The MHC class II transactivator (CIITA) activates the expression of multiple genes involved in Ag presentation, but inhibits Th2-type cytokine production, including IL-4, during Th1 cell differentiation. Th1 cells derived from CIITA-deficient mice produce both Th1- and Th2-type cytokines, and the introduction of CIITA to Th2 cells down-regulates Th2-type cytokine gene transcription. Here we show that the IL-4 promoter is regulated by multiple protein-protein interactions among CIITA, NF-AT, and coactivator CBP/p300. The introduction of CBP/p300 and NF-AT enhances the IL-4 promoter activity, and this activation was repressed by CIITA. Furthermore, our data show that CIITA competes with NF-AT to bind CBP/p300 and that this competition dramatically influences transcriptional activation of the IL-4 promoter. We identified two domains of CIITA that interact with two distinct domains of CBP/p300 that are also recognized by NF-AT. CIITA mutants that retain the ability to interact with CBP/p300 are sufficient to inhibit NF-AT-mediated IL-4 gene expression.
Class II transactivator (CIITA)3 is a critical transcription factor activating genes involved in Ag presentation, such as MHC class II, invariant chain, and H-2M genes (1, 2, 3, 4). The deficiency in CIITA results in immunodeficiency in both human and mouse (4, 5). Recently, we have demonstrated that CIITA inhibits the expression of the IL-4 gene during CD4 T cell differentiation, and introduction of CIITA to Th2 cells is sufficient to down-regulate IL-4 production (6). CIITA is not a DNA binding protein. The exact mechanism of CIITA action is not known, but the interaction of CIITA with sequence-specific DNA binding proteins as well as the basal transcriptional machinery may be required for its function (7, 8, 9, 10). Therefore, it is not surprising to observe the different outcome of CIITA function with a different set of regulatory factors depending on the target promoter.
Cyclic AMP response element binding protein (CREB) binding protein (CBP) and p300 are highly homologous nuclear proteins originally identified for their ability to interact with the transcription factor CREB and with adenovirus E1A proteins, respectively (reviewed in Ref. 11). CBP/p300 also plays a critical role during development, and the deficiency of CBP/p300 is fatal (12). The ability of CBP/p300 to interact with multiple, signal-dependent transcription factors, CREB, STATs, MyoD, nuclear hormone receptors, and the basal transcriptional machinery, has led to the proposal that these coactivators function as signal integrators by coordinating complex signal transduction events at the transcriptional level (13, 14, 15, 16, 17, 18, 19). Depending on the context, specific transcription factors can either cooperate or interfere with each other. Another function of CBP/p300 appears to be the modification of chromatin structure. Intrinsic histone acetyltransferase activity has been found in CBP/p300, which is considered to play a crucial role in transcription from tightly packed chromatin (20). CBP/p300 also acetylates transcription factors (21, 22). GATA-1 and p53 are acetylated by CBP/p300, and the DNA binding ability and transactivation potential of these transcription factors are greatly enhanced upon acetylation.
In MHC class II-specific transcription, CBP/p300 functions as a coactivator. CBP/p300 binds to CIITA and stimulates MHC class II promoter activity (23, 24). Adenovirus E1A inhibits MHC class II transcription by interacting directly with CBP/p300 (24). E1A interacts with the same CBP/p300 region that binds CIITA, suggesting that E1A interferes with MHC class II gene expression by targeting interactions between CIITA and coactivators (24). CBP/p300 is also a coactivator for NF-AT-mediated transcription of the IL-2 promoter and the synthetic promoter containing the NF-AT binding sites (25, 26). However, a role for CBP/p300 in IL-4 gene transcription has not been demonstrated.
The observation that CBP/p300 interacts with both CIITA and NF-AT raises the possibility of competitive interactions of two transcription factors for CBP/p300 binding to regulate gene expression. We tested this hypothesis to determine whether the inhibition of IL-4 gene transcription by CIITA is due to the competition between NF-AT and CIITA for CBP/p300 binding. Here, we demonstrate that an interaction of CBP/p300 with NF-AT can lead to dramatic activation of the IL-4 promoter. Furthermore, our data indicate that CIITA interferes with NF-AT binding to CBP/p300, resulting in the down-regulation of IL-4 gene transcription. This study provides further insight into the mechanisms by which IL-4 gene transcription is controlled by multiple protein-protein interactions.
Materials and Methods
Cell culture and transfections
Both 293T human embryonic kidney epithelial cells and the 68-41 T cell hybridoma were maintained in Clicks medium supplemented with 10% FBS, 2 mM glutamine, 100 μg/ml of penicillin, and 100 μg/ml of streptomycin. To generate stable transfectants of 68-41 T cells (27), cells were electroporated with DNA encoding CIITA or neomycin gene and selected with 1 mg/ml of G418 (Life Technologies, Gaithersburg, MD). For the transient transfection of 68-41 cells, 1 × 107 cells were mixed with 10–30 μg of DNA as described in the figure legends. Cells were then electroporated (0.25 kV and 960 μF) using Gene Pulser (Bio-Rad, Hercules, CA) followed by stimulation with PMA (25 ng/ml) and ionomycin (1.5 μM) overnight. Cells were harvested and analyzed for luciferase and β-galactosidase activity as previously described (6).
293T cells were transfected using a standard calcium phosphate method with 2.5 × 105 cells and 1 μg of DNA unless indicated otherwise. Cells were analyzed for luciferase activity 2 days after transfection. The CMV promoter-driven β-galactosidase expression vector was cotransfected in all transfections, and luciferase values were normalized to β-galactosidase activity as described previously (6). Relative luciferase activity (RLA) was calculated using the luciferase activity of cells transfected with the reporter DNA alone as 1 unless noted. Values in all transfections represent the average of at least three independent experiments.
The following DNA constructs were described previously: FLAG-tagged wild-type CIITA (28), antisense CIITA (6), NF-AT1 and NF-AT2 (29), wild-type p300 and p3001–347 (30), CBP (14), E1A wild-type and the mutant lacking the CBP/p300 binding site (31), the luciferase reporter constructs containing the 3-kb or the 157-bp fragment of the IL-4 promoter or the MHC class II Eα promoter (6), the trimerized NF-AT and the minimal IL-4 promoter (32), and the trimerized AP-1 (33). All CIITA mutants were made using the expression vector pcDNA3 containing a FLAG epitope at the N-terminus. CIITA1–331 containing the acidic and the P/S/T domains was generated using the EcoRI-HindIII fragment of the wild-type gene. CIITA408–857 was amplified by PCR using primers 5′-TGCTCTAGACACCGGCGGCCGCGTGAGACACGAGTG-3′ containing XbaI site at the 5′ end and 5′-TCTTGGTGCTCTGTCATCCCT-3′. The PCR product was then digested with XbaI and NcoI and ligated with a fragment encompassing NcoI to the nucleotide position 2702 of CIITA where the stop codon was introduced. CIITA980–1130 was generated using a fragment from BamHI to the 3′ end of cDNA for the cloning. Inferior numbers denote amino acids positions. The integrity of the mutants was confirmed by sequencing.
Nuclear extracts from 68-41 T cells were prepared as previously described (34) after 5 h of stimulation with PMA (25 ng/ml) and ionomycin (1.5 μM). Gel-shift reactions were conducted in 20-μl volumes containing 50 mM NaCl, 5 mM EDTA, 10 mM Tris-HCl (pH 7.5), 5% glycerol, 5–10 μg nuclear extract, 1.5 μg poly(dI-dC) (Roche, Indianapolis, IN), and 30,000 cpm of 32P end-labeled NF-AT oligonucleotides (5′-CGCCCAAAGAGGAAAATTTGTTTCATA-3′, Santa Cruz Biotechnology, Santa Cruz, CA). The following protease inhibitors were added to the reactions: 1 mM DTT, 1 mM PMSF, 2 μg/ml leupeptin, and 2 μg/ml aprotinin. The reactions were incubated on ice for 20 min to allow DNA-protein complexes to form. Reactions were run on a 4.5% nondenaturing polyacrylamide gel at room temperature for 3 h at 150 V in 1× TBE. IFN-responsive sequence oligonucleotides (5′-TCGAATCTCCACAGTTTCACTTCTGCACCTG3′) were used as a nonspecific competitor.
Immunoprecipitation and Western blot
Typically, 1 × 106 293T cells were transfected for immunoprecipitation experiments. Cells were harvested 2 days after transfection, washed with 1× PBS, lysed in 200 μl of ice-cold lysis buffer (1% Triton X-100, 25 mM Tris-HCl (pH 7.5), 150 mM NaCl, 1 mM EDTA, 1 mM PMSF, 1 μg/ml leupeptin, and 1 μg/ml aprotinin), and incubated on ice for 1 h with occasional vortexing. Lysates were clarified through centrifugation at 20,800 × g for 30 min at 4°C. The protein concentration was determined using the Lowry assay (Bio-Rad). p300 coimmunoprecipitation was performed with the anti-p300 Ab (N-15, Santa Cruz Biotechnology) for 1 h with occasional rocking. Immune complexes were collected on protein A-Sepharose beads, then the beads were washed three times with lysis buffer, and bound proteins were resolved on SDS-PAGE and Western blotted. NF-AT and CIITA were detected with polyclonal anti-NF-AT Ab (Santa Cruz Biotechnology) and a monoclonal anti-FLAG Ab (M2, anti-FLAG, Sigma, St. Louis, MO), respectively. HRP-conjugated Ab (Jackson ImmunoResearch Laboratories, West Grove, PA) was used as a secondary Ab followed by enhanced chemiluminescence assay for visualization (ECL, NEN Life Science Products, Boston, MA).
CIITA inhibits CBP/p300-mediated activation of the IL-4 promoter
The significance of CBP/p300 has been demonstrated for the regulation of transcription, including genes encoding molecules that are essential for the proper immune response such as MHC class II and IL-4 (23, 24, 25, 26). Therefore, we first tested whether p300 can activate the IL-4 promoter. 293T human embryonic kidney epithelial cells were transfected with the 3.0-kb IL-4 promoter-driven luciferase and the p300 expression vector or the empty control vector. As shown in Fig. 1,A, the luciferase activity was enhanced by cotransfecting p300 in a dose-dependent manner, indicating that p300 is sufficient to activate the IL-4 promoter in the absence of T cell-specific transcription factors. Studies have demonstrated that the adenovirus E1A protein inhibits CBP/p300-mediated transcription by binding to CBP/p300 (22, 23). Consistent with this, a wild-type E1A, but not a mutant that has deletion of the p300 binding domain, abolished p300-mediated activation of the IL-4 promoter (Fig. 1 B).
We have reported that CIITA inhibits IL-4 gene expression during Th1 cell differentiation and in committed Th2 cells (6). Because CIITA interacts with CBP/p300 (23, 24), we examined whether CIITA represses p300-mediated activation of the IL-4 promoter. When the CIITA expression vector was cotransfected with p300, the luciferase activity driven by the IL-4 promoter was greatly reduced (Fig. 1,C, left panel). This reduction is specific for the IL-4 promoter because the MHC class II promoter was not inhibited by CIITA (Fig. 1,C, right panel). Instead, CIITA together with p300 activated the MHC class II promoter synergistically, consistent with previous studies (23, 24). The inhibition of the IL-4 promoter by CIITA is dose dependent (Figs. 1,D and 4), and the luciferase activity was not affected when the antisense CIITA was cotransfected (Fig. 4, lane 7). These data indicate that p300 activates the IL-4 promoter in nonlymphoid cells, and CIITA inhibits p300-mediated activation of the IL-4 promoter.
The NF-AT binding site of the IL-4 promoter is sufficient to mediate the activation by p300
CBP/p300 activates transcription by interacting with other transcription factors; thus, we wanted to examine which proteins are recognized by CBP/p300 on the IL-4 promoter. To do this, we delineated a cis-acting element(s) of the IL-4 promoter that mediates the activation by p300. Luciferase reporters driven by three different lengths of the IL-4 promoter, 3 kb, 748 bp, and 157 bp, were tested. As shown in Fig. 2,A, all three promoters were activated at comparable levels when p300 was cotransfected (lanes 2, 5, and 8). This activation was repressed by CIITA (lanes 3, 6, and 9). Therefore, the 157-bp fragment of the IL-4 promoter contains a sufficient cis-acting element(s) for the activation and the repression by p300 and CIITA, respectively. There are multiple cis-acting elements within the 157-bp fragment, and studies have demonstrated that NF-AT/AP-1 binding sites are the major element required to activate IL-4 gene transcription (32). In addition, NF-AT has been shown to interact with CBP/p300 (25, 26). Therefore, we tested whether the NF-AT/AP-1 binding site is responsible for the activation by p300. The luciferase construct driven by the IL-4 minimal promoter (−58 to +60) with the trimerized NF-AT/AP-1 binding site of the IL-4 promoter was transfected with p300. The trimerized NF-AT/AP-1 binding site was sufficient to be activated by p300 and repressed by CIITA (Fig. 2,B, lanes 4–6). Because the NF-AT/AP-1 site is a composite site for both NF-AT and AP-1 binding, it is necessary to distinguish whether the activation is mediated by NF-AT or AP-1. p300 did not activate the promoter driven by the trimerized AP-1 binding sites or the minimal IL-4 promoter without the NF-AT binding site (Fig. 2 B, lanes 7–12). Together, these data demonstrate that the NF-AT binding site is the primary target of p300 and CIITA.
CIITA inhibits NF-AT-mediated activation of the IL-4 promoter
It has been shown that NF-AT and CBP/p300 activate the IL-2 promoter or a synthetic promoter containing the NF-AT binding site (25, 26). Moreover, our results, shown in Fig. 2, indicate that the NF-AT binding site of the IL-4 promoter is sufficient to be activated by p300. Therefore, we examined whether protein-protein interaction between NF-AT and p300 mediates activation of the IL-4 promoter. When 293T cells were transfected with the NF-AT1 or the NF-AT2 expression vector, the IL-4 promoter was activated significantly (Fig. 3,A, lanes 2 and 4), and this activation was inhibited by CIITA (Fig. 3,A, lanes 3 and 5). Cotransfection of NF-AT1 or NF-AT2 with p300 further enhanced the luciferase activities, demonstrating the synergistic effect of NF-AT and p300 interaction (Fig. 3,A, lanes 8 and 10). Again, CIITA inhibited the activation by NF-AT and p300 (Fig. 3 A, lanes 9 and 11).
Next, we investigated whether CIITA can inhibit NF-AT-mediated activation of the IL-4 promoter in T cells. We used the 68-41 Th1 cell hybridoma for the transfections. 68-41 cells were transfected with DNA, followed by stimulation with PMA and ionomycin. Consistent with data from 293T cells, the IL-4 promoter was activated by NF-AT1 or NF-AT2, and this activation was inhibited by CIITA (Fig. 3,B, left panel). The addition of CBP and NF-AT augmented the luciferase activity, and it was repressed by CIITA (Fig. 3 B, right panel). Therefore, CIITA inhibits IL-4 gene transcription activated by the interaction of NF-AT and CBP/p300 in both nonlymphoid and T cells.
CIITA and NF-AT compete for p300 binding
The binding of NF-AT to its DNA motif is required to activate the IL-4 promoter. Therefore, we asked whether CIITA inhibits NF-AT binding to the DNA template, causing repression of the IL-4 promoter. To test this, EMSA was performed using nuclear extracts prepared from 68-41 cells stably transfected with the CIITA expression vector or the control DNA. The DNA binding activity of NF-AT was comparable regardless of CIITA expression, suggesting that CIITA does not modulate the DNA binding ability of NF-AT (Fig. 4 A).
Studies have shown that NF-AT and CIITA bind to the same regions of CBP/p300 (23, 24, 25, 26). These observations raise the possibility that the inhibition of IL-4 gene transcription by CIITA is caused by competition between CIITA and NF-AT for CBP/p300 binding. To test this hypothesis, we used two experimental approaches. First, IL-4 promoter activity was measured to determine the effect of CIITA dose. Second, using the same transfected cells protein-protein interactions among NF-AT, CIITA, and p300 were assessed. 293T cells were cotransfected with a constant amount of vector expressing p300 and NF-AT1, and a luciferase reporter driven by the IL-4 promoter. As indicated in Fig. 4 B, an increasing amount of the CIITA expression vector was cotransfected. After cells were harvested, we measured the luciferase activity and performed immunoprecipitation assays.
The luciferase activity was greatly enhanced when p300 and NF-AT1 were cotransfected, confirming the synergistic interaction between p300 and NF-AT1 (Fig. 4,B, lanes 1 and 2). This activation of the IL-4 promoter by p300 and NF-AT1 was repressed by CIITA in a dose-dependent manner (Fig. 4,B, lanes 3–6). The inhibition is specific, because antisense CIITA at the highest dose did not inhibit the IL-4 promoter (Fig. 4 B, lane 7).
If the inhibitory effect is due to the competition of CIITA with NF-AT to bind p300, we should be able to detect more CIITA but less NF-AT protein bound by p300 as the CIITA level increases. To test this, coimmunoprecipitation experiments were conducted using an anti-p300 Ab followed by Western blotting using an anti-NF-AT Ab. The level of NF-AT protein coimmunoprecipitated by p300 was diminished as the amount of the CIITA expression vector increased for transfection (Fig. 4,C, upper panel). The protein level of transfected NF-AT remained constant (Fig. 4 C, lower panel). The membrane was then stripped and reprobed with an anti-FLAG Ab to detect CIITA. As expected, the level of CIITA protein bound to p300 was increased as more CIITA protein was expressed. Taken together, our data demonstrate that the inhibition of IL-4 gene transcription by CIITA is at least partly due to the competition between NF-AT and CIITA for CBP/p300 binding.
CIITA mutants capable of interacting with CBP/p300 inhibit IL-4 gene expression
The acidic domain of CIITA is shown to interact with the CH3 domain of CBP/p300 (23, 24). This predicts that a CIITA mutant lacking the acidic domain should not repress the IL-4 promoter, because it cannot compete with NF-AT to bind CBP/p300. Unexpectedly, a CIITA mutant lacking the acidic domain inhibited the IL-4 promoter at a comparable level as the wild type (data not shown). This result indicates that the mutant may be able to interact with CBP/p300 through another domain. To identify the domain, we divided CIITA into three regions (Fig. 5,A); CIITA1–331 containing the acidic and the proline-serine-threonine rich domain, CIITA408–857 with the GTP binding domain, and CIITA980–1130 that has leucine-rich repeats. Each mutant was then tested for the repression function on the IL-4 promoter that was activated by NF-AT2 and CBP/p300. The three mutants showed different degrees of inhibition, although none of them was able to activate the MHC class II promoter (Fig. 5,B). CIITA1–331 mutant repressed better than the wild type, while CIITA980–1130 or the antisense CIITA could not repress the IL-4 promoter (Fig. 5,B, middle panel, lanes 9, 11, and 12). Interestingly, CIITA408–857 also behaved as a repressor (Fig. 5,B, lane 10). A similar pattern of repression was observed in T cells as well (Fig. 5 B, right panel). These data demonstrate that either the acidic or the GTP binding domain of CIITA is sufficient for the repression of IL-4 gene expression.
We next determined which region of p300 is recognized by CIITA408–857. The CH3 domain is already identified as an interaction site for the acidic domain of CIITA. Therefore, we thought that p3001–347 containing N-terminal 347 aa would be a reasonable candidate, because N-terminus p300 is recognized by other transcription factors, including NF-AT (11, 25, 26). 293T cells were transfected with CIITA mutants with either the full-length p3001–2414 or p3001–347 followed by immunoblot. CIITA408–857 was able to interact with both the wild-type and the amino-terminal regions of p300 (Fig. 5,C, lanes 1 and 2). CIITA980–1130 containing leucine-rich repeats did not interact with p300 (Fig. 5,C, lane 3). The wild-type CIITA also interacted with N-terminus as well as the full-length p300 (Fig. 5 D). These results together with the previous studies imply bivalent interaction between CIITA and CBP/p300: the acidic domain of CIITA with the CH3 region of CBP/p300, and the GTP binding domain of CIITA with N-terminus CBP/p300. Not surprisingly, both regions of CBP/p300 interacting with CIITA overlap with NF-AT recognition sites.
The proper regulation of IL-4 gene expression during Th2 cell differentiation is critical to a host to regulate the immune response upon infection. We have shown previously that CIITA plays a role in the regulation of Th cell differentiation (6). The CIITA gene is expressed in developing Th1 cells, but not in Th2 cells. The deficiency of CIITA during Th1 cell differentiation allows activation of the IL-4 gene, and the introduction of CIITA to Th2 cells down-regulates IL-4 gene transcription. Here, our data showed that the inhibition of IL-4 by CIITA is at least in part due to the competition between CIITA and NF-AT to bind a limited amount of the coactivators CBP/p300.
NF-AT is an essential transcription factor regulating IL-4 gene transcription. There are four NF-AT genes encoding the cytoplasmic subunit, NF-AT1, NF-AT2, NF-AT3, and NF-AT4, although at the protein level NF-AT1 and NF-AT2 are expressed in peripheral T cells and T cell lines (35). In vitro, all can bind to and activate multiple cytokine genes, but the studies in vivo suggest functional differences among NF-AT family members; NF-AT1 and NF-AT2 are negative and positive regulators of the IL-4 gene, respectively (29, 36, 37, 38, 39, 40, 41, 42). In this study we showed that IL-4 gene transcription is synergistically enhanced by the interaction of CBP/p300 with NF-AT1 or NF-AT2 by transfection experiments. The transactivation mediated by CBP/p300 and NF-AT is reduced by CIITA activity. However, CIITA does not affect the DNA binding ability of NF-AT, as demonstrated by EMSA in vitro. In addition, CIITA does not bind to NF-AT (T. J. Sisk and C.-H. Chang, unpublished observations). Instead, our data imply that the repression effect of CIITA is probably due to the reduced availability of CBP/p300 to interact with NF-AT on the IL-4 promoter. Because both CIITA and NF-AT bind to the same domains of CBP/p300, it is not surprising to observe the competition between the two.
CBP/p300 seems to behave as a bridging factor to assemble a competent transcriptional machinery rather than act as a acetyltransferase on the exogenous IL-4 promoter. A CBP mutant lacking the enzymatic activity was able to transactivate the IL-2 promoter (26), and the same mutant activated the IL-4 promoter (S. Roys and C.-H. Chang, unpublished observations). Unfortunately, CBP or p300 homozygous knockout mice are embryonic lethal (12), and thus the role of CBP/p300 in transcription of the IL-4 gene in vivo remains unclear.
None of the CIITA mutants tested activated the MHC class II promoter, indicating that the integrity of CIITA domains is critical for the activation function on the MHC class II promoter. On the contrary, CIITA mutants showed different degrees of IL-4 inhibition. Mutants capable of interacting with CBP/p300 can repress IL-4. This further supports the idea that the inhibition observed in IL-4 gene expression is due to the competition between transcription factors. It would be ideal to show the interaction of CBP/p300 with NF-AT or CIITA present in T cells. However, due to low levels of CBP/p300 in T cells, we were unable to detect immunoprecipitation product by CBP/p300 under a variety of conditions. Therefore, like others, we relied on overexpression systems (25, 26).
Our study also indicates that the interaction between CIITA and CBP/p300 is bivalent. A similar bivalent interaction was demonstrated between NF-κB and CBP/p300; interestingly, this interaction is regulated by the conformational change in NF-κB upon phosphorylation (43). CIITA activity as a repressor also depends on its conformational change, which results in the alteration of its potential to interact with CBP/p300 (T. J. Sisk and C.-H. Chang, unpublished observations). We observed that the CIITA mutant containing the acidic and P/S/T domain inhibited the IL-4 promoter more efficiently than the full-length CIITA even at a lower concentration (S. Roys and C.-H. Chang, unpublished observations). This may be due to a different conformation of the CIITA mutant that could have better accessibility to CBP/p300 or a higher affinity to CBP/p300. CIITA1–331 nearly abolishes IL-4 promoter activity, whereas CIITA408–857 leads to only a modest reduction in promoter activity in 68-41 T cells (Fig. 5 B). This suggests that the acidic domain has a higher affinity for CBP/p300 than does CIITA408–805; however, experiments that will determine the relative affinities of these specific protein-protein interactions remain to be performed.
It should be noted that CIITA is not a global inhibitor of genes that can be activated by CBP/p300. For example, IFN-γ gene expression was not altered in Th1 cells derived from CIITA-deficient mice, although CBP/p300 interaction with STAT1 induces the IFN-γ gene transcription (16, 18). The subtle variance in the recognition of CBP/p300 domain by a different transcription factor in the context of promoter elements may explain the final consequence. It remains to be demonstrated that CIITA controls the binding of NF-AT to CBP/p300 during Th1 cell differentiation to prevent IL-4 gene transcription.
We thank Drs. Roland Kwok, Mike Imperiale, Gary Nabel, Ken Murphy, Mercedes Rincon, Jenny Ting, and Gerald Crabtree for proving reagents. We are very much indebted to Dr. Wes Dunnick for valuable suggestions and the critical reading of the manuscript.
This work was supported in part by National Institutes of Health Grant AI41510 (to C.-H.C.).
Abbreviations used in this paper: CIITA, class II transactivator; CREB, cAMP response element binding protein; CBP, CREB binding protein; RLA, relative luciferase activity.