Cyclic AMP-responsive element binding protein (CREB) is a transcription factor that was originally discovered and characterized from studies of hormone-induced cAMP-dependent regulation of cellular functions. In particular, CREB was found to be critical for stimulus-induced somatostatin upregulation downstream of glucagon and epinephrine. The mechanism of hormone-induced cAMP-dependent increase in somatostatin gene transcription was narrowed down to a 8-bp element in the somatostatin promoter region, which was named cAMP-responsive element (CRE). CREB was identified as a protein factor that bound to CRE sequence and was purified to homogeneity from nuclear extracts using a CRE affinity chromatography. CREB binds to a specific palindromic CRE sequence, 5′-TGACGTCA-3′ as well as a half-site CRE motif (5′-TGACG-3′) (Montminy and Bilezikjian 1987; Mayr and Montminy 2001). Early experiments identified that CRE-bound CREB was in a dimerized state through binding of the C-terminal basic and leucine zipper domain (bZIP), and when the intracellular level of the second-messenger cAMP increased upon various hormonal stimuli, this in turn triggered cAMP-dependent gene transcription in the nucleus via cAMP/PKA-dependent phosphorylation of CREB (Gonzales and Montminy 1989; Mayr and Montminy 2001). Several other intracellular signaling pathways, such as CaMKs and several MAPK-dependent cascades, have also been shown to determine CREB phosphorylation status (Shaywitz and Greenberg 1999; Bito and Takemoto-Kimura 2003). Through regulation of CRE-dependent gene expression, CREB mediates cell growth, survival, death, proliferation, and differentiation, in response to a variety of stimuli in many different cell types. CREB function has especially been studied in the nervous system, among other organs. In the central nervous system, CREB is considered to have a major role in long-term memory formation.
More than 5% of mammalian genes appear to be potentially regulated by CREB-dependent transcription, as determined by genome-wide analyses of CREB-bound promoters (Impey et al. 2004; Zhang et al. 2005). The presence of CRE sequences have been reported within putative promoter elements that drives transcription of numerous functional proteins, such as transcription factors (e.g., c-Fos, Egr-1, Per1, and CCAAT-enhancer-binding protein β), cellular metabolic enzymes (e.g., cytochrome c, phosphoenolpyruvate carboxykinase, cyclooxygenase-2, superoxide dismutase 2, and Bcl-2), growth factors and neuropeptides (e.g., somatostatin, enkephalin, brain-derived neurotrophic factor, insulin-like growth factor, fibroblast growth factor 6, and vasopressin), as well as many neuronal proteins (e.g., synapsin I, TrkB). Recently, CRE sequences are also reported in the regulatory regions of several brain-related miRNAs (e.g., miR-124, miR-132, miR-9). These miRANs, in turn, regulate the expression of CREB itself.
Gene and Protein Structure
Creb1 gene consists of at least 12 exons, several of which are alternatively spliced, resulting in a variety of isoforms. CREB expression is rather ubiquitous. The major isoforms of CREB are α- and Δ-CREB in most tissues. The β isoform is also expressed but at much lower levels and can be upregulated when other isoforms are downregulated. α-CREB is longer than Δ-CREB, by the presence of 14-residue-long insert termed α-peptide. Alternative splicing of several 5′-exons generates β-CREB, which lacks the first 40 amino acid residues of CREB protein. α-CREB, Δ-CREB, and β-CREB are all able of activating CRE-dependent transcription in response to cAMP increase, while several other splice variants, lacking the C-terminal bZIP domain, have been shown to generate inhibitory forms of CREB.
The characteristic functional domains of CREB are: an N-terminal glutamine-rich domain (Q1), a kinase-inducible domain (KID) with a phosphorylation site at Ser-133, another glutamine-rich domain (Q2), and a C-terminal bZIP domain (Mayr and Montminy 2001). Activating transcription factor 1 (ATF1) and cAMP-responsive element modulator (CREM) share structural and functional similarities with CREB as regulators of CRE-dependent gene transcription and form a CREB/ATF family.
CREM has an extensive sequence identity with CREB and has a notably interesting function. This gene generates both activator and repressor forms of CREM from the same gene by alternative splicing. The CREM isoforms α, β, and γ function as inhibitors of CREB- and cAMP-mediated transcription. However, another variant which is most similar to CREB with regards to its amino acid sequence, CREMτ, functions as an activator of CRE-mediated transcription.
Mechanistic Basis of CREB-Mediated Transcriptional Activation
Compared with the variety of kinase-signaling pathways activating CREB, much less information is available about phosphatases that dephosphorylate pSer133. Either Ser/Thr protein phosphatase 1 (PP-1) or PP-2A are shown to dephosphorylate CREB directly at Ser133 after pCREB dissociates from CBP, depending on the cell types. Several studies indicate that other phosphatases, such as calcineurin (PP2B), PTEN, and protein tyrosine phosphatase 1B (PTP1B), may indirectly regulate dephosphorylation of phospho-Ser133, either by suppressing upstream CREB kinases or activating CREB phosphatases.
CREB Co-activators and CREB-Interacting Proteins
It has been shown that CREB constitutively binds functional CRE sites of many genes, even prior to cAMP elevation or Ser-133 phosphorylation events. Thus, the dominant effect of CREB phosphorylation at Ser-133 is to potentiate its transcriptional activity via enhanced interaction of CREB with CBP and its paralog p300. The binding of CBP/p300 to CREB leads to the recruitment of the RNA polymerase II complex to initiate transcription from transcriptional start sites that are adjacent to the CRE sites. Hence, the regulated CREB-CBP interaction via the KID and KIX domains plays a major role in determining the level of CRE-dependent transcription activity. Additionally, the CREB Q2 domain recruits TAF II130 present in the TFIID complex and may contribute to regulating CRE transcription in a rather stimulus-independent manner (Ferreri et al. 1994). An N-terminal glutamine-rich domain (Q1) may also contribute to transcription activity, but the extent of this has not been extensively demonstrated. While CREMτ, a transcription activator among the CREM isoforms and which is highly similar to CREB protein, includes both Q1 and Q2 domains, CREM repressor isoforms possess no domains sharing sequence identities with either Q1 or Q2 regions. These contrasting facts also support the notion that both Q1 and Q2 may play additional roles in CREB transcription activation.
The CREB coactivator role of CRTC is stimulated by CRTC dephosphorylation, which enables its nuclear entry. On the other hand, phosphorylated CRTC, which is excluded from the nucleus, localizes to the cytoplasm and remains in an inactive state. The nucleocytoplasmic shuttling of CRTC is strongly regulated by a balance between calcineurin-dependent dephosphorylation that occurs as a result of cellular calcium mobilization and various regulated kinases (such as SIK, AMPK, and MARK) that inhibit CRTC function (Screaton et al. 2004; Bittinger et al. 2004). For example, among the many phosphorylation sites of CRTC1, Ser151, and Ser245 appeared to critically participate in the activity-dependent dephosphorylation process and contribute to the nuclear translocation in the central nervous system (Nonaka et al. 2014). Thus, CRTC appears to be a valuable sensor of intracellular milieu that can influence CREB function in a manner that is independent of phospho-CREB formation (Fig. 2). Interestingly, CRTC1-CREB signaling activated region-specifically in vivo at the contextual fear memory task (Nonaka et al. 2014). Therefore, CRTC1 may play an important role in determining CREB specificity at spatiotemporal levels in the brain.
Indicator for CREB Activity in Living Cells and Animals
As noted above, Ser-133 phosphorylation is an important factor for CREB-mediated gene expression. For that reason, spatiotemporal profile of CREB activation has been examined by immunostainings using specific antibodies for Ser-133 phosphorylation. However, there are increasing demands for visualizing CREB activity in living cells or animals. Therefore, several genetically encoded indicators have been developed to visualize real-time CREB activity. Using the variants of GFP, ART and ICAP were reported as FRET indicators (Nagai et al. 2000; Friedrich et al. 2010). In the case of ART, FRET signals change when phosphorylation causes a conformation change of KID domain. In ICAP which developed later, FRET signals change by the interaction of KID and KIX domain, followed by phosphorylation. KID/KIX-BEAM, unlike the further two, is a single fluorescent probe using split β-lactamase reporter (Spotts et al. 2002). These indicators can detect the CREB activity for seconds to minutes order in the living cells. In living animal imaging, bioluminescent probes are also reported by using split luciferase and KID-KIX interaction (Ishimoto et al. 2015). Further exploitation is needed for enough spatiotemporal resolution and dynamic range to achieve in vivo CREB activity imaging.
Physiological Functions of CREB
Physiological functions of CREB have been investigated in various organs using genetically modified model animals. In particular, in vivo CREB actions have been intensively studied in the developing and mature nervous and immune systems, as well as in the control of hepatic gluconeogenesis, insulin resistance in obesity, and oncogenesis (Silva et al. 1998; Kandel 2001; Bito and Takemoto-Kimura 2003; Conkright and Montminy 2005; Nonaka 2009).
A full CREB-null phenotype was obtained in knockout mice in which the bZIP domain of CREB gene was deleted. These CREB-null mutants were smaller than wild-type littermates and died immediately after birth due to respiratory distress. The commissural fibers in the brain were markedly reduced. There were also severe impairments in the development of T cells of the α/β lineage but not those of the γ/δ lineage (Lonze et al. 2002). A double mutant mouse line harboring mutations in both CREB and CREM genes showed a more severe neurodegeneration phenotype during brain development, suggesting that CREM gene products may compensate for the loss of CREB gene, especially in the developing central nervous system (Mantamadiotis et al. 2002). A dominant-negative CREB inhibitory peptide (A-CREB) that interferes with the bZIP domain function of CREB (Ahn et al. 1998) has also been successfully used to probe CREB function in various organs, such as liver, adipocyte, bone, skin, and brain, using tissue-specific promoters. In particular, liver-targeted A-CREB expression has demonstrated the critical involvement of CREB in regulating hepatic gluconeogenesis (Dentin et al. 2007).
Interestingly, a CREB mutant mouse line, in which the exon 1 of CREB1 gene was deleted, showed a loss of both α- and Δ-isoforms of CREB, but a compensatory expression of both β-CREB (transcribed in this case from a cryptic alternate promoter) and CREMτ was found in a wide range of organs, except for the brain. These CREB mutant mice showed no obvious developmental abnormality in the body and also no obvious anatomical disorders in the brain. However, when hippocampus-dependent memory tasks were carried out, these adult mice exhibited a profound specific impairment in long-term memory while the short-term memory remained unaltered (Silva et al. 1998).
In order to investigate the roles of CREB activity in hippocampus, brain-targeted VP16-CREB transgenic mice have been generated (Barco et al. 2002). These mice express the constitutively active form of CREB. The hippocampal CA1 neurons of these transgenic mice exhibited a strong facilitation of a persistent late phase of long-term potentiation (L-LTP), which could be elicited even with weak stimuli that would otherwise only induce an early phase of LTP (E-LTP) in the wild-type mice. The results indicate that elevation of CRE-driven gene products in the brain might be sufficient to accelerate consolidation of LTP and are consistent with a “synaptic tagging and capture” hypothesis.
Due to the relationship between CREB and long-term memory, there are several reports of the therapeutic attempts to cure Alzheimer’s disease (AD) by regulating CREB signaling pathway (Teich et al. 2015). For example, inhibitors of phosphodiesterases (PDEs) which hydrolyze cAMP into 5’AMP and lead to downregulate the CREB activity have been reported to rescue the memory deficits in the AD model mice. It is also proposed that histone acetyltransferase (HAT) activity of CBP is another candidate of AD treatment. The histone deacetylases (HDAC) inhibitors and HAT activators are reported to enhance the long-term memory via CREB-CBP dependent transcription. CREB signaling pathway is also suggested to be involved in Parkinson disease, Huntington’s disease, and depression.
Finally, from recent studies, it appears that aberrant CREB activity may be linked to oncogenesis and the etiology of certain leukemia (Conkright and Montminy 2005). Enhanced CREB expression or dominant active mutation in CRTC1 was found in primary human cancer or leukemic cells. Consistently, a transgenic mouse line, which overexpressed CREB specifically in the macrophage/monocyte lineage, exhibited increased proliferation of bone marrow cell in a growth factor–independent manner.
CREB is a well-studied transcription factor that is ubiquitously expressed and is activated by a variety of cellular signaling pathways, downstream of cell growth, hormonal or neurotransmitter stimuli. Mobilization of various second-messenger pathways leads to CREB activation via phosphorylation of Ser-133, and this event is regulated by many kinases, such as PKA, CaMKIV, or MSK, as a function of the activation context. CREB has many coactivators that potentiate its transcriptional activity. The phosphorylation of Ser-133 in the KID domain of CREB facilitates the binding of the KIX domain of CBP/p300 to KID, thereby stabilizing the recruitment of RNA polymerase II complex and allowing transcriptional initiation. CRTC is another important coactivator and interacting protein of CREB. CREB plays a key role in determining cellular proliferation in the developing nervous and immune systems, while also controlling protein synthesis-dependent forms of long-term memory in the mature brain. Furthermore, CREB is essential in the control of hepatic gluconeogenesis, insulin resistance in obesity, and also in oncogenesis.
- Iourgenko V, Zhang W, Mickanin C, Daly I, Jiang C, Hexham JM, Orth AP, Miraglia L, Meltzer J, Garza D, Chirn GW, McWhinnie E, Cohen D, Skelton J, Terry R, Yu Y, Bodian D, Buxton FP, Zhu J, Song C, Labow MA. Identification of a family of cAMP response element-binding protein coactivators by genome-scale functional analysis in mammalian cells. Proc Natl Acad Sci U S A. 2003;100:12147–52.CrossRefPubMedPubMedCentralGoogle Scholar
- Zhang X, Odom DT, Koo SH, Conkright MD, Canettieri G, Best J, Chen H, Jenner R, Herbolsheimer E, Jacobsen E, Kadam S, Ecker JR, Emerson B, Hogenesch JB, Unterman T, Young RA, Montminy M. Genome-wide analysis of cAMP-response element binding protein occupancy, phosphorylation, and target gene activation in human tissues. Proc Natl Acad Sci U S A. 2005;102:4459–64.CrossRefPubMedPubMedCentralGoogle Scholar