Abstract
Stimulation of Nb2 cells with PRL results in the rapid phosphorylation of a 120-kDa protein identified as the adapter protein cbl on tyrosine residues. Maximal phosphorylation of cbl occurs at 20 min after PRL stimulation and declines thereafter. Stimulation with as little as 5 nm PRL resulted in the phosphorylation of cbl; increasing the concentration of PRL to 100 nm had only a minimal effect upon the phosphorylation of cbl. The cbl protein appears to be constitutively associated with grb2 and the p85 subunit of phosphatidylinositol 3-kinase (PI 3-kinase). The constitutive association of cbl with the p85 subunit of PI 3-kinase was observed in Nb2 cells as well as in 32Dcl3 cells transfected with either the rat Nb2 (intermediate) form of the PRL receptor or the long form of the human PRL receptor. A glutathione S-transferase fusion protein encoding the SH3 domain of the p85 subunit of PI 3-kinase bound to cbl in lysates of both unstimulated and PRL-stimulated Nb2 cells; however, neither of the SH2 domains of p85 bound to cbl under the same conditions. PRL stimulation increased the cbl-associated PI kinase activity. The majority of PI kinase activity appeared to be cbl-associated after PRL stimulation. These results suggest that cbl may function as an adapter protein in PRL-mediated signaling events and regulate activation of PI 3-kinase. Our model suggests that the p85 subunit of PI 3-kinase is constitutively associated with cbl through binding of the p85 SH3 domain to a proline-rich sequence in cbl. After the tyrosine phosphorylation of cbl, an SH2 domain(s) of p85 binds to a specific phosphorylation site(s) in cbl, leading to the activation of PI 3-kinase.
INTRODUCTION
Cbl is the cellular homolog of the oncogene present in the Cas-NS-1 retrovirus, which induces B cell lymphomas (1, 2). Sequence analysis of the cbl cDNA revealed that the protein contains 913 amino acids, a putative nuclear localization sequence in its N-terminal region, a RING finger motif typical of DNA-binding proteins, and several proline-rich sequences in its C-terminal half that may serve as SH3-binding sites (1, 3). There is no evidence, however, that cbl is present in the nucleus or that it binds to DNA. Amino acid sequence analysis of cbl does not predict the presence of any catalytic domain of a known signaling molecule or enzyme. The fact that cbl becomes tyrosine-phosphorylated after activation of a variety of receptors, however, does suggest that it may serve as an adapter molecule in a manner analogous to that of insulin-regulated substrate-1 in signal transduction mediated by the insulin receptor.
Numerous recent studies have demonstrated that cbl becomes tyrosine-phosphorylated after stimulation of a wide variety of receptors including: the T cell receptor (4–6), the B cell antigen receptor (7, 8), the Fc receptor (9, 10), the epidermal growth factor receptor (11–14), the colony-stimulating factor-1 receptor (15), the erythropoietin receptor (16), the receptor for granulocyte-macrophage colony-stimulating factor (16), and the receptor for interleukin-3 (17). The cbl protein is also phosphorylated in cells expressing activated oncogenes such as v-abl or BCR-ABL (18–20). In these studies, cbl was observed to be associated with a variety of proteins by either coimmunoprecipitation studies or binding to bacterial fusion proteins. Association of the p85 subunit of phosphatidylinositol 3-kinase (PI 3-kinase) with tyrosine-phosphorylated cbl has been described to occur in a phosphotyrosine-dependent manner via the SH2 domains of p85, and in a phosphotyrosine-independent manner involving the SH3 domain of p85 (4, 6, 11, 19). It is likely, however, that both domains are important in the interaction of cbl with p85 leading to activation of PI 3-kinase. cbl has also been observed to interact with the SH3 domain of lyn (9, 10), with the SH2 and SH3 domains of fyn (6, 9), with the SH2 domain of crk (19), and with the SH3 domain of grb2 in a constitutive manner (5, 14, 16). Interactions involving the SH2 domains are largely thought to occur in a phosphotyrosine-dependent manner, although one study has suggested that this may not be the case (17). The association of cbl with the p85 subunit of PI 3-kinase suggests that phosphorylation of cbl may regulate activation of PI 3-kinase. The constitutive association of the p85 subunit of PI 3-kinase with cbl, mediated by the binding of the p85 SH3 domain to a proline-rich sequence in cbl, would allow the rapid activation of PI 3-kinase after the binding of the SH2 domain(s) of p85 bind to a specific phosphorylation site(s) in cbl. Supporting this hypothesis is the observation that activated PI 3-kinase activity is associated with cbl in stimulated T- and B lymphocytes (4, 6, 7) as well as in interleukin-3-stimulated myeloid cells (17).
The PRL receptor is a member of the cytokine receptor superfamily (21–25). Stimulation of the receptors for erythropoietin, granulocyte-macrophage colony-stimulating factor, and interleukin-3, three other cytokine receptor family members, results in tyrosine phosphorylation of cbl, suggesting that PRL stimulation might also result in the phosphorylation of cbl. In this manuscript we demonstrate that cbl is phosphorylated after PRL stimulation of Nb2 cells as well as in 32Dcl3 cells transfected with the long form of the human PRLR or the Nb2 (intermediate) form of the rat PRLR. Cbl is constitutively associated with the SH2-containing adapter protein grb2 and the p85 subunit of PI 3-kinase. Quantitative studies suggest that the majority of PI kinase activity is associated with cbl after PRL stimulation of Nb2 cells. This suggests that cbl may function as an adapter or docking molecule in a manner similar to that observed with insulin-regulated substrate-1 in insulin receptor signaling. After the ligand-induced tyrosine phosphorylation of cbl, multiple signaling molecules could bind to specific phosphorylation sites in cbl, presumably through the binding of their SH2 domains to specific phosphorylation sites, leading to the activation of these signaling molecules.
RESULTS
Cbl Is Rapidly Phosphorylated on Tyrosine Residues after PRL Stimulation
To determine whether cbl was phosphorylated on tyrosine residues after PRL stimulation, Nb2 cells were stimulated with 10 nm PRL for 10 min and immunoprecipitated with either antiphosphotyrosine or anti-cbl antibodies. The results indicate that cbl becomes tyrosine phosphorylated after stimulation of Nb2 cells with PRL (Fig. 1). The major tyrosine-phosphorylated protein present in antiphosphotyrosine immunoprecipitates of PRL-stimulated Nb2 cells has a molecular mass of 130,000 and corresponds to JAK2 (26–28). By densitometry there is a 6-fold increase in the phosphorylation of JAK2 after PRL stimulation (Fig. 1, first and second lanes). Immunoblotting with anti-cbl antibody indicates that PRL stimulation did not change the amount of cbl present in the cell (Fig. 1, seventh and eighth lanes). The anti-cbl immunoblot also indicates that only a small amount of the cbl protein appears in the antiphosphotyrosine immunoprecipitate of PRL-stimulated Nb2 cells (Fig. 1, sixth and eighth lanes). Densitometric measurements indicate that only about 5% of the cbl in the anti-cbl immunoprecipitate could be detected in the anti-phosphotyrosine immunoprecipitate of PRL-stimulated Nb2 cells (Fig. 1, lane 6 vs. lane 8). Only 25% of the anti-cbl immunoprecipitate was loaded in lanes 7 and 8 of Fig. 1 to allow for a more accurate quantification of the amounts of cbl in the anti-cbl and anti-phosphotyrosine immunoprecipitates; however, the figure of 5% takes into account this difference in loading. This result was consistent in four different studies with Nb2 cells. From the study shown in Fig. 1, it is clear that cbl does not comigrate with any of the major tyrosine-phosphorylated proteins present in the antiphosphotyrosine immunoprecipitate (Fig. 1, lanes 1–4). Therefore, the phosphorylation of cbl in PRL-stimulated Nb2 cells would not have been anticipated by an examination of the sizes of tyrosine-phosphorylated proteins in anti-phosphotyrosine immunoblots prepared from these cells. Furthermore, there are no detectable tyrosine-phosphorylated proteins that coprecipitate with cbl (Fig. 1, lane 4). Three major tyrosine-phosphorylated proteins are detected in the 4G10 immunoprecipitate with mol wts of 130,000, 66,000, and 42,000 (Fig. 1). Based upon immunoblot studies with anti-JAK2 antiserum, we believe that 130 kDa protein to be JAK2 (data not shown). The 66 and 42 kDa proteins may correspond to shc and mitogen-activated protein kinase, respectively.
Cbl Is Phosphorylated on Tyrosine Residues after Stimulation with PRL Nb2 cells were cultured overnight in RPMI 1640 with 5% charcoal-stripped serum, then stimulated with 10 nm rat PRL for 0 (lanes marked −) or 10 min (lanes marked +). Cells were lysed with modified RIPA and immunoprecipitated with either anti-phosphotyrosine monoclonal antibody 4G10 (lanes 1, 2, 5, and 6) or anti-cbl antibody (lanes 3, 4, 7, and 8). Each immunoprecipitate contained lysate prepared from 2 × 107 cells. After display of the immunoprecipitated proteins on a 7% SDS polyacrylamide gel, the proteins were transferred to Immobilon and immunoblotted with either antiphosphotyrosine antibody 4G10 (lanes 1–4) or anti-cbl antibody (lanes 5–8). The position of prestained molecular mass markers are indicated on the left side of the figure, and the position of cbl is indicated in the center of the figure.
The kinetics of cbl phosphorylation were also examined by stimulating Nb2 cells with 10 nm PRL for 0–120 min. Maximal phosphorylation was detected approximately 20 min after PRL stimulation and decreased after this time (Fig. 2A, top panel). The phosphorylation of cbl was significantly decreased by 1 h, and no increased phosphorylation could be detected by 2 h (Fig. 2). No coimmunoprecipitating tyrosine-phosphorylated proteins were noted in the anti-cbl immunoprecipitates at any of the times examined (data not shown). The time course of cbl phosphorylation described in Fig. 2 was consistently observed in four different studies with Nb2 cells. Immunoblotting with anti-cbl antiserum indicated that the amount of cbl protein did not change over the time period examined (Fig. 2A, bottom panel).
Kinetics and Dose-Response of cbl Phosphorylation A, Nb2 cells were cultured overnight as described in Fig. 1, then stimulated for 0–120 min with 10 nm rat PRL. Cells were lysed in EB and immunoprecipitated with anti-cbl antibody. The time of PRL stimulation, in minutes, is indicated at the top of each lane. Immunoblotting was with antiphosphotyrosine antibody 4G10 (top panel). The immunoblot was then stripped as described (17 ) and reprobed with anti-cbl antibody to indicate the amount of cbl protein in each immunoprecipitate. B, Nb2 cells were cultured overnight in RPMI 1640 with 5% charcoal-stripped serum, then stimulated with 0–200 ng/ml rat PRL for 10 min. Cells were lysed in EB and immunoprecipitated with anti-cbl antibody, after which extent of phosphorylation was examined by antiphosphotyrosine immunoblotting. The immunoblot was reprobed with anti-cbl antibody to indicate the amount of cbl in each immunoprecipitate. The concentration of rat PRL used to stimulate the Nb2 cells, in nanograms/ml is indicated at the top of each lane.
The effect of increasing concentrations of PRL upon the phosphorylation of cbl was also examined. Nb2 cells were stimulated with 0–200 ng/ml PRL for 15 min before lysis and analysis of cbl phosphorylation by antiphosphotyrosine immunoblotting (Fig. 2B). Although a low level of cbl phosphorylation could be observed in unstimulated cells (Fig. 2B, lane 1), PRL stimulated the tyrosine-phosphorylation of cbl (Fig. 2B, lanes 5–6). The amount of tyrosine phosphorylation remained fairly constant over a range of 10–100 ng/ml, and in some studies there appears to be a slight decrease in the extent of cbl phosphorylation in cells stimulated with 200 ng/ml PRL. There was no significant change in the level of cbl protein after PRL stimulation over this dose range (Fig. 2B, bottom panel). This result was consistently observed in four independent studies of Nb2 cells.
Association of cbl with the SH2/SH3-Containing Adapter Protein grb2
As noted in the introduction above, cbl has been noted to associate with numerous proteins, including src-like tyrosine kinases, the SH2-containing adapter protein grb2, and the p85 subunit of PI 3-kinase (4–6, 9–11, 16, 19). The ability of cbl to coimmunoprecipitate with some of these proteins was examined in lysates of both unstimulated and PRL-stimulated Nb2 cells.
The adapter protein grb2 couples the GTP exchange protein SOS to activated growth factor receptors, leading to the activation of SOS and subsequent activation of ras (29–33). We observed the constitutive association of grb2 with cbl (Fig. 3). Unstimulated and stimulated Nb2 cells were immunoprecipitated with either anti-grb2 or anti-cbl antibodies, and the immunoprecipitates were subjected to immunoblotting with either anti-phosphotyrosine, anti-cbl, or anti-grb2 antibodies. Antiphosphotyrosine immunoblotting confirmed the PRL-stimulated phosphorylation of cbl (Fig. 3, lanes 3 and 4), although there was a detectable amount of phosphorylated cbl in the unstimulated Nb2 cells. No cbl protein was detected in anti-grb2 immunoprecipitates immunoblotted with anti-cbl antibody (Fig. 3, lanes 5 and 6); however, grb2 protein was present in anti-cbl immunoprecipitates from both unstimulated and stimulated cells (Fig. 3, lanes 11 and 12). Note that the gels used for analysis of grb2 were 12% gels, whereas those probed with anti-phosphotyrosine or anti-cbl were 7% gels. There was no difference in the amount of grb2 protein present in anti-cbl immunoprecipitates from unstimulated vs. PRL-stimulated Nb2 cells. By densitometry, the amount of grb2 protein detected in the anti-cbl immunoprecipitates represented only 12% (range of 9–14% was observed in three different studies of Nb2 cells) of the amount observed in anti-grb2 immunoprecipitates (Fig. 3, compare lanes 9–12). When lanes 9–12 of Fig. 3 were immunoblotted with the anti-cbl antibody, no cross-reaction of the antiserum with grb2 was detected. This suggests that the anti-cbl antibody does not nonspecifically react with grb2. These results suggest that a small portion of cbl is constitutively associated with grb2, and the amount of this complex does not change after PRL stimulation. We do not understand why we were unable to detect cbl in the anti-grb2 immunoprecipitates by immunoblotting; however, this may reflect the small amount of cbl associated with grb2. There was no evidence that grb2 became tyrosine-phosphorylated after PRL stimulation (data not shown). Anti-grb2 immunoprecipitates also contained a phosphotyrosine-containing protein larger than cbl, with a mol wt of approximately 130,000–140,000 (Fig. 3, lanes 1 and 2). This protein remains to be identified.
Constitutive Association of cbl with grb2 Nb2 cells were cultured as described in Fig. 1, then stimulated for 0 (lanes marked −) or 10 (lanes marked +) minutes with 10 nm PRL. The cells were lysed with EB and immunoprecipitated with either anti-grb2 (lanes 1, 2, 5, 6, 9, and 10) or anti-cbl antibody (lanes 3, 4, 7, 8, 11, and 12). The immunoprecipitates were resolved on 7% (left and center panels) or 12% (right panel) SDS polyacrylamide gels and transferred to Immobilon, and the immunoblots were probed with antiphosphotyrosine antibody 4G10 (lanes 1–4), anti-cbl antibody (lanes 5–8), or anti-grb2 antibody (lanes 9–12). The positions of prestained molecular mass markers are indicated. The position of cbl is indicated on the left, and the position of grb2 is indicated on the right.
Constitutive Association of the p85 Subunit of PI 3-Kinase with cbl
To further examine the role of cbl as an adapter protein, the association of cbl with PI 3-kinase was also examined. Unstimulated and stimulated cells were immunoprecipitated with antiphosphotyrosine, anti-cbl, or an anti-p85 subunit antibody that immunoprecipitates both the α- and β-isoforms of p85. Immunoblotting with an anti-p85 subunit antibody detected p85 in both the anti-cbl and anti-p85 immunoprecipitates of both unstimulated and PRL-stimulated Nb2 cells (Fig. 4A, lanes 4 and 6). The p85 protein was also detected in the anti-phosphotyrosine immunoprecipitate of unstimulated cells; however, there was a 4-fold increase in the amount of p85 in the antiphosphotyrosine immunoprecipitate after PRL stimulation of Nb2 cells (Fig. 4, lanes 1 and 2). Densitometric analysis of the p85 bands in Fig. 4 indicate that the anti-cbl immunoprecipitates contain 10–20% of the protein present in the anti-p85 immunoprecipitate obtained with the monoclonal anti-p85 antibody used in this study. The anti-cbl immunoprecipitates were also observed to contain the 110-kDa catalytic subunit of PI 3-kinase as determined by immunoblotting with a monoclonal antibody directed against this protein (data not shown). The antiphosphotyrosine immunoprecipitate of PRL-stimulated cells contained 20–30% of the p85 subunit present in the anti-p85 immunoprecipitate. These numbers reflect the range of values observed in four different studies with Nb2 cells. Although p85 was present in the antiphosphotyrosine immunoprecipitate, we could not detect the tyrosine phosphorylation of p85 (Fig. 1 and data not shown). Reprobing the immunoblot with anti-cbl antibody revealed the present of cbl protein in the anti-cbl immunoprecipitates but not the anti-p85 immunoprecipitates (Fig. 4A, lanes 3 and 4, bottom panel). Longer exposures of the immunoblot, which revealed the presence of cbl in the antiphosphotyrosine immunoprecipitate, failed to demonstrate the presence of cbl in the anti-p85 immunoprecipitate (data not shown). Stimulation of Nb2 cells with up to 200 ng/ml PRL did not change the amount of p85 associated with cbl (data not shown).
Association of cbl with the p85 Subunit of PI 3-Kinase Nb2 or 32D/hPRLR cells were cultured as described in Fig. 1, then stimulated for 0 or 10 min with 10 nm PRL. The cells were lysed with EB, immunoprecipitated with antiphosphotyrosine antibody 4G10 (lanes 1 and 2), anti-cbl (lanes 3 and 4), or anti-p85 subunit of PI 3-kinase (Transduction Laboratory No. P13030), and the immunoprecipitates were resolved on a 7% SDS polyacrylamide gel. The top panel was probed with anti-p85 antiserum (Transduction Laboratories Catalog No. P13030), and the bottom panel was probed with anti-cbl antiserum.
The results obtained with Nb2 were confirmed by conducting the same study on 32Dcl3 cells transfected with either the long form of the human PRL receptor (hPRLR) or the Nb2 (intermediate) form of the rat PRLR. The data in Fig. 4, lanes 7–12, show the results obtained with the 32D/hPRLR cells. Consistent with the results obtained with Nb2 cells, the constitutive association of p85 with cbl was obtained (Fig. 4, lanes 9 and 10). Compared with the results observed with Nb2 cells, a higher percentage of p85 appeared to be cbl-associated; approximately 50–70% of the amount of p85 present in the anti-p85 immunoprecipitate was present in the anti-cbl immunoprecipitate. Also consistent with the Nb2 cell study, p85 was detected in the antiphosphotyrosine immunoprecipitate of both unstimulated and PRL-stimulated cells, and there was a 1.5- to 2-fold increase in the amount of p85 present in the antiphosphotyrosine immunoprecipitates of lysates of PRL-stimulated 32D/hPRLR cells compared to the amount in lysates of unstimulated cells. Consistent results were observed in three independent studies.
Under the conditions used, the anti-cbl antibody precipitated all of the cbl protein present in the cell extract; reimmunoprecipitation of the cell lysate with a second aliquot of anti-cbl antibody did not precipitate any additional protein (data not shown). Similar studies conducted with the anti-p85 antibody suggested that greater than 90% of the p85 was removed from the cell lysate during the first immunoprecipitation (data not shown). When a monoclonal antibody specific for the β-isoform of p85 was used, no association of p85 with cbl was noted (data not shown). This is consistent with results reported elsewhere with the T cell receptor (4). When cells were lysed with RIPA [150 mm NaCl, 50 mm Tris (pH 7.4), 2 mm EGTA, 1% Triton X-100, 0.25% sodium deoxycholate, 1 mm sodium orthovanadate] instead of EB [50 mm NaCl, 10 mm Tris (pH 7.4), 5 mm EDTA, 50 mm NaF, 1% Triton X-100, 1 mm sodium orthovanadate], no association of p85 with cbl was detected, indicating that this complex of proteins is very sensitive to the presence of detergents (data not shown). These results are consistent with those previously described for cbl after stimulation of myeloid cells with interleukin-3 (17).
The p85 subunit of PI 3-kinase contains an SH3 domain and two SH2 domains that regulate the interaction of p85 with activated growth factor receptors and other signaling molecules (34, 35). SH2 domains mediate binding to phosphorylated tyrosine residues in a sequence-specific context (36, 37), while SH3 domains bind to proline-rich sequences (38). The constitutive association of p85 with cbl suggested that this interaction was mediated by SH3 domains. This was examined with a binding assay in which the amount of cbl that bound to GST fusion proteins encoding the SH3 domain and each of the two SH2 domains (GST-N-SH2 and GST-C-SH2, the N-terminal and C-terminal SH2 domains of p85, respectively), was examined. As can be seen in Fig. 5, only the SH3 domain of p85 was able to bind to cbl in lysates of either unstimulated or PRL-stimulated Nb2 cells (Fig. 5, lanes 7 and 8). No cbl was observed to bind to GST alone, or to either the N- or C-terminal SH2 domains. Stimulation of Nb2 cells with PRL did not alter the amount of cbl that bound to the SH3 domain of p85. These results are consistent with those described in Fig. 4 and suggest a constitutive association of p85 with cbl.
The SH3 Domain of p85 Binds to cbl Unstimulated Nb2 cells, lysed in EB, were prepared as described in Fig. 1. GST, GST-N-SH2, GST-C-SH2, or GST-SH3 was added to the cell lysates at a final concentration of 2 μm. After a 1 h incubation at 4 C on a rocking platform, 40 μl of a 50% suspension of glutathione agarose beads were added for 20 min. The bound protein complexes were washed three times with EB and solubilized in 30 μl of 2× SDS gel sample buffer, and the protein was resolved by electrophoresis on a 7% SDS gel. Immunoblotting was with anti-cbl antibody. The fusion protein used is indicated at the top of each lane, and the position of cbl is indicated on the right side of the panel.
To demonstrate that cbl immunoprecipitates contained PI kinase activity in addition to the p85 subunit of PI 3-kinase, the PI kinase assay was used to demonstrate the presence of enzymatic activity in the various immunoprecipitates. Nb2 cells were stimulated with PRL for 0–20 min and immunoprecipitated with either anti-cbl, anti-p85 subunit, or a nonimmune rabbit serum. A very small amount of PI kinase activity was detected in the nonimmune serum control; however, the amount of activity did not increase after PRL stimulation (Fig. 6, lanes 9 and 10). The anti-p85 subunit antibody detected a small amount of PI kinase activity in unstimulated cells, and PRL stimulation induced a 4-fold increase in PI kinase activity at 5 min (Fig. 6, lanes 5–8). Anti-cbl immunoprecipitates contained PI kinase activity that increased after PRL stimulation (Fig. 6, lanes 1–4). Increased PI kinase activity was detectable as early as 2 min after PRL stimulation and was still evident 20 min later. Densitometric analysis indicated that there was a 10-fold increase in PI kinase activity at 2 min compared to unstimulated cells. The difference in the increase of PI kinase activity at 2 min between the anti-cbl and anti-p85 immunoprecipitates may reflect the increased basal activity present in the anti-p85 immunoprecipitate compared with the anti-cbl immunoprecipitate. In addition to the expected product of phosphatidylinositol phosphate (PIP), a second spot that migrated halfway between the origin and PIP was observed (indicated by the arrow on the right side of Fig. 6). We suspect that this spot may correspond to an oxidation product of PIP as the amount of it increases directly in proportion to the amount of PIP. Another spot of unknown identity was observed only in the reactions of anti-p85 immunoprecipitates. A comparison of the amount of PI kinase activity in anti-cbl immunoprecipitates at 5 and 20 min with the amount in anti-p85 immunoprecipitates at the same time points indicates that most, if not all, of the PI kinase activity is cbl-associated after PRL stimulation. PI kinase activity was observed in anti-phosphotyrosine immunoprecipitates; however, the amount of activity did not increase after PRL stimulation (data not shown).
PRL-Induced Association of cbl with PI 3-Kinase Activity Cells were stimulated for 0–20 min with 10 nm rat PRL, lysed with EB, and immunoprecipitated with either anti-cbl (lanes marked cbl), anti-p85 antibody (Transduction Laboratory No. P13030, lanes marked p85), or nonimmune rabbit serum (lanes marked NIS). PI Kinase assay was conducted as described in Materials and Methods. The origin and the position of the reaction product, PIP, are indicated on the right side of the figure. Lane numbers are indicated across the top of the figure.
DISCUSSION
In this manuscript we have demonstrated the phosphorylation of cbl after stimulation of Nb2 cells with PRL. The phosphorylation of this protein was not anticipated because we have not previously detected a protein of this size in our antiphosphotyrosine immunoblots of either total cell lysates of PRL-stimulated cells or antiphosphotyrosine immunoprecipitates of PRL-stimulated cells. Tyrosine-phosphorylated JAK2 migrates with a slower electrophoretic mobility and was readily separable from cbl. In the absence of PRL stimulation, cbl was associated with the SH2-containing adapter protein grb2, and the amount of grb2 complexed with cbl did not change after PRL stimulation. The constitutive association of grb2 with cbl is consistent with our previous results obtained in studies on the phosphorylation of cbl after stimulation of myeloid cells with interleukin-3 (17).
These studies have also described the constitutive association of the p85 subunit of PI 3-kinase with cbl. This association was detected in Nb2 cells and 32Dcl3 cells transfected with either the Nb2 form of the rat PRLR or the long form of the human PRLR. The percent of p85 associated with cbl varied from 10–70% of the amount of p85 present in an anti-p85 immunoprecipitate, depending upon the cell line examined. Binding studies with GST fusion proteins indicate that the SH3 domain of p85 binds to cbl, which would explain the constitutive association between these two proteins. This result is consistent with studies describing the interaction between cbl and p85 as mediated by the SH3 domain of p85 (5, 6, 14, 16). The association of cbl with the p85 subunit of PI 3-kinase has been observed in studies of the T cell receptor, the B cell receptor, the Fc receptor, the EGF receptor, the interleukin-3 receptor, and the BCR-ABL oncogene (6, 7, 11, 17, 19). The fact that multiple receptor systems appear to stimulate association of p85 and cbl suggests that cbl may have a universal role as an adapter protein in coupling the activation of PI 3-kinase to activated growth factor receptors and oncogenes. The binding of the p85 subunit of PI 3-kinase to the receptor for platelet-derived growth factor results in the tyrosine phosphorylation of p85 and the activation of PI 3-kinase. We have not observed tyrosine phosphorylation of p85 after PRL stimulation, suggesting that this is not required to occur for activation of PI 3-kinase. Several studies have indicated that the binding of p85 to cbl is mediated by the binding of the SH2 domain(s) of p85 to tyrosine-phosphorylated cbl (6, 7, 11, 19). Two major cbl phosphorylation sites have been mapped using cbl purified from v-abl-transformed cells (39). Both of these phosphorylation sites fall into the pYXXP consensus class and would appear to be binding sites for crk and crkL (36, 37, 39). The p85 subunit of PI 3-kinase binds to cbl in BCR/ABL-and v-abl-transformed cells (19), suggesting that either p85 binds to one of these phosphorylation sites, or that additional sites of phosphorylation remain to be identified. It will be interesting to determine whether cbl is phosphorylated at these two sites in PRL-stimulated cells, and whether cbl bearing tyrosine to phenylalanine mutations at these two sites will function as a dominant-negative mutant in blocking PRL-stimulated activation of PI 3-kinase. It is reasonable to expect that both the SH2 and SH3 domains of p85 are involved in the interaction of p85 with cbl. Our model described below suggests the SH3 and SH2 domains play different roles in the interaction of p85 with cbl and that both interactions are important.
While this work was in progress, another group described the PRL-stimulated activation of PI 3-kinase in Nb2 cells (40). PI kinase activity was observed in antiphosphotyrosine immunoprecipitates of Nb2 cells stimulated with 100 ng/ml ovine PRL; however, a substantial amount of PI kinase activity was observed in immunoprecipitates of unstimulated cells. No activation was observed at lower concentrations of PRL. A modest increase in PI kinase activity was detected after 5 min; however, maximal activity was not observed until 20–40 min of stimulation. In contrast to our results, these investigators also described the tyrosine phosphorylation of p85 after PRL stimulation (40). We have consistently failed to observe the tyrosine phosphorylation of p85, even when concentrations of rat PRL as high as 200 nm were used to stimulate the cells. We cannot account for the difference between our results and those of Al-Sakkaf et al. (40) since Nb2 cells were used in both studies. Belanga et al. (41) have also recently described the phosphorylation of the p85 subunit of PI 3-kinase in 293 cells transfected with an epitope-tagged version of the rat PRLR after stimulation of these cells with rat PRL (41). We have not observed the tyrosine phosphorylation of p85 in either Nb2 cells or 32Dcl3 cells transfected with either the hPRLR or the rat Nb2 form of the PRLR. We suspect that the phosphorylation of p85 described by Belanga et al. may be specific to the 293 cells used in their studies.
The model that we would favor is that grb2-cbl complexes, bound by the SH3 domain of grb2 binding to cbl, exist in unstimulated cells. A small amount of the cbl protein is also complexed with p85; however, it is not clear whether the cbl-p85 complexes also contain grb2. After binding of PRL to the receptor, JAK2 becomes activated and phosphorylates the receptor at multiple sites. The grb2-cbl complex may bind to one of these phosphorylation sites, resulting in the phosphorylation of cbl by JAK2 or perhaps by fyn. After the tyrosine phosphorylation of cbl, one of the SH2 domains of the p85 subunit of PI 3-kinase presumably binds to a specific phosphotyrosine residue in cbl, resulting in the activation of PI 3-kinase. Our model predicts that both phosphotyrosine-independent and phosphotyrosine-dependent interactions between p85 and cbl, mediated by the SH3 and SH2 domains of p85, respectively, are critical in the regulation of PI 3-kinase by cbl. Alternatively, PI 3-kinase could be activated by the binding of the fyn SH3 domain to a proline-rich sequence in p85 as has been described by Pleiman et al. (42); however, this would not be consistent with the association of PI kinase activity with cbl. This model leads to several specific predictions: 1) that cbl associates with the PRLR via a grb2 SH2 domain-binding site; 2) that the SH3 domain of p85 binds to cbl; 3) that cbl becomes tyrosine phosphorylated after PRL stimulation; 4) after tyrosine phosphorylation of cbl, one of the SH2 domains of p85 binds to this phosphorylation site, resulting in the catalytic activation of PI 3-kinase; and 5) that fyn might be responsible for the phosphorylation of cbl. Each of these predictions can be tested by mapping the phosphorylation sites in cbl and mutating these phosphorylation sites to phenylalanine residues that cannot be phosphorylated. These tyrosine to phenylalanine point mutants would be expected to block activation of PI 3-kinase and perhaps block PRL-induced mitogenesis.
MATERIALS AND METHODS
Cells and Cell Culture
The Nb2 cell line was obtained from D. Li-yuan Yu-Lee (Baylor College of Medicine, Houston, TX) through the courtesy of Dr. Peter Gout. The cells were maintained in RPMI 1640 media supplemented with 10% FCS, 10% horse serum, 1 mml-glutamine, 100 U/ml penicillin, and 100 μg/ml streptomycin. All media components were obtained from GIBCO/BRL (Gaithersburg, MD). Charcoal-stripped FCS was obtained from HyClone (Logan, UT). Rat PRL, lots AFP-6452B and AFP-3697A, was obtained from the National Hormone and Pituitary Program (Rockville, MD).
Immunoprecipitation and Immunoblotting
Immunoprecipitation was as previously described (43). Cells were lysed in either RIPA or with EB. Both lysis buffers were supplemented with 100 U/ml kallikrein inhibitor (CalBiochem. La Jolla, CA). Rabbit anti-p85 subunit of PI 3-kinase and antiphosphotyrosine monoclonal antibody 4G10 coupled to agarose beads were obtained from Upstate Biotechnology, Inc. (Lake Placid, NY). Antibodies to cbl, grb2, and the p110 subunit of PI 3-kinase were obtained from Santa Cruz Biotechnology (Santa Cruz, CA). An additional antibody to the p85 subunit of PI 3-kinase was obtained from Transduction Labs (Lexington, KY). Immunoprecipitated proteins were resolved on SDS polyacrylamide gels and electro-transferred to Immobilon membrane (Millipore, Bedford, MA). Immunoblotting was conducted as described (17, 43) using the Enhanced Chemiluminescence Lighting (ECL) system according to manufacturer’s recommendations (Amersham, Arlington Heights, IL).
PI Kinase Assay
Reactions were a modification of previously described protocol (17, 44, 45). The immunoprecipitated proteins were washed three times with RIPA, twice with PAN [20 mm piperazine-N,N′-bis[2-ethanesulfonic acid] (pH 7.0), 20 μl/ml aprotinin, 100 mm NaCl], and resuspended in 50 μl PAN. A 5-μl aliquot of each sample was removed and placed in a new tube and 1 μl of 2 mg/ml PI in 4.5 mm EGTA, 90% dimethylsulfoxide was added to each reaction. The tubes were incubated at room temperature for 10 min before addition of the reaction mixture containing ATP and incubation at 30 C for 15 min. The final reaction mixture contained 20 mm HEPES (pH 7.4), 5 mm MgCl2, 0.45 mm EGTA, 10μ m ATP (5 μCi γ-[32P]ATP), and 0.2 mg/ml PI. Reactions were terminated by the addition of 0.1 ml 1 m HCl and extracted with 0.2 ml CHCl3-methanol (1:1). After the aqueous phase was discarded, the organic phase was reextracted with 1 m HCl-methanol (1:1), and the organic phase was dried in a Savant SpeedVac. The samples were dissolved in 10μ l CHCl3-methanol and spotted in Silica gel 60 plates (E. Merck) that had been impregnated with sodium tartrate, and the plate was developed in CHCl3-methanol-4 m NH4OH (9:7:2). After chromatography, the plate was allowed to dry before autoradiography.
GST Fusion Protein and Binding Assays
GST fusion proteins containing the amino-terminal SH2 domain, the carboxyl-terminal SH2 domain, and the SH3 domain of p85 were kindly provided by L. C. Cantley and B. Duckworth (Beth Isreal Hospital, Boston, MA) (46). GST fusion proteins were purified as previously described (17). Binding assays were conducted using lysates prepared in RIPA from equal numbers of cells (2 × 107 cells per binding assay) in 1 ml lysis buffer. Two nanomoles of the indicated GST fusion protein were added to each lysate to yield a final concentration of 2 μm. The remainder of the binding assay was conducted as described previously (17).
Densitometry
Gels were scanned with a Hewlett Packard ScanJet IIcx/T scanner (Palo Alto, CA) and images analyzed with Sigma Gel software (Sigma Chemical Co, St. Louis, MO). Care was taken not to overexpose the film such that the densitometer measurements were made in the linear range of the film.
Acknowledgments
Richard Klinghoffer and Andrius Kazlauskas kindly provided assistance with PI kinase assays. GST fusion proteins encoding the SH3 and SH2 domains of p85 were kindly provided by Brian Duckworth and Lewis Cantley. The authors also acknowledge the services of the University of Colorado Cancer Center DNA Sequencing Core Facility in support of this research. We thank Elizabeth Burton and Drs. Mary Reyland and Arthur Gutierrez-Hartmann for their comments on this manuscript.
This work was supported by NIH Grant DK-48879. The University of Colorado Cancer Center is supported by NIH Grant CA-46934.
