Skip to main content
NCBI home page
Proceedings of the National Academy of Sciences of the United States of America logoLink to Proceedings of the National Academy of Sciences of the United States of America
. 2004 Feb 13;101(8):2345–2350. doi: 10.1073/pnas.0308065101

The Shc adaptor protein is critical for VEGF induction by Met/HGF and ErbB2 receptors and for early onset of tumor angiogenesis

Caroline Saucier *, Hanane Khoury *,†, Ka-Man Venus Lai ‡,§, Pascal Peschard *,†, David Dankort , Monica A Naujokas *, Jocelyn Holash ∥, George D Yancopoulos ∥, William J Muller *,†, Tony Pawson ‡,§, Morag Park *,†,**,††,‡‡
PMCID: PMC356953  PMID: 14983012

Abstract

The etiology and progression of a variety of human malignancies are linked to the deregulation of receptor tyrosine kinases (RTKs). To define the role of RTK-dependent signals in various oncogenic processes, we have previously engineered RTK oncoproteins that recruit either the Shc or Grb2 adaptor proteins. Although these RTK oncoproteins transform cells with similar efficiencies, fibroblasts expressing the Shc-binding RTK oncoproteins induced tumors with short latency (≈7 days), whereas cells expressing the Grb2-binding RTK oncoproteins induced tumors with delayed latency (≈24 days). The early onset of tumor formation correlated with the ability of cells expressing the Shc-binding RTK oncoproteins to produce vascular endothelial growth factor (VEGF) in culture and an angiogenic response in vivo. Consistent with this, treatment with a VEGF inhibitor, VEGF-Trap, blocked the in vivo angiogenic and tumorigenic properties of these cells. The importance of Shc recruitment to RTKs for the induction of VEGF was further demonstrated by using mutants of the Neu/ErbB2 RTK, where the Shc, but not Grb2, binding mutant induced VEGF. Moreover, the use of fibroblasts derived from ShcA-deficient mouse embryos, demonstrated that Shc was essential for the induction of VEGF by the Met/hepatocyte growth factor RTK oncoprotein and by serum-derived growth factors. Together, our findings identify Shc as a critical angiogenic switch for VEGF production downstream from the Met and ErbB2 RTKs.

Among the 58 members of the receptor tyrosine kinase (RTK) family identified to date, deregulation of at least 31 of them have been linked to various human malignancies (1). The mechanisms that lead to deregulation of RTKs may differ, but in all cases, the normally tightly regulated intracellular signaling of the RTK is perturbed (1). Deregulation of a receptor or physiological stimulation by ligand, promotes activation of the intracellular kinase and subsequent phosphorylation of the receptor on tyrosine residues, some of which act as binding sites for a variety of signaling proteins. These proteins contain Src homology 2 (SH2) or phosphotyrosine-binding domains that recognize phosphorylated tyrosine residues in the context of their surrounding amino acids (2). The combination of proteins recruited to RTKs dictates a series of downstream signals within the interior of the cell that culminate in distinct biological effects.

To discriminate the role of proximal-binding partners of RTKs and their downstream signaling pathways in various cancer cell behaviors, we have previously engineered RTK oncoproteins that recruit a signaling protein of choice (3). Using these tools, we have shown that the direct recruitment of the Grb2 or Shc adaptor proteins to a RTK oncoprotein is sufficient to induce similar parameters of cell transformation, including foci of morphologically transformed fibroblasts, anchorage-independent growth, and experimental metastasis (3).

Several studies have implicated the recruitment of the Grb2 or Shc adaptor proteins as important mediators of cell transformation downstream from RTKs (4–9). Grb2 and Shc associate with tyrosine phosphorylated RTKs through their respective SH2 and phosphotyrosine-binding domains (10–14). In addition, the recruitment of Shc to activated RTKs results in its phosphorylation on tyrosine residues Y239/240 and Y317, which provide optimal binding sites for the SH2 domain of Grb2 (15–18). In turn, Grb2 through protein–protein interactions involving its SH3 domains, links receptors with multiple downstream signaling proteins, such as activation of the Ras/ERK and PI3′K/AKT signaling pathways (10, 13, 16, 19–22). Although many RTKs can bind directly to Grb2, some RTKs rely on Shc to indirectly recruit Grb2 (23, 24). Thus, the association of these adaptor proteins to RTKs has often been shown to fulfill redundant biological functions. However, the Shc adaptor protein can recruit signaling molecules in addition to Grb2 (25). Although this has not been extensively examined, it suggests that the direct recruitment of Shc or Grb2 to RTKs could activate independent downstream signaling pathways and consequently regulate distinct biological functions.

Solid tumors will not expand beyond a size of ≈2 mm3 if new blood vessels from the preexisting host vasculature are not attracted to supply the oxygen and nutrients required to sustain their growth (26, 27). This process, defined as tumor angiogenesis, is the product of a complex series of interactions between the tumor and its host microenvironment. The tumor-surrounding stroma and infiltrating blood-derived cells are known sources of proangiogenic factors. However, in many cases, cancer cells themselves produce proangiogenic factors (26). Among many factors known to promote angiogenesis, vascular endothelial growth factor (VEGF) is one of the most potent pro-angiogenic factors that is frequently up-regulated in human tumors (26, 28). The production of VEGF by cancer cells can be triggered by limited oxygen, i.e., hypoxia, found within the microenvironment of solid tumors, but its expression is also regulated downstream from RTKs (29–39).

In the present study, we have compared the role of the direct recruitment of Grb2 or Shc to a RTK oncoprotein in tumorigenesis. We show that the recruitment of Shc, but not of Grb2, to a RTK is sufficient to augment the expression of VEGF. This finding correlates with a rapid tumor formation and a robust angiogenic response in vivo of fibroblasts expressing the Shc-binding RTK oncoproteins, which we show are dependent on VEGF. We used fibroblasts derived from ShcA-deficient mouse embryos (40) to demonstrate that Shc is essential for VEGF production induced by the Met/hepatocyte growth factor (HGF) RTK or serum-derived growth factors. These studies have unveiled a previously unsuspected role for the Shc adaptor protein in RTK-mediated VEGF production and tumor angiogenesis.

Materials and Methods

Antibodies. Antibody 144 was raised against a peptide in the C terminus of the Met protein (41). Antibodies for phosphotyrosine and Grb2 were purchased from Transduction Laboratories (Lexington, KY), and the Neu and VEGF antibody from Santa Cruz Biotechnology. The Shc antibody was provided by John Bergeron (McGill University).

DNA Constructs and Cell Lines. The cloning and characterization of the Tpr-Met, CSF-Met, the signal specific RTK oncoproteins, and activated Neu/ErbB2 add-back mutants have been reported (3, 4, 6, 42). The cloning of the ligand-activated Grb2- and Shc-binding RTKs were performed as described for the Tpr-Met variants (3), but using the Y1349/1356F CSF-Met receptor mutant as a recipient (42). All cells were cultured at 37°C in DMEM supplemented with 10% FBS. Expression of Tpr-Met in wild-type mouse embryo fibroblasts (MEF) or ShcA-deficient MEF cells (40) was obtained by cotransfection of Tpr-Met cDNA with pLXSH vector by using geneporter (Gene Therapy System). Colonies resistant to hygromycin (150 μg/ml) were picked and expanded into cell lines. Transfection of 293T and Rat-1 fibroblast cells were performed by calcium phosphate method.

Tumorigenesis and in Vivo Angiogenesis Assays. For tumorigenesis assay, fibroblasts (105 cells per 100 μl) were injected s.c. into 4- to 5-week-old female nude mice (CD1 nu/nu, Charles River Breeding Laboratories). Tumors were measured periodically and allowed to grow until they reached ≈1 cm3 or before ulceration, at which time the mice were killed and the tumors were collected. For the in vivo angiogenesis assay, 105 cells mixed with 250 μl of serum-depleted Matrigel (Becton Dickinson Labware, Bedford, MA) were injected s.c. into mice. The resulting Matrigel plugs were photographed and collected after 10 days (six Matrigel plugs per cell line). For VEGF-Trap treatment (Regeneron Pharmaceuticals, Tarrytown, NY; ref. 43), tumorigenesis and in vivo angiogenesis assays were performed as described above with the exception that 6- to 7-week-old nude mice were used. After implantation of the cells, vehicle (5 mM phosphate/5 mM citrate, pH 6/100 mM NaCl/20% glycerol/0.1% Tween) or 25 mg/kg VEGF-Trap was administrated to mice twice a week by s.c. injection at the nape of the neck for the duration of the experiments. Tumor specimens were fixed overnight in 3.7% formaldehyde at 4°C, and then embedded in paraffin and sectioned for hematoxylin and eosin staining.

Cell Lysate, Immunoprecipitation, in Vivo Association Assay, and Immunoblotting. Preparation of cell lysates, immunoprecipitations, in vitro association assays, or immunoblots were performed as described (4). Proteins were visualized by using enhanced chemiluminescence (ECL, Amersham Pharmacia).

Detection of VEGF Protein in Cell-Conditioned Media. Fibroblasts, seeded at a density of 1–2 × 106 cells per 100-mm culture dish, were, on the next day, incubated for 48 h in 4 ml of medium free of phenol red and serum. For stimulations, colony-stimulating factor (CSF; 100 ng/ml) was added to the medium. To detect VEGF protein, 1–1.4 ml of cleared conditioned media was incubated for 1 h at 4°C with 25 μl of 50% heparin–Sepharose. The heparin–sepharose protein complex was rinsed three times with buffer (20 mM Tris, pH 7.5/200 mM NaCl/1 mM DTTand proteinase inhibitors) and bound proteins were eluted with Laemmli sample buffer for immunoblot detection of VEGF.

Northern Blot Analysis. Total RNA was isolated from serum-starved (24 h) cells expressing variant or control proteins by the Trizol method. RNAs (40 μg) were resolved by electrophoresis in formaldehyde-containing agarose gels and transferred to a N+-Hybond filter (Amersham Pharmacia). The blots were hybridized with a 32P-labeled cDNA probe corresponding to the full-length VEGF transcript, kindly provided by Daniel J. Dumont (University of Toronto).

Results

Fibroblasts Expressing Shc-Binding RTK Oncoproteins Induce Tumors with Short Latency. To define the contribution of Grb2 and Shc recruitment to a RTK in diverse oncogenic processes, we have previously engineered RTK oncoproteins capable of directly recruiting only the Grb2 or Shc adaptor proteins (3). These were generated by using a mutant of the Met receptor oncoprotein, Tpr-Met, in which we have substituted the two C-terminal tyrosine residues (Y482, Y489) with nonphosphorylateable phenylalanine residues (Fig. 1A). These tyrosine residues provide a multisubstrate binding region required for the recruitment of all known signaling proteins to the Tpr-Met oncoprotein (3, 4, 7, 19). This mutant, which is catalytically active but nononcogenic (3, 4), was used as a recipient to substitute the multisubstrate binding region of the Met RTK oncoprotein (amino acids 486–494) with a cassette containing either a Grb2-binding site from the epidermal growth factor receptor (Y-Grb2), or Shc-binding sites from the TrkA (Y-Shc-1) or EGF (Y-Shc-2) receptors (Fig. 1 A and ref. 3).

Fig. 1.

Fig. 1.

Open in a new tab

RTK oncoproteins coupling to Shc confer on fibroblasts the ability to form tumors in nude mice with short latency. (A) Diagram of the RTK oncoproteins specific for recruitment of Grb2 or Shc. The amino acid sequences substituted within the Tpr-Met Y482/489F cassette mutant and the inserted binding motifs are shown. The Grb2-binding site from the epidermal growth factor receptor was inserted to generate the Y-Grb2 RTK oncoprotein, whereas the Y-Shc-1 and Y-Shc-2 RTK oncoproteins contain, respectively, the Shc-binding sites from the TrkA or EGF receptors (3). (B) Expression and phosphorylation levels of RTK oncoproteins and control proteins in fibroblast cell lines. Lysates (500 μg) of cells expressing RTK oncoproteins or control proteins were subjected to immunoprecipitation with an antibody specific for Met (Ab 144) and immunoblotted with the same antibody or anti-pTyr. (C) The growth of tumor (mm3) over time (day) was measured after s.c. injection of 105 cells (see Table 1). The results represent the mean tumor volume obtained from two independent experiments in which at least three mice were injected for each cell line.

We previously used these RTK oncoproteins to establish that the direct recruitment of Grb2 or Shc is sufficient to induce cell transformation, anchorage-independent growth, and experimental metastasis (3). However, the individual contribution of Grb2 or Shc signals in tumorigenesis was unknown. To discriminate this, we have established stable fibroblast cell lines expressing the Grb2 (Y-Grb2)- or Shc (Y-Shc-1 or Y-Shc-2)-binding RTK oncoproteins or control proteins where the inserted tyrosine residue was substituted with a nonphosphorylateable phenylalanine residue (Fig. 1 A, Y-Grb2 Y/F or Y-Shc-1 Y/F). As expected, because every signaling-specific RTK oncoprotein and control protein was derived from the constitutively activated recipient Tpr-Met mutant (Fig. 1 A, Y482/489F, and ref. 3), these were equally phosphorylated on tyrosine residues relative to their levels of expression in stable cell lines (Fig. 1B).

The tumorigenicity of fibroblast cell lines was evaluated after their s.c. injection into the flank of nude mice. Cells expressing controls for the Grb2 or Shc RTK oncoproteins (Y-Grb2 Y/For Y-Shc-1 Y/F) or Tpr-Met cassette mutant (Y482/489F) failed to develop tumors 90 days after inoculation (Table 1). In contrast, fibroblasts transformed with Grb2- or Shc-binding RTK oncoproteins grew as tumors, but with distinct latencies (Fig. 1C). Cells expressing the Shc RTK oncoproteins (Y-Shc-1 or Y-Shc-2) induced palpable tumors with a short latency of ≈7 days, whereas the appearance of tumors with cells expressing the Grb2 RTK oncoprotein was delayed to ≈24 days after their implantation (Fig. 1C and Table 1).

Table 1. In vivo tumorigenicity of Fr3T3 cells expressing signal-specific binding RTK oncoproteins.

Cell line No. of tumors/no. of injections Tumor latency, day
Y-Grb2 6/6 24 ± 3.1
Y-Grb2 (Y/F) 0/6 >90
Y-Shc-1 6/6 6.0 ± 1.1
Y-Shc-1 (Y/F) 0/6 >90
Y-Shc-2 6/6 7.3 ± 1.2
Y482/489F 0/6 >90
Open in a new tab

The tumorigenecity of cells expressing signal protein binding RTK oncoproteins was evaluated after their injection subcutaneously into nude mice.

Fibroblasts Expressing Shc-Binding RTK Oncoproteins Induce an Angiogenic Response and Produce VEGF. Because the transforming activity of the Grb2- and Shc-binding RTK oncoproteins in culture assays are similar (3), the difference observed in the latency of tumor formation suggested a host–tumor cell response critical for the expansion of solid neoplasms, such as the induction of angiogenesis (26, 27). To test this, and to circumvent problems associated with the distinct tumor latencies, an in vivo angiogenesis Matrigel plug assay was performed (44). Fibroblasts expressing the Grb2- or Shc-binding RTK oncoproteins, or control proteins, were mixed with a Matrigel solution depleted of growth factors and injected s.c. into nude mice, hence allowing the maintenance of cells within the Matrigel. After 10 days, we observed that Matrigel plugs of cells expressing the Shc-binding RTK oncoproteins (Y-Shc-1 or Y-Shc-2) were red and contained many blood vessels (Fig. 2A and Fig. 7, which is published as supporting information on the PNAS web site). In contrast, Matrigel plugs containing fibroblasts expressing the Grb2 RTK oncoprotein or control proteins (Tpr-Met Y482/489F, Y-Grb2 Y/F or Y-Shc-1 Y/F) remained clear and were poorly vascularized (Figs. 2A and 7 and data not shown).

Fig. 2.

Fig. 2.

Open in a new tab

Fibroblasts expressing Shc-binding RTK oncoproteins promote angiogenesis in vivo and enhance VEGF mRNA and protein. (A) Representative photographs of Matrigel plugs formed 10 days after injection of fibroblasts (105 cells) expressing the Grb2 or Shc RTK oncoproteins, or the Tpr-Met Y482/489F control mixed with 250 μl of growth factor-depleted Matrigel solution. (B) Cells were seeded at a density of 106 per 100-mm plate. The following day, medium was replaced with serum-free media, and CM was collected after 48 h. The level of VEGF was detected after enrichment with heparin by immunoblot analysis using a VEGF antibody. (C) The level of VEGF mRNA was detected by Northern blot analysis of total RNA (40 μg) isolated from serum-starved (24 h) cells expressing variant or control proteins.

The in vivo angiogenic properties of cells expressing the two independent Shc-binding RTK oncoproteins suggested that these cells, in contrast to cells expressing the Grb2-binding RTK, produced proangiogenic factors. VEGF is one of the most potent proangiogenic factors commonly up-regulated in human tumors (26, 28), and is induced by several RTKs (29–39). Therefore, the ability of cells expressing the Grb2- or Shc-binding RTK oncoproteins to produce VEGF protein in their conditioned media (CM) was examined. The VEGF protein (VEGF165, ≈23 kDa) was readily detected in the CM of cells expressing either of the Shc RTK oncoproteins (Fig. 2B, Y-Shc-1 and Y-Shc-2), whereas the level of VEGF protein produced by fibroblasts expressing the Grb2 RTK oncoprotein or controls was barely detectable or absent (Fig. 2B, Y-Grb2, Y-Grb2 Y/F, Y-Shc-1 Y/F, and Y482/489F). Consistent with an increase in VEGF protein, the level of VEGF mRNA detected in cells expressing the Shc RTK oncoproteins (Y-Shc-1 or Y-Shc-2) was significantly enhanced when compared to cells expressing the Grb2 RTK oncoprotein or controls (Fig. 2C). Hence, the angiogenic properties of fibroblasts expressing the Shc-binding RTK oncoproteins reflect their ability to produce VEGF protein, which correlates with the rapid growth of these cells as tumors in vivo (Fig. 1C and Table 1).

The Tumorigenic and Angiogenic Properties of Cells Expressing Shc-Binding RTK Oncoproteins Are VEGF-Dependent. We next wanted to define whether VEGF-mediated angiogenesis underlines the in vivo angiogenic and tumorigenic responses of cells expressing the Shc-binding RTK oncoproteins. For this, we tested the ability of these cells to induce angiogenesis and tumor growth after their implantation in animals treated or not with the soluble VEGF inhibitor, VEGF-Trap (25 mg/kg; ref. 43). We observed that Matrigel plugs of VEGF-Trap-treated mice remained pale and poorly vascularized when compared to Matrigel plugs of vehicle-treated animals, which were red and infiltrated by many blood vessels (Fig. 3A). Similarly, we observed that the tumor growth of cells expressing either of the Shc RTK oncoproteins was drastically hindered (Fig. 3B). Although within 5 days, cells expressing the Shc-binding RTK oncoproteins were able to form small tumors in both VEGF-Trap- and vehicle-treated mice, further growth of these tumors was marginal in animals treated with VEGF-Trap (Fig. 3B). In contrast, in vehicle-treated mice these small tumors quickly expanded into large and highly vascularized tumors (Fig. 3B). These experiments conducted with VEGF-Trap demonstrate that the ability of cells expressing Shc RTK oncoproteins to induce angiogenesis and tumor growth is VEGF dependent.

Fig. 3.

Fig. 3.

Open in a new tab

VEGF-Trap abrogates the in vivo angiogenic response and tumor growth of cells expressing Shc-binding RTK oncoproteins. (A) Nude mice were treated with either vehicle or 25 mg/kg VEGF-Trap 2 days before the injection of fibroblasts expressing Shc-binding RTK oncoproteins mixed with Matrigel (105 cells per 250 μl), and treatment was continued on a twice-weekly regimen for a period of 10 days. Representative photographs of the Matrigel plugs are shown. (B) Nude mice were treated twice a week with 25 mg/kg VEGF-Trap or vehicle after s.c. implantation of cells expressing Shc-binding RTK oncoproteins. Results represent the mean tumor volume ± SEM (cm3) over time (day) from two independent experiments, each conducted with three mice per treatment group.

Shc-Binding Neu/ErbB2 RTK Oncoprotein Induces VEGF. The ErbB2/HER2 RTK is a member of the epidermal growth factor receptor family, which is often deregulated in human malignancies (45, 46). The ErbB2 RTK recruits the Grb2 and Shc adaptor proteins to different tyrosines (Fig. 4A, YB and YD, respectively), which independently mediate transforming signals (6). Studies have shown that activation of ErbB2/HER2 RTK induces VEGF expression (32, 34), but the importance of Shc versus Grb2 in ErbB2/HER2-mediated up-regulation of VEGF was unknown. To evaluate this, we tested the ability of Rat-1 fibroblasts to produce VEGF in their CM when expressing either the activated wild-type rat Neu/ErbB2 RTK oncoprotein (NT), the nontumorigenic mutant that lacks all five of the tyrosine autophosphorylation sites (NYPD), or add-back mutants, which possess only the Grb2 (NT-YB)- or Shc (NT-YD)-binding sites (Fig. 4 A and B, and ref. 6). The VEGF protein was produced in the CM of cells expressing the NT or Shc-binding Neu add-back mutant (NT-YD), where, in contrast, no detectable level of VEGF protein was observed in the CM of cells expressing the Grb2-binding mutant (NT-YB), or control NYPD mutant (Fig. 4C). Hence, in a similar manner, Shc- but not Grb2-dependent signals are sufficient for the induction of VEGF downstream from the Neu/ErbB2 RTK.

Fig. 4.

Fig. 4.

Open in a new tab

A Neu/ErbB2 RTK mutant that recruits Shc but not Grb2 increases VEGF protein. (A) Diagram of the activated Neu/ErbB2 RTK add-back mutants that retain the Grb2- or Shc-binding site. (B) Expression and phosphorylation levels of RTK add-back mutant or control proteins in Rat-1 fibroblast cell lines. (C) Levels of VEGF produced in the culture media of cells expressing the Neu/ErbB2 RTK add-back mutant or control proteins were detected by immunoblot analysis using a VEGF antibody.

Ligand Stimulation of a Shc-Binding RTK Induces VEGF. The Grb2 and Shc RTK oncoproteins, derived from the Met/HGF receptor oncoprotein, Tpr-Met, as well as the Neu add-back mutants, are constitutively activated RTK oncoproteins (3, 6). We next tested whether constitutive activation of the Shc-binding RTKs was required for VEGF production, or whether this could be regulated by a transient activation after ligand stimulation. For this, a Met receptor mutant lacking the two C-terminal tyrosines (Fig. 5A, Met-Y1349/1356F) critical for recruitment of signaling proteins (42) was engineered to specifically bind upon stimulation either to Grb2 or Shc (Fig. 5A, RTK-Grb2 or RTK-Shc). The binding specificity of the Grb2 or Shc-binding RTKs was confirmed after transient transfection in 293T cells, as well as in stable Rat-1 fibroblast cell lines by in vitro association or coimmunoprecipitation assays (Fig. 8, which is published as supporting information on the PNAS web site). The ability of the Grb2- or Shc-binding RTKs to induce the production of VEGF protein upon ligand stimulation was tested in Rat-1 fibroblasts (Fig. 5B). After 48 h of stimulation, VEGF was detected in the CM of two independent cell lines expressing the RTK-Shc, but not in cells expressing the RTK that recruits Grb2 or the Met Y1349/1356F signaling-deficient mutant (Fig. 5C). Importantly, VEGF production by cells expressing the Shc-binding RTK was not observed in absence of stimulation. These results demonstrate that the recruitment of Shc, but not of Grb2, to a RTK upon ligand stimulation is sufficient to enhance the production of VEGF.

Fig. 5.

Fig. 5.

Open in a new tab

The binding of Shc but not of Grb2 to a ligand-activated transmembrane RTK induces VEGF. (A) Schematic representation of the design of RTKs specific for the binding of Grb2 or Shc (their binding specificities are shown in Fig. 8). (B) Expression levels of the Grb2 or Shc signaling specific RTK and control proteins in populations of Rat-1 fibroblast cells. (C) Levels of VEGF in CM of Rat-1 fibroblasts (106 per 100-mm plate) after 48-h stimulation or not (CSF 100 ng/ml) was detected by immunoblot analysis using a VEGF antibody.

The Met Receptor Oncoprotein or Serum-Derived Growth Factors Fail to Induce VEGF in Shc Knockout Cells. Although the activation of the Met receptor by its ligand HGF has been shown to induce angiogenesis and VEGF production in several cell types, the proximal signaling proteins implicated have not been defined (29–31, 33). The results obtained with the signaling specific RTKs and Neu/ErbB2 add-back mutants indicate that the recruitment of Shc to RTKs plays a critical role in the induction of VEGF. To define the requirement for Shc in Met-induced VEGF expression, we stably expressed the wild-type Tpr-Met oncogene in MEFs derived from wild-type (+/+), or ShcA-deficient (–/–) mouse embryos, as well as in ShcA-deficient MEFs reexpressing the p52/p46 ShcA isoforms (–/– p52Shc, ref. 40). Consistent with Shc being a downstream target of the Met receptor (4, 7), when expressed, its level of phosphorylation on tyrosine residues was elevated by expression of Tpr-Met (Fig. 6A). An increase in the production of VEGF was induced by Tpr-Met in wild-type MEFs, but not in the ShcA-deficient cells (Fig. 6B). Similarly, VEGF was increased by serum stimulation in wild-type MEFs, but not in the ShcA-deficient cells (Fig. 6C). Notably, the induction of VEGF protein by Tpr-Met and serum-derived growth factors was rescued in ShcA-deficient MEFs transfected with the p52ShcA gene (Fig. 6 B and C). These results identify Shc, as a critical intermediate required for VEGF production downstream from the Met RTK oncoprotein and serum-derived growth factors.

Fig. 6.

Fig. 6.

Open in a new tab

The induction of VEGF mediated by the Met RTK oncoprotein and serum-derived growth factors depends on Shc. (A) Expression and phosphorylation levels of the Tpr-Met and Shc proteins in wild-type (+/+) or ShcA-deficient (–/–) MEFs, or in ShcA-deficient MEF stably transfected with p52 ShcA cDNA that express or not the Tpr-Met oncoprotein. Lysates (250 μg) prepared from the different cell lines were subjected to immunoprecipitation with anti-Met and immunoblotted with Met or pTyr antibody. To detect the phosphorylation levels of Shc proteins, cell lysates (500 μg) were subjected to immunoprecipitation with an antibody specific for Shc and immunoblotted with anti-pTyr. The expression level of Shc proteins is shown by immunoblot analysis of whole-cell lysates with anti-Shc. (B) Levels of VEGF produced after 48 h in the CM of wild-type (+/+) or ShcA-deficient (–/–) MEFs, or in ShcA-deficient MEF stably transfected with p52 ShcA cDNA, which express or not Tpr-Met, was detected by immunoblot analysis. (C) The level of VEGF in the CM of wild-type (+/+) or ShcA-deficient (–/–) MEFs, or in ShcA-deficient MEF stably transfected with p52 ShcA cDNA after 48 h of serum stimulation.

Discussion

Receptor tyrosine kinases modulate a wide range of cellular processes and their deregulation contributes to many hallmarks of cancer, including unrestrained cell proliferation, morphological transformation, anchorage-independent growth, evasion of apoptosis, cell motility, invasion, and angiogenesis (1). To better understand the contribution Grb2 and Shc recruitment to RTKs in oncogenesis, we have used RTK oncoproteins engineered to bind to a single signaling protein (3). We have previously shown that fibroblasts expressing a RTK oncoprotein engineered to recruit only the Grb2 or Shc adaptor proteins show similar transforming activities (3).

Many of the RTKs deregulated in human malignancies contribute to angiogenesis by promoting the expression of VEGF (29–39). However, the required receptor proximal events were unknown. In this study, we show that a RTK oncoprotein engineered to recruit Shc, but not a RTK binding to Grb2, promotes the production of VEGF (Fig. 2). The importance of Shc recruitment for the induction of VEGF was further demonstrated in the context of a ligand-activated RTK derived from the Met/HGF RTK and for the Neu/ErbB2 RTK oncoprotein, where in each case the recruitment of Shc, but not of Grb2, was sufficient for induction of VEGF (Figs. 4 and 5). Moreover, a critical requirement for Shc in the production of VEGF downstream from the Met/HGF RTK oncoprotein, Tpr-Met, and serum-derived growth factors was established by using Shc null MEFs (Fig. 6). In contrast to wild-type MEFs, the Tpr-Met RTK oncoprotein and serum growth factors were unable to induce VEGF production in MEFs derived from ShcA-deficient mice. Notably, the ability of Tpr-Met and serum-derived growth factors to induce VEGF was rescued by reexpression of the ShcA gene in these cells (Fig. 6).

Consistent with a role for Shc in vascular remodeling, ShcA-deficient animals die at embryonic day 11.5 of development due to severe heart defects and improper angiogenesis (40). These angiogenic defects were in part attributed to a role of Shc to potentiate activation of the mitogen-activated protein kinase pathway by RTKs in the developing vasculature (40). However, proper dosage of VEGF is critical during early development, because embryonic lethality due to vascular defects is observed by inactivation of a single VEGF allele (26, 28). Thus, our findings suggest that the angiogenic defects observed in ShcA-deficient mice might also be attributed to a role of Shc as a regulator of VEGF production.

Considering the potency of VEGF at stimulating angiogenesis and the strict dependence of solid tumors on neovascularization for their growth (26, 28), cancer cells harboring deregulated RTKs that recruit Shc, could be at an advantage through their ability to promote early initiation of tumor vascularization. In agreement with this, in contrast to cells expressing a Grb2-binding RTK oncoprotein, cells expressing RTK oncoproteins engineered to bind Shc had the capacity to promote a robust angiogenic response in mice when seeded in Matrigel, as well as to form tumors with short latency (Figs. 1 and 2 and Table 1). The angiogenic response and short tumor latency of these cells were dependent on VEGF, because these were blocked by treatment of animals with the VEGF inhibitor, VEGF-Trap (Fig. 3 and ref. 43). Furthermore, these results are consistent with the short latency of mammary tumor development and enhanced tumor burden in transgenic mice expressing an activated Neu/ErbB2 RTK mutant in which only the Shc-binding site was reintroduced (YD), when compared to a mutant of Neu/ErbB2 that binds Grb2 (YB) (8).

Mechanisms regulating VEGF expression are complex and have been shown to vary depending on the cell context and the receptor investigated (28). These include enhanced stability of VEGF mRNA, as well as transcriptional activation of the VEGF gene, through activation of SP1 and/or AP2 transcription factors, or by the stabilization of the hypoxia-induced factor, HIF-1α (28). In all cases studied, RTKs promote the production of VEGF protein by increasing levels of VEGF mRNA (29–39). Consistent with this, we have observed that the enhanced production of VEGF protein induced by the Shc-binding RTKs or the Tpr-Met oncoprotein, correlated with an increase in the level of VEGF mRNA (Fig. 2 and data not shown).

The production of VEGF induced by RTKs has been reported to depend on extracellular signal-related kinase (ERK) and/or phosphatidylinositol 3-kinase (PI3′K) (33, 36, 37, 39). The Shc adaptor protein has the capacity to link the Met and ErbB2 RTKs with both of these signaling pathways. Downstream of the Met or ErbB2 receptor, Shc is phosphorylated on tyrosine residues (3, 4, 6–8), which provide binding sites for the SH2 domain of Grb2 (15–18). This, in turn, can lead to activation of the Ras/ERK pathway, through the Ras exchange factor Sos (10, 13, 16), and the PI3′K/Akt signaling pathway, through recruitment of the docking protein, Gab1 (19–22). However, the chemical inhibition of PI3′K or mitogen-activated protein/ERK pathways had little effect on the level of VEGF produced downstream from the Shc-binding RTK or Tpr-Met oncoproteins, and Gab1 phosphorylation and its indirect recruitment to the Shc- or Grb2-binding RTKs are similar (refs. 3 and 47 and data not shown). This argues that additional signaling pathways targeted by Shc may be involved in the production of VEGF, and further experiments will be required to define these mechanisms at the molecular level.

The development of new and more efficient therapeutic drugs targeting RTK signaling in cancer requires a better understanding of the individual contributions of signaling proteins recruited to RTKs in biological events, which, if not strictly coordinated, may contribute to the development and progression of cancer. Our findings raise the possibility that the activation of Shc-dependent signaling pathways may be a key and common signaling intermediate for RTKs to promote angiogenesis through an enhanced production of VEGF. Because RTKs that promote VEGF production share the capacity to recruit and/or to phosphorylate Shc, this study identifies Shc or Shc-dependent signals as important therapeutic targets for the treatment of cancers harboring deregulated RTKs.

Supplementary Material

Supporting Figures
pnas_0308065101_index.html (1KB, html)

Acknowledgments

We thank members of the Park laboratory for helpful comments. C.S. was a recipient of the Royal Victoria Hospital Research Institute (RVH-RI) and Fonds de la Recherche en Sante du Quebec postdoctoral fellowships. H.K. and P.P., respectively, were recipients of the RVH-RI and Terry Fox research studentships. M.P. is a recipient of a Canadian Institutes of Health Research scientist award. This work was supported by an operating grant awarded to M.P. from the Canadian Institutes of Health Research.

Abbreviations: RTK, receptor tyrosine kinase; SH2, Src homology domain 2; VEGF, vascular endothelial growth factor; MEF, mouse embryo fibroblasts; CM, conditioned medium; HGF, hepatocyte growth factor.

References

  • 1.Blume-Jensen, P. & Hunter, T. (2001) Nature 411, 355–365. [DOI] [PubMed] [Google Scholar]
  • 2.Pawson, T. & Nash, P. (2000) Genes Dev. 14, 1027–1047. [PubMed] [Google Scholar]
  • 3.Saucier, C., Papavasiliou, V., Palazzo, A., Naujokas, M. A., Kremer, R. & Park, M. (2002) Oncogene 21, 1800–1811. [DOI] [PubMed] [Google Scholar]
  • 4.Fixman, E. D., Fournier, T. M., Kamikura, D. M., Naujokas, M. A. & Park, M. (1996) J. Biol. Chem. 271, 13116–13122. [DOI] [PubMed] [Google Scholar]
  • 5.Pelicci, G., Lanfrancone, L., Grignani, F., McGlade, J., Cavallo, F., Forni, G., Nicoletti, L., Pawson, T. & Pelicci, P. G. (1992) Cell 70, 93–104. [DOI] [PubMed] [Google Scholar]
  • 6.Dankort, D. L., Wang, Z., Blackmore, V., Moran, M. F. & Muller, W. J. (1997) Mol. Cell. Biol. 17, 5410–5425. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Ponzetto, C., Zhen, Z., Audero, E., Maina, F., Bardelli, A., Basile, M. L., Giordano, S., Narsimhan, R. & Comoglio, P. (1996) J. Biol. Chem. 271, 14119–14123. [DOI] [PubMed] [Google Scholar]
  • 8.Dankort, D., Maslikowski, B., Warner, N., Kanno, N., Kim, H., Wang, Z., Moran, M. F., Oshima, R. O., Cardiff, R. D. & Muller, W. J. (2001) Mol. Cell. Biol. 21, 1540–1551. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Asai, N., Murakami, H., Iwashita, T. & Takahashi, M. (1996) J. Biol. Chem. 271, 17644–17649. [DOI] [PubMed] [Google Scholar]
  • 10.Lowenstein, E. J., Daly, R. J., Batzer, A. G., Li, W., Margolis, B., Lammers, R., Ullrich, A., Skolnik, E. Y., Bar-Sagi, D. & Schlessinger, J. (1992) Cell 70, 431–442. [DOI] [PubMed] [Google Scholar]
  • 11.Blaikie, P., Immanuel, D., Wu, J., Li, N., Yajnik, V. & Margolis, B. (1994) J. Biol. Chem. 269, 32031–32034. [PubMed] [Google Scholar]
  • 12.Kavanaugh, W. M. & Williams, L. T. (1994) Science 266, 1862–1865. [DOI] [PubMed] [Google Scholar]
  • 13.Rozakis-Adcock, M., Fernley, R., Wade, J., Pawson, T. & Bowtell, D. (1993) Nature 363, 83–85. [DOI] [PubMed] [Google Scholar]
  • 14.van der Geer, P., Wiley, S., Lai, V. K. M., Olivier, J. P., Gish, G. D., Stephens, R., Kaplan, D., Shoelson, S. & Pawson, T. (1995) Curr. Biol. 5, 404–412. [DOI] [PubMed] [Google Scholar]
  • 15.Salcini, A. E., McGlade, J., Pelicci, G., Nicoletti, I., Pawson, T. & Pelicci, P. G. (1994) Oncogene 9, 2827–2836. [PubMed] [Google Scholar]
  • 16.Rozakis-Adcock, M., McGlade, J., Mbamalu, G., Pelicci, G., Daly, R., Li, W., Batzer, A., Thomas, S., Brugge, J., Pelicci, P. G., et al. (1992) Nature 360, 689–692. [DOI] [PubMed] [Google Scholar]
  • 17.Gotoh, N., Toyoda, M. & Shibuya, M. (1997) Mol. Cell. Biol. 17, 1824–1831. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.van der Geer, P., Wiley, S., Gish, G. D. & Pawson, T. (1996) Curr. Biol. 6, 1435–1444. [DOI] [PubMed] [Google Scholar]
  • 19.Fixman, E. D., Holgado-Madruga, M., Nguyen, L., Kamikura, D. M., Fournier, T. M., Wong, A. J. & Park, M. (1997) J. Biol. Chem. 272, 20167–20172. [DOI] [PubMed] [Google Scholar]
  • 20.Lock, L. S., Royal, I., Naujokas, M. A. & Park, M. (2000) J. Biol. Chem. 275, 31536–31545. [DOI] [PubMed] [Google Scholar]
  • 21.Ong, S. H., Dilworth, S., Hauck-Schmalenberger, I., Pawson, T. & Kiefer, F. (2001) EMBO J. 20, 6327–6336. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Gu, H., Maeda, H., Moon, J. J., Lord, J. D., Yoakim, M., Nelson, B. H. & Neel, B. G. (2000) Mol. Cell. Biol. 20, 7109–7120. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Basu, T., Warne, P. H. & Downward, J. (1994) Oncogene 9, 3483–3489. [PubMed] [Google Scholar]
  • 24.Batzer, A. G., Rotin, D., Urena, J. M., Skolnik, E. Y. & Schlessinger, J. (1994) Mol. Cell. Biol. 14, 5192–51201. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Ravichandran, K. S. (2001) Oncogene 20, 6322–6330. [DOI] [PubMed] [Google Scholar]
  • 26.Carmeliet, P. & Jain, R. K. (2000) Nature 407, 249–257. [DOI] [PubMed] [Google Scholar]
  • 27.Folkman, J. (1971) N. Engl. J. Med. 285, 1182–1186. [DOI] [PubMed] [Google Scholar]
  • 28.Ferrara, N. (1999) J. Mol. Med. 77, 527–543. [DOI] [PubMed] [Google Scholar]
  • 29.Toyoda, M., Takayama, H., Horiguchi, N., Otsuka, T., Fukusato, T., Merlino, G., Takagi, H. & Mori, M. (2001) FEBS Lett. 509, 95–100. [DOI] [PubMed] [Google Scholar]
  • 30.Gille, J., Khalik, M., Konig, V. & Kaufmann, R. (1998) J. Invest. Dermatol. 111, 1160–1165. [DOI] [PubMed] [Google Scholar]
  • 31.Wojta, J., Kaun, C., Breuss, J. M., Koshelnick, Y., Beckmann, R., Hattey, E., Mildner, M., Weninger, W., Nakamura, T., Tschachler, E. & Binder, B. R. (1999) Lab. Invest 79, 427–438. [PubMed] [Google Scholar]
  • 32.Petit, A. M., Rak, J., Hung, M. C., Rockwell, P., Goldstein, N., Fendly, B. & Kerbel, R. S. (1997) Am. J. Pathol. 151, 1523–1530. [PMC free article] [PubMed] [Google Scholar]
  • 33.Dong, G., Chen, Z., Li, Z. Y., Yeh, N. T., Bancroft, C. C. & Van Waes, C. (2001) Cancer Res. 61, 5911–5918. [PubMed] [Google Scholar]
  • 34.Yen, L., You, X. L., Al Moustafa, A. E., Batist, G., Hynes, N. E., Mader, S., Meloche, S. & Alaoui-Jamali, M. A. (2000) Oncogene 19, 3460–3469. [DOI] [PubMed] [Google Scholar]
  • 35.Calza, L., Giardino, L., Giuliani, A., Aloe, L. & Levi-Montalcini, R. (2001) Proc. Natl. Acad. Sci. USA 98, 4160–4165. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Maity, A., Pore, N., Lee, J., Solomon, D. & O'Rourke, D. M. (2000) Cancer Res. 60, 5879–5886. [PubMed] [Google Scholar]
  • 37.Wang, D., Huang, H. J., Kazlauskas, A. & Cavenee, W. K. (1999) Cancer Res. 59, 1464–1472. [PubMed] [Google Scholar]
  • 38.Seghezzi, G., Patel, S., Ren, C. J., Gualandris, A., Pintucci, G., Robbins, E. S., Shapiro, R. L., Galloway, A. C., Rifkin, D. B. & Mignatti, P. (1998) J. Cell Biol. 141, 1659–1673. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Miele, C., Rochford, J. J., Filippa, N., Giorgetti-Peraldi, S. & Van Obberghen, E. (2000) J. Biol. Chem. 275, 21695–21702. [DOI] [PubMed] [Google Scholar]
  • 40.Lai, V. K.-M. & Pawson, T. (2000) Genes & Dev. 14, 1132–1145. [PMC free article] [PubMed] [Google Scholar]
  • 41.Rodrigues, G. A., Naujokas, M. A. & Park, M. (1991) Mol. Cell. Biol. 11, 2962–2970. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Zhu, H., Naujokas, M. A., Fixman, E. D., Torossian, K. & Park, M. (1994) J. Biol. Chem. 269, 29943–29948. [PubMed] [Google Scholar]
  • 43.Holash, J., Davis, S., Papadopoulos, N., Croll, S. D., Ho, L., Russell, M., Boland, P., Leidich, R., Hylton, D., Burova, E., et al. (2002) Proc. Natl. Acad. Sci. USA 99, 11393–11398. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Passaniti, A., Taylor, R. M., Pili, R., Guo, Y., Long, P. V., Haney, J. A., Pauly, R. R., Grant, D. S. & Martin, G. R. (1992) Lab. Invest. 67, 519–528. [PubMed] [Google Scholar]
  • 45.Slamon, D. J., Godolphin, W., Jones, L. A., Holt, J. A., Wong, S. G., Keith, D. E., Levin, W. J., Stuart, S. G., Udove, J., Ullrich, A., et al. (1989) Science 244, 707–712. [DOI] [PubMed] [Google Scholar]
  • 46.Olayioye, M. A., Neve, R. M., Lane, H. A. & Hynes, N. E. (2000) EMBO J. 19, 3159–3167. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Lock, L. S., Frigault, M. M., Saucier, C. & Park, M. (2003) J. Biol. Chem. 278, 30083–30090. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supporting Figures
pnas_0308065101_index.html (1KB, html)
pnas_0308065101_1.html (898B, html)
pnas_0308065101_08065Fig7.jpg (40.3KB, jpg)
pnas_0308065101_2.html (2.4KB, html)
pnas_0308065101_08065Fig8.jpg (69.4KB, jpg)

Articles from Proceedings of the National Academy of Sciences of the United States of America are provided here courtesy of National Academy of Sciences

RESOURCES