Cloning and characterization of RAP250, a novel nuclear receptor coactivator.

Ligand-induced transcriptional activation of gene expression by nuclear receptors is dependent on recruitment of coactivators as intermediary factors. The present work describes the cloning and characterization of RAP250, a novel human nuclear receptor coactivator. The results of in vitro and in vivo experiments indicate that the interaction of RAP250 with nuclear receptors is ligand-dependent or ligand-enhanced depending on the nuclear receptor and involves only one short LXXLL motif called nuclear receptor box. Transient transfection assays further demonstrate that RAP250 has a large intrinsic glutamine-rich activation domain and can significantly enhance the transcriptional activity of several nuclear receptors, acting as a coactivator. Interestingly, Northern blot and in situ hybridization analyses reveal that RAP250 is widely expressed with the highest expression in reproductive organs (testis, prostate and ovary) and brain. Together, our data suggest that RAP250 may play an important role in mammalian gene expression mediated by nuclear receptor.

Ligand-induced transcriptional activation of gene expression by nuclear receptors is dependent on recruitment of coactivators as intermediary factors. The present work describes the cloning and characterization of RAP250, a novel human nuclear receptor coactivator. The results of in vitro and in vivo experiments indicate that the interaction of RAP250 with nuclear receptors is ligand-dependent or ligand-enhanced depending on the nuclear receptor and involves only one short LXXLL motif called nuclear receptor box. Transient transfection assays further demonstrate that RAP250 has a large intrinsic glutamine-rich activation domain and can significantly enhance the transcriptional activity of several nuclear receptors, acting as a coactivator. Interestingly, Northern blot and in situ hybridization analyses reveal that RAP250 is widely expressed with the highest expression in reproductive organs (testis, prostate and ovary) and brain. Together, our data suggest that RAP250 may play an important role in mammalian gene expression mediated by nuclear receptor.
The nuclear receptor (NR) 1 superfamily is a large group of structurally related transcription factors that regulate target gene transcription in response to ligands. The complex genetic programs regulated by NRs include biological processes such as growth, cell differentiation, and homeostasis (1). They can be divided into several subfamilies on the basis of characteristics such as dimerization status, nature of the ligand, or structure of the DNA response element. NRs are characterized by a common domain structure, including a highly variable N-terminal region that contains a constitutive activation function (AF-1), a highly conserved DNA binding domain (DBD) responsible for recognition of specific DNA response elements, a conserved multifunctional C-terminal ligand binding domain (LBD), containing a dimerization and a ligand-dependent transactivation function (AF-2) (2). The liganded NRs bind to their cognate hormone response elements, located in the promoter or enhancer regions of target genes, and stimulate transcriptional activation by transmitting signals to the transcriptional machinery via direct protein-protein interactions (3)(4)(5). In addition, another class of proteins, called coactivators, are recruited and serve as bridging molecules between the transcription initiation complex and NRs (for reviews, see Refs. 6 and 7). Most of the coactivators interact with the AF-2 domain of NRs through one or several LXXLL motifs called NR boxes (8 -12). Bona fide AF-2 coactivators include the three related members of the p160/SRC family, as well as the cointegrators CBP and p300 (for review, see Ref. 6). Because these coactivators possess intrinsic histone acetyltransferase activity and function in complex with other acetyltransferases, such as P/CAF, it has been proposed that functional connections exist between NR activation and the histone acetylation status. Evidence for the existence of NR-coactivator complexes came from biochemical studies identifying the TRAP/DRIP complex (13)(14)(15)(16), which may function more directly through contacts to the basal machinery. In addition to coactivators, other AF-2-binding proteins, such as RIP140 (17,18) or the nuclear orphan receptor SHP (19), may serve important regulatory functions by inhibiting NR activation.
To identify new potential coactivators, a mouse embryo cDNA library was screened using the yeast two-hybrid system with PPAR␣ as bait. Here, we report the cloning and characterization of RAP250, a new NR coactivator.

EXPERIMENTAL PROCEDURES
Plasmids-All constructs were generated using standard cloning procedures and verified by restriction enzyme analysis and DNA sequencing. Details of each construction are available upon request. The partial mouse RAP250 cDNA fragment encoding amino acids 782-1138 was released from the EcoRI site of the pGAD10 clone isolated by the yeast two-hybrid system and subcloned into the EcoRI site of pGEX-4T1 vector (Amersham Pharmacia Biotech). GST-hER␣ (aa 249 -595) and GST-hTR␣ (aa 122-410) have been described previously (19,20). Mutated variant of RAP250 was constructed by two independent PCRs with the flanking primers and two mutagenesis primers: 5Ј-TTAACGAGCCCATTGGCGGTCAACGCACTACAGAGT-GAC-3Ј and 5Ј-GTCACTCTGTAGTGCGTTGACCGCCAATGGGCTC-GTTAA-3Ј. The corresponding PCR products were isolated and combined together by an additional PCR with the flanking primers. The product of this PCR was digested by EcoRI and cloned into the corresponding site of pGEX-4T1. GST-RAP250 (aa 818 -931) was generated by PCR and cloned into the EcoRI/SalI sites of pGEX-4T1.
Yeast Two-hybrid Screening and 5Ј-RACE PCR-To isolate cDNAs encoding proteins that interact with PPAR␣, yeast two-hybrid screening was carried out as described previously for the isolation of hRIP140 (18). As bait, Gal4-PPAR␣ LBD/AF-2 was used to screen a mouse embryo cDNA library (CLONTECH) in the vector pGAD10. One clone revealed no homology with any characterized proteins but a strong homology with a human EST sequence of 6504 bp named KIAA0181 (24). The Kazusa DNA Research Institute provided us the human homologue cDNA. We used RACE PCR to obtain the remaining 5Јend sequence of the human RAP250. This PCR amplification was performed using human testis Marathon-ready cDNA (CLONTECH) as template. The first amplification was performed using the adaptor primer 1 and the gene-specific primer 5Ј-ATAGGAAATCCCGCCTCCATCCTA-3Ј for 30 cycles followed by a final elongation of 7 min. Each cycle consisted of 10 s at 94°C, 10 s at 63°C, and 1 min 30 s at 68°C; 1 l of the PCR product was used as a template for the second amplification with the adaptor primer 2 and the nested gene-specific primer 5Ј-CTGGTTGT-TGCTCTGAGCAAGGAT-3Ј for 30 cycles, using essentially the same conditions as those used for the first amplification. The PCR product was cloned into pGEM-T (Promega), and 10 independent clones were sequenced.
Northern Blot-Human multiple tissue Northern blots with approximately 2 g of poly(A) ϩ RNA per lane were hybridized according to the protocol of the manufacturer (CLONTECH). The probe used to detect RAP250 expression was a 0.8-kb EcoRI fragment (nucleotides 1104 -1828) of hRAP250 cDNA radioactively labeled by the random-prime method (Rediprime, Amersham Pharmacia Biotech). GAPDH cDNA was also used as a control probe. Some variations of the control GAPDH levels were observed, showing a strong expression in skeletal muscle and heart, as it is often the case with these two tissues. Nevertheless, according to manufacturer, these variations in GAPDH expression reflect a tissue-specific expression rather than a nonequal loading of the samples.
In Situ Hybridization-Adult male and female Harlan Sprague-Dawley rats, NMRI mice, and postnatal mice (1.5, 4, 8, and 14 days) were decapitated, and the tissues were excised and frozen on dry ice. Embryonic mice (e9 -e17) and rats (e12-e21) were excised from pregnant females and frozen. The tissues were sectioned with Microm HM-500 cryostat at 14 m and thawed on Polysine glasses (Menzel, Germany). In situ hybridization was carried out as described previously. Two oligonucleotide probes directed against the mouse RAP250 mRNA (nucleotides 3185-3219 and 3448 -3479 as numbered on human sequence) were used. The sequences had 100% ( 3185 GCACCCCCACCA-CAGCCACCACAGCAGCAGCCACA 3219 ) and 96% ( 3448 GCAAGGACCT-GCCTCTGTGCCACCATCACCTG 3479 ) homology to human RAP250 and less than 70% homology with any other known gene compared with the known sequences in the GenBank TM data base. Both probes produced similar results when used separately and were usually combined to intensify the hybridization signal. Several probes to nonrelated mRNAs with known expression patterns, with similar length and GC content, were used as controls to verify the specificity of the hybridizations.
In Vitro Translation of RAP250 cDNA-The human full-length RAP250 cDNA cloned in pSG5 was in vitro transcribed and translated using rabbit reticulocyte lysate (TNT coupled in vitro system Promega) according to the manufacturer's recommendations with [ 35 S]methionine. The radiolabeled protein was separated by 7% SDS-polyacrylamide gel electrophoresis, and the gel was then dried and autoradiographed.
Expression and Purification of GST Fusion Proteins-Cultures of Escherichia coli BL21(pLys) carrying the pGEX fusion constructs (RAP250, ER␣, and TR␣) were grown at 37°C in Luria-Bertani medium containing 100 g/ml ampicillin and 34 g/ml chloramphenicol and supplemented with 2% glucose. At A 600 0.6, the cultures were induced with 0.5 mM isopropyl-␤-D-thiogalactopyranoside for 2-3 h at 30°C. GST and GST-RAP250 proteins were purified as described previously (18), and the bacteria expressing GST fused to NRs were harvested by centrifugation and resuspended in STE buffer (10 mM Tris-HCl (pH 8), 150 mM NaCl, 1 mM EDTA) and frozen on dry ice. After thawing, lysozyme were added to a final concentration of 0.1 mg/ml, and the suspension was rotated at 4°C for 15 min. Then, dithiothreitol was added to a final concentration of 5 mM and Sarcosyl to 1.5%. After centrifugation at 10,000 ϫ g for 30 min at 4°C the lysates were added to glutathione-Sepharose 4B (Amersham Pharmacia Biotech) for 2 h at 4°C and washed three times with phosphate-buffered saline (PBS). To produce pure GST fusion protein for electrophoretic mobility shift assay, the proteins were eluted with 4 volumes of 20 mM glutathione in 50 mM Tris-HCl (pH 8). Protein concentrations were determined by the Bradford dye binding procedure (Bio-Rad).
In Vitro Protein-Protein Interaction Assay (GST Pull-down Assay)-All the NRs that we tested in pull-down assays were in vitro transcribed and translated using rabbit reticulocyte lysate (TNT-coupled in vitro system Promega) according to manufacturer's recommendations with and washed three times for 15 min with incubation buffer without bovine serum albumin. Washed beads were resuspended in 60 l of 1ϫ SDS sample buffer, and an aliquot was subject to SDS-polyacrylamide gel electrophoresis. Before autoradiography, gels were stained with Coomassie Blue to control for the stability of the GST fusion proteins and equal loading.
Electrophoretic Mobility Shift Assays-TR␣ and RXR␣ were synthesized in rabbit reticulocyte lysate by using the TNT-coupled in vitro transcription-translation system (Promega). Double-stranded synthetic oligonucleotides DR4-TRE 5Ј-TCGATCAGGTCATTTCAG-GTCAGAG-3Ј were radioactively labeled with [␣-32 P]dCTP. Binding reactions were performed in a total volume of 20 l in 1ϫ reaction buffer (5% glycerol, 5 mM dithiothreitol, 5 mM EDTA, 250 mM KCl, 100 mM HEPES (pH 7.5), 1 g of poly(dI-dC), 25 mM MgCl 2 , 1 mg of bovine serum albumin per ml, 1 g of salmon sperm DNA, 0.05% Triton X-100), 0.5 ng of labeled probe, 2 l of each in vitro translated receptor protein, and, when indicated, 2 l of the appropriate ligand in Me 2 SO. Finally, 0.5 g per reaction of the purified GST-RAP250 was added as indicated in results. The binding reaction was allowed to proceed for 20 min on ice before the reaction mixtures were loaded on a 4% nondenaturing polyacrylamide gel. After 3 h of electrophoresis in 0.5ϫ Tris-borate-EDTA (TBE) buffer at 4°C, the gels were dried and autoradiographed.
Cell Cultures and Transient Transfections-COS-7 cells were maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum, 100 g/ml penicillin, and 100 g/ml streptomycin (Life Technologies Inc.). For transient transfection assays, COS-7 cells were plated onto 6-well plates (Falcon) 24 h prior to transfection. Cells were transfected using Lipofectin as instructed by the manufacturer (Life Technologies Inc.). For each well, 0.5 g of reporter plasmid and 0.5 g of Gal4 expression plasmid were transfected. 24 h after transfection, cells were harvested and cell extracts were analyzed for luciferase activity as described (18). Nuclear extracts were prepared from transfected cells, and Western blotting using a mouse anti-Gal4-DBD monoclonal antibody (Santa Cruz Biotechology) was carried out as described previously (19). In the mammalian interaction and/or coactivation assays, 100 ng of wild type NR plasmids (TRs, RXR, and PPARs) and 0.5 g of VP16 or VP16-mRAP250 plasmid (interaction assay) or 0.5 g of pSG5 or pSG5-hRAP250 plasmid (coactivation assay) were transfected into COS-7 cells together with 0.5 g of the appropriate reporter plasmid. Transfections were carried out for 4 -6 h; the cells were then grown in presence of appropriate ligand or Me 2 SO for 36 h and then harvested and analyzed as described previously (18).  (24), starts at methionine ϩ78, whereas the cDNA sequence encoding the first 77 amino acids was obtained by RACE PCR. RAP250 contains two LXXLL motifs (boxed residues 887-891 and 1491-1495). The boxed part on protein sequence represents the location on human sequence of the partial clone originally isolated from the mouse embryo cDNA library (aa 782-1138). RAP250 contains two polyglutamine stretches (underlined) and, in between, a region rich in glutamine (20% Gln). The complete human protein sequence contains 2063 residues with a predicted molecular mass of 220 kDa. The nucleotide sequence contains an in-frame stop codon upstream of the first methionine (data not shown). B, in vitro translation of RAP250 cDNA. The human full-length RAP250 cDNA cloned in pSG5 was in vitro transcribed and translated using rabbit reticulocyte lysate with [ 35 S]methionine. Radiolabeled protein was fractionated on a 7% SDS-polyacrylamide gel electrophoresis gel and visualized using autoradiography. The size of hRAP250 was estimated to be 250 kDa by comparison to the size of mPBP (TRAP220) (27) and the molecular markers. C, alignment of the mouse RAP250 partial clone and the homologue region on human protein. The residues shown in black boxes in mouse RAP250 differ from the corresponding human residues. The two RAP250 sequences share 90% identity. The 55 amino acids surrounding the LXXLL motif (LVNLL) are identical in both species, providing a large consensus sequence within the interacting region.
GenBank TM Accession Numbers-The human RAP250 nucleotide and protein sequences have been submitted to GenBank TM data base with accession number AF128458, and the partial mouse RAP250 nucleotide and protein sequences have been submitted with accession number AF135169.

RESULTS
Isolation and Cloning of RAP250 cDNA-We used the yeast two-hybrid system to screen a mouse embryo cDNA library with PPAR␣-LBD as a bait as described previously (18,20). Of the isolated clones, more than 50% were isoforms of RXR, and a majority of the other clones were interacting parts of SMRT, N-CoR, TIF-2, and TRAP220 as described by Treuter et al. (18). However, one of the interacting clones revealed no homology with any described protein, and data base searches revealed a strong homology with a human EST sequence of 6504 bp named KIAA0181 (24). Compared with the human clone, this positive mouse clone only contained a partial cDNA sequence of 1.1 kb encoding 355 amino acids, corresponding to amino acids FIG. 2. Tissue distribution of RAP250 mRNA. A, a human multiple tissue Northern blot (CLONTECH) containing 2 g of poly(A ϩ ) RNA from each tissue was probed with 32 P-labeled RAP250 and GAPDH (control) cDNAs. Northern blotting revealed a major RAP250 transcript of approximately 7.5 kb, and this messenger was differentially transcribed among the various tissues. Two alternative transcripts of different sizes are also present in peripheral blood leukocyte (approximately 6 kb) and testis (approximately 4.5 kb). RNA molecular weights are indicated on the left in kilobases. B, RAP250 mRNA expression in mouse embryo (Ba-Bc), postnatal and adult mouse tissues (Bd), and adult rat tissues (Be-Bh). Ba, E9 mouse embryo: neural epithelium (ne) and placenta (pl). Bb, E15 mouse embryo: developing neocortex (co), thalamus (th), basal ganglia (bg), cerebellum (ce), spinal cord (sc), olfactory epithelium (oe), whisker follicles (wf), thymus (th), heart (he), liver (li), intestine (in), and kidney (ki). Bc, 1.5-day-old mouse: hippocampus (hc), tooth (to), salivary gland (sg), lung (lu), stomach (st), spleen (sp), and skin (sk). Bd, 8-day-old mouse eye: retina (arrowheads) and lens (le). Be, rat testis: seminiferous tubules (arrowheads). Bf, rat prostate. Bg, rat ovary: follicles (arrowheads), interstitial tissue (it), and corpus luteum (cl). Bh, mouse brain: cerebral cortex (co), hippocampus (hc), and piriform cortex (pco). 782-1138 of the human sequence (Fig. 1A) and showing 90% identity with the human protein (Fig. 1C). The human clone contained a long open reading frame starting at nucleotide position 60 and a stop codon at nucleotide 6020. A downstream polyadenylation signal was also present in this cDNA, indicating that the 3Ј-end of the gene was intact. However, the first in-frame ATG codon of this human clone was not in an optimal context for translational initiation (no perfect Kozak site (26) and no in-frame stop codon upstream of this first methionine), suggesting that it perhaps was not the real initiation codon. In order to get a longer 5Ј-end sequence, we performed a 5Ј-RACE PCR and amplified an additional sequence of 433 nucleotides, which included 231 nucleotides encoding 77 additional amino acids. The nucleotide sequence of the reconstituted full-length cDNA is 6878 bp in length. It contains a short 5Ј-untranslated region of 202 bp with an upstream stop codon in frame with the first methionine, a longer 3Ј-untranslated region of 484 bp, and a 6189-bp open reading frame that encodes a protein of 2063 amino acids with a calculated mass of 220 kDa (Fig. 1A). The beginning of the coding sequence was defined by the first ATG downstream of an in-frame stop codon at position Ϫ51. The sequence (ACCATGGTTTTG) surrounding the ATG essentially conforms to the Kozak site ((A/G)cc ATG Gat) (26). To determine the size of the protein, we performed an in vitro translation after synthesis of mRNA via a bacteriophage DNA-dependent RNA polymerase. As shown in Fig. 1B, the estimated M r of the most prevalent in vitro synthesized protein is 250,000, which is consistent with the calculated size. This protein was designated RAP250 (nuclear receptor-activating protein 250) to signify both its potential coactivation function and its size in kilodaltons. The human RAP250 shows some specific features such as a Gln-rich region flanked by two poly-Gln stretches in the N-terminal region (see Fig. 1A) and two copies of the LXXLL motif (LVNLL, aa 887-891, and LSQLL, aa 1494 -1495) that are consistent with a coactivator function. Of the two LXXLL motifs present in the human sequence, the mouse partial clone only contains the first one (Fig. 1C).
Human RAP250 mRNA Is Present in Many Tissues-North-ern blot analysis of human RNAs revealed a widespread major RAP250 transcript of approximately 7.5 kb in length, which was present at different levels depending on the tissue ( Fig.  2A). High levels were detected in reproductive organs, such as ovary, testis, and prostate, as well as in peripheral blood leukocytes, brain, and heart, and intermediate levels were observed in pancreas, kidney, liver, colon, spleen, and placenta. RAP250 mRNA levels were low but still detectable in small intestine, thymus, and skeletal muscle. Interestingly, in testis, a second transcript of approximately 4.5 kb in length was also detected. cDNA cloning and sequence analysis of this shorter mRNA indicated that it was an alternatively spliced form of RAP250 with an open reading frame encoding a 1070-amino acid-protein and encompassing amino acids 1-971 and 1965-2063 (data not shown). Thus, considering mRNA levels, testis appears to be the main RAP250 expressing organ. In peripheral blood leukocytes, a second 6-kb transcript was also detected that might represent either another alternatively spliced mRNA or a closely related but different gene product.
Ontogeny of RAP250 mRNA Expression-The levels of RAP250 mRNA in mouse and rat embryos were quite similar. RAP250 mRNA was widely detected during ontogeny. At embryonic day 9 (e9), clear signal was present in placenta, and lower expression could be seen in uterus (Fig. 2, Ba). At this stage, neural tube expressed high levels of RAP250 mRNA and the expression in central nervous system was high throughout ontogeny (Fig. 2, Ba-Bc). The expression in spinal cord and in cerebrum was high and became more restricted during later stages of development (e17 onwards) and postnatal life. High expression was seen in cerebellum during development of this subregion of the brain (Fig. 2, Bc). Also, sensory ganglia and retina showed high expression from e11 onwards (Fig. 2, Bd). In the alimentary tract (oral cavity, stomach, and intestine), expression was seen from e13 and thereafter (Fig. 2, Bb and Bc). The developing teeth and salivary gland were also labeled (Fig. 2, Bc). Olfactory epithelium was strongly labeled from e13 onwards (Fig. 2, Bb and Bc). Strong expression was present in liver (from e11) and kidney (from e13 onwards), and these levels decreased at later stages of development (Fig. 2, Bb and Bc). Lung had moderate signal from e13 and this level decreased during postnatal life (Fig. 2, Bc). Prominent signal was seen in thymus from e15 onwards, and in spleen from e17 and during early postnatal life, and subsequently, the expression decreased (Fig. 2, Bb and Bc). Low to moderate signal was seen in brown fat, as well as developing muscles, bones, and intervertebral discs. In adult mouse and rat, expression of RAP250 mRNA was more restricted than during embryonic development. High expression was observed in male and female rat genital organs. In testis, seminiferous tubules exhibited a strong signal and expression in separate tubules varied, indicating that RAP250 may be expressed in a stage-specific manner during spermatogenesis (Fig. 2, Be). In dipped sections, RAP250 mRNA could be seen in primary spermatocytes. Prom-inent expression was also seen in the epithelium of prostate, whereas epididymis and seminal vesicles had low signals (Fig.  2, Bf). In ovary, the strongest signal was seen in interstitial cells and in the granulosa cells of different size follicles (Fig. 2,  Bg). In the central nervous system, high expression was present in olfactory bulb, piriform cortex, hippocampus and cerebellar cortex, whereas other areas exhibited lower levels of RAP250 mRNA (Fig. 2, Bh).
RAP250 Interacts with PPARs, TRs, and ERs in Vitro via the LXXLL Motif-The partial mouse RAP250 clone was originally isolated via its interaction with the PPAR␣-LBD bait in the yeast two-hybrid system. To determine whether the mouse RAP250 could interact with other NRs, we set up an in vitro protein-protein interaction assay with the mouse partial RAP250 clone (aa 782-1138) fused to GST, referred to here as GST-RAP250, and radioactively labeled in vitro translated NRs. As shown in Fig. 3, PPARs, TRs and ERs specifically interacted with GST-RAP250 but not with GST. The addition of appropriate ligands in the binding buffer increased the interaction between GST-RAP250 and most of NRs (Fig. 3B, lanes 3,  and Fig. 4), indicating a ligand-dependent interaction in the cases of TRs, and a ligand-enhanced interaction for ERs and PPARs. To investigate whether the LXXLL motif (LVNLL, aa 891-895) of mouse RAP250 was responsible for the interaction  9 -12). GST protein alone does not shift the receptors (lanes 1, 2, 7, and 8). B, RAP250/TR␤/RXR␣ oligomeric complex formation on a TRE (DR4) in the presence of ligand using TR␣-mut/RXR␣ (lanes 3-6) but not using TR␣-mut/RXR␣-mut (lanes 9 -12). GST protein alone does not shift the receptors (lanes 1, 2, 7, and  8). The mutations TR␣-mut and RXR␣-mut are helix 12 deletions known to not bind coactivators. Arrows indicate TR␣/RXR␣ heterodimer (TR/RXR) and the ternary complex (TC) formed on DNA.

FIG. 6. RAP250 strongly interacts with NRs in mammalian cells.
A, COS-7 cells were transfected with NR expression vectors (TR␣/RXR␣, PPAR␥/RXR␣, or RXR␣) and luciferase reporter genes (DR4, PPRE, or DR1) together with either VP16 (gray bars) or VP16-mRAP250 (black bars). The activity of the luciferase reporter gene was measured 24 h after addition of appropriate ligands (T 3 /9-cis RA, BRL 9653/9-cis RA, or 9-cis RA, respectively). Results represent mean Ϯ S.D. of at least two separate experiments carried out in duplicate and were normalized to the activity of each NR/reporter combination transfected with VP16, which was set as 1. RLU, relative light units.
with NRs, the leucine core motif was mutated to AVNAL, as shown in Fig. 3A. Mutation of this motif abolished the interaction of GST-RAP250 with all tested NRs both in absence or presence of ligands (Fig. 3, lanes 5 and 6), indicating that the interaction between RAP250 and NRs was mediated by an LXXLL motif, the integrity of which is required to function as an NR box.
RAP250 Contains Only One Functionally Active NR Box-Whereas the NR-interacting RAP250 mouse fragment contains only one NR box domain, which is identical between mouse and human proteins (see Fig. 1C), analysis of the human RAP250 protein sequence revealed a second LXXLL motif (LSQLL, aa 1491-1495) that might possibly serve as an additional receptor recognition sequence. First, the full-length RAP250 clone was translated in vitro and used in a GST pull-down assay with ER␣ or TR␣ expressed as GST fusion proteins, in both the absence and presence of appropriate ligand. As expected, this transcript interacted in a ligand-dependent manner with the tested NRs (Fig. 4A). To further characterize the interaction between RAP250 and these NRs, eight overlapping cDNA fragments covering the human RAP250 sequence (Fig. 4B) translated in vitro and used in a GST pull-down assay. As seen in Fig. 4C, only fragment 4 (aa 819 -1096), containing the first LXXLL motif, interacted strongly with NRs, whereas no interaction was observed with fragment 6 (aa 1491-1495), which contained the second LXXLL motif. The sequence of this region, SLSQLL, does not fit closely the consensus motif found in coactivators, which often contains a hydrophobic amino acid upstream of the first conserved leucine residue. This sequence difference might explain why the second motif did not function as an NR box. With the exception of fragment 5 (aa 1061-1338), which showed a very weak interaction with ER␣, but not with TR␣, none of the other RAP250 fragments interacted with the tested NRs. However, the strength of the interaction between fragment 4 and ER was manyfold higher, and thus, we consider this NR box-containing region to be the interacting region. As observed in Fig. 3, RAP250 interacts in a ligand-dependent manner with TR␣ and in a ligand-enhanced manner with ER␣.

FIG. 7. Localization of a transcription activation domain in RAP250.
A, COS-7 cells were transfected with plasmid encoding the indicated human Gal4-RAP250 fragment (columns 1-8), and the transcriptional activity from the cotransfected Gal4-luciferase reporter plasmid was measured. The numbers below the figure represent the human RAP250 amino acids fused to Gal4-DBD. For each well, 0.5 g of reporter plasmid and 0.5 g of each Gal4-RAP250 expression plasmid were transfected. 24 h after transfection, cells were harvested and cell extracts were analyzed for luciferase activity as described under "Experimental Procedures." The results represent mean Ϯ S.D. of at least three experiments. B, further mapping of the transcriptional activity using the indicated human Gal-RAP250 constructs in a transient transfection experiment with COS-7 cells. The experiment was carried out and data are presented as described for A. C, Western blot analysis of Gal4 fusion protein levels in transfected cell extracts. Nuclear extracts was separated by 12% SDS-polyacrylamide gel electrophoresis, transferred to nitrocellulose, and probed with a mouse anti-Gal4-DBD monoclonal antibody (Santa Cruz Biotechnology). After incubation with a horseradish peroxidase-conjugated anti-mouse antibody, the bands were visualized by enhanced chemiluminescence. Detected proteins of expected size are indicated with an asterisk.
Taken together, these data indicate that the human RAP250, although containing two LXXLL motifs, may possess only one major NR-interacting-domain.
RAP250 Interacts with NRs Bound to DNA-To determine whether RAP250 interacted with NRs bound to their DNA response elements, we performed electrophoretic mobility shift assays. Ternary complex formation with NR heterodimers was assessed using in vitro translated TR␣/RXR␣ and the purified mouse RAP250 (aa 818 -931) fused to GST. As seen in Fig. 5A, GST-protein alone did not interact with DNA bound TR␣/RXR␣ in either the presence or absence of ligands (Fig. 5A, lanes 1  and 2). Addition of GST-RAP250 to the binding reaction resulted in a weak supershift in the absence of ligands (Fig. 5A,  lane 3). However, the addition of ligands (T3 or 9-cis RA or both) resulted in a pronounced mobility shift of the TR␣/RXR␣ dimer (Fig. 5A, lanes 4 -6), which indicated the formation of an oligomeric complex containing GST-RAP250.
To further characterize the stoichiometry of this interaction, we used helix 12 mutated forms of TR␣ and RXR␣. These mutants have previously been shown not to interact with coactivators (21,22). When the mutated RXR␣ was used, the ligand-dependent interaction of GST-RAP250 in response to 9-cis RA was lost (Fig. 5A, lane 11). However, GST-RAP250 still interacted with the NR heterodimer in the presence of T 3 (Fig.  5A, lanes 10 and 12). In a similar manner, when the mutated form of TR␣ was used, the ligand-dependent interaction of GST-RAP250 in response to T 3 was lost (Fig. 5B, lane 4), but GST-RAP250 still interacted with the NRs in response to 9-cis RA (Fig. 5B, lanes 5 and 6). When the mutated receptors were used in combination, receptor heterodimers were detected on DNA, but no interaction of GST-RAP250 was detected (Fig. 5B,  lanes 7-12). These results demonstrated that RAP250 was able to interact with DNA-bound NRs in a ligand-and AF-2-dependent manner. Because the oligomeric complexes detected with GST-RAP250 and wild type receptors were located at the same position as those detected when one NRs was mutated, it seemed likely that one GST-RAP250 molecule bound per receptor heterodimer.
RAP250 Differentially Interacts with Wild Type NRs in Transfected Mammalian Cells-To determine whether RAP250 interacted with NRs in vivo, we used a mammalian transient transfection assay derived from the two-hybrid assay, which employed a herpesvirus VP16 activation domain fused to the mouse RAP250 clone (aa 782-1138). The VP16-RAP250 expression vector or a VP16 activation domain vector without RAP250 was transfected into COS-7 cells together with expression vectors for wild type NRs and a luciferase reporter gene containing appropriate response elements. As seen in Fig. 6, TR␣/RXR␣-mediated activity of the reporter gene was stimulated 3.8-fold by VP16-RAP250, as compared with VP16. A similar stimulation was observed for PPAR␥/RXR␣, (3.3-fold), whereas RXR␣-mediated reporter gene expression was stimulated 10.5-fold. These in vivo data supported the observation made in vitro (Fig. 3) that RAP250 interacted with all tested NRs.
RAP250 Activates Transcription via a Large Intrinsic Transcription Activation Domain-In order to identify activation and/or repression domains within human RAP250, fragments of RAP250 were fused to the Gal4 DNA binding domain (DBD) and analyzed for transcription activation potential using transient transfection assays in COS-7 cells. In a first set of experiments, eight Gal4-RAP250 constructs containing about 300 amino acids each were assayed for their putative transcriptional activity in mammalian COS-7 cells (Fig. 7A). Two of the constructs were able to activate transcription, whereas none had significant repression activity. The construct containing amino acids 335-630 activated transcription approximately 5-fold as compared with Gal4-DBD alone, and the construct containing amino acids 577-855 activated transcription about 10-fold. Because these two fragments partially overlapped each other, a second set of constructs was made to determine whether there were two independent activation domains or whether the two fragments had one common activation domain in the overlapping region. Removal of 52 amino acids from the Gal4-RAP250-(335-630) construct generated a Gal4-RAP250-(335-577) construct that reduced the activation capacity from 5-fold to 3-fold, as shown in Fig. 7B. In a similar manner, 52 amino acids were removed from the Gal4-RAP250-(577-855) construct, generating Gal4-RAP250-(630 -855), which reduced the activation capacity from 10-fold to 6-fold (Fig. 7B). Moreover, the fragment spanning amino acids 577-630, which overlaps the first set of fusion constructs (constructs 2 and 3), retained a weak activation potential of about 2-fold. An additional construct that contains all sequences showing activation potential, i.e. Gal4-RAP250-(335-855), activated transcription about 18-fold. Control studies demonstrated that all constructs were expressed and, with the exception of construct 7 (aa 1539 -1771), at similar levels (Fig. 7C). From these data, we concluded that RAP250 contained one large activation domain localized between amino acids 335 and 855 and that removal of sequences within this activation domain gradually reduced its strength.
RAP250 Functions as a Coactivator for NRs-Based on the fact that fragments of RAP250 bind in vitro to several nuclear hormone receptors, it was important to investigate whether full-length RAP250 could also function as a coactivator for the transcriptional activity of these receptors. To investigate this hypothesis, expression vectors for TR␣ and RXR␣ were cotransfected with pSG5-hRAP250 or an empty vector together with a luciferase reporter gene, DR4-Tk-Luc for TR␣, or DR1-Tk-Luc FIG. 8. Human RAP250 enhances the transcriptional activities of wild type TR␣ and RXR␣. COS-7 cells were transiently co-transfected with 0.5 g of luciferase reporter plasmid (containing DR4 for TR␣ and DR1 for RXR␣), 0.5 g of each receptor construct (TR␣ and RXR␣), and 0.5 g of pSG5-RAP250 or 0.5 g of pSG5 in the absence or presence of the appropriate ligand (20 nM T 3 , or 10 M 9-cis retinoic acid). The activity of the luciferase reporter gene was measured 24 h after addition of ligands. Results represent mean Ϯ S.D. of at least two separate experiments carried out in duplicate and were normalized to the activity of each NR/reporter combination transfected with pSG5 without ligand, which was set as 1.
for RXR␣. As shown in Fig. 8, addition of appropriate ligand alone induced expression of the reporter genes 4.6-fold for TR␣/RXR␣ and 3.2-fold for RXR. Overexpression of hRAP250 stimulated transcription activation 7.2-fold for TR␣/RXR␣ and 7.7-fold for RXR␣. Thus, these results suggested that fulllength RAP250 also acted as a coactivator. DISCUSSION This study describes the structural and functional properties of RAP250, a novel NR coactivator isolated from a mouse embryo library using the yeast two-hybrid system. Our results show that RAP250 interacts with multiple members of the NR family in a ligand-dependent manner, indicating that RAP250 is a coactivator of all these NRs. We also show that this interaction is dependent on an NR box and an intact AF-2 domain. Furthermore, RAP250 possesses an intrinsic transcriptional activation domain and functions as a NR coactivator in mammalian cells. A schematic representation of RAP250 is shown in Fig. 9. Together these results suggested that RAP250 may be a new AF-2 NR coactivator. Based on sequence analysis, RAP250 appears to be different from other NR coactivators characterized to date. For example, it shows no significant sequence homology with any known NR coactivator and has no bHLH/ PAS domain as found in the p160 proteins or any motifs that would suggest histone acetylase or deacetylase activity.
Out of the two LXXLL motifs found in RAP250, only the first one functions as an NR box, in contrast to the p160 proteins that contain multiple NR boxes. In that respect, RAP250 resembles coactivators, such as TIF-1 (28,29) or PGC-1 (30), that also have only one functional NR box. One model of coactivators binding to NR heterodimers proposes a bridging function of the p160 proteins by simultaneous binding of the two dimer-subunits with two adjacent NR boxes (31). However, this possibility clearly does not exist for RAP250, which has only one NR box. Another model, suggested by the presence of only one functional NR box, could involve two molecules of coactivator per NR-dimer, each of the subunits of the dimer binding one coactivator molecule via the NR box (20). In the case of RAP250, it would mean that each of the subunits of the NRdimer binds one RAP250 molecule via the NR box. However, the electrophoretic mobility shift assay analysis (Fig. 5) does not suggest that this is the case for RAP250, because the complexes containing RAP250 and wild type NRs migrate at the same speed as when one of the NRs in the dimer has a mutation that prevents the coactivator from binding. An alternative could be that RAP250 is part of a larger complex of proteins, in which it would bind a NR via its NR box, in a manner similar to TRAP220/DRIP205 in the TRAP/DRIP complex (13,14,32,33). Because RAP250 does not correspond to any of the subunits that have been identified (16,34), it is possible that RAP250 either is part of a new coactivator complex or represents an unidentified member of the TRAP/DRIP complex.
The RAP250 activation domain is large and glutamine-rich with approximately 20% of Q residues (Figs. 1A and 9) but is not active when tested in yeast (data not shown). This is in agreement with previous findings demonstrating that glutamine-rich activation domains of transcription factors Oct-1, Oct-2, and Sp1, which activate transcription in mammalian cells, are inactive in yeast, probably reflecting some basic difference between the organization of yeast and mammalian promoters or transcription complexes (35). Interestingly, the enhancer-binding protein Sp1, which is a prototype for glutamine-rich transcription factors, was recently shown to interact with a transcription complex called CRSP, which contains TRAP220 (25).
RAP250 mRNA is widely expressed, with the apparently highest expression in brain and reproductive organs, such as ovary, testis, and prostate (Fig. 2, A and B). Further work is required to determine whether RAP250 is of particular importance in these organs.