ROR α regulates the expression of genes involved in lipid homeostasis in skeletal muscle cells: Caveolin-3 and CPT-1 are direct target of ROR.

used to standardize the quantity of DNA to a total of ~ 0.65 µ g DNA per well. Fold activation is expressed relative to luciferase activity obtained after cotransfection of the reporter with the pSG5 vector, alone, arbitrarily set at 1. The mean Luciferase fold activation values and standard deviations (bars) were derived from a minimum of 2-3 independent experiments comprised of six replicates.


INTRODUCTION
Members of the nuclear hormone receptor (NR) superfamily bind specific DNA elements and function as transcriptional regulators (1,2). This group includes the 'orphan NRs' which have no known ligands in the 'classical sense'. The orphan receptor, ROR/RZR (retinoic acid receptor related orphan receptor), is closely related to Rev-erbAα, RVR/Reverbβ/BD73 and the Drosophila orphan receptor, E75A, particularly in the DNA-binding domain (DBD) and the putative ligand-binding domain (LBD). ROR, Rev-erbAα, and RVR bind as monomers to an asymmetric ( A / T ) 6 RGGTCA motif. ROR functions as a constitutive trans-activator of gene expression, in contrast, Rev-e r b A α a n d RVR, do not activate transcription, mediate transcriptional repression, and can repress constitutive trans-activation from this motif by RORα (3)(4)(5)(6)(7)(8)(9).
Decreases in specific lipoprotein compartments namely ApoAI, the major constituent of HDL, and ApoAII leads to a pronounced hypoalphalipoproteinemia (16). Accordingly, it has been shown that RORα regulates the expression of ApoAI and ApoCIII in cell culture (16,17). Susceptibility to atherosclerosis in this animal model is linked to a complex phenotype that includes aberrant vascular physiology, lipid profiles and hyper-sensitive inflammatory responses.
Studies examining motor capabilities and coordination in staggerer mice have shown a reduction of muscular strength in these mice (18). Furthermore, the studies of Lau et. al. (19) demonstrated that RORα potentiates skeletal muscle myogenesis and the comments in a review by Jarvis et. al. (20) suggest that staggerer mice have skeletal muscle irregularities.
The association between RORα, lipid homeostasis and skeletal muscle is not surprising. RORα is abundantly expressed in skeletal muscle, which is one of the most metabolically demanding tissues that relies heavily on fatty acids as an energy source, and accounts for 75% of glucose disposal. Moreover, this suggest that muscle has an important role in obesity, which is primarily controlled by increased food intake or decreased energy expenditure. However, the fundamental role of RORα in skeletal muscle cholesterol, lipid, glucose and energy homeostasis has not been examined. Moreover, the contribution of skeletal muscle, a major mass tissue, to the ROR knockout phenotype has not been resolved.
The involvement of RORα in the progression of atherosclerosis, identifies RORα as a therapeutic target in the treatment of cardiovascular disease. In addition, skeletal muscle is rapidly emerging as a critical target tissue in the battle against obesity, type II diabetes, by guest on July 10, 2020 http://www.jbc.org/ Downloaded from dyslipidemia, syndrome X and atherosclerosis. NRs in skeletal muscle, for example LXR, PPARα, β/δ and γ, have been shown to be involved in enhancing insulin stimulated glucose disposal rate, decreasing triglycerides and increasing lipid catabolism, cholesterol efflux and plasma HDL-C levels (21)(22)(23). Hence, orphan NRs that regulate lipid and glucose metabolism, cholesterol homeostasis, energy expenditure and thermogenesis in skeletal muscle have enormous pharmacological utility for the treatment of dyslipidemia, and syndrome X.
We hypothesize that RORα regulates lipid, glucose and energy homeostasis in this major mass lean tissue. This hypothesis was addressed by examining the effect of ectopically expressing the staggerer form of RORα in muscle cells, and investigating the subsequent effect on gene expression involved in lipid metabolism. Our investigation reveals that ROR has a central role in the regulation of lipid metabolism in muscle cells. Moreover, we suggest that the atherogenic and dyslipidemic phenotype of the staggerer mouse is related to aberrant RORα function in skeletal muscle. mls of DMEM supplemented with 5% charcoal stripped fetal calf serum. 16-24h later, the culture medium was changed, and the cells were subsequently harvested for the assay of Luciferase activity 24-48h after the transfection period as described previously. Fold activation is expressed relative to LUC activity obtained after cotransfection of the reporter, and pSG5 vector only, arbitrarily set at 1. The mean fold activation values and standard deviations (bars) were derived from a minimum of 2 independent experiments comprised of six replicates.
2.6 kb mouse caveolin-3 promoter was isolated from a BAC clone 23N20 (Incyte Genetics) and cloned into HindIII site of pGL3basic. Deletion constructs of the Cav3 promoter were obtained using native restriction sites or by PCR with pfu DNA polymerase. The DNA sequences of these constructs were confirmed by cycle sequencing (ABI). GLUT4-pGL2 was obtained by PCR cloning of a mouse skeletal muscle genomic DNA into pGL2basic. The DNA sequence contains 2.2kb sequence upstream of the transcription start site was confirmed by cycle sequencing. pGEX-RORα1 were previously described (19); GAL4-ROR constructs were cloned in-frame into the multiple cloning site of vector pGAL0 and confirmed by dideoxy sequencing (Pharmacia Uppsala, Sweden).
Site directed mutagenesis: pGL3-Cav3Luc (-2595/+35) was mutated by quickchange PCR site directed mutagenesis (Stratagene) according to manufacturer instruction. 125ng oligonucleotides containing GG to TT mutation and its complementary strand (see sequences below) were added to a PCR reaction containing 50ng pGL3-Cav3Luc (-2595/+35), dNTP, 1x pfu buffer and pfu or pfu ultra. After restriction digestion with dpnI, the reaction was transformed into DH5α. The colonies were screened and sequence confirmed by cycle sequencing after DNA extraction. The oligonucleotides to mutate the mouse caveolin 3

RESULTS:
ROR α and γ mRNA expression are induced during myogenic differentiation RORα and γ mRNAs are expressed in skeletal muscle tissue. To elucidate the functional role of RORα in skeletal muscle we initially investigated the expression of RORα mRNA relative to the 18S rRNA in the mouse C2C12 myoblast cell line. This skeletal muscle cell line has proven to be a reliable system to study skeletal muscle lipid homeostasis in cell culture (21)(22)(23).
Proliferating C2C12 myoblasts can be induced to biochemically and morphologically differentiate into post-mitotic multinucleated myotubes by serum withdrawal in culture over a 48-96 h period. This transition from a non-muscle phenotype to a contractile phenotype is associated with the repression of non-muscle proteins and the activation/expression of a structurally diverse group of genes responsible for contraction and to cope with the extreme metabolic demands on this organ ( Figure 1A) Using the Genbank sequences of ROR α and γ we designed specific primers for the amplification of mouse ROR by quantitative real time PCR from total RNA isolated from proliferating and differentiated C2C12 cells ( Figure 1A). Total polyA + RNA was isolated from proliferating myoblasts (PMB), confluent myoblasts (CMB) and post-mitotic myotubes after 2 and 5 days (MT2 & MT5) of serum withdrawal and examined by real time PCR and 'Northern Blot' analysis.
We observed that RORα and γ are expressed in proliferating myoblasts, however, these transcripts are induced 2.5-4 fold, relative to 18S rRNA and GAPDH mRNA (data not shown), as the cells exit the cell cycle and fuse to form differentiated multinucleated myotubes that have acquired a muscle-specific phenotype ( Fig. 1B and D). RORα is seen transcribed as a double band ( Figure 1C) in C2C12 as described in differentiated P19 cells (3). Concomitant with this increase in ROR encoding mRNAs, was the striking induction of myogenin mRNA (figure 1E), that encodes the hierarchical bHLH regulator, and repression of the cytoskeletal non-muscle β-actin mRNA ( Figure 1F), that confirmed that these cells had terminally differentiated. Moreover, we observed the induction of the slow and fast isoforms of the contractile protein, Troponin I ( Figure 1G  Ectopic ROR∆E expression represses endogenous levels of ROR α and γ, and attenuates ROR dependent gene expression. To understand the biological role of RORα in skeletal muscle lipid homeostasis and to identify the metabolic target(s) of this orphan receptor in muscle cells, we proceeded to examine the effect of attenuating ROR function in C2C12 cells.
In order to perturb ROR function, and disrupt ROR mediated gene expression we utilized the ROR∆DE plasmid. This construct encodes amino acids 1-235, but lacks the entire E region and part of the hinge/D-region of RORα. This was chosen because McBroom et al (30) reported that deletion of this region preserved DNA binding but destroyed trans-activation, and operated in a dominant negative manner. Furthermore, the 'staggerer' phenotype in mice is due to a frameshift mutation in RORα (29), that produces a similar non-functional C-terminal domain. Finally, we noted that GAL4 hybrid analysis rRNA controls. This was demonstrated rigorously by quantitative real time PCR with primers that specifically detect the human specific transcript expressed from the expression vector ( Figure 3A and B) and northern analysis ( Figure 3C). Furthermore, the total pool size of RORα mRNA increased ~2-3 fold when measured by real time PCR with primers that detected both mouse and human transcripts ( Figure 3D). Interestingly, we observed by real time PCR and northern analysis that the endogenous levels of the mouse RORα mRNA transcripts were reduced in both polyclonal pools of the C2-RORα1 (1-235) cells ( Figure 3E and F) relative to the wild type levels. The down regulation of the endogenous RORα transcripts is not surprising. During myogenesis mRNA pool sizes in muscle tissue are under strict control (31), mechanisms exist that sense total output from exogenous and endogenous genes (32). Furthermore, exogenous expression of a number of different contractile protein transgenes in the mouse (eg myosin light chain 2, troponin I fast, skeletal and cardiac actin ) results in the decline in the expression of the corresponding endogenous gene (32)(33)(34)(35)(36).
Finally, we observed that ectopic expression of the RORα∆DE in muscle C2 cells completely ablated RORγ mRNA expression ( Figure 3G).
In conclusion, the cell lines ectopically expresses a dominant negative RORα expression vector, that compromises ROR function, and ROR dependent gene expression.
Moreover, RORα mRNA expression is attenuated, and RORγ mRNA expression is ablated. RORα, and -γ receptors, and that the stable cell line produced by ectopic ROR∆E overexpression has compromised endogenous ROR α/γ expression, and function.
We isolated total RNA from the native (wild type) and ROR∆DE differentiated myotubes, and analysed the expression levels of several myogenic mRNAs by quantitative real time PCR. We demonstrated as previously reported that these cells retain the potential to differentiate (although the initial rate is impaired slightly), exit the cell cycle, and express all genes associated with contraction, and lipid metabolism (data not shown). This suggests that ectopic ROR∆DE expression does not significantly affect proliferation, cell cycle withdrawal, differentiation and/or phenotypic acquisition of these skeletal muscle cells.
Subsequently, we utilized quantitative real-time PCR to investigate the expression Finally, the expression of the closely related, but opposingly acting orphan receptors, Rev-erbα and Rev-erbβ/RVR was further repressed by ROR∆E expression ( Figures   5A and B). This was expected as RORα has been demonstrated to regulate the Rev-erbα gene.

PGC-1 efficiently coactivates the synthetic and native RORE heterologous reporter genes.
We then investigated the ability of ROR to activate the synthetic and wild type-ROR dependent reporter in COS-1 cells, and the ability of the cofactors p300, SRC-2/GRIP-1, and PGC1 to coactivate ROR dependent activation of gene expression. We utilised the synthetic ROREx5-tk-LUC reporter that contains five copies of a consensus binding site (24), and the native mPCP-2x4-tk-LUC containing four copies of the mouse purkinje cell protein-2 RORE motif (25) cloned upstream of the heterologous herpes simplex virus thymidine kinase (tk) promoter linked to the luciferase reporter gene, respectively. As shown RORalpha expression efficiently activated the expression of the synthetic ( Figure 6A) and native RORE ( Figure 6B)-containing reporters in COS-1 cells, several hundred fold. In contrast, RORα relatively insignificantly (i.e. <10-fold) regulated the basal tk-LUCbackbone, lacking the RORE ( Figure 6C).
Furthermore, we observed that RORα dependent activation of the synthetic RORE was significantly enhanced by the coactivator, PGC1, relative to p300 and SRC-2/GRIP-1.
Similarly, ROR dependent activation of the native RORE was most efficiently coactivated by PGC-1, relative to p300 and SRC-2/GRIP-1. In summary, we demonstrate that PGC-1 expression selectively coactivates RORα mediated activation of the heterologous synthetic and native RORE-tk-LUC reporters.

M-CPT1 And Cav-3 Genes Are Primary Targets Of RORα In Skeletal Muscle.
We  Figure 7C).
Interestingly, the p-mCav3-2595-LUC promoter that encompasses 2595 bp immediately upstream of the murine Cav-3 transcription start site (that drives the expression of the muscle-specific caveolin-3 gene) was also efficiently activated by RORα expression ( Figure 8A).
In an attempt to further elucidate which sequences were necessary for the trans- RORα dependent activation ( Figure 8A). Moreover, we observed that p300 and SRC-2/GRIP-1 significantly enhanced the transactivation of the mCav-3 promoter. ( Figure 8B).
To rigorously define the molecular basis of direct ROR action, we focused on one of the promoters that we defined as a primary ROR target gene. We selected and scanned the Subsequently, we examined the ability of the motifs 1-6 to interact with GST-ROR, and observed that motifs 2, 4, 5 and 6 bound ROR with different efficiencies in an EMSA assay, relative to our control ROR recognition motif, the mouse PCP-2 RORE ( Figure 8D).
However, we observed that only the m2, and m6, (and the synthetic RORE), but not the m1, M3-M5 motifs) efficiently competed against the efficient GST-ROR-mPCP-2 RORE interaction ( Figure 8E). This suggested that either motif 2 or 6 mediated ROR activation of the mouse caveolin-3 gene. Hence, we independently mutated the GG nt in the nuclear receptor halfsites to TT in the m2 and m6 putative RORE's within the native caveolin-3 promoter, and examined the ability of ROR (in the presence of the cofactor p300) to transactivate the caveolin-3 promoter. We observed that the mutation of motif 6 in the mouse caveolin-3 promoter compromised the ability of ROR to trans-activate the promoter ( Figure   8F). This was also consistent with the analysis of several additional 5' unidirectional deletion mutant of the caveolin-3 promoter between -2595 and -984 that were examined (data not shown).
In summary, these transfections demonstrated that RORα directly regulates the CPT-1 and Cav3 genes that potentially encode proteins involved in beta-oxidation and cholesterol homeostasis. Moreover, we defined the specific sequences in a muscle-specific gene (i.e. caveolin-3) that mediated the direct regulation of expression by ROR. This suggests that RORα has a central role in the regulation of lipid homeostasis.

DISCUSSION:
Genetic Decreases in specific lipoprotein compartments namely ApoAI, the major constituent of HDL, and ApoAII leads to a pronounced hypoalphalipoproteinemia (16). Accordingly, it has been shown that RORα regulates the expression of ApoAI and ApoCIII in cell culture (16,17). Further studies in staggerer mice, correlate atherogenic susceptibility in this animal  In addition, we demonstrate that M-CPT-1 an established target for PPARα in cardiac muscle (26,48,49) and PPARβ/δ in skeletal muscle cells, is also directly regulated by RORα, and coativated by . p300, and PGC-1. Moreover, it suggests crosstalk may occur between PPAR and ROR α signalling in a tissue specific manner.
The fact that RORα is involved in all of these processes highlights the crucial role

Glycogenin / GYG1
Initiates the synthesis of glycogen, the principal storage form of glucose in skeletal muscle.

LPL
Lipoprotein lipase. Hydrolysis of lipoprotein triglycerides into free fatty acids and responsible for the uptake of free fatty acids.

SCD-1 & -2
Stearoyl CoA desaturase-1, & -2. Enzymes associated with adiposity, i.e. storage and esterification of cholesterol, and responsible for the cis saturation of stearoyl and palmitoyl-CoA converting them to oleate and palmitoleate, which are the monounsaturated fatty acids of triglycerides.

UCP-1, -2 & -3
Uncoupling proteins. Mitochondrial proteins that uncouple metabolic fuel-oxidation from ATP-synthesis, regulating energy expenditure.        TT below the line denotes mutagenesis of the bases above from GG to TT by Quickchange site directed mutagenesis referred to in Figure 8F.