Orphan Nuclear Hormone Receptor RevErbα Modulates Expression from the Promoter of the Hydratase-dehydrogenase Gene by Inhibiting Peroxisome Proliferator-activated Receptor α-Dependent Transactivation*

Peroxisome proliferator-activated receptor α (PPARα) heterodimerizes with the 9-cis-retinoic acid receptor (RXRα) to bind to peroxisome proliferator-response elements (PPRE) present in the upstream regions of a number of genes involved in metabolic homeostasis. Among these genes are those encoding fatty acyl-CoA oxidase (AOx) and enoyl-CoA hydratase/3-hydroxyacyl-CoA dehydrogenase (HD), the first two enzymes of the peroxisomal β-oxidation pathway. Here we demonstrate that the orphan nuclear hormone receptor, RevErbα, modulates PPARα/RXRα- dependent transactivation in a response element-specific manner. In vitro binding analysis showed that RevErbα bound the HD-PPRE but not the AOx-PPRE. Determinants within the HD-PPRE required for RevErbα binding were distinct from those required for PPARα/RXRα binding. In transient transfections, RevErbα antagonized transactivation by PPARα/RXRα from an HD-PPRE luciferase reporter construct, whereas no effects were observed with an AOx-PPRE reporter construct. These data identify the HD gene as a target for RevErbα and illustrate cross-talk between the RevErbα and PPARα signaling pathways on the HD-PPRE. Our results suggest a novel role for RevErbα in regulating peroxisomal β-oxidation.

Peroxisome proliferator-activated receptors (PPAR) 1 are members of the steroid/thyroid hormone receptor superfamily that act to regulate a number of genes involved in differentiation and lipid metabolism (1)(2)(3)(4)(5). These ligand-activated transcription factors respond to a class of chemical agents termed peroxisome proliferators that include the fibrate family of hypolipidemic drugs, phthalate ester plasticizers, herbicides, pesticides, antidiabetic thiazolidinediones, as well as certain fatty acids (6 -11). PPARs heterodimerize with the 9-cis-retinoic acid receptor, RXR␣, and bind to specific response elements termed peroxisome proliferator-response elements (PPRE) found upstream of target genes. PPREs have been identified in the promoter regions of a number of genes involved in fatty acid metabolism, including those coding for liver fatty acid-binding protein (12,13), mitochondrial medium chain acyl-CoA dehydrogenase (14), malic enzyme (15), 3-hydroxy-3-methylglutaryl-CoA synthase (16), apolipoprotein A1 (17), and the first two enzymes of the peroxisomal ␤-oxidation pathway, fatty acyl-CoA oxidase (AOx) (1) and enoyl-CoA hydratase/3-hydroxyacyl-CoA dehydrogenase (HD) (2). PPREs consist of a direct repeat of the hexameric motif TGACC(T/C) separated by one nucleotide (DR1), and serve to bind not only PPARs but also several other nuclear hormone receptors that differentially modulate PPAR function. Nuclear hormone receptors shown to modulate PPAR function include chicken ovalbumin upstream promoter transcription factor (18), hepatocyte nuclear factor-4 (19), thyroid hormone receptor (20), LXR␣ (21), and RZR␣ (22). Therefore, transcriptional regulation from PPREs is a net aggregate response generated in part by the availability of PPARs and other factors that bind PPREs, the complexity of response elements, and the interplay of PPARs with other nuclear hormone receptors and cofactors, including corepressors and coactivators. This complex interaction of various factors and elements ensures that the correct transcriptional response to extra-and intracellular stimuli will be elicited from appropriate target genes.
RevErb␣ (also known as Ear1) (23) is an orphan member of the steroid/thyroid hormone receptor superfamily (24) that responds to the fibrate family of hypolipidemic drugs (17). RevErb␣ is expressed in a variety of tissue types, including adipocytes (24), skeletal muscle (25), and liver (25). Although the biological function of RevErb␣ remains elusive, it has been implicated in adipogenesis (24), thyroid hormone signaling (26 -28), and muscle differentiation (29). Recently, the gene for rat apolipoprotein A1, as well as the CYP4A6 gene encoding a member of the cytochrome P450 fatty acid -hydroxylase family, have been shown to contain PPREs that bind, and are regulated by, RevErb␣ (17,30). Human RevErb␣ has also been shown to mediate the transcriptional repression of its own promoter in vitro (31).

EXPERIMENTAL PROCEDURES
Plasmids-The luciferase reporter plasmids pCPSluc, pHD(x3)luc, pM2(x3)luc, and pAOx(x2)luc, and the expression plasmids for rat PPAR␣ and human RXR␣, have been described previously (2,18,22,39,40). The expression plasmid for human RevErb␣ (RevErb␣:SG5) was constructed by excision of the human RevErb␣ cDNA as a BamHI fragment from the plasmid pCMX:hRevErb␣ (a kind gift of H. P. Harding and M. A. Lazar) (23), followed by cloning of the fragment in the correct orientation into the BamHI site of the plasmid pSG5 (Promega).
In Vitro Transcription/Translation-Transcription/translation of cDNAs encoding PPAR␣, RXR␣, and RevErb␣ was performed using the TNT T7-coupled rabbit reticulocyte lysate system according to the manufacturer's protocol (Promega). Translation products labeled with L-[ 35 S]methionine were analyzed on 15% SDS-polyacrylamide gels. Synthesis of proteins for use in electrophoretic mobility shift analysis (EMSA) was carried out in parallel with unlabeled methionine.
Electrophoretic Mobility Shift Analysis-EMSA using in vitro translated proteins and radiolabeled oligonucleotide probes was carried out essentially as described (18). Oligonucleotides corresponding to the AOx-PPRE (5Ј-gatCCTTTCCCGAACGTGACCTTTGTCCTGGTCCCC-TTTTGCTa and its complement) and to the HD-PPRE (5Ј-gatCCTCT-CCTTTGACCTATTGAACTATTACCTACATTTGA and its complement) were annealed and end-labeled with the Klenow fragment of DNA polymerase I and [␣-32 P]dATP. The underlined sequences indicate the TGACCT-like DRs. Combinations of lysates programmed for RXR␣, PPAR␣, or RevErb␣ were incubated at room temperature with labeled probes in a final volume of 20 l containing 6 mM Hepes (pH 7.9), 120 mM NaCl, 0.4 mM MgCl 2 , 0.1 mM EDTA, 7% (v/v) glycerol, 4 g of bovine serum albumin, 4 g of nonspecific competitor DNA (poly(dI⅐dC) and sonicated salmon sperm DNA, 1:1 weight ratio), 150 M phenylmethylsulfonyl fluoride, and 0.2 mM dithiothreitol. The total amount of reticulocyte lysate in each reaction was kept constant by addition of unprogrammed lysate. Binding reactions were analyzed by electrophoresis at 4°C on prerun 2.5 or 3.5% (as indicated) polyacrylamide gels (30:1 acrylamide/N,NЈ-methylene bisacrylamide weight ratio) with 0.5 ϫ Tris borate-EDTA (22 mM Tris base, 22 mM boric acid, 1 mM EDTA) or 1.0 ϫ Tris borate-EDTA as running buffers, followed by autoradiography.
Transient Transfection and Measurement of Luciferase Activity-Transfections were performed in African green monkey kidney BSC40 monolayer cells by the calcium phosphate method (2). Cells at 60 -70% confluence were transfected in medium without phenol red and supplemented with 5% (v/v) charcoal-stripped fetal bovine serum. Transfections were carried out with luciferase reporter constructs (pCPSluc, pHD(x3)luc, pM2(x3)luc, or pAOx(x2)luc) and expression plasmids for RXR␣, PPAR␣, and RevErb␣, as indicated. 2 g of the ␤-galactosidase expression plasmid pCH110 (Amersham Pharmacia Biotech) was included in transfections as an internal reference. Effector plasmid dosage was kept constant by the addition of empty expression plasmid. Total DNA was kept constant at 20 g/10-cm plate by the addition of sonicated salmon sperm DNA. The peroxisome proliferator Wy-14,643 in dimethyl sulfoxide was added to fresh medium to a final concentration of 0.1 mM. An equivalent amount of dimethyl sulfoxide was added to control medium. Cells were harvested and lysates prepared 48 h post-transfection, and luciferase activity was quantitated as described previously (2).

RevErb␣ Binds to the HD-PPRE-Both the HD-and AOx-
PPREs contain potential binding sites for RevErb␣ (Fig. 1A). We therefore examined whether RevErb␣ is capable of binding to the HD-and AOx-PPREs by performing EMSA with radiolabeled PPRE probes and in vitro translated receptors. As expected, PPAR␣ and RXR␣ bound as a heterodimer on the HD-PPRE (Fig. 1B). RevErb␣ also bound the HD-PPRE, forming a complex with a mobility slightly less than that formed by the PPAR␣/RXR␣ heterodimer (Fig. 1B). Inclusion of RXR␣ or PPAR␣ with RevErb␣ in the binding reactions had no effect on the formation of the RevErb␣⅐HD-PPRE complex. Moreover, coincubation of all three receptors with the HD-PPRE produced only two distinct complexes corresponding to PPAR␣/RXR␣ heterodimers and RevErb␣ monomers (Fig. 1B). Therefore, the three receptors do not co-occupy the HD-PPRE in some higher order complex, and RevErb␣ does not form complexes with PPAR␣ or RXR␣ on this element in vitro under the EMSA conditions used. In contrast to the results obtained with the HD-PPRE, RevErb␣ was unable to bind to the AOx-PPRE. Furthermore, only the characteristic PPAR␣/RXR␣ heterodimer was generated on the AOx-PPRE when RevErb␣ was coincubated with PPAR␣ and RXR␣ (Fig. 1B). Binding of RevErb␣ to the HD-PPRE was specific, since the radiolabeled complex was refractory to competition by nonspecific unlabeled oligonucleotide but was eliminated by addition of unlabeled HD-PPRE oligonucleotide (data not shown).
We next investigated whether binding by PPAR␣/RXR␣ to the HD-PPRE could be influenced by RevErb␣ and vice versa. EMSA was first performed using radiolabeled HD-PPRE, constant amounts of PPAR␣/RXR␣, and increasing amounts of RevErb␣ ( Fig. 2A). Increasing the amount of RevErb␣ did not affect binding of PPAR␣/RXR␣ to the HD-PPRE, but did result in greater amounts of RevErb␣ monomeric complex forming on the HD-PPRE. Conversely, when EMSA was performed with a constant amount of RevErb␣, increasing the amounts of PPAR␣ and RXR␣ led to increased formation of PPAR␣⅐RXR␣ complexes on the HD-PPRE, while the binding of RevErb␣ monomers was unaffected (Fig. 2B). Similar findings were observed over a wide range of HD-PPRE probe concentration (Fig.  2, A and B). These results suggest that RevErb␣ and PPAR␣/ RXR␣ bind independently to the HD-PPRE, and that RevErb␣ monomers and PPAR␣/RXR␣ heterodimers do not simultaneously occupy the HD-PPRE.
RevErb␣ and PPAR␣/RXR␣ Bind to Distinct Sites on the HD-PPRE-The HD-PPRE consists of four half-sites (sites I-IV) related to the consensus TGACC(T/C) hexameric half-site. Each half-site of the HD-PPRE could potentially serve as a binding site for RevErb␣. In order to determine the sequence requirements for RevErb␣ interaction, oligonucleotide probes harboring mutations in each of the four hexameric half-sites of the HD-PPRE were used in binding studies (Fig. 3). EMSA showed that mutations in sites I, III, and IV did not affect binding of RevErb␣, either in the absence (data not shown) or presence of PPAR␣/RXR␣. However, disruption of site II effectively abrogated RevErb␣ binding. This observation is consistent with the fact that site II most closely resembles the consensus sequence for RevErb␣ binding (Fig. 1A). Binding of PPAR␣/RXR␣ was found to require the integrity of sites III and IV but not of sites I and II (Fig. 3), as previously demonstrated (22,40). These results show that RevErb␣ and PPAR␣/RXR␣ target distinct half-sites on the HD-PPRE. RevErb␣ Antagonizes Transactivation by PPAR␣/RXR␣ from the HD-PPRE but Not the AOx-PPRE-To investigate the in vivo properties of RevErb␣ on transcriptional regulation, we carried out transient transfection assays using luciferase reporter plasmids containing either the HD-PPRE (pHD(x3)luc) or the AOx-PPRE (pAOx(x2)luc), along with expression plasmids for RevErb␣, PPAR␣, and/or RXR␣ in BSC40 African monkey kidney cells. These cells were chosen because they contain low levels of endogenous PPAR␣ and RXR␣ (2,18). Co-transfection of PPAR␣ and RXR␣ with the HD-PPRE reporter plasmid led to a 2-fold induction of transcription over basal levels in the absence of the peroxisome proliferator, Wy-14,643 (Fig. 4A). Addition of proliferator led to a potent induction of transcription (10 -15-fold) over basal levels. Increasing amounts of RevErb␣ expression plasmid inhibited transactivation from the HD-PPRE by PPAR␣/RXR␣ both in the presence and absence of peroxisome proliferator. Transactivation by PPAR␣/RXR␣ was reduced by 80% at the highest concentration of RevErb␣ expression plasmid used (2 g). PPAR␣/RXR␣ also activated transcription of a reporter gene that contained the HD-PPRE harboring a disruption in site II; however, in this case, RevErb␣-dependent inhibition was not observed (Fig.  4B). This finding is in agreement with in vitro binding data for RevErb␣ (see Fig. 3) and indicates that inhibition of PPAR␣/ RXR␣-mediated activation is dependent upon RevErb␣ binding to the HD-PPRE.
RevErb␣ did not significantly affect transcriptional activation by PPAR␣/RXR␣ on the AOx-PPRE, in either the absence or presence of Wy-14,643 (Fig. 4C). These data are in keeping with in vitro binding data showing that RevErb␣ failed to bind the AOx-PPRE (Fig. 1B). Control transfections with the parental reporter construct pCPSluc, which lacks a PPRE, showed that the presence of RevErb␣, PPAR␣, and RXR␣ did not influence basal levels of luciferase activity (data not shown), demonstrating the need for a functional PPRE for receptor activity. Together, these data show that RevErb␣ antagonizes transactivation by PPAR␣/RXR␣ specifically from the HD-PPRE.
Increased Amounts of PPAR␣ and RXR␣ Can Overcome Repression by RevErb␣-We were interested in determining whether increased amounts of PPAR␣ and RXR␣ could modulate the repression exerted by RevErb␣ on transcription from the HD-PPRE. Transient transfections with a constant amount of RevErb␣ expression plasmid and increasing amounts of PPAR␣ and RXR␣ expression plasmids demonstrated that the repressive effects of RevErb␣ could be alleviated in a dose-dependent manner by increasing amounts of PPAR␣ and RXR␣, either in the presence or absence of Wy-14,643 (Fig. 5). These results suggest that the net transcriptional response from the HD-PPRE is influenced by the relative levels of PPAR␣, RXR␣, and RevErb␣ in vivo. DISCUSSION A number of recent observations have pointed to possible interplay between the RevErb␣ and PPAR␣ signaling pathways. First, there is strong sequence similarity between the RevErb␣ consensus binding site and the PPREs of the AOx and HD genes. Second, a role for RevErb␣ in PPAR-mediated signaling has been suggested by studies showing that the PPRE of the CYP4A6 gene binds both RevErb␣ and PPAR␣ (30) and that the PPAR␥ isoform, a key regulator of adipogenesis, may in turn be regulated by RevErb␣, whose mRNA levels are dramatically increased during differentiation of preadipocytes to adipocytes (24). Third, RevErb␣ has been shown to be encoded on the opposite strand of thyroid hormone receptor ␣ and to be able to bind the thyroid hormone, triiodothyronine (27), and we have previously demonstrated cross-talk between thy-roid hormone receptor ␣ and PPAR␣ in regulating transcription from the AOx-PPRE (20). As a result of these observations, we considered the possibility of a role for RevErb␣ in regulating transcription from the AOx-and HD-PPREs.
We have demonstrated here that the RevErb␣ and PPAR␣ signaling pathways converge and that RevErb␣ serves to repress transcriptional activation specifically from the HD-PPRE. Interestingly, RevErb␣ had no effect on PPAR␣/RXR␣mediated activation via the AOx-PPRE. Consistent with this observation, RevErb␣ was shown to bind specifically to the HD-PPRE but not the AOx-PPRE. The HD-PPRE is comprised of four hexameric direct repeats arranged as two tandem DR1 arrays separated by two nucleotides (a DR2). This complex arrangement is thought to permit the interaction of a diverse array of nuclear hormone receptors with the HD-PPRE, thereby increasing the complexity of transcriptional regulation from this PPRE. RevErb␣ and PPAR␣/RXR␣ target distinct half-sites on the HD-PPRE. The integrity of site II is required for RevErb␣ binding, while sites III/IV serve to bind RXR␣/ PPAR␣. Site II was also required for the RevErb␣-dependent repressive effects on transcriptional activation by PPAR␣/ RXR␣, indicating that inhibition requires binding of RevErb␣ to the HD-PPRE. Although sites III/IV have been shown to be essential and sufficient for PPAR␣/RXR␣ binding and activity (22,40), an arrangement in which PPAR␣/RXR␣ heterodimers are bound to both DR1 sites has been suggested to yield the highest level of transactivation (41). Since RevErb␣ occupies site II within the HD-PPRE, this may preclude binding of PPAR␣/RXR␣ to the upstream DR1 element, resulting in reduced levels of transactivation from the HD-PPRE by PPAR␣/RXR␣.
Although RevErb␣ and PPAR␣/RXR␣ use distinct determinants on the HD-PPRE, we did not observe a higher order complex containing all three receptors in vitro. However, the limitation of our in vitro binding analysis does not preclude the possibility of a higher order complex forming among RevErb␣, PPAR␣, and RXR␣ in vivo, perhaps through the cooperativity or association of regulatory cofactors such as SRC-1, p300, N-CoR, and SMRT-1 (42)(43)(44)(45). The involvement of such cofactors in transcriptional regulation by nuclear hormone receptors is well established. Indeed, N-CoR has been shown to interact with RevErb␣ in mammalian cells (38). A model can be proposed in which repression of transactivation by RevErb␣ is the result of a shift from active to repressive states of the receptor through its association with corepressors and dissociation from coactivators, respectively. RevErb␣ can also be envisioned to be subjected in vivo to post-translational modifications such as phosphorylation that could initiate its repressive state. Interestingly, the amino terminus of RevErb␣ contains numerous serine and threonine residues that could potentially be phosphorylated (27).
The ligand for RevErb␣ remains unknown. It is therefore impossible to ascertain at this time whether an endogenous ligand exists in BSC40 cells that could induce transcriptional repression upon binding RevErb␣, as has been demonstrated for androstanol and the mCAR␤ receptor (46). It has been suggested that RevErb␣ lacks the AF2 transactivation domain that is responsible for ligand binding (37), thus precluding the possibility of RevErb␣ having any capacity for ligand-dependent activation or repression (38). Orphan receptors lacking AF2 domains could instead act as competitors for ligand-inducible receptors (37), and such a scenario has been proposed to explain the blocking of RZR␣-mediated transactivation by RevErb␣ (25,35). Nevertheless, the absence of an AF2 domain does not preclude the possibility that RevErb␣ contains undetected, and yet undefined, activation domains that could be revealed through interaction with a novel ligand (36). RevErb␣ could also potentially activate transcription by cooperative interaction with a non-AF2-dependent coactivator or, indirectly, by recruiting corepressors away from other nuclear receptors (38). TR␣ and PPAR␣ have been reported to form nonbinding heterodimers in vivo (20,47), and RevErb␣ could similarly form inactive, nonbinding complexes (36) with PPAR␣, RXR␣, or other nuclear receptors, effectively sequestering these receptors and preventing them from forming heterodimers that normally potentiate transcription, leading to an overall repression of transcription. However, we consider the latter scenario unlikely, since inhibition by RevErb␣ required the integrity of the HD-PPRE and had no effect on PPAR␣/RXR␣-mediated transactivation via the AOx-PPRE.
In summary, our results identify the gene encoding enoyl-CoA hydratase/3-hydroxyacyl-CoA dehydrogenase, the second enzyme of the peroxisomal ␤-oxidation pathway, as a target for RevErb␣ and demonstrate that this orphan nuclear hormone receptor serves as a negative modulator of PPAR␣/RXR␣-mediated transactivation from the PPRE of this gene. Interestingly, Gervois et al. (48) have recently demonstrated that fibrate drugs, which are potent activators of PPAR␣, induce expression of the RevErb␣ gene through competition between RevErb␣ and PPAR␣/RXR␣ for binding to the autoregulatory RevDR2 site of the RevErb␣ gene. These results are in agreement with our findings that the repressive effects of RevErb␣ on the HD-PPRE can be overcome by increasing the concentration of PPAR␣/RXR␣, and together, illustrate a convergence of the RevErb␣ and PPAR␣ signaling pathways in gene regulation. Transcriptional control of peroxisomal ␤-oxidation involves a complex network of interacting regulatory factors that integrate a diverse array of host signaling pathways to determine a net transcriptional to a particular environmental or physiological cue. An understanding of the various transcriptional factors that control peroxisomal ␤-oxidation could provide for the development of pharmacological agents that specifically target this metabolic pathway as a means to modulate overall lipid metabolism.