The Co-repressor Hairless Protects RORα Orphan Nuclear Receptor from Proteasome-mediated Degradation*

RORα is a constitutively active orphan nuclear receptor essential for cerebellar development and is previously shown to regulate genes involved in both myogenesis and adipogenesis. The transcriptional activity of RORα is dependent on the presence of a ubiquitous ligand and can be abolished by interaction with Hairless (Hr), a ligand-oblivious nuclear receptor co-repressor. In this study, we first demonstrate that RORα is a short-lived protein and that treatment with the MG-132 proteasome inhibitor results in the accumulation of ubiquitin-conjugated receptor and inhibition of transcription. These data show that RORα transcriptional activity and degradation are intrinsically linked. In addition, the introduction of inactivation mutations in the ligand-binding pocket and co-regulator-binding surface of RORα significantly increases protein stability, indicating that ligand and/or co-regulator binding perpetuates RORα degradation. Strikingly, expression of the co-repressor Hr results in the stabilization of RORα because of an inhibition of proteasome-mediated degradation of the receptor. Stabilization of RORα by Hr requires intact nuclear receptor recognition LXXLL motifs within Hr. Interestingly, the co-repressor nuclear receptor co-repressor (NCoR) has no effect on RORα protein turnover. This study shows that stabilization of RORα is an essential component of Hr-mediated repression and suggests a molecular mechanism to achieve transcriptional repression by a liganded receptor-co-repressor complex.

The ubiquitin-proteasome pathway is the major system employed by eukaryotes for the selective degradation of cellular proteins that play key roles in cellular processes such as cell cycle regulation, differentiation, signal transduction, transcription, and chromosomal stabilization (reviewed in Refs. 1 and 2). Proteolytic degradation by the ubiquitin-proteasome system involves ATP-dependent covalent attachment of a macromolecular chain of ubiquitin (Ub) 1 molecules to the target protein, followed by degradation through the multicatalytic 26 S pro-teasome. The conjugation of Ub, a highly conserved 8.6-kDa protein, to its target protein is mediated by the serial action of three enzymes: E1, the Ub-activating enzyme, activates Ub in an ATP-dependent manner; E2, the Ub-conjugating enzyme, catalyzes the attachment of Ub to the substrate protein; and E3, the Ub-ligases, serves as a scaffold between E2 and the substrate and provides recognition specificity of the substrate. Ubiquitinylation of a substrate is reversible, and Ub moieties can be cleaved from a target protein by deubiquitinating enzymes. These enzymes assure that the cell is not depleted of a Ub pool. A protein tagged with a polyubiquitin chain is recognized and degraded by the 26 S proteasome complex. This complex is composed of a 19 S regulatory subcomplex, consisting of a "lid" subunit and a "base" subunit, the latter containing the six ATPases required for the degradation executed by the 20 S catalytic subcomplex (2).
The Ub-proteasome pathway has recently emerged as a key regulator of transcription controlling the level, location, and activity of transcription factors and associated co-factors (3). Nuclear receptors are short-lived transcription factors whose turnover is mediated by the Ub-proteasome complex (reviewed in Ref. 4). A number of nuclear receptors, including the estrogen, progesterone, glucocorticoid, retinoic acid, and thyroid hormone receptors, as well as peroxisome proliferator-activated receptor ␥, are degraded in a ligand-dependent fashion (5)(6)(7)(8)(9)(10)(11). Degradation of the vitamin D and pregnane X receptors reportedly occurs in a ligand-independent fashion, signaled by unstable interactions of the receptors with heat shock proteins (12,13). In addition to ligand binding, phosphorylation of nuclear receptors by signal transduction pathways and coregulator binding also serve as signals to the Ub-proteasome complex, targeting the receptor for degradation (9,14). Corepressors such as NCoR and co-activators such as members of the steroid receptor co-activator (SRC) family and CREB-binding protein are also substrates for proteasome-mediated degradation (5,15,16).
The ligand influences the stability of nuclear receptors by inducing a conformational change that permits co-factor docking. A number of these co-factors have been identified as Ubproteasome or Ub-like pathway enzymes, with a role in both proteasomal degradation and transcriptional activation. The E3 Ub-ligases RSP5/RPF1 and the E6-associated protein and the ATPase subunit of the 26 S proteasome SUG1 (suppressor of Gal4) all participate in nuclear receptor transactivation while simultaneously mediating their degradation (17)(18)(19)(20)(21)(22). Recently, p300/CREB-binding protein has been shown to mediate the polyubiquitination of the p53 transcription factor through its intrinsic Ub-ligase activity in addition to its function as a transcriptional co-activator (23). This dual action of p300/ CREB-binding protein adds strong evidence that the Ub-proteasome pathway plays a direct regulatory role in transcription. Indication that transcriptional activation and protein degradation occur concomitantly is further supported by the loss of nuclear receptor-mediated transcriptional activation observed upon inhibition of the 26 S proteasome function (5,24). This observation suggests that the Ub-proteasomal complex is integral to nuclear receptor-mediated transcription. Down-regulation of an activation complex may be required for the exchange of co-activator complexes leading to disruption of the preinitiation complex, thereby allowing transcriptional elongation to proceed. The cell can then recycle components of the activation complex necessary for the initiation of a second round of transcription. This pathway provides a means of preventing the overstimulation by hormone (25).
ROR␣ (retinoic acid-related orphan receptor ␣; NR1F1) is a member of the ROR subfamily, which also includes ROR␤ (NR1F2) and ROR␥ (NR1F3), each regulating diverse physiological processes (26). Genetic ablation of the rora gene mimics the staggerer phenotype that is caused by massive neurodegeneration of Purkinje cells in the cerebellum (27)(28)(29). ROR␣ knock-out and staggerer mice also serve as a model for agerelated degenerative pathologies because they exhibit greater susceptibility to atherosclerosis, immunodeficiencies linked to an overexpression of inflammatory cytokines, abnormal formation and maintenance of bone, and changes in muscle differentiation (reviewed in Ref. 30). ROR␣ is a potent transcriptional activator even in the absence of exogenously added ligand. However, the recent resolution of the crystal structures of ROR␣ and ROR␤ ligand-binding domains (LBDs), in combination with mutagenesis assays of the ROR␣ LBD, suggest that members of the ROR family require that their LBD be occupied by a ligand for transcriptional activation to occur (31)(32)(33). Regulation of ROR␣ transcriptional activity is also mediated through co-regulator recruitment. ROR␣ has been shown to interact with members of the SRC family and the p300/CREBbinding protein co-integrators (32,34,35). Repression of ROR␣ activity can be achieved by displacement from its binding site by the transcriptionally inactive orphan nuclear receptors Re-vErbA␣ and RVR (36 -38) and possibly by an active mechanism through interaction with the co-repressors NCoR and SMRT, although the effect of NCoR expression on ROR␣ activity is negligible (39). However, more recently, ROR␣ transcriptional activity was shown to be completely abrogated by the co-repressor Hairless (Hr) (32). Hr is a nuclear protein that also mediates transrepression by thyroid hormone receptor and vitamin D receptor through the recruitment of histone deacetylases (40,41). In contrast to the ubiquitous expression pattern and promiscuity of NCoR and SMRT, Hr expression is restricted to the skin and brain and is nuclear receptor-selective.
We demonstrate in this study that the Ub-proteasome pathway regulates ROR␣-mediated gene transcription. We observed that the ROR␣ protein expression level increases upon inhibition of the 26 S proteasome complex with the MG-132 peptide aldehyde. ROR␣ degradation occurs following Ub conjugation, likely signaled by recruitment of co-factors to the activated receptor. Strikingly, we show that the Hr co-repressor protects ROR␣ from degradation, a mechanism not shared by the co-repressor NCoR. Proteasomal inhibition is thus unfavorable to ROR␣ transcriptional activity. Taken together, these results show that protection of ROR␣ from the activity of the Ub-proteasome pathway by Hr is an integral component of ROR␣-regulated transcription.
Cell Culture and Transient Transfection-Cos-1 cells obtained from the American Type Culture Collection were cultured in Dulbecco's minimal essential medium containing penicillin (25 units/ml), streptomycin (25 units/ml), and 10% fetal calf serum at 37°C with 5% CO 2 . Twenty-four hours prior to transfection, the cells were split and seeded in 12-well plates. The cells were transfected with FuGENE 6 Transfection Reagent (Roche Applied Science), following the protocol supplied by the manufacturer. A total of 1 g of DNA/well was transfected including 0.05 g of pCMX-hROR␣1 or mutant derivatives, 0.5 g of pCMX-hNCoR or pRK5-Myc-rHR, 0.5 g of reporter plasmid, and 0.25 g of internal control pCMV␤Gal. The cells were treated with ethanol (vehicle) or 0.1, 0.5, or 1.0 M MG-132 for 6 -24 h, as specified in the figure legends. The cells were harvested and assayed for luciferase and ␤-galactosidase activities. The normalized values are expressed in terms of relative luciferase units. The error bars represent the standard deviations between duplicate samples. Each graph is one representative experiment of a total of three independent experiments.
Co-immunoprecipitation and Immunoblotting Assays-Cos-1 cells in 10-cm dishes were transiently transfected as described above with 10 g of FLAG-ROR␣ and HA-Ub and treated with ethanol (vehicle) or 1 M MG-132 for 24 h. The cells were lysed in IP buffer (1% Nonidet P-40, 10% glycerol, 150 mM NaCl, 50 mM Tris-HCl, pH 7.5) supplemented with protease inhibitor mixture (Complete Mini EDTA-free; Roche Applied Science). The lysates (containing a total of 250 g of protein) were incubated with 5 g of FLAG antibody (Sigma) overnight at 4°C with gentle rotation. The proteins were collected on protein G-Sepharose for 2 h at 4°C with mild rotation and then washed three times with ice-cold low salt buffer (1% Nonidet P-40, 50 mM Tris-HCl, pH 8.0). The immunoprecipitates were resolved by SDS-PAGE, transferred to a hydrophobic polyvinylidene difluoride membrane (Amersham Biosciences), and immunoblotted with FLAG antibody or HA antibody (HA.11; Berkeley Antibody Company). The proteins were visualized with the POD chemiluminescence kit following the manufacturer's instructions (Roche Applied Science). Immunoblotting for detection of ROR␣ wild type and mutants, Myc-Hr, SRC-1, hNCoR, or actin was similarly done using anti-ROR␣ antibody (C-16; Santa Cruz Biotechnology), anti-c-Myc antibody (Roche Applied Science), anti-SRC-1 antibody (M-341; Santa Cruz Biotechnology), anti-hNCoR antibody (H-303; Santa Cruz Biotechnology), and anti-actin antibody (I-19; Santa Cruz Biotechnology), respectively. The lysates were prepared from transiently transfected Cos-1 harvested in modified RIPA buffer (50 mM Tris-HCl, pH 7.4, 1% Nonidet P-40, 0.25% sodium deoxycholate, 150 mM NaCl, 1 mM EDTA, 1 mM phenylmethylsulfonyl fluoride, 1 g/ml leupeptin, 1 mM Na 3 VO 4 , 1 mM NaF), resolved by SDS-PAGE, transferred, and immunoblotted as described above.
In Vitro Degradation Assay-The cell extract was prepared from Cos-1 cells harvested in modified RIPA buffer. 5 l of in vitro translated [ 35 S]methionine-labeled ROR␣, using TNT rabbit reticulocyte lysate (Promega, Madison, WI), was incubated with 50 g of cell extract, ethanol (vehicle), or 50 M MG-132, 20 M lactacystin, 50 g/ml expressed sequence tag, 2 mM phenylmethylsulfonyl fluoride in a final volume of 50 l of degradation buffer (20 mM Tris-HCl, pH 7.4, 50 mM NaCl, 0.2 mM dithiothreitol) for 2 h at 37°C. The samples were resolved by SDS-PAGE. The gels were fixed, treated with the fluorographic reagent Amplify (Amersham Biosciences), dried, and exposed.
Pulse-Chase Assay-Cos-1 cells in 10-cm dishes were transiently transfected as described above, with 5 g of pCMX-hROR␣1 and pRK5-Myc-rHr or pCMX-hSRC-1 for 24 h as specified in figure legends. The cells were washed carefully with 1ϫ phosphate-buffered saline, and the medium was replaced with Dulbecco's modified Eagle's medium (Ϫmethionine/Ϫcysteine) for 2 h, followed by the addition of [ 35 S]methionine (100 Ci/ml) for an additional 1 h. The cells were washed with 1ϫ phosphate-buffered saline and chased with Dulbecco's modified Eagle's medium for the times indicated in the figure legends. The cells were harvested and lysed in IP buffer. 400 g of lysate was immunoprecipitated with 5 l of anti-ROR␣ antibody (C-16; Santa Cruz Biotechnology) for 30 min at 4°C with gentle rotation, followed by incubation with 50% slurry of protein G-Sepharose for an additional 30 min. The beads were then washed with low salt buffer (1% Nonidet P-40, 50 mM Tris-HCl, pH 8.0), followed by a wash in high salt buffer (500 mM NaCl, 1% Nonidet P-40, 50 mM Tris-HCl, pH 8.0), and resuspended in 2ϫ SDS sample buffer. The samples were boiled for 3 min and resolved by SDS-PAGE. The gels were fixed and treated with fluorographic reagent, dried, and exposed. Quantification was performed using the Typhoon 8600 Phos-phorImager (Amersham Biosciences).

ROR␣, a Target of the Ub-Proteasome Complex-Nuclear
receptors are short-lived proteins that are rapidly turned over by the Ub-proteasome complex. Using pulse-chase analysis, we first determined that the half-life of ROR␣ in transiently transfected Cos-1 cells is ϳ1.3 h (Fig. 1A). This rapid turnover is reminiscent of liganded nuclear receptors, which have a shorter half-life than their unliganded counterparts. We next investigated whether ROR␣ degradation is mediated by the Ub-proteasome complex using pharmacological inhibitors. Peptide aldehydes (MG-132) or natural products (lactacystin) act as pseudosubstrates that become covalently linked to the 26 S proteasome and inactivate its chymotryptic and tryptic-like activities (45). As shown in Fig. 1B, ROR␣ is not expressed endogenously in Cos-1 cells, although ROR␣ expressed through transient transfection is detected by immunoblotting with anti-ROR␣ antibody. Blocking of the 26 S proteosome with MG-132 leads to a substantial increase of ROR␣ protein, suggesting that ROR␣ is a likely substrate of the Ub-proteasome complex (Fig. 1B). We next used Cos-1 extract as a source of Ub-proteasome complex in an in vitro degradation assay to determine whether in vitro translated and labeled ROR␣ is proteolytically degraded. Loss of labeled ROR␣ was observed upon incubation with Cos-1 extracts, a process that was not observed with extracts obtained from Cos-1 cells previously treated with the inhibitor MG-132, indicating that ROR␣ is degraded by the 26 S proteasome in vitro ( Fig. 2A). In addition to MG-132, treatment with lactacystin, an irreversible specific inhibitor of the 20 S proteasome, also blocked ROR␣ degradation as demonstrated by a marked increase in protein expression in comparison with the control sample (Fig. 2B). In contrast, the lysosomal-specific cysteine protease inhibitor expressed sequence tag, as well as the nonspecific serine protease inhibitor phenylmethylsulfonyl fluoride, failed to stabilize ROR␣ protein levels, providing further evidence that ROR␣ is specifically degraded by the 26 S proteasome. Substrates destined for proteasomal degradation are tagged by covalent attachment of a macromolecular Ub chain. Co-immunoprecipitation of ROR␣ and Ub resulted in the appearance of high molecular weight Ub-conjugated ROR␣ complexes in cells treated with MG-132 (Fig. 2C). Given the absence of Ub-ROR␣ complexes in untreated cells, Ub-tagged ROR␣ is likely rapidly degraded by the 26 S proteasome under normal conditions (Fig. 2C).
A Putative PEST Motif Is Not Involved in Degradation-Proteins targeted for degradation by the Ub-proteasome complex often contain a short hydrophilic stretch of at least 12 amino acids termed a PEST motif. A PEST region serves as a proteolytic signal leading to rapid destruction of the protein (46,47). We first used a PESTfind program (at.embnet.org/ embnet/tools/bio/PESTfind/about.htm) to identify putative PEST sequences in ROR␣. This algorithmic program scores the hydrophilicity in a range of Ϫ50 to ϩ50, and scores above ϩ5.0 are considered more probable PEST motif candidates. A putative PEST motif with a score of ϩ6.88 was located in the N-terminal region of the protein (Fig. 3A). To determine the involvement of this putative PEST motif in signaling ROR␣ degradation, we tested three N-terminal deletion mutants referred to as ROR␣⌬12, ROR␣⌬25, and ROR␣⌬35 (Fig. 3B). These constructs were transiently transfected in Cos-1 cells, and their transactivation potential was assessed on a ROREdriven reporter. All deletion proteins displayed potent transcriptional activity, which was inhibited by treatment of the cells with the MG-132 (Fig. 3B). Given that protein expression of ROR␣⌬25 is stabilized upon treatment with MG-132 ( Fig.   FIG. 1. ROR␣ protein is rapidly turned over. A, a pulse- Proteasomal Degradation Is an Integral Part of ROR␣ Transactivation Potential-Ub-mediated degradation of nuclear receptors and other transcription factors is tightly coupled to their transactivation potential, providing the cell with a mechanism that protects it against possible deleterious prolonged periods of transcription at specific genes. In particular, it has been demonstrated that this pathway is imperative for a functional hormone-mediated transcriptional response of the estrogen receptor (5,24). However, it is not known whether this finding could be extended to orphan nuclear receptors such as ROR␣ that display potent and constitutive transcriptional activity. As shown in Fig. 4A, treatment of Cos-1 cells transiently transfected with ROR␣ and a RORE-driven reporter with increasing concentrations of MG-132 results in a progressive inhibition of ROR␣-mediated transactivation. Blocking of the 26 S proteasome most likely leads to the accumulation of transcriptionally inactive Ub-conjugated ROR␣. These data show that a functional Ub-proteasome pathway is critical for efficient transcriptional activation by ROR␣. We next investigated whether the presence of a ligand could influence ROR␣ degradation. As mentioned above, recent crystallographic and mutational studies have shown that the presence of a ligand, possibly cholesterol or a close derivative trapped within the ROR␣ LBD, is essential for its transcriptional activity (32,33). However, the harsh methodology used to manipulate cholesterol levels in the cells necessary to observe a cholesterol-mediated regulation of ROR␣ activity could indirectly affect the Ubproteasome pathway (33). We have therefore used the LBD mutants ROR␣ V364G and ROR␣ L361F to simulate unliganded and liganded receptor conditions, respectively (32). The valine residue at position 364 lines the ligand-binding pocket and is in close proximity to the bound cholesterol molecule, whereas the leucine residue at position 361 is not within the vicinity of the putative ligand. Transiently transfected Cos-1 cells were treated with the proteasome inhibitor MG-132, and the transactivation potential as well as protein expression levels of both wild type and mutant ROR␣ were assayed. As expected, ROR␣ V364G displays considerably reduced transcriptional activation, whereas ROR␣ L361F is functional and activates transcription from the reporter gene (Fig. 4B). Interestingly, ROR␣ V364G expression is greater than that of wild type ROR␣ in the absence of the proteasomal inhibitor, demonstrating that the transcriptional inactivity is independent of protein expression levels of this mutant (Fig. 4C). ROR␣ L361F expression is similar to that of wild type ROR␣, indicating that this mutation does not affect the function of the receptor (Fig. 4C). Treatment of cells with MG-132 increases the expression levels of both wild type ROR␣ and ROR␣ L361F , whereas no significant effect can be observed on ROR␣ V364G expression (Fig. 4C). These data suggest that transcriptionally inactive ROR␣ V364G is not degraded by the Ub-proteasome complex and indicates that ligand binding is required for rapid degradation of the receptor. We and others have demonstrated that ROR␣ transcriptional activity is dependent on the integrity of a functional co-activator binding surface (32,34). To assess the importance of this interface in proteasomal-mediated ROR␣ degradation, we tested the expression levels of an AF-2-deficient mutant (ROR␣ E509K ), as well as a functional hydrophobic cleft mutant (ROR␣ K357A ) as a positive control. As shown in Fig. 4B, the ROR␣ K357A mutant exhibits constitutive transcriptional activity, whereas ROR␣ E509K is transcriptionally inactive. We have previously shown that their transactivation potentials correlate to their ability to interact with SRC co-activators (32). In a manner analogous to wild type ROR␣, ROR␣ K357A degradation is blocked by MG-132, and its transcriptional activity is inhibited (Fig. 4B). The transcriptionally inactive AF-2-deficient mutant ROR␣ E509K exhibits a higher protein expression level than wild type ROR␣, whereby treatment with MG-132 has no effect. This suggests that the proteasomal-mediated degradation of ROR␣ requires an intact AF-2 domain and the concomitant recruitment of co-factor proteins.
The Co-repressor Hr Protects ROR␣ from Proteasome-mediated Degradation-The results presented above show that proteasomal degradation of ROR␣ is closely linked to its activation state and therefore provides a mechanism regulating ROR␣mediated transcription. Evidence that the mechanisms for destroying active nuclear receptors involved direct participation of co-activator proteins is also accumulating. However, there is little evidence for the involvement of co-repressor proteins in this process. Recently, we have identified Hr as a potent repressor of ROR␣ transcriptional activity (32).We therefore investigated whether co-expression of Hr and mutant derivatives (Fig. 5A) affects ROR␣ protein stability. As shown in Fig. 5B, co-expression of increasing amounts of Hr together with ROR␣ in Cos-1 cells leads to a 2-fold increase in ROR␣ stability. We next tested whether direct interaction between Hr and ROR␣ was required for protein stabilization. We had previously demonstrated that Hr possesses two functional nuclear receptor recognition LXXLL motifs and that ablation of both motifs was necessary to abrogate Hr/ROR␣ interaction (32). As shown in Fig. 5C, mutation of an individual Hr LXXLL motif (Hr m1 and Hr m2 ) does not significantly hinder stabilization of ROR␣ by Hr (Fig 5C, compare lanes 3 and 4 with lane 1), whereas mutation of both motifs (Hr m3 ) leads to a complete loss of stabilization, even in the presence of increasing amount of the mutant protein (Fig. 5D). These data indicate that Hr protects ROR␣ from degradation as a result of direct interaction. As further controls for these experiments, we tested the effects of co-expression of NCoR and SRC-1 on ROR␣ stability. NCoR expression had no effect on ROR␣ protein stability (Fig. 5C, lane 6), whereas expression of SRC-1 led to a decrease of ROR␣ protein detected (Fig. 5C, compare lanes 1 and 7 in top panel). The effect or lack thereof of the expression of both repressor proteins on ROR␣ stability correlates with their respective effect on transcriptional activity (Fig. 5E). Interestingly, the SRC-1-mediated destabilization of ROR␣ is blocked by MG-132 (Fig. 5C, lane 7,  second panel). Finally, using pulse-chase analyses, we demonstrate that Hr stabilizes ROR␣ and increases the receptor half-life from ϳ1.3 to ϳ2.8 h (Fig. 5F), whereas SRC-1 does not significantly affect ROR␣ half-life (Fig. 5G). These data suggest that the potent Hr repression of ROR␣-mediated transcriptional activity may result as a consequence of stabilized ROR␣-Hr complex.

DISCUSSION
The Ub-proteasome system is involved in regulating the turnover of many transcription factors, including members of the nuclear receptor superfamily (4). Transcription factor activation and destruction are closely linked, providing the cell with an efficient mechanism for attenuating transcription (48). The more potently a given transcription factor activates transcription, the more rapidly it is Ub-tagged and degraded. An inverse correlation has been established between the strength of an activation domain and the protein half-life (49 -51). Given that the ROR␣ orphan nuclear receptor is a strong transcriptional activator, we investigated whether its potent activation domain is involved in the down-regulation of the receptor by signaling to the Ub-proteasome complex. In this study, we show that ROR␣ is Ub-conjugated and rapidly degraded by the Ub-proteasome pathway with a half-life of only 1.3 h. Treatment of cells with MG-132, a pseudosubstrate that inhibits the catalytic function of the 26 S proteasome, results in a marked increase of ROR␣ protein levels both in vivo and in vitro. Interestingly, blocking the Ub-proteasome pathway also impairs ROR␣ transcriptional activity, indicating that degradation is an integral part of ROR␣-mediated transcription.
Recognition of target proteins by the Ub-proteasome complex is mediated through specific motifs that signal to E3 Ub ligase enzymes that a given protein is to be tagged with a polyubiquitin chain. Many rapidly degraded regulatory proteins contain PEST motifs, regions rich in proline, glutamic acid, serine, and threonine residues. The glucocorticoid receptor encodes a PEST motif, essential for ligand-mediated degradation, because point mutation of this motif abrogates down-regulation (9). ROR␣ encodes a putative PEST sequence in its N-terminal domain. We have demonstrated that this motif is not required for proteasomal degradation of this protein. Similarly, RXR encodes putative PEST motifs in the N-terminal and hinge domains, although mutation of these motifs does not affect proteolytic degradation of this receptor (6). The absence of a functional PEST motif is not uncommon given that a number of receptors, including estrogen receptor ␣ and thyroid hormone receptor, are down-regulated by the Ub-proteasome complex despite the lack of this consensus signaling motif (5,7). Proteolysis may also be triggered by phosphorylation and/or recruitment of co-factors. For example, ligand-dependent degradation of retinoic acid receptor ␥ is dependent on its phosphorylation and dimerization states (14). Ligand-dependent glucocorticoid receptor degradation is also signaled by phosphorylation, because a phosphorylation-deficient mutant does not undergo proteolysis (9). It is currently unknown whether ROR␣ serves as a substrate for specific kinases and whether such post-translational modification plays a role in ROR␣ stabilization.
To date, studies of nuclear receptor degradation by the Ubproteasome complex had been limited to receptors that are regulated by a known ligand (5,7,9,10,12,52). The ligand plays a key role in mediating substrate recognition by inducing the transconformation of the LBD that allows docking of proteins involved in the Ub-proteasome pathway. One caveat in the study of orphan nuclear receptors is the absence of a bona fide ligand. Although cholesterol or a cholesterol derivative has recently been suggested as a physiological ROR␣ ligand (33), we found it difficult to manipulate endogenous cellular cholesterol levels without affecting the general state of the cells, making it impractical to use such ligands to study specific protein degradation. Mutagenesis assays have therefore been instrumental to the understanding of the mechanisms involved in ROR␣ transcriptional activity. Mutations in the ligand-binding pocket render the receptor transcriptionally inactive, providing additional evidence that ROR␣ activity is regulated by an endogenous ubiquitous ligand (32). The ligand-binding pocket mutant ROR␣ V364G provides a means of mimicking unliganded receptor conditions. Strikingly, ROR␣ V364G exhibits greater protein expression than wild type ROR␣ and is unaffected by inhibition of the 26 S proteasome complex, showing that proteolytic degradation requires an intact ligand-binding pocket and is ligand-dependent. The AF-2-deficient mutant, ROR␣ E509K , is highly expressed irrespective of treatment with the proteasome inhibitor MG-132, suggesting that only a transcriptionally active receptor undergoes rapid proteasomal degradation. Moreover, these results also demonstrated that an intact co-activator-binding surface is required for proteolytic degradation. Given that both activation and degradation are regulated by ligand, binding of a putative ROR␣ ligand may not only recruit co-activator proteins necessary for transcriptional activation but may also recruit proteins of the Ub-proteasome complex. This is the first demonstration of an orphan nuclear receptor that is degraded by the Ub-proteasome pathway.
Recently, a number of Ub-proteasome and Ub-like pathway enzymes have been shown to be nuclear receptor co-activators. The E3 Ub ligases E6-associated protein and RPF1/RSP5 play a dual role as nuclear receptor co-activators and may be integral components of the RNA polymerase II machinery (17)(18)(19). FIG. 5. Hr co-repressor blocks proteasome-mediated ROR␣ degradation. A, schematic representation of the co-repressor tagged with the Myc epitope (gray box). Hr encodes two LXXLL motifs as indicated by black boxes. The LXXLL motifs were mutated individually (Hr m1 or Hr m2 ) or in combination (Hr m3 ). B, Cos-1 cells were transiently transfected with ROR␣ and increasing amounts of Hr. Western blot analysis of ROR␣, Hr, and actin, as a loading control, was performed using anti-ROR␣, anti-Myc, and anti-actin antibodies, respectively. C, Cos-1 cells were transiently transfected with ROR␣ and Hr, Hr m1 , Hr m2 , Hr m3 , NCoR, or SRC-1 and treated with ethanol (Ϫ) or 10 M MG-132 (ϩ) for 6 h prior to harvesting. Western blot analysis of ROR␣, Hr, NCoR, SRC-1, and actin, as a loading control, was performed using anti-ROR␣, anti-Myc, anti-NCoR, anti-SRC-1, and anti-actin antibodies, respectively. D, Cos-1 cells were transiently transfected with ROR␣ and increasing amounts of Hr m3 and analyzed as described for B. E, Cos-1 cells were transiently transfected with NCoR and Hr in the absence (open columns) or presence of ROR␣ (closed columns), and transcriptional activity on a RORE␣2 3 -TkLuc reporter was assayed. The normalized values are expressed in relative luciferase units (RLU). F and G, Cos-1 cells were transiently transfected with ROR␣ and Hr (F) or SRC-1 (G) expression plasmids, and pulse-chase analysis was used to determine the half-life of ROR␣ in the presence of Hr and SRC-1, respectively. Representative experiments are shown in the insets of F and G.
RNA polymerase II recruits E3 Ub ligases through phosphorylation of its C-terminal domain. In addition, the ATPase enzymes of the 19 S proteasome subcomplex, namely SUG1 and SUG2, have recently been shown to associate with actively transcribed genes (53,54). SUG1 binds directly to the activation domains of GAL4 and other transcription factors and functions as a nuclear receptor co-activator (20 -22, 55, 56). ROR␣ has been shown to recruit SUG1 to its LBD in a yeast twohybrid assay, although it is not yet known whether this putative co-activator can potentiate ROR␣ transcriptional activity or recruit the Ub-proteasome complex for ROR␣ degradation (34). The involvement of SUG1 in the regulation of ROR␣ transcriptional activity and stability warrants further investigation. Recently, p300 was shown to contain intrinsic E3 Ubligase activity catalyzing the ubiquitination of the p53 transcription factor (23). Given that ROR␣ recruits p300 via its LBD, p300 may, in addition to functioning as a co-activator, participate in the ubiquitinylation and subsequent degradation of this orphan receptor (35).
Interestingly, we have also demonstrated that expression of the AF-2-dependent Hr co-repressor leads to the stabilization of ROR␣ and a prolonged half-life of the protein. Mutation of the LXXLL nuclear receptor recognition motifs that are involved in mediating Hr-ROR␣ interaction leads not only to an inability of Hr to repress ROR␣ transcriptional activity as previously demonstrated (32) but to the loss of the protective effect against ROR␣ degradation. This suggests that to effectively repress transcription, the co-repressor-nuclear receptor must stably exist on the promoter for a prolonged period. Similarly, Sin3 protects the p53 transcription factor from proteasome-mediated degradation, thus increasing its efficacy as a repressor (57). Moreover, the E2F-1 and c-Myc transcription factors are also stabilized by their transcriptional repression partners, pRB and miz-1, respectively (58,59). Co-repressormediated stabilization of a transcription factor is therefore a mechanism that extends beyond Hr-ROR␣ and may be a general requirement for efficient transcriptional repression. However, despite previous demonstration of interaction between ROR␣ and NCoR in vitro (39), the current study has shown that co-repressor-mediated stabilization of ROR␣ does not extend to NCoR. It thus remains to be determined whether the co-repressor NCoR can participate in the stabilization of other classes of nuclear receptors or whether NCoR utilizes other mechanisms to achieve its repressive function.
In conclusion, we have shown that expression of the corepressor Hr results in stabilization of ROR␣ caused by an inhibition of proteasome-mediated degradation of the receptor. It thus appears that by protecting ROR␣ from degradation, Hr may be enhancing the efficacy of the Hr-ROR␣ complex as a transrepressor, either by prolonging the action of associated histone deacetylases or preventing receptor turnover to coactivator-bound receptor. The effective interference of Hr on the interdependency of ROR␣-mediated transcription and Ubproteasome degradation may explain its potent repressive action on ROR␣ transcriptional function, a mechanism rendered essential for shutting off a powerful and constitutively active nuclear receptor.