In addition to the well-known function of ACTH as the main regulator of adrenal steroidogenesis, we have previously demonstrated its effect on the transcriptional stimulation of HO-1 expression, a component of the cellular antioxidant defense system. In agreement, we hereby demonstrate that, in adrenocortical Y1 cells, HO-1 induction correlates with a significant prevention of the generation of reactive oxygen species induced by H2O2/Fe2+. ACTH/cAMP-dependent activation of redox-imbalanced related factors such as NRF2 or NFκB and the participation of MAPKs in this mechanism was, however, discarded based on results with specific inhibitors and reporter plasmids. We suggest the involvement of CREB in HO-1 induction by ACTH/cAMP, as transfection of cells with a dominant-negative isoform of CREB (DN-CREB-M1) decreased, while overexpression of CREB increased HO-1 protein levels. Sequence screening of the murine HO-1 promoter revealed CRE-like sites located at −146 and −37 of the transcription start site and ChIP studies indicated that this region recruits phosphorylated CREB (pCREB) upon cAMP stimulation in Y1 cells. In agreement, H89 (PKA inhibitor) or cotransfection with DN-CREB-M1 prevented the 8Br-cAMP-dependent increase in luciferase activity in cells transfected with pHO-1[−295/+74].LUC. ACTH and cAMP treatment induced the activation of the PI3K/Akt signaling pathway in a PKA-independent mechanism. Inhibition of this pathway prevented the cAMP-dependent increase in HO-1 protein levels and luciferase activity in cells transfected with pHO-1[−295/+74].LUC. Finally, here we show a crosstalk between the cAMP/PKA and PI3K pathways that affects the binding of p-CREB to its cognate element in the murine promoter of the Hmox1 gene.
ACTH is the main regulator of glucocorticoid synthesis in the adrenal cortex where it also affects other biological functions such as differentiation, mitogenesis and apoptosis (Gallo-Payet & Payet 2003). Many of these effects are mediated by ACTH binding to the MC2R receptor and downstream activation of the cAMP/PKA pathway (Gallo-Payet 2016). In turn, stimulation of PKA activity is involved in the phosphorylation and activation of several transcription factors engaged in the transcriptional stimulation of steroidogenic-related genes. Among them, the cAMP response element binding protein (CREB), the GATA-4 binding protein, the steroidogenic factor 1 (SF1) and the CAAT enhancer binding protein (C/EBP) could be included (Manna et al. 2003a, b, Stocco et al. 2005).
Glucocorticoid synthesis in the adrenal cortex is a multistep process where mitochondrial and microsomal cytochrome P450 enzymes, mixed function oxidases, catalyze the hydroxylation of steroids and the reduction of O2 to water. Even under physiological conditions, ACTH-stimulated steroidogenesis generates oxidative stress, as incomplete reduction in O2 leads to the production of reactive oxygen species (ROS), lipoperoxides and other oxygenated species (Hanukoglu 2006, Prasad et al. 2014). To cope with the potentially damaging effects associated with oxidative stress, adrenal cells activate several antioxidant mechanisms (Chinn et al. 2002, Hanukoglu 2006, Battista et al. 2009). Among them, we have previously shown that the ACTH-dependent induction of heme oxygenase 1 (HO-1) in adrenocortical cells correlated with the prevention of the generation of lipoperoxides and protein carbonyls triggered by H2O2 treatment (Pomeraniec et al. 2004).
Heme oxygenases catalyze the rate-limiting step in heme catabolism, the conversion of heme to biliverdin (and then bilirubin) with the concomitant release of iron and CO. In addition to this well-known role in heme catabolism, the activity of the inducible isoform of HO has been recognized as a cellular defense mechanism activated by stressful signals (Elbirt & Bonkovsky 1999, Ryter & Tyrrell 2000, Bach 2002). Heme oxygenase activity has been also involved in the modulation of steroid production. By attenuating the hormone-induced glucocorticoid response, heme oxygenase activity has been recently included among the modulatory systems involved in the fine-tuning regulation of adrenal steroidogenesis (Pomeraniec et al. 2004, Grion et al. 2007, Astort et al. 2009).
The induction of HO-1 expression is primarily controlled at the transcriptional level (Maines 1988, Ryter & Choi 2002, Alam & Cook 2003). Both mouse and human HO-1 promoters are subject to a complex regulation that involves multiple transcription factors, including NRF2 (nuclear factor-erythroid 2-related factor 2), heat-shock factor proteins, NF-κB and members of the AP-1(activating protein-1) bZIP(basic region/leucine zipper) families, with response elements often embedded in antioxidant response elements (AREs) (Ryter & Choi 2005, Surh et al. 2008). Activation of the NRF2 transcription factor, as a consequence of cellular redox imbalance, has been linked to the induction of HO-1 and other cytoprotective genes. In basal conditions, NRF2 is maintained in the cytoplasm by interacting with Klech-like ECH-associated protein 1 (KEAP 1) and targeted to proteosomal degradation by ubiquitin modification. Upon stimulation, NRF2 dissociates from KEAP 1 in the cytosol and translocates to the nucleus where it promotes the induction of antioxidant genes (Kensler et al. 2007). Activation of the HO-1 promoter has also been linked to the PKA-dependent phosphorylation of CREB (Lonze & Ginty 2002, Kronke et al. 2003, Wright et al. 2009). CREB activation has been shown to play a role in regulating ROS detoxification as a direct modulator of antioxidant gene expression (Bedogni et al. 2003, Kronke et al. 2003).
Although we have previously shown that induction of HO-1 by ACTH involves the activation of the cAMP/PKA pathway (Pomeraniec et al. 2004), the underlying molecular mechanism has remained elusive. In the present report, we analyzed the involvement of PKA and CREB activation in the transcriptional stimulation of HO-1 by ACTH, and identified a CREB-responsive binding region in the murine Hmox1 gene proximal promoter. CREB binding to this region is also affected by activation of the PI3K/AKT pathway.
Materials and methods
The antibody raised against HO-1 was acquired from StressGen Biotechnologies Corp. (Victoria, Canada, and those for pAkt and total Akt were bought from New England Biolabs (Ipswich, MA, USA). CREB and Phospho-CREB (Ser 133) antibodies were from Santa Cruz Biotechnology and Cell Signaling Technology respectively. Actin and peroxidase-conjugated goat antirabbit-IgG antibodies were obtained from Santa Cruz Biotechnology and Bio-Rad, respectively. Fetal calf serum, penicillin, streptomycin, MMVL reverse transcriptase, DNAse I and TAQ polymerase were from Invitrogen (Life Technologies), while 8Br-cAMP was purchased from Sigma. ACTH was obtained from ELEA Laboratories (Buenos Aires, Argentina). All other chemicals were of the highest quality available.
Cell culture and treatments
The adrenal Y1 cell line (CCL-79, American Type Culture Collection) is an ACTH- and cAMP-responsive mouse adrenocortical tumor cell line established by Yasumura and coworkers (1966). Cells were grown in monolayers in Ham’s F10 medium containing heat-inactivated fetal bovine (2.5%) and horse (12.5%) serum, 200U/mL penicillin G and 270μg/mL streptomycin sulfate, in a humidified atmosphere of 5% CO2 in air at 37°C as described previously (Schimmer 1979). Cells were washed twice with PBS and medium was replaced with serum-free Ham’s F10 2h before the initiation of the experiments. Cell viability was determined by trypan blue dye exclusion test.
Determination of intracellular ROS generation
ROS generation was measured using the nonfluorescent probe 2′,7′-dichlorodihydrofluorescein diacetate (H2DCF diacetate, Molecular Probes, Invitrogen) as described previously (Astort et al. 2014, Mercau et al. 2014). After treatments, cells were washed twice with PBS and then incubated in PBS containing 10µmol/L of 2ʹ,7ʹ-H2DCF for 1h at 37°C. Cells were washed twice with PBS. Upon uptake by the cells and hydrolysis by cellular esterases, H2DCF is rapidly oxidized by ROS to the fluorescent 2ʹ7ʹ-dichlorofluorescein. Fluorescence intensity (excitation 485nm, emission 535nm) was determined in a multimode microplate reader, FLUOstar Omega (BMGLabtech, Ortemberg, Germany). ROS generation levels are expressed in arbitrary units relative to total protein content.
The HO-1 luciferase reporter plasmid containing the (−4045/+74) fragment from the murine HO-1 promoter cloned into the basic pGL2 reporter vector [pHO-1(−4045;+74).LUC] was generously provided by Dr M Perrella (Pulmonary and Critical Care Division, Brigham and Women’s Hospital, Boston, MA, USA) (Chung et al. 2005). A pGL2 reporter plasmid containing only the (−295/+74) fragment from the murine HO-1 promoter was generated by PCR amplification using the pHO-1(−4045/+74).LUC construct as template [pHO-1(−295;+74).LUC]. The distal enhancer E1 from the murine HO-1 promoter was amplified by PCR (−4045/−3524, 523bp) using the pHO-1(−4045/+74).LUC construct as a template and cloned into a pGL3 promoter plasmid [pHO-1(E1).LUC] (Promega). The identity of reporter plasmids was confirmed by DNA sequencing (MACROGEN, Korea). Both p3xARE.LUC (pNRF2.LUC) reporter plasmid, pcDNA3.1/V5HisB-mNRF2 (NRF2 FL) (Innamorato et al. 2008, McMahon et al. 2004) and PGL3-NFκB-LUC (pκB-LUC), containing five NFκB response elements, were kindly provided by Dr O Coso (FCEyN, UBA, CABA, Argentina) (Squarize et al. 2006). An expression vector containing the entire wild-type coding region of CREB (RSV-CREB), the dominant-negative isoform (DN-CREB-M1) carrying a mutation in the phosphorylation site S133A obtained from RSV-CREB (Gonzalez & Montminy 1989) and a luciferase reporter for CREB p4xCRE.LUC (pCRE.LUC), were kindly supplied by Dr E Cánepa (FCEyN, UBA, CABA, Argentina) (Giono et al. 2001).
Transfection of Y1 cells and luciferase assay
For overexpression assays, 105 cells per well were seeded in 24-well plates. Twenty four hours later, 0.5µg of expression plasmids (RSV-CREB or DN-CREB-M1 or pGEX-4T-3 as control) were transiently transfected with Lipofectamine 2000 Transfection Reagent (Invitrogen) according to manufacturer’s instructions in serum-free OptiMEM medium; pCMV-βGAL (β-galactosidase expression plasmid) was included as control of transfection. Three hours later, transfection medium was aspirated and replaced with fresh Ham’s F10 medium with sera, and the cells were further incubated for additional 24h before any treatment.
For luciferase reporter assays, 104 cells were seeded on 96-well plates. One day after, cells were transiently transfected with luciferase 0.2µg of the following reporter plasmids: pHO-1(−4045/+74).LUC or pHO-1(−295;+74).LUC or pHO-1(E1).LUC or pNRF2(ARE).LUC or p4xCRE.LUC), and pCMV-βGAL in Lipofectamine 2000 according to manufacturer’s instructions in serum-free OptiMEM medium, and the following day were treated accordingly. In another set of experiments, cells were seeded in 96-well plates and transfected with 0.2µg of total DNA per well in a combination of a reporter plasmid, an expression plasmid (RSV-CREB or DN-CREB-M1 or pGEX-4T-3) and pCMV-βGAL (ratio 8:1:1).
Luciferase activity was determined in cell homogenates using the Steady-Glo Luciferase Assay System (Promega) and values were normalized to β-galactosidase activity. Luminescence was measured in a FLUOstar Omega plate luminometer (BMG LABTECH, Germany).
Y1 cells were washed twice in PBS and lysed in 20mM Tris–HCl pH 7.4, 250mM NaCl, 1% Triton X-100, 1mM Na3VO4, 100mM NaF, 1mM EGTA, with 1X proteases and phosphatase inhibitors cocktails (Sigma-Aldrich). After electrophoresis, proteins were transferred to polyvinylidinefluoride membranes for 1h at 15V in a Bio-Rad Trans-Blot SD system with 25mM Tris–HCl pH 9.2, 192mM glycine and 20% methanol. Membranes were blocked in TBS-Tween (50mM Tris–HCl pH 7.4, 0.15M NaCl, 0.05% Tween 20) with 5% skimmed milk for 60min at room temperature and then incubated overnight with the following antisera: HO-1, p-AKT, AKT, p-CREB, CREB or β-ACTIN at 4°C. Specific protein bands were visualized by chemiluminescence (Enhanced Chemiluminescence Reagent, Amersham Pharmacia Biotech, Little Chalfont, Bucks, UK) in an ImageQuant Imaging System (GE Healthcare) and quantified with the AlphaEase Fluorchem software (V. 4.1.0, Alpha Innotech Corporation, San Jose, CA, USA).
Chromatin immunoprecipitation (ChIP) assay
Chromatin immunoprecipitation assays were performed as described previously (Bertucci et al. 2013, Rocha-Viegas et al. 2014) with chromatin obtained from adrenocortical Y1 cells, using ChIPAb+Phospho-CREB (Ser133) from Millipore (EMD Millipore) and IgG (kch-504-250) from Diagenode (Denville, NJ, USA) as a negative control. Quantification of chromatin immunoprecipitation was performed by real-time PCR using DNA Engine OpticonTM System cycler (MJ Research) with the following primers specific for the HO-1 promoter region (−42;+30), forward: 5ʹ-ACGTGACCCGCGTACTTAAA-3ʹ; reverse: 5ʹ-AAACTCTGGAGGCGA CTGG-3ʹ and for the distal enhancer E1(−3979; −3843), forward: 5ʹ-GCTGGA ATGC TGAGTTGTGAT-3ʹ; reverse: 5ʹ-CTGAGGCTGAGG GAACAGAG-3ʹ. Chromatin immunoprecipitation signal was normalized to input and was expressed as percentage of input (mean±s.e.m., n=3).
Data are expressed as mean±s.e.m. Statistical analysis was performed by one-way ANOVA followed by Tukey’s post hoc t-test or Student’s t-test as considered necessary using GraphPad InStat version 3.06 for Windows (GraphPad Software). Data were considered significantly different when P<0.05.
Taking into consideration that HO-1 belongs to a family of antioxidant enzymes induced by the generation of a cellular redox imbalance and our previous results demonstrating the induction of HO-1 by ACTH in adrenocortical cells, we first evaluated the generation of reactive oxygen species (ROS) upon ACTH treatment. Results from a time course experiment, performed in murine adrenocortical Y1 cells incubated in the presence of ACTH, showed that ROS generation was significantly increased already 30min after hormone addition and this parameter remained elevated for at least 2h (Fig. 1A), confirming the ACTH-dependence of ROS generation. Moreover, when cells were preincubated with ACTH during 5h, in order to induce HO-1 expression, and then treated with H2O2/Fe2+, ROS generation was significantly lower compared with cells that were preincubated in the absence of hormone (Fig. 1B). These results prompted us to analyze the involvement of the Nrf2 pathway in the induction of HO-1 by ACTH. Thus, transient transfections with an NRF2 reporter, pNRF2.LUC, were performed in Y1 cells. Our results showed that while overexpression of NRF2 (NRF2 FL) significantly increased luciferase activity, neither ACTH nor a permeable analog of cAMP, 8Br-cAMP, treatments were effective (Fig. 1C), suggesting that this transcription factor was not activated by the cAMP pathway. In another set of experiments, we demonstrated that ACTH treatment of Y1 cells did not activate the NFκB pathway (Fig. 1D). Since sequence analysis of the murine HO-1 promoter indicated the presence of an NRF2 consensus binding motif located 4Kbp upstream the transcription start site (TSS), we also performed transient transfections with a reporter plasmid containing 4Kbp of the murine HO-1 promoter (−4045/+74). Results showed that stimulation of these cells with 8Br-cAMP or with ACTH significantly increased luciferase activity (Fig. 1E). However, neither 8Br-cAMP nor ACTH treatments were able to stimulate pHO-1(E1).LUC, a reporter plasmid containing the murine distal enhancer E1 and the Nrf2 consensus site (−4045/−3524, Fig. 1F) further supporting the idea that ACTH-dependent induction of HO-1 is not mediated by NRF2 activation. In addition, stimulation by 8Br-cAMP was conserved in a reporter plasmid containing only the proximal fragment of the HO-1 promoter (−295/+74) (Fig. 1G). 8Br-cAMP-stimulated luciferase activity of pHO-1(−295;+74).LUC was not affected by the addition of specific inhibitors for ERK, JNK or p38 MAPK signaling pathways (Fig. 1H).
It is well known that stimulation of the cAMP/PKA pathway has been associated with the phosphorylation and activation of CREB. Therefore, we sought to analyze the involvement of CREB on the induction of HO-1 in adrenal cells. To that end, we first evaluated the induction of HO-1 by cAMP in Y1 cells transfected with a plasmid expressing a dominant-negative form of CREB (DN-CREB-M1). Our results showed an inhibition of the 8Br-cAMP-induced increase in HO-1 protein levels in cells transfected with increasing levels of DN-CREB-M1 expression vector (Fig. 2A). Conversely, overexpression of CREB in cells transfected with increasing amounts of RSV-CREB significantly upregulated HO-1 protein levels (Fig. 2B). The effect of DN-CREB-M1 or RSV-CREB on the activity of a pCRE.LUC reporter in Y1 cells was also assayed (Fig. 2C and D).
In the search of cAMP-response elements (CRE) along 4Kbp of the murine HO-1 promoter, we performed an analysis of transcription factor binding motifs with the MatInspector Release 8.2 software (Genomatix Software Suite, V3.1, Munich, Germany). Results showed in Fig. 2E indicate two different putative CRE binding sites located at −146 and −37.
Therefore, we performed a ChIP analysis in order to assess CREB recruitment to the endogenous gene. Results showed that stimulation of Y1 cells with 8Br-cAMP significantly increased the recruitment of phospho-CREB to the proximal promoter region of the Hmox1 gene, while no significant binding was observed at the distal enhancer E1 (around 4 Kbp upstream the TSS) (Fig. 2F). These results support the notion that the proximal promoter region is involved in the induction of HO-1 by a cAMP/PKA/CREB-dependent pathway.
In agreement, transfection of Y1 cells with DN-CREB-M1 prevented the increase in luciferase activity of pHO-1 (−295/+74).LUC induced by ACTH or 8Br-cAMP (Fig. 3A). Similar results were obtained when the cells were incubated in the presence of the PKA inhibitor, H89 (Fig. 3B). A sequence comparison of the proximal promoter region of the murine Hmox1 gene with the human, monkey, bovine and rat sequences indicates that the −37 CRE site is highly conserved among species, while a significant homology was observed in the −146 CRE site between mouse and rat promoter regions (Fig. 3C).
As several antioxidant and protective effects have been associated with the stimulation of the PI3K/AKT pathway in different systems, we analyzed the involvement of this pathway in the induction of HO-1 by cAMP in Y1 cells. Figure 4 shows a time course of AKT phosphorylation by ACTH (Fig. 4A) and 8Br-cAMP (Fig. 4B) indicating the activation of this signaling pathway. As shown in Fig. 4C, this effect was inhibited by the PI3K inhibitor, wortmannin, but not by H89, ruling out the involvement of PKA on Akt activation. As for the expression of HO-1, wortmannin treatment also inhibited HO-1 induction by 8Br-cAMP, as determined by western blotting (Fig. 4D). Accordingly, the PI3K inhibitor, LY204002, also prevented the increase in the pAKT/AKT ratio and in the expression levels of HO-1 triggered by 8Br-cAMP (Fig. 4E). Luciferase activity of the reporter plasmid containing the proximal fragment of the HO-1 promoter pHO-1(−295/+74).LUC was also inhibited by wortmannin and LY294002 (Fig. 4F). These results suggest the establishment of a crosstalk between the cAMP and PI3K pathways affecting the stimulation of HO-1transcription.
In order to test this hypothesis, we first incubated Y1 cells with 8Br-cAMP in the presence of wortmannin and assessed CREB activation by determining Ser133-CREB phosphorylation. Results presented in Fig. 5A indicated that wortmannin slightly but significantly prevented the increase in pCREB/CREB ratio triggered by 8Br-cAMP. This treatment also prevented the 8BrcAMP-dependent increase in the pAKT/AKT ratio (insert) and the stimulation of luciferase activity in cells transfected with pCRE.LUC, further supporting the crosstalk hypothesis (Fig. 5B).
When ChIP analysis was performed to assess the effect of wortmannin on the binding of pCREB to the proximal promoter of HO-1, we observed that the PI3K inhibitor completely blocked the 8Br-cAMP-dependent increase in pCREB binding to the proximal promoter region, suggesting that an active PI3K signaling pathway is necessary for the proper recruitment of pCREB to this region (Fig. 5C).
In this study, we report that the transcriptional stimulation of the Hmox1 gene by ACTH/cAMP relies on the phosphorylation of CREB and its binding to a functional CRE localized within the proximal promoter region of HO-1. We also suggest the involvement of the PI3K pathway in the recruitment of pCREB to the HO-1 promoter.
Induction of HO-1 has been widely recognized as an antioxidant response upon exposure to stressful stimuli. In adrenal cells, HO activity has also been included in the category of autocrine/paracrine modulators of steroidogenesis (Bhasker et al. 1989, Bohm et al. 1993, Pomeraniec et al. 2004, Spiga et al. 2011). As HO-1 expression is induced in adrenocortical cells 5–8h after ACTH addition (Pomeraniec et al. 2004), we hypothesize that induction of HO-1 by ACTH in adrenocortical cells could serve two purposes: (a) protects the cells from the deleterious effects associated with an increased generation of ROS and (b) inhibits steroidogenesis by a mechanism involving heme oxygenase activity (CO generation, heme catabolism etc.), thus allowing the cells to avoid an unregulated steroidogenic response. In fact, we have previously demonstrated that inhibition of HO-1 expression or activity results in significantly higher steroid levels in both Y1 cells and rats (Pomeraniec et al. 2004, Grion et al. 2007).
ACTH is the main regulator of glucocorticoid synthesis in adrenal cells, a process that, even under physiological conditions, releases reactive oxygen species by leaking incompletely reduced products of O2 (Battista et al. 2009). Accordingly, we have previously demonstrated an increase in ROS generation in adrenocortical Y1 cells incubated with ACTH. The role of HO-1 as an antioxidant enzyme is suggested, as ROS generation induced by H2O2 was significantly lower in cells where HO-1 was induced by pretreatment with ACTH. Although the involvement of NRF2 in HO-1 induction in adrenocortical cells stimulated with NO-donors has been previously demonstrated (Astort et al. 2014), several evidences led us to discard its participation in the induction of HO-1 by ACTH/cAMP. First, neither ACTH nor 8Br-cAMP treatment stimulated luciferase activity of an NRF2 reporter plasmid, and although 8Br-cAMP was able to stimulate luciferase activity in Y1 cells transfected with a reporter plasmid containing 4Kbp of the HO-1 promoter of the murine gene (including the E1 enhancer containing an ARE element), stimulation was conserved in a reporter plasmid carrying only 300bp of the proximal promoter that lacks the ARE. Finally, 8Br-cAMP treatment failed to induce luciferase activity of a reporter plasmid containing only the E1 region.
Several effects of ACTH in adrenal cells are mediated by the activation of the cAMP/PKA pathway (Kimura 1986, Schimmer 1995, Gallo-Payet & Payet 2003, Sabban et al. 2004, Gallo-Payet 2016). As stated previously, among them we have demonstrated the participation of the cAMP/PKA pathway in the transcriptional stimulation of HO-1 by ACTH (Pomeraniec et al. 2004). Similar results were obtained in hepatic cells, where glucagon treatment led to the induction of the Hmox1 gene through a cAMP/PKA-dependent mechanism (Immenschuh et al. 1998b).
Phosphorylation of CREB on Ser133 by PKA regulates its transcriptional activity and promotes the recruitment of the co-activator protein CBP (CREB-binding protein)/p300 linking the CRE/CREB complex to components of the basal transcriptional machinery. It was therefore a suitable candidate as a mediator for the ACTH effect at the transcriptional level. Moreover, CREB phosphorylation and binding to CREs has been involved in the induction of HO-1 by different stimuli (Kronke et al. 2003, Wright et al. 2009, Park et al. 2013, Barnett et al. 2015). In our experimental system, targeting CREB expression with a dominant-negative isoform of CREB (DN-CREB-M1) decreased, while overexpressing CREB (RSV-CREB) increased HO-1 protein levels in 8Br-cAMP-stimulated Y1 cells, suggesting the involvement of CREB in the ACTH/cAMP-mediated induction of HO-1.
Involvement of CRE sites in the induction of HO-1 has been linked to the stimulation of cGMP/PKG, to the inhibition of protein phosphatases 1 and 2A and to the overexpression of biliverdin reductase via ATF-2 (Immenschuh et al. 1998a, 2000, Kravets et al. 2004). In rat hepatocytes, cGMP/PKG-dependent induction of HO-1 involves a CRE/AP-1 site located −665/−654 upstream the TSS (Immenschuh et al. 1998a), while in human endothelial cells, a distal promoter fragment (between 3.8 and 4.9kb upstream the TSS) containing a CRE site was responsible for the PKA-mediated induction of HO-1 by oxidized phospholipids (Kronke et al. 2003). This CRE site has been also shown to act in concert with a proximal E-box located within the proximal promoter of the human Hmox1 gene (Wright et al. 2009).
Analysis of 4kb from the murine HO-1 promoter showed the presence of only two putative CREB binding sites located at −146 and −37 upstream the TSS. These CRE-like sites present a high homology with those in the proximal promoter of the rat Hmox1 gene, while the −37 site is conserved in all the species analyzed. In particular, this last site contains the Ebox sequence reported by Wright and coworkers in the human HO-1 promoter (Wright et al. 2009).
Our results showed that this proximal promoter region of the murine Hmox1 gene recruited phospho-CREB in cAMP-stimulated Y1 cells, as determined by ChIP analysis. In agreement, both pharmacological (PKA inhibitor H89) and molecular (DN-CREB-M1) inhibition of CREB activation resulted in a significant decrease in the activity of the pHO-1(−295/+74).LUC reporter plasmid. These results further support the involvement of this proximal promoter region in the PKA-mediated induction of Hmox1 gene. This is, to our knowledge, the first report on the induction of HO-1 mediated by CREB recruitment within the proximal promoter region of the murine Hmox1 gene.
It is well known that activation of PI3K/AKT signaling functions as a cytoprotective mechanism against oxidative stress (Mo et al. 2012, Ryu et al. 2014, Yang et al. 2013). In this sense, our results demonstrated that, in addition to stimulating Hmox1 gene expression, treatment of Y1 cells with either ACTH or 8Br-cAMP also leads to an increase in Akt phosphorylation levels. As this effect was prevented by the addition of specific PI3K inhibitors (Bain et al. 2007), but not by H89, we ruled out the involvement of PKA on AKT phosphorylation. As an alternative mechanism, Baviera and coworkers described an Epac-dependent PI3K activation by cAMP in rat skeletal muscle (Baviera et al. 2010).
Our results also indicated that the PI3K signaling pathway could modulate the transcriptional stimulation of the Hmox1 gene triggered by cAMP, as wortmannin and LY 294002 treatments inhibited HO-1 induction and the activation of the pHO-1(−295/+74).LUC reporter induced by 8Br-cAMP. Based on these results, we hypothesized that PI3K activation could affect the phosphorylation status of CREB and/or the binding of p-CREB to the HO-1 promoter. Although PI3K inhibition, only slightly (but significantly) decreased CREB phosphorylation by 8Br-cAMP, this pathway appears to play a significant role in the recruitment of pCREB to the HO-1 promoter, suggesting that protein kinases activated downstream of PI3K (such as AKT, PKCξ, RSK, SGK, p70S6K etc.) could modulate the binding of CREB to the HO-1 promoter. We do not discard the possibility that activation of the PI3K pathway affects the binding of CREB co-activators, thus influencing HO-1 transcription. In this sense, an increase in the recruitment of CBP to NRF2 or NFκB upon activation of the PI3K pathway has been previously demonstrated (Tang et al. 2007, Sakamoto et al. 2009).
In summary, binding of ACTH to the MC2R induces steroidogenesis and triggers the transcriptional stimulation of Hmox1 gene. In the present work, we demonstrate that HO-1 induction involves the classical cAMP/PKA pathway that leads to the phosphorylation and activation of CREB and to p-CREB recruitment to a proximal region of theHO-1 promoter. Our results also show that HO-1 induction is modulated by a functional crosstalk between the cAMP/PKA pathway and the PI3K-dependent cascade (Fig. 6).
Declaration of interest
The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the research reported.
This study was supported by grants from Agencia Nacional de Promoción de Ciencia y Tecnología (PICT 2008 N 1034) and University of Buenos Aires (UBACyT 2011-2014).
BedogniBPaniGColavittiRRiccioABorrelloSMurphyMSmithREboliMLGaleottiT2003Redox regulation of cAMP-responsive element-binding protein and induction of manganous superoxide dismutase in nerve growth factor-dependent cell survival. Journal of Biological Chemistry27816510–16519. (doi:10.1074/jbc.M301089200)
BertucciPYNachtASAlloMRocha-ViegasLBallareCSoronellasDCastellanoGZaurinRKornblihttARBeatoM2013Progesterone receptor induces bcl-x expression through intragenic binding sites favoring RNA polymerase II elongation. Nucleic Acids Research416072–6086. (doi:10.1093/nar/gkt327)
ChinnAMCiaisDBaillySChambazELaMarreJFeigeJJ2002Identification of two novel ACTH-responsive genes encoding manganese-dependent superoxide dismutase (SOD2) and the zinc finger protein TIS11b [tetradecanoyl phorbol acetate (TPA)-inducible sequence 11b]. Molecular Endocrinology161417–1427. (doi:10.1210/mend.16.6.0844)
GrionNRepettoEMPomeraniecYCalejmanCMAstortFSanchezRPignataroOPAriasPCymeryngCB2007Induction of nitric oxide synthase and heme oxygenase activities by endotoxin in the rat adrenal cortex: involvement of both signaling systems in the modulation of ACTH-dependent steroid production. Journal of Endocrinology19411–20. (doi:10.1677/JOE-06-0199)
ImmenschuhSHinkeVOhlmannAGifhorn-KatzSKatzNJungermannKKietzmannT1998aTranscriptional activation of the haem oxygenase-1 gene by cGMP via a cAMP response element/activator protein-1 element in primary cultures of rat hepatocytes.Biochemical Journal334141–146. (doi:10.1042/bj3340141)
KronkeGBochkovVNHuberJGruberFBlumlSFurnkranzAKadlABinderBRLeitingerN2003Oxidized phospholipids induce expression of human heme oxygenase-1 involving activation of cAMP-responsive element-binding protein. Journal of Biological Chemistry27851006–51014. (doi:10.1074/jbc.M304103200)
MannaPREubankDWLalliESassone-CorsiPStoccoDM2003aTranscriptional regulation of the mouse steroidogenic acute regulatory protein gene by the cAMP response-element binding protein and steroidogenic factor 1.Journal of Molecular Endocrinology30381–397. (doi:10.1677/jme.0.0300381)
McMahonMThomasNItohKYamamotoMHayesJD2004Redox-regulated turnover of Nrf2 is determined by at least two separate protein domains, the redox-sensitive Neh2 degron and the redox-insensitive Neh6 degron. Journal of Biological Chemistry27931556–31567. (doi:10.1074/jbc.M403061200)
MercauMEAstortFGiordaninoEFMartinez CalejmanCSanchezRCaldareriLRepettoEMCosoOACymeryngCB2014Involvement of PI3K/Akt and p38 MAPK in the induction of COX-2 expression by bacterial lipopolysaccharide in murine adrenocortical cells. Molecular and Cellular Endocrinology38443–51. (doi:10.1016/j.mce.2014.01.007)
MoLYangCGuMZhengDLinLWangXLanAHuFFengJ2012PI3K/Akt signaling pathway-induced heme oxygenase-1 upregulation mediates the adaptive cytoprotection of hydrogen peroxide preconditioning against oxidative injury in PC12 cells. International Journal of Molecular Medicine30314–320. (doi:10.3892/ijmm.2012.1002)
RyuMJKangKAPiaoMJKimKCZhengJYaoCWChaJWChungHSKimSCJungE20147,8-Dihydroxyflavone protects human keratinocytes against oxidative stress-induced cell damage via the ERK and PI3K/Akt-mediated Nrf2/HO-1 signaling pathways. International Journal of Molecular Medicine33964–970. (doi:10.3892/ijmm.2014.1643)
SpigaFLiuYAguileraGLightmanSL2011Temporal effect of adrenocorticotrophic hormone on adrenal glucocorticoid steroidogenesis: involvement of the transducer of regulated cyclic AMP-response element-binding protein activity. Journal of Neuroendocrinology23136–142. (doi:10.1111/j.1365-2826.2010.02096.x)
TangCHLuDYYangRSTsaiHYKaoMCFuWMChenYF2007Leptin-induced IL-6 production is mediated by leptin receptor, insulin receptor substrate-1, phosphatidylinositol 3-kinase, Akt, NF-kappaB, and p300 pathway in microglia. Journal of Immunology1791292–1302. (doi:10.4049/jimmunol.179.2.1292)