Reciprocal effects of C/EBPalpha and PKCdelta on JunB expression and monocytic differentiation depend upon the C/EBPalpha basic region

Twist-2 controls myeloid lineage development and function

Basic helix-loop-helix (bHLH) transcription factors play critical roles in lymphoid and erythroid development; however, little is known about their role in myeloid lineage development. In this study, we identify the bHLH transcription factor Twist-2 as a key negative regulator of myeloid lineage development, as manifested by marked increases in mature myeloid populations of macrophages, neutrophils, and basophils in Twist-2-deficient mice. Mechanistic studies demonstrate that Twist-2 inhibits the proliferation as well as differentiation of granulocyte macrophage progenitors (GMP) by interacting with and inhibiting the transcription factors Runx1 and C/EBPalpha. Moreover, Twist-2 was found to have a contrasting effect on cytokine production: inhibiting the production of proinflammatory cytokines such as interleukin-12 (IL-12) and interferon-gamma (IFNgamma) while promoting the regulatory cytokine IL-10 by myeloid cells. The data from further analyses suggest that Twist-2 activates the transcription factor c-Maf, leading to IL-10 expression. In addition, Twist-2 was found to be essential for endotoxin tolerance. Thus, this study reveals the critical role of Twist-2 in regulating the development of myeloid lineages, as well as the function and inflammatory responses of mature myeloid cells.

The MADS transcription factor Mef2c is a pivotal modulator of myeloid cell fate

Mef2c is a MADS (MCM1-agamous–deficient serum response factor) transcription factor best known for its role in muscle and cardiovascular development. A causal role of up-regulated MEF2C expression in myelomonocytic acute myeloid leukemia (AML) has recently been demonstrated. Due to the pronounced monocytic component observed in Mef2c-induced AML, this study was designed to assess the importance of Mef2c in normal myeloid differentiation. Analysis of bone marrow (BM) cells manipulated to constitutively express Mef2c demonstrated increased monopoiesis at the expense of granulopoiesis, whereas BM isolated from Mef2cΔ/− mice showed reduced levels of monocytic differentiation in response to cytokines. Mechanistic studies showed that loss of Mef2c expression correlated with reduced levels of transcripts encoding c-Jun, but not PU.1, C/EBPα, or JunB transcription factors. Inhibiting Jun expression by short-interfering RNA impaired Mef2c-mediated inhibition of granulocyte development. Moreover, retroviral expression of c-Jun in BM cells promoted monocytic differentiation. The ability of Mef2c to modulate cell-fate decisions between monocyte and granulocyte differentiation, coupled with its functional sensitivity to extracellular stimuli, demonstrate an important role in immunity—and, consistent with findings of other myeloid transcription factors, a target of oncogenic lesions in AML.

Transcriptional dysregulation during myeloid transformation in AML

The current paradigm on leukemogenesis indicates that leukemias are propagated by leukemic stem cells. The genomic events and pathways involved in the transformation of hematopoietic precursors into leukemic stem cells are increasingly understood. This concept is based on genomic mutations or functional dysregulation of transcription factors in malignant cells of patients with acute myeloid leukemia (AML). Loss of the CCAAT/enhancer binding protein-alpha (CEBPA) function in myeloid cells in vitro and in vivo leads to a differentiation block, similar to that observed in blasts from AML patients. CEBPA alterations in specific subgroups of AML comprise genomic mutations leading to dominant-negative mutant proteins, transcriptional suppression by leukemic fusion proteins, translational inhibition by activated RNA-binding proteins, and functional inhibition by phosphorylation or increased proteasomal-dependent degradation. The PU.1 gene can be mutated or its expression or function can be blocked by leukemogenic fusion proteins in AML. Point mutations in the RUNX1/AML1 gene are also observed in specific subtypes of AML, in addition to RUNX1 being the most frequent target for chromosomal translocation in AML. These data are persuasive evidence that impaired function of particular transcription factors contributes directly to the development of human AML, and restoring their function represents a promising target for novel therapeutic strategies in AML.

Difference in gene expression of macrophage between normal spleen and portal hypertensive spleen idendified by cDNA microarray

AIM: To identify the difference in gene expression of microphage (Mφ) between normal spleen and portal hypertensive spleen using cDNA microarrays and find new gene functions associated with hypersplenism in portal hypertension.

METHODS: The Biostar-H140s chip containing 14 112 spots of cDNAs were used to investigate the difference of the expression. The total RNA extracted from macrophages isolated from both normal spleen and portal hypertensive spleen was reversely transcribed to cDNA with the incorporation of fluorescent (cy3 and cy5) labeled dCTP to prepare the hybridization probes. After hybridization, the gene chip was scanned for the fluorescent intensity. The differentially expressed genes were screened. That was repeated three times, and only the genes which had differential expression in all three chips were considered to be associated with hypersplenism in portal hypertension.

RESULTS: Eight hundred and ninety-six, 1330 and 898 genes were identified to be differentially expressed in three chips, respectively. One hundred and twenty-one genes (0.86%) were identified to be differentially expressed in all three chips, including 21 up-regulated genes and 73 down-regulated genes. The differentially expressed genes were related to ionic channel and transport protein, cyclin, cytoskeleton, cell receptor, cell signal conduct, metabolism, immune, and so on. These genes might be related to the hypersplenism in portal hypertension.

CONCLUSION: The investigations based on cDNA microarray can screen differentially expressed genes of macrophages between normal spleen and portal hypertensive spleen, thus may provide a new idea in studying the pathogenesis of hypersplenism in portal hypertension.

Proteomic discovery of Max as a novel interacting partner of C/EBPalpha: a Myc/Max/Mad link

The transcription factor CCAAT/enhancer binding protein a (C/EBPalpha) is important in the regulation of granulopoiesis and is disrupted in human acute myeloid leukemia. In the present study, we sought to identify novel C/EBPalpha interacting proteins in vivo through immunoprecipitation using mass spectrometry-based proteomic techniques. We identified Max, a heterodimeric partner of Myc, as one of the interacting proteins of C/EBPalpha in our screen. We confirmed the in vivo interaction of C/EBPalpha with Max and showed that this interaction involves the basic region of C/EBPalpha. Endogenous C/EBPalpha and Max, but not Myc and Max, colocalize in intranuclear structures during granulocytic differentiation of myeloid U937 cells. Max enhanced the transactivation capacity of C/EBPalpha on a minimal promoter. A chromatin immunoprecipitation assay revealed occupancy of the human C/EBPalpha promoter in vivo by Max and Myc under cellular settings and by C/EBPalpha and Max under retinoic acid induced granulocytic differentiation. Interestingly, enforced expression of Max and C/EBPalpha results in granulocytic differentiation of the human hematopoietic CD34(+) cells, as evidenced by CD11b, CD15 and granulocyte colony-stimulating factor receptor expression. Silencing of Max by short hairpin RNA in CD34(+) and U937 cells strongly reduced the differentiation-inducing potential of C/EBPalpha, indicating the importance of C/EBPalpha-Max in myeloid progenitor differentiation. Taken together, our data reveal Max as a novel co-activator of C/EBPalpha functions, thereby suggesting a possible link between C/EBPalpha and Myc-Max-Mad network.

Myc target transcriptomes

The c-Myc oncogenic transcription factor plays a central role in many human cancers through the regulation of gene expression. Although the molecular mechanisms by which c-Myc and its obligate partner, Max, regulate gene expression are becoming better defined, genes or transcriptomes that c-Myc regulate are just emerging from a variety of different experimental approaches. Studies of individual c-Myc target genes and their functional implications are now complemented by large surveys of c-Myc target genes through the use of subtraction cloning, DNA microarray analysis, serial analysis of gene expression (SAGE), chromatin immunoprecipitation, and genome marking methods. To fully appreciate the differences between physiological c-Myc function in normal cells and deregulated c-Myc function in tumors, the challenge now is to determine how the authenticated transcriptomes effect the various phenotypes induced by c-Myc and to define how c-Myc transcriptomes are altered by the Mad family of proteins.

C/EBPalpha directs monocytic commitment of primary myeloid progenitors

C/EBPalpha is required for generation of granulocyte-monocyte progenitors, but the subsequent role of C/EBPalpha in myeloid lineage commitment remains uncertain. We transduced murine marrow cells with C/EBPalpha-estradiol receptor (ER) or empty vector and subjected these to lineage depletion just prior to culture in estradiol with myeloid cytokines. This protocol limits biases due to lineage-specific effects on developmental kinetics, proliferation, and apoptosis. Also, lowering the dose of estradiol reduced activated C/EBPalpha-ER to near the physiologic range. C/EBPalpha-ER increased Mac1(+)/Gr1(-)/MPO(-)/low monocytes 1.9-fold while reducing Mac1(+)/Gr1(+)/MPO(hi) granulocytes 2.5-fold at 48 hours, even in 0.01 microM estradiol. This pattern was confirmed morphologically and by quantitative polymerase chain reaction (PCR) assay of lineage markers. To directly assess effects on immature progenitors, transduced cells were cultured for 1 day with and then in methylcellulose without estradiol. A 2-fold increase in monocytic compared with granulocytic colonies was observed in IL-3/IL-6/SCF or GM-CSF, but not G-CSF, even in 0.01 microM estradiol. C/EBPalpha-ER induced PU.1 mRNA, and PU.1-ER stimulated monocytic development, suggesting that transcriptional induction of PU.1 by C/EBPalpha contributes to monopoiesis. A C/EBPalpha variant incapable of zippering with c-Jun did not induce monopoiesis, and a variant unable to bind NF-kappaB p50 stimulated granulopoiesis, suggesting their cooperation with C/EBPalpha during monocytic commitment.

Recruitment of nuclear factor Y to the inverted CCAAT element (ICE) by c-Jun and E1A stimulates basal transcription of the bone sialoprotein gene in osteosarcoma cells

Bone sialoprotein (BSP), a major protein in the extracellular matrix of bone, is expressed almost exclusively by bone cells and by cancer cells that have a propensity to metastasize to bone. Previous studies have shown that v-src stimulates basal transcription of bsp in osteosarcoma (ROS 17/2.8) cells by targeting the inverted CCAAT element (ICE) in the proximal promoter. To identify possible downstream effectors of Src we studied the effects of the proto-oncogene c-jun, which functions downstream of Src, on basal transcription of bsp using transient transfection assays. Increased expression of endogenous c-Jun induced by the tumor promoter 12-O-tetradecanoyl-phorbol 13-acetate and ectopic expression of c-Jun increased basal transcription of chimeric reporter constructs encompassing the proximal promoter by 1.5-3-fold in ROS 17/2.8 osteosarcoma cells, with more modest effects in a normal bone cell line, RBMC-D8. The effects of c-Jun were abrogated by mutations in the ICE box and by co-expression of dominant negative nuclear factor Y, subunit A (NF-YA). The increase in bsp transcription did not require phosphorylation of c-Jun and was not altered by trichostatin treatment or by ectopic expression of p300/CREB-binding protein (CBP) or mutated forms lacking histone acetyltransferase (HAT) activity. Similarly, ectopic expression of p300/CBP-associated factor (P/CAF), which transduces p300/CBP effects, or of HAT-defective P/CAF did not influence the c-jun effects. Surprisingly, E1A, which competes with P/CAF binding to p300/CBP, also stimulated BSP transcription through NF-Y independently of c-jun, p300/CBP, and P/CAF. Collectively, these studies show that c-Jun and E1A regulate basal transcription of bsp in osteosarcoma cells by recruiting the NF-Y transcriptional complex to the ICE box in a mechanism that is independent of p300/CBP and P/CAF HAT activities.

A dominant-negative mutant of C/EBPα, associated with acute myeloid leukemias, inhibits differentiation of myeloid and erythroid progenitors of man but not mouse

The CCAAT/enhancer binding protein alpha (C/EBPα) is an essential transcription factor for granulocytic differentiation. C/EBPα mutations are found in approximately 8% of acute myeloid leukemia (AML) patients. Most of these mutations occur in the N-terminal coding region, resulting in a frame shift and the enhanced translation of a dominant-negative 30-kDa protein, which may be responsible for the differentiation block observed in AML. To test this hypothesis, we introduced a cDNA encoding an N-terminal mutated C/EBPα (mut10) into primary hematopoietic progenitors using a retroviral vector. Expression of mut10 in human CD34+ cord blood cells dramatically inhibited differentiation of both myeloid and erythroid lineages. Immunohistochemical analysis demonstrated coexpression of both myeloid and erythroid markers in the immature transformed cells. Surprisingly, mut10 did not block myelocytic differentiation in murine progenitors but did alter their differentiation kinetics and clonogenicity. Experiments were performed to confirm that the differential effect of mut10 on murine and human progenitors was not due to species-specific differences in C/EBPα protein sequences, expression levels, or inefficient targeting of relevant cells. Taken together, our results underline the intrinsic differences between hematopoietic controls in mouse and human and support the hypothesis that mutations in CEBPA are critical events in the disruption of myeloid differentiation in AMLs. (Blood. 2004;103:2744-2752)

C/EBPalphap30, a myeloid leukemia oncoprotein, limits G-CSF receptor expression but not terminal granulopoiesis via site-selective inhibition of C/EBP DNA binding

Heterozygous mutations of the CEBPA gene are present in 5% of acute myeloid leukemia (AML) cases and often lead to the expression of an N-terminally truncated, 30 kDa isoform, C/EBPalphap30, from an internal translation start site. We have assessed the effect of C/EBPalphap30 on granulopoiesis utilizing C/EBPalphap30-ER, containing the estradiol receptor ligand-binding domain. In contrast to C/EBPalpha-ER, C/EBPalphap30-ER did not induce 32Dcl3 myeloid cell differentiation in IL-3. However, both isoforms, when expressed at high levels, were capable of inhibiting E2F activity in 32Dcl3 cells and of slowing their G1 to S progression. C/EBPalphap30 repressed expression of the endogenous G-CSF receptor several-fold. To facilitate investigation of the effect of C/EBPalphap30-ER on granulopoiesis downstream of G-CSF signalling, we coexpressed exogenous G-CSF receptor. C/EBPalphap30-ER/GR cells expressed several granulocytic markers in G-CSF and demonstrated nuclear maturation. Rat C/EBPalpha-ER and C/EBPalphap30-ER, expressed in 293T cells, bound the C/EBP site from the NE gene with similar affinity, as did human C/EBPalpha and C/EBPalphap30. In contrast, C/EBPalphap30 bound the C/EBP sites in the PU.1 or GR gene with 3-6-fold reduced affinity. Thus, the selective inhibition of GR expression by C/EBPalphap30-ER is due in part to its variable affinity for C/EBP sites. Variation in affinity for selected cis elements among isoforms may affect the biology of basic region-leucine zipper (bZIP) proteins.

Regulation of granulocyte and monocyte differentiation by CCAAT/enhancer binding protein α

CCAAT/enhancer binding protein alpha (C/EBPalpha)-ER induces 32Dcl3 neutrophilic differentiation and inhibits 32DPKCdelta maturation to macrophages in response to phorbol ester. In 32Dcl3 cells, C/EBPalpha-ER rapidly induces the PU.1 and C/EBPalpha RNAs even in the presence of cycloheximide, suggesting that these are direct C/EBPalpha genetic targets. C/EBPalpha strongly binds and modestly activates the murine PU.1 promoter via an evolutionarily conserved binding site. C/EBPalpha-ER variants incapable of binding DNA still slow G1 progression but do not induce differentiation. N-terminally truncated C/EBPalpha variants, including the p30 isoform expressed in a subset of AMLs, also retain the ability to slow 32D cl3 proliferation, whereas the C/EBPalpha(BRM2)-ER variant does not slow G1 progression, has a reduced capacity to induce early granulocytic markers, and does not induce terminal maturation. In 32DPKCdelta cells, C/EBPalpha-ER strongly inhibits endogenous or exogenous JunB induction, dependent upon the outer surface of the C/EBPalpha basic region, but does not inhibit c-Jun, PU.1, or C/EBPbeta expression. Exogenous JunB restores AP-1 DNA binding but does not overcome inhibition of monopoiesis by C/EBPalpha-ER. In summary, we propose that while C/EBPalpha is required for development of immature granulocyte-monocyte progenitors, C/EBPalpha subsequently inhibits monopoiesis, via inhibition of JunB express and via additional activities, and induces granulopoiesis, via induction of PU.1, C/EBPepsilon, and cell cycle arrest.

Cross-talk between regulators of myeloid development: C/EBPalpha binds and activates the promoter of the PU.1 gene

CCAAT/enhancer-binding protein (C/EBP)alpha and PU.1 are required for myelopoiesis. Examination of the murine PU.1 promoter revealed several potential C/EBP-binding sites. Gel-shift assay demonstrated that C/EBPalpha expressed in 293T cells bound the site centered at -68 most potently. C/EBPalpha from 32D cl3 myeloid cell nuclear extracts also bound this site strongly, and endogenous C/EBPbeta did so to a lesser extent, whereas these C/EBP isoforms bound the neutrophil elastase promoter with equal affinity. The -68 site in the murine PU.1 promoter is conserved in the human PU.1 promoter. Mutation of the -68 C/EBP-binding site in a -85/+152 promoter segment linked to the luciferase cDNA reduced promoter activity fourfold in 293T cells in the presence of cotransfected C/EBPalpha and twofold in 32D cl3 myeloid cells. Induction of endogenous PU.1 RNA by C/EBPalpha-estradiol receptor (ER) in the presence of cycloheximide is obviated by mutation of the C/EBPalpha DNA-binding domain, and chromosomal immunoprecipitation demonstrated specific interaction of C/EBPalpha and C/EBPalpha-ER with the PU.1 promoter. Finally, PU.1 RNA is reduced several-fold in immortalized C/EBPalpha (-/-) compared with (+/-) cells. Together, these findings indicate that C/EBPalpha binds and activates the endogenous PU.1 gene in myeloid cells. Induction of PU.1 by C/EBPalpha may account for increased levels of PU.1 in myeloid as compared with B lymphoid cells and in this way, may contribute to the specification of myeloid progenitors.