one-way ANOVA and Bonferroni’s multiple comparison test where **P!0.001. (C) Loss of function RORg suppresses the RORg-mediated transactivation of heterologous reporter containing RORg-responsive elements. COS-1 cells were co-transfected with 0.5 mg synthetic reporter, 1.0 mgofeachofpSG5,pSG5-RORg and pSG5-RORgDH12. pSG5 was added to normalize the final amount of DNA transfected to 3 mg in three wells of 24-well plate. Significance was calculated using one-way ANOVA and Bonferroni’s multiple comparison test where **P!0.001. (D) Dominant negative RORg suppresses the RORg-mediated transactivation of heterologous reporter containing RORg-responsive elements in dose-dependent manner. COS-1 cells were co-transfected with 0.5 mg synthetic reporter, 1.0 mg of each of pSG5, pSG5-RORg and increasing concentrations of pSG5-RORgDH12. pSG5 was added to normalize the final amount of DNA transfected to 3 mg in three wells of 24-well plate. Fold activation is expressed (meanGS.E.M. and derived from three independent (nZ3); each transfection was composed of three replicates). Significance was calculated using Student’s t-test where ***P!0.0001.

In conclusion, we have produced stable cell lines that sarcomeric slow troponin I (TNNI1) and fast troponin I ectopically over-express dominant negative RORg (TNNI2) contractile proteins was observed in stable and native RORg and constitutively active RORg. wild-type myotubes (Fig. 4A and B). Similar levels of myogenin transcripts were noticed in the lines that over- express the dominant negative and native forms of RORg. Stable ectopic VP16-RORg expression modulates Curiously, the line that over-expressed constitutively myogenin mRNA expression active RORg showed significantly increased levels of The cell lines that ectopically express the dominant myogenin mRNA expression (Fig. 4C).However,as negative, native and activated forms of RORg cell lines observed above the increased levels of myogenin mRNA were differentiated and retained the potential to did not impact on contractile protein mRNA expression. morphologically differentiate. In order to confirm that Further, we tested the potential of ROR (a and g)to the changes observed in these cell lines were not due to transactivate the myogenin promoter. We transiently aberrant differentiation, we measured the levels of transfected COS-1 cells with the reporter plasmid myogenic markers in RNA from differentiated cells. containing K1565 bp of the myogenin promoter and Comparable expression of the mRNAs encoding the examined the effect of co-transfected (exogenous)

Journal of Molecular Endocrinology (2007) 39, 29–44 www.endocrinology-journals.org

Downloaded from Bioscientifica.com at 10/01/2021 04:14:37PM via free access RORg and skeletal muscle metabolism . S RAICHUR and others 35

g a Total RORγ ROR and ROR expression. Interestingly, the myo- A genin promoter was significantly transactivated by P<0·001 RORg (not RORa)(Fig. 4D). P<0·001 P<0·001 Ectopic VP16-RORg expression induces RORa and 350 Rev-erba mRNA expression ** ** 300 ) ** We subsequently profiled the expression of several NRs –4 250 involved in metabolism (Table 1). This was performed 200 to investigate whether ectopic over-expression of dominant negative, native and activated RORg vectors 150 altered the expression of other NRs.

mRNA/GAPDH 100 Interestingly, the cell lines that ectopically expressed Normalized (x10 native and constitutively active RORg were characterized 50 by significantly increased expression of RORa transcripts 0 (Fig. 5A).Furthermore,ectopicVP16-RORg induced significantly greater levels of RORa,relativetonative γ ROR RORg expression (Fig. 5A). Previous studies have Western blot demonstrated that RORa regulates Rev-erba expression GAPDH (Delerive et al. 2002). Moreover, VP16-RORg expression resulted in a corresponding increase in Rev-erba mRNA γ γ expression. This suggested a threshold of RORa γ∆H12 expression may be required for the induction of Rev-erba C2: ROR C2C12 (WT) mRNA expression. Surprisingly, the dominant negative C2: ROR C2: VP16-ROR form of RORg did not alter RORa or Rev-erba expression. The profiling of the other NRs involved in metabolism indicated that the effects of ectopic RORg were relatively γ Endogenous ROR specific. For example, profiling of many NRs involved in B the regulation of lipid homeostasis demonstrated the P<0·01 majority of relevant NRs were refractory to ectopic RORg P<0·01 expression (Table 1). For example, the expression of (LXR)a and LXRb, peroxisome proliferator 3 * * activated receptor (PPAR)a,PPARd and PPARg was refractory to the effects of increased RORg (and the P<0·001

) corresponding changes in RORa and Rev-erba). –4 Previous studies have demonstrated that RORa 2 transactivates the Rev-erba promoter (Delerive et al. 2002). Hence, we co-transfected the Rev-erba promoter with RORa and RORg. This demonstrated that unlike 1 RORa, RORg does not significantly transactivate the mRNA/GAPDH

Normalized (x10 Rev-erba promoter (Fig. 5C). Moreover, it suggests that ** the increased Rev-erba mRNA expression in these cells g a 0 (i.e. C2:VP16-ROR ) is a reflection of increased ROR g γ γ (not ROR ) mRNA and protein expression. γ∆H12 OR C2: ROR C2C12 (WT) C2: ROR C2: VP16-R Stable over-expression of dominant negative, native Figure 3 (A) qPCR analysis of total RORg mRNA and western and activated RORg vectors perturbs the expression blot analysis of total RORg protein in wild-type C2C12 myotubes of several genes in lipid and carbohydrate metabolism and stably transfected dominant negative, native and activated RORg C2C12 myotubes. (B) qPCR analysis of endogenous We then examined each of the stable cell lines for changes RORg mRNA in wild-type C2C12 myotubes and stably trans- in expression of critical genes involved in metabolism fected dominant negative, native and activated RORg C2C12 myotubes. Normalized mRNA expression is expressed as the relative to wild-type C2C12 cells. Surprisingly, expression meanGS.E.M. of three (nZ3) independent samples, each assayed profiling analysis of the cell line that over-expressed the in triplicate. Significance was calculated using one-way ANOVA dominant negative form of RORg (and resulted in and Tukey’s multiple comparison test where **P!0.001. suppression of endogenous RORg mRNA expression) www.endocrinology-journals.org Journal of Molecular Endocrinology (2007) 39, 29–44

Downloaded from Bioscientifica.com at 10/01/2021 04:14:37PM via free access 36 S RAICHUR and others . RORg and skeletal muscle metabolism

A TNNI 1 B TNNI 2

1250 70 ) ) 60 –4 1000 –4 50 750 40

500 30 mRNA/GAPDH mRNA/GAPDH 20 Normalized (x10 250 Normalized (x10 10 0 0 γ γ γ γ H12 γ∆ γ∆H12 C2: ROR C2: ROR C2C12 (WT) C2C12 (WT) C2: ROR C2: VP16-ROR C2: ROR C2: VP16-ROR

C Myogenin D Myogenin promoter P<0·001 10 P<0·001 80 ** ** ) 70 8 –4 60 50 6 40 4 30

mRNA/GAPDH 20 Normalized (x10

Relative fold activity 2 10 0 0 γ γ H12 pGL2-Myogenin(–1·6kb) + + + γ∆ pSG5 + – – C2: ROR C2C12 (WT) pSG5-RORγ – + – C2: ROR C2: VP16-ROR pSG5-RORα – – + Figure 4 (A–C) qPCR analysis of mRNA expression of markers indicative of the acquisition of the muscle contractile (TNNI1 and TNNI2) and muscle specific (myogenin), during differentiation respectively in wild- type C2C12 myotubes and stably transfected dominant negative, native and activated RORg C2C12 myotubes. Normalized mRNA expression is expressed as the meanGS.E.M. of three (nZ3) independent samples, each assayed in triplicate. (D) Promoter analysis of myogenin was performed by co-transfecting with ROR (a and g) in non-skeletal muscle cell lines. COS-1 cells were co-transfected with 1.8 mg reporter (myogenin promoter), and 0.7 mg of each of pSG5, pSG5-RORg and pSG5-RORa. pSG5 was added to normalize the final amount of DNA transfected to 3 mg in three wells of 24-well plate. Fold activation is expressed (meanGS.E.M. and derived from three independent transfections (nZ3); each transfection was composed of three replicates) relative to luciferase activity obtained after co-transfection of the reporter, and pSG5 vector only was arbitrarily set at 1. Significance was calculated using the one-way ANOVA and Tukey’s multiple comparison test where **P!0.001.

did not significantly affect programmes of gene from diabetic patients (Stuart et al. 2007)and expression that control metabolism of lipids and normalized after pioglitazone treatment. carbohydrates (Table 2). Interestingly, expression profil- Interestingly, exogenous expression of activated RORg ing of the cell lines that exogenously expressed native (but not the native RORg) resulted in increased RORg and activated RORg did reveal a functional (and expression of the adiponectin receptor 2 (AdipoR2) distinct) role of RORg in the regulation of metabolism. mRNA (Fig. 6B). AdipoR2 is the receptor for adiponec- Exogenous expression of the native RORg and the tin, an ‘adipokine’ that regulates glucose uptake and fatty activated RORg leads to concordant effects on specific acid oxidation in skeletal muscle (Yamauchi et al. 2003). genes. For example, we observed significant activation of Subsequently, we observed significant activation of GLUT5 mRNA expression in the lines that exogenously IL-15 mRNA expression in the lines that ectopically expressed native RORg and activated RORg (Fig. 6A). over-express native RORg and activated RORg cell lines GLUT5 mRNA encodes the ‘fructose-specific’ glucose (Fig. 6C). IL-15 is highly expressed in skeletal muscle transporter, and its expression is elevated in the muscle and previous studies have demonstrated that IL-15

Journal of Molecular Endocrinology (2007) 39, 29–44 www.endocrinology-journals.org

Downloaded from Bioscientifica.com at 10/01/2021 04:14:37PM via free access RORg and skeletal muscle metabolism . S RAICHUR and others 37

Table 1 Relative mRNA expression of selected nuclear receptors from wild-type C2C12, stably over expressed dominant negative (retinoid-related orphan receptor) RORg C2C12, native RORg C2C12 and activated RORg C2C12 cell lines

C2C12-RORgDH12 C2C12-native RORg C2C12-VP16-RORg Wild-type Transcript Fold change and Transcript Fold change Transcript Fold change and NRs C2C12s levelsa P valueb levelsa and P valueb levelsa P valueb LXRa 3.67G0.06 1.87G0.08 1.9; PZ0.001 2.12G0.08 1.7; PZ0.002 1.5G0.07 2.39; PZ0.001 LXRb 67.4G2.7 123.9G9.81.8; PZ0.007 107.3G3.61.5; PZ0.056 118.9G5.61.7; PZ0.001 PPARa 0.445G0.17 0.261G0.048 1.6; PZ0.050 0.49G0.07 1.0; PZ0.050 0.223G0.025 2.0; PZ0.050 PPARd 33.8G0.97 47.5G2.61.4; PZ0.010 45.5G2.81.3; PZ0.021 52.9G2.71.5; PZ0.003 PPARg 8.0G0.03 17.9G1.52.2; PZ0.004 18.7G1.02.3; PZ0.001 17.8G0.72.2; PZ0.001 Rev-erba 19.0G2.934.0G2.31.78; PZ0.022 26.0G2.81.3; PZ0.021 124.4G6.9 [6.5; PZ0.001c Rev-erbb 30.6G1.440.8G4.01.33; PZ0.140 34.7G2.11.1; PZ0236 33.5G0.89 1.3; PZ0.039 RORa 41.2G2.443.2G2.81.04; PZ0.407 96.13G4.92.3; PZ0.001 181.0G5.3 [4.4; PZ0.001c RORg 1.81G0.19 0.38G0.12 Y4.7; PZ0.001c 2.8G0.16 1.5; PZ0.015 2.7G0.19 1.5; PZ1.000 Nur77 35.6G2.348.7G4.61.36; PZ0.016 35.5G1.11.0; PZ1.000 45.5G1.51.28; PZ0.034 Nurr1 10.4G1.310.5G0.71.0; PZ0.050 7.05G0.31.4; PZ0.010 12.6G1.06 1.2; PZ0.050 NOR1 0.469G0.053 0.514G0.034 1.1; PZ0.050 0.514G0.042 1.1; PZ0.050 0.554G0.071 1.1; PZ0.050 COUP-TFI 9.2G1.322.0G3.02.3; PZ0.001 15.5G2.81.6; PZ0.050 23.3G2.82.53; PZ0.001 COUP-TFII 68.2G7.785.1G5.21.2; PZ0.050 74.6G3.51.0; PZ0.050 98.5G7.21.4; PZ0.010 a K4 Relative expression expressed as a number of transcripts per GAPDH transcript (!10 )GS.D. bFold change and associated P value. cDenotes significant up- or downregulation. increases glucose utilization (Busquets et al. 2006) and negative and native RORg vectors resulted in increased administration of IL-15 to rats and mice inhibits white CD36 and FABP4 mRNA expression (Fig. 7A and B). adipose tissue deposition (Quinn et al. 2005). In In summary, analysis of the lines that over-express summary, several genes involved in carbohydrate ectopic native RORg and activated RORg suggested that metabolism were modulated by ectopic expression of this orphan NR is involved in the regulation of lipid and native and activated RORg. carbohydrate homeostasis, and the control of muscle In addition, we observed significant increases in mass. However, the latter observations (Fig. 7A and B) myostatin (and IL-15) mRNA expression in the lines that present us with two apparent paradoxes. First, over- ectopically over-express native RORg and activated RORg expression of the dominant negative and the native NR cell lines (Fig. 6C and D). Myostatin is a negative regulator produced a similar effect on several target genes (FABP4 of muscle mass and positive regulator of adiposity and CD36). Secondly, ectopic expression of the native (McPherron et al. 1997, Thomas et al. 2000). Moreover, RORg and the activated RORg lead to either opposite or IL-15 has anabolic effects on skeletal muscle protein similar effects (for example, see 6A vs 7A) on specific synthesis both in vivo and in vitro (Quinn et al. 2005). genes. There are a number of possible explanations. As Furthermore, we observed that the expression of the discussed previously, it still remains unknown whether a mRNA encoding LPL (Fig. 6E) was significantly natural agonist/ligand exists for RORg and this coupled enhanced in the lines transfected with the native and to the observation that agonist-dependent NRs can constitutively active RORg. LPL is involved in prefer- function as repressors in the absence of ligand ential lipid utilization and TG hydrolysis. Subsequently, complicates the interpretation of native NR over- to understand the molecular mechanisms involved in expression. The issue is further convoluted by several the significant increase in LPL transcripts in the RORg observations that underscore the permutations associ- over-expressing lines, we analyzed the LPL promoter. We ated with NR-mediated regulation: i) loss of LXR in transiently transfected COS-1 cells with reporter plasmid mice leads to derepression (increased expression) containing K1980 bp of LPL promoter and examined of ABCA1, a classical target gene, whereas another the reporter activity after co-transfection of exogenous target gene, SREBP-1c remains silenced (Wagner et al. RORg and RORa plasmids. Interestingly, LPL was 2003); ii) agonist activation of PPARg and PPARd can significantly transactivated by RORg (not RORa; lead to transcriptional repression (Lee et al. 2003, Fig. 6F). Thus, we speculate that LPL expression is Ghisletti et al. 2007); and iii) orphan NRs (e.g. chicken modulated by RORg expression in skeletal muscle cells. ovalbumin upstream promotor transcriptional factor The expression of the mRNAs encoding CD36 and (COUP-TF)) have been reported to function as either FABP4 (involved in fatty acid and oxysterol absorption) activators and repressors of transcription, or accessory was significantly attenuated in the stably over-expres- factors. These observations can be partially explained by sing activated RORg cell line relative to the wild type. differential co-repressor/coactivator recruitment and Surprisingly, the over-expression of the dominant displacement in gene- and cell-specific manner. www.endocrinology-journals.org Journal of Molecular Endocrinology (2007) 39, 29–44

Downloaded from Bioscientifica.com at 10/01/2021 04:14:37PM via free access 38 S RAICHUR and others . RORg and skeletal muscle metabolism

RORg gain and loss of function studies identify a role RORα for RORg in the regulation of UCP3 mRNA expression and ROS production A P<0·001 200 P<0·001 ** We further investigated the physiological role of RORg in skeletal muscle cells by measuring steady-state ATP )

–4 levels, the rate of fatty acid oxidation and ROS P<0·001 production. Surprisingly, we did not observe any 100 ** significant change in fatty acid oxidation and total ATP content (data not shown) in the stably over- expressing dominant negative, native and activated mRNA/GAPDH

Normalized (x10 RORg cell lines relative to wild-type C2C12 cells. Additional expression profiling revealed that the 0 expression of the mRNA encoding UCP3 (Fig. 8A) was RORα significantly increased in the cell line over-expressing Western blot g GAPDH VP16-ROR . UCPs are inner mitochondrial membrane transporters that uncouple substrate oxidation from γ T) γ ATP synthesis, converting fuel to heat. However, many γ∆H12 C2C12 (W C2: ROR studies have suggested that UCP3 is involved in C2: ROR C2: VP16-ROR preferential lipid utilization and modulation of ROS production in skeletal muscle cells. UCP3 knockout Rev-erbα mice and adenoviral over-expression in skeletal muscle B cells have demonstrated that aberrant UCP3 expression P<0·001 150 effects ROS production (Vidal-Puig et al. 2000, ** MacLellan et al. 2005). We investigated whether VP16-

) RORg mediated increases in UCP3 mRNA expression –4 100 in skeletal muscle cells, altered ROS levels relative to the wild-type (and dominant negative) RORg expression in C2C12 cells. Interestingly, the line that 50 over-expressed ectopic-activated RORg showed a signi- mRNA/GAPDH

Normalized (x10 ficant decrease in endogenous and (hydrogen peroxide (H2O2)) induced ROS production relative to the 0 γ dominant negative and wild-type cells (Fig. 8B). T) γ R γ∆H12 C2C12 (W C2: ROR C2: ROR C2: VP16-RO Discussion Rev-erbα promoter P<0·001 Genetic, molecular and biochemical studies have C demonstrated that RORa has a distinct role in lipid ** metabolism. For example, RORa-deficient mice develop 6 severe atherosclerosis, hypo-a-lipoproteinaemia, 5 3 C2C12 myotubes. (B) qPCR analysis of Rev-erba mRNA in wild- 4 type C2C12 myotubes and stably transfected dominant negative, native and activated RORg C2C12 myotubes. Normalized mRNA 3 expression is expressed as the meanGS.E.M. of three (nZ3) independent samples, each assayed in triplicate. (C) Promoter 2 analysis of Rev-erba was performed by co-transfecting with ROR Relative fold activity 1 (a and g) in non-skeletal muscle cell lines. COS-1 cells were co-transfected with 1.8 mg reporter (Rev-erba promoter), 0.7 mgof 0 each of pSG5, pSG5-RORg and pSG5-RORa. pSG5 was added to normalize the final amount of DNA transfected to 3 mg in three pGL2-Rev-erbα(–1·7kb) + + + wells of 24-well plate. Fold activation is expressed (meanGS.E.M. pSG5 + – – and derived from three independent transfections (nZ3); each pSG5-RORγ – + – transfection was composed of three replicates) relative to pSG5-RORγ – – + luciferase activity obtained after co-transfection of the reporter, Figure 5 (A) qPCR analysis of total RORa mRNA and western and pSG5 vector only was arbitrarily set at 1. Significance was blot analysis of RORa protein in wild-type C2C12 myotubes and calculated using one-way ANOVA Tukey’s multiple comparison stably transfected dominant negative, native and activated RORg test where **P!0.001.

Journal of Molecular Endocrinology (2007) 39, 29–44 www.endocrinology-journals.org

Downloaded from Bioscientifica.com at 10/01/2021 04:14:37PM via free access RORg and skeletal muscle metabolism . S RAICHUR and others 39

Table 2 Relative mRNA expression of genes involved in metabolism from wild-type C2C12, stably over expressed dominant negative (retinoid-related orphan receptor) RORg C2C12, native RORg C2C12 and activated RORg C2C12 cell lines

C2C12-RORgDH12 C2C12-native RORg C2C12-VP16-RORg Wild-type Transcript Fold change Transcript Fold change Transcript Fold change and Genes C2C12s levelsa and P valueb levelsa and P valueb levelsa P valueb Energy expenditure and balance UCP2 458.4G22.0 445.1G11.01.0; PZ0.418 643.4G38.01.4; PZ0.035 464.0G48.01.01; PZ0.536 UCP3 2.89G0.31 7.01G0.49 2.4; PZ0.003 8.48G0.72 2.9; PZ0.001 23.3G2.5 [8.06; PZ0.001c AMPKb1 26.6G0.523.9G1.61.1; PZ0.049 18.8G0.81.4; PZ0.036 21.5G1.31.2; PZ0.032 AMPKg3 61.2G5.386.7G2.81.41; PZ0.008 100.2G6.91.63; PZ0.025 147.8G14.02.41; PZ0.006 Fatty acid and lipid absorption CD36 718.6G55.0 1654.7G15.32.3; PZ0.001 1740.3G90.02.42; PZ0.001 174.9G18.3 Y4.1; PZ0.001c FABP4 0.18G0.01 0.32G0.05 1.77; PZ0.089 0.3G0.04 1.6; PZ0.012 0.03G0.001 Y6.0; PZ0.001c FABP3 12.06G0.616.03G1.01.3; PZ0.040 12.29G0.61.0; PZ0.256 11.29G0.41.06; PZ0.203 Lipid catabolism ACS4 69.5G3.3 136.0G8.21.95; PZ0.002 140.0G12.02.01; PZ0.014 158.3G6.52.27; PZ0.001 LPL 43.9G1.739.2G2.41.34; PZ0.008 411.2G31.0 [9.5; PZ 322.6G27.2 [7.34; PZ0.001c 0.001c AdipoR1 529.6G19.5 752.7G66.31.3; PZ0.050 688.2G22.81.3; PZ0.025 768.3G41.31.45; PZ0.008 AdipoR2 2.1G0.73.0G0.91.4; PZ0.159 3.7G0.37 1.76; PZ0.031 8.2G0.7 [3.9; PZ0.001c CPT-1 2.4G0.14 2.6G0.26 1.08; PZ0.358 2.5G0.25 1.04; PZ0.574 1.44G0.14 1.6; PZ0.011 PDK-2 48.8G1.284.61G4.51.73; PZ0.002 83.45G5.11.71; PZ0.025 94.38G4.07 1.9; PZ0.001 PDK-4 10.4G1.115.2G1.11.4; PZ0.023 17.3G2.11.6; PZ0.021 39.2G2.6 [3.9; PZ0.001c PGC-1 4.3G0.26 5.0G0.24 1.16; PZ0.026 3.8G0.23 1.13; PZ0.025 5.1G0.51.1; PZ0.698 MCAD 44.7G2.252.9G4.51.1; PZ0.041 59.6G4.11.29; PZ0.024 62.4G5.21.39; PZ0.056 Lipid efflux and homeostasis ABCA1 15.9G2.424.3G1.51.5; PZ0.068 16.7G1.51.0; PZ0.258 15.0G1.31.06; PZ0.985 ABCG1 2.98G0.24 5.1G0.51.71; PZ0.027 3.5G0.21 1.1; PZ0.754 4.0G0.36 1.34; PZ0.151 ApoE 27.9G0.423.9G2.61.1; PZ0.580 88.5G5.5 [3.1; PZ 50.1G2.51.8; PZ0.001 0.001c CAV-3 199.8G4.9 184.1G6.21.08; PZ0.063 166.2G8.41.2; PZ0.241 218.5G9.21.09; PZ0.0360 Glucose and lipid storage GYG1 29.02G0.965.7G4.22.24; PZ0.001 58.8G4.92.06; PZ0.002 73.9G3.82.5; PZ0.001 ADRP 189.1G11.2 380.1G16.12.01; PZ0.001 340.9G20.91.7; PZ0.025 253.5G9.31.3; PZ0.016 Lipogenesis FAS 133.0G9.9 228.0G20.41.71; PZ0.019 153.0G10.41.15; PZ0.252 222.1G23.71.6; PZ0.039 SCD-1 3899.6G234.0 6194.3G293.01.58; PZ0.004 5694.3G262.01.4; PZ0.265 6746.0G362.01.7; PZ0.003 SCD-2 892.0G24.0 1723.0G65.01.93; PZ0.001 1619.0G146.01.81; PZ0.015 339.5G37.32.7; PZ0.001 SREBP1c 18.3G1.722.7G1.41.24; PZ0.258 24.5G1.81.33; PZ0.154 12.56G1.41.4; PZ0.107 Myokines IL-15 11.41G1.215.99G1.21.4; PZ0.145 36.5G2.2 [3.17; 28.03G0.55 2.4; PZ0.002 PZ0.001c Myostatin 0.38G0.21.1G0.22.8; PZ0.026 9.62G0.7 [25.3; 9.21G0.29 [24.2; PZ0.001c PZ0.001c Sugar uptake GLUT4 0.12G0.003 0.218G0.02 1.8; PZ0.011 0.204G0.01 1.7; PZ0.154 0.25G0.02 2.0; PZ0.004 GLUT5 0.03G0.0008 0.04G0.01 1.3; PZ0.025 0.11G0.02 [3.6; 0.12G0.002 [4.0; PZ0.001c PZ0.001c Others VCAM-1 86.5G3.393.4G2.71.07; PZ0.487 145.9G5.31.68; PZ0.015 134.3G6.41.5; PZ0.003 a K4 Relative expression expressed as number of transcripts per GAPDH transcript (!10 )GS.D. bFold change and associated P value. cDenotes significant up- or downregulation. hypotriglyceridaemia, muscular atrophy and heigh- Although, RORg is highly expressed in skeletal muscle tened inflammatory responses (Vu-Dac et al. 1997, (a major mass peripheral metabolic tissue), the Mamontova et al. 1998, Raspe et al. 1999, Jetten & Ueda functional role of this orphan NR in metabolism has 2002, Jetten 2004). In concordance, in vitro cell culture not been explored. We utilized the C2C12 skeletal muscle studies indicated that RORa regulates the expression of cell culture model to investigate whether RORg regulates genes involved in lipid homeostasis in skeletal muscle genetic programmes implicated in metabolism. This cells (Lau et al. 2004). in vitro model system is robust and data derived from www.endocrinology-journals.org Journal of Molecular Endocrinology (2007) 39, 29–44

Downloaded from Bioscientifica.com at 10/01/2021 04:14:37PM via free access 40 S RAICHUR and others . RORg and skeletal muscle metabolism

A GLUT5 B AdipoR C IL 15 P<0·001 P<0·001 P<0·001 P<0·001 0.15 P<0·001 P<0·001 9 ** 35 **

** ) ) 8 ) ** –4 –4 ** 30 7 –4 0.10 6 25 5 20 4 15 0.05 3 10 mRNA/GAPDH mRNA/GAPDH 2 mRNA/GAPDH Normalized (x10 Normalized (x10 1 Normalized (x10 5 0.00 0 0 γ γ γ γ γ γ T) 12 R R T) 12 R R T) R R γ∆H O γ∆H O γ∆H12 O RO RO 6-RO 12 (W OR C2: R 12 (W OR C2: R 12 (W R C2: R C2C P16- C2C P16- C2C C2: R C2: V C2: R C2: V C2: RO C2: VP1

D Myostatin E LPL F LPL promoter

P<0·001 P<0·001 P<0·001 2.5 ** P<0·001 12.5 P<0·001 ) 450 ** 2.0 –4

** ) 10.0 ** 400 –4 350 ** 1.5 7.5 300 250 5.0 200 1.0 150 mRNA/GAPDH 2.5 0.5 Normalized (x10 Relative fold activity mRNA/GAPDH 100 0.0 Normalized (x10 50 γ γ 0 0.0 T) 12 R R γ γ γ∆H O RO T) 12 R R 12 (W R 2: R γ∆H O O pGL2-LPL(–2.0kb) + + + 2C O C P16- 12 (W R 2: R 6-R pSG5 C 2C O C + – – C2: R C2: V C pSG5-RORγ – + – C2: R C2: VP1 pSG5-RORα – – + Figure 6 qPCR analysis of (A) GLUT5 mRNA, (B) AdipoR mRNA, (C) IL-15 mRNA, (D) myostatin mRNA and (E) LPL mRNA from wild- type C2C12 myotubes and stably transfected dominant negative, native and activated RORg C2C12 myotubes. Normalized mRNA expression is expressed as the meanGS.E.M. of three (nZ3) independent samples, each assayed in triplicate. (F) Analysis of the LPL promoter was performed by co-transfecting with ROR (a and g) in non-skeletal muscle cell lines. COS-1 cells were co-transfected with 1.8 mg reporter (LPL promoter), 0.7 mg of each of pSG5, pSG5-RORg and pSG5-RORa. pSG5 was added to normalize the final amount of DNA transfected to 3 mg in three wells of 24-well plate. Fold activation is expressed (meanGS.E.M. and derived from three independent transfections (nZ3); each transfection was composed of three replicates) relative to luciferase activity obtained after co-transfection of the reporter, and pSG5 vector only was arbitrarily set at 1. Significance was calculated using one-way ANOVA and Tukey’s multiple comparison test where **P!0.001.

the analysis of differentiated post-mitotic multinucleated gene expression involved in metabolism. This is myotubes with LXR and PPARd agonists have been consistent with the phenotype of RORg-deficient mice validated/reproduced in mice (Muscat et al. 2002, Dressel that appear normal in the context of lipid and et al. 2003, Holst et al. 2003, Wang et al. 2004). carbohydrate homeostasis. We probed RORg function in skeletal muscle cells by Interestingly, we observed that ectopic over- stable ectopic expression of vectors that encode expression of activated RORg significantly increased dominant negative, native and activated RORg. The the expression of RORa and Rev-erba mRNA. These experiments indicated that the RORg dominant observations are consistent with the previous reports negative vector in skeletal muscle cells significantly demonstrating that RORa directly transactivates the repressed the endogenous levels of the RORg tran- Rev-erba promoter (Delerive et al. 2002). scripts, and RORg-dependent gene expression. In the context of carbohydrate metabolism, we However, we did not observe any significant change in observed the lines expressing native and activated

Journal of Molecular Endocrinology (2007) 39, 29–44 www.endocrinology-journals.org

Downloaded from Bioscientifica.com at 10/01/2021 04:14:37PM via free access RORg and skeletal muscle metabolism . S RAICHUR and others 41

A CD36 B FABP4 P<0·001 P<0·001

P<0·001 P<0·001 P<0·001 P<0·001 0·35 ** 2000 **

** )

) ** 0·30 –4 –4 0·25 10 10 × × 0·20 1000 0·15 0·10 mRNA/GAPDH mRNA/GAPDH Normalized ( Normalized ( ** 0·05 ** 0 0·00 γ γ ) γ γ H12 OR OR H12 OR OR (WT) γ∆ γ∆ C2: R C2: R C2C12 C2C12 (WT C2:ROR C2: VP16-R C2:ROR C2: VP16-R Figure 7 qPCR analysis of (A) CD36 mRNA and (B) FABP4 mRNA from wild-type C2C12 myotubes and stably transfected dominant negative, native and activated RORg C2C12 myotubes. Normalized mRNA expression is expressed as the mean G S.E.M. of three (nZ3) independent samples, each assayed in triplicate. Significance was calculated using one-way ANOVA Tukey’s and Dunnett’s multiple comparison test where **P!0.001.

RORg expressed significantly elevated levels of GLUT5 because of a secondary effect on the integrity of the mRNA. This transcript is increased in skeletal muscle mitochondria (Schrauwen & Hesselink 2002). and repressed in adipose in the insulin-resistant We also demonstrated that the expression of mRNA diabetic state (Litherland et al. 2004) and normalized which encodes myostatin, a negative regulator of by pioglitazone therapy. Concordantly, we noted a skeletal muscle mass and positive regulator of adiposity, significant increase in IL-15 mRNA expression that is significantly increased in the stable lines which encodes a cytokine expressed in skeletal muscle which ectopically express native and activated RORg induces glucose uptake and oxidation (Busquets et al. (McPherron et al. 1997). In this context, we observe 2006). elevated levels of IL-15 that have been demonstrated to Exogenous expression of the VP16-RORg in skeletal have anabolic effects in muscle, and decrease white muscle cells resulted in the significant elevation of adipose tissue mass in rodents (Quinn et al. 2005). In transcripts encoding UCP3, which correlated with summary, it appears that RORg modulates genes attenuated production of ROS. UCPs are inner involved in the regulation of muscle and adipose mass. g mitochondrial membrane transporters that uncouple Notably, over-expression of native ROR and g substrate oxidation from ATP synthesis, converting fuel chimeric VP16-ROR implicated the orphan NR in the regulation of lipid metabolism. For example, in to heat. However, UCP3 has been implicated in the the RORg gain of function cell lines, we observed transport of fatty acid anions (Himms-Hagen & Harper elevated LPL mRNA expression. LPL is a rate- 2001) and regulating the accumulation of ROS limiting enzyme, which hydrolyses the TG-rich (Vidal-Puig et al. 2000). For example, reduced UCP3 lipoproteins into fatty acids in skeletal muscle. expression has been linked to increased ROS pro- Furthermore, we demonstrated that RORg (but duction (Vidal-Puig et al. 2000). Moreover, increased not RORa) transactivates the LPL promoter. In the expression minimizes ROS production (MacLellan et al. context of lipid homeostasis, we observed significant 2005). In this context, decreased UCP3 expression in repression of CD36 and FABP4 mRNAs (that encode skeletal muscle is associated with insulin resistance in proteins involved in fatty acid absorption) in the diabetic individuals (Schrauwen et al. 2001, Patti et al. VP16-RORg cell line. Whether RORg regulates the 2003, Houstis et al. 2006). Increased ROS production LPL promoter by primary (direct) and/or secondary may critically damage cell membranes and reduce mechanisms remain unclear at present. Future mitochondrial number. In the course of mitochondrial studies utilizing ChIP, deletion and EMSA analysis respiration, increased ROS production may lead to will be used to address these questions. deleterious changes in transmembrane potential. These latter observations raise questions about the Therefore, UCP3 may influence energy expenditure complex mechanisms that mediate the sometimes www.endocrinology-journals.org Journal of Molecular Endocrinology (2007) 39, 29–44

Downloaded from Bioscientifica.com at 10/01/2021 04:14:37PM via free access 42 S RAICHUR and others . RORg and skeletal muscle metabolism

A UCP3 NRs that respectively functionasactivatorsandrepressors P<0·001 of transcription. This expression pattern of NRs may account for the contrasting effects of native and VP16- 25 ** RORg over-expression in skeletal muscle cells. In addition, differential effects of either NR gain or loss of function on 20 specific targets are not unforeseen paradoxes. As discussed )

–4 previously, it is not clear whether RORg is modulated by 10

× 15 natural compounds (Stehlin-Gaon et al. 2003), and this impacts on whether the NR can be additionally modulated. For example, oestrogen-related receptors can function 10 independently of ligands, but retain the capacity to

mRNA/GAPDH modulate by agonists and antagonists (Kamei et al. 2003, Normalized ( 5 Rodriguez-Calvo et al. 2006). Furthermore, gain and loss of NR function can differentially effect target genes (LXR, 0 PPARd and COUP-TF) due to differential cofactor recruit- γ γ ment and displacement in diverse cells and tissues (Lee γ∆ H12 C2: ROR et al. 2003, Wagner et al. 2003). C2C12 (WT) C2: ROR C2: VP16-ROR In conclusion, we suggest that RORg in the skeletal muscle cells controls several programmes of gene B Reactive oxygen species expression that have important roles in the control of P<0·001 muscle growth, lipid and carbohydrate metabolism. Moreover, we present evidence that this orphan NR may 5000000 modulate mitochondrial function via the UCP3- mediated regulation of ROS production. 4000000

P<0·001 ** 3000000 Acknowledgements We thank Dr Mary Wang, Rebecca Fitzsimmons, Rachel 2000000 Burow, Shayama Wijedasa and Dr Stephen Myers for ** their excellent assistance. This work was supported by a National Health and Medical Research Council H2DCFDA Fluoresence 1000000 (NHMRC) of Australia project grant. G E O M is a Principal Research Fellow of the NHMRC, Australia. S R is a recipient of International Postgraduate Research 0 Scholarship (IPRS), Australia. The authors declare that Endogenous ROS Induced (H2O2) ROS ) γ ) γ there is no conflict of interest that would prejudice the H12 H12 γ∆ γ∆ impartiality of this scientific work. The IMB is an 12 (WT 6-ROR 6-ROR Australian Research Council Special Research Centre in C2C C2C12 (WT C2: RORC2: VP1 C2: RORC2: VP1 Functional and Applied Genomics. Figure 8 (A) qPCR analysis of UCP3 mRNA from wild-type C2C12 myotubes and stably transfected dominant negative, native and activated RORg C2C12 myotubes. Normalized References mRNA expression is expressed as the meanGS.E.M.ofthree (nZ3) independent samples, each assayed in triplicate. (B) Measurement of endogenous and induced ROS in wild-type Adelmant G, Begue A, Stehelin D & Laudet V 1996 A functional Rev- C2C12 myotubes and stably transfected dominant negative erba responsive element located in the human Rev-erba promoter and activated RORg C2C12 myotubes. Significance was mediates a repressing activity. PNAS 93 3553–3558. calculated using one-way ANOVA and Tukey’s multiple Baron AD, Brechtel G, Wallace P & Edelman SV 1988 Rates and tissue comparison test where **P!0.001. sites of non-insulin- and insulin-mediated glucose uptake in humans. American Journal of Physiology 255 E769–E774. Becker-Andre M, Andre E & DeLamarter JF 1993 Identification of nuclear receptor mRNAs by RT-PCR amplification of conserved apparently disparate effectofectopicnativeRORg versus zinc-finger motif sequences. Biochemical and Biophysical Research activated RORg expression in skeletal muscle cells on a Communications 194 1371–1379. subset of target genes. One clear explanation is that Bonnelye E, Vanacker JM, Desbiens X, Begue A, Stehelin D & Laudet V 1994 Rev-erbb, a new member of the nuclear receptor superfamily, VP16-RORg (but not native RORg) expression activates is expressed in the nervous system during chicken development. both RORa and Rev-erba. These are opposingly acting Cell Growth & Differentiation 5 1357–1365.

Journal of Molecular Endocrinology (2007) 39, 29–44 www.endocrinology-journals.org

Downloaded from Bioscientifica.com at 10/01/2021 04:14:37PM via free access RORg and skeletal muscle metabolism . S RAICHUR and others 43

Bookout AL & Mangelsdorf DJ 2003 Quantitative real-time PCR Jetten AM & Ueda E 2002 Retinoid-related orphan receptors (RORs): protocol for analysis of nuclear receptor signaling pathways. Nuclear roles in cell survival, differentiation and disease. Cell Death and Receptor Signaling 1 e012. Differentiation 9 1167–1171. Busquets S, Figueras M, Almendro V, Lopez-Soriano FJ & Argiles JM Kamei Y, Ohizumi H, Fujitani Y, Nemoto T, Tanaka T, Takahashi N, 2006 Interleukin-15 increases glucose uptake in skeletal muscle. An Kawada T, Miyoshi M, Ezaki O & Kakizuka A 2003 PPARg coactivator antidiabetogenic effect of the cytokine. Biochimica et Biophysica Acta 1beta/ERR ligand 1 is an ERR protein ligand, whose expression 1760 1613–1617. induces a high-energy expenditure and antagonizes obesity. PNAS Carlberg C, Hooft van Huijsduijnen R, Staple JK, DeLamarter JF & 100 12378–12383. Becker-Andre M 1994 RZRs, a new family of retinoid-related orphan Kurebayashi S, Ueda E, Sakaue M, Patel DD, Medvedev A, Zhang F & receptors that function as both monomers and homodimers. Jetten AM 2000 Retinoid-related orphan receptor gamma (RORg) Molecular Endocrinology 8 757–770. is essential for lymphoid organogenesis and controls apoptosis Coumans JV, Yeoh T, Seeto RK, Keogh A, Brennan K, Gunning P, during thymopoiesis. PNAS 97 10132–10137. Hardeman E & dos Remedios CG 1997 Variations in the relative Lau P, Bailey P, Dowhan DH & Muscat GE 1999 Exogenous expression mRNA levels of actins and myosin heavy chains do not produce of a dominant negative RORa1 vector in muscle cells impairs corresponding differences in their proteins in the adult human differentiation: RORa1 directly interacts with p300 and myoD. heart. Journal of Molecular and Cellular Cardiology 29 895–905. Nucleic Acids Research 27 411–420. Delerive P, Chin WW & Suen CS 2002 Identification of Reverb(alpha) Lau P, Nixon SJ, Parton RG & Muscat GE 2004 RORa regulates the as a novel ROR(alpha) target gene. Journal of Biological Chemistry 277 expression of genes involved in lipid homeostasis in skeletal muscle 35013–35018. cells: caveolin-3 and CPT-1 are direct targets of ROR. Journal of Dressel U, Allen TL, Pippal JB, Rohde PR, Lau P & Muscat GE 2003 Biological Chemistry 279 36828–36840. The peroxisome proliferator-activated receptor beta/delta agonist, Lee CH, Chawla A, Urbiztondo N, Liao D, Boisvert WA, Evans RM & GW501516, regulates the expression of genes involved in lipid Curtiss LK 2003 Transcriptional repression of atherogenic inflam- catabolism and energy uncoupling in skeletal muscle cells. mation: modulation by PPARdelta. Science 302 453–457. Molecular Endocrinology 17 2477–2493. Litherland GJ, Hajduch E, Gould GW & Hundal HS 2004 Fructose Dumas B, Harding HP, Choi HS, Lehmann KA, Chung M, Lazar MA & transport and metabolism in adipose tissue of Zucker rats: Moore DD 1994 A new orphan member of the nuclear hormone diminished GLUT5 activity during obesity and insulin resistance. receptor superfamily closely related to Rev-Erb. Molecular Endo- Molecular and Cellular Biochemistry 261 23–33. crinology 8 996–1005. MacLellan JD, Gerrits MF, Gowing A, Smith PJ, Wheeler MB & Harper Edmondson DG, Cheng TC, Cserjesi P, Chakraborty T & Olson EN ME 2005 Physiological increases in uncoupling protein 3 augment 1992 Analysis of the myogenin promoter reveals an indirect pathway fatty acid oxidation and decrease reactive oxygen species for positive autoregulation mediated by the muscle-specific production without uncoupling respiration in muscle cells. Diabetes enhancer factor MEF-2. Molecular and Cellular Biology 12 3665–3677. 54 2343–2350. Forman BM, Chen J, Blumberg B, Kliewer SA, Henshaw R, Ong ES & Mamontova A, Seguret-Mace S, Esposito B, Chaniale C, Bouly M, Evans RM 1994 Cross-talk among RORa1 and the Rev-erb family of Delhaye-Bouchaud N, Luc G, Staels B, Duverger N, Mariani J et al. orphan nuclear receptors. Molecular Endocrinology 8 1253–1261. 1998 Severe atherosclerosis and hypoalphalipoproteinemia in the Ghisletti S, Huang W, Ogawa S, Pascual G, Lin ME, Willson TM, staggerer mouse, a mutant of the nuclear receptor RORa. Rosenfeld MG & Glass CK 2007 Parallel SUMOylation-dependent Circulation 98 2738–2743. pathways mediate gene- and signal-specific transrepression by LXRs McPherron AC, Lawler AM & Lee SJ 1997 Regulation of skeletal and PPARg. Molecular Cell 25 57–70. muscle mass in mice by a new TGF-b superfamily member. Nature Giguere V, Tini M, Flock G, Ong E, Evans RM & Otulakowski G 1994 387 83–90. Isoform-specific amino-terminal domains dictate DNA-binding Medvedev A, Yan ZH, Hirose T, Giguere V & Jetten AM 1996 Cloning properties of RORa, a novel family of orphan hormone nuclear of a cDNA encoding the murine orphan receptor RZR/RORg and receptors. Genes and Development 8 538–553. characterization of its response element. Gene 181 199–206. Harding HP & Lazar MA 1993 The orphan receptor Rev-ErbAa Muscat GE, Wagner BL, Hou J, Tangirala RK, Bischoff ED, Rohde P, activates transcription via a novel response element. Molecular and Petrowski M, Li J, Shao G, Macondray G et al. 2002 Regulation Cellular Biology 13 3113–3121. of cholesterol homeostasis and lipid metabolism in skeletal He YW, Deftos ML, Ojala EW & Bevan MJ 1998 RORg t, a novel isoform muscle by liver X receptors. Journal of Biological Chemistry 277 of an orphan receptor, negatively regulates Fas ligand expression 40722–40728. and IL-2 production in T cells. Immunity 9 797–806. Ortiz MA, Piedrafita FJ, Pfahl M & Maki R 1995 TOR: a new orphan Himms-Hagen J & Harper ME 2001 Physiological role of UCP3 may be receptor expressed in the thymus that can modulate retinoid and export of fatty acids from mitochondria when fatty acid oxidation thyroid hormone signals. Molecular Endocrinology 9 1679–1691. predominates: an hypothesis. Experimental Biology and Medicine 226 Patti ME, Butte AJ, Crunkhorn S, Cusi K, Berria R, Kashyap S, Miyazaki 78–84. Y, Kohane I, Costello M, Saccone R et al. 2003 Coordinated Hirose T, Smith RJ & Jetten AM 1994 RORg: the third member of reduction of genes of oxidative metabolism in humans with insulin ROR/RZR orphan receptor subfamily that is highly expressed in resistance and diabetes: potential role of PGC1 and NRF1. PNAS skeletal muscle. Biochemical and Biophysical Research Communications 100 8466–8471. 205 1976–1983. Quinn LS, Strait-Bodey L, Anderson BG, Argiles JM & Havel PJ 2005 Holst D, Luquet S, Nogueira V, Kristiansen K, Leverve X & Grimaldi PA Interleukin-15 stimulates adiponectin secretion by 3T3-L1 adipo- 2003 Nutritional regulation and role of peroxisome proliferator- cytes: evidence for a skeletal muscle-to-fat signaling pathway. Cell activated receptor delta in fatty acid catabolism in skeletal muscle. Biology International 29 449–457. Biochimica et Biophysica Acta 1633 43–50. Ramakrishnan SN, Lau P, Burke LJ & Muscat GE 2005 Rev-erbb Houstis N, Rosen ED & Lander ES 2006 Reactive oxygen species have a regulates the expression of genes involved in lipid absorption in causal role in multiple forms of insulin resistance. Nature 440 skeletal muscle cells: evidence for cross-talk between orphan 944–948. nuclear receptors and myokines. Journal of Biological Chemistry 280 Jetten AM 2004 Recent advances in the mechanisms of action and 8651–8659. physiological functions of the retinoid-related orphan receptors Raspe E, Madsen L, Lefebvre AM, Leitersdorf I, Gelman L, Peinado- (RORs). Current Drug Targets Inflammation and Allergy 3 395–412. Onsurbe J, Dallongeville J, Fruchart JC, Berge R & Staels B 1999 www.endocrinology-journals.org Journal of Molecular Endocrinology (2007) 39, 29–44

Downloaded from Bioscientifica.com at 10/01/2021 04:14:37PM via free access 44 S RAICHUR and others . RORg and skeletal muscle metabolism

Modulation of rat liver apolipoprotein gene expression and serum Stuart CA, Howell ME & Yin D 2007 Overexpression of GLUT5 lipid levels by tetradecylthioacetic acid (TTA) via PPARa activation. in diabetic muscle is reversed by pioglitazone. Diabetes Care 30 Journal of Lipid Research 40 2099–2110. 925–931. dos Remedios CG, Berry DA, Carter LK, Coumans JV, Heinke MY, Thomas M, Langley B, Berry C, Sharma M, Kirk S, Bass J & Kambadur Kiessling PC, Seeto RK, Thorvaldson T, Trahair T, Yeoh T et al. 1996 R 2000 Myostatin, a negative regulator of muscle growth, functions Different electrophoretic techniques produce conflicting data in by inhibiting myoblast proliferation. Journal of Biological Chemistry the analysis of myocardial samples from dilated cardiomyopathy 275 40235–40243. patients: protein levels do not necessarily reflect mRNA levels. Vidal-Puig AJ, Grujic D, Zhang CY, Hagen T, Boss O, Ido Y, Szczepanik Electrophoresis 17 235–238. A, Wade J, Mootha V, Cortright R et al. 2000 Energy metabolism in Retnakaran R, Flock G & Giguere V 1994 Identification of RVR, a novel uncoupling protein 3 gene knockout mice. Journal of Biological orphan nuclear receptor that acts as a negative transcriptional Chemistry 275 16258–16266. regulator. Molecular Endocrinology 8 1234–1244. Vu-Dac N, Gervois P, Grotzinger T, De Vos P, Schoonjans K, Fruchart JC, Rodriguez-Calvo R, Jove M, Coll T, Camins A, Sanchez RM, Alegret M, Auwerx J, Mariani J, Tedgui A & Staels B 1997 Transcriptional Merlos M, Pallas M, Laguna JC & Vazquez-Carrera M 2006 PGC-1b regulation of apolipoprotein A-I gene expression by the nuclear down-regulation is associated with reduced ERRa activity and receptor RORa. Journal of Biological Chemistry 272 22401–22404. MCAD expression in skeletal muscle of senescence-accelerated Wagner BL, Valledor AF, Shao G, Daige CL, Bischoff ED, Petrowski M, mice. Journals of Gerontology. Series A, Biological Sciences and Medical Jepsen K, Baek SH, Heyman RA, Rosenfeld MG et al. 2003 Sciences 61 773–780. Promoter-specific roles for liver X receptor/corepressor complexes Schoonjans K, Peinado-Onsurbe J, Lefebvre AM, Heyman RA, Briggs M, in the regulation of ABCA1 and SREBP1 gene expression. Molecular Deeb S, Staels B & Auwerx J 1996 PPARa and PPARg activators direct a and Cellular Biology 23 5780–5789. distinct tissue-specific transcriptional response via a PPRE in the Wang YX, Zhang CL, Yu RT, Cho HK, Nelson MC, Bayuga- lipoprotein lipase gene. EMBO Journal 15 5336–5348. Ocampo CR, Ham J, Kang H & Evans RM 2004 Regulation of Schrauwen P & Hesselink M 2002 UCP2 and UCP3 in muscle controlling muscle fiber type and running endurance by PPARdelta. PLoS body metabolism. Journal of Experimental Biology 205 2275–2285. Biology 2 e294. Schrauwen P, Hesselink MK, Blaak EE, Borghouts LB, Schaart G, Saris Xie H, Sadim MS & Sun Z 2005 RORgt recruits steroid receptor WH & Keizer HA 2001 Uncoupling protein 3 content is decreased coactivators to ensure thymocyte survival. Journal of Immunology 175 in skeletal muscle of patients with type 2 diabetes. Diabetes 50 3800–3809. 2870–2873. Yamauchi T, Kamon J, Ito Y, Tsuchida A, Yokomizo T, Kita S, Sugiyama Schreiber SN, Emter R, Hock MB, Knutti D, Cardenas J, Podvinec M, T, Miyagishi M, Hara K, Tsunoda M et al. 2003 Cloning of Oakeley EJ & Kralli A 2004 The estrogen-related receptor alpha adiponectin receptors that mediate antidiabetic metabolic effects. (ERRa) functions in PPARg coactivator 1alpha (PGC-1alpha)- Nature 423 762–769. induced mitochondrial biogenesis. PNAS 101 6472–6477. Stehlin-Gaon C, Willmann D, Zeyer D, Sanglier S, Van Dorsselaer A, Renaud JP, Moras D & Schule R 2003 All-trans retinoic acid is a Received in final form 20 March 2007 ligand for the orphan nuclear receptor RORb. Nature Structural Accepted 28 April 2007 Biology 10 820–825. Made available online as an Accepted Preprint 3 May 2007

Journal of Molecular Endocrinology (2007) 39, 29–44 www.endocrinology-journals.org

Downloaded from Bioscientifica.com at 10/01/2021 04:14:37PM via free access