Repression of Interleukin-5 Transcription by the Glucocorticoid Receptor Targets GATA3 Signaling and Involves Histone Deacetylase Recruitment*

Glucocorticoids are the mainstay of asthma therapy and mediate the repression of a number of cytokine genes, such as Interleukin (IL)-4, -5, -13, and granulocyte macrophage colony-stimulating factor (GM-CSF), which are central to the pathogenesis of asthmatic airway inflammation. The glucocorticoid receptor (GR) mediates repression by a number of diverse mechanisms. We have previously suggested that one such repressive activity is by direct binding of GR to elements within the GM-CSF enhancer that are recognized by the nuclear factor of activated T cells·activator protein 1 (NF-AT·AP-1) complex. We reasoned that, because many cytokine genes activated in asthma are transcriptionally regulated by the recruitment of this complex to DNA, their binding sites might provide a target for GR to mediate its repressive effects. Here, we show that transcriptional repression of the Interleukin-5 gene involves recruitment of GR to a DNA region located within the IL-5 proximal promoter, which is bound by NF-AT and AP-1 proteins. GR recruitment had a profound effect upon the activation capacity of GATA3, which has a binding site close to the NF-AT·AP-1 domain in both IL-5 and IL-13 promoters. Repression by GR involves co-repressor recruitment, because treatment of transfected cells with the deacetylase inhibitor trichostatin A caused a partial relief of repression. Additionally, repression could be augmented by co-transfection of cells with a histone deacetylase (HDAC1). These data suggest that the local recruitment of GR causes repression by inhibiting transcriptional activation by GATA3, a key tissue-specific determinant of expression of Th2 cytokines.

Transcriptional regulation by the glucocorticoid receptor (GR) 1 is known to take place by a number of diverse mechanisms. GR was initially demonstrated to mediate transcriptional activation of the murine mammary tumor virus (MMTV) gene by binding as a dimer to a consensus response element, the GRE (1). Subsequently, however, the array of mechanisms utilized by GR in controlling gene expression has expanded, and in particular, details of transcriptional repression by GR have emerged. These mechanisms involve both DNA binding (2,3), the demonstration of the negative GRE (nGRE) (4), the recruitment of either coactivator or co-repressor complexes to a common subunit (5), the ability to sequester non-DNA-bound proteins (6,7), and the regulation of kinase activity (8). It is now clear that some of these systems may operate in conjunction with one another, thereby imparting both variability and complexity of transcriptional regulation upon a target promoter.
The genes encoding IL-4, -5, -13, and GM-CSF lie in close proximity on human chromosome 5q (9). The products of these genes are pro-inflammatory cytokines, which mediate both asthmatic and allergic responses. Our previous work has described a mechanism for transcriptional repression of the GM-CSF gene by glucocorticoids, which involves binding of GR to NF-AT⅐AP-1 sites within the GM-CSF enhancer (10). The NF-AT⅐AP-1 complex mediates signaling through the T cell receptor and acts upon numerous genes including IL-4, -5, -13, and GM-CSF (11). Specificity of transcriptional regulation is governed, at least in part, by factors such as GATA3 (12) and c-Maf (13), which precisely target selected genes rather than having ubiquitous functions. To provide further evidence that NF-AT⅐AP-1 sites might be sites of action for glucocorticoids, we have used the IL-5 promoter. This promoter has a number of NF-AT⅐AP-1 sites in addition to a GATA3 response element (14). GATA3 is a key determinant of T cell differentiation, which has been proposed to control chromatin conformation at the 5q locus. In addition to IL-5, GATA3 also binds the IL-13 proximal promoter (15). As GATA and AP-1 proteins have previously been demonstrated to functionally interact (16), we tested the hypothesis that glucocorticoid repression might effect GATA3-dependent transcriptional activation.
Here we demonstrated that GR mediates IL-5 repression, at least in part by acting through an NF-AT⅐AP-1 site in the proximal promoter. This site has a consensus NF-AT recognition sequence yet lacks an obvious AP-1 site. GR was able to bind to this site in vitro. Repressive activity could be augmented by co-expression of a histone deacetylase, whereas treatment of cells with the HDAC inhibitor TSA partially relieved repression. These data suggested that repression was an active process. This NFAT⅐AP-1 site juxtaposes a crucial GATA site in the proximal promoter of IL-5. This proximity prompted us to examine the effect of glucocorticoids on GATA3-mediated activation. Our data demonstrated that glucocorticoids had a profound repressive effect and suggest that GATA3, a key driver of T cell differentiation and IL-5 and -13 transcription, may be a target for the repressive effect of glucocorticoids on pro-inflammatory Th2 cytokine genes.

EXPERIMENTAL PROCEDURES
Purification of CD4 ϩ ve T Cells-Venous blood was taken from healthy human volunteers using heparin as an anticoagulant. Peripheral blood mononuclear cells were isolated by density-gradient centrifugation using Lymphoprep (Nycomed, Oslo, Norway) according to the manufacturer's instructions. CD4 ϩ T cells were purified using positive selection (Dynal, Oslo, Norway). CD4 ϩ purity was Ͼ95%. The cells were resuspended at 1 ϫ 10 6 /ml in RPMI 1640 medium supplemented with 10% fetal calf serum and IL-2 (50 units/ml). Ethical approval for the use of human volunteers for this study was provided by the Institutional Ethical Review Committee.
Stimulation of CD4 ϩ ve T Cells-CD4 ϩ ve T cells were stimulated in 24-well microtitre plates coated with anti-CD3 (OKT-3) at 1 g/ml. Where indicated, dexamethasone (Sigma) was added to a final concentration of 10 Ϫ7 M immediately following plating. Cultures were incubated for 18 h at 37°C, and then the cells were harvested for RNA extraction.
Differentiation of Human Th2 Cells-Th2 cells were generated from naïve T cells following the protocol outlined in Cousins et al. (17). Day 28 Th2 cells were used in this study. The cells were activated using anti-CD3/anti-CD28 either alone or in the presence of 10 Ϫ7 M dexamethasone for 18 h and then harvested for RNA extraction.
Cell lines and Culture Conditions-Jurkat cells were grown in RPMI 1640 (Invitrogen) medium, HEK293 cells in Dulbecco's modified Eagle's medium, and HeLa cells were grown in minimal essential medium, all supplemented with 10% fetal calf serum (Sigma), L-glutamine (2 mM), penicillin (100 IU/ml), and streptomycin (100 g/ml) at 37°C and 5% CO 2 in humidified air. Where indicated, cells were activated with 100 ng/ml phorbol dibutyrate and 1 g/ml ionomycin (Calbiochem). Dexamethasone (Sigma) was stored at a concentration of 10 Ϫ2 M in ethanol and then further diluted in the appropriate culture medium and added to the cells to give the relevant final concentration. TSA (Sigma) was used at a final concentration of 33 ϫ 10 Ϫ9 M.
Transfections-Transfections and CAT assays were carried out as previously described (10). Jurkat T cells were grown in 10% fetal calf serum until the day prior to transfection and then transferred to medium containing serum that had been stripped with charcoal to remove endogenous steroid and maintained in this medium posttransfection. 4 ϫ 10 6 Jurkat cells were transfected with 5 g of reporter plasmid DNA Ϯ1-3 g of expression plasmid DNA. Electroporation was carried out at 300 mV, 960 microfarads, ϱ ohms, with a Gene Pulser (Bio-Rad). The samples were activated and treated with dexamethasone 10 min post-transfection as indicated. Cells were incubated at 37°C, 5% CO 2 , in humidified air for 20 h, harvested by centrifugation, and cell lysates assayed for CAT activity. HEK293 and HeLa cells were transfected by calcium phosphate precipitation as previously described (21). For HeLa cells, 1 g of reporter plasmid was used along with the same amount of expression vectors for the different transcription factors. Plasmid concentrations were equilibrated with an empty CMV expression vector. For histone deacetylase inhibition assays, TSA was added to cell cultures at a final concentration of 33 nM where indicated.
Electrophoretic Mobility Shift Assays-Electrophoretic mobility shift assays were performed as previously described using purified FLAG-tagged GR protein (10). Oligonucleotide probes were labeled with [␥-32 P]ATP with T4 polynucleotide kinase according to standard procedures (22), and 0.5 l of GR was incubated with 7.6 fmol 32 P end-labeled oligonucleotide duplex plus 100 ng of poly(dI⅐dC) (500fold excess) in binding buffer (10 mM Tris-HCl, pH 7.5, 1 mM MgCl 2 , 0.5 mM Na 3 EDTA, 50 mM NaCl, 0.5 mM dithiothreitol, 4% glycerol), for 15 min at room temperature. Where indicated, 152 fmol (20-fold excess) unlabeled specific competitor DNA was added to the binding reaction. Complexes were resolved on 5% polyacrylamide gels run in 0.3ϫ Tris borate-EDTA buffer (22). The glucocorticoid response element (GRE) corresponding to Ϫ187 to Ϫ161 from the mouse mammary tumor virus long terminal repeat was used as a positive control for GR binding and has the following sequence: 5Ј-GATCGTTTATG-GTTACAAACTGTTCTTAAAACA-3Ј (23). The IL-5 Ϫ130 to Ϫ90 oligo used in electrophoretic mobility shift analysis was 5Ј-ctagTAAGA-TATAAGGCATTGGAAACATTTAGTTTCACGATATGC-3Ј.
Immunoprecipitations-Immunoprecipitations were performed as previously described (10) with modifications. HEK293 cells were transfected with CMV.hGR␣, CS2ϩMT.Sin3A, and CMV.FLAG.H-DAC1 or CMV.FLAG.GATA3 as a negative control. Whole cell lysates were prepared in IPH buffer (50 mM Tris-Cl, pH 8.0, 150 mM NaCl, 5 mM EDTA, 0.5% NP40 (IGEPAL, Sigma), 1 mM phenylmethylsulfonyl fluoride), and FLAG-tagged proteins were immunoprecipitated by incubation with anti-FLAG-conjugated agarose beads (Sigma) at 4°C. Immunoprecipitates were captured by centrifugation, and the beads were washed four times with IPH buffer containing 500 mM NaCl, with thorough mixing of samples at each wash. Immunoprecipitates were then resuspended and electrophoresed through 8% polyacrylamide gels. Proteins were transferred to nitrocellulose membranes, blocked in TBS/Tween 20/4% milk, and then probed with anti-GR antiserum (Transduction Laboratories), anti-Myc tag (New England Biolabs), or anti-FLAG (Sigma). The location of immunoreactive complexes was revealed by ECL (Amersham Biosciences) and autoradiography.

Th2 Cytokine Transcription Is Repressed by Glucocorticoids-
The effects of the synthetic glucocorticoid dexamethasone on expression of genes within the human 5q locus was analyzed using primary human CD4 ϩ T cells and primary human Th2 cells. The cells were stimulated with anti-CD3/ anti-CD28 for 18 h, either alone or in the presence of 10 Ϫ7 M dexamethasone before harvest and cDNA preparation. Reverse transcription-PCR analysis demonstrated that dexamethasone treatment inhibited IL-4, -5, and -13 and GM-CSF, although having no effect on either RAD50 RNA or 18 S ribosomal RNA expression (Fig. 1A). These data were confirmed using quantitative real-time PCR using cDNA derived from the same source of cells. PCR reactions were normalized for 18 S ribosomal RNA expression, and cytokine gene expression was significantly repressed by glucocorticoid treatment compared with RAD50 (Fig. 1C). Data derived from freshly isolated CD4 ϩ cells, containing both naïve and memory populations, was similar to that generated from differentiated Th2 cells. These data demonstrated that glucocorticoids are able to act broadly to mediate repression of 5q cytokine genes.
Repression of IL-5 Transcription Is Mediated by the Proximal Promoter-Although glucocorticoids have numerous potential functional targets, we wished to assess effects mediated through interactions of GR with DNA. The IL-5 promoter was therefore functionally tested for its capacity to mediate glucocorticoid responsiveness. Two CAT reporter vectors were generated, driven by either 447-or 130-bp of IL-5 promoter DNA ( Fig. 2A). Jurkat T cells were transiently co-transfected with NF-ATc, c-Fos, and c-Jun expression vectors, treated with or without 10 Ϫ7 M dexamethasone, and activated with phorbol dibutyrate/ionomycin for 20 h. A combination of NF-AT and AP-1 proteins could transcriptionally up-regulate both IL-5 promoter fusions, and this up-regulation could be repressed by treatment with dexamethasone. By contrast, a consensus GRE, driving CAT expression under control of the minimal SV40 promoter, was potently activated by glucocorticoids (Fig. 2B). These data suggest that a GR-responsive domain resides within the most proximal 130 bp to the IL-5 transcriptional start site. domain also contains previously described response elements for GATA3, as well as CLE0 binding proteins. These three subdomains were isolated from the IL-5 promoter and were cloned upstream of the TATA box of the E1b minimal promoter (Fig. 3A). In these experiments, we asked whether a given IL-5 region, once activated by NF-AT/AP-1, could be repressed by glucocorticoids. The transcriptional activity of each activated construct was ascribed a value of 100%, and the effect of glucocorticoids was assessed relative to this value. The data demonstrated that the minimal domain mediating repression mapped between Ϫ130 and Ϫ90 relative to the IL-5 transcriptional start site (Fig. 3B). By contrast, the GATA.CLE0 element (Ϫ85 to Ϫ39) and the CLE0 (Ϫ63 to Ϫ39) were slightly activated by the addition of glucocorticoids. The potent down-regulation of the endogenous message demonstrated in Fig. 1 suggests that the inhibitory effect of glucocorticoids overrides any local activation at the GATA/CLE0 site.

Repression of the IL-5 Promoter Is
Glucocorticoid Receptor Binds to the IL-5 NF-AT/AP-1 Response Element in Vitro-To understand the mechanism of repression, we determined whether purified GR was capable of directly binding to this Ϫ130 to Ϫ90 NF-AT/AP-1 element using a band retardation assay. In this experiment, we compared the capacity of GR to bind to IL-5 and to a consensus murine mammary tumor virus GRE and also the capacity of the IL-5 element to act as a competitor of GR binding to the GRE and vice versa. We observed comparable DNA binding ability of GR to the two sites, and each individual site showed almost equivalent activity as a competitor (Fig. 4). These data  2 and 6, respectively). GR binding to each oligonucleotide was competed with a 20ϫ excess of either self-oligo- (lanes 3 and 7) or non-self-specific competitor (lanes 4 and 8). Gels were dried and the position of radioactive complexes revealed by autoradiography. suggest that in this in vitro scenario, GR was capable of binding directly to the Ϫ130 to Ϫ90 IL-5 element.
Repression of IL-5 Is Mediated by Histone Deacetylases-GR is known to mediate its repressive effects by a variety of mechanisms. Many classes of transcription factor impose their transcriptional effects through the recruitment of proteins that have the ability to modify nucleosomal histone tails, particularly the amino termini of H3 and H4. Acetylation and deacetylation have been demonstrated to be involved in gene activation and repression, respectively. We asked whether the repressive activity of GR at the IL-5 promoter was active or whether it merely blocked the accessibility of the positively acting NF-AT⅐AP-1 complex to DNA. We addressed this in two ways: first, by the inclusion of the HDAC inhibitor TSA in our repression experiments and second, by HDAC overexpression. We reasoned that if there was no active component, TSA would have no effect on repressive capacity. TSA was, however, able to relieve GR-mediated transcriptional repression of reporter constructs driven by the minimal 41-bp NF-AT/AP-1 response element (Ϫ130 -Ϫ90.IL-5.CAT), the IL-5 promoter (-130-ϩ35.IL-5.CAT), and the GM-CSF enhancer driving CAT expression from the minimal promoter E1b (GM-CSF.E1b.CAT) (Fig. 5A). These data imply that histone deacetylases are involved in the repression of IL-5. HDAC1 was able to augment GR-mediated repression when overexpressed in transient transfections using both IL-5 and GM-CSF reporter constructs (Fig. 5B). Taken together, these data suggest that GR promotes active repression that is facilitated by co-repressor recruitment.
GR Interacts with HDAC1-To determine whether GR was detectable in a complex with HDAC1, we overexpressed both proteins and Sin3A, a HDAC co-factor, in HEK293 cells and performed co-immunoprecipitation experiments. GR immunoreactivity was present in FLAG immunoprecipitates from FLAG HDAC1-expressing cells, as was Myc-tagged Sin3A. By contrast, no GR was detected in FLAG immunoprecipitates from cells expressing the negative control, FLAG.GATA3 (Fig.  6). These data suggest that GR may form a complex with histone deacetylases. The interaction between GR and HDAC1 appeared to be stable, because it was able to withstand vortexing in 500 mM NaCl/IPH wash buffer.
Glucocorticoids Potently Repress GATA3-mediated Activation of the IL-5 and IL-13 Promoters-PCR of cDNA derived from the primary CD4 ϩ and Th2 cells demonstrated that cytokine messenger RNA expression was potently repressed by glucocorticoids (Fig. 1). By contrast, the transfections demonstrated only a partial repressive activity mediated by GR binding to NF-AT/AP-1 response elements. In these experiments, we envisaged competition between NF-AT/AP-1 and GR for binding to the same DNA element. The consequence of this competition would be that the two distinct classes of transcription factors would recruit co-activator (via NF-AT⅐AP-1) or co-repressor (via GR) complexes to the same DNA site, thereby imparting conflicting regulatory signals upon the transcriptional apparatus. To address how this site occupancy would influence repression by GR, we used the fact that the IL-5 promoter can be activated by GATA3, which binding site juxtaposes the Ϫ130 to Ϫ90 NF-AT/AP-1 site. We reasoned that binding of GATA3 to its site and activation via this route would leave the NF-AT/AP-1 site vacant to permit unimpeded DNA binding by GR. This was done in parallel with transfection of a combination of both NF-AT/AP-1 and GATA3. We performed these co-transfection experiments on both IL-5 Ϫ130 to ϩ35 and IL-13 Ϫ227 to ϩ50 promoters, which show similar organization with respect to the proximity of NF-AT/AP-1 and GATA3 response elements. With both promoters, NF-AT/AP-1-mediated activation was only partially repressible by glucocorticoids (Fig. 7). When the NF-AT response element was vacant, as when the constructs were activated by GATA3, treatment with glucocorticoids caused an almost complete repression. We observed no ability of GR to interact with GATA3 as assessed by co-immunoprecipitation (Fig. 6), and we had previously shown that GR does not mediate repression by acting through the GATA.CLE0 site (Fig. 3). When both GATA3 and NFAT/AP-1 proteins were used as the activator, the constructs reverted to being only partially repressible by steroids. The data suggest that, at both IL-5 and IL-13 proximal promoters, the most profound effect of glucocorticoids is upon GATA3-dependent transcriptional activation and that NF-AT/AP-1 impairs this repression by competing with GR for DNA binding.

DISCUSSION
The cytokines, IL-4, -5, and -13 are expressed by the Th2 subclass of helper lymphocytes (24). Data from intracellular cytokine staining of primary Th2 cells has shown a rapid induction of synthesis upon cell activation, suggesting that the chromosome 5q cytokine locus is in a conformation that is permissive for gene expression, yet requisite of an external stimulus (17). The establishment of either a permissive or non-permissive environment for transcription in Th1 and Th2 cells during their differentiation from a naïve precursor provides a detailed paradigm for regulation of gene loci. Data from naïve primary murine T cells suggests that they are poised to permit rapid expression of small quantities of both IL-4 and IFN-␥ (25). During differentiation to Th2 cells, under the influence of external IL-4, the ability to express IFN-␥ is diminished and is accompanied by loss of local histone acetylation (25). By contrast, development of the Th2 lineage is accompanied by increased histone acetylation at the IL-4 locus and an increase in locus accessibility as measured by the appearance of novel DNaseI hypersensitivity sites. One factor in particular, GATA3, has been postulated to be involved in the establishment of these Th2 hypersensitive sites and in maintaining histone acetylation at the murine cytokine gene locus (26). In general, these hypersensitive sites tend not be located at promoters, although it is suggested that they co-localize with regions rich in potential GATA3 binding sites. In addition, the proximal promoters of IL-5 and IL-13 are bound by and transcriptionally up-regulated by GATA3 (14,15). These data suggest that GATA3 has roles both proximal to and distal from transcriptional start sites of the target genes, perhaps reflecting an influence both on the establishment and maintenance of a locus-wide chromatin structure, as well as having a local, permissive influence on the transcription of specific genes.
Th2 cytokine synthesis in vivo is up-regulated during both asthmatic and allergic responses. Glucocorticoids are the mainstay of therapeutic intervention for these conditions (27), and although steroids are known to alleviate the pro-inflammatory actions of these cytokines, their precise mode of action in regulating cytokine function has not been fully dissected. Repressive roles of the receptor have been shown to encompass a wide variety of mechanistic actions, including DNA binding and tethered and sequestering activities (28). Additionally, GR has been shown to up-regulate transcription of regulatory partners of transcription factors mediating pro-inflammatory responses (IB) (29) and to inhibit activatory kinases such as c-Jun Nterminal kinase (8). It is likely that these activities can act either in isolation or in combination, depending upon the DNA context and the admixture of factors recruited to the promoter. It has therefore been difficult to construct a common paradigm for repressive functions of GR; rather, an alternative approach has been to dissect individual response elements.
The nuclear hormone receptors appear to cycle on and off their target promoters once bound by ligand (30). Studies on the estrogen receptor also illustrated an exchange of the cofactor complement recruited during successive periods of estrogen receptor DNA binding (31). Because many of these cofactors exert their effects by mediating covalent modifications of nucleosomal histone tails, it is postulated that such an exchange facility might permit a chain of different modifications to be implemented during successive stages of a transcriptional event. Cycling on and off DNA of the receptor might permit scanning of the local steroid ligand concentration, such that any influence on transcriptional activity may be maintained if ligand is present or rapidly uncoupled to terminate effects when ligand levels fall (32).
Our previous data on the GM-CSF gene suggests that one effect of GR localized to the GM-CSF enhancer, which acts as a composite response element for NF-AT⅐AP-1 complexes and has four binding sites for these factors (33). In addition, this enhancer contains a number of potential GATA sites and is the location of DNaseI-hypersensitive site formation (34). We demonstrated the capacity for GR to bind directly to these NF-AT/AP-1 sites in vitro. Such an activity could potentially repress transcription by a number of mechanisms. First, it may prevent binding of NF-AT⅐AP-1 complexes and consequent recruitment of co-activators. Second, GR may prevent co-activator recruitment, or it may induce a conformation on the NF-AT/AP-1 proteins that favors co-repressor recruitment. Third, GR may directly recruit co-repressor complexes. We chose to determine whether a similar mechanism of action to that observed in the GM-CSF gene was operative in another gene located within the human 5q locus, IL-5, which plays a pivotal role in eosinophil function. We showed that IL-5 mRNA expression is repressed by glucocorticoids and that a component of this repression maps to a 41-bp response element that mediates NF-AT/AP-1 signaling. This region has also recently been suggested to be the target of the Wolf-Hirschorn syndrome candidate protein, which acts as a transcriptional repressor and has homology to the SET domain proteins (35), indicative of potential histone methyltransferase activity. It has previously been suggested that the IL-4 gene is regulated by glucocorticoids and that this regulation involved HDAC recruit-ment (36). Furthermore, Adcock and colleagues (37) were the first to suggest the ability of GR to recruit the HDAC complex to cytokine promoters. Although our data also suggest HDAC recruitment, the mode of recruitment and the consequences of this co-repressor recruitment differ from previous studies.
We have shown that the HDAC inhibitor TSA relieves GRmediated repression of IL-5, suggesting that HDAC complexes were recruited to the promoter. This view is supported by the finding that GR co-immunoprecipitates with FLAG-tagged HDAC1, although whether this was a direct interaction was not established. We also demonstrated that co-expression of HDAC1 augmented GR-dependent transcriptional repression. The recruitment of GR to both IL-5 and -13 proximal promoters profoundly influenced transcription mediated by GATA3. The local recruitment of GR may alter the ability of GATA3 either to bind to its target site, to cause transcriptional up-regulation, or maintain an environment that is permissive for transcription. We are currently investigating these possibilities with chromatin immunoprecipitation in primary Th2 cells. These activities present an interesting paradox; that is, what happens to GR to permit recruitment of co-activator complexes following ligand binding in some circumstances, yet co-repressor complexes in others. This ability has been addressed previously. Yamamoto and colleagues (3) demonstrated that recruitment of the co-activator GRIP1 to a tethered GR at the col3A promoter led to repression. These data suggested that the architecture of this previously designated activator could be altered to permit repression. A second line of evidence involves the steroid RNA co-activator SRA. Evans and colleagues reported that a co-factor SHARP (SMRT/HDAC-associated repressor protein) bound directly to SRA and repressed GR-and SRA-mediated transcriptional activation of a GR-responsive reporter construct (5).
Our studies suggest that the ability to recruit particular co-factors may be related to the architecture of GR in a particular context. The emerging data on transcriptional regulation by nuclear hormone receptors serves to illustrate the complexity of co-activator and co-repressor complexes that can assemble at a given promoter, the variety of enzymatic activities that these factors carry, and the dynamic nature of the transcription process.