Abstract
We have analyzed the promoter of human gilz (glucocorticoid-induced leucine zipper), a dexamethasone-inducible gene that is involved in regulating apoptosis, and identified six glucocorticoid (GC)-responsive elements and three Forkhead responsive elements (FHREs). Promoter deletion analysis and point mutations showed that individual mutation of the GC-responsive elements does not affect GC-induced transcription and that FHRE-1 and FHRE-3 elements contribute to the effects of GCs. Furthermore, overexpression of the Forkhead transcription factor FoxO3 enhances GC-induced gilz mRNA expression. The functional significance of the interaction between FoxO3 and GC receptor was established in T lymphocytes. Indeed, we show that GCs failed to induce GILZ expression in the presence of IL-2, a cytokine known to antagonize GC effects in T cells. Using a constitutive active mutant of protein kinase B that inactivates FoxO3 or a FoxO3 mutant that cannot be inactivated by protein kinase B, we demonstrate that IL-2 inhibitory effects on GILZ expression are mediated through inhibition of FoxO3 transcriptional activity. Therefore, FoxO3 appears to be a key factor mediating GC and IL-2 antagonism for gilz regulation in T lymphocytes. This regulation of GILZ expression was placed in a meaningful context in evaluating the effects of GILZ on GC-induced apoptosis in T lymphocytes.
GLUCOCORTICOIDS (GCs) EXERT their biological effects through activation of an intracellular receptor, the GC receptor (GR), causing its translocation from the cytoplasm to the nucleus where it regulates gene expression (for review see Ref. 1). Hormone-bound GR regulates transcription of target genes through direct binding of GR homodimers to the GC-responsive elements (GREs) present in the promoter of target genes, or through DNA binding-independent mechanisms involving protein to protein interactions (1–7).
In immune cells, cross-talks between GC and cytokine pathways play a major role in the therapeutic effects of GCs but also affect fundamental processes such as proliferation, differentiation or apoptosis. IL-2 has been shown to selectively protect CD4+CD25+ regulatory T cells from dexamethasone (DEX)-induced cell death, whereas IL-7 and IL-15 do not exert preferential protective effects (8). In mouse T helper cell lines, DEX completely inhibits IL-2-induced cell proliferation and reduces IL-4-mediated cell growth but has no effect on IL-9 response (9). These observations likely result from the antagonistic regulation of specific prosurvival or proapoptotic genes by GCs and cytokines.
Glucocorticoid-induced leucine zipper (GILZ) was initially isolated as a DEX-responsive gene from a thymus subtraction cDNA library (10). Little is known on the function of the GILZ protein; however, GILZ appears to modulate the response of peripheral T lymphocytes to antigenic stimulation. GILZ delays apoptosis of activated T lymphocytes through different mechanisms. Overexpression of GILZ in the murine lymphoid 3 DO cell line inhibits T cell receptor-induced apoptosis via down-regulation of Fas/FasL expression. GILZ also significantly affects IL-2 production and IL-2 receptor expression. Both observations suggest a role for GILZ in the inhibition of activation-induced cell death (AICD) (11). Furthermore, we recently reported that GILZ delays IL-2 withdrawal-induced apoptosis of activated T cells through a mechanism involving down-regulation of Bim expression, a proapoptotic member of the Bcl-2 family (12).
In monocyte/macrophage cells, IL-4 and IL-10 induce GILZ expression. However, in activated T lymphocytes, gilz transcripts are induced upon GC treatment or IL-2 deprivation (12, 13), suggesting that multiple regulatory pathways are involved in controlling GILZ expression (14).
We have recently cloned and characterized the human gilz promoter (12). This promoter contains a TATA box and putative binding sites for transcription factors: six GREs, three Forkhead-responsive elements (FHREs), three STAT6s (signal transducers and activators of transcription), three nuclear factor of activated T cells, two Octamer, and two c-myc binding sites. We previously reported that two of the three FHREs were responsible for IL-2 withdrawal-induced gilz expression.
In the immune system, IL-2 is a cytokine involved in proliferation, survival, and also death in the AICD of T lymphocytes, thus regulating the magnitude and the duration of T lymphocyte activation (15, 16). Binding of IL-2 to its receptor triggers a signaling cascade that induces, among others, the phosphoinositide 3-kinase (PI3K) pathway involved in T cell proliferation and survival (17, 18). One of the downstream effector of PI3K signaling is protein kinase B (PKB)/Akt, a Serine/Threonine kinase promoting cell survival by targeting proapoptotic proteins (19). PKB exerts one of its effects by phosphorylating members of the Forkhead family of transcription factors (FoxO1, FoxO3, and FoxO4), which contain three RXRXXS/T consensus phosphorylation sites for PKB (20). Activation of the PI3K/PKB pathway after IL-2 stimulation results in the phosphorylation of FoxO proteins, leading to their nuclear exclusion and therefore in the inhibition of their transcriptional activity.
In the present report we aimed to study the molecular mechanisms regulating expression of GILZ upon GC treatment of activated T cells. Six putative GREs are present in the gilz promoter, two of them being in the proximal region and four others being clustered in the distal region. Individual mutation of these GREs or mutation of the cluster of four GREs did not alter significantly GC-induced transcription. Our results suggest also that FoxO3 participates in GC-induced GILZ expression in the absence of IL-2. GILZ expression results in a delay in GC-induced apoptosis in CTLL-2 cells. Moreover, GCs failed to induce GILZ expression in the presence of IL-2, and we showed that IL-2 inhibition of FoxO3 transcriptional activity prevents GC-induced GILZ expression.
RESULTS
DEX-Induced GILZ Production in T Lymphocytes Does Not Require New Protein Synthesis
We first investigated GILZ regulation by GCs in CTLL-2 cells, a murine T lymphocyte cell line, dependent on IL-2 for their survival and growth. A 4-h treatment with DEX, in the absence of IL-2, induced GILZ expression in a dose-dependent manner as evaluated by Western blot analysis (Fig. 1A). Full expression was found at 10−7m DEX, and this concentration will be used for the other experiments. Upon longer exposure of the film (data not shown), GILZ is detectable in untreated cells, consistent with our previous report indicating that IL-2 withdrawal induced GILZ synthesis (12); however, the levels of GILZ in DEX-treated cells is much higher than in untreated cells. Therefore, film exposure to evaluate GC-induced GILZ synthesis did not allow for the detection of GILZ in untreated CTLL-2 cells.
DEX Induces GILZ Expression in T Lymphocytes A, CTLL-2 cells were deprived of IL-2 for 2 h and then treated with the indicated concentrations of DEX for 4 h. Western blot (WB) was performed using an anti-GILZ antibody. As a loading control, the membrane was blotted again with an anti-β-tubulin antibody. GILZ expression was quantified by densitometric analysis and normalized to the densitometric value of the β-tubulin (ratio GILZ/β-Tubulin). B, CTLL-2 cells were deprived of IL-2 for 2 h and then treated with the indicated period of times with DEX (10−7m). Western blot was performed using an anti-GILZ antibody. As a loading control, the membrane was blotted again with an anti-β-tubulin antibody. GILZ expression was quantified by densitometric analysis and normalized to the densitometric value of the β-tubulin (ratio GILZ/β Tubulin). C, CTLL-2 cells were deprived of IL-2 for 2 h and then treated for the indicated periods of time with DEX 10−7m. Total RNA was extracted and used as a template for RT-PCR using primers specific for gilz or β-actin. gilz expression was quantified by densitometric analysis and normalized to the densitometric value of the β-actin (ratio gilz/β actin). D, CTLL-2 cells were deprived of IL-2 for 2 h, pretreated with or without CHX (1 μm) for 1 h, and then DEX (10−7m) was added or not for 3 h. Total RNA was extracted and used as a template for RT-PCR using primers specific for gilz or β-actin.
We then studied the expression of GILZ during DEX treatment. Figure 1B shows that GILZ is induced in a time-dependent manner to reach a maximum after 6 h of DEX treatment. The level of gilz mRNA expression was then evaluated. RT-PCR experiments showed that gilz is expressed as soon as 30 min after DEX treatment (Fig. 1C). The level of gilz mRNA increased in a time-dependent manner to reach a steady state after 4 h of stimulation by GCs and was still present 8 h after DEX addition (Fig. 1C).
To assess whether GC-induced GILZ production required new protein synthesis, CTLL-2 cells were treated for 4 h with or without DEX in the presence or not of cycloheximide (CHX). CHX caused an increase of basal and DEX-induced gilz mRNA levels (Fig. 1D). This phenomenon has been previously described for early responsive gene and has been termed “superinduction” (21). Furthermore, RU 486, a glucocorticoid receptor (GR) antagonist, totally inhibited DEX-induced GILZ expression (data not shown), suggesting that GILZ induction required a functional GR.
Thus, the data shown above indicate that gilz typically behaves as an early response gene and that its induction by GCs occurs independently of new protein synthesis and likely depends upon transcriptional regulation.
Functional Analysis of the 5′-Regulatory Region of the gilz Gene in Response to DEX
We have previously described the human gilz promoter (12), and a scheme of the main regulatory sites located upstream of the TATA box is presented in Fig. 2A. Since our description of the gilz promoter, Wang et al. (22) published a paper identifying an additional GRE. gilz promoter should now contain six potential GREs, and four of them have a conserved half-palindrome TGTTCT known to be critical for GR binding to DNA and for transcriptional activity (Fig. 2B) (23). Three FHREs are also present in the gilz promoter.
Functional Analysis of the gilz Promoter in Response to DEX A, Analysis of the transcriptional activity of various gilz promoter constructs. Left, Schematic representation of gilz promoter deletion constructs driving the luciferase reporter gene. Numbers indicate the various binding site positions relative to the transcription start site (+1). Right, After 2 h of IL-2 deprivation, cells were transfected with the indicated promoter deletion constructs (10 μg) and 0.1 μg pCMV-Renilla. Cells were then treated with or without DEX (10−7m) for 6 h. Results are expressed in normalized RLU and represent the mean ± semof three independent experiments. *, P < 0.05 compared with untreated cells. NT, Not Treated. B, Alignment of the consensus core GRE with the six GREs present in the gilz promoter. The mutations that were introduced for the experiments described below are underlined. C, After 2 h of IL-2 deprivation, CTLL-2 cells were transfected with the reporter plasmid p-1940 wild type (wt) or with either p-1940GRE-1mut, p-1940GRE-2mut, p-1940GRE-3mut, p-1940GRE-3′mut, p-1940GRE-4mut, p-1940GRE-5mut (10 μg) containing one mutation in GRE-1, one mutation in GRE-2, one mutation in GRE-3, one mutation in GRE-3′, one mutation in GRE-4, one mutation in GRE-5, respectively, and treated with or without DEX (10−7m) for 6 h. Results are expressed as percentage of DEX-induced p-1940 wt (wild type) activity (DEX treatment corresponds to 100%) and represent the mean ± sem of three independent experiments performed in duplicate. D, After 2 h of IL-2 deprivation, CTLL-2 cells were transfected with the reporter plasmid p-1940 wt or with either p-1940GRE-3,3′,4,5mut or p-1940GRE-1,2,3,3′,4,5mut and treated with or without DEX (10−7m) for 6 h. Results are expressed in normalized RLU and represent the mean ± sem of a representative experiment performed in duplicate.
To further study the molecular mechanisms involved in GC-induced gilz promoter activity, a series of constructs containing successive 5′-deletions of the gilz promoter coupled to a firefly luciferase reporter gene was generated (Fig. 2A). Promoter activities of these constructs were assessed in CTLL-2 cells treated with or without DEX (10−7m) in the absence of IL-2. Activity of the p-1940 construct was induced upon DEX treatment (3.5-fold), whereas the activities of the three shortest constructs were not affected by DEX (Fig. 2A), suggesting that the region located between nucleotides −1940 and −1526 is important for GC-induced gilz promoter activity. Single mutations of each GRE did not alter significantly p-1940 activity (Fig. 2C). Furthermore, single mutation of all the GREs present in the distal cluster (GRE-3, GRE-3′, GRE-4, and GRE-5) showed a slight but not significant reduction of DEX induction (Fig. 2D).
Together these results show that none of the GREs by itself is critical for GC induction of the gilz promoter. However, mutation of all the GREs abolished DEX induction, supporting a critical role of the GREs for GC-induced gilz expression.
FoxO3 Binding to the gilz Promoter Is Necessary for Full GC Responsiveness of the gilz Promoter
Although CTLL-2 cells are dependent on IL-2 for their survival and their growth, we first studied the mechanism of GILZ regulation upon a unique treatment by GCs. We and others have previously reported that the Forkhead protein FoxO3 is transcriptionally active in the absence of IL-2 in CTLL-2 cells (12, 24). Cooperation between FoxO3 and GR has already been described in the regulation of genes involved in gluconeogenesis (25–27).
We first investigated whether DEX treatment of CTLL-2 cells affects FoxO3 binding to the FHREs present in the gilz promoter. We used a DNA affinity precipitation method with double-stranded oligonucleotides corresponding to FHRE-1, FHRE-2, or FHRE-3 sequences of the gilz promoter coupled to streptavidin agarose beads. Nuclear proteins from DEX-treated or untreated CTLL-2 cells transfected with a FoxO3 expression plasmid were incubated with 5′-biotinylated FHRE-1, FHRE-2, and FHRE-3 oligonucleotides. Figure 3A shows that FoxO3 binding to FHRE-1 and FHRE-3 was not affected by DEX treatment and that no FoxO3 binding to FHRE-2 was detected in either condition. DEX treatment did not affect the amounts of FoxO3 expressed in CTLL-2 cells (data not shown).
FHREs Present in the gilz Promoter Participate in GC Induction A, DNA affinity purification of FHRE-binding proteins. CTLL-2 cells were transiently transfected with 20 μg pcDNA3-FoxO3 wild type (WT). After 16 h of expression, cells were deprived of IL-2 for 2 h and then treated with or without DEX (10−7m) for 3 h. 5′-Biotinylated FHRE-1, FHRE-2, and FHRE-3 and mutated FHRE-1 (FHRE-1mut) and FHRE-3 probes (FHRE-3mut), coupled to agarose beads, were incubated with nuclear extracts, and bound proteins were resolved by SDS-PAGE and identified by Western blotting (WB) using anti-FoxO3-specific antibody. N.E., Nuclear extract. B, Mutations of FHRE-1 and FHRE-3 in the gilz promoter affect its regulation by GCs. CTLL-2 cells were deprived of IL-2 for 2 h and then transiently transfected with the reporter plasmid p-1940 wild type (wt) or with either p-1940FHRE-1mut, p-1940FHRE-3mut, or p-1940FHRE-1+3mut (10 μg) containing one mutation in FHRE-1, or one mutation in FHRE-3, or both mutations, respectively, and 0.1 μg pCMV-Renilla. Cells were then treated with or without DEX (10−7m) for 6 h. Results are expressed in DEX fold induction (DEX-treated cells/control cells) and represent the mean ± sem of three independent experiments performed in duplicate. *, P < 0.05 and **, P < 0.01 compared with cells transfected with the p-1940 wild type construct.
To address whether FoxO3 binding to these FHREs could contribute to the regulation of gilz promoter by GCs, we then evaluated the effects of point mutations of FHRE-1 (p-1940FHRE-1mut), FHRE-3 (p-1940FHRE-3mut), and of both sites (p-1940FHRE-1+3mut) on DEX-induced p-1940 activity. We first verified that these point mutations indeed invalidated FoxO3 binding to FHRE-1 and FHRE-3 (Fig. 3A). Figure 3B shows that the sole mutation of FHRE-3, but not of FHRE-1, down-regulated significantly the activity of p-1940 upon DEX treatment. When both sites were mutated, a 67% reduction of DEX inducibility was observed.
These data suggest that binding of FoxO3 proteins is required for the full activation of gilz promoter in response to GCs in CTLL-2 cells.
Forkhead Transcription Factors Participate in GC-Induced Endogenous gilz Expression
To further explore our hypothesis of a role for FoxO3 in the regulation of GILZ expression by GCs, we then investigated whether FoxO3 overexpression would modify gilz promoter regulation and endogenous gilz expression.
FoxO3 was overexpressed in CTLL-2 cells, and luciferase activity of p-1940 was measured in cells treated with increasing concentrations of DEX. Figure 4A shows that FoxO3 overexpression significantly increased basal promoter activity. At a suboptimal dose of DEX (5 × 10−9m), promoter activity was higher when FoxO3 was overexpressed; however, at higher concentrations of DEX this effect was no longer observed. The effect of FoxO3 overexpression was then assessed in another model of T lymphocytes, the human Jurkat T cell line that does not express a functional GR. Jurkat cells were transfected with increasing amounts of a GR expression vector in the presence or not of FoxO3. Figure 4B shows that FoxO3 significantly enhanced GC-induced p-1940 activity when cells were transfected with 1, 5, and 10 μg of GR expression vector, suggesting a role for FoxO3 in the regulation of the gilz promoter by GCs. In this model, basal activity of p-1940 was not affected by FoxO3 overexpression.
FoxO3 Enhances GC-Induced gilz Promoter Activity and Endogenous gilz Expression A, FoxO3 enhances p-1940 transactivation by GCs in CTLL-2 cells. CTLL-2 cells were transiently transfected with 10 μg p-1940 luciferase construct, 10 μg pcDNA3, or pcDNA3-FoxO3 wild type and 0.1 μg pCMV-Renilla. Cells were deprived of IL-2 for 2 h and then treated with or without the indicated concentrations of DEX for 6 h. Results are expressed in normalized RLU and represent the mean ± sem of three independent experiments performed in duplicate. *, P < 0.05 compared with control cells for each concentration of DEX. B, GR and FoxO3 overexpression cooperates to transactivate gilz promoter in Jurkat cells. Jurkat cells were transiently transfected with 10 μg p-1940 luciferase construct, 1 μg pcDNA3 or pcDNA3-FoxO3 wild type, 0.1 μg pCMV-Renilla, and the indicated quantity of pcDNA3-GR. The total amounts of transfected DNA were kept constant by addition of empty control vector. Cells were then treated with DEX (10−7m) for 12 h. Results are expressed in normalized RLU and represent the mean ± sem of three independent experiments performed in duplicate. *, P < 0.05 and ***, P < 0.001 compared with control cells for each quantity of GR expression vector. C, FoxO3 cooperates with DEX to enhance gilz mRNA expression in CTLL-2 cells. CTLL-2 cells were transiently transfected with the indicated dose of pcDNA3-FoxO3 wild type. The total amounts of transfected DNA were kept constant by addition of empty control vector. Cells were maintained in culture in the absence of IL-2 for 6 h and then treated for 1 h with the indicated concentration of DEX. Total RNA was extracted and used as a template for RT-PCR and real-time quantitative PCR using primers specific for gilz or 18S. Results are presented as relative gilz mRNA expression compared with 18S expression. *, P < 0.05; and **, P < 0.01 compared with control cells for each concentration of DEX.
We then measured, by real-time quantitative PCR, the levels of endogenous gilz mRNA in CTLL-2 cells transfected with increasing doses of FoxO3 and treated with two concentrations of DEX for 1 h. Results showed that in these experimental conditions, overexpression of FoxO3 (1 and 10 μg) did not affect the basal level of gilz mRNA but significantly enhanced gilz mRNA expression at low doses of DEX (5.10−9m) (Fig. 4C). When cells were treated with 10−7m of DEX, GC-induced gilz expression was increased only when cells were transfected with 10 μg of FoxO3 expression vector (Fig. 4C), suggesting that at a higher level of GR activation more FoxO3 protein is required. A discrepancy exists between results obtained in Figs. 4A and 4C when using 10 μg of FoxO3 plasmid. These differences could be due to the experimental systems (transient transfection of a luciferase construct vs. real-time PCR) in which endogenous FoxO3 may differentially impact the endogenous gene from the p-1940 construct.
Together these results further suggest that FoxO3 participates in GC-induced GILZ expression.
GILZ Protects CTLL-2 Cells from GC-Induced Apoptosis
We then asked what would be the physiological consequence of GILZ expression induced by GCs in T lymphocytes. Given that GCs induce apoptosis in CTLL-2 cells (28), we investigated the levels of apoptosis upon DEX treatment in CTLL-2 cells stably transfected with a pSUPER-GILZ vector to induce cellular expression of small interfering RNA (siRNA)-GILZ (12, 29). When cells were treated with 10−7m DEX, GILZ expression was only slightly reduced due probably to an insufficient expression of gilz siRNA in this case (data not shown). Consequently, cells were treated with 10−8m DEX for 16, 20, and 24 h, and the percentage of apoptotic cells was determined by quantification of cells in sub-G1. Results showed that inhibition of GILZ expression in siRNA-GILZ-expressing cells resulted in a dose-dependent increase of apoptosis compared with control cells expressing a random siRNA (Fig. 5A). Indeed, in clone 30, which expresses sufficient levels of siRNA to significantly inhibit GC-induced GILZ expression (Fig. 5A, right panel), the percentage of apoptotic cells is 2-fold higher than in control cells for the 16-h time point, whereas the number of apoptotic cells in clone 31, in which silencing of GILZ expression was not as efficient, was lower than in clone 30.
GILZ Protects CTLL-2 Cells from GC-Induced Apoptosis A, Randomly selected siRNA-Rd (siRNA-random) or siRNA-GILZ CTLL-2 clones, siRNA-rd clone B and C and siRNA-GILZ clone 30 and 31, were treated with DEX (10−8m) for the indicated period of time. Cells were then stained with propidium iodide and analyzed for DNA hypodiploidy by flow cytometry. The data shown are representative of three independent experiments. Right panel, Western blot (WB) analysis of GILZ expression in the various siRNA-GILZ clones cultured in the presence of DEX (10−8m) for 16 h. B, Randomly selected control and Myc-GILZ overexpressing CTLL-2 clones were treated with DEX (10−7m) for the indicated period of time. Cells were then stained with propidium iodide and analyzed for DNA hypodiploidy by flow cytometry. The data shown are representative of three independent experiments. Right panel, Western blot analysis of Myc-GILZ-expressing clones cultured in the presence of DEX (10−7m) for 12 h.
As a mirror of these experiments, we assessed the percentage of apoptosis in GILZ-overexpressing cells. Because we expected these cells to be more resistant to apoptosis, cells were treated with 10−7m DEX. Figure 5B shows that, in these cells, the number of cells in apoptosis is lower than in control cells, suggesting that GILZ hindered DEX-induced apoptosis.
Therefore, in T lymphocytes, GCs induce the expression of GILZ, resulting in a delay of GC proapoptotic activity.
GC-Induced GILZ Synthesis Is Repressed by IL-2 in Activated T Lymphocytes
IL-2 is a cytokine that plays a major role in T cell proliferation and in the regulation of T cell survival and apoptosis. Furthermore, IL-2 inhibits GC-induced antiproliferative effects and apoptosis of CTLL-2 cells (30). Therefore, we investigated the effects of IL-2 stimulation on GC-induced GILZ expression.
CTLL-2 cells were treated with DEX in the presence or not of IL-2, and results showed that GILZ expression was dramatically reduced (5-fold reduction at 4 h) in cells cotreated with GC and IL-2 at 4 and 8 h of induction as compared with cells stimulated with DEX alone (Fig. 6A). To rule out that GILZ regulation by DEX and IL-2 was a unique feature of the CTLL-2 cell line, we next investigated GILZ expression in human T cell blasts derived from healthy blood donors. Peripheral blood lymphocytes stimulated for 48 h with phytohemagglutinin (PHA) were cultured with IL-2 for 3 d, and then deprived of IL-2 for 24 h and treated with or without DEX and IL-2 for 4 h. Our results showed that GILZ expression was up-regulated in DEX-treated cells after 4 h of treatment, and that IL-2 reduced DEX-induced GILZ expression by 2-fold (Fig. 6B).
Regulation of GILZ Expression by DEX and IL-2 in Murine CTLL-2 Cells and in Human PHA-Stimulated Peripheral Blood Lymphocytes A, CTLL-2 cells were deprived of IL-2 for 2 h and then stimulated for the indicated period of time with or without DEX (10−7m) and IL-2 (1 ng/ml). B, PHA-stimulated peripheral blood lymphocytes were deprived of IL-2 for 24 h and then stimulated for 4 h with or without DEX (10−7m) and IL-2 (7.5 ng/ml). Western blot (WB) analysis was performed using polyclonal anti-GILZ antibody. As a loading control, membranes were blotted again with an antibody specific for β-tubulin. GILZ expression was quantified by densitometric analysis and normalized to the densitometric value of the β-tubulin (ratio GILZ/β Tubulin). C, After 2 h of IL-2 deprivation, cells were transiently transfected with the indicated promoter constructs (10 μg) and 0.1 μg pCMV-Renilla. Cells were then treated with or without DEX (10−7m) and IL-2 (1 ng/ml) for 6 h. Results are expressed in normalized RLU and represent the mean ± sem of three independent experiments. NT, Not treated. DEX fold inductions are indicated in brackets. *, P < 0.05 compared with untreated cells.
We then investigated whether IL-2-mediated inhibition of GILZ expression would also be observed in the regulation of the gilz promoter. CTLL-2 cells were transfected with the p-1940 reporter construct and treated with or without DEX and IL-2. Results showed that IL-2 totally abrogated DEX-induced luciferase activity of the p-1940 construct (Fig. 6C).
These results show that IL-2 down-regulates endogenous GC-induced GILZ expression as well as gilz promoter activity, suggesting that transcriptional mechanisms are involved in the inhibitory effects mediated by IL-2.
FoxO3 Participates in IL-2-Mediated Inhibition of GC-Induced gilz Promoter Activity
It was reported previously that IL-2 stimulation resulted in the inhibition of FoxO3 transcriptional activity in CTLL-2 cells (24, 31). Because we showed that FoxO3 plays a role in GC-induced GILZ expression, we asked whether FoxO3 inhibition through activation of the PI3K/PKB pathway by IL-2 participates in the down-regulation of GC-induced gilz promoter activity observed upon IL-2 stimulation.
FoxO3 subcellular localization was first assessed by Western blot in cells treated with or without DEX and IL-2. Results showed that in untreated and in DEX-treated cells, endogenous FoxO3 was present in the nucleus, whereas upon IL-2 treatment, in the presence of DEX or not, FoxO3 was present in the cytoplasm (Fig. 7A). These results show that FoxO3 might be transcriptionally active only in untreated or DEX-treated cells.
A Constitutive Active Form of PKB Reduces DEX-Induced gilz Promoter Activity A, Subcellular localization of FoxO3 in CTLL-2 cells. Cells were deprived of IL-2 for 3 h and then stimulated with or without DEX (10−7m) and IL-2 (1 ng/ml) for 1 h. After dosing, equal amounts of cytosolic and nuclear proteins were loaded in polyacrylamide gel, and Western blotting (WB) was performed using a specific anti-FoxO3 antibody. Membrane was blotted again with an anti-β-tubulin antibody to control the possible nuclei contamination with cytoplasmic material. B, Overexpression of activated PKB (gag-PKB) inhibits Forkhead protein transcriptional activity on a FHRE-Luc reporter construct. CTLL-2 cells were transiently transfected with 5 μg FHRE-Luc, 0.1 μg pCMV-Renilla, and 20 μg of empty vector or pSG5-gag-PKB. Luciferase activity was measured 8 h after transfection. Results are expressed in normalized RLU and represent the mean ± sem of three independent experiments performed in duplicate. C, gag-PKB expression leads to transcriptional repression of GC-induced gilz promoter activity. CTLL-2 cells were transiently transfected with 5 μg p-1940 luciferase construct, 0.1 μg pCMV-Renilla, and 20 μg of vector or pSG5-gag-PKB. Cells were deprived of IL-2 for 2 h and then treated with or without DEX (10−7m) for 6 h. Results are expressed in normalized RLU and represent the mean ± sem of three independent experiments performed in duplicate. NT, Not treated. DEX fold inductions are indicated in brackets. *, P < 0.05.
To evaluate the effects of the PI3K/PKB pathway in the regulation of GC-induced gilz promoter activity, transcriptional activity of Forkhead proteins was measured in cells by overexpressing a constitutive active form of PKB, gag-PKB (32). Figure 7B shows that, in the absence of IL-2, gag-PKB overexpression considerably reduced FoxO3 transcriptional activity, as assessed with a FHRE-Luc reporter construct. Cotransfection of gag-PKB with the p-1940 luciferase construct in CTLL-2 cells was then performed. In gag-PKB-overexpressing cells, a 52% decrease of DEX-induced p-1940 luciferase activity was observed (Fig. 7C), in accordance with the reduced induction observed in the p-1940FHRE1+3mut construct (Fig. 3B). The decrease of p-1940 luciferase activity in untreated cells transfected with gag-PKB is consistent with our previous study showing that FoxO3 activity enhanced gilz promoter basal activity (12).
To further investigate the role of FoxO3 in the IL-2-mediated inhibitory effect, FoxO3 TM, a triple mutant (T32A, S253A, S315A) that cannot be phosphorylated by PKB (33), was overexpressed in CTLL-2 cells. IL-2 did not affect FoxO3 TM subcellular localization (Fig. 8A) or transcriptional activity (Fig. 8B). CTLL-2 cells were then cotransfected with the p-1940 reporter construct and FoxO3 TM. In control cells transfected with the empty vector, IL-2 reduced DEX-induced transcription by 60%, whereas in cells overexpressing FoxO3 TM the inhibitory effect of IL-2 was totally abrogated (Fig. 8C).
Overexpression of a Constitutive Active FoxO3 Mutant in CTLL-2 Cells Suppresses IL-2 Inhibition of GC-Induced gilz Promoter Activity A, Subcellular localization of FoxO3 in CTLL-2 cells overexpressing FoxO3 TM. CTLL-2 cells were transfected with pcDNA3 or pcDNA3-FoxO3 TM, deprived of IL-2 for 3 h, and then stimulated with or without IL-2 (1 ng/ml) for 1 h. Equal amounts of cytosolic (C) and nuclear (N) extracts were loaded in polyacrylamide gel, and Western blot (WB) was performed using a specific anti-FoxO3 antibody. After stripping, membranes were blotted again with anti-β-tubulin antibody to control the possible nuclei contamination with cytoplasmic material. B, FoxO3 TM is transcriptionally active. CTLL-2 cells were transiently transfected with 5 μg FHRE-Luc, 0.1 μg pCMV-Renilla, and 20 μg of pcDNA3 or pcDNA3-FoxO3 TM. Cells were then treated with or without IL-2 (1 ng/ml) for 8 h. Results are expressed in normalized RLU and represent the mean ± sem of two independent experiments performed in duplicate. NT, Not treated. C, CTLL-2 cells were transiently transfected with 5 μg p-1940 luciferase construct, 0.1 μg pCMV-Renilla, and 20 μg pcDNA3 or pcDNA3-FoxO3 TM. Cells were deprived of IL-2 for 2 h and then treated with or without DEX 10−7m and IL-2 (1 ng/ml) for 6 h. Results are expressed as a percentage of DEX activity where DEX treatment alone corresponds to 100%. Data represent the mean ± sem of three independent experiments performed in duplicate.
These data suggest that phosphorylation of FoxO3 by PKB contributes to the inhibitory effect of IL-2 on GC-induced gilz promoter activation.
DISCUSSION
GILZ has been previously reported to be induced by GCs in hematopoietic cells (13). GC effects are mainly the consequence of gene regulation through either GR binding to GREs on target gene promoters or protein-protein interactions with several transcription factors. Other mechanisms, such as recruitment of coactivators or corepressors, also regulate GR transcriptional activity (for review see Ref. 1).
GILZ expression is induced by GCs in lymphocytes in a dose-dependent manner, and gene expression does not require new protein synthesis. Individual mutations of the six GREs show no modifications of DEX-induced GILZ promoter activity. However, mutations of all the GREs suppress DEX effects, suggesting a critical role of the GREs in gilz induction by GCs. It is not unusual to find multiple GREs within a promoter as in the phenylethanolamine N-methyltransferase gene where four GREs are present in the distal region of the promoter (34). Similarly, in the tyrosine amino-transferase (tat) promoter, two functional GREs are clustered but located 2.5 kb upstream from the +1 (35). However, the affinity of the GR for half-palindromes depends on their neighboring sequences: whereas the affinity is very low for an isolated half-site, it can be high when this site is located at an appropriate distance from other GREs, as seen in the mmtv promoter (23), or from accessory factor binding sites, as described in the pepck (phosphoenol pyruvate carboxykinase) promoter (36).
In this work we identify gilz as a new example of a gene for which binding of FoxO3 to its target DNA sequence is needed for full activation by GCs. Indeed, when FHRE-1 and FHRE-3 were mutated, the transactivation of the p-1940 construct by GCs was significantly reduced. Moreover, quantification of gilz mRNA showed that FoxO3 overexpression led to a significant increase of gilz expression upon GC treatment. Several reports have mentioned previously that GR transcriptional activity requires binding of accessory factors of the Forkhead family to target sequences surrounding the GC-responsive elements (GREs). In the pepck, the igfbp-1 (IGF-binding protein-1), the tat, and the g6pt (glucose 6 phosphate transporter) promoters, accessory factor-responsive elements are in close proximity to the GREs and therefore constitute a GC-responsive unit necessary for their full activation upon GC treatment (25, 26, 37, 38). Whether the same mechanism is involved in the regulation of the gilz promoter and whether the GREs and the FHREs may constitute a GC-responsive unit remains to be evaluated.
Many groups have investigated whether direct protein-protein interaction between the GR and FoxO could explain their cooperative effects, as shown for the estrogen receptor and the progesterone receptor, which interact with FoxO1 and FoxO3 through their ligand-binding domain (39, 40). Previous reports have failed to detect interaction between in vitro translated FoxO1 and a glutathione-S-transferase fusion protein coupled to the ligand-binding domain of the GR, and we were also unable to detect such an interaction in coimmunoprecipitation experiments (data not shown), suggesting that other mechanisms are responsible for the cooperation (39, 40). Stafford et al. (41) have shown that binding of accessory factors to the pepck promoter enhanced GR affinity toward its GREs and reduced the dissociation rate of GR. Forkhead proteins are known to participate in the recruitment of coactivators such as members of the steroid receptor coactivator family (steroid receptor coactivator-1) and cAMP response element-binding protein-binding protein that are essential for the transcriptional machinery (42, 43). Therefore, FoxO3 might act as an accessory factor necessary for fine tuning the physiological response to GCs.
GCs are well known inducers of apoptosis in T lymphocytes (44, 45). Because GILZ is induced by GC treatment, we hypothesized that it could participate in the regulation of apoptosis. Our observations showed that GILZ hindered GC-induced apoptosis of CTLL-2 cells. These results reinforce the fact that GILZ is induced by proapoptotic signals, GC and IL-2 withdrawal (12), not to contribute to the apoptotic mechanism but rather to exert antiapoptotic effects, therefore delaying T cell death. In vivo treatment by GCs has been previously shown to promote a Th2 response that is characterized by lymphocytes secreting high amounts of IL-4 and involved in IgE isotype switching and allergy (46). Therefore, although not investigated in this work, we can speculate that GILZ could play a role in survival and/or differentiation of lymphocyte subpopulations such as Th2 cells upon antigenic stimulus. Another paradoxical point is that IL-2 known to promote growth and survival of T lymphocytes down-regulates GILZ expression. However, in the context of AICD observed in T lymphocytes upon antigen stimulation, it could be hypothesized that GILZ previously described to inhibit AICD must be down-regulated by T cell receptor activation and IL-2 to avoid survival of unwanted lymphocytes (11). Recently, Delfino et al. (47) showed that in GILZ transgenic mice, the number of CD4+CD8+ cells in the thymus was decreased, suggesting a proapoptotic function for GILZ. This work suggests that GILZ may not play the same role in the thymus and in the peripheral lymphocyte compartment.
We have previously reported that GILZ is induced upon IL-2 deprivation and that IL-2 treatment inhibits FoxO3 transcriptional activity, resulting in the inhibition of GILZ expression (12). These previous results suggested that IL-2 might be an important regulator of GILZ in lymphocytes. Having shown that FoxO3 binding to the FHREs is required for full GC-induced p-1940 activity, we investigated whether IL-2 stimulation would inhibit DEX-induced gilz promoter transactivation through the inhibition of FoxO3 activity. Indeed, the PI3K/PKB pathway, which is activated by IL-2 in lymphocytes, is known to inhibit transcriptional activity of a number of related Forkhead transcription factors (FoxO1, FoxO3, and FoxO4) (48–50). Our results show that IL-2-induced FoxO3 inhibition might explain at least part of the inhibitory effect of IL-2 on GC-induced GILZ expression.
Independent studies have shown that, on one hand, erythropoietin (Epo), a survival and terminal differentiation factor of erythroid progenitors, activated the PI3K/PKB pathway resulting in FoxO3 inactivation in the U7/Epo cell line (51) and, on the other hand, a microarray study, complemented by quantitative PCR analysis, has shown that Epo inhibited basal and DEX-induced gilz transcript expression in erythroid progenitors (52). These observations could be paralleled with our results obtained with IL-2 and GCs, further suggesting the key role of FoxO3 in the regulation of GILZ expression.
The antagonism between GC and cytokine pathways has been clearly described and is based, in part, on suppression of cytokine gene transcription by GCs (53) but also on cross-talk between cytokine transduction pathways and activated GR. GR transcriptional activity can be regulated through protein-protein interactions with transcription factors known to be activated by cytokines such as activator protein 1, nuclear factor-κB, STAT5, and STAT6 (3, 4–7, 54–56). The present study describes a new mechanism that mediates the opposite effects of GCs and IL-2, placing FoxO3 as a mediator of their antagonism, and reinforces previous observations that IL-2 can counteract GC-mediated transcription in T cells by different pathways (7).
MATERIALS AND METHODS
Chemical and Reagents
Recombinant human IL-2 (Proleukin) was a gift of Chiron (Emeryville, CA). DEX and CHX were purchased from Sigma (Les Mureaux, France). The dual luciferase reporter assay system was purchased from Promega Corp. (Madison, WI).
Cell Culture and Transfection
The murine IL-2-dependent cytotoxic T-cell line CTLL-2 and human Jurkat T cell line were cultured in RPMI 1640 medium (Invitrogen, Carlsbad, CA) containing 0.1 mg/ml streptomycin, 100 U/ml penicillin (Invitrogen), 1% sodium pyruvate (Invitrogen), and 10% fetal calf serum (Invitrogen). For CTLL-2 cells, 50 μm 2-mercaptoethanol (Sigma) and 1 ng/ml of human recombinant IL-2 were added to the culture medium. Transfections were performed using the electroporation method as described previously (7). Human peripheral blood mononuclear cells were obtained from anonymous healthy volunteers who were clearly informed about the aim of the study and gave their informed consent. Mononuclear cells were separated by Ficoll-Hypaque gradient centrifugation. Nonadherent lymphocytes were stimulated with 1 μg/ml PHA for 48 h and further expanded for 3 d in the presence of 7.5 ng/ml IL-2. T cell blasts were then washed to remove IL-2 and incubated 24 h at 37 C in culture medium without IL-2.
Plasmid Constructs
Reporter Plasmids.
Human genomic DNA was purified from the bacterial clone RP13–364K23 (The Sanger Centre, Hinxton, UK) and used as a template for PCR. The p-1940 reporter construct was described previously (12). For construction of the 5′-truncated gilz promoter, fragments were amplified using the following primer: forward primer for p-1526, cggggtacAGGCCTTCAGACTGCATTTG; for p-854, cggggtacGATCCTGGATAATGTTGAAT; for p-416, cggggtacGGGCAGTGACTGGGA-AGAGG; and cccaagcttCGCCAGTCCAACCCAGACTC as a reverse primer for all constructs named above. The PCR-amplified fragments were digested by KpnI/HindIII and inserted in the pGL3basic vector. Mutations of the GRE of the p-1940 construct were performed with the QuikChange Site-Directed Mutagenesis kit (Stratagene, La Jolla, CA) using the mutated primers GRE-1mut (forward, TTCAAACTAGCGCTCATCTTTA-CTGAATGC); GRE-2mut (forward, TGGTGGAACCCAATCATCTCCTTTGGTCCT); GRE-3mut (forward, AGTTAACAGAATCACCTGAGA TCATTATTG); GRE-3′ mut (forward, TAGTGCAAACACCGTCATCAGAGAGG); GRE-4mut (forward, ATAGCCTGCACTTTCATCTGTCTACTACAC); and GRE-5mut (forward, TTAGTGAATCATCTTGATGACCCATAAGTA). The p-1940 reporter construct was also mutated on GREs 3, 3′, 4, and 5 (p-1940 GRE 3,3′,4,5 mut) and GREs 1, 2, 3, 3′, 4, and 5 (p-1940 GRE 1,2,3,3′,4,5 mut). Mutations of the FHRE of the p-1940 reporter construct were described previously (12). All the constructs cited above were sequenced (MWG Biotech, Courtaboeuf, France). FHRE-Luc was a kind gift of Dr. M. Greenberg (Harvard Medical School, Boston, MA) (33).
Expression Plasmids.
pcDNA3-FoxO3 WT (wild type) was described previously (12). pECE-FoxO3 TM (T32A, S253A, S315A) was a kind gift of Dr. M. Greenberg (33). Plasmid was digested by HindIII/XbaI, and the FoxO3-containing fragments were subcloned into pcDNA3.1 vector. p-SG5-gagPKB first described by Genot et al. (57) was kindly provided by Dr. E. Genot (INSERM U441, Pessac, France). For pcDNA3-GR, construct p-GEX-5X1-human GR, a kind gift of Dr. D. K. Granner (Department of Molecular Physiology and Biophysics, Vanderbilt University School of Medicine, Nashville, TN), was digested by BamHI/XhoI, and the human GR-containing fragment was subcloned into pcDNA3-Myc vector.
Measurement of Transcriptional Activity Using Dual-Luciferase Reporter Assay
Cells were transiently transfected with 10 μg of firefly luciferase reporter plasmid and 0.1 μg of Renilla luciferase reporter plasmid (pCMV-Renilla) used as an internal control for normalization. Each electroporation cuvette cellular content was resuspended in 10 ml of complete media and then divided to perform the various treatments as indicated. Protein extracts were prepared using the Passive Lysis Buffer provided in the Dual-Luciferase Assay (Promega). Equal amounts of protein extracts were plated into a 96-well plate. Firefly luciferase activity was measured for 12 sec using the Microlumat Plus LB 96V luminometer (Berthold Technologies, Thoiry, France). To assess the internal standard activity, Stop and Glo reagent was added (Promega), and the peak of the Renilla luciferase activity was then measured. Normalized relative luciferase units (RLU) were then calculated as firefly luciferase units of protein extracts of treated or untreated cells divided by Renilla luciferase units of protein extracts of untreated cells. DEX fold induction was calculated as normalized RLU of DEX-treated cells divided by normalized RLU of untreated cells. Data represent the mean ± sem of three independent experiments, each performed in duplicate.
RT-PCR Analysis
Total RNA was extracted using Trizol (Invitrogen). PCRs were performed using 1 U of Taq Polymerase (Qbiogen, Illkirch, France). Primers specific for gilz were ATGGAGGTGGCGGTCTATCA and TTACACCGCAGAACCACCAG (25 cycles at 95 C for 1 min, at 60 C for 1 min, and at 72 C for 1 min) and for β-actin GGGTCAGAAGGATTCCTATG and GGTCTCAAACATGATCTGGG (20 cycles at 95 C for 1 min, at 55 C for 1 min, and at 72 C for 1 min). Densitometric analysis was performed using ImageQuant software (Amersham Biosciences, Buckinghamshire, UK).
Real-Time PCR
Total RNA from CTLL-2 cells was isolated by Trizol extraction technique (Invitrogen). Total RNA (1 μg) was treated with DNase I Amplification Grade procedure (Invitrogen). Random primed cDNA was prepared from total RNA with 200 U of reverse transcriptase using the Superscript II kit (Invitrogen). Real-time PCR was carried out on an ABI 7700 Sequence Detector (Applied Biosystems, Foster City, CA) using the qPCR Mastermix Plus for Sybr Green I (Eurogentec, Seraing, Belgium). PCRs were performed with 50 ng cDNA in the presence of 2.5 mm MgCl2, 200 μm deoxynucleoside triphosphates, 1.25 U Hot goldstar DNA polymerase, and 300 nm specific primers. Reaction parameters were 95 C for 10 min followed by 40 cycles at 95 C for 15 sec and at 60 C for 1 min. Standard curves were generated using serial dilutions of linearized standard plasmids, spanning 5 orders of magnitude and yielding correlation coefficients greater than 0.98 and efficiencies of at least 0.70 in all experiments. Amplification of ribosomal 18S RNA was used as internal control for data normalization expressed as the ratio (attomoles of GILZ per attomole of 18S). Primers were as follows: mGILZ forward (5′-GTGGCTCTGTCCTTAGGGTGG-3′); mGILZ reverse (5′-CCAGATGGGCATGTGCTTG-3′); mouse ribosomal 18S forward (5′-CCCTGCCCTTTGTACACACC-3′); mouse ribosomal 18S reverse (5-CGATCCGAGGGCCTCACTA-3′).
Antibodies and Western Blot
Polyclonal antibodies directed against GILZ were prepared by immunizing two rabbits with a glutathione-S-transferase-GILZ fusion protein. Polyclonal anti-FoxO3 antibody was purchased from Upstate Biotechnology, Inc. (Lake Placid, NY; no. 06–951). Monoclonal anti-β tubulin antibody was purchased from Sigma (no. T4026). For Western blotting, cells were lysed in Nonidet P-40 buffer (50 mm Tris HCl, pH 8; 0.5% Nonidet P-40; 150 mm NaCl; 0.1 mm EDTA; 10 mm NaF; 1 mm phenylmethylsulfonylfluoride; 1 μg/ml aprotinin; 1 μg/ml leupeptin). Proteins were resolved by SDS-PAGE, electroblotted onto Amersham polyvinylidine difluoride membranes, and visualized using standard enhanced chemiluminescence (Amersham Biosciences). Densitometric analyses of the blot were performed on several exposures to determine the limits of the linear response of the films using ImageQuant software (Amersham Biosciences).
DNA Affinity Precipitation of FoxO3 Proteins
CTLL-2 cells (10 ×106) were transfected with 20 μg pcDNA3-FoxO3 and cultured with IL-2 for 16 h. Cells were then deprived of IL-2 and treated with or without DEX (10−7m) for 3 h. Nuclear extracts were performed using a Kontes all-glass Dounce homogenizer as described previously (7). The double-stranded 5′-biotinylated oligonucleotides FHRE-1, FHRE-1mut, FHRE-2, FHRE-3, and FHRE-3mut (described previously in Ref. 12) were first coupled to streptavidin agarose beads (Sigma) for 1 h at 4 C. Nuclear extracts were then incubated for 2 h with the precoated beads. Beads were then washed three times and boiled in reducing sample buffer to elute the bound proteins. Proteins were then separated on an 8% polyacrylamide gel and electroblotted onto polyvinylidine difluoride membranes (Amersham Biosciences). Western blots were performed using the specific anti-FoxO3 antibody.
Measurement of Apoptosis
CTLL-2 clones stably expressing gilz siRNA or overexpressing GILZ have been described previously (12). Cells were treated with DEX for increasing periods of time, permeabilized with ethanol 100%, and stained with 50 μg/ml propidium iodide (Boehringer Mannheim, Mannheim, Germany). Apoptosis was determined by quantification of DNA hypodiploidy using flow cytometry (sub-G1 peak). Data acquisition was performed using Cellquest software (Becton Dickinson, Franklin Lakes, NJ).
Statistical Analysis
An unpaired Student’s t test was used to evaluate differences between means. Differences were considered significant when P < 0.05.
Acknowledgments
We thank Dr. M. Greenberg for pECE-FoxO3 TM and FHRE-Luc plasmids, Dr. E. Genot for providing the pSG5-gag-PKB vector, and Dr. D. K. Granner for providing the pGEX-5X1-GR plasmid. We also thank Dr. S. Kerdine-Römer for helpful discussion and Isabelle Cantaloube, Damiana Lecoeuche, and Say Viengchareun for excellent technical assistance.
This work was supported by a fellowship from the Association pour la Recherche sur le Cancer (to M.-L.A.-L.).
Abbreviations:
- AICD,
Activation-induced cell death;
- CHX,
cycloheximide;
- DEX,
dexamethasone;
- Epo,
erythropoietin;
- FHRE,
Forkhead-responsive element;
- GC,
glucocorticoid;
- GR,
GC receptor;
- GRE,
GC-responsive element;
- PHA,
phytohemagglutinin;
- PI3K,
phosphoinositide 3-kinase;
- PKB,
protein kinase B;
- siRNA,
small interfering RNA;
- RLU,
relative luciferase units;
- STAT,
signal transducer and activator of transcription.
