Loading...
Page 2
Loading...
Page 3
Loading...
Page 4
Loading...
Page 5
Loading...
Page 6
Loading...
Page 7
Loading...
Page 8
Loading...
Page 9
Loading...
Page 10
Loading...
Page 11
Loading...
Page 12
/lp/unpaywall/mppar-gamma-2-tissue-specific-regulator-of-an-adipocyte-enhancer-cWQhMqbhrB
References
References for this paper are not available at this time. We will be adding them shortly, thank you for your patience.
- Publisher
- Unpaywall
- ISSN
- 0890-9369
- DOI
- 10.1101/gad.8.10.1224
- Publisher site
- See Article on Publisher Site
Abstract
Downloaded from genesdev.cshlp.org on October 23, 2021 - Published by Cold Spring Harbor Laboratory Press mPPAR /2: tissue-specific regulator of an adipocyte enhancer Peter Tontonoz, Erding Hu, Reed A. Graves, ~ Adriane I. Budavari, and Bruce M. Spiegelman e Dana-Farber Cancer Institute and the Department of Cell Biology, Harvard Medical School, Boston, Massachusetts 02115 USA Previously, we have isolated and characterized an enhancer from the 5'-flanking region of the adipocyte P2 (aP2) gene that directs high-level adipocyte-specific gene expression in both cultured cells and transgenic mice. The key regulator of this enhancer is a cell type-restricted nuclear factor termed ARF6. Target sequences for ARF6 in the aP2 enhancer exhibit homology to a direct repeat of hormone response elements (HREs) spaced by one nucleotide; this motif (DR-l) has been demonstrated previously to be the preferred binding site for heterodimers of the retinoid X receptor (RXR) and the peroxisome proliferator-activated receptor (PPAR). We have cloned a novel member of the peroxisome proliferator-activated receptor family designated mPPART2, and we demonstrate that a heterodimeric complex of mPPAR~/2 and RXR~ constitute a functional ARF6 complex. Expression of mPPAR,/2 is induced very early during the differentiation of several cultured adipocyte cell lines and is strikingly adipose-specific in vivo. mPPAR-/2 and RXRc~ form heterodimers on ARF6-binding sites in vitro, and antiserum to RXRc~ specifically inhibits ARF6 activity in adipocyte nuclear extracts. Moreover, forced expression of mPPAR~/2 and RXR~ activates the adipocyte-specific aP2 enhancer in cultured fibroblasts, and this activation is potentiated by peroxisome proliferators, fatty acids, and 9-cis retinoic acid. These results identify mPPAR~/2 as the first adipocyte-specific transcription factor and suggest mechanisms whereby fatty acids, peroxisome proliferators, 9-cis retinoic acid, and other lipids may regulate adipocyte gene expression and differentiation. [Key Words: Adipocyte transcription factor; adipocyte P2 enhancer; ARF6; peroxisome proliferator-activated receptor; retinoid X receptor; adipocyte differentiation] Received March 1, 1994; revised version accepted April 1, 1994. The murine adipocyte P2 (aP2) gene encodes an intra- The development of functionally and morphologically cellular lipid-binding protein and is expressed exclu- distinct cell types in higher eukaryotes results from the execution of an elaborate program of differential gene sively in adipose cells. We and others have been studying expression. In contrast with certain other cell types, rel- the transcriptional regulation of this gene in an effort to atively little is known about the regulatory mechanisms understand the molecular basis for its adipocyte-specific that direct adipocyte-specific gene expression. Although expression. Early studies demonstrated that the proxi- several factors, including CCAAT/enhancer-binding mal promoter region (168 bp of the 5' flank), which con- protein (C/EBPa)(Christy et al. 1989; Herrera et al. 1989) tains binding sites for transcription factors AP-1 and and the recently described basic helix-loop-helix C/EBP, could direct low-level differentiation-dependent (bHLH)-leucine zipper protein ADD1 (Tontonoz et al. gene expression in cultured 3T3-F442A adipocytes 1993), may play roles in the regulation of certain adipo- (Christy et al. 1989; Herrera et al. 1989) but could not cyte genes, transcription factors exhibiting true specific- support adipose expression in transgenic mice (Ross et ity for adipose cells have not been described. Several al. 1990). We subsequently searched farther upstream lines of evidence indicate that C/EBPot may play a role in and identified a 518-bp DNA fragment extending from terminal adipocyte differentiation (Freytag and Geddes -5.4 to -4.9 kb that could direct high level adipose- 1992; Umek et al. 1991). However, because C/EBPa specific expression of a chloramphenicol acetyl trans- mRNA is induced rather late in the time course of dif- ferase (CAT) reporter gene in transgenic mice as well as ferentiation and is expressed in liver, lung, kidney, and cultured cells (Ross et al. 1990; Graves et al. 1991). Since intestine, as well as fat, it is unlikely to play a substan- its identification, this enhancer has been used in trans- tial role in either the initiation of the adipocyte program genic animals to direct adipose-specific expression of a or the establishment of tissue specificity. number of genes that alter the biological function of this tissue, including SV40 large T antigen (Ross et al. 1992), diphtheria toxin A chain (Ross et al. 1993), and the in- ~Present address: Department of Medicine, University of Chicago, Chi- sulin-sensitive glucose transporter GLUT4 (Shepherd et c.ago, Illinois 60637 USA. 2Corresponding author. al. 1994). 1224 GENES & DEVELOPMENT 8:1224-1234 © 1994 by Cold Spring Harbor Laboratory Press ISSN 0890-9369/94 $5.00 Downloaded from genesdev.cshlp.org on October 23, 2021 - Published by Cold Spring Harbor Laboratory Press Adipocyte-specific transcription factor PPAR7 Multiple cis-acting elements within the 518-bp en- ARE6 5' hancer have been shown to be important for its differen- tiation-dependent activity in cultured 3T3-F442A prea- dipocytes and adipocytes. Initial studies identified a ARE7 5, binding site (ARE1) for a member of the NF-1 transcrip- tion factor family. This NF-1 site appears to contribute to the overall activity of the enhancer, but it is not func- tional as an isolated element, and the enhancer retains differentiation-dependent activity in the absence of this ARE7: GTGAACTCTGATCCAG site (Graves et al. 1991). Further analysis established the existence of four additional cis-acting elements, each of mutation: cacttgagactaggtc which is also required for full enhancer activity in tran- sient transfection assays (Graves et al. 1992). One pair of -1 relative elements, ARE2 and ARE4, binds an apparently ubiqui- activity: tous positive-acting factor termed ARF2. A second pair -2 of elements, ARE6 and ARE7, binds a separate nuclear factor, termed ARF6, that is detected only in nuclear Figure 1. Definition of a functional ARF6-binding site. {A} Nu- extracts derived from adipocytes. Multiple copies of ei- cleotide sequence comparison of the ARE6 and ARE7 sites from the aP2 enhancer. Previously identified regions of homology are ther ARE6 or ARE7 activate transcription only in differ- shaded. DNA-protein contact points identified by methylation entiated adipocytes. The observation that the ARF6- interference analysis are indicated by asterisks (*}. (B) Muta- binding sites are both necessary and sufficient for adipo- tional analysis of the ARE7 site. The effect of individual trans- cyte-specific expression suggests that ARF6 functions as version mutations on ARF6 binding is diagramed schemati- a differentiation-dependent and tissue-specific switch for cally. the aP2 enhancer. In this paper we report the identification and cloning of a nuclear hormone receptor, mPPAR72, that appears posed of two HRE half-sites, arranged with a particular to be an important component of the ARF6 complex. orientation and spacing. Various arrangements of HREs Expression of mPPAR~/2 is highly specific for adipose have been shown to preferentially recognize different nu- tissue and is sufficient to activate the adipocyte-specific clear hormone receptor combinations (Umesono et al. aP2 enhancer in heterologous cell types. We suggest that 1991; Kliewer et al. 1992b). The ARF6 recognition se- this novel peroxisome proliferator-activated receptor quence can be interpreted as an imperfect version of a (PPAR) family member may function as a key regulator type of nuclear hormone receptor-binding site known as of adipogenic gene expression. DR-1 (direct repeat with 1-nucleotide spacer). This motif has been shown to preferentially bind homodimers of retinoic acid X receptor (RXR) (Mangelsdorf et al. 1991) Results or the liver-restricted transcription factor HNF4 (Sladek et al. 1990), heterodimers of RXR and COUP-TF (Kliewer ARF6-binding sequences in the aP2 enhancer are DR-l- et al. 1992a), and heterodimers of RXR and the PPARs type hormone response elements (Kliewer et al. 1992c). An alignment of the consensus We have used partially purified ARF6 derived from adi- DR-1 site with ARE6 and ARE7 is shown in Figure 2A. pocyte nuclear extracts to characterize the DNA recog- DNA mobility retardation experiments using various nition sequence for this factor in detail. Initial compar- HRE sequences as competitor demonstrated that ARF6 ison of the two binding sites for ARF6 in the aP2 en- preferentially recognizes DR-1 sites. As shown in Figure hancer (ARE6 and ARE7, Fig. 1A) identified a 10- out of 2B, the consensus DR-1 sequence and the fatty acyl-CoA 12-bp identity (Graves et al. 1992). Methylation interfer- oxidase peroxisome proliferator response element (PPRE) ence analysis performed on the ARE6 and ARE7 sites are optimal competitors for ARF6 binding to the ARE7 indicated that the actual region of DNA-protein inter- site. action extends significantly 5' to the region of obvious homology. Guanosine residues identified as sites of RXRa is a component of the ARF6 complex DNA-protein contact are indicated by asterisks in Fig- ure 1A. Extensive mutational analysis of the ARE7-bind- The above data strongly suggest that at least one com- ing site further defined the bases important for ARF6 ponent of ARF6 is a member of the nuclear hormone binding. The effect of individual transversion mutations receptor superfamily. To determine whether the ARF6 on ARF6 binding to the ARE7 site is diagramed schemat- complex contained proteins known to bind DR-1 sites, ically in Figure lB. we obtained antisera to several such factors. DNA mo- This functional analysis of the ARE7 site reveals a bility retardation assays using labeled ARE7 oligonucle- homology between the ARF6 recognition sequence and otide as probe were performed with nuclear extract and the consensus nuclear hormone response element (HRE) partially purified ARF6 in the presence of these antisera. half-site TGAA/cCT. Binding sites for members of the As shown in Figure 3, antisera to the retinoic acid recep- nuclear hormone receptor superfamily are typically com- tor o~ (RARcx) and to murine PPARcx have no effect on GENES & DEVELOPMENT 1225 III M..I.I I III Downloaded from genesdev.cshlp.org on October 23, 2021 - Published by Cold Spring Harbor Laboratory Press Tontonoz et al. RXRo~, because such dimers bind to DNA relatively poorly and bind only in the presence of the ligand 9-cis- acting retinoic acid (Zhang et al. 1992). Although the antisera supershift data above indicated that mPPARa DRI TGAACTnTGAAC~ C C was not present in the ARF6 complex, three distinct iiiii~/~ PPAR family members with different tissue distribu- ARE7 tions have been described in Xenopus, including one ARE6 TTGCACATTT.CA~CCAG (xPPART) that is expressed at high levels in the fat body i:!i: : i!: (Dreyer et al. 1992). We hypothesized that the partner for PPRE RXRo~ in the ARF6 complex might be an adipose-ex- pressed member of the PPAR family. We therefore de- TRE AGGTCATGACC~ signed primers for the polymerase chain reaction (PCR) to specifically amplify sequences encoding RXR- and ~RARE TGAACTTTTCGGTGAACC PPAR-related receptors from adipocyte cDNA (see Ma- terials and methods). Products from these PCR reactions were subcloned and sequenced. Of 20 clones isolated, 2 encoded mPPARot, three encoded the murine homolog of DR1 PPRE TRE RARI~ NS ARE7 competitor: the PPAR family member NUC-1, and 15 were found to 1 I I I I [~l I I I '1 encode a novel member of the PPAR family subse- 5 20 S 20 5 20 5 20 5 20 fold excess: 5 20 quently named PPAR72. Full-length mPPAR72 cDNA clones were subsequently isolated from an adipocyte hZaplI cDNA library using the PCR-amplified fragment as probe. The DNA and predicted amino acid sequence of the longest of these clones is shown in Figure 4. It poten- tially encodes a 505-amino-acid protein with a predicted molecular mass of 55.6 kD. The 5'-most ATG codon in this sequence is in an appropriate context for transla- tional initiation. Primer extension experiments indi- cated that there are only an additional 25 bp 5' in the mRNA that are not present in this clone (data not shown). Data base searches revealed that the receptor encoded by this cDNA is a murine homolog of Xenopus Figure 2. Relationship of ARF6-binding sites to HREs. {A) Se- quence comparison of ARE6 and ARE7 with the consensus DR-1 HRE, fatty acyl-CoA oxidase PPRE, palindromic thyroid protein: N. extract purified I i J~l hormone response element (TREp), and the retinoic acid re- PPRu PPRa sponse element from RAR~ ([3RARE). (B) ARF6 preferentially Antisera: - RXR(~ RXRa RARu RARa binds DR-l-type HREs. Double-stranded radiolabeled ARE7 oli- gonucleotide was used as probe in a DNA mobility retardation assay of partially purified ARF6 from adipocyte nuclear extract. DNA protein complexes were resolved from free DNA on a 5% polyacrylamide gel. The gel was dried and exposed to film for 2 hr at - 70°C. The HREs shown in A were used as competitors in the binding reaction in 5- and 20-fold molar excess as indicated. The ARE2 site from the aP2 enhancer was used a nonspecific competitor (NS). (See Materials and methods for complete se- quence of the oligonucleotides used.) ARF6 binding. Antisera to RXRc~, however, specifically inhibit ARF6 complex formation and give a characteris- tic "supershift" when incubated with either nuclear ex- tract or partially purified ARF6. This antisera were raised Figure 3. RXRa is a component of the ARF6 complex. Double- to a single peptide specific to RXRoL (Kliewer et al. stranded radiolabeled ARE7 oligonucleotide was used as probe in a DNA mobility retardation assay of adipocyte nuclear ex- 1992c}. Thus, at least one of the polypeptides comprising tract and partially purified ARF6. Antisera to RARa, RXRa, or ARF6 is RXR~ itself or a very closely related molecule. PPARa were included in the binding reaction as indicated. DNA protein complexes were resolved from free DNA on a 5% poly- Cloning of a novel PPAR from adipose tissue acrylamide gel. The gel was dried and exposed to film for 2 hr at The ARF6 complex is unlikely to be a homodimer of - 70°C. 1226 GENES & DEVELOPMENT !!!!!!!i!i!!!:i!;;i!i!~!!!;!!ii!!i!~!il!!i!!!!iiii~!iiiiii!il ~il,;ii~iiil;;~il;i!!iii!iiiiiii,iiiiiii~iii~i;!!!ii!!!!il :i!!:iiii!iiiiiiii!i'ii:iii~i~ii'iiiiii!i!iii!iiiiiiii~?!ii~!~i!il Downloaded from genesdev.cshlp.org on October 23, 2021 - Published by Cold Spring Harbor Laboratory Press Adipocyte-specific transcription factor PPAR7 CCC AGT GTG AAT TAC AGC AAA TCT CTG TTT TAT GCT GTT ATG GGT GAA ACT CTG GGA GAT TCT CCT GTT GAC 72 M G E T L G D S P V D ii CCA GAG CAT GGT GCC TTC GCT GAT GCA CTG CCT ATG AGC ACT TCA CAA GAA ATT ACC ATG GTT GAC ACA GAG 144 P E H G A F A D A L P M S T S Q E I T M V D T E 35 ATG CCA TTC TGG CCC ACC AAC TTC GGA ATC AGC TCT GTG GAC CTC TCC GTG ATG GAA GAC CAC TCG CAT TCC 216 M P F W P T N F G I S S V D L S V M E D H S H S 59 TTT GAC ATC AAG CCC TTT ACC ACA GTT GAT TTC TCC AGC ATT TCT GCT CCA CAC TAT GAA GAC ATT CCA TTC 288 F D I K P F T T V D F S S I S A P H Y E D I P F 83 AcA A0A GCT GAC tea A TG ~ GC~ ~AT T AC AM ~AT ~C C~G ~0 ~ CAA ~ TAC ~ A~ ~0 A~C A~ 360 T R A D P M V A D Y K Y D L K L Q E Y Q S A I K 107 V E P A S P P Y Y $ E K T Q L Y N R P H E E P S 131 AAC TCC CTC ATG GCC ATT GAG TGC CGA GTC TGT GGG GAT AAA GCA TCA GGC TTC CAC TAT GGA GTT CAT GCT 504 N S L M A I E C R V C G D K A S G F H Y G V H A 155 TGT GAA GGA TGC AAG GGT TTT TTC CGA AGA ACC ATC CGA TTG AAG CTT ATT TAT GAT AGG TGT GAT CTT AAC 576 C E G C K G F F R R T I R L K L I Y D R C D L N 179 C R I H K K S R N K C Q Y C R m Q K C h i V G M 203 TCT CAC AAT GCC ATC AGG TTT GGG CGG ATG CCA CAG GCC GAG AAG GAG AAG CTG TTG GCG GAG ATC TCC AGT 720 S H N A I R F G R M P Q A E K E K h L A E I S S 227 GAT ATC GAC CAG CTG AAC CCA GAG TCT GCT GAT CTG CGA GCC CTG GCA AAG CAT TTG TAT GAC TCA TAC ATA 792 D I D Q L N P E S A D L R A L A K H L Y D S Y I 251 AAG TCC TTC CCG CTG ACC AAA GCC AAG GCG AGG GCG ATC TTG ACA GGA AAG ACA ACG GAC AAA TCA CCA TTT 864 K S F P L T K A K A R A I L T G K T T D K S P F 275 GTC ATC TAC GAC ATG AAT TCC TTA ATG ATG GGA GAA GAT AAA ATC AAG TTC AAA CAT ATC ACC CCC CTG CAG 936 V I Y D M N S L M M G E~ D K I K F K H I T P L Q 299 GAG CAG AGC AAA GAG GTG GCC ATC CGA ATT TTT CAA GGG TGC CAG TTT CGA TCC GTA GAA GCC GTG CAA GAG i008 E Q S K E V A I R I F Q G C Q F R S V E A V Q E 323 ATC ACA GAG TAT GCC AAA AAT ATC CCT GGT TTC ATT AAC CTT GAT TTG AAT GAC CAA GTG ACT CTG CTC AAG 1080 I T E Y A K N I P G F I N L D L N D Q V T L L K 347 TAT GGT GTC CAT GAG ATC ATC TAC ACG ATG CTG GCC TCC CTG ATG AAT AAA GAT GGA GTC CTC ATC TCA GAG 1152 Y G V H E I I Y T M L A S L M N K D G V L I S E 371 GGC CAA GGA TTC ATG ACC AGG GAG TTC CTC AAA AAC CTG CGG AAG CCC TTT GGT GAC TTT ATG GAG CCT AAG ]224 G Q G F M T R E F L K N L R K P F G D F M E P K 395 TTT GAG TTT GCT GTG AAG TTC AAT GCA CTG GAA TTA GAT GAC AGT GAC TTG GCT ATA TTT ATA GCT GTC ATT 1296 F E F A V K F N A L E L D D S D L A I F I A V I 419 ATT CTC AGT GGA GAC CGC CCA GGC TTG CTG AAC GTG AAG CCC ATC GAG GAC ATC CAA GAC AAC CTG CTG CAG 1368 I L S G D R P G L L N V K P I E D I Q D N L L Q 443 GCC CTG GAA CTG CAG CTC AAG CTG AAT CAC CCA GAG TCC TCT CAG CTG TTC GCC AAG GTG CTC CAG AAG ATG 1440 A L E L Q L K L N H P E S S Q L F A K V L Q K M 467 ACA GAC CTC AGG CAG ATC GTC ACA GAG CAC GTG CAG CTA CTG CAT GTG ATC AAG AAG ACA GAG ACA GAC ATG 1512 T D L R Q I V T E H V Q L L H V i K K T E T D M 491 Figure 4. Nucleotide and corresponding AGC CTT CAC CCC CTG CTC CAG GAG ATC TAC AAG GAC TTG TAT TAG CAGGAAAGTCCCACCCGCTGACAACGTGTTCCTTC 1592 amino acid sequence of the mPPAR72 S L H P L L Q E I Y K D L Y * 505 cDNA. The putative polyadenylation sig- TATTGATTGCACTATTATTTTGAGGGAAAAAAATC TGACAC C TAAGAAATTTAC TGTGAAAAAGCATTTAAAAACAAAAGTTTTAGAAGATGATG 1687 nal and initiation codons are shown in ATCTATTTTATGCATATTGTTTATAAAGATACATTTACAATTTACTTTTAATATTAAAAATTACCACATTATAAAATTAAAAAAAAAAAAAAA 1780 boldface type. PPAR7 (Dreyer et al. 1992). A cDNA encoding a different mPPAR 7 mRNA is expressed at high levels specifically isoform of this PPAR was described while this manu- in adipose tissue script was in preparation (Zhu et al. 1993). We have therefore designated this novel isoform mPPAR72. The It has been demonstrated previously that ARF6-binding mPPAR~/2 cDNA encodes an additional 30 amino acids activity is differentiation-dependent in cultured adipo- amino-terminal to the first ATG codon of mPPAR~/1 and cytes (Graves et al. 1992). Because RXRa mRNA is present in both preadipocytes and adipocytes (data not has a different 5'-untranslated sequence (diagramed shown), it was likely that another component of the schematically in Fig. 5A). The ATG codon at position 31 represents the 5'-most ATG codon in the PPART1 ARF6 complex would exhibit differentiation-dependent expression. We therefore examined the time course of eDNA. In vitro translation of the mPPAR72 cDNA yields two major polypeptide species of relative molec- mPPAR7 expression during the differentiation of cul- ular mass 56 and 52 kD (Fig. 5B, mPPAR72), indicating tured 3T3-F442A and 3T3-L1 preadipocytes. 3T3- that the ATG codons at positions 1 and 31 can both F442A cells differentiate spontaneously at confluence in function as translational initiators. Translation of presence of insulin and permissive serum, whereas dif- the PPAR~/1 cDNA yields only the 52-kD species ferentiation of 3T3-L 1 cells requires induction with me- (mPPART1). We sequenced the 5' ends of five separate thylisobutylxanthine and dexamethasone (see Materials cDNA clones, all of which encoded the mPPAR72 iso- and methods). Cultured adipocytes derived from these form, suggesting it is the predominant form expressed in cell lines accumulate lipid and express most of the genes adipose tissue (see below). characteristic of adipose cells in vivo (Green and Ke- GENES & DEVELOPMENT 1227 Downloaded from genesdev.cshlp.org on October 23, 2021 - Published by Cold Spring Harbor Laboratory Press Tontonoz et al. 71 72 Figure 5. Structure of mPPAR~ isoforms. (A) Schematic representation of the mP- PAR~ 1 and mPPAR~2 cDNAs. Identical se- quences are indicated by similar shading. A 56kDa ATG (B) In vitro translation of the mPPAR-yl and I DNA LIGAND -- 52 kDa 5' UTR / 3' UTR mPPAR~2 cDNAs, mPPAR-~I-SPORT and mPPAR~ 2.1 kb mPPAR~2-SPORT plasmids were trans- lated by rabbit reticulocyte lysate in the ATG ATG presence of L[3SS]methionine. Translation products were resolved on an 9.0% SDS- 1.8 kb mPPAR'~2 polyacrylamide gel. hinde 1974). As shown in Figure 6, the mPPAR~ mRNA in adipocytes migrates as a single species of 2.0 kb. Al- though the mPPAR~ eDNA probe used for Northern A 3T3-F442A analysis hybridizes to both the mPPAR~ 1 and mPPAR~/2 days postconfluence: 0 1 2 3 5 7 9 11 isoforms, S1 nuclease analysis indicates that the mP- PAR~ mRNA present in adipose cells encodes almost PPARy exclusively the mPPAR~2 isoform (data not shown), mP- PAR~2 mRNA levels are induced very early (1-2 days postconfluence) in the course of differentiation of both LPL 3T3-F442A and 3T3-L 1 preadipocytes. This induction is coincident with the induction of lipoprotein lipase (LPL), one of the earliest markers of adipocyte differentiation, and clearly precedes the induction of aP2. The tissue distribution of mPPARy and RXRa was ex- aP2 amined using RNA prepared from a variety of mouse tissues. RXRoL mRNA has been demonstrated previously to be expressed in a number of tissues (Mangelsdorf et al. 1992). Figure 7A (bottom) shows that that RXRo~ expres- 36B4 sion is especially prominent in adipose tissue and liver. In contrast, mPPAR~ is expressed at very high levels B 3T3-L1 only in adipose tissue [(F) fat, Fig. 7A, (top)]. As noted days post induction: 12357911 above, this mRNA encodes predominently the PPAR~2 isoform. Expression is at least 20-fold higher in adipose rnPPARy tissue than all other tissues examined. A prolonged ex- posure of this blot also reveals minimal expression in heart, skeletal muscle, kidney, and liver (Fig. 7B, top). At this level of signal, however, one must be concerned LPL about unavoidable contamination of these tissues with adipose cells. To address this issue, this same blot was probed with the cDNA for aP2, which has been shown previously to be expressed exclusively in fat (Bernlohr et aP2 al. 1984). Figure 7B (bottom) demonstrates that the heart, kidney, and skeletal muscle samples do contain signifi- cant amounts of aP2 mRNA, suggesting contamination with small numbers of adipose cells. It is therefore likely 36B4 that even some of the minimal mPPAR~/expression ob- served in nonadipose tissues actually results from fat cells present in these tissues. Northern analysis of rat Figure 6. Regulation of mPPAR~/ mRNA in two cell culture tissue RNA indicates that PPAR~2, like aP2, is expressed models of adipocyte differentiation. Total RNA (10 ~g per lane) in both white and brown adipose tissue (data not shown). was isolated, blqtted to nylon, and hybridized with 32p-labeled mPPAR~/2 cDNA as described in Materials and methods. Equiv- alent amounts of intact RNA were run in each lane as indicated rnPPAR 72/RXRa heterodimers bind to ARF6-binding by hybridization to a 36B4 cDNA probe. (A) Time course of sites in the aP2 enhancer differentiation of 3T3-F442A preadipocytes; (B) Time course of To determine whether PPAR~/2 could bind to ARF6- differentiation of 3T3-L1 preadipocytes. 1228 GENES & DEVELOPMENT Downloaded from genesdev.cshlp.org on October 23, 2021 - Published by Cold Spring Harbor Laboratory Press Adipocyte-specific transcription factor PPAR~/ complex before differentiation, in vitro-translated pro- teins were added to preadipocyte nuclear extract. No Br F H In Ki Li M P Sp complex is formed on ARE7 with preadipocyte nuclear extract plus in vitro-translated RXR~ (lane 9). However, mPPARy the addition of in vitro-translated mPPAR~/2 to preadi- 2.0 kb 6hm pocyte nuclear extract generates an ARF6-1ike complex (lane 10), and this complex is disrupted by antisera to RXRet {lane 11). These data strongly suggest that mPPAR~/2 is the stoichiometrically limiting component of the ARF6 complex in preadipocytes. RXRG 5.6 kb 24 hrs PPAR 72 activates the aP2 enhancer through ARF6- binding sites The above data demonstrate that the DNA-binding ac- tivity of the PPAR~/2/RXRet heterodimer is virtually BrF H In Ki Li M P Sp identical to that of the adipocyte factor ARF6. Because ARF6 has been suggested previously to be the key regu- lator of the aP2 enhancer (Graves et al. 1992), we inves- mPPARy 2.0 kb tigated whether PPAR~/2 and RXRot could activate the 48 hrs aP2 enhancer in nonadipose cells. Each of these cDNAs was introduced into pSV-SPORT, an expression vector driven by the SV40 enhancer/promoter (see Materials and methods), and transient transfection experiments were performed in NIH-3T3 cells. As shown in Figure 9, aP2 = 0.9 kb cotransfection of either the RXRet expression vector (3.5- 4 hrs fold activation above basal, lane 2) or the mPPAR~/2 ex- pression vector (4-fold activation above basal, lane 3) stimulated transcription of a CAT reporter construct Figure 7. mPPAR~/ is expressed with adipose specificity. (A) containing the 518-bp aP2 enhancer linked to the SV40 mPPAR~/, 6-hr exposure; RXRet, 24-hr exposure. (B) mPPAR% enhancerless promoter (Graves et al. 1992). Transfection 48-hr exposure; aP2, 4-hr exposure. Total RNA (10 ~g per lane) was isolated from adult mouse tissues, blotted to nylon, and hybridized with 3~P-labeled mPPAR~/2 or RXRa eDNA as de- scribed in Materials and methods. Note: The mPPAR-y2 probe ARF6: + recognizes both the mPPAR-yl and mPPAR~/2 isoforms. (B) antisera: RX PPa RX PPa Brain; (F) epidymal white fat; (H) heart; (In) intestine; (Ki) kid- PreAdextract: + + + + + ney; (Li} liver; (MI skeletal muscle; (PI pancreas; (Spl spleen. PPAR~2: + + + + + + + Equivalent amounts of intact RNA were run in each lane as RXRa: + + + + + indicated by ethidium bromide staining of the membrane after transfer and hybridization to a 36B4 eDNA probe (not shown). binding sequences from the aP2 enhancer, DNA mobil- ity retardation assays were performed with in vitro- translated protein and labeled ARE7 oligonucleotide. Be- cause RXRs have been demonstrated to be required cofactors for binding of PPARs to DNA (Keller et al. 1993; Kliewer et al. 1992c), RXRa was analyzed in par- allel. As shown in Figure 8, neither RXR~ nor mPPAR~2 binds independently to ARE7 (lanes 3,4); however, a prominent complex is formed when both mPPAR~/2 and 1 2 3 4 5 6 7 8 9 10 11 12 RXR~ are present in the binding reaction (lane 5). This Figure 8. mPPAR~/2 and RXRoL bind to the ARE7 site from the complex migrates with precisely the same mobility as aP2 enhancer as a heterodimer. Double-stranded 3ZP-labeled the ARF6 complex from adipocyte nuclear extract (lane ARE7 oligonucleotide was used as probe in a DNA mobility 1), and, as expected, complex formation can be inhibited retardation assay with partially purified ARF6, preadipocyte nu- by antisera to RXR~ but not by antisera to mPPAR~ clear extract, and in vitro-translated mPPAR~/2 and RXRa. An- (lanes 6,7]. As has been demonstrated previously (Graves tisera to RXRoL (RXet) or mPPARot (PPoL) were included in the et al. 1992), ARF6-binding activity is not present in prea- binding reaction where indicated. DNA protein complexes were dipocyte nuclear extracts (lane 8 ). To determine whether resolved from free DNA on a 5% polyacrylamide gel. The gel RXRa or mPPAR~2 is the limiting component of the was dried and exposed to film for 1 hr at -70°C. GENES & DEVELOPMENT 1229 Downloaded from genesdev.cshlp.org on October 23, 2021 - Published by Cold Spring Harbor Laboratory Press Tontonoz et al. ETYA: + + + family members have been shown to be activated by fatty acids (Keller et al. 1993), we investigated whether 9-Cis-RA: + + + mPPAR~2 could be similarly activated. Figure 10 dem- receptor: -- RXR PPRy both -- RXR PPRy both -- both both onstrates that mPPAR~/2 can be effectively activated by linoleic acid (lane 3) and by the peroxisome proliferator clofibric acid (lane 4), as well as by ETYA. No significant activation is observed when the basal pSVKS1 CAT con- struct is used as the reporter (lanes 8,9). These data dem- onstrate that mPPAR~/2 can function as a dominant, activator-dependent regulator of an adipose-specific en- hancer, presumably through formation of a hetero- 1 2 3 4 5 6 7 8 9 10 11 I II I dimeric DNA-binding complex with RXRet. aP2 enhancer CAT pSV CAT Figure 9. mPPAR~/2 and RXRa activate the aP2 enhancer in Discussion nonadipose cells. NIH-3T3 cells were cotransfected with either the aP2 enhancer CAT contruct or the parental pSVKS1 CAT The PPARs comprise a subfamily of nuclear hormone construct (2 ~g) and mPPAR~2 and/or RXR~ expression vector receptors first identified by virtue of their ability to be (2 lag) as described in Materials and methods. After transfection, activated by certain hypolipidemic drugs, plasticizers, cells were treated for 24 hr with 1 ~M 9-cis retinoic acid or 1 ~M and herbicides that cause proliferation of peroxisomes in ETYA as indicated. The level of CAT gene expression resulting rodent liver (Isseman and Green 1990). Such compounds from each transfection was determined by measuring CAT en- are also known to induce enzymes of the peroxisomal zyme activity. fatty acid [3-oxidation system. PPARs have subsequently been shown to directly activate transcription of the acyl- CoA oxidase gene, which catalyzes the rate-limiting step of both constructs simultaneously had an apparently in the [3-oxidation pathway (Dreyer et al. 1992). In addi- synergistic effect (15-fold activation above basal, lane 4). tion to the original murine PPAR (mPPAR~), three It has been shown previously that the transcriptional PPARs have been identified in Xenopus (xPPAR ~,[3, and activating function of both RXRs and PPARs is strongly ~), one in rat (rPPAR, an a homolog), and two in human dependent on the presence of specific ligands and/or ac- (hPPARc~ and hPPAR/NUC-1, a [3 homolog). We have tivators (Keller et al. 1993, Kliewer et al. 1992c). Tran- identified a novel murine PPAR termed mPPAR~2 that scriptional activation of the aP2 enhancer by RXRa is is expressed at high levels in adipose tissue. A different enhanced when the cells are treated with its high-affin- isoform of mPPAR% which we propose be referred to as ity ligand 9-cis retinoic acid (lane 6); activation by mPPAR~/1, has recently been described independently by PPAR~2 is enhanced when cells are treated with Zhu et al. (1993). It remains to be determined whether 5,8,11,14-eicosatetraynoic acid (ETYA, lane 7), a syn- thetic arachadonic acid analog and potent activator of PPAR family members. Simultaneous addition of both ETYA: + + + receptors and their activators has a maximal effect (lane clofibric: + 8). No significant activation of the basal SV40 promoter linoleic: + CAT construct (pSV CAT) is observed even in the pres- 9-cis-RA: + + + ence of both activators (lanes 9-11). both receptor: PPRy RXR PPRy PPRy PPRy both The 518-bp aP2 enhancer is a complex regulator con- taining multiple cis-acting elements. The above data demonstrate that mPPAR~/2 and RXRot can dominantly activate the aP2 enhancer but do not establish that this activation is mediated by the ARF6-binding sites. We therefore tested the ability of mPPAR~/2 and RXRa to activate transcription of a CAT construct containing multiple copies of an isolated ARF6-binding sequence. I 2 3 4 5 6 7 8 9 We have shown previously that mutiple copies of these I II sequences (ARE6 and ARE7) stimulate adipocyte-spe- pSV CAT ARE7 CAT cific expression of a CAT reporter gene in cultured cells Figure 10. mPPAR~2 and RXRa activate transcription through (Graves et al. 1992). Figure 10 demonstrates that RXRa the ARE7 site from the aP2 enhancer. NIH-3T3 cells were and mPPAR~/2 can each independently stimulate tran- cotransfected with either the ARE7 CAT contruct or the paren- scription of the ARE7/CAT reporter construct in the tal pSVKS1 CAT construct (2 ~g) and mPPAR~2 and/or RXRc~ presence of their appropriate activators (lanes 2,5). expression vector (2 ~g). After transfection, cells were treated Again, simultaneous expression of both receptors yields for 24 hr with 1 ~M 9-cis retinoic acid, 50 ~M linoleic acid, 50 a maximal effect (lane 6). Previous studies have sug- ~M clofibric acid, or 1 tzM ETYA as indicated. The level of CAT gested that fatty acids may stimulate transcription of gene expression resulting from each transfection was deter- aP2 in adipocytes (Amri et al. 1991). Because other PPAR mined by measuring CAT enzyme activity. 1230 GENES & DEVELOPMENT Downloaded from genesdev.cshlp.org on October 23, 2021 - Published by Cold Spring Harbor Laboratory Press Adipocyte-specific transcription factor PPARv each isoform has a specific biological role. Multiple iso- ykinase (PEPCK) gene, which is expressed in a differen- tiation-dependent manner in adipose cells (P. Tontonoz, forms resulting from alternative promoter usage and dif- E. Beale, and B.M. Spiegelman, unpubl.}. This binding ferential splicing have been described for numerous site is located within a region demonstrated previously other nuclear hormone receptors, including the proges- to be important for adipose expression in transgenic terone receptor and RARe, RARf~, and RARe/(Kastner et mice (Short et al. 1992j. Our definition of the ARF6 rec- al. 19901 Chambon et al. 1991). Because mPPAR~/1 and mPPAR~/2 have different 5'-untranslated sequences, ognition sequence as a DR-l-type HRE should facilitate their mRNAs are most likely transcribed from different the identification of ARF6-binding sites in regulatory re- promoters. While mPPAR~/1 and mPPAR~/2 represent gions of other adipocyte genes as well. the first examples of isoforms in the PPAR family, it is Identification of the endogenous ligand(s) for likely that multiple isoforms will be identified for other mPPAR~/2 will be an important goal of future studies. PPARs as well. Several groups have demonstrated previously that poly- unsaturated fatty acids and arachadonic acid are endog- Previously, we identified a trans-acting factor termed enous activators of PPARs; however, it is not yet clear ARF6 as the key factor controlling the fat-specific ex- pression of the aP2 gene (Graves et al. 1992). In the whether such molecules actually bind to PPARs directly present work we demonstrate that mPPAR~2 forms het- (Gottlicher et al. 1992; Keller et al. 1993). Previous ob- servations that the ARE6 and ARE7 sites of the aP2 en- erodimers with RXRo~ and that the DNA-binding and hancer are constitutively active in adipocytes suggest transcriptional activation properties of the mPPAR~/2/ RXR~ complex are identical to those of the adipocyte that either mPPAR~/2 is capable of ligand-independent transcription factor ARF6. The mPPARv2/RXRa hetero- activation in vivo or that the endogenous ligand for dimer binds to ARF6 target sequences in vitro and can mPPARv2 is normally present at significant levels in activate expression of the fat-specific aP2 enhancer in differentiated adipocytes. Studies have shown that fatty nonadipose cells in transient transfections. Maximal ac- acids can induce the expression of aP2 by both transcrip- tivation is observed when both receptors and their acti- tional and post-transcriptional mechanisms (Amri et al. vators are present, suggesting that the mPPAR~/2/RXRe~ 1991, Distel et al. 1992). Our demonstration here that heterodimer is the functional ARF6 complex in vivo. fatty acids can activate mPPAR~/2 suggests that While all members of the PPAR family described thus mPPAR~/2 is likely to mediate the transcriptional acti- vation of aP2 by fatty acids in vivo. far have identical DNA recognition sequences, and thus The observation that liver- and kidney-expressed would all be expected to bind to ARF6 sites in vitro, several lines of evidence strongly suggest that mPPAR~/2 PPARs can be activated by fatty acids correlated well in particular is a major regulator, if not the dominant with the finding that high fat diets can induce the per- regulator of the aP2 enhancer in adipocytes. First, mP- oxisomal B-oxidation system (Osmundsen et al. 1991). PAR~/ mRNA is very abundant as determined by both PPARs have therefore been postulated to be responsible Northern blotting and the screening of cDNA libraries. for activation of a safeguard system of non-energy-gen- Second, it exhibits adipose cell specificity consistent erating fatty acid oxidation in response to physiologic with the expression of ARF6 and the aP2 mRNA itself. lipid overload (Keller and Wahli 1993). Our data suggest mPPAR~/mRNA is at least 20 times more abundant in that certain members of the PPAR family may play a white and brown fat than any other tissue. Third, the more central physiologic role as key regulators of gene mPPAR~/mRNA expressed in adipose tissue encodes al- expression in adipocytes, an important cell type respon- most exclusively the mPPAR~/2 isoform. Fourth, mP- sible for lipid storage and the maintenance of energy bal- PAR~/2 forms a complex with RXR~ in vitro that mi- ance. The identification of a PPAR family member that grates with precisely the same mobility in electropho- exhibits specificity for adipose cells and functions as a retic DNA-binding assays as ARF6 from cultured key regulator of an adipocyte enhancer suggests a mo- adipocytes. Finally, the other two known murine PPARs, lecular link between lipid homeostasis and adipose de- mPPAR~ and mNUC-1, are expressed at much lower velopment. Obviously, lipid metabolism must be linked levels in adipose tissue and are not tissue-specific (Isse- to fat cell differentiation in that marked overfeeding man and Green 1990; E. Hu, unpubl.). While the exis- leads to increased adiposity in vertebrate animals. This tence of additional adipose-specific PPAR family mem- increase in adipose tissue mass results largely from de bers cannot be ruled out, extensive PCR analysis of fat novo differentiation of new adipocytes from preadipo- cell mRNA with oligonucleotides designed to amplify cytes (Aihaud et al. 1992). If fatty acids or other lipids such molecules has failed to identify additional PPARs. derived from the diet can serve to activate mPPAR~/2, Preliminary analysis of peptide sequences derived from this may stimulate adipogenic gene expression or even affinity-purified ARF6 preparations indicates that a ma- fat cell differentiation per se. Although this paper pre- jor polypeptide present in this material is PPAR~/ (P. sents no data that directly implicate mPPAR~/2 in the Tontonoz, unpubl.). regulation of the differentiation process, PPAR activa- If mPPAR~2 plays a broad role in the regulation of tors such as the fibrate hypolidipemic drugs have been adipogenesis, it will likely be implicated in the regula- reported previously to induce and/or potentiate the dif- tion of other adipocyte-specific genes in addition to aP2. ferentiation of certain preadipocyte cell lines (Brandes et We have recently identified an ARF6-binding site in the al. 1987; Gharbi-Chihi et al. 1993). It is reasonable to 5'-flanking region of the phosphoenolpyruvate carbox- hypothesize that this effect may be mediated by an adi- GENES & DEVELOPMENT 1231 Downloaded from genesdev.cshlp.org on October 23, 2021 - Published by Cold Spring Harbor Laboratory Press Tontonoz et al. were prepared as described (Dignam et al. 1983), except that pocyte-specific family member such as mPPAR~/2. Fur- extracts were not dialyzed but added directly to the binding ther studies, including the experimental alteration of reaction. The purification of ARF6 from adipocyte nuclear ex- mPPAR~/2 expression in various adipose and nonadipose tract will be described elsewhere (P. Tontonoz and B.M. Spiegel- cells, will be necessary to determine whether this factor man, unpubl.). plays such a central role in adipocyte development. Oligonucleotides Materials and methods The sequences of double-stranded oligonucleotides used were as Cloning of the mPPAR72 cDNA follows (only one strand shown): ARE7, 5'-GATCTGTGAAC- TCTGATCCAGTAAG-3'; TREp, 5'-GATCTCTCAGGTCAT- First-strand cDNA was prepared from 3T3-F442A adipocyte to- GACCTGAATG-3'; PPRE, 5'-GATCTGTGACCTTTGTCCT- tal RNA (10 ~g) using the cDNA Cycle Kit (Invitrogen). The AGTAAG-3'; DR- 1, 5'-GATCTGTGACCTCTGACCTAGTA- sequences of the PCR primers used were 5'-GAT/CAAA/ AG-3'; [~RARE, 5'-GATCTGGGTTCACCGAAAGTTCACG-3'. GGCITCIGGCTTT/CCA-3' and 5'-CGGATIGCA/GTTA/ The nonspecific oligonucleotide was the ARE2 oligonucleotide GTGIGACAT-3'. PCR was performed for 35 cycles of 94°C/2 described in Graves et al. (1992). rain, 60°C/1 min, and 72°C/2 min. PCR products were isolated following electrophoresis through an agarose gel, subcloned into pCRII using the TA Cloning Kit (Invitrogen) and sequenced Plasmids, cell culture, and transfections using the Sequenase Kit (U.S. Biochemical). Full-length cDNA The mPPAR~/2 expression vector was contructed by ligating the clones were isolated from a ~ZAPII 3T3-F442A adipocyte entire 1.8-kb cDNA as an XbaI-XhoI fragment into the XbaI- cDNA library (custom made by Stratagene), by high stringency HindIII sites of pSV-SPORT (GIBCO-BRLI. The RXRa expres- hybridization using the subcloned PCR fragment as probe (Ma- sion vector was constructed by ligating the 1.8-kb EcoRI frag- niatis et al. 1982). eDNA sequences were subcloned into pBlue- ment from the RXRa hXR3-1 clone {Mangelsdorf et al. 1990} script SK (Stratagene) by in vivo excision from the XZAPII phage into the EcoRI site of pSV-SPORT. 3T3-F442A and 3T3-L1 as described by the manufacturer and sequenced using a series preadipocytes were cultured and differentiated as described of internal oligonucleotide primers. (Green and Kehinde 1974; Graves et al. 1991}. The 518-bp aP2 enhancer CAT construct was described previously (Graves et al. RNA analysis 1992). The ARE7 CAT construct was constructed by ligating three copies of the ARE7 oligonucleotide (see abovel into the Total RNA was isolated from cultured cells and tissues of adult CAT reporter vector pSVKS1 (Graves et al. 1992). NIH-3T3 cells mice by guanidine isothiocyanate extraction as described IChir- were cultured in Dulbecco modified Eagle medium (DMEMI gwin et al. 1979}. Ten micrograms of RNA was denatured in containing 10% bovine calf serum (HyClone) and transfected 1 formamide and formaldehyde at 55°C for 15 min and electro- day postconfluence by the modified calcium phosphate method phoresed through formaldehyde-containing agarose gels as de- described previously {Graves et al. 1991}. Each 90-mm dish re- scribed (Maniatis et al. 1982). RNA was blotted to BioTrans ceived 2 ~g of pSVKSl-derived reporter, 1 ~g of pSV-SPORT- nylon membranes (ICN), and membranes were cross-linked, hy- derived expression vector, 1 ~g of ~-actin-LacZ plasmid as an bridized, and washed as directed by the manufacturer, cDNA internal control (Oliviero et ~1. 1992), and 15 t~g of sonicated probes were labeled with [a-32P]dCTP (6000 Ci/mmole) by the salmon sperm DNA {Sigma). Twelve hours after transfection, random priming method (Fineberg and Vogelstein 1984) to a cells were refed with DMEM containing 10% charcoal-stripped specific activity of at least 109 cpm/~g. Autoradiographs were calf serum and activators as indicated. ETYA, clofibric acid, and digitally scanned and quantitated using Digital Darkroom and linoleic acid were from Sigma; 9-cis retinoic acid was obtained Image software for Apple Macintosh computers. from A. Levin (Hoffmann-La Roche, Nutley, NJ). Activators were dissolved in dimethylsulfoxide and applied to the cells in DNA-t~nding assays a volume of 10 ~1. Transfections were performed in duplicate and repeated a minimum of four times with quantitatively and DNA mobility retardation assays were performed as described qualitatively similar results. CAT activity was assayed as de- (Graves et al. 1991) except that the reaction buffer consisted of scribed (Graves et al. 1991} and quantitated using a Phosphor- 20 mM HEPES (pH 7.9), 150 mM NaC1, 5% glycerol, and 0.1% Imaging device IMolecular Dynamics}. NP-40. 0.5-1.0 ~g of poly[d(I-C)] (Pharmacia) was included in each reaction as nonspecific competitor. When antisera were used, binding reactions were incubated with antisera for 15 rain Acknowledgments at room temperature before the addition of the probe. RXRe~, mPPAR(~, and RAR~ antisera were obtained from R. Evans [Salk We thank members of the Spiegelman laboratory for helpful Institute, La Jolla, CA). The methylation interference footprint- discussions and Jae Bum Kim for providing the 3T3---L1 RNA. ing assay was performed under standard DNA mobility retarda- We are grateful to l- Reddy [Northwestern University) for the tion conditions using methylated DNA probes as described (Sie- mPPAR~I cDNA, R. Evans (Salk Institute) for the antisera to benlist and Gilbert 1980). In vitro translation of RXRa-SPORT, RXRct, RARa, and mPPAR, and A. Levin IHoffmann-La Rochel mPPAR~/1-SPORT and PPAR~/2-SPORT plasmids were per- for providing 9-cis retinoic acid. P.T. was supported by National formed using the TNT SP6-coupled reticulocyte lysate system Research Service award T32 GM07753-14 from the National (Promega) as recommended by the manufacturer. Five microli- Institute of General Medical Sciences. E.H. was supported by a ters of the 50 ~1 translation product was used in each binding post-doctoral fellowship from Juvenile Diabetes Association In- reaction. For aSS-labeling, in vitro translations were carried out ternational. This work was funded under grant DK31405-11 in the presence of 50 ~Ci of translation grade L-[aSS]methionine from the National Institutes of Health {to B.M.S). (1100 Ci/mmole}, and 5 ~1 of each reaction was resolved on a The publication costs of this article were defrayed in part by 9.0% SDS-polyacrylamide gel. Preadipocyte nuclear extracts payment of page charges. This article must therefore be hereby 1232 GENES & DEVELOPMENT Downloaded from genesdev.cshlp.org on October 23, 2021 - Published by Cold Spring Harbor Laboratory Press Adipocyte-specitic transcription tactor PPAR7 marked "advertisement" in accordance with 18 USC section 1991. Identification of a potent adipocyte-specific enhancer: 1734 solely to indicate this fact. Involvement of an NF-l-like factor. Genes & Dev. 5: 428- Graves, R.A., P. Tontonoz, and B.M. Spiegelman. 1992. Analysis Note added in proof of a tissue-specific enhancer: ARF6 regulates adipogenic gene expression. Mol. Cell. Biol. 12: 1202-1208. The nucleotide sequence data for PPAR~/2 DNA has been sub- Green, H. and O. Kehinde. 1974. Sublines of mouse 3T3 cells mitted to the EMBL data library under accession number that accumulate lipid. Cell 1:113-116. U09138. Herrera, R., H.S. Ro, G.S. Robinson, K.G. Xanthopoulos, and B.M. Spiegelman. 1989. A direct role for C/EBP and the AP-1 binding site in gene expression linked to adipocyte differen- References tiation. Mol. Cell. Biol. 9: 5331-5339. Aihaud, G., P. Grimaldi, and R. Negrel. 1992. Cellular and mo- Isseman, I. and S. Green. 1990. Activation of a member of the lecular aspects of adipose tissue development. Annu. Rev. steriod hormone receptor superfamily by peroxisome prolif- Nutr. 12: 207-233. erators. Nature 347: 645--649. Amri, E.Z., G. Ailhaud, and P.A. Grimaldi. 1991. Regulation of Kastner, P., A. Krust, B. Turcotte, U. Stropp, L. Tora, H. Grone- adipose cell differentiation I. Fatty acids are inducers of the meyer, and P. Chambon. 1990. Two distinct oestrogen-reg- aP2 gene expression. J. Lipid Res. 22: 1457-1463. ulated promoters generate transcripts encoding the two Bemlohr, D.A., C.W. Angus, M.D. Lane, M.A. Bolanowski, and functionally distinct human progesterone receptor forms A T.J. Kelly. 1984. Expression of specific mRNAs during adi- and B. EMBO J. 9: 1603-1614. pose differentiation: identification of an mRNA encoding a Keller, H. and W. Wahli. 1993. Peroxisome proliferator-acti- homologue of myelin P2 protein. Proc. Natl. Acad. Sci. vated receptors: A link between endocrinology and nutri- 81: 5468-5472. tion? Trends Endocrinol. Metab. 4: 291-296. Brandes, R., R. Hertz, R. Arad, S. Naishtat, S. Weil, and J. Bar- Keller, H., C. Dreyer, J. Medin, A. Mahfoudi, K. Ozato, and W. Tana. 1987. Adipocyte conversion of cultured 3T3-L1 prea- Wahli. 1993. Fatty acids and retinoids control lipid metabo- dipocytes by bezafibrate. Life Sci. 40: 935-941. lism through activation of peroxisome proliferator-activated Chambon, P., A. Zelent, M. Petkovich, C. Mendelsohn, P. Le- receptor-retinoid X receptor heterodimers. Proc. Natl. Acad. roy, A. Krust, P. Kastner, and N. Brand. 1991. The family of Sci. 90: 2160-2164. retinoic acid nuclear receptors. In Retinoids ten years on (ed. Kliewer, S.A., K. Umesono, R. Heyman, D. Mangelsdorf, J. J.H. Saurat), pp. 10-27. S. Karger, Basel, Switzerland. Dyck, and R. Evans. M. 1992a. Retinoid x receptor-COUP- Chirgwin, J.M., A.E. Przbyla, R.J. MacDonald, and W.J. Rutter. TF interactions modulate retinoic acid signaling. Proc. Natl. 1979. Isolation of biologically active ribonucleic acid from Acad. Sci. 89: 1448-1452. sources enriched in ribonuclease. Biochemistry 18: 5294- Kliewer, S.A., K. Umesono, D.J. Mangelsdorf, and R.M. Evans. 5299. 1992b. Retinoid X receptor interacts with nuclear receptors Christy, R.J., V.W. Yang, J.M. Ntambi, D.E. Geiman, W.M. in retinoic acid, thyroid hormone, and vitamin D3 signal- Landschulz, A.D. Friedman, Y. Nakabeppu, T.J. Kelly, and ling. Nature 355: 446--449. M.D. Lane. 1989. Differentiation-induced gene expression in Kliewer, S.A., K. Umesono, D. Noonan, R. Heyman, and R.M. 3T3-L1 preadipocytes: CCAAT/enhancer binding protein Evans. 1992c. Convergence of 9-cis retinoic acid and perox- interacts with and activates the promoters of two adipocyte- isome proliferator signalling pathways through heterodimer specific genes. Genes & Dev. 3: 1323-1335. formation of their receptors. Nature 358: 771-774. Dignam, J.D., R.M. Lebowitz, and R.G. Roeder. 1983. Accurate Mangelsdorf, D.J., E. Ong, J.A. Dyck, and R.M. Evans. 1990. transcription initiation by RNA polymerase II in soluble ex- Nuclear receptor that identifies a novel retinoic acid re- tract from isolated mammalian nuclei. Nucleic Acids Res. sponse pathway. Nature 345: 224-229. 11: 1475-1489. Mangelsdorf, D.J., K. Umesono, S.A. Kliewer, U. Borgmeyer, Distel, R.J., G.S. Robinson, and B.M. Spiegelman. 1992. Fatty E.S. Ong, and R.M. Evans. 1991. A direct repeat in the cel- acid regulation of gene expression: Transcriptional and post- lular retinol-binding protein type II gene confers differential transcriptional mechanisms. J. Biol. Chem. 267: 5937-5941. regulation by RXR and RAR. Cell 66: 555-561. Dreyer, C., G. Krey, H. Keller, F. Givel, G. Helftenbein, and W. Mangelsdorf, D.J., U. Borgmeyer, R. Heyman, J. Zhou, E. Ong, Wahli. 1992. Control of the peroxisomal p-oxidation path- A. Oro, A. Kakizuka, and R.M. Evans. 1992. Characteriza- way by a novel family of nuclear hormone receptors. Cell tion of three RXR genes that mediate the action of 9-cis 68: 879-887. retinoic acid. Genes & Dev. 6: 329-344. Fineberg, A.P. and B. Vogelstein. 1984. A technique for radiola- Maniatis, T., E.F. Fritsch, and J. Sambrook. 1982. Molecular beling DNA restriction fragments to high specific activity. cloning. A laboratory manual. Cold Spring Harbor Labora- Anal. Biochem. 137: 266--267. tory, Cold Spring Harbor, New York. Freytag, S.O. and T.J. Geddes. 1992. Reciprocal regulation of Oliviero, S., G.S. Robinson, K. Struhl, and B.M. Spiegelman. adipogenesis by Myc and C/EBP alpha. Science 956: 379- 1992. Yeast GCN4 as a probe for oncogenesis by AP-1 tran- scription factors: transcriptional activation through AP-1 Gharbi-Chihi, J., M. Teboul, J. Bismuth, J. Bonne, and J. Torre- sites is not sufficient for cellular transformation. Genes & sani. 1993. Increase of adipocyte differentiation by hypolip- Dev. 6: 1799-1809. idemic fibrate drugs in Ob17 preadipocytes: Requirement for Osmundsen, H., J. Bremer, and J.I. Pedersen. 1991. Metabolic thyroid hormone. Biochem. Biophys. Acta 1177: 8-14. aspects of peroxisomal beta-oxidation. Biochem. Biophys. Gottlicher, M., E. Widmark, Q. Li, and J.A. Gustafsson. 1992. Acta. 1085: 141-158. Fatty acids activate a chimera of the clofibric acid-activated Ross, S.R., R.A. Graves, A. Greenstein, K.A. Platt, H.-L. Shyu, B. receptor and the glucocorticoid receptor. Proc. Natl. Acad. Mellovitz, and B.M. Spiegelman. 1990. A fat-specific en- Sci. 89: 4653-4657. hancer is the primary determinant of gene expression for Graves, R.A., P. Tontonoz, S.R. Ross, and B.M. Spiegelman. adipocyte P2 in vivo. Proc. Natl. Acad. Sci. 87:9590-95941 GENES & DEVELOPMENT 1233 Downloaded from genesdev.cshlp.org on October 23, 2021 - Published by Cold Spring Harbor Laboratory Press Tontonoz et al. Ross, S.R., L. Choy, R.A. Graves, N. Fox, V. Solevjeva, S. Klaus, D. Ricquier, and B.M. Spiegelman. 1992. Hibemoma forma- tion in transgenic mice and isolation of a brown adipocyte cell line expressing the uncoupling protein gene. Proc. Natl. Acacl. Sci. 89: 7561-7565. Ross, S.R., R.A. Graves, and B.M. Spiegelman. 1993. Targeted expression of a toxin gene to adipose tissue: Transgenic mice resistant to obesity. Genes & Dev. 7: 1318-1324. Shepherd, P.R., L. Gnudi, E. Tozzo, H. Yang, F. Leach, and B.B. Kahn. 1994. Adipose cell hyperplasia and enhanced glucose disposal in transgenic mice overexpressing GLUT4 selec- tively in adipose tissue. ]. Biol. Chem. 268: 22243-22246. Short, M.K., D.E. Clouthier, M. Schaefer, R.E. Hammer, M.A. Magnuson, and E.G. Beale. 1992. Tissue-specific, develop- mental, hormonal and dietary regulation of a rat phosphoe- nolpyruvate carboxykinase-human growth hormone fusion gene in transgenic mice. Mol. Ceil. Biol. 12: 1007-1020. Siebenlist, U. and W. Gilbert. 1980. Contacts between Esche- richia coli RNA polymerase and an early promoter of phage TT. Proc. Natl. Acad. Sci. 77: 122-126. Sladek, F.M., W. Zhong, E. Lai, and J.E. Damell. 1990. Liver- enriched transcription factor HNF-4 is a novel member of the steroid hormone receptor superfamily. Genes & Dev. 4: 2353-2365. Tontonoz, P., J.B. Kim, R.A. Graves, and B.M. Spiegelman. 1993. ADD 1: A novel helix-loop-helix transcription factor associ- ated with adipocyte determination and differentiation. Mol. Cell. Biol. 13: 4753-4759. Umek, R.M., A.D. Friedman, and S.L. McKnight. 1991. CCAAT/enhancer binding protein: A component of a differ- entiation switch. Science 251: 288-292. Umesono, K., K.K. Murakami, C.C. Thompson, and R.M. Evans. 1991. Direct repeats as selective response elements for the thyroid hormone, retinoic acid, and vitamin D3 re- ceptors. Ceil 65: 1255-1266. Zhang, X.K., J. Lehmann, B. Hoffmann, M.I. Dawson, J. Cam- eron, G. Graupner, T. Hermann, P. Tran, and M. Pfahl. 1992. Homodimer formation of retinoic acid X receptor induced by 9-cis-retinoic acid. Nature 358: 587-591. Zhu, Y., K. Alvares, Q. Huang, M.S. Rao, and J.K. Reddy. 1993. Cloning of a new member of the peroxisome proliferator- activated receptor gene family from mouse liver. 1. Biol. Chem. 268: 26817-26820. 1234 GENES & DEVELOPMENT Downloaded from genesdev.cshlp.org on October 23, 2021 - Published by Cold Spring Harbor Laboratory Press P Tontonoz, E Hu, R A Graves, et al. Genes Dev. 1994, 8: Access the most recent version at doi:10.1101/gad.8.10.1224 This article cites 40 articles, 20 of which can be accessed free at: References http://genesdev.cshlp.org/content/8/10/1224.full.html#ref-list-1 License Receive free email alerts when new articles cite this article - sign up in the box at the top Email Alerting right corner of the article or click here. Service Copyright © Cold Spring Harbor Laboratory Press
Journal
Genes & Development – Unpaywall
Published: May 15, 1994