Peroxisome Proliferator-activated Receptor Mediates Cross-talk with Thyroid Hormone Receptor by Competition for Retinoid X Receptor

The peroxisome proliferator-activated receptors (PPAR) and thyroid hormone receptors (TR) are members of the nuclear receptor superfamily, which regulate lipid metabolism and tissue differentiation. In order to bind to DNA and activate transcription, PPAR requires the formation of heterodimers with the retinoid X receptor (RXR). In addition to activating transcription through its own response elements, PPAR is able to selectively down-regulate the transcriptional activity of TR, but not vitamin D receptor. The molecular basis of this functional interaction has not been fully elucidated. By means of site-directed mutagenesis of hPPARα we mapped its inhibitory action on TR to a leucine zipper-like motif in the ligand binding domain of PPAR, which is highly conserved among all subtypes of this receptor and mediates heterodimerization with RXR. Replacement of a single leucine by arginine at position 433 of hPPARα (L433R) abolished heterodimerization of PPAR with RXR and consequently its trans-activating capacity. However, a similar mutation of a leucine residue to arginine at position 422 showed no alteration of heterodimerization, DNA binding, or transcriptional activation. The dimerization deficient mutant L433R was no longer able to inhibit TR action, demonstrating that the selective inhibitory effect of PPAR results from the competition for RXR as well as possibly for other TR-auxiliary proteins. In contrast, abolition of DNA binding by a mutation in the P-box of PPAR (C122S) did not eliminate the inhibition of TR trans-activation, indicating that competition for DNA binding is not involved. Additionally, no evidence for the formation of PPAR:TR heterodimers was found in co-immunoprecipitation experiments. In summary, we have demonstrated that PPAR selectively inhibits the transcriptional activity of TRs by competition for RXR and possibly non-RXR TR-auxiliary proteins. In contrast, this functional interaction is independent of the formation of PPAR:TR heterodimers or competition for DNA binding. The peroxisome proliferator-activated receptors (PPAR) and thyroid hormone receptors (TR) are members of the nuclear receptor superfamily, which regulate lipid metabolism and tissue differentiation. In order to bind to DNA and activate transcription, PPAR requires the formation of heterodimers with the retinoid X receptor (RXR). In addition to activating transcription through its own response elements, PPAR is able to selectively down-regulate the transcriptional activity of TR, but not vitamin D receptor. The molecular basis of this functional interaction has not been fully elucidated. By means of site-directed mutagenesis of hPPARα we mapped its inhibitory action on TR to a leucine zipper-like motif in the ligand binding domain of PPAR, which is highly conserved among all subtypes of this receptor and mediates heterodimerization with RXR. Replacement of a single leucine by arginine at position 433 of hPPARα (L433R) abolished heterodimerization of PPAR with RXR and consequently its trans-activating capacity. However, a similar mutation of a leucine residue to arginine at position 422 showed no alteration of heterodimerization, DNA binding, or transcriptional activation. The dimerization deficient mutant L433R was no longer able to inhibit TR action, demonstrating that the selective inhibitory effect of PPAR results from the competition for RXR as well as possibly for other TR-auxiliary proteins. In contrast, abolition of DNA binding by a mutation in the P-box of PPAR (C122S) did not eliminate the inhibition of TR trans-activation, indicating that competition for DNA binding is not involved. Additionally, no evidence for the formation of PPAR:TR heterodimers was found in co-immunoprecipitation experiments. In summary, we have demonstrated that PPAR selectively inhibits the transcriptional activity of TRs by competition for RXR and possibly non-RXR TR-auxiliary proteins. In contrast, this functional interaction is independent of the formation of PPAR:TR heterodimers or competition for DNA binding. The peroxisome proliferator-activated receptors (PPAR)1 1The abbreviations used are: PPARperoxisome proliferator-activated receptorT3L-triiodothyronineTRT3 receptorRXRretinoid X receptorTRET3-response elementTRAPTR-auxiliary proteinsPPREperoxisome proliferator-activator response elementBFEbifunctional enzymeVDRvitamin D receptorVD1,25(OH)2-vitamin DEMSAelectrophoretic gel mobility shift assayETYA5,8,11,14-eicosatetraynoic acid. are a novel subfamily of the steroid/thyroid hormone nuclear receptor proteins involved in the ligand-inducible regulation of lipid metabolism, adipose tissue differentiation, and possibly hepatocarcinogenesis in rodents(1Green S. Wahli W. Mol. Cell. Endocrinol. 1994; 100: 149-153Google Scholar). Their closest relatives in the superfamily are the type II nuclear hormone receptors, such as the retinoic acid, vitamin D, and thyroid hormone receptors. In particular, the P-box of the first zinc-finger responsible for DNA-binding specificity is fully conserved between hPPARα and the L-triiodothyronine (T3)-receptors (TR) α1 and β1 (Fig. 1A), reflecting the preferential binding of both receptors to differently spaced AGGTCA half-sites(2Krey G. Keller H. Mahfoudi A. Medin J. Ozato K. Dreyer C. Wahli W. J. Steroid Biochem. Mol. Biol. 1993; 47: 65-73Google Scholar). Additionally, a high degree of sequence conservation is found between a putative leucine zipper motif in the ligand-binding domain of PPARα and TR(3Au-Fliegner M. Helmer E. Casanova J. Raaka B.M. Samuels H.H. Mol. Cell. Biol. 1993; 13: 5725-5737Google Scholar). This motif is very highly conserved among various species and PPAR subtypes (Fig. 1B). Besides the structural homology, TRs and PPARs require heterodimerization with the retinoid X receptor (RXR) for optimal DNA binding and both receptors are co-expressed in brain, liver, and adipocytes, where they are involved in the regulation of lipid metabolism(1Green S. Wahli W. Mol. Cell. Endocrinol. 1994; 100: 149-153Google Scholar). We and others have recently reported that the PPARα is able to modulate TR activity either positively or negatively, depending on the T3-response element (TRE)(4Bogazzi F. Hudson L.D. Nikodem V.M. J. Biol. Chem. 1994; 269: 11683-11686Google Scholar, 5Meier-Heusler S.C. Zhu X. Juge-Aubry C. Pernin A. Burger A.G. Cheng S.Y. Meier C.A. Mol. Cell. Endocrinol. 1995; 107: 55-66Google Scholar). Although it has been suggested that rTRα is able to form heterodimers with rPPARα in solution, it has not been demonstrated whether this interaction quantitatively accounts for the observed transcriptional changes. Here we examined the mechanism by which hPPARα structurally and functionally interacts with TR and RXR. We show that hPPARα is an efficient competitor for RXR and most likely for other TR-auxiliary proteins (TRAPs), thereby specifically inhibiting TR activity by disrupting the formation of TR:RXR heterodimers. A series of point mutations in hPPARα allowed the mapping of a region that is indispensable for this cross-talk to a carboxyl-terminal leucine zipper-like motif, which is highly conserved among all subtypes (PPARα, β, γ, and δ) of this receptor. peroxisome proliferator-activated receptor L-triiodothyronine T3 receptor retinoid X receptor T3-response element TR-auxiliary proteins peroxisome proliferator-activator response element bifunctional enzyme vitamin D receptor 1,25(OH)2-vitamin D electrophoretic gel mobility shift assay 5,8,11,14-eicosatetraynoic acid. The pSG5-hPPAR plasmid was kindly provided by Dr. F. Gonzalez(6Sher T. Yi H.F. McBride O.W. Gonzalez F.J. Biochemistry. 1993; 32: 5598-5604Google Scholar). The pBL2-BFE-CAT containing the peroxisome proliferator-response element (PPRE) from the bifunctional enzyme (BFE) promoter (position −2950 to −2925) as well as the pSG5-VDR vector for the human vitamin D receptor and the pBL2-DR3-CAT with a DR3 vitamin D response element were generously provided by Dr. C. Carlberg(7Bardot O. Aldridge T.C. Latruffe N. Green S. Biochem. Biophys. Res. Commun. 1993; 192: 37-45Google Scholar, 8Carlberg C. Bendik I. Wyss A. Meier E. Sturzenbecker L.J. Grippo J.F. Hunziker W. Nature. 1993; 361: 657-660Google Scholar). The pSV2-hTRβ1 and pMTV-TRElap-CAT plasmids are described elsewhere(9Meier C.A. Dickstein B.M. Ashizawa K. McClaskey J.H. Muchmore P. Ransom S.C. Menke J.B. Hao E.H. Usala S.J. Bercu B.B. Cheng S.Y. Weintraub B.D. Mol. Endocrinol. 1992; 6: 248-258Google Scholar, 10Meier C.A. Parkison C. Chen A. Ashizawa K. Meier-Heusler S.C. Muchmore P. Cheng S.Y. Weintraub B.D. J. Clin. Invest. 1993; 92: 1986-1993Google Scholar). Mutant hPPARα (pSG5-hPPARα-L433R and -L422R) were created by the polymerase chain reaction-mediated splice donor site overlap extension method and by subsequently replacing the HindIII/XbaI fragment of pSG5-hPPARα with the mutated polymerase chain reaction product(10Meier C.A. Parkison C. Chen A. Ashizawa K. Meier-Heusler S.C. Muchmore P. Cheng S.Y. Weintraub B.D. J. Clin. Invest. 1993; 92: 1986-1993Google Scholar). The clones were verified by dideoxy sequencing to rule out spurious mutations. hPPARα-L433R and -L422R have a T to G point mutation in codons 433 and 422 at nucleotide positions 1514 and 1400, respectively, changing a leucine to arginine. pSG5-hPPARα-C122S has a point mutation, replacing a C by a G at nucleotide position 581, changing codon 122 from cysteine to serine. This mutant polymerase chain reaction fragment was used to replace the wild-type AvaI/AvaI fragment in pSG5-hPPARα. Rabbit polyclonal anti-TRβ1 antibodies were raised against a synthetic peptide corresponding to the unique hTRβ1 amino acid sequence 61-81 as described by Falcone et al.(11Falcone M. Miyamoto T. Fierrorenoy F. Macchia E. DeGroot L.J. Endocrinology. 1992; 131: 2419-2429Google Scholar). The peptide was coupled to the maleimide-activated keyhole limpet hemocyanin (Pierce) for immunization. Fifty μg of protein was injected with Specol (Central Veterinary Institute, Lelystad, the Netherlands). The specificity of the antibody was confirmed by immunoprecipitation of 35Smethionine-labeled in vitro translated hTRα1, hTRα2, and hTRβ1 as already described(12Schwartz H.L. Strait K.A. Ling N.C. Oppenheimer J.H. J. Biol. Chem. 1992; 267: 11794-11799Google Scholar). To prepare an anti-PPAR antibody the cDNA encoding the 101 first amino acids of the mouse PPARα were cloned into the pQE-9 bacterial expression vector. The expressed polypeptide was purified on a Ni-NTA-agarose column under native conditions according to the manufacturer's instructions (Quiagen, Hilden, Germany), and injected subcutaneously into KOBU rabbits. After the primary injection (200 μg of polypeptide with Freund's adjuvant), the rabbits were boosted 4 times (200 μg/boost). The serum was collected 10 days after the final boost. 35SMethionine-labeled and unlabeled receptors were synthesized using the rabbit reticulocyte lysate transcription/translation kit TnT/T7 (Promega, Madison, WI) according to the manufacturer's instructions. The labeled receptors were analyzed for appropriate size by electrophoresis on a 12.5% sodium dodecyl sulfate-polyacrylamide gel and quantitated by the trichloroacetic acid precipitation method as described(10Meier C.A. Parkison C. Chen A. Ashizawa K. Meier-Heusler S.C. Muchmore P. Cheng S.Y. Weintraub B.D. J. Clin. Invest. 1993; 92: 1986-1993Google Scholar). In vitro translated 35Smethionine-labeled and unlabeled receptors were brought to a final volume of 20 μl with EMSA binding buffer and incubated with the appropriate rabbit polyclonal antibodies or preimmune sera overnight at 4°C. Complexes were precipitated with 50 μl of protein A-agarose (slurry 50%) (Boehringer Mannheim, Germany) previously washed with phosphate-buffered saline containing 0.3% Tween 20, 0.5 mM methionine, and 1% bovine serum albumin. Samples were incubated for 1 h at 4°C with regular shaking. After microcentrifugation the pellet was washed 4 times with 1 ml of phosphate-buffered saline containing 0.3% Tween 20 and 0.5 mM methionine. 50 μl of SDS-PAGE denaturing sample buffer was added to the final pellet and boiled for 5 min at 94°C. The supernatant was subjected to electrophoresis on a 12.5% polyacrylamide gel. Single stranded oligonucleotides were synthesized by Microsynth (Balgach, Switzerland) and annealed with the complementary strands. The following sequences were used(7Bardot O. Aldridge T.C. Latruffe N. Green S. Biochem. Biophys. Res. Commun. 1993; 192: 37-45Google Scholar, 10Meier C.A. Parkison C. Chen A. Ashizawa K. Meier-Heusler S.C. Muchmore P. Cheng S.Y. Weintraub B.D. J. Clin. Invest. 1993; 92: 1986-1993Google Scholar) : TRE-LAP, 5’-AAGGGGATCCAGCTTGACCTGACGTCAGGTCAAGTC-3’; and PPRE (BFE), 5’-AGGGCTTTGACCTATTGAACTATTACCTAC-3’. The ends were filled in using Taq polymerase (Promega, Madison, WI) in the presence of α-32PdCTP (Amersham, United Kingdom). In vitro translated receptors were incubated with 20,000 cpm of the labeled double stranded oligonucleotides in the presence of 2 μg of polyd(I-C) and EMSA binding buffer (50 mM KCl, 20 mM Hepes, 20% glycerol, 0.05% Nonidet P-40, 10 mM β-mercaptoethanol, pH 7.5) added to a final volume of 25 μl. Incubation was performed at room temperature for 20 min and the mixture was then loaded on a 5% polyacrylamide gel. The electrophoresis was performed at 4°C and 300 V for 80 min. HepG2 cells were plated 24 h before transfection in modified Eagle's medium containing 10% (v/v) hormone-depleted fetal calf serum(13Flug F. Copp R.P. Casanova J. Horowitz Z.D. Janocko L. Plotnick M. Samuels H.H. J. Biol. Chem. 1987; 262: 6373-6382Google Scholar), penicillin (100 units/ml), streptomycin (100 μg/ml), and amphotericin B (0.25 μg/ml) in 6-well plates at a density of 0.5 × 106 cells/well. The medium was changed 4 h before transfection. Using the calcium-phosphate method (CellPhect kit, Pharmacia Biotech Inc., Piscataway, NJ) the cells were transfected with the appropriate plasmids. Eighteen hours later the plates were washed once with phosphate-buffered saline, and fresh medium was added together with 500 nM of either L-triiodothyronine, the peroxisome proliferator, and arachidonic acid analogue ETYA or 1,25(OH)2-vitamin D (VD). After another 24 h the cells were harvested, lysed, and the chloramphenicol acetyltransferase activity determined in the extract as described(14Steinfelder H.J. Hauser P. Nakayama Y. Radovick S. McClaskey J.H. Taylor T. Weintraub B.D. Wondisford F.E. Proc. Natl. Acad. Sci. U. S. A. 1991; 88: 3130-3134Google Scholar). Chloramphenicol acetyltransferase activity was normalized for the protein concentration as measured by the Coomassie Blue method. Experiments were performed in triplicate and repeated two to four times. We have previously reported an inhibitory effect of hPPARα on hTRs in HeLa cells(5Meier-Heusler S.C. Zhu X. Juge-Aubry C. Pernin A. Burger A.G. Cheng S.Y. Meier C.A. Mol. Cell. Endocrinol. 1995; 107: 55-66Google Scholar). In contrast to the dominant negative effect of mutant hTRβ1 from kindreds with thyroid hormone resistance, this effect was observed on all types of TREs, i.e. palindromic, inverted palindromic, and direct repeat arrangements of the half-sites. The present experiments were performed in the human hepatocarcinoma cell line HepG2 representing a tissue of high expression of both TR and PPAR(15Lemberger T. Staels B. Saladin R. Desvergne B. J. Wahli W. J. Biol. Chem. 1994; 269: Scholar, M. Miyamoto T. Fierrorenoy F. Macchia E. DeGroot L.J. J. Endocrinol. 1994; Scholar). the inhibition by PPAR of transcription of hTRβ1 on a results were on containing a or not However, this inhibitory effect of hPPARα was for no alteration of vitamin was observed on a element in the presence of of hPPARα (Fig. no in the transcriptional activity of hTRβ1 on a was present of were transfected (Fig. experiments of 35Smethionine-labeled hPPARα with unlabeled hTRβ1 in the presence of an antibody did not show evidence for the formation of of PPAR:TR heterodimers in (Fig. However, the precipitation of after with hTRβ1 the of the antibody to hTRβ1 in its using an anti-PPAR together with 35Smethionine-labeled hTRβ1 and no PPAR:TR heterodimers were (Fig. In order to for inhibitory other the formation of heterodimers which we were to we performed EMSA on a (Fig. The addition of a of hPPARα did not inhibit the formation and DNA binding of hTRβ1 with the of formation of heterodimers (Fig. However, hPPARα was able to the formation of TR:RXR heterodimers already at receptor that hPPARα with high for this of the TR:RXR was and the of observed and This that PPAR for for DNA binding. However, for that have not been the of the in not that observed in the of PPAR and RXR in is likely that the TR:RXR as well as possibly TR:RXR heterodimers the concentration to the formation of The of the on this element was previously by and other S.C. Zhu X. Juge-Aubry C. Pernin A. Burger A.G. Cheng S.Y. Meier C.A. Mol. Cell. Endocrinol. 1995; 107: 55-66Google Scholar, M. J. Biol. Chem. 1992; 267: Scholar, J. Biol. Chem. 1993; Scholar). The of the and of hTRβ1 on was confirmed by quantitatively both with the anti-TRβ1 not residue 433 of hPPARα mediates the interaction with interaction of wild-type and mutant hPPARα with by co-immunoprecipitation with an anti-PPAR the hPPARα wild-type 1 and 5 and and and were able to form heterodimers with the L433R mutant its to with RXR and the of PPAR:TR heterodimers with an anti-PPAR antibody 35Smethionine-labeled hTRβ1 was used of 4 of 35Smethionine-labeled are incubated with of in vitro translated wild-type or mutated PPARs and 2 ml anti-PPAR and no was volume is by preimmune serum and were precipitated by protein A-agarose and binding of wild-type and mutant hPPARα to the PPRE from in In the of the wild-type mutant PPARs were able to bind to DNA addition of PPAR wild-type and the mutant were of with the element and the repeat and P-box showed no binding. In all the volume of reticulocyte lysate was and with reticulocyte lysate where In order to whether hPPARα for for DNA binding, mutant hPPARα were created as in is in the P-box at the of the first zinc-finger (Fig. The amino acid sequence of the ligand-binding of hTRβ1 and hPPARα the presence of a highly conserved leucine zipper-like corresponding to the repeat in hTRβ1 (Fig. M. Helmer E. Casanova J. Raaka B.M. Samuels H.H. Mol. Cell. Biol. 1993; 13: 5725-5737Google Scholar, T. J. Biol. Chem. 1993; Scholar). In order to its in the of the leucine was mutated to an arginine To a effect of this mutation on the or protein of PPAR, a mutation was created amino acids The hPPARα were with to to with RXR in solution, to bind to and to through a PPRE and to modulate TR experiments with the anti-PPAR antibody the presence of heterodimers in or the and -L422R where In contrast, the to mutation at position 433 abolished the heterodimerization with (Fig. In the wild-type of the was able to with hTRβ1 (Fig. analyzed in an EMSA using the PPRE from the no binding to DNA was observed wild-type or mutant hPPARα was used (Fig. However, the wild-type and mutant to DNA RXR was present (Fig. and The was to to bind to DNA to its mutation in the DNA-binding domain the hPPARα-L433R was of with DNA to its in with RXR results that the leucine residue 433 of the leucine zipper motif in the ligand-binding domain of PPAR is for the heterodimerization with the cysteine residue 122 in the P-box is for binding to DNA the heterodimerization with RXR. The mutation to this putative leucine zipper repeat the formation of heterodimers DNA binding. The of mutant hPPARα was in a transfection for the to as well as to modulate TR on the PPRE from the wild-type hPPARα and showed transcriptional as from the EMSA experiments (Fig. The inhibitory effect of hPPARα-L433R and in the presence of ETYA on the activity of the was not analyzed In order to the of the mutant PPARs to inhibit TR experiments with hTRβ1 and mutant hPPARα were performed as in the leucine zipper mutant hPPARα-L433R its to down-regulate TR activity (Fig. the DNA mutant (Fig. and the mutant (Fig. efficient of that the leucine residue at position 433 of hPPARα is for cross-talk with by a mechanism heterodimerization with RXR. Besides structural and the to with TR and PPAR are both involved in the regulation of lipid We show that PPAR is able to selectively inhibit TR activity by for RXR. In that a highly conserved leucine zipper-like repeat in the ligand-binding domain for PPAR to with RXR. However, a similar mutation amino acids the of protein in the mutation L433R of transfected HepG2 cells show that all the PPAR are expressed to the wild-type protein not possibly for the action of PPAR on TR, such as the formation of PPAR:TR heterodimers the competition for DNA binding, were in the present In contrast to has been reported for and F. Hudson L.D. Nikodem V.M. J. Biol. Chem. 1994; 269: 11683-11686Google Scholar), we found no evidence for the formation of of heterodimers. the formation of PPAR:TR heterodimers it is not likely to quantitatively as is already from the The present results with the P-box mutant show that competition for DNA binding is not a mechanism for the inhibition of TR action by the was a efficient of thyroid hormone action, its dimerization with RXR was and binding to a has been by gel shift This of the mutant has not been analyzed but are in expression or The of TR by PPAR is highly the activity was not by However, the transcriptional activity of on RXR on the vitamin element PPAR to inhibit the formation of has a for RXR EMSA that PPAR the binding of heterodimers to but to a TR:RXR heterodimers not This together with the that the transfection of of RXR not TR activity in the presence of PPAR in of PPAR and with as non-RXR S.C. Zhu X. Juge-Aubry C. Pernin A. Burger A.G. Cheng S.Y. Meier C.A. Mol. Cell. Endocrinol. 1995; 107: 55-66Google Scholar). However, this putative is to with PPAR by means of a similar dimerization region as hPPARα-L433R the inhibitory effect of PPAR on A effect of PPAR on thyroid hormone in cell has been recently J. H. H. T. Biochem. Biophys. 1994; Scholar), a for this In the of thyroid hormone resistance, which is to dominant negative mutations in hTRβ1 the of a nuclear to modulate thyroid hormone action in a dominant negative in and in S. Usala S.J. 1993; Scholar). In TRs and PPARs are co-expressed in such as and brain, and both receptors regulate similar in lipid metabolism, such as the by enzyme and bifunctional The effect of PPAR on TR on the PPAR protein as well as its for the by and the possibly by as demonstrated for other nuclear T. Staels B. Saladin R. Desvergne B. J. Wahli W. J. Biol. Chem. 1994; 269: Scholar). In summary, we have demonstrated that PPAR is able to selectively inhibit the transcriptional activity of TRs by for RXR and possibly for a but as non-RXR In the results that a highly conserved leucine zipper-like motif in the ligand-binding domain of PPAR not for PPAR to with RXR. We are to F. Gonzalez Institute, and C. for the PPAR and expression and We B. Desvergne of for