Suppression of Apoptosis Induced by Growth Factor Withdrawal by an Oncogenic Form of c-Cbl*

The v-Cbl oncogene induces myeloid and B-cell leukemia; however, the mechanism by which transformation occurs is not understood. An oncogenic form of c-Cbl (Cbl-ΔY371) was expressed in the interleukin-3 (IL-3)-dependent cell line 32Dcl3 to determine whether it was able to induce growth factor-independent proliferation. We were unable to isolate clones of transfected 32Dcl3 cells expressing Cbl-ΔY371 that proliferated in the absence of IL-3. In contrast, 32Dcl3/Cbl-ΔY371 cells did not undergo apoptosis like parental 32Dcl3 cells when cultured in the absence of IL-3. Both 32Dcl3 and 32D/CblΔY371 cells arrested in G1 when cultured in the absence of IL-3. Approximately 18% of the 32Dcl3 cells cultured in the absence of IL-3 for 24 h were present in a sub-G1 fraction, while only 4% of the 32D/Cbl-ΔY371 and 2% of the 32D/Bcl-2 cells were found in a sub-G1 fraction. There was no difference in the pattern of tyrosine-phosphorylated proteins observed following stimulation of either cell type with IL-3. The phosphorylation of JAK2, STAT5, and endogenous c-Cbl was identical in both cell types. No differences were detected in the activation of Akt, ERK1, or ERK2 in unstimulated or IL-3-stimulated 32D/Cbl-ΔY371 cells compared with parental 32Dcl3 cells. Likewise, there was no difference in the pattern of phosphorylation of JAK2, STAT5, ERK1, ERK2, or Akt when 32Dcl3 and 32D/CblDY371 cells were withdrawn from medium containing IL-3. The protein levels of various Bcl-2 family members were examined in cells grown in the absence or presence of IL-3. We observed a consistent increased amount of Bcl-2 protein in five different clones of 32D/Cbl-ΔY317 cells. These data suggest that the Cbl-ΔY371 mutant may suppress apoptosis by a mechanism that involves the overexpression of Bcl-2. Consistent with this result, activation of caspase-3 was suppressed in 32D/Cbl-ΔY371 cells cultured in the absence of IL-3 compared with 32Dcl3 cells cultured under the same conditions.

The c-Cbl proto-oncogene has attracted considerable attention in recent years, since it becomes phosphorylated on tyrosine residues following activation of a wide variety of cell sur-face receptors. Receptors whose activation induces the phosphorylation of Cbl include the T cell receptor (1)(2)(3)(4)(5)(6), the B cell receptor (7,8), the Fc receptor (9), and the receptors for numerous growth factors including epidermal growth factor (10 -14), colony-stimulating factor-1 (15), erythropoietin (16,17), interleukin-3 (IL-3) 1 (16,18), granulocyte-macrophage colony-stimulating factor (17), thrombopoietin (19), and prolactin (20). Cbl is a 120-kDa protein with 22 tyrosine residues, an amino-terminal region able to bind to phosphotyrosine (21,22), and a proline-rich region with numerous Src homology 3 binding sites. The large size of the protein suggests that it may function as a large adapter or scaffolding molecule that could regulate the activation of other downstream signaling molecules in a manner similar to that of insulin-receptor substrate-1 (23). Other studies have suggested that Cbl functions as a negative regulator of signaling, perhaps by stimulating the down-regulation of growth factor receptors (24,25). Several recent studies have suggested that down-regulation of growth factor receptors requires the RING finger motif of Cbl functioning as an E2-dependent ubiquitin-protein ligase inducing the ubiquitination of growth factor receptors (24, 26 -29). Src-like kinases (3, 7, 18, 30 -32), Syk and/or ZAP-70 (4,21,33), phosphatidylinositol 3-kinase (PI 3-kinase) (2,3,7,18,31,34,35), and several small adapter proteins including Crk, Shc, and Grb2 (1,3,7,8,(35)(36)(37) have been observed to be associated with Cbl in either a constitutive or ligand-induced manner. It has been suggested that either the Syk/ZAP-70 family of tyrosine kinases (38) or Src-like kinases (38 -41) might be responsible for the phosphorylation of Cbl. We have suggested that the phosphorylation of Cbl could regulate the activation of PI 3-kinases in response to certain stimuli such as cytokines like IL-3 and prolactin (18,20,41). Since phosphatidylinositol 3-kinase is thought to regulate the antiapoptotic protein kinase Akt (42)(43)(44)(45)(46)(47)(48), Cbl might represent an integration point for proliferative and antiapoptotic signaling pathways. Evidence demonstrating that Cbl plays a critical role in these functions is lacking, particularly since mice bearing homologous loss of both alleles encoding c-Cbl have a relatively mild phenotype (49).
Cbl was first discovered as the oncogene present in the Cas NS-1 retrovirus, which has been shown to induce B-cell lymphomas and myeloid leukemia in mice (50). The v-cbl oncogene is also capable of transforming NIH3T3 cells in vitro, and these transformed cells are tumorigenic when injected into nude mice (51). The v-Cbl oncogene consists of the amino-terminal end of the gag gene fused to the amino-terminal end of c-Cbl encompassing the amino-terminal phosphotyrosine binding region but lacking the ring finger motif, the proline-rich region, and the C-terminal end, which contains several major phosphorylation sites (50,52). Other oncogenic forms of c-Cbl have been discovered (51). The murine pre-B cell line 70Z expresses an oncogenic form of Cbl in which 17 amino acids, from 366 to 382, have been deleted (51). Studies by Langdon and colleagues have demonstrated that deletion of either of two single amino acids within this 17-amino acid region results in the oncogenic activation of c-Cbl (51). In this study, we have utilized one of these mutants, Cbl⌬Y371, in which Tyr 371 has been deleted.
The expression of oncogenes in hematopoietic cells results in numerous changes, including induction of growth factor-independent proliferation (53)(54)(55)(56), inhibition of terminal differentiation (53,57,58), and suppression of apoptosis (59,60). Recent studies have demonstrated that the expression of oncogenic forms of Cbl in fibroblasts results in an increase in the number of tyrosine-phosphorylated proteins, suggesting that one or more tyrosine kinases have been activated by v-Cbl (61). We have previously studied the ability of two activated tyrosine kinases, v-Src and BCR-ABL, to induce growth factorindependent proliferation of growth factor-dependent cell lines (53,54). In both cases, we were able to provide evidence that these oncogenes induced the autocrine production of growth factors such as IL-3 or granulocyte-macrophage colony-stimulating factor. These oncogenes are also able to block the ability of granulocyte colony-stimulating factor to induce terminal differentiation of hematopoietic progenitor cell lines such as 32Dcl3 (53,57). Therefore, we hypothesized that oncogenic forms of Cbl would induce growth factor-independent proliferation of factor-dependent cells in a manner like that observed with v-Src. Contrary to our expectation, we observed instead that the Cbl⌬Y371 deletion mutant suppressed apoptosis induced by growth factor withdrawal. The mechanism by which apoptosis is suppressed appears to involve an increase in the level of the antiapoptotic protein Bcl-2.

MATERIALS AND METHODS
Cells and Cell Culture-The 32Dcl3 cell line was obtained from Dr. Joel Greenberger (University of Pittsburgh), and their cultivation has been described recently (18). Fetal calf serum was from Summit Biotechnology (Fort Collins, CO). Recombinant murine IL-3 (mIL-3) was obtained from Becton-Dickinson/Collaborative Biotechnology, Inc. (Franklin Lakes, NJ). All other media components were from Life Technologies, Inc.
32Dcl3 cells expressing the various mutants of c-Cbl were generated by introducing the plasmid DNAs of interest into 32Dcl3 cells with a Cell-Porator (1000 V/cm at 800 microfarads) (Life Technologies). Electroporation chambers used had a 0.4-cm gap between the electrodes, and cells were resuspended in Dulbecco's modified Eagle's medium without any additional supplements at room temperature. The Cbl⌬Y371 mutant was expressed in the pZEN-Neo vector (51), which contains the neomycin resistance marker as a dominant selectable marker. A Bcl-2 expression vector was constructed in which the coding sequence of Bcl-2 was inserted into the pcDNA3 expression vector (InVitrogen, Carlsbad, CA). Following electroporation, the cells were cultured in the presence of IL-3 for 2-3 days. Transfected cells were then selected by the addition of 1.0 mg/ml G418 (Life Technologies). G418-resistant cells were allowed to expand in the presence of IL-3 for 3 days, after which the drug was removed. Conditions that allow the isolation of growth factor-independent cells have been previously described (53,54). Single cell clones were then isolated following growth of the cells at limiting dilution in semisolid media containing 0.6% Sea-Plaque agarose (FMC Corp., Freeport, ME). Isolated colonies were picked from soft agar, expanded in liquid culture, and used in the described experiments. Multiple clones of 32D/Cbl⌬Y371cells were obtained from two separate electroporations to eliminate the possibility of examining "sister" clones. All clones used in these studies were negative for mycoplasma; this is important, since a recent report indicates that mycoplasma can suppress apoptosis of 32Dcl3 and induce growth factor-independent proliferation (62).
Cells were cultured for 16 h in media supplemented with 7.5% fetal calf serum to reduce the basal level of tyrosine-phosphorylated proteins prior to stimulation with recombinant mIL-3 for the indicated periods of time. In one set of experiments, cells were removed from IL-3 and starved for 0 -8 h to determine the kinetics with which phosphorylated proteins disappeared.
Immunoprecipitation and Immunoblotting-Cells to be immunoprecipitated were lysed in EB (50 mM NaCl, 10 mM Tris, pH 7.4, 5 mM EDTA, 50 mM NaF, 1% Triton X-100, 1 mM sodium orthovanadate with 100 units/ml Kallikrein inhibitor), and the lysates were clarified by spinning at 13,000 rpm in a Savant RCF13K refrigerated microcentrifuge for 30 min. A 1-g amount of the indicated antibody was added to a cell lysate made from 2 ϫ 10 7 32Dcl3 cells in a final volume of 1 ml and placed on a rocking platform for 1 h at 4°C. The immune complexes were collected by adding 30 l of Pansorbin (Calbiochem) to each immunoprecipitate for 1 h. The bound proteins were washed three times with lysis buffer, and the immunoprecipitated proteins were resolved by SDS-polyacrylamide gel electrophoresis. The resolved proteins were electrotransferred to Immobilon membranes (Millipore Corp., Bedford, MA). Detection of proteins by immunoblotting was conducted using the enhanced chemiluminescence lighting (ECL) system according to the manufacturer's recommendations (Amersham Pharmacia Biotech). Agarose-conjugated anti-phosphotyrosine monoclonal antibody 4G10, rabbit anti-JAK2, and sheep anti-Akt were obtained from Upstate Biotechnology, Inc. (Lake Placid, NY). Anti-phospho-Akt was obtained from New England Biolabs (Beverly, MA), and anti-phospho-ERK was obtained from Promega Biotechnology (Madison, WI). Polyclonal antibodies directed against Cbl, ERK1/2, Bcl-2, Bax, Bcl-X L , Bak, and A1 were obtained from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA). Polyclonal antibodies to Mcl1 and Bad were obtained from Transduction Laboratories (Lexington, KY). A monoclonal antibody directed against the influenza HA epitope tag (clone 12CA5) was obtained from Roche Molecular Biochemicals. A monoclonal antibody 4G10 directed against phosphotyrosine was kindly provided by Dr. Brian Druker (University of Oregon Health Sciences Center, Portland, OR). A rabbit polyclonal antibody directed against STAT5 was provided by Andrew Larner (Cleveland Clinic, Cleveland, OH). Nonimmune rabbit serum was obtained from our own nonimmunized animals.
Proliferation Assays-Cultures were initiated at 2 ϫ 10 5 cells/ml in medium that either contained or lacked 50 units/ml recombinant murine IL-3 (rmIL-3). The proliferation of 32Dcl3 cells was monitored by direct counting of viable cells, which excluded trypan blue. Results were plotted as the number of viable cells at each time point examined. In one study, 100 nM wortmannin (Calbiochem) was added to each set of cells, and the proliferation of cells was monitored as described. Controls for this study included cultures with the same concentration of ethanol, the solvent in which the wortmannin was dissolved, as that present in the wortmannin-treated cultures.
Cell Cycle Analysis-Cells were cultured in the presence or absence of 50 units/ml rmIL-3 for varying periods of time. Samples of cells were withdrawn at varying time points, and the cells were pelleted by centrifugation at 1,500 rpm for 5 min in a refrigerated centrifuge. The supernatant fluid was removed, and the pellet was washed once with phosphate-buffered saline. The supernatant fluid was removed, and the pellet was vortexed briefly. A 0.5-ml volume of saponin/propridium iodide solution was added (PBS containing 0.3% saponin, 25 g/ml propidium iodide, 10 mM EDTA, 0.2 mg/ml RNase) and incubated for 10 min at room temperature. Stained cells were stored for up to 24 h at 4°C before analysis. Ten thousand stained cells were analyzed with a Coulter XL flow cytometer (Hialeah, FL). Cell cycle modeling was performed with the ModFit software package (Verity House Software, Topsham, ME).
Assay for Caspase-3 Activity-The Caspase-3 Cellular Activity Assay Kit PLUS (Catalogue number AK-703) from BIOMOL (Plymouth Meeting, PA) was used according to the manufacturer's suggestions. At varying time points, a sample of cells used was removed and pelleted by centrifugation. The cells were lysed in cell lysis buffer (50 mM HEPES, pH 7.4, 100 mM NaCl, 0.1% CHAPS, 10 mM dithiothreitol, 1 mM EDTA, 10% glycerol). Following complete lysis of the cells, the lysate was clarified by centrifugation at 10,000 ϫ g for 10 min at 4°C, and the supernatant was frozen at Ϫ70°C until all the samples were collected. Caspase assays were performed as outlined by the manufacturer, and the cleavage of the substrate, N-acetyl-Asp-Glu-Val-Asp-p-nitroanaline), was monitored at 1-min intervals on a microtiter plate reader at 405 nm. Data were plotted as A 405 versus time for each sample. The initial time period over which change in OD versus time was linear was used to calculate the specific activity of the caspase present in each sample, in comparison with standard samples of caspase-3.

RESULTS
To determine whether oncogenic forms of Cbl could induce 32Dcl3 cells to become growth factor-independent, these cells were transfected with vectors encoding a panel of Cbl mutants, and single cell clones were isolated following growth in soft agar. The oncogenic forms used included v-Cbl, which is present in the Cas-NS-1 retrovirus (50); the 70Z form of c-Cbl, which lacks a 17-amino acid region at the N-terminal of the ring finger motif (51); and two deletion mutants lacking either tyrosine residue Tyr 368 or Tyr 371 , which are also oncogenic (51) (Fig. 1). In this study, we focused upon the Cbl⌬Y371 mutant, although it appears that the other oncogenic mutants behave in a similar manner (data not shown). Expression of the epitopetagged Cbl⌬Y371 protein could be observed by probing an immunoblot of whole cell lysates with an anti-HA monoclonal antibody (Fig. 2).
Our attempts to isolate cells that were able to proliferate in growth factor-free media were completely unsuccessful; however, we did notice that 32D/Cbl⌬Y371 cells did not die as rapidly as the untransfected control 32Dcl3 cells. Normally, we cannot detect any viable 32Dcl3 cells after 3 days of culture in media lacking IL-3; however, viable 32D/Cbl⌬Y371 cells could be found after 2 weeks, when cultured in the absence of IL-3. This observation suggested that we quantitate the proliferation of these cells in media that either contained or lacked exogenous IL-3. As shown in Fig. 3, 32Dcl3 cells proliferate when cultured in the presence of 50 units/ml recombinant murine IL-3; however, there is a rapid decrease in the number of viable cells when they are cultured in the absence of IL-3. The 32D/ Cbl⌬Y371 clone 9 cells cultured in the presence of IL-3 proliferated in a manner similar to that seen with 32Dcl3 cells; however, when cultured in the absence of IL-3, the number of viable 32D/Cbl⌬Y371 clone 9 cells did not decrease as rapidly. As a control for these studies, we compared the proliferation of the 32D/Cbl⌬Y371 cells to 32Dcl3 cells transfected with a Bcl-2 expression vector. Bcl-2 has been reported to suppress apoptosis induced by growth factor withdrawal (63-65). The 32D/ Bcl-2 clone 2 cells behaved in a manner similar to that observed with the 32D/Cbl⌬Y371 clone 9 cells in that they proliferated in the presence of IL-3, and the number of viable cells did not rapidly decrease when these cells were cultured in the absence of IL-3. This suggests that the Cbl⌬Y371 mutant is able to suppress apoptosis induced by growth factor withdrawal in a manner reminiscent of that observed in cells that overexpress Bcl-2; however, there must be some difference in the effect of Bcl-2 versus Cbl⌬Y371, since there was a transient 2-fold increase in the number of 32D/Bcl-2 cells when cultured in the absence of IL-3 that was not observed in 32D/Cbl⌬Y371 cells. The data shown in Fig. 3 are representative of three different studies, and although only one clone of 32D/Cbl⌬Y371 cells is shown in Fig. 3, similar results were obtained with four other independently derived clones (data not shown). 32Dcl3 cells transfected with vector alone (pLJ, pZEN-Neo, or pcDNA3) behaved in a manner identical to that observed with parental 32Dcl3 cells; i.e. they underwent apoptosis when cultured in the absence of IL-3 (53, 54, 66) (data not shown). A recent study has suggested that mycoplasma can suppress apoptosis of 32Dcl3 cells induced by growth factor withdrawal and allows these cells to grow in the absence of IL-3 (62). All of the different clones used in this study have been demonstrated to be free of mycoplasma (data not shown), and none of the 32D/ Cbl⌬Y371 cells proliferate in the absence of IL-3 (Fig. 3). There-  The structural domains of c-Cbl are noted. From the N-terminal end these motifs include a glycine rich-region, seven histidines, a putative nuclear localization signal within a basically charged region, a ring finger motif, two acidic domains that flank a proline-rich region, and a leucine zipper motif near the C-terminal end of the protein. Amino acids 1-357 encode a phosphotyrosine-binding domain with a cryptic Src homology 2 domain-like structure (80). The end of the v-Cbl oncogene is noted; the amino-terminal region of Cbl usptream of this stop site is included in v-Cbl. Also noted is the location of the 17 amino acids (aa) that are deleted in the 70Z form of Cbl. The location of tyrosine 371, which is deleted in the oncogenic form of Cbl used in this study, is indicated by the boldface Y at the bottom. Plus signs, the basic region; minus signs, the acidic region; PP, the proline-rich region.
fore, the results we have described cannot be due to mycoplasma contamination.
Similar studies were conducted with three other oncogenic forms of c-Cbl: v-Cbl, the 70Z mutant of c-Cbl, and Cbl⌬Y368. All of these oncogenic forms of c-Cbl were also able to suppress apoptosis as described above (data not shown). Similar results have also been obtained with v-Cbl in the laboratory of S. J. Corey. 2 It is of interest to note that the G306E mutant of v-cbl, which is not able to transform NIH3T3 cells (61), was able to suppress apoptosis of 32Dcl3 cells following IL-3 withdrawal as effectively as v-cbl (data not shown). This point mutation inactivates the phosphotyrosine-binding activity of v-Cbl, suggesting that binding to phosphotyrosine-containing proteins is not required for suppression of apoptosis.
To determine the effect of IL-3 withdrawal upon the cell cycle status of the cells examined in Fig. 3, 32Dcl3, 32D/Cbl⌬Y371 clone 9, and 32D/Bcl-2 clone 2 cells were cultured in the absence of IL-3 for 0 -24 h. Cells were withdrawn at 8, 16, and 24 h, fixed, and stained with saponin/propridium iodide for flow cytometric analysis. As shown in Fig. 4 and Table I, the withdrawal of IL-3 caused an increase in the number of cells in the G 1 /G 0 phase of the cell cycle for all three cell lines. There was a very dramatic increase in the number of 32Dcl3 cells present in the sub-G 1 fraction after 24 h, compared with the number of 32D/Cbl⌬Y371 clone 9 and 32D/Bcl-2 clone 2 cells present in this fraction at the same time. ModFit analysis indicates that 18% of the 32Dcl3 cells were apoptotic after 24 h in the absence of IL-3, while only 4 and 2% of the 32D/Cbl⌬Y371 clone 9 and 32D/Bcl-2 clone 2 cells, respectively, were apoptotic at that same time. This provides additional evidence that Cbl⌬Y371 functions to suppress apoptosis induced by IL-3 withdrawal.
Oncogenic Cbl Does Not Induce an Increase in Phosphotyrosine-containing Proteins in 32Dcl3 Cells-Bonita et al. (61) have reported that the expression of oncogenic forms of Cbl in NIH3T3 cells results in the dysregulation of cellular tyrosine kinases and that one of the kinases that was activated by oncogenic Cbl was the receptor for platelet-derived growth factor. They suggested that there may be other tyrosine kinases that are activated by oncogenic Cbl in these cells (61). Based upon these results, we were interested in determining whether there was an increase in the number of phosphotyrosine-containing proteins in 32Dcl3 cells expressing the Cbl⌬Y371 mutant or an alteration in IL-3-induced phosphorylation of cellular proteins. Both 32Dcl3 and 32D/Cbl⌬Y371 clone 9 cells were cultured overnight in the absence of IL-3 and then stimulated with 100 units/ml recombinant murine IL-3 for 0 -30 min. The cells were then lysed, immunoprecipitated with anti-phosphotyrosine monoclonal antibody, and immunoblotted with a monoclonal antibody to phosphotyrosine (Fig. 5). There was essentially no difference in the number of phosphotyrosinecontaining proteins present in 32Dcl3 cells from the number in 32D/Cbl⌬Y371 cells, although there was a slight decrease in the phosphorylation of a band with an approximate molecular weight of 120,000 in the latter cells (Fig. 5). It does appear that phosphorylation of the major proteins that appear following IL-3 stimulation may be slightly delayed in the 32D/Cbl⌬Y371 cells compared with the 32Dcl3 cells ( Activation of Signaling Molecules Downstream of the IL-3 Receptor Is Not Altered in Cells Expressing the Cbl⌬Y371 Mutant-Previous studies on oncogenic forms of Cbl have demonstrated that they are able to enhance the kinase activity of the receptors for epidermal growth factor (13) and platelet-derived growth factor (61). Activation of cytokine receptors has been shown to activate the Janus kinase (JAK)/signal transducer and activator of transcription (STAT) pathway (67,68). Therefore, it seems logical to determine whether the Cbl⌬Y731 mutant would enhance activation of the JAK2/STAT signaling pathway. The activation/phosphorylation of both JAK2 and STAT5 was examined by anti-phosphotyrosine immunoblotting. Activation of JAK2 requires the phosphorylation of tyrosine residues in the activation loop of JAK2 (69), and therefore the tyrosine phosphorylation of JAK2 can be directly correlated with catalytic activation of this kinase. There was no difference in the amount of phosphorylated JAK2 or the kinetics with which phosphorylated JAK2 appeared in 32D/Cbl⌬Y371 when compared with 32Dcl3 cells (Fig. 6A). Two different clones of 32D/Cbl⌬Y371 cells are examined in Fig. 6A. Compared with 32Dcl3 cells, there was no difference in the amount of JAK2 protein as shown by reprobing the immunoblot with anti-JAK2 antibody (data not shown). Although not shown here, overexpression of Bcl-2 in 32Dcl3 cells did not alter IL-3-induced activation of JAK2 or the amount of JAK2 protein in these cells (data not shown). Thus, the observed resistance of 32D/ Cbl⌬Y371 cells to apoptosis induced by growth factor withdrawal cannot be explained by an increase in basal JAK2 kinase activity. In addition to JAK2, cytokines such as IL-3 also induce activation of Src-related tyrosine kinases such as Fyn, Hck, and Lyn (70,71). Although we have not observed the activation of the Syk-related tyrosine kinases by IL-3, other investigators have observed its activation by cytokines such as granuclocyte colony-stimulating factor (72). We have examined that activa-tion of both Fyn and Syk in IL-3-stimulated 32Dcl3 and 32D/ Cbl⌬Y371 cells, and, consistent with the results shown with JAK2 above, we have not detected any alteration in the activation of either Fyn or Syk (data not shown). Although we have not conducted an exhaustive examination of all different tyrosine kinases present in 32Dcl3 cells, our data indicate that there is no alteration in at least three different tyrosine kinases, which represent three different classes of tyrosine kinases (JAK2, Fyn, and Syk).
Activation of JAK family tyrosine kinase is required for activation of STAT molecules (67,68). The activation of STAT5 was examined in three different clones of cells expressing the Cbl⌬Y371 mutant and two clones of cells overexpressing Bcl-2 (Fig. 6B). Consistent with the results shown in Fig. 6A, there was very little difference in the amount or the kinetics of STAT5 phosphorylation in any of the different cells examined. Although it appears that there may be significantly less phosphorylated STAT5 in 32D/Cbl⌬Y371 clone 15 cells at the 5-min time point, the anti-STAT5 immunoblot indicates that there is less STAT5 protein in this sample (Fig. 6B, bottom panel, lane  11). These data suggest that the activation of both JAK2 and STAT5 is unaltered in 32Dcl3 cells expressing the Cbl⌬Y371 mutant and in cells overexpressing Bcl-2, compared with the parental 32Dcl3 cell line. Thus, the diminished apoptosis observed in Figs. 3 and 4 above does not result from an alteration in the JAK/STAT signaling pathway as detected by these approaches.
Numerous studies have demonstrated that activation of extracellular signal-regulated kinases (ERKs) is required for proliferation and morphological transformation by activated oncogenes. We therefore wanted to determine whether expression of the Cbl⌬Y371 mutant altered IL-3-induced activation of ERKs   FIG. 4. Cell cycle analysis of 32Dcl3, 32D/Cbl⌬Y371 clone 9, and 32D/Bcl-2 clone 2 cells cultured in the absence of IL-3. 32Dcl3, 32D/Cbl⌬Y371 clone 9, and 32D/Bcl-2 clone 2 cells were cultured in the absence of IL-3 for 0 -24 h. Cells were withdrawn at varying times, fixed, stained with saponin/propidium iodide, and subjected to flow cytometric analysis. The cytometric data were then subjected to analysis with the ModFit program. The data shown are from the cells that were cultured in the absence of IL-3 for 24 h. A, 32Dcl3 cells; B, 32D/Cbl⌬Y371 clone 9 cells; C, 32D/Bcl-2 clone 2 cells. The G 1 /G 0 and G 2 ϩ M peaks are shown in hatched light gray; the S phase peak is unshaded; and the peak representing apoptotic cells is shown in black. The stippled area indicates cell debris.

TABLE I Cell cycle analysis of cells cultured in the absence of IL-3ϩ
The percentage of cells in G 1 /G 0 , S, and G 2 /M represents the number of cells in each phase as a percentage of the number of cells in the cycle and does not include the cells modeled as apoptotic. The percentage apoptotic represents the number of cells modeled as apoptotic as a percentage of the total number of cells modeled, minus debris.  in 32Dcl3 cells. ERK activation was examined by immunoblotting whole cell lysates of IL-3-stimulated cells with an antibody that was specific for the activated forms of ERK1 and ERK2. Stimulation of 32Dcl3 cells with IL-3 was observed to result in the rapid activation of ERK1 and ERK2 as shown in lanes 1-3 of Fig. 6C. The activation of ERK in these cells generally peaks between 5 and 15 min after stimulation and declines after this time (data not shown). The IL-3-induced activation of ERKs was examined in three different clones of 32D/Cbl⌬Y371 cells (Fig. 6C, lanes 4 -12). The activation of ERK in clones 13 and 15 appeared to be identical, with a peak of active ERK at 5 min and then a decline after that time. The amount of activated ERK in both clone 13 and clone 15 cells appeared to be equivalent to that present in 32Dcl3 cells (compare lanes 2, 8, and 11 in Fig. 6C). In contrast, the activation of ERK in 32D/Cbl⌬Y371 clone 9 cells appeared to be more transient, and the amount of activated ERK did not appear to be as great as that detected in the other cell lines (Fig. 6C, lanes 4 -6). IL-3-induced activation of ERK was also examined in one clone of 32D/Bcl-2 cells (Fig.  6C, lanes 13-15). ERK activation in these cells did not appear to differ significantly from that detected in the 32D/Cbl⌬Y371 clone 13 and 15 cells. These results suggest that the expression of the Cbl⌬Y371 protein in 32Dcl3 did not induce growth factorindependent activation of ERK and did not block IL-3-induced ERK activation, although the activation of ERK might be somewhat more transient in some clones of Cbl⌬Y371 cells.
Apoptosis is a genetically regulated cell death process that involves a series of biochemical changes to a cell (73)(74)(75). Several molecules have been identified that are able to regulate this process by suppressing cellular death; these include the antiapoptotic members of the Bcl-2 family (60), and the serine/ threonine kinase Akt (also known as protein kinase B, or PKB) (43, 46 -48, 76). Recent studies have indicated that cytokines such as IL-3 activate Akt and that activation of Akt is critical for suppression of apoptosis induced by growth factor withdrawal (43, 46 -48, 76). Thus, an increase in the amount of activated Akt could explain the decreased apoptosis observed with 32D/Cbl⌬Y371 cells. Activated Akt can be quantitated using an antibody that recognizes the phosphorylated activated form of Akt.
The amount of activated Akt in unstimulated and IL-3-stimulated 32Dcl3 and 32D/Cbl⌬Y371 clone 9 cells was examined by immunoblotting whole cell lysates with an anti-phospho-Akt antibody. As shown in Fig. 6D, activation of Akt was maximal 5 min after stimulation with IL-3 in both the 32Dcl3 and 32D/Cbl⌬Y371 clone 9 cells and declined after that time. The decline in phosphorylated Akt appears to be slightly faster in the 32D/Cbl⌬Y371 clone 9 cells compared with the 32Dcl3 cells, although this was not consistently observed in all studies with these cells. There was no difference in the amount of activated Akt present in either cell line over the time course examined in this study or in the total amount of Akt (Fig. 6D, bottom panel). It is clear from this study that there is not a pool of activated Akt present in unstimulated 32D/Cbl⌬Y371 clone 9 cells that could account for the lower amount of apoptosis observed in these cells following their cultivation in media lacking IL-3. This suggests that some molecule other than Akt is responsible.
Activation of Akt is dependent upon activation of PI 3-kinase and can be inhibited by treatment of cells with wortmannin (42,43,45,46,48). The ability of wortmannin to inhibit PI 3-kinase may explain in part why wortmannin treatment can induce apoptosis of several cell types. As a second means to determine whether Akt activation contributed to the resistance of the 32D/Cbl⌬Y371 cells to apoptosis induced by growth factor withdrawal, these cells were treated with 100 nM wortmannin upon removal from IL-3-containing media. When the parental 32Dcl3 cells were cultured in the absence of IL-3, they underwent apoptosis regardless of whether wortmannin was added to the media (Fig. 7). Consistent with the results shown in Fig.  3, 32D/Cbl⌬Y371 clone 9 cells remained viable when cultured in the absence of IL-3; however, the addition of 100 nM wortmannin to the media resulted in the 2-3-fold decrease in the number of viable cells as measured by trypan blue exclusion  (Fig. 7). This suggests that there may be a PI 3-kinase-dependent pathway that contributes to the suppression of apoptosis in these cells. Therefore, it is possible that Akt could contribute to this process; however, the results of the wortmannin study clearly indicate that there must also be an Akt-independent component to this process. The fact that there is a wortmanninsensitive component suggests that either there is a small pool of activated Akt in 32D/Cbl⌬Y371 cells that cannot be detected by immunoblotting with the anti-phospho-Akt antibody or that PI 3-kinase can activate antiapoptotic molecules other than Akt.
Expression of Cbl⌬Y371 Does Not Alter the Phosphorylation of Signaling Molecules following IL-3 Withdrawal-The data presented in Figs. 5 and 6 indicate that Cbl⌬Y371 does not increase the phosphorylation/activation of signaling molecules following IL-3 stimulation. It is possible, however, that Cbl⌬Y371 could increase or prolong the phosphorylation of these signaling molecules following IL-3 withdrawal. To examine this point, 32Dcl3 and 32D/Cbl⌬Y371 clone 9 cells that had been cultured in IL-3 were removed from IL-3-containing media, and cell lysates were prepared at times varying from 0 to 8 h. The phosphorylation of JAK2, STAT5, ERK1, ERK2, and Akt was examined at 0 min, 15 min, 30 min, 1 h, 2 h, 4 h, and 8 h following IL-3 withdrawal. To demonstrate that these proteins could still be phosphorylated/activated after an 8-h starvation, a sample of both cells was stimulated with 100 units/ml IL-3 and also analyzed.
There was no difference in the time at which phosphorylated JAK2, STAT5, ERK1, ERK2, and Akt disappeared from 32Dcl3 cells versus 32D/Cbl⌬Y371 clone 9 cells (Fig. 8, A, B, C, and D,  top panels). Withdrawal from IL-3-containing media had no effect upon the amounts of these proteins over the time period examined (Fig. 8, A, B, C, and D, bottom panels). Furthermore, all of these signaling molecules could be phosphorylated after 8 h of withdrawal from IL-3, by stimulation of these cells with IL-3 for 15 min (Fig. 8, A, B, C, and D, lanes 6 and 12). These data indicate that the ability of Cbl⌬Y371 to suppress apoptosis induced by growth factor withdrawal cannot be explained by the prolonged phosphorylation/activation of these signaling molecules.
Elevated Levels of Bcl-2 Are Present in 32D/Cbl⌬Y371 Clone 9 Cells-As noted above, some members of the Bcl-2 family of proteins are able to suppress apoptosis (Bcl-2, Bcl-x L , Mcl1, A1, and Bak), while others are able to induce apoptosis (Bad, Bclx S , and Bax) (60). Immunoblot analysis was used to determine whether the presence of the Cbl⌬Y7371 protein altered the level of different Bcl-2 family members. For each cell line to be analyzed, one set of cells was cultured in the presence of IL-3 for 16 h, while the second group of cells was cultured in the absence of IL-3 for the same period of time. Whole cell lysates FIG. 7. Effect of wortmannin upon cell viability. 32Dcl3 and 32D/Cbl⌬Y371 clone 9 cells were washed and then placed in culture media lacking IL-3 in the presence and absence of 100 nM wortmannin. At 24-h intervals, the number of viable cells was determined by counting the cells that excluded trypan blue. All cultures were initiated with 2 ϫ 10 5 cells/ml. q, 32Dcl3; E, 32Dcl3 in the presence of 100 nM wortmannin; f, 32D/Cbl⌬Y371 clone 9; Ⅺ, 32D/Cbl ⌬Y371 clone 9 in the presence of 100 nM wortmannin. A, 32Dcl3 (lanes 1-6) and 32D/Cbl⌬Y371 clone 13 cells (lanes 7-12) were cultured in media containing IL-3 and then pelleted and cultured in media lacking IL-3 for 0 -8 h (lanes 1-5 and 7-11). One sample of each cell type was cultured in the absence of IL-3 for 8 h and then stimulated with 100 units/ml IL-3 for 15 min prior to lysis (lanes 6 and 12). The lysates were immunoprecipitated with a polyclonal rabbit antibody to JAK2. The immunoprecipitated proteins were resolved on a 7.5% SDS-polyacrylamide gel and immunoblotted with antiphosphotyrosine antibody 4G10; then the blot was stripped and reprobed with anti-JAK2 antibody. B, 32Dcl3 (lanes 1-6) and 32D/Cbl⌬Y371 clone 13 (lanes 7-12) cells were treated as described for A. The cell lysates were immunoprecipitated with a polyclonal rabbit antibody to STAT5, and the immunoblot was probed with anti-phosphotyrosine antibody 4G10 (top panel) and then stripped and reprobed with anti-STAT5 antibody (bottom panel). C, 32Dcl3 (lanes 1-6) and 32D/Cbl⌬Y371 clone 13 (lanes 7-12) cells were treated as described for A. 25 g of the whole cell lysate was resolved on a 10% SDS-polyacrylamide gel, and the immunoblot was probed with an antibody that recognizes the activated forms of ERK1 and ERK2 (top panel). The immunoblot was stripped and reprobed with an anti-ERK antibody that recognizes both ERK1 and ERK2 (bottom panel) .  D, 32Dcl3 (lanes 1-6) and 32D/Cbl⌬Y371 clone 13 (lanes 7-12) cells were treated as described in A, and 50 g of the whole cell lysate was resolved on a 10% SDS-polyacrylamide gel. The filter was probed with anti-phospho-Akt antibody ( were prepared and subjected to immunoblot analysis. There was a dramatic increase in the amount of Bcl-2 in 32D/ Cbl⌬Y371 clone 9 cells compared with that present in 32Dcl3 cells (Fig. 9, lanes 3 and 4 versus lanes 1 and 2). Under the conditions used in this study and the time exposure used, Bcl-2 was barely detectable in 32Dcl3 cells grown in the presence or absence of IL-3 (Fig. 9A, lanes 1 and 2). In contrast, the amount of Bcl-2 present in 32D/Cbl⌬Y371 clone 9 cells was almost the same as that detected in 32D/Bcl-2 clone 2 cells (Fig. 9A, lanes  3 and 4 versus lanes 5 and 6). In other studies, we have observed that the Bcl-2 protein can only be detected when 32Dcl3 cells are grown in the presence of IL-3 and that IL-3 stimulation of 32Dcl3 cells that had been cultured in the absence of IL-3 resulted in the appearance of Bcl-2 protein (data not shown). This was not observed in Fig. 9A, perhaps because the time the film was exposed was too short to detect the presence of Bcl-2 in 32Dcl3 cells grown in the presence of IL-3. In this context, it is interesting to note that Bcl-2 was detected in both the 32D/Cbl⌬Y371 clone 9 and 32D/Bcl-2 clone 2 cells cultured in the absence of IL-3. The presence of Bcl-2 in the latter cell line is not surprising, since the Bcl-2 cDNA is overexpressed through the use the cytomegalovirus promoter present in the pcDNA3 expression vector. The high level of the Bcl-2 protein in the 32D/Cbl⌬Y371 clone 9 cells was not expected, however, and could potentially represent a gene whose expression is induced by the Cbl⌬Y371 mutant protein.

FIG. 8. Expression of the Cbl⌬Y371 protein does not alter the dephosphorylation/inactivation of signaling molecules following withdrawal of cells from IL-3.
There was no Mcl1 protein detected in any of the three cell lines grown in the absence of IL-3; however, the protein could be detected in all three cell lines cultured in the presence of IL-3. Slightly more Mcl1 protein was detected in both the 32D/Cbl⌬Y371 clone 9 and 32D/Bcl-2 clone 2 cells compared with that detected in 32Dcl3 cells (Fig. 9C). The significance of this difference is not clear at this time. There did not appear to be any significant difference in the amount of either Bcl-x L or Bax between the three cell lines (Fig. 9, B and D). The levels of these latter two proteins were not altered when any of the three cell lines were cultured in the absence of IL-3. These data suggest that the apparent resistance of the 32D/Cbl⌬Y371 cell lines to apoptosis induced by growth factor withdrawal could be explained by altered levels of the antiapoptotic protein Bcl-2. To demonstrate that the increased level of Bcl-2 protein was not unique to just the 32D/Cbl⌬Y371 clone 9 cells, we also examined the level of Bcl-2 protein in three other independent clones of 32D/Cbl⌬Y371 cells (Fig. 9E). An elevated level of Bcl-2 protein was observed in 32D/Cbl⌬Y371 clones 3, 4, and 5 when they were compared with the parental 32Dcl3 cells (Fig.  9E). Thus, this observation is not unique to a single clone of cells.
Transfection of 32Dcl3 and 32D/Cbl⌬Y371 with a bcl-2 promoter reporter vector did not reveal any significant difference in the expression of the bcl-2 promoter between these two cell lines (data not shown). Although it is possible that the promoter did not include the required responsive region of the bcl-2 promoter, we believe that this suggests that transcriptional induction of the bcl-2 gene does not explain the data presented in Fig. 9. Other possibilities include that there is an increase in the half-life of the bcl-2 mRNA or the half-life of the Bcl-2 protein.
Caspase Activation Is Suppressed in 32D/Cbl⌬Y371 Clone 9 Cells Cultured in the Absence of IL-3-As noted above, apoptosis is a genetically controlled process that involves many different molecules. A central event in the apoptotic process is the activation of a specific set of proteases referred to as caspases (77). At least nine different caspases are currently known. Some caspase family members are activated very early during apoptosis, while others are activated late in the death process. We have examined the activation of caspase 3 as a function of the time in which the different cells were cultured in the absence of IL-3. The 32Dcl3, 32D/Cbl⌬Y371 clone 9, and 32D/ Bcl-2 clone 2 cells were cultured in the presence or absence of 50 units/ml rmIL-3 for up to 48 h. Four hours after placing all three cell types in growth factor-free media, the levels of caspase 3 was essentially the same in all cultures. After this time point, however, there was a dramatic increase in the amount of caspase activity in the 32Dcl3 cells but not in 32D/ Cbl⌬Y371 clone 9 or 32D/Bcl-2 clone 2 cells (Fig. 10). Although there was a slow but detectable increase in the amount of caspase detected by this assay in 32D/Cbl⌬Y371 clone 9 and 32D/Bcl-2 clone 2 cells cultured in the absence of IL-3 for 48 h, there was still at least 4-fold more caspase activity in the 32Dcl3 cells cultured in the absence of IL-3 for 48 h (Fig. 10). This suggests that there is a dramatic difference in the response of these different cell types to growth factor withdrawal and that the responses of 32D/Cbl⌬Y371 clone 9 and 32D/Bcl-2 clone 2 cells are quite similar. DISCUSSION In this study, we have examined the effects of an oncogenic form of the c-Cbl proto-oncogene upon the proliferation of the growth factor-dependent murine myeloid cell line, 32Dcl3. The oncogenic form of Cbl we have used was constructed by Andoniou et al. (51) and contains a deletion of a single tyrosine residue, Tyr 371 . This activated form of Cbl has been previously reported to induce the transformation of NIH3T3 cells, and these transformed cells were tumorigenic when injected into nude mice (51). The deletion of this single tyrosine residue must result in a dramatic change in the conformation of this protein such that it becomes oncogenic. Mutation of the tyro- sine to a phenylalanine residue, which cannot be phosphorylated, did not result in the oncogenic activation of this protein.
The structure of a c-Cbl-UbcH7 complex has recently been published and indicates that both Tyr 368 and Tyr 371 lie in a linker region between the phosphotyrosine-binding domain and the RING finger motif and that these residues are in a buried environment and make multiple van der Waals contacts with hydrophobic residues in the phosphotyrosine-binding domain (78). Therefore, it is likely that deletion of either Tyr 368 or Tyr 371 would result in a conformational change.
Prior to beginning these studies, it had been demonstrated that expression of an oncogenic form of Cbl in NIH3T3 cells resulted in the disregulation of one or more tyrosine kinases present in those cells, and one of the deregulated kinases appeared to be the platelet-derived growth factor receptor (61). We therefore expected that expression of the Cbl⌬Y371 protein in 32Dcl3 cells would result in the activation of one or more tyrosine kinases in these cells. Based upon our studies that indicated that two different activated tyrosine kinases, v-Src and BCR-ABL (53), could induce growth factor-independent proliferation of factor-dependent cells, we predicted that the Cbl⌬Y371 mutant would also induce growth factor-independent proliferation. Contrary to our predictions, we did not observe an increase in the basal level of phosphotyrosinecontaining proteins in 32D/Cbl⌬Y371 cells; nor was the phosphorylation/activation of signaling molecules potentiated in 32D/Cbl⌬Y371 cells following withdrawal from IL-3-containing media. We also did not identify any constitutively activated tyrosine kinases, and the cells remained growth factor-dependent for cellular proliferation. Although we have not examined all tyrosine kinases present in 32Dcl3 cells, we have examined three different kinases that represent three different classes of tyrosine kinases: JAK2, Fyn, and Syk.
To our surprise, we observed that expression of the Cbl⌬Y371 protein in 32Dcl3 cells resulted in a suppression of apoptosis. Suppression of apoptosis was characterized by the continued viability of transfected cells when cultured in the absence of IL-3, arrest of the cells in the G 1 phase of the cell cycle when cultured in the absence of IL-3, and decreased caspase activation when the cells were cultured in the absence of IL-3. The mechanism by which apoptosis was suppressed appears to result from an increase in the level of the antiapoptotic Bcl-2 protein. We have not detected an increase in the level of any other antiapoptotic Bcl-2 family member; nor have we detected a decrease in the amount of a proapoptotic Bcl-2 family member. However, we have not exhaustively examined all of the members of this family of proteins, since we have not been able to obtain good antibodies to each of the different Bcl-2 family members. We have also not detected the prolonged phosphorylation/activation of signaling molecules in 32D/Cbl⌬Y371 following withdrawal from IL-3-containing media.
The bcl-2 oncogene was discovered because it mapped to a chromosomal breakpoint present in B-cell leukemia (79). In this case, the translocation of a promoter for the IgG upstream of the coding region of the bcl-2 gene resulted in the disregulated expression of the bcl-2 gene. The juxtaposition of the IgG promoter upstream of the bcl-2 gene explains why there is a specific increase in the amount of Bcl-2 in B-lymphocytes. The tissue specificity of this promoter also explains why neoplasia is not observed in other tissues. Based upon the observation that disregulated expression of bcl-2 results in the B-cell leukemia described above, it would be expected that the ability of oncogenic forms of Cbl to increase Bcl-2 protein levels could partially explain the leukemogenic potential of oncogenic forms of c-Cbl.
At this time, we do not understand the basis for the increased level of Bcl-2 protein in 32Dcl3 cells expressing the Cbl⌬Y371 mutant protein. Transfection of a bcl-2 promoter reporter vector into 32D/Cbl⌬Y371cells did not reveal an increase in promoter expression in these cells compared with normal 32Dcl3 cells. This suggests that other mechanisms may explain the increased level of Bcl-2 protein. Possibilities include that there is an increase in the half-life of the Bcl-2 mRNA, or an increase in the half-life of the Bcl-2 protein. To our knowledge, this is the first example of an oncogene inducing a change in the level of an antiapoptotic Bcl-2 family member.
The results obtained in our study are very similar to those described by Corey and colleagues. 2 These investigators have expressed an oncogenic truncation mutant of c-Cbl, which is similar to v-Cbl, in 32Dcl3 cells, and they have determined that it can suppress apoptosis in a manner similar to that described in this paper. 2 In addition, they have shown that their oncogenic form of Cbl blocks the terminal different of 32Dcl3 cells induced by granulocyte colony-stimulating factor. 2 If it can be demonstrated that oncogenic forms of Cbl are all able to suppress apoptosis by modifying the levels of Bcl-2 family members, then another mechanism by which oncogenes can transform cells will be established. It will also be of importance to FIG. 10. Activation of caspase 3 is suppressed in 32D/Cbl⌬Y371 clone 9 cells following cultivation in media lacking IL-3. 32Dcl3, 32D/Cbl⌬Y371 clone 9, and 32D/Bcl-2 clone 2 cells were cultured in the absence of rmIL-3 for 0 -48 h. At 0, 4,8,16,24,36, and 48 h, cells were withdrawn and lysed, and the amount of activated caspase 3 was determined using a Caspase 3 Cellular Activity Assay Kit from Biomol. The specific activity of caspase 3 was determined at each time point from the initial slope of the reaction. The specific activity in pmol of p-nitroanaline/min/g of total cellular protein is plotted versus the time the different cells were cultured in the absence of IL-3. q, 32Dcl3 cells; f, 32D/Cbl⌬Y371 clone 9 cells; OE, 32D/Bcl-2 clone 2 cells. determine whether mutation of c-Cbl plays a role in the genesis of human cancer or leukemia.