Transcription Factor ZBP-89 Cooperates with Histone Acetyltransferase p300 during Butyrate Activation of p21 waf1 Transcription in Human Cells*

Inducible p53-independent regulation of the cyclin-dependent kinase inhibitor p21 waf1 transcription is mediated through proximal GC-rich sites. Prior studies have shown that Sp1, Sp3, and the histone acetylase co-activator p300 are components of the complexes binding to these sites. Although Sp1 and Sp3 collaborate with p300, a direct interaction between Sp1 and p300 does not occur. This study sought to determine whether ZBP-89 rather than Sp1 is the direct target of p300 during butyrate induction of p21 waf1 . ZBP-89 (BFCOL1, BERF-1, ZNF 148) is a Krüppel-type zinc finger transcription factor that binds to GC-rich elements and represses or activates known target genes. Adenoviral-mediated expression of ZBP-89 in HT-29 cells revealed that ZBP-89 potentiates butyrate induction of endogenous p21 waf1 gene expression. Further, cotransfection of a ZBP-89 expression vector with a 2.3-kilobase p21 waf1 reporter recapitulated the potentiation by butyrate. DNase I footprinting analysis of the human p21 waf1 promoter with recombinant ZBP-89 identified a binding site at −245 to −215. Electrophoretic mobility shift assays confirmed that both recombinant and endogenous ZBP-89 and Sp1 bind to this element. The potentiation was abolished in the presence of adenoviral protein E1A. Deletion of the N-terminal domain of ZBP-89 abolished the potentiation mediated by butyrate treatment. This same deletion mutant abolished the ZBP-89 interaction with p300. Cotransfection of p300 with ZBP-89 stimulated the p21 waf1 promoter in the absence of butyrate. p300 co-precipitated with ZBP-89 but not with Sp1, whereas ZBP-89 co-precipitated with Sp1. Together, these findings demonstrate that ZBP-89 also plays a critical role in butyrate activation of the p21 waf1 promoter and reveals preferential cooperation of this four-zinc finger transcription factor with p300.

The cyclin-dependent kinase inhibitor p21 waf1 controls cell cycle progression through binding to G 1 cyclin/CDK complexes (1)(2)(3). DNA damage stimulates p21 waf1 transcription through p53-dependent mechanisms (4), whereas agents that regulate cellular differentiation may regulate p21 waf1 transcription through p53-independent mechanisms (5). Many of the studies reporting p53-independent regulation of p21 waf1 transcription demonstrate a requirement for GC-rich sites located within the first 100 bp 1 of its promoter (6,7). These sites have consistently been shown to bind members of the Sp family of transcription factors (8 -13). Several studies have shown that Sp1 or Sp3 mediate activation of the p21 waf1 promoter by these extracellular regulators such as nerve growth factor and transforming growth factor-␤; however, these same signals do not stimulate Sp1 binding or gene expression (10). The transcriptional coactivator p300 mediates growth arrest by catalyzing histone acetylation and subsequent chromatin rearrangements through its endogenous acetyltransferase enzyme activity (14). Taken together, these results raised the possibility that Sp1 transcriptional activity may be regulated by its association with a co-activator. As a result, the p300 co-activator was shown to co-precipitate in complexes with Sp1 (10). Moreover, activation of the p21 waf1 promoter by butyrate and nerve growth factor has been shown to require a functional collaboration between Sp1 and p300 (10,15). Yet, p300 does not interact directly with Sp1 or Sp3 (15). Thus, although p300 and Sp1 are components of the complex activating p21 waf1 , the interaction is indirect raising the possibility that other factors are likely to participate in this transcriptional regulatory complex. Presumably, at least one or more of these other factors are capable of direct contact with p300.
ZBP-89 (BFCOL1, BERF1, ZNF 148) is a widely expressed four-zinc finger transcription factor that binds to GC-rich DNA elements in a variety of promoters involved in growth regulation, e.g. promoters for gastrin, T-cell ␣and ␤receptors, ornithine decarboxylase (ODC), enolase, type I procollagen, cyclin-dependent inhibitor p21 waf1 , vimentin, and stromelysin (16 -23). In many instances, ZBP-89 appears to repress promoter activity by opposing the effect of Sp1, which also binds to the same or overlapping DNA element. Thus, competitive binding to the shared promoter elements may mediate transcriptional regulation by Sp1 and ZBP-89 (16,19). ZBP-89 and Sp1 may also regulate transcription cooperatively because it has been shown that ZBP-89 directly binds Sp1 in co-precipitation assays (22).
Studies by Hasegawa et al. (24) showed binding of ZBP-89 to a proximal Sp1 element from the mouse p21 waf1 promoter. We hypothesized that ZBP-89 may be present in the p300/Sp1 activation complex and participate in the activation of p21 waf1 transcription. p21 waf1 is required for butyrate-mediated growth inhibition in HT-29 colorectal adenocarcinoma cells (25,26).
Thus the goals of this study were to examine the role of ZBP-89 in p21 waf1 activation by butyrate. The results demonstrate that both ZBP-89 and Sp1 recognize the same p21 waf1 regulatory sequences and that ZBP-89-dependent activation depends upon elevated histone acetyltransferase activity. We also found that Sp1, as reported (15), does not directly bind p300. Instead, we find that p300 cooperates directly with ZBP-89.
The N-terminal Myc-tagged and C-terminal FLAG-tagged full-length rat ZBP-89 expression vector was prepared using the following primers: 5Ј-TATACCATGGAACAAAAACTCATCTCAGAAGAGGATCTGGAAC-AAAAACTCATCTCAGAAGAGGATCTGAACATTGACGACAAACTG-G-3Ј as the forward primer and 5Ј-ATACTCGAGTCACTTGTCATCGT-CGTCCTTGTAGTCCTTGTCATCGTCGTCCTTGTAGTCGCCAAAAG-TCTGGCCAG -3Ј as the reverse primer. This Myc-ZBP-89-FLAG PCR fragment was blunt end-ligated into the EcoRV and NotI sites of pcDNA3 (Invitrogen, Carlsbad, CA). The orientation was verified by restriction analysis, and the expression was confirmed by anti-FLAG antibody. The resulting plasmid was labeled as pCMV/Myc-ZBP-89-FLAG.
Construction of Recombinant Adenovirus-To construct a ZBP-89 expressing recombinant adenovirus, the 3.5-kb DNA fragment that contains the ZBP-89 cDNA, CMV promoter, and bovine growth hormone poly(A) signal sequence was excised from pCMV/Myc-ZBP-89-FLAG with NruI and PvuII and then blunt end-ligated into the EcoRV site of the shuttle plasmid pAdMCSloxP (obtained from the University of Michigan Cancer Center Vector Core) (27). Recombinant replicationdeficient adenovirus was produced by the Vector Core using the method of Aoki et al. (27). Briefly, the ZBP-89 adenoviral shuttle plasmid was recombined with the adenovirus type 5 cosmid, and then transfected into 293T cells. The resultant recombinant adenoviral particles were harvested from the cells and called Ad5-ZBP-89. The Ad5-ZBP-89 viral particles were purified by CsCl centrifugation and titered.
Cell Culture and Transfections-The HT-29 human colorectal adenocarcinoma (HTB-38) and HeLa human cervix adenocarcinoma cell lines (CCL-2) were purchased from American Type Culture Collection (Manassas, VA) and cultured in Dulbecco's modified Eagle's medium (Life Technologies, Inc.) containing 10% fetal calf serum, 100 g/ml penicillin, and 100 g/ml streptomycin in a humidified atmosphere of 5% CO 2 and 95% air at 37°C. The 293/Tet-on cells were purchased from CLONTECH. Sodium butyrate and trichostatin A (TSA) were purchased from Sigma and used at final concentrations of 5 mM and 0.3 M, respectively. The cells were plated in six-well plates and transiently transfected using FuGENE 6 (Roche Molecular Biochemicals). The p21 waf1 reporter plasmid p21 waf1 /2300-Luc and ODC reporter plasmid pODC2300-Luc were cotransfected with the pCMV/rZBP-89 expression vector (16). 48 h after transfection, the cells were harvested for luciferase, ␤-galactosidase, and protein assays. At least three transfections in triplicate were performed with each promoter construct. In experiments using butyrate or TSA, the luciferase activity was normalized to cell protein because these treatments stimulated the CMV promoter expressing ␤-galactosidase. However, in the absence of butyrate or TSA treatment, ␤-galactosidase was used to normalize transfections, and no significant differences in plasmid transfection efficiency were detected. Induction of the Tet-on promoter was accomplished with 2 g/ml doxycycline.
Adenovirus Infection of HT-29 Cells-HT-29 cells were cultured in McCoy's 5A medium, grown to 50% confluence, and then infected with the recombinant adenoviral particles (Ad5 vector or Ad5-ZBP-89) in F12 serum-free medium (Life Technologies, Inc.) at 5 ϫ 10 7 viral particles/5 ϫ 10 6 cells/100-mm plate (equivalent to 10 multiplicities of infection) for 6 h. The Ad5 vector, which contains the CMV promoter alone and the poly(A) sequence, was used as a control at the same multiplicity of infection. After infection, the viral particles were washed off, and fresh medium containing serum was added. 36 h later, the cells were treated with or without 5 mM sodium butyrate for another 12 h before they were processed for immunoblot analysis. 50 g of whole cell extracts were separated on a 4 -12% NuPAGE Bis-Tris gradient gel (Novex) and then transferred to polyvinylidene fluoride membrane for immunoblot analysis with designated antibodies. Enhanced chemiluminescence was used to detect the antigen-antibody complexes. The anti-FLAG M1 monoclonal antibody was purchased from Sigma, and the anti-actin (C-2) and anti-p21 waf1 (F-5) monoclonal antibodies were purchased from Santa Cruz Biotechnology.
Ribonuclease Protection Assay-Riboprobes were generated from antisense templates for human ZBP-89 and human cyclophilin. The human ZBP-89 antisense template was constructed using the human ZBP-89 cDNA fragment (28), from ϩ382 to ϩ639 amplified with 5Ј-G-ATGAGAGACAAAAAACAAATCAGAGAGCCAGTAGAC-3Ј as the forward primer and 5Ј-GGTACTTCTGTATGAAACGCATGTACAATT-GACTAC-3Ј as the reverse primer. The PCR product generated was inserted into the pCR2.1 vector (InVitrogen) and sequenced to confirm the orientation. The template was linearized with AccI and used to transcribe a 290-nucleotide ZBP-89 antisense riboprobe that protected a 257-nucleotide fragment. The human pTRI-cyclophilin template (Ambion) generated a 165-nucleotide probe and a 103-nucleotide protected fragment. All riboprobes were prepared using MAXIscript In Vitro Transcription Kit (Ambion). Total RNA was isolated from butyratetreated HT-29 cells using TRIZOL reagent (Life Technologies, Inc.) and then mRNA extracted using poly(A)Ttract mRNA Isolation System (Promega). Total RNA was hybridized for 16 h with riboprobes at 45°C in hybridization buffer (1 mM EDTA, 300 mM sodium acetate, pH 6.4, 100 mM sodium citrate, pH 6.4, 80% deionized formamide). After hybridization, the samples were digested at 37°C for 30 min in an RNase A/T1 mixture containing 60 units/ml RNase A, 250 units RNase T1 (Ambion) in digestion buffer (300 mM NaCl, 10 mM Tris, pH 7.4, 5 mM EDTA). The protected fragment was precipitated in isopropanol, dissolved in loading buffer (95% formamide, 0.025% xylene cyanol, 0.025% bromphenol blue, 0.5 mM EDTA, 0.025% SDS), and then resolved on a 6% polyacrylamide, 8 M urea gel. Protected fragments were quantified on a PhosphorImager and normalized to cyclophilin mRNA levels.
Coimmunoprecipitation Studies-Coprecipitation of ZBP-89 with p300, CBP, E1A, or Sp1 antibody was performed in 293T cells stably transfected with the pTet-on repressor (CLONTECH) and then transiently transfected with pBI-G/rZBP-89 using FuGENE6. The cells were treated with or without 2 g/ml doxycycline to induce ZBP-89 production. The 293T cells were lysed in RIPA buffer (1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% SDS, 1 mM dithiothreitol in phosphate-buffered saline), and the protein concentration was determined. Whole cell extracts were precleared with rabbit preimmune serum and protein A/G-agarose (Santa Cruz) at 4°C for 30 min. The supernatant was incubated with the primary antibody (5-10 g) for 1 h at 4°C followed by the addition of 20 l of Protein A/G-Agarose for another 16 h. The pellets were collected and washed with RIPA buffer three times. The proteins were separated on a 4 -12% NuPAGE Bis-Tris gel (Novex) and then transferred to polyvinylidene fluoride membrane for immunoblot analysis with designated antibodies. Enhanced chemiluminescence was used to detect the antigen-antibody complex. E1A, p300, CBP, and Sp1 antibodies were purchased from Santa Cruz Biotechnology.
DNase I Footprinting Assays-Recombinant ZBP-89 protein was prepared as described previously (16). Protein concentration was determined by the method of Bradford (29). To footprint ZBP-89-binding sites from Ϫ556 to ϩ6, the 560-bp fragment was PCR amplified from p21 waf1 /2300-Luc with forward primer 5Ј-TTCTGGGAGAGGTGAC-CTAGT-3Ј and reverse primer 5Ј-CTTCGGCAGCTGCTCACACC-3Ј. The fragment was end labeled with T4 polynucleotide kinase then digested with BstEII to make the labeled antisense strand. To footprint the proximal p21 waf1 promoter elements, p21 waf1 /300-Luc corresponding to Ϫ291 to ϩ16 was digested with XbaI, end labeled with T4 polynucleotide kinase at ϩ16, and digested with HindIII. The resulting probe was used to footprint the antisense strand. To footprint the sense strand, the same DNA fragment was restricted with SacI and HindIII, end-labeled with T4 polynucleotide kinase, and then digested with XbaI. Footprinting assays were performed with recombinant ZBP-89 prepared as described above. Recombinant ZBP-89 was incubated first with probe on ice for 20 min in the binding buffer (25 mM Tris-HCl, pH 8.0, 50 mM KCl, 6.25 mM MgCl 2 , 0.5 mM EDTA, 10% glycerol, 0.5 mM dithiothreitol, 5% polyvinyl alcohol, 2 g/ml poly(dI-dC). DNase I digestion was then carried out for 1 min, and phenol/chloroform was extracted prior to resolving on 6% polyacrylamide, 8 M urea gel. The Maxim-Gilbert GϩA ladder was prepared using the probe DNA fragments (30).
Electrophoretic Mobility Shift Assay-HT-29 nuclear extract was prepared by the Dignam method without dialysis (31). A doublestranded oligonucleotide probe corresponding to the p21 waf1 promoter Ϫ245 to Ϫ215 (5Ј-GAGGGACTGGGGGAGGAGGGAAGTGCCCTC-3Ј) was end labeled with T4 polynucleotide kinase. The probe was purified using Quick Spin G-25 columns (Roche Molecular Biochemicals). Recombinant ZBP-89, affinity-purified Sp1 (Promega), or nuclear extract was incubated with radiolabeled probe (30,000 cpm/0.2-1 ng) in a final volume of 20 l containing 10 mM Tris-HCl, pH 8.0, 1 mM ZnCl 2 , 120 mM KCl, 1 mM EDTA, 1 g poly(dI-dC), 1 g of bovine serum albumin, 1 mM dithiothreitol, 5 mM MgCl 2 , 10% glycerol at room temperature for 15 min before resolving on a 4% nondenaturing polyacrylamide gel containing 45 mM Tris base, 45 mM boric acid, and 1 mM EDTA. The EMSAs were resolved at 200 V for 2-6 h at 4°C. To perform supershift EMSAs, the protein was incubated first with each antibody for 30 min on ice followed by the addition of probe and incubation for 15 min at room temperature. Antibodies to Sp1 were purchased from Santa Cruz Biotechnology. Rabbit ZBP-89 antiserum was prepared as described previously (32), and the IgG fraction was prepared using protein A-agarose (Santa Cruz).
DNA Affinity Precipitation-Quantitation of the changes in ZBP-89 and Sp1 binding to the p21 waf1 promoter element was achieved by DNA affinity precipitation assays (DAPA) according to the method of Billon et al. (10). Briefly, oligonucleotides biotinylated at the 5Ј-termini and corresponding to the sense Ϫ242 to Ϫ212 (5Ј-GGACTGGGGGAG-GAGGGAAGTGCCCTCCCT), and antisense strands of the p21 waf1 element were annealed. The DAPA was performed by incubating 2 g of biotinylated DNA probe with 300 g of HT-29 whole cell extracts in binding buffer containing 20 mM HEPES, pH 7.9, 10% glycerol, 50 mM KCl, 0.2 mM EDTA, 1.5 mM MgCl 2 , 20 M ZnCl 2 , 1 mM dithiothreitol, and 0.25% Triton X-100. The mixture was incubated on ice for 30 min prior to adding 100 l of streptavidin-agarose (Novagen). 2 h later, the agarose beads were collected and rinsed with binding buffer three times. Protein was eluted from the DNA probe by adding Laemmli loading buffer and heating to 90°C for 5 min. The eluted protein was resolved on a SDS 4 -12% polyacrylamide gel (Novex) transferred to polyvinylidene fluoride membrane followed by immunoblot analysis for ZBP-89 or Sp1 protein levels. To analyze changes in ZBP-89 and Sp1 protein after butyrate treatment over 1 h, immunoblot analysis was performed on the same extracts used for the DAPA. Whole cell or nuclear extracts were prepared from HT-29 cells, separated on a 4 -12% NuPAGE Bis-Tris gel (Novex), and transferred to polyvinylidene fluoride membrane. The membrane was blocked for 1 h in 0.5ϫ UniBlock (Analytical Genetic Testing Center) at room temperature. ZBP-89 and Sp1 antibodies were used to detect the respective proteins.
A potential mechanism of ZBP-89 growth repression may be due to direct binding to the p21 waf1 promoter. To determine whether ZBP-89 stimulates endogenous p21 waf1 gene expression, HT-29 cells were infected with adenoviral particles expressing ZBP-89 ( Fig. 2A). Anti-FLAG antibody was used to distinguish transfected ZBP-89 expression from endogenous ZBP-89 protein. The same membrane was reprobed with antibodies to ZBP-89, p21 waf1 and actin. There was little effect of ZBP-89 on the endogenous level of p21 waf1 protein in the absence of butyrate. However, in the presence of butyrate, there was a 3-fold enhancement of elevated p21 waf1 protein levels in the presence of elevated ZBP-89 protein expression. Cotransfection of the p21 waf1 promoter with a ZBP-89 expression vector slightly repressed basal promoter activity (Fig. 2B). However, in the presence of sodium butyrate, ZBP-89 potentiated the induction of p21 waf1 nearly 3-fold greater than observed with the control vector, consistent with the studies using adenoviralmediated expression of ZBP-89. Activation of the ODC promoter was not observed in the presence of butyrate and ZBP-89 expression, demonstrating that the butyrate effect on the p21 waf1 promoter was specific. Both p21 waf1 and ODC are Radiolabeled probe (10 9 cpm/g specific activity) was hybridized in a reaction mixture containing 150 g of total RNA. Cyclophilin mRNA was detected simultaneously as the control by including its probe (10 6 cpm/g specific activity) in the reaction mixture. A representative autoradiogram is shown. B, ZBP-89 and cyclophilin mRNA levels were quantified on a PhosphorImager and expressed as a ratio relative to the untreated control for three independent experiments. The means Ϯ S.E. are shown.
Identification of ZBP-89-binding Sites within the p21 waf1 Promoter-To identify the preferred site of ZBP-89 binding, we performed footprinting analysis on the first 500 bp of the p21 waf1 promoter (Fig. 3A). The results showed that a sequence at Ϫ245 to Ϫ215 was the major site footprinted. That sequence relative to the other Sp1 sites within the proximal p21 waf1 promoter is shown (Fig. 3B). To study the complexes binding to the ZBP-89 footprint, a probe corresponding to Ϫ245 to Ϫ215 was end labeled and incubated with HT-29 nuclear extracts. Purified ZBP-89 or Sp1 bound to the Ϫ245 to Ϫ215 DNA element (Fig. 3C). Incubation with Sp1, Sp3, or ZBP-89 antibodies alone or in combination revealed the identity of the native complexes in nuclear extracts (Fig. 3D). Competition with WT versus mutant or known ZBP-89-binding sites demonstrated preference of the p21 waf1 element for ZBP-89 binding (Fig. 3, C and D). Scanning mutations within the 30-bp footprinted element identified the ZBP-89 consensus site as GG-GAGG or its converse CCCTCC (Fig. 4A). Both elements were also high affinity binding sites for Sp1. When that site was mutated within the context of the 2.3-kb p21 waf1 promoter, the potentiation by ZBP-89 overexpression was blocked (Fig. 4B). Moreover, mutation of the downstream ZBP-89 consensus site CCCTCC had some effect on the enhancement. However, the double mutation of both sites resulted in the most significant decrease in the potentiation with the induction by butyrate suppressed below what was observed in the absence of ZBP-89 (Fig. 4B). Complete inhibition of the butyrate effect was not achieved perhaps because of intact downstream Sp1 sites that are known to mediate butyrate induction (5).
To determine whether an increase in ZBP-89 binding to the p21 waf1 promoter occurred within the same time frame as activation of p21 waf1 gene expression, both EMSAs and DAPA were performed. Antibody to Sp1 or ZBP-89 was used to supershift their respective antigens (Fig. 5A). Quantitation of the supershifted bands showed that there was no increase in ZBP-89 or Sp1 binding within 60 min of butyrate treatment. These results were confirmed by using DAPA (10) to detect DNA-protein interactions (Fig. 5B). Moreover, there was no significant increase in total ZBP-89 and Sp1 protein within the first 60 min of butyrate treatment (Fig. 5B). The lack of increased Sp1 binding to a p21 waf1 element in response to nerve growth factor has been previously reported (10). Thus our results confirm other reports demonstrating no significant increase in complex binding to account for p21 waf1 promoter induction.
ZBP-89-dependent Transcriptional Activation of p21 waf1 Requires the N-terminal Domain-To determine which domain of ZBP-89 was required for the butyrate-dependent activation, a series of ZBP-89 mutations were prepared (Fig. 6A). An immunoblot was performed to demonstrate that the constructs were expressed in nuclear extracts, and EMSAs were performed to demonstrate that the mutants could bind DNA (Fig. 6, B and C). Cotransfection of full-length ZBP-89 versus C-terminal and N-terminal truncated forms showed that the expected potentiation mediated by butyrate was significantly reduced with any combination of deletions that did not include the N-terminal acidic domain. There was less reduction of the enhanced activation with the C-terminal truncated form. Moreover, the zinc finger domain alone significantly reduced p21 waf1 promoter induction below what was expected in the absence of the ZBP-89 expression vector (Fig. 6D). The N-terminal truncated forms did not mediate the enhanced butyrate induction yet were still capable of binding the p21 waf1 promoter (Fig. 6C), probably preventing endogenous ZBP-89 from binding to the target promoter elements. The reduced potentiation reveals the contribution made by endogenous ZBP-89 to activation of the 2.3-kb p21 waf1 promoter by butyrate. Hasegawa and co-workers (24) have shown that ZBP-89 binds to the third proximal GCrich site of the mouse p21 waf1 promoter that also binds Sp1. Thus the mutant ZBP-89 construct likely affects both the high affinity ZBP-89 site footprinted as well as lower affinity downstream sites. However, there are at least five other downstream Sp1 sites that do not appear to be affected by overexpression of the ZBP-89 ZF mutant. Thus as observed with the mutations of the upstream ZBP-89-binding sites (Fig. 4B), there still remained some residual activation of the p21 waf1 promoter mediated presumably by Sp family members.
The Histone Acetyltransferase p300 Enhances ZBP-89 Acti-   3. ZBP-89, Sp1, and Sp3 bind p21 waf1 promoter elements. A, DNase I footprint of ZBP-89 binding to the human p21 waf1 promoter. A 300-bp p21 waf1 promoter fragment (from -285 to ϩ16) was end labeled at Ϫ285 or at ϩ16, to create the sense and antisense footprinting probes, respectively. Lanes 1 and 10, Maxam-Gilbert GϩA ladder; lanes 2 and 6, 0.4 g of GST protein; lanes 3 and 7, 0.8 g of GST-ZBP-89; lanes 4 and 8, 0.8 g of GST protein; lanes 5 and 9, 1.6 g of GST-ZBP-89. B, the p21 waf1 promoter sequence from Ϫ290 to ϩ16. The ZBP-89 footprint is indicated in bold type. The six known GC-rich/Sp1-binding sites are underlined, and the TATA box is italicized. C, recombinant ZBP-89 and Sp1 protein binding to p21 waf1 element. A kinase end labeled double-stranded oligonucleotide probe corresponding to Ϫ245 to Ϫ215 of human p21 waf1 promoter and 200 ng of recombinant ZBP-89 or 10 ng of affinity-purified Sp1 (Promega) were used in EMSAs. The compounds (Fig. 7). In addition, the activation appeared to be due to the involvement of a co-activator because E1A reduced the expected potentiation (Fig. 7).
To determine whether ZBP-89 forms protein-protein interactions with co-activators, co-immunoprecipitation studies were performed. The results showed that ZBP-89 interacts with p300 but not CBP (Fig. 8A), whereas Sp1 did not interact with either p300 or CBP as reported previously (15). Moreover ZBP-89 also co-precipitated with E1A as did p300 and Sp1 (Fig.  8A). This result demonstrated preferential complex formation between p300, ZBP-89, and E1A. Sp1 is apparently part of this complex through its association with ZBP-89 because ZBP-89 co-precipitated with Sp1 antibody (Fig. 8B). The potentiation by ZBP-89 was most significantly affected by removal of the N-terminal domain (see Fig. 6D). Therefore to determine whether the interaction between p300 and ZBP-89 occurs through the N-terminal or C-terminal domain, co-precipitation studies were performed with full-length and ZBP-89 deletion mutants (Fig. 8C). Lanes 1-4 show that the proteins transfected into 293T cells were specifically detected with the designated antibodies. Lanes 6 and 8 show co-precipitation of the full-length and C-terminal truncated form of ZBP-89 with p300. However, the N-terminal truncated form of ZBP-89 did not co-precipitate with p300 (Fig. 8C, lane 7). To examine directly whether the co-activator p300 cooperates with ZBP-89 to activate the p21 waf1 promoter, co-transfection experiments were performed in HeLa cells with the ZBP-89 construct without the C-terminal domain. There was no activation with p300 and the N-terminal ZBP-89 deletion mutant (data not shown). As shown in Fig. 8D, co-expression of p300 and ZBP-89 stimulated p21 waf1 promoter activity. Moreover, an E1A expression vector blocked this activation. The E1A RG 2 mutant, which contains a point mutation of E1A within the p300-binding site, failed to inhibit p300 activation, whereas the ⌬CR 2 deletion mutant did not prevent the expected E1A-mediated inhibition (Fig. 8D). DISCUSSION Regulation of p21 waf1 transcription by several extracellular signals is p53-independent and targets GC-rich sites within the proximal promoter (5). Sp1 and Sp3 almost invariably bind to these sites. Histone deacetylase inhibitors, e.g. butyrate or TSA, are examples of extracellular signals that induce growth arrest and cellular differentiation while stimulating p21 waf1 gene expression (9,25,34). An increase in the level of histone acetylation and subsequent relaxation of chromatin at sites of active transcription is thought to be one mechanism by which butyrate activates gene expression (35). Site-specific transcription factors assist in the recruitment of co-activators, e.g. p300, to specific regulatory sites within targeted promoters (14). Although butyrate-dependent activation of p21 waf1 transcription requires Sp1, the interaction between Sp1 and p300 is indirect raising the likelihood that other factors participate in the direct interaction with p300 (15). The novel result reported here is that ZBP-89 is one of those transcription factors interacting preferentially with p300. Both zinc finger transcription factors essentially recognize the same DNA-binding domain (16,19) and interact with each other (22). Thus, a plausible model of p21 waf1 promoter activation by butyrate is through direct association of ZBP-89 with the histone acetyltransferase co-activator p300 within the transcription regulatory complex that includes Sp1 (Fig. 9).
ZBP-89 is a four-zinc finger transcription factor that binds and represses or activates several target genes; yet, little is known about its regulation. Clues to its function have been deduced from observations implicating its role in cell growth (32,36). In addition, all target genes that are repressed by ZBP-89 are themselves regulated by mitogens or developmental signals. To study its role in proliferation, the HT-29 colon cancer cell line was used because of its ability to undergo morphologic and biochemical differentiation with butyrate treatment. If the transcriptional repression mediated by ZBP-89 was related to its growth regulatory effects, then we predicted that extracellular mediators of growth arrest should stimulate ZBP-89 gene expression. Indeed, we found that butyrate stimulates ZBP-89 gene expression.
Focusing on a known target of ZBP-89 that mediates growth arrest, i.e. p21 waf1 , we proceeded to define a mechanism by which ZBP-89 may inhibit cell growth. Two high affinity ZBP-89-binding sites were distal to the proximal GC-rich sites bound by Sp1. Mutating these sites effectively abolished enhanced p21 waf1 activation mediated by butyrate, but complete suppression of the induction was not achieved. This was likely due to proximal Sp1 sites that are also recognized by ZBP-89 (24). Removing the N-terminal domain also abolished the potentiation, whereas overexpression of the zinc finger DNAbinding domain alone further reduced the induction of the promoter by butyrate perhaps because of its ability to displace endogenous ZBP-89 and weakly bound Sp1 from these proximal sites. The N-terminal domain was not only important for butyrate-mediated transactivation but was also required for enhancement by p300. Moreover, E1A adenoviral protein, which binds p300, inhibited the ZBP-89-mediated enhancement of p21 waf1 induction by both butyrate and TSA. Consistent with the functional data, co-precipitation experiments revealed that p300 co-precipitated with ZBP-89 and E1A but not Sp1. However, E1A co-precipitated with both ZBP-89 and Sp1. It has been reported previously that Sp1 can co-immunopre- cipitate with E1A (37). Our results also concur with those of Xiao et al. (15), who clearly demonstrated no direct interaction between p300 and Sp1. Moreover, co-precipitation of Sp1 with ZBP-89 documented in this report is consistent with the direct interaction between these two proteins as previously reported by Wieczorek et al. (22), implicating p300, ZBP-89, and Sp1 as components of a multi-factor regulatory complex.
The role of the ZBP-89 acidic domain is consistent with other Krü ppel-type factors targeted by histone acetyltransferase coactivators. Passantino et al. (20) demonstrated an active regulatory domain within the N terminus of ZBP-89. We show here that ZBP-89-dependent activation as well as p300 potentiation requires an intact N-terminal domain. The first 70 amino acids of ZBP-89 are comprised of nearly 25% acidic residues (16). Thus, the N terminus of ZBP-89 like other Krü ppel-like factors, e.g. gut-enriched Krü ppel-like factor and erythroid Krü ppellike factor, is highly acidic and binds p300 (38,39). In contrast, Sp1, which does not directly bind p300, contains a glutamine- rich activation domain (40). Thus, the charge differences may direct the interaction of site specific transcription factors to their respective co-activator.
A variety of transcription factors bind CBP/p300 (41), and ZBP-89 appears to be one of several transcription factors that also forms a complex with at least one of these coactivators. The co-precipitation studies show that ZBP-89 binds p300 but not CBP. Differences in co-activator targets have been reported (14). p300 does not acetylate CREB despite their tight association (42). Retinoic acid-dependent differentiation requires p300 but not CBP (43). p300-deficient and CBP-deficient mice exhibit different phenotypes (44). Mice null for p300 die in midgestation and exhibit defects in cell proliferation in addition to specific tissue defects (44). Recently, it has been demonstrated that CBP may regulate proliferation specifically in hematopoietic tissue (45). Point mutations in p300 have been found in gastric cancer (46). Collectively, these results suggest that p300 essentially behaves as a bifunctional co-activator mediating both growth-suppressing and -activating activities (14). The fact that CBP/p300 are both negative regulators of the cell cycle, mediate differentiation or apoptosis, and may be tumor suppressors (47) is consistent with the proposed role for ZBP-89 in growth inhibition. Although p53 binds p300 and DNA activation of p21 waf1 requires p53, HT-29 cells are p53deficient (25). ZBP-89 appears to have a preference for p300 over CBP, which may reflect the direct involvement of p300 in growth suppression.
Through their ability to activate human proliferating cell nuclear antigen that interact with p53, CBP/p300 may sense DNA damage and activate DNA repair (47,48). By this mechanism, the reported interaction of an 80-kDa GADD34-like (growth arrest DNA damage) protein with the mouse ZBP-89 homologue BFCOL1 (24) may have relevance because the GADD protein family senses DNA damage and arrests growthindependent of p53 status (49,50). To our knowledge, association of a histone acetyltransferase gene product with GADD proteins has not been reported. Moreover, the role of GADD34 and ZBP-89 in p21 waf1 gene expression is not known (24).
In summary, these data support the aggregation of a large complex of factors binding to the p21 waf1 promoter that include ZBP-89, Sp family members and coactivators. Contact with p300 and subsequent transcriptional activation may depend upon the extracellular signal. In the case of butyrate, our data would support butyrate induction of p21 waf1 through promotion of a p300 and ZBP-89 interaction with the subsequent recruitment of Sp1 to the complex through its association with ZBP-89 (Fig. 9). Histone deacetylase 1 binds Sp1 in G 0 raising the possibility that cofactors other than histone acetyltransferases may target this transcription factor in the resting state (51).
Although not evaluated in this study, it is tempting to speculate on the possibility of butyrate inhibiting histone deacetylases associated with specific transcription factors on specific  9. Proposed mechanism for ZBP-89-dependent regulation of p21 waf1 transcription by butyrate. Histone deacetylase inhibitors, e.g. butyrate or TSA, create an environment of elevated histone acetylation by inhibiting histone deacetylases, thus favoring histone acetyltransferase activity at specific sites. The p300 co-activator preferentially associates with ZBP-89 rather than Sp1. The Sp1/ZBP-89 containing complexes regulate the p21 waf1 promoter at GC-rich sites.
promoters. In this way, the level of histone acetylation would rise like a rheostat to switch transcription on at specific promoter sites.