Introduction

Hematopoietic differentiation proceeds in a largely irreversible fashion and the role of transcription factors in regulating hematopoiesis has been well documented. This is particularly true for CCAAT/enhancer binding protein a (C/EBP), one of the lineage-specific transcription factors that is essential for commitment to and development of the granulocytic lineage.^{1, 2} Recent data have indicated that C/EBP may also regulate hematopoietic stem cell activity³ and act as a tumor suppressor gene in acute myeloid leukemias (AMLs), indicating an important role for C/EBP in the control of cellular proliferation in vivo.⁴ Inactivation of C/EBP is an important event in AML, and ectopic overexpression of C/EBP leads to differentiation and growth arrest in AML.⁵ It is therefore suggested that C/EBP has a crucial role in regulating the balance between cell proliferation and differentiation, which is crucial for lineage commitment of any cell type. These findings and data from our laboratory indicate that for AML to develop, the activity of C/EBP must be curbed by either mutations or antagonistic protein–protein interactions.

C/EBP can form protein–protein interactions with other bZIP and non-bZIP factors. Among them, c-Jun and PU.1,^{6, 7} E2F, p21, and cyclin-dependent kinases CDK2 and CDK4 have been well characterized.^{8, 9, 10} Thus, it has become increasingly clear that like most proteins, C/EBP might not work alone, but in association with other factors regulates gene transcription. However, studies involving protein–protein interactions of C/EBP at the global proteomic level are lacking. We therefore took advantage of high-throughput proteomics by mass spectrometry (LC-MS/MS) to identify proteins that specifically associate with C/EBP in vivo. In our screen, Max was identified as a novel interacting partner of C/EBP in addition to other new and known partners of C/EBP.

Max is a member of the basic region-helix–loop–helix-leucine zipper protein that belongs to a network of transcription factors, which includes the Myc and Mad families of protein (commonly referred to as a Myc–Max–Mad network).¹¹ The Myc–Max–Mad proteins can affect different aspects of cell behavior, including cell cycle, proliferation and differentiation, by modulating distinct target genes.^{12, 13, 14, 15} Max can form a homo- or a heterodimer and bind specifically to E-box DNA elements in target promoters (consensus CACGTG).^{16, 17} To function as transcriptional regulators, the members of the Myc and Mad families must heterodimerize with Max. Whereas Myc–Max activates transcription, Mad–Max and Mnt–Max repress transcription.^{18, 19, 20} Indirect evidences to the fact that C/EBP could be a part of the Myc–Max–Mad network do exist in the literature.^{21, 22} However, no direct evidence has been reported so far.

In this study, we have characterized the role of Max as an interacting partner of C/EBP. We show that Max is an important co-activator of C/EBP and the stable silencing of Max inhibits the differentiation-inducing potential of C/EBP. C/EBP and Max not only colocalize but also the heterocomplex is preferentially formed on the human C/EBP (hC/EBP) promoter in vivo during granulocytic differentiation, thereby contributing to increased transactivation and differentiation capacity of C/EBP.

Materials and methods

Transfection of human hematopoietic CD34+ progenitors

Human CD34+ hematopoietic cells were selected, using a magnetic CD34 selection kit system (Milteny Biotec, Bergisch, Gladbach, Germany), from small aliquots of leukapheresis products collected from either healthy donor or a patient undergoing stem/progenitor cell collection after granulocyte-colony stimulating factor treatment for non-hematologic malignancy at Klinikum Krollwitz Hospital Halle, Germany, following their informed consent. After magnetic selection, more than 85% of the cells expressed the CD34 antigen. An aliquot containing 5 10⁵ CD34+ cells was cultured in Iscove's modified Dulbecco's medium with 20% heat-inactivated fetal calf serum, 100 ng/ml Flt3-ligand, 100 ng/ml of stem cell factor, 100 ng/ml thrombopoietin, 100 ng/ml of interleukin-6 (IL-6) and 50 ng/ml of IL-3, 100 U/ml penicillin/streptomycin and 2 mM L-glutamine. The cells were transfected with various expression constructs using AMAXA nucleofection technology essentially as described by the manufacturer and analyzed for CD11b and CD15 expression by flow cytometry.

Cell lines, antibodies and treatments

Human myeloid cell lines U937 and K562-ER-C/EBP were cultured under standard conditions. -Estradiol and retinoic acid (RA) (Sigma-Aldrich, Munich, Germany) were used at a concentration of 1–5 M and 10^-6 M, respectively. The antibodies used in this study were purchased from Santa Cruz (Heidelberg, Germany); for C/EBP, SC-61 (14AA), SC-9315 (N-19) Max, SC-765 (C-124) and c-Myc, SC-42 (C-33) and Molecular Probes, Gmbh, Karlsruhe, Germany).

Immunoprecipitation and immunoblotting

The immunoprecipitation (IP) was performed from 500–1000 g nuclear extracts of U937 cells in an IP buffer (50 mM Tris pH 7.5, 150 mM NaCl, 0.5% NP-40, 0.25% sodium deoxycholate), followed by washing in the buffer (50 mM Tris pH 7.5, 0.1% NP-40, 0.05% sodium deoxycholate) with respective antibodies (Santa Cruz) and the corresponding IgGs as controls. A Western blot analysis was used to confirm the identity of immunoprecipitated and/or co-precipitated proteins as described previously.²³ Alternatively, the immunocomplexes were incubated with urea lysis buffer for further proteomic analysis.

Proteomic analysis: two-dimensional gel electrophoresis and protein identification by mass spectrometry

The proteomics methodology was used essentially as described recently by our group.³⁴

Transient transfections using AMAXA and effectene

Effectene transfection reagent (Qiagen, Gmbh, Hilden, Germany) and lipofectamine (Invitrogen, Gmbh, Karlsruhe, Germany) were used for transient transfections according to the manufacturer's instructions. Transient transfections were carried out with minimal promoter/luciferase construct, which has been derived from an oligo 5'-GATCCAGATTGCGCAATCG-3' by self-annealing, followed by ligation into a BamHI site of the thymidine kinase (TK) promoter and co-transfected with expression plasmids for hC/EBP, Renilla Luciferase-null and/or Max as described.²³ The Nucleofector kit (AMAXA, Gmbh, Cologne, Germany) was used essentially as described by the manufacturer. A 5 g portion of plasmid DNA constructs was used for each transfection and the transfection efficiency was analyzed using a plasmid with eGFP marker (2 g). For CD34+ and U937 cells, nucleofector solution kits used were VPA-1003 and VCA-1003 with nucleofection programs U-08 and V-01, respectively. The voltages are automatically adjusted according to the program and are essentially 110 V AC with a frequency of 50–60 Hz and a power consumption of 16 VA/fuse.

Immunofluorescence and flow cytometry

U937 cells (3 10⁵), under uninduced condition or induced with RA (Sigma-Aldrich), were cytocentrifuged on glass slides with coverslips, fixed using 1:1 methanol/acetone and permeabilized using 0.3% Triton X. After blocking in PBG (0.5% BSA, 0.045% Fish–gelatin in phosphate-buffered saline) containing 5% FBS, the fixed cells were incubated with anti-C/EBP (anti-goat; Santa Cruz), anti-Max (anti-rabbit; Santa Cruz) and anti-Myc (anti-mouse; Santa Cruz) antibodies, followed by incubation with corresponding Alexa Fluor 488 chicken anti-goat, Alexa Fluor 594 chicken anti-rabbit and anti-mouse IgG secondary antibodies (Molecular Probes) and 4',6'-diamidino-2-phenylindole dihydrochloride (DAPI, 1 g/ml) for 15 min. The cells were mounted in aqueous mounting medium and the images were acquired and analyzed using a fluorescence microscope ( 100, 60). Flow cytometry was performed for CD11b, CD14 and CD15 expression on Bectin Dikinson flow cytometer, using the supplied analysis software.

Quantitative real-time PCR

RNA isolation from CD34+ and U937 cells, transfected with different expression constructs, by TRIZOL (Invitrogen, Germany) was followed by cDNA synthesis using standard conditions. Equal amount of cDNA was taken for a subsequent quantitative real-time PCR (Q-RT-PCR) using the Quantitech SyBR Green PCR kit (Qiagen, Germany) in a Rotor-Gene RG-3000 (Corbett Research, Sydney, Australia). The delta ct value (ct) was then calculated from the given ct value by the formula ct=ct_sample-ct_control). The fold change was calculated as fold change=2^-ct. The following primer sequences were used: myeloperoxidase (MPO), 5'-TCG GTA CCC AGT TCA GGA AG-3' (forward) and 5'-CCA GGT TCA ATG CAG GAA GT-3' (reverse); neutrophilelastase (NE), 5'-TGC TCA ACG ACA TCG TGA TT-3' (forward) and 5'-CTC ACG AGA GTG CAG ACG TT-3' (reverse); GCSFR, 5'-AAG AGC CCC CTT ACC CAC TAC ACC ATC TT-3' (forward) and 5'-TGC TGT GAG CTG GGT CTG GGA CAC TT-3' (reverse); CD14, 5'-CAA CTT CTC CGA ACC TCA GC-3' (forward) and 5'-CCA GTA GCT GAG CAG GAA CC-3' (reverse).

Chromatin immunoprecipitation assay

Logarithmically growing and differentiating U937 cells (1 10⁸ cells) were fixed with formaldehyde (final concentration 1% (v/v)) in serum free RPMI-1640 medium, at 4°C for 1 h. Glycine was added to a final concentration of 0.125 M to stop cross-linking. Fixed cells were pelleted by centrifugation and sequentially washed and sonicated (five times for 20 s each) to make soluble chromatin. Samples of total chromatin were taken at this point to use as a positive control in the PCRs (input chromatin). Antibodies against C/EBP, Max and c-Myc were used overnight at 4°C. After serial elution, washing and cross-link reverse, the samples were extracted twice with phenol/chloroform and precipitated with ethanol overnight in the presence of 20 g glycogen as a carrier. DNA fragments were recovered by centrifugation, resuspended in ddH₂O, and used for PCR amplification. For detection of immunoprecipitated C/EBP promoter region, two primers, forward (5'-ACCGCTACCGACCACGTGGGCG-3') and reverse (5'-AGCACCTCCGGGTCGCGAATGG-3'), specific for a 280 bp region in the cellular C/EBP promoter that encompasses the C/EBP site were used for Q-RT-PCR amplification.

Results

Identification of Max, a heterodimeric partner of Myc, as a novel interacting protein of C/EBP

To identify interacting proteins of C/EBP in vivo under physiological conditions on a global level, we applied proteomics technique coupled with mass spectrometry using the IP conditions of endogenous C/EBP from myeloid U937 cells as a model system.

Under our experimental conditions, we could specifically immunoprecipitate endogenous C/EBP from the nuclear extracts of U937 cells (Figure 1a) and co-immunoprecipitate other endogenous proteins (as positive controls) such as c-Jun and CDK4 (Figure 1b and data not shown) that were not present in the isotype IgG control. Immunocomplexes were further processed for proteomic analysis. The protein spots excised from the 2D gels (Figure 1c, spots are numbered) were identified by MALDI-TOF MS. Additionally, the individual bands were excised from Coomassie/silver-stained sodium dodecyl sulfate-polyacrylamide gel electrophoresis gels (Figure 1d) and processed for LC-MS/MS. From both screens, we were able to reveal the identity of 10 proteins by MS, which specifically interact with C/EBP (Table 1). Among these proteins, we identified Max as one interacting partner of C/EBP. C/EBP was also identified by MS analysis of the corresponding band (Figure 1d), thereby serving as a control for our experimental setup. Proteins in other bands could not be determined because of the poor quality of the spectrum. The discovery of Max as a novel C/EBP partner is intriguing because of the role Max plays in switching of the complexes during myeloid differentiation.²⁴ We therefore selected Max for further functional and biological characterization.

Figure 1.

MS-based proteomics identifies proteins specifically interacting with C/EBP in vivo after its immunoprecipitation from myeloid U937 cells. (a) C/EBP IP from nuclear extracts of U937 cells and a corresponding immunoblotting (IB) with anti-C/EBP antibody to confirm the presence of C/EBP protein in the IP complex. In vitro-translated C/EBP (ivt, lane 1) was used as a positive control in the Western blot. (b) C/EBP IP and corresponding IB with anti-c-Jun antibody to show endogenous proteins co-precipitated with C/EBP. ivt c-Jun was used as a positive control for c-Jun. (c) Silver-stained 2D gels showing proteins specifically interacting with C/EBP. C/EBP was immunoprecipitated from nuclear extracts using anti-C/EBP antibody (anti-rabbit; Santa Cruz) and the immunocomplex separated in the first dimension by pH 4–7 IPGphor strips followed by their separation in the second dimension using 12% SDS-PAGE. As a specificity control, we used immunoprecipitation with IgG under similar conditions. (d) Silver-stained SDS-PAGE gels after IP with anti-C/EBP and anti-IgG. The bands were excised and peptide mixture after trypsin digestion was run on a reverse-phase high-pressure liquid chromatography and the peptides identified by MALDI-TOF-TOF (Applied Biosystems, Darmstadt, Germany).

Full figure and legend (142K)

Table 1 - MS results of the proteins interacting with C/EBP: MALDI-TOF Reflex III (Bruker Daltonics) and LC-MS/MS.

Full table

C/EBP and Max interact in a cellular setting: confirmation of proteomics data

To confirm the observed interaction of Max with CEBP by an alternative technique, we performed reciprocal immunoprecipitation. Our results demonstrate that C/EBP interacts with Max and vice versa (Figure 2a) in vivo, and thereby confirm proteomic results. It is important to note that for the same amounts of nuclear extracts used (5 and 10 g) as input controls, the levels of the two transcription factors are dramatically different, which is likely due to Max being more stable than C/EBP.

Figure 2.

In vivo interaction of C/EBP with Max confirmed by reciprocal IP involves the DNA-binding domain of C/EBP. (a) Reciprocal IPs: C/EBP and Max were immunoprecipitated (IP C/EBP, IP Max) from nuclear extracts of U937 cells by incubation with anti-C/EBP and anti-Max, respectively, and respective IgG as controls. The blot was first probed with anti-C/EBP antibody, stripped and reprobed with anti-Max antibody. (b) Basic region of C/EBP is involved in its interaction with Max. Schematic representation of wild-type hC/EBP and different mutants used in this study. TAD, transactivation domains 1 and 2; BR, basic region; LZ, leucine zipper domain; HLH, helix–loop–helix. (c) hC/EBP wild type and its mutants were transfected in 293T cells and co-transfected with wild-type Max expression plasmid. At 24 h post-transfection, the nuclear extracts were prepared and IP of Max performed for the samples followed by immunoblot for C/EBP or Max using anti-C/EBP and HA antibodies, respectively.

Full figure and legend (73K)

BR3 region of C/EBP is involved in its interaction with Max

To investigate the protein domains that might be involved in C/EBP–Max interaction, we performed co-immunoprecipitation studies using different mutants of C/EBP as shown. C/EBP and its various mutants (kind gift from Dr Alan Friedman; Figure 2b) were transiently transfected into 293 cells, and co-transfected with an expression plasmid for Max (a kind gift from Dr Dirk Eick) containing a carboxy-terminal HA tag.²⁵ Max was then immunoprecipitated from nuclear extracts using anti-Max antibody. The associated complexes were assayed by immunoblotting for C/EBP using anti-C/EBP antibody. Our results demonstrate that C/EBP could be co-immunoprecipitated when IP was performed using anti-Max antibody in samples in which wild-type C/EBP: wild-type Max, GZ/LZ C/EBP: wild-type Max and L1-2V C/EBP: wild-type Max were coexpressed (Figure 2c, lanes 4, 3, 1, respectively). However, C/EBP could not be co-immunoprecipitated in immunoprecipitated samples in which basic region mutant BR3-C/EBP: wild-type Max was co-expressed (Figure 2c, lane 2). We also show that Max could be specifically immunoprecipitated (as controls) with immunoblot for Max using anti-HA antibody (Figure 2c, lower panel). The relative expression of C/EBP mutants was the same (data not shown). These data show that the basic region of C/EBP is involved in its interaction with Max in a cellular setting. Furthermore, we observed that wild-type Max and its basic region mutants have the same ability to interact with C/EBP (Supplementary Figure S1a and b).

C/EBP and Max colocalize

Given the fact that C/EBP and Max are nuclear transcription factors and the observation that they interact in vivo, we next investigated the localization of these proteins by indirect immunofluorescence in myeloid U937 cells. We observed both endogenous C/EBP and Max to be localized in intranuclear structures (Figure 3a) and the overlay of the two images shows that both proteins colocalize in these intranuclear structures (Figure 3a,panel 4; yellow signal).

Figure 3.

Endogenous C/EBP-Max but not Myc–Max remains colocalized during granulocytic differentiation of U937 cells. (a) Indirect immunofluorescence staining for C/EBP (anti-goat; Santa Cruz), Max (anti-rabbit; Santa Cruz) and Myc (anti-mouse, Santa Cruz) using respective conjugated secondary antibodies (Molecular Probes). U937 cells were cytocentrifuged on glass slide cover slips, fixed with methanol/acetone, permeabilized with 0.3% Triton X stained with respective antibodies (Alexa Fluor, Molecular Probes) and DAPI. The morphology of the cells was visualized under fluorescence microscope (X60–100). Colocalization is demonstrated by the yellow signals. Indirect immunofluorescence staining for (b) C/EBP-Max and (c) Myc–Max using conjugated antibodies (Molecular Pobes) in U937 cells after RA treatment. (d) Immunoblot analysis showing expression of c-Myc, Max and C/EBP under RA-induced and uninduced conditions from various fractions. Blots were stripped and reprobed with specific antibody. Upper panel: whole-cell lysates; NF: nuclear fraction; CF: cytoplasmic fraction.

Full figure and legend (200K)

C/EBP–Max but not Myc–Max remains colocalized during granulocytic differentiation of myeloid U937 cells

We next investigated the effect on C/EBP–Max colocalization when the cells were triggered for granulocytic differentiation by RA for 24 h. We observed intranuclear staining with C/EBP and Max antibodies, and the overlay of the two images shows that both proteins remain colocalized even after RA treatment of the cells (Figure 3b, panel 4; yellow signal). As Max is associated with Myc, we also analyzed their localization in U937 cells. We observed that endogenous Myc–Max colocalize in the nucleus under uninduced condition (Figure 3a, panels 5 and 6). On the other hand, no intranuclear c-Myc signal could be detected after RA treatment (Figure 3c, panel 4; only green signal from Max). We next investigated the expression of c-Myc, Max and C/EBP before and after RA treatment from various fractions (whole-cell lysates, nuclear fraction (NF) and cytoplasmic fraction (CF)) by Western blotting, using specific antibodies (Figure 3d). Our results revealed that the c-Myc protein level was drastically decreased in all the three fractions (Figure 3d, upper and lower panels) by RA. However, C/EBP was undetectable in the CF and slightly increased in the NF by RA when analyzed by immunoblotting. Dot blot analysis revealed the presence of CEBP in the CF as well. This indicates that the concentration of C/EBP in the CF is quite low, so as not to be detected by immunoblotting (data not shown). Max, on the other hand, was relatively unchanged under induced and uninduced conditions. These data demonstrate that retention/colocalization of C/EBP–Max, and not Myc–Max heterocomplexes, in the nucleus might be important events during granulocytic differentiation of U937 cells.

Max enhances the ability of C/EBP to transactivate a minimal thymidine kinase promoter

To investigate the functional importance of C/EBP–Max interaction and their colocalization, we performed transient transfection assays in the fibroblast 293T and the myeloid U937 cells using a minimal TK promoter containing two CCAAT binding sites cloned upstream of the luciferase reporter gene. Transfection of a Max expression construct significantly enhanced the ability of C/EBP to transactivate a minimal TK promoter containing two CCAAT binding site in a dose-dependent manner (Figure 4a). In control experiments, no effect of Max on C/EBP activity was observed when promoter with no CCAAT binding sites was used, whereas C/EBP alone was able to transactivate the minimal promoter construct ninefold. Similar results were obtained with myeloid U937 cells (Figure 4b). Interestingly, co-transfection studies with the human 2200 bp C/EBPa promoter (which has intact E-box site and no CCAAT site) revealed that C/EBP alone was unable to transactivate the promoter, whereas, co-transfection of Max led to a significant increase in the promoter activity (Figure 4c). It is important to point out that Max itself does show some activation.

Figure 4.

Max enhances the transactivation capacity of C/EBP in transient transfection assays. (a, b) Transient transfection in 293T and U937 cells with a reporter construct of a minimal TK promoter with CEBP binding sites only p(CEBP)2TK and expression plasmids for hC/EBP and Max. pTK (without CEBP sites) was used as control. Luciferase activities were measured 24 h after transfection and the values normalized by using Renilla luciferase PRL0. (c) Transient transfection in 293T cells with a 2200 bp hC/EBP promoter showing increased promoter activation when Max is co-expressed. Histogram on the right shows promoter activation by hC/EBP on a minimal promoter, used as a positive control in this experiment.

Full figure and legend (43K)

C/EBP and Max associate in vivo: a Myc–Max–Mad link

To further elucidate the mechanism by which Max augments the transcriptional activity of C/EBP, we hypothesized that Max might associate with the hC/EBP promoter in vivo because similar to C/EBP, Max also possesses a DNA binding basic region. To test this possibility, we performed quantitative radioactive and non-radioactive chromatin immunoprecipitation (ChIP) in U937 cells (Figure 5). Chromatin was subjected to IP by using antibodies directed against C/EBP, c-Myc and Max. The presence of C/EBP promoter was detected by amplifying a promoter region using primers specific for a 280 bp region in the C/EBP promoter that encompasses the CACGTG site (commonly referred to as E-box; Figure 5a). The E-box is conserved in the human and mouse C/EBP promoter (Figure 5a). We observed that under normal physiological conditions (uninduced), endogenous c-Myc and Max appeared on C/EBP promoter and there was undetectable endogenous C/EBP occupancy on the hC/EBP promoter (Figure 5b). IP using an isotype-matched IgG served as a negative control.

Figure 5.

Max is associated at the hC/EBP promoter in vivo and Max–C/EBP associate strongly during granulocytic differentiation. A ChIP assay was performed on logarithmically growing and RA-treated U937 cells, and the precipitated chromatin was PCR-amplified using specific primers. (a) Comparison of the human and mouse C/EBP promoters encompassing a consensus CACGT sequence, commonly referred to as E-box and known to be occupied by Myc+Max heterodimers. (b) In vivo occupancy by Myc and Max at the hC/EBP promoter in logarithmically growing and (c) by Max-C/EBP in RA-treated U937 cells. Input: Radioactive and Q-RT-PCR performed on total chromatin. The histograms beneath show the Q-RT-PCR average ct values from two independent experiments normalized with the control sample. (d, e) ChIP assay using GAPDH promoter and human TERT promoter as controls with a non-radioactive RT-PCR. (f) Sheared DNA from U937 cells following 10 sonication pulses shows the optimal size range for IP (200–1000 bp). Lane: 1, unsheared; lanes: 2 and 3, sheared DNA.

Full figure and legend (101K)

We next investigated the affect on heterocomplex formation at the hC/EBP promoter upon differentiation by RA. We observed that both Max and C/EBP appeared on C/EBP promoter and in fact, more C/EBP was associated with the promoter in the context of chromatin upon differentiation induction (Figure 5c, lane 4). The amount of Max bound to the promoter was fairly constant. DNA recovery was quantified as a percentage of the total input chromatin (lanes 5–7). Q-RT-PCR confirmed this observation and the histograms shown represent the average values from two independent experiments (Figure 5c, lower panel). A promoter without the CACGTG site, such as GAPDH promoter (Figure 5d), was used as a negative control for C/EBP and Max occupancy and hTERT promoter (Figure 5e) as a positive control for Myc and Max interaction on the CACGTG site (E-box). The size of the DNA fragments before and after sonication is also shown (Figure 5f). Thus, C/EBP and Max associate in vivo in the context of chromatin and are associated together more strongly on the hC/EBP promoter when the cells are induced towards granulocytic differentiation.

Overexpression of Max and C/EBP promotes differentiation along the granulocytic pathway in human hematopoietic CD34+ cells

We next asked whether interaction of Max with C/EBP is biologically important for C/EBP functions. Hence, we performed overexpression studies using three different experimental systems: human hematopoietic CD34+ cells, estradiol-inducible K562-C/EBP-ER cells and U937 cells. Our results revealed that overexpression of Max or C/EBP alone in CD34+ cells leads to a significant increase in the proportion of CD11b+ (Figure 6a, dot plot 44 vs 20%) and CD15+ (Figure 6a, dot plot 29 vs 13%) cells compared with the mock-transfected control, respectively. The histograms represent the average values from three different experiments, and the viable cell count data (Trypan blue staining) under different conditions are also shown for days 1 and 4 (Figure 6b). Q-RT-PCR in these cells revealed increased GCSF receptor expression (Figure 6c). Similar results were observed with U937 and K562-C/EBP-ER cells (Supplementary Figure S2a and data not shown). The morphology of the cells was observed to correlate with the surface marker expression (Supplementary Figure S2b). Q-RT-PCR in U937 cells for various granulocytic/ monocytic markers was also performed to complement the fluorescence-activated cell sorting results (Supplementary Figure S2c).

Figure 6.

Overexpression of Max induces differentiation along granulocytic pathway in human hematopoietic CD34+ cells. (a) The expression plasmids for human C/EBP and Max were transfected into human hematopoietic CD34+ cells by using AMAXA technology. The surface expression of CD11b and CD15 was analyzed by flow cytometry at day 4. The histograms underneath represent data from three different experiments. (b) Trypan blue staining, showing the number of viable cells under different conditions. (c) Q-RT-PCR for GCSF receptor expression under the conditions shown from two experiments. (d) Stable silencing of Max by shRNA inhibits C/EBP-induced differentiation in human hematopoietic CD34+ cells. The expression plasmid for human C/EBP and/or expression Arrest shRNA_Max plasmid (Open Biosystems) were transfected into human hematopoietic CD34+ cells or U937 cells by using the AMAXA technology. After their selection in puromycine, the cells were analyzed for the surface expression of CD15 by flow cytometry and the data shown as dot plot with percentage of positive cells representative of one experiment. shRNA control was also used in all the experiments and is shown. The histograms represent the data from three different experiments. (e) A Western blot for Max using anti-Max antibody showing silencing of Max at the protein level by shRNA_MAX. The blot was stripped and reprobed with C/EBP antibody. (f) Model, a summary of our data showing the importance of Max as a co-activator of C/EBP in the differentiation of myeloid progenitors. Enforced expression of Max and CEBP induces differentiation along the granulocytic pathway, and stable silencing of Max inhibits CEBP-induced differentiation.

Full figure and legend (300K)

Stable silencing of Max by short hairpin RNA reduces the differentiation-inducing capacity of C/EBP in human hematopoietic CD34+ cells

If Max is a biologically important co-activator of C/EBP, silencing of Max should inhibit differentiation induction by C/EBP. To address this, we performed RNA interference experiments in human hematopoietic CD34+ cells and myeloid U937 cells (Supplementary Figure S2d) by using short hairpin RNA (shRNA) against Max (cat. no. RHS1764-9690535; Open Biosystems, Heidelberg, Germany) and control shRNA (cat. no. RHS1707; Open Biosystems). Cells were transfected with expression plasmids for C/EBP alone and/or co-expressed with shRNA against Max, control shRNA, and the cells cultured in media containing puromycine. After selection, the cells were analyzed for granulocytic differentiation, using CD15 expression as a marker. Our results revealed that C/EBP alone induces granulocytic differentiation (CD15+) five- to six-fold as compared with the mock-transfected CD34+ (Figure 6d). Coexpression of Max shRNA led to a significant decrease in CD15+ population (about twofold), whereas control shRNA did not lead to any significant reduction in CD15+ population (Figure 6d, compare histograms). The reduction of Max protein level with shRNA was confirmed by Western blotting and Max shRNA did not affect the expression of C/EBP (Figure 6e). In conclusion, we propose a model shown as Figure 6f. Thus, Max is important for C/EBP-mediated effects on granulocytic differentiation and might have an important role in stem cell development.

Discussion

It has become increasingly clear that interaction of C/EBP with other nuclear proteins plays an important role not only in lineage commitment and differentiation in the hematopoietic system but also in the pathogenesis of AML. Although the lineage commitment decision by C/EBP was proposed by our laboratory to involve the functional inactivation of the myeloid master regulator PU.1 and/or its co-activator c-Jun through protein–protein interactions,^{6, 7} relatively little is known about how C/EBP interacts with other nuclear proteins to activate gene transcription. The results presented in this article provide evidence that Max, a heterodimerization partner of Myc, is a novel, functionally and biologically important co-activator of CEBP. C/EBP and Max not only colocalize but also the heterocomplex is preferentially formed on the hC/EBP promoter during granulocytic differentiation, thereby contributing to increased transactivation and differentiation capacity of C/EBP.

We used MS-based proteomic analysis as a means of identifying the interacting partners of C/EBP, utilizing IP of C/EBP from myeloid U937 cells as a model system. U937 cells are a good model system for studying myeloid differentiation in general, as they are bipotential and can be differentiated into granulocytic lineage by RA and in particular, with respect to the functions of C/EBP, as a threefold level of C/EBP protein (above the level of endogenous C/EBP) in U937 cells is sufficient for their granulocytic differentiation.²⁶ In addition to nine other proteins (see Table), we identified Max, an essential heterodimerization partner of Myc,¹⁶ as a novel interacting partner of C/EBP in our screen (Figure 1). The discovery of Max as a novel C/EBP partner is intriguing because of the role Max plays in switching of the complexes during myeloid differentiation.²⁴ Of particular importance is the fact that transgenic mice carrying an inserted transgene encoding Max have been shown to exhibit a 50- to 60-fold elevation of blood neutrophils.²⁷ Additionally, Max is an essential heterodimerization partner of Myc family members to regulate transcription¹¹ and c-Myc is an important target of C/EBP.²⁶ We confirmed the in vivo interaction of C/EBP with Max by IP technique and showed that the basic DNA-binding region of C/EBP is involved in this interaction, as the mutant of CEBP (C/EBP BR3), which lacks DNA-binding region, could not be co-precipitated with Max (Figure 2). C/EBP BR3 carries mutations in four amino acids, residues Arg297, Lys298, Arg300 and Lys302.²⁸ Of these, only Arg300 is expected to contact DNA. Neither the BR3 nor the Leu12Val variants bind DNA, suggesting that interaction with Max is likely via Arg297, Lys298 and/or Lys302. Arg297 is known to participate in the interaction between C/EBP and E2F.⁸ Further study is required to pin point the exact amino acid involved in the C/EBP and Max interaction.

The endogenous C/EBP and Max proteins are not distributed evenly throughout the nucleoplasm (Figure 3), but are localized in intranuclear structures within the nucleus. These structures represent, presumably, centromeres, which are chromosomal structures associated with intranuclear chromosome positioning and cell cycle regulation. Interestingly, C/EBP is associated with cell cycle regulation.^{29, 30} In other cell systems, such as pituitary progenitor GHFT1-5 cells, C/EBP has been shown to concentrate at chromatin surrounding the centromeres.³¹ The observation that C/EBP–Max but not Myc–Max remain colocalized during granulocytic differentiation (Figure 3) indicates that these intranuclear structures (centromeres) are selectively targeted by C/EBP–Max during granulocytic differentiation. We observed the occupancy of the hC/EBP promoter by Max in vivo under physiological conditions, and recruitment of more C/EBP whereas Max is retained on the promoter during granulocytic differentiation. It is possible that the C/EBP-Max heterocomplex regulates the balance of acetylated histones to modify chromatin structure at the hC/EBP promoter and lead to transcriptional activation, as was shown by our results. In fact, TIP60, a histone acetyl transferase, was identified as an interacting partner of C/EBP to regulate histone acetylation at the hC/EBP promoter in an alternative approach (Bararia et al., manuscript submitted for publication). To our knowledge, this is a first report showing occupancy of the hC/EBP promoter by Max in vivo.

The occupancy by Max of the hC/EBP promoter raises a possibility that Myc could also form a part of the complex under physiological conditions, as Max requires dimerization with Myc for efficient DNA binding. In fact, it was shown that purified Myc+Max heterodimers form stable complexes on the mouse C/EBP promoter that includes the USF binding site.²¹ The USF DNA recognition site CACGTG (which is the same as the E-box, occupied by Myc–Max) is found in both the human and the mouse C/EBP promoter, and the USF binding site (for HLH-bZIP) is crucial for activation of the hC/EBP promoter by C/EBP.³² Our colocalization and ChIP data (Figures 4 and 5) and the data that C/EBP is co-precipitated with Myc IP (unpublished observation) support this Myc–Max link. Thus, it is tempting to speculate that C/EBP exists in association with the Myc–Max–Mad network to regulate differentiation under cellular settings. Given that the C/EBP–Max heterocomplex is formed on hC/EBP promoter, specifically during granulocytic differentiation, this would mean that the balance between such complexes, under the influence of growth and differentiation signals, could be an important part of a molecular switch that is regulating genes important for growth and differentiation.

By using overexpression studies, we have demonstrated that enforced expression of C/EBP and Max in human hematopoietic CD34+ cells induces granulocytic differentiation. The role of C/EBP in the transition from CMPs to GMPs in myeloid progenitors has been recently characterized.³ The role of Max in inducing granulocytic differentiation indicates that Max can activate myeloid differentiation program either independent of C/EBP or in association with it. In vivo interaction and retention of C/EBP–Max heterocomplex in myeloid cells (Figures 2, 4 and 5) and inhibition of differentiation-inducing capacity of C/EBP by stable silencing of Max using shRNA against MAX in CD34+ cells (Figure 6) suggest CEBP–Max association likely plays an important role in this process of myeloid progenitor differentiation. A very recent data from Alan Friedman's group has shown the role of C/EBP in monopoiesis.³³ This means that the commitment decisions do not necessarily depend upon a single transcription factor but, in fact, on a number of cooperating factors.

In summary, we conclude that Max is a biologically and functionally important and relevant interacting partner of C/EBP and has important co-activator functions for C/EBP-induced granulocytic differentiation in myeloid progenitors.

References

1. Scott LM, Civin CI, Rorth P, Friedman AD. A novel temporal expression pattern of three C/EBP family members in differentiating myelomonocytic cells. Blood 1992; 80: 1725–1735. | PubMed | ISI | ChemPort |
2. Zhang DE, Zhang P, Wang ND, Hetherington CJ, Darlington GJ, Tenen DG. Absence of granulocyte colony-stimulating factor signaling and neutrophil development in CCAAT enhancer binding protein alpha-deficient mice. Proc Natl Acad Sci USA 1997; 94: 569–574. | Article | PubMed | ChemPort |
3. Zhang P, Iwasaki-Arai J, Iwasaki H, Fenyus ML, Dayaram T, Owens BM et al. Enhancement of hematopoietic stem cell repopulating capacity and self-renewal in the absence of the transcription factor C/EBP alpha. Immunity 2004; 21: 853–863. | Article | PubMed | ChemPort |
4. Pabst T, Mueller BU, Zhang P, Radomska HS, Narravula S, Schnittger S et al. Dominant-negative mutations of CEBPA, encoding CCAAT/enhancer binding protein-alpha (C/EBPalpha), in acute myeloid leukemia. Nat Genet 2001; 27: 263–270. | Article | PubMed | ISI | ChemPort |
5. Pabst T, Mueller BU, Harakawa N, Schoch C, Haferlach T, Behre G et al. AML1-ETO downregulates the granulocytic differentiation factor C/EBPalpha in t(8;21) myeloid leukemia. Nat Med 2001; 7: 444–451. | Article | PubMed | ISI | ChemPort |
6. Rangatia J, Vangala RK, Treiber N, Zhang P, Radomska H, Tenen DG et al. Downregulation of c-Jun expression by transcription factor C/EBPalpha is critical for granulocytic lineage commitment. Mol Cell Biol 2002; 22: 8681–8694. | Article | PubMed | ChemPort |
7. Reddy VA, Iwama A, Iotzova G, Schulz M, Elsasser A, Vangala RK et al. Granulocyte inducer C/EBPalpha inactivates the myeloid master regulator PU.1: possible role in lineage commitment decisions. Blood 2002; 100: 483–490. | Article | PubMed | ISI | ChemPort |
8. Porse BT, Pedersen TA, Xu X, Lindberg B, Wewer UM, Friis-Hansen L et al. E2F repression by C/EBPalpha is required for adipogenesis and granulopoiesis in vivo. Cell 2001; 107: 247–258. | Article | PubMed | ISI | ChemPort |
9. Timchenko NA, Wilde M, Nakanishi M, Smith JR, Darlington GJ. CCAAT/enhancer-binding protein alpha (C/EBP alpha) inhibits cell proliferation through the p21 (WAF-1/CIP-1/SDI-1) protein. Genes Dev 1996; 10: 804–815. | PubMed | ISI | ChemPort |
10. Wang H, Iakova P, Wilde M, Welm A, Goode T, Roesler WJ et al. C/EBPalpha arrests cell proliferation through direct inhibition of Cdk2 and Cdk4. Mol Cell 2001; 8: 817–828. | Article | PubMed | ISI | ChemPort |
11. Grandori C, Cowley SM, James LP, Eisenman RN. The Myc/Max/Mad network and the transcriptional control of cell behavior. Annu Rev Cell Dev Biol 2000; 16: 653–699. | Article | PubMed | ISI | ChemPort |
12. Atchley WR, Fernandes AD. Sequence signatures and the probabilistic identification of proteins in the Myc–Max–Mad network. Proc Natl Acad Sci USA 2005; 102: 6401–6406. | Article | PubMed | ChemPort |
13. Holzel M, Kohlhuber F, Schlosser I, Holzel D, Luscher B, Eick D. Myc/Max/Mad regulate the frequency but not the duration of productive cell cycles. EMBO Rep 2001; 2: 1125–1132. | Article | PubMed | ChemPort |
14. Hurlin PJ, Queva C, Koskinen PJ, Steingrimsson E, Ayer DE, Copeland NG et al. Mad3 and Mad4: novel Max-interacting transcriptional repressors that suppress c-myc dependent transformation and are expressed during neural and epidermal differentiation. EMBO J 1995; 14: 5646–5659. | PubMed | ISI | ChemPort |
15. Nilsson JA, Maclean KH, Keller UB, Pendeville H, Baudino TA, Cleveland JL. Mnt loss triggers Myc transcription targets, proliferation, apoptosis, and transformation. Mol Cell Biol 2004; 24: 1560–1569. | Article | PubMed | ISI | ChemPort |
16. Blackwood EM, Luscher B, Kretzner L, Eisenman RN. The Myc:Max protein complex and cell growth regulation. Cold Spring Harb Symp Quant Biol 1991; 56: 109–117. | PubMed | ChemPort |
17. Kato GJ, Lee WM, Chen LL, Dang CV. Max: functional domains and interaction with c-Myc. Genes Dev 1992; 6: 81–92. | PubMed | ISI | ChemPort |
18. James L, Eisenman RN. Myc and Mad bHLHZ domains possess identical DNA-binding specificities but only partially overlapping functions in vivo. Proc Natl Acad Sci USA 2002; 99: 10429–10434. | Article | PubMed | ChemPort |
19. Luscher B. Function and regulation of the transcription factors of the Myc/Max/Mad network. Gene 2001; 277: 1–14. | Article | PubMed | ISI | ChemPort |
20. Kretzner L, Blackwood EM, Eisenman RN. Myc and Max proteins possess distinct transcriptional activities. Nature 1992; 359: 426–429. | Article | PubMed | ISI | ChemPort |
21. Legraverend C, Antonson P, Flodby P, Xanthopoulos KG. High level activity of the mouse CCAAT/enhancer binding protein (C/EBP alpha) gene promoter involves autoregulation and several ubiquitous transcription factors. Nucleic Acids Res 1993; 21: 1735–1742. | PubMed | ISI | ChemPort |
22. D'Alo' F, Johansen LM, Nelson EA, Radomska HS, Evans EK, Zhang P et al. The amino terminal and E2F interaction domains are critical for C/EBP alpha-mediated induction of granulopoietic development of hematopoietic cells. Blood 2003; 102: 3163–3171. | Article | PubMed | ChemPort |
23. Zada AA, Singh SM, Reddy VA, Elsasser A, Meisel A, Haferlach T et al. Downregulation of c-Jun expression and cell cycle regulatory molecules in acute myeloid leukemia cells upon CD44 ligation. Oncogene 2003; 22: 2296–2308. | Article | PubMed | ISI | ChemPort |
24. Ayer DE, Eisenman RN. A switch from Myc:Max to Mad:Max heterocomplexes accompanies monocyte/macrophage differentiation. Genes Dev 1993; 7: 2110–2119. | PubMed | ISI | ChemPort |
25. Kohlhuber F, Hermeking H, Graessmann A, Eick D. Induction of apoptosis by the c-Myc helix–loop–helix/leucine zipper domain in mouse 3T3-L1 fibroblasts. J Biol Chem 1995; 270: 28797–28805. | Article | PubMed | ChemPort |
26. Johansen LM, Iwama A, Lodie TA, Sasaki K, Felsher DW, Golub TR et al. c-Myc is a critical target for c/EBPalpha in granulopoiesis. Mol Cell Biol 2001; 21: 3789–3806. | Article | PubMed | ISI | ChemPort |
27. Metcalf D, Lindeman GJ, Nicola NA. Analysis of hematopoiesis in max 41 transgenic mice that exhibit sustained elevations of blood granulocytes and monocytes. Blood 1995; 85: 2364–2370. | PubMed | ISI | ChemPort |
28. Liu H, Keefer JR, Wang QF, Friedman AD. Reciprocal effects of C/EBPalpha and PKCdelta on JunB expression and monocytic differentiation depend upon the C/EBPalpha basic region. Blood 2003; 101: 3885–3892. | Article | PubMed | ChemPort |
29. Timchenko NA, Harris TE, Wilde M, Bilyeu TA, Burgess-Beusse BL, Finegold MJ et al. CCAAT/enhancer binding protein alpha regulates p21 protein and hepatocyte proliferation in newborn mice. Mol Cell Biol 1997; 17: 7353–7361. | PubMed | ISI | ChemPort |
30. Timchenko NA, Wilde M, Darlington GJ. C/EBPalpha regulates formation of S-phase-specific E2F–p107 complexes in livers of newborn mice. Mol Cell Biol 1999; 19: 2936–2945. | PubMed | ISI | ChemPort |
31. Schaufele F, Enwright III JF, Wang X, Teoh C, Srihari R, Erickson R et al. CCAAT/enhancer binding protein alpha assembles essential cooperating factors in common subnuclear domains. Mol Endocrinol 2001; 15: 1665–1676. | Article | PubMed | ChemPort |
32. Timchenko N, Wilson DR, Taylor LR, Abdelsayed S, Wilde M, Sawadogo M et al. Autoregulation of the human C/EBP alpha gene by stimulation of upstream stimulatory factor binding. Mol Cell Biol 1995; 15: 1192–1202. | PubMed | ISI | ChemPort |
33. Wang D, D'costa J, Civin CI, Friedman AD. C/EBP{alpha} directs monocytic commitment of primary myeloid progenitors. Blood 2006; 108: 1223–1229. | Article | PubMed | ChemPort |
34. Peer Zada AA, Geletu HM, Pulikkan JA, Müller CT, Reddy VA, Christopeit M et al. Proteomic analysis of acute promyelocytic leukemia: PML-RARalpha leads to decreased phosphorylation of OP18 at Ser63. Proteomics 2006 (in press).

Acknowledgements

We thank Dr Dirk Eick for providing Max expression plasmids, Dr Robert Eisenman for in vitro-translatable Max plasmid and Dr Alan Friedman for C/EBP constructs. This work was financially supported by Grant F03/04 of the Deutsche-Jose-Carreras Leukemia Stiftung, Muenchen Germany to Peer Zada AA.

Supplementary Information accompanies the paper on the Leukemia website (http://www.nature.com/leu)