Co-expression with BCR induces activation of the FES tyrosine kinase and phosphorylation of specific N-terminal BCR tyrosine residues.

The human BCR gene encodes a protein with serine/threonine kinase activity and regulatory domains for the small G-proteins RAC and CDC42. Previous work in our laboratory has established that BCR is a substrate for c-FES, a non-receptor tyrosine kinase linked to myeloid growth and differentiation. Tyrosine phosphorylation led to the association of BCR with the RAS guanine nucleotide exchange complex GRB2-SOS in vivo via the GRB2 SH2 domain, linking BCR to RAS signaling (Maru, Y., Peters, K. L., Afar, D. E. H., Shibuya, M., Witte, O. N., and Smithgall, T. E. (1995) Mol. Cell. Biol. 15, 835-842). In the present study, we demonstrate that BCR Tyr-246 and at least one of the closely spaced tyrosine residues, Tyr-279, Tyr-283, and Tyr-289 (3Y cluster), are phosphorylated by FES both in vitro and in 32Pi-labeled cells. Mutagenesis of BCR Tyr-177 to Phe completely abolished FES-induced BCR binding to the GRB2 SH2 domain, identifying Tyr-177 as an additional phosphorylation site for FES. Co-expression of BCR and FES in human 293T cells stimulated the tyrosine autophosphorylation of FES. By contrast, tyrosine phosphorylation of BCR by FES suppressed BCR serine/threonine kinase activity toward the 14-3-3 protein and BCR substrate, BAP-1. These data show that tyrosine phosphorylation by FES affects the interaction of BCR with multiple signaling partners and suggest a general role for BCR in non-receptor protein-tyrosine kinase regulation and signal transduction.

The human BCR gene encodes a 160-kDa protein (BCR) with multiple biochemical functions. The N-terminal portion of BCR is a structurally distinct protein kinase capable of autophosphorylation on serine and threonine residues (1,2). This region of BCR also binds and phosphorylates BAP-1, a member of the 14-3-3 protein family that has been implicated in BCR and BCR-ABL function (3). The central domain of BCR is homologous to guanine-nucleotide exchange factors for RHO-related GTPases (4), while the C-terminal region exhibits GTPaseactivating protein activity toward these small G-proteins (5,6). Thus, BCR may regulate multiple small GTPases involved in mitogenic signaling, cytoskeletal organization, and regulation of NADPH oxidase activity in phagocytes (7)(8)(9)(10).
BCR was first discovered in the context of BCR-ABL, the transforming tyrosine kinase associated with chronic myelogenous leukemia (11). N-terminal, BCR-derived sequences are essential for BCR-ABL transforming activity and serve several functions. The extreme N-terminal portion of BCR-ABL encodes a coiled-coil oligomerization domain that may promote BCR-ABL activation and is indirectly required for BCR-ABL cytoskeletal localization (12). The C-terminal portion of the BCR kinase domain binds to the ABL SH2 domain in a phosphotyrosine-independent manner (13). This interaction has been proposed to release the ABL tyrosine kinase from negative regulation within BCR-ABL. BCR-derived sequences are also involved in BCR-ABL signal transduction. For example, tyrosine phosphorylation of BCR Tyr-177 within BCR-ABL leads to direct interaction with the RAS guanine nucleotide exchange complex GRB2-SOS via the GRB2 SH2 domain (14,15). BCR-ABL has also been linked to the SHC adaptor protein (15,16), although the specific mechanism of BCR-ABL/SHC interaction is unknown. Both pathways may contribute to the activation of RAS, which is required for transformation by BCR-ABL (17).
Accumulating evidence implicates normal BCR as a tyrosine kinase substrate and possible signaling intermediate. For example, BCR forms heteromeric complexes with BCR-ABL and is phosphorylated by BCR-ABL on multiple tyrosine residues including Tyr-177, the GRB2 binding site (18 -20). Recent work from our laboratory has shown that BCR is a major transformation-related substrate for the v-FPS tyrosine kinase and its normal human homolog, c-FES (21). Tyrosine phosphorylation led to the association of BCR with GRB2-SOS in v-FPS-transformed fibroblasts via the GRB2 SH2 domain. Furthermore, tyrosine phosphorylation of BCR by c-FES strongly enhanced the binding of BCR to multiple SH2 domains in vitro, including those from GRB2, RAS GTPase-activating protein, phospholipase C-␥, and the p85 subunit of phosphatidylinositol 3Јkinase (21). 1 These data strongly suggest that tyrosine phosphorylation of BCR induces interaction with downstream effectors that contain SH2 domains and implicate BCR as a key intermediate in signaling pathways regulated by BCR-ABL, FPS/FES, and possibly other non-receptor tyrosine kinases.
In the present study, we have mapped the specific tyrosine residues that are phosphorylated by c-FES within the BCR N-terminal region both in vitro and in vivo. We observed that tyrosine phosphorylation of BCR creates specific binding sites for the GRB2 and SHC SH2 domains, suggesting that BCR may couple FES to RAS signaling in a manner analogous to BCR-ABL. Unexpectedly, we observed that co-expression of FES and BCR in human cells stimulated FES tyrosine kinase activity while inhibiting BCR serine/threonine kinase activity toward the 14-3-3 protein, BAP-1. By contrast, co-expression of FES with a BCR mutant lacking the FES tyrosine phosphorylation sites completely blocked FES autophosphorylation in human cells. These data provide new evidence for BCR as an effector and regulatory protein for tyrosine kinases of the FPS/FES family and show that BCR is subject to regulation by tyrosine kinases in vivo.
Phosphorylation, Tryptic Phosphopeptide Mapping, and Phosphoamino Acid Analysis of GST-BCR Fusion Proteins in Vitro-Recombinant FES was expressed as a C-terminal FLAG fusion protein using a baculovirus/Sf9 cell system and purified using the anti-FLAG M2 affinity gel (21). Phosphorylation of GST-BCR 162-413 fusion proteins by recombinant FES was conducted in 40 l of kinase buffer (50 mM HEPES, pH 7.4, and 10 mM MgCl 2 ) containing 1 g of GST-BCR fusion protein and 10 Ci of [␥-32 P]ATP (3000 Ci/mmol, DuPont NEN). Phosphorylation reactions were incubated for 10 min at 30°C and stopped by heating at 95°C for 5 min in SDS-PAGE sample buffer. Phosphoproteins were resolved by SDS-PAGE and visualized by storage phosphor technology (Molecular Dynamics PhosphorImager). Two-dimensional tryptic phosphopeptide mapping and phosphoamino acid analysis are described elsewhere (24,25).
Construction and Expression of Full-length BCR Mutants-The cDNA encoding full-length BCR was subcloned into pSP70 (Promega). A unique SacI-StuI BCR fragment was cut from pSP70/BCR and subcloned into pLSMA4 (a gift of Dr. Solon Rhode, Eppley Institute, University of Nebraska Medical Center). A unique BamHI-BglII BCR fragment was cut from pLSMA4/BCR and subcloned into pSP72 (Promega). BCR sequences encoding Tyr mutations in the 162-413 region were cut from the pGEX-2T constructs described above with NcoI and BglII and swapped with the corresponding wild-type BCR fragment in the pSP72/ BCR construct. The resulting mutant fragments were cloned back through pLSMA4/BCR and pSP70/BCR to generate the full-length mutants. Full-length BCR wild-type or single Tyr to Phe mutants were then subcloned into the baculovirus transfer vector pVL1393 (Pharmingen) and the mammalian expression vector pcDNA3 (Invitrogen). The combination mutant Y177F/Y246F was made by replacing the StuI/ RsrII fragment of full-length BCR Y246F with the corresponding fragment of BCR Y177F. BCR Y177F/⌬3Y, Y246F/⌬3Y, and Y177F/Y246F/ ⌬3Y were made by replacing the ⌬3Y BssHII/SfiI fragment with the corresponding Y177F, Y246F, or Y177F/Y246F fragment. Preparation of recombinant baculoviruses and expression of BCR in Sf9 cells are described elsewhere (21,23).
In Vitro SH2 Domain Binding Assay-pGEX vectors containing the coding sequences of the GRB2 and SHC SH2 domains were provided by Dr. Yoshiro Maru (Institute of Medical Science, University of Tokyo). pGEX vectors for expression of the ABL and phosphatidylinositol 3-kinase p85 subunit SH2 domains and the anti-BCR antibody Rb-1 were provided by Dr. Owen Witte (Howard Hughes Medical Institute, UCLA). Details of the BCR-SH2 domain binding assay are described elsewhere (21). Briefly, subconfluent monolayers of Sf9 cells were infected with recombinant BCR wild-type or mutant baculoviruses either alone or with a FES baculovirus. Forty-eight hours postinfection, the cells were sonicated in 0.5 ml of lysis buffer (20 mM HEPES, pH 7.0, 150 mM NaCl, 5 mM EDTA, 1% Triton X-100, 1.0 mM Na 3 VO 4 , 0.05 mM Na 2 MoO 4 , 20 mM NaF, 1 mM phenylmethylsulfonyl fluoride, 50 g/ml aprotinin, and 25 g/ml leupeptin). Cell lysates were clarified by mi-crocentrifugation for 10 min at 4°C, and 0.1 ml aliquots were diluted with 0.9 ml incubation buffer (20 mM HEPES, pH 7.0, containing 150 mM NaCl, 0.1% Triton X-100, 10% glycerol, 1.0 mM Na 3 VO 4 , 0.05 mM Na 2 MoO 4 , 20 mM NaF, 1 mM phenylmethylsulfonyl fluoride, 50 g/ml aprotinin, and 25 g/ml leupeptin) and mixed with 100 nM immobilized GST-SH2 fusion protein. After 2 h of incubation at 4°C and then washing, the SH2-BCR protein complexes were separated on SDS-PAGE, transferred to polyvinylidene difluoride membranes, and immunoblotted with the anti-BCR antibody Rb-1.

Expression of FES and BCR Proteins in 293T Cells and in Vivo
Labeling-BCR wild-type or mutant proteins were transiently expressed either alone or with FES in 293T cells and then labeled with 32 P i as described elsewhere (26). BCR proteins were immunoprecipitated from the cell lysates with an anti-BCR monoclonal antibody (Santa Cruz Biotechnology) and protein G-Sepharose, separated by SDS-PAGE, and visualized by storage phosphor technology. Alternatively, the FES protein, which carries a C-terminal FLAG epitope tag (21), was immunoprecipitated from the transfected 293T cell lysates with the M2 anti-FLAG antibody affinity gel (Kodak Scientific Imaging Systems). FES autophosphorylation in the M2 immunoprecipitates and in the crude cell lysates was assessed by immunoblotting with antibodies to phosphotyrosine (PY20; Transduction Laboratories).
Phosphorylation of BAP-1 in Vitro-BCR was expressed either alone or with FES in human 293T cells and immunoprecipitated from cell lysates using the BCR monoclonal antibody. Phosphorylation of GST-BAP-1 was conducted in 50 l of kinase buffer containing 2 g of a GST-BAP-1 fusion protein and 10 Ci of [␥-32 P]ATP. Phosphorylation reactions were incubated for 10 min at 30°C and stopped by heating at 95°C for 5 min in SDS-PAGE sample buffer. Phosphoproteins were resolved by SDS-PAGE and visualized by storage phosphor technology. A pGEX vector for expression of GST-BAP-1 was generously provided by Dr. Yoshiro Maru.

Phosphorylation of the BCR 162-413 Region by FES in
Vitro-Recent work in our laboratory established that BCR is a target for the v-FPS and c-FES tyrosine kinases (21). Tyrosine phosphorylation by these kinases occurs within a region of the BCR N-terminal Ser/Thr kinase domain defined by amino acids 162-413 (21). 1 This BCR region contains nine tyrosine residues, which represent potential phosphorylation sites, including Tyr-177, which is the presumptive GRB2 binding site (14,15,20,21). To determine which of these Tyr residues are targeted by FES, we created a family of GST-BCR 162-413 fusion proteins with individual mutations of tyrosines 177, 231, 246, 316, 328, and 360 as well as a deletion of the closely spaced tyrosines 276, 279, and 283 (3Y cluster; see Fig. 1). The GST-BCR 162-413 fusion proteins were phosphorylated in vitro with recombinant FES and [␥-32 P]ATP. As shown in Fig. 2, the wild-type GST-BCR 162-413 fusion protein was readily phosphorylated by FES as observed previously with the FES hom- phosphorylation studies is also shown. Numbering above the diagram indicates the potential BCR tyrosine phosphorylation sites that were individually mutated to Phe (YF mutants). Numbering below the diagram indicates three closely spaced tyrosine residues (3Y cluster) that were deleted (⌬3Y mutant). Arrows show the tryptic digestion sites that separate these tyrosines in the BCR 162-413 region. GAP, RHO GTPase-activating protein domain. olog, v-FPS (21). On the other hand, the Y246F and ⌬3Y mutants were phosphorylated to a lesser extent than the wild-type fusion protein, suggesting that Tyr-246 and the 3Y cluster (Tyr-276, Tyr-279, and Tyr-283) may represent potential phosphorylation sites for FES. This result was confirmed by twodimensional tryptic phosphopeptide mapping (see below). All of the other mutant fusion proteins were phosphorylated to approximately the same extent as the wild-type. GST itself was not phosphorylated by FES in vitro (data not shown), indicating that all of the phosphorylation sites are located within the BCR-derived sequence. Phosphoamino acid analysis showed that the fusion proteins were phosphorylated exclusively on tyrosine by FES (data not shown).
Two-dimensional Tryptic Mapping of BCR Tyrosine Phosphorylation Sites for FES in Vitro-Two-dimensional tryptic phosphopeptide analysis was performed to identify the GST-BCR 162-413 tyrosine residues phosphorylated by FES in vitro. As shown in Fig. 3, FES-phosphorylated GST-BCR 162-413 wild type gave rise to two phosphopeptides. By contrast, the Y246F and ⌬3Y mutants each yielded only one phosphopeptide, consistent with the prediction from Fig. 2 that these are phosphorylation sites for FES. The positions of the individual phosphopeptides observed with the mutants correspond to those observed in the wild-type tryptic map as determined by phosphopeptide mixing experiments (data not shown). All other point mutants resulted in a two-dimensional tryptic map identical to the wild type. These data indicate that Tyr-246 and one or more tyrosines in the 3Y cluster (Tyr-276, Tyr-279, and Tyr-283) are phosphorylated by FES in vitro.
Phosphorylation of BCR by v-FPS and c-FES induces GRB2 binding via the GRB2 SH2 domain (21), which is predicted to occur via Tyr-177 as demonstrated previously with BCR-ABL (14,15). To our surprise, FES did not detectably phosphorylate the GST-BCR fusion protein on Tyr-177 in vitro (Fig. 3). To determine whether the lack of GST-BCR Tyr-177 phosphorylation was unique to FES, we phosphorylated the wild-type and Y177F mutant fusion proteins with the p185 form of BCR-ABL and performed two-dimensional tryptic phosphopeptide analysis. The resulting phosphopeptide maps were identical, indicating that BCR-ABL also failed to phosphorylate GST-BCR on Tyr-177 in vitro (data not shown). These results suggest that Tyr-177 is only phosphorylated to a minor extent or may not be accessible for phosphorylation in the context of the fusion protein. However, GRB2 SH2 domain binding experiments with full-length BCR clearly indicate that Tyr-177 is phosphorylated by FES in living cells (see below).
Tyrosine Phosphorylation of BCR by FES in Intact Cells-To verify that the same BCR sites phosphorylated by FES in vitro are also utilized in living cells, full-length BCR was expressed either alone or with FES in human 293T cells and labeled with 32 P i . Labeled BCR was immunoprecipitated and subjected to phosphoamino acid analysis and two-dimensional tryptic phosphopeptide mapping. As shown in Fig. 4A, BCR was phosphorylated primarily on Ser when expressed alone but was additionally phosphorylated on tyrosine when co-expressed with FES. Fig. 4B shows that BCR alone gives rise to nine Ser/Thr phosphopeptides, which are likely to arise from BCR autophosphorylation. Co-expression of BCR with FES gave rise to three additional phosphopeptides (Fig. 4B, peptides a, b, and c). Phosphoamino acid analysis of these three new phosphopeptides showed that they contain phosphotyrosine (data not shown).
To identify the BCR tyrosines phosphorylated by FES in vivo, FES was co-expressed with full-length BCR proteins containing the same series of tyrosine mutations shown in Fig. 1. The co-transfected cells were labeled with 32 P i , and BCR was immunoprecipitated and analyzed by two-dimensional tryptic mapping. As shown in Fig. 4B Two-dimensional tryptic mapping also showed that co-expression of BCR with FES caused the loss of serine phosphopeptide 9 from BCR (Fig. 4B). This observation suggests that tyrosine phosphorylation of BCR by FES may affect BCR autophosphorylation in vivo. As described in more detail below, Tyr phosphorylation of BCR by FES also reduced the Ser/Thr kinase activity of BCR toward the 14-3-3 protein, BAP-1. Alternatively, phosphorylation of this peptide on tyrosine may affect its position in the two-dimensional map.
Characterization of BCR-SH2 Domain Binding Specificity in Vitro-Previous work from our laboratory has shown that transformation of 3Y1 cells with the FES homolog v-FPS led to BCR/GRB2-SOS interaction via the GRB2 SH2 domain (21). Furthermore, tyrosine phosphorylation of BCR by FES strongly enhanced BCR binding to the SH2 domains of GRB2, ABL, p85, and other signaling proteins in vitro (21). 1 To identify the tyrosine residues responsible for recruitment of specific SH2 domains, SH2 binding assays were conducted with wildtype and tyrosine phosphorylation site mutants of BCR. Recombinant SH2 domains from ABL, GRB2, p85 (C-terminal), and SHC were incubated with Sf9 cell lysates expressing fulllength BCR or the Y177F, Y246F, and ⌬3Y mutants either alone or with FES. Following incubation and washing, bound BCR proteins were visualized by immunoblotting. As shown in Fig. 5, tyrosine phosphorylation induced strong association of BCR with all of these GST-SH2 fusion proteins. Note that the concentrations of the SH2 fusion proteins used in these experiments was 100 nM, which is within the range of binding constants for physiological SH2-target protein interactions (27).
Mutagenesis of Tyr-177 to Phe completely abolished the FES-induced binding of BCR to the SH2 domain of GRB2 but did not affect binding to the other SH2 domains (Fig. 5). This result clearly identifies BCR Tyr-177 as the FES-induced site of BCR-GRB2 interaction and indicates that Tyr-177 is an in vivo phosphorylation site for FES. Deletion of the 3Y cluster specifically abolished binding of BCR to the SHC SH2 domain, suggesting that phosphorylation of this BCR region by FES creates a binding site for the SHC SH2 domain. By contrast, all of the single tyrosine mutants as well as ⌬3Y bound to the ABL and p85 SH2 domains following FES-mediated phosphorylation, indicating that more than one phosphotyrosine residue or a phosphotyrosine residue outside of the BCR 162-413 region mediates these binding interactions.
Data shown in Fig. 5 demonstrate that tyrosine phosphorylation of BCR by FES greatly enhanced ABL and p85 SH2 domain binding, possibly by creating multiple phosphotyrosinedependent sites. To test this hypothesis, SH2 binding assays were conducted using BCR proteins with all possible combinations of mutations of the FES phosphorylation sites (Y177F/ Y246F, Y177F/⌬3Y, Y246F/⌬3Y, and Y177F/Y246F/⌬3Y mutants). As shown in Fig. 6, all of the BCR double mutants exhibited diminished binding to the ABL and p85 SH2 domains while the triple mutant (Y177F/Y246F/⌬3Y) did not bind to either of these SH2 domains. These results demonstrate that FES-mediated phosphorylation of multiple BCR tyrosine residues in the 162-413 region is required for maximal ABL and p85 SH2 domain binding.
Stimulation of FES Tyrosine Kinase Activity by BCR in Vivo-During the analysis of BCR phosphorylation in 32 P ilabeled 293T cells, we observed that co-expression of BCR with FES enhanced the phosphotyrosine content of FES. This finding suggested that FES-BCR interaction may activate FES in vivo. To test this idea directly, FES was expressed alone or with BCR in 293T cells, and tyrosine autophosphorylation was assessed both in anti-FES immunoprecipitates and clarifed cell lysates by anti-phosphotyrosine immunoblot analysis. As shown in Fig. 7, autophosphorylation of FES is very weak in vivo when expressed alone, consistent with published findings from other systems (28,29). However, co-expression of FES with BCR strongly activated FES autophosphorylation, leading to extensive BCR phosphorylation. The BCR mutants Y177F, Y246F, ⌬3Y, Y177F/Y246F, and Y177F/⌬3Y, all of which are phosphorylated by FES, activated FES autophosphorylation to FIG. 4

. Tyrosine phosphorylation of BCR by FES in vivo.
Human 293T cells expressing wild-type or mutant forms of BCR either alone or together with FES were labeled with 32 P i . Labeled BCR was immunoprecipitated from the clarified cell lysates with anti-BCR monoclonal antibodies, separated by SDS-PAGE, and digested with trypsin. A portion of the tryptic digest was hydrolyzed with HCl for phosphoamino acid analysis. A, phosphoamino acids were separated by two-dimensional electrophoresis on thin layer plates. B, tryptic phosphopeptides were separated by electrophoresis (right to left) and chromatography (bottom to top). BCR alone consistently yielded 9 serine phosphopeptides (numbered; top left). Three new BCR phosphopetides were observed in the presence of FES (peptides a, b, and c; top right). Peptide b was not observed with the BCR Y246F mutant in the presence of FES (bottom left), while peptide a was not observed with the BCR ⌬3Y mutant (bottom right).

FIG. 5. Phosphorylation of BCR Tyr-177 and the 3Y cluster by FES creates specific binding sites for the GRB2 and SHC SH2
domains, respectively. SH2 domain binding assays were conducted with immobilized GST fusion proteins containing the SH2 domains of GRB2, SHC, ABL, and the phosphatidylinositol 3-kinase p85 subunit (C-terminal SH2). Immobilized GST-SH2 domain fusion proteins were mixed with cell lysates from Sf9 cells expressing wild-type (WT) or mutant BCR proteins in the absence and presence of FES-FLAG. Following incubation and washing, SH2-bound BCR proteins were visualized by immunoblotting (top four panels). Expression of BCR proteins, FES-FLAG, and tyrosine phosphorylation of BCR were verified in the clarified cell lysates by immunoblotting with the anti-BCR antibody Rb-1, the anti-FLAG antibody M2, and the anti-phosphotyrosine antibody, PY20, respectively (bottom three panels).
almost the same extent as wild-type BCR. However, co-expression with Y246F/⌬3Y, which is weakly phosphorylated by FES, did not activate FES. Co-expression with Y177F/Y246F/⌬3Y mutant, which lacks all known Tyr phosphorylation sites for FES, completely suppressed FES autophosphorylation. These results suggest that BCR can stimulate FES tyrosine kinase activity in vivo and that tyrosine phosphorylation of BCR is required for this effect.
To determine if the effect of BCR on FES is unique to mammalian cells, the same experiment was conducted in Sf9 insect cells. As shown in Fig. 8, FES autokinase activity was very strong when expressed alone, and co-expression with wild-type and mutant forms of BCR had no additional activating effect. These findings suggest that FES kinase activity may be suppressed by a factor present in mammalian cells and that interaction with BCR overcomes this inhibitory effect, leading to activation (see "Discussion").
To verify that the effect of BCR on FES tyrosine phosphorylation was direct and not mediated by activation of another tyrosine kinase, we co-expressed BCR and a kinase-defective mutant of FES (K590E mutant) (24) in 293T cells. No autophosphorylation of this FES mutant was observed in the presence or absence of BCR, indicating that stimulation of FES autophosphorylation by BCR requires the kinase activity of FES (data not shown). However, our data cannot rule out the possibility that BCR could inhibit a phosphotyrosine phosphatase unique to mammalian cells.
Tyrosine Phosphorylation by FES Suppresses BCR Ser/Thr Kinase Activity Toward the 14-3-3 Protein, BAP-1-Tryptic phosphopeptide analysis shows that tyrosine phosphorylation may affect BCR autophosphorylation (Fig. 4). To determine whether tyrosine phosphorylation of BCR affects its Ser/Thr kinase activity toward a substrate, BCR was expressed either alone or with FES in 293T cells followed by in vitro kinase assay with the 14-3-3 protein and BCR substrate, BAP-1 (3). As shown in Fig. 9, co-expression with FES suppressed BAP-1 phosphorylation by BCR by more than 60%. Control experiments showed that equal amounts of BCR are present in each immunoprecipitate and that the end point of the reaction shown falls on the linear portion of the progress curve for the phosphorylation reaction (data not shown). We also verified that BAP-1 is not a substrate for FES and does not bind to FES FIG. 6. Phosphorylation of more than one tyrosine residue is required for maximal binding of BCR to the ABL and phosphatidylinositol 3-kinase p85 SH2 domains. Immobilized ABL and p85 C-terminal GST-SH2 fusion proteins were mixed with cell lysates from Sf9 cells expressing wild-type (WT) or mutant BCR proteins and FES-FLAG. Following incubation and washing, bound BCR proteins were visualized by immunoblotting (top two panels). Expression of BCR proteins, FES-FLAG, and tyrosine phosphorylation of BCR proteins were verified in the clarified cell lysates by immunoblotting with the anti-BCR antibody Rb-1, the anti-FLAG antibody M2, and the antiphosphotyrosine antibody, PY20, respectively (bottom three panels). No SH2 domain binding of these BCR mutants was observed in the absence of FES co-expression (data not shown).

FIG. 7. Co-expression with BCR stimulates FES tyrosine kinase activity in human 293T cells.
293T cells were transfected with a pcDNA3/FES-FLAG expression construct either alone or with pcDNA3 constructs containing the wild-type (WT) or mutant BCR sequences shown. FES was immunoprecipitated from transfected cell lysates using the anti-FLAG antibody affinity gel, and FES protein levels and autophosphorylation were assessed by immunoblotting the precipitates with the anti-FLAG antibody M2 and PY20, respectively (top two panels). Expression of BCR proteins, FES-FLAG, and tyrosine phosphorylation of BCR were verified in the clarified cell lysates by immunoblotting with the anti-BCR antibody Rb-1, the anti-FLAG antibody M2, and the anti-phosphotyrosine antibody, PY20, respectively (bottom three panels). (data not shown). These results indicate that Tyr phosphorylation of BCR by FES leads to suppression of BCR Ser/Thr kinase activity. DISCUSSION Using a combination of tryptic phosphopeptide mapping and SH2 domain binding assays, we have identified BCR Tyr-177, Tyr-246, and one or more of three closely spaced tyrosine residues (Tyr-276, Tyr-279, Tyr-283, the 3Y cluster) as phosphorylation sites for FES in living cells. Phosphorylation of Tyr-177 and the 3Y cluster creates specific binding sites for the SH2 domains of GRB2 and SHC, suggesting that BCR may serve as an intermediate linking FES to the RAS signal transduction pathway. These results are consistent with our previous finding that tyrosine-phosphorylated BCR complexes with GRB2-SOS in fibroblasts transformed with the FES homolog, v-FPS (21). BCR also possesses regulatory domains for small G-proteins of the RHO family, including RAC and CDC42 (5,6). Recent studies have shown that activation of these GTPases is required for transformation by RAS and for normal and oncogenic signal transduction by tyrosine kinases (reviewed in Ref. 8). In this regard, BCR may serve to integrate tyrosine kinase signaling through the RAS and RHO signaling pathways.
A recent study has mapped several BCR tyrosine residues that are phosphorylated within BCR-ABL (20). Although FES shares some of these phosphorylation sites (Tyr-177 and possibly Tyr-283), both kinases phosphorylate unique BCR sites as well. For example, BCR-ABL but not FES phosphorylates BCR Tyr-360 (20), while FES uniquely phosphorylates BCR Tyr-246. Although the result of specific phosphorylation of BCR Tyr-246 by FES is unclear at present, it may create an additional site for the recruitment of SH2 domain effectors to the BCR N-terminal domain. Alternatively, phosphorylation of this site could affect SH2 domain binding indirectly by altering the conformation of the BCR N-terminal domain or influence BCR Ser/Thr kinase activity (see below).
Data presented here suggest that BCR may serve as a positive regulator of FES tyrosine kinase activity in vivo. Like other cytoplasmic protein-tyrosine kinases, FES has tyrosine autokinase activity that is readily detectable in an in vitro immune complex kinase assay. Autophosphorylation is also observed in vivo when FES is expressed in non-mammalian systems, such as Sf9 cells (Fig. 7). However, Tyr autophosphorylation of FES is strongly inhibited in mammalian cells (26,28,29), suggesting that it is regulated in trans by a factor unique to mammalian cells. Our results show that co-expression of FES with BCR strongly stimulated FES tyrosine kinase activity in human cells, suggesting that BCR is a positive regulatory factor for FES. Previous work from our laboratory has shown that the FES unique N-terminal and SH2 domains bind to BCR (21). FES-BCR interaction may lead to displacement of a negative regulatory protein and release of FES tyrosine kinase activity. The ability of BCR to activate FES is in some ways analogous to the activation of the ABL tyrosine kinase within BCR-ABL. Like FES, the c-ABL tyrosine kinase may be negatively regulated by non-covalent association with a cellular factor, such as the recently described ABL binding protein, Abi-2 (30). Fusion to BCR releases the tyrosine kinase activity of ABL, an effect that may be dependent upon direct interaction of the ABL SH2 domain with BCR-derived sequences (13). Such an interaction may prevent interaction with the ABL regulatory factor.
Tyrosine phosphorylation is also likely to influence the serine/threonine kinase activity of BCR in vivo. Results shown in Fig. 9 demonstrate that tyrosine phosphorylation of BCR by FES suppresses BCR serine/threonine kinase activity toward the 14-3-3 protein BAP-1 (3). The mechanism of this suppression may involve decreased affinity of BCR for BAP-1 as a result of tyrosine phosphorylation. A recent study has shown that 14-3-3 proteins bind to serine-phosphorylated sequences within target proteins with high affinity and specificity but show no affinity for the unphosphorylated sequence (31). Tyrosine phosphorylation of BCR by FES may inhibit the serine autophosphorylation of BCR in vivo (Fig. 4), resulting in the loss of a binding site for BAP-1 and other 14-3-3 proteins. The 14-3-3 proteins have been shown to link BCR and RAF in vivo, which may alter the subcellular localization and function of these kinases (22). Decreased phosphorylation of BAP-1 as a result of FES-induced tyrosine phosphorylation may influence signal transduction by both BCR and BCR-ABL and affect their ability to interact with RAF or other signaling partners via 14-3-3 in vivo. FIG. 9. Tyrosine phosphorylation by FES suppresses BCR Ser/ Thr kinase activity toward the 14-3-3 protein BAP-1 in vitro. BCR was expressed alone or with FES in 293T cells and immunoprecipitated with a monoclonal antibody. Aliquots of the BCR immunoprecipitates (20 or 40 l) were incubated with a purified GST-BAP-1 fusion protein and [␥-32 P]ATP. A, phosphorylated proteins were separated by SDS-PAGE and visualized by storage phosphor technology. B, 32 P incorporation into GST-BAP-1 was quantitated directly from the imaged gel shown in A and plotted as a percent of control. Control immunoblots verified equivalent levels of BCR proteins in the two immunoprecipitates, and no detectable phosphorylation of GST by BCR was observed under these conditions (data not shown).