Introduction

The v-ErbB oncogene is related in structure to the epidermal growth factor receptor (EGF-R) (Bruskin et al., 1990; McMahon et al., 1991; Navolanic et al., 2003). v-ErbB lacks most of the external ligand binding domain; however, it is constitutively activated and delivers a growth signal in the absence of ligand (McMahon et al., 1991; Navolanic et al., 2003). The v-ErbB oncogene was initially discovered in the genome of avian erythroblastosis virus (AEV) and is the viral homologue of one of the first cellular oncogenes discovered, c-ErbB (Weiss et al., 1984). v-ErbB will abrogate the requirement of erythroid progenitor cells for erythropoietin and stem cell factor (SCF) and block terminal differentiation (von Lindern et al., 2001). Overexpression of the EGF-R has been observed in many pathological situations (Navolanic et al., 2003). The EGF-R gene is amplified in certain tumors and this genetic abnormality is associated with malignancy. In certain human tumors (e.g. gliomas), there is a truncated form of the EGF-R referred to as EGFvIII, which is activated and resembles v-ErbB in its biological properties (Kuan et al., 2001). Elucidating the mechanisms of EGF-R-mediated signal transduction will increase our understanding of how deregulated signaling can result in malignant transformation.

The proliferation of many hematopoietic precursor cells is promoted by interleukin-3 (IL-3), granulocyte/macrophage colony-stimulating factor (GM-CSF), SCF (a.k.a., c-Kit-L) and Flt-3L (the ligand for the tyrosine kinase receptors Flt-2 and Flt-3 (Blalock et al., 1999; Chang et al., 2002; Lee and McCubrey, 2002). IL-3 and GM-CSF exert their biological activity by binding to the IL-3 and GM-CSF receptors, respectively (Kitamura et al., 1991; Sato et al., 1993; Ihle et al., 1995; Pritchard et al., 1995; Steelman et al., 1996; Blalock et al., 1999; McCubrey et al., 2000). These receptors are comprised of a ligand-specific -subunit, and a common -subunit (_c), which is essential for signal transduction (Kitamura et al., 1991; Sato et al., 1993; Ihle et al., 1995; Pritchard et al., 1995; Steelman et al., 1996; Blalock et al., 1999; McCubrey et al., 2000). After receptor ligation, the Shc protein becomes phosphorylated and forms a complex with the receptor (Larson et al., 2003; Steelman et al., 2004). Shc then recruits the Grb2/Sos complex to the _c chain, which subsequently results in Ras activation, leading to the stimulation of Raf and MEK and mobilization of the MAP (ERK) kinase cascade (Bosch et al., 1997; Blalock et al., 1999, 2000a, 2000b, 2001; Chang et al., 2003b, 2003c). Receptor activation also creates a docking site on the receptor for phosphatidylinositol 3-kinase (PI3K), which phosphorylates certain membrane lipids in the cell membrane. The phospholipids recruit certain kinases via their pleckstrin (PH) homology domains including phosphoinositol-dependent kinase (PDK) and Akt. PDK phosphorylates and activates Akt, a key kinase that regulates many proteins involved in the control of apoptosis (Blalock et al., 1999; Chang et al., 2003a).

In the following study, the effects of a conditional v-ErbB construct on the morphological transformation of NIH-3T3 cells and the factor dependence of hematopoietic cell lines were investigated. These studies used a novel v-ErbB:ER construct where its effects on growth and prevention of apoptosis could be directly measured. v-ErbB activity was required for the morphological transformation of NIH-3T3 cells and abrogation of cytokine dependence as these cells reverted to their nontransformed and cytokine-dependent states upon removal of -estradiol or 4-OH tamoxifen (4HT). Understanding how deregulated v-ErbB expression activates signaling pathways and prevents apoptosis will further our comprehension of this critical pathway in normal and neoplastic cell growth.

Results

Hormone-dependent transformation by conditionally active v-ErbB:ER

Transduction of chicken c-Erb-B/EGFR into AEV ES-4 led to the generation of a constitutively active, oncogenic form of the protein, v-ErbB. The AEV ES-4 v-ErbB protein differs from c-ErbB/EGFR by a 584 amino-acid N-terminal deletion, a 73 amino-acid C-terminal deletion and a 21 amino-acid internal deletion (Figure 1). To generate a conditionally active form of v-ErbB, DNA sequences encoding v-ErbB ES-4 were fused in frame to the sequences encoding the hormone binding domain of the human estrogen receptor (hbER) to generate v-ErbB:ER. The strategy used to generate v-ErbB:ER led to the further loss of 11 amino acids from the C-terminus of v-ErbB. Sequences encoding v-ErbB were subcloned into a series of replication-defective retrovirus vectors for expression in mammalian cells (Bruskin et al., 1990; McMahon et al., 1991; McMahon, 2001). MLV-based vectors expressing ErbB were originally described as having the capacity to transform fibroblasts (NIH-3T3 and Rat1) and certain hematopoeitic cell lines (Bruskin et al., 1990; Miller et al., 1990; McMahon et al., 1991). In separate, but unpublished experiments conducted by Dr Arthur Bruskin and JM Bishop in 1986, a deletion mutant of ErbB truncated at the same EcoRI site used to make ErbB:ER was uncompromised in fibroblast-transforming potential. Moreover, the experiments described below clearly demonstrate that the removal of 11 amino acids from the C-terminus of ErbB-ES4 that occurred in the construction of ErbB:ER did not compromise the ability of this oncogene to transform fibroblastic and hematopoietic cells in a hormone-dependent manner.

Figure 1.

Schematic diagram of chicken c-Erb-B/EGF-R, v-ErbB (ES-4) and construction of v-ErbB:ER. Details of the construction of v-ErbB:ER are described in the text

Full figure and legend (12K)

To test for hormone-dependent oncogenic transformation, v-ErbB:ER was expressed in NIH-3T3 cells (named NIH-3T3/v-ErbB:ER), a cell line permissive for v-ErbB transformation (Bruskin et al., 1990; McMahon et al., 1991). NIH-3T3/v-ErbB:ER cells were treated with -estradiol, 4HT or ethanol as a solvent control. Ethanol-treated cells had a normal flattened, nonrefractile cell morphology and displayed contact inhibition of growth (Figure 2, panel a). In contrast, addition of -estradiol or 4HT (panels b and c) for 16–24 h resulted in cells displaying a rounded, refractile morphology with abundant dendritic processes and a loss of contact inhibition.

Figure 2.

Hormone-dependent oncogenic transformation of NIH-3T3 cells by v-ErbB:ER. NIH-3T3 cells were infected with a retrovirus encoding v-ErbB:ER and resistance to geneticin (G418). These cells are named NIH-3T3/v-ErbB:ER. G418-resistant cells (10⁶ cells/60 mm Petri dish) were selected and treated with: (a) 0.1% (v/v) ethanol, (b) -estradiol or (c) 1 M 4HT for 24 h in phenol red free DMEM medium containing 5% CS FBS. In (d–f), 500 NIH-3T3/v-ErbB:ER cells were initially seeded in 60 mm plates and cultured for 2 weeks in phenol red free medium containing 5% CS FBS and 1 M -estradiol to allow focus formation to occur. After focus formation became visible, the cells were treated with: (d) 0.25% (v/v) DMSO, (e) 5 M U0126 or (f) 5 M LY294002 for 48 h in medium containing 1 M -estradiol. Cells were photographed using a Nikon TMS microscope at 100 magnification

Full figure and legend (86K)

NIH-3T3/v-ErbB:ER cells were also cultured, at initially lower cell densities, for 2 weeks in the presence of -estradiol in phenol red free medium containing charcoal-stripped (CS) fetal bovine serum (FBS). Under these conditions, the NIH-3T3/v-ErbB:ER cells formed foci of cells consistent with the morphological transformation (panel d), whereas in the absence of -estradiol these foci were not present. The elevated foci of cells could be reversed by treatment with the MEK inhibitor U0126 for 2 days (panel e). In contrast, treatment with the PI3K inhibitor LY294002 did not result in the reversal of elevated foci (panel f). These results demonstrated the importance of the MEK pathway in the maintenance of morphological transformation of NIH-3T3 cells by the v-ErbB:ER oncogene. Similar effects of MEK and PI3K inhibitors on v-ErbB-transformed fibroblasts have been observed (Fan et al., 2002).

Expression of the 97 kDa v-ErbB:ER chimera was assessed by Western blotting cell extracts from control (Neo) or v-ErbB:ER-expressing cells that were treated with ethanol, -estradiol or 4HT for 16 h prior to preparation of cell extracts (Figure 3, panel a). Extracts of cells expressing the 69 kDa Raf-1:ER protein were used a positive control (Samuels et al., 1993). The expression of v-ErbB:ER was readily detected in ethanol-treated cells; however, the addition of -estradiol or 4HT led to the appearance of forms of v-ErbB:ER with reduced mobility. The v-ErbB:ER protein did not display the same extent of hormone-dependent stabilization as that observed with Raf-1:ER (Samuels et al., 1993).

Figure 3.

Expression of v-ErbB:ER in NIH-3T3 cells. (a) Cell extracts were prepared from control virus-infected NIH-3T3 cells (Neo) or from NIH-3T3 cells expressing Raf-1:ER or v-ErbB:ER that were either treated with 0.1% (v/v) ethanol, 1 M -estradiol or 1 M 4HT for 24 h. Expression of kinase:ER fusion proteins was detected by Western blotting with anti-hbER antisera. The expression of ERK1/2 was used as an appropriate loading control. (b and c) NIH-3T3/v-ErbB:ER cells were treated with 1 M -estradiol for different periods of time. (b) Hormone-dependent tyrosine phosphorylation of v-ErbB:ER was assessed by Western blotting of anti-ErbB immunoprecipitates with an anti-phosphotyrosine monoclonal antibody (4G10). (c) Total tyrosine phosphorylation in whole-cell lysates was assessed by Western blotting of with the same anti-phosphotyrosine mAb

Full figure and legend (67K)

The species of v-ErbB:ER with reduced mobility observed after hormone addition are most likely not dimers, since this gel was run under reducing/denaturing conditions such that dimers should not survive the sample preparation. Since addition of hormone to cells expressing v-ErbB:ER leads to induced tyrosine phosphorylation of the fusion protein (see below), it seems highly likely that the species of v-ErbB:ER with reduced mobility observed after hormone addition are a consequence of phosphorylation. The mechanism of conditional activation of v-ErbB:ER most likely reflects hormone-dependent dimerization in intact cells. The major hormone-dependent dimerization domain in the ER is contained within the hormone binding domain and this domain displays hormone-dependent dimerization even when removed from the rest of the protein. Moreover, fusion of EGFR/v-ErbB or Raf to other conditional dimerization systems (FKBP12 and derivatives, GyrB) gives rise to conditionally active forms of the cognate protein kinase that have been published by several other research groups (Farrar et al., 1996; Luo et al., 1996; Muthuswamy et al., 1999).

The addition of estrogen analogs to cells expressing various protein kinase:ER fusion proteins leads to the rapid, post-translational activation of protein kinase activity (Samuels et al., 1993; McCubrey et al., 1998; McMahon, 2001). To determine if v-ErbB:ER was also rapidly activated upon hormone addition, NIH-3T3 cells expressing v-ErbB:ER were treated with -estradiol for different periods of time. Cell extracts were prepared and the tyrosine phosphorylation of v-ErbB:ER and other cellular proteins assessed by immunoprecipitation and Western blotting. The addition of -estradiol led to rapid tyrosine phosphorylation of v-ErbB:ER that was detected 60 min after hormone addition (panel b). This is consistent with the rapid activation of a number of other protein kinase:ER fusion proteins such as Raf-1:ER. Analysis of whole-cell extracts revealed a modest stimulation of tyrosine phosphorylation in response to v-ErbB:ER activation that is entirely consistent with the extent of tyrosine phosphorylation observed in NIH-3T3 cells transformed by constitutively active v-ErbB (panel c) (Bruskin et al., 1990; McMahon et al., 1991). These data clearly indicate that the v-ErbB:ER chimera is hormone dependent for biochemical activation of protein tyrosine kinase activity and phosphorylation, as well as for oncogenic cell transformation.

Conditional v-ErbB:ER expression abrogates the cytokine dependence of hematopoietic cells

Hematopoietic cell lines have been isolated that require IL-3 or in some cases GM-CSF for survival (Dexter et al., 1980; McKearn et al., 1985; Kitamura et al., 1989; Shelton et al., 2004a, 2004b and 2004c). Hence, IL-3 and GM-CSF are sometimes referred to as survival factors. The human TF-1 line is a cytokine-dependent (IL-3, GM-CSF) line isolated from a patient with erythroleukemia (Kitamura et al., 1989). FDC-P1 is an IL-3/GM-CSF-dependent cell line recovered from the bone marrow of normal DBA/2 mice and is representative of early hematopoietic precursor cells (Dexter et al., 1980). The IL-3-dependent FL5.12 cell line was isolated from fetal liver and resembles early hematopoietic cells (McKearn et al., 1985). Either IL-3 or GM-CSF is required for the survival of these cell lines and their removal results in apoptosis, programmed cell death. While these cell lines proliferate continuously in culture as long as IL-3 or GM-CSF is provided, they are non-tumorigenic upon injection into immunocompromised mice (McCubrey et al., 1989). The loss of cytokine dependency by hematopoietic cells is an important factor in the development of leukemias. Spontaneous factor-independent cells are rarely recovered from these cell lines, which makes them attractive model systems to analyse the effects various genes have on signal transduction and leukemogenesis (McCubrey et al., 1989, 1993). The following studies sought to investigate the effects of the conditional v-ErbB:ER oncoprotein on the cytokine dependence of hematopoietic cells. The human TF-1 and murine FDC-P1 and FL5.12 cells were infected with the retroviral vector encoding the v-ErbB:ER oncogene. Some cells were infected with a retroviral vector encoding the BCR-ABL-activated oncoprotein or an empty retroviral vector (LNL6) as positive and negative controls, respectively. As an additional negative control, cells were mock infected (Table 1).

Table 1 - Effects of activated oncogene expression on the cytokine dependence of hematopoietic cells^a.

Full table

At 2 days post-infection, the cells were plated in medium containing cytokines (IL-3 or GM-CSF) and G418, in medium containing G418 and -estradiol or in medium containing G418 but lacking cytokines and -estradiol. After a mock infection, no colonies growing in the presence of cytokine and G418 or -estradiol and G418 were recovered, indicating that the G418 concentration was sufficient to kill uninfected cells (Table 1). After infection of these cells with the empty pLNL6 retroviral vector, colonies were observed growing in the presence of cytokines and G418 (Table 1). However, no clones were isolated when the LNL6-infected cells were plated in medium lacking cytokines, indicating that infection with the retroviral vector was not sufficient to isolate cytokine-independent cells.

As expected, when the cells were infected with the v-ErbB:ER retrovirus, colonies growing in the presence of cytokine and G418 were readily observed. Moreover, v-ErbB:ER abrogated the cytokine dependence of these hematopoietic cells, as cells growing in the presence of -estradiol and G418 were recovered. Furthermore, v-ErbB:ER-responsive cells were readily recovered from the cytokine and G418 selected cells at an efficiency approaching 1, as determined by limiting dilution analysis, indicating that the v-ErbB:ER oncogene readily abrogated the cytokine dependence of these cells (data not presented). The activation of the conditional v-ErbB:ER oncogene was required for the isolation of cytokine-independent cells as no colonies were recovered when the cells were plated in the absence of -estradiol.

The effects of the BCR-ABL-activated tyrosine kinase on the cytokine dependence of these cells were determined as a control. BCR-ABL abrogated the cytokine dependence of all three cell lines. However, unlike the v-ErbB:ER-infected cells, BCR-ABL-transformed cells were recovered in the absence of -estradiol.

Conditional growth of v-ErbB:ER-transformed cells

The growth characteristics of the v-ErbB:ER-responsive hematopoietic cells were examined (Figure 4). The v-ErbB:ER-responsive FDC-P1, FL5.12 and TF-1 cells proliferated in response to cytokines or -estradiol. However, the cells did not proliferate in the absence of cytokines or -estradiol. Thus, the v-ErbB:ER protein conditionally abrogated the cytokine dependence of hematopoietic cells.

Figure 4.

Conditional growth of v-ErbB:ER-transformed cells. Growth of: (a) FD/v-ErbB:ER, (b) FL/v-ErbB:ER and (c) TF/v-ErbB:ER cells. Symbols: , solid squares=IL-3 or GM-CSF; , solid triangles=1 M -estradiol; and , solid circles=absence of both cytokine and -estradiol. These experiments were repeated three times and similar results were obtained

Full figure and legend (24K)

v-ErbB:ER activates the Raf/MEK/ERK and PI3K/Akt pathways

The Raf/MEK/ERK and PI3K/Akt pathways are central in the transmission of proliferative and antiapoptotic effects. To determine whether v-ErbB:ER induced the activation of the Raf/MEK/ERK and PI3K/Akt pathways, FL/v-ErbB:ER cells were deprived of -estradiol for 24 h to synchronize the cells, and then treated with IL-3 or 4HT for varying periods of time. The presence of activated forms of the MEK1, ERK and Akt were examined by Western blotting (Figure 5). Activated MEK, ERK and Akt were detected with phospho-specific antibodies (Ab) 15 min after IL-3 treatment. Activated forms of MEK and ERK were detected 30 min after 4HT stimulation. Activated Akt was detected 2 h after 4HT stimulation.

Figure 5.

v-ERBB:ER activates MEK, ERK and AKT phosphorylation. FL/v-ErbB:ER cells were cultured in the absence of 4HT for 24 h in phenol red free RPMI 1640 containing 5% CS FBS to avoid estrogenic effects of phenol red and endogenous estrogens present in FBS. The cells were then treated with DMSO, 5 M U0126, 5 M LY294002, 4 M AG1478 for 1 h and a portion of the cells were collected. Then, the cells were treated with either IL-3 or 500 nM 4HT for the indicated time points and the remaining protein lysates made. Western blots were performed with Ab, which recognized activation-specific and total proteins. This experiment was repeated three times and similar results were obtained. Similar results were observed with FD/v-ErbB:ER cells

Full figure and legend (169K)

The specificity of induction of these molecules was also examined by pretreatment of the cells with EGF-R, MEK and PI3K inhibitors. To determine whether the EGF-R inhibitor suppressed ERK and Akt activation, the FL/v-ErbB:ER cells were deprived of 4HT for 24 h, and then pretreated with the EGF-R inhibitor (4 M, AG1478) for 1 h before the addition of either IL-3 or 4HT, and then protein lysates were prepared from samples isolated at different time points after stimulation (Figure 5). Treatment with the EGF-R inhibitor suppressed MEK, ERK and Akt activation in response to 4HT treatment but not IL-3.

Treatment of FL/v-ErbB:ER cells with the MEK inhibitor U0126 suppressed the appearance of activated forms of ERK after either IL-3 or 4HT treatment. Interestingly, pre-treatment with the MEK inhibitor U0126 did not suppress the appearance of activated forms of MEK after either IL-3 or 4HT treatment. U0126 has been reported to suppress MEK activity and not MEK activation (Favata et al., 1998). Pretreatment of FL/v-ErbB:ER cells with U0126 did not suppress the activation of Akt in response to either IL-3 or 4HT treatment.

The effects of inhibition of the PI3K pathway were also examined. Pre-treatment of FL/v-ErbB:ER cells with LY294002 suppressed the appearance of activated forms of Akt after IL-3 treatment, but did not inhibit the appearance of activated forms of MEK or ERK after IL-3 treatment. Inhibition of the PI3K pathway by pretreatment of FL/v-ErbB:ER cells with LY294002 suppressed Akt activation and reduced the levels of activated MEK and ERK detected after 4HT treatment. In summary, the EGF-R inhibitor suppressed MEK, ERK and Akt activation in response to v-ErbB:ER activation, but not after IL-3 stimulation. In contrast the MEK and PI3K inhibitors suppressed both ERK and Akt activation in response to either IL-3 or v-ErbB:ER activation.

v-Erb-B:ER prevents apoptosis

The effects of v-ErbB:ER expression on the prevention of apoptosis were examined (Figure 6). In these experiments, the cytokine-dependent parental FDC-P1 and factor-independent FD/BCR-ABL cells were used as controls. When FDC-P1 or FD/v-ErbB:ER cells were cultured in the absence of IL-3 or 4HT for 48 h, 19 and 22% of the cells registered as viable (panels a and d, bottom left quadrants) and 80 and 70% of the cells registered as apoptotic (sum of annexin-positive cells in upper left and right quadrants in panels a and d). In contrast, when the cells were cultured in the presence of IL-3 for 48 h, 85 and 60% of the cells were viable (panels b and e). When these cells were cultured in medium containing 4HT, approximately 18 and 61% of the cells were viable and 81 and 34% of the cells were apoptotic (panels c and f). Thus, the expression of the activated v-ErbB:ER oncoprotein hindered apoptosis in the FD/v-ErbB:ER cells, but as expected, 4HT did not prevent apoptosis in FDC-P1 cells. In contrast, cells transformed with the BCR-ABL oncoprotein were viable in the absence and presence of IL-3 (panels g–i), demonstrating the effects of constitutive BCR-ABL oncoprotein expression on the prevention of apoptosis.

Figure 6.

Effects of v-Erb-B:ER and BCR-ABL on the prevention of apoptosis. FDC-P1, FD/v-ErbB:ER and FD/BCR-ABL cells were cultured for 2 days in the absence of growth factors, in the presence of IL-3 or in the presence 4HT for 2 days, and then annexin V/PI binding analysis was performed. Key: bottom left quadrant, annexin-negative, PI-negative 'viable cells'; top left quadrant, annexin-positive, PI-negative 'early apoptotic cells'; top right quadrant, annexin-positive, PI-positive, 'necrotic, late apoptotic cells'; bottom right quadrant, annexin-negative, PI-positive 'dead cells'. The apoptotic cells are the sum of the annexin-positive cells that are the top left and top right-hand quadrants. These experiments were repeated four times and similar results were obtained

Full figure and legend (69K)

Effects of EGFR, MEK and PI3K inhibitors on the induction of apoptosis

Previously, it was shown that v-ErbB:ER induced MEK, ERK and Akt activation, and pretreatment with an EGF-R inhibitor inhibited the appearance of activated forms of MEK, ERK and Akt (Figure 5). To determine the consequences EGF-R, MEK and PI3K inhibition on the induction of apoptosis, FDC-P1 and FD/v-ErbB:ER cells were treated with EGF-R, MEK and PI3K inhibitors and the extent of apoptosis was determined by annexin V/PI binding assays. As controls for the specificity of these inhibitors, the effects of AG1478, U0126 and LY294002 on cells transformed by BCR-ABL and Raf-1:ER (Hoyle et al., 2000) were also determined. The percentage of viable cells vs inhibitor concentration is plotted graphically in Figure 7. FDC-P1 cells were grown in medium (RPMI+10% FBS) supplemented with IL-3, FD/BCR-ABL were cultured in the absence of additional supplements and FD/v-ErbB:ER and FD/Raf-1:ER were cultured in medium supplemented with 4HT.

Figure 7.

Effects on EGFR, MEK and PI3K inhibitors on the induction of apoptosis. (a, c and e) FD/v-ErbB:ER cells in 4HT (solid triangles, ) and FDC-P1 cells in IL-3 (solid squares,) were cultured for 2 days in the presence of different concentrations of EGF-R (AG1478), MEK (U0126) and PI3K (LY294002) inhibitors, and the extent of apoptosis was determined by annexin V/PI binding as described in Figure 6. (b, d and f) FD/BCR-ABL cells in medium with no additional supplement (solid diamonds ) and FD/Raf-1:ER cells in 4HT (solid circles, ) were cultured for 2 days in the presence of different concentrations of EGF-R (AG1478), MEK (U0126) and PI3K (LY294002) inhibitors, and the extent of apoptosis was determined by annexin V/PI binding as described in Figure 6. These experiments were performed four to six times and averaged together. The statistically significant differences between the FD/v-ErbB:ER and FDC-P1 cells treated with the AG1478 inhibitor are indicated by ^**P<0.01, which were determined by a Student's t-test. The statistically significant differences between the FD/BCR-ABL and FD/Raf-1:ER cells treated with the AG1478 inhibitor are indicated by ^**P<0.01, which were determined by a Student's t-test

Full figure and legend (30K)

At concentrations approaching the IC₅₀ for AG1478 (3 nM), apoptosis was observed in the FD/v-ErbB:ER, but not in the FDC-P1 parental cell line (panel a) or cells transformed by either BCR-ABL or activated Raf-1 (panel b). Induction of apoptosis in FD/v-ErbB:ER cells was observed at as low as 400 pM AG1478. In contrast, apoptosis was only observed in FDC-P1, FD/BCR-ABL and FD/Raf-1:ER cells when they were cultured in 40 M AG1478. At concentrations approaching 5 M, the MEK (panels c and d) and PI3K (panels e and f) inhibitors induced apoptosis in all cell lines; however, less apoptosis was observed in FD/BCR-ABL cells when they were treated with the MEK inhibitor.

The effects of the EGF-R inhibitor on apoptosis and DNA synthesis in FD/v-ErbB:ER cells were further examined (Figure 8). AG1478 induced apoptosis and inhibited DNA synthesis when the FD/v-ErbB:ER cells were cultured in 4HT-containing medium, which activated the v-ErbB:ER, but not when they were cultured in response to IL-3 (Figure 8). Even when the FD/v-ErbB:ER cells were cultured in the presence of both IL-3 and 4HT, the EGF-R inhibitor did not induce apoptosis or inhibit DNA synthesis, indicating that this EGF-R inhibitor did not suppress cytokine-mediated prevention of apoptosis or DNA synthesis when the cells were grown in the presence of an activated v-ErbB:ER protein (data not presented). Thus, this inhibitor targeted Erb-B-mediated growth but had little effects on normal cytokine-mediated proliferation.

Figure 8.

Effects of the EGF-R inhibitor on the induction of apoptosis and DNA synthesis in FD/v-ErbB:ER cells. FD/v-ErbB:ER cells were cultured for 2 days in the presence of IL-3 (a), 4HT (b), IL-3 and 40 nM AG1478 (c), 4HT and 40 nM AG1478 (d) and annexin V/PI binding experiments performed. The extent of apoptosis (e) and fold inhibition of DNA syntheses (f) on the same cells treated with different concentrations of AG1478 in the presence and absence of IL-3 and 4HT were determined. The induction of apoptosis was determined when the cells were treated with the EGF-R inhibitor in the presence of 4HT (solid triangles, ) or IL-3 (solid squares, ). These experiments were performed three times and averaged together. The statistically significant differences between EGF-R inhibitor-treated cells and either IL-3- or 4HT-treated cells are indicated by ^**P<0.01, which were determined by a Student's t-test

Full figure and legend (52K)

Discussion

One of the goals of this research was to investigate the effects of conditional ErbB kinase activity upon the morphological transformation of NIH-3T3 cells and the abrogation of cytokine-dependence of hematopoietic cells. The novel v-ErbB:ER conditional oncogene used in these studies conditionally transformed NIH-3T3 cells and abrogated the cytokine-dependence of hematopoietic cells. FDC-P1, FL5.12 and TF-1 hematopoietic cells, which grew in response to v-ErbB:ER in the absence of exogenous cytokines were readily and efficiently isolated. Cell proliferation and prevention of apoptosis were determined to be dependent on v-ErbB, PI3K/Akt and Raf/MEK/ERK activities as treatment of the hematopoietic cells with inhibitors to these pathways prevented proliferation and promoted apoptosis. v-Erb-B:ER did not appear to induce autocrine cytokines, which exerted effects on DNA synthesis or ERK or Akt activation in either FDC-P1 or FL5.12 cells (data not presented).

The v-ErbB:ER- and BCR-ABL-activated tyrosine kinases were able to transform efficiently FL5.12 hematopoietic cells to cytokine independence. This may result from their ability to induce both the Raf/MEK/ERK and PI3K/Akt, as well as other downstream pathways. Recently, it was shown that overexpression of activated Raf or Akt by themselves did not abrogate the cytokine dependence of FL5.12 cells (Shelton et al., 2003); however, upon activation of both pathways, a synergistic effect on abrogation of cytokine dependence was observed. Thus, while FDC-P1 cells are transformed by activated Raf (Hoyle et al., 2000), FL5.12 cells require activated Raf and Akt to become factor independent (Shelton et al., 2003). In this study, we have demonstrated that v-ErbB:ER, which can activate both the Raf/MEK/ERK and PI3K/Akt pathways, can also readily abrogate the cytokine dependence of FL5.12 cells.

v-ErbB:ER activated downstream MEK, ERK and Akt. The EGF-R inhibitor AG1472 inhibited MEK, ERK and Akt activation when the cells were grown in response to v-ErbB:ER, but not when they were grown in response to IL-3. Furthermore, AG1478 induced apoptosis when the cells were grown in response to v-ErbB:ER, but not when they were grown in response to either IL-3 or both IL-3 and v-ErbB:ER (presence of both IL-3 and 4HT in the culture medium). Thus, IL-3 can induce downstream signal transduction pathways, which promote proliferation in the presence of an activated v-ErbB:ER and EGF-R inhibitor.

The MEK inhibitor U0126 inhibited the appearance of activated forms of ERK when the cells were cultured in either 4HT- or IL-3-containing medium. While the MEK inhibitor U0126 inhibited the appearance of activated ERK, it did not inhibit the appearance of activated MEK that was detected with the phospho-specific MEK1 Ab purchased from Cell Signaling. Similar results have been observed in cells that grow in response to activated Raf and Akt. Thus, our results indicate that U0126 does not prevent the detection of phosphorylated forms of MEK1 by the phospho-specific MEK1 Ab, but it does block downstream ERK activation, as activated forms of ERK1 and ERK2 were not detected. Indeed similar results with 40126 not blocking MEK phosphorylation but having an effect on ERK phosphorylation have been reported (Ahn et al., 2001). Some other studies with this inhibitor have indicated that U0126 blocks MEK1 activation (Davies et al., 2000). However, the major difference lies in how activation of MEK1 was detected. In the study by Davies et al., an in vitro kinase assay was performed with immunoprecipitated proteins, whereas our studies have used a phospho-specific Ab, which has only recently become commercially available. In contrast, when the FL/v-ErbB:ER cells were grown in response to v-ErbB:ER, the EGF-R inhibitor blocked both MEK and ERK activation, indicating that the EGF-R inhibitor blocked EGF-R-mediated MEK activation.

The PI3K inhibitor LY294002 inhibited the appearance of activated forms of Akt. This occurred when the cells were grown in response to either 4HT or IL-3. Thus, both the MEK and PI3K inhibitors had effects on cells when they were grown in response to either IL-3 or 4HT. Although MEK and PI3K inhibitors may eventually prove to be effective in inducing apoptosis of malignant clones, they may also effect the growth of normal cells which rely on functional MEK and PI3K activity for proliferation.

The v-ErbB:ER-transformed hematopoietic cells were very sensitive to the EGF-R inhibitor AG1478 when they were grown in response to v-ErbB:ER. In contrast, cytokine-dependent FDC-P1 cells or FD/v-ErbB:ER cells grown in response to IL-3 were 100- to 1000-fold less sensitive to the EGF-R inhibitor AG1478. Thus in this system, the v-ErbB:ER-responsive cells were very sensitive to the EGF-R inhibitor. Humans with lung cancers expressing mutated EGF-R are in many cases more responsive to inhibition of EGF-R using Iressa. In other transformed human cells, the situation is more complex as there may be multiple gene mutations, which are necessary for malignant growth. Furthermore, these studies point to the need to identify the precise mutation(s) that allows the malignant clones to proliferate in the absence of the previously required survival signals. Targeting these oncogenic proteins may prove more effective in therapy.

Materials and methods

Cell lines and growth factors

Cells were maintained in a humidified 5% CO₂ incubator. The human TF-1 line (Kitamura et al., 1989) was purchased from the ATCC (Rockville, MD, USA) and grown in RPMI 1640 medium (RPMI) containing 5% FBS (Atlanta Biologicals, Atlanta, GA, USA) supplemented with 1 ng/ml rGM-CSF (R&D Systems, Minneapolis, MN, USA). The IL-3-dependent murine FDC-P1 (Dexter et al., 1980) and FL5.12 cell lines (McKearn et al., 1985) were cultured in RPMI+5% FBS supplemented with 10% WEHI-3B(D^-) conditioned medium as a source of IL-3. v-ErbB:ER-responsive cells were grown in RPMI+5%, FBS+1 M, -estradiol or 500 nM 4HT (Sigma, St Louis, MO, USA), an ER antagonist that activates the v-ErbB:ER construct. v-ErbB:ER-transformed cells were also treated with the EGF-R inhibitor (AG 1478, Calbiochem, San Diego, CA, USA), the MEK inhibitor U0126 (Promega, Madison, WI, USA) or the PI3K inhibitors LY294002 (Calbiochem, San Diego, CA, USA), which were dissolved in dimethyl sulfoxide (DMSO, Sigma).

Assays of [³H]thymidine incorporation

Cells were incubated for 48 h in the presence or absence of either IL-3 or 4HT and different concentrations of the EGFR inhibitor as indicated. [³H]thymidine (6.7 Ci/mmol, NEN, Boston, MA, USA) was added for the last 4 h and then the levels of [³H]thymidine incorporation determined as described (McCubrey et al., 1993, 1995).

Construction of v-ErbB:ER

A plasmid containing the HE14 form of the hormone binding domain of the hbER (gift from Dr Martin Eilers) (Eilers et al., 1989) was digested with BamHI, which cuts uniquely at the 5'-end of the hbER coding sequences (Kumar et al., 1986; McMahon 2001). This BamHI site was filled in with the Klenow fragment of DNA polymerase and ligated to EcoRI octamer linkers (GGAATTCC). The resulting product was digested with EcoRI to liberate a 1 kb fragment encoding the human hbER. pGEM3Zf(+) encoding the 551 amino-acid Gag:ErbB fusion protein from AEV ES-4 (v-ErbB) (Bruskin et al., 1990; McMahon et al., 1991) was digested with EcoRI, which cuts 45 nucleotides upstream of the normal v-ErbB stop codon. This plasmid was ligated with the EcoRI fragment encoding hbER to generate a plasmid encoding a Gag:ErbB:ER fusion protein (v-ErbB:ER). The resulting fusion protein consists of amino acids 1–540 of v-ErbB ES-4 fused to amino acids 282–595 of human hbER. Note that the EcoRI site downstream of the v-ErbB:ER stop codon was converted to a XmnI site by partial digestion with EcoRI followed by end filling with Klenow and religation. DNA sequences encoding v-ErbB:ER were subcloned into pMLV-SVNeo3 as well as a number of additional replication-defective retrovirus vectors for expression in mammalian cells (Bruskin et al., 1990; McMahon et al., 1991).

Retroviral infection of cells

The neomycin-resistant (neo^r) pLNCv-ErbB:ER retrovirus was used in this study. As a control for some experiments, some cells were infected with a retrovirus encoding BCR-ABL (Clark et al., 1988).

NIH-3T3, TF-1, FDC-P1 and FL5.12 cells were infected with viral stocks as described (Samuels et al., 1993; McCubrey et al., 1998). Neo^r NIH-3T3 cells were isolated by selection in medium containing 500 g/ml G418. Neo^r TF-1, FDC-P1 or FL5.12 cells were isolated by selection in medium containing 2 mg/ml G418 (Sigma) in the presence of either cytokine (GM-CSF or IL-3) or -estradiol as described (McCubrey et al., 1998). The cells used for the following study were selected in cytokine and G418. They were determined to be v-ErbB:ER responsive by limiting dilution analysis by comparing their plating efficiency in the presence of -estradiol or IL-3. The nomenclature of the v-ErbB:ER-infected cells is TF/v-ErbB:ER, FD/v-ErbB:ER and FL/v-ErbB:ER. The nomenclature of BCR-ABL-infected cells is FD/BCR-ABL, FL/BCR-ABL and TF/BCR-ABL. FD/Raf-1:ER cells infected with the conditionally active Raf-1:ER gene have been described previously (Hoyle et al., 2000).

Preparation of cell extracts and analysis by Western blotting

Protein extracts and Western immunoblotting with phospho-specific and control Ab was performed as described (Blalock et al., 2003). ER Ab was purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Phospho-specific MEK, ERK, Akt Ab as well as Ab, which recognize total MEK, ERK and Akt, were purchased from Cell Signaling (Beverly, MA, USA). The -phosphotyrosine monoclonal 4G10 Ab was purchased from Upstate Bioltechnology (Lake Placid, NY, USA).

Apoptosis assays using annexin V/PI

Annexin V/PI binding experiments were preformed as described (Shelton et al., 2003).

References

1. Ahn NG, Nahreini TS, Tolwinski NS and Resing KA. (2001). Methods Enzymol., 332, 417–431. | Article | PubMed | ISI | ChemPort |
2. Blalock WL, Moye PW, Chang F, Pearce M, Steelman LS, McMahon M and McCubrey JA. (2000b). Leukemia, 14, 1080–1096. | Article | PubMed |
3. Blalock WL, Pearce M, Chang F, Lee J, Pohnert S, Burrows C, Steelman LS, Franklin RA, McMahon M and McCubrey J. (2001). Leukemia, 15, 794–807. | Article | PubMed | ChemPort |
4. Blalock WL, Pearce M, Steelman LS, Franklin RA, McCarthy SA, Cherwinski H, McMahon M and McCubrey JA. (2000a). Oncogene, 19, 526–536. | Article | PubMed | ISI | ChemPort |
5. Blalock WL, Steelman LS, Shelton JG, Moye PW, Lee JT, Franklin RA, Mirza A, McMahon M, White MK and McCubrey JA. (2003). Leukemia, 17, 1058–1068. | Article | PubMed | ChemPort |
6. Blalock WL, Weinstein-Oppenheimer C, Chang F, Hoyle PE, Wang XY, Algate PA, Franklin RA, Oberhaus SM, Steelman LS and McCubrey JA. (1999). Leukemia, 13, 1109–1166. | Article | PubMed | ISI | ChemPort |
7. Bosch E, Cherwinski H, Peterson D and McMahon M. (1997). Oncogene, 11, 1021–1034. | Article |
8. Bruskin A, Jackson J, Bishop JM, McCarley DJ and Schatzman RC. (1990). Oncogene, 5, 15–24. | PubMed |
9. Chang F, Lee JT, Navolanic PM, Steelman JG, Blalock WL, Franklin RA and McCubrey JA. (2003a). Leukemia, 17, 590–603. | Article | PubMed | ISI | ChemPort |
10. Chang F, Steelman LS, Lee JT, Shelton JG, Navolanic PM, Blalock WL, Franklin RA and McCubrey JA. (2003c). Leukemia, 17, 1263–1294. | Article | PubMed | ISI | ChemPort |
11. Chang F, Steelman LS and McCubrey JA. (2002). Cell Cycle, 1, 220–227. | PubMed | ChemPort |
12. Chang F, Steelman LS, Shelton JG, Lee JT, Navolanic PN, Blalock WL, Franklin RA and McCubrey JA. (2003b). Int. J. Oncol., 22, 469–480. | PubMed | ISI | ChemPort |
13. Clark SS, McLaughin J, Timmons M, Pendergast AM, Ben-Neriah Y, Dow LW, Crist W, Rovera G, Smith SD and Witte ON. (1988). Science, 239, 775–777. | Article | PubMed | ISI | ChemPort |
14. Davies SP, Reddy H, Caivano M and Cohen P. (2000). Biochem. J., 351, 95–105. | Article | PubMed | ISI | ChemPort |
15. Dexter TM, Garland J, Scott D, Scolnick E and Metcalf D. (1980). J. Exp. Med., 152, 1036–1047. | Article | PubMed | ISI | ChemPort |
16. Eilers M, Picard D, Yamamoto KR and Bishop JM. (1989). Nature, 340, 66–68. | Article | PubMed | ISI | ChemPort |
17. Fan QW, Zhang C, Shokat KM and Weiss WA. (2002). Curr. Biol., 12, 1386–1394. | Article | PubMed | ISI | ChemPort |
18. Farrar MA, Alberol-Ila and Perlmutter RM. (1996). Nature, 383, 178–181. | Article | PubMed | ISI | ChemPort |
19. Favata MF, Horiuchi KY, Manos EJ, Daulerio AJ, Stradley DA, Feeser WS, Van Dyk DE, Pitts WJ, Earl RA, Hobbs F, Copeland RA, Magolda RL, Scherle PA and Trzaskos JM. (1998). J. Biol. Chem., 273, 18623–18632. | Article | PubMed | ISI | ChemPort |
20. Hoyle PE, Moye PW, Steelman LS, Blalock WL, Franklin RA, Pearce M, Cherwinski H, Bosch E, McMahon M and McCubrey JA. (2000). Leukemia, 14, 642–656. | Article | PubMed | ISI | ChemPort |
21. Ihle JN, Witthuhn BA, Quelle FW, Yamamoto K and Silvennoinen O. (1995). Annu. Rev. Immunol., 13, 369–399. | Article | PubMed | ISI | ChemPort |
22. Kitamura T, Sato M, Arai K-I and Miyajima A. (1991). Cell, 66, 1165–1174. | Article | PubMed | ISI | ChemPort |
23. Kitamura T, Tange T, Terasawa T, Chiba S, Kuwaki T, Miyagawa K, Piao YF, Miyazono K, Urabe A and Takaku F. (1989). J. Cell. Physiol., 140, 323–334. | Article | PubMed | ISI | ChemPort |
24. Kuan CT, Wilstrand CJ and Bigner DD. (2001). Brain Tumor Pathol., 17, 71–78.
25. Kumar V, Green S, Staub A and Chambon P. (1986). EMBO J., 5, 2231–2236. | PubMed | ISI | ChemPort |
26. Larson RA, Daley GQ, Schiffer CA, Porcu P, Pui CH, Marie JP, Steelman LS, Bertrand FE and McCubrey JA. (2003). Leukemia, 17, 2358–2382. | Article | PubMed |
27. Lee Jr JT and McCubrey JA. (2002). Leukemia, 16, 486–507. | Article | PubMed | ISI | ChemPort |
28. Luo Z, Tzivion G, Belshaw PJ, Vavvas D, Marshall M and Avruch J. (1996). Nature, 383, 181–185. | Article | PubMed | ISI | ChemPort |
29. McCubrey JA, Holland G, McKearn J and Risser R. (1989). Oncogene Res., 4, 97–109.
30. McCubrey JA, May WS, Duronio V and Mufson A. (2000). Leukemia, 14, 9–21. | Article | PubMed | ISI | ChemPort |
31. McCubrey JA, Smith SR, Algate PA, deVente JE, White MK and Steelman LS. (1993). Oncogene, 8, 2905–2915. | PubMed | ChemPort |
32. McCubrey JA, Steelman LS, Hoyle PA, Blalock WL, Weinstein-Oppenheimer CR, Franklin RA, Cherwinski H, Bosch E and McMahon M. (1998). Leukemia, 12, 1903–1929. | Article | PubMed | ISI | ChemPort |
33. McCubrey JA, Steelman LS, Wang XY, Davidian EW, Hoyle PE, White C, Prevost KD, Algate PA, Robbins P, Mylott D and White MK. (1995). Int. J. Oncol., 7, 573–586.
34. McKearn JP, McCubrey JA and Fagg B. (1985). Proc. Natl. Acad. Sci. USA, 82, 7414–7418.
35. McMahon M. (2001). Methods Enzymol., 332, 401–417. | Article | PubMed | ISI | ChemPort |
36. McMahon M, Schatzman RC and Bishop JM. (1991). Mol. Cell. Biol., 11, 4760–4770. | PubMed |
37. Miller M, Kennewell A, Takayama Y, Bruskin A, Bishop JM, Johnson G and Symonds G. (1990). Oncogene, 5, 1125–1131. | PubMed |
38. Muthuswamy SK, Gilman M and Brugge JS. (1999). Mol. Cell. Biol., 19, 6845–6857. | PubMed | ISI | ChemPort |
39. Navolanic PM, Steelman LS and McCubrey JA. (2003). Int. J. Oncol., 22, 237–252. | PubMed | ISI | ChemPort |
40. Pritchard CA, Samuels ML, Bosch E and McMahon M. (1995). Mol. Cell. Biol., 15, 9430–9442.
41. Samuels ML, Weber MJ, Bishop JM and McMahon M. (1993). Mol. Cell. Biol., 13, 6241–6252. | PubMed | ISI | ChemPort |
42. Sato N, Sakamaki K, Terada N, Arai K and Miyajima A. (1993). EMBO J., 12, 4181–4189. | PubMed | ISI | ChemPort |
43. Shelton JG, Blalock WL, White ER, Steelman LS and McCubrey JA. (2004a). Cell Cycle, 3, 503–512. | PubMed | ChemPort |
44. Shelton JG, Chang F, Lee JT, Franklin RA, Steelman LS and McCubrey JA. (2004b). Cell Cycle, 3, 189–196.
45. Shelton JG, Steelman LS, Lee JT, Knapp SL, Blalock WL, Moye PM, Franklin RA, Pohnert SC, Mizra AM, McMahon M and McCubrey JA. (2003). Oncogene, 24, 2478–2492. | Article |
46. Shelton JG, Steelman LS, White ER and McCubrey JA. (2004c). Cell Cycle, 3, 372–379.
47. Steelman LS, Algate PA, Blalock WL, Wang XY, Prevost KD, Hoyle PE and McCubrey JA. (1996). Leukemia, 10, 528–542. | PubMed |
48. Steelman LS, Pohnert SC, Shelton JG, Franklin RA, Bertrand FE and McCubrey JA. (2004c). Leukemia, 14, 189–218. | Article |
49. Von Lindern M, Deiner EM, Dolznig H, Parren-Van Amelsvoort M, Hayman MJ, Mullner EW and Beug H. (2001). Oncogene, 21, 3651–3664.
50. Weiss RA, Teich NM, Varmus HE and Coffin JM. (1984). RNA Tumor Viruses 2nd edn. Cold Spring Harbor Press: Cold Spring Harbor, New York, 1292pp.

Acknowledgements

This work was supported in part by Grants (R01CA51025 and R01CA98185) from the NCI (National Cancer Institute, National Institutes of Health) to JAM and Grant (9930099N) from the American Heart Association to RAF. MM is especially grateful to J Michael Bishop, who provided unwavering enthusiasm, support and advice both at the initiation of these studies and on an ongoing basis since.