Spatially Defined EGF Receptor Activation Reveals an F-Actin-Dependent Phospho-Erk Signaling Complex

      Abstract

      We investigated the association of signaling proteins with epidermal growth factor (EGF) receptors (EGFR) using biotinylated EGF bound to streptavidin that is covalently coupled in an ordered array of micron-sized features on silicon surfaces. Using NIH-3T3 cells stably expressing EGFR, we observe concentration of fluorescently labeled receptors and stimulated tyrosine phosphorylation that are spatially confined to the regions of immobilized EGF and quantified by cross-correlation analysis. We observe recruitment of phosphorylated paxillin to activated EGFR at these patterned features, as well as β1-containing integrins that preferentially localize to more peripheral EGF features, as quantified by radial fluorescence analysis. In addition, we detect recruitment of EGFP-Ras, MEK, and phosphorylated Erk to patterned EGF in a process that depends on F-actin and phosphoinositides. These studies reveal and quantify the coformation of multiprotein EGFR signaling complexes at the plasma membrane in response to micropatterned growth factors.

      Introduction

      Epidermal growth factor receptor (EGFR) belongs to the family of receptor tyrosine kinases (RTK) that are key contributors to cancer progression in multiple cell types (
      • Blume-Jensen P.
      • Hunter T.
      Oncogenic kinase signalling.
      ). EGFR signaling is known to be an important regulator of cellular responses, including proliferation, migration, and apoptosis (
      • Lemmon M.A.J.S.
      • Schlessinger J.
      • Ferguson K.M.
      The EGFR family: not so prototypical receptor tyrosine kinases.
      ,
      • Schlessinger J.
      Receptor tyrosine kinases: legacy of the first two decades.
      ,
      • Pines G.
      • Köstler W.J.
      • Yarden Y.
      Oncogenic mutant forms of EGFR: lessons in signal transduction and targets for cancer therapy.
      ). Numerous studies have shown that overexpression and mutations in EGFRs can make them potent oncoproteins (
      • Gschwind A.
      • Fischer O.M.
      • Ullrich A.
      The discovery of receptor tyrosine kinases: targets for cancer therapy.
      ). EGF binding and dimerization of EGFR activates the cytoplasmic kinase domain with consequent transphosphorylation of tyrosine residues in this region of EGFR (
      • Lemmon M.A.J.S.
      • Schlessinger J.
      • Ferguson K.M.
      The EGFR family: not so prototypical receptor tyrosine kinases.
      ). These phosphorylated tyrosine residues act as docking sites for signaling adaptors, including Grb2, Shc, and enzymes such as phospholipase Cγ (PLCγ), thus linking EGFR to its downstream signaling partners. EGFR activation results in initiation of several signaling cascades, including those activating extracellular signal-regulated kinases (Erk) (
      • Lemmon M.A.J.S.
      • Schlessinger J.
      • Ferguson K.M.
      The EGFR family: not so prototypical receptor tyrosine kinases.
      ,
      • McKay M.M.
      • Morrison D.K.
      Integrating signals from RTKs to ERK/MAPK.
      ) via Ras, as well as Akt signaling through phosphatidylinositol 3-kinase (PI3K) activation. EGF binding also causes internalization of EGFR, which further regulates receptor signaling capacity (
      • Sorkin A.
      • Goh L.K.
      Endocytosis and intracellular trafficking of ErbBs.
      ).
      Although the protein participants and signaling sequelae are well established, less is understood about the structural organization of signaling partner interactions with activated EGFR. Previous results using beads conjugated with EGF have shown that localized stimulation of EGFR can cause lateral propagation of EGFR activation at the plasma membrane at a length scale of several microns (
      • Verveer P.J.
      • Wouters F.S.
      • Bastiaens P.I.H.
      • et al.
      Quantitative imaging of lateral ErbB1 receptor signal propagation in the plasma membrane.
      ). Effects of RTK localization on signaling consequences have been a focus of some recent studies. Evidence that EGFRs can continue signaling from endosomes has been described, and clathrin-mediated endocytosis of EGFR has been suggested to play a role in the activation of Erk (
      • Sigismund S.
      • Argenzio E.
      • Di Fiore P.P.
      • et al.
      Clathrin-mediated internalization is essential for sustained EGFR signaling but dispensable for degradation.
      ,
      • Taub N.
      • Teis D.
      • Huber L.A.
      • et al.
      Late endosomal traffic of the epidermal growth factor receptor ensures spatial and temporal fidelity of mitogen-activated protein kinase signaling.
      ). In contrast, other studies revealed that a reduction in EGF-stimulated endocytosis of EGFR leads to more sustained activation of Akt, suggesting that EGFR signaling occurs primarily at the plasma membrane (
      • Sousa L.P.
      • Lax I.
      • Schlessinger J.
      • et al.
      Suppression of EGFR endocytosis by dynamin depletion reveals that EGFR signaling occurs primarily at the plasma membrane.
      ). Recent studies in prostate cancer cells provided evidence that an adaptor protein, paxillin, participates in Erk activation and its subsequent nuclear translocation downstream of EGFR signaling (
      • Sen A.
      • O’Malley K.
      • Hammes S.R.
      • et al.
      Paxillin regulates androgen- and epidermal growth factor-induced MAPK signaling and cell proliferation in prostate cancer cells.
      ,
      • Sen A.
      • De Castro I.
      • Hammes S.R.
      • et al.
      Paxillin mediates extranuclear and intranuclear signaling in prostate cancer proliferation.
      ). Physical association of activated EGFR with paxillin and other downstream signaling partners, including MEK and Erk, have not been described for intact cells. Although F-actin association with activated EGFR was previously described (
      • den Hartigh J.C.
      • van Bergen en Henegouwen P.M.
      • Boonstra J.
      • et al.
      The EGF receptor is an actin-binding protein.
      ,
      • Rijken P.J.
      • Post S.M.
      • Boonstra J.
      • et al.
      Actin polymerization localizes to the activated epidermal growth factor receptor in the plasma membrane, independent of the cytosolic free calcium transient.
      ), the role of F-actin in the recruitment of signaling partners to activated EGFR at the plasma membrane has not been defined.
      We previously demonstrated that micropatterned ligand surfaces are useful tools to study the spatiotemporal aspects of FcεRI signaling (
      • Wu M.
      • Holowka D.
      • Baird B.
      • et al.
      Visualization of plasma membrane compartmentalization with patterned lipid bilayers.
      ,
      • Torres A.J.
      • Vasudevan L.
      • Baird B.A.
      • et al.
      Focal adhesion proteins connect IgE receptors to the cytoskeleton as revealed by micropatterned ligand arrays.
      ). We showed that FcεRI bound to anti-DNP-IgE on the surface of mast cells is recruited and activated by spatially defined patterns of DNP-presenting features. This activation leads to the corecruitment of Lyn, a Src family tyrosine kinase that initiates the phosphorylation of FcεRI. In addition, we found that micron-scale recruitment of Lyn to FcεRI complexes requires polymerization of actin, although this is not necessary for sufficient recruitment of Lyn to initiate FcεRI phosphorylation (
      • Wu M.
      • Holowka D.
      • Baird B.
      • et al.
      Visualization of plasma membrane compartmentalization with patterned lipid bilayers.
      ,
      • Torres A.J.
      • Vasudevan L.
      • Baird B.A.
      • et al.
      Focal adhesion proteins connect IgE receptors to the cytoskeleton as revealed by micropatterned ligand arrays.
      ). Previous studies have shown that EGF attached to surfaces can stimulate EGFR, as detected by tyrosine phosphorylation (
      • Gonçalves R.
      • Martins M.C.L.
      • Barbosa M.A.
      • et al.
      Bioactivity of immobilized EGF on self-assembled monolayers: optimization of the immobilization process.
      ,
      • Ito Y.
      • Chen G.
      • Imanishi Y.
      Micropatterned immobilization of epidermal growth factor to regulate cell function.
      ,
      • Stabley D.
      • Retterer S.
      • Salaita K.
      • et al.
      Manipulating the lateral diffusion of surface-anchored EGF demonstrates that receptor clustering modulates phosphorylation levels.
      ).
      Here, we use NIH-3T3 cells stably expressing EGFR and micron-sized features of surface-attached EGF to locally activate EGFR in well-defined geometrical arrays. Under these conditions, we find that Ras, as well as MAP kinase proteins MEK and phospho-Erk (pErk), are recruited to the pattern-localized EGFR signaling complexes at the plasma membrane in an actin polymerization-dependent manner. Furthermore, we find that paxillin is recruited to these activated EGFR signaling complexes as tyrosine- and serine-phosphorylated species. We find that F-actin colocalizes with these EGFR signaling complexes in a PI(4,5)P2-dependent manner that also correlates with the preferential localization of β1 integrin to more peripheral sites of cell engagement with EGF. Our results provide strong evidence that EGFR signaling complexes are established by surface-attached EGF, and that these complexes are stabilized by coupling to the actin cytoskeleton.

      Materials and Methods

      All cell culture reagents, EGF and N-terminal-labeled EGF-biotin, Lipofectamine 2000, and precast gels for blotting were from Invitrogen (Carlsbad, CA), as were Alexa488-labeled goat antimouse IgG1, Alexa488-labeled goat antirabbit (H+L), Alexa488-labeled donkey antirat IgG (H+L), Alexa488- and Alexa647-phalloidin, Alexa647-labeled goat antimouse IgG (H+L), Alexa633-labeled goat antimouse IgG2b (γ2b), and Alexa568-labeled streptavidin. FuGene HD was from Roche Applied Sciences (Indianapolis, IN). Quercetin and phenylarsine oxide (PAO) were purchased from Sigma-Aldrich (St. Louis, MO). PP2 was purchased from Enzo LifeSciences (Farmingdale, NY). Cytochalasin D, Iressa, and rapamycin were purchased from Calbiochem (EMD Chemicals, San Diego, CA). Rabbit anti-pErk and anti-phosphotyrosine 1068-EGFR antibodies were purchased from Cell Signaling Technology (Danvers, MA). Rabbit anti-MEK antibody was from Dr. R. Cerione (Cornell University). Mouse monoclonal anti-phosphotyrosine, clone 4G10, FITC-labeled rat anti-β1 integrin, and rabbit anti-C-terminal EGFR were obtained from Millipore (Temecula, CA). Anti-phosphotyrosine paxillin was purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Dynamin 2-EGFP cDNA was obtained from Dr. M. McNiven (The Mayo Clinic). Rabbit anti-ezrin and anti-moesin antibodies were from Dr. A. Bretscher (Cornell University). Avian paxillin-EGFP cDNA was from Dr. A. Horwitz (University of Virginia). EGFR-EGFP cDNA was from Dr. J. Koland (University of Iowa) (
      • Monick M.M.
      • Cameron K.
      • Hunninghake G.W.
      • et al.
      Activation of the epidermal growth factor receptor by respiratory syncytial virus results in increased inflammation and delayed apoptosis.
      ). Preparation of Lyn-mRFP cDNA was described previously (
      • Hammond S.
      • Wagenknecht-Wiesner A.
      • Baird B.
      • et al.
      Roles for SH2 and SH3 domains in Lyn kinase association with activated FcepsilonRI in RBL mast cells revealed by patterned surface analysis.
      ). The cDNA construct of PLCγ1-EGFP (
      • Wang X.J.
      • Liao H.J.
      • Carpenter G.
      • et al.
      EGF-dependent translocation of green fluorescent protein-tagged PLC-gamma1 to the plasma membrane and endosomes.
      ) was from Dr. G. Carpenter (Vanderbilt University). EGFP-H-Ras was from Dr. A. Kenworthy (Vanderbilt University). Plasma membrane-targeted CFP-FRB domain (PM-CFP-FRB) and mRFP-tagged FKBP12- phosphoinositide 5-phosphatase (mRFP-FKBP12-5ptase) cDNA (
      • Varnai P.
      • Thyagarajan B.
      • Balla T.
      • et al.
      Rapidly inducible changes in phosphatidylinositol 4,5-bisphosphate levels influence multiple regulatory functions of the lipid in intact living cells.
      ) were obtained from Dr. M. Korzeniowski (Cornell University). The mRFP sequence was excised using polymerase chain reaction.

      Cell culture and transfection

      NIH-3T3 cells stably overexpressing wild-type EGFR, hereafter referred to as NIH-3T3 (EGFR) cells, were cultured as monolayers in Dulbecco’s modified Eagle’s medium containing 10% (v/v) calf serum (
      • Schlessinger J.
      Receptor tyrosine kinases: legacy of the first two decades.
      ) as described elsewhere (
      • Bryant K.L.
      • Antonyak M.A.
      • Holowka D.
      • et al.
      Mutations in the polybasic juxtamembrane sequence of both plasma membrane- and endoplasmic reticulum-localized epidermal growth factor receptors confer ligand-independent cell transformation.
      ). Chemical transfection of cells was carried out using a complex of 2 μg cDNA with 8 μL Fugene HD (Promega, Madison, WI) in 100 μL Opti-MEM (Life Technologies, Grand Island, NY) that was added to plated cells in 1 mL Opti-MEM per MatTek (Ashland, MA) well. After 3–5 h, cells were returned to medium for overnight culture. For use in experiments, NIH-3T3 (EGFR) cells were serum starved for 12–14 h before harvesting.

      Microfabrication of patterned EGF surfaces

      Surfaces with 1.5 to 4 μm features were patterned with a parylene layer as described previously (
      • Ilic B.
      • Craighead H.
      Topographical patterning of chemically sensitive biological materials using a polymer-based dry lift off.
      ,
      • Orth R.
      • Wu M.
      • Baird B.
      • et al.
      Mast cell activation on patterned lipid bilayers of subcellular dimensions.
      ). An 8 × 8 mm parylene-patterned silicon substrate was prepared for functionalization by plasma cleaning for 5 min at room temperature under vacuum in a glass dish. The surface was rinsed with acetone and dried under nitrogen, followed by treatment with 3-mercaptopropyltrimethyoxysilane for 30 min at room temperature. The substrate was then washed with absolute ethanol (3×) and incubated in 2 mM N-(γ-maleimidobutyryloxy)succinimide in absolute ethanol for 1 h at room temperature (
      • Torres A.J.
      • Holowka D.
      • Baird B.A.
      Micropatterned ligand arrays to study spatial regulation in Fc receptor signaling.
      ). The samples were rinsed with phosphate buffered saline (PBS, pH 7.4) and placed on parafilm, followed by incubation with 50 μg/mL Alexa568-streptavidin for 1 h at room temperature. After washing with PBS, samples were incubated with 500 ng/mL EGF-biotin for 30 min at room temperature and rinsed thoroughly with PBS. The parylene was then mechanically peeled away to yield patterned EGF on the silicon substrate. For negative controls, incubation with EGF-biotin was omitted. Nonspecific binding to the micropatterned substrate was minimized by incubation with 1 mg/mL bovine serum albumin (BSA) in PBS for 5 min before plating cells for experiments.

      Fluorescence microscopy

      For experiments involving patterned EGF surfaces, ∼100 μL of cells suspended at a concentration of 0.5 × 106 cells per mL in buffered salt solution (BSS: 135 mM NaCl, 5.0 mM KCl, 1.8 mM CaCl2, 1.0 mM MgCl2, 5.6 mM glucose, and 20 mM Hepes pH 7.4) with 1 mg/ml BSA were added to a patterned substrate (∼8 × 8 mm) in the center of a 35-mm petri dish with a coverglass insert (0.16–0.19 mm; MatTek). After 40 min of incubation at 37°C (or as otherwise indicated), cells were fixed with 4% paraformaldehyde in PBS for 20 min at room temperature followed by quenching with 10 mg/ml BSA in PBS with 0.01% NaN3 (PBS/BSA). For immunofluorescence after fixation, cells were labeled with a primary antibody at room temperature for 1 h in presence of 0.1% Triton-X100 in PBS/BSA. After washing with PBS/BSA, a fluorophore-labeled secondary antibody was incubated with samples at room temperature for 1 h in PBS/BSA. For labeling of F-actin, fixed cells were incubated with 5 μg/mL Alexa488-phalloidin or Alexa647-phalloidin in PBS-BSA with 0.1% Triton-X for 30 min at room temperature before rinsing with PBS.
      Labeled silicon chips with adherent fixed cells were inverted in a coverslip dish for imaging on a Zeiss LSM 710 inverted confocal microscope and a 63× Oil Plan-Apochromat objective. A DF 488/561/647 filter set was used to perform sequential 1/2/3 color imaging of the samples. The width of the focal plane was adjusted for optimal image quality. For experiments in which the phosphoinositide-5-phosphatase was recruited to the plasma membrane, rapamycin (100 nM) was added to NIH-3T3 (EGFR) cells coexpressing PM-CFP-FRB and FKBP12-5ptase just before allowing them to settle on EGF patterns and were fixed as described previously.

      Quantification of colocalization

      • 1)
        Pearson’s cross-correlation coefficients were determined on a cell-by-cell basis using a MATLAB (The MathWorks, Natick, MA) code as described elsewhere (
        • Torres A.J.
        • Vasudevan L.
        • Baird B.A.
        • et al.
        Focal adhesion proteins connect IgE receptors to the cytoskeleton as revealed by micropatterned ligand arrays.
        ). The equation used to calculate Pearson’s cross-correlation coefficient (ρ) is
      ρ=(xix)(yiy)(xix)2(yiy)2,


      where, xi and yi are the pixel intensity values for the two color channels being compared, and x and y are the average values of xi and yi in the respective channels. A mask is drawn around the edge of the cell of interest using a MATLAB script, and ρ values are calculated for each individual cell. Statistical analyses were carried out by calculating the p values using Student’s t-test, and a value of p ≤ 0.05 was considered to be statistically significant.
      • 2)
        Radial analysis of fluorescence intensity, averaged over selected patterned features, was used in cases where the labeled cellular component was not uniformly distributed across the underlying features. In these studies, for example, patterned features at the cell edge were compared to features in the cell middle to evaluate differential distributions of labeled components. As depicted in Fig. S1 in the Supporting Material, a MATLAB code is used to identify patterned features in a selected region of interest (e.g., cell edge), based on the localization of fluorescent streptavidin. The fluorescence intensities of the labeled cellular response component (e.g., pTyr or β1 integrin) associated with features located in this same region are then grouped together as an average, representing the cellular response distribution in that region. Pixel intensity values radiating from the feature center are averaged to yield a radial plot, which delineates the feature edge and thereby specifies locations On versus Off the patterned feature. The ratio of averaged fluorescence intensity values, On Pattern:Off Pattern, is taken as a quantitative measure of the relative cellular response distribution in terms of a particular labeled component.
      This versatile analysis scheme offers a more detailed assessment of fluorescence for distinct regions of a cell. For relevant experiments in our studies, an average value for the cellular response distribution was determined separately for those patterned features contacting only the interior of the cell (Cell Middle), or for just those features located at the cell periphery (Cell Edge), or for both of these inner and outer regions combined (Entire Cell).
      For both of these methods, analysis depends on confocal imaging of labeled cell components at the ventral plasma membrane, and is largely independent of cell morphology in the vicinity of the patterned features. Occasionally some cells (<10%) appear markedly distorted because dynamic membrane extensions are anchored by membrane EGFR binding to the patterned EGF. Although in extreme cases this could enhance apparent values for colocalization of labeled components with patterned features, we expect this effect to be small for our comparisons of values averaged over many cells.

      Immunoblot analysis

      Cells were lysed in 25 mM Tris, 100 mM NaCl, 1 mM EDTA, 1% (v/v) Triton X-100, 1 mM dithiothreitol, 1 mM sodium orthovanadate, 1 mM β-glycerol phosphate, 1 μg/mL of leupeptin, and 1 μg/mL aprotinin (lysis buffer), and whole-cell lysates (WCL) were recovered in the supernatant following microfuge centrifugation. Protein concentrations of the WCL were determined using the Bio-Rad DC protein assay. WCL (40 μg) were resolved by sodium dodecyl sulfate polyacrylamide gel electrophoresis, and proteins were transferred to poly(vinylidene fluoride) membranes. Membranes were blocked by 10% (w/v) BSA in 20 mM Tris, 135 mM NaCl, and 0.02% Tween 20 (TBST) and incubated with the indicated primary antibodies diluted in TBST. The primary antibodies were detected with horseradish peroxidase-conjugated antimouse IgG (GE Healthcare, Buckinghamshire, UK) diluted in TBST followed by exposure to enhanced chemiluminescence (ECL) reagent (Invitrogen).

      Results

      Cellular EGFR is activated on micropatterned EGF surfaces

      Micropatterned surfaces, with A568-streptavidin covalently attached to silicon substrates, were prepared as described in Materials and Methods. Fig. 1 A schematically depicts the preparation of micropatterned surfaces with the parylene lift-off method and association of cells with these surfaces. NIH-3T3 cells endogenously express a small amount of EGFR that is difficult to detect by immunocytochemistry (A. Singhai, unpublished results). To visualize EGFR more clearly, we used NIH-3T3 cells stably overexpressing EGFR (NIH-3T3 (EGFR) cells), which is localized primarily at the plasma membrane in these cells (
      • Bryant K.L.
      • Antonyak M.A.
      • Holowka D.
      • et al.
      Mutations in the polybasic juxtamembrane sequence of both plasma membrane- and endoplasmic reticulum-localized epidermal growth factor receptors confer ligand-independent cell transformation.
      ). As shown in Fig. 1 B, green channel, we find that EGFR, labeled with an anti-EGFR, is maximally recruited to the EGF patches within 40 min at 37°C. Similar results are obtained with RBL-2H3 cells transiently transfected with EGFP-tagged EGFR (A. Singhai, unpublished results). Under these conditions, we detect robust tyrosine phosphorylation colocalized at patterned features with clustered EGFR, most likely due to EGFR kinase activity (Fig. 1 B, blue channel and overlay). This receptor recruitment and tyrosine phosphorylation can be detected as early as 10 min after plating in >90% of the imaged cells, with increasing intensity at times up to 40 min after plating at 37°C (A. Singhai, unpublished results). This time course for binding and activation appears slower than observed with an optimal dose of soluble EGF and Western blot detection (e.g., see Fig. 6). This apparent kinetic difference may be due in part to more limited accessibility or density of the pattern-associated EGF, but it is primarily limited by the time of cell settling on the patterns, as previously determined for FcεRI engagement of liganded patterns by RBL mast cells (
      • Wu M.
      • Holowka D.
      • Baird B.
      • et al.
      Visualization of plasma membrane compartmentalization with patterned lipid bilayers.
      ). Incubation of the patterns with a 10-fold higher dose of EGF-biotin did not enhance these kinetics. Clustering of EGFR and enhanced tyrosine phosphorylation depend on the presence of EGF at the patterned sites (Fig. 1 B).
      Figure thumbnail gr1
      Figure 1EGFR and phosphotyrosine labeling concentrate with patterned EGF. (A) A schematic shows the fabrication of EGF patterned surfaces and subsequent incubation of cells on the patterned substrate. (B) NIH-3T3 cells stably overexpressing EGFR were plated onto micropatterned, covalently immobilized EGF (A568-streptavidin, red channel), fixed, and immunolabeled with anti-EGFR primary antibody and A488-tagged secondary antibody (green channel), together with an anti-phosphotyrosine primary antibody (4G10) and A647-tagged secondary antibody (blue channel). The bottom panel shows NIH-3T3 (EGFR) cells plated on micropatterned surfaces of covalently immobilized A568-streptavidin without EGF-biotin. (C) Quantification of colocalization from Pearson’s cross-correlation coefficient analysis compares the green channel (EGFR) with the red channel (streptavidin), and the blue channel (anti-phosphotyrosine) with the red channel, as described in Materials and Methods; N > 40 cells for three independent experiments. Error bars represent mean ± SE; ∗∗∗∗ indicates p ≤ 0.0001. Scale bar 20 μm. To see this figure in color, go online.
      Recruitment of labeled components to the micron-sized features can be quantified using Pearson’s cross-correlation analysis, which evaluates the spatial coincidence of two distinct labels (
      • Torres A.J.
      • Vasudevan L.
      • Baird B.A.
      • et al.
      Focal adhesion proteins connect IgE receptors to the cytoskeleton as revealed by micropatterned ligand arrays.
      ). Consistent with our qualitative observations in confocal images, we find that Pearson’s cross-correlation coefficient, ρ, is ∼0.7 for EGFR localized to EGF features, and ∼0.05 in the absence of EGF-biotin (Fig. 1 C). Similarly, ρ for tyrosine phosphorylation at the patterns is ∼0.8 in presence of EGF and ∼0.06 in its absence.
      We also detect phosphorylation of EGFR by labeling with an antibody specific for EGFR-phospho-tyrosine 1068 (Fig. S2 A). Iressa (gefitinib) is a potent inhibitor of the tyrosine kinase activity of EGFR (
      • Wakeling A.E.
      • Guy S.P.
      • Gibson K.H.
      • et al.
      ZD1839 (Iressa): an orally active inhibitor of epidermal growth factor signaling with potential for cancer therapy.
      ). When NIH-3T3 (EGFR) cells are serum-starved overnight and plated in the presence of 10 μM Iressa for 40 min at 37°C on patterned EGF substrates, there is little or no tyrosine phosphorylation at EGF features, even though EGFR is recruited to the EGF features under these conditions (Fig. S2, A and B). These observations are compiled together with others described in Table 1.
      Table 1Summary of observations
      Yes, indicates that proteins could be optically resolved as colocalizing with micropatterned EGF bound to EGFR; None, indicates no concentration of label at EGF patterns was observed; and –, indicates not tested.
      ProteinControl+PP2+Iressa+PAO or quercetin+Cyto D
      EGFRYesYesYesYesYes
      p-EGFRYesYesNoneYesYes
      p-TyrYesYesVery weakYesYes
      Paxillin-EGFPYesYesNoneNoneNone
      p-Tyr-PaxillinYesNone
      p-Ser-PaxillinYesYes-
      F-actinYesYesNoneNoneNone
      EzrinNone
      MoesinNone
      EGFP-H-RasYesYesNone
      MEKYesNone
      pErkYesNoneNone
      PLCγ1-EGFPYes
      β1 integrinYes
      Dynamin 2-EGFPYesNoneYes
      a Yes, indicates that proteins could be optically resolved as colocalizing with micropatterned EGF bound to EGFR; None, indicates no concentration of label at EGF patterns was observed; and –, indicates not tested.

      Paxillin is recruited to and phosphorylated at pattern-localized EGFR complexes

      Paxillin, a multidomain adaptor protein, is known to participate in integrin-mediated signaling as a part of focal adhesions (
      • Brown M.C.
      • Turner C.E.
      Paxillin: adapting to change.
      ), as well as in other cell signaling contexts (
      • Antoniades I.
      • Stylianou P.
      • Skourides P.A.
      Making the connection: ciliary adhesion complexes anchor basal bodies to the actin cytoskeleton.
      ). In this role, paxillin interacts with several actin cytoskeleton regulating proteins, including vinculin and talin (
      • Salgia R.
      • Li J.L.
      • Griffen J.D.
      • et al.
      Molecular cloning of human paxillin, a focal adhesion protein phosphorylated by P210BCR/ABL.
      ). We previously found that paxillin is recruited to clustered IgE receptor complexes, and it contributes to the regulation of IgE receptor signaling (
      • Torres A.J.
      • Vasudevan L.
      • Baird B.A.
      • et al.
      Focal adhesion proteins connect IgE receptors to the cytoskeleton as revealed by micropatterned ligand arrays.
      ). Prior studies showed paxillin plays a role in the activation of Erk due to EGFR stimulation in prostate cancer cells (
      • Sen A.
      • O’Malley K.
      • Hammes S.R.
      • et al.
      Paxillin regulates androgen- and epidermal growth factor-induced MAPK signaling and cell proliferation in prostate cancer cells.
      ,
      • Sen A.
      • De Castro I.
      • Hammes S.R.
      • et al.
      Paxillin mediates extranuclear and intranuclear signaling in prostate cancer proliferation.
      ). We examined the distribution of paxillin-EGFP in NIH-3T3 (EGFR) cells under conditions of EGFR clustering at the patterned features, and we observed that paxillin-EGFP is visibly concentrated at these features (Fig. 2 A, top panel). This concentration is not observed in the absence of EGF-biotin (Fig. 2 A, middle panel). Quantification of this colocalization using Pearson’s cross-correlation analysis gives a value of ∼0.25 for paxillin-EGFP colocalization with A568-streptavidin patches when EGF is present on the surface, a substantially higher value than when EGF is absent from the surface (ρ ∼0.05, Fig. 2 B). Furthermore, this recruitment of paxillin-EGFP is prevented by treatment of cells with Iressa, demonstrating its dependence on tyrosine kinase activation (Fig. 2 A, bottom panel, and Fig. 2 B).
      Figure thumbnail gr2
      Figure 2Paxillin is recruited to EGF patterns in a ligand-dependent manner. (A) NIH-3T3 (EGFR) cells transiently transfected with paxillin-EGFP (green channel), were plated on micropatterned, covalently attached A568-streptavidin (red channel) surfaces with EGF-biotin (top panel), or without EGF-biotin (middle panel). The bottom panel shows the NIH-3T3 (EGFR) cells transiently transfected with paxillin-EGFP and pretreated with 10 μM Iressa overnight, before incubation on micropatterned EGF surfaces. Scale bar 20 μm. (B) Quantification by Pearson’s cross-correlation coefficient compares the red and green channels; N > 20 cells for each sample in at least two independent experiments. Error bars represent mean ± SE; ∗∗∗∗ indicates p ≤ 0.0001, ∗∗∗ indicates p ≤ 0.001. To see this figure in color, go online.
      Paxillin has multiple sites of phosphorylation that mediate its binding to several different proteins (
      • Brown M.C.
      • Turner C.E.
      Paxillin: adapting to change.
      ). Paxillin is phosphorylated at Tyr-31 and Tyr-118 by Src kinase (
      • Schaller M.D.
      • Schaefer E.M.
      Multiple stimuli induce tyrosine phosphorylation of the Crk-binding sites of paxillin.
      ), and at Ser-83 (
      • Huang C.
      • Borchers C.H.
      • Jacobson K.
      • et al.
      Phosphorylation of paxillin by p38MAPK is involved in the neurite extension of PC-12 cells.
      ) and Ser-126 (
      • Woodrow M.A.
      • Woods D.
      • McMahon M.
      • et al.
      Ras-induced serine phosphorylation of the focal adhesion protein paxillin is mediated by the Raf—>MEK—>ERK pathway.
      ) in an Erk-dependent manner. We find that paxillin recruited to EGF patterns is phosphorylated at Tyr-118 (and possibly Tyr-31), identified by labeling with anti-paxillin phospho-Tyr-118 (Fig. S3 A). Paxillin recruited to clustered EGFR is also phosphorylated at Ser-126 as detected by labeling with an antibody specific for that phosphorylated residue. Quantification of these results using Pearson’s cross-correlation analysis shows statistically significant association of tyrosine phosphorylated paxillin (ρ ∼0.41) and serine phosphorylated paxillin (ρ ∼0.28) at EGF features (Fig. S3 B). As expected, tyrosine phosphorylation of recruited paxillin is inhibited by pretreatment with 20 μM PP2, a well-known Src kinase inhibitor. However, PP2 does not inhibit recruitment of paxillin-EGFP or paxillin phosphorylated at serine residues (Fig. S3 B). Together, these results indicate that paxillin is recruited to clustered EGFR at the plasma membrane in a ligand-dependent and EGFR kinase activity-dependent manner. Furthermore, recruited paxillin is phosphorylated on tyrosine residues, likely by Src kinase, but paxillin recruitment and serine phosphorylation do not depend on Src activity.

      F-actin is recruited together with β1 integrin to pattern-localized EGFR complexes

      Our previous study showed that F-actin colocalizes with IgE receptors (FcεRI) clustered in micron-scale patterns (
      • den Hartigh J.C.
      • van Bergen en Henegouwen P.M.
      • Boonstra J.
      • et al.
      The EGF receptor is an actin-binding protein.
      ). Similarly, we found that F-actin labeled with Alexa488-phalloidin in NIH-3T3 (EGFR) cells localizes at EGF features (Fig. 3 A). This concentration of F-actin can be observed as soon as 10 min after plating the cells and is maximal after 30 min at 37°C (A. Singhai, unpublished results). Inhibition of actin polymerization by 2 μM cytochalasin D during cell attachment substantially reduces this localized accumulation (Fig. 3 B and Fig. S4). We also found that Iressa inhibits stimulated F-actin recruitment (Fig. S5), consistent with its inhibition of tyrosine phosphorylation stimulated by the patterned EGF-EGFR (Fig. S2) We evaluated whether ezrin or moesin, which connect the actin cytoskeleton to the plasma membrane under some conditions (
      • Bretscher A.
      • Edwards K.
      • Fehon R.G.
      ERM proteins and merlin: integrators at the cell cortex.
      ), become concentrated with EGFR at EGF features. We detected no significant concentration of these proteins in three separate experiments (Fig. 3, A and B).
      Figure thumbnail gr3
      Figure 3F-actin coredistributes with EGFR complexes at the plasma membrane. (A) NIH-3T3 (EGFR) cells were plated onto micropatterned EGF surfaces (red channel) and immunolabeled with A488-phalloidin (green channel, top panel), anti-moesin (middle panel), or anti-ezrin (bottom panel). Scale bar 20 μm. (B) Quantification of colocalization by Pearson’s cross-correlation coefficient compares the green channel with the red channel; N > 20 cells per sample from two independent experiments. Error bars represent mean ± SE; ∗∗∗∗ indicates p ≤ 0.0001, ∗∗∗ indicates p ≤ 0.001. To see this figure in color, go online.
      Because paxillin and F-actin are often concentrated in focal adhesions nucleated by integrins (
      • Turner C.E.
      Paxillin and focal adhesion signalling.
      ), we examined whether integrins also localize to the EGFR signaling complexes. Integrin α5β1 is an abundant integrin in NIH-3T3 cells (
      • Woods D.
      • Cherwinski H.
      • McMahon M.
      • et al.
      Induction of β3-integrin gene expression by sustained activation of the Ras-regulated Raf-MEK-extracellular signal-regulated kinase signaling pathway.
      ), and we investigated the distribution of this integrin in NIH-3T3 (EGFR) cells using an anti-β1 mAb. As illustrated in Fig. 4 A, we find that this integrin localizes to EGFR clusters at the patterned surfaces, but preferentially at peripheral regions compared to the middle of adherent cells. In contrast, phosphotyrosine detected by mAb 4G10 concentrates similarly at patterned EGF in the middle and periphery of attached cells (Fig. 4 A). Interestingly, β1 clustering at patterned features often appears more asymmetric than that for EGFR or phosphotyrosine (see Fig. 4 A and Fig. S1). We sought to quantify this difference in distribution with respect to the patterned features and to distinguish different regions of the cell. We chose to analyze averaged fluorescence distributions according to grouped features using a radial analysis approach that quantifies the fluorescence intensity as a function of the distance from the center of the patterned feature (described in Materials and Methods and Fig. S1). The results from this analysis for both β1 integrin and phosphotyrosine label are shown in Fig. 4 B. Consistent with the representative images in Fig. 4 A, we find that β1-containing integrins are more highly enriched at patterned EGF features that are at the cell periphery compared to those that are located under the middle of these cells. In contrast, stimulated tyrosine phosphorylation is more uniformly localized at all patterned EGF features in contact with the cells. These results suggest that integrin concentration in regions where EGFR is clustered depends on processes stimulated by EGF that are downstream of tyrosine phosphorylation.
      Figure thumbnail gr4
      Figure 4Integrin β1 exhibits preferential localization at peripheral EGF-bound EGFR signaling complexes. (A) NIH-3T3 (EGFR) cells were plated onto micropatterned EGF surfaces (red channel) and immunolabeled with an anti-β1 integrin primary antibody and A488-tagged secondary antibody (green channel) and subsequently immunolabeled with an anti-phosphotyrosine 4G10 antibody and A633-tagged secondary antibody (blue channel). Panels display two representative cells. Scale bar 10 μm. (B) Radial fluorescence analysis, as described in Materials and Methods, was performed on β1 integrin and phosphotyrosine labeling to provide a measure of the average cell response for protein recruitment at patterned EGF; N = 25 cells from two independent experiments. Error bars represent mean ± SE. To see this figure in color, go online.

      The Erk signaling pathway is recruited to patterned EGF

      EGFR activated by EGF initiates a signaling cascade leading to activation of Ras, MEK, and Erk (
      • Buday L.
      • Downward J.
      Epidermal growth factor regulates p21ras through the formation of a complex of receptor, Grb2 adapter protein, and Sos nucleotide exchange factor.
      ,
      • Morrison D.K.
      • Cutler R.E.
      The complexity of Raf-1 regulation.
      ). EGFR phosphotyrosine residues are docking sites for binding of Grb2, an adaptor protein that recruits and activates Son of sevenless, which is a guanine nucleotide exchange factor that activates Ras (
      • Gale N.W.
      • Kaplan S.
      • Bar-Sagi D.
      • et al.
      Grb2 mediates the EGF-dependent activation of guanine nucleotide exchange on Ras.
      ). To investigate whether proteins in the Erk signaling pathway colocalize with clustered, activated EGFR, EGFP-H-Ras was transiently expressed in NIH 3T3 (EGFR) cells, and these cells were plated on EGF-patterned surfaces for 40 min at 37°C. Confocal imaging shows concentration of EGFP-H-Ras fluorescence with patterned EGF in >95% of imaged cells (Fig. 5 A). We observe similar recruitment of transiently transfected EGFP-N-Ras to patterned EGF (A. Singhai, unpublished results). Recruitment of EGFP-H-Ras to patterned EGF is not reduced by 20 μM PP2, consistent with Src kinase-independent recruitment (A. Singhai, unpublished results).
      Figure thumbnail gr5
      Figure 5Proteins in the MAP kinase signaling cascade: Ras, MEK, and pErk coredistribute with EGFR signaling complexes at the plasma membrane. (A) NIH-3T3 (EGFR) cells transiently transfected with EGFP-H-Ras (green channel, top panel), were plated on micropatterned EGF surfaces. NIH-3T3 (EGFR) cells plated on micropatterned EGF surfaces, were fixed and subsequently immunolabeled with anti-MEK (green channel, middle panel) or anti-phospho-Erk (green channel, bottom panel), followed by A488-tagged secondary antibodies. Scale bar 20 μm. (B) Quantification with Pearson’s cross-correlation coefficient is reported for the green and red channels; N > 30 cells from three independent experiments. Error bars represent mean ± SE; ∗∗∗ indicates p ≤ 0.001, ∗∗ indicates p ≤ 0.01. To see this figure in color, go online.
      Similar to results with EGFP-H-Ras, we also observed colocalization of MEK and pErk with patterned EGF by antibodies specific for these MAP kinase signaling components (Fig. 5). Quantification of colocalization by Pearson’s cross-correlation analysis shows limited but statistically significant recruitment of these more downstream partners in the Erk activation pathway in a ligand-dependent manner (Fig. 5 B). In addition to their recruitment to clustered EGFR, a noticeable fraction of the fluorescent label for MEK and pErk becomes localized in the nucleus under these conditions of cell activation (observed with z-sectioning; A. Singhai, unpublished results). Together, these results provide evidence for activation of the MAP kinase signaling cascade in response to patterned EGF, and the principal components of this cascade are detected as part of a macromolecular signaling complex colocalizing with EGFR at patterned EGF features.

      Recruitment of EGFR signaling partners to patterned EGF depends on the actin cytoskeleton

      We found that cytochalasin D does not inhibit tyrosine phosphorylation of EGFR clustered at EGF features, as detected by anti-phosphotyrosine 1068 or anti-phosphotyrosine mAb 4G10 (Fig. S4), consistent with cytoskeletal independence of these early events. In contrast, cytochalasin D inhibits recruitment of paxillin and pErk to patterned EGF, as summarized by quantitative Pearson’s cross-correlation analysis in Fig. 6 A. Consistent with these results, we observe that cytochalasin D inhibits Erk activation by soluble EGF in NIH-3T3 (EGFR) cells, as detected by Western blotting (Fig. 6 B). Quantification of Erk phosphorylation stimulated by soluble EGF in multiple experiments indicates ∼50% inhibition with an optimal dose (2 μM) of cytochalasin D (Fig. 6 C). These results suggest that F-actin stabilizes EGFR signaling complexes formed by both soluble and surface-immobilized EGF, thereby participating in Erk activation by EGFR in these cells.
      Figure thumbnail gr6
      Figure 6Recruitment of pErk and paxillin depend on F-actin polymerization. (A) NIH-3T3 (EGFR) cells transiently transfected with or without paxillin-EGFP were pretreated with or without 2 μM cytochalasin D before incubation on the micropatterned EGF surfaces. Untransfected NIH-3T3 (EGFR) cells were immunolabeled with anti-phospho-Erk primary and A488-tagged secondary antibodies. Quantification of colocalization between paxillin and patterned EGF or phosphorylated Erk and patterned EGF were carried out using Pearson’s cross-correlation coefficient analysis; N > 30 cells over two independent experiments. (B) NIH-3T3 (EGFR) cells, preincubated with or without 2 μM cytochalasin D for 5 min, and then stimulated or not with EGF (50 ng/ml, 4 min) were solubilized and prepared for Western blotting. Proteins were fractionated on sodium dodecyl sulfate gels and labeled with anti-phospho-Tyr-1068, anti-EGFR, anti-phospho-Erk, or anti-Erk antibodies. (C) Average intensities of phospho-Erk bands are from Western blots in three independent experiments. The intensities were normalized for each individual experiment (relative to pErk stimulated with EGF in the absence of cytochalasin D). Error bars represent mean ± SE; ∗∗∗ indicates p ≤ 0.001, indicates p ≤ 0.05.
      We found that several other proteins are recruited to EGFR signaling complexes in micron-scale patterns. Dynamin 2 is a GTPase commonly implicated in the endocytosis of receptors on cell surfaces by facilitating fission of the forming endosomes from the plasma membrane (
      • Damke H.
      • Binns D.D.
      • Baba T.
      • et al.
      Dynamin GTPase domain mutants block endocytic vesicle formation at morphologically distinct stages.
      ). We recently found that dynamin 2 plays a role in organizing the actin cytoskeleton at the sites of IgE receptor endocytosis in response to cross-linking by soluble antigens (A. Torres, D. Holowka, and B. Baird, unpublished results). We find that dynamin 2-EGFP expressed in NIH-3T3 (EGFR) cells is recruited to patterned EGF in >90% of the expressing cells (Fig. S6). This recruitment is not inhibited by cytochalasin D, similar to observations with IgE receptors concentrated in micron-scale patterns, suggesting that dynamin 2 is recruited in the absence of F-actin localization (A. Singhai, unpublished results).
      PLCγ1 is an enzyme that mediates hydrolysis of PI(4,5)P2 to inositol 1,4,5-trisphosphate and diacylglycerol, and it was previously shown to bind directly to activated EGFR via its tandem SH2 domains (
      • Margolis B.
      • Li N.
      • Schlessinger J.
      • et al.
      The tyrosine phosphorylated carboxy-terminus of the EGF receptor is a binding site for GAP and PLC-gamma.
      ). NIH-3T3 (EGFR) cells transiently transfected with EGFP-tagged PLCγ1 show clear recruitment of this enzyme to patterned EGF (Fig. S6). Because tyrosine phosphorylation of paxillin by a Src kinase was previously shown to be stimulated by EGF (
      • Sen A.
      • O’Malley K.
      • Hammes S.R.
      • et al.
      Paxillin regulates androgen- and epidermal growth factor-induced MAPK signaling and cell proliferation in prostate cancer cells.
      ), we investigated whether the Src family kinase Lyn tagged with mRFP colocalizes with EGFR signaling complexes in micron-scale EGF patterns. As shown in Fig. S6, we observe Lyn-mRFP concentrating with the patterned features in ∼60% of the expressing cells. This suggests Src family kinases can be recruited to regions of activated EGFR complexes, providing a mechanism by which these kinases can mediate phosphorylation of Tyr-118 in paxillin associated with the activated EGFR signaling complexes (Fig. S2).

      Inhibition of PI(4,5)P2 synthesis interferes with recruitment of F-actin, Erk, and dynamin 2 to pattern-localized EGFR complexes

      It is now well established that PI(4,5)P2 plays multiple roles in regulating cell signaling (
      • Di Paolo G.
      • De Camilli P.
      Phosphoinositides in cell regulation and membrane dynamics.
      ). We recently showed that two inhibitors of phosphoinositide synthesis, PAO and quercetin, effectively inhibit multiple processes in IgE receptor signaling in mast cells at concentrations that are selective for inhibition of PI(4)P and PI(4,5)P2 synthesis (
      • Santos M. de S.
      • Naal R.M.
      • Holowka D.
      • et al.
      Inhibitors of PI(4,5)P2 synthesis reveal dynamic regulation of IgE receptor signaling by phosphoinositides in RBL mast cells.
      ). We further found that under these conditions, these compounds inhibit recruitment of F-actin to IgE receptor complexes localized to micropatterned antigen (A. Singhai, unpublished results). In NIH-3T3 (EGFR) cells, we find that pretreatment of cells with 2 μM PAO or 20 μM quercetin leads to substantial reduction in F-actin recruitment to EGF patterns, as detected by A488-phalloidin labeling (Fig. 7, A and B). In addition, treatment with PAO or quercetin inhibits recruitment of pErk at EGFR signaling complexes on patterned EGF surfaces (Fig. 7 C). In other experiments we found that these compounds prevent concentration of dynamin 2 with clustered EGFR bound to patterned EGF in >95% of the imaged cells (A. Singhai, unpublished results), suggesting that phosphoinositides play a role in this co-localization that further leads to actin organization in these regions.
      Figure thumbnail gr7
      Figure 7PI(4,5)P2 synthesis is implicated in the recruitment of F-actin and phosphorylated Erk to EGFR signaling complexes. (A) NIH-3T3 (EGFR) cells were pretreated with 2 μM PAO (middle panel) or 20 μM quercetin (bottom panel) or not treated (control; top panel) before plating on micropatterned EGF surfaces. Cells were labeled with A488-phalloidin (green channel) or anti-phospho-Erk antibody, followed by A488-tagged secondary antibody (images not shown) after fixation. Scale bar 20 μm. (B) Pearson’s cross-correlation analysis is shown for F-actin colocalization with patterned EGF; N > 30 cells for at least two independent experiments. (C) Similar analysis for colocalization of antibody-labeled phosphorylated Erk with patterned EGF; N > 30 cells for at least two independent experiments. Error bars represent mean ± SE; ∗∗∗∗ indicates p ≤ 0.0001, ∗∗∗ indicates p ≤ 0.001. To see this figure in color, go online.
      To further evaluate the role of phosphoinositides in assembly of EGFR signaling complexes, we coexpressed a rapamycin-recruitable inositol 5-phosphatase, FKBP12-5ptase, together with the plasma membrane-targeted rapamycin binding domain, PM-CFP-FRB (
      • Varnai P.
      • Thyagarajan B.
      • Balla T.
      • et al.
      Rapidly inducible changes in phosphatidylinositol 4,5-bisphosphate levels influence multiple regulatory functions of the lipid in intact living cells.
      ), in NIH 3T3 (EGFR) cells. We acutely recruited the inositol 5-phosphatase to the plasma membrane by the addition of 100 nM rapamycin just before allowing these cells to settle on EGF patterns. As shown in representative images in Fig. 8 A and quantified by radial analysis in Fig. 8 B, we find that inositol 5-phosphatase recruitment significantly reduces the localization of F-actin at EGF patterns. These results provide strong evidence that PI(4,5)P2, and possibly other phosphoinositides, play a role in the stabilization of F-actin-associated signaling complexes at patterned EGF sites.
      Figure thumbnail gr8
      Figure 8Rapamycin-dependent recruitment of an inositol 5-phosphatase supports a role for phosphoinositides in F-actin localization to EGFR signaling complexes at patterned EGF features. (A) NIH-3T3 (EGFR) cells transiently transfected with PM2-CFP-FRB (cyan channel) and FKBP12-5ptase are briefly treated with 100 nM rapamycin (Rapa) for 5 min at 37°C before plating on EGF patterns (red channel). The cells are fixed and subsequently labeled for F-actin using A647-phalloidin (blue channel). Representative images are shown for a cell not treated with Rapa (top panel) and a cell treated with 100 nM Rapa (bottom panel). Scale bar: 10 μm. (B) Radial fluorescence analysis for F-actin recruitment at patterned EGF was performed for both the untreated and Rapa treated cases; N = 22 from two independent experiments. Error bars represent mean ± SE; ∗∗∗ indicates p ≤ 0.001, ∗∗indicates p ≤ 0.01. To see this figure in color, go online.

      Discussion

      RTKs, including EGFR, are key regulators of cellular processes such as proliferation, migration, and differentiation (
      • Lemmon M.A.J.S.
      • Schlessinger J.
      • Ferguson K.M.
      The EGFR family: not so prototypical receptor tyrosine kinases.
      ). EGF binding to the extracellular domain of EGFR causes dimerization leading to transphosphorylation of tyrosine residues in the C-terminal segments of these receptors. Signal transduction propagates by recruitment of adaptor proteins, such as Grb2 and Gab2, and enzymes, such as PLCγ1, to these phosphotyrosine residues. Recruited adaptor proteins mediate activation of downstream signaling cascades, including the MAP kinase cascade and PI3-kinase-dependent activation of Akt. The signaling pathways leading to activation of these kinases have been established (
      • Gale N.W.
      • Kaplan S.
      • Bar-Sagi D.
      • et al.
      Grb2 mediates the EGF-dependent activation of guanine nucleotide exchange on Ras.
      ,
      • Zheng J.
      • Cahill S.M.
      • Cowburn D.
      • et al.
      Identification of the binding site for acidic phospholipids on the pH domain of dynamin: implications for stimulation of GTPase activity.
      ,
      • Ramos J.W.
      The regulation of extracellular signal-regulated kinase (ERK) in mammalian cells.
      ), but detectable association of these signaling proteins in EGFR complexes has not been previously demonstrated. Micropatterned ligand arrays have been valuable tools to study the interactions of cytoplasmic and membrane-bound proteins with ligand-activated receptors at the plasma membrane. We previously used such arrays to probe the recruitment of signaling partners of activated IgE/FcεRI complexes (
      • Wu M.
      • Holowka D.
      • Baird B.
      • et al.
      Visualization of plasma membrane compartmentalization with patterned lipid bilayers.
      ,
      • Torres A.J.
      • Vasudevan L.
      • Baird B.A.
      • et al.
      Focal adhesion proteins connect IgE receptors to the cytoskeleton as revealed by micropatterned ligand arrays.
      ). In this study, we show for the first time, to our knowledge, that EGF engagement of EFGR can cause stable association of Erk signaling proteins in a signaling complex.
      In previous studies from other laboratories, covalently immobilized EGF was shown to stimulate tyrosine phosphorylation in Chinese hamster ovary (CHO) cells overexpressing EGFR (
      • Ito Y.
      • Chen G.
      • Imanishi Y.
      Micropatterned immobilization of epidermal growth factor to regulate cell function.
      ,
      • Chen G.
      • Ito Y.
      Gradient micropattern immobilization of EGF to investigate the effect of artificial juxtacrine stimulation.
      ) and in breast cancer BT-20 cells (
      • Gonçalves R.
      • Martins M.C.L.
      • Barbosa M.A.
      • et al.
      Bioactivity of immobilized EGF on self-assembled monolayers: optimization of the immobilization process.
      ). These studies used EGF patterned on surfaces with feature sizes of dimensions ∼10–100 μm, and they found that cells cultured on these surfaces show higher growth rates compared to the cells plated on surfaces without attached EGF. Our experiments use patterns of EGF-biotin tightly bound to covalently attached A568-streptavidin in features with subcellular dimensions of 1–4 μm, and we observe for individual cells that EGFR accumulates at these features and that tyrosine phosphorylation occurs in the same patterned regions (Fig. 1). We further observe that EGFR recruited to patterned EGF is directly tyrosine phosphorylated at Tyr-1068 (Fig. S2 A), consistent with previous observations for EGFR stimulated with soluble EGF (
      • Bryant K.L.
      • Antonyak M.A.
      • Holowka D.
      • et al.
      Mutations in the polybasic juxtamembrane sequence of both plasma membrane- and endoplasmic reticulum-localized epidermal growth factor receptors confer ligand-independent cell transformation.
      ,
      • Ullrich A.
      • Schlessinger J.
      Signal transduction by receptors with tyrosine kinase activity.
      ). This strategy has permitted us to visualize interactions with downstream signaling partners and F-actin under conditions in which receptor endocytosis is prevented. In initial experiments we found that EGFP-tagged EGFR mediates endocytosis of rhodamine-EGF when this ligand is simply adsorbed to the surface in a patterned array (A. Singhai, unpublished results), demonstrating the importance of covalently attaching the ligand complex in micropatterned surfaces to visualize robust recruitment of EGFR in spatially resolved clusters.
      In response to binding soluble EGF, cell surface EGFR dimerizes in the process of activating its tyrosine kinase domain leading to transphosphorylation (
      • Lemmon M.A.J.S.
      • Schlessinger J.
      • Ferguson K.M.
      The EGFR family: not so prototypical receptor tyrosine kinases.
      ). Our results show that the local concentration of EGF at micropatterned surfaces similarly causes tyrosine phosphorylation, a process that is dependent on EGFR kinase activity and inhibited by 10 μM Iressa. This treatment however does not inhibit colocalization of EGFR with patterned EGF. This observation shows that EGFR kinase activity is not required for EGFR binding to immobilized EGF, but that kinase activity is required for activation leading to downstream signaling. Using a resonance energy transfer fluorescence lifetime imaging-based method, Verveer et al. (
      • Verveer P.J.
      • Wouters F.S.
      • Bastiaens P.I.H.
      • et al.
      Quantitative imaging of lateral ErbB1 receptor signal propagation in the plasma membrane.
      ) detected stimulated EGFR tyrosine phosphorylation distal to the region of attachment to micron-sized EGF-conjugated beads on MCF7 cells. A subsequent study provided evidence that long-range propagation of EGFR activation depends on high densities of receptor expression (
      • Sawano A.
      • Takayama S.
      • Miyawaki A.
      • et al.
      Lateral propagation of EGF signaling after local stimulation is dependent on receptor density.
      ). In our stably transfected NIH 3T3 cells, EGFR is moderately overexpressed (
      • Bryant K.L.
      • Antonyak M.A.
      • Holowka D.
      • et al.
      Mutations in the polybasic juxtamembrane sequence of both plasma membrane- and endoplasmic reticulum-localized epidermal growth factor receptors confer ligand-independent cell transformation.
      ). We observe that tyrosine phosphorylation is largely confined to the patterned EGF features, as determined both by cross correlation and by radial analysis, but we cannot exclude some component of more distal phosphorylation, such as the relatively low levels of off-pattern labeling evident in Fig. 4 A.
      Activated EGFR initiates a signaling cascade that leads to activation of Ras, and our results show recruitment of EGFP-H-Ras to clustered EGFR at the plasma membrane (Fig. 5). Ras activates Raf kinase (
      • Downward J.
      Control of ras activation.
      ,
      • Leevers S.J.
      • Paterson H.F.
      • Marshall C.J.
      Requirement for Ras in Raf activation is overcome by targeting Raf to the plasma membrane.
      ), which further activates MEK to phosphorylate Erk in the MAP kinase cascade (
      • Zheng C.F.
      • Guan K.L.
      Activation of MEK family kinases requires phosphorylation of two conserved Ser/Thr residues.
      ). Postactivation, Erk and MEK are translocated as a complex into the nucleus to initiate gene transcription, which leads to cell division and proliferation among other responses (
      • Burack W.R.
      • Shaw A.S.
      Live cell imaging of ERK and MEK: simple binding equilibrium explains the regulated nucleocytoplasmic distribution of ERK.
      ). Our experiments show that MEK and Erk colocalize with activated EGFR clustered at the plasma membrane by the micropatterned EGF surfaces (Fig. 5 and Table 1). Previously, these interactions have been difficult to detect, possibly due in part to limitations in optical resolution or to EGFR endocytosis after activation by EGF. Stabilization of EFGR signaling complexes in large (micron-scale) clusters enhances detection of signaling partners. Our results demonstrate that macromolecular complexes of EGFR formed on the plasma membrane can activate Erk colocalized with these complexes, along with other members of this MAP kinase cascade, Ras and MEK.
      Recent studies pointed to a role for paxillin in EGFR signaling (
      • Sen A.
      • O’Malley K.
      • Hammes S.R.
      • et al.
      Paxillin regulates androgen- and epidermal growth factor-induced MAPK signaling and cell proliferation in prostate cancer cells.
      ,
      • Sen A.
      • De Castro I.
      • Hammes S.R.
      • et al.
      Paxillin mediates extranuclear and intranuclear signaling in prostate cancer proliferation.
      ), and we find that paxillin-EGFP expressed in NIH-3T3 cells colocalizes with the microclustered EGFR at the plasma membrane in a process that depends on EGFR kinase activity (Fig. 2). Paxillin is known be phosphorylated on Tyr-31 and Try-118 by Src kinase, and on Ser-83 and Ser-126 by Erk (
      • Brown M.C.
      • Turner C.E.
      Paxillin: adapting to change.
      ). In agreement, we observe tyrosine phosphorylation of paxillin recruited to EGFR signaling complexes at the plasma membrane by labeling with anti-phospho-paxillin antibodies (Fig. S2). Consistent with tyrosine phosphorylation of paxillin by Src (
      • Schaller M.D.
      • Schaefer E.M.
      Multiple stimuli induce tyrosine phosphorylation of the Crk-binding sites of paxillin.
      ), we observe recruitment of Lyn, a Src family kinase, to clustered EGFR at the plasma membrane (Fig. S3). Treatment with the Src family kinase inhibitor PP2 inhibits tyrosine phosphorylation of paxillin colocalized with these clusters, and this treatment causes a small increase in the recruitment of paxillin-EGFP and phospho-Ser-paxillin as quantified by Pearson’s cross-correlation analysis (Fig. S2). This suggests enhanced association of paxillin with clustered EGFR in the absence of Src kinase activity. Under these conditions, the activity of Src kinase is not critical for the activation of Erk, as observed in experiments with patterned EGF, as well as with soluble EGF (K. Bryant, unpublished results). In contrast, others have shown reduced Erk activation in the presence of PP2 in prostate cancer cells (
      • Sen A.
      • O’Malley K.
      • Hammes S.R.
      • et al.
      Paxillin regulates androgen- and epidermal growth factor-induced MAPK signaling and cell proliferation in prostate cancer cells.
      ). This distinction is possibly due to differences in cell types and differential contributions of other signaling pathways, such as the formation of focal adhesions, which may be more important for activation of Erk kinase in prostate cancer cells.
      Our previous experiments that characterized IgE-FcεRI signaling using micropatterned ligands showed that F-actin is recruited to the clustered FcεRI complexes (
      • Wu M.
      • Holowka D.
      • Baird B.
      • et al.
      Visualization of plasma membrane compartmentalization with patterned lipid bilayers.
      ,
      • Torres A.J.
      • Vasudevan L.
      • Baird B.A.
      • et al.
      Focal adhesion proteins connect IgE receptors to the cytoskeleton as revealed by micropatterned ligand arrays.
      ). We further observed that the Src family kinase Lyn, chiefly responsible for the phosphorylation of FcεRI, is visibly recruited to the clustered FcεRI. In these experiments, micron-scale accumulation of Lyn with FcεRI at patterned ligands was prevented by cytochalasin D, even though tyrosine phosphorylation of FcεRI was enhanced, indicating the presence of catalytically active Lyn in subdetectable amounts (
      • Wu M.
      • Holowka D.
      • Baird B.
      • et al.
      Visualization of plasma membrane compartmentalization with patterned lipid bilayers.
      ). Based on these previous experiments, we hypothesized that actin polymerization stabilizes EGFR signaling complexes at the plasma membrane, even though a high level of stabilization is not essential for signal initiation. Our present results show recruitment of F-actin to microclustered EGFR, whereas the actin-binding proteins ezrin and moesin do not visibly colocalize with the EGF features (Fig. 3). Consistent with our observations with FcεRI, we find that cytochalasin D prevents detectable recruitment of more downstream partners such as paxillin, H-Ras, and Erk to the EGFR signaling complexes at the plasma membrane (Fig. 6 and A. Singhai, unpublished results). In contrast, pretreatment with cytochalasin D does not visibly inhibit recruitment or tyrosine phosphorylation of EGFR or the recruitment of dynamin 2 in an EGF-dependent manner. Although a previous study showed no effect of cytochalasin D on EGF-dependent activation of Erk in Rat 1a fibroblasts (
      • Luttrell L.M.
      • Daaka Y.
      • Lefkowitz R.J.
      • et al.
      G protein-coupled receptors mediate two functionally distinct pathways of tyrosine phosphorylation in rat 1a fibroblasts. Shc phosphorylation and receptor endocytosis correlate with activation of Erk kinases.
      ), we find that cytochalasin D causes ∼50% inhibition of Erk activation in NIH-3T3 (EGFR) cells stimulated with soluble EGF (Fig. 6). These results suggest some cell type-specific differences in the role of F-actin in stabilizing macromolecular signaling complexes formed with clustered EGFR at the plasma membrane.
      Paxillin, F-actin, and integrins are known to participate in focal adhesion complexes. These macromolecular structures are formed at sites of cell interactions with the extracellular matrix (
      • Turner C.E.
      Paxillin and focal adhesion signalling.
      ) and have the capacity to activate Erk through integrin-fibronectin-mediated signaling (
      • Ishibe S.
      • Joly D.
      • Cantley L.G.
      • et al.
      Phosphorylation-dependent paxillin-ERK association mediates hepatocyte growth factor-stimulated epithelial morphogenesis.
      ). Moreover, studies have shown that integrins can associate with EGFR, and integrin-dependent cell adhesion contributes to EGFR activation in an F-actin-dependent manner (
      • Yu X.
      • Miyamoto S.
      • Mekada E.
      Integrin α 2 β 1-dependent EGF receptor activation at cell-cell contact sites.
      ,
      • Alexi X.
      • Berditchevski F.
      • Odintsova E.
      The effect of cell-ECM adhesion on signalling via the ErbB family of growth factor receptors.
      ). We find that β1-containing integrins become localized to patterned-recruited EGFR signaling complexes, but exhibit higher concentrations at the more peripheral EGF patterned features in contact with individual NIH-3T3 (EGFR) cells. This suggests recruitment that correlates with a more dynamic engagement of the cells with these patterns, possibly due to filopodia extension mediating initial attachments to these peripheral sites. F-actin recruitment sometimes appears more concentrated at these peripheral sites, especially in cases of cellular extensions, suggesting possible focal adhesion complex formation. Preferential recruitment of β1 integrin to peripheral EGF features is consistently more pronounced compared to F-actin recruitment, suggesting that F-actin association with EGF-clustered EGFR does not depend on integrin recruitment or filopodia extension. Furthermore, downstream signaling proteins that stably associate with EGFR at EGF features in an F-actin-dependent manner, including paxillin and pErk, do not appear to associate preferentially with more peripheral features. Future studies will assess the extent to which focal adhesion formation contributes to the more downstream signaling consequences of EGFR activation by patterned EGF.
      After EGFR is activated by binding to soluble EGF, it is normally endocytosed (
      • Sorkin A.
      • Goh L.K.
      Endocytosis and intracellular trafficking of ErbBs.
      ). With micropatterned EGF surfaces, we find that clathrin is visibly concentrated at EGF features in ∼35% of the cells (A. Singhai, unpublished results). By comparison, we observe that dynamin 2, a GTPase implicated for its role in both clathrin-dependent (
      • Vieira A.V.
      • Lamaze C.
      • Schmid S.L.
      Control of EGF receptor signaling by clathrin-mediated endocytosis.
      ) and clathrin-independent endocytosis (
      • Mayor S.
      • Pagano R.E.
      Pathways of clathrin-independent endocytosis.
      ), colocalizes with patterned EGF in >90% of cells. The actin cytoskeleton has been implicated in endocytosis of EGFR (
      • Benesch S.
      • Polo S.
      • Rottner K.
      • et al.
      N-WASP deficiency impairs EGF internalization and actin assembly at clathrin-coated pits.
      ), and its accumulation at EGFR complexes may contribute to this process with soluble EGF, in addition to its role in stabilizing associations with downstream signaling partners described previously. Our results, taken together, suggest that some of the proteins recruited to clustered EGFR may be a part of the EGFR endocytic pathway, even though endocytosis of EGFR is prevented by the attachment of EGF to the silicon substrate. Furthermore, our results support previous evidence that endocytosis of EGFR is not necessary for activation of the MAP kinase cascade (
      • Sousa L.P.
      • Lax I.
      • Schlessinger J.
      • et al.
      Suppression of EGFR endocytosis by dynamin depletion reveals that EGFR signaling occurs primarily at the plasma membrane.
      ).
      Phosphoinositides are known to play important roles in cell signaling in all cell types, as well as in receptor endocytosis (
      • Di Paolo G.
      • De Camilli P.
      Phosphoinositides in cell regulation and membrane dynamics.
      ). Our results in Fig. 7 indicate that recruitment of F-actin and pErk to patterned EGF are inhibited by pretreatment with PAO or quercetin at concentrations that selectively inhibit phosphoinositide synthesis (
      • Santos M. de S.
      • Naal R.M.
      • Holowka D.
      • et al.
      Inhibitors of PI(4,5)P2 synthesis reveal dynamic regulation of IgE receptor signaling by phosphoinositides in RBL mast cells.
      ). Furthermore, acute recruitment of an inositol 5-phosphatase to the plasma membrane by rapamycin also inhibits recruitment of F-actin to EGF patterns, and this provides even stronger evidence that phosphoinositides play an important role in the F-actin stabilization of signaling complexes in this situation.
      Consistent with these results, PI(4,5)P2 has been implicated in the organization of the actin cytoskeleton (
      • Di Paolo G.
      • De Camilli P.
      Phosphoinositides in cell regulation and membrane dynamics.
      ,
      • Engqvist-Goldstein Å.E.Y.
      • Drubin D.G.
      Actin assembly and endocytosis: from yeast to mammals.
      ). Inhibition of pErk recruitment to patterned EGF by inhibitors of phosphoinositide synthesis may be a consequence of inhibition of F-actin recruitment, as cytochalasin D also inhibits pErk recruitment (Fig. 6). Our results further show that dynamin 2, which binds PI(4,5)P2 via its PH domain (
      • Zheng J.
      • Cahill S.M.
      • Cowburn D.
      • et al.
      Identification of the binding site for acidic phospholipids on the pH domain of dynamin: implications for stimulation of GTPase activity.
      ,
      • Ramos J.W.
      The regulation of extracellular signal-regulated kinase (ERK) in mammalian cells.
      ), is prevented from colocalization with clustered EGFR signaling complexes by PAO or quercetin, even though dynamin 2 recruitment is not inhibited by cytochalasin D (A. Torres and A. Singhai, unpublished results). This suggests that dynamin 2 recruitment to the patterned EGFR signaling complexes depends on association with PI(4,5)P2 in a step that is upstream of F-actin recruitment. Participation of dynamin 2 in the organization of F-actin with clustered EGFR signaling complexes is a subject of continuing interest.
      In conclusion, we have shown that micropatterned EGF surfaces with quantitative image analysis schemes provide new, to our knowledge, insights into the structural organization of EGFR signaling complexes. This approach complements traditional biochemical methods such as immunoprecipitation of solubilized EGFR to detect interactions of signaling partners. As summarized in Table 1, our experiments show direct evidence for the formation of macromolecular EGFR signaling complexes that include PLCγ1, H-Ras, MEK, and pErk, in addition to paxillin, F-actin, and dynamin 2. Our results provide evidence that phosphoinositide synthesis and the actin cytoskeleton participate in the stabilization of these EGFR signaling complexes, in addition to possible roles for these in EGFR endocytosis. Our results reveal how stabilization of EGFR signaling complexes by patterned EGF ligand contributes to robust downstream signaling.
      We thank Carol Bayles for maintaining the Cornell Imaging Facility and the Cornell NanoScale Science and Technology Facility (CNF) for maintaining the microfabrication facility.
      This work was supported by the National Institutes of Health from the National Institute of Allergy and Infectious Diseases (grant R01-AI018306) and by the Nanobiotechnology Center (NSF: ECS9876771). The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH or the NSF.

      Supporting Material

      References

        • Blume-Jensen P.
        • Hunter T.
        Oncogenic kinase signalling.
        Nature. 2001; 411: 355-365
        • Lemmon M.A.J.S.
        • Schlessinger J.
        • Ferguson K.M.
        The EGFR family: not so prototypical receptor tyrosine kinases.
        Cold Spring Harb. Perspect. Biol. 2014; 6: a020768
        • Schlessinger J.
        Receptor tyrosine kinases: legacy of the first two decades.
        Cold Spring Harb. Perspect. Biol. 2014; 6: a008912
        • Pines G.
        • Köstler W.J.
        • Yarden Y.
        Oncogenic mutant forms of EGFR: lessons in signal transduction and targets for cancer therapy.
        FEBS Lett. 2010; 584: 2699-2706
        • Gschwind A.
        • Fischer O.M.
        • Ullrich A.
        The discovery of receptor tyrosine kinases: targets for cancer therapy.
        Nat. Rev. Cancer. 2004; 4: 361-370
        • McKay M.M.
        • Morrison D.K.
        Integrating signals from RTKs to ERK/MAPK.
        Oncogene. 2007; 26: 3113-3121
        • Sorkin A.
        • Goh L.K.
        Endocytosis and intracellular trafficking of ErbBs.
        Exp. Cell Res. 2009; 315: 683-696
        • Verveer P.J.
        • Wouters F.S.
        • Bastiaens P.I.H.
        • et al.
        Quantitative imaging of lateral ErbB1 receptor signal propagation in the plasma membrane.
        Science. 2000; 290: 1567-1570
        • Sigismund S.
        • Argenzio E.
        • Di Fiore P.P.
        • et al.
        Clathrin-mediated internalization is essential for sustained EGFR signaling but dispensable for degradation.
        Dev. Cell. 2008; 15: 209-219
        • Taub N.
        • Teis D.
        • Huber L.A.
        • et al.
        Late endosomal traffic of the epidermal growth factor receptor ensures spatial and temporal fidelity of mitogen-activated protein kinase signaling.
        Mol. Biol. Cell. 2007; 18: 4698-4710
        • Sousa L.P.
        • Lax I.
        • Schlessinger J.
        • et al.
        Suppression of EGFR endocytosis by dynamin depletion reveals that EGFR signaling occurs primarily at the plasma membrane.
        Proc. Natl. Acad. Sci. USA. 2012; 109: 4419-4424
        • Sen A.
        • O’Malley K.
        • Hammes S.R.
        • et al.
        Paxillin regulates androgen- and epidermal growth factor-induced MAPK signaling and cell proliferation in prostate cancer cells.
        J. Biol. Chem. 2010; 285: 28787-28795
        • Sen A.
        • De Castro I.
        • Hammes S.R.
        • et al.
        Paxillin mediates extranuclear and intranuclear signaling in prostate cancer proliferation.
        J. Clin. Invest. 2012; 122: 2469-2481
        • den Hartigh J.C.
        • van Bergen en Henegouwen P.M.
        • Boonstra J.
        • et al.
        The EGF receptor is an actin-binding protein.
        J. Cell Biol. 1992; 119: 349-355
        • Rijken P.J.
        • Post S.M.
        • Boonstra J.
        • et al.
        Actin polymerization localizes to the activated epidermal growth factor receptor in the plasma membrane, independent of the cytosolic free calcium transient.
        Exp. Cell Res. 1995; 218: 223-232
        • Wu M.
        • Holowka D.
        • Baird B.
        • et al.
        Visualization of plasma membrane compartmentalization with patterned lipid bilayers.
        Proc. Natl. Acad. Sci. USA. 2004; 101: 13798-13803
        • Torres A.J.
        • Vasudevan L.
        • Baird B.A.
        • et al.
        Focal adhesion proteins connect IgE receptors to the cytoskeleton as revealed by micropatterned ligand arrays.
        Proc. Natl. Acad. Sci. USA. 2008; 105: 17238-17244
        • Gonçalves R.
        • Martins M.C.L.
        • Barbosa M.A.
        • et al.
        Bioactivity of immobilized EGF on self-assembled monolayers: optimization of the immobilization process.
        J. Biomed. Mater. Res. A. 2010; 94: 576-585
        • Ito Y.
        • Chen G.
        • Imanishi Y.
        Micropatterned immobilization of epidermal growth factor to regulate cell function.
        Bioconjug. Chem. 1998; 9: 277-282
        • Stabley D.
        • Retterer S.
        • Salaita K.
        • et al.
        Manipulating the lateral diffusion of surface-anchored EGF demonstrates that receptor clustering modulates phosphorylation levels.
        Integr. Biol. (Camb). 2013; 5: 659-668
        • Monick M.M.
        • Cameron K.
        • Hunninghake G.W.
        • et al.
        Activation of the epidermal growth factor receptor by respiratory syncytial virus results in increased inflammation and delayed apoptosis.
        J. Biol. Chem. 2005; 280: 2147-2158
        • Hammond S.
        • Wagenknecht-Wiesner A.
        • Baird B.
        • et al.
        Roles for SH2 and SH3 domains in Lyn kinase association with activated FcepsilonRI in RBL mast cells revealed by patterned surface analysis.
        J. Struct. Biol. 2009; 168: 161-167
        • Wang X.J.
        • Liao H.J.
        • Carpenter G.
        • et al.
        EGF-dependent translocation of green fluorescent protein-tagged PLC-gamma1 to the plasma membrane and endosomes.
        Exp. Cell Res. 2001; 267: 28-36
        • Varnai P.
        • Thyagarajan B.
        • Balla T.
        • et al.
        Rapidly inducible changes in phosphatidylinositol 4,5-bisphosphate levels influence multiple regulatory functions of the lipid in intact living cells.
        J. Cell Biol. 2006; 175: 377-382
        • Bryant K.L.
        • Antonyak M.A.
        • Holowka D.
        • et al.
        Mutations in the polybasic juxtamembrane sequence of both plasma membrane- and endoplasmic reticulum-localized epidermal growth factor receptors confer ligand-independent cell transformation.
        J. Biol. Chem. 2013; 288: 34930-34942
        • Ilic B.
        • Craighead H.
        Topographical patterning of chemically sensitive biological materials using a polymer-based dry lift off.
        Biomed. Microdevices. 2000; 2: 317-322
        • Orth R.
        • Wu M.
        • Baird B.
        • et al.
        Mast cell activation on patterned lipid bilayers of subcellular dimensions.
        Langmuir. 2003; 19: 1599-1605
        • Torres A.J.
        • Holowka D.
        • Baird B.A.
        Micropatterned ligand arrays to study spatial regulation in Fc receptor signaling.
        Methods Mol. Biol. 2011; 748: 195-207
        • Wakeling A.E.
        • Guy S.P.
        • Gibson K.H.
        • et al.
        ZD1839 (Iressa): an orally active inhibitor of epidermal growth factor signaling with potential for cancer therapy.
        Cancer Res. 2002; 62: 5749-5754
        • Brown M.C.
        • Turner C.E.
        Paxillin: adapting to change.
        Physiol. Rev. 2004; 84: 1315-1339
        • Antoniades I.
        • Stylianou P.
        • Skourides P.A.
        Making the connection: ciliary adhesion complexes anchor basal bodies to the actin cytoskeleton.
        Dev. Cell. 2014; 28: 70-80
        • Salgia R.
        • Li J.L.
        • Griffen J.D.
        • et al.
        Molecular cloning of human paxillin, a focal adhesion protein phosphorylated by P210BCR/ABL.
        J. Biol. Chem. 1995; 270: 5039-5047
        • Schaller M.D.
        • Schaefer E.M.
        Multiple stimuli induce tyrosine phosphorylation of the Crk-binding sites of paxillin.
        Biochem. J. 2001; 360: 57-66
        • Huang C.
        • Borchers C.H.
        • Jacobson K.
        • et al.
        Phosphorylation of paxillin by p38MAPK is involved in the neurite extension of PC-12 cells.
        J. Cell Biol. 2004; 164: 593-602
        • Woodrow M.A.
        • Woods D.
        • McMahon M.
        • et al.
        Ras-induced serine phosphorylation of the focal adhesion protein paxillin is mediated by the Raf—>MEK—>ERK pathway.
        Exp. Cell Res. 2003; 287: 325-338
        • Bretscher A.
        • Edwards K.
        • Fehon R.G.
        ERM proteins and merlin: integrators at the cell cortex.
        Nat. Rev. Mol. Cell Biol. 2002; 3: 586-599
        • Turner C.E.
        Paxillin and focal adhesion signalling.
        Nat. Cell Biol. 2000; 2: E231-E236
        • Woods D.
        • Cherwinski H.
        • McMahon M.
        • et al.
        Induction of β3-integrin gene expression by sustained activation of the Ras-regulated Raf-MEK-extracellular signal-regulated kinase signaling pathway.
        Mol. Cell. Biol. 2001; 21: 3192-3205
        • Buday L.
        • Downward J.
        Epidermal growth factor regulates p21ras through the formation of a complex of receptor, Grb2 adapter protein, and Sos nucleotide exchange factor.
        Cell. 1993; 73: 611-620
        • Morrison D.K.
        • Cutler R.E.
        The complexity of Raf-1 regulation.
        Curr. Opin. Cell Biol. 1997; 9: 174-179
        • Gale N.W.
        • Kaplan S.
        • Bar-Sagi D.
        • et al.
        Grb2 mediates the EGF-dependent activation of guanine nucleotide exchange on Ras.
        Nature. 1993; 363: 88-92
        • Damke H.
        • Binns D.D.
        • Baba T.
        • et al.
        Dynamin GTPase domain mutants block endocytic vesicle formation at morphologically distinct stages.
        Mol. Biol. Cell. 2001; 12: 2578-2589
        • Margolis B.
        • Li N.
        • Schlessinger J.
        • et al.
        The tyrosine phosphorylated carboxy-terminus of the EGF receptor is a binding site for GAP and PLC-gamma.
        EMBO J. 1990; 9: 4375-4380
        • Di Paolo G.
        • De Camilli P.
        Phosphoinositides in cell regulation and membrane dynamics.
        Nature. 2006; 443: 651-657
        • Santos M. de S.
        • Naal R.M.
        • Holowka D.
        • et al.
        Inhibitors of PI(4,5)P2 synthesis reveal dynamic regulation of IgE receptor signaling by phosphoinositides in RBL mast cells.
        Mol. Pharmacol. 2013; 83: 793-804
        • Zheng J.
        • Cahill S.M.
        • Cowburn D.
        • et al.
        Identification of the binding site for acidic phospholipids on the pH domain of dynamin: implications for stimulation of GTPase activity.
        J. Mol. Biol. 1996; 255: 14-21
        • Ramos J.W.
        The regulation of extracellular signal-regulated kinase (ERK) in mammalian cells.
        Int. J. Biochem. Cell Biol. 2008; 40: 2707-2719
        • Chen G.
        • Ito Y.
        Gradient micropattern immobilization of EGF to investigate the effect of artificial juxtacrine stimulation.
        Biomaterials. 2001; 22: 2453-2457
        • Ullrich A.
        • Schlessinger J.
        Signal transduction by receptors with tyrosine kinase activity.
        Cell. 1990; 61: 203-212
        • Sawano A.
        • Takayama S.
        • Miyawaki A.
        • et al.
        Lateral propagation of EGF signaling after local stimulation is dependent on receptor density.
        Dev. Cell. 2002; 3: 245-257
        • Downward J.
        Control of ras activation.
        Cancer Surv. 1996; 27: 87-100
        • Leevers S.J.
        • Paterson H.F.
        • Marshall C.J.
        Requirement for Ras in Raf activation is overcome by targeting Raf to the plasma membrane.
        Nature. 1994; 369: 411-414
        • Zheng C.F.
        • Guan K.L.
        Activation of MEK family kinases requires phosphorylation of two conserved Ser/Thr residues.
        EMBO J. 1994; 13: 1123-1131
        • Burack W.R.
        • Shaw A.S.
        Live cell imaging of ERK and MEK: simple binding equilibrium explains the regulated nucleocytoplasmic distribution of ERK.
        J. Biol. Chem. 2005; 280: 3832-3837
        • Luttrell L.M.
        • Daaka Y.
        • Lefkowitz R.J.
        • et al.
        G protein-coupled receptors mediate two functionally distinct pathways of tyrosine phosphorylation in rat 1a fibroblasts. Shc phosphorylation and receptor endocytosis correlate with activation of Erk kinases.
        J. Biol. Chem. 1997; 272: 31648-31656
        • Ishibe S.
        • Joly D.
        • Cantley L.G.
        • et al.
        Phosphorylation-dependent paxillin-ERK association mediates hepatocyte growth factor-stimulated epithelial morphogenesis.
        Mol. Cell. 2003; 12: 1275-1285
        • Yu X.
        • Miyamoto S.
        • Mekada E.
        Integrin α 2 β 1-dependent EGF receptor activation at cell-cell contact sites.
        J. Cell Sci. 2000; 113: 2139-2147
        • Alexi X.
        • Berditchevski F.
        • Odintsova E.
        The effect of cell-ECM adhesion on signalling via the ErbB family of growth factor receptors.
        Biochem. Soc. Trans. 2011; 39: 568-573
        • Vieira A.V.
        • Lamaze C.
        • Schmid S.L.
        Control of EGF receptor signaling by clathrin-mediated endocytosis.
        Science. 1996; 274: 2086-2089
        • Mayor S.
        • Pagano R.E.
        Pathways of clathrin-independent endocytosis.
        Nat. Rev. Mol. Cell Biol. 2007; 8: 603-612
        • Benesch S.
        • Polo S.
        • Rottner K.
        • et al.
        N-WASP deficiency impairs EGF internalization and actin assembly at clathrin-coated pits.
        J. Cell Sci. 2005; 118: 3103-3115
        • Engqvist-Goldstein Å.E.Y.
        • Drubin D.G.
        Actin assembly and endocytosis: from yeast to mammals.
        Annu. Rev. Cell Dev. Biol. 2003; 19: 287-332

      Linked Article

      • Pinning Down the EGF Receptor
        • Thomas M. Jovin
        Biophysical JournalDecember 02, 2014
        • In Brief
          According to leading investigators in the field of cellular signal transduction, the epidermal growth factor (EGF) receptor (EGFR, ErbB1, HER1), ubiquitously encountered in signaling mechanisms and thus in human tumors, is the best studied yet least prototypic of receptor tyrosine kinases in general (1). This perception arises primarily from the fact that activation of the EGFR is thought to be conformational/allosteric, i.e., not requiring covalent modifications at or near the active site (2). However, a number of outstanding, perplexing questions exist that might still place the EGFR in the well known but poorly understood category.
        • Full-Text
        • PDF
        Open Archive
      Advertisement