Calcineurin controls gene transcription following stimulation of a Gαq- coupled designer receptor
Abstract
Stimulation of Gaq-coupled receptors triggers the activation of gene transcription via a rise of intracellular Ca2+. To investigate the role of the Ca2+/calmodulin-dependent phosphatase calcineurin in regulating transcription following Gαq-coupled receptor stimulation, we used a gain-of-function approach and expressed ΔCnA, a constitutively active mutant of calcineurin A. Furthermore, we expressed hM3Dq, a designer receptor that is specifically coupled to Gαq and can be activated by the pharmacological compound clozapine-N-oxide. Stimulation of hM3Dq or expression of ΔCnA induced transcription of a reporter gene controlled by the calcineurin substrate nuclear factor of activated T cells (NFAT), suggesting that calcineurin increased NFAT-regulated gene tran- scription. In contrast, expression of ΔCnA attenuated hM3Dq-induced biosynthesis of the transcription factors c- Fos and Egr-1 and reduced both c-Fos and Egr-1 promoter activities. A dissection of the c-Fos and Egr-1 pro- moters revealed that calcineurin inhibited serum response element-mediated transcription. In particular, the expression of ΔCnA reduced the transcriptional activity of the ternary complex factor Elk-1 following stimulation of hM3Dq receptors. Furthermore, ΔCnA reduced the transcriptional activity of the transcription factor CREB and thus attenuated transcription mediated by the cAMP response element. In summary, we show that calci- neurin functions as a positive and negative modulator of gene transcription.
1. Introduction
Calcineurin, also known as protein phosphatase 2B, is a Ca2+/cal- modulin-regulated serine/threonine phosphatase that is composed of two polypeptides, the catalytic calcineurin A-subunit and the regulatory B-subunit. Calcineurin A has a catalytical domain, binding sites for calmodulin and calcineurin B, and an autoinhibitory domain. Calcineurin B is tightly bound to calcineurin A and is structurally re- lated to calmodulin, having four Ca2+ binding EF-hands. The activation of calcineurin requires the binding of Ca2+ to calcineurin B. In addition, Ca2+ binding to calmodulin triggers the interaction of the Ca2+/cal- modulin complex with calcineurin A. The autoinhibitory domain of calcineurin A blocks the catalytic domain in the absence of Ca2+. Binding of Ca2+ to calcineurin B and interaction of the Ca2+/calmo- dulin complex with calcineurin A abrogate this autoinhibition [1].
Calcineurin has been described to function as a regulator of gene transcription via dephosphorylation of phosphorylated transcription factors or signaling proteins that are essential for transcription factor activation. Proteins of the NFAT family of transcription factors are substrates for calcineurin. The cytoplasmic subunits of NFATc tran- scription factors translocate into the nucleus upon calcineurin-catalyzed dephosphorylation, bind to their cognate DNA binding sites, and sti- mulate transcription of NFAT-regulated genes together with nuclear partner proteins [2]. In contrast to the stimulation of NFAT-regulated gene transcription, we have described that calcineurin negatively regulates transcription mediated by the transcription factor AP-1 [3–5]. We have shown in previous studies that stimulation of Gαq-coupled M3 acetylcholine receptors, gonadotropin-releasing hormone receptors, AT1 angiotensin II receptors, Ca2+ sensing receptors, and Gαq-coupled designer receptors induces the activation of stimulus-responsive tran- scription factors, including AP-1, c-Jun, Egr-1, Elk-1 and CREB [3,6–9]. Thus, the regulation of gene transcription is an integral part of Gαq- coupled receptor signaling. Stimulation of a Gαq-coupled receptor triggers an activation of phospholipase C, and the generation of IP3 and diacylglycerol. IP3 binds to the ionotropic IP3 receptor of the en- doplasmic recticulum and promotes an influx of Ca2+ ions into the cytosol and a rise in cytosolic Ca2+. As a result, the Ca2+/calmodulin- regulated phosphatase calcineurin is activated.
Here, we studied the impact of calcineurin in regulating gene transcription following stimulation of a Gαq-coupled receptor. We ex- pressed a Gαq-coupled designer receptor termed hM3Dq in HEK293 cells. This receptor is a mutated type 3 muscarinic acetylcholine receptor that is not more responsive to acetylcholine, but can be efficiently activated by the pharmacological compound clozapine-N- oxide (CNO) [10]. The signaling pathway of G protein-coupled re- ceptors is complex, because a particular ligand may bind and activate several distinct G protein-coupled receptors that differ in their binding preferences for particular G proteins. In addition, some G protein-cou- pled receptors have been reported to couple to more than one G-protein. The use of a Gαq-coupled designer receptors ensured that the
signaling pathway mediated by Gαq was specifically activated in our experiments. A comparison of the signaling cascade induced by either
CNO-treated hM3Dq designer receptors or acetylcholine-stimulated type 3 muscarinic acetylcholine receptor revealed that the functional consequences were very similar, including receptor-induced Ca2+ mo- bilization and activation of extracellular signal-regulated protein ki- nase, receptor internalization, and receptor phosphorylation [11,12].
As intracellular targets, we analyzed the expression of the tran- scription factors c-Fos and Egr-1, because both the c-Fos and Egr-1 genes are controlled by several stimulus-induced transcription factors. We asked whether calcineurin influences gene transcription following stimulation of a Gαq-coupled designer receptor. For the experiments, we used a gain-of-function approach by expressing a constitutively
active mutant of calcineurin A (ΔCnA) that lacked the C-terminal au- toinhibitory domain and an intact calmodulin binding site. ΔCnA functions independently of the cytosolic Ca2+ concentration. The results show that calcineurin either positively or negatively regulates gene transcription. Expression of ΔCnA upregulated transcription of a NFAT-responsive reporter gene. In contrast, the biosynthesis and ac-
tivities of Egr-1 and c-Fos were negatively influenced by the expression of ΔCnA. Likewise, expression of ΔCnA reduced the transcriptional activity of the transcription factors Elk-1 and CREB.
2. Materials and methods
2.1. Cell culture and reagents
HEK293 cells were cultured as described [8,9]. HEK293 cells ex- pressing hM3Dq were incubated for 24 h in DMEM containing 0.05% fetal bovine serum. Stimulation with clozapine-N-oxide (Enzo Life sci- ences, # NS-105-0005) was performed in medium containing 0.05% fetal bovine serum. CNO was dissolved in ethanol, and used at a final concentration of 1 μM.
2.2. Lentiviral gene transfer
The lentiviral transfer vectors pFUW-REST-Elk-1ΔC, pFUW-hM3Dq, pFUW-ΔCnA, pFUW-FLAG-C2/CREB, pFUW-GAL4-Elk-1, pFUW-GAL4-c-Fos, pFUW-GAL4-CREB, pFUW-REST/CREB have been described elsewhere [3,6,7,9,13–16]. Viral particles were produced as previously described [17,18].
2.3. Reporter assays
The lentiviral transfer vectors pFWEgr-1.2. luc, pFWEgr-1SRE.luc, pFWEBS24. luc, pFWColl.luc, pFWhc-Fos. luc, pFWmc-Fos. luc, pmc- Fos. lucΔEts, pFWmc-Fos. lucΔCArG, pFWc-FosCRE4. luc, pFW9E3/ cCAF.luc and pFWUAS5Sp12. luc have been described elsewhere [3,5,7,14,15,19–21]. Plasmid (NFAT)3TK.luc was a kind gift of Edgar Serfling, Würzburg University, Germany. The promoter region containing three copies of the NFAT binding site derived from the IL-2 promoter upstream of a minimal TK promoter was cloned into a lentiviral transfer vector upstream of the luciferase reporter gene. The plasmid was termed pFW(NFAT)3TK.luc. Infected cells were maintained in medium containing 0.05% fetal bovine serum for 24 h and then sti- mulated with CNO [1 μM] for 24 h. Cell extracts were prepared using reporter lysis buffer (Promega, Mannheim, Germany) and analyzed for luciferase activities. Luciferase activity was normalized to the protein concentration.
2.4. Western blots
Nuclear extracts were prepared as described [22]. 30 μg of nuclear proteins were separated by SDS-PAGE and the blots were incubated with antibodies directed against Egr-1 (Santa Cruz, Heidelberg, Germany, # sc-189), c-Fos (Santa Cruz, Heidelberg, Germany, # sc-52) or HDAC1 (Santa Cruz, Heidelberg, Germany, # sc-81598). Im- munoreactive bands were detected via enhanced chemiluminescence as described [23,24].
3. Results
3.1. Stimulation of a Gαq-coupled designer receptor or expression of a constitutively active mutant of calcineurin activates NFAT-responsive gene transcription Phosphorylated NFAT transcription factors are localized in the cy- tosol. Upon dephosphorylation by the Ca2+/calmodulin-regulated protein phosphatase calcineurin, NFAT translocates into the nucleus and activates transcription via binding to NFAT-responsive elements [2,25]. To assess whether stimulation of Gαq-coupled designer re- ceptors (hM3Dq) activates calcineurin, we measured NFAT-regulated gene transcription in HEK293 cells expressing hM3Dq receptors. We used a chromosomally embedded NFAT-responsive luciferase reporter gene (Fig. 1A) to ensure that the reporter gene was included in the nucleosomal architecture of the chromatin. The results show that sti- mulation of HEK293 cells expressing hM3Dq with CNO significantly increased transcription of the NFAT-responsive reporter gene (Fig. 1B), indicating that calcineurin-mediated dephosphorylation of NFAT occured as a response to hM3Dq stimulation. To confirm these data, we expressed a constitutively active mutant of calcineurin A (ΔCnA) in HEK293 cells harbouring an NFAT-responsive luciferase reporter gene. The domain structures of wild type and mutant calcineurin A are de- picted in Fig. 1C showing that the calcineurin A mutant lacked the C-terminal autoinhibitory domain and an intact calmodulin binding site. ΔCnA does not require Ca2+ ions for its activation. Expression of ΔCnA led to a significant increase in NFAT-regulated transcription (Fig. 1D).
3.2. Calcineurin increases phosphorylation of extracellular signal-regulated protein kinase (ERK1/2) following stimulation of a Gαq-coupled designer receptor
Calcineurin has been described to increase the phosphorylation of the protein kinase ERK1/2 by reversing a negative feedback [26]. We measured the phosphorylation status of ERK1/2 in the presence or absence of the constitutively active calcineurin mutant ΔCnA in order to corroborate these results and to support the view that stimulation of a Gαq-coupled designer receptor activates calcineurin. Supplemental Fig. S1 shows that stimulation of hM3Dq receptors with CNO triggered the phosphorylation of ERK1/2. Phosphorylation of ERK1/2 was increased in the presence of ΔCnA.
3.3. Expression of ΔCnA interferes with c-Fos expression following stimulation of hM3Dq
Stimulation of the designer receptor hM3Dq induces the biosynth- esis of c-Fos [9]. Fig. 2A shows that expression of ΔCnA significantly reduced the stimulus-induced biosynthesis of c-Fos. Likewise, ΔCnA expression interfered with the upregulation of c-Fos promoter activity (Fig. 2B and C). Moreover, the transcriptional activation potential of c- Fos, measured with a GAL4-c-Fos fusion protein and a GAL4-responsive promoter (Fig. 2D and E) was significantly reduced in the presence of ΔCnA (Fig. 2F). These data indicate that calcineurin functions as a negative regulator of hM3Dq-induced c-Fos biosynthesis and function.
3.4. Major role of SRF in hM3Dq-induced activation of the c-Fos promoter
A landmark genetic element of the c-Fos promoter is the serum re- sponse element (SRE) that provides binding sites for a SRF dimer and a ternary complex factor (TCF), e.g. Elk-1. The SRF dimer binds to the CArG box (CC[A/T]6GG), while TCF interacts with the adjacent Ets consensus core sequence GGAA/T (Fig. 3A). Mutational inactivation of either the CArG box or the Ets consensus site revealed that the CArG box is important to induce c-Fos gene transcription following stimula- tion of hM3Dq receptors. Mutations within the ternary complex factor binding site did not significantly reduce the fold-stimulation of the c- Fos promoter in HEK293 cells that expressed CNO-activated hM3Dq designer receptors (Fig. 3B).
3.5. Role of Elk-1 in hM3Dq-induced activation of the c-Fos promoter
The previous results suggest that TCFs play no role in the coupling of hM3Dq designer receptor stimulation with the activation of c-Fos gene transcription. However, TCF proteins require the presence of SRF within the SRE, which is not the case for the ΔCArG mutant. Moreover, expression of a dominant-negative mutant of Elk-1, termed REST/Elk- 1ΔC (Fig. 3C), significantly reduced basal as well as stimulated c-Fos promoter activity (Fig. 3D), confirming previous observations [9].
These data suggest that Elk-1 is involved in the regulation of c-Fos gene transcription in HEK293 cells expressing activated Gαq-coupled de- signer receptors. In a previous study we had shown that stimulation of hM3Dq designer receptors strongly increased the transcriptional acti- vation potential of Elk-1, as measured with a GAL4-Elk-1 fusion protein (Fig. 3E [9]). As Elk-1 has been described as a major substrate for calcineurin [27–29], we measured the outcome of ΔCnA expression on Elk-1 activity. Fig. 3F shows that ΔCnA significantly reduced Elk-1 activity in HEK293 cells that expressed CNO-activated hM3Dq designer receptors. Thus, Elk-1 is necessary, but not an intact DNA binding site for Elk-1.
3.6. Expression of ΔCnA interferes with the biosynthesis of Egr-1 following stimulation of hM3Dq designer receptors
To confirm the previous data we analyzed the expression of Egr-1, encoding a zinc finger transcription factor. Egr-1 expression is mainly regulated by TCF and SRF transcription factors that bind to multiple sites within the Egr-1 regulatory region. In a previous study, we showed that stimulation of hM3Dq designer receptors induced the biosynthesis of Egr-1 and increased the transcriptional activity of Egr-1 [9]. Fig. 4A shows that the upregulation of Egr-1 biosynthesis, induced by the sti- mulation of hM3Dq receptors, was attenuated in the presence of the constitutively active mutant of calcineurin A (ΔCnA). Next, we analyzed the transcriptional activity of Egr-1, using an Egr-1-responsive reporter gene termed EBS24. luc (Fig. 4B). ΔCnA expression significantly re- duced the cellular Egr-1 activity following stimulation of hM3Dq de- signer receptors (Fig. 4C). Expression experiments involving the dominant-negative mutant of Elk-1 (REST/Elk-1ΔC) revealed that Elk-1 is required to connect designer receptor stimulation with the increase in Egr-1 activity (Fig. 4D). In a previous study we showed that stimulation of hM3Dq receptors triggers the activation of the Egr-1 promoter. We repeated these experiments in the presence of the constitutively active mutant of calcineurin (ΔCnA). Fig. 4E schematically shows the chro- matin-embedded Egr-1 promoter/luciferase reporter genes used in the experiments. The results, depicted in Fig. 4F show that the activation of the Egr-1 promoter following hM3Dq receptor stimulation is sig- nificantly reduced in the presence of ΔCnA. Moreover, the analysis of a reporter gene that was solely controlled by the proximal SREs derived from the Egr-1 promoter (Fig. 4E) revealed that the activation of SRE- dependent transcription is impaired in the presence of ΔCnA (Fig. 4G).
3.7. Calcineurin blocks transcription of an Elk-1-regulated reporter gene following stimulation of a Gαq-coupled designer receptor
The previous results showed that calcineurin attenuated transcrip- tion of the c-Fos and Egr-1 genes following stimulation of hM3Dq de- signer receptors. Moreover, they showed that the SRE is important for the stimulus-induced biosynthesis of c-Fos and Egr-1, indicating that the SRE is the target for the calcineurin action. Additionally, it had been shown that stimulation of hM3Dq designer receptors activates tran- scription of the 9E3/cCAF gene that has two binding sites for Elk-1 (Fig. 5A) [30,31], but does not contain a binding site for SRF. Fig. 5B shows that expression of the calcineurin A mutant reduced 9E3/cCAF promoter activity in HEK293 cells expressing CNO-activated hM3Dq
designer receptors. Together, we conclude that calcineurin functions as a negative regulator of SRE and Elk-1-regulated gene transcription.
3.8. Calcineurin interferes with cAMP response element (CRE)-regulated transcription and the upregulation of the transcriptional activation potential of CREB in HEK293 cells expressing an activated Gαq-coupled designer receptor
Several reports described that calcineurin regulates CREB-mediated transcription either positively or negatively [32,33]. We used a chro- matin-embedded CREB-responsive reporter gene (c-FosCRE4. luc, Fig. 6A) to investigate the role of calcineurin in CRE-regulated gene transcription. Stimulation of hM3Dq designer receptors activated transcription of the CRE-responsive reporter gene (Fig. 6B, left panel),
confirming previous results [9]. Reporter gene transcription was sig- nificantly reduced in the presence of ΔCnA (Fig. 6B, right panel). Cal- cineurin couples elevated intracellular Ca2+ concentrations with the dephosphorylation of substrate proteins. Therefore, as a control, we studied transcription of the c-FosCRE4. luc reporter gene induced by the expression of C2/CREB, a phosphorylation-independent mutant of CREB. The CREB mutant contains the transcriptional activation domain of CREB2 instead of its native phosphorylation-dependent activation domain (Fig. 6C). C2/CREB is therefore constitutively active. Fig. 6D (left panel) shows that expression of C2/CREB stimulates transcription of the CREB-responsive reporter gene c-FosCRE4. luc. However, ex- pression of ΔCnA did not reduce reporter gene transcription (Fig. 6D, right panel).
Next, we asked whether calcineurin interferes with the biological activity of CREB. We used a GAL4-CREB fusion protein to measure the transcriptional activation potential of CREB (Fig. 6E). Stimulation of Gαq-coupled designer receptors increased the transcriptional activity of CREB (Fig. 6F, left panel), confirming previous results [9]. Expression of ΔCnA significantly reduced the transcriptional activation potential of CREB in the presence of activated hM3Dq designer receptors (Fig. 6F, right panel). We conclude that calcineurin functions as a negative regulator of CREB and CREB-regulated gene transcription.
3.9. Stimulation of the c-Fos promoter via hM3Dq activation is independent of the presence of the CRE and the activation of CREB
Stimulation of hM3Dq designer receptors requires the SRE within the c-Fos and Egr-1 genes to activate c-Fos and Egr-1 expression. The c- Fos promoter contains a CRE that mediates cAMP and Ca2+-regulated expression of c-Fos [21,34]. Given the fact that stimulation of Gαq- coupled designer receptors activates CREB and increases CRE-regulated transcription [9], we asked whether the CRE within the context of the c- Fos promoter plays a role in connecting designer receptor stimulation with c-Fos gene transcription. A proximal portion of the human c-Fos promoter including the CRE was deleted from the c-Fos promoter (Fig. 7A). A reporter gene under the control of this mutated c-Fos promoter (hc-FosTK.luc) was shown to be no longer responsive to ele- vated cAMP levels in the cells [21]. HEK293 cells containing a reporter gene under the control of either the wild-type or the mutated c-Fos promoter were infected with a lentivirus expressing hM3Dq. The re- sults, depicted in Fig. 7B, show that the deletion of the CRE within the c-Fos promoter did not reduce c-Fos promoter-controlled reporter gene transcription following activation of hM3Dq with CNO. Likewise, ex- pression of a dominant-negative mutant of CREB (REST/CREB, Fig. 7C) did not reduce c-Fos promoter activity following stimulation of hM3Dq designer receptors. Together, we conclude that calcineurin controls CREB-regulated transcription, but CREB and the CRE within the c-Fos gene do not contribute to the upregulation of c-Fos gene transcription as a result of designer receptor activation.
4. Discussion
Stimulation of Gαq-coupled receptors triggers a signaling cascade leading to changes in the gene expression pattern of the cells [3,6–9]. The signaling cascade induced by stimulation of Gαq-coupled receptors involves the activation of phospholipase C, the generation of IP3, and an influx of Ca2+ ions from the endoplasmic reticulum into the cytosol, mediated by the stimulation of IP3 receptors. The rise in cytosolic Ca2+ activates the Ca2+/calmodulin-dependent phosphatase calcineurin as measured by the induction of NFAT-regulated gene transcription [35,36]. Calcineurin is known to be involved in the regulation of gene transcription via dephosphorylation of transcription factors. The best studied physiological substrate of calcineurin is NFAT, a transcription factor that translocates from the cytoplasm to the nucleus upon de- phosphorylation and triggers transcription of NFAT-responsive genes [2,25]. The ternary complex factor (TCF) Elk-1 that binds to a dimer of the serum-response factor (SRF) and forms a ternary complex mediating serum response element (SRE)-regulated gene transcription has also been identified as a substrate of calcineurin [27–29].
We asked whether calcineurin has an impact on gene transcription following stimulation of Gαq-coupled designer receptors. Many studies have been performed with the immunosuppressive drugs tacrolimus (FK506) and cyclosporin A, used as calcineurin inhibitors that both
trigger a loss-of-function phenotype. Tacrolimus and cyclosporin A function as complexes with the immunophilins cyclophilin A and FKBP12 [37]. However, neither compound is calcineurin-specific and other activities have been reported. The immunophilins are cis-trans peptidyl-prolyl isomerases that regulate many biological functions, in- cluding protein folding, intracellular Ca2+ release, and assembly and
activity of ion channels [38–40]. Cyclosporin A inhibits mitochondrial Ca2+ uptake and blocks the mitochondrial permeability transition pore [41,42] and modulates innate immunity [43]. We therefore decided to use a gain-of-function approach by expressing a constitutively active mutant of calcineurin A (ΔCnA) that lacked the C-terminal auto- inhibitory domain and the calmodulin binding site. ΔCnA did not
require Ca2+ ions for activation. Transgenic mice have been generated expressing the constitutivly active mutant of calcineurin A to in- vestigate the role of calcineurin in vivo [44,45].
As transcriptional targets for analyzing hM3Dq receptor-induced gene transcription we analyzed the genes encoding the transcription factors c-Fos and Egr-1. Both genes are highly responsive to in- tracellular signaling cascades and are termed immediate-early response genes. While Egr-1 gene transcription is mainly controlled by multiple SREs, the c-Fos gene is regulated by multiple transcription factors, in- cluding CREB, AP-1, SRF, and TCF. Expression of the constitutively active mutant of calcineurin attenuated both c-Fos and Egr-1 expression following stimulation of hM3Dq designer receptors. Likewise, the ac- tivities of the c-Fos and Egr-1 promoters were significantly lower in the presence of ΔCnA. In line with this, it has been shown that pharmacological inhibition of calcineurin increased Egr-1 and c-Fos mRNA
concentrations in murine erythroleukaemia cells [46], supporting our view that calcineurin has a negative effect on c-Fos and Egr-1 expres- sion. Likewise, expression of a constitutively active mutant of calci- neurin A was shown to inhibit KCl-induced c-Fos expression in neurons [33]. A molecular dissection of transcription factor binding sites within the c-Fos gene revealed that the binding site for SRF is important to couple designer receptor stimulation with an induction of c-Fos ex- pression. In contrast, mutation of the adjacent binding site for TCF had no impact on stimulation of c-Fos expression as a result of hM3Dq de- signer receptor activation. These data are in agreement with earlier observations [47]. However, according to our knowledge, SRF has not been described as a substrate for calcineurin.
In contrast, the TCF protein Elk-1 has been identified as a major substrate for calcineurin [27–29]. Expression of a dominant-negative mutant of Elk-1 inhibited Egr-1 and c-Fos expression and significantly reduced Egr-1 and c-Fos promoter activities following Gαq-coupled designer receptor activation [9], indicating that Elk-1 plays a major role in connecting designer receptor stimulation with c-Fos and Egr-1 expression. Interestingly, expression of the dominant-negative mutant of Elk-1 and expression of the constitutively active mutant of calcineurin showed similar results concerning the regulation of c-Fos and Egr-1. In this context, it is sur- prising that pharmacological inhibition of calcineurin stimulated Egr-1 expression in neurons whereas c-Fos expression remained unchanged [33], although calcineurin targets the SREs in both genes. These puz- zeling results may reflect the problems associated with a pharmacolo- gical loss-of-function strategy. To further support our hypothesis con- cerning the major role of Elk-1 as a target for calcineurin, we showed that expression of the constitutively active mutant of calcineurin A at- tenuated the activation of Elk-1 in cells expressing CNO-actived hM3Dq designer receptors. These data fit very well with the observation that a constitutively active mutant of calcineurin blocked the upregulation of the transcriptional activation potential of Elk-1 mediated by the protein kinase Raf-1 [28]. Moreover, we have shown here that expression of the calcineurin mutant ΔCnA interfered with the designer receptor-induced upregulation of 9E3/cCAF promoter activity. The 9E3/cCAF gene is
regulated by Elk-1, but is independent of SRF. Thus, we think that Elk-1 is recruited to the c-Fos promoter via protein-protein interaction with SRF, as described for the GnRH-induced activation of c-Fos expression in gonadotropes [48], the neurokinin B-stimulated c-Fos transcription in immortalized GnRH neurons [49], and the pregnenolone sulfate/ TRPM3-induced c-Fos expression in HEK293 cells [21]. This scenario explains why the Elk-1 mutant REST/Elk-1ΔC reduces stimulus-induced c-Fos expression, although the binding of Elk-1 to DNA is dispensable in this scenario.
Calcineurin has been described as a positive or negative regulator of CREB-mediated gene transcription. Based on pharmacological inhibi- tion of calcineurin it has been proposed that calcineurin functions as a positive regulator for CREB-induced transcription [33,50]. Surprisingly, pharmacological inhibition of calcineurin not only attenuated KCl-in- duced transcription of a CREB-responsive reporter gene, but also in- terfered with forskolin and Br-cAMP-stimulated reporter gene tran- scription [50] which activates CREB in a Ca2+-independent manner. Calcineurin has also been described as a negative regulator of CREB via dephosphorylation of inhibitor-1 and the subsequent activation of protein phosphatase-1 [32]. Using a chromatin-embedded CREB-re- sponsive reporter gene having four copies of the c-Fos CRE in its reg- ulatory region, we showed here that calcineurin negatively regulates CRE-mediated transcription. These data were corroborated by experi- ments showing that the transcriptional activation potential of CREB was significantly reduced in the presence of a constitutively active mutant of calcineurin A. Calcineurin had no effect on the transcriptional ac- tivity of C2/CREB, a CREB mutant that functions as an activator of CRE- mediated transcription independent of phosphorylation. Deletion mu- tagenesis as well as expression of a dominant-negative mutant of CREB revealed that CREB and the CREB binding site CRE play no role in hM3Dq receptor-induced activation of c-Fos. Nevertheless, transcrip- tion of a CREB-responsive reporter gene was activated following sti- mulation of Gαq-coupled designer receptors. These observations in- dicate that the promoter context is important for the transcriptional activation via CRE/CREB.
In summary, we have shown that calcineurin functions as a positive and negative modulator of gene transcription. While expression of ΔCnA, a constitutively active mutant of calcineurin, stimulated NFAT- controlled gene transcription, the expression and/or transcriptional
activities of the transcriptional factors Egr-1, Elk-1, c-Fos and CREB were reduced in the presence of ΔCnA. The important genetic element for influencing Egr-1 and c-Fos expression by calcineurin is the SRE. The binding of SRF is necessary for recruiting a TCF protein to the SRE. The TCF protein Elk-1 has been described as a substrate for calcineurin.
In this study, we showed that calcineurin controls the biological activity of Elk-1 that functions within the SRE but also in genes solely regulated by Elk-1. Calcineurin also functions as a negative regulator of CREB, regulating CREB activity and CREB-mediated transcriptional induction. However, the transcriptional regulation via CREB depends on the pro- moter context.
Stimulus-responsive transcription factors such as Egr-1, Elk-1, c-Fos, and CREB are quickly and transiently activated by different means. Egr- 1 and c-Fos are activated via the induction of the biosynthesis of these proteins. Elk-1 and CREB are activated via phosphorylation. Shut-off devices reduce the transcriptional activity of stimulus-responsive tran- scription factors. The data presented here identified calcineurin as an important negative Clozapine N-oxide regulator of Egr-1, Elk-1, c-Fos and CREB.