Tyrosine Kinase Activity of the EGF Receptor Is Enhanced by the Expression of Oncogenic 70Z-Cbl

Tyrosine kinase activity of the EGF receptor is enhanced by the expression of oncogenic 70Z-Cbl

Christine BF Thien and Wallace Y Langdon

Department of Pathology, The University of Western Australia, Queen Elizabeth II Medical Centre, Nedlands, Western Australia 6907, Australia

The 120 kD product of the c-Cbl oncogene is a signal transduction across the cell membrane into the prominent substrate of protein tyrosine kinases that cytoplasm. Furthermore, Cbl has been found to lacks a known catalytic activity but possesses an array associate constitutively or inductively with many of binding sites for cytoplasmic signalling proteins. An signalling proteins such as the Grb2, Crk and Nck oncogenic form of Cbl was recently identi®ed in the 70Z/ adaptor proteins, members of the Src, Abl and Syk 3 pre-B cell lymphoma which has a small deletion at the tyrosine kinase families, the p85 regulatory subunit of N-terminus of the Ring ®nger domain. This form of Cbl, PI 3-kinase and 14-3-3 proteins (Donovan et al., 1994; termed 70Z-Cbl, exhibits an enhanced level of tyrosine de Jong et al., 1995; Buday et al., 1996; Rivero- phosphorylation compared with c-Cbl. Here we demon- Lezcano et al., 1994; Ribon et al., 1996; Andoniou et strate that the expression of 70Z-Cbl induces a tenfold al., 1994, 1996; Smit et al., 1996; Tanaka et al., 1996; enhancement in the kinase activity of the EGF receptor Tsygankov et al., 1996; Fournel et al., 1996; Ota et al., in serum-starved and EGF-stimulated cells. In serum- 1996; Kim et al., 1995; Meisner et al., 1995; Solto and starved cells this results in EGF receptor autopho- Cantley, 1996; Liu et al., 1996). To date however a sphorylation and the recruitment of Grb2, Shc and Sos1 de®nite function for Cbl has not emerged from these but does not induce a corresponding increase in MAP studies. The most revealing clue about the function of kinase activity. Furthermore the expression of 70Z-Cbl Cbl has come from genetic studies in C. elegans where greatly enhances EGF-induced tyrosine phosphorylation the Cbl homologue, Sli-1, has been identi®ed as a of the protein tyrosine phosphatase SHP-2. We also negative regulator of the Let-23 receptor tyrosine show that the Cbl/EGF receptor complex is predomi- kinase (Jongeward et al., 1995; Yoon et al., 1995). nantly associated with CrkII and is distinct to the Grb2/ Importantly these experiments have demonstrated that Shc/Sos1 complex that associates with the EGF Sli-1 acts at the level of Let-23 and the Sem5 adaptor receptor. These ®ndings therefore demonstrate a bio- (Jongeward et al., 1995), a ®nding consistent with chemical eect of an oncogenic Cbl protein and support mammalian studies that place Cbl at an initiating point predictions from C. elegans that Cbl functions as in tyrosine kinase mediated signal transduction. regulator of receptor tyrosine kinases. Studies of oncogenic forms of Cbl have also provided clues about Cbl function. Cbl can be Keywords: growth factor; oncogene; receptor; signal converted to an oncogenic protein either by a large transduction; tyrosine kinase carboxy truncation that generated v-Cbl, or by a small internal deletion at the amino terminus of the Ring ®nger that was identi®ed in a mutant allele of Cbl from the mouse pre-B cell lymphoma line, 70Z/3 (Figure 1a) Introduction (Langdon et al., 1989; Blake et al., 1991; Andoniou et al., 1994). The v-Cbl truncation removes a large The product of the c-Cbl proto-oncogene has been proline-rich SH3-binding region and the Ring ®nger identi®ed as a ubiquitous substrate of protein tyrosine domain to reveal a novel phosphotyrosine binding kinases. Cbl lacks a de®ned catalytic domain but is (PTB) domain that forms a direct association with the rapidly phosphorylated on tyrosine residues following ZAP-70 tyrosine kinase and the EGF receptor (Lupher the stimulation of a wide range of cell surface receptors et al., 1996; Thien and Langdon, 1997). These studies which include growth factor receptors, immunoglobu- suggest that v-Cbl competes with c-Cbl for binding lin receptors, antigen receptors and integrin receptors sites on activated receptor complexes, and that (Donovan et al., 1994; Tanaka et al., 1995; Galisteo et transformation could involve a dominant negative al., 1995; Bowtell and Langdon, 1995; Odai et al., mechanism that blocks the putative regulatory role of 1995; Marcilla et al., 1995; Wang et al., 1996; Cory et c-Cbl. Indeed we have found that v-Cbl transformation al., 1995; Panchamoorthy et al., 1996; Kontani et al., is dependent on the expression of this protein at high 1996; Ota et al., 1996; Ojaniemi et al., 1997). Indeed in levels (WL, unpublished). In contrast, transformation many cell types Cbl appears to be one of the most by 70Z-Cbl appears to involve a positive signalling prominent and rapidly phosphorylated substrates of mechanism. This protein exhibits a markedly enhanced protein tyrosine kinases (Andoniou et al., 1994; level of tyrosine phosphorylation under conditions of Donovan et al., 1994; Panchamoorthy et al., 1996). minimal growth factor stimulation, and has the This suggests a pivotal role in early events following capability to transform ®broblasts at protein levels where v-Cbl transformation is not evident (Andoniou et al., 1994; Bowtell and Langdon, 1995). Recently the Correspondence: WY Langdon eects of 70Z-Cbl on the transcriptonal activation of Received 2 June1997; revised 1 August 1997; accepted 1 August 1997 nuclear factor of activated T cells (NFAT) were 70Z-Cbl enhancement of EGF receptor kinase activity CBF Thien and WY Langdon 2910 The mechanism by which 70Z-Cbl mediates fibro- a blast transformation and Ras-dependent activation of NFAT remains to be resolved. However the elevated level of tyrosine phosphorylation of 70Z-Cbl protein compared to c-Cbl has provided direction for further investigation. We originally hypothesized that the enhanced tyrosine phosphorylation was due to an intrinsic property of 70Z-Cbl which increased its accessibility to tyrosine kinases (Andoniou et al., 1994). In this study we have further investigated the mechanism of 70Z-Cbl tyrosine phosphorylation using b NIH3T3 ®broblasts expressing the human epidermal growth factor receptor (EGFR). The ®ndings presented here show that the 70Z-Cbl protein markedly enhances the kinase activity of the EGFR and that this results in an increase in the tyrosine phosphorylation of the EGFR, Shc, Cbl and SHP-2.

Results

Tyrosine phosphorylation of 70Z-Cbl in serum-starved cells To determine the mechanism for the enhanced tyrosine phosphorylation of 70Z-Cbl we examined NIH3T3 cells that co-express the EGFR and HA-tagged constructs of c-Cbl, 70Z-Cbl or v-Cbl (Figure 1a). The morphology of these cells is shown in Figure 1b c d where cells with no introduced Cbl (a), or cells overexpressing HA-c-Cbl (b) exhibit a ¯at morphol- -- c-Cbl 70Z-Cbl v-Cbl -- c-Cbl 70Z-Cbl v-Cbl -- c-Cbl 70Z-Cbl v-Cbl -- c-Cbl 70Z-Cbl v-Cbl ogy, and cells expressing HA-70Z-Cbl (c) and HA-v- EGF : – – – – + + + + EGF : – – – – + + + + Cbl (d) are refractile and rounded in appearance 202 – – EGFR consistent with a transformed phenotype. These cells were grown to *90% con¯uency in complete media – c-Cbl – c-Cbl 103 – before replacement with media containing 0.5% FCS for 24 h. The cells were either left quiescent or were 68 – stimulated with 1 ng/ml of EGF for 2 min at 378C before lysis and immunoprecipitation with anti-HA antibodies. The immunoprecipitated proteins were 44 – – v-Cbl separated by SDS polyacrylamide gel electrophoresis and immunoblotted with anti-phosphotyrosine or anti- I.P. : Anti-HA I.P. : Anti-HA Blot : Anti-HA Blot : Anti-P-Tyr HA antibodies. The anti-HA immunoblot in Figure 1c shows the expression of the tagged Cbl proteins and Figure 1 (a) Diagrammatic representation of the HA-tagged c- reveals that 70Z-Cbl has the lowest level of protein Cbl, 70Z-Cbl and v-Cbl constructs used in this study. (b) Morphology of NIH3T3 cells expressing the human EGFR with expression. The anti-phosphotyrosine blot of these (a) no introduced Cbl; (b) HA-c-Cbl; (c) HA-70Z-Cbl and (d) HA- samples shows that in quiescent cells (Figure 1d, v-Cbl. (c) and (d) Enhanced tyrosine phosphorylation of 70Z-Cbl EGF7) the 70Z-Cbl protein has a markedly higher in serum-starved cells. NIH3T3 cells expressing the human EGFR level of tyrosine phosphorylation compared to c-Cbl and HA-tagged c-Cbl, 70Z-Cbl or v-Cbl constructs were grown in media containing 0.5% FCS for 24 h. Cells were left quiescent protein, even though it is expressed at a lower level, (EGF7) or were stimulated with 1 ng/ml of EGF for 2 min at and that it is associated with tyrosine phosphorylated 378C (EGF+) before lysis and immunoprecipitation with anti-HA EGFR. These results are consistent with our previous antibodies. The immunoprecipitated proteins were separated by reports that have shown a high level of 70Z-Cbl SDS polyacrylamide gel electrophoresis and immunoblotted with tyrosine phosphorylation in minimally stimulated cells anti-HA antibodies (c) or anti-phosphotyrosine antibodies (d). NIH3T3 cells expressing the human EGFR but with no (Andoniou et al., 1994; Bowtell and Langdon, 1995). introduced form of Cbl is represented by ± Following EGF stimulation there is a large increase in the amount of tyrosine phosphorylated c-Cbl protein and this protein is recruited to the activated EGFR (Figure 1d,EGF+). There is also a slight increase in investigated in Jurkat T cells. Transient expression of 70Z-Cbl tyrosine phosphorylation, however the 70Z-Cbl, but not c-Cbl or v-Cbl, was found to induce amount of immunoprecipitated protein was markedly an increase in the basal activity of NFAT which was reduced after EGF stimulation since it was not further enhanced by calcium ionophore (Liu et al., detected with anti-HA antibodies. The explanation 1997). Furthermore coexpression of a dominant for this reduction is not known, but 70Z-Cbl protein negative Ras abrogated the 70Z-Cbl mediated activa- could not be detected by anti-HA immunoblotting of tion of NFAT. total cell lysates which suggests it is susceptible to 70Z-Cbl enhancement of EGF receptor kinase activity CBF Thien and WY Langdon 2911 degradation after EGF treatment. As previously shown v-Cbl protein is also recruited to the activated receptor, a but unlike c-Cbl and 70Z-Cbl it is not tyrosine phosphorylated (Bowtell and Langdon, 1995). -- c-Cbl 70Z-Cbl v-Cbl A noteable feature of the anti-phosphotyrosine blot EGF : – – – – + -- + c-Cbl + 70Z-Cbl +v-Cbl of EGF-stimulated cells was the high molecular weight smear associated with the EGFR that co-immunopre- – EGFR cipitated with c-Cbl but not with 70Z-Cbl or v-Cbl. The same eect on the EGFR in cells overexpressing c- I.P. : Anti-EGFR Cbl is seen in Figures 2 and 4 where lysates were Blot : Anti-P-Tyr immunoprecipitated with antibodies to the EGFR, Grb2, Shc and Sos1. This observation raises the possibility that overexpression of c-Cbl can aect the – EGFR level of EGFR ubiquitination, a possibility that is currently under investigation. I.P. : Anti-EGFR Blot : Anti-EGFR Tyrosine phosphorylation of the EGFR in serum-starved cells transformed by 70Z-Cbl b 0 0.01 0.1 1 To examine whether the enhanced tyrosine phosphor- EGF : ylation of 70Z-Cbl could be a result of an increase in (ng/ml) Cbl 7OZ Cbl 7OZ Cbl 7OZ Cbl 7OZ kinase activity of the EGFR we investigated EGFR tyrosine phosphorylation in serum-starved and EGF- – EGFR stimulated ®broblasts. Since autophosphorylation provides an indication of EGFR kinase activity we immunoprecipitated the EGFR from quiescent and I.P. : Anti-EGFR EGF stimulated cells expressing HA-tagged c-Cbl, 70Z- Blot : Anti-P-Tyr Cbl or v-Cbl and immunoblotted with anti-phosphotyrosine antibodies. Figure 2a shows a marked – EGFR enhancement in EGFR tyrosine phosphorylation in serum-starved cells that express 70Z-Cbl. In contrast we could not detect EGFR tyrosine phosphorylation in I.P. : Anti-EGFR serum-starved cells that express c-Cbl or v-Cbl. Blot : Anti-EGFR Following EGF stimulation the receptors from all Figure 2 Enhanced tyrosine phosphorylation of the EGFR in four cell lines are tyrosine phosphorylated at levels that minimally stimulated cells expressing 70Z-Cbl. (a) NIH3T3 cells appear equivalent. Therefore at this level of growth expressing the human EGFR and HA-tagged c-Cbl, 70Z-Cbl or factor stimulation (i.e. 1 ng/ml of EGF for 2 min at v-Cbl constructs were grown in media containing 0.5% FCS for 24 h. Cells were left quiescent (EGF7) or were stimulated with 378C) 70Z-Cbl does not induce an obvious enhance- 1 ng/ml of EGF for 2 min at 378C (EGF+) before lysis and ment in EGFR phosphorylation. Furthermore we did immunoprecipitation with anti-EGFR antibodies. The immuno- not observe a suppression in EGFR tyrosine phos- precipitates were analysed by immunoblotting with anti-phospho- phorylation in cells overexpressing c-Cbl. This was tyrosine or anti-EGFR antibodies. (b) NIH3T3 cells expressing surprising in view of the putative negative regulatory the human EGFR and HA-tagged c-Cbl or 70Z-Cbl constructs were grown in media containing 0.5% FCS for 24 h. Cells were role of c-Cbl and recent ®ndings that EGFR left quiescent (EGF:0) or were stimulated with a range of EGF phosphorylation is suppressed in cells overexpressing concentrations for 2 min at 378C before lysis and immunopreci- c-Cbl (Ueno et al., 1997). At present it is not clear why pitation with anti-EGFR antibodies. The immunoprecipitates these two studies dier in c-Cbl's eect on the EGFR, were analysed by immunoblotting with anti-phosphotyrosine or however we have also failed to see suppression of anti-EGFR antibodies EGFR tyrosine phosphorylation in BALB 3T3 ®broblasts overexpressing c-Cbl (data not shown). Since 1 ng/ml of EGF induced a uniformly high level of EGFR phosphorylation we investigated the possibility was tested directly by an in vitro kinase eects of stimulating c-Cbl and 70Z-Cbl cells with assay. Immunoprecipitated EGFR from serum-starved lower concentrations of EGF. Figure 2b show that the cells and cells stimulated with 0.1 or 1.0 ng/ml of EGF eect of 70Z-Cbl on EGFR tyrosine phosphorylation is was incubated with an optimized peptide substrate that most notable at low levels of stimulation (i.e. 0 and has a single tyrosine residue which is readily 0.01 ng/ml of EGF). These results provide the ®rst phosphorylated by a variety of protein tyrosine evidence that 70Z-Cbl increases the tyrosine phosphor- kinases including the EGFR (Schaefer and Hsaio, ylation of the EGFR under conditions of minimal 1996). The biotin labeled peptide was isolated by growth factor stimulation. membrane capture and its incorporation of (g-32P)ATP was measured by scintillation counting. The results in Figure 3 show that the kinase activity of the EGFR in Enhanced kinase activity of the EGFR in 70Z-Cbl serum-starved cells is tenfold higher in 70Z-Cbl transformed cells transformed cells than cells expressing c-Cbl or v-Cbl. An explanation for the high level of EGFR tyrosine A similar level of enhancement in EGFR kinase phosphorylation in 70Z-Cbl transformed cells is that activity in 70Z-Cbl transformed cells was maintained the kinase activity of the receptor is enhanced. This when the cells were stimulated with 0.1 or 1.0 ng/ml of 70Z-Cbl enhancement of EGF receptor kinase activity CBF Thien and WY Langdon 2912 70Z-Cbl, but not c-Cbl or v-Cbl (Figure 4a, b and c). Thus a consequence of 70Z-Cbl expression is the association of Grb2, Shc and Sos1 with the EGFR under conditions of minimal growth factor stimulation where these interactions are not normally evident. Following EGF stimulation there is recruitment of Grb2, Shc and Sos1 to the phosphorylated EGFR in all four cell lines, and as was shown by EGFR immunoprecipitation (Figure 2a), there is no obvious variation between cell lines in the amount of tyrosine phosphorylated EGFR that coimmunoprecipitates with these proteins (Figure 4a, b and c). A feature of the Grb2, Shc and Sos1 immunoprecipitations from EGF-stimulated cells was the reduced level of tyrosine phosphorylated Shc in v-Cbl transformed cells (Figure 4). This raised the possibility that v-Cbl may have an inhibitory eect on the tyrosine phosphorylation of substrates downstream of the EGFR. Figure 3 The expression of oncogenic 70Z-Cbl enhances EGFR kinase activity in serum-starved and EGF-stimulated cells. NIH3T3 cells expressing the human EGFR alone. or with HA- Opposing eects of 70Z-Cbl and v-Cbl on EGF-induced tagged c-Cbl, 70Z-Cbl or v-Cbl constructs were grown in media containing 0.5% FCS for 24 h. Cells were left quiescent (EGF:0) tyrosine phosphorylation of SHP-2 or were stimulated with 0.1 or 1.0 ng/ml of EGF for 2 min at To examine the eects of c-Cbl, 70Z-Cbl and v-Cbl 378C before lysis and immunoprecipitation with anti-EGFR antibodies. The immunoprecipitates were washed four times in expression on protein tyrosine phosphorylation we lysis buer and 5% of the protein A Sepharose beads were immunoblotted total cell lysates with anti-phosphotyr- incubated with 25 ml of a kinase reaction mix containing 0.5 mCi osine antibodies (Figure 5a), and anti-Shc antibodies to 32 of (g- P)ATP and 2.5 mM of biotinylated substrate #2 (Promega con®rm equal loading (Figure 5b). The results in SignaTECT PTK Assay System) for 15 min at 308C. The reaction Figure 5a show the spectrum of tyrosine phosphory- was terminated with 12.5 ml of 7.5 M guanidine hydrochloride and 30% of the reaction was added to streptavidin ®lters which were lated proteins in serum-starved and EGF-stimulated washed with 2 M NaCl, 1% H3PO4 to remove unincorporated cells. In serum-starved cells expressing 70Z-Cbl and v- ATP. The radioactivity of individual ®lters was measured by Cbl there is an increase in tyrosine phosphorylated liquid scintillation counting. The results shown above are the proteins at *75 kD and 52 kD. The 52 kD protein is means from three experiments likely to be p52 Shc since this protein was found to be more highly phosphorylated in these cells, although more so with 70Z-Cbl (see Figure 4b). The identity of the 75 kD protein is not known. The enhanced EGF. This high level of enhancement after EGF phosphorylation of these proteins in v-Cbl trans- stimulation was unexpected since results in Figure 2b formed cells is surprising since the kinase activity of showed no marked dierences in EGFR tyrosine the EGFR is not elevated suggesting that the phosphorylation between c-Cbl and 70Z-Cbl cells mechanisms mediating these eects dier between the stimulated with 0.1 and 1.0 ng/ml of EGF. Thus it two oncogenic forms of Cbl. It is also noteworthy that appears that EGFR phosphorylation in these cells had the phosphorylated EGFR and 70Z-Cbl can be clearly reached a plateau and that the enhancement in kinase identi®ed in total cell lysates from serum-starved 70Z- activity was only evident in the presence of excess Cbl transformed cells. substrate. These ®ndings therefore demonstrate that Following stimulation with EGF there is a the expression of 70Z-Cbl can markedly enhance the quantitative and qualitative increase in tyrosine enzymatic activity of the EGFR. phosphorylation in all four cell lines. However in v- Cbl transformed cells this eect is markedly suppressed and this is evident for a range of proteins from Recruitment of Grb2, Shc and Sos1 to the EGFR in *150 kD to p46 Shc. An exception appears to be the serum-starved cells transformed by 70Z-Cbl EGFR which, as we showed in Figure 2a, is The phosphorylation of tyrosine residues in the C- phosphorylated to a similar level in cells expressing c- terminal region of the EGFR provides binding sites for Cbl or 70Z-Cbl. Proteins in v-Cbl transformed cells the Grb2 and Shc adaptor proteins and this mediates with reduced tyrosine phosphorylation are evident in the recruitment of the guanine nucleotide exchange the region of the 68 kD molecular weight marker factor Sos1 which can then promote the activation of raising the possibility that this may involve the protein Ras (Egan et al., 1993; Rozakis-Adcock et al., 1993; Li tyrosine phosphatase SHP-2. It has been demonstrated et al., 1993; Buday and Downward, 1993a). It was that the SH2 domains of SHP-2 bind to the activated therefore important to investigate whether these EGFR and this leads to its phosphorylation on proteins are associated with the tyrosine phosphory- tyrosine and stimulation of its catalytic activity lated EGFR in serum-starved cells expressing 70Z-Cbl. (Lechleider et al., 1993; Vogel et al., 1993). SHP-2 The immunoprecipitation of Grb2, Shc and Sos1 and tyrosine phosphorylation was therefore examined in immunoblotting with anti-phosphotyrosine antibodies quiescent cells and cells stimulated with either 1 ng/ml revealed that all three proteins are associated with the or 50 ng/ml of EGF. Anti-SHP-2 immunoprecipitation phosphorylated EGFR in quiescent cells expressing and anti-phosphotyrosine immunoblotting revealed 70Z-Cbl enhancement of EGF receptor kinase activity CBF Thien and WY Langdon 2913

a b c - c-Cbl 70Z-Cbl v-Cbl - c-Cbl 70Z-Cbl v-Cbl - c-Cbl 70Z-Cbl v-Cbl - c-Cbl 70Z-Cbl v-Cbl - c-Cbl 70Z-Cbl v-Cbl EGF: – – – – + + + + EGF: – – – – + - + c-Cbl + 70Z-Cbl + v-Cbl EGF: – – – – + + + + 199 – EGFR 202– 202– – EGFR – EGFR 106 – Cbl – Cbl – Cbl 103– 103– 69 – p52 44 – p46 68– 68– I.P. : Anti-Grb2 – p52 – p46 – p52 Blot : Anti-P-Tyr 44– – p46 44– 199 I.P. : Anti-Shc – EGFR Blot : Anti-P-Tyr I.P. : Anti-Sos1 Blot : Anti-P-Tyr I.P. : Anti-Grb2 Blot : Anti-EGFR – p66 – p52 Ig 202 – Grb 2 – p46 – Sos I.P. : Anti-Grb2 I.P. : Anti-Shc I.P. :Anti-Sos 1 Blot : Anti-Grb2 Blot : Anti-Shc Blot : Anti-Sos 1 Figure 4 The expression of 70Z-Cbl promotes the recruitment of Grb2, Shc and Sos1 to the EGFR in serum-starved cells. NIH3T3 cells expressing the human EGFR and HA-tagged c-Cbl, 70Z-Cbl or v-Cbl constructs were grown in media containing 0.5% FCS for 24 h. Cells were left quiescent (EGF7) or were stimulated with 1 ng/ml of EGF for 2 min at 378C (EGF+) before lysis and immunoprecipitation with anti-Grb2 antibodies (a), anti-Shc antibodies (b) or anti-Sos1 antibodies (c). The immunoprecipitates were analysed by immunoblotting with the antibodies indicated above

a SHP-2 tyrosine phosphorylation was only detectable after EGF stimulation, although a number of SHP-2 associated proteins did show enhanced tyrosine - C-Cbl 70Z-Cbl v-Cbl - C-Cbl 70Z-Cbl v-Cbl phosphorylation in quiescent 70Z-Cbl transformed EGF : – – – – + + + + cells (data not shown). Following EGF stimulation the anti-phosphotyrosine immunoblot revealed a 202 – – EGFR number of features that demonstrate the contrasting eects of 70Z-Cbl and v-Cbl (Figure 6). Figure 6a and b shows exposure times of 10 and 20 min respectively where the shorter exposure clearly reveals the enhanced 103 – level of SHP-2 tyrosine phosphorylation in 70Z-Cbl transformed cells. An 80 kD protein that co-immunoprecipitated with SHP-2 is also highly phosphorylated 68 – as a result of 70Z-Cbl expression and there is a slight increase in the level of phosphorylated EGFR – p52 associated with SHP-2. In contrast, Figure 6b shows – p46 that SHP-2 tyrosine phosphorylation is not detectable 44 – in v-Cbl cells stimulated with 1 ng/ml of EGF, and is Total Lysates markedly suppressed even after stimulation with 50 ng/ Blot :Anti-P-Tyr ml of EGF. The v-Cbl transformed cells also show a b marked reduction in the tyrosine phosphorylation of proteins that associated with SHP-2. These proteins – p66 include the EGFR, a smeared protein or proteins at 110 ± 130 kD and p46 Shc. Immunoblotting with anti- – p52 HA antibodies did not detect Cbl protein in the 110 ± – p46 130 kD smear associated with SHP-2, a ®nding consistent with the fact that tyrosine phosphorylation Total Lysates at 120 kD in the SHP-2 immunoprecipitates is not Blot :Anti-Shc enhanced in cells overexpressing c-Cbl or 70Z-Cbl. Figure 5 Suppressed tyrosine phosphorylation of downstream The reduction in phosphorylated EGFR associating substrates following EGF stimulation of cells expressing v-Cbl. with SHP-2 in v-Cbl transformed cells contrasts with NIH3T3 cells expressing the human EGFR and HA-tagged c-Cbl, the ®ndings of Grb2 and Shc where associations with 70Z-Cbl or v-Cbl constructs were grown in media containing 0.5% FCS for 24 h. Cells were left quiescent (EGF7) or were the EGFR appear unaected (Figure 4). These ®ndings stimulated with 1 ng/ml of EGF for 2 min at 378C (EGF+) are of particular interest since they raise the possibility before lysis. 50 mg of each cell lysate was analysed by that the v-Cbl PTB domain may compete with SHP-2 immunoblotting with anti-phosphotyrosine antibodies (a), or for the same binding site or sites on the EGFR (Case anti-Shc antibodies (b). The uniform signal of the anti-Shc immunoblot indicates equivalent loading of protein from each et al., 1994) and therefore block its access to the lysate sample receptor. 70Z-Cbl enhancement of EGF receptor kinase activity CBF Thien and WY Langdon 2914 70Z-Cbl does not constitutively activate MAP kinase cells expressing 70Z-Cbl suggested that a concomitant increase in MAP kinase activity may occur in 70Z-Cbl The increase in EGFR kinase activity and the cells. To test this possibility we analysed the MAP recruitment of Sos1 to the EGFR in serum-starved kinases ERK1 and ERK2 in lysates from serum- starved and EGF-stimulated cells by a mobility shift assay. The mobility shift assay in Figure 7 shows that the basal levels of p44 ERK1 and p42 ERK2 -- c-Cbl 70Z-Cbl v-Cbl -- c-Cbl 70Z-Cbl v-Cbl phosphorylation are equivalent in serum-starved cells expressing c-Cbl, 70Z-Cbl and v-Cbl. MAP kinase a EGF 1 ng/ml 50ng/ml activity was also analysed by the in vitro phosphoryla- 202 – tion of a GST-Elk-1 fusion protein at serine 383. MAP — EGFR kinase immunoprecipitated from serum-starved cell lines induced a low level of Elk-1 phosphorylation, and this phosphorylation was equivalent between cells expressing c-Cbl, 70Z-Cbl and v-Cbl (data not shown). 103 – Prolonged activation of MAPK has been shown to promote its nuclear localization but no increase in the levels of nuclear ERKs was observed in 70Z-Cbl cells (data not shown). These ®ndings demonstrate that the 68 – = SHP-2 increase in EGFR kinase activity in 70Z-Cbl transformed cells does not constitutively activate MAP kinase activity in serum-starved cells. — Shc Surprisingly EGF stimulation of these cells revealed — Shc that the activation of both ERK1 and ERK2 was 44 – markedly suppressed in 70Z-Cbl and v-Cbl trans- b formed cells (Figure 7a, EGF+). The activation of 202 – the ERK kinase MEK was also suppressed in these — EGFR cells as shown by immunoblotting with a phosphospeci®c MEK antibody (Figure 7b). These unexpected ®ndings are consistent with a recent report that found MAP kinase and MEK activities were suppressed in 103 – NIH3T3 ®broblasts transformed by v-Src, v-Ras and v-Raf (Stofega et al., 1997). Furthermore this study also found that ERK activity was not constitutively 68 – = SHP-2 activated in these transformed cells. From this study it was hypothesized that negative regulatory mechanisms

— Shc — Shc 44 – - c-Cbl 70Z-Cbl v-Cbl a - c-Cbl 70Z-Cbl v-Cbl EGF – + I.P. : Anti-SHP-2 46 – = p44ERK1 Blot : Anti-P-Tyr = p44ERK2 c Blot : Anti-ERK1/2 68 – = SHP-2 b 46 – – phosphoMEK I.P. : Anti-SHP-2 Blot : Anti-SHP-2 Blot : Anti-phosphoMEK Figure 6 SHP-2 tyrosine phosphorylation is enhanced in 70Z- c Cbl transformed cells but suppressed in v-Cbl transformed cells. 46 – NIH3T3 cells expressing the human EGFR and HA-tagged c-Cbl, – MEK 70Z-Cbl or v-Cbl constructs were grown in media containing 0.5% FCS for 24 h. Cells were stimulated with 1 ng/ml or 50 ng/ Blot : Anti-MEK ml of EGF for 2 min at 378C before lysis and immunoprecipita- Total Lysates tion with anti-SHP-2 antibodies. The immunoprecipitates were Figure 7 70Z-Cbl expression does not constitutively activate analysed by immunoblotting with anti-phosphotyrosine antibodies MAP kinases ERK1 and ERK2. (a) Analysis of MAP kinase (a) and (b) or anti-SHP-2 antibodies (c). The exposure time of activation by electrophoretic mobility shift in SDS-polyacrylamide panel A is 10 min and most clearly shows the enhanced tyrosine gels. NIH3T3 cells expressing the human EGFR and HA-tagged phosphorylation of SHP-2 in 70Z-Cbl cells following treatment c-Cbl, 70Z-Cbl or v-Cbl constructs were grown in media with 50 ng/ml of EGF and the reduced tyrosine phosphorylation containing 0.5% FCS for 24 h. Cells were left quiescent of SHP-2 associated proteins in v-Cbl transformed cells. The (EGF7) or were stimulated with 1 ng/ml of EGF for 2 min at exposure time in panel B is 20 min and reveals the enhanced 378C (EGF+) before lysis and immunoblotting of cytoplasmic tyrosine phosphorylation of SHP-2 in 70Z-Cbl transformed cells lysates with ERK antibodies. The activation of the two forms of stimulated with 1 ng/ml of EGF and the reduced tyrosine MAP kinase is indicated by the phosphorylation-mediated shift in phosphorylation of SHP-2 in v-Cbl transformed cells stimulated mobility of p42 (ERK2) and p44 (ERK1). (b) Analysis of MEK with 50 ng/ml of EGF. An indication of the relative enhancement activation by immunoblotting with a phospho-speci®c MEK and suppression SHP-2 phosphorylation can also be assessed by antibody. (c) Analysis of MEK protein levels by immunoblotting the width of the band in the anti-SHP-2 immunoblot in panel C with a polyclonal MEK antibody 70Z-Cbl enhancement of EGF receptor kinase activity CBF Thien and WY Langdon 2915 a we examined whether this association is aected by overexpression of c-Cbl, 70Z-Cbl or v-Cbl. This was tested by immunoprecipitation of lysates from serum-

- c-Cbl 70Z-Cbl v-Cbl starved and EGF-stimulated cells with anti-CrkII - c-Cbl 70Z-Cbl v-Cbl EGF : – – – – + + + + antibodies and immunoblotting with anti-phosphotyr- 1 2 3 4 5 6 7 8 osine antibodies. Figure 8a shows a low level of 202 – tyrosine phosphorylated protein at 120 kD in serum- starved cells overexpressing c-Cbl and a slightly higher – EGFR level of tyrosine phosphorylated p120 in 70Z-Cbl expressing cells (lanes 2 and 3). Although these – Cbl proteins could not be detected by immunoblotting 103 – with anti-Cbl antibodies (Figure 8b) it is likely that they represent low levels of c-Cbl and 70Z-Cbl protein I.P. : Anti-Crk since the 120 kD phosphoprotein is not evident in Blot : Anti-P-Tyr serum-starved cells expressing endogenous c-Cbl (lane 1). The slight dierence in electrophoretic mobility b characteristically observed between c-Cbl and 70Z-Cbl is additional evidence for the identity of these proteins. – Cbl The anti-phosphotyrosine blot in Figure 8a also shows that following EGF stimulation there is a large increase 103 – in the level of 120 kD and 170 kD phosphoproteins I.P. : Anti-Crk associated with CrkII in cells overexpressing c-Cbl and Blot : Anti-Cbl (R2) 70Z-Cbl. Immumoblotting with anti-Cbl antibodies and anti-EGFR antibodies showed these proteins to c 44 – be c-Cbl, 70Z-Cbl and the EGFR thus revealing that overexpression of Cbl greatly enhances the level of Crk =Crk protein that can associate with the EGFR (Figure 8b and data not shown). These experiments therefore con®rm the inducible association between Cbl and Crk I.P. : Anti-Crk Blot : Anti-Crk proteins following EGF stimulation and demonstrate that c-Cbl may function as an adaptor molecule that is Figure 8 Overexpression of c-Cbl and 70Z-Cbl enhances the recruitment of CrkII to the EGFR. NIH3T3 cells expressing the utilized by Crk for its recruitment to the EGFR. It is human EGFR and HA-tagged c-Cbl, 70Z-Cbl or v-Cbl constructs also interesting that the amount of endogenous c-Cbl were grown in media containing 0.5% FCS for 24 h. Cells were associated with Crk is reduced in v-Cbl transformed left quiescent (EGF7, lanes 1 ± 4) or were stimulated with 1 ng/ cells (Figure 8a, compare lanes 5 and 8). Since v-Cbl is ml of EGF for 2 min at 378C (EGF+, lanes 5 ± 6) before lysis not detectably associated with Crk, presumably and immunoprecipitation with anti-CrkII antibodies. The immunoprecipitates were immunoblotted with (a) anti-phospho- because it is lacking Y700 and Y774, this suggests tyrosine, (b) anti-Cbl or (c) anti-CrkII antibodies that v-Cbl may compete with c-Cbl for EGFR binding site(s) and therefore block the recruitment of Crk. have evolved to prevent the constitutive activation of Discussion these pathways, and that these negative feedback mechanisms can override a growth factor induced Although Cbl's role in cellular signalling has yet to be stimulus (Stofega et al., 1997). Thus cells transformed elucidated evidence has accumulated in recent years by activated forms of Cbl appear to have an equivalent that suggests Cbl may have multiple functions. The suppressive eect on MAP kinase as those oncogenes most convincing evidence in mammalian cells points that are direct participants of the Ras pathway. towards Cbl functioning as a modular docking protein that provides binding sites for numerous signalling proteins, and that this promotes their involvement with Recruitment of Crk to the EGF receptor is mediated by activated receptor-coupled tyrosine kinases. Some of Cbl the molecular aspects of these interactions have been A consistently observed eect of Cbl tyrosine determined, e.g. Cbl's binding to SH3 and SH2 phosphorylation in numerous cell types is the domains and 14-3-3 proteins but whether Cbl also prominent association that occurs between Cbl and in¯uences signalling by directly regulating proteins with Crk proteins (de Jong et al., 1995; Sattler et al., 1996; enzymatic activity remains to be determined. Interest- Andoniou et al., 1996; Reedquist et al., 1996; Buday et ingly two recent studies have provided the initital al., 1996; Khwaja et al., 1996; Smit et al., 1996; evidence that Cbl may regulate the activity of proteins Sawasdikosol et al., 1996; Ingham et al., 1996). This with which it interacts. Overexpression of c-Cbl in mast association is mediated via the SH2 domain of CrkII cells was found to suppress Syk kinase activity (Ota or CrkL and tyrosine phosphorylated residues in Cbl and Samelson, 1997) and treatment of cells with at 700 and 774 that conform to the consensus binding antisense Cbl can enhance the activation of the JAK- site of pYXXP (Andoniou et al., 1996). An induced STAT pathway (Ueno et al., 1997). These ®ndings are association between Cbl and Crk proteins has also consistent with genetic studies from C. elegans where been found to occur following EGF stimulation the Cbl homologue, Sli-1, has been identi®ed as a (Fukazawa et al., 1996; Khwaja et al., 1996) and here negative regulator of the Let-23 receptor tyrosine 70Z-Cbl enhancement of EGF receptor kinase activity CBF Thien and WY Langdon 2916 kinase. In an attempt to provide biochemical evidence Characterization of downstream eects of 70Z-Cbl and to explain the observations in C. elegans we have v-Cbl: relevance to Cbl function and transformation examined the eects of c-Cbl and oncogenic forms of Cbl on the activity of the EGFR. Here we show that A consequence of 70Z-Cbl enhancement of EGFR while the overexpression of c-Cbl and v-Cbl do not kinase activity is that it results in receptor autopho- appear to directly aect receptor activity the expression sphorylation under conditions of minimal growth of 70Z-Cbl results in a signi®cant increase in the factor stimulation (Figure 2) which in turn promotes tyrosine kinase activity of the EGFR. This study recruitment of Grb2, Shc and Sos1 to the EGFR therefore provides a novel example of an intracellular (Figure 4). Since this event is not normally observed protein enhancing the kinase activity of the EGFR. in serum-starved cells, and these steps are involved in An important issue to emerge from this ®nding is to Ras activation, it is feasible that a downstream eect determine the mechanism that enables 70Z-Cbl to of 70Z-Cbl expression is the prolonged activation of achieve this eect. 70Z-Cbl lacks 17 amino acids at the Ras resulting in cellular transformation. Indeed the amino terminus of the Ring ®nger domain which is recent ®nding in Jurkat T cells that showed the centrally located in the Cbl protein (Figure 1a). This activation of NFAT by 70Z-Cbl expression was mutation was initially observed to induce three eects; abrogated by a dominant negative form of Ras has (i) it converted Cbl to an oncogenic protein; (ii) it provided convincing evidence that 70Z-Cbl acts greatly enhanced Cbl tyrosine phosphorylation under upstream of the Ras pathway (Liu et al., 1997). conditions of minimal growth factor stimulation and Furthermore, if the function of 70Z-Cbl does re¯ect a (iii) the electrophoretic mobility of 70Z-Cbl was found deregulation of normal Cbl function then Cbl may to be retarded in SDS-polyacrylamide gels (Andoniou determine the strength or duration of signals initiated et al., 1994; Bowtell and Langdon, 1995). More from tyrosine kinases and therefore the degree of Ras recently the expression of 70Z-Cbl in Jurkat T cells activation. was found to induce an increase in the basal activity of Additional downstream eects of the enhanced NFAT and this activation was abrogated by coexpres- kinase activity of the EGFR were increases in the sion of a dominant negative form of Ras (Liu et al., tyrosine phosphorylation of Shc (Figure 4b), SHP-2 1997). These ®ndings have clearly demonstrated the (Figure 6) and a protein of *75 kD (Figure 5). profound eects of this form of Cbl on signal Whether these eects are involved in 70Z-Cbl transduction, and here we have provided evidence to transformation, and whether this activity and tyrosine suggest that they are initiated by the ability of 70Z-Cbl phosphorylation are normally regulated by c-Cbl, to enhance the activation of a protein tyrosine kinase. remains to be resolved. These points are of interest in It will therefore be important to determine whether this answering the broader question of whether Cbl's eect requires a direct association between 70Z-Cbl putative role as a regulator of tyrosine kinases also and the EGFR (or T lymphoid tyrosine kinases) and involves a degree of substrate speci®city. why 70Z-Cbl, but not overexpressed c-Cbl, can mediate In contrast to 70Z-Cbl we ®nd no evidence of enhanced tyrosine activity. A clue to answering these enhanced EGFR tyrosine kinase activity in v-Cbl questions may come from the retarded electrophoretic transformed cells (Figure 3). Indeed the most striking mobility of 70Z-Cbl which suggests the protein is eect of v-Cbl expression is that following EGF structurally altered. A structural alteration could stimulation there is a reduction in the tyrosine enhance the accessibility of 70Z-Cbl to tyrosine phosphorylation of a range of proteins, notably p46 kinases either by the exposure of its PTB domain or and p52 Shc, SHP-2 and a number of SHP-2 associated by increasing the binding capacity of an adaptor proteins (Figures 4 and 6). This eect occurs without a protein. Indeed in a recent examination of interactions reduction in EGFR tyrosine phosphorylation or kinase between the EGFR and Cbl's PTB domain we found activity and suggests that v-Cbl may compete with evidence that this domain is masked in c-Cbl but can SHP-2, and possibly Shc, for binding sites on the be utilized following the deletion of sequences EGFR and thus perturb their tyrosine phosphoryla- encompassing the Ring ®nger domain (Thien and tion. These observations are interesting in view of Langdon, 1997). It is also possible that tyrosine recent ®ndings that demonstrate v-Cbl binding to the phosphorylation of c-Cbl may induce the transient EGFR correlates with an ability to transform since unmasking of the PTB domain and allow an both functions are abrogated by an amino acid interaction between amino-terminal sequences of Cbl substitution equivalent to a loss-of-function mutation and the EGFR. We hypothesize that the PTB domain from Sli-1 (Yoon et al., 1995; Thien and Langdon, is constitutively accessible in 70Z-Cbl and that this 1997). Therefore v-Cbl transformation may involve the promotes an interaction between the amino-terminal competitive inhibition of EGFR binding proteins. region of Cbl and the EGFR kinase domain. The Characterization of the v-Cbl binding sites on the potential signi®cance of the amino-terminal region of EGFR and other receptor complexes should allow us Cbl has been evident since the isolation of Sli-1 (Yoon to determine whether these sites are also recognized by et al., 1995), indicating that this highly conserved 400 the SH2 domains of SHP-2 and provide direction for amino acid region performs an important function in future studies to investigate SHP-2 involvement in 70Z- controlling the growth and fate of cells. However Cbl and v-Cbl transformation. further studies will be required to elucidate the In view of the increase in recruitment to the EGFR mechanism by which 70Z-Cbl mediates its eect on of proteins that regulate Ras in serum-starved cells EGFR tyrosine kinase activity. It is also feasible that transformed by 70Z-Cbl it was surprising that we the eect is indirect or involves the protection of key found no evidence of enhanced activation of the MAP receptor autophosphorylation sites that allow the kinases ERK1 and ERK2 (Figure 7). Indeed our assays kinase domain to remain active. clearly demonstrate ERK activation in response to 70Z-Cbl enhancement of EGF receptor kinase activity CBF Thien and WY Langdon 2917 EGF stimulation is suppressed in 70Z-Cbl and v-Cbl studies will therefore be aimed at de®ning the role of transformed cells which may be a consequence of a this interaction and will be part of wider studies to negative feedback mechanism that suppresses the Ras dissect the biochemical questions arising from the novel pathway. This result is consistent with a recent study eect of oncogenic 70Z-Cbl on the kinase activity of that investigated ERK activation in NIH3T3 fibro- the EGFR. blasts tranformed by oncogenic forms of Src, Ras and Raf (Stofega et al., 1997). These ®ndings suggest that dierences exist between the rapid stimulatory signal induced by EGF and the prolonged signalling mediated Materials and methods by constitutively active Cbl, Src, Ras and Raf. EGF stimulation is known to induce a rapid activation of Cell lines and viruses MAP kinases which reaches a peak within 2 ± 5 min NIH3T3 ®broblasts were obtained from ATCC and and returns to a basal level by 1 h (Buday and cultured in DMEM supplemented with 10% fetal calf Downward, 1993b) whereas sustained activation is serum (FCS) and 2 mM L-glutamine. NhEGFR is a stable associated with translocation of MAP kinases to the cell line derived from NIH3T3 ®broblasts and established nucleus (Traverse et al., 1992, 1994; Nguyen et al., by infection with helper-free stocks of the pBABE 1993; Dikic et al., 1994). In this study however we did retroviral vector containing the human EGFR cDNA and not detect an increase in nuclear ERKs in 70Z-Cbl selected in 2 mg/ml puromycin (Sigma). HA-tagged Cbl transformed cells (data not shown). constructs were stably expressed in NhEGFR cells using the pJZenNeo retroviral vector and selected in 400 mg/ml geneticin (G418, Gibco BRL0) as previously described Cbl mediated-recruitment of Crk to the EGFR (Blake et al., 1993). A recent study has shown that the oncogenic growth signal from the EGFR to Ras is predominantly Antibodies mediated by CrkII in rat ®broblasts (Kizaka-Kondoh Polyclonal rabbit anti-HA and anti-Cbl (R2) antibodies et al., 1996), and since EGF stimulation induces an have been described previously (Bowtell and Langdon, association between Cbl and CrkII (Fukazawa et al., 1995; Blake et al., 1993). Monoclonal anti-HA antibody 1996; Khwaja et al., 1996) we investigated whether was isolated from 12CA5 culture supernatants and the CrkII is involved in Cbl transformation. Findings in monoclonal anti-phosphotyrosine antibody 4G10 was this study demonstrate that overexpression of both c- kindly provided by Dr Brian Druker. Sheep anti-human Cbl and 70Z-Cbl greatly enhances the recruitment of EGFR antibodies and rabbit anti-human Shc antibodies were purchased from UBI. Rabbit anti-Grb2, Sos1 and CrkII to the activated EGFR (Figure 8a, lanes 6 and 7) SHP-2 antibodies, and the monoclonal anti-EGFR anti- indicating that increasing CrkII association with the body 528 were purchased from Santa Cruz. Monoclonal EGFR via Cbl does not induce cellular transformation anti-CrkII antibody was purchased from Transduction or enhance EGFR kinase activity. However we cannot Laboratories, phospho-speci®c MEK1/2 from New Eng- discount that CrkII is required for 70Z-Cbl transfor- land Biolabs and polyclonal rabbit anti-MEK was a gift mation and studies utilizing 70Z-Cbl constructs with from Dr P Tilbrook. altered CrkII SH2 binding sites should determine this possibility. An interesting observation from this study EGF stimulation was the clear distinction between the pool of EGFR associating with Cbl and CrkII and that associating Cells were grown to 90% con¯uency and made quiescent by culturing in DMEM containing 0.5% FCS for 24 h with Grb2, Shc and Sos1 (compare the amount of prior to EGF stimulation. Cells were either left quiescent tyrosine phosphorylated Cbl in the CrkII immunopre- or stimulated with a range of EGF concentrations for cipitates in Figure 8a with the amount of tyrosine 2 min at 378C and lysates prepared immediately. phosphorylated Cbl in the Grb2, Shc and Sos1 immunoprecipitates in Figure 4). It will therefore be Immunoprecipitations and immunoblotting important to compare the relative kinase activity of these pools of EGFR and to determine whether 70Z- Cells were washed in Tris-buered saline (TBS; 150 mM Cbl has a greater or lesser eect on their enhancement. NaCl, 20 mM Tris pH 7.4) and then lysed in 1 ml of ice- Cbl's induced association with Crk proteins has also cold TBS containing 1% Triton X-100, 1 mM EDTA, 1 mM sodium orthovanadate (Na V0 ), 10 mM NaF, 10 mg/ml created interest because of the role of Crk in recruiting 3 4 aprotinin and 1 mg/ml each of chymostatin, leupeptin, the nucleotide exchange factor C3G (Tanaka et al., antipain and pepstatin. Cell debris and nuclei were 1994; Matsuda et al., 1994) and therefore the activation removed by centrifugation at 6000 r.p.m. in a microfuge of the Rap1A and Rap1B G-proteins (Gotoh et al., for 5 min at 48C. The protein concentrations of the cleared 1995). The putative role of Rap proteins as negative lysates were measured using the Bradford Bio-Rad protein regulators of Ras (Hariharan et al., 1991; Cook et al., assay and adjusted to 1 mg/ml of protein before incubation 1993) has obvious implications for providing a with primary antibodies for 3 h at 48C followed by 2 h biochemical explanation for Sli-1's role as a negative with protein-A Sepharose (Pharmacia). Immune complexes regulator of receptor tyrosine kinases. Furthermore it were washed 4 times in lysis buer before suspension and has recently been shown that v-Crk transformation of boiling in Laemmli sample buer. Immune complexes were rat ®broblasts is augmented by the overexpression of c- electrophoretically separated on 8% or 10% SDS-polyacrylamide gels and transferred to nitrocellulose mem- Src and that transformation involves an increase in the branes. Bound proteins were detected by immunoblotting kinase activity of c-Src (Sabe et al., 1992, 1995). In and primary antibodies were visualized using horse-radish view of these ®ndings and the observations in this peroxidase conjugated antibodies and ECL reagents report it is likely that Cbl and Crk work in concert to (Amersham) as previously described (Andoniou et al., regulate the activity of tyrosine kinases. Ongoing 1994) 70Z-Cbl enhancement of EGF receptor kinase activity CBF Thien and WY Langdon 2918 In vitro protein kinase assay 12 cm in length until the 44 kD marker was 2.5 cm from the bottom. The proteins were transferred to a nitrocellu- The EGFR was immunoprecipitated from lysates of serum- lose membrane and immunoblotted with a rabbit anti- starved or EGF-stimulated cells cultured in 100 mm tissue MAPK antibody which recognizes ERK1 and ERK2 culture dishes using the procedures described above. The (Santa Cruz, SC-94). MAP kinase activity was also immunoprecipitates were washed four times in lysis buer measured by the in vitro phosphorylation of the MAP and 5% of the beads were incubated with 25 mlofkinase kinase substrate Elk-1 at serine 383. A phospho-speci®c reaction mix for 15 min at 308C. The reaction mix antibody to MAP kinase (New England Biolabs) was used consisted of 8 mM b-glycerophosphate, 8 mM imidazole to immunoprecipitate active MAP kinase from cell lysates hydrochloride pH 7.3, 0.2 mM EGTA, 20 mM MgCl ,1mM 2 which was incubated with 1 mg of GST-Elk-1 fusion MnCl ,0.1mg/mlBSA,20mMATP, 0.5 mCi of (g-32P)ATP 2 protein (New England Biolabs) at 308Cfor30mininthe and 2.5 mM of biotinylated substrate #2 (Promega presence of kinase buer and ATP. Phosphorylation of SignaTECT PTK Assay System). The reaction was Elk-1 was then measured by immunoblotting using a terminated with 12.5 mlof7.5M guanidine hydrochloride phospho-speci®c Elk-1 (serine 383) antibody (New Eng- and 30% of the reaction was added to streptavidin ®lters land Biolabs). The activation of the ERK kinase MEK was which were washed with 2 M NaCl, 1% H PO to remove 3 4 analysed by immunoblotting with a phosphospeci®c MEK unincorporated ATP. The radioactivity of individual ®lters antibody (New England BioLabs). was measured by liquid scintillation counting.

MAP kinase assays Activation of p44 MAPK(ERK1) and p42 MAPK(ERK2) Acknowledgements was examined by a mobility shift assay to detect the We thank Dr Ed Nice and Dr Brian Druker for the decrease in electrophoretic mobility which occurs after generous provision of EGF and 4G10 antibodies. This phosphorylation of tyrosine and threonine residues. study was supported by grants from NH and MRC Lysates from serum-starved and EGF-stimulated cells (Canberra), The Cancer Foundation of Western Australia were separated on an 8.5% SDS-polyacrylamide gel of and the Raine Medical Research Foundation.

References

Andoniou CE, Thien CBF and Langdon WY. (1994). EMBO Hariharan IK, Carthew RW and Rubin GM. (1991). Cell, 67, J., 13, 4515 ± 4523. 717 ± 722. Andoniou CE, Thien CBF and Langdon WY. (1996). Ingham RJ, Krebs DL, Barbazuk SM, Turck CW, Hirai H, Oncogene, 12, 1981 ± 1989. Matsuda M and Gold MR. (1996). J. Biol. Chem., 271, Blake TJ, Heath KG and Langdon WY. (1993). EMBO J., 32306 ± 32314. 12, 2017 ± 2026. Jongeward GD, Clandinin TR and Sternberg PW. (1995). Blake TJ, Shapiro M, Morse HC III and Langdon WY. Genetics, 139, 1553 ± 1566. (1991). Oncogene, 6, 653 ± 657. Khwaja A, Hallberg B, Warne PH and Downward J. (1996). Bowtell DDL and Langdon WY. (1995). Oncogene, 11, Oncogene, 12, 2491 ± 2498. 1561 ± 1567. Kim TJ, Kim Y-T and Pillai S. (1995). J. Biol. Chem., 270, Buday L and Downward J. (1993a). Cell, 73, 611 ± 620. 27504 ± 27509. Buday L and Downward J. (1993b). Mol. Cell. Biol., 13, Kizaka-Kondoh S, Matsuda M and Okayama H. (1996). 1903 ± 1910. Proc.Natl.Acad.Sci.USA,93, 12177 ± 12182. BudayL,KhwajaA,SipekiS,FaragoAandDownwardJ. Kontani K, Kukimoto I, Nishina H, Hoshino S, Hazeki O, (1996). J. Biol. Chem., 271, 6159 ± 6163. Kanaho Y and Katada T. (1996). J. Biol. Chem., 271, CaseRD,PiccioneE,WolfG,BenettAM,LechleiderRJ, 1534 ± 1537. Neel BG and Sholelson SE. (1994). J. Biol. Chem., 269, Langdon WY, Hartley JW, Klinken SP, Ruscetti SK and 10467 ± 10474. Morse HC III. (1989). Proc. Natl. Acad. Sci. USA, 86, Cook SJ, Rubinfeld B, Albert I and McCormick F. (1993). 1168 ± 1172. EMBO J., 12, 3475 ± 3485. Lechleider RJ, Sugimoto S, Bennett AM, Kashishian AS, Cory GOC, Lovering RC, Hinshelwood S, MacCarthy- Cooper JA, Shoelson SE, Walsh CT and Neel BG. (1993). Morrogh L, Levinski RJ and Kinnon C. (1995). J. Exp. J. Biol. Chem., 268, 21478 ± 21481. Med., 182, 611 ± 615. Li N, Batzer A, Daly R, Yajnik V, Skolnik E, Chardin P, deJongR,tenHoeveJ,HeisterkampNandGroenJ. Bar-Sagi D, Margolis B and Schlessinger J. (1993). Nature, (1995). J. Biol. Chem., 270, 21468 ± 21471. 363, 85 ± 88. Dikic I, Schlessinger J and Lax I. (1994). Curr. Biol., 4, 702 ± Liu Y-C, Elly C, Langdon WY and Altman A. (1997). J. Biol. 708. Chem., 272, 168 ± 173. Donovan JA, Wange RL, Langdon WY and Samelson LE. Liu Y-C, Elly C, Yoshida H, Bonnefoy-Berard N and (1994). J. Biol. Chem., 269, 22921 ± 22924. Altman A. (1996). J. Biol. Chem., 271, 14591 ± 14595. Egan SE, Giddings BW, Brooks MW, Buday L, Sizeland AM Lupher ML Jr, Reedquist KA, Miyake S, Langdon WY and and Weinberg RA. (1993). Nature, 363, 45 ± 51. Band H. (1996). J. Biol. Chem., 271, 24063 ± 24068. Fournel M, Davidson D, Weil R and Veillette A. (1996). J. Marcilla A, Rivero-Lezcano OM, Agarwal A and Robbins Exp. Med., 183, 301 ± 306. KC. (1995). J. Biol. Chem., 270, 9115 ± 9120. Fukazawa T, Miyake S, Band V and Band H. (1996). J. Biol. Matsuda M, Hashimoto Y, Muroya K, Hasegawa H, Kurata Chem., 271, 14554 ± 14559. T, Tanaka S, Nakamura S and Hattori S. (1994). Mol. Galisteo ML, Dikic I, Batzer AG, Langdon WY and Cell. Biol., 14, 5495 ± 5500. Schlessinger J. (1995). J. Biol. Chem., 270, 20241 ± 20245. Meisner H, Conway BR, Hartley D and Czech MP. (1995). Gotoh T, Hattori S, Nakamura S, Kitayama H, Noda M, Mol. Cell. Biol., 15, 3571 ± 3578. Takai Y, Kaibuchi K, Matsui H, Hatase O, Takahashi H, Nguyen TT, Scimeca J-C, Filloux C, Peraldi P, Carpentier J- Kurata T and Matsuda M. (1995). Mol. Cell. Biol., 15, L and van Obberghen E. (1993). J. Biol. Chem., 268, 6746 ± 6753. 9803 ± 9810. 70Z-Cbl enhancement of EGF receptor kinase activity CBF Thien and WY Langdon 2919 Odai H, Sasaki K, Iwamatsu A, Hanazono Y, Tanaka T, Solto SP and Cantely LC. (1996). J. Biol. Chem., 271, 563 ± Mitani K, Yazaki Y and Hirai H. (1995). J. Biol. Chem., 567. 270, 10800 ± 10806. Stofega MR, Yu C-L and Jove R. (1997). Cell Growth Ojaniemi M, Martin SS, Dol® F, Olefsky JM and Vuori K. Dieren., 8, 113 ± 117. (1997). J. Biol. Chem., 272 3780 ± 3787. Tanaka S, Amling M, Ne L, Peyman A, Uhlmann E, Levy Ota Y, Beitz LO, Scharenberg AM, Donovan JA, Kinet J-P JB and Baron R. (1996). Nature, 383, 528 ± 531. and Samelson LE. (1996). J. Exp. Med., 184, 1713 ± 1723. Tanaka S, Morishita T, Hashimoto Y, Hattori S, Nakamura Ota Y and Samelson LE. (1997). Science, 276, 418 ± 420. S, Shibuya M, Matuoka K, Takenawa T, Kurata T, Panchamoorthy G, Fukazawa T, Miyake S, Solto S, Nagashima K and Matsuda M. (1994). Proc. Natl. Acad. Reedquist K, Druker B, Shoelson S, Cantley L and Band Sci. USA, 91, 3443 ± 3447. H. (1996). J. Biol. Chem., 271, 3187 ± 3194. Tanaka S, Ne L, Baron R and Levy J. (1995). J. Biol. Reedquist KA, Fukazawa T, Panchamoorthy G, Langdon Chem., 270, 14347 ± 14351. WY, Shoelson SE, Druker BJ and Band H. (1996). J. Biol. Thien CBF and Langdon WY. (1997). Oncogene, 14, 2239 ± Chem., 271, 8435 ± 8451. 2249. Ribon V, Hubbell S, Herrera R and Saltiel AR. (1996). Mol. Traverse S, Gomez N, Paterson H, Marshall C and Cohen P. Cell. Biol., 16, 45 ± 52. (1992). Biochem. J., 288, 351 ± 355. Rivero-Lezcano OM, Sameshima JH, Marcilla A and Traverse S, Seedorf K, Paterson H, Marshall CJ, Cohen P Robbins KC. (1994). J. Biol. Chem., 269, 17363 ± 17366. and Ullrich A. (1994). Curr. Biol., 4, 694 ± 701. Rozakis-Adcock M, Fernley R, Wade J, Pawson T and Tsygankov AY, Mahajan S, Fincke JE and Bolen JB. (1996). Bowtell D. (1993). Nature, 363, 83 ± 85. J. Biol. Chem., 271, Sabe H, Okada M, Nakagawa H and Hanafusa H. (1992). Ueno H, Sasaki K, Miyagawa K, Honda H, Mitani K, Mol. Cell. Biol., 12, 4706 ± 4713. Yazaki Y and Hirai H. (1997). J. Biol. Chem., 272, 8739 ± Sabe H, Shoelson SE and Hanafusa H. (1995). J. Biol. 8743. Chem., 270, 31219 ± 31224. Vogel W, Lammers R, Huang J and Ullrich A. (1993). Sattler M, Salgia R, Okuda K, Uemura N, Durstin MA, Science, 259, 1611 ± 1614. Pisick E, Xu G, Li J-L, Prasad KV and Grin JD. (1996). Wang Y, Yeung YG, Langdon WY and Stanley ER. (1996). Oncogene, 12, 839 ± 846. J. Biol. Chem., 271, 17 ± 20. Sawasdikosol S, Chang J-H, Pratt JC, Wolf G, Shoelson SE Yoon CH, Lee J, Joneward GD and Sternberg PW. (1995). and Burako SJ. (1996). J. Immunol., 157, 110 ± 116. Science, 269, 1102 ± 1105. Schaefer E and Hsaio K. (1996). Promega Notes, 59, 2±9. Smit L, van der Horst G and Borst J. (1996). J. Biol. Chem., 271, 8564 ± 8569.