Up-regulation of p300 Binding and p50 Acetylation in Tumor Necrosis Factor-α-induced Cyclooxygenase-2 Promoter Activation*

It is well established that p300 plays an important role in mediating gene expressions. However, it is less clear how its binding is influenced by physiological stimuli and how its altered binding affects transactivator acetylation and binding. In this study, we determined p300 binding to a core cyclooxygenase-2 (COX-2) promoter region by chromatin immunoprecipitation and streptavidin-agarose pull-down assays in basal and tumor necrosis factor-α (TNFα)-treated human foreskin fibroblasts. We found basal binding of p300, p50/p65 NF-κB, cyclic AMP regulatory element-binding protein-2, CCAAT/enhancer-binding protein β, and c-Jun. p50/p65 and p300 binding was selectively increased by TNFα. Immunoprecipitation confirmed direct interaction of p300 with NF-κB and the other involved transactivators. p50 acetylation was detected in resting cells and was increased by TNFα or lipopolysaccharide. Overexpression of p300 augmented p50 acetylation, which was attenuated by deletion of its histone acetyltransferase domain. Enhanced p50 acetylation correlated with increased p50 binding to COX-2 promoter and transcriptional activation. Co-transfection of E1A with p300 abrogated p50 acetylation and p50 binding. These findings suggest that up-regulation of p300 binding and its acetylation of NF-κB occupies a central position in COX-2 promoter activation.

Cyclooxygense-2 (COX-2) 1 catalyzes the synthesis of robust prostaglandins and thromboxane (1). It is highly inducible in many cell types by cytokines and oncogenic and mitogenic factors (2). COX-2 has been shown to play an important role in inflammation, angiogenesis, and tumorigenesis (3)(4)(5)(6)(7). Its involvement in these pathophysiological processes depends largely on its transcriptional activation by diverse stimuli. Its promoter activation by pro-inflammatory mediators has been extensively investigated. Several regulatory elements located at the 5Ј-flanking untranslated region including a cyclic AMP response element (CRE) at Ϫ53 to Ϫ59, a CCCAAT/enhancerbinding protein (C/EBP) element at Ϫ124 to Ϫ132, and two NF-B sites at Ϫ438 to Ϫ447 and Ϫ213 to Ϫ222 are involved in human COX-2 transactivation (8,9). We have recently demonstrated binding of CREB, C/EBP␤, and NF-B to their respective binding sites in this region (10,11). Our results indicate that COX-2 induction in human fibroblasts and endothelial cells requires binding of multiple transactivators. p300/CREBbinding protein overexpression has been shown to up-regulate COX-2 transcription (12)(13)(14), suggesting an important role for p300 in bridging the multiple DNA-bound transactivators with general transcription factors to initiate COX-2 transcription. p300 belongs to a large class of transcription co-activators, which serve as adaptors for transcriptional activation of diverse genes (for a review see Refs. [15][16][17]. It is a 2414-amino acid protein containing several domains for binding to transactivators, adenoviral E1A, and general transcription factors. It is a histone acetyltransferase (HAT) that acetylates histone and induces chromatin remodeling to facilitate transactivation (18). p300 has been shown to acetylate transactivators and enhance their binding to DNA (19,20). Because p300 is capable of binding to CREB (21), C/EBP␤ (22), and NF-B (23), it is likely to be a major co-activator for COX-2 transcriptional activation. However, there was little reported data about p300 binding to COX-2 promoter, nor was there information about p300 acetylation of COX-2-bound transactivators. In this study, we determined p300 interaction with transactivators bound to the core COX-2 promoter region and assessed the role of p300 HAT in regulating COX-2 transcriptional activity in human fibroblasts stimulated with tumor necrosis factor-␣ (TNF␣). Our results show that TNF␣ up-regulated binding of p50/p65 NF-B, which correlated with enhanced p300 recruitment to the promoter complex. Deletion mutation of p300 HAT reduced p300 recruitment and severely attenuated its ability to enhance basal and TNF␣-induced COX-2 promoter activity. Our results further show that p300 acetylated p50 NF-B but not p65, C/EBP␤, or CREB-2. These findings indicate that p300 mediates and regulates COX-2 transactivation by multiple mechanisms including a selective acetylation of p50 NF-B.

EXPERIMENTAL PROCEDURES
Plasmids-A promoter region of human COX-2 gene (Ϫ891 to ϩ9 from the transcription start site) was constructed into the luciferase reporter vector pGL3 as previously described (8). The expression vectors containing full-length p300 (pCL.p300) and its HAT deletion mutant (CL.p300⌬HAT, ⌬1472-1522) (19) were provided by Dr. Joan Boyes. The E1A expression vector (24)  serum-free medium for 24 h, washed with phosphate-buffered saline, and incubated in fresh medium in the presence or absence of 10 ng/ml TNF␣ (Sigma), 100 nM of phorbol 12-myristate 13-acetate (Sigma), or 2 g/ml of lipopolysaccharide (Escherichia coli 026:B6, Sigma) at 37°C for 4 h. After washing with chilled phosphate-buffered saline three times, the cells were harvested and processed to prepare cell lysates or nuclear extracts. All of the tissue culture reagents were obtained from Invitrogen.
Transient Transfection-The transfection procedure was performed as previously described (10). In brief, 10 l of LipofectAMINE 2000 reagent (Invitrogen) and 4 g of luciferase expression constructs were mixed, and the mixture was slowly added to each well of HFb grown in a 6-well plate and incubated for 24 h. The cells were washed, incubated in serum-free medium, and treated with TNF␣ (10 ng/ml). The expressed luciferase activity was measured in a luminometer (TD-20/20). To evaluate the effect of p300 or p300 ⌬HAT on COX-2 promoter activation, 2 g of pCL.p300, pCL.p300 ⌬HAT, or pCL.vector plus 4 g of luciferase expression constructs were mixed with 15 l of Lipo-fectAMINE 2000 reagent, and the mixture was added slowly to cells cultured in a 6-well plate. To evaluate the effect of p300 overexpression on p50 acetylation or binding, 10 g of pCL constructs were mixed with 25 l of LipofectAMINE 2000 reagent, and the mixture was added to cells cultured in a 10-cm dish. Co-transfection of E1A and p300 was performed by mixing 4 g of E1A plasmid construct with 6 g of p300 construct and 25 l of LipofectAMINE, and the mixture was slowly added to ϳ90% confluent cells in serum-free medium in a 10-cm dish and incubated for 3 h. 10% fetal bovine serum was added and incubated for an additional 21 h. The cells were washed and incubated in serumfree medium for 24 h prior to the addition of TNF␣ or vehicle control.
Western Blot Analysis-Western blot analysis was performed as previously described with minor modifications (25). In brief, cell pellets were lysed with lysis buffer containing 50 mM Tris-HCl, pH 7.5, 150 mM NaCl, 1 mM EDTA, 1 mM phenylmethylsulfonyl fluoride, 1 g/ml leupeptin, 5 g/ml aprotinin, 1% Nonidet P-450, 0.5% sodium deoxycholae, and 0.1% SDS. The lysate was centrifuged, and the supernatant was collected and boiled for 5 min. The protein concentration was determined. The lysate proteins were separated by electrophoresis in a 4 -15% SDS-PAGE minigel (Bio-Rad) and then electrophoretically transferred to a nitrocellulose membrane (Amersham Biosciences). Western blots were probed with a specific rabbit polyclonal anti-p300 antibody (Santa Cruz Biotechnology). The protein bands were detected by enhanced chemiluminescence (Pierce).
Immunoprecipitation-Nuclear extracts were prepared from HFb by a method previously described (26). 800 g of nuclear extracts were incubated with a specific rabbit polyclonal antibody against p300, c-Jun, CREB-2, C/EBP␤, p50, or p65 (all from Santa Cruz Biotechnology) at a final concentration of 4 g/ml each overnight at 4°C. Protein A/G FIG. 1. TNF␣ increased p300 and p50/p65 NF-B binding to chromatin COX-2 promoter region. Chromatin fragments prepared from HFb treated without or with TNF␣ were immunoprecipitated (IP) with specific antibodies against p300, c-Jun, CREB-2, C/EBP␤, p50, or p65, and the COX-2 promoter region (Ϫ32 to Ϫ709) in the chromatin precipitate was amplified by PCR under identical conditions. Rabbit nonimmune IgG was included as a negative control. a, a representative set of results. b, densitometric analysis of basal versus TNF␣-treated samples from three experiments. Each bar denotes the mean Ϯ S.D. Only the difference in p300, p50, and p65 binding was statistically significant (p Ͻ 0.05 for each).
FIG. 2. TNF␣ increased p300, p50, and p65 binding to a biotinylated COX-2 promoter probe. Nuclear extracts from HFb were incubated with streptavidin-agarose beads and a biotinylated COX-2 promoter sequence Ϫ30 to Ϫ453. Proteins in the complex were analyzed by Western blots. a, results from a representative experiment. b, comparison of the densitometry of basal versus TNF␣-treated samples in three experiments. Each bar is the mean Ϯ S.D. Only the difference in p300, p50, and p65 binding levels was statistically significant (p Ͻ 0.05).
plus agarose (Santa Cruz) was then added for 2 h at 4°C. The beads were washed four times in RIPA buffer (50 mM Tris, pH 8.0, 150 mM NaC1, 2 mM NaF, 1 mM EDTA, 1 mM EGTA, 1 mM NaVO 4 , 1% Triton X-100) containing protease inhibitors (Roche Molecular Biochemicals), and the immunoprecipitated proteins were separated by SDS-PAGE and analyzed by Western blotting. Control immunoprecipitation was performed with a nonimmune rabbit normal immunoglobulin (Santa Cruz).
DNA-Protein Binding Assay-Binding of p300 to COX-2 promoter DNA sequence was assayed by a technique recently described (12). 80 -90% confluent HFb were incubated in serum-free medium with 10 ng/ml TNF␣ for 4 h before the nuclear extracts were prepared. The biotin-labeled double-stranded oligonucleotides were synthesized by integrated DNA technologies based on human COX-2 promoter se- FIG. 6. Deletion of HAT reduced p300 binding. Nuclear extracts from HFb treated with or without TNF␣ stimulation were incubated with biotinylated COX-2 promoter probe and streptavidin-agarose beads. p300 in the complex was detected by Western blots. Control denotes the use of a nonrelated biotinylated sequence as the probe. a, Western blot of a representative experiment. b, densitometric analysis of three experiments. Each bar is the mean Ϯ S.D. quence Ϫ30 to Ϫ453 (8). A nonrelevant biotinylated sequence 5Ј-AGAGTGGTCACTACCCCCTCTG-3Ј was included as a control. The binding assay was performed by mixing 600 g of HFb nuclear extract proteins, 6 g of biotin-labeled DNA oligonucleotides, and 60 l of streptavidin-agarose beads (4%) with 70% slurry. The mixture was incubated at room temperature for 1 h with shaking. The beads were pelleted and washed with cold phosphate-buffered saline three times. The binding proteins were separated by SDS-PAGE followed by Western blot analysis probed with specific antibodies.
Chromatin Immunoprecipitation (ChIP)-The assay was done as described with minor modifications (27). 80 -90% confluent HFb were serum-starved for 24 h and treated with or without TNF␣ (10 ng/ml) at 37°C for 4 h. 1% formaldehyde was added to the culture medium, and after incubation for 20 min at 37°C, the cells were washed twice in phosphate-buffered saline, scraped, and lysed in lysis buffer (1% SDS, 10 mM Tris-HCl, pH 8.0, with 1 mM phenylmethylsulfonyl fluoride, pepstatin A, and aprotinin) for 10 min at 4°C. The lysates were soni-cated five times for 10 s each time, and the debris was removed by centrifugation. One-third of the lysate was used as DNA input control. The remaining two-thirds of the lysate were diluted 10-fold with a dilution buffer (0.01% SDS, 1% Triton X-100, 1 mM EDTA, 10 mM Tris-HCl, pH 8.0, and 150 mM NaCl) followed by incubation with antibodies against p300, c-Jun, C/EBP␤, CREB-2, p50, p65 NF-B, or a nonimmune rabbit IgG (Santa Cruz) overnight at 4°C. Immunoprecipitated complexes were collected by using protein A/G plus agarose beads. The precipitates were extensively washed and incubated in an elution buffer (1% SDS and 0.1 M NaHCO 3 ) at room temperature for 20 min. Cross-linking of protein-DNA complexes was reversed at 65°C for 5 h, followed by treatment with 100 g/ml proteinase K for 3 h at 50°C. DNA was extracted three times with phenol/chloroform and precipitated with ethanol. The pellets were resuspended in TE buffer and subjected to PCR amplification using specific COX-2 promoter primers: 5Ј primer, Ϫ709 CTGTTGAAAGCAACTTAGCT Ϫ690 , and 3Ј primer Ϫ32 A-GACTGAAAACCAAGCCCAT Ϫ51 . The resulting product of 678 bp in length was separated by agarose gel electrophoresis.
Acetylation of Transactivators-p50, p65, C/EBP␤, or CREB-2 in nuclear extracts was immunoprecipitated with a specific antibody, and the immunoprecipitates were collected by using protein A/G plus agarose beads. After extensive washing the proteins were separated by SDS-PAGE, and acetylated transactivators were detected on Western blots using a monoclonal antibody against acetylated lysine (1:1000 dilution; Cell Signaling Technology).

RESULTS
Up-regulation of p300 Binding to COX-2 Promoter by TNF␣-To determine whether p300 recruitment to COX-2 promoter-transactivator complex was altered by TNF␣ stimulation, we evaluated p300 binding in unstimulated as well as FIG. 7. TNF␣ and lipopolysaccharide increased p50 acetylation. a, nuclear extracts from HFb treated with or without TNF␣ for 4 h were immunoprecipitated with CREB-2, C/EBP␤, p50, or p65 antibodies or a nonimmune rabbit IgG, and the acetylated transactivator in the precipitate was detected by Western blot analysis. IP control denotes immunoprecipitation of nuclear extracts from TNF␣ treated cells with nonimmune IgG. Only acetylated p50 was detected in unstimulated cells which was increased by TNF␣. Densitometric analysis of acetylated p50 (Ac-p50) shows a significant increase in Ac-p50 in TNF␣treated cells (n ϭ 3, p Ͻ 0.05). b, nuclear extracts from cells treated with or without phorbol 12-myristate 13-acetate (PMA, 100 nM) or lipopolysaccharide (LPS, 2 g/ml) for 4 h were immunoprecipitated with a normal rabbit IgG (IP control) or a specific p50 antibody. Acetylated p50 was detected by Western blots using an acetylated lysine antibody. LPS significantly increased Ac-p50 over the basal level (n ϭ 3, p Ͻ 0.05), whereas phorbol 12-myristate 13-acetate did not significantly increase acetylated p50.
FIG. 8. p300 overexpression increased p50 acetylation. Nuclear extracts were immunoprecipitated with an anti-p50 antibody. Acetylated p50 (Ac-p50, a) and total p50 (b) were analyzed by Western blots using an anti-acetylated lysine antibody and an anti-p50 antibody, respectively. The upper panels show a representative Western blot, and the lower panels show densitometric analysis of three experiments. p300 Binding and p50 Acetylation TNF␣-stimulated HFb by ChIP. Chromatin was immunoprecipitated with a p300 antibody, and a COX-2 promoter-enhancer region (Ϫ32 to Ϫ709) containing the essential binding sites for promoter activation was amplified by PCR. Vectortransfected HFb, like native HFb, shows trace p300 binding at basal state that was increased by TNF␣ treatment (Fig. 1). ChIP assays using specific transactivator antibodies also detected binding of c-Jun, CREB-2, C/EBP␤, p50, and p65 NF-B to the core COX-2 promoter region in chromatin structure in unstimulated cells, and TNF␣ treatment resulted in a significant increase only in p50 and p65 NF-B binding (Fig. 1). Binding of p300 and transactivators to COX-2 promoter was specific as immunoprecipitation with a normal rabbit IgG did not show detectable COX-2 promoter fragment (Fig. 1).
We next used the streptavidin-agarose pull-down assay, which provides quantitative information of transactivator binding, to evaluate the effect of TNF␣ on transactivator and p300 binding to a biotinylated COX-2 promoter sequence (Ϫ453 to Ϫ30). Nuclear extracts from HFb treated with and without TNF␣ were incubated with the biotinylated probe and streptavidin-agarose beads. Transactivators and p300 present in the complex were analyzed by Western blots. Consistent with the ChIP assays, c-Jun, CREB-2, C/EBP␤, and p50/p65 NF-B as well as p300 were detected in resting cells, and only p300 and p50/p65 NF-B binding was increased by TNF␣ stimulation (Fig. 2).
Interaction of p300 with COX-2 promoter-bound transactivators was determined by immunoprecipitation of nuclear extract proteins with antibodies against c-Jun, CREB-2, C/EBP␤, p50, p65, or nonimmune IgG and analysis of p300 in the complex by Western blots. p300 was co-precipitated with each of the transactivators in resting and TNF␣-treated cells (Fig. 3a). p300 complexion with each transactivator tended to be increased by TNF␣. However, quantitation of binding is difficult because of large p300 molecular mass and slow mobility in gel electrophoresis. Interaction of p300 with these transactivators was confirmed by immunoprecipitation of nuclear extract proteins with anti-p300 antibodies and identification of transactivators by Western blot analysis (Fig. 3b). Together, these results provide direct evidence for recruitment and binding of p300 to COX-2 promoter in resting cells and an up-regulation of p300 binding as a result of increased p50/p65 binding to the promoter following TNF␣ stimulation.
Attenuation of COX-2 Promoter Activation by p300 HAT Deletion Mutation-HFb expressed a low basal level of p300 (Fig.  4). Transient transfection of p300 increased its level, which was accompanied by a large increase in basal COX-2 promoter activity (Fig. 5), a result consistent with the reported data (12)(13)(14). Overexpression of a p300 HAT deletion mutant by transient transfection to a level similar to that of the overexpressed wild type (WT) protein also increased the basal COX-2 promoter activity, but the increase was only about 30% of that induced by WT p300 (Fig. 5). TNF␣ did not significantly increase native or transduced p300 levels (Fig. 4). However, p300 overexpression augmented COX-2 promoter activity stimulated with TNF␣, and this augmenting effect was greatly attenuated when p300 HAT domain was deleted (Fig. 5). Because the HAT domain of p300 is situated close to transactivator binding domains, its deletion (⌬1472-1522) may influence p300 binding. We therefore compared recruitment of WT and ⌬HAT p300 to the biotinylated probe by the streptavidin bead pull-down assay. The ⌬HAT mu- p300 Binding and p50 Acetylation tant did not bind as well as WT p300, and its binding was close to the basal p300 binding (Fig. 6). Together, these results suggest that p300 plays a major role in regulating COX-2 promoter activity, and its HAT activity is crucial for the regulation of COX-2 transactivation and p300 recruitment.
Selective p50 Acetylation by p300 -It is well established that p300 HAT contributes to promoter activation by acetylating core histones in chromatin structure. In this study, we investigated whether it acetylates COX-2 promoter-bound transactivators. p50, p65, CREB-2, and C/EBP transactivators were immunoprecipitated with their respective antibodies, and the acetylated proteins were detected by Western blots using an antibody specific for acetylated lysine. A low level of acetylated p50 was detected at basal state and was increased by TNF␣ stimulation (Fig. 7a). Neither p65 nor other transactivators were acetylated. Acetylated p50 was increased slightly by stimulation with phorbol 12-myristate 13-acetate (100 nM) for 4 h and was significantly increased by stimulation with lipopolysaccharide (2 g/ml) for 4 h (Fig. 7b). To determine whether p50 acetylation was mediated by p300, we overexpressed WT or ⌬HAT p300 and assayed p50 acetylation. At the basal state, p300 transfection increased p50 acetylation, whereas ⌬HAT caused a much less increase (Fig. 8a). p300 overexpression augmented TNF␣-induced p50 acetylation, which was attenuated by HAT deletion mutation (Fig. 8). As expected, p50 protein levels at basal state were low and were increased by TNF␣ stimulation. Neither p300 nor p300 ⌬HAT transfection altered p50 protein levels (Fig. 8b).
To determine whether p50 acetylation was correlated with an increased p50 binding to the COX-2 promoter, we measured p50 binding to the biotinylated COX-2 promoter in HFb transduced by WT, ⌬HAT, or its control vector. At the basal state, p300 overexpression increased p50 binding by more than 2-fold, whereas ⌬HAT overexpression exerted a lesser increase (Fig. 9). TNF␣ increased p50 binding, which was augmented by WT but not ⌬HAT p300 overexpression (Fig. 9). ChIP assays were performed to evaluate the effect of p300 overexpression on p50 binding to chromatin COX-2 promoter region. Corresponding to the results of in vitro binding experiments shown in Fig.  9, overexpression of wild type p300 augmented p50 binding to COX-2 promoter region in the chromatin structure of resting and TNF␣-stimulated cells (Fig. 10). ⌬HAT overexpression attenuated the increase in p50 binding (Fig. 10). Inhibition of p50 Acetylation by E1A-As adenoviral E1A is a potent inhibitor of p300 co-activator activities (28), we evaluated its effect on p50 acetylation and binding. Its overexpression suppressed COX-2 promoter activities stimulated by TNF␣ and p300 overexpression (data not shown). Overexpression of E1A by transient transfection abrogated p50 acetylation induced by p300 overexpression in the presence or absence of TNF␣ without altering the p50 level (Fig. 11). Inhibition of p50 acetylation by E1A was correlated with reduction of p50 bind- FIG. 11. E1A transfection abrogated acetylated p50 stimulated by p300. HFb co-transfected with E1A and p300 were treated with or without TNF␣. Nuclear extracts were immunoprecipitated (IP) with a p50 antibody or a control IgG. Acetylated p50 (Ac-p50, a) and p50 levels (b) were detected by Western blot analysis using antibodies specific for acetylated lysine (a) or p50 (b). Densitometric analysis shows the means Ϯ S.D. from three experiments. ing to a COX-2 promoter probe (Fig. 12a) and to the chromatin COX-2 promoter region (Fig. 12b). DISCUSSION Results from this study show a low level of p300 recruited to the COX-2 promoter-bound transactivators in resting human fibroblasts. p300 recruitment was enhanced in cells treated with TNF␣ and further augmented by p300 overexpression. Enhanced p300 binding was correlated with an up-regulation of COX-2 promoter activities. These enhancing activities of p300 were abrogated by E1A, an inhibitor of p300. These results indicate that p300 plays a crucial role in regulating COX-2 transcription. Our findings further indicate that p300-mediated COX-2 transcriptional activation depends on HAT. Deletion of the HAT domain resulted in a more than 60% reduction in COX-2 promoter activity induced by p300. p300 HAT is capable of acetylating the N-terminal lysine residues of core histones, thereby modifying the chromatin structure (15)(16)(17). This property of p300 is likely contributing to accessibility of COX-2 promoter regulatory elements for transactivator binding. In this study, our results shed light on another important property of p300 that enhances NF-B dependent COX-2 transcription. Our results indicate that p300 HAT is capable of acetylating p50, thereby increasing p50/p65 NF-B binding to COX-2 promoter. p300 has previously been shown to acetylate p53 (19) and GATA (20). To our knowledge, this is the first report of p300-mediated p50 acetylation under physiological conditions. Interestingly, p50 acetylation by p300 was demonstrated as a mechanism by which human immunodeficiency virus-1 replicates in host lymphocytes (29). Binding of host p50/p65 NF-B to a viral long terminal repeat region has been shown to play a key role in viral genome transcription. In the presence of a viral protein Tat, p300 HAT acetylates p50 at several lysine residues, which results in enhanced p50 binding to its cognate sites. Our results show p50 acetylation in TNF␣-treated cells in the absence of the viral Tat. It is unclear whether under physiological conditions, p50 acetylation by p300 also depends on a Tat-like co-factor. That p300 HAT acetylates p50 without a concurrent acetylation of C/EBP␤, CREB-2, c-Jun, or p65 suggests a stringent requirement of an appropriate lysine structural environment in p50 and possibly also p53 and GATA-1 for the action of p300 HAT.
NF-B is a heterodimer typically comprising a p50 subunit that binds to promoter and a p65 subunit that interacts with p300. NF-B is sequestered in cytosol, and upon stimulation, it translocates to the nucleus where it binds to its cognate sites. NF-B has been shown to play a key role in COX-2 transcriptional activation stimulated by TNF␣, lipopolysaccharide, and other pathophysiological stresses. There are two NF-B sites at the core promoter region of human COX-2, and mutation of either site results in loss of response to TNF␣ stimulation (10). In this study, our results show that TNF␣ selectively increased NF-B binding to these sites, which was accompanied by enhanced p300 recruitment and binding to the COX-2 promoter complex. In view of augmented p50/p65 binding and p300 recruitment by p300 overexpression and the requirement of HAT for the p300-induced binding activities, we propose that p300 binding to DNA-bound NF-B is limited by a low level of p300 in cells and that TNF␣ is capable of augmenting p300 binding by a positive feedback loop driven by p50 acetylation. p50 acetylation by HAT of p300 bound to the complex leads to an increased p50/p65 binding, which in turn recruits additional p300 to the complex. This autoregulatory loop ensures up-regulation of NF-B-mediated gene expression. Because NF-B plays a key role in transcriptional activation of myriad proinflammatory genes, this regulatory mechanism has im-portant implications in inflammation, tissue injury, and tumorigenesis.
There is a large body of data supporting the notion that p300 recruitment to DNA-bound transactivators is essential for transcription of many genes. However, there is little information about the role of increased p300 binding in gene transcription under physiological conditions. In this study, our results demonstrate increased p300 binding to chromatin as well as naked COX-2 promoter sequence by TNF␣ stimulation. Increased p300 binding correlated closely with an enhanced p50/ p65 binding. Our results provide further evidence for complex formation between p300, and each involved transactivator and an up-regulation of p300 and p50/p65 in the complex by TNF␣. Together with our previously reported results (10 -12), these findings indicate that at the basal state, p300 is recruited to COX-2 promoter by interacting with constitutively bound CREB and c-Jun at the CRE site, C/EBP␤ at CRE and C/EBP sites, and p50/p65 NF-B at both NF-B sites. TNF␣ stimulation increases p50/p65 binding via p50 acetylation, and the increased NF-B bound to its specific sequence recruits additional p300 leading to amplified transcriptional activation. TNF␣ stimulation could therefore serve as a model for understanding COX-2 transcriptional stimulation by diverse cytokines, growth factors, angiogenic factors, and environmental stress. These agonists induce COX-2 transcription by up-regulating the binding of distinct groups of transactivators, which in turn recruit p300. p300 binding to transactivators drives the transcriptional initiation complex via interaction with general transcriptional factors. Furthermore, p300 acetylates core histone, notably H3 and H4 lysines, to modify chromatin structure and increase transactivator binding. In addition, p300 acetylates selective classes of transactivators, such as p50, thereby further increasing transactivator binding and amplifying transactivation of genes. Increased transactivator binding results in recruiting additional p300, thus creating a positive regulatory loop for COX-2 transactivation. p300 thus occupies a central position in regulating COX-2 promoter activation through its pleiotropic actions.