Differential Utilization of ShcA Tyrosine Residues and Functional Domains in the Transduction of Epidermal Growth Factor-induced Mitogen-activated Protein Kinase Activation in 293T Cells and Nerve Growth Factor-induced Neurite Outgrowth in PC12 Cells

By transient expression of both truncated forms of p52 SHCA and those with point mutations in 293T cells, it has been shown that, in addition to Tyr-317, Tyr-239/240 is a major site of phosphorylation that serves as a docking site for Grb2·Sos1 complexes. In addition, analysis of epidermal growth factor (EGF)-induced activation of mitogen-activated protein kinase in 293T cells showed that the overexpression Shc SH2 or phosphotyrosine binding (PTB) domains of ShcA alone has a more potent negative effect than the overexpression of the forms of ShcA lacking Tyr-317 or Tyr 239/240 or both. In transiently transfected PC12 cells, the ShcA PTB domain and tyrosine phosphorylation in the CH1 domain, especially on Tyr-239/240, are crucial for mediating nerve growth factor (NGF)-induced neurite outgrowth. These findings suggest that the EGF and NGF (TrkA) receptor can utilize Shc in different ways to promote their activity. For EGF-induced mitogen-activated protein kinase activation in 293T cells, both Shc PTB and SH2 domains are essential for optimal activation, indicating that a mechanism independent of Grb2 engagement with Shc may exist. For NGF-induced neurite outgrowth in PC12 cells, Shc PTB plays an essential role, and phosphorylation on Tyr-239/240, but not on Tyr-317, is required.

By transient expression of both truncated forms of p52 SHCA and those with point mutations in 293T cells, it has been shown that, in addition to Tyr-317, Tyr-239/240 is a major site of phosphorylation that serves as a docking site for Grb2⅐Sos1 complexes. In addition, analysis of epidermal growth factor (EGF)-induced activation of mitogen-activated protein kinase in 293T cells showed that the overexpression Shc SH2 or phosphotyrosine binding (PTB) domains of ShcA alone has a more potent negative effect than the overexpression of the forms of ShcA lacking Tyr-317 or Tyr 239/240 or both. In transiently transfected PC12 cells, the ShcA PTB domain and tyrosine phosphorylation in the CH1 domain, especially on Tyr-239/240, are crucial for mediating nerve growth factor (NGF)-induced neurite outgrowth. These findings suggest that the EGF and NGF (TrkA) receptor can utilize Shc in different ways to promote their activity. For EGF-induced mitogen-activated protein kinase activation in 293T cells, both Shc PTB and SH2 domains are essential for optimal activation, indicating that a mechanism independent of Grb2 engagement with Shc may exist. For NGF-induced neurite outgrowth in PC12 cells, Shc PTB plays an essential role, and phosphorylation on Tyr-239/240, but not on Tyr-317, is required.
Activation of the Ras/MAP 1 kinase pathway, triggering proliferation and differentiation, is one of the major early events induced by receptor tyrosine kinases (reviewed in Ref. 1). Grb2 and Sos1 have been linked to the mediation of this activation (2,3). Grb2 is a 23-kDa adaptor protein composed of an SH2type phosphotyrosine recognition domain, flanked by two SH3 domains that recognize proline-rich sequences (reviewed in Ref. 4). Sos1 is a 170-kDa protein with a guanine nucleotide exchanging domain that can catalyze the conversion of inactive Ras, with bound GDP, into its active form, with bound GTP (reviewed in Ref. 5). Beside its catalytic domain, Sos1 contains a proline-rich region enabling it to couple with Grb2 (6). Activation of Ras can be achieved when cytosolic Sos translocates to the plasma membrane, where cellular Ras normally resides (7). This translocation is mediated by the recruitment of Grb2 onto the intracellular domain of an activated receptor tyrosine kinase, such as the epidermal growth factor receptor (EGFR) (8). Grb2 can dock through its SH2 domain with specific phosphotyrosine residues of the activated EGFR which subsequently allows relocalization of Sos1 to the plasma membrane level (reviewed in Ref. 9). Activated Ras allows a cascade of events starting with the activation of the serine kinase Raf that phosphorylates and activates a MAP kinase kinase (also denoted MEK), which in turn activates MAP kinase by dual phosphorylation on tyrosine and threonine residues (reviewed in Ref. 1).
Although the EGFR can utilize Grb2 directly to mediate MAP kinase activation, it can also recruit Grb2 and Sos1 indirectly, which involves the Shc family of adaptor proteins (10,11). The first member identified and the prototype of this family of molecules, ShcA, is composed of a phosphotyrosine binding (PTB) domain at its N terminus and one C-terminal SH2 domain. These domains are separated by a collagen homologue (CH1) domain (12,13). ShcA is present in two major forms of 46 and 52 kDa and a less abundant one of 66 kDa. p46 shc and p52 SHCA derive from alternative transcription initiation (12), whereas p66 shc arises from alternative splicing of the messenger, resulting in the N-terminal addition of a second CH domain (CH2) (14). EGFR transphosphorylation of tyrosine residues in the intracellular domain provides specific docking sites for both Shc PTB and SH2 domains (15)(16)(17). ShcA recruitment to the activated receptor is followed by Shc phosphorylation of tyrosine (18,19) creating a recognition site for the Grb2 SH2 domain and inducing the formation of a Shc⅐Grb2⅐Sos1 complex leading to MAP kinase activation (2, 18, 20 -22).
In PC12 cells, MAP kinase activation is necessary (but possibly not sufficient) for NGF-induced neurite outgrowth (reviewed in Refs. 23 and 24). ShcA binds to the activated NGF receptor, TrkA, through its PTB domain (25) and is in turn phosphorylated on tyrosine allowing recruitment of Grb2 and Sos1 (reviewed in Ref. 13). It has been shown that a tyrosine to phenylalanine mutation of the Shc binding site on TrkA dramatically decreased MAP kinase activation and suppresses NGF-induced neurite outgrowth in PC12 cells (26,27).
Recently, several other adaptor proteins, displaying the same architecture and a high level of identity in their PTB and SH2 domains with those of Shc, were identified (28 -30). Interestingly, the sequence alignment of the Shc proteins revealed the conservation of tyrosines in positions 239 and 240 in the CH1 domain (of ShcA). This suggests that these residues could have an important function in the recruitment of SH2 containing molecules and the starting point of a new signaling pathway.
In the present report, mutated and truncated forms of p52 SHCA have been expressed in 293T (human kidney) and PC12 cells and their effects on Shc⅐Grb2⅐Sos complex formation, MAP kinase activation, and NGF-induced neurite outgrowth in PC12 cells determined.

MATERIALS AND METHODS
Chemical and Reagents-Mouse 2.5 S NGF and EGF were purified according the methods of Mobley et al. (31) and Savage and Cohen (32).
GST-SHC Construct and Transient Transfection-The cDNA of the human 52-kDa form of ShcA was obtained by reverse transcriptasepolymerase chain reaction as described previously (34). Full-length p52 SHCA cDNA and its truncated forms were initially cloned into the pGEX-3X vector (Pharmacia Biotech Inc.). Point mutations in the sequences encoding the GST-fusion proteins were made using the USE system (Pharmacia) and appropriate mutagenesis primers. The sequences of the oligonucleotides used for the substitution of Tyr-317 and Tyr-239/240 to phenylalanine is CTGGACGTTGACAAAGGAGGGATC-CAC and CCGGGAAGTCATTAAACAACTGATGGTCAGG, respectively. Mutated and truncated sequence were amplified by polymerase chain reaction and cloned into the EcoRI site of the pCMV-1 expression vector (35). The structure and mutations of the constructs used are summarized in Fig. 1.
Thirty percent confluent 293T cells in 100-mm dishes were transfected with 10 g of DNA using the calcium phosphate co-precipitation method (36). The precipitate was incubated for 12 h, and cells were allowed to grow in fresh complete medium overnight. Cells were then starved in serum-free medium for the next 24 h and stimulated with 50 ng of EGF per ml for the indicated period. PC12 cells were grown at low density on poly-L-lysine-coated glass coverslips in 6-well plates and transfected with 2.5 g of DNA using Lipofectin reagent (Life Technologies, Inc.) following the manufacturer's recommendations. After a 5-h incubation, cells were allowed to recover for 24 h in complete medium and then treated with 100 ng of NGF per ml in Dulbecco's modified Eagle's medium containing 1% horse serum for 48 h.
Immune Complex Kinase Assay-293T cells plated in 100-mm dishes transfected and treated as above were stimulated for 5 min with 50 ng of EGF per ml. Cells were harvested in ice-cold phosphate-buffered saline (PBS), pelleted at 800 ϫ g, and lysed for 20 min in 1 ml of Triton X-100 lysis buffer (1% w/v), 50 mM Tris-HCl (pH 7.5), 100 mM NaCl, 50 mM NaF, 5 mM EDTA, 40 mM ␤-glycerophosphate, 200 mM sodium orthovanadate with 1 mM phenylmethylsulfonyl fluoride, 20 g/ml aprotinin, and 20 g/ml leupeptin. Lysates were centrifuged at 13,000 ϫ g for 15 min at 4°C. Protein concentration was determined, and 1.5 mg of protein (1-ml final volume) was incubated with gentle rocking at 4°C for 3 h with 10 l of agarose-conjugated anti-Erk1 polyclonal antibody (Santa Cruz Biotechnology). Agarose beads were washed 3 times in lysis buffer and once in kinase buffer (20 mM Hepes (pH 7.4), 10 mM MgCl 2 , 1 mM dithiothreitol, 10 mM p-nitrophenyl phosphate) and 20 l of kinase buffer containing 2 mg/ml myelin basic protein (MBP) (Sigma) as substrate, 50 M ATP, and 5 Ci of [␥-32 P]ATP. Kinase reactions were performed for 10 min at 30°C and stopped by adding 40 l of 2 ϫ Laemmli sample buffer. Samples were boiled for 5 min and separated on a 12% SDS-PAGE. Gels were stained with Coomassie Blue, and bands were excised from the gel, and radioactivity was counted by liquid scintillation.
Immunocytochemistry-Following 48 h of NGF stimulation, PC12 cells were washed in PBS and fixed in 3% paraformaldehyde (in PBS) for 10 min at room temperature. After 2 washes in PBS, cells were permeabilized and blocked for 1 h at room temperature in PBS, 0.1% Triton X-100, 1% bovine serum albumin (PBS/Triton X-100/bovine serum albumin) and incubated with anti-GST polyclonal antibody (Santa Cruz Biotechnology) overnight at 4°C. Cells were washed 3 times in PBS/Triton X-100/bovine serum albumin, and incubated with fluorescein isothiocyanate-conjugated anti-rabbit IgG (Molecular Probes) for 1 h at room temperature. After 3 washes in PBS/Triton X-100/bovine serum albumin and once in PBS alone, coverslips were mounted in ProLong anti-fading medium (Molecular Probes) and observed with an epifluorescence-equipped microscope. Immunolabeled PC12 cells were scored and cellular extensions having at least twice the length of the cell body were counted as neurites. Fig. 2b, GST-SHC (see Fig. 1 for description of SHC constructs) expressed in 293T cells stimulated with EGF became robustly phosphorylated on tyrosine residues and was able to complex with Grb2 ( Fig. 2c) and Sos1 (Fig. 2d) as early as 1 min following EGF stimulation. A major tyrosinephosphorylated band of 180 kDa was also found associated with GST-SHC and was identified immunologically as being the EGFR receptor (solid triangle in Fig. 2b) (data not shown), demonstrating that the construct binds the EGFR upon ligand stimulation. When EGF stimulation was prolonged for 10 and 30 min, the GST-SHC tyrosine phosphorylation level began to decrease. The amount of Grb2 found in the GST-SHC precipitates also decreased (Fig. 2c), and an even more significant decrease in the amount of Sos1 was observed (Fig. 2d).

Mutational Analysis of SHC EGF-induced Tyrosine Phosphorylation and SHC⅐Grb2⅐Sos1 Complex Formation-As shown in the immunoblot in
GST-SHCY317F (in which Tyr-317 has been substituted by a phenylalanine) was also clearly phosphorylated on tyrosine residue(s) following EGF stimulation and bound the activated EGF receptor, as shown in the anti-phosphotyrosine immuno- blot (Fig. 2b). As with GST-SHC, Grb2 was found associated with GST-SHCY317F in a growth factor stimulation-dependent manner (Fig. 2c). The level of Grb2 associated with tyrosine-phosphorylated GST-SHCY317F did not vary for the different time points of EGF stimulation. More importantly, although present in a lower amount than when GST-SHC was expressed, Sos1 was also detected in GST-SHCY317F precipitates (Fig. 2d), and the level of Sos1 present with GST-SHCY317F was significantly decreased after 10 and 30 min EGF treatment.
When GST-SHC protein, in which Tyr-239 and -240 had been changed to phenylalanine residues (GST-SHCY239/240F), was expressed, the anti-phosphotyrosine immunoblot, depicted in Fig. 2b, showed that it was only weakly phosphorylated on tyrosine following EGF stimulation and bound to the activated EGFR. Furthermore, Grb2 was also detected in the GST-SHCY239/240F precipitate in lower amounts (Fig. 2c), and the presence of Sos1 protein was detectable only after long exposures of the immunoblot with the enhanced chemiluminescence detection system (data not shown). When the triple mutant, SHCY239/240/317F was expressed, neither tyrosine phosphorylation of the fusion protein and Grb2 (Fig. 2c) nor Sos1 association (Fig. 2d) could be detected. This construct, however, normally associates with the activated EGFR (Fig. 2b).
GST-SH2 and GST-N constructs were also expressed in 293T cells. In contrast to the Shc full-length constructs, GST-SH2 association to the activated EGF receptor was never superior to the background level observed for GST alone. On the other hand, GST-N clearly bound strongly following EGF stimulation. It was also phosphorylated on tyrosine (Fig. 2b).
Beside their ability to form complexes with Grb2⅐Sos1 and to bind to the activated EGF receptor, the GST-SHC constructs, when expressed in 293T cells, also associated with other phosphotyrosine-containing proteins. Unidentified 65-and 120-kDa phosphotyrosine-containing proteins were complexed with GST-SH2 in a growth factor-independent manner, since both proteins were found in resting cells as well as in EGF-stimulated cells (open triangles in Fig. 2b). A phosphotyrosine-containing protein co-migrating with p65 was also found associated with all the GST-SHC constructs in a growth factorindependent manner. The 120-kDa phosphotyrosine protein was not bound to the Shc and SHCY317F constructs, but it was found with SHCY239/240F and SHCY239/240/317F in resting cells; it disappeared in EGF-stimulated cells. GST-N was also bound to a 140-kDa phosphotyrosine-containing protein in resting cells (upper solid diamond in Fig. 2b). Similarly, a phosphotyrosine-containing protein, co-migrating with the 140-kDa protein, was also found with all full-length Shc constructs that bound in a growth factor-independent manner.
A few other phosphotyrosine-containing proteins were found specifically associated with the full-length GST-SHC constructs, suggesting that their interaction with Shc takes place on the CH1 domain. This appears to be the case for a 60-kDa protein detectable in GST-SHC and GST-SHCY317F precipitates from resting cells (lower open diamond in Fig. 2b). Interestingly, upon EGF stimulation, the amount of tyrosine-phosphorylated p60 was dramatically decreased both in GST-SHC and in GST-SHCY317F constructs. However, the level of the phosphotyrosine-containing 60-kDa protein was not altered upon EGF stimulation in GST-SHC precipitates when Tyr-239 and -240 or when Tyr-239, -240, and -317 were substituted by phenylalanine.
SHCA PTB and SH2 Domains Are Necessary for Optimal EGF-induced MAP Kinase Activation in 293T Cells-To determine which domain and tyrosine residue of p52 SHCA was essential for the transduction of MAP kinase activation upon EGF stimulation, 293T cells transiently expressing the Shc constructs ( Fig. 1) were stimulated with EGF; MAP kinases were immunoprecipitated, and their kinase activity was assayed on MBP. Three independent transfections and sets of kinase assays were performed; the results are shown in Fig. 3. Overexpression of GST-SHC in EGF-treated 293T cells was without significant enhancing effect compared with GST alone. When the SHC SH2 and N-terminal domains were expressed alone, a negative effect on EGF-induced MAP kinase activation that could reach, depending on the experiment, 50% of the control "fold activation" was observed. On the other hand, the overexpression of the mutated forms of SHC, SHCY317F, and SHCY239/240F, were only able to slightly down-regulate EGFinduced MAP kinase activation in the three different experiments. In each case, overexpression of GST-SHCY239/240/ 317F did not alter EGF-induced MAP kinase activation.
ShcA PTB Domain and Tyr-239/240 Are Required for NGFinduced Neurite Outgrowth in PC12 Cells-To determine which tyrosine residues and domains of Shc are important for TrkA signaling, PC12 cells transiently expressing the various GST-SHC constructs and stimulated for 48 h with NGF to induce differentiation were examined. The effects of the expression of GST-SHC proteins on NGF-induced PC12 cell differentiation were assayed by immunocytochemistry and neurite scoring. As shown in Fig. 4, approximately 50% of the PC12 cells expressing GST alone and treated with NGF for 48 h grew neurites (see Fig. 4 and Fig. 5a). This value was significantly increased to 80% when PC12 cells expressed the full-length and wild type GST-SHCA ( Fig. 4 and Fig. 5b). In contrast GST-N, which lacks the CH1 and SH2 domains, allowed only 20% of the cells that expressed the construct to respond to NGF (Fig. 4 and Fig. 5c). GST-SH2 expression was apparently without effect on the number of PC12 cells responsive to NGF in comparison to GST alone; however, a negative effect on neurite branching in cells expressing SH2 domain alone was noticeable (Fig. 5d). GST-SHCY317F expressed in PC12 cells had a significant enhancing effect (70%) on PC12 cell differentiation, and branched neurites were well developed (Fig. 5e). On the other hand, GST-SHCY239/240F expression had no apparent enhancing effect on NGF-induced neurite outgrowth. Transient expression of GST-SHCY239/240/317F in NGF-treated PC12 cells had a significant negative effect allowing only 12% of the cells to grow neurites ( Fig. 4 and Fig. 5f).

DISCUSSION
Overexpression of Shc has been shown to induce neoplastic transformation in fibroblasts and cause differentiation on PC12 cells, and these two phenomena depend on the coupling of Grb2 to Shc Tyr-317 and Ras activation, respectively (18,19). Early studies on Shc have also indicated that Tyr-317 was the major site of tyrosine phosphorylation responsible for Grb2 association (18,19). From the data presented herein, it is clear that Shc can also undergo growth factor-induced phosphorylation on other tyrosine residues. Indeed, mutation of this residue (Tyr-317) still allows Shc to retain its capability to be phosphorylated on a tyrosine residue(s) and to form a complex with Grb2 in 293T cells following EGF stimulation, and comparable substitutions demonstrate that Tyr-239/240 can be phosphorylated in vivo and can act as a Grb2 binding site. Although SHCY239/240/317F contains additional tyrosine residues in conserved positions in other Shc isoforms, especially in the N-terminal portion (29), no residual tyrosine phosphorylation or associated Grb2 was detected in the triple mutant following EGF stimulation. This observation suggests that Tyr-239/240 and Tyr-317 likely account for all SHC tyrosine phosphorylation and Grb2 binding sites in vivo. However, it is interesting to note that the truncated SHC construct GST-N (see Fig. 1) undergoes tyrosine phosphorylation upon EGF stimulation. This is likely an artifact arising from the elimination of steric constraints and enables tyrosine kinases to phos-

FIG. 3. Effect of the expression of truncated and mutated GST-SHC fusion proteins on EGF-induced MAP kinase activity in 293T cells.
The cells were transiently transfected with 10 g of pCMV-1 expression vector containing the cDNA of the truncated and mutated GST-SHC fusion proteins. Cells were starved for 24 h in serum-free medium and stimulated for 5 min with 50 ng/ml EGF. Cells were lysed in 1% Triton X-100 buffer, and MAP kinases were immunoprecipitated with anti-Erk1 agarose-conjugated polyclonal antibodies. MAP kinase activity was assayed using MBP as a substrate in the presence of 50 M ATP and 5 Ci of [␥-32 P]ATP. Samples were separated by SDS-PAGE; gels were stained with Coomassie Blue, and the bands corresponding to MBP were excised and quantitated by scintillation counting. The fold activation of MAP kinase was determined by comparing the level of MBP kinase activity found in unstimulated and EGF-stimulated cells transfected with the same construct. The three histograms represent the fold activations obtained in three independent experiments.

FIG. 4. Effect of the expression of truncated and mutated GST-SHC fusion proteins on NGF-induced neurite outgrowth in PC12 cells. PC12 cells plated on poly-L-lysine-coated glass coverslips
were transfected with the truncated and mutated forms of p52 SHCA cloned into the pCMV-1 expression vector using Lipofectin reagent. PC12 cells were treated for 48 h with 100 ng/ml NGF and processed for detection of the GST antigen, as described under "Materials and Methods." The histogram represents the mean of the percentages of immunolabeled cells responsive to NGF obtained in two independent experiments. The obtained percentages of NGF-responsive cells for the two experiments were 40 and 60% for GST, 75 and 85% for GST-SHC, 20 and 23% for GST-N, 50 and 55% for GST-SH2, 60 and 76% for GST-SHCY317F, 39 and 66% for GST-SHCY239/240F, and 6 and 18% for GST-SHCY239/240/317F. phorylate tyrosine residue(s) otherwise unavailable, probably in the most C-terminal part of GST-N. Importantly, tyrosinephosphorylated GST-N does not form a complex with Grb2, which supports the view that this N-terminal phosphorylation is not physiologically important.
Grb2 SH2 domain interactions have been shown to be specified by asparagine in the position ϩ2 to the phosphorylated tyrosine (37). Therefore, Tyr-239 appears to be an good candidate for an alternate specific Grb2 SH2 domain binding site, since it occurs in a Tyr(P)-Xaa-Asn-Xaa sequence, a context similar to Tyr-317. As a consequence, phosphorylation on Tyr-240 is less likely to be involved in Grb2 SH2 recognition. A more detailed point-mutation analysis will be required to determine whether Tyr-240 can be phosphorylated. If so, it would probably be in a sequence recognized by SH2 domain-containing molecules other than Grb2. It is possible, of course, that a steric competition could occur between Grb2 and one or another SH2 domain-containing molecules, which remain to be identified. Recently, two other studies have reported similar results concerning the identification of Tyr-239/240 as a major phosphorylation site on Shc (38,39).
Of major importance is the finding that SHCY317F and SHCY239/240F can form complexes with Grb2, whereas substitution of Tyr-239, -240, and 317 by phenylalanine totally abolishes SHC⅐Grb2 complex formation. This is consistent with results obtained previously, establishing that phosphorylation on Tyr-239/240 and Tyr-317 is responsible for Grb2 engagement by Shc (18,19,39). More importantly, the data presented herein clearly show that Sos1 is also present with Grb2 in GST-SHCY317F precipitates following EGF stimulation of 293T cells, indicating that the phosphorylation on Shc Tyr-239/ 240 is in part responsible for the formation of Shc⅐Grb2⅐Sos1 complexes. In addition, and in agreement with previous observations (40) showing that EGFR signaling induces a destabilization of the Shc⅐Grb2⅐Sos1 complex by MAP kinase phosphorylation of Sos1 on serine (41), we observed a similar situation for both GST-SHC and GST-SHCY317F constructs in EGFstimulated 293T.
Based on the observation that the level of the Shc⅐Grb2⅐Sos1 complex is optimal with wild type Shc and significantly lower with SHCY317F, it follows that phosphorylation on Tyr-317 alone in SHCY239/240F should also engage Sos1, thus accounting for this difference. Interestingly, it was observed that, although GST-SHCY239/240F binds to the activated EGFR, tyrosine phosphorylation, following EGF stimulation, was weak in 293T cells (see Fig. 2b). Consistent with this weak phosphorylation, a lower level of Grb2 and very little Sos1 were found associated with SHCY239/240F. This observation suggests that, in EGF-stimulated 293T cells, phosphorylation of Shc Tyr-317 may depend to some degree on the phosphorylation of Tyr-239/240 and/or on the proteins that are recruited to Tyr(P)-239/240 (and the signaling pathway(s) they transduce). It also suggests that the kinase domain of the activated EGF receptor may not be responsible for Shc phosphorylation on Tyr-317, since although being very weakly phosphorylated following EGF stimulation, the GST-SHCY239/240F forms a complex with the activated EGFR. Previous studies have reported that Shc can undergo tyrosine phosphorylation in the absence of specific binding sites on the EGFR (42,43) and that phosphorylation on Tyr-317 does not depend on Shc binding to the activated EGFR (44). In addition, it has been shown that in v-Src or v-Sea transformed cells, Shc is phosphorylated and associates with Grb2 (45)(46)(47). This suggests that in vivo, growth factor-induced Shc tyrosine phosphorylation can be achieved directly by the activated growth factor receptor kinase domain and that Shc tyrosine residues can be phosphorylated by activated non-receptor tyrosine kinases, such as the Src family. Further studies will be necessary to identify the kinases that are responsible in vivo for Shc Tyr-239/240 and Tyr-317 phosphorylations and to understand how the phosphorylation on Tyr-239/240 can regulate the phosphorylation of Tyr-317. Nevertheless, the data reported here suggest that two Grb2⅐Sos1 complexes, instead of one, can be recruited at the same time on a single Shc molecule.
Having demonstrated that Grb2 and Sos1 can complex with Shc on an additional site, it was of interest to examine the function in vivo of the different phosphorylatable sites in different cellular milieus. In 293T cells, MAP kinase activation assays showed that overexpression of the SH2 and N terminus domains of ShcA has a negative effect on EGF-induced MAP kinase activation, consistent with recent studies that have used microinjection of various Shc domains fused to GST to point out the importance of the Shc SH2 domain for EGFRinduced DNA synthesis and mitogenesis (44,48,49). Our data clearly indicate that the function of the Shc N terminus that contains the PTB domain is to couple Shc to the activated EGFR. In addition, they indicate that in vivo Shc SH2 interaction with the activated EGFR does not occur to a significant degree, in comparison to what was observed with the Shc PTB domain. Therefore, the negative effect of the overexpression in 293T cells of the GST-SH2 construct on the EGF-induced MAP kinase activation cannot be explained by a participation of this domain in the coupling of Shc to the EGFR. Interestingly, the Shc SH2 domain associates with a 65-kDa tyrosine-phosphorylated protein in a growth factor-independent manner, suggesting that this protein and the Shc SH2 domain are constitutively associated in vivo. This suggests that Shc SH2 domain function may be less in the targeting of Shc to the EGF receptor in vivo than in the coupling of effectors that contribute to the activation of MAP kinase. The 65-kDa protein observed here behaves differently since its phosphorylation on tyrosine and its association to Shc SH2 domain are apparently not regulated by growth factor stimulation. Isolation and characterization of this p65 is currently in progress.
As Shc⅐Grb2⅐Sos1 complex formation mediates Ras activation (2, 18, 20 -22), it would be expected that overexpression of SHCY239/240/317F in 293T cells would have a negative effect on EGF-induced MAP kinase activation, since this triple mutant is unable to complex with Grb2. Surprisingly, overexpression of the triple mutant did not have a negative effect and even enhanced EGF-induced MAP kinase activation in 293T cells. These unexpected results are likely due to other structural features in Shc PTB and SH2 domains, which are consistent with the fact that overexpression of the GST-SHCY239/240/ 317F construct rescues the cells unable to form the Shc⅐Grb2⅐ Sos1 complex. Furthermore, although unphosphorylated in the triple mutant, Shc CH1 domain may conserve some important functions in the mediation of the EGFR signaling. The CH1 domain contains proline-rich sequence motifs (12) that are responsible for the coupling of Shc with SH3 domain-containing signaling molecules such as the Esp8 protein (50) and tyrosine kinases such as Src, Lyn, and Fyn (51). Therefore, it seems likely that Shc can initiate other pathways than the one leading to Ras activation. The observations reported herein suggest that one of these pathways also can lead to MAP kinase activation.
Our data also clearly indicate that Shc PTB domain is essential for the transduction of NGF signals through TrkA in the differentiation of PC12 cells since overexpression of the GST-N protein has a dramatic negative effect. This result is consistent with previous studies reporting that ShcA binds to Tyr-490 on TrkA (26). However, and contrary to what was observed in 293T cells and the EGFR, Shc SH2 domain overexpression did not have any apparent effect on the number of NGF-responsive cells, suggesting that Shc SH2 domain function is not as important as the Shc PTB domain in this activity. Nonetheless, that Shc SH2 domain does affect this process is indicated by the observation that PC12 cells overexpressing this domain alone grew shorter neurites with poor branching. Interestingly, we were unable to find the p65 protein associated with the Shc SH2 domain in PC12 cells (data not shown). Instead, a 118-kDa phosphotyrosine-containing protein was observed (data not shown) that may correspond to the p115 protein previously described by Dikic et al. (25).
Overexpression of GST-SHC in PC12 cells has a clear enhancing effect on NGF neurotrophic activity in agreement with previous observations (19). Substitution of Tyr-239/240/317 by phenylalanine does have a similar negative effect as seen with GST-N. Thus, Shc mediation of NGF neurotrophic activity in PC12 cells appears to strictly depend on the formation of a Shc⅐Grb2⅐Sos1 complex in agreement with previous reports showing that PC12 cell differentiation induced by Shc overexpression depends on the formation of Shc⅐Grb2⅐Sos1 complexes and Ras activity (10,18).
Substitution of Tyr-317 by phenylalanine appears to have a very limited negative effect on the ability of overexpressed Shc to enhance neurite outgrowth. In contrast, substitution of Tyr-239/240 by phenylalanine dramatically impairs the Shc enhancing effect. These results are in agreement with the data obtained in 293T and indicate that, in PC12 cells, Tyr(P)-239/ 240 also is a potent site for the recruitment of Grb2⅐Sos1 complexes. They confirm that in the absence of phosphorylation of these tyrosine residues, phosphorylation on Tyr-317 is weak and Shc⅐Grb2⅐Sos1 complexes are formed in only very limited amounts. As a consequence, the data suggest that in PC12 cells stimulated with NGF, Shc undergoes phosphorylation on Tyr-239/240/317 but that the Tyr-239/240 makes a more important contribution to the activation of the Ras/MAP kinase pathway.
Altogether, the data reported in the present study indicate for the first time that the way growth factor receptors utilize different Shc tyrosine phosphorylatable residues and modular domains to mediate their biological activities depends on the cellular context. In 293T cells, EGF-induced MAP kinase activity depends on the presence of Shc but not strictly on the formation of Shc⅐Grb2⅐Sos1 complexes. On the other hand, in PC12 cells, NGF neurotrophic activity, for which MAP kinase activity is fundamental, strictly depends on the formation of the Shc⅐Grb2⅐Sos1 complexes and less on Shc SH2 domain. It will now be interesting to determine the role of the different Shc tyrosine residues and modular domains in the EGF-dependent stimulation of MAP kinase activity in PC12 cells overexpressing EGFR.