Developmentally Regulated N-terminal Variants of the NuclearReceptor Hepatocyte Nuclear Factor 4� Mediate MultipleInteractions through Coactivator and Corepressor-HistoneDeacetylase Complexes*

Received for publication, July 26, 2002, and in revised form, August 26, 2002Published, JBC Papers in Press, August 29, 2002, DOI 10.1074/jbcM207545200

Maria Elena Torres-Padilla‡, Frances M. Sladek§ and Mary C. Weiss‡¶

From the ‡Unite de Genetique de la Differenciation, FRE 2364 du CNRS, Departement de Biologie du Developpement,Institut Pasteur, 75724 Paris Cedex 15, France, and the §Department of Cell Biology and Neuroscience,University of California, Riverside, California 92521

To understand the mechanisms governing the regula-tion of nuclear receptor (NR) function, we compared theparameters of activation and repression of two isoformsof the orphan receptor hepatocyte nuclear factor (HNF)4�. HNF4�7 and HNF4�1 differ only in their N-terminaldomains, and their expression in the liver is regulateddevelopmentally. We show that the N-terminal activa-tion function (AF)-1 of HNF4�1 possesses significant ac-tivity that can be enhanced through interaction withglucocorticoid receptor-interacting protein 1 (GRIP-1)and cAMP response element-binding protein-bindingprotein (CBP). In striking contrast, HNF4�7 possessesno measurable AF-1, implying major functional differ-ences between the isoforms. Indeed, although HNF4�1and HNF4�7 are able to interact via AF-2 with GRIP-1,p300, and silencing mediator for retinoid and thyroidreceptors (SMRT), only HNF4�1 interacts in a synergis-tic fashion with GRIP-1 and p300. Although both iso-forms interact physically and functionally with SMRT,the repression of HNF4�7 is less robust than that ofHNF4�1, which may be caused by an increased ability ofthe latter to recruit histone deacetylase (HDAC) activityto target promoters. Moreover, association of SMRTwith HDACs enhanced recruitment of HNF4�1 but not ofHNF4�7. These observations suggest that NR isoform-specific association with SMRT could affect activity ofthe SMRT complex, implying that selection of HDACpartners is a novel point of regulation for NR activity.Possible physiological consequences of the multiple in-teractions with these coregulators are discussed.

Members of the nuclear receptor (NR)1 superfamily regulatea broad array of physiological processes (1). Their effects are

elicited by either activation or repression of a set of genes thatharbor DNA motifs recognized by NRs. Although mechanismsof activation are well understood today, repression is beginningto be recognized as a key player in homeostasis (2–4).

Gene activation is generally associated with relaxed chroma-tin structure, which is facilitated by acetylation and methyla-tion of histone N-terminal tails through the recruitment ofacetyltransferases or methyltransferases (5, 6). Recruitment ofthese coregulators to target promoters occurs mainly throughdirect physical interaction with one of the two activation func-tion modules of NRs, the ligand-dependent AF-2 (7). The role ofthe other AF, AF-1, is less well studied. Nevertheless, increas-ing evidence suggests that AF-1 is involved in recruitment ofcoactivators such as Src-1, GRIP-1, and CBP (8–11). The mech-anisms of cooperative action of the two AFs within a single NRare by and large poorly understood, but it has been shown forestrogen receptor � that the transcription intermediary factorTIF2 mediates such cooperation (12).

Gene repression is associated with recruitment of histonedeacetylases (HDACs) that remove acetyl groups from his-tones, provoking the recondensation of chromatin (5, 6). Re-pression is achieved by recruiting corepressors to regulatoryregions through direct physical association with NRs or othertranscription factors (13–17). The corepressors SMRT (silenc-ing mediator for retinoid and thyroid receptors) and NCoR(nuclear receptor corepressor) contain two receptor interactingdomains (RID1 and RID2) that mediate interactions with NRs(18, 19). In most cases, interaction with the RIDs occurs in theabsence of the corresponding ligand (20). Moreover, differentNRs exhibit a preference for association with NCoR or SMRT(21) and for one RID within the corepressors (18, 22, 23). SMRTand NCoR tether additional components of the corepressorcomplexes, including HDACs, through their silencing domains(4, 16, 24). The search for partners of SMRT and NCoR hasrevealed an increasing number of HDACs.

HDACs have been classified on the basis of their relatednessto yeast homologs. The mammalian class I HDACs (HDAC1, 2,and 3) are related to RPD3, and HDACs of class II (HDAC4, 5,6, and 7) share similarities with HDA1 (25–28). There seems tobe a selectivity for association with HDACs because it hasrecently been established that SMRT functions as an activatingcofactor for HDAC3 but not for HDAC4, 5, or 7 (29, 30). Instead,HDAC4 needs to be bridged to a SMRT�HDAC3 complex to

* This work was supported by Association pour la Recherche contre leCancer and the Human Frontiers Science Program Grant RG0303/2000-M (to M. C. W.) and National Institutes of Health Grant DK53892(to F. M. S.). The costs of publication of this article were defrayed inpart by the payment of page charges. This article must therefore behereby marked “advertisement” in accordance with 18 U.S.C. Section1734 solely to indicate this fact.

¶ To whom correspondence should be addressed: Unite de Genetiquede la Differenciation, Departement de Biologie du Developpement, In-stitut Pasteur, 25, Rue du Dr. Roux, 75724 Paris Cedex 15, France. Tel.:33-1-4568-8500; Fax: 33-1-4061-3231; E-mail: mweiss@pasteur.fr.

1 The abbreviations used are: NR, nuclear receptor; AF, activationfunction; �-gal, �-galactosidase; CBP, cAMP response element-bindingprotein; CMV, cytomegalovirus; DBD, DNA binding domain; GRIP-1,glucocorticoid receptor-interacting protein 1; GST, glutathione S-trans-ferase; HA, hemagglutinin; HDAC, histone deacetylase; HIV, humanimmunodeficiency virus; HNF, hepatocyte nuclear factor; LTR, long

terminal repeat; Luc, luciferase; NCoR, nuclear receptor corepressor;RID, receptor interacting domains; SMRT, silencing mediator for reti-noid and thyroid receptors; TSA, trichostatin A; VSV, vesicular stoma-titis virus.

THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 277, No. 47, Issue of November 22, pp. 44677–44687, 2002© 2002 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S.A.

This paper is available on line at http://www.jbc.org 44677

by guest on February 14, 2018http://w

ww

.jbc.org/D

ownloaded from

become active (29). Regulated binding of HDACs might play arole in determining SMRT corepressor activity on promoters,thereby constituting a critical level of regulation for NR activitywhich integrates the diversity of HDACs in the current modelfor repression of transcription. No data concerning the occur-rence and consequences of recruitment of HDAC family mem-bers to different promoters have been documented.

HNF4� is a member of the NR superfamily and is highlyexpressed in the liver (31). A key role for HNF4� in develop-ment and hepatic differentiation and homeostasis has beendemonstrated unambiguously (32–34). A number of isoforms ofHNF4� resulting from alternative splicing and alternative pro-moter usage have been identified (35), but their roles in devel-opment and differentiation have not been examined. The activ-ity of HNF4�1, the first isoform identified, has been shown tobe repressed by SMRT (36) and enhanced by association withCBP/p300 and GRIP-1 (37–39). HNF4�2, a C-terminal splicevariant, exhibits enhanced interaction with coactivators (38).We showed recently that HNF4�7, resulting from alternativepromoter usage, is expressed during the early stages of liverdevelopment, prior to HNF4�1, and that it exhibits transcrip-tional activity different from that of HNF4�1 (40). Beyondthese findings, nothing is known about functional differences inthe HNF4� isoforms.

To contribute to understanding the mechanisms governingthe regulation of NR function, we have compared the parame-ters of activation and repression of the two isoforms of HNF4�which differ only in their N-terminal domains. We show herethat the N-terminal AF-1 of HNF4�1 possesses significantactivity that can be enhanced through interaction with GRIP-1and CBP. In striking contrast, HNF4�7 possesses no measur-able AF-1, implying major functional differences between iso-forms. Indeed, we were able to show that although HNF4�1and HNF4�7 are both able to interact via the AF-2 withGRIP-1, p300, and SMRT, only HNF4�1 is able to interact in asynergistic fashion with both GRIP-1 and p300. Although bothfactors show physical and functional interaction with SMRTand HDACs, the repression of HNF4�7 is less robust than thatof HNF4�1. Our determination that they recruit HDAC activ-ity in a different fashion suggests a biochemical basis for theirdifferent roles in transactivation of target genes. Our observa-tions strongly suggest that both qualitative and quantitativedifferences ranging from the nature of the recruited coregula-tor to the selectivity of its association with direct effectors ofhistone modification such as HDACs can act simultaneously tomodulate transcriptional activity of NRs.

EXPERIMENTAL PROCEDURES

Plasmids—The empty pCB6 vector (41) and the CMV.HNF4�1.VSV(42) and CMV.HNF4�7.VSV expression vectors (40) containing thefull-length rat cDNAs for the corresponding isoforms are describedelsewhere. The apoA-I reporter pZL.HIV.LTR.AI-4 (38) and the apoC-III.Luc reporter (43) have been described. The pCMV.SMRT (44),pSG5.GRIP-1 (45), pSG5.GRIP-1.NRmut (10), CMV.CBP (46), pCMV.HA.p300 (47), pEV589.HDAC3.FLAG (48), and pL67.HDAC4.FLAGplasmids (49) are described elsewhere. The pSG424 plasmid containsthe DNA binding domain (DBD) of the yeast GAL4 protein (50), and thereporter plasmid Tp1.GAL4.Luc (51) harbors 10 GAL4 binding sites infront of the Tp1 minimal promoter driving expression of the luciferasegene. The two GAL4.HNF4�.A/B constructs were prepared by insertingthe sequence encoding the A/B domains of rat HNF4�1 or HNF4�7in-frame into the EcoRI sites of the pSG424 vector. For theGAL4.HNF4�.�F constructs, rat cDNAs encoding truncated HNF4�1and HNF4�7 proteins were cloned into the EcoRI sites of the pSG424plasmid. The constructions were verified by sequencing. The CMV.�-galplasmid (California Biotechnology Inc.) contains the lacZ gene undercontrol of the CMV promoter.

The pSET.HNF4�1.6His and pSET.HNF4�7.6His vectors containthe full-length cDNA of rat HNF4�1 and HNF4�7, respectively. Theseplasmids were constructed as follows. The cDNAs for each of the iso-

forms was amplified by PCR and ligated into the HindIII site of thepGEX-3x-6His vector (Amersham Biosciences). The cDNA with the6His tag in 3� derived from the pGEX-3x-6His vector was then excisedby EcoRI and XhoI and ligated into the pRSET5d vector (52). Theconstructs were verified by sequencing. The GST-SMRT fusion proteinsconstructs, GST-SMRT.RID1 (amino acids 1055–1291) and GST-SMRT.RID2 (1291–1495), from human SMRT have been describedpreviously (19).

Transient Transfections—HNF4�-deficient NIH3T3 and 293T cells(40, 53) were grown in Dulbecco’s modified Eagle’s medium supple-mented with 5% fetal calf serum at 37 °C under a 5% CO2 atmosphere.For transient transfection assays, 2 � 105 cells were seeded in six-wellplates and transfected by the calcium phosphate coprecipitation proce-dure the next day using 1 �g of reporter, 100 ng of CMV.�-gal, andvarious amounts of expression vectors as indicated in the figure leg-ends. Equivalent molar amounts of empty vector were added to equalizethe DNA amount. Glycerol shock was carried out 5 h later, and cellswere harvested 24 h after the shock. �-Galactosidase activity wasmeasured by the standard colorimetric method, and luciferase activitywas determined with a Berthold Luminometer 9501 as described (54).For the one-hybrid assay, the indicated amounts of the pSG424 or theGAL4.HNF4� plasmids along with 1 �g of Tp1.GAL4.Luc, variousamounts of CMV.SMRT, and 100 ng of CMV.�-gal were used. For TSAtreatment, 100 ng of trichostatin A (Sigma) or vehicle (dimethyl sulf-oxide) was added 8 h before harvesting. For immunoprecipitations, 6 �106 cells were seeded 1 day before transfection and were transfected bythe same procedure with 15 �g of expression vectors as indicated in thefigure legends.

Immunoprecipitations and Western Blot Analysis—Transfected cellswere rinsed with ice-cold phosphate-buffered saline, resuspended in 1ml of lysis buffer (150 mM NaCl, 1 mM EDTA, 20 mM Tris-HCl, pH 8.0,0.4% Nonidet P-40, 5% glycerol), and incubated for 30 min at 4 °C.Lysates were cleared by centrifugation and incubated for 1 h with eitherthe �445 antiserum (31) directed against the C-terminal region ofHNF4�1/�7, or with an anti-SMRT antibody (sc-1612, Santa Cruz Bio-technology) and then overnight with protein A or protein G PLUS-agarose (Santa Cruz Biotechnology). Beads were pelleted by centrifu-gation and washed with the same buffer. For the double pulldownexperiments, the lysis buffer was 200 mM NaCl, 1 mM EDTA, 20 mM

Tris-HCl, pH 8.0, 0.5% Nonidet P-40, 5% glycerol. Beads were washedfour times with lysis buffer and twice with 0.01% Nonidet P-40 NET-N(20 mM Tris-Cl, pH 8.0, 1 mM EDTA, 100 mM NaCl). Whole cell extractsand immunocomplexes were subjected to Tris-glycine SDS-PAGE andblotted onto a polyvinylidene difluoride membrane (Millipore). Theprimary antibodies were a monoclonal anti-VSV antibody coupled toperoxidase (Roche Molecular Biochemicals), the �445 antiserum, theanti-SMRT antibody, and an anti-GAL4 DBD antibody (sc-577, SantaCruz Biotechnology). When required, goat anti-rabbit or donkey anti-goat horseradish peroxidase antibodies were used. Peroxidase activitywas detected by ECL (Amersham Biosciences).

GST Pulldown Assays and Protease Digestion—All protein-proteininteraction analyses were performed as described previously (38) usingagarose-immobilized GST fusion proteins. [35S]Methionine-labeled pro-teins were produced with the TNT system (Promega) usingpSET.HNF4�1.6His, pSET.HNF4�7.6His, pEV589.HDAC3.FLAG (48),or pL67.HDAC4.FLAG (49) plasmid. Where indicated, 2 �l of the TNTproduct was incubated for 10 min on ice with 200 ng of double-strandedoligonucleotides in 100 mM HEPES, 10 mM MgCl2, 200 mM KCl, 2.5 mM

EGTA, 5 mM dithiothreitol, and 20% Ficoll. The sequence of the apoC-III oligonucleotide used is 5�-TCGAGCGCTGGGCAAAGGTCACCTGC.Protease digestion assays of 35S-HNF4�1 and 35S-HNF4�7 with L-1-tosylamido-2-phenylethyl chloromethyl ketone-treated trypsin (Sigma)were performed by the addition of the indicated amounts of trypsin in50 mM Tris-HCl, pH 7.0. The reaction was carried out at 25 °C for 30min and was stopped by adding SDS loading buffer. Samples wereseparated by SDS-PAGE (10% gel) and analyzed by autoradiography.

HDAC Assays—Extracts of �8 � 108 transfected 293T cells wereimmunoprecipitated with the �445 antiserum. Immunocomplexes wereresuspended in TBS buffer, and HDAC activity was determined asdescribed by Taunton et al. (28).

RESULTS

HNF4�1 and HNF4�7 are isoforms of HNF4� resulting fromalternative promoter usage leading to the transcription of al-ternative coding first exons (Fig. 1A). The two promoters aresequentially transcribed during development (40). Distal to theamino acids encoded by the first exon, the two proteins are

HNF4� Isoforms and Coregulator Interaction44678

by guest on February 14, 2018http://w

ww

.jbc.org/D

ownloaded from

identical, and they differ in their amino acid compositions onlywithin the A/B domain (Fig. 1B) (55). We showed previouslythat HNF4�1 and HNF4�7 exhibit different transactivationactivities in transient transfection assays depending on thetarget gene (40). We now explore the mechanisms whereby

these isoforms behave differently as transactivators. Becauseboth isoforms share the same DBD, likely showing the sameDNA binding properties, we reasoned that functional interac-tions with coregulators might account for the differences.

The A/B Domain of HNF4�7 Lacks the AF-1 Function

The A/B domain of NRs is located at the N-terminal regionand harbors AF-1 (Fig. 1B), one of the two activation functionmodules. The AF-1 of HNF4�1 has been shown to interact withmembers of the basal machinery and with CBP (8), providing asurface of contact with transcriptional coregulators. The A/Bdomain of HNF4�1 consists of 49 amino acids, and its AF-1 hasbeen mapped to amino acids 1–24 (56). The AF-1 exhibits anoverall negative charge and shows autonomous transactivationaccounting for 40% of HNF4�1 activity as tested by the one-hybrid assay (56). As shown in Fig. 1B, the correspondingregion in HNF4�7 contains completely unrelated nonchargedamino acids. Thus, we first questioned whether the A/B domainof HNF�7 possesses an AF-1 function.

The regions corresponding to the A/B domains of HNF4�1and HNF4�7 were fused to the DBD of GAL4, and one-hybridassays were performed to test transactivation potentials. 293Tcells were transfected with 300 ng of one or the other of theGAL4.HNF4�.A/B constructs and a reporter for GAL4. As hasbeen shown previously, the A/B domain of HNF�1 showed arobust transactivation activity in this assay, provoking a 120-fold activation compared with the GAL4 construction (Fig. 1C).This activity is caused entirely by the presence of the AF-1domain (8). In striking contrast, the HNF4�7 A/B constructshowed no activity, and increasing the GAL4.HNF�7.A/B plas-mid amount up to 5 �g still did not result in transactivation ofthe reporter gene (data not shown). Nonetheless, it is possiblethat the A/B domain of HNF4�7 is constitutively repressedby endogenous HDAC activity. To exclude this possibilitywe measured the transcriptional activity of theGAL4.HNF4�7.A/B construct in the presence of the HDACinhibitor TSA and still failed to detect transactivation (data notshown). The lack of activity was not the result of the absence ofthe protein because the two fusion proteins were detected byWestern blot analysis using an anti-GAL4 DBD antibody (Fig.1D). These experiments demonstrate unambiguously thatHNF4�7 does not contain an AF-1 function.

It could not be ruled out, however, that the A/B region ofHNF4�7 tethers coactivators to target promoters. To test thishypothesis we performed the same one-hybrid experiments inthe presence of CBP or GRIP-1, which belong to different fam-ilies of coactivators.

Coactivators Increase HNF4�7 Activity Onlythrough Its AF-2

CBP and GRIP-1 Coactivate the AF-1 of HNF4�1—NIH3T3cells were cotransfected with 500 ng of pSG424, GAL4.HNF4-�1.A/B, or GAL4.HNF4�7.A/B plasmid, the Tp1.GAL4.Luc re-porter, and 5 �g of an expression vector for GRIP-1 or CBP.CBP elicited a modest but reproducible increase in the tran-scriptional activity of the GAL4.HNF4�1.A/B construct (Fig.2A, left panel). In contrast, even in the presence of CBP, the A/Bdomain of HNF4�7 failed to display substantial activity (Fig.2A). In the case of the p160 family member GRIP-1, an increaseof 5.9-fold of activity was elicited for the construct containingHNF4�1 A/B, but no stimulation of activity of the GAL4.HNF-4�7.A/B construct was observed (Fig. 2A, right panel).

GRIP-1 Coactivates HNF4�7 through Its AF-2—The AF-2module of NRs is located in the E domain and, in the case ofHNF4�1, has been shown to interact functionally with GRIP-1(38). To verify that the AF-2 of HNF4�7 is functional and can

FIG. 1. The A/B domain of HNF4�7 is missing the AF-1 func-tion. A, representation of exons 1A and 1D and A/B domains of HNF4�isoforms. HNF4�1 and HNF4�7 are encoded, respectively, by exons 1Aand 1D and a part of exon 2 so that their amino acid compositions at theN terminus differ. Bent arrows indicate alternative promoters. Exons1A and 1D code for 30 and 17 amino acids, respectively. The AF-1domain in HNF4�1 is indicated. B, amino acid sequence of theN-terminal regions of HNF4�1 and HNF4�7. Negatively charged aminoacids are indicated in bold characters. The AF-1 is localized to the first24 amino acids of HNF4�1 and is entirely encoded by exon 1A. C,HNF4�7 lacks the AF-1 function. A/B domains from both isoformscorresponding to amino acids 1–49 for HNF4�1 and 1–36 for HNF4�7were fused to the DBD of GAL4. 300 ng of the GAL4 constructs wascotransfected with 1 �g of the Tp1.GAL4.Luc reporter plasmid for GAL4in 293T cells. Luciferase activity was normalized to �-gal activity and isexpressed as activity relative to the GAL4 DBD plasmid alone. Resultsare the mean � S.E. of three independent experiments. D, Western blotanalysis using an anti-GAL4 DBD antibody on extracts of 293T cellstransfected with the GAL4 constructs used in C.

HNF4� Isoforms and Coregulator Interaction 44679

by guest on February 14, 2018http://w

ww

.jbc.org/D

ownloaded from

be coactivated, we used a longer fusion protein, GAL4.HNF-4�7.�F, which includes the A/B through the E domains (Fig.2B). NIH3T3 cells were cotransfected with GAL4,GAL4.HNF4�7.A/B, or GAL4.HNF4�7.�F with variousamounts of GRIP-1 expression vector and the Tp1.GAL4.Lucplasmid. Again, the A/B domain fused to GAL4 failed to induceany transcriptional activity (Fig. 2C). However, theGAL4.HNF4�7.�F construct did elicit a transcriptional activ-ity that was augmented in a dose-dependent manner uponaddition of increasing amounts of GRIP-1. Thus, GRIP-1 is able

to coactivate HNF4�7, but only through the E domain, andpresumably AF-2.

For the estrogen receptor � and glucocorticoid receptor, forwhich the agonists are well characterized, the AF-1 and AF-2modules have been shown to synergize through the action ofcoregulators (9, 12). We wondered whether the AF-1 and AF-2of HNF4� are able to synergize via the p160 and p300 coacti-vators. Moreover, if the AF-1 is necessary for the simultaneousaction of these two families of coregulators, then no synergyshould be observed for HNF4�7 in the presence of GRIP-1 andp300.

HNF4�1, but Not HNF4�7, Responds Synergisticallyto the p160 and p300 Coactivators

293T cells were cotransfected with the CMV.HNF4�1.VSV orCMV.HNF4�7.VSV plasmid encoding full-length proteins, ex-pression vectors for p300 and/or GRIP-1, and the apoA-I re-porter plasmid. Both HNF4�1 and HNF4�7 transactivated theapoA-I reporter (Fig. 3). p300 coactivated both isoforms by9-fold and 8-fold for HNF4�1 and HNF4�7, respectively (Fig.3). GRIP-1 showed similar behavior, coactivating both iso-forms, HNF4�1 by 2-fold and HNF4�7 by 4-fold (Fig. 3). Whenboth coactivators were cotransfected with the reporter,HNF4�1 activity was increased 15-fold, which represents asynergistic rather than an additive effect compared with theeffects provoked by either coactivator alone. In the case ofHNF4�7, activity was increased 12-fold, which corresponds to

FIG. 2. p160 coactivators increase HNF4�7 activity onlythrough its AF-2. A, CBP and GRIP-1 coactivate the AF-1 of HNF4�1.5 �g of CMV.CBP or pSG5.GRIP-1 plasmid was transfected intoNIH3T3 cells together with the GAL4.HNF4�.A/B plasmids and theTp1.GAL4.Luc reporter. Basal activity of the Tp1.GAL4.Luc reportervaried depending on the coactivator cloning vector used to equalize theDNA amount in transfections. B, schematic representation of full-length HNF4�7 and the GAL4.HNF4�7 fusion proteins used in C. C,GRIP-1 does not coactivate the A/B domain of HNF4�7, but it doescoactivate a GAL4 HNF4�7 fusion protein containing the AF-2 domain.100 ng of the GAL4 plasmids was cotransfected with increasingamounts of the pSG5.GRIP-1 plasmid and the Tp1.GAL4.Luc reporter.Luciferase activity was normalized to �-gal activity and is expressedrelative to the activity of the GAL4 plasmid alone. Numbers above thebars represent the -fold induction relative to transcriptional activity ofGAL4.HNF4�.A/B constructs in the absence of coactivators. Results arethe mean � S.E. of three independent experiments.

FIG. 3. Synergistic effects of p160 and p300 coactivators onAF-1 and AF-2 modules are isoform-limited. 293T cells were co-transfected with 20 ng of empty (pCB6), HNF4�1, or HNF4�7 expres-sion vectors and the apoA-I reporter. In addition, members of eachgroup received 100 ng of the CMV.p300.HA, the pSG5.GRIP-1, or thepSG5.GRIP-1.NRmut plasmid or combinations thereof indicated in thefigure. Within each graph, above the name of each coactivator(s), fourconditions, including pCB6 and HNF4�, minus and plus coactivator(s)each, were carried out. The raw luciferase unit (RLU) values obtainedwith empty pCB6 vector are not visible because they were below thelimits of resolution of the scale of the axis. The apoA-I reporter consistsof four site A elements responsive to HNF4� from the human apoA-Ipromoter in front of the HIV LTR driving expression of the fireflyluciferase gene. Luciferase activity was normalized to �-gal activity. Onthe apoA-I reporter, HNF4�1 and HNF4�7 induced a 175- and 88-foldactivation, respectively, in the absence of coactivators. Numbers abovethe bars represent the -fold induction relative to transcriptional activityof HNF4�1 (solid bars) or HNF4�7 (open bars) where 1� is the activityin the absence of coactivators. Shown is the mean � S.E. of a repre-sentative experiment performed in triplicate.

HNF4� Isoforms and Coregulator Interaction44680

by guest on February 14, 2018http://w

ww

.jbc.org/D

ownloaded from

the additive value for the induction elicited by each coactivatoralone. Thus, through the combined action of GRIP-1 and p300,the two AF modules of HNF4�1 appear to synergize. However,this is not the case for HNF4�7, which possesses only oneactivation function module. Hence, synergy seems to be de-pendent on the presence of the AF-1 in HNF4�1. Moreover, thesynergistic effect was lost when a construct bearing a mutationin the NR boxes of GRIP-1 was used (Fig. 3, GRIP-1 mut), thusindicating that the synergy is also dependent on the NR boxesthat interact with the AF-2 (10). From these experiments wecan conclude that in the context of the full-length HNF4�7,p300 and GRIP-1 act independently of the A/B domain.

SMRT Corepresses HNF4�1 More Robustly than HNF4�7

Previous work from one of our laboratories showed that thecorepressor SMRT represses HNF4�1 activity in transienttransfection assays (36). In that work, a potential role in core-pressor activity for the A/B domain of HNF4�1 was suggestedbecause deletion of this domain restored responsiveness toSMRT of an HNF4�1 protein truncated in the F domain. We

therefore compared the ability of SMRT to repress activity ofHNF4�1 and HNF4�7. Expression vectors containing full-length HNF4�1 or HNF4�7 were cotransfected along with in-creasing amounts of CMV.SMRT in 293T cells. We chose to usethe apoA-I and the apoC-III reporters that are more avidlyactivated by HNF4�1 than by HNF4�7 (Fig. 4A and Ref. 40).The relevance of the HNF4 sites for the activation of thesepromoters is underscored by recent findings that HNF4� con-trols the constitutive expression of both the apoA-I and apoC-III genes in vivo (33, 57), neither of which is expressed in fetalliver (58) when HNF4�7 activity is predominant (40).

Cotransfection with SMRT had no effect on the activity ofeither of the reporters in the presence of the pCB6 empty vectoronly, but it induced a dose-dependent decrease of the activityelicited by HNF4�1 and by HNF4�7 on both of the reporters(Fig. 4A). The effect of SMRT on the activity of both HNF4�

isoforms was seen even when only 25 ng of CMV.SMRT wasused in the transfections. The maximal amount of SMRT re-pressed the HNF4�1-mediated activation of the apoC-III andapoA-I reporters by 87 and 55%, respectively (Fig. 4A, top

FIG. 4. SMRT corepresses HNF4�1 more robustly than HNF4�7 and is recruited by both isoforms. A, 293T cells were cotransfectedwith 200 ng of empty (pCB6), HNF4�1, or HNF4�7 expression vector, increasing amounts of CMV.SMRT, and 1 �g of the indicated reporter. TheapoC-III reporter contains the region �821 to �24 from the human apoC-III gene driving the expression of the firefly luciferase gene. Luciferaseactivity was normalized to internal control �-gal activity. The mean � S.E. of two independent experiments performed in triplicate is shown. Beloweach graph, repression by increasing amounts of SMRT for each isoform is expressed as percent activity, 100 being basal activity of HNF4�1 orHNF4�7 on each reporter. B, Western blot analysis of transfected HNF4� isoforms using the �445 antiserum. 15 �l of the same protein extractsused for determining luciferase activity was separated by SDS-PAGE and analyzed by Western blot. C, SMRT coimmunoprecipitates with HNF4�1and HNF4�7. 293T cells were transfected with 15 �g of CMV.HNF4�1.VSV, CMV.HNF4�7.VSV, and/or CMV.SMRT as indicated. Whole cellextracts were immunoprecipitated with the �445 antiserum. Immunoprecipitate complexes (IP) were separated by SDS-PAGE and analyzed byWestern blot with an anti-VSV or an anti-SMRT antibody. D, HNF4�1 and HNF4�7 interact with the RID2 of SMRT in vitro. A GST pulldownassay was performed using in vitro translated 35S-HNF4�1 or 35S-HNF4�7 and bacterially purified GST.RID1 and GST.RID2 proteins. Shown areSDS-PAGE autoradiographies representative of several independent experiments. The percentage of labeled HNF4� bound to the GST fusionprotein was calculated with a PhosphorImager using 10% of the input of the reaction as reference. We systematically observe a second, fastmigrating band as a result of the in vitro translation of HNF4�1 which probably corresponds to partial translation products that were not takeninto account for quantification.

HNF4� Isoforms and Coregulator Interaction 44681

by guest on February 14, 2018http://w

ww

.jbc.org/D

ownloaded from

panel). However, in the case of HNF4�7, the maximal amountof SMRT repressed transactivation of the two reporters by only65 and 35%, respectively (Fig. 4A, top panel). Furthermore, thedose-response curves showed differences in the kinetics of re-pression of the two isoforms. The repression of HNF4�7 bySMRT was much more gradual than that of HNF4�1. Westernblot analysis of extracts from transfected cells revealed thatthese differences are not the result of different amounts ofprotein because both HNF4�1 and HNF4�7 are present atequivalent amounts in transfected 293T cells (Fig. 4B).

These results show that NR isoforms can be repressed dif-ferently by SMRT. We then examined whether both isoformsinteract in vivo and in vitro directly with SMRT and whetherdifferences in the affinity with which they interact with SMRTcould account for the results observed in the cotransfectionexperiments.

To explore whether physical interaction between the HNF4�isoforms and SMRT takes place in vivo, coimmunoprecipitationexperiments were performed. 293T cells were cotransfectedwith the same amount of either the CMV.HNF4�1.VSV or theCMV.HNF4�7.VSV expression vector with or withoutCMV.SMRT. Whole cell extracts were prepared and immuno-precipitated with the �445 antiserum directed against theC-terminal region of HNF4�1 (which is shared with HNF4�7).Immunoprecipitates were analyzed by Western blot using ananti-VSV and an anti-SMRT antibody. As shown in Fig. 4C,SMRT coprecipitated with both HNF4�1 and HNF4�7 (Fig. 4C,lanes 6 and 10).

SMRT contains two silencing domains responsible for tran-scriptional repression and two independent RIDs that displaydifferent affinities for members of the NR family (18, 23). RID1is contained within the region spanning amino acids 1055–1291, and RID2 is delimited by amino acids 1291–1495 (23).The ability of HNF4�1 and HNF4�7 to bind to RID1 or RID2was tested by GST pulldown assays. GST.RID1 and GST.RID2proteins were bacterially produced, purified, and incubatedwith in vitro translated 35S-HNF4�1 or 35S-HNF4�7. As shownin Fig. 4D, both HNF4�1 and HNF4�7 bind to RID2 but not toRID1 of SMRT. Indeed, previous studies have shown thatHNF4�1 interacts with RID2 of SMRT in vitro (36). Moreover,the fraction of HNF4�1 and HNF4�7 protein bound toGST.RID2 is equivalent to 14% of the input for each of theisoforms, indicating that in vitro, HNF4�1 and HNF4�7 bindwith the same affinity to RID2 of SMRT.

Our results demonstrate that SMRT interacts with both ofthe HNF4� isoforms in vivo and in vitro. However, these re-sults do not explain the differences observed regarding therepression of transactivation by HNF4�1 or HNF4�7. There-fore, to determine whether the AF-1 domain of HNF4�1 playsa role in SMRT repression we examined whether SMRT is ableto interact functionally with this domain and repress itsactivity.

SMRT Represses the Activity of AF-2 but NotAF-1 of HNF4�

One-hybrid experiments using the GAL4.HNF4�1.A/Bconstruct were performed using increasing amounts ofCMV.SMRT. Addition of the SMRT expression vector did notdecrease the activity elicited by the A/B domain of HNF4�1(Fig. 5A). Even when a 20-fold excess of CMV.SMRT comparedwith the GAL4.HNF4�1.A/B construct was used (400 ng), norepression was observed. These results show that SMRT doesnot repress the activity of the AF-1 domain. On the contrary,when the GAL4.HNF4�1.�F and GAL4.HNF4�7.�F constructsthat include the AF-2 of both isoforms were cotransfected withincreasing amounts of CMV.SMRT, repression of transcrip-

tional activation of both HNF4�1 and HNF4�7 was observed(Fig. 5B). Thus, SMRT action is mediated solely through re-pression of the AF-2 activity within full-length HNF4�1 andHNF4�7.

Because the difference in repression of HNF4�1 andHNF4�7 by SMRT cannot be explained by differences in theAF-1 or by recruitment of SMRT by the two isoforms, andbecause SMRT has been shown to tether different HDACs toconstitute the corepressor complex (4, 16, 26, 30, 59), we con-sidered two hypotheses to explain our results. One, HNF4�7could preferentially recruit inactive SMRT complexes to targetpromoters. Alternatively, the interaction of the NR isoformwith SMRT could affect the transcriptional repression elicitedby the recruited SMRT complex. In both cases, the ability of thetwo isoforms to recruit HDAC activity to target promotersshould be different.

Endogenous HDAC Activity Is Recruited to TargetPromoters Differently by HNF4�1 and HNF4�7

As a first approximation, we sought to determine whetherHNF4�1 and HNF4�7 differ in their ability to recruit HDACactivity by analyzing their transcriptional activities on HNF4reporters in the absence or presence of an inhibitor of HDACactivity, TSA. 293T cells were transiently transfected withexpression vectors encoding full-length HNF4�1 or HNF4�7together with the apoC-III or the apoA-I reporters and treatedwith TSA. The basal activity of both reporters was increasedslightly upon TSA treatment (2.9- and 3-fold for the apoC-IIIand apoA-I reporters, respectively) as seen from the results

FIG. 5. SMRT represses the activity of the AF-2 but not theAF-1 of HNF4�. A, SMRT does not repress the AF-1 activity ofHNF4�1. 293T cells were cotransfected with 1 �g of the Tp1.GAL4.Lucreporter plasmid, 20 ng of the GAL4 constructs, and increasingamounts of CMV.SMRT as indicated. B, the same experiment as in Ausing the GAL4.HNF4�1.�F and GAL4.HNF4�7.�F constructs. Lucif-erase activity was normalized to �-gal activity. The results obtainedfrom a representative of two experiments performed in triplicate areshown.

HNF4� Isoforms and Coregulator Interaction44682

by guest on February 14, 2018http://w

ww

.jbc.org/D

ownloaded from

obtained with the empty pCB6 vector (Fig. 6A). The addition ofTSA had a dramatic effect on transactivation of the apoC-IIIreporter by HNF4�1. The activity was induced 44-fold com-pared with HNF4�1 activity in the absence of TSA (Fig. 6A, leftpanel). This result indicates that HNF4�1 is highly repressedby association with endogenous HDAC activity. On the con-trary, the addition of TSA increased only slightly, by 7.7-fold,the HNF4�7 activity on the apoC-III promoter. Thus, HNF4�7recruits less HDAC activity than HNF4�1 to the apoC-III pro-moter. For the apoA-I reporter, the effects elicited by TSA weresmaller than with the apoC-III reporter (Fig. 6A, right panel),and its addition had the same impact on the transcriptionalactivity of both isoforms (4.6- and 5-fold for HNF4�1 andHNF4�7, respectively). The same results were obtained whensmaller amounts of HNF4� expression vectors were used forcotransfection (data not shown).

Thus, HNF4�7 recruits the same or less HDAC activity thanHNF4�1 to target promoters. Furthermore, the ability to re-cruit HDAC activity varies depending on the reporter used, andthis holds for both NR isoforms. Most importantly, the relief ofinhibition by TSA was stronger for HNF4�1 compared withHNF4�7 in the context of a native promoter.

Because a fold activation of HNF4� activity in the presenceof TSA was observed with both of the reporters tested, it ispossible that HNF4� is constitutively associated with HDACs.To verify the constitutive association of HDAC activity withHNF4� isoforms, HDAC assays were performed on HNF4�

immunocomplexes.HDAC Activity Is Associated Constitutively with HNF4�—

293T cells were transfected with expression vectors encodingeither full-length HNF4�1 or HNF4�7, and cell extracts wereprecipitated with the �445 antiserum. Immunocomplexes wereassayed for HDAC activity in the absence and presence of thedeacetylase inhibitor sodium butyrate as a control for the spec-ificity of the reaction. Crude lysate from 293T cells (equivalentto 1/250 of the volume used for each immunoprecipitation re-action) was used as a positive control. As expected, HDACactivity was found to be associated with both HNF4�1 andHNF4�7 and this activity was inhibited by sodium butyrate(Fig. 6B). Thus, these results confirm that NRs may be consti-tutively associated with HDAC activity within the cell (60).

It has recently been shown that recruitment of HDAC4 to aSMRT/NCoR complex does not result in activation of the com-plex unless a class I HDAC, HDAC3, is also associated with thecomplex (29). These findings suggest that the recruitment ofdifferent HDACs could result in NR�SMRT�HDAC complexesdisplaying different activities.

HNF4� Associates with SMRT�HDAC Complexes inVitro—To test the ability of HNF4� to associate with differentSMRT�HDAC complexes, we prepared class I and class IISMRT�HDAC complexes and tested the ability of in vitro trans-lated 35S-HNF4�1 and 35S-HNF4�7 to interact with these com-plexes (Fig. 6C). SMRT�HDAC complexes were prepared usingimmunoprecipitated SMRT from 293T cells and subsequentincubation with in vitro translated HDAC3 or HDAC4. 293Tcells were transfected with the CMV.SMRT plasmid. Whole cellextracts were immunoprecipitated under stringent conditionswith an anti-SMRT antibody, and protein G-agarose beadswere extensively washed. The presence of SMRT in the beadswas verified by Western blot analysis using an anti-SMRTantibody (Fig. 6D, bottom). The SMRT-containing beads werethen incubated with in vitro translated HDACs (lanes 3–6 and8–10). The SMRT�HDAC complexes were then washed andincubated with in vitro translated 35S-HNF4�1 (Fig. 6D, lanes4 and 9) or 35S-HNF4�7 (Fig. 6D, lanes 5 and 10). As controls,SMRT-containing beads were incubated with in vitro trans-

lated 35S-HNF4�1 (Fig. 6D, lane 1) or 35S-HNF4�7 (Fig. 6D,lane 2). HNF4�1 as well as HNF4�7 bound to SMRT in theabsence of any HDAC (Fig. 6D, lanes 1 and 2). These results arein line with the coimmunoprecipitation and GST pulldownexperiments in Fig. 4. In vitro translated class I HDAC,HDAC3, and class II HDAC, HDAC4 both formed stable com-plexes with SMRT in vitro (Fig. 6D, lanes 6 and 8).

In lanes 3–6 of Fig. 6D, SMRT�HDAC3 complexes were used.Because HDAC3 migrates at the same position as HNF4�7(compare input lanes for HDAC3 and HNF4�7), we used non-labeled HDAC3 (lanes 3–5) for further incubations with 35S-HNF4�. An 35S-HDAC3 control (lane 6) was performed in par-allel. Both HNF4�1 and HNF4�7 bound to the SMRT�HDAC3complex (Fig. 6D, lanes 4 and 5). Surprisingly, however, thepresence of HDAC3 greatly increased the amount of HNF4�1bound to the complex (from 0.1 to 1.5% of the input, comparelanes 1 and 4 in Fig. 6D) and provoked no change for HNF4�7(compare lanes 2 and 5). Thus, it appears that the presence ofHDAC3 either renders more stable the HNF4�1�SMRT�HDAC3complex or induces an increase in affinity for HNF4�1.

In lanes 7–10 of Fig. 6D, SMRT�35S-HDAC4 complexes wereused for incubation with 35S-HNF4�1 and 35S-HNF4�7. BothHNF4� isoforms were able to bind to these complexes (Fig. 6D,lanes 9 and 10), and no differences were observed concerningthe binding of the two isoforms. Again, the presence of thedeacetylase appears to stabilize or increase the affinity forHNF4�1 only (compare lanes 1, 2, and 9, 10).

Thus, both HNF4� isoforms can associate to the two types ofSMRT/HDAC class I and class II complexes in vitro. Moreover,association of SMRT with HDACs enhanced recruitment ofHNF4�1 but not of HNF4�7. Thus, the presence of an HDACaffected the binding of one but not the other NR isoform.

The association of HNF4� with HDACs was constitutive asrevealed by the HDAC assays on immunoprecipitated HNF4�(Fig. 6B). It is possible that once HNF4� translocates to itstarget, association with a SMRT/HDAC complex is determinedby the promoter context. In that case, the affinity of the com-plex could be affected by specific DNA binding mediated by theNR partner. Indeed, binding of a NR to its target site on DNAcould allosterically modulate its interaction with other proteinsby provoking the acquisition of a different “DNA-bound” con-formation. We tested this hypothesis indirectly by analyzingthe sensitivity of both isoforms to protease digestion in theabsence and presence of DNA.

HNF4�1 Exhibits Enhanced Protease Resistance uponBinding to DNA

Because the amino acid composition of the two isoforms isdifferent at the A/B domain, it could be anticipated that differ-ences occur in the folding of the two proteins. We analyzed theprotease sensitivity exhibited by both isoforms alone or afterincubation with the apoC-III oligonucleotide (Fig. 7). Diges-tions were carried out with trypsin on in vitro translated 35S-HNF4�1 and 35S-HNF4�7. No major differences concerningprotection of the longer fragments of HNF4�1 compared with-HNF4�7 were observed in the absence of DNA (compare rela-tive band intensities in lanes digested with 1 ng/�l in Fig. 7, noDNA). Thus, no significant differences in resistance to trypsindigestion were observed between HNF4�1 and HNF4�7 in theabsence of DNA. However, a slight protection of long fragmentswas observed upon trypsin digestion in the presence of theapoC-III oligonucleotide for HNF4�1 but not for the HNF4�7isoform (compare lanes digested with 1 and 5 ng/�l with andwithout DNA in Fig. 7). These results suggest that HNF4�1may fold differently upon DNA binding compared with itsnative “unbound” conformation. Such a change could confer a

HNF4� Isoforms and Coregulator Interaction 44683

by guest on February 14, 2018http://w

ww

.jbc.org/D

ownloaded from

FIG. 6. HNF4�1 and HNF4�7 recruit different HDAC activities to target reporters and associate with SMRT�HDAC complexes invitro. A, 293T cells were cotransfected with 200 ng of empty (pCB6), HNF4�1, or HNF4�7 expression vector and 1 �g of the indicated reporter.Cells were incubated with or without 100 ng of TSA for 8 h before harvesting. Luciferase activity was measured and normalized to the internalcontrol �-gal activity. Numbers above the bars indicate the fold difference of transactivation obtained in the presence of TSA. Shown arerepresentatives of two independent experiments performed in triplicate. B, endogenous HDAC activity coprecipitates with HNF4�. 293T cells weretransfected with HNF4�1 or HNF4�7 expression vectors. Whole cell extracts were precipitated with the �445 antiserum, and immunocomplexesor crude lysate (equivalent to 1/250 of the volume used for each immunoprecipitation reaction) were assayed for HDAC activity in the absence orpresence of the nonspecific inhibitor sodium butyrate. C, diagram for the experiment performed in D. SMRT�HDAC complexes were prepared byincubating immunoprecipitated (IP) SMRT first with in vitro translated HDACs and then with in vitro translated HNF4�. As controls,

HNF4� Isoforms and Coregulator Interaction44684

by guest on February 14, 2018http://w

ww

.jbc.org/D

ownloaded from

different affinity for corepressors and/or coactivators uponbinding to target genes.

Taken together, our results indicate that NR isoforms can berepressed differently by SMRT, likely because of quantitativeand qualitative differences in their abilities to recruit HDACactivity to target promoters. Although we have demonstrated asignificant functional difference within the A/B domain of bothHNF4� isoforms, this domain does not seem to be directlyinvolved in the transcriptional repression mediated by SMRT.Rather, the different amino acid composition of the A/B domaincould provoke conformational changes that are responsible forthe association of HNF4� isoforms with distinct SMRT/HDACcomplexes upon DNA binding.

DISCUSSION

The transcriptional activities of NR are the consequence ofthe combined activity of the AF-1 and AF-2 modules. Themechanisms of action of the AF-1 domain have been only partlyelucidated. In view of the lack of homology in this region amongNRs, diverse modes of action may be involved. It has beensuggested that AF-1 of the progesterone and estrogen receptorsis responsible for the cell- and promoter-specific activity (61,62). AF-2 is generally ligand-dependent and requires the bridg-ing of coactivators for transcriptional activity.

Here we have compared the parameters of activation andrepression of two isoforms of the orphan receptor HNF4� whichdiffer only in their N-terminal domains (Table I). We haveshown that the N-terminal AF-1 of HNF4�1 possesses signifi-cant activity that can be enhanced through interaction with

GRIP-1 and CBP. In striking contrast, HNF4�7 possesses nomeasurable AF-1 (Figs. 1 and 2), implying major functionaldifferences between the isoforms. Although HNF4�1 andHNF4�7 are both able to interact via the AF-2 with GRIP-1,CBP, and SMRT, only HNF4�1 is able to interact in a syner-gistic fashion with both GRIP-1 and p300 (Figs. 3–5). Moreover,enhanced binding to SMRT in the presence of HDACs is char-acteristic of HNF4�1 only, suggesting a novel level of regula-tion of NR activity via selection of HDAC partners (Fig. 6).

Because HNF4�7 does not contain the AF-1 function, itsactions are presumably mediated solely through the AF-2,which is functionally coactivated by the p160 and p300 cointe-grators. The AF-1 and AF-2 modules of HNF4� have beenshown to act cooperatively (56) and to interact physically withcoactivators as well as with members of the basal machinery oftranscription (8, 38, 39). We conclude that the AF-1 and AF-2 ofHNF4�1 are able to synergize through the action of the p160and p300 coactivators and that the AF-1 is necessary for thesimultaneous action of these two families of coregulators. In-deed, previous studies documented AF-2-independent interac-tion between HNF4�1 and GRIP-1 in vitro (38). The ability ofthe AF-1 and AF-2 to interact with separate surfaces of coac-tivators could be important for the ability of both transactiva-tion functions to synergize. Indeed, the AF-1 could constitutean additional point of contact in HNF4�1 to stabilize the p160-p300-NR interaction or to facilitate the formation of a stablepreinitiation complex on target promoters (63). p160 coactiva-tors have been shown to interact with the AF-1 of other NRs(10, 11, 64), and this family of coactivators may thus be impli-cated in determining the cell or promoter-specific actions ofNRs.

It has recently been shown that recruitment of HDAC4 to aSMRT/NCoR complex does not result in activation of the com-plex unless another class I HDAC, HDAC3, is associated withthe complex (29). Indeed, these observations suggest that se-lective recruitment of HDACs may be mediated by SMRT andcould be a mechanism to regulate specific cell differentiationprograms (24). Moreover, SMRT serves not only as a platformthat recruits the components of the complex to the DNA-boundNR, but actively participates in repression by inducing deacety-lase activity of HDACs through its deacetylase activating do-

immunoprecipitated SMRT was incubated with in vitro translated HNF4�1 and HNF4�7 in the absence of HDACs. D, HNF4� associates with classI and class II HDAC�SMRT complexes in vitro. 293T cells were transfected with the CMV.SMRT plasmid. Whole cell extracts were precipitatedwith an anti-SMRT antibody, and protein G-agarose beads were washed extensively. First, the SMRT-containing beads were incubated with invitro translated HDAC3 (lanes 3–5), 35S-HDAC3 (lane 6), or 35S-HDAC4 (lanes 8–10). After this first incubation, beads were washed extensivelyto remove nonbound proteins and were subsequently incubated with either 35S-HNF4�1 (lanes 4 and 9), 35S-HNF4�7 (lanes 5 and 10), or emptyvector-TNT product (lanes 1–3, 6–8). The SMRT interaction with 35S-HNF4�1 and 35S-HNF4�7 in the absence of HDACs is shown in lanes 1 and2. Complexes were resolved by SDS-PAGE and exposed for autoradiography and PhosphorImager quantification. The percentage of bound proteinwas calculated using the values corresponding to the 10% input as reference. We systematically observe a second, fast migrating band as a resultof the in vitro translation of HNF4�1 (which probably corresponds to partial translation products) which was not taken into account forquantification. The presence of SMRT was revealed in the same filter using an anti-SMRT antibody (bottom).

FIG. 7. HNF4�1 exhibits enhanced protease resistance in thepresence of DNA. In vitro translated 35S-labeled HNF4�1 or HNF4�7was incubated for 10 min on ice with or without 200 ng (equivalent toa 50 molar-fold excess) of double-stranded oligonucleotide containingthe HNF4� binding site from the apoC-III human promoter. Afterincubation with DNA, HNF4� proteins were subjected to protease di-gestion with increasing amounts of trypsin. Digestion products wereseparated by SDS-PAGE and exposed to autoradiography. The arrowshows a long fragment for HNF4�1 which is protected only upon incu-bation with DNA.

TABLE IPhysical and functional interactions with coregulators of AF-1 and

AF-2 modules of HNF4� isoforms

AF1 (N-terminal) AF2 (C-terminal)

HNF4�1Activity � �Recruitment � �Coactivatorsa � �Corepressorb � �

HNF4�7Activity � �Recruitment � �Coactivators � �Corepressor � �

a Coactivators, GRIP-1, CBP, and p300.b Corepressor, SMRT.

HNF4� Isoforms and Coregulator Interaction 44685

by guest on February 14, 2018http://w

ww

.jbc.org/D

ownloaded from

main (30). Nevertheless, the deacetylase activating domain ofSMRT activates HDAC3 but not HDAC4 upon recruitment tothe complex (30). Another mechanism of regulation of the core-pressor complex activity has been documented in which bind-ing of the NR to SMRT through domains other than RIDprevents the assembly of the other partners of the complex (65).Here, even though recruitment of SMRT takes place, the activ-ity of the NR is not repressed, a phenomenon known as “anti-repression” (65).

Thus, SMRT-mediated repression must be considered a mul-tistep process, and each step is potentially open to regulation.First, SMRT is recruited to the DNA-bound NR, and then thecomplex is assembled by association of mSin3A (16) or TBL1 (4)and selective recruitment of HDACs. The HDACs are in turnspecifically activated by interaction with SMRT, and repres-sion can then occur.

We showed that the HNF4�1 and HNF4�7 isoforms arerepressed in a different fashion by the corepressor SMRT.Although both isoforms interact with the same region of SMRTin vitro and with apparently similar affinity, the activity ofHNF4�7 is significantly less repressed in cotransfection exper-iments. It can be considered that HNF4�7 exhibits a mild formof antirepression compared with HNF4�1. Using the criterionof inhibition of endogenous HDAC activity by TSA, we foundthat both isoforms are able to recruit HDAC activity to twotarget promoters, but the degree of relief of inhibition was bothisoform- and target promoter-dependent. Such differences ofrecruitment of HDAC activity could be explained by associationwith SMRT complexes displaying different composition/activi-ties. In other words, HNF4�7 could preferentially recruit inac-tive SMRT complexes such as SMRT�HDAC4 (29) instead ofSMRT�HDAC3. In support of this hypothesis, our data indicatethat within a NR�SMRT�HDAC complex, the presence of class Ior class II HDAC may increase binding of a particular isoformof a NR. Moreover, we showed that both isoforms are able toassociate with SMRT�HDAC3 and SMRT�HDAC4 complexes invitro. Because the activity displayed by these deacetylases dif-fers (29), association with specific SMRT�HDAC complexeswithin the cell could represent a critical step in the regulationof NR activity.

Differences in protein folding resulting from DNA binding ofthe two isoforms could also selectively affect the recruitment ofSMRT complexes. In this sense, activity of the corepressorcomplex may be dependent not only on the NR isoform, but onthe DNA binding site as well. Indeed, HNF4�1, unlikeHNF4�7, exhibited enhanced protease resistance upon specificbinding to DNA (Fig. 7).

The AF-1 of NR provides a surface of contact for coactivators,but its role in repression remains to be established. Removal ofthe A/B domain does not affect SMRT binding to HNF4�1 invitro, but a truncated protein lacking both the A/B and Fdomains shows increased binding of SMRT (36). It can bespeculated that the A/B domain is involved in regulating con-formation of the F domain, which could then affect functionalinteractions with SMRT (36). Indeed, changes in the conforma-tion of the C-terminal domain of other NRs can regulate asso-ciation of the receptor with SMRT (18, 22, 66). We have shownthat the AF-1 of HNF4�1 is not affected by SMRT, whose maintarget for repression is the AF-2 of both isoforms. For theNRs whose ligands are well characterized, repression bySMRT�NCoR occurs only for the unliganded receptor, providinga mechanism to reinforce ligand inducibility. For the orphanreceptors such as HNF4� or COUP-TF, the unresponsivenessof the AF-1 to repression could provide a basis to explain theirconstitutive activity (67). Moreover, in the case of HNF4� iso-

forms, only HNF4�1 possesses a functional AF-1 that could actto antagonize SMRT repression.

Transcription factors whose activity is modulated by coacti-vators and corepressors can be controlled by the physiologicallevels of expression of coregulators. Messenger levels of SMRT,NCoR, and coactivators can vary from one cell type to anotherand during differentiation (68, 69). Indeed, in the liver, SMRTtranscripts are entirely absent for a transitory period at birth,2

at the very stage when expression of HNF4�1 is dramaticallyinduced (40) and the liver must assume the functions necessaryto render metabolism of the newborn independent of maternalcirculation. Thus, a number of genes involved mainly in lipidand glucose metabolism, which are also HNF4� targets (35),must be activated within a short time. We proposed previouslythat HNF4�1 could be involved in activation of expression ofhepatic neonatal functions because it is expressed mainly inadult liver and because it transactivates robustly reportergenes for adult hepatic functions (40). The down-regulation ofSMRT could facilitate the activation of these genes through theaction of HNF4�1 in the absence of a corepressor that inhibitsits activity. Further, in adult liver, when rapid metabolicchanges are required in response to nutritional state, SMRT ispresent in concert with the HNF4� isoform that is more sensi-tive to SMRT repression.

We have compared the functional and physical interactionswith coregulators of HNF4� isoforms whose expression is de-velopmentally regulated. The embryonic isoform, HNF4�7,possesses only one activation function, AF-2, that can intereactwith both coactivators and corepressors. The adult isoform,HNF4�1, possesses two activation functions, AF-1 being able tointeract only with coactivators and AF-2 with both classes ofcoregulators (Table I). What could be the utility to the orga-nism of expressing NR isoforms with only one or with twoactivation domains? One potentially important difference wasdocumented: synergy between two different families of coacti-vators occurs only for HNF4�1. If coregulators can be associ-ated simultaneously with two domains, cases not only of syn-ergy but also of antagonism between coactivators andcorepressors should be possible.

The fetal liver is in a relatively constant environment, butthe adult liver is constantly aggressed, by nutritional needsand xenobiotics, such that under some circumstances HNF4target genes may need to respond simultaneously to inductionand to repression. In this context, the HNF4�1 bipartite factorthat is subject to synergy or to antagonism in its transcrip-tional capacities may be optimal. In any case, the existence ofHNF4� isoforms that show differences in the recruitment ofcoregulators and their effectors in a site/promoter-specific con-text for activation and repression adds a new level of complex-ity and of flexibility to hepatocyte-specific gene regulation by aNR.

Acknowledgments—We thank G. Hautbergue for providing the pGEXand pSET, M. L. Privalsky for all of the SMRT constructs, S. Emilianifor the HDAC3 and HDAC4 constructs, R. Goodman for the CBP ex-pression vector, M. R. Stallcup for the GRIP-1 plasmids, and MarionMathieu for kindly providing the histone H4 acetylated peptide. Wethank C. Mulet for DNA plasmid preparation and A. Israel for gener-ously sharing facilities. We are particularly indebted to M. R. Stallcup,C. Deschatrette, D. Faust, G. Hayhurst, and R.Weil for helpful discus-sions and to M. D. Ruse for sharing results prior to publication.

REFERENCES

1. Mangelsdorf, D. J., Thummel, C., Beato, M., Umesono, K., Blumberg, B.,Kastner, P., Mark, M., Chambon, P., and Evans, R. M. (1995) Cell 83,835–839

2. Gelmetti, V., Zhang, J., Fanelli, M., Minucci, S., Pelicci, P. G., and Lazar, M. A.(1998) Mol. Cell. Biol. 18, 7185–7191

2 M. E. Torres-Padilla and M. C. Weiss, manuscript in preparation.

HNF4� Isoforms and Coregulator Interaction44686

by guest on February 14, 2018http://w

ww

.jbc.org/D

ownloaded from

3. Grignani, F., De Matteis, S., Nervi, C., Tomassoni, L., Gelmetti, V., Cioce, M.,Fanelli, M., Ruthardt, M., Ferrara, F. F., Zamir, I., Seiser, C., Lazar, M. A.,Minucci, S., and Pelicci, P. G. (1998) Nature 391, 815–818

4. Guenther, M. G., Lane, W. S., Fischle, W., Verdin, E., and Lazar, M. (2000)Genes Dev. 14, 1048–1057

5. Stallcup, M. R. (2001) Oncogene 20, 3014–30206. Sterner, D. E., and Berger, S. L. (2000) Microbiol. Mol. Biol. Rev. 64, 435–4597. Torchia, J., Glass, C. K., and Rosenfeld, M. G. (1998) Curr. Opin. Cell Biol. 10,

373–3838. Green, V. J., Kokkotou, E., and Ladias, J. A. (1998) J. Biol. Chem. 273,

29950–299579. Hittelman, A. B., Burakov, D., Iniguez-Lluhi, J. A., Freedman, L. P., and

Garabedian, M. J. (1999) EMBO J. 18, 5380–538810. Ma, H., Hong, H., Huang, S. M., Irvine, R. A., Webb, P., Kushner, P. J.,

Coetzee, G. A., and Stallcup, M. R. (1999) Mol. Cell. Biol. 19, 6164–617311. Tremblay, G. B., Tremblay, A., Labrie, F., and Giguere, V. (1999) Mol. Cell.

Biol. 19, 1919–192712. Benecke, A., Chambon, P., and Gronemeyer, H. (2000) EMBO Rep. 1, 151–15713. Chen, J. D., and Evans, R. M. (1995) Nature 377, 454–45714. Heinzel, T., Lavinsky, R. M., Mullen, T. M., Soderstrom, M., Laherty, C. D.,

Torchia, J., Yang, W. M., Brard, G., Ngo, S. D., Davie, J. R., Seto, E.,Eisenman, R. N., Rose, D. W., Glass, C. K., and Rosenfeld, M. G. (1997)Nature 387, 16–17

15. Horlein, A. J., Naar, A. M., Heinzel, T., Torchia, J., Kurokawa, R., Ryan, A.,Kamei, Y., Soderstrom, M., Glass, C. K., and Rosenfeld, M. G. (1995) Nature377, 397–404

16. Nagy, L., Kao, H. Y., Chakravarti, D., Lin, R. J., Hassig, C. A., Ayer, D. E.,Schreiber, S. L., and Evans, R. M. (1997) Cell 89, 373–380

17. Bailey, P., Downes, M., Lau, P., Harris, J., Chen, S. L., Hamamori, Y.,Sartorelli, V., and Muscat, G. E. (1999) Mol. Endocrinol. 13, 1155–1168

18. Li, H., Leo, C., Schroen, D. J., and Chen, J. D. (1997) Mol. Endocrinol. 11,2025–2037

19. Wong, C. W., and Privalsky, M. L. (1998) Mol. Cell. Biol. 18, 5500–551020. Privalsky, M. L. (2001) Curr. Top. Microbiol. Immunol. 254, 117–13621. Lavinsky, R. M., Jepsen, K., Heinzel, T., Torchia, J., Mullen, T. M., Schiff, R.,

Leon Del-Rio, A., Ricote, M., Ngo, S. D., Gemsch, J., Hilsenbeck, S. G.,Osborne, C. K., Glass, C. K., Rosenfeld, M. G., and Rose, D. W. (1998) Proc.Natl. Acad. Sci. U. S. A. 95, 2920–2925

22. Lin, B. C., Hong, S. H., Krig, S., Yoh, S. M., and Privalsky, M. L. (1997) Mol.Cell. Biol. 17, 6131–6138

23. Wong, C. W., and Privalsky, M. L. (1998) Mol. Cell. Biol. 18, 5724–573324. Wu, X., Li, H., Park, E. J., and Chen, J. D. (2001) J. Biol. Chem. 276,

24177–2418525. Carmen, A. A., Rundlett, S. E., and Grunstein, M. (1996) J. Biol. Chem. 271,

15837–1584426. Kao, H. Y., Downes, M., Ordentlich, P., and Evans, R. M. (2000) Genes Dev. 14,

55–6627. Rundlett, S. E., Carmen, A. A., Kobayashi, R., Bavykin, S., Turner, B. M., and

Grunstein, M. (1996) Proc. Natl. Acad. Sci. U. S. A. 93, 14503–1450828. Taunton, J., Hassig, C. A., and Schreiber, S. L. (1996) Science 272, 408–41129. Fischle, W., Dequiedt, F., Hendzel, M. J., Guenther, M. G., Lazar, M., Voelter,

W., and Verdin, E. (2002) Mol. Cell 9, 45–5730. Guenther, M. G., Barak, O., and Lazar, M. (2001) Mol. Cell. Biol. 21,

6091–610131. Sladek, F. M., Zhong, W. M., Lai, E., and Darnell, J. E., Jr. (1990) Genes Dev.

4, 2353–236532. Chen, W. S., Manova, K., Weinstein, D. C., Duncan, S. A., Plump, A. S.,

Prezioso, V. R., Bachvarova, R. F., and Darnell, J. E., Jr. (1994) Genes Dev.8, 2466–2477

33. Hayhurst, G. P., Lee, Y. H., Lambert, G., Ward, J. M., and Gonzalez, F. J.(2001) Mol. Cell. Biol. 21, 1393–1403

34. Li, J., Ning, G., and Duncan, S. A. (2000) Genes Dev. 14, 464–47435. Sladek, F. M., and Seidel, S. D. (2001) in Nuclear Receptors and Genetic

Diseases (Burris, T. P., and McCabe, E., eds) Academic Press, London36. Ruse, M. D., Jr., Privalsky, M. L., and Sladek, F. M. (2002) Mol. Cell. Biol. 22,

1626–163837. Dell, H., and Hadzopoulou-Cladaras, M. (1999) J. Biol. Chem. 274, 9013–902138. Sladek, F. M., Ruse, M. D., Jr., Nepomuceno, L., Huang, S. M., and Stallcup,

M. R. (1999) Mol. Cell. Biol. 19, 6509–652239. Yoshida, E., Aratani, S., Itou, H., Miyagishi, M., Takiguchi, M., Osumu, T.,

Murakami, K., and Fukamizu, A. (1997) Biochem. Biophys. Res. Commun.241, 664–669

40. Torres-Padilla, M. E., Fougere-Deschatrette, C., and Weiss, M. C. (2001) Mech.Dev. 109, 183–193

41. Brewer, C. (1994) Methods Cell Biol. 43, 233–24442. Spath, G. F., and Weiss, M. C. (1997) Mol. Cell. Biol. 17, 1913–192243. Sladek, F. M., Dallas-Yang, Q., and Nepomuceno, L. (1998) Diabetes 47,

985–99044. Hong, S. H., and Privalsky, M. L. (2000) Mol. Cell. Biol. 20, 6612–662545. Ding, X. F., Anderson, C. M., Ma, H., Hong, H., Uht, R. M., Kushner, P. J., and

Stallcup, M. R. (1998) Mol. Endocrinol. 12, 302–31346. Chrivia, J. C., Kwok, R. P., Lamb, N., Hagiwara, M., Montminy, M. R., and

Goodman, R. H. (1993) Nature 365, 855–85947. Eckner, R., Ewen, M. E., Newsome, D., Gerdes, M., DeCaprio, J. A., Lawrence,

J. B., and Livingston, D. M. (1994) Genes Dev. 8, 869–88448. Emiliani, S., Fischle, W., Van Lint, C., Al-Abed, Y., and Verdin, E. (1998) Proc.

Natl. Acad. Sci. U. S. A. 95, 2795–280049. Grozinger, C. M., Hassig, C. A., and Schreiber, S. L. (1999) Proc. Natl. Acad.

Sci. U. S. A. 96, 4868–487350. Sadowski, I., and Ptashne, M. (1989) Nucleic Acids Res. 17, 753951. Manet, E., Allera, C., Gruffat, H., Mikaelian, I., Rigolet, A., and Sergeant, A.

(1993) Gene Expression 3, 49–5952. Schoepfer, R. (1993) Gene (Amst.) 124, 83–8553. Maeda, Y., Seidel, S. D., Wei, G., Liu, X., and Sladek, F. M. (2002) Mol.

Endocrinol. 16, 402–41054. Thompson, J. F., Hayes, L. S., and Lloyd, D. B. (1991) Gene (Amst.) 103,

171–17755. Nakhei, H., Lingott, A., Lemm, I., and Ryffel, G. U. (1998) Nucleic Acids Res.

26, 497–50456. Hadzopoulou-Cladaras, M., Kistanova, E., Evagelopoulou, C., Zeng, S.,

Cladaras, C., and Ladias, J. A. (1997) J. Biol. Chem. 272, 539–55057. Kan, H. Y., Georgopoulos, S., and Zannis, V. (2000) J. Biol. Chem. 275,

30423–3043158. Karathanasis, S. K. (1985) Proc. Natl. Acad. Sci. U. S. A. 82, 6374–637859. Hassig, C. A., Fleischer, T. C., Billin, A. N., Schreiber, S. L., and Ayer, D. E.

(1997) Cell 89, 341–34760. Lizcano, F., Koibuchi, N., Fukuda, H., Dangond, F., and Chin, W. W. (2001)

Mol. Cell. Endocrinol. 172, 13–2061. Berry, M., Metzger, D., and Chambon, P. (1990) EMBO J. 9, 2811–281862. Tora, L., Gronemeyer, H., Turcotte, B., Gaub, M. P., and Chambon, P. (1988)

Nature 333, 185–18863. Elliston, J. F., Fawell, S. E., Klein-Hitpass, L., Tsai, S. Y., Tsai, M. J., Parker,

M. G., and O’Malley, B. W. (1990) Mol. Cell. Biol. 10, 6607–661264. Webb, P., Nguyen, P., Shinsako, J., Anderson, C., Feng, W., Nguyen, M. P.,

Chen, D., Huang, S. M., Subramanian, S., McKinerney, E.,Katzenellenbogen, B. S., Stallcup, M. R., and Kushner, P. J. (1998) Mol.Endocrinol. 12, 1605–1618

65. Yang, Z., Hong, S. H., and Privalsky, M. L. (1999) J. Biol. Chem. 274,37131–37138

66. Sande, S., and Privalsky, M. L. (1996) Mol. Endocrinol. 10, 813–82567. Guigere, V. (1999) Endocr. Rev. 20, 689–72568. Misiti, S., Schomburg, L., Yen, P. M., and Chin, W. W. (1998) Endocrinology

139, 2493–250069. Soderstrom, M., Vo, A., Heinzel, T., Lavinsky, R. M., Yang, W. M., Seto, E.,

Peterson, D. A., Rosenfeld, M. G., and Glass, C. K. (1997) Mol. Endocrinol.11, 682–692

HNF4� Isoforms and Coregulator Interaction 44687

by guest on February 14, 2018http://w

ww

.jbc.org/D

ownloaded from

Maria Elena Torres-Padilla, Frances M. Sladek and Mary C. WeissCorepressor-Histone Deacetylase Complexes

Mediate Multiple Interactions through Coactivator andαNuclear Factor 4Developmentally Regulated N-terminal Variants of the Nuclear Receptor Hepatocyte

doi: 10.1074/jbc.M207545200 originally published online August 29, 20022002, 277:44677-44687.J. Biol. Chem. 

  10.1074/jbc.M207545200Access the most updated version of this article at doi:

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

  http://www.jbc.org/content/277/47/44677.full.html#ref-list-1

This article cites 68 references, 36 of which can be accessed free at

by guest on February 14, 2018http://w

ww

.jbc.org/D

ownloaded from