Vitamin D-interacting Protein 205 (DRIP205) Coactivation of Estrogen Receptor α (ERα) Involves Multiple Domains of Both Proteins*

Vitamin D-interacting protein 205 (DRIP205) is a mediator complex protein that anchors the complex to the estrogen receptor (ER) and other nuclear receptors (NRs). In ZR-75 breast cancer cells treated with 17β-estradiol (E2) and transfected with a construct containing three tandem estrogen responsive elements (pERE3), DRIP205 coactivates ERα-mediated transactivation. DRIP205Δ587–636 is a DRIP205 mutant in which both NR boxes within amino acids 587–636 have been deleted and, in parallel transfection studies, DRIP205Δ587–636 also coactivates ERα. Moreover, both wild-type and variant DRIP205 also colocalize with ERα in the nuclei of transfected cells. Extensive deletion analysis of DRIP205 shows that multiple domains of this protein play a role in coactivation of ERα and in interactions with ERα. Coactivation of ERα by DRIP205 does not require NR boxes, and variants with deletion of N-terminal (amino acids 1–639) and C-terminal (amino acids 576–1566) significantly coactivate ERα. DRIP205 resembles p160 coactivators that also interact with multiple regions of ERα; however, unlike p160 coactivators, DRIP205 coactivation of ERα does not require NR boxes.

The nuclear receptor (NR) 1 superfamily of transcription factors contains both ligand-activated and orphan receptors that interact with genomic cis-elements in target gene promoters to induce gene expression (1)(2)(3)(4)(5). NRs have a conserved modular structure that features a DNA-binding domain (DBD) C, which has high homology among different classes of NRs. In contrast, there are major differences among NRs in their A/B, E/F, and D domains that contain activation function 1, activation function 2, and hinge regions, respectively. The C-terminal E/F region of NRs also contains the ligand binding domain, which undergoes ligand-dependent conformational changes that are important for subsequent transcriptional activation of target genes. Steroid hormone receptors such as estrogen receptor ␣ (ER␣) have been extensively used as models for determining the mecha-nisms of ligand-dependent receptor-mediated transactivation, which requires the assembly and recruitment of a nuclear complex of coactivator/coregulatory proteins (5)(6)(7)(8)(9)(10). The p160 steroid receptor coactivators (SRC) were first discovered as nuclear proteins that specifically interact with ligand-bound hormone receptors (11,12), and subsequent studies have identified a multitude of structurally diverse coactivators that enhance receptor-mediated transactivation (13)(14)(15)(16)(17)(18)(19). Other coregulatory proteins such as p300/CBP and pCAF are also components of the receptor-coregulatory complex, and these proteins, in part, modify chromatin structure and promote accessibility through their histone acetyltransferase activities.
Mediator complex proteins also enhance transactivation through recruitment of RNA polymerase II to target gene promoters (20 -28), and several studies report ligand-dependent interactions of mediator complex proteins with NRs (29 -40). Mediator-NR interactions involve direct binding of the 200 -220-kDa mediator protein to the receptor and vitamin D-interacting protein 205 (DRIP205, also known as thyroid hormone receptor-associated protein 220 (TRAP220), mediator 220 (Med220), and peroxisome proliferator-activated receptor-binding protein) anchors the complex to the NRs. Several studies have investigated coactivation of NRs including ER␣ and ER␤ by DRIP205 and these results highlight the complexity of NRcoactivator interactions (36 -41). For example, Warnmark and co-workers (37) showed that TRAP220 interacted preferentially with ER␤ compared with ER␣ and interactions were dependent on the two LXXLL NR box motifs (NR1 and NR2) in TRAP220. The NR box motifs have been identified in multiple NR coactivators and are structural features that facilitate ligand-dependent NR-coactivator interactions (41,42). In transfection studies, TRAP220 did not enhance E2dependent transactivation in HeLa cells (40), whereas two reports suggest that DRIP205 coactivates ER␣ in vitro and in U2OS cells (10,36).
The mechanisms of ER␣-and ER␣/Sp1-mediated gene expression in breast cancer cells have been studied in this laboratory, and we have shown that coactivation of these responses are highly variable in different cell lines (41). This study investigates ligand-dependent coactivation of ER␣ by DRIP205 in breast cancer cell lines transfected with pERE 3 . The results demonstrate that coactivation of ER␣ by DRIP205 in ZR-75 breast cancer cells is complex and involves multiple regions of DRIP205 and ER␣. In mammalian two-hybrid and coimmunoprecipitation assays, interactions of DRIP205 and ER␣ also involved multiple domains of both proteins. Moreover, coactivation of ER␣ by DRIP205 and interaction of these proteins do not require the NR box motifs of DRIP205. I). The first step was to amplify several cycles of annealed fragments A and B. Then, the primers were added when the reactions were paused, and the products of the first step would stand for the template of the second step. The final products were digested with NdeI and NotI, and ligated back to pcDNA3-DRIP205, which was digested with NotI and then partially digested with NdeI. This construct was called pcDNA3-DRIP205mB and used as a template generating other constructs by PCR. The expression plasmid with deletion of both NR boxes, namely pcDNA3-Dm3 (DRIP205⌬587-636), was generated using twostep PCR (Table I). The final PCR products and plasmid pcDNA3-DRIP205mB were separately digested with NheI and SspI, followed by ligation.
pM and pVP16 vectors were purchased from Clontech (Palo Alto, CA). pcDNA3.1/His A, B, and C was purchased from Invitrogen (Carlsbad, CA). pMD, pMDm3, pMDm4, pMDm1, pMDm5, pMDm6, pMDm7, pMDm7⌬, pMDm13, pMDm12, and pMDm8 expression plasmids were generated by PCR. PCR products were digested with restriction enzymes described in Table I, and ligated back into the pM vector, which was digested with the same set of restriction enzymes.
XDm5 expression plasmids were constructed by PCR. PCR products were digested with BamHI and EcoRV and ligated to pcDNA 3.1/His C. XDm7, XDm7⌬, XDm12, and XDm8 expression plasmids were generated by PCR. PCR products were digested with BamHI, EcoRV, and ligated to pcDNA3.1/His A. VP16-hER, VP16-TAF1, and VP16-HE19 expression plasmids were generated by digestion of pM-hER, pM-TAF1 expression plasmids with EcoRI, SalI, and ligated to pVP vector digested with EcoRI and SalI. pfu Turbo DNA polymerase from Statagene (La Jolla, CA) was used in PCR. All the primers were ordered from IDT, Coralville, IA. All constructs were sequenced afterward to confirm the cloning.
Transient Transfection Assays-ZR-75 or MDA-MB-231 cells were seeded in 12-well plates in DMEM/F-12 medium without phenol red (Invitrogen) supplemented with 2.2 g/liter sodium bicarbonate, 10 ml/ liter AAS, and 2.5% charcoal-stripped FBS. After incubation for 12 h, at 37°C in 5% CO 2 , 95% air, cells were co-transfected with DNA using the calcium phosphate method. In coactivation experiments, cells were co-transfected with 250 ng of pcDNA3.1-␤-galactosidase (used as an internal control), 1 g of pERE 3 -LUC, 5 ng of pcDNA3-hER␣, and various amounts of DRIP205 or deletion mutant constructs. The pcDNA3 empty vector was used to maintain DNA mass balance among different treatment groups. In the mammalian two-hybrid assay, cells were co-transfected with 250 ng of pcDNA3.1-␤-galactosidase (used as an internal control), 500 ng of pGAL4 5 -LUC, 500 ng of pM or pM-DRIP wild-type or deletion mutants, 500 ng of pVP, VP-hER␣, VP-HE19, or VP-TAF1. Six h after transfection, cells were shocked with 25% glycerol/ PBS for 1 min, washed with PBS (2ϫ), and then treated with Me 2 SO or 10 nM E2 for another 30 -48 h. Each treatment was replicated three or four times. Cells were then washed twice in PBS and harvested with 100 l of reporter lysis buffer (Promega Corp., Madison, MI). After one freeze-thaw cycle, cell lysates were centrifuged 30 s, and the supernatant was used for determination of protein activity. Luciferase (Promega Corporation, Madison, MI) and ␤-galactosidase (Applied Biosystems, Foster City, CA) activity were read by a Packard Luminometer. Relative luciferase activity was calculated by dividing luciferase activity by ␤-galactosidase activity for each well. In coactivation experiments, -fold induction was calculated by dividing the relative luciferase activity of E2 treated groups by relative luciferase activity in controls (Me 2 SO-treated). In the mammalian two-hybrid assay, -fold induction was calculated by dividing -fold induction of E2/Me 2 SO of VP-hER␣ or VP-hER␣ deletion mutants by -fold induction of E2/Me 2 SO obtained using the empty vector pVP.
Western Blot Analysis-For determination of ER␣ protein level, ZR-75 cells were seeded, transfected, and harvested as described above. After luciferase and ␤-galactosidase activity were read, 6 l of 5 M NaCl was added to the remaining ϳ60 l of lysates to obtain maximal protein yield. Lysates were incubated on ice for 1 h with occasional vortexing followed by centrifugation (16,000 ϫ g, 10 min, 4°C). Equal amounts of total protein from each treatment group were separated by SDS-PAGE and transferred to polyvinylidene difluoride membrane (transfer buffer: 48 mM Tris, 39 mM glycine, 0.025% SDS).
For GAL4-DRIP fusion protein confirmation, COS-7 cells were seeded in 6-well plates with DMEM/F-12 medium without phenol red supplemented with 2.5% charcoal-stripped FBS. After overnight incubation, cells were transfected with 2 g of each fusion protein expression plasmids using Lipofectamine 2000 reagent (Invitrogen). Medium was changed 12 h after the transfection, and, after 48 h, cells were trypsinized, transferred, and washed with PBS (3ϫ). Cellular and nuclear extracts were obtained using NE-PER TM Nuclear and Cytoplasmic Extraction Reagents (Pierce Biotechnology). Cellular and nuclear lysates were separated by 7.5% SDS-PAGE, and transferred to polyvinylidene difluoride membrane.
Coimmunoprecipitation Assay-[ 35 35 S-labeled DRIP205 wild-type or variant protein (0.5 l) were incubated in 0.5 ml of coimmunoprecipitation buffer (PBS ϩ 0.001% IGEPAL CA630). The coimmunoprecipitation buffer was freshly supplemented with 1 M E2, 1:100 dilution of protease inhibitor mixture (Sigma), and 1 mM phenylmethylsulfonyl fluoride (Sigma). After incubation for 1 h on a rocker at 4°C, 400 ng of antibody was added. DRIP205 antibody c-19 was purchased from Santa Cruz Biotechnology. This antibody was raised against the C terminus of TRAP220 of human origin, and was used to pull down DRIP205⌬587-636. ER␣ HC 20 antibody was used for ER␣ immunoprecipitation. For Xpress-tagged DRIP205 deletion mutants, anti-Xpress antibody from Invitrogen was used. After incubation for 1 h at 4°C, 20 l of a 50% slurry of protein G-Sepharose beads (Amersham Biosciences) were added to the reaction solution, followed by incubation for 2 h on a rocker at 4°C. Samples were then centrifuged at 12,000 ϫ g at 4°C for 1 min. The supernatant was carefully removed, and the pellet was washed with PBS ϩ 1% IGEPAL CA630 (3ϫ), and finally washed with PBS. The final pellet was boiled in 30 l of 2ϫ SDS sample buffer, and proteins were separated on a 6% SDS-PAGE and visualized by autoradiography.
Immunocytochemistry-COS-7 cells were seeded onto 2-well glass chamber slides at 75,000 cells per well in DMEM/F-12 medium supplemented with 5% charcoal-stripped FBS. After 12 h incubation in a 37°C incubator with 5% CO 2 , cells were transfected with 500 ng of pMD or pMDm3, and 500 ng of hER␣ expression plasmids using Genejuice transfection reagent (Novagen). After another 24-h incubation, cells were treated with Me 2 SO vehicle or 10 nM E2 for 1 h. Slides were then washed with PBS, fixed with Ϫ20°C methanol, air-dried, and washed with PBS ϩ 0.3% Tween 20 (PBS/Tween). Slides were blocked for 1 h with 5% donkey serum in antibody dilution buffer (1% bovine serum albumin in 0.02 M PBS ϩ 0.3% Tween 20), washed with PBS/Tween briefly, and incubated with anti-DRIP205 c-19 antibody (Santa Cruz) at 1:50 dilution in antibody dilution buffer at 4°C for 12 h. Slides were washed (goat serum for controls) with PBS/Tween (3 ϫ 10 min), incubated with donkey anti-goat IgG fluorescein isothiocyanate (Santa Cruz) at 1:200 dilution in antibody dilution buffer for 1 h and washed with PBS/Tween (3 ϫ 10 min). Slides were subsequently blocked with 5% donkey serum in antibody dilution buffer for 1 h, washed with PBS/Tween briefly, incubated with anti-ER␣ H-184 antibody (Santa Cruz) at 1:50 dilution in antibody dilution buffer at 4°C for 12 h (rabbit serum for controls), washed with PBS/Tween (3 ϫ 10 min), incubated with donkey anti-rabbit IgG Alexa Fluor 594 (Molecular Probes) at 1:500 dilution in antibody dilution buffer for 1 h, and washed with PBS/Tween (3 ϫ 10 min). Slides were finally washed in deionized water, and coverglass mounted using Prolong Gold antifade reagent with 4,6-diamidino-2-phenylindole (Molecular Probes). Immunofluorescence images of DRIP205 and ER␣ were examined using a Zeiss Axioplan 2 microscope (Carl Zeiss, Thornwood, NY) fitted with an Axiocam high resolution digital camera. Digital images were captured using Axiovision 3.0 software.
Statistical Analysis-Statistical differences between different groups were determined by analysis of variance and Scheffe's test for significance. The data are presented as mean Ϯ S.E. for at least 3 separate determinations for each treatment.

Coactivation of Wild-type and Variant ER␣ by DRIP205-
Coactivation of ER␣-dependent transactivation by DRIP205 was initially examined in ER␣-positive ZR-75 and ER␣-negative MDA-MB-231 cells. The full-length DRIP205 expression plasmid (Fig. 1A) encodes for 1566 amino acids, which is identical to amino acids 16 -1581 of the TRAP220 coding sequence. Both cell lines were transfected with pERE 3 , which contains three tandem EREs in a minimal TATA-luciferase construct, and an ER␣ expression plasmid. The transfected pERE 3 construct is overexpressed in the transfected cells and minimal E2 inducibility is observed in ZR-75 cells in the absence of cotransfected ER␣. This system is ideal for investigating coactivation of ER␣ and determining receptor and coactivation domain requirements for transactivation in breast cancer cell context. The results illustrated in Fig. 1, B and C, show that E2 induces transactivation 6-and 4-fold in ZR-75 and MDA-MB-231 cells, respectively, and this was enhanced Ͼ3and Ͼ5-fold by DRIP205. Fig. 1D shows that transfected DRIP205 has no significant effect on ER␣ protein expression levels in the presence or absence of E2. ER␣ levels were lower in cells treated with E2 because of degradation of ER␣ through the proteasome pathway.
E2 also induced transactivation in ZR-75 cells transfected with pERE 3 and wild-type or variant ER␣ expression plasmids with mutations in helix 12 (ER␣-TAF1) ( Fig. 2A) or deletion of activation function (AF) 1 (HE19) (Fig. 2B). However, studies with DRIP205 showed that coactivation was observed only with wild-type ER␣, suggesting that AF1 and wild-type AF2 were both necessary for functional DRIP205-ER␣ interactions. The requirements for both AFs of ER␣ for coactivation by DRIP205 were further investigated in competition experiments by transfecting GRIP1-NR box peptide and ER␣-AF1 peptide expression plasmids. Previous studies show that these peptides competitively squelch AF2 and AF1, respectively (41)(42)(43)(44). The results show that both peptides inhibit DRIP205 coactivation of ER␣ (Fig. 2, C and D), indicating that coactivation by DRIP205 is associated with both AF1 and AF2 of ER␣.
Coactivation of ER␣ by DRIP205 Deletion Mutants-Previous studies have shown that the NR boxes in DRIP205/ TRAP220 contribute to the physical and functional interactions of these coactivators with ER and other NRs (34, 36, 40 -42). Their role in coactivation of ER␣ was further investigated in ZR-75 cells transfected with pERE 3 and DRIP205⌬587-636, in which both NR boxes were deleted. The results (Fig. 3A) show that like wild-type DRIP205, DRIP205⌬587-636 also coactivates ER␣. These studies demonstrate that coactivation of ER␣ in ZR-75 cells by DRIP205 does not require NR1 or NR2. Moreover, coactivation of ER␣ by DRIP205⌬587-636 was also decreased by overexpression of GRIP1-NR and ER␣-AF1 (Fig.  3, B and C), and this complements their inhibition of wild-type DRIP205 coactivation of ER␣ (Fig. 2, B and C). These results suggest that coactivation of ER␣ is dependent on both C-and N-terminal sequences of DRIP205 and/or their interactions between these domains of DRIP205. Coimmunoprecipitation of hER␣ with DRIP205 or DRIP205⌬587-636 were investigated by incubation of in vitro translated Xpress-tagged wild-type and variant [ 35 S]DRIP205, [ 35 S]ER␣, and DRIP205 and ER␣ antibodies followed by SDS-PAGE (Fig. 3D). Lanes 1, 3, and 6 show 35 S-labeled ER␣, DRIP205, and DRIP205⌬587-636, and the latter two bands were indistinguishable on the gel because of only small differences in molecular mass. We also investigated colocalization of ER␣ with DRIP205 (Fig. 4A) and DRIP205⌬587-636 (Fig. 4B) by transfection studies in COS-7 cells using the GAL4-DRIP205 expression plasmids. This cell line is ER␣-negative and only expresses low levels of DRIP205. In cells transfected with ER␣ and DRIP205/ DRIP205⌬587-636 expression plasmids, their corresponding expressed proteins were primarily localized in the nucleus; however, some cytoplasmic staining was observed for ER␣. In cells transfected with DRIP205/ER␣ and treated with Me 2 SO (solvent control) or 10 nM E2 for 1 h, immunostaining showed that ER␣ and DRIP205 were extensively colocalized (Fig. 4A,  right panels). Estrogen treatment induced a punctate pattern of DRIP205 nuclear staining, whereas, this was not observed for ER␣. In the corresponding experiment with DRIP205⌬587-636 (Fig. 4B), punctate staining of the deletion mutant was observed in solvent-treated cells and this was enhanced in E2-treated cells (Fig. 4B, left panel). DRIP205⌬587-636 also colocalized with ER␣ (Fig. 4B, right panel), confirming that interaction of DRIP205 with ER␣ did not require the NR boxes.
Initial coactivation studies with N-and C-terminal deletion mutants of DRIP205 were determined in ZR-75 cells transfected with a series of GAL4-DRIP205 deletion constructs, which are readily expressed in the nucleus because of the GAL4 DNA binding domain component of the fusion protein. The results in Fig. 5A summarize coactivation studies with a series of constructs that express amino acids 1-575 (pMDm5), 486 -1051 (pMDm7), 486 -586 ϩ 637-1051 (pMDm7⌬), 640 -1566 (pMDm6), 641-1051 (pMDm12), and 1052-1566 (pMDm8). Significant coactivation was observed for several GAL4-DRIP205 chimeras (pMDm5, pMDm7, pMDm7⌬, and pMDm6), two of which express N-terminal (pMDm5) or C-terminal (pMDm6) regions of DRIP205 but do not contain the central NR box sequences. These results confirm that the NR boxes of DRIP205 are not required for coactivation of ER␣, and that the coactivation includes sequences within the N-and C-terminal regions of this protein. Results in Fig. 5B summarize Western blot analysis of nuclear extracts from cells transfected with several of the GAL4-DRIP205 fusion proteins and the results show that all of the proteins were expressed in the nucleus. pMDm5, which contains N-terminal amino acids 1-575, gave a less intense band than the other proteins but was not detected in nuclear extracts. Coactivations of ER␣ by Xpress-tagged DRIP205 variants were also investigated in ZR-75 cells. Transfection with XDm7, XDm7⌬, and XDm6 coactivated ER␣ (Fig.  5C) and these data were consistent with results obtained with the corresponding GAL4-DRIP205 chimeras. XDm5 does not contain a nuclear localization signal and did not coactivate ER␣ (data not shown). Interactions of Xpress-tagged DRIP205 deletion proteins with ER␣ were also investigated using in vitro translated proteins and Xpress antibody (Fig. 5D). Expression of 35 S-labeled ER␣ and Xpress-tagged DRIP205 proteins alone (lanes 1, 3, 6, 9, and 12) are shown and the Xpress antibody immunoprecipitates 35 S-labeled DRIP205 chimeras (lanes 4, 7, 10, and 13) but not ER␣ (lane 2). However, the Xpress antibody immunoprecipitates ER␣ after coincubation with Xpresstagged DRIP205 chimeras (lanes 5, 8, 11, and 14). These results show that ER␣ interacts with multiple domains of DRIP205 and the interactions do not require NR boxes.
DRIP205 Activation Function-Previous studies showed that DRIP205 interacts with ER and other NRs and is a component of the mediator complex of proteins and related nuclear factors (20 -28). A series of GAL4-DRIP205 constructs containing multiple deletions in DRIP205 were used to investigate their AF activity (i.e. interactions with other coregulatory proteins and the basal transcription machinery) in a mammalian one-hybrid assay and interactions with VP16-ER␣ (wild-type and variants) in a mammalian two-hybrid assay. ZR-75 cells were transfected with chimeric GAL4-DRIP205 constructs, pGal4 and transactivation was determined. The highest activity was observed for the following constructs where pMDm7⌬ Ͼ Ͼ pMDm7 Ͼ pMDm12 Ͼ pMDm4 (Fig. 6A). Maximal AF activity was observed for pMDm7⌬, in which the NR boxes (amino acids 587-636), N-terminal amino acids 1-485, and C-terminal amino acids 1052-1566 were deleted. Insertion of the NR boxes (to give pMDm7) resulted in significant loss of AF activity. The NR boxes of DRIP205 also exhibited activity (pMDm4), regions flanking the NR boxes (i.e. amino acids 486 -586 and amino acids 640 -1051) cooperatively contribute to the AF of DRIP205. This pattern of activity for the GAL4-DRIP205 constructs was also observed in MCF-7 breast cancer cells (data not shown), suggesting that in both cell lines, similar regions in DRIP205 are important for interacting with other coregulatory factors or the basal transcription machinery.
DRIP205 Interactions with ER␣-Mammalian Two-hybrid Experiments-E2-dependent interactions of the same GAL4-DRIP205 constructs with ER␣ were determined in ZR-75 cells, transfected with VP16-ER␣, where wild-type ER␣ is fused to the VP16 activation domain. The results (Fig. 6B) show that VP16-ER␣ interacts with multiple regions of DRIP205. E2 significantly induced activity in cells transfected with VP16-ER␣ and GAL4-DRIP205 (wild-type) or seven other constructs including pMDm4, pMDm1, pMD, pMDm5, pMDm8, pMDm7, and pMDm3. pMDm4, which expresses the NR box sequences (amino acids 528 -714) exhibited the highest E2-dependent activity, suggesting that this sequence facilitates but is not required for ER␣-DRIP205 interaction. All GAL4-DRIP205 chimeras that contain the NR boxes exhibit E2-induced interactions with ER␣ and addition of flanking sequences decreased hormone-induced activity. However, constructs such as pMDm8, pMDm5, and pMDm3, which do not contain the NR boxes also exhibit E2-dependent interactions with VP16-ER␣, confirming that multiple domains of DRIP205 are involved in interactions with ER␣. A similar pattern of E2-dependent interactions of FIG. 3. NR box deletion mutant DRIP205⌬587-636 coactivates ER␣. A, coactivation of ER␣ by DRIP205⌬587-636 in ZR-75 cells. Cells were cotransfected with 1000 ng of pERE 3 -LUC, 250 ng of ␤-galactosidase, 5 ng of hER␣, and various amount of DRIP205⌬587-636 expression plasmids, treated with Me 2 SO or 10 nM E2, and luciferase activity was determined as described under "Materials and Methods." Results are expressed as mean Ϯ S.E. for three separate experiments for each treatment group and significant (p Ͻ 0.05) enhancement is indicated (*). AF2 (B) and AF1 (C) are required for DRIP205⌬587-636 coactivation of ER␣. Cells were cotransfected with 1000 ng of pERE 3 -LUC, 250 ng of ␤-galactosidase, 5 ng of hER␣, 0 or 1 ng of DRIP205 and increasing amounts of GRIP1-NR (B) or ER␣-AF1 (C) peptide expression plasmids, treated with Me 2 SO or 10 nM E2, and luciferase activity was determined as described under "Materials and Methods." Significant coactivation (p Ͻ 0.05) (*) and significant inhibition by competing peptides are indicated (**). D, coimmunoprecipitation of ER␣ and DRIP205 or DRIP205⌬587-636. In vitro translated proteins (lanes 1, 3, 6, 9, 10, and 11; 30% of input) were incubated, immunoprecipitated with specific antibodies, and analyzed by SDS-PAGE as described under "Materials and Methods." Protein bands were identified by comparison with in vitro translated proteins and molecular markers (250, 150, 100, 75, and 50 kDa) (data not shown). DMSO, dimethyl sulfoxide.
Helix 12 of ER␣ is a major site for recruitment of the LXXLL box coactivators (reviewed in Refs. [13][14][15][16][17][18][19]. However, E2-dependent activation of ER␣-TAF1 ( Fig. 2A), which contains helix 12 mutations is observed in several cell lines, suggesting that other sites within ER␣ are sufficient for interaction with coactivators. We therefore used VP16-TAF1 to investigate ligand-dependent interactions with GAL4-DRIP205 constructs (Fig. 6C). The results showed that Ͼ2fold induced activity was only observed for 4 chimeric proteins, namely pMDm4 and pMDm7, which contain NR boxes, and pMDm5 and pMDm12, which express N-and C-terminal regions flanking the NR boxes. These results confirm that multiple regions of DRIP205 (ϮNR boxes) interact with ER␣ in regions outside helix 12 of ER␣. A comparison of results in Fig. 6, B and C, suggest that mutations of helix 12 did not affect ligand-dependent interactions of ER␣ with pMDm4, pMDm5, pMDm7, or pMDm12. In contrast, E2 did not induce activity in cells transfected with pMD (full-length), pMDm3, pMDm1, pMDm8 plus VP16-ER␣-TAF1, suggesting that some regions of DRIP205 required an intact helix 12 for E2-dependent interactions with ER␣; however, these interactions were not necessarily LXXLL box-dependent (e.g. pMDm8 and pMDm3). These results suggest that DRIP205 interaction with ER␣ in breast cancer cells are highly complex and are dependent on multiple domains of both proteins. This was further illustrated by comparing interactions of the GAL4-DRIP205 constructs with VP16-HE19 (C, D, and E/F domains) (Fig. 6D). E2-dependent GAL4-DRIP205 interactions with VP16-HE19 in a mammalian two-hybrid assay gave significant induction in ZR-75 cells transfected with pM and pVP-HE19 (compared with pM plus pVP). Therefore, a comparison of transactivation results obtained for VP16-HE19 plus GAL4-DRIP205 constructs, with VP16-HE19 plus GAL4 (pM) alone, shows that enhanced activity was observed only for pMDm3, which does not contain the NR boxes. A comparison of these data with DRIP205 interactions observed using VP16-ER␣ suggests that DRIP205 interactions are primarily AF1-dependent or require both AF regions of ER␣. The results confirm that DRIP205-ER␣ interactions and coactivation of ER␣ by DRIP205 are complex and dependent on multiple regions of both proteins but do not require NR boxes. , and pMDm4 (lane 6) and nuclear lysates were analyzed by Western blot analysis as described under "Materials and Methods." Lane 1 was shown as a negative control. GAL4 DBD monoclonal antibody RK5C1 was used. C, coactivation of ER␣ by Xpress-tagged DRIP205 deletion mutants. ZR-75 cells were cotransfected with 1000 ng of pERE 3 -LUC, 250 ng of ␤-galactosidase, 5 ng of hER␣, and various amounts of XDm6, XDm7, or XDm7⌬ expression plasmids, treated with Me 2 SO or 10 nM E2, and luciferase activity was determined as described under "Materials and Methods." Significant (p Ͻ 0.05) coactivation is indicated (*). D, coimmunoprecipitation of ER␣ and various Xpress-tagged DRIP205 deletion mutants. In vitro translated proteins (lanes 1, 3, 6, 9, and 12; 30% of input) were incubated, immunoprecipitated, and analyzed by SDS-PAGE as described under "Materials and Methods." Protein bands were identified by comparing them to in vitro translated proteins and molecular markers as indicated in the legend to Fig. 3. DMSO, dimethyl sulfoxide. that DRIP205 and/or related mediator proteins interact with NRs and appear to be the critical linkage protein between NRs and other protein components of the mediator complex (29 -40). For example, the mediator complex of proteins binds ER␣ and ER␤ (ligand-dependent), whereas extracts from TRAP220 Ϫ/Ϫ embryo fibroblasts do not bind ER, demonstrating the critical role of TRAP220 in ER-mediator complex interactions and transactivation (39). Previous reports have investigated coactivation of ER␣ by DRIP205 and interactions of these proteins and the results are variable and dependent on the assay and cell context. For example, TRAP220 did not coacti-vate ER␣-mediated transactivation in HeLa cells and exhibited weak interactions with ER␣ that were dependent on extended LXXLL motifs (40). In contrast, peroxisome proliferator-activated receptor-binding protein potentiated E2-dependent transactivation in CV-1 cells (45). In a mammalian two-hybrid assay in COS-7 cells, GAL4-TRAP220 interacted with VP16-ER (DEF) (ER␣ and ER␤); and in a GST-ER (E/F) pull down assay, interactions with TRAP220 were dependent on NR boxes and ER␤ was the preferred interacting ER subtype (37). Burakov and co-workers (36,38) reported that in GST pull down assays, DRIP205 did not bind the N-terminal ABCD FIG. 6. Mapping of DRIP205 activation function and interactions with ER␣. A, DRIP205 activation. ZR-75 cells were transfected with 250 ng of ␤-galactosidase, 500 ng of pGAL4-LUC, 500 ng of pM or pM DRIP wild-type or deletion mutants, and luciferase activity was determined as described under "Materials and Methods." Interactions of wild-type and variant pM-DRIP205 with VP16-ER␣ (B), VP16-TAF1 (C), and VP16-HE19 (D). ZR-75 cells were transfected with 250 ng of ␤-galactosidase, 500 ng of pGAL4-LUC, 500 ng pM or pM DRIP wild-type or deletion mutants, and 500 ng of pVP16 empty or VP16-hER␣ (B), VP16-TAF1 (C), and VP16-HE19 (D), treated with Me 2 SO or 10 nM E2, and luciferase activity was determined as described under "Materials and Methods." -Fold induction by E2 for each treatment group was calculated by dividing the -fold induction observed for E2/Me 2 SO in cells transfected with VP16-hER␣ (wild-type or variant) by the -fold induction (E2/Me 2 SO) observed in cells transfected with pVP16 (empty vector) alone. domain of ER␣, and interactions with the ligand binding domain (E/F) were dependent on intact DRIP205 NR boxes, and L539A/L540A mutation in helix 12 of ER␣ abrogated interactions. Previous studies in this laboratory showed that in ERpositive and ER-negative breast cancer cells, several coactivators including SRC-1, SRC-2, SRC-3, and p68 RNA helicase did not enhance ER␣-dependent transactivation (41). This study investigates coactivation of ER␣ by DRIP205 and interactions of these proteins in a breast cancer cell context.
The results consistently show that DRIP205 coactivates ER␣ in ZR-75 and MDA-MB-231 cells transfected with pERE 3 (Fig.  1), and although the -fold enhancement of induction is variable, coactivation by DRIP205 was consistently observed in both cell lines. Wild-type DRIP205 did not coactivate HE19 (A/B domain deletion) or ER␣-TAF1 (helix 12 mutations) (Fig. 2, A and B), suggesting that both AF1 and AF2 of ER␣ are required for coactivation by DRIP205. Moreover, overexpression of ER␣ AF1 and GRIP1-NR box peptides decreased DRIP205 coactivation of ER␣ (Fig. 2, C and D), confirming that coactivation involves both AF1 and AF2 of ER␣. These results are consistent with previous reports showing that both AF1 and AF2 play a role in coactivation by SRCs (46 -48).
DRIP205 coactivates ER and other NRs and also anchors interactions of the DRIP complex with NRs and other transcription cofactors. Therefore, we examined deletion variants of DRIP205 to determine domains required for ER␣ coactivation and interactions with ER␣ and other nuclear cofactors. The deletion mutant DRIP205⌬587-636 does not contain the NR boxes, but like wild-type DRIP205, this variant coactivates ER␣ (Fig. 3A) and coimmunoprecipitates with ER␣ (Fig. 3D). Moreover, coactivation of ER␣ by DRIP205⌬587-636 is decreased in cells transfected with ER␣-AF1 and GRIP1-NR box peptides (Fig. 3, B and C).
Previous studies using cellular imaging techniques have demonstrated that E2 induces colocalization of ER␣ and SRCs in cells treated with E2, and ligand-induced ER␣-SRC interactions were dependent on LXXLL box motifs in SRCs (49,50). We observed colocalization of ER␣ with DRIP205 and DRIP205⌬587-636 in COS-7 cells treated with Me 2 SO or E2 (Fig. 4); however, treatment with E2 enhanced the punctate pattern of wild-type/variant DRIP205 staining alone and in colocalization with ER␣. These results complement the functional coactivation studies (Figs. 2 and 3) and show that in contrast to SRCs, coactivation of ER␣ by DRIP205 does not require the NR boxes. We also examined coactivation of ER␣ by other GAL4-or Xpress-tagged DRIP205 deletion mutants, and show that coactivation can be observed with both N-and Cterminal domains of DRIP205 and coactivation did not require the NR boxes (Fig. 5).
Because multiple domains within DRIP205 are sufficient for coactivation of ER␣, we also examined interactions of GAL4-DRIP205 (wild-type and variants) with VP16-ER␣ (wild-type and variants) in a mammalian two-hybrid assay (Fig. 6B). The results were compared with the AF activity of the GAL4-DRIP205 constructs, which would reflect their interactions with other nuclear cofactors or basal transcriptional machinery required for transactivation (Fig. 6A). Maximal AF activity was observed for pMDm7⌬, which contains amino acids 486 -586 and 637-1051. The NR box sequence (amino acids 587-636) is deleted from pMDm7⌬ and inclusion of this region (pMDm7) decreased transactivation. A comparison between the AF activity of GAL4-DRIP constructs and their ligand-dependent interactions with VP16-ER␣ (wild-type and variant) in mammalian two-hybrid assays provides some insights on regions within DRIP205 that interact with ER␣ and/or other nuclear cofactors. pMDm7⌬ exhibited high AF activity, but minimal interactions with ER␣, ER␣-TAF1, and HE19 in mammalian two-hybrid assays (Fig. 5). However, pMDm7⌬ coimmunoprecipitates with ER␣ (Fig. 5D) and coactivates ER␣, suggesting that this variant DRIP205 protein facilitates interactions with ER␣ and other nuclear cofactors. Interestingly, pMDm4, which contains both NR boxes, interacts with VP16-ER␣ and VP16-ER-TAF1 but not VP16-HE19. These results suggest that interactions of DRIP205 NR boxes with ER␣ may also be dependent on AF1. Results of the mammalian two-hybrid studies indicate that multiple domains of ER␣ and DRIP205 are involved in protein-protein interactions and in coactivation of ER␣. The lack of specificity for DRIP205-ER␣ binding in the two-hybrid assay is not unlike the complex interactions of p160 coactivators with ER and other NRs (46 -48). DRIP205 and p160 coactivators differ with respect to the role of their respective NR boxes, which are important for p160-ER␣ interactions and AF2-dependent coactivation (13)(14)(15)(16)(17)(18)(19). However, these characteristics of DRIP205 may be consistent with the critical role for this protein as a component of several nuclear cofactor complexes involved in transcriptional activation of genes through interactions with multiple transcription factors. Current studies are focused on coactivation of ER␣ and ER␣/Sp1 by DRIP205 and other mediator proteins to determine their individual and cooperative activities as coactivators of ER␣ and other NRs.