CREB is one component of the binding complex of the Ces-2/E2A-HLF binding element and is an integral part of the interleukin-3 survival signal

Abstract

The Ces-2/E2A-HLF binding element (CBE) is recognized by Caenorhabditis elegans death specification gene product Ces-2 and human acute lymphocytic leukemia oncoprotein E2A-HLF. In an attempt to identify a cellular CBE-binding protein(s) that may be involved in apoptosis regulation in mammals, multiple nuclear binding complexes of CBE were identified in various mammalian cell lines and tissues by electrophoretic mobility shift assay. Cyclic AMP (cAMP)-responsive element (CRE)-binding protein (CREB) was present in one major CBE complex of Ba/F3 and TF-1 cells, and both in vitro-translated and Escherichia coli-synthesized CREB bound to CBE. Activation of CREB by cAMP-elevating chemicals or the catalytic subunit of protein kinase A (PKAc) resulted in induction of the CBE-driven reporter gene. Stimulation of Ba/F3 cells with interleukin-3 (IL-3) promptly induced phosphorylation of CREB at serine(133) partially via a PKA-dependent pathway. Consistently, Ba/F3 cell survival in the absence of IL-3 was prolonged by activation of PKA. Conversely, treatment of cells with a PKA inhibitor or expression of the dominant negative forms of the regulatory subunit type I of PKA and CREB overrode the survival activity of IL-3. Last, the bcl-2 gene was demonstrated to be one candidate cellular target of the CREB-containing CBE complex, as mutations in the CRE and CBE sites significantly reduced the IL-3 inducibility of the bcl-2 promoter. Together, our results suggest that CREB is one cellular counterpart of Ces-2/E2A-HLF and is part of IL-3 dependent apoptosis regulation in hematopoietic cells.

FIG. 1

Multiple CBE-binding complexes exist in mammalian cells. Nuclear extracts were prepared from various cell lines and tissues (as indicated at the top) and subjected to EMSA with a ³²P-labeled CBE probe. A 100-fold excess of unlabeled CBE probe (lane 2) or mutant CBE probe M4 (lane 3) was used to compete (Comp.) for the formation of specific complexes. Multiple binding complexes are categorized as slow-migrating (S) and fast-migrating (F′) groups. Each group is composed of at least two to three complexes. Thy, thymocytes; Spl., splenocytes.

FIG. 2

CREB exists in CBE-binding complexes. (A) Molecular weight determination of CBE-binding proteins by UV cross-linking. The nuclear extract of Ba/F3 cells was mixed with a labeled CBE probe and irradiated with (+) or without (−) short-wavelength UV for 10 min (see Materials and Methods) in the presence (lanes 3 and 5) or absence (lanes 2 and 4) of competitor (Comp.) CBE or M4 (lane 6). One half of the sample was analyzed by EMSA (lanes 1 to 3), and the other half was analyzed by SDS-PAGE (lanes 4 to 6). Positions of molecular weight markers are marked at the left of lane 4. (B) Nuclear extracts prepared from Ba/F3 (lanes 1 to 3) and TF-1 (lanes 4 to 6) cells or in vitro-synthesized CREB (lanes 7 to 9) were subjected to EMSA. Rabbit preimmune serum (Pre.) or polyclonal anti-CREB antibody (Ab) (CREB) was used to perform supershift experiments. The position of the supershifted binding complex is indicated by a thick arrow. In the preimmune serum, there is an undetermined CBE-binding activity that did not supershift the CREB-containing complex but comigrated with the supershifted CREB-CBE complex. The positions for slow- and fast-migrating groups (S and F) are indicated as described in the legend to Fig. 1.

FIG. 3

Transcriptional activation of CBE reporter genes by the cAMP/PKA/CREB signaling pathway. (A) Activation of CBE reporter genes by cAMP-elevating agents. 293T and PC12 cells were transfected with p3xCBE-Luc (1 μg) and treated with or without forskolin plus IBMX (as indicated at the bottom). (B) CREB-dependent activation of the CBE reporter gene by PKA in PC12 cells. PC12 cells were transfected with the PKAc expression plasmid (0.1 μg) and various luciferase reporter plasmids (1 μg of each), together with dominant negative mutant CREBR287L (1 μg). (C) CREB-dependent activation of CBE reporter genes by PKA in Ba/F3 cells. Ba/F3 cells were electroporated with the PKAc expression plasmid (0.45 μg) and CBE luciferase reporter plasmid (3 μg), together with CREBR287L (0, 1.5 and 4 μg; shown as a triangle). Twenty hours after transfection, cell lysates were prepared and analyzed by luciferase assays. Data shown are representative results from three independent experiments performed in duplicate (in PC12 cells) or quadruplicate (in Ba/F3 cells). Luciferase activities are plotted in arbitrary units per microgram of total protein.

FIG. 4

IL-3 induces CREB phosphorylation and survival function partly via the PKA pathway. (A) IL-3 induces PKA-dependent CREB Ser¹³³ phosphorylation. Cytokine-starved Ba/F3 cells were pretreated with dimethyl sulfoxide (lanes 1 and 2) or PKI (lanes 3 and 4) for 30 min before being restimulated with IL-3 (IL3). After 5 min of stimulation, cells were harvested and analyzed by Western blot analysis with antibodies specific to phosphorylated CREB at Ser¹³³ (-CREB) and to all forms of CREB (CREB). The star indicates the position of the phosphorylated ATF-2. (B) Increase of survival of Ba/F3 cells by cAMP-elevating agents. Viable cell number was measured by trypan blue staining after 18 h of incubation in cytokine-free medium without (−) or with (+) forskolin (FsK 100 μM) plus IBMX. Numbers of viable cells are presented as percentages of the number of cells initially seeded. Results shown are means ± standard deviations from two independent experiments performed in duplicate. (C) Decrease of cell viability by PKI. Ba/F3 cells were cultured in IL-3-containing medium supplemented with various doses of Myr-PKI (lanes 2 and 3, 10 and 20 μM). Cells cultured in cytokine-free medium were included as a negative control (lane 4). Numbers of viable cells under various treatments were determined and are presented as a percentage of the number of cells initially seeded. Shown are representative results from three independent experiments performed in duplicate.

FIG. 5

Dominant negative forms of CREB and PKA reverse the antiapoptotic effect of IL-3 in Ba/F3 cells. (A and C) Expression of CREBR287L (A) and PKArDN (C) in Ba/F3 cells. Vectors pcDNA-3, pcDNA-3/CREBR287L, and MT-REV_AB-neo (indicated as PKArDN) (7.5 μg of each construct) were transfected into Ba/F3 cells along with 2.5 μg of pCMV-GFP. Cells were allowed to express these proteins for 20 h in IL-3-containing medium without (A; C, top) or with (C, bottom) 2 μM cadmium chloride. Solid peaks indicate cells stained with control immunoglobulin G (IgG). Open peaks indicate cells stained with anti-CREB antibody (A) or anti-PKAr antibody (C). Open peaks in dotted lines are cells not expressing GFP (GFP −), and solid lines are cells expressing GFP (GFP +). Transfection efficiency was 10 to 12%. (B and D) Dominant negative forms of CREB (B) and PKAr (D) enhance apoptosis of Ba/F3 cells in the presence of IL-3. After transfection with the marker pCD16/7/Stop (2.5 μg) and the pcDNA-3 vector (7.5 or 18 μg), pcDNA-3/CREBR287L (18 μg), or MT-REV_AB-neo (7.5 μg), cells were cultured in IL-3 containing medium with or without 2 μM cadmium chloride for 20 h. The percentages of apoptotic (annexin V-positive) cells in the transfected populations (CD16 positive) were analyzed and quantified as described in Materials and Methods. One set of representative data from three independent experiments is shown.

FIG. 6

Inducible expression of CREBR287L reduces CBE binding and viability of Ba/F3 cells. (A) Two Ba/F3 derivatives (clones 4 and 8) stably overexpressing CREBR287L under a Tet-off inducible system were established. The expression level of HA-CREBR287L (indicated by an arrowhead) in clones 4 and 8 in medium with (+) or without (−) tetracycline (Tet) was analyzed by immunoblotting with anti-HA and anti-CREB antibodies. C3, a Ba/F3 derivative expressing pTRE vector alone, is included as a negative control. (B) Decrease of CBE-binding activity in cells overexpressing CREBR287L. Nuclear extracts were prepared from each cell clone under conditions (with or without tetracycline) as indicated, and EMSA was then performed as described for Fig. 1 with a CBE (top) or GATA-1 (as an internal control; bottom) probe. The slow- and fast-migrating complexes are indicated as S and F, respectively. The asterisk indicates the position of the GATA-1 complex. (C) Suppression of cell viability by CREBR287L in stable lines. Cells cultured in the presence (+) or absence (−) of tetracycline for 72 h were washed and seeded in medium without IL-3. Sixteen hours after IL-3 depletion numbers of viable cell in each culture group were determined and are presented as percentages of the number of cells initially seeded. The data presented are averages from three independent experiments done in duplicate. ∗, 0.001 < P < 0.01 compared to lane 3; ∗∗, 0.001 < P < 0.01 compared to lane 5.

FIG. 7

Expression of PKAc suppresses CWIA of Ba/F3 cells. (A) Expression of PKAc in transiently transfected Ba/F3 cells. Ba/F3 cells were transfected with pCMV vector or PKAc (7.5 μg of each), as indicated together with pCMV-GFP (2.5 μg). At 20 h after transfection, cells were analyzed for expression of PKAc by flow cytometry. The solid peaks indicate cells stained with rabbit normal serum, and the open peaks indicate cells stained with rabbit anti-PKAc antibody. The open peaks in dotted lines indicate the GFP-negative populations (GFP −), and solid lines show the GFP-positive populations (GFP +). Transfection efficiency was 10 to 12%. IgG, immunoglobulin G. (B) Suppression of CWIA by expression of PKAc. After transfection with the marker pCD16/7/Stop (2.5 μg) and pCMV vector or PKAc expression plasmid (2 μg of each) with or without CREBR287L (18 μg), cells were cultured in the absence of IL-3 for 20 h prior to being stained with anti-CD16 antibody and annexin V and analyzed by flow cytometry. Apoptotic percentages are as indicated at the upper right. Transfection efficiency was around 15%. One set of representative data from three independent experiments is shown.

FIG. 8

The CBE site in the bcl-2 promoter is important for IL-3 inducibility in hematopoietic cells. (A and B) Competition for ΔHLF (A) or CREB (B) binding to the CBE consensus by various CBE-like sequences. The consensus CBE sequence was labeled with ³²P and incubated with either ΔHLF (A) or CREB (B) in the presence of a 10-, 100-, or 1,000-fold molar excess (indicated as a triangle) of competitive (Comp.) CBE-like sequences. The genes from which the competitive CBE-like sites were are indicated above the triangles, and their sequences are shown in Table 1. The competitive EMSAs were repeated three times, and a representative set of data is shown. (C and D) Both CRE and CBE contribute to PKA (C) or IL-3 (D) stimulation of bcl-2 reporter activity. (C) Ba/F3 cells were electroporated with various bcl-2 reporters as indicated together with a PKAc (+PKA) or control (−PKA) expression vector. Twenty hours after transfection, cell lysates were prepared and analyzed by luciferase assays. (D) Same as panel C except that the PKA expression vector was omitted, and the transfected cells were left in medium with or without IL-3 prior to being analyzed for luciferase activity. WT, mC, mCB denote reporter plasmids Bcl-2WT, Bcl-2mC, and Bcl-2mCB, respectively (see Materials and Methods). The data presented are averages from three independent experiments performed in duplicate.