¹Fels Institute for Cancer Research and Molecular Biology, Department of Biochemistry, Temple University School of Medicine, 3307 N. Broad St., Philadelphia, Pennsylvania, PA 19140, USA

²Cell and Molecular Biology Graduate Group, University of Pennsylvania, Philadelphia, Pennsylvania, PA 19104, USA

Correspondence to: B Hoffman , Fels Institute for Cancer Research and Molecular Biology, Department of Biochemistry, Temple University School of Medicine, 3307 N. Broad St., Philadelphia, Pennsylvania, PA 19140, USA

Previously we have shown that deregulated expression of c-myc in M1 myeloid leukemic cells blocked IL-6-induced differentiation and its associated growth arrest; however, the cells proliferated at a significantly reduced rate compared to untreated cells. The basis for the increased doubling time of IL-6-treated M1myc cells was found to be due to the induction of a p53-independent apoptotic pathway. The apoptotic response was not completely penetrant; in the same population of cells both proliferation and apoptosis were continuously ongoing. Down-regulation of Bcl-2 was insufficient to account for the apoptotic response, since deregulated expression of Bcl-2 delayed, but did not block, the onset of apoptosis. Furthermore, our results indicated that the IL-6-induced partial hypophosphorylation of the retinoblastoma gene product (Rb), observed in M1myc cells, was not responsible for the apoptotic response. Finally, the findings in M1 cells were extended to myeloid cells derived from the bone marrow of wild type and p53-deficient mice, where the deregulated expression of c-myc was also shown to block terminal differentiation and induce apoptosis independent of p53. These findings provide new insights into how myc participates in the neoplastic process, and how additional mutations can promote more aggressive tumors. Oncogene (2000) 19, 2967-2977

myeloid; c-myc; p53; differentiation; apoptosis

Introduction

The proto-oncogene c-myc has been shown to play a pivotal role in the control of proliferation, differentiation, and apoptosis, and its deregulated expression participates in the progression of a wide range of neoplasias (Eilers, 1999; Facchini and Penn, 1998; Nesbit et al., 1999; Schmidt, 1999; Spencer and Groudine, 1991). Repression of c-myc is required for terminal differentiation of many cell types, including myeloid cells (Freytag, 1988; Hoffman-Liebermann and Liebermann, 1991a; Packham and Cleveland, 1995; Selvakumaran et al., 1993).

The murine M1 myeloid leukemic cell line proliferates autonomously and can be induced with the physiological inducer interleukin-6 (IL-6) to undergo terminal differentiation and growth arrest. As is the case for many cell types (Evan and Littlewood, 1993), c-myc is highly expressed in proliferating M1 myeloblasts and is suppressed upon induction of terminal differentiation (Liebermann and Hoffman-Liebermann, 1989). Deregulated and continued expression of c-myc in M1 cells blocked terminal myeloid differentiation induced by IL-6 at an intermediate stage in the progression from immature blasts to mature macrophages (Hoffman-Liebermann and Liebermann, 1991a). In addition, IL6-treated M1myc cells failed to exit the cell cycle and continued to proliferate, albeit with an increased doubling time relative to untreated M1 and M1myc cells.

The proto-oncogene c-myb, whose transcripts are found primarily in tissues of hematopoietic origin (Graf, 1992), is downregulated more rapidly than c-myc following induction of differentiation of M1 cells (Hoffman-Liebermann and Liebermann, 1991b). Similar to the effect of c-myc, the deregulated expression of c-myb also blocked M1 myeloid differentiation, although at an earlier stage (Selvakumaran et al., 1992). Furthermore, M1myb cells failed to exit the cell cycle in response to IL-6 and continued to proliferate. In contrast to IL6-treated M1myc cells, the doubling time of IL6-treated M1myb cells closely resembled untreated M1, M1myc, and M1myb cells (Figure 1). Thus, M1myb cells provided a counterpoint to dissect the molecular basis for the increased doubling time of M1myc cells treated with IL6.

To understand how the c-myc oncogene participates in the neoplastic process by influencing proliferation rates, in this work we have extended the analysis of c-myc function in the regulation of terminal myeloid differentiation by examining the molecular basis for the altered growth kinetics of IL-6-treated M1myc. We have shown that deregulated expression of c-myc during differentiation of myeloid leukemic cells, as well as myeloid cells derived from the bone marrow of wild type and p53-deficient mice, not only blocks terminal differentiation and its associated growth arrest, but also induces apoptosis.

Results

Deregulated c-myc induced apoptosis in IL-6-treated M1myc cells in a p53-independent manner

Experiments were carried out to determine if the increased doubling time of proliferating M1myc cells in the presence of IL-6 (Figure 1a) was due to loss of survival by apoptosis. Analysis of M1myc cell lines revealed that there was a sharp reduction in the relative number of viable cells subsequent to stimulation with IL-6 (Figure 1b). The percentage of viable M1myc cells decreased to 65-70% by day 3, but then stabilized and remained uniform over the next 8 weeks, the time the M1myc cells were maintained in culture in the presence of IL-6. It should be pointed out that IL-6-treated M1myc cell cultures had to be diluted every 3 days to prevent overgrowth.

Apoptotic cell death requires protein synthesis and is frequently accompanied by the activation of a Ca²⁺/Mg²⁺-dependent endonuclease which generates single-stranded DNA breaks, preferentially between nucleosomes, resulting in laddering of genomic DNA when analysed on agarose gels (Liebermann et al., 1995; Meikrantz and Schlegel, 1995). To ascertain if loss of survival was due to apoptosis, the genomic DNA from cells treated with IL-6 was examined (Figure 1c). The DNA from M1myc cells exhibited the characteristic laddering pattern of apoptosis as early as 1 day following treatment with IL-6, with the intensity of the bands getting progressively stronger up to 3 days. Furthermore, a significant number of IL6-treated M1myc cells displayed the typical apoptotic morphology characterized by nuclear and cytoplasmic condensation and formation of apoptotic bodies (not shown). Taken together, this indicated that IL-6-stimulated M1myc cells were indeed undergoing programmed cell death. Since the parental M1 and M1myc cells, untreated and treated with IL-6, are devoid of expression of the tumor suppressor p53 (Yonish-Rouach et al., 1991; Figure 5d), the observed c-myc-mediated apoptosis in M1myc cells must proceed via a mechanism that is independent of p53.

Consistent with the percent viability (Figure 1b), the laddering of genomic DNA in the eighth week was of similar intensity as on day 3 (Figure 1c). These results indicated, therefore, that the apoptotic effect induced by deregulated c-myc in IL-6-treated M1 cells was not completely penetrant; the cells that survived continued to proliferate until some proportion of them succumbed to apoptotic cell death. Thus, both proliferation and apoptosis are ongoing in the same population of cells (Figure 1b,c). Similar results were obtained with four other independent clones of M1myc.

In contrast to the early induction of apoptosis in IL-6-treated M1myc cells, similarly treated M1myb cells, as well as untreated M1, M1myb, and M1myc cells all showed no DNA laddering or loss of viability (Figure 1c). In the parental M1 and M1-neo control cells stimulated with IL-6, however, there was both apoptosis and complete loss of cell viability at day 7 or 8 post-treatment (Figure 1b,c). This cell death is known to be associated with the end of the normal program of terminal myeloid differentiation in M1 cells (Liebermann and Hoffman, 1994). Since endogenous c-myc expression has been down-regulated, the apoptosis in these cells occurred by a myc-independent process.

Fluorescence activated cell sorter (FACS) analysis of the cell cycle in M1, M1myb, and M1myc cell lines was carried out (Figure 2). The data indicated that M1 cells treated with IL-6 quickly became growth arrested and accumulated predominantly in the G0/G1 phase of the cell cycle. M1myb cells were not affected at all by IL-6. IL-6-stimulated M1myc cells, as expected, did not show growth arrest in any phase, but the FACS analysis clearly demonstrated that there were two subpopulations of cells; one that was proliferating and one that was undergoing cell death, as evidenced by the sub-G0/G1 peak which is indicative of cells with less than 2N DNA content. In addition, the FACS data was in agreement with the cell viability counts (compare Figures 1b and 2).

Expression analysis of the differential response of M1myb and M1myc cells to interleukin-6

An important difference between the effects of deregulated expression of c-myc and c-myb on myeloid differentiation is that, whereas both were capable of disrupting terminal differentiation and its associated growth arrest in M1 cells, only c-myc also caused these cells to enter into an apoptotic cell death program. Comparing M1myc and M1myb cells for the expression patterns of key gene products known to be involved in cellular proliferation as well as differentiation, growth arrest, and apoptosis, provided an opportunity to dissect the molecular machinery underlying the c-myc-mediated apoptotic response in myeloid cells treated with the differentiation inducer IL6. Among the genes whose regulation was investigated were some which have been implicated as positive regulators (cdk2, cdk4, cdc2, PCNA, and the cyclins A, E, and D1) and negative regulators (the retinoblastoma gene product and the cyclin-dependent kinase inhibitors, p15 and p16) of the cell cycle, others which were postulated to be regulators of apoptosis (bcl-2, bax, and mdm-2), and yet others that were associated with the induction of terminal myeloid differentiation (MyD88, MyD116, MyD118, and JunB).

Analysis of the mRNA expression (Figure 3a) and protein expression (Figure 3b) patterns of M1, M1myb, and M1myc cell lines stimulated with IL-6 showed that for M1myb and M1myc cells there were no significant differences in the expression profiles for most of the genes examined. However, they did differ in the expression of three genes. These were the inhibitor of cyclin-dependent kinases, p15 (Figure 3a), the bcl-2 proto-oncogene which has the ability to suppress apoptosis (Bissonnette et al., 1992; Fanidi et al., 1992; Oltavi and Korsmeyer, 1994; Oltvai et al., 1993), and the retinoblastoma gene product, Rb (Figure 3b), which functions as a tumor suppressor and has been implicated in the control of a critical checkpoint in the G1 phase of the cell cycle (Sherr, 1996).

The role of p15 and Rb in c-myc-mediated apoptosis of IL-6-treated M1myc cells

p15 transcripts were up-regulated in both M1 and M1myc cells, but not in M1myb cells, upon stimulation with IL-6 (Figure 3a). The induction of active p15 inhibits the kinase activity of cdk4-cdk6-cyclin D, which in turn prevents the phosphorylation of Rb, leading to growth inhibition (Peter and Herskowitz, 1994; Sherr, 1996). p15 induction in IL6-treated M1myc cells was surprising since these cells did not growth arrest.

The Rb gene product was not only down-regulated but also shifted from the hyperphosphorylated to the hypophosphorylated form as differentiation proceeded in M1 cells (Figure 3b). In M1myb cells there was no detectable change in either the quantity or phosphorylation status of Rb. By contrast, in IL6-treated M1myc cells, Rb levels were slightly decreased and by day 3 there was an equal amount of the hyperphosphorylated and hypophosphorylated forms. Similar levels of increased p15 expression and decreased Rb phosphorylation were observed after 7 days treatment with IL-6 as were seen at 3 days post-treatment (data not shown).

The idea that the induction of p15 and its possible role in the hypophosphorylation of Rb underlie the c-myc mediated apoptosis in IL-6-treated M1myc cells was tested by preventing Rb hypophosphorylation. This was achieved by overexpressing a cdk4 transgene in M1myc cells (M1myc/Cdk4). Figure 4a shows the levels of deregulated cdk4 expression in one M1myc/Cdk4 clone. The cells were then induced for terminal differentiation with IL-6 to see if the presence of exogenous cdk4 abrogated c-myc-mediated apoptosis.

To determine if the exogenous cdk4 was functional, the phosphorylation status of the target of the cdk4-cdk6-cyclin D complex, the retinoblastoma gene product was examined. In IL6-treated M1myc cells there was an equal amount of the hyperphosphorylated and hypophosphorylated Rb, but in M1myc/Cdk4 cells all of the Rb stayed in the hyperphosphorylated form (Figure 4b). Thus, these data demonstrate that the exogenous cdk4 was, indeed, active and able to prevent even partial hypophosphorylation of Rb.

Analysis of parental M1myc, M1myc/Cdk4, and M1myc/puro cells treated with IL-6 for growth kinetics (Figure 4c) and DNA ladders (data not shown) revealed that there was no significant difference in the response of any of these cell lines. This implied that the underlying c-myc-mediated apoptosis was proceeding unaffected by the deregulated expression of cdk4.

These results negate the hypothesis that hypophosphorylation of Rb, possibly resulting from the induction of p15, was participating in the myc-mediated apoptotic response induced by IL-6.

The effect of deregulated expression of bcl-2 on the apoptotic response of IL-6-treated M1myc cells

Bcl-2 was down-regulated in both IL-6-treated M1 and M1myc cells, but not in M1myb cells (Figure 3b). In contrast, the level of bax protein, a dimerization partner of bcl-2, did not change in IL-6-treated M1myc and M1myb cells, but was down-regulated in similarly treated M1 cells (Figure 3b). Since it is thought that the ratio of bcl-2 to bax can influence the intracellular set point for apoptosis (Chao and Korsmeyer, 1998; Oltvai and Korsmeyer, 1994; Oltvai et al., 1993; Zamzami et al., 1998), the observed down-regulation of bcl-2, together with the unchanged basal levels of bax, may account for the apoptosis seen in M1myc cells. Therefore, this hypothesis was tested by deregulating the expression of bcl-2 in M1myc cells (M1myc/Bcl-2) and ascertaining whether IL-6-induced apoptosis is suppressed.

Figure 5a shows the sustained deregulated high expression of the bcl-2 transgene in an IL-6-treated M1myc/Bcl-2 clone. For 4 days IL-6-treated M1myc/Bcl-2 cells continued to proliferate with growth kinetics that were similar to those of untreated M1myc cells (Figure 5b). To ascertain if this restoration of the normal doubling time in M1myc/Bcl-2 cells was due to suppression of c-myc-mediated apoptosis, genomic DNA was analysed (Figure 5b). Whereas in M1myc cells apoptotic DNA laddering was apparent at day 1 post-treatment with IL-6, there were no detectable ladders in M1myc/Bcl-2 cells even by 4 days following addition of IL-6. However, DNA laddering was observed at 5 days (Figure 5c). In agreement with these results, further analysis of the growth kinetics of IL-6-treated M1myc/Bcl-2 cells beyond day 3 revealed that the doubling time of these cells increased after 5 days, and ultimately assumed the same time as that of M1myc cells stimulated with IL-6 (Figure 5c).

Taken together, these results demonstrated that while the deregulated expression of bcl-2 initially suppressed the myc-mediated apoptosis in IL-6-treated M1myc cells, its expression was not sufficient to maintain the suppression of cell death. Hence, the down-regulation of bcl-2 observed in M1myc cells stimulated with IL-6 cannot by itself account for the myc-mediated apoptotic response. Additionally, the morbidity of IL-6-treated M1myc/Cdk4 and M1myc/Bcl-2 cells was p53-independent, as shown by the absence of p53 expression in these cells (Figure 5d). This ruled out the possibility that manipulations of M1 cells by the transfection of myc, cdk4, or bcl-2 resulted in the inadvertent induction of a p53-dependent apoptotic pathway.

Finally, since it was possible that a rescue of both the bcl-2 and cdk4-sensitive pathways was required to protect IL-6-treated M1myc cells from cell death, stable M1myc cell lines that co-expressed bcl-2 and cdk4 were established. IL-6 treatment of M1myc/Bcl-2/Cdk4 cells resulted in a response that was indistinguishable from that of IL-6-treated M1myc/Bcl-2 cells, that is, myc-mediated apoptosis was only delayed (data not shown). As with M1myc/Cdk4 cells, it was confirmed that in IL-6-treated M1myc/Bcl-2/Cdk4 cells the cdk4 transgene was expressed in a deregulated mode and and was functional (data not shown).

Deregulated expression of c-myc in myeloid cells from normal and p53^-/- bone marrow (BM) blocked terminal differentiation and induced apoptosis

To assess the effect of deregulated expression of c-myc on myeloid bone marrow cell differentiation and survival, myeloblast enriched BM cells from wild type (BALBc) and p53^-/- (BALB/cJ-Trp53^tm1Tyj) mice were infected with retroviral vectors MSCVmyc/neo and MCSVneo, as described in Materials and methods. Cells were seeded in methyl cellulose with IL-3 in the presence of G418 (400 ug/ml), and after 7 days colonies appeared. It was observed that BM infected with the c-myc-expressing vector yielded more compact colonies (86»plus;5%) compared to following infection with the control vector (8»plus;4%). One hundred colonies from each infection were picked and transferred to liquid culture; half were maintained in IL-3 and half were maintained in GM-CSF. Several colonies from each infection and set of growth conditions were expanded and, after 5 days, analysed for cell morphology, per cent viable cells, DNA ladders and growth kinetics. Deregulated expression of c-myc blocked terminal differentiation and promoted apoptosis (Figure 6a), as determined by cell morphology and appearance of DNA ladders. Furthermore, both wild type and p53^-/- BM-myc cells displayed significantly increased doubling times as compared to BM-neo controls (Figure 6b) in response to the induction of differentiation by the hematokine, GM-CSF.

Discussion

In this study we have demonstrated that, in addition to preventing terminal myeloid differentiation and its associated growth arrest, deregulated c-myc induced an ancillary program of apoptosis that is independent of the tumor suppressor, p53. Interestingly, this effect of c-myc was not completely penetrant. In the presence of IL-6, M1myc cells could be maintained in culture for extended periods of time, and analysis of this cell population indicated that both proliferation and apoptosis were continuously ongoing. Additionally, normal myeloid cells obtained from wild type as well as from p53^-/- bone marrow which expressed deregulated c-myc and were induced for differentiation behaved similarly to IL-6-treated M1myc cells in that they were also blocked in differentiation and underwent apoptosis. In contrast to M1 leukemic cells, which proliferate autonomously and undergo growth arrest in association with terminal differentiation, normal myeloid cells treated with hematokines are induced for both proliferation and differentiation. The cells undergoing differentiation become growth arrested. If deregulated myc only blocked differentiation and growth arrest, the doubling time of the myeloid cell population would have decreased (i.e. increased proliferation) relative to BM-neo controls. However, the doubling time was observed to increase (Figure 6b, compare GM-CSF-treated BM-neo and BM-myc cells), consistent with our data that deregulated c-myc induces apoptosis, in addition to blocking the terminal differentiation program.

Our results indicate, therefore, that the p53-independent apoptosis associated with the myc-mediated block in differentiation is not peculiar to the M1 cell line, but is a feature of normal myeloid cells as well and, perhaps, may be a general feature of cells undergoing terminal differentiation. However, it appeared from our data that the BMp53^-/- myc cells were a little less susceptible to cell death than their wild type counterpart (BM-myc cells; see relative ladder intensity and viability data, Figure 6a). This may suggest a p53-dependent component in both the apoptosis associated with the normal differentiation program (compare BM-neo and BMp53^-/- neo, Figure 6a), as well as cell death resulting from the deregulated expression of myc.

Although a role for c-myc in the regulation of apoptosis has been implicated in growth arrested fibroblasts (Evan et al., 1992), factor deprived IL-3 dependent 32D myeloid cells (Askew et al., 1991), anti- TCR antibody stimulated T-cell hybridomas (Shi et al., 1992), and in several other examples (reviewed in Hoffman and Liebermann, 1998; Packham and Cleveland, 1995; Prendergast, 1999), where its deregulated expression has been shown to either accelerate or induce apoptosis, this appears to be the first demonstration that c-myc can regulate p53-independent apoptosis during terminal differentiation.

Perhaps the most interesting and puzzling feature of this phenomenon is that the program of cell death is not completely penetrant, that is, not all the cells die after differentiation is initiated. Instead, the surviving cells continue to proliferate until some proportion of them die in a second round of apoptosis. The survivors once again go through mitotic growth and again some of them undergo apoptotic cell death, and so on. This was observed up to 8 weeks. The process appears to be a stochastic one, where individual cells commit to the death program over an extended period of time. The likelihood that any given cell will die probably depends on whether a threshhold for apoptosis is crossed in that cell. Identifying the apoptotic pathway activated in IL6-treated M1myc cells will be necessary to understand the incomplete penetrance of the apoptotic response.

Two recent studies have, for the first time, tied myc-induced apoptosis to the well characterized CD95/Fas/APO-1 apoptotic pathway (Hueber et al., 1997; Kagaya et al., 1997). In serum deprived fibroblasts, c-myc-mediated apoptosis required interaction between the transmembrane cell surface receptor CD95 and its CD95L ligand. Furthermore, both Bcl-2 and IGF-1 (insulin-like growth factor-1) could suppress this apoptosis by their action downstream of CD95 (Hueber et al., 1997). If the activation of the CD95 pathway is a requirement for the c-myc-induced apoptosis in differentiating myeloid cells, there are models to explain the stochastic nature of the observed apoptosis in M1myc cells. The idea of an apoptotic threshhold discussed above could be represented by the concentrations of CD95 or its ligand. In this regard, it is interesting that a recent study suggests that c-myc activates the expression of the CD95 ligand in T lymphocytes (Wang et al., 1998). The possible involvement of CD95 receptor and ligand as mediators of myc-mediated apoptosis in IL-6-treated M1myc cells is currently under investigation.

These studies should increase our knowledge about how deregulated c-myc promotes apoptosis in conjunction with blocking terminal myeloid differentiation and its associated growth arrest. This information will provide further insights into the role of c-myc in tumorigenesis and how additional mutations blocking the apoptotic response can result in more aggressive tumors.

Materials and methods

Cells, cell culture and cytokines

M1 cells (M1D+ clone 6; Hoffman-Liebermann and Liebermann, 1991b; Lord et al., 1990) were cultured in Dulbecco's modified Eagle's medium (DMEM; Gibco BRL, Gaithersburg, MD, USA) supplemented with 10% heat inactivated horse serum (Gibco-BRL), and 1% penicillin and streptomycin (Gibco-BRL). M1myc (Hoffman-Liebermann and Liebermann, 1991a) and M1myb (Selvakumaran et al., 1992) cell lines were similarly cultured, except that the culture medium was supplemented with 400 ug/ml G418 (Gibco-BRL) and 5 uM methotrexate, respectively, in order to maintain selection of the transgenes. Five independent clones of each cell type were tested for their response to IL-6. The control M1neo and M1DHFR cell lines have been described previously (Hoffman-Liebermann and Liebermann, 1991a, Selvakumaran et al., 1992). All cells were cultured in a humidified atmosphere with 10% CO₂ at 37°C. Cells were induced for differentiation with IL-6 at 50 ng/ml (a generous gift from Amgen Inc., Thousand Oaks, CA, USA) after being seeded at a concentration of 0.15´10⁶ cells/ml.

Bone marrow infection, selection, and expansion

Myeloblast enriched bone marrow (BM) cells, obtained from femurs of sodium caseinate injected mice [BALB/c (wild type) and BALB/cJ-Trp53^tm1Tyj mice (p53^-/-) from Jackson Laboratory, Bar Harbor, Maine, USA], consist primarily of cells of the myeloid lineage (95»plus;4%), with 33»plus;3% myeloid precursors at the myeloblast to promyelocyte stage (Liebermann and Hoffman-Liebermann, 1989). MSCV retroviral vectors, packaged as helper-free infectious ecotropic retroviruses using Bosc23 packaging cells (Pear et al., 1993), were particularly efficient in infecting myeloblast enriched BM cells. Briefly, 10 ug of pMSCVneo and pMSCVmyc/neo were transfected into Bosc23 cells using the calcium phosphate-DNA precipitation method (Pellicer et al., 1978) and after 48 h the cells were treated with mitomycin C (10 ug/ml) for 3 h. After repeated washings the cells were refed and BM was added (5´10⁶ cells) in the presence of polybrene (8 ug/ml), 10% interleukin-3 (IL-3) and recombinant rat stem cell factor (SCF, 200 ng/ml; a generous gift from Amgen Inc., Thousand Oaks, CA, USA). The source of IL-3 was WEHI-3B conditioned medium (Ymer et al., 1985). The BM was cocultivated for 48-72 h, and then seeded in methylcellulose (StemCell Technologies, Inc.) with IL-3, and with or without G418 (650 ug/ml). After 7 days, colonies were assessed, both for number and morphology (compact, diffuse). Cells from colonies were transferred to liquid culture for further analysis, in the presence of either IL-3 (10%) or murine recombinant GM-CSF (100 ng/ml; a generous gift from Amgen Inc., Thousand Oaks, CA, USA).

Assays for differentiation-associated properties

Cells were collected at the indicated times, and following cytocentrifugation were subjected to May-Grunwald-Giemsa staining. Morphological differentiation was determined by counting at least 300 cells on such stained cytospin smears, and scoring the proportion of immature blast cells, cells at intermediate stages of differentiation, and mature macrophages (Lord et al., 1990). Viability was determined by counting cells in a hemocytometer using the trypan blue dye exclusion method. Results of all experiments represent the mean of at least three independent determinations, with standard deviations up to»plus;15% (i.e., 20%=20»plus;3%).

RNA extraction and Northern blotting

Total RNA was prepared from 3-5´10⁶ cells using TRIzol reagent (Gibco-BRL) as described in the manufacturer's specifications. Ten g per lane of each RNA sample was electrophoresed on a 1% agarose gel containing formaldehyde (0.7%). Northern blot analysis and stripping blots of probe to rehybridize were done as described previously (Krishnaraju et al., 1995; Lord et al., 1990).

Protein extraction and immunoblotting

Preparation of protein extracts and subsequent Western blot analysis were carried out using standard techniques. Fifty ug of each protein extract sample was fractionated on SDS-PAGE gels and equal loading of protein was verified by staining the blots with 0.1% Ponceau S solution (Sigma) prior to incubation with antibody. Signals were developed by using the enhanced chemiluminescence (ECL) Western blotting system (Amersham). Primary antibody against the retinoblastoma gene product was obtained from Pharmingen, San Diego, CA, USA (clone G3-245, cat # 14001A). Antibodies against murine bcl-2, bax, cyclin E, cdk-2, cdk-4 and p53 were from Santa Cruz Biotechnology Inc., Santa Cruz, CA, USA (cat # sc-492, sc-526, sc-481, sc-163, sc-260, and sc-1314 respectively). The cyclin A, cyclin D1 and cdc2 antibodies were a kind gift from X Grana, Temple University.

General recombinant DNA techniques, expression vectors and DNA probes

Plasmids and DNA probes were prepared as previously described (Hoffman-Liebermann and Liebermann, 1991a; Selvakumaran et al., 1992). The retroviral plasmid expression vectors, MSCV-neomycin (neo) and MSCV-puromycin (puro), were a gift from Dr Robert G Hawley (Hawley, 1994); (University of Toronto, Toronto, Canada). To construct the MSCV-puro-Bc12 and MSCV-puro-cdk4 vectors, full length human Bcl-2 and murine cdk4, respectively, were cloned into the EcoR1 site of the 6.3 Kb MSCV-puromycin plasmid. The 0.9 Kb human Bcl-2 fragment was excised from the puc19-Bc12 vector (Tsujimoto and Croce, 1986), and the 1.6 Kb murine cdk4 fragment was isolated from a pBS-cdk4 vector (obtained from C Sherr, HHMI, St Jude Children's Research Hospital, Memphis, TN, USA). The murine p15 and p16 DNA probes were excised from pBS-p15 and pBS-p16 vectors, respectively, also from C Sherr. DNA probes for murine junB, MyD88, MyD116, MyD118, c-myc, c-myb, and -actin, have been previously described (Abdollahi et al., 1991a,b; Lord et al., 1990; Hoffman-Liebermann and Liebermann, 1991a; Krishnaraju et al., 1995; Lord et al., 1993). The PCNA and mdm2 fragments were excised from the pT7PCNA (B Stillman, Cold Spring Harbor Laboratory, NY, USA) and p11B-mdm2 (D George, University of Pennsylvania, Philadelphia, PA, USA) vectors, respectively. All probes were labeled by random priming (Gibco-BRL, RadPrime DNA labeling kit, cat # 18428-011) to a specific activity equal to or greater than 10⁹ c.p.m./ug.

Establishment of M1myc/bcl-2 and M1myc/cdk4 cell lines

Virus was generated from the plasmid forms of the retroviral vectors, MSCV-puro, MSCV-puro-Bcl-2 and MSCV-puro-cdk4 by transient transfection (48 h) of the packaging cell line PA317 using calcium phosphate-DNA precipitation (Pellicer et al., 1978). The resulting viral conditioned medium (VCM) was used to infect M1 and M1myc cells as previously described (Selvakumaran et al., 1993). For puromycin resistant colony selection, infected cells were seeded at 100 cells/ml in growth medium (DMEM+10% horse serum) containing puromycin at 4 ug/ml. At least five independent clones expressing each transgene were examined for their response to IL-6. For each engineered cell line, all the clones behaved similarly. Control cell lines (M1/MSCV-puro and M1myc/MSCV-puro clones) were indistinguishable from that of their respective parental cells.

Autoradiographic analysis of apoptotic DNA fragmentation

To facilitate the 3' end-labeling of apoptotically fragmented DNA, 500 ng of genomic DNA were incubated with 25 units of terminal deoxynucleotidyl transferase enzyme (Boehringer-Mannheim, BMB) and 50 uCi of radiolabeled dideoxy-nucleotide, [³²P]-ddATP (3000 Ci/mmol; Amersham, Arlington Heights, IL, USA) in the presence of reaction buffer (0.2 M potassium cacodylate, 2.5 mM cobalt chloride, 25 mM Tris-HCl, 1.25 mg/ml bovine serum albumin; pH 6.6) for 60 min at 37°C. After removal of unincorporated radionucleotide by ethanol precipitation, the labeled DNA was resuspended in TE buffer and subjected to electrophoretic analysis in a 2% agarose gel. Finally, the dried gel was exposed to Kodak X-Omat film for 30 min or more at -80°C.

Flow cytometric analysis

Cells were harvested by centrifugation, fixed in 70% ice-cold ethanol, and were subsequently stained with propidium iodide (100 g/ml in 7.6 mM sodium citrate; Sigma, St. Louis, MO, USA) prior to analysis using a Coulter Epics Elite system (Miami, FL, USA). An appropriate window was chosen in the fluorescence-activated cell sorter (FACS) such that both living and apoptotic cells were analysed. Cell cycle analysis was performed at least three times with similar results.

Acknowledgements

This work was supported by National Institute of Health grant IROI CA81168 to B Hoffman.

Abdollahi A, Lord KA, Hoffman-Liebermann B and Liebermann DA. (1991a). Cell Growth Differ 2, 401-407.

Abdollahi A, Lord KA, Hoffman-Liebermann B and Liebermann DA. (1991b). Oncogene 6, 165-167. MEDLINE

Askew DS, Ashmun RA, Simmons BC and Cleveland JL. (1991). Oncogene 6, 1915-1922. MEDLINE

Bissonnette RP, Echeverri F, Mahboubi A and Green DR. (1992). Nature 359, 552-554. MEDLINE

Chao DT and Korsmeyer SJ. (1998). Annu Rev Immunol 16, 395-419. MEDLINE

Eilers M. (1999). Mol Cells 9, 1-6. MEDLINE

Evan GI and Littlewood TD. (1993). Curr Opin Genet Dev 3, 44-49. MEDLINE

Evan GI, Wyllie AH, Gilbert CS, Littlewood TD, Land H, Brooks M, Waters CM, Penn LZ and Hancock DC. (1992). Cell 69, 119-128. MEDLINE

Facchini LM and Penn LZ. (1998). Faseb J 12, 633-651. MEDLINE

Fanidi A, Harrington EA and Evan GI. (1992). Nature 359, 554-556. MEDLINE

Freytag SO. (1988). Mol Cell Biol 8, 1614-1624. MEDLINE

Graf T. (1992). Curr Opin Genet Dev 2, 249-255. MEDLINE

Hawley RG. (1994). Ann N Y Acad Sci 716, 327-330. MEDLINE

Hoffman B and Liebermann DA. (1998). Oncogene 17, 3351-3357. MEDLINE

Hoffman-Liebermann B and Liebermann DA. (1991a). Mol Cell Biol 11, 2375-2381. MEDLINE

Hoffman-Liebermann B and Liebermann DA. (1991b). Oncogene 6, 903-909. MEDLINE

Hueber AO, Zornig M, Lyon D, Suda T, Nagata S and Evan GI. (1997). Science 278, 1305-1309. Article MEDLINE

Kagaya S, Kitanaka C, Noguchi K, Mochizuki T, Sugiyama A, Asai A, Yasuhara N, Eguchi Y, Tsujimoto Y and Kuchino Y. (1997). Mol Cell Biol 17, 6736-6745. MEDLINE

Krishnaraju K, Nguyen HQ, Liebermann DA and Hoffman B. (1995). Mol Cell Biol 15, 5499-5507. MEDLINE

Liebermann DA and Hoffman B. (1994). Stem Cells Dayt 12, 352-369. MEDLINE

Liebermann DA, Hoffman B and Steinman RA. (1995). Oncogene 11, 199-210. MEDLINE

Liebermann DA and Hoffman-Liebermann B. (1989). Oncogene 4, 583-592. MEDLINE

Lord KA, Abdollahi A, Hoffman-Liebermann B and Liebermann DA. (1990). Cell Growth Differ 1, 637-645. MEDLINE

Lord KA, Abdollahi A, Hoffman-Liebermann B and Liebermann DA. (1993). Mol Cell Biol 13, 841-851. MEDLINE

Meikrantz W and Schlegel R. (1995). J Cell Biochem 58, 160-174. MEDLINE

Nesbit CE, Tersak JM and Prochownik EV. (1999). Oncogene 18, 3004-3016. MEDLINE

Oltvai ZN and Korsmeyer SJ. (1994). Cell 79, 189-192. MEDLINE

Oltvai ZN, Milliman CL and Korsmeyer SJ. (1993). Cell 74, 609-619. MEDLINE

Packham G and Cleveland JL. (1995). Biochim Biophys Acta 1242, 11-28. MEDLINE

Pear W, Nolan G, Scott M and Baltimore D. (1993). Proc Natl Acad Sci USA 90, 8392-8396. MEDLINE

Pellicer A, Wigler M, Axel R and Silverstein S. (1978). Cell 14, 133-141. MEDLINE

Peter M and Herskowitz I. (1994). Cell 79, 181-184. MEDLINE

Prendergast GC. (1999). Oncogene 18, 2967-2987. MEDLINE

Schmidt EV. (1999). Oncogene 18, 2988-2996. MEDLINE

Selvakumaran M, Liebermann D and Hoffman-Liebermann B. (1993). Blood 81, 2257-2262. MEDLINE

Selvakumaran M, Liebermann DA and Hoffman-Liebermann B. (1992). Mol Cell Biol 12, 2493-2500. MEDLINE

Sherr CJ. (1996). Science 274, 1672-1677. Article MEDLINE

Shi Y, Glynn JM, Guilbert LJ, Cotter TG, Bissonnette RP and Green DR. (1992). Science 257, 212-214. MEDLINE

Spencer CA and Groudine M. (1991). Adv Cancer Res 56, 1-48. MEDLINE

Tsujimoto Y and Croce CM. (1986). Proc Natl Acad Sci USA 83, 5214-5218. MEDLINE

Wang R, Brunner T, Zhang L and Shi Y. (1998). Oncogene 17, 1503-1508. MEDLINE

Ymer S, Tucker WQ, Sanderson CJ, Hapel AJ, Campbell HD and Young IG. (1985). Nature 317, 255-258. MEDLINE

Yonish-Rouach E, Resnitzky D, Lotem J, Sachs L, Kimchi A and Oren M. (1991). Nature 352, 345-347. MEDLINE

Zamzami N, Brenner C, Marzo I, Susin SA and Kroemer G. (1998). Oncogene 16, 2265-2282. MEDLINE

Figure 1 (a) Growth kinetics and (b) viability of M1, M1myc and M1myb cells treated with IL-6. Cells were seeded at 0.15´10⁶/ml in the presence or absence of IL-6 (50 ng/ml) and the number of viable cells was determined at the indicated times by trypan blue dye exclusion and counting in a hemocytometer. Each time point is the average of at least three experiments, with a standard deviation of up to 15%. Untreated M1, M1myc and M1myb cells proliferated with the same kinetics; therefore, only M1 is shown. On day 3 cells were reseeded at 0.15´10⁶/ml. For long term cultures, cells were split 1 : 3 every 3 days and supplemented with fresh medium containing IL-6. M1-neo control cells (not shown) consistently behaved like the parental M1 cells. (c) Analysis of genomic DNA from M1, M1myc, and M1myb cells stimulated with IL-6. Genomic DNA was extracted from 1´10⁷ cells at the indicated time points, end-labeled, and resolved on an agarose gel (500 ng/lane) as described in Materials and methods

Figure 2 Cell cycle analysis by quantitative flow cytometry. Untreated and IL-6-treated (50 ng/ml) M1, M1myb, and M1myc cells were collected at 3 days post-treatment and subjected to FACS analysis as described in Materials and methods

Figure 3 (a) Analysis of mRNA expression in M1, M1myc and M1myb cell lines treated with IL-6. RNA blots were hybridized to the appropriate ³²P-labeled probe, washed, and subjected to autoradiography for 24-48 h at -80°C. Significant differences in the expression pattern between IL-6-treated M1myc and M1myb cells are indicated by an '*'. (b) Analysis of protein expression in IL-6-treated M1, M1myc, and M1myb cells. Protein extracts were prepared before and after induction for differentiation with IL-6 (50 ng/ml) at the indicated time points as described in Materials and methods. Fifty g per lane of total protein extract were fractionated on a 10% SDS-PAGE gel, transferred to a PVDF membrane, and probed with the appropriate antibody (detailed in Materials and methods). Signals were developed by using the enhanced chemiluminescence (ECL) Western blotting system (Amersham). Again, significant differences in the expression pattern between IL-6-treated M1myc and M1myb cells are indicated by an '*'

Figure 4 Analysis of M1myc/cdk4 cell lines. (a) Deregulated expression of the cdk4 transgene in M1myc/cdk4 cells. Fifty ug per lane of total protein extract were fractionated on a 12% SDS-PAGE gel, transferred to a PVDF membrane, and probed with an anti-murine cdk4 antibody (Santa Cruz Biotech, CA, USA; diluted 1 : 2000). (b) Western blot analysis of the expression and phosphorylation status of the retinoblastoma gene product (Rb) in IL-6-treated M1, M1myb, M1myc, and M1myc/cdk4 cells. Fifty ug per lane of total protein extract were fractionated on a 6% SDS-PAGE gel, transferred to a PVDF membrane, and probed with an anti-human Rb monoclonal antibody (Pharmingen, CA, USA; diluted 1 : 1000). Signals were developed by using the enhanced chemiluminescence (ECL) Western blotting system (Amersham). ppRb indicates the hyperphosphorylated form and Rb is the hypophosphorylated form of the retinoblastoma gene product. (c) Growth kinetics after induction for differentiation with IL-6. Cells were seeded at 0.15´10⁶/ml in the presence of IL-6 (50 ng/ml) and the number of viable cells was determined at the indicated times. Each time point is the average of at least three experiments, with a standard deviation of up to 15%

Figure 5 Effect of Bcl-2 expression on survival of M1myc cells. (a) Deregulated expression of the Bcl-2 transgene in M1myc/Bcl-2 cells. Fifty ug per lane of total protein extract were fractionated on a 15% SDS-PAGE gel, transferred to a PVDF membrane, and probed with an anti-murine Bcl-2 antibody (Santa Cruz Biotech, CA, USA; diluted 1 : 2000). (b) Analysis up to day 3 following addition of IL6. (c) Analysis after 3 days following addition of IL6. For growth kinetics analysis, cells were seeded at 0.15´10⁶/ml in the presence or absence of IL-6 and the number of viable cells was determined at the indicated times. On day 3, M1, M1myc/Bc12+IL6, and M1myc+IL6 cells were diluted and reseeded at a concentration of 0.15´10⁶/ml in fresh medium with IL-6. Viable cell numbers were again determined at 1-day intervals. Genomic DNA was extracted from 1´10⁷ cells at the indicated time points, end-labeled, and resolved on an agarose gel (500 ng/lane) as described in Materials and methods. (d) Western blot of IL-6-treated cell lines indicating the absence of detectable levels of endogenous p53. Cell lysate from M1p53^ts cells was used as a positive control for the p53 antibody (Santa Cruz Biotech; diluted 1 : 1000)

Figure 6 (a) Wild type (BALBc) and p53^-/-(BALB/cJ-Trp53^tm1Tyj) BM was infected with either MSCVneo or MSCVmyc/neo, as described in Materials and methods. Cells were seeded in methyl cellulose with IL-3 (10%) and G418 (400 ug/ml). After 7 days colonies were picked and transferred to liquid culture in presence of G418 and either IL-3 (10%) or GM-CSF (100 ng/ml). After 5 days cells in liquid culture were analysed for per cent differentiated cells, per cent viable cells and DNA ladders. (b) MSCVneo and MSCVmyc/neo infected BM cells were established as in a. Colonies were picked from methylcellulose and transferred to liquid culture supplemented with G418 and GM-CSF (100 ng/ml). Cells were seeded at 0.15´10⁶/ml in the presence of 100 ng/ml GM-CSF and the number of viable cells was determined at the indicated times by trypan blue dye exclusion and counting in a hemocytometer. On day 3 cells were reseeded at 0.15´10⁶/ml. BM-neo and BMp53^-/-neo, cells infected with MSCV-neo; BM-myc and BMp53^-/-myc, cells infected with MSCV-myc-neo