Tip60 Is a Nuclear Hormone Receptor Coactivator*

The androgen receptor (AR) is a member of the nuclear hormone receptor superfamily. Recent work in this field has been focused upon defining the mechanisms of transcriptional control exacted by members of this superfamily. Using a COOH-terminal region of the human AR in a yeast two-hybrid screen, we have identified Tip60 as an AR-interacting protein. In this report, we show that Tip60, which was originally identified as a coactivator for the human immunodeficiency virus TAT protein, can enhance AR-mediated transactivation in a ligand-dependent manner in LNCaP and COS-1 cell lines. In addition, our experiments show that Tip60 can also enhance transactivation through the estrogen receptor and progesterone receptor in a ligand-dependent manner; thus identifying Tip60 as a nuclear hormone receptor coactivator. Our studies also demonstrate that Tip60 co-immunoprecipitates with the full-length ARin vitro and that, in our system, Tip60 enhances transactivation to levels observed with the coactivators steroid receptor coactivator 1, p300, and CREB-binding protein. The importance of such proteins in enhancing nuclear hormone receptor-mediated transcriptional activation is widely accepted, and this work suggests that Tip60 may have an equally important role to play.

The androgen receptor (AR) is a member of the nuclear hormone receptor superfamily. Recent work in this field has been focused upon defining the mechanisms of transcriptional control exacted by members of this superfamily. Using a COOH-terminal region of the human AR in a yeast two-hybrid screen, we have identified Tip60 as an AR-interacting protein. In this report, we show that Tip60, which was originally identified as a coactivator for the human immunodeficiency virus TAT protein, can enhance AR-mediated transactivation in a ligand-dependent manner in LNCaP and COS-1 cell lines. In addition, our experiments show that Tip60 can also enhance transactivation through the estrogen receptor and progesterone receptor in a ligand-dependent manner; thus identifying Tip60 as a nuclear hormone receptor coactivator. Our studies also demonstrate that Tip60 co-immunoprecipitates with the full-length AR in vitro and that, in our system, Tip60 enhances transactivation to levels observed with the coactivators steroid receptor coactivator 1, p300, and CREB-binding protein.
The importance of such proteins in enhancing nuclear hormone receptor-mediated transcriptional activation is widely accepted, and this work suggests that Tip60 may have an equally important role to play.
The AR 1 is a member of the nuclear hormone receptor superfamily of ligand-dependent transcription factors. In the nucleus, the ligand-activated receptors bind to their cognate response elements in or near promoter regions of target genes to positively or negatively regulate gene expression (reviewed in Ref. 1). The mechanism by which specific modulation of target gene expression is achieved by nuclear receptors (NRs) is not clearly defined. Early research indicated that there might be factors common to different receptors required for efficient transcriptional activation. This was demonstrated by the observation that different ligand-bound steroid receptors affected or inhibited transactivation by other NRs. This suggested the sequestration of mutually required cofactors (2,3). Subsequent work has led to the identification of several coactivators of NR-mediated transactivation. These coactiva-tors enhance levels of NR transactivation severalfold (reviewed in Refs. 4 and 5) and display a broad specificity across the superfamily, with the exception of ARA70, which appears to be specific for the AR (6). The NR-bound coactivator proteins are believed to enhance the stability of the pre-initiation complex, resulting in increased rates of transcription initiation. Direct contacts between pre-initiation complex components and coactivators have been demonstrated, supporting this hypothesis; for example, SRC-1 has been shown to interact with TATA box-binding protein and TFIIB (7). In addition, members of the NR superfamily have also been shown to interact with components of the pre-initiation complex (8 -10).
The discovery that the coactivators SRC-1, p300, and CBP (CREB-binding protein) possess intrinsic histone acetyltransferase (HAT) activity provided further insight into the regulation of NR transactivation (11,12). Histone acetylation has long been associated with transcriptional activity (13), and the subsequently altered nucleosomal structure is believed to facilitate access for additional factors necessary for activation (14). Histone deacetylases can reverse the effects on nucleosomal disruption of HAT proteins, leading to repression of gene expression (15)(16)(17). The subsequent demonstration that a histone deacetylase, HDAC1, interacted with the NR corepressors SMRT (18) and NCoR (19) provoked the realization that the transcriptional output directed by a NR from a given gene may be reflected in the position of equilibrium between histone acetylation and deacetylation. Coactivators and corepressors may possess such activities or recruit proteins that do to exact their function.
Recent work investigating the progesterone receptor (PR) has shown that when bound by ligand, it exists in stable ternary complexes with SRC-1 and TIF-2 (20). Such complexes may exist with other NRs, enhancing the potential for diverse interaction upon ligand binding. In addition, many coactivators have been shown to interact with one another, such as CBP/ PCAF (21), SRC-1/CBP (22), SRC-1/p300 (23), and SRC-1/ PCAF (11). Also, multiple NR-interacting domains have been identified in certain coactivators such as SRC-1 (24). The multiplicity of such protein interactions therefore suggests that maximal transcriptional activation is likely to require the targeted coupling of hyperacetylated histones with stabilized preinitiation complexes and that this is likely to be mediated by multimeric, coactivating, HAT-containing protein complexes. The observed complexity of gene expression may be reflected in the composition of these higher-order, activating complexes.
In this study, we sought to identify AR-interacting proteins in an attempt to dissect the mechanism by which the AR exacts transcriptional control. Due to the multitude of potential interacting proteins, we utilized the yeast two-hybrid system (25). This led to the identification of the Tat-interacting protein (Tip60) as an AR-interacting protein. Tip60 was originally reported as a human coactivator for the human immunodeficiency virus type I-encoded TAT protein (26), and more recent work has shown that it is a member of a family of related genes involved in transcriptional regulation (27). Subsequently, we demonstrated that Tip60 acts as a ligand-dependent coactivator for the AR, PR, and ER (estrogen receptor). The level of enhancement observed is comparable to that of SRC-1, p300, and CBP in our system, and, taken together, these results suggest that Tip60 is potentially an important member of the NR coactivator family.

Yeast Two-hybrid Screening
Residues 559 -918 of the hAR (GenBank accession number m23263) containing the DNA and ligand binding domains were amplified by PCR from pGEM3Z-AR (a gift from S. Liao, University of Chicago, Chicago, IL) using primers GGACTCTAGAATGCCCATTGACTATTACTT and TGTTTCTAGAGCTTCACTGGGTGTGGAA and subcloned into pT7-7 via XbaI sites. This fragment was released by NdeI-BamHI restriction of pT7-7hAR559 -918 and subcloned downstream of the GAL4 DNA binding domain (DBD) in pAS2-1 (CLONTECH). pAS2-1hAR559 -918 was transformed into the yeast strain PJ69-4A (a gift from P. James, University of Wisconsin). PJ69-4A is an improved two-hybrid strain with an additional nutritional reporter gene that serves to reduce false positives (28). The subsequent tryptophan-positive strain was transformed with 50 g of a human brain cDNA:GAL4 activation domain fusion library constructed in pACT2 (CLONTECH). 200,000 transformed cells were plated onto media lacking tryptophan, leucine, and adenine, supplemented with 1 M dihydrotesterone (DHT). Adeninepositive, LacZ-positive, histidine-positive colonies were isolated, and plasmid DNA was prepared and transformed into Escherichia coli according to the protocol provided by CLONTECH. pACT2 plasmids containing cDNA were identified by colony PCR using primers specific for the LEU2 gene contained in pACT2. Specificity of interaction for the hAR was determined by examining the liquid LacZ activity (see below) of cDNA clones in the presence of the GAL4DBD:hAR559 -918 fusion protein versus the activity observed in the presence of the GAL4DBD alone. After sequencing and examination of the GenBank TM database, we determined that one interacting clone was encoded by a 1.7-kb cDNA fragment with a 96% identity to Tip60 (GenBank accession number u40989), which maintained the reported open reading frame for Tip60 (25). Sequence alignment revealed that the identified clone began 70 residues downstream of the NH 2 terminus. Subsequent two-hybrid screens in our laboratory have identified full-length Tip60 cDNAs as AR-interacting clones, suggesting that the 70 residues absent from our construct did not significantly interfere with the interaction. All Tip60 constructs used in this work are derived from the original 1.7-kb clone.

Analysis of Tip60-AR Interaction Using the Yeast Two-hybrid System
pAS2-1hAR constructs were co-transformed with pACT2Tip60 or pACT2 into PJ69-4A according to the manufacturer's guidelines (CLONTECH). Subsequent Trp-positive, Leu-positive colonies were inoculated in triplicate into selective media and grown at 30°C overnight in the presence or absence of 1 M DHT. Samples were diluted to an A 600 of 0.2 and re-grown to an A 600 of 0.6 -0.8. Samples were divided into three 1-ml aliquots. Cells were recovered by centrifugation at 14,000 rpm for 5 min, washed once with buffer Z (0.1 M sodium phosphate, pH 7.0, 10 mM KCl, and 10 mM MgSO 4 ) and resuspended in 800 l of buffer Z containing 21 l of ␤-mercaptoethanol. 10 l of 0.1% SDS were added, followed by 50 l of chloroform. Samples were vortexed for 1 min and placed at 30°C, and 200 l of o-nitrophenyl-␤-D-galactopyranoside (4 mg/ml in buffer Z) were added. Reactions were timed and terminated upon observing an obvious yellow color or after 1 h via the addition of 500 l of 1 M Na 2 CO 3 . A 420 of the samples was determined, and activity was calculated as follows: (A 420 ϫ 1000)/(A 600 ϫ time). All assays were performed in triplicate and repeated at least three times. hAR fusion proteins for these experiments were constructed as follows; hAR559-918 was constructed as described above. Additional constructs were amplified by PCR using the following primer combinations: hAR1-150 (p1:TCAAGGATCCAAGTGCAGTTAGGGCTGGGA -p3:AC-GTGGATCCGGCAGCTGAGTCATCCTCGT), hAR1-300 (p1-p5:CCTA-GGATCCCGCCTTCTAGCCCTTTGGTG), hAR559 -624 (p6:TATTGG-ATCCCACCCCAGAAGACCTGCCTG-p9:TCCGGGATCCCAGAGTCA-TCCCTGCTTCAT), and hAR624 -918 (p8:GCCCGGATCCTGAAGAAA-CTTGGTAATCTG-p10:TAGGGGATCCAATGCTTCACTGGGTGTGG-A). PCR products were cloned into the BamHI site of pAS2-1. All fusion constructs were sequenced to confirm maintenance of the open reading frame, and expression was confirmed by Western blotting using anti-AR antibodies (Novocastra).

In Vitro Interaction of AR and Tip60
Templates were prepared using a Geneclean kit (Anachem) and resuspended in RNase-free distilled H 2 O. An in vitro coupled transcription and translation kit (T7-TNT; Promega) was used according to the manufacturer's instructions. After completion of the 90-min reaction, samples were combined equally on ice, as indicated. 1 ml of immunoprecipitation buffer (50 mM Tris, pH 7.5, 150 mM NaCl, 0.2 mM Na 3 VO 4 , 0.5% Nonidet P-40, 1 mM phenylmethylsulfonyl fluoride, 1 mM dithiothreitol, 25 g/ml leupeptin, 25 g/ml aprotinin, and 25 g/ml pepstatin) was added to each sample, mixed, and incubated on ice for 30 min. 20 l of protein G-Sepharose (PGS) pre-washed three times in immunoprecipitation buffer were added to each sample and incubated for an additional 4 h at 4°C with rotation to remove any proteins that interacted nonspecifically with PGS. PGS was removed by centrifugation at 14,000 rpm for 3 min. The supernatant was incubated with 2.5 g of penta-His antibody (Qiagen) overnight at 4°C with rotation. 20 l of PGS were added to each sample and incubated at 4°C for an additional 60 min. PGS antibody conjugates were recovered by centrifugation at 14,000 rpm for 3 min, resuspended in 1 ml of buffer A (PBS, 0.2% Triton X-100, and 350 mM NaCl), and re-spun. Samples were resuspended in 1 ml of buffer B (PBS and 0.2% Triton X-100), re-spun, and resuspended in SDS sample buffer. Samples were resolved on 10% polyacrylamide gels at 200 V for 45 min with equal loading. The gel was fixed for 30 min (10% propanol and 10% acetic acid), soaked in Amplify (Amersham Pharmacia Biotech) for 30 min, dried under vacuum, and exposed to x-ray film for 4 -24 h at Ϫ70°C. Templates used were pT7-7hAR1-918 and full-length Sap-1a in pASG4 (a gift from A. D. Sharrocks, Newcastle University). pRSET-C-Tip60 was constructed via excision of a BamHI fragment from pACT2Tip60 utilizing a BamHI site in the 3Ј untranslated region of Tip60 cDNA and cloned in-frame with the hexa-His tag of pRSET-C (R&D Systems). Orientation was determined via restriction analysis and confirmed by sequencing.

Transient Transfections of COS-1 and LNCaP Cells
COS-1 Cells-Cells were cultured for at least 48 h before transfection in Dulbecco's modified Eagle's medium (Life Technologies, Inc.) supplemented with 10% fetal calf serum that had been stripped of steroids by treatment with dextran-coated charcoal. Cells were routinely grown to 60 -80% confluence, washed, removed, and seeded at a density of 5 ϫ 10 4 cells/ml in fresh medium in 35-mm wells (Corning). After 20 h, cells were washed once with serum-free medium and transfected using Lipofectin (Life Technologies, Inc.) according to the manufacturer's recommendations. The amounts of DNA used are described in the figure legends. After 7 h, cells were washed twice with the appropriate medium and incubated in 2 ml of medium containing steroid or vehicle, as indicated. Mibolerone is used in mammalian experiments with the AR because it is a specific androgen analogue. After 48 h, cells were recovered and assayed for luciferase activity according to the manufacturer's instructions (Promega). Luciferase activity was corrected for the corresponding ␤-galactosidase activity to give relative activity. ␤-Galactosidase assays were typically performed in a 96-well plate (Corning) as follows: 10 l of sample extract were incubated with 80 l of buffer Z and 10 l of o-nitrophenyl-␤-D-galactopyranoside (4 mg/ml) at 30°C for 2 h. The reaction was terminated via the addition of 50 l of 1 M Na 2 CO 3 . A 420 values were obtained using a MR5000 plate reader (Dynatech), and activity was calculated as described above. Transfections were performed in triplicate and repeated at least three times.
LNCaP Cells-These cells were routinely subcultured in RPMI 1640 medium (Life Technologies, Inc.) supplemented with 10% fetal calf serum. Cells were seeded at 1 ϫ 10 5 cells/well in 35-mm wells at least 24 h before transfection. For studies examining the ER and PR, cells were cultured in phenol red-free RPMI 1640 medium supplemented with 10% dextran-coated charcoal-fetal calf serum. Cells were transfected using Lipofectin for a period of 8 h and washed, and medium containing vehicle or steroid was added as indicated. Cells were harvested at 72 h after transfection and analyzed as described above. pCMV5Tip60 was constructed via cloning of a BglII fragment from pACT2Tip60 into pCMV5. Expression was confirmed using an anti-HA antibody (Santa Cruz Biotechnology). hAR1-918 was cloned into the XbaI site of pCDNA3. The following mammalian expression plasmids were gratefully received: RSVhPRB and full-length hSRC-1a in pCR3. . In addition, pPSAluc, a reporter construct derived from the PSA promoter, was used for AR-based experiments in LNCaP cells. A 658-base pair fragment containing two androgen-responsive elements was amplified by PCR from the PSA promoter (GenBank accession numbers u37672 and x13940) using primers GTGGAGCTGGATTCT-GGGT and CTCAGCTTGACAGTCAGGG. This product was cloned into the TA vector (Invitrogen), released via KpnI/XhoI digestion, and inserted into the corresponding sites in the luciferase reporter plasmid pGL3basic (Promega). For Fig. 6, pSGp53 expressing GAL4DBDp53 1-42 was obtained from A. D. Sharrocks (Newcastle University). The reporter plasmid TK-GAL4UASLuc and the GAL4DBD-expressing plasmid pSG424 were gifts from R. Clifton-Bleigh (Cambridge University). The Tip60:GAL4DBD fusion construct was created by the insertion of the BamHI fragment from pACT2Tip60 into the BamHI site of pSG424 downstream of the GAL4DBD. Correct orientation was determined by sequencing, and expression was confirmed via Western analysis with an anti-GAL4DBD antibody (CLONTECH).

Tip60 Interacts with the Ligand Binding Domain of the AR-
Yeast two-hybrid screening, which was originally described by Fields and Song (25), using the AR construct AR559 -918 comprising the DNA and ligand binding domains (DBD and LBD, respectively) as bait identified Tip60 as an AR-interacting protein (see "Experimental Procedures"). In an attempt to define the specificity of the interaction with the AR, we examined the association of Tip60 with various GAL4DBD:AR constructs in the yeast two-hybrid system in the presence of 1 M DHT. AR constructs spanning the first 300 residues (AR1-150 and AR1-300) and the DBD of the AR (AR559 -624) failed to interact with the Tip60GAL4AD fusion protein (Fig. 1a), whereas both AR559 -918 and AR624 -918 interacted specifically with Tip60. Therefore, Tip60 appears to interact specifically with the LBD of the AR. The interaction appeared to be increased significantly with the attachment of the DBD to the LBD, although the DBD of the AR fails to interact by itself. A potential explanation for this observation is that the LBD may be prevented from folding into the optimal conformation for interaction when positioned directly at the COOH terminus of the GAL4DBD. The presence of the DBD may serve to distance the LBD from the GAL4DBD in this scenario, or, alternatively, it may be that the AR DBD is essential for directing or influencing the conformational change upon ligand binding.
The interaction of AR and Tip60 is not ligand-dependent ( Fig. 1b) but is enhanced in the presence of ligand in the yeast two-hybrid system. We examined the interaction of the AR regions 559 -918 and 624 -918 with Tip60 in the presence and absence of 1 M DHT and found that the interaction was significantly enhanced in the presence of DHT. Values obtained in the absence of DHT for AR559 -918 and AR624 -918 were similar, indicating that the AR DBD may have a role to play in altering the conformation of the AR upon ligand binding in this system. The majority of interactions reported to date, such as ARA70 with the AR (6), have been ligand-dependent. It is therefore interesting that the AR:Tip60 interaction is not liganddependent. Significantly, PCAF, a human homologue of the yeast Gcn5 (21), has been shown to bind to the AR and ER in a ligand-independent manner and to function as a coactivator for nuclear receptors (29). The interaction reported in the yeast two-hybrid system between PCAF and the AR is similar to the interaction observed between Tip60 and the AR in that it is enhanced in the presence of ligand (29). Thus, Tip60 and PCAF may represent a class of coactivators that are complexed with the aporeceptor. It is possible that the ligand-independent Tip60:AR complex could hold the AR in a state of readiness, thus facilitating a rapid activation response upon ligand binding, enabling the directed recruitment of other cofactors to AR-regulated genes.
Tip60 Interacts with the AR in Vitro-To confirm the interaction data obtained from the two-hybrid system, we sought to determine whether Tip60 could interact with the AR in vitro. Using proteins transcribed and translated in vitro in the presence of [ 35 S]methionine, we attempted to co-immunoprecipitate the AR with Tip60 using an anti-His antibody recognizing His-tagged Tip60. The transcription factor Sap-1a (Elk4) (30), an Ets family-related protein, was included as a control for this experiment. Before immunoprecipitation, labeled HexaHis-Tip60 was incubated with rabbit reticulocyte lysate extracts containing Sap-1a or AR, as indicated (Fig. 2, lanes 2 and 4,  respectively). Lanes 1 and 3 show the control samples, in which immunoprecipitations were carried out on extracts containing Sap-1a or AR in the absence of Tip60 to determine the levels immunoprecipitated that are not specifically associated with Tip60. Confirmation of the two-hybrid results was obtained via the observation that full-length AR co-immunoprecipitated with Tip60 (lane 8), whereas very low amounts of nonspecific AR were visible in the absence of Tip60 (lane 7). This interac-  35 Slabeled proteins were mixed after in vitro translation and before immunoprecipitation (Pre-IP) as indicated in lanes 1-4. Samples were incubated with anti-His antibody at 4°C overnight, which recognized the His-tagged Tip60. Antibody complexes were recovered with PGS, washed, and subjected to SDS-polyacrylamide gel electrophoresis. Resultant gels were soaked in Amplify (Amersham Pharmacia Biotech) and exposed to x-ray film (lanes 5-8 , Post-IP). Significant amounts of AR co-immunoprecipitate with Tip60 (lane 8) versus immunoprecipitation in the absence of Tip60 (lane 7). Incubation of Tip60 with Sap-1a (lane 6) did not alter the amount of Sap-1a recovered (lane 5), demonstrating that the Tip60:AR interaction is specific. tion was not affected by the presence of ligand (data not shown), again a feature previously described for PCAF (29). Importantly, Sap-1a did not interact with Tip60 in this experiment, confirming the specificity of the immunoprecipitation procedure because no additional Sap-1a is recovered in the presence of Tip60 (lane 6) versus the absence of Tip60 (lane 5).
We also attempted to demonstrate the interaction using anti-AR. In this system, the only anti-AR that works consistently in our hands for the purposes of immunoprecipitation is sc-805 (Santa Cruz Biotechnology). However, this proved unsuccessful. Subsequent investigation leads us to believe that the antibody directed against the ligand binding domain and Tip60 compete for AR binding. This phenomenon has been reported for other studies in which antibody against an AR-interacting protein, BAG-1L, failed to co-immunoprecipitate the AR (31). However, our experiment, in combination with the two-hybrid data, provides sufficient evidence for the AR:Tip60 interaction.
Tip60 Enhances AR-mediated Transactivation-As stated earlier, Tip60 was previously identified as a coactivator for the human immunodeficiency virus type I-encoded TAT protein. This fact led us to believe that it could act in a similar manner for the AR. We investigated the effect of increasing levels of Tip60 upon AR-dependent transactivation in transiently transfected LNCaP and COS-1 cells (Fig. 3). In LNCaP cells, pPSAluc, a PSA-derived reporter plasmid containing two androgenresponsive elements, was utilized. Co-transfection of the indicated plasmids demonstrated a ligand-dependent enhancement of AR-mediated transactivation by Tip60 (Fig. 3a). In these experiments, Tip60 expression produced a 5-fold enhancement of AR transactivation without any effect on basal transcription. We also investigated the effect of Tip60 on ARmediated transactivation in COS-1 cells using a reporter plasmid derived from the murine mammary tumor virus promoter pMTVluc (Fig. 3b). Expression of Tip60 in this context resulted in a ligand-dependent, 3-4-fold enhancement of transcriptional activation without any effect on basal transcription levels. These results demonstrate that Tip60 is a ligand-dependent coactivator for the AR. This effect was observed using two different cell lines and reporter constructs, which provides confidence in the coactivating properties of Tip60.
Tip60 Enhances PR-and ER-dependent Transactivation- Having outlined the role of Tip60 in AR-regulated promoters, we wished to investigate the role of Tip60 in transcriptional coactivation with other nuclear receptors. pMTVluc was used to study the effect of Tip60 on PR-dependent transactivation. Co-transfection of LNCaP cells with pMTVluc, PR, and Tip60 in the presence of progesterone led to a 4-fold increase in PR-mediated transactivation (Fig. 4a) without any effects on basal levels of transcription. This Tip60-mediated increase was comparable to levels observed with the AR (Fig. 3a) and was also observed with the PR in COS-1 cells (data not shown). We also investigated the effect of Tip60 on ER-dependent transactivation (Fig. 4b). Using pVITEREluc, a reporter vector derived from the vitellogenin gene, a 3-fold increase in ER-dependent transactivation was noted. Levels of enhancement were not as high as those observed with the PR; however, this may reflect the promoter constructs studied. Tip60 Coactivates to Levels Observed with SRC-1, CBP, and p300 -Having established that Tip60 is a nuclear receptor coactivator, we attempted to make a comparison of the relative extents of coactivation exacted by SRC-1, CBP, p300, and Tip60 upon the AR. SRC-1, CBP, and p300 are established as important coactivators (reviewed in Refs. 4 and 5), and due to the variability generally observed in reporter assays, LNCaP cells were transiently transfected with differing coactivators to confirm the effects observed with Tip60, as indicated in Fig. 5. In all instances, mean activity is enhanced at least 2-fold. It is clear from these results that Tip60 enhances transcription and therefore coactivates to levels comparable with SRC-1, CBP, and p300. Hence, this experiment suggests that Tip60 could be an equally important AR coactivator and could be potentially as influential upon transactivation by nuclear receptors as SRC-1, CBP, and p300.
Tip60 Contains No Intrinsic Transcriptional Activity-In an attempt to define the mechanism by which Tip60 acts as a coactivator, we sought to determine the existence, if any, of a transcriptional activation function within Tip60. The presence of an activation function has been demonstrated for SRC-1, whereas PCAF was shown to possess no activation function (24). The determination of this property is essential for the characterization of a coactivator. Although Tip60 contains an atypical zinc finger region (26), no DNA binding ability has been demonstrated. Thus, we fused Tip60 to a GAL4DBD and pSK was added to a total of 1 g of DNA/transfected well. After transfection, medium containing 10 nM progesterone (P) or vehicle was added. Cells were harvested 72 h later, and relative activity was determined as described earlier. b, effect of increasing amounts of Tip60 on ER-dependent transactivation. pCMV5-Tip60 was added as indicated to cells transfected as described in a, with the exception that medium contained 1 nM estradiol or vehicle after transfection. Cells were harvested 72 h later, and relative activity was determined as described earlier. Data are expressed as the mean of at least three separate experiments performed in triplicate Ϯ S.D. examined its effect upon a GAL4-responsive reporter plasmid (Fig. 6). The observed activity was compared with the activity of a p53 fragment containing an identified activation domain, p53 1-42 , fused to a GAL4DBD, and the GAL4DBD alone. It can clearly be seen that Tip60 appears to contain no intrinsic activation function, producing no additional activity when compared with GAL4DBD alone, thus distinguishing it from SRC-1. This feature groups Tip60 with PCAF with respect to coactivator properties. DISCUSSION In this study, we have identified Tip60 as a nuclear receptor coactivator. Tip60 is a member of the MYST/SAS family of genes that is implicated in transcriptional control and conserved between species (32). This family includes the yeast SAS2 and SAS3 genes implicated in transcriptional silencing (32), yeast gene product ESA1 with HAT activity that is essential for growth (33), and the Drosophila MOF gene required for X-linked dosage compensation (27). A significant member of the family is the human MOZ protein, which exhibits strong homology to Tip60, originally identified via a common translocation found in acute myeloid leukemia. Other related human cDNA sequences contain the MYST domain but have no defined function at present (34). Thus, it is apparent that Tip60 is a member of a family of proteins that are likely to be important in the normal and abnormal control of cellular growth.
The MYST/SAS family genes contain HAT-like domains, which are conserved between members, including Tip60 (32). However, no importance can yet be attached to this domain in Tip60 with respect to its role in coactivation. Tip60 has been shown to acetylate free histones but failed to acetylate nucleosomal histones (35), thus raising concerns about the physiological substrates, if any, for this domain. If it is not required for histone acetylation, there may be other non-histone substrates, for example p53 is acetylated by p300, resulting in a 20-fold increase in affinity for a p53 binding site (36). Thus, an important future task is to determine the significance, if any, of the HAT-like domain within Tip60.
An interesting feature contained within the amino acid sequence of Tip60 is the presence of a LXXLL motif in the COOH region of the protein (amino acids 458 -462, GenBank accession number u40989). This LXXLL region is present in other NR coactivators such as GRIP1, CBP, and SRC-1 and has been shown to mediate hormone-dependent binding to nuclear receptors in some detail, with the deletion of this region abolishing the interaction between a NR and coactivator in some instances (37)(38)(39). Hence, peptide regions containing single or multiple LXXLL signatures are often designated as NR boxes. The presence of this NR box in Tip60 may be responsible for its ability to act as a coactivator for several NRs. Other unidentified motifs may account for the ligand-independent binding observed between Tip60 and the AR.
The lack of intrinsic transcriptional activity within Tip60 raises mechanistic questions. This feature is similar to PCAF (24), which is known to interact with several other coactivators and is therefore thought to recruit additional necessary factors once it is bound to a NR. PCAF has also recently been shown to enhance transcription when tethered to a promoter in the presence of an enhancer element and its cognate transcription factor (40), but not in its absence, suggesting that it may well recruit factors necessary for the action of a particular transcription factor. It is interesting to hypothesize that Tip60 can act in a similar mode, and we are currently investigating this possibility.
Tip60 and the recently identified nuclear receptor coactivator NCoA-62 (41) both display broad specificity, like most coactivators (5). CBP interacts with many other transcription factors (reviewed in Ref. 42), and it is becoming increasingly evident that SRC-1 is a signal integrator not unlike CBP and p300. For example, it has recently been shown that SRC-1 can mediate AP-1-dependent transcription by directly interacting with c-Jun and c-Fos (43). The combination of SRC-1 and p300 at the AP-1 sites studied resulted in a synergistic response, highlighting cooperation between coactivators. Also, SRC-1 has been demonstrated to act as a coactivator for nuclear factor B and is able to bind specifically to the p50 subunit of nuclear factor B (44). However, as these features become more evident, the lack of specificity demonstrated by coactivators provokes questions regarding the manner in which distinct responses are elicited by nuclear receptors. The phosphorylation status of the coactivators is not well studied, although TIF1␣ was recently shown to be hyperphosphorylated in the presence of the ER upon estradiol treatment (45). TIF1␣ was also shown to possess the ability to phosphorylate TFIIE␣, TAF II 28, and TAF II 55. It would therefore seem that specific transcriptional control is unlikely to be achieved by coactivators exclusive to a particular NR. This is more likely to be achieved by posttranslational modification of transcription factors by coactivators and of coactivators themselves by other cellular factors. This will serve to alter the dynamic state of flux between distinct activating complexes that share many cofactors in order to achieve specificity.
In conclusion, we have demonstrated that Tip60 is a liganddependent coactivator for the AR and that it interacts directly via the LBD of the AR. The demonstration of this function has been achieved using transiently transfected reporter constructs. It would perhaps be more pertinent to examine the FIG. 5. Tip60 enhances AR-dependent transactivation to levels comparable with SRC-1, CBP, and p300. Effect of 100 ng of plasmids expressing SRC-1, CBP, p300, or Tip60, as indicated, upon LNCaP cells transfected with 0.5 g of pPSAluc and 0.2 g of pCMV-␤gal per 35-mm well. pSK was added to a total of 1 g of DNA/transfected well. After transfection, medium containing 10 nM mibolerone (Mib) or vehicle was added. Relative activity was calculated as described previously. Data are expressed as the mean of at least three separate experiments performed in triplicate Ϯ S.D.
FIG. 6. Tip60 contains no intrinsic transcriptional activity. Effect of 200 ng of plasmids expressing GAL4DBD alone, GAL4DBDp53 1-42 or GAL4DBDTip60, as indicated, co-transfected with 0.2 g of GAL4-responsive reporter plasmid into COS-1 cells. Cells were recovered after 48 h, and relative activity was calculated as described earlier. Data are expressed as the mean of at least three separate experiments performed in triplicate Ϯ S.D. effects of Tip60 on AR-responsive genes, such as ornithine decarboxylase (46) and prostate specific antigen (47), in their normal chromosomal context to truly assess its significance. However, it is clear that Tip60, in our system, appears to be as important as SRC-1 and CBP in AR-dependent transactivation and is thus worthy of further investigation into its mechanism of action.