C/EBPbeta: a major PML-RARA-responsive gene in retinoic acid-induced differentiation of APL cells

Abstract

In acute promyelocytic leukemia (APL), the translocation t(15;17) induces a block at the promyelocytic stage of differentiation in an all-trans-retinoic acid (ATRA)-responsive manner. Here we report that upon treatment with ATRA, t(15;17) cells (NB4) reveal a very rapid increase in protein level and binding activity of C/EBPbeta, a C/EBP family member, which was not observed in an ATRA-resistant NB4 cell line. We further provide evidence that ATRA mediates a direct increase of C/EBPbeta, only in PML-RARA (promyelocytic leukemia-retinoic acid receptor alpha)-expressing cells. In addition, transactivation experiments indicate that the PML-RARA fusion protein, but not PML-RARA mutants defective in transactivation, strongly transactivates the C/EBPbeta promoter. These results suggest that PML-RARA mediates ATRA-induced C/EBPbeta expression. Finally, we demonstrate the importance of C/EBPbeta in granulocytic differentiation. We show that not only does C/EBPbeta induce granulocytic differentiation of non-APL myeloid cell lines independent of addition of ATRA or other cytokines, but also that C/EBPbeta induction is required during ATRA-induced differentiation of APL cells. Taken together, C/EBPbeta is an ATRA-dependent PML-RARA target gene involved in ATRA-induced differentiation of APL cells.

Fig. 1. ATRA induces changes in the C/EBP DNA binding profile during promyelocytic differentiation. EMSA analysis of nuclear extracts from NB4 (A) and NB4-R2 cells (B) treated with ATRA. A double-stranded oligonucleotide including the C/EBP-binding site from the G-CSF receptor was radiolabeled with ³²P and incubated in the absence (free probe; lane 1) or presence of 10 µg of nuclear extracts from cells which were untreated (lanes 3–6), or treated for 24 (lanes 7–10), 48 (lanes 11–14) and 72 h (lanes 15–18) with ATRA. For competition analysis, a 50-fold molar excess of unlabeled oligonucleotide was used (lane 2). Antisera against C/EBPα (α), C/EBPβ (β) or C/EBPε (ε) were added to the binding reaction. Supershifted complexes are indicated with an arrow. Free probe was present in excess in each lane (shown for A only, but similar results apply to the EMSA shown in B and Figure 2).

Fig. 2. C/EBPβ DNA binding activity is rapidly enhanced during ATRA-mediated differentiation of APL cells. (A) NB4 cells were treated with 1 µM ATRA for 2 (lanes 4–6), 4 (lanes 7–9) and 8 h (lanes 10–12). EMSAs were performed using the same conditions as in Figure 1. DNA–C/EBP complexes were characterized by supershift experiments using antisera against C/EBPα (α), C/EBPβ (β) or C/EBPε (ε). (B) C/EBPα and C/EBPβ binding profile in primary APL cells. Untreated cells and cells treated for 15 h with 1 µM ATRA were subjected to EMSA analysis using the same conditions as in (A).

Fig. 3. The increase in C/EBPβ protein after ATRA treatment reflects the increase in DNA binding activity. Nuclear extracts from ATRA-sensitive (NB4) and ATRA-resistant (NB4-R2) cells were tested for C/EBPβ and C/EBPε expression by western blotting following ATRA treatment. The membranes were incubated with antiserum against C/EBPβ (upper panel) and then stripped and incubated with an antiserum against C/EBPε (lower panel). The arrows designate the bands corresponding to the 43 kDa C/EBPβ isoform, as well as the 32 and 27 kDa C/EBPε isoforms. On the left of the figure, specificity of the C/EBPβ antibody is shown using extracts from untransfected CV1cells and from CV1 cells transfected with a human C/EBPβ cDNA.

Fig. 4. The presence of PML–RARA is required to rapidly induce C/EBPβ RNA. (A) Northern blot analysis of total RNA isolated from NB4, NB4-R2 and HT93 cells untreated or treated with 1 µM ATRA for the indicated time. The 28S rRNA level is shown as a loading control. (B) Comparison by northern blot of C/EBPβ induction after ATRA treatment between PML–RARA- and PLZF–RARA-expressing cells. PML–RARA and PLZF–RARA were induced respectively by pre-treating the U937-PR9 cells or the U937-B412 cells with zinc for 15 h. (C) Northern blot analysis of total RNA from NB4, U937-PR9 or zinc-pre-treated U937-PR9 cells untreated or treated with 1 µM ATRA for 1, 2 or 4 h in the absence (–) or presence (+) of cycloheximide (CHX). The 28S rRNA level is shown as a loading control. (D) Level of PML–RARA protein after a short period of ATRA treatment is revealed by western blot analysis of NB4 nuclear extracts using an anti-RARA antibody. (E) Northern blot analysis of the half-life of C/EBPβ mRNA in uninduced and ATRA-induced NB4 cells. NB4 cells were pre-treated with 1 µM ATRA for 4 h and then actinomycin D (10 µg/ml) was added for the indicated times to the cells prior to extraction of RNA. Results are presented as RNA level normalized against actin with respect to time.

Fig. 5. C/EBPβ promoter is specifically activated by PML–RARA in the presence of ATRA. (A) Schematic representation of promoter constructs used in transient transfections shown in (B) and (C). The white rectangle labeled ‘TATAA’ represents the putative TATA box of the C/EBPβ promoter. (B) U937-PR9 cells were transiently transfected with the different deletions of the C/EBPβ reporter constructs. PML–RARA induction was obtained by pre-treating U937-PR9 cells with zinc for 15 h. Relative luciferase activity in the absence (white bar) or presence (black bar) of 2 µM ATRA is presented. U937 cells (without PML–RARA) were treated with zinc for 15 h to control for the effect of zinc treatment. The RAREx3LUC reporter vector was used as a control for ATRA treatment. (C) C-terminal PML–RARA deletion mutants were tested for their ability to activate the C/EBPβ promoter in U937-PR9 cells. The BetaP0.3LUC reporter was co-transfected in non-zinc-stimulated U937-PR9 cells with expression vector for PML–RARA wild-type and mutants (lanes 3–5). Relative luciferase activity in the absence (white bar) or presence (black bar) of 2 µM ATRA is presented.

Fig. 6. C/EBPβ directs differentiation of a multipotent hematopoietic cell line. (A) β-Estradiol induces nuclear localization of the C/EBPβ-ER fusion protein followed by granulocytic differentiation. (i) Stably transfected K562 cells were subjected to immunofluorescence using an anti-C/EBPβ antibody after treatment for 24 h with vehicle only (left panel) or β-estradiol (right panel). Cell maturation was evaluated by morphology (ii) and by NBT reduction assay (iii) following 48 h of treatment with vehicle only (left panel) or β-estradiol (right panel). (B) Cell surface expression of CD11b was determined by FACS. Cells transfected with empty ER vector or with C/EBPβ-ER were treated with vehicle only or with β-estradiol for 72 h. (C) RNA expression of the G-CSF receptor (upper panel) or C/EBPε (lower panel) was assessed using northern blot analysis. As indicated in the figure, cells infected with empty vector (ER) or C/EBPβ-ER were treated for different times with β-estradiol.

Fig. 7. Inhibition of ATRA-induced granulocytic differentiation by a dominant-negative C/EBP. U937-PR9 (left panel) and NB4 (right panel) cells were infected with a virus expressing the dominant- negative form of C/EBP, A-C/EBP (bottom two panels) or mock virus (top panels). The infected cells were treated with ATRA to induce differentiation. CD11b expression in the GFP-negative cell population (bottom panels) or GFP-positive cell population (top two panels) was examined by flow cytometry 72 h after induction of differentiation.

Fig. 8. Reduction of C/EBPβ expression induces reduction of ATRA-induced differentiation of NB4 cells. (A) C/EBPβ expression was evaluated by northern blot after 4 h of ATRA treatment in two different pools of cells transfected with the RNAi vector MSCV-U6-beta (pool1 and pool2) in comparison with cells transfected with an empty vector MSCV-U6 (ev). Cell maturation after 5 days of ATRA treatment was evaluated by morphology using MGG staining (B) and/or by NBT reduction assay (C). Results are presented as the percentage of NBT-positive cells and correspond to a count of 500 cells.