Adaptor-mediated recruitment of RNA polymerase II to a signal-dependent activator.

The second messenger cAMP stimulates the expression of a number of target genes via the protein kinase A-mediated phosphorylation of CREB at Ser-133 (Gonzalez, G. A., and Montminy, M. R.(1989) Cell 59, 675-680). Ser-133 phosphorylation enhances CREB activity by promoting interaction with a 265-kDa CREB binding protein referred to as CBP (Arias, J., Alberts, A., Brindle, P., Claret, F., Smeal, T., Karin, M., Feramisco, J., and Montminy, M.(1994) Nature 370, 226-228; Chrivia, J. C., Kwok, R. P., Lamb, N., Hagiwara, M., Montminy, M. R., and Goodman, R. H.(1993) Nature 365, 855-859). The mechanism by which CBP in turn mediates induction of cAMP-responsive genes is unknown but is thought to involve recruitment of basal transcription factors to the promoter. Here we demonstrate that CBP associates specifically with RNA polymerase II in HeLa nuclear extracts. This association in turn permits RNA polymerase II to be recruited to CREB in a phospho-(Ser-133)-dependent manner. As anti-CBP antiserum, which inhibits recruitment of CBP and RNA polymerase II to phospho-(Ser-133) CREB, attenuates transcriptional induction by protein kinase A in vitro, our results demonstrate that the CBP•RNA polymerase II complex is critical for expression of cAMP-responsive genes.

A number of hormones and growth factors stimulate the expression of target genes by inducing the reversible phosphorylation of specific transcription factors (4). Although phosphorylation has been shown to regulate a number of nuclear factors by inducing their nuclear targeting or DNA binding activities, the cAMP-responsive transcription factor CREB belongs to a group whose transactivation potential is affected (1,5). In this regard, Chrivia et al. (3) have characterized a nuclear CREB binding protein, termed CBP, which binds to CREB in a phospho-(Ser-133)-dependent manner. The requirement for CBP in mediating cAMP-dependent transcription has been demonstrated by cellular microinjection experiments in which CBP antisera blocked transcriptional induction by cAMP (2) and by transient transfection experiments in which overexpression of CBP could potentiate CREB activity in response to agonist (2,6). Here we examine the mechanism by which CBP interacts with the transcriptional apparatus to induce target gene expression in response to hormonal stimulation. Our results suggest that CBP is constitutively associated with specific components of the transcriptional apparatus and that this association in turn permits recruitment of certain basal factors to promoters of cAMP-responsive genes.

Preparation of Nuclear Extracts and in Vitro Transcription Assays-
Nuclear extract preparations and in vitro transcription assays were carried out as described previously (5). To evaluate the effect of PKA 1 on in vitro transcription reactions, purified recombinant PKA catalytic subunit (1 g) (kindly provided by S. Taylor) was added to HeLa nuclear extracts during transcription reactions. CREB activity was monitored with an adenovirus major late promoter template containing three cAMP-responsive elements (CREs) from the rat somatostatin promoter (Ϫ56 to Ϫ32). Affinity-purified antisera were added to in vitro transcription assays as reported (5).
Immunoprecipitation Assays-For immunoprecipitation assays, HeLa nuclear extract (100 g) was precleared with protein A-Sepharose for 30 min at 4°C. Precleared extract was incubated with primary antibody for 1 h at 4°C. To detect ␣ RNA polymerase II, a monoclonal antiserum, raised against a C-terminal domain polypeptide of the large subunit (Promega), was used. Antibody complexes were recovered by incubation with protein A-Sepharose beads for 1 h at 4°C. Beads were washed three times with buffer (B100) containing 100 mM KCl, 10 mM Tris, 1% Nonidet P-40, resuspended in 2 ϫ SDS loading buffer, and analyzed by SDS-polyacrylamide gel electrophoresis.
Glutathione-Sepharose Chromatography-A CREB cDNA fragment encoding the kinase-inducible domain (aa 88 -160) was fused in-frame to the glutathione S-transferase (GST) cDNA in the pGEX-2T plasmid (Promega). GST⅐KID fusion protein was expressed, and purification from BL21 Escherichia coli was carried out as described previously (3). GST⅐KID fusion protein was phosphorylated at Ser-133 using purified PKA catalytic subunit as described previously (7). For GST affinity chromatography, HeLa extracts (100 g) were added to glutathione-Sepharose beads (50 l) containing GST⅐KID or GST phospho-(Ser-133)⅐KID polypeptides and co-incubated for 1 h at 4°C. Beads were washed three times with B100 buffer (see above) and then evaluated by Western blot analysis.

RESULTS AND DISCUSSION
Preliminary evidence suggesting that CBP migrates as a high molecular mass complex of 2000 kDa during gel filtration chromatography (not shown) prompted us to examine whether CBP might stimulate cAMP-responsive genes by virtue of its association with specific basal transcription factors. When purified from HeLa nuclear extracts by phosphocellulose chromatography ( Fig. 1, top), CBP was detected predominantly in the 0.3 M KCl "B" fraction, which also contained RNA polymerase II. Following subsequent fractionation over a Mono-S ion exchange resin (Fig. 1, bottom), CBP again eluted with peak fractions of RNA polymerase II. In contrast to the relatively sharp elution profile for CBP, however, RNA polymerase II appeared to be more broadly distributed, indicating that only a fraction of RNA polymerase II may be associated with CBP. By contrast with RNA polymerase II, CBP did not co-elute from the Mono-S column with TFIIB, a basal factor that has been reported to interact with CBP in GST pull-down assays (6).
To test whether CBP in fact associates with RNA polymerase II, we performed co-immunoprecipitation studies with two distinct anti-CBP antisera directed against aa 1-100 (5729) and aa 455-679 (5614) of the protein ( Fig. 2A). Although neither antiserum was capable of recognizing purified RNA polymerase II directly by Western blot or immunoprecipitation assay (not shown), the large subunit of RNA polymerase II was detected in immunoprecipitates of HeLa extracts with both antisera under non-denaturing conditions. In agreement with the broad elution profile of RNA polymerase II following Mono-S chromatography, only a limited fraction (10 -20%) of the RNA polymerase II large subunit appeared to be associated with CBP in HeLa extracts. By contrast with RNA polymerase II, other basal transcription factors such as TBP and TFIIB did not appear to associate detectably with CBP in co-fractionation or co-immunoprecipitation assays (Fig. 1), indicating that the CBP-RNA polymerase II interaction was indeed specific. In this regard, RNA polymerase II was found to co-precipitate with CBP even at high concentrations of KCl (0.8 M), suggesting that this complex was also stable.
Previous reports showing that CBP interacts with a KID (aa 88 -160) in CREB (2, 3) prompted us to examine whether CBP mediates the PKA-dependent recruitment of RNA polymerase II to this region (Fig. 2B). Following affinity chromatography of crude HeLa nuclear extracts over glutathione-Sepharose resin containing either GST⅐KID or GST-phospho-(Ser-133)⅐KID fusion proteins, CBP was specifically bound to phospho-(Ser-133)⅐KID resin. Similarly, RNA polymerase II was detected on resins containing phospho-(Ser-133)⅐KID but not unphosphorylated KID peptide. As purified RNA polymerase II was unable to bind to phospho-(Ser-133)⅐KID directly (not shown), these results indicate that RNA polymerase II is recruited to phospho-(Ser-133)⅐KID via CBP.
To test whether PKA simulates formation of a heteromeric complex consisting of phospho-(Ser-133) CREB⅐CBP⅐RNA po-lymerase II, as predicted by GST affinity chromatography experiments, we performed immunoprecipitation assays on crude HeLa nuclear extracts (Fig. 3). Using a CREB antiserum (253) that can recognize the CREB⅐CBP complex (8), we detected CBP in immunoprecipitates from PKA-treated but not untreated HeLa nuclear extracts. Similarly, the large subunit of RNA polymerase II was recovered specifically from immunoprecipitates of PKA-treated HeLa extracts, demonstrating that PKA induces formation of a phospho(Ser-133) CREB⅐CBP⅐RNA polymerase II complex.
In order to determine whether recruitment of the CBP⅐polymerase II complex to phospho-(Ser-133) CREB is critical for transcriptional induction by PKA, we performed in vitro transcription assays on crude HeLa nuclear extracts (Fig. 4). Addition of PKA to nuclear extracts induced transcription from a cAMP-responsive template containing three consensus cAMP-responsive elements (3 ϫ CRE) approximately 4-fold. But PKA treatment had no effect on an internal control ade- novirus major late promoter template lacking CRE sites. Addition of affinity-purified CBP antiserum, which blocks recruitment of the CBP⅐RNA polymerase II complex to CREB (2), specifically inhibited PKA-inducible transcription from the 3 ϫ CRE template. Unrelated antiserum (anti-corticotropin-releasing factor binding protein) had no effect on PKA induction of the 3 ϫ CRE template, however, demonstrating that the inhibition by anti-CBP antiserum was indeed specific.
Our observation that CBP is found in a high molecular weight complex is supported by recent findings of Maldonado et al., 2 suggesting that CBP is contained within purified prepa-rations of a mammalian RNA polymerase II holoenzyme. In contrast to results of Kwok et al. (6), TFIIB did not appear to associate detectably with CBP in HeLa extracts. These results would suggest that TFIIB may interact with CBP only after being recruited to the promoter.
Our results do not address whether CBP interacts directly with RNA polymerase II. In preliminary GST pull-down assays, purified recombinant CBP polypeptides are unable to associate directly with the large subunit of RNA polymerase II, 3 suggesting that the interaction between CBP and RNA polymerase II may either require post-translational modification (i.e. phosphorylation) or may involve other proteins within the RNA polymerase II holoenzyme complex.
In a previous study, we found that the activity of purified phospho-(Ser-133) CREB in cell-free transcription assays was indistinguishable from that of unphosphorylated CREB (9). In this report, we found the addition of PKA to HeLa nuclear extracts during the transcription reaction was critical for Ser-133 phosphorylation-dependent activity. These results are consistent with recent findings that other PKA-dependent events in addition to CREB phosphorylation are required for transcriptional induction by cAMP in vivo (8). In this regard, CBP contains a consensus PKA phosphorylation site at Ser-1772, and it is tempting to speculate that the interaction between CBP and RNA polymerase II may itself be regulated by CBP phosphorylation. In yeast, interaction between an upstream activator (GAL4) and a component of the yeast RNA polymerase II holoenzyme complex (GAL11) is sufficient for transcriptional induction (10). The importance of CBP in mediating not only cAMP but also mitogen-inducible transcription (2) indicates that CBP may similarly provide contact points for recruitment of mammalian RNA polymerase II holoenzyme by multiple signal-dependent activators.