Distinct effects of cAMP and mitogenic signals on CREB-binding protein recruitment impart specificity to target gene activation via CREB.

Ser-133 phosphorylation of the cAMP-responsive element-binding protein (CREB) is sufficient to induce cellular gene expression in response to cAMP, but additional promoter-bound factors are required for target gene activation by CREB in response to mitogen/stress signals. To compare the relative effects of different signals on recruitment of the coactivator CREB-binding protein (CBP) to CREB in living cells, we developed a fluorescence resonance energy transfer (FRET) assay. cAMP promoted the interaction of CREB with CBP in a phosphorylation-dependent manner by FRET analysis, but mitogen/stress signals were far less effective in stimulating complex formation even though they induced comparable levels of Ser-133 phosphorylation. cAMP and non-cAMP stimuli were comparably active in promoting this interaction in the cytosol; the formation of CREB x CBP complexes in response to non-cAMP signals was specifically inhibited in the nucleus. Non-cAMP signals had no effect on intrinsic CREB- or CBP-binding activities by Far Western blot assay, thereby supporting the presence of a distinct CREB x CBP antagonist. Our studies indicate that the relative effects of cAMP and mitogen/stress signals on CREB x CBP complex formation impart selectivity to gene activation through CREB phosphorylated at Ser-133.

Minimal activators that bind to the KIX domain of p300/CBP identified by phage display screening.

Human gene therapy approaches involving transcription factors often rely on artificial activation domains for transcriptional activation. These domains are often large (e.g., 80 amino acids for VP16), recruit multiple co-activation complexes at once, and offer no fine control over the level of transcription. In an attempt to understand the sequence and structural requirements of a minimal mammalian activator, we employed a molecular diversity approach with a peptide phage display library composed of random eight-amino acid peptides. Using the KIX domain of the mammalian co-activators p300 and CBP as target, we discovered a family of synthetic binding peptides. These peptides share significant homology with natural KIX domain ligands, and are shown to bind an overlapping, yet distinct, surface of p300/CREB-binding protein (CBP). When fused to a heterologous DNA binding domain, these synthetic peptides function as titratable, modular, and potent transcriptional activators in living cells through specific recruitment of p300/CBP, with the level of transcriptional activation proportional to the affinity of the synthetic peptide for the KIX domain. Taken together, our data demonstrate that a molecular diversity approach can be used to discover minimal, co-activator domain-specific synthetic activators, and that transcriptional activation can be modulated as desired at the level of co-activator recruitment.

Structural analyses of CREB-CBP transcriptional activator-coactivator complexes by NMR spectroscopy: implications for mapping the boundaries of structural domains.

A number of signal-dependent and development-specific transcription factors recruit CREB binding protein (CBP) for their transactivation function. The KIX domain of CBP is a common docking site for many of these transcription factors. We recently determined the solution structure of the KIX domain complexed to one of its targets, the Ser133-phosphorylated kinase inducible transactivation domain (pKID) of the cyclic AMP response element binding protein. The NMR studies have now been extended to a slightly longer KIX construct that, unlike the original KIX construct, is readily amenable to structural analysis in both the free and pKID-bound forms. This addition of six residues (KRRSRL) to the C terminus of the original construct elongates the C-terminal alpha3 helix of KIX by about eight residues. On the basis of the NMR structure of the original KIX construct, residues in the extended helix are predicted to be solvent exposed and thus are not expected to contribute to the hydrophobic core of the domain. Their role appears to be in the stabilization of the alpha3 helix through favorable electrostatic interactions with the helix dipole, which in turn confers stability on the core of the KIX domain. These results have important implications for the identification of novel protein domain boundaries. Chemical shift perturbation mapping firmly establishes a similar mode of pKID binding to the longer KIX construct and rules out any additional intermolecular interactions between residues in the C-terminal extension and pKID.

Solution structure of the KIX domain of CBP bound to the transactivation domain of CREB: a model for activator:coactivator interactions.

The nuclear factor CREB activates transcription of target genes in part through direct interactions with the KIX domain of the coactivator CBP in a phosphorylation-dependent manner. The solution structure of the complex formed by the phosphorylated kinase-inducible domain (pKID) of CREB with KIX reveals that pKID undergoes a coil-->helix folding transition upon binding to KIX, forming two alpha helices. The amphipathic helix alphaB of pKID interacts with a hydrophobic groove defined by helices alpha1 and alpha3 of KIX. The other pKID helix, alphaA, contacts a different face of the alpha3 helix. The phosphate group of the critical phosphoserine residue of pKID forms a hydrogen bond to the side chain of Tyr-658 of KIX. The structure provides a model for interactions between other transactivation domains and their targets.

Distinct roles for P-CREB and LEF-1 in TCR alpha enhancer assembly and activation on chromatin templates in vitro.

The distal enhancer of the T-cell receptor (TCR) alpha chain gene has become a paradigm for studies of the assembly and activity of architectural enhancer complexes. Here we have reconstituted regulated TCR alpha enhancer activity in vitro on chromatin templates using purified T-cell transcription factors (LEF-1, AML1, and Ets-1) and the cyclic AMP-responsive transcription factor CREB. When added in combination, these factors activate the TCR alpha enhancer in a highly synergistic manner. Alternatively, the enhancer could also be activated in vitro by high levels of either CREB or a complex containing all of the T-cell proteins (LEF-1, AML1, and Ets-1). Phosphorylation of CREB by protein kinase A enhanced transcription 10-fold in vitro, and this effect was abolished by a point mutation affecting the CREB PKA phosphorylation site (Ser-133). Interestingly, LEF-1 strongly enhanced the binding of the AML1/Ets-1 complex on chromatin, but not nonchromatin, templates. A LEF-1 mutant containing only the HMG DNA-binding domain was sufficient to form a higher-order complex with AML1/Ets-1, but exhibited only partial activity in transcription. We conclude that the T cell-enriched proteins assemble on the enhancer independently of CREB and function synergistically with CREB to activate the TCR alpha enhancer in a chromatin environment.

The signal-dependent coactivator CBP is a nuclear target for pp90RSK.

We have examined the mechanism by which growth factor-mediated induction of the Ras pathway interferes with signaling via the second messenger cAMP. Activation of cellular Ras with insulin or NGF stimulated recruitment of the S6 kinase pp90RSK to the signal-dependent coactivator CBP. Formation of the pp90RSK-CBP complex occurred with high stoichiometry and persisted for 6-8 hr following growth factor addition. pp90RSK specifically recognized the E1A-binding domain of the coactivator CBP. In addition, like E1A, binding of pp90RSK to CBP was sufficient to repress transcription of cAMP-responsive genes via the cAMP-inducible factor CREB. By contrast with its effects on the cAMP pathway, formation of the pp90RSK-CBP complex was required for induction of Ras-responsive genes. These results provide a demonstration of cross-coupling between two signaling pathways that occurs at the level of a signal-dependent coactivator.

Phosphorylation of CREB at Ser-133 induces complex formation with CREB-binding protein via a direct mechanism.

We have characterized a phosphoserine binding domain in the coactivator CREB-binding protein (CBP) which interacts with the protein kinase A-phosphorylated, and hence activated, form of the cyclic AMP-responsive factor CREB. The CREB binding domain, referred to as KIX, is alpha helical and binds to an unstructured kinase-inducible domain in CREB following phosphorylation of CREB at Ser-133. Phospho-Ser-133 forms direct contacts with residues in KIX, and these contacts are further stabilized by hydrophobic residues in the kinase-inducible domain which flank phospho-Ser-133. Like the src homology 2 (SH2) domains which bind phosphotyrosine-containing peptides, phosphoserine 133 appears to coordinate with a single arginine residue (Arg-600) in KIX which is conserved in the CBP-related protein P300. Since mutagenesis of Arg-600 to Gln severely reduces CREB-CBP complex formation, our results demonstrate that, as in the case of tyrosine kinase pathways, signal transduction through serine/threonine kinase pathways may also require protein interaction motifs which are capable of recognizing phosphorylated amino acids.

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.

Pancreatic islet expression of the homeobox factor STF-1 relies on an E-box motif that binds USF.

The commitment of cells to specific lineages during development is determined in large part by the relative expression of various homeodomain (HOX) selector proteins, which mediate the activation of distinct genetic programs. But the mechanisms by which individual HOX genes are themselves targeted for expression in different cell types remain largely uncharacterized. Here, we demonstrate that STF-1, a homeodomain protein that functions in pancreatic morphogenesis and in glucose homeostasis is encoded by an "orphan" homeobox gene on mouse chromosome 5. When fused to a beta-galactosidase reporter gene, a 6.5-kilobase genomic fragment of 5'-flanking sequence from the STF-1 gene shows pancreatic islet specific activity in transgenic mice. Two distinct elements within the STF-1 promoter are required for islet-restricted expression: a distal enhancer sequence located between -3 and -6.5 kilobases and a proximal E-box sequence located at -104, which is recognized primarily by the helix loop helix/leucine zipper nuclear factor USF. As point mutation within the -104 E-box that disrupt USF binding correspondingly impair STF-1 promoter activity, our results demonstrate that USF is an important component of the regulatory apparatus which directs STF-1 expression to pancreatic islet cells.

Expression of murine STF-1, a putative insulin gene transcription factor, in beta cells of pancreas, duodenal epithelium and pancreatic exocrine and endocrine progenitors during ontogeny.

The XlHbox 8 homeodomain protein of Xenopus and STF-1, its mammalian homolog, are selectively expressed by beta cells of adult mouse pancreatic islets, where they are likely to regulate insulin expression. We sought to determine whether the expression of the homeobox protein/s during mouse embryonic development was specific to beta cells or, alternatively, whether XlHbox 8/STF-1 protein/s were initially expressed by multipotential precursors and only later became restricted to the insulin-containing cells. With two antibodies, we studied the localization of STF-1 during murine pancreatic development. In embryos, as in adults, STF-1 was expressed by most beta cells, by subsets of the other islet cell types and by mucosal epithelial cells of the duodenum. In addition, most epithelial cells of the pancreatic duct and exocrine cells of the pancreas transiently contained STF-1. We conclude that in mouse, STF-1 not only labels a domain of intestinal epithelial cells but also provides a spatial and temporal marker of endodermal commitment to a pancreatic and subsequently, to an endocrine beta cell fate. We propose a model of pancreatic cell development that suggests that exocrine and endocrine (alpha, beta, delta and PP) cells arise from a common precursor pool of STF-1+ cells and that progression towards a defined monospecific non-beta cell type is correlated with loss of STF-1 expression.

Insulin expression in pancreatic islet cells relies on cooperative interactions between the helix loop helix factor E47 and the homeobox factor STF-1.

The development of endocrine cell types within the pancreas is thought to involve the progressive restriction of pluripotential stem cells, which gives rise to the four major cell types: insulin-, glucagon-, somatostatin-, and pancreatic polypeptide-expressing cells. The mechanism by which these peptide hormone genes are induced and then either maintained or repressed during development is unknown, but their coexpression in early precursor cells suggests the involvement of common regulatory factors. Here we show that the somatostatin transcription factor STF-1 is also a principal regulator of insulin expression in beta-cells of the pancreas. STF-1 stimulates the insulin gene by recognizing two well defined islet-specifying elements on the insulin promoter and by subsequently synergizing in trans with the juxtaposed helix-loop-helix protein E47. Within the STF-1 protein, an N-terminal trans-activation domain functions cooperatively with E47 to stimulate insulin transcription. As truncated STF-1 polypeptides lacking the N-terminal activation domain strongly inhibit insulin promoter activity in beta-islet cells, our results suggest that the specification of islet cell types during development may be in part determined by the expression of STF-1 relative to other islet cell factors.

Recombinant cyclic AMP response element binding protein (CREB) phosphorylated on Ser-133 is transcriptionally active upon its introduction into fibroblast nuclei.

To date, it has not been possible to determine whether the single phosphorylation of the cyclic AMP response element binding factor (CREB) at Ser-133 is sufficient for the transcriptional activation by cAMP-mediated pathways. Previous in vivo studies investigating this point have relied upon transfection of cyclic AMP-dependent kinase (cAPK) or its activation by treatment of cells with cell-permeable cAMP analogs. However, as numerous cellular proteins, including CREB, are substrates for activated cAPK, the possibility remains that cAPK substrates other than CREB are required for the transcriptional activity of CRE-containing promoters. To further address this, we compared the activity of recombinant CREB phosphorylated on Ser-133 in both cell-free transcription assays and in vivo after introduction of the same preparations into fibroblasts by microinjection. The activity of phosphorylated CREB, nonphosphorylated CREB, and a mutant form of CREB, containing Ala substituted for Ser at position 133, was found to be nearly identical in cell-free in vitro transcription assays. In contrast, we found that only the phosphorylated CREB microinjected into fibroblasts resulted in the stimulation of expression of CRE-regulated genes. These results suggest that phosphorylation of CREB on Ser-133 directly stimulates its ability to transactivate gene expression in intact cells.

Phosphorylated CREB binds specifically to the nuclear protein CBP.

Cyclic AMP-regulated gene expression frequently involves a DNA element known as the cAMP-regulated enhancer (CRE). Many transcription factors bind to this element, including the protein CREB, which is activated as a result of phosphorylation by protein kinase A. This modification stimulates interaction with one or more of the general transcription factors or, alternatively, allows recruitment of a co-activator. Here we report that CREB phosphorylated by protein kinase A binds specifically to a nuclear protein of M(r) 265K which we term CBP (for CREB-binding protein). Fusion of a heterologous DNA-binding domain to the amino terminus of CBP enables the chimaeric protein to function as a protein kinase A-regulated transcriptional activator. We propose that CBP may participate in cAMP-regulated gene expression by interacting with the activated phosphorylated form of CREB.

Characterization of somatostatin transactivating factor-1, a novel homeobox factor that stimulates somatostatin expression in pancreatic islet cells.

The endocrine pancreas consists of several differentiated cell types that are distinguished by their selective expression of peptide hormones such as insulin, glucagon, and somatostatin. Although a number of homeobox-type factors have been proposed as key regulators of individual peptide genes in the pancreas, their cellular distribution and relative abundance remain uncharacterized. Also, their overlapping DNA binding specificities have further obscured the regulatory functions these factors perform during development. In this report we characterize a novel homeobox-type somatostatin transactivating factor termed STF-1, which is uniformly expressed in cells of the endocrine pancreas and small intestine. The 283-amino acid STF-1 protein binds to tissue-specific elements within the somatostatin promoter and stimulates somatostatin gene expression both in vivo and in vitro. Remarkably, STF-1 comprises the predominant tissue-specific element-binding activity in nuclear extracts from somatostatin-producing pancreatic islet cells, suggesting that this protein may have a primary role in regulating peptide hormone expression and specifying endocrine cell lineage in the developing gut.

Coupling of hormonal stimulation and transcription via the cyclic AMP-responsive factor CREB is rate limited by nuclear entry of protein kinase A.

Cyclic AMP (cAMP) regulates a number of eukaryotic genes by mediating the protein kinase A (PKA)-dependent phosphorylation of the CREB transcription factor at Ser-133. In this study, we test the hypothesis that the stoichiometry and kinetics of CREB phosphorylation are determined by the liberation and subsequent translocation of PKA catalytic subunit (C subunit) into the nucleus. Using fluorescence imaging techniques, we observed that PKA was activated in a stimulus-dependent fashion that led to nuclear entry of C subunit over a 30-min period. The degree of CREB phosphorylation, assessed with antiserum specific for CREB phosphorylated at Ser-133, correlated with the amount of PKA liberated. The time course of phosphorylation closely paralleled the nuclear entry of the catalytic subunit. There was a linear relationship between the subsequent induction of the cAMP-responsive somatostatin gene and the degree of CREB phosphorylation, suggesting that each event--kinase activation, CREB phosphorylation, and transcriptional induction--was tightly coupled to the next. In contrast to other PKA-mediated cellular responses which are rapid and quantitative, the slow, incremental regulation of CREB activity by cAMP suggests that multifunctional kinases like PKA may coordinate cellular responses by dictating the kinetics and stoichiometry of phosphorylation for key substrates like CREB.

The CREB family of transcription activators.

A diverse family of transcription factors bind to the cAMP-response elements found in a variety of mammalian and viral gene promoters. One of the members of this family, CREB, is being intensively studied so as to elucidate the mechanisms by which second messenger signal transduction pathways act to positively and negatively regulate transcription.

Somatotroph hypoplasia and dwarfism in transgenic mice expressing a non-phosphorylatable CREB mutant.

Most of the transcriptional effects of cyclic AMP are mediated by the cAMP response element binding protein (CREB). After activation of cAMP-dependent protein kinase A, the catalytic subunits of this enzyme apparently mediate the phosphorylation and activation of CREB. As cAMP serves as a mitogenic signal for anterior pituitary somatotrophic cells, we investigated whether CREB similarly regulates proliferation of these cells. We prepared transgenic mice expressing a transcriptionally inactive mutant of CREB (CREBM1), which cannot be phosphorylated, in cells of the anterior pituitary. If CREB activity is required for proliferation, the overexpressed mutant protein would effectively compete with wild-type CREB activity and thereby block the response to cAMP. As predicted, the CREBM1 transgenic mice exhibited a dwarf phenotype with atrophied pituitary glands markedly deficient in somatotroph but not other cell types. We conclude that transcriptional activation of CREB is necessary for the normal development of a highly restricted cell type, and that environmental cues, possibly provided by the hypothalamic growth hormone-releasing factor, are necessary for population of the pituitary by somatotrophic cells.

Characterization of motifs which are critical for activity of the cyclic AMP-responsive transcription factor CREB.

Cyclic AMP mediates the hormonal stimulation of a number of eukaryotic genes by directing the protein kinase A (PK-A)-dependent phosphorylation of transcription factor CREB. We have previously determined that although phosphorylation at Ser-133 is critical for induction, this site does not appear to participate directly in transactivation. To test the hypothesis that CREB ultimately activates transcription through domains that are distinct from the PK-A site, we constructed a series of CREB mutants and evaluated them by transient assays in F9 teratocarcinoma cells. Remarkably, a glutamine-rich region near the N terminus appeared to be important for PK-A-mediated induction of CREB since removal of this domain caused a marked reduction in CREB activity. A second region consisting of a short acidic motif (DLSSD) C terminal to the PK-A site also appeared to synergize with the phosphorylation motif to permit transcriptional activation. Biochemical experiments with purified recombinant CREB protein further demonstrate that the transactivation domain is more sensitive to trypsin digestion than are the DNA-binding and dimerization domains, suggesting that the activator region may be structured to permit interactions with other proteins in the RNA polymerase II complex.