| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J. Biol. Chem., Vol. 282, Issue 23, 16764-16775, June 8, 2007
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
¶1
From the
Department of Neuroscience and Cell Biology, ¶Sealy Center for Cancer Cell Biology, University of Texas Medical Branch, Galveston, Texas 77555-1074 and Department of Molecular Biology and Microbiology, Biomolecular Science Center, University of Central Florida, Orlando, Florida 32826-3227
Received for publication, November 22, 2006 , and in revised form, April 18, 2007.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
Endocytic studies on other RTKs, most notably the epidermal growth factor receptor (EGFR), suggest that modification of the endocytic trafficking machinery by the activated receptor and/or its associated signaling molecules regulates receptor internalization and degradation. For example, the signaling adaptor Grb2 has been shown to function as an important initiator of EGFR endocytosis (17–20). Small interfering RNA (siRNA)-mediated depletion of Grb2 and overexpression of Grb2-SH3 mutants deficient in binding downstream signaling molecules blocks the recruitment of EGFR into clathrin-coated pits and subsequent receptor internalization, with negligible effects on the downstream activation of mitogen-activated protein kinase or phosphatidylinositol 3-kinase (19). Grb2 functions to regulate EGFR internalization through the recruitment of the E3 ubiquitin ligase Cbl (20–22). However, it remains unclear if the requirement for Grb2 on EGFR endocytosis is mediated solely through interactions with Cbl, functions at the level of EGFR ubiquitination, involves additional Grb2-dependent interactions, or extends to other RTKs.
cMet and the EGFR are prototypic members for distinct RTK subfamilies and as such exhibit unique structural and functional characteristics. In contrast to the EGFR, cMet is a heterodimer composed of an extracellular 
chain that is disulfide-linked to a transmembrane 
chain that contains cytoplasmic tyrosine kinase activity (23–25). The extracellular domain of cMet contains a distinctive -propeller fold that mediates ligand binding (26). Ligand binding induces the autophosphorylation of cMet and EGFR at multiple tyrosines in their respective C-terminal regions enabling the recruitment and subsequent activation of shared downstream adaptor and signaling molecules. The EGFR possesses six tyrosine transphosphorylation sites dispersed throughout the C-terminal region (for review, see Ref. 27), whereas cMet contains two key tyrosine phosphorylation sites at positions 1349 and 1356 in a unique multisubstrate-docking site, that function to concomitantly activate multiple downstream signaling pathways (28, 29). In contrast to cMet, which contains one binding site for the signaling adaptor Grb2 at phosphotyrosine 1356 (28, 30), Grb2 binds EGFR directly at two sites (31). Moreover, Grb2 appears to play a minor role in signaling events downstream of ligand activated cMet in vivo (32–34) when compared with the EGFR (34). The adaptor protein Gab1 binds to two sites in the multisubstrate-docking site of cMet; directly at phosphotyrosine 1349 or indirectly via the adaptor Grb2 to phosphotyrosine 1356 (35–38). Gab1 recruitment is essential for the induction of cell motility, branching morphogenesis, and the formation of crypts in three-dimensional matrices, hallmarks of cMet signaling (29, 33, 39–41). Given these differences in adaptor recruitment and biological activity, some differences may exist in the regulatory mechanisms controlling cMet versus EGFR internalization.
The E3 ubiquitin ligase Cbl has been shown to mediate the ubiquitination of ligand-activated cMet (4, 5, 14, 42, 43), receptor internalization (4), and degradation (13, 14). Cbl recruitment via its tyrosine kinase binding domain to phosphotyrosine 1003 of cMet has been shown to contribute to receptor ubiquitination (13, 14, 42) and degradation in lysosomes (13, 14, 43). Cbl has been shown to function as an adaptor important for cMet internalization through clathrin-coated pits via the recruitment of the endophilin-CIN85 complex (4). However, a requirement for Cbl ubiquitin ligase activity and its mode of recruitment for cMet internalization was not examined. In this study we report that the ubiquitin ligase activity of Cbl is important for clathrin-mediated cMet internalization. Moreover, Grb2 bound directly to the cMet multisubstrate docking site acts as an intermediary for Cbl ubiquitin ligase activity in these events.
|
EXPERIMENTAL PROCEDURES |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
-actin and anti-phosphotyrosine clone PY-20 (Sigma-Aldrich), ubiquitin (P4D1), Cbl (C-15), Grb2 (C-23), cMet C-12, and cMet C-28 (Santa Cruz Biotechnologies), anti-Gab1 CT, phospho-Met Tyr-1234, Tyr-1235 (UpState Biotechnology), Gab1, phospho-Gab1, c-Jun N-terminal kinase (JNK), phospho-JNK, p42/p44 MAPK, phospho p42/p44, MEK (mitogen-activated protein kinase/extracellular signal-regulated kinase kinase), and phospho-MEK (Cell Signaling Technology), the monoclonal GFP antibody (JL-8) (BD Biosciences Clontech), peroxidase-conjugated goat anti-mouse immunoglobulin G (IgG) and goat anti-rabbit IgG (Jackson ImmunoResearch Laboratories, Inc.), and Alexa488 or Alexa594-labeled goat anti-mouse, anti-rabbit, or anti-rat secondary antibodies (Invitrogen). A pLXSN plasmid encoding wild type (wt)-cMet was previously reported (44). pLXSN plasmids encoding the full-length Met mutants N1358H and Y1349F, Y1356F were generously provided by Pascal Peschard and Morag Park (McGill University). The K1110A allele encodes a Met receptor lacking kinase activity (45, 46). The K1110A mutation was generated through a PCR-based SOEing approach (47) using wild type cMet cDNA as a template, Pfu DNA polymerase, and the primers 5'-CAAGCTAGCCACAGCACAGTG-3',5'-CCCATCCTAACTAGTGGGGAC-3',5'-CACTGTGCTGTGGCTAGCTTG-3', and 5'-GCTCTAGAACTAGTGGATCCC-3'. The final 
1.3-kilobase PCR product was subcloned between the SpeI and XcmI sites of pLXSN-Met-wt, resulting in a full-length Met gene with the K1110A mutation and verified by DNA sequencing. The various Met alleles were subcloned from the plasmid pLXSN into the murine stem cell virus vector pMSCVpuro (BD Biosciences) for retroviral-mediated transfection in the target cell lines. pYFP-N1 plasmids encoding Grb2 and Grb2-mSH3 were kindly provided by Lawrence Samelson (National Institutes of Health) (48). The pEGFP plasmids encoding Cbl, Cbl-SH2, Cbl-SH2/R86A, Grb2-SH2, and Grb2/R86A were kindly provided by Alexander Sorkin (University of Colorado Health Sciences Center) (17, 22). pcDNA3-Myc-UbR and pcDNA3-Myc-UbRL8A/I44A were kind gifts from Inger Madshus (Norwegian Radium Hospital) (18). Plasmids encoding hemagglutinin-tagged Cbl (wt), 70Z-Cbl, and Cbl-G306E were kindly provided by Jannie Borst (The Netherlands Cancer Institute) (20). Cell Lines—The H10 cell line derived from kidney epithelial cells from embryos of cMet null –/– mice (49) was a generous gift of Dr. Lloyd Cantley (Yale University School of Medicine). Polyclonal H10 cell lines expressing cMet alleles were generated by retroviral transfection essentially as described (50), except that 1–5 µg/ml puromycin was used for selection, and limiting dilution was not performed. Cell lines were analyzed for cMet expression by Western blotting with the anti-cMet antibody DL-21 (Upstate%20Biotechnology">Upstate Biotechnology), and those cell lines that exhibited similar cMet levels were chosen for further analysis. The parental H10 cMet null –/– cell line was maintained in Dulbecco's modified Eagle's medium/F-12(1:1) with 10% fetal bovine serum. H10 derivatives stably expressing wild type or mutant cMet alleles were grown in the same medium containing 5 µg/ml puromycin. Human mammary epithelial cells (T47D) stably expressing full-length human cMet (T47D/cMet) were a generous gift of Morag Park (McGill University) and were maintained in Dulbecco's modified Eagle's medium containing 10% fetal bovine serum supplemented with 400 µg/ml G418.
siRNAs and Cell Transfections—For siRNAs depletion studies, cells were grown on coverslips in 60-mm plates and transfected with SmartpoolsTM containing four siRNA duplexes for Gab1, Grb2, Shc, or a control siRNA using Lipofectamine 2000 reagent (Invitrogen). Typically each coverslip of cells was transfected with 200 pmol of siRNA and 8 µl of Lipofectamine 2000. T47D/cMet cells were routinely cultured on coverslips coated with 100 µg/ml polylysine before experimentation. All experiments were routinely performed 72 h after siRNA transfection. Cell transfections using plasmids were performed as previously described (3).
Immunoprecipitation and GST Pulldown Assays—For immunoprecipitation (IP) studies cells were washed twice with ice-cold PBS (137 mM NaCl, 2.7 mM KCl, 8.2 mM Na2HPO4, 1.5 mM KH2PO4) and lysed on ice in radioimmune precipitation assay buffer (50 mM Tris-HCl, pH 7.4, 1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% SDS, 150 mM NaCl, 1 mM EDTA) with protease inhibitors (2 µg/ml aprotinin, 2 µg/ml leupeptin, 2 µg/ml pepstatin) and phosphatase inhibitors (10 mM NaF, 2 mM Na3VO4). Cell lysates were cleared by centrifugation at 12,000 x g at 4 °C, and then 500–1000 µg of each lysate was incubated with 5 µg of primary antibody overnight at 4 °C. Antibody-protein complexes were precipitated with 50 µl of protein A/G-agarose solution (Pierce) by rotation at 4 °C for 4 h. The protein-beads complex were collected by centrifugation at 1000 x g for 5 min, washed with lysis buffer 3 times, and then resuspended in SDS loading buffer and fractionated by SDS/PAGE. For GST pulldown assays, the cell lysates were prepared in lysis buffer (1% Triton X-100, 50 mM Tris-HCl, pH 7.4, 150 mM NaCl, 2 mM EDTA, 2 µg/ml aprotinin, 2 µg/ml leupeptin, 2 µg/ml pepstatin, 10 mM NaF, 2 mM Na3VO4). Lysates (0.5–1.0 mg of protein) were incubated with the appropriate GST fusion proteins prebound to glutathione-Sepharose beads (Sigma) overnight at 4 °C. Beads were washed three times in lysis buffer followed by a final wash in 10 mM Tris-HCl, pH 7.4, then analyzed by SDS/PAGE.
Cell Surface Biotinylation and Western Analysis—The biotinylation and receptor internalization assays were described previously (3). Briefly, cells were surface-biotinylated at 4 °C for 30 min with EZ-link NHS-SS-biotin (Pierce). Surface-expressed proteins were isolated directly using streptavidin-agarose beads as described by the manufacturer (Pierce) and identified by Western analysis. Western analysis was performed using ECL (GE Healthcare), and the resulting digitized blots were quantified using AlphaEase Version 3.1.2 Software (Alpha Innotech Corp.).
Flow Cytometry Analysis—cMet internalization from the plasma membrane was measured by assaying the residual amount of immunoreactive receptor accessible at the cell surface to a monoclonal antibody recognizing an extracellular epitope. Briefly, cells grown to 80% confluence on 60-mm dishes were sera-starved for 4–6 h and then stimulated with 100 ng/ml internalin B (InlB) at 37 °C under steady state conditions to drive cMet internalization. At the indicated times, receptor endocytosis was terminated by placing the cells on ice. The cells were immediately rinsed with ice-cold PBS, and residual surface-bound InlB was stripped using three consecutive 5-min ice-cold acid washes (Dulbecco's modified Eagle's medium, pH 3.5). Cells were detached with PBS containing 5 mM EDTA and then resuspended in ice-cold FACS buffer (PBS containing 2% fetal bovine serum). Cells were incubated with anti-cMet antibody (AF276, R&D System, MN) in FACS buffer at 4 °C for 1 h, washed 3 times with ice-cold FACS buffer, and then incubated with Alexa488-conjugated secondary antibody in FACS buffer at 4 °C for 30 min. After 3 ice-cold washes with FACS buffer, the cells were analyzed using a FACS Canto flow cytometer or fixed with 2% paraformaldehyde, PBS for later analysis. Cell viability, as determined by the exclusion of propidium iodide, routinely exceeded 95%. 20,000 cells were analyzed for each sample in triplicate for each condition in each experiment. The mean fluorescence intensity (MFI) of the cells at each time point was analyzed and compared with a negative control using control antibody only. The MFI values for each condition were averaged, and the S.E. was calculated across all experiments. The relative percentage of residual cell surface cMet at each time point (tx) was calculated relative to the MFI of cells without internalization (t0) as (MFI tx – MFI control antibody only/MFI t0 – MFI control antibody only) x 100.
InlB Labeling, Confocal Microscopy, and Analysis—The purification and labeling of recombinant His6-tagged InlB has been described in detail elsewhere (3). For confocal microscopy cells grown on coverslips were incubated in media containing 5.0 µg/ml Alexa594-labeled Tfn with 100 ng/ml InlB or HGF for 10 min at 37 °C, then fixed immediately after each experiment using 4% paraformaldehyde (Ted Pella Inc.) in PBS. Residual paraformaldehyde was quenched using 50 mM NH4Cl/PBS, and cells were then permeabilized with 0.05% (w/v) saponin (or 0.1% Triton X-100, PBS for HGF staining) for 20 min and then blocked with 10% goat serum in PBS. All antibody dilutions were performed in 5% goat serum, PBS with 0.05% saponin for 1 h. In studies staining for HGF, saponin was omitted from the buffers. Coverslips were mounted and observed using a Zeiss LSM 510 confocal microscope with a 63x oil (1.4 NA) immersion objective. Samples were visualized with the 488- and 543-nm laser lines and emission filter sets at 505–530 nm for YFP and Alexa488 detection or 585–615 nm for Texas Red and Alexa594 detection, respectively. Figure presentation was accomplished in Adobe Photoshop Version 6.0. Quantification of co localization and internal fluorescence intensity were accomplished using Metamorph v5.0 (Molecular Devices) as described previously (3). All pixel intensity levels were normalized relative to control values, expressed as a percentage ± S.E., and differences were statistically verified by analysis of variance using GraphPad Prism Software (GraphPad Software Inc.).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
but not HGF, indicating that cMet signaling is specifically abrogated in these cells (49). We generated several cell lines stably expressing moderate levels of wt or mutant cMet, including tyrosine kinase-deficient K1110A cMet (KinD-cMet) and the multisubstrate-docking site mutant Y1349F, Y1356F (YF-cMet), which is deficient in Gab1 and Grb2 binding. siRNA-mediated depletion of clathrin heavy chain inhibited wt cMet internalization (supplemental Fig. S1), confirming that cMet internalization is clathrin-dependent in these cells, consistent with our previous studies using T47D/cMet and Vero cells (3). We previously reported that the soluble form of the InlB protein of Listeria monocytogenes mimics HGF-induced cMet internalization and degradation (3). In addition to being presented as a soluble protein, InlB is also found anchored to the cell surface of bacteria (51). The surface-anchored form of InlB mediates bacterial uptake into host cells (5), whereas soluble InlB is thought to modulate host transcriptional responses (52, 53). We previously reported that HGF and soluble InlB are internalized with cMet via a clathrin-dependent mechanism (3). Interestingly, evidence indicates that the internalization of soluble InlB or HGF probably occurs through a mechanism that differs from that utilized by bacteria expressing surface-anchored InlB. Although the uptake of Listeria and cMet activated by soluble ligands involves select components of clathrin-coated pits (5), Listeria entry requires phosphatidylinositol 3-kinase activity (54), whereas the internalization of soluble HGF or InlB does not depend on this kinase (3). Because soluble InlB and HGF are internalized through identical pathways, we used InlB as an initial ligand to examine the molecular control of cMet endocytosis and then validated key findings using the physiological ligand HGF.
Western analysis of cell lysates prepared from the stable cell lines expressing wt or mutant cMet confirmed that treatment with InlB for 15 min at 37 °C caused phosphorylation of wt and the YF-cMet mutant on tyrosine residues 1234 and 1235 in the activation loop but not on kinase-dead cMet (KinD-cMet) (Fig. 1A). Ubiquitination of wt cMet but not KinD-cMet was increased in response to InlB. A low level of ubiqitinated YF-cMet was detected under these conditions, consistent with the direct association of Cbl to phosphotyrosine 1003 of cMet, independently of Grb2 binding (14). Cbl phosphorylation was not detected in unstimulated cells or in cells stably expressing KinD-cMet after treatment with ligand (Fig. 1A). In keeping with previous results (14, 36), cMet-mediated phosphorylation of Cbl was reduced in cells expressing YF-cMet relative to cells expressing wt cMet. This is consistent with a role for Grb2 in recruiting Cbl to cMet via phosphotyrosine 1356. Surface biotinylation assays were used to confirm the cell surface expression of wt and mutant cMet in each of the polyclonal cell lines (Fig. 1B). Cells were surface-biotinylated at 4 °C for 30 min using NHS-SS-biotin. Cell lysates were then prepared, and biotinylated proteins were isolated by streptavidin pulldown and analyzed by Western blots. Comparable levels of wt and mutant cMet were detected at the cell surface of the cell lines (Fig. 1B). No cMet was detected in the parental null cells, consistent with previous reports (49).
Using a GST pulldown assay, we verified that Grb2 bound to wt cMet in response to InlB but not to YF-cMet or KinD-cMet (Fig. 1C). No Grb2 binding was detected in the absence of ligand, confirming that Grb2 binding to wt cMet was specific and dependent on cMet autophosphorylation. Similarly, Gab1 phosphorylation in response to InlB was detected in null cells expressing wt cMet but not in cells expressing the YF- or KinD-cMet mutants (Fig. 1D), consistent with previous reports using fibroblasts and HEK293 cells (35, 55). Together, these data confirm that the recruitment of Gab1 and Grb2 to cMet is disrupted in the YF- and KinD-cMet mutants but not in cells expressing wt cMet.
|
Grb2 Is Required for cMet Internalization—Our data indicating the dependence on phosphotyrosines 1349 and 1356 in the multisubstrate docking region for cMet internalization raised the possibility that the adaptors Grb2 and/or Gab1 could be required for this process. To test this idea we used specific siRNAs to deplete T47D/cMet cells of endogenous Grb2. T47D/cMet cells are a human mammary epithelial cell line that stably expresses human cMet on their surface and have been extensively used on studies examining cMet signaling and endocytosis in the past (3, 13, 14, 44). Endogenous Grb2 was weakly detected by Western analysis in T47D/cMet cells transfected with Grb2-siRNAs (Fig. 2A). Conversely, comparable levels of endogenous transferrin receptor, actin, and cMet were observed in Grb2-depleted T47D/cMet cells as well as untransfected cells, and T47D/cMet cells transfected with a control siRNA, demonstrating the specificity of the siRNA treatment. Western analysis confirmed that siRNA-mediated depletion of Grb2 did not reduce cMet phosphorylation in response to InlB (10 min at 37 °C) (Fig. 2B). Conversely, cMet ubiquitination and Cbl phosphorylation were reduced by Grb2 depletion. siRNA-treated cells were coincubated with Tfn and HGF or Alexa-InlB for 10 min at 37 °C, and ligand internalization was measured by confocal microscopy. cMet internalization in response to HGF or InlB was reduced 51.3 ± 4.85 and 74.4 ± 2.87%, respectively, in Grb2-depleted cells versus cells transfected with a control siRNA (Fig. 2C). Comparable levels of co-internalized Tfn were detected in control and Grb2-depleted T47D/cMet cells, indicating that loss of Grb2 led to a specific block in cMet internalization and not a general defect in clathrin-mediated endocytosis (Fig. 2C and supplemental Fig. S3). siRNA-mediated Grb2 depletion caused a comparable block in cMet endocytosis in cMet null cells stably expressing wild type receptor, indicating that endogenous Grb2 is required for cMet uptake in different cell types (supplemental Fig. S4). We confirmed the requirement for Grb2 in cMet internalization using flow cytometry. As expected, we observed higher levels of surface cMet in Grb2-depleted T47D/cMet cells relative to cells transiently transfected with a nonspecific control siRNA after receptor activation with InlB (Fig. 2D). Thus, loss of Grb2 led to a specific block in HGF- and InlB-induced cMet internalization and not a general inhibition in clathrin-mediated endocytosis.
|
|
We tested the idea that the ubiquitin ligase activity of Cbl was important for cMet internalization using confocal microscopy. T47D/cMet cells were transfected with wt Cbl or Cbl mutants defective in either ubiquitin ligase activity (70Z-Cbl) or binding to phosphotyrosine 1003 (Cbl-G306E) (Fig. 4A). Previous studies have demonstrated that Cbl binds cMet and induces its ubiquitination in response to HGF, whereas the ubiquitin ligase-deficient mutant Cbl-70Z interacts with cMet but is unable to induce receptor ubiquitination (4, 13, 14, 43). The cells were allowed to co internalize Tfn with Alexa-InlB or HGF for 10 min at 37 °C, and the relative amount of internalized ligand was examined by confocal microscopy (supplemental Fig. S7). cMet internalization was unaffected in mock-transfected control cells and in cells expressing wt Cbl or G306E, a mutant unable to bind RTKs directly. However, in cells overexpressing the ubiquitin ligase-deficient mutant 70Z-Cbl, internalized HGF and Alexa-InlB staining was reduced to 33.5 ± 2.4 and 50.4 ± 2.6% of control levels, respectively (Fig. 4B). To confirm the requirement of Cbl-mediated ubiquitination for cMet internalization, we took advantage of well characterized ubiquitin mutants that have altered conjugation and binding properties (Fig. 4C). Cbl functions to covalently attach monoubiquitin to substrate proteins via the C-terminal glycine residues of ubiquitin. The ubiquitin mutant UbR lacks these glycine residues important for substrate conjugation. However, UbR retains the ability to interact with intracellular proteins harboring ubiquitin binding domains, including the ubiquitin-interacting motif. Thus, when expressed in cells, UbR functions to block the interaction of ubiquitinated proteins with proteins harboring a ubiquitin-interacting motif (18). Using confocal microscopy, we examined the effect of transiently expressing mutant ubiquitin (UbR) on cMet internalization in T47D/cMet cells. UbR containing the double mutation L8A,I44A (UbR-L8A,I44A) that blocks the interaction of UbR with ubiquitin interacting motifs was used as a control (supplemental Fig. S8). As shown in Fig. 4D, expression of UbR reduced cMet internalization to 33.7 ± 3.4% relative to mock-transfected control cells or cells expressing the double mutant UbR-L8A,I44A. Tfn uptake was unaffected in cells expressing UbR or UbR-L8A,I44A. This is the first report indicating that the ubiquitin ligase activity of Cbl is important for cMet internalization in response to soluble InlB as well as HGF.
|
|
The Signaling Adaptor Gab1 Is Not Required for cMet Internalization—In addition to binding Cbl, Grb2 has been shown to bind Gab1, an adaptor protein that mediates almost all of the signaling cascades downstream from activated cMet including phosphatidylinositol 3-kinase, c-Jun N-terminal kinase, and extracellular signal-regulated kinase signaling (for review, see Ref. 63). Gab1 is recruited to cMet by two mechanisms. The primary mechanism involves the indirect recruitment of Gab1 to cMet via Grb2 (35, 38). In addition, Gab1 can bind directly to phosphotyrosine 1349 in the multisubstrate docking site of cMet (64). To examine the importance of Gab1 in cMet internalization, a Gab1-specific siRNA was used to deplete T47D/cMet cells of endogenous Gab1. Western analysis confirmed the specificity of the siRNA-mediated Gab1 knockdown in cells transfected with Gab1 but not a scrambled control (Con) siRNA (Fig. 6A). Comparable levels of Cbl, actin, Grb2, transferrin receptor, and cMet were detected in Gab1-depleted cells as well as mock-transfected cells and cells transfected with a control siRNA. Western analysis confirmed that siRNA depletion of Gab1 did not inhibit ligand-induced cMet phosphorylation and ubiquitination or Cbl phosphorylation (Fig. 6B). To examine the effect of Gab1 depletion on cMet internalization, siRNA-treated cells were incubated with InlB (100 ng/ml for 10 min at 37 °C) to activate cMet signaling, and the relative amounts of residual cell surface cMet were measured using flow cytometry (Fig. 6C). Comparable levels of cell surface cMet were detected in control and Gab1-depleted cells, indicating that siRNA-mediated knockdown of Gab1 did not inhibit cMet internalization in response to InlB. Confocal microscopy studies confirmed that cMet internalization in response to InlB or HGF was not decreased in Gab1 depleted T47D/cMet cells or cells transfected with a control siRNA (data not shown). Thus, loss of Gab1 does not block clathrin-mediated cMet internalization.
The Direct Binding of Grb2 to the Multisubstrate Docking Site of cMet Promotes Receptor Endocytosis—A similar role for Grb2 and Cbl in EGFR internalization has been reported (17, 18, 21, 22), suggesting a general role for these proteins in RTK endocytosis. However, it is unclear whether Grb2 functions by directly interacting with cMet at phosphotyrosine 1356. Alternatively, the ability of the adaptor Shc to recruit Grb2 indirectly to cMet (56) may underlie the requirement for Grb2 in cMet endocytosis. To distinguish between these possibilities, we generated a polyclonal cell line using cMet–/– null cells expressing moderate levels of the cMet mutant N1358H (NH-cMet). N1358H is a well characterized mutation that specifically interferes with the direct binding of Grb2 to cMet with no measurable effect on the recruitment of other signaling or adaptor molecules including Shc (29, 65). Surface biotinylation studies confirmed that, like wt cMet, the mutant receptor NH-cMet was expressed at the cell surface (Fig. 7A). GST pulldown assays confirmed that recombinant Grb2 bound to wt cMet in response to InlB. Conversely, Grb2 binding to NH-cMet was inhibited under these conditions (Fig. 7B). cMet ubiquitination and Cbl phosphorylation in response to InlB was detectable in null cells expressing wt cMet but reduced in cells expressing the mutant NH-cMet (Fig. 7C). These data support the conclusion that the direct binding of Grb2 to the docking site of cMet is inhibited by the mutation N1358H. We next compared the internalization properties of NH-cMet with the wild type receptor. Using Alexa-InlB to stimulate cMet internalization (10 min, 37 °C), we confirmed that cMet internalization was reduced in cMet null cells stably expressing NH-cMet but not in cMet null cells stably expressing wt cMet by confocal microscopy (Fig. 7D). To confirm our confocal studies, we used flow cytometry analysis to compare the internalization properties of wt and NH-cMet. Cells were incubated in media containing 100 ng/ml InlB for increasing periods of time at 37 °C and processed for flow cytometry. Under these conditions, the cell surface levels of NH-cMet remained unaltered over the course of the experiment. Conversely, cell surface levels of wt cMet decreased to 35.7 ± 7.6% after treatment with 100 ng/ml InlB for 20 min (Fig. 7E).
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
A similar role for Grb2 has been reported for the EGFR. EGFR internalization through clathrin-coated pits was specifically blocked in Grb2-depleted cells (17). The block in EGFR internalization imposed by Grb2 depletion was rescued by transient expression of a Grb2·Cbl chimera (22). Overexpression of Grb2-SH3 mutants deficient in binding downstream signaling molecules inhibited the recruitment of EGFR to clathrin-coated pits (18) and prevented the translocation of Grb2 with activated EGFR into endocytic structures, presumably endocytic transport vesicles (48). In a separate study, overexpression of Grb2 with inactivating point mutations in the SH3 domains inhibited EGF-induced coated pit formation, with negligible effects on the downstream activation of mitogen-activated protein kinase or phosphatidylinositol 3-kinase (19). Our studies showing that Grb2-mediated internalization of cMet was dependent on the ubiquitin ligase activity of Cbl are consistent with several models for EGFR internalization in which the Grb2·Cbl complex likely functions to couple distinct stages of RTK endocytosis (20, 22, 68).
The multiadaptor Gab1 plays a critical role in cMet signaling by providing a scaffold for the simultaneous activation of several downstream signaling cascades including the activation of phosphatidylinositol 3-kinase, c-Jun N-terminal kinase, and extracellular signal-regulated kinase signaling (for review, see Ref. 63). Our data showing that the recruitment of Gab1 to cMet is dispensable for receptor internalization are consistent with our previous report that phosphatidylinositol 3-kinase signaling was not required for cMet endocytosis (3). cMet internalization was unaffected by overexpression of a dominant negative mutant for the p85 subunit of phosphatidylinositol 3-kinase. Similarly, no differences in the internalization properties of cMet were observed in T47D/cMet cells treated with the phosphatidylinositol 3-kinase inhibitors LY294002 or wortmannin (3). Thus, although Gab1 likely regulates the intensity and duration of cMet signaling, this does not appear to occur at the level of receptor internalization.
The recruitment of Grb2 directly or indirectly via the adaptor Shc to phosphotyrosine 1356 of cMet is essential for cell transformation and experimental metastases in response to oncogenic Met (69). The ability of Shc to form a stable complex with Grb2 in response to HGF (56) and EGF (70, 71) suggested that Grb2 recruited indirectly via Shc could regulate RTK endocytosis. Our data showing that cMet internalization was impaired by the N1358H mutation suggested that the direct recruitment of Grb2 to cMet rather than Grb2 recruited indirectly through Shc mediates cMet internalization, since Shc binding is unaffected by the N1358H mutation (29, 65). Consistent with this idea, siRNA-mediated Shc depletion did not interfere with cMet internalization through clathrin-coated pits. Thus, in light of our observations, we propose that Grb2 bound directly to cMet is indispensable for receptor endocytosis. In contrast to cMet, Shc has been reported to play a role in EGF-induced EGFR internalization (66, 72). EGF has been shown to promote the formation of a complex between wild type Shc and the adapter complex AP2 that is recruited to phosphotyrosine 1148 of the EGFR. Consistent with a role for Shc in EGFR endocytosis, the internalization rate constant of an EGFR truncation mutant lacking the region encompassing phosphotyrosine 1148 was significantly slower than that observed for the wild type receptor (73). Thus, the mode of Grb2 recruitment appears to diverge for cMet versus EGFR internalization, suggesting that distinct subsets of Grb2-mediated interactions are involved in the clathrin-dependent internalization of different RTKs.
The precise mechanism by which Cbl regulates cMet internalization and RTK endocytosis in general remains controversial. Cbl can be viewed as a bimodal molecule that functions as an adaptor protein that physically links Cbl/cMet complexes to the endocytic machinery (4) and as an E3 ubiquitin ligase to ubiquitinate cMet (13, 14) and its associated endocytic components. Although Grb2 and Cbl ubiquitin ligase activity was recently reported to be important for L. monocytogenes infection (5), these studies focused on the internalization of Listeria expressing surface anchored InlB and not soluble InlB. Moreover, these studies did not examine cMet internalization directly but instead monitored bacterial engulfment. A role for Cbl as an adaptor that links HGF-activated cMet to the endophillin-CIN85 complex (4) has been clearly demonstrated. In these studies inhibition of complex formation using dominant interfering forms of endophilin, CIN85, or Cbl was sufficient to inhibit cMet internalization and enhance cMet signaling. However, these studies did not address the requirement for Cbl ubiquitin ligase activity in this event or the mode of Cbl recruitment to cMet. Our studies showing that expression of a Cbl mutant deficient in ubiquitin ligase activity blocks cMet internalization and that a Grb2·Cbl chimera rescues receptor endocytosis in cells depleted of endogenous Grb2 are the first to show a definitive role for Cbl ubiquitin ligase activity in this process. Moreover, our studies indicate that Grb2 functions as a critical intermediary for Cbl ubiquitin ligase activity in the internalization of soluble InlB·cMet and HGF·cMet complexes. Cbl has been shown to mediate ubiquitination of cMet in response to HGF (43) in a process that involves the phosphotyrosine 1003 residue in the juxtamembrane region of cMet (13, 14, 42, 74). The observation that internalization of an ubiquitination-deficient Y1003F-cMet mutant is unaffected implies that receptor ubiquitination may not be a critical determinant for receptor internalization. Rather, Cbl-mediated ubiquitination of cMet via phosphotyrosine 1003 is important for degradation of the internalized receptor (13). Because the Y1003F-cMet was weakly ubiquitinated, we cannot rule out the possibility that different levels of, or sites of cMet ubiquitination contribute to distinct steps for cMet endocytic trafficking. Alternatively, clathrin-dependent internalization and receptor degradation may be regulated by the indirect recruitment of Cbl via Grb2 to phosphotyrosine 1356 versus Cbl recruited directly to phosphotyrosine 1003.
In conclusion our data demonstrate a strict requirement for a specific subset of Grb2-mediated interactions for cMet internalization from the cell surface. The direct binding of Grb2 to cMet is essential for receptor internalization. Moreover, endocytosis of soluble InlB·cMet and HGF·cMet complexes is dependent on the E3 ubiquitin ligase activity of Cbl and does not involve the Grb2 binding partner Gab1. Thus, Grb2 functions as an important intermediary for linking ligand-activated cMet with the ubiquitin ligase activity of Cbl for receptor internalization via clathrin-coated pits.
|
FOOTNOTES |
|---|
The on-line version of this article (available at http://www.jbc.org) contains supplemental Figs. S1–S8.
1 To whom correspondence should be addressed: Dept. of Neuroscience and Cell Biology, University of Texas Medical Branch, Galveston, TX 77555-1074. Tel.: 409-772-2775; Fax: 409-747-1938; E-mail: laelferi{at}utmb.edu.
2 The abbreviations used are: RTK, receptor-tyrosine kinase; HGF, hepatocyte growth factor; cMet, HGF receptor; InlB, internalin B; Alexa-InlB, Alexa-labeled InlB; EGFR, epidermal growth factor (EGF) receptor; Tfn, transferrin; siRNA, small interfering RNA; wt, wild type; IP, immunoprecipitation; PBS, phosphate-buffered saline; GST, glutathione S-transferase; FACS, fluorescence-activated cell sorter; MFI, mean fluorescence intensity; EYFP, enhanced yellow fluorescent (YFP) protein; KinD-cMet, kinasedead cMet; UbR, mutant ubiquitin.
3 N. Li, M. Lorinczi, K. Ireton, and L. A. Elferink, unpublished data.
|
ACKNOWLEDGMENTS |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
