Peroxisome Proliferator-activated Receptor α (PPARα) Turnover by the Ubiquitin-Proteasome System Controls the Ligand-induced Expression Level of Its Target Genes*

Peroxisome proliferator activated-receptor α (PPARα) is a ligand-activated transcription factor belonging to the nuclear receptor family. PPARα is implicated in the regulation of lipid and glucose metabolism and in the control of inflammatory response. Recently, it has been demonstrated that a number of nuclear receptors are degraded by the ubiquitin-proteasome pathway. Since PPARα exhibits a circadian expression rhythm and since PPARα is rapidly regulated under certain pathophysiological conditions such as the acute phase inflammatory response, we hypothesized that PPARα protein levels must be under tight control. Here, we studied the mechanisms controlling PPARα protein levels and their consequences on the transcriptional control of PPARα target genes. Using pulse-chase experiments, it is shown that PPARα is a short-lived protein and that addition of its ligands stabilizes this nuclear receptor. By transient cotransfection experiments using expression vectors for PPARα and hemagglutinin-tagged ubiquitin, it is demonstrated that PPARα protein is ubiquitinated and that its ligands decrease the ubiquitination of this nuclear receptor, thus providing a mechanism for the ligand-dependent stabilization observed in pulse-chase experiments. In addition, treatment with MG132, a selective proteasome inhibitor, increases the level of ubiquitinated PPARα and inhibits its degradation in transfected cells. Furthermore, MG132 treatment enhances the level of endogenous PPARα in HepG2 cells. Finally, transient transfection and quantitative reverse transcription-PCR show that inhibition of PPARα degradation increases its transcriptional activation and expression of target genes such as apoA-II and fatty acid transport protein (FATP). Taken together, these data demonstrate that PPARα is degraded by the ubiquitin-proteasome system in a ligand-dependent manner. Regulation of its degradation provides a novel regulatory mechanism of transcriptional activity of this nuclear receptor.

The peroxisome proliferator-activated receptors (PPARs) 1 are members of the nuclear receptor superfamily that act as ligand-dependent transcription factors. PPAR␣ is highly expressed in liver, skeletal and cardiac muscle, and proximal tubular epithelium of kidney. A significant expression of PPAR␣ has also been shown in endothelial cells, smooth muscle cells, and cells involved in the inflammatory process (1). The ligands of PPAR␣ are natural fatty acids and derivatives such as leukotriene B4 and 8-S-hydroxyeicosatetraenoic acid or oxidized phospholipids from oxidized low density lipoprotein. The fibrates hypolipidemic drugs are synthetic PPAR␣ ligands (1). PPAR␣ plays a role in intracellular fatty acid metabolism and in triglyceride metabolism by regulating genes involved in the transport and degradation of fatty acids in mitochondria and peroxisomes (2). PPAR␣ is also implicated in the metabolism of lipids and lipoproteins. As a result, PPAR␣ activation decreases the hepatic very low density lipoprotein secretion and plasma triglyceride levels (3). Furthermore, PPAR␣ agonists increase plasma concentration of high density lipoprotein particles by regulating the transcription of the major high density lipoprotein apolipoproteins (apo) in liver including apoA-II (4). PPAR␣ regulates gene expression by binding, as a heterodimer with the retinoid X receptor, to specific DNA sequences, called PPAR response elements (PPRE), resulting in the transcriptional activation of target genes (5). More recently, PPAR␣ has also been shown to play a negative role in the inflammatory response by interfering negatively with the AP-1 and NF-B signalization pathway (6).
Physiological responses to nuclear receptor ligands not only depend on the potency of the ligand but also on the expression levels of the nuclear receptors in a given tissue. Regulation of nuclear receptor expression occurs at both protein and mRNA levels. For instance, expression of certain nuclear receptors is repressed in the acute phase inflammatory response (7). This effect is very rapid, which suggests a tight control not only of nuclear receptor mRNA but also of protein levels, likely via a control of the degradation and stability of the nuclear receptors. Indeed, several nuclear receptors, such as the retinoid X receptor ␣, the retinoic acid receptor ␥ (8,9), the retinoic acid receptor ␣ (9), the thyroid hormone receptor (10), and PPAR␥ (11), have been shown to be degraded by the ubiquitin-proteasome system. This degradation pathway is implicated in the regulation of many short-lived proteins involved in essential functions of the cells, including cell cycle control, transcription regulation, and signal transduction (12). The proteins degraded by this pathway are covalently modified by fixation of an 8-kDa polypeptide, called ubiquitin, on lysine residues in a three-step process. In the first step, ubiquitin is activated by a ubiquitinactivating enzyme (E1). Then, the activated ubiquitin is transferred to a ubiquitin carrier protein (E2). Finally, ubiquitinprotein isopeptide ligase (E3) catalyzes the covalent bond of ubiquitin to the target protein. Following this process, multiubiquitinated proteins are rapidly degraded by the 26 S proteasome (13).
We demonstrated previously that PPAR␣ mRNA and protein levels follow a circadian rhythm (14). More recently, it was reported that PPAR␣ mRNA is rapidly down-regulated in the acute phase inflammatory response (7). Since these responses imply a rapid regulation also at the levels of PPAR␣ protein, the present study was designed to test whether PPAR␣ protein levels are controlled by the ubiquitin-proteasome degradation pathway. Our results demonstrate that PPAR␣ is an unstable protein that is rapidly degraded and that ligand activation stabilizes this nuclear receptor. Moreover, we show that the degradation of PPAR␣ involves the ubiquitin-proteasome pathway and that its stabilization observed in the presence of the ligand is due to a decrease of PPAR␣ ubiquitination. Inhibition of the proteasome increases the amount of PPAR␣ protein and consequently the transcriptional activation of PPAR␣-dependent promoters. These results indicate that the proteasome plays an important role in the regulation of PPAR␣ protein level, a mechanism contributing to the magnitude of ligand response.

EXPERIMENTAL PROCEDURES
Materials-Dulbecco's modified Eagle's medium (DMEM) without methionine was purchased from ICN (Orsay, France), and DMEM and fetal calf serum (FCS) were purchased from Invitrogen. [ 35 S]methionine was obtained from PerkinElmer Life Sciences. Wy 14,643 was from Chemsyn Science Laboratories (Lenexa, KS); fenofibric acid was from Laboratoires Fournier (Dijon, France); ciprofibrate was from Sanofi (Aramon, France); and GW7647 was kindly provided by Dr. Peter J. Brown (GlaxoSmithKline) (15) and cerivastatin was provided by Bayer (Wuppertal, Germany). MG132 was from Calbiochem. ExGen 500 was obtained from Euromedex (Souffelweyersheim, France). The anti-PPAR␣ antibody (H98) and the monoclonal antibody against the HA epitope was obtained from Santa Cruz Biotechnology, Inc. (Le Perray en Yvelines, France), and the secondary antibody against rabbit IgG was purchased from Sigma.
Pulse-chase Experiment-COS 7 cells were grown in DMEM medium containing 10% FCS. Cells were transfected with 2 g of PPAR␣ expression vector (pSG5hPPAR␣) (16) using ExGen 500 according to the manufacturer's protocol. Twenty-four h after transfection, COS 7 cells were deprived from methionine for 30 min, labeled with [ 35 S]methionine for 1 h, and after washes, treated with 50 M Wy 14,643 or Me 2 SO (vehicle) in DMEM medium containing 2% FCS for the indicated periods of time. Cells were lysed in RIPA buffer (10 mM Tris-HCl (pH 8), 150 mM NaCl, 2 mM EDTA, 0.5% Nonidet P-40, 0.5% deoxycholate, 1% SDS) containing 4 mM orthovanadate, 20 mM ␤-glycerophosphate, 1% aprotinin, and 1 mM phenylmethylsulfonyl fluoride. The lysates were centrifuged at 10,000 ϫ g for 30 min at 4°C. Labeling efficiency was determined by measuring the radioactivity contained in the pellet by precipitation of the proteins with 20% trichloroacetic acid at 4°C for 30 min. The precipitated proteins were loaded on a filter under aspiration. After washes with 5% trichloroacetic acid, scintillated liquid was added to the filter, and the radioactivity was measured on a Beckman counter. PPAR␣ was immunoprecipitated from 7 ϫ 10 6 cpm using 4 g of an antibody directed against amino acids 1-98 of PPAR␣. After one night of incubation at 4°C under agitation, 20 l of protein G-Sepharose (Sigma) were added for 30 min at 4°C under agitation. The beads were pelleted by centrifugation for 1 min at 3,000 ϫ g and were successively washed in 1 ml of RIPA, 1 ml of RIPA/1 M NaCl, 1 ml of RIPA/TNE buffer(v/v) (TNE buffer, 10 mM Tris-HCl (pH 8), 150 mM NaCl, 2 mM EDTA), and 1 ml of TNE buffer. Laemmli loading buffer was added to the beads. After boiling, the proteins were separated by SDS-PAGE and analyzed by autoradiography.
Ubiquitination Detection-COS 7 cells were grown in DMEM containing 10% FCS. Cells were transfected for 3 h with 2 g of expression vectors for PPAR␣ (pSG5hPPAR␣) and for HA-tagged ubiquitin (MT123-Ubiquitine-HA) (17) using ExGen 500. After 24 h, cells were treated with different compounds for 5 h in the absence of fetal calf serum. Then, cells were lysed on ice in RIPA buffer containing 4 mM orthovanadate, 20 mM ␤-glycerophosphate, 1% aprotinin, and 1 mM phenylmethylsulfonyl fluoride. The lysates were centrifuged at 10,000 ϫ g for 30 min at 4°C, and the protein concentrations of the supernatants were determined using the BCA kit (Uptima Interchim, Montluçon, France). Immunoprecipitation of PPAR␣ protein was performed on 150 g of protein extract as described above. The immunoprecipitated proteins were analyzed by Western blot using a monoclonal antibody against the HA epitope. The blot was revealed with the ECL (enhanced chemiluminescence) (Amersham Biosciences) reagent according to the manufacturer's protocol.
Preparation of Nuclear Extracts and Western Blot Analysis-HepG2 cells were treated with 40 M MG132 for 2 or 4 h in the absence of serum and subsequently trypsinized and washed. Then, cells were resuspended in DMEM medium containing 10% FCS and 10% Me 2 SO, frozen in liquid nitrogen, and conserved at Ϫ80°C. To prepare cytoplasmic extracts, cells were centrifuged for 5 min at 800 ϫ g and resuspended in 5 ml of HB buffer (15 mM Tris-HCl (pH 8), 15 mM NaCl, 60 mM KCl, 0,5 mM EDTA, 1 mM phenylmethylsulfonyl fluoride), centrifuged at 800 ϫ g for 5 min, resuspended in 200 l of HB buffer supplemented with 0.05% Triton X-100 (Sigma), and centrifuged for 10 min at 1,000 ϫ g, and the supernatant was collected. The pellet containing the nuclei was washed with 5 ml of HB buffer containing 0.05% Triton X-100 and 5 ml of HB buffer. Nuclei were incubated at 4°C for 30 min in 50 l of HB buffer containing 360 mM KCl and centrifuged for 5 min at 10,000 ϫ g, and the supernatant corresponding to the nuclear extract was collected. The concentration of protein in the extracts was determined using the BCA kit. Twenty g of nuclear extracts were analyzed by Western blot using the anti-PPAR␣ antibody. The blot was revealed with the ECL reagent.
Transfection Experiments-HepG2 cells were cultured in 24-well plates. Cells were transfected with 10 ng of PPRE-containing reporter plasmid J6-TK-pGL3 (4) or the control TK-pGL3 and 50 ng of pSV␤galactosidase control vectors using ExGen 500. After 24 h, cells were treated with 40 M MG132 for 2 or 4 h prior to adding 50 M Wy 14,643 or Me 2 SO for 24 h. Cell extracts were prepared, and luciferase and ␤-galactosidase assays were performed as described previously (16).
RNA Analysis-HepG2 cells were grown in DMEM medium containing 10% FCS supplemented with 1 mM sodium pyruvate and nonessential amino acids. Cells were treated with 40 M MG132 for 2 h prior to adding 50 M Wy 14,643 or Me 2 SO for 24 h. RNA extraction was performed using TRIzol (Invitrogen) reagent according to the manufacturer's protocol. One g of RNA was reverse-transcribed using random hexamer primers and Moloney murine leukemia virus reverse transcriptase (Invitrogen). RNA levels were measured by quantitative PCR using the LightCycler-FastStar DNA SYBR Green I kit (Roche Diagnostics) on the LightCycler system (Roche Diagnostics) and as primers for human FATP, 5Ј-GGC GCC ACC CCG ACA AGA C-3Ј and 5Ј-CGG GCT GGC ATG GAC CTC AC-3Ј (fragment size: 325 bp), and as primers for human apoA-II, 5Ј-CAT GAA GCT GCT CGC AGC AAC TG-3Ј and 5Ј-CTG GGC TCT TGA CCT TCT CCA TC-3Ј (fragment size: 166 bp). As control, cyclophilin mRNA was measured using GCA TAC GGG TCC TGG CAT CTT GTC C sense and ATG GTG ATC TTC TTG CTG GTC TTG C antisense primers. For each primer pair, the linearity of the reaction was confirmed to have a correlation coefficient of at least 0.98 over the detection area by measuring a 10-fold dilution curve with cDNA isolated from HepG2 cells. Samples were analyzed in triplicate in three independent runs. Ct values, defined as the cycle number in which the detected fluorescence exceeds the threshold value (18,19), were determined for FATP and apoA-II and normalized to the Ct of cyclophilin using the following equation: relative values ϭ 2 (Ct cyclophilin Ϫ Ct target gene) .

PPAR␣ Is a Short-lived Protein That Is Stabilized by Its
Ligand, Wy 14,643-To study the stability of PPAR␣ protein, pulse-chase experiments were performed in COS 7 cells transfected with the pSG5hPPAR␣ expression vector. PPAR␣ protein was immunoprecipitated after a chase of 0, 2, 5, 10, and 24 h. In the absence of ligand, PPAR␣ protein is rapidly degraded in cells (Fig. 1A). After 2 h of chase, the quantity of [ 35 S]-labeled PPAR␣ obtained was drastically decreased, indicating that PPAR␣ is a short-lived protein. Since ligand activation has been shown to influence the stability of nuclear receptor proteins (20 -24), the effect of treatment with Wy 14,643, a PPAR␣ ligand, was tested next. After 2 and 5 h of chase, a significantly higher amount of [ 35 S]-labeled PPAR␣ protein was observed in the presence of Wy 14,643 than in the presence of Me 2 SO (Fig. 1A), indicating that Wy 14,643 induces a stabilization of PPAR␣ protein. Next, a more detailed time course was performed to determine PPAR␣ protein stability in the presence and absence of ligand (Fig. 1B). This experiment showed that PPAR␣ presents a half-life of approximately 1 h in the absence of ligand and of approximately 2 h in the presence of Wy 14,643. Again, when compared with the vehicle-treated cells, slower degradation of PPAR␣ protein was observed in the presence of Wy 14,643. Interestingly, this protective effect of the ligand was observed only during the first 3 h of activation (Fig. 1B). These data indicate that PPAR␣ protein is rapidly degraded in cells and that ligand activation stabilizes this protein in a transitory manner.
PPAR␣ Protein Is Degraded by the Proteasome-To demonstrate the implication of the proteasome in the degradation of PPAR␣, a pulse-chase experiment was performed in COS 7 cells transfected with the PPAR␣ expression vector and treated or not with MG132. PPAR␣ protein was immunoprecipitated after 0, 2, and 4 h. In line with the data above, the results obtained show that PPAR␣ protein is rapidly degraded in the cells. After 2 h of chase, a very low amount of [ 35 S]-labeled PPAR␣ proteins was observed, which became undetectable after 4 h (Fig. 1C). In the presence of MG132, a stabilization of the protein was observed. After 2 h of chase, the quantity of PPAR␣ protein was not reduced (Fig. 1D). Since HepG2 cells express significant amounts of endogenous PPAR␣ protein (25), the influence of MG132 on endogenous PPAR␣ level was analyzed. Western blot analysis using a specific PPAR␣ antiserum demonstrated that treatment of HepG2 cells with MG132 (Fig.  2, lanes 2 and 3) resulted in a significant increase of endogenous PPAR␣ proteins levels (Fig. 2, lane 1). These results indicate that both endogenously and exogenously expressed PPAR␣ proteins are degraded by the ubiquitin-proteasome pathway.
PPAR␣ Protein Is Ubiquitinated in a Ligand-dependent Manner-A number of nuclear receptors are degraded by the ubiquitin-proteasome pathway (8 -11). To determine whether PPAR␣ is also degraded by this system, COS 7 cells were cotransfected with expression vectors for PPAR␣ and an HA epitope-tagged ubiquitin. After treatment with Wy 14,643 or vehicle for 5 h, PPAR␣ proteins were immunoprecipitated, and the ubiquitinated PPAR␣ proteins were revealed with an anti-HA antibody. The results from this experiment show that PPAR␣ is ubiquitinated (Fig. 3, lane 3). The high molecular weight of the ubiquitinated PPAR␣ proteins suggests the presence of numerous ubiquitination sites. Addition of Wy 14,643 decreased the amount of ubiquitinated PPAR␣ protein (Fig. 3,  lane 4), data that are in line with the stabilization effect of the ligand obtained in the pulse-chase experiment (Fig. 1). To determine whether the ubiquitinated PPAR␣ protein is degraded by the proteasome, COS 7 cells transfected with an expression vector for PPAR␣ were treated with the proteasome inhibitor, MG132, which inhibits the degradation of ubiquitinconjugated proteins by the 26 S proteasome complex. In the presence of this inhibitor, an increase of ubiquitinated PPAR␣ protein was observed in cells treated or not with Wy 14,643 (Fig. 3, lanes 5 and 6). These results show that ubiquitinated PPAR␣ protein is degraded by the proteasome. To confirm that the effect of Wy 14,643 on PPAR␣ ubiquitination is a ligandspecific effect, the influence of other PPAR␣ ligands, including clinically used fibrates and the highly specific PPAR␣ agonist GW7647, were analyzed next. In addition, the influence of cerivastatin, which is not a PPAR␣ ligand but which is known to activate this nuclear receptor by modulating its phosphorylation status (25), was tested. The results obtained show that the decrease in PPAR␣ ubiquitination was observed only with the PPAR␣ ligands but not with cerivastatin (Fig. 4). These data demonstrate that PPAR␣ is ubiquitinated and that ligand activation decreases PPAR␣ protein ubiquitination.
Inhibition of PPAR␣ Degradation by the Proteasome Increases Its Transcriptional Activity-To determine the consequence of inhibition of PPAR␣ degradation on its transcriptional activity, HepG2 cells were transfected with a reporter vector containing the J site PPRE of the apoA-II gene promoter (J6-TK-pGL3) or with the control reporter vector TK-pGL3 and treated or not with MG132 for 2 and 4 h prior to activation with the Wy 14,643 compound. In cells transfected with the TK-pGL3, no modification of the reporter activity was induced by the treatment with Wy 14,643 and MG132 (Fig. 5). In the untreated cells transfected with the J6-TK-pGL3, a 2-fold activation of reporter activity was observed in the presence of Wy 14,643 as compared with vehicle (Fig. 5). Pretreatment of HepG2 cells with MG132 increased both the basal and ligandstimulated transcriptional activation of the reporter gene. This effect was already observed when the cells were pretreated for 2 h with MG132 and was even more pronounced after 4 h (Fig. 5).
To determine whether this stabilization of PPAR␣ protein by MG132 resulted in changes in PPAR␣ target gene expression, HepG2 cells were treated or not with MG132 for 2 h prior to activation with Wy 14,643, and the expression level of two established PPAR␣ target genes, apoA-II (4) and FATP (26), was analyzed by quantitative PCR. Pretreatment with MG132 increased Wy 14,643-induced expression of both the apoA-II (Fig. 6A) and FATP (Fig. 6B) genes. These data show that MG132 stabilization of PPAR␣ protein levels results in an enhanced ligand-induced expression of PPAR␣ target genes in HepG2 cells.

DISCUSSION
Eukaryotic cells exhibit rigorous control over gene expression by tightly regulating the expression and activity of transcription factor proteins. The concentrations of these transcriptional regulators are controlled, at least in part, through proteasome-mediated protein degradation. The first step in this process, the ubiquitination of proteins, which are subsequently degraded by the 26 S proteasome complex, is a highly regulated process leading to the modulation of transcription factor activity (13). In this report, it is shown that the ubiquitin-proteasome pathway controls the degradation of PPAR␣ protein and as such modulates the concentration of PPAR␣ in hepatocytes. Results from pulse-chase experiments demonstrate that PPAR␣ protein has a short half-life, which is extended by liganding of the receptor, results that confirm and extend data that appeared when this work was in progress (27). In addition, we show that PPAR␣ protein stabilization by its ligand is associated with a reduction in ubiquitination of PPAR␣ protein. Furthermore, a highly specific inhibitor of the proteasome, MG132, blocks PPAR␣ protein degradation, and the resulting enhanced expression of PPAR␣ leads to a higher transcriptional activation of a reporter gene driven by a PPARresponsive element. In addition, treatment of HepG2 cells with MG132 results in an enhanced ligand-induced expression of endogenous PPAR␣-responsive genes. Recently, it was demonstrated that elevating expression of PPAR␣ in HepG2 cells by overexpressing exogenous PPAR␣ results in an increased expression of endogenous PPAR␣ target genes (28), whereas genetic ablation of PPAR␣ expression results in decreased basal and/or ligand-induced expression of PPAR␣ target genes (29), indicating that PPAR␣ protein levels are a determinant of the response to its ligands. The results from this study, demonstrating that PPAR␣ protein expression and activity are regulated at the level of degradation, thus provide a novel mechanism of control of PPAR␣ activity.
Although only a limited number of studies have addressed PPAR␣ protein regulation, the concentration of PPAR␣ has been shown to be critical under a number of physiological situations. PPAR␣ expression has been shown to oscillate with a circadian rhythm in liver (14). The diurnal variations of PPAR␣ mRNA is closely followed by a parallel cycling of PPAR␣ protein. This rapid diurnal cycling of PPAR␣ protein levels implied that the half-life of the protein should be short enough to allow its levels to significantly decrease over a period of 12 h. Our results from pulse-chase experiments demonstrate that PPAR␣ protein half-life is approximately 1 h due to its rapid degradation by the proteasome, which provides a molecular mechanism potentially contributing to the rapid circadian cycling of this protein. Furthermore, our data indicate that regulation of PPAR␣ protein level by controlling its degradation modulates the expression of different PPAR␣ target genes in response to its ligand. For example, the expression of the apoA-II and FATP genes, two well characterized PPAR␣ target genes (4,26), by MG132 treatment is increased in a liganddependent manner. Thus, in addition to PPAR␣ control by the level of its gene transcription as well as by the potency of the ligand, the magnitude of the physiological response to PPAR␣ activation is also regulated at the level of its stability. It will be of interest to identify physiological factors that influence PPAR␣ degradation and as such affect the PPAR␣ signaling pathway.
Interestingly, our study shows that ligand activation protects PPAR␣ from ubiquitination and degradation in a rapid but transient manner. Previous studies demonstrated that the estrogen and vitamin D3 receptors are also degraded by the proteasome and that this is accelerated after ligand exposure (20,21). Similarly, progesterone receptor protein levels are down-regulated after progesterone treatment (22,23). Other results clearly indicate a hormone-mediated destabilization of the glucocorticoid receptor (24). Under basal conditions, glucocorticoid receptor has a fairly long half-life of 18 h, whereas dexamethasone-treatment decreases its half-life to 8 -9 h. In contrast to the previous receptors, PPAR␣ has a very short half-life, and ligand activation prolongs its half-life. We therefore propose a model in which the ubiquitin-proteasome pathway may contribute to the regulation of duration and magnitude of the response to PPAR␣ activators. The interaction with its ligand reduces the ubiquitination of the PPAR␣ protein and consequently its degradation. This protective effect appears transitory, which may be due either to a rapid metabolization of the ligand in liver cells and/or to a ligand-dependent recruitment of coactivators that could induce the degradation of the PPAR␣ protein. Indeed, it has been shown previously that the AF-2 domain of certain nuclear receptors binds a component of the proteasome in a hormone-dependent manner (30), which may result in the arrest of the transcriptional activation by the liganded receptor in a temporally defined manner. Further investigation is necessary to identify the mechanisms explaining the transitory stabilization of PPAR␣ protein by its ligand.