The PTB domain of ShcA couples receptor activation to the cytoskeletal regulator IQGAP1
2010; Springer Nature; Volume: 29; Issue: 5 Linguagem: Inglês
10.1038/emboj.2009.399
ISSN1460-2075
AutoresMatthew J. Smith, W. Rod Hardy, Guangyao Li, Marilyn Goudreault, Steven J. Hersch, Pavel Metalnikov, Andrei Starostine, Tony Pawson, Mitsuhiko Ikura,
Tópico(s)Cell Adhesion Molecules Research
ResumoArticle14 January 2010free access The PTB domain of ShcA couples receptor activation to the cytoskeletal regulator IQGAP1 Matthew J Smith Matthew J Smith Division of Signaling Biology, Ontario Cancer Institute, MaRS TMDT, Toronto, Ontario, Canada Search for more papers by this author W Rod Hardy W Rod Hardy Samuel Lunenfeld Research Institute, Mount Sinai Hospital, Toronto, Ontario, Canada Search for more papers by this author Guang-Yao Li Guang-Yao Li Division of Signaling Biology, Ontario Cancer Institute, MaRS TMDT, Toronto, Ontario, Canada Search for more papers by this author Marilyn Goudreault Marilyn Goudreault Samuel Lunenfeld Research Institute, Mount Sinai Hospital, Toronto, Ontario, Canada Search for more papers by this author Steven Hersch Steven Hersch Division of Signaling Biology, Ontario Cancer Institute, MaRS TMDT, Toronto, Ontario, Canada Search for more papers by this author Pavel Metalnikov Pavel Metalnikov Samuel Lunenfeld Research Institute, Mount Sinai Hospital, Toronto, Ontario, Canada Search for more papers by this author Andrei Starostine Andrei Starostine Samuel Lunenfeld Research Institute, Mount Sinai Hospital, Toronto, Ontario, Canada Search for more papers by this author Tony Pawson Corresponding Author Tony Pawson Samuel Lunenfeld Research Institute, Mount Sinai Hospital, Toronto, Ontario, Canada Department of Molecular and Medical Genetics, University of Toronto, Toronto, Ontario, Canada Search for more papers by this author Mitsuhiko Ikura Corresponding Author Mitsuhiko Ikura Division of Signaling Biology, Ontario Cancer Institute, MaRS TMDT, Toronto, Ontario, Canada Department of Medical Biophysics, University of Toronto, Toronto, Ontario, Canada Search for more papers by this author Matthew J Smith Matthew J Smith Division of Signaling Biology, Ontario Cancer Institute, MaRS TMDT, Toronto, Ontario, Canada Search for more papers by this author W Rod Hardy W Rod Hardy Samuel Lunenfeld Research Institute, Mount Sinai Hospital, Toronto, Ontario, Canada Search for more papers by this author Guang-Yao Li Guang-Yao Li Division of Signaling Biology, Ontario Cancer Institute, MaRS TMDT, Toronto, Ontario, Canada Search for more papers by this author Marilyn Goudreault Marilyn Goudreault Samuel Lunenfeld Research Institute, Mount Sinai Hospital, Toronto, Ontario, Canada Search for more papers by this author Steven Hersch Steven Hersch Division of Signaling Biology, Ontario Cancer Institute, MaRS TMDT, Toronto, Ontario, Canada Search for more papers by this author Pavel Metalnikov Pavel Metalnikov Samuel Lunenfeld Research Institute, Mount Sinai Hospital, Toronto, Ontario, Canada Search for more papers by this author Andrei Starostine Andrei Starostine Samuel Lunenfeld Research Institute, Mount Sinai Hospital, Toronto, Ontario, Canada Search for more papers by this author Tony Pawson Corresponding Author Tony Pawson Samuel Lunenfeld Research Institute, Mount Sinai Hospital, Toronto, Ontario, Canada Department of Molecular and Medical Genetics, University of Toronto, Toronto, Ontario, Canada Search for more papers by this author Mitsuhiko Ikura Corresponding Author Mitsuhiko Ikura Division of Signaling Biology, Ontario Cancer Institute, MaRS TMDT, Toronto, Ontario, Canada Department of Medical Biophysics, University of Toronto, Toronto, Ontario, Canada Search for more papers by this author Author Information Matthew J Smith1, W Rod Hardy2, Guang-Yao Li1, Marilyn Goudreault2, Steven Hersch1, Pavel Metalnikov2, Andrei Starostine2, Tony Pawson 2,3 and Mitsuhiko Ikura 1,4 1Division of Signaling Biology, Ontario Cancer Institute, MaRS TMDT, Toronto, Ontario, Canada 2Samuel Lunenfeld Research Institute, Mount Sinai Hospital, Toronto, Ontario, Canada 3Department of Molecular and Medical Genetics, University of Toronto, Toronto, Ontario, Canada 4Department of Medical Biophysics, University of Toronto, Toronto, Ontario, Canada *Corresponding authors: Department of Medical Biophysics and Division of Signaling Biology, University of Toronto and Ontario Cancer Institute, MaRS TMDT, 101 College Street, Toronto, Ontario, Canada M5G 1L7. Tel.: +14165817550; Fax: +14165817564; E-mail: [email protected] of Molecular and Medical Genetics, University of Toronto and Samuel Lunenfeld Research Institute, Mount Sinai Hospital, 600 University Avenue, Toronto, Ontario Canada, M5G 1X5. Tel.: +14165868262; Fax: +14165868869; E-mail: [email protected] The EMBO Journal (2010)29:884-896https://doi.org/10.1038/emboj.2009.399 PDFDownload PDF of article text and main figures. Peer ReviewDownload a summary of the editorial decision process including editorial decision letters, reviewer comments and author responses to feedback. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info Adaptor proteins respond to stimuli and recruit downstream complexes using interactions conferred by associated protein domains and linear motifs. The ShcA adaptor contains two phosphotyrosine recognition modules responsible for binding activated receptors, resulting in the subsequent recruitment of Grb2 and activation of Ras/MAPK. However, there is evidence that Grb2-independent signalling from ShcA has an important role in development. Using mass spectrometry, we identified the multidomain scaffold IQGAP1 as a ShcA-interacting protein. IQGAP1 and ShcA co-precipitate and are co-recruited to membrane ruffles induced by activated receptors of the ErbB family, and a reduction in ShcA protein levels inhibits the formation of lamellipodia. We used NMR to characterize a direct, non-canonical ShcA PTB domain interaction with a helical fragment from the IQGAP1 N-terminal region that is pTyr-independent. This interaction is mutually exclusive with binding to a more conventional PTB domain peptide ligand from PTP–PEST. ShcA-mediated recruitment of IQGAP1 may have an important role in cytoskeletal reorganization downstream of activated receptors at the cell surface. Introduction The Shc proteins are a family of adaptors containing N-terminal PTB and C-terminal SH2 domains (Ravichandran, 2001), each of which bind phosphorylated tyrosine (pTyr) residues in the context of specific amino acid sequences. ShcA is the most widely expressed of four mammalian Shc proteins, and its primary role is in signalling from activated receptors at the cell surface. These include receptors for growth factors (Ravichandran et al, 1995; Ricketts et al, 1996; Ganju et al, 1998; Meakin et al, 1999; Hennige et al, 2000; Finlayson et al, 2003; Motegi et al, 2004), antigens (Ravichandran et al, 1993; Pratt et al, 1999; Patrussi et al, 2005; Fukushima et al, 2006), and cytokines (Dorsch et al, 1994; Bates et al, 1998; Hunt et al, 1999), as well as integrins (Mainiero et al, 1995; Wary et al, 1996; Mauro et al, 1999; Cowan et al, 2000; Dans et al, 2001; Weyts et al, 2002) and GPCRs (Luttrell et al, 1997; Schafer et al, 2004; Natarajan and Berk, 2006). Recruitment of ShcA to assorted membrane proteins results in its phosphorylation on three tyrosine residues (Y239/40 and Y313) in a central CH1 domain that subsequently activate the Ras/MAPK pathway through the recruitment of Grb2 and SOS (Salcini et al, 1994; van der Geer et al, 1996). PTB domain interactions typically mediate ShcA association with surface receptors. Selectivity is achieved through recognition of linear motifs with the consensus sequence ΦXNPXpY (Φ being a hydrophobic residue and X representing any amino acid) (Uhlik et al, 2005). A structure of the ShcA PTB domain complexed with a 12 residue phosphopeptide from the juxtamembrane of TrkA was solved by NMR spectroscopy (Zhou et al, 1995b). PTB domains, now identified in over 50 human proteins, share a similar core fold consisting of a central β-sandwich capped on one end by a conserved C-terminal α-helix, and on the other by a variable length α-helix found between strands β1 and β2 (or β2 and β3). The canonical peptide-binding groove is located between the fifth β-strand and the C-terminal α-helix (α3 in Shc); however, PTB domains possess surprisingly flexible binding properties including a variable dependence on phosphorylation of the NPXY tyrosine (Uhlik et al, 2005). Numerous PTB domains (i.e. Dab or Numb) bind irrespective of phosphorylation or preferentially recognize unphosphorylated ligands, but members of the Shc and IRS-1/Dok families bind with higher affinity to phosphorylated motifs and serve as adaptors in normal and oncogenic receptor tyrosine kinase (RTK) signalling. Although stimulation of the Ras pathway is a major component of Shc function, there are reasons to believe that Shc is involved in diverse processes. Though ShcA is an evolutionarily conserved protein, there is an intriguing absence of Y239/40 and Y313 tyrosines in more ancient orthologs. None of the three Shc-like proteins in nematodes contain these sites, and Drosophila Shc (dShc) has Y239/40 but not Y313; it is only in the chordate lineage that these are fixed. As Y313 appears to be the more potent Ras/MAPK activator, the absence of this phosphorylation site could limit the ability of dShc to stimulate the Ras/ERK pathway (Lai et al, 1995; Velazquez et al, 2000). These data imply a basic function for ShcA independent of its tyrosine phosphorylation, association with Grb2, or activation of MAPK. In support of this idea, ShcA has been reported to regulate c-Myc expression (Gotoh et al, 1997), play a role in cell survival (Gotoh et al, 1996), and control aspects of cytoskeletal architecture (Khoury et al, 2001). Mice lacking ShcA die at E11.5 with cardiovascular defects, including lack of cardiac trabeculation, deficiencies in angiogenesis, and maintenance between endothelial and mesenchymal cell contacts (Lai and Pawson, 2000). Whole mount immunostaining also revealed a loss of MAPK activity using pERK antibodies. However, recent work has shown that a significant fraction of transgenic mice expressing ShcA proteins in which Y239/40 and Y313 have been replaced with phenylalanine (henceforth referred to as ShcA3F) are viable (Hardy et al, 2007). Though ShcA3F/3F mice that live to adulthood exhibit severe ataxia, this is a much milder phenotype than that observed for mice expressing ShcA lacking PTB domain function (E11.5 lethal, similar to ShcA−/−). These results support the notion that the central pTyr motifs in ShcA are dispensable for some of its biological activity, and hint that ShcA may have entirely independent signalling properties. To search for unidentified pathways in the ShcA signalling network, we took a mass spectrometry-based approach to isolate new binding proteins. In this paper we show that ShcA interacts with the versatile scaffolding protein IQGAP1 (IQ-motif containing with homology to RasGAPs). This interaction is mediated through the ShcA PTB domain via a non-canonical binding mode, and we provide evidence that ShcA and IQGAP1 function downstream of activated receptors. Results Identification of novel ShcA-binding proteins To begin exploring alternative pathways of ShcA signalling, we used a mass spectrometry-based approach to reveal unidentified binding partners. As bait, Flag-tagged wild-type p52 ShcA or ShcA3F were stably expressed in Rat1 fibroblasts (Figure 1A). The 3F protein carries Tyr-to-Phe mutations in the three CH1 domain residues that are targets for Grb2 (Y239/40 and Y313). We co-expressed in these cells an activated RTK ErbB2-YD, which carries a V664E mutation in the transmembrane region that increases homodimerization (Bargmann and Weinberg, 1988), and has only the ShcA PTB domain-binding site in the C-terminal tail (as defined by Dankort et al, 1997). This induced a transformed phenotype to these fibroblasts (Figure 1B). Proteins bound to ShcA after immunoprecipitation were separated by SDS–PAGE and stained with colloidal coomassie (Figure 1C). Cells expressing ErbB2-YD alone were used as a control. Proteins that associated specifically with both wild type and ShcA3F were excised and identified by tandem mass spectrometry (LC-MS–MS). A protein of 197 kDa, equally visible in precipitations of both mutant and wild-type ShcA, was identified as IQGAP1. No IQGAP1 peptides were found from a corresponding region of the control lane. IQGAP1 is known to affect MEK/ERK activation (Roy et al, 2005) and has been linked to growth factor signalling and cytoskeletal rearrangements (Yamaoka-Tojo et al, 2004; Bensenor et al, 2007). Thus, we sought to further confirm this interaction and determine whether it was significant to ShcA signalling. Figure 1.Identification and verification of IQGAP1 as an ShcA-binding protein. (A) Schematic diagram of ShcA proteins used as bait for mass spectrometry-based analysis. p52 isoform of ShcA was tagged with a Flag-GFP sequence at its N-terminus. (B) Rat1 fibroblasts containing activated ErbB2 exhibit characteristics of transformation, and were used to produce recombinant ShcA proteins. Cells stably expressing ErbB2 (right) no longer display contact inhibition like wild-type cells (left). (C) Proteins interacting with ShcA were identified by mass spectrometry. Immunoprecipitation of ShcA (lane 2) or ShcA3F (lane 3) with anti-Flag antibodies was used to isolate bound proteins. Cells expressing only ErbB2 were used as control (lane 1). Precipitated proteins were detected with colloidal coomassie. Position of 197 kDa IQGAP1 is marked. (D) Endogenous IQGAP1 and ShcA co-precipitate in Rat1 cells. Anti-Shc antibodies were used to immunoprecipitate ShcA. Anti-RFP antibodies served as a control. Cells were grown in standard media, were stimulated with EGF, grown in the absence/presence of extracellular Ca2+, or were stably expressing ErbB2. Immunoblotting with anti-Shc or anti-IQGAP1 determined expression levels (bottom). IQGAP1 precipitated by ShcA or controls was detected with anti-IQGAP1 (top). (E) Recombinant IQGAP1 co-precipitates with wildtype and 3F ShcA. HA-tagged IQGAP1 was co-expressed with Flag-tagged ShcA in HEK 293T cells. Immunoprecipitation with anti-Flag antibodies and immunoblotting with anti-HA revealed co-precipitated IQGAP1 (top). Anti-Flag immunoblot confirmed ShcA protein expression (bottom). Download figure Download PowerPoint Verification of an ShcA interaction with IQGAP1 To validate the mass spectrometry result, we tested whether ShcA and IQGAP1 could be co-purified from Rat1 fibroblasts. We used wild-type Rat1 cells grown in standard media (10% serum), cells starved overnight and stimulated with EGF, or cells stably expressing ErbB2-YD. As several IQGAP1 interactions are mediated by Ca2+/calmodulin, we also grew cells in media either lacking or supplemented with Ca2+ (Figure 1D). After the immunoprecipitation of endogenous Shc, we observed co-precipitation of IQGAP1 from all conditions (Figure 1D). Subsequently, we tested the requirement for the CH1 domain residues Y239/40 and Y313 using HEK 293T cells co-transfected with HA-tagged IQGAP1 and Flag-tagged wild type or 3F ShcA. IQGAP1 was co-precipitated with both proteins (Figure 1E). These data corroborate an interaction, provide evidence that the Shc CH1 tyrosines are not required for association with IQGAP1, and indicate that ErbB2 activity is not strictly required for formation of the complex. Membrane recruitment of ShcA and IQGAP1 We next considered whether ShcA and IQGAP1 interact in cells. ShcA is generally cytoplasmic or ER localized, and is re-distributed upon receptor activation to the cell surface (Lotti et al, 1996). IQGAP1 is found in a Ca2+-regulated complex with calmodulin (Ho et al, 1999), but in multiple cell types can accumulate at the leading edge of migrating cells (Mataraza et al, 2003, 2007; Watanabe et al, 2004; Yamaoka-Tojo et al, 2004). We initially utilized the Rat1 cell lines expressing ErbB2-YD and EGFP/Flag-ShcA (Figure 1A and B). The ErbB2-YD receptor is able to signal downstream through ShcA, but unlike wild-type ErbB2 has no direct Grb2-binding sites (Dankort et al, 1997, 2001). Wild-type cells showed pools of endogenous IQGAP1 localized generally in the cytoplasm, as well as to the cell cortex in cells nearing the edge of a monolayer (Figure 2A). However, we observed a clear loss of cytoplasmic IQGAP1 and its substantial recruitment to membrane ruffles along the leading edges of cellular protrusions in those cells which express ErbB2-YD and EGFP-ShcA. Although we detected exogenously expressed EGFP-ShcA or EGFP-ShcA3F in these regions, they also remained prominent throughout the cytoplasm. In contrast to IQGAP1, a control protein that is also cytoplasmic in unstimulated cells (GAPDH) was not significantly recruited to membrane ruffles (Figure 2B). As ErbB2-YD signals through ShcA, these data signified that receptor activation may promote a ShcA interaction with IQGAP1 that is not dependant on the phosphorylation of its CH1 tyrosines. Figure 2.ShcA and IQGAP1 localization in Rat1 fibroblasts. (A) Cells stably expressing ErbB2-YD and EGFP-ShcA (green) were stained for endogenous IQGAP1 (red). Wild-type cells (column 1) served as a control. Merged images (bottom panels) show overlapping regions in yellow. Membrane ruffles are clearly seen in phase images (top panels). Bar=10 μm. (B) A control cytoplasmic protein, GAPDH, still resides in the cytoplasm in cells expressing ErbB2-YD. Rat1 cells were stained with antibody against endogenous GAPDH (red). Both wild-type cells (column 1) and those expressing ErbB2-YD and EGFP-ShcA (green; column 2) show GAPDH localized throughout the cytoplasm. Bar=10 μm. (C) IQGAP1 is also localized to membrane regions with endogenous ShcA. Cells were stained with anti-IQGAP1 (red) and antibody against endogenous Shc (green). Shc and IQGAP1 are generally cytoplasmic in wildtype cells grown in standard media (column 1), or are localized to membrane ruffles induced by expression of ErbB2-YD (column 2) or stimulation with EGF (column 3). Merged images (bottom panels) show overlapping localizations in yellow. Bar=10 μm. Download figure Download PowerPoint To address the dependence of IQGAP1 localization on exogenously expressed ShcA, wild-type Rat1 cells were immunostained for the endogenous proteins (Figure 2C). Both IQGAP1 and Shc were generally cytoplasmic in unstimulated cells, particularly localized in the perinuclear region (with IQGAP1 also in the cell cortex in cells nearing the edge of a monolayer). We then examined cells expressing ErbB2-YD, and once more found IQGAP1 in membrane regions, wherein it co-localized with a fraction of Shc. The Shc proteins remained prominent throughout the cytoplasm. Finally, we monitored the effect of EGF on IQGAP1 and Shc localization in wild-type fibroblasts. After 20 min of stimulation, Shc was clearly detected in EGF-induced endosomes (see also Supplementary Figure S1A). It is interesting to note that IQGAP1 was not observed in these punctate vesicles and remained in the cytoplasm as well as concentrated at membrane ruffles. This evidence suggested that IQGAP1 membrane recruitment by activated receptors does not require the overexpression of ShcA, and that IQGAP1 does not interact with ShcA after receptor internalization. IQGAP1 has a role in regulating epithelial cell junctions that is distinct from its activity in fibroblasts (Kuroda et al, 1998). We therefore chose to examine ShcA and IQGAP1 in 5637 bladder carcinoma cells, which show moderate but stable cell–cell contacts (Supplementary Figure S1B). Although IQGAP1 showed a basolateral distribution in these cells (similar to E-cadherin or β-catenin), Shc was localized in the cytoplasm. Stimulation of these highly metastatic cells with EGF-induced rapid and extensive membrane ruffling around the edges of the monolayer. Immunostaining for IQGAP1 and ShcA revealed their presence in these lamellipodia, along with actin, after 5 min of EGF treatment (Figure 3A). Similar results were obtained in the A431 skin carcinoma cell line (Supplementary Figure S1C), but not with MDCK cells in which EGF-induced membrane ruffling was notably diminished or even absent (Supplementary Figure S2). Previous work has demonstrated similar dynamics upon EGF stimulation for fluorescently tagged Grb2 adaptor proteins (Sorkin et al, 2000), and to some extent ShcA (Sato et al, 2000). We consequently tested whether exogenously expressed, Venus-tagged IQGAP1 could be recruited to membranes upon EGF stimulation in 5637 cells. Indeed, although Grb2, ShcA, IQGAP1, and the Venus control are equally localized throughout the cytoplasm in unstimulated cells, only the Grb2, ShcA, and IQGAP1 proteins demonstrated clear membrane recruitment 5 min after the addition of EGF (Figure 3B). The Venus control remained cytoplasmic and displayed no or very little localization to the membrane. Moreover, we once more failed to observe IQGAP1 in the punctate endosomal vesicles that are evident in ShcA or Grb2 expressing cells following 20 min of EGF stimulation. This work corroborates our biochemical data, and establishes ShcA and IQGAP1 co-recruitment to membrane regions in both fibroblast and epithelial cell lines. Figure 3.EGFR activation results in ShcA and IQGAP1 recruitment to membrane ruffles in 5637 epithelial cells. (A) Cells were stained for endogenous IQGAP1 (red) or Shc (green), along with actin (blue). 5 min after EGF stimulation massive ruffling was observed along the border of 5637 monolayers (column 2), subsiding after 10–15 min (column 3). ShcA is observed in EGFR-containing endosomes at 20 min mark (column 4). Bar=10 μm. (B) Exogenously expressed ShcA, IQGAP1, and Grb2 show similar activity during the early stages of EGF stimulation. Venus-tagged proteins, as well as Venus alone, were expressed in 5637 cells, where they are localized throughout the cytoplasm and nucleus (apart from ShcA, which is not detected in the nucleus; column 1). Cells were EGF stimulated for 5 min (column 2) or 20 min (column 3). Grb2, Shc, and IQGAP1 all showed early membrane localization, whereas the Venus control protein did not. Only Grb2 and ShcA were detected in EGF-containing endosomes at 20 min. Bar=10 μm. Download figure Download PowerPoint To further explore the role of ShcA in recruiting IQGAP1 to the cell cortex, we examined cells with reduced levels of ShcA protein for changes in IQGAP1 localization. Clonal lines of 5637 cells stably expressing shRNA against all three isoforms of ShcA were established, along with control lines expressing a scrambled shRNA sequence (Figure 4A). These cells were subjected to a time course of EGF stimulation, and control cells exhibited strong lamellipodia formation 5 min post-induction, in a manner analogous to wild-type cells. However, cells with reduced levels of ShcA showed greatly diminished levels of membrane ruffling, in most cases presenting no perceptible lamellipodia (Figure 4B and Supplementary Figure S3A). To determine whether IQGAP1 was still being recruited to the outer cortex in ShcA-knockdown cells, we immunostained these cell lines with antibodies against endogenous IQGAP1. Though control cells clearly retained the capacity to recruit IQGAP1 to membrane ruffles, there was no detectable change in IQGAP1 localization in ShcA-reduced cells upon stimulation with EGF (Figure 4C). Even treatment with significantly higher concentrations of EGF (100 ng/ml) failed to induce strong membrane ruffling in ShcA-knockdown cells (Supplementary Figure S3B/C). This lack of lamellipodia formation in ShcA-depleted cells likely correlates with a decrease in Rac1 activity. The loss of IQGAP1 has been highly associated with reduced migration in numerous cell types, and it is a powerful effector of Rac1 (Mataraza et al, 2003; Noritake et al, 2004; Watanabe et al, 2004). These data support a role for ShcA in recruiting IQGAP1 to membrane proximal regions in growth factor-stimulated cells. Figure 4.Reduction in ShcA protein levels leads to a loss of EGF-stimulated lamellipodia formation and IQGAP1 membrane recruitment. (A) Vector-based shRNA sequences directed against all three isoforms of ShcA were stably expressed in 5637 cells. Lysates from wild-type cells, as well as expanded clonal lines expressing either 'scrambled' shRNA control or shRNA against ShcA were immunoblotted with anti-Shc antibodies to determine protein levels (bottom). Probing with anti-IQGAP1 (top) served as a loading control and to ensure its level expression. Two lines for both the control and ShcA knockdown are shown. (B) 5637 cells with reduced levels of ShcA do not exhibit strong membrane ruffling. Control and ShcA knockdown lines were stimulated with 10 ng/ml EGF for 0, 5, or 20 min. Phase contrast images demonstrate a clear induction of lamellipodia in control cells at 5 min (middle, arrows). Similar results were obtained on multiple clonal cell lines. Bar=40 μm. (C) Control and ShcA knockdown cells were stained for endogenous IQGAP1 (red) to establish its localization upon EGF stimulation. As with wild-type cells, 5 min after addition of EGF membrane ruffles containing IQGAP1 were observed in the control (top). These regions are clearly visible in DIC images and are marked by arrows. ShcA knockdown lines were highly deficient in lamellipodia formation and IQGAP1 was subsequently not recruited to the cell cortex (bottom). Bar=10 μm. Download figure Download PowerPoint Establishment of binding requirements for the ShcA–IQGAP1 complex Previous reports have shown that IQGAP1 is tyrosine phosphorylated upon stimulation with EGF, PDGF, or VEGF (Blagoev et al, 2004; Yamaoka-Tojo et al, 2004; Kratchmarova et al, 2005), suggesting that IQGAP1 could directly associate with the ShcA PTB or SH2 domain. To begin examining this, we first chose to study IQGAP1 phosphorylation and its relationship with ShcA. We have detected a basal IQGAP1 tyrosine phosphorylation using anti-pTyr antibodies (Figure 5A). To determine whether phosphorylation is further stimulated by RTKs, we monitored IQGAP1 pTyr levels in cells expressing ErbB2-NT (carrying the V664E activating mutation and all five C-terminal pTyr sites; as defined in Dankort et al (1997)). Not only was IQGAP1 phosphorylation significantly increased by ErbB2, but overexpression of ShcA could further augment this (Figure 5B). Moreover, there was an enhanced capacity for hyperphosphorylated IQGAP1 to co-precipitate ShcA compared to IQGAP1 with only basal levels of phosphorylation. These data support a relationship between IQGAP1 pTyr levels and its interaction with ShcA, but do not yet clarify whether the PTB or SH2 domains might mediate a direct interaction. Figure 5.IQGAP1 is tyrosine phosphorylated and interacts directly with the ShcA PTB domain. (A) IQGAP1 is tyrosine phosphorylated in HEK 293T cells. Endogenous IQGAP1 was precipitated with anti-IQGAP1, and immunoblotting with anti-pTyr revealed phosphorylation (bottom panel). Anti-IQGAP1 immunoblot confirmed protein level (top panel), and precipitation with anti-Myc served as control. (B) Overexpression of ShcA induces phosphorylation of IQGAP1 in cells expressing activated ErbB2. Flag-tagged ShcA was co-expressed in HEK 293T cells with RFP-tagged ErbB2-NT or RFP alone as control. pTyr levels were determined by anti-pTyr immunoblot after immunoprecipitation of endogenous IQGAP1 (IP: top). Western blotting with anti-IQGAP1 confirmed protein levels (IP: middle), and with anti-Shc to determine co-precipitation (IP: bottom). Immunoblots of cell lysates with anti-RFP confirmed expression of ErbB2 (Lysate: top), while anti-Flag verified ShcA expression (Lysate: bottom). (C) Both the SH2 and PTB domains of ShcA precipitate full-length IQGAP1 from cell lysates. Recombinant domains were expressed in Escherichia coli as GST-fusion proteins, along with GST as a control. A coomassie stained gel shows loading of the purified proteins (bottom). HA-tagged IQGAP1 was expressed in HEK 293T cells, and GST pull downs performed with purified domains bound to glutathione beads. Precipitated IQGAP1 was identified by anti-HA immunoblot (top). Cells were also stimulated with the phosphatase inhibitor pervanadate for 30 min to induce tyrosine phosphorylation (middle). (D) Schematic showing IQGAP1 truncations. Fragments of IQGAP1 lacking the C-terminus (ΔC), the N-terminus (ΔN), the IQ motifs (ΔIQ), or the WW domains (ΔWW) were cloned and expressed with HA tags. Amino acid numbers are indicated according to murine IQGAP1. (E) The PTB domain of ShcA directly interacts with the N-terminal region of IQGAP1. Constructs expressing full-length IQGAP1, or the truncations, were expressed in HEK 293T cells. Proteins were immunoprecipitated with anti-HA and transferred to nitrocellulose membrane. Probing with GST–SH2 domain and anti-GST antibodies revealed no direct interactions (bottom panels). The PTB domain bound all truncations except ΔN (top panels). Reprobing with anti-HA confirmed expression of the IQGAP1 fragments (top panels). Download figure Download PowerPoint To map the IQGAP1-binding site, we carried out a series of in vitro-binding experiments. GST-tagged ShcA PTB and SH2 domains were expressed and purified from Escherichia coli. Glutathione beads carrying the recombinant domains were incubated with cell lysates expressing HA-tagged IQGAP1, and co-precipitating proteins were identified by anti-HA immunoblot. Though both domains consistently pulled down IQGAP1, the PTB domain interaction was considerably more robust (Figure 5C). This experiment is complicated by the ability of IQGAP1 to oligomerize, as well as its affinity for actin and microtubules (Bashour et al, 1997; Fukata et al, 2002; Ren et al, 2005). We therefore sought to narrow the region of IQGAP1 targeted by ShcA using a set of truncati
Referência(s)