c-Myc Is Essential but Not Sufficient for c-Myb-mediated Block of Granulocytic Differentiation*

The c-myb proto-oncogene plays a central role in hematopoiesis and encodes a major translational product of 75 kDa. c-Myb is highly expressed in immature hematopoietic cells, and its expression is down-regulated during terminal differentiation. Deregulated expression of c-Myb has been shown to block terminal differentiation of hematopoietic cells. Here we have studied the mechanism of action and the nature of target genes through which c-Myb mediates the block in differentiation of 32Dcl3 murine myeloid cells. We show that the ectopic overexpression of c-Myb in 32Dcl3 cells results in the overexpression of c-Myc. However, enforced expression of c-Myc in 32Dcl3 cells did not alter the normal pattern of differentiation. In addition, expression of dominant-negative mutants of c-Myc relieved c-Myb-mediated block in differentiation. These results led us to conclude that c-myc is a target gene of c-Myb and activation of the c-myc gene is a necessary event in Myb-mediated transformation. However, c-Myc expression alone is inadequate to elicit the phenotypic effects seen with Myb-mediated block in differentiation of myeloid cells, suggesting that activation of additional transcriptional targets by c-Myb plays a critical role in this process.

The hematopoietic system is composed of a variety of cells that range from fully differentiated nonproliferating cells to pluripotent stem cells with extensive capacity to self-renew, proliferate, and differentiate (1). In response to a demand by the body for mature blood cells, the stem cells generate a group of progenitor cell population, that express specific cytokine receptors on their surface and proliferate or differentiate in response to the cognate cytokine binding (1). It is becoming increasingly evident that the proto-oncogene c-myb plays the role of a master regulator of hematopoietic cell development. Thus, in mammalian cells, c-myb is highly expressed in immature hematopoietic cells, and its expression is down-regulated during terminal differentiation of these cells (2)(3)(4)(5)(6)(7). Constitutive expression of c-Myb has been shown to block Me 2 SOinduced terminal differentiation of erythroleukemic cells (8) as well as GCSF-or IL 1 -6-induced terminal differentiation of mouse myeloid cell lines (9 -12). In addition, inhibition of c-Myb expression using antisense oligonucleotides has been shown to cause growth arrest of several hematopoietic cell lines and interfere with normal human hematopoiesis in vitro (13)(14)(15). Targeted disruption of c-myb locus results in neonatal death of mice because of a failure of fetal hepatic hematopoiesis (16). While the role of c-Myb in hematopoiesis is well established, its mechanism of action is yet to be defined.
The myb gene product codes for a major translational product of 75 kDa, which is localized to the nucleus and acts as a mediator of transcriptional regulation (7,(17)(18)(19). Like most transcription factors, the protein encoded by this gene exhibits a modular structure, with an N-terminal DNA binding domain, a central transactivation domain and a C-terminal negative regulatory domain (reviewed in Ref. 20). Many target genes have been identified for c-Myb, which include genes such as mim-1, TCR␤, the neutrophil elastase gene, the homeobox gene GBX2 as well as several proto-oncogenes such as c-kit and c-myc (reviewed in Refs. 20 -22). Of these the most interesting is the c-myc gene, which also codes for a nuclear transcription factor and has been shown to be intimately associate with cell growth, differentiation, and apoptosis (reviewed in Refs. 23 and 24). The c-Myc protein belongs to the family of transcription factors that contain the basic helix-loop-helix leucine zipper (bHLH-LZ) domains at the C-terminal end, which mediate the DNA binding and homo-and heterodimer formation (reviewed in Ref. 25). Thus the c-Myc protein was found to require the formation of heterodimers with its partner, Max, for sequencespecific DNA binding as well as for biological activity (reviewed in Ref. 25). While the c-myc and c-myb genes are co-expressed in many hematopoietic cells, the c-myc gene expression is seen more ubiquitously than that of c-myb, suggesting that c-myb does not regulate c-myc expression in all cell types (reviewed in Refs. 20 and 26).
The interrelationship between c-myb and c-myc was best studied in the myeloid leukemic M1 cell line, which is a transformed mouse myeloid cell line that proliferates autonomously in a factor-independent manner and forms tumors in nude mice (12,27). However, this cell line was found to undergo terminal differentiation and growth arrest, when cultured in the presence of IL-6 or leukemia inhibitory factor (LIF) (28 -30). Under conditions that allow continuous proliferation, these cells were found to express high levels of c-Myc and c-Myb, and the expression of these two proto-oncogenes is down-regulated following treatment with IL-6 or LIF, which induce terminal differentiation (31). In addition, it was found that constitutive expression of c-Myb or c-Myc in these cells results in a block to IL-6-or LIF-induced terminal differentiation (4,12). However, while constitutive expression of c-Myb in these cells lead to a block in terminal differentiation at an early stage, similar expression of c-Myc was found to result in a block leading to an intermediate stage myeloid cells, suggesting that the two protooncogenes mediate differentiation by different mechanisms (4,12).
However, in a recent study, it was observed that conditional expression of a Myb-estrogen receptor fusion protein (Myb-ER) in IL-6-treated M1 cells resulted in the restoration of expression of c-Myc expression (that was down-regulated by IL-6) and restored the proliferative activity of IL-6-treated M1 cells (27). Similarly, expression of a dominant-negative mutant of c-Myb in a mouse myeloid leukemic cell line, RI-4-11 resulted in the down-regulation of c-Myc expression leading to the suggestion that c-Myc may be solely responsible for Myb-mediated proliferation (27).
While these studies conducted in the M1 cell line are very informative for our understanding of the role of myc and myb in the differentiation of myeloid cells along the monocytic pathway, it is at present unclear whether they play a similar role in the differentiation of myeloid cells along the granulocytic pathway. In addition, these studies suffer from the problem that M1 cells are tumorigenic and, unlike normal hematopoietic cells, proliferate in a factor-independent manner. To gain an understanding of the relationships that exist between myb and myc in myeloid cell differentiation along the granulocytic pathway, we investigated the role of these two genes in GCSF-induced terminal differentiation of 32Dcl3 cells. These cells, derived from normal mouse bone marrow, have been found to be strictly dependent on IL-3 for growth and undergo terminal differentiation when placed in an IL-3-free medium containing GCSF (32,33). However, when the cells were incubated in the absence of IL-3 or GCSF, they were found to become arrested in the G 1 phase of the cell cycle and rapidly enter apoptotic pathways loosing viability in 48 -96 h (34). We and others (9 -11) had previously shown that overexpression of v-Myb or c-Myb in these cells results in a block to their ability to terminally differentiate into granulocytes in the presence of G-CSF. This cell line was particularly useful for studies aimed at understanding the effects of oncogenes and proto-oncogenes on myeloid differentiation as it is not predisposed to develop factor-independent clones (34). Results presented in this communication show that while ectopic overexpression of c-Myb in 32Dcl3 cells (32D/Myb) resulted in a block to the GCSF-induced terminal differentiation of these cells, ectopic overexpression of c-Myc (32D/Myc) failed to achieve the same effect. However, ectopic overexpression of c-Myb was found to result in elevated expression of c-Myc, and expression of dominant-negative mutants of c-Myc in 32D/Myb cells was found to result in a block to Myb-mediated effects on the terminal differentiation program of these cells. Similarly, in M1 cells, c-Myc dominantnegative mutants were found to relieve c-Myb-mediated block in terminal differentiation. These results led us to conclude that c-myc is a target of c-Myb, and activation of the c-myc gene is a necessary event in Myb-mediated transformation. However, ectopic overexpression of c-Myc expression alone was inadequate to elicit the phenotypic effects seen with Myb-mediated block in differentiation of myeloid cells along the granulocytic pathway, suggesting that activation of additional transcriptional targets by c-Myb is necessary for this process.

MATERIALS AND METHODS
Construction of Plasmids-The c-myc dominant-negative constructs, mycRX and max2RX, were generously provided by Dr. Bruno Amati and described previously (35). The constructs LK444A (36) and LK444A/c-myc (28) were kindly provided by Dr. Dan Liebermann. The constructs pMT-neo and pMT/c-myb have been described previously (37). The pOPRSVI-Puro/MCS was constructed by replacing the neomycin cassette in pOPRSVI/MCS (Stratagene) with a puromycin cassette (38). The myc dominant-negative constructs, mycRX and max2RX, were subcloned into pOPRSVI-Puro/MCS to express these constructs in 32Dcl3 cells.
Cell Lines and Culture Conditions-The murine myeloid progenitor cell line, 32Dcl3 (32,33), was maintained in Isocove's modified Dulbecco's medium (IMDM) supplemented with 10% fetal bovine serum, 0.5% penicillin/streptomycin, and 10% WEHI3B cell-conditioned medium as a source of IL-3 (39). Since all the cell lines established had metallothionein promoter-based constructs, 100 M ZnCl 2 was also used in all experiments. The M1 cell line was maintained in RPMI 1640 with 10% horse serum and 0.5% penicillin/streptomycin.
Growth and Differentiation Assays-To induce granulocyte differentiation, parental or transfected 32Dcl3 cells were washed twice in IL-3-free medium and plated at a density of 1 ϫ 10 5 cells/ml in IMDM containing 10% fetal bovine serum and 10% GCSF (32). The viability and proliferation of the cell cultures were monitored for the indicated days. Morphological analysis of GCSF-treated cells was performed using an aliquot of cells that was cytospun and stained with May-Grunwald-Geimsa stain.
The antibiotic-resistant M1 cells were induced to differentiate with conditioned medium containing IL-6. After 4 days, cytospin smears of the IL-6-treated cells were prepared and stained with Wright-Giemsa stain.
Transient Transfection and Luciferase Assay-For transfection into QT6 fibroblasts, cells were seeded into 6-well plates at a density of 0.5 ϫ 10 5 cells/well. The following day, DNA were transfected by calcium phosphate method (Invitrogen) using the manufacturer's protocol. In each transfection, 5 g of reporter plasmid (LK444A/c-myc) and 5 g of effector plasmid were transfected along with 0.5 g of pRL-CH110 as an internal standard. The effector plasmid, pODC (228 -281), was constructed by ligating firefly luciferase gene with the ODC promoter region Ϫ228 to Ϫ281 (51). The internal control plasmid, pRL-CH110, was constructed by ligating the Renilla luciferase gene to the RSV promoter. For transient transfections with the myc dominant-negatives, 15 g of the corresponding dominant-negative construct was also added to the DNA mixture above to get a final ratio of 1:3 between myc transcriptional activator and dominant-negative transcriptional suppressors. Following incubation for 24 h, cells were washed and incubated in fresh medium for additional 24 h. After 48 h of post-transfection, cells were lysed in 500 l/well of Passive Lysis buffer (Promega). Luciferase activity was assayed using Dual-Luciferase Reporter Plasmid System (Promega). Firefly luciferase activities were normalized against Renilla luciferase activity to determine relative luciferase activity, and the activation fold was obtained by setting the value of the empty vector control as 1.0.
Establishment of Stable Cell Lines Expressing Transgenes-32Dcl3 cells expressing pMT-Neo (empty vector) and pMT/c-myb have been described previously (37). For stable transfection of 32Dcl3 cells, exponentially growing 32Dcl3 cells were electroporated with various linearized plasmids DNAs using a Gene-Pulser (BioRad) at a pulse of 230 V, 960 mF. The surviving cells were selected in 1 mg/ml of G418 (pMT-neo based constructs) or in 1 mg/ml of G418 and 2 g/ml of puromycin (pMT-neo/pOPRSVI-Puro-based constructs) for 2-3 weeks.
For M1 cells, the cDNA encoding for c-Myb was subcloned into the retroviral vector MSCVneo, while the cDNAs encoding for MycRX and Max2RX were subcloned into MSCVpac (38). The vectors were transfected into packaging cell line Phenix together with the expression vector VSV-G (Stratagene). Vector, MycRX, and Max2RX retroviral supernatants were harvested to infect M1 cells. The infected cells were selected in puromycin (2 g/ml). The puromycin-resistant cells were further infected with c-Myb retroviral supernatant. The infected cells were then selected in puromycin and G418 (0.4 mg/ml).
Northern Blot Analysis-Total RNA from each cell line in the presence of IL-3 was isolated and purified using the Trizol Reagent (Invitrogen). 20 g of total RNA was used per sample, and Northern blot analysis was performed as described previously (37). Full-length c-myb, c-myc, mycRX, and max2RX cDNAs were used to detect c-myb, c-myc, mycRX, and max2RX mRNA transcripts. As an internal control for RNA loading, gels were stained with ethidium bromide to compare levels of 18 S/28 S ribosomal RNAs across lanes.
Western Blot Analysis-Anti-Myc and anti-Max primary antibodies were purchased from Santa Cruz. To analyze protein products of the transfected c-myb gene in 32Dcl3 cells, normalized amounts of proteins from each cell lysate were separated by SDS-polyacrylamide gel electrophoresis (37), and separated proteins were transferred to a nitrocellulose membrane in transfer buffer (10 mM CAPS, 10% methanol, pH 10.5). The filter was blocked with 5% nonfat milk in TBS-T buffer (10 mM Tris, pH 7.5, 100 mM NaCl, 0.1% Tween-20) for 1 h at room temperature, incubated with primary anti-c-Myb polyclonal antibody (37) overnight at 4°C and washed three times in TBS-T solution. The secondary antibody reaction was performed by incubating the filters with horseradish peroxidase-conjugated anti-rabbit Ig (Amersham Biosciences) and washed in a similar manner as was described for the primary antibody reaction. The Amersham Biosciences ECL detection system was used for visualization as specified by the manufacturer. To detect protein products of c-Myc, MycRX and Max2RX, anti-c-Myc N262, anti-Max C124, and anti-c-Myc 9E10 (Santa Cruz Biotechnology) primary antibodies and 3% nonfat milk were used in the protocol described above.
Analysis of DNA Fragmentation-DNA fragments released from 10 ϫ 10 6 cells were extracted and separated by electrophoresis in agarose gels (40). After experiments, cells were lysed in a buffer containing 0.5% Triton X-100, 50 mM Tris, 10 mM EDTA, pH 7.4, and 200 mg/ml proteinase K. Cell lysates were then incubated at 42°C for 90 min, treated with 0.2 mg/ml RNase A and incubated at 37°C for an additional 30 min and centrifuged at 14,000 ϫ g for 20 min. The resulting supernatant was then extracted twice with phenol/chloroform (1:1). DNA fragments were precipitated with one-tenth volume of 3 M NaOAc and 2.5 volumes of ethanol at Ϫ20°C overnight and resuspended in 10 mM Tris/HCl, 1 mM EDTA, pH 8.0. 2 mg of DNA were then analyzed on a 2% agarose gel.
Viability Assays-Mock-transfected and c-myc-transfected 32Dcl3 cells were analyzed for cell viability in the absence of IL-3 at different time points. At the indicated time points, viable and non-viable cells were identified by trypan blue exclusion/inclusion, counted in triplicates on a hemocytometer and percentage of viable cells calculated.

Constitutive Expression of c-Myb and c-Myc in 32Dcl3 Cells-
The role of c-Myb on myeloid cell growth and differentiation has been studied previously (37). To study the effect of c-Myc on myeloid differentiation along the granulocytic pathway and compare it with that of c-Myb, we used the 32Dcl3 cell line. This cell line is derived from normal mouse bone marrow and was found to be strictly dependent on IL-3 for proliferation (33,41). In the absence of IL-3, the cells undergo death via apoptosis over a period of 24 -72 h. Upon addition of GCSF, these cells terminally differentiate into granulocytes over a period of 10 -12 days (10,32,33). The linearized c-myb cDNA construct in pMT-Neo (under the control of metallothionein promoter) and c-myc cDNA construct in LK444A (under the control of actin promoter) were transfected into 32Dcl3 cells by electroporation. As negative controls, empty vector DNAs, pMT-Neo, and LK444A, was similarly introduced into 32Dcl3 cells. Following selection with G418, mass cultures were established from stable transfections. The derived cell lines were tested for the expression of each transgene. In the case of pMT-Neo based constructs, the expression levels were analyzed in the presence of Zn 2ϩ , which acts as an inducer of transcription from the metallothionein promoter.
The profiles of mRNA induction in these cell lines are shown in Fig. 1, A and B. The parental 32Dcl3 cells (32D/wild-type) and those transfected with empty vector DNA (32D/pMTneo) showed the presence of a 3.4-kb endogenous c-myb band (Fig.  1A). In cells transfected with c-myb (32D/pMT/c-myb in Fig.  1A), transgenic c-myb expression was seen as a 2.3-kb transcript. Similarly, the parental 32Dcl3 cells (32D/wild-type) and the cells transfected with empty vector DNA (32D/LK444A) showed the presence of a 2.4-kb endogenous c-myc band (Fig.  1B). In cells transfected with c-myc (32D/LK444A/c-myc/1 and 32D/LK444A/c-myc/2 in Fig. 1B), transgenic c-myc expression was seen as a 1.4-kb transcript. The smaller size of the transgenic transcripts are due to the absence of the 3Ј-untranslated sequence, which was deleted during the construction of these expression plasmids. To definitely demonstrate expression of c-Myb and c-Myc proteins from the transgenes, we compared the protein levels following incubation of 32D cells in the pres-ence of GCSF for 10 days. It was shown previously that, in 32Dcl3 cells, the endogenous levels of c-Myb and c-Myc proteins remain relatively high until day 6 -8 of GCSF treatment followed by a downward regulation of mRNA and protein, such that they become undetectable by day 10 of GCSF treatment (9,10). Taking advantage of this observation, we analyzed the levels of c-Myb and c-Myc proteins on the 10th day of GCSF treatment in the transfected cells to determine the levels of the transgenic proteins expressed in the absence of the endogenous protein product. As shown in Morphological analysis of the cells treated with GCSF is presented in Fig. 3. Cells transfected with empty vector plasmids (32D/pMT-Neo and 32D/LK444A), when treated with GCSF, showed a progressive increase in the appearance of mature forms, which included the myelocytes, metamyelocytes, and granulocytes. By day 10, the entire population was completely replaced by terminally differentiated cells, consisting of metamyelocytes and granulocytes (Fig. 3). In contrast, 32D/cmyb cells showed little evidence of progression to a differentiated phenotype and continued to proliferate indefinitely in GCSF-containing medium as myelocytes/promyelocytes. Interestingly and in agreement with data presented in Fig. 2, both the vector-transfected cells and 32D/c-myc cells progressed through the normal pattern of differentiation when treated with GCSF and by day 10 were terminally differentiated into granulocytes. Again, c-Myb and c-Myc had different effects on 32D cell differentiation program. Whereas ectopic expression of c-Myb completely blocks the ability of these cells to differentiate in response to GCSF, c-Myc by itself apparently had no effect on 32D cell differentiation.
Effect of Ectopic Overexpression of c-Myc on the Apoptotic Program of 32Dcl3 Cells-Since c-Myc does not appear to have any phenotypic changes on 32Dcl3 cell differentiation, we then investigated if the overexpressed c-Myc protein is in fact functional in these cells. It has been previously demonstrated that ectopic overexpression of c-Myc in 32Dcl3 cells results in accelerated apoptotic death of cells following IL-3 withdrawal (34). c-Myc has also been shown to induce apoptosis in M1 myeloid leukemia cells following treatment with TGF␤1 (42). As reported by other groups, the 32D/c-myc cells showed accelerated kinetics of cell death (Fig. 4A). This observation could be fur-ther verified by examining the degradation of DNA into oligonucleosomal fragments, which is shown in Fig. 4B. In empty vector-transfected cells, the onset of DNA ladder formation did not occur for 24 h and the most intense ladder formation was seen at the 48-h time point. In contrast, in 32Dcl3 cells transfected with c-myc, the DNA ladders appeared much earlier with high intensity ladders appearing around 18 h following IL-3 withdrawal.
Induction of c-myc by c-myb in 32Dcl3 Cells in Response to GCSF-As noted earlier, when 32Dcl3 cells are treated with GCSF, the expression of c-myc is detected for 6 -8 days. Thereafter the expression of c-myc declines and by day 10 is detectable only at minimal levels. Using Atlas Gene Arrays (Clontech), we had noticed high levels of c-myc mRNA transcript in 32D/c-myb cells even at day 10 of GCSF treatment (data not shown). Sequence analysis of the c-myc promoter/enhancer region had earlier revealed the presence of multiple Myb-binding sites in this region and in vitro transcriptional transactivation assays had shown that c-Myb can mediate transcriptional transactivation of this promoter leading to the suggestion that c-myc could be a target gene of c-Myb (43)(44)(45). Recently, the regulation of resident chromosomal copy of c-myc by c-Myb was demonstrated in M1 myeloid leukemia cells where c-Myb prevented IL-6 induced growth arrest and restored expression of c-Myc in nearly terminally differentiated M1 cells (27).
To determine the expression of c-myc in 32D/c-myb cells in the presence of IL-3 and GCSF, we analyzed the protein lysates from these cells by Western blot. As shown in Fig. 5A, in the presence of IL-3, parental, mock-transfected and 32D/c-myb cells had approximately equal levels of c-Myc protein expression. However, when the cells were incubated in GCSF for 10 days, the expression of c-Myc was markedly down-regulated in wild-type and empty vector-transfected cells while its expression was elevated in cells that were transfected with c-myb expression vectors (Fig. 5B). It should be noted that after 10 days of incubation in GCSF, the parental and empty vectortransfected cells were terminally differentiated into granulocytes, while the cells transfected with c-Myb expression plasmid were found to be promyelocytes/metamyelocytes. Thus, as seen in M1 leukemia cells, ectopic expression of c-Myb induces To address the question as to whether or not c-Myc is necessary for c-Myb-mediated block in 32D cell differentiation, we constitutively expressed c-Myc dominant-negatives mutants, MycRX and Max2RX, in 32D/c-myb cells. These dominantnegative mutants were previously constructed based on the observation that the biological and biochemical activity of Myc is dependent on its ability to heterodimerize with Max, which facilitates the binding of the Myc protein to DNA via the bHLH-LZ motif (35). As depicted in Fig. 6A, the dominantnegative mutants were generated by reciprocal exchange of HLH-LZ regions. Thus MycRX contains the HLH-LZ region of Max and Max2RX contains the HLH-LZ region of Myc. MycRX sequesters wild-type Myc into inactive complexes whereas Max2RX forms stable binding complexes with wild-type Max (46). These mutants were found to behave as strong dominant suppressors of Myc activity by decreasing the levels of functional endogenous Myc-Max complexes (35).
To further verify that the c-Myc dominant negatives sup-press the transcriptional activity of c-Myc, we performed transactivation assays using QT6 fibroblasts. As a reporter, we used a fragment of ODC promoter, pODC (228 -281) shown to have maximal transactivation by c-Myc (51). As depicted in Fig. 6B, exogenous c-myc gives ϳ2-fold transactivation over that of empty vector alone. A high level of activity was observed in empty vector-transfected cells due to the presence of high level of endogenous c-myc in these cells. When c-myc was transfected along with the c-myc dominant-negative constructs, the transactivation activity of c-Myc was suppressed and found to almost the same as that observed for empty vector alone. Thus the dominant-negative mutants of c-myc inhibit its transcriptional activity.
The linearized mycRX and max2RX constructs in pOPRSVI-Puro/MCS (under the control of RSV promoter) were transfected into the c-Myb-expressing 32Dcl3 cells via electroporation and their expression by Northern (Fig. 6, C and D) and Western blot analysis (Fig. 6, E and F).
Effect of c-Myc Dominant Negatives on c-Myb-induced Proliferation of 32D Cells-Following the verification of the expression of c-Myc dominant negatives in 32D/c-myb cells, we tested the effects of constitutive expression of these Myc mutants on c-Myb-mediated block in GCSF-induced terminal differentiation of 32Dcl3 cells. As described before, these cells were removed from IL-3-containing medium, washed twice, and cultured in the presence of GCSF. The growth curves are shown in Fig. 7. 32D/c-myb cells transfected with empty vector (32D/pOPRSVI-MCS/1 and 32D/pOPRSVI-MCS/2) continued to proliferate beyond day 10 as seen in 32D/c-myb cells. In contrast, 32D/c-myb cells transfected with c-Myc dominantnegative mutants (32D/MycRX/1 and 32D/MycRX/2 in Fig. 7A; 32D/Max2RX/1 and 32D/Max2RX/2 in Fig. 7B) underwent several rounds of cell division until about day 8 -9 after which they ceased proliferation. The growth pattern of these cells was very similar to that observed in parental 32D cells. Thus, inhibition of the activity of c-Myc in 32D/c-myb cells results in a reduction of the proliferative potential conferred by c-Myb. Fig. 8 shows the morphological analysis of cells treated with GCSF. As expected, 32D/c-myb cells transfected with empty vector (32D/pMT-c-myb/ pOPRSVI/MCS/1 and 32D/pMT-c-myb/pOPRSVI/MCS/2) were blocked in their progression to a differentiated phenotype and continued to proliferate in GCSF containing medium at the myelocytic/promyelocytic stage. Interestingly, 32D/c-myb cells transfected with c-Myc dominant-negative mutants (32D/ pMT-c-myb/pOPRSVI/MycRX/1 and 32D/pMT-c-myb/pOPRSVI/ MycRX/2; 32D/pMT-c-myb/pOPRSVI/Max2RX/1 and 32D/pMTc-myb/pOPRSVI/Max2RX/2) resumed normal pattern of differ- entiation, and by day 10, the cell population predominantly consisted of growth-arrested and terminally differentiated granulocytes (Fig. 8). These results suggest that c-Myc is an essential mediator of the property of c-Myb to block terminal differentiation of 32D myeloid cells.

Effect of c-Myc Dominant Negatives on c-Myb-induced Block of Differentiation of 32D Cells-
Effect of c-Myc Dominant Negatives on c-Myb-induced Block of Differentiation of M1 Cells-It has been previously shown that ectopic overexpression c-myb also blocks IL-6-induced terminal differentiation of M1 cells into monocytes. To verify whether this process is also dependent on the expression of c-Myc, we expressed c-Myb together with either MycRX or Max2RX in M1 cells by retroviral transduction. Following selection, parental M1 cells together with the infected M1 cells V/c-Myb were treated with IL-6 for 4 days. The parental M1 cells differentiated into mature cells with monocytic morphology confirming the previous observations that IL-6 treatment was sufficient to induce differentiation in this cell line (Fig. 9). When c-Myb was expressed in these cells, differentiation was blocked and cells (V/c-Myb) remained as blasts (Fig. 9). We next expressed the dominant-negative mutants of Myc (MycRX and Max2RX) in these cells using a retroviral expression system. Following the verification of expression of c-Myb as well as MycRX and Max2RX, we exposed the cells to IL-6 for 4 days. These results presented in Fig. 9 show that when the dominant-negative mutant MycRX or Max2RX was expressed in the presence of c-Myb, the blockage of differentiation by c-Myb was inhibited and the cells (MycRX/c-Myb and Max2RX/c-Myb) differentiated into mature cells (Fig. 9). The results were consistent with those observed in 32D cells and suggest that in both systems c-myc gene expression is essential for c-myb-induced transformation.

DISCUSSION
In this article, we have examined the molecular mechanisms associated with Myb-mediated transformation of the myeloid precursor cell line, 32Dcl3. This cell line, which is derived from normal mouse bone marrow is strictly dependent on IL-3 for growth (32,33) and is not predisposed to developing factor-independent clones (34). These cells have been found to undergo terminal differentiation when placed in an IL-3-free medium containing GCSF (32,33). However, when the cells were incubated in the absence of IL-3 or GCSF, they were found to become arrested in the G 1 phase of the cell cycle and rapidly enter apoptotic pathways loosing viability in 48 -96 h (34). Results presented here as well those published earlier show that ectopic overexpression of v-Myb or c-Myb in these cells results in a block to their ability to terminally differentiate into granulocytes in the presence of GCSF (9 -11). In addition, the results presented in this communication show that ectopic overexpression of c-Myb in these cells results in an elevated expression of c-Myc further supporting the notion that at least some of the effects mediated by c-Myb are achieved via the activation of c-Myc. A number of earlier studies have shown that the promoter/ enhancer region of c-myc contains several Myb-binding sites and in vitro transactivation studies showed that the c-myc promoter is activated by c-Myb (43,44,45). However, the expression pattern of c-Myc and c-Myb in mammalian tissues is very different suggesting that the regulation of c-myc by c-Myb is not universal and may occur, if any, in a limited number of tissues. However, at least in one cell line, the myeloid leukemic cell line, M1, it was found that expression of c-Myc is dependent on the expression of c-Myb, suggesting that in myeloid cells, this relationship might exist (27). This notion is further confirmed by our observations presented here, which show that ectopic overexpression of c-Myb results in a block to GCSF-induced down-regulation of c-myc.
To determine whether elevation of c-Myc expression alone is adequate for Myb-mediated effects on GCSF-induced granulocytic differentiation program, we ectopically overexpressed c-Myc in 32Dcl3 cells and studied the effects of this gene on the differentiation program. Our results show that unlike with c-Myb, ectopic overexpression of c-Myc had little or no effect on GCSF-induced terminal differentiation of 32Dcl3 cells, suggesting that Myb-mediated effects are not solely manifested through c-Myc regulation. This is in sharp contrast to the results obtained with the M1 leukemic cell line, whose terminal differentiation program was found to be blocked by ectopic overexpression of c-Myb and c-Myc (12,31). Such differences are seen in other cell types as well and the role of c-Myc in differentiation appears to be tissue-dependent. For example, c-Myc-expressing primary quail myoblasts are unable to form myotubes or express muscle-specific genes, such as myosin, and show severely reduced expression of the myogenic regulatory factors myoD, myogenin, and myf5 (47). However, activation of c-Myc in primary human keratinocytes caused a progressive reduction in growth rate and a marked stimulation of terminal differentiation (48). In a recent study, it was shown that, although c-Myc expression was downregulated in human leukemia K562 cells during differentiation and ectopic overexpression of c-Myc in these cells did not affect the monocytic-macrophagic and magakaryocytic differentiation induced by 12-O-tetradecanoyl phorbol-13 acetate (TPA) and staurosporine respectively (49). However, c-Myc is able to inhibit the erythoid differentiation of K562 cells when induced with 1-␤-D-arabinofuranosylcytosine (50).
To gain a further understanding of the role of c-Myc in Myb-mediated transformation of myeloid precursor cells, we expressed dominant-negative mutants of Myc in 32Dcl3 cells that have been earlier transfected with a c-myb expression vector. These dominant-negative mutants take advantage of the fact that the Myc and Max proteins heterodimerize and bind to nuclear DNA via the b-HLH-LZ motifs. As depicted in Fig. 5A, the dominant-negative mutants were generated by reciprocal exchange of HLH-LZ regions (35). As a result of this exchange, these mutant molecules sequester endogenous wildtype Myc and Max into inactive complexes (35,46). These mutants were previously shown to act as strong dominant suppressors of Myc activity by decreasing endogenous active Myc-Max complexes (35). Interestingly, expression of these dominant-negative mutants of Myc in Myb-transfected cells resulted in a block to the effects of c-Myb and resulted in the re-initiation of the GCSF-induced terminal differentiation program. To investigate whether this relationship between c-Myc and c-Myb holds true in another cell line, we used the M1 myeloid leukemic cells. As shown before, expression of c-Myb in M1 cells blocked IL-6-mediated terminal differentiation of M1 cells. Moreover, when c-Myc dominant negatives were expressed in Myb-transfected M1 cells, Myb-induced block was relieved, and the cells were able to differentiate in a normal manner. These results taken together suggest that while the activation of c-Myc by itself is inadequate to mimic the phenotypic effects of c-Myb, its expression is essential for the Mybmediated block to GCSF-induced terminal differentiation of 32Dcl3 cells. These observations suggest that c-Myb activates additional signal transduction pathways that in concert with c-Myc bring about the phenotypic effects exerted by c-Myb. Identification of these transcriptional targets of c-Myb and a detailed understanding of the signal transduction pathways that these genes are associated with is likely to provide an understanding of the role played by proto-oncogenes in the regulation of hematopoiesis.