Identification of Endosomal Epidermal Growth Factor Receptor ...

Research

Identification of Endosomal Epidermal Growth Factor Receptor Signaling Targets by Functional Organelle Proteomics*□ S

Taras Stasyk‡, Natalia Schiefermeier‡, Sergej Skvortsov§, Heinz Zwierzina§, Johan Pera¨nen¶, Guenther K. Bonn储, and Lukas A. Huber‡** Epidermal growth factor (EGF) receptor (EGFR) signal transduction is organized by scaffold and adaptor proteins, which have specific subcellular distribution. On a way from the plasma membrane to the lysosome EGFRs are still in their active state and can signal from distinct subcellular locations. To identify organelle-specific targets of EGF receptor signaling on endosomes a combination of subcellular fractionation, two-dimensional DIGE, fluorescence labeling of phosphoproteins, and MALDITOF/TOF mass spectrometry was applied. All together 23 EGF-regulated (phospho)proteins were identified as being differentially associated with endosomal fractions by functional organelle proteomics; among them were proteins known to be involved in endosomal trafficking and cytoskeleton rearrangement (Alix, myosin-9, myosin regulatory light chain, Trap1, moesin, cytokeratin 8, septins 2 and 11, and CapZ␤). Interestingly R-Ras, a small GTPase of the Ras family that regulates cell survival and integrin activity, was associated with endosomes in a ligand-dependent manner. EGF-dependent association of R-Ras with late endosomes was confirmed by confocal laser scanning immunofluorescence microscopy and Western blotting of endosomal fractions. EGFR tyrosine kinase inhibitor gefitinib was used to confirm EGF-dependent regulation of all identified proteins. EGF-dependent association of signaling molecules, such as R-Ras, with late endosomes suggests signaling specification through intracellular organelles. Molecular & Cellular Proteomics 6:908 –922, 2007.

The classical view on the mechanism of epidermal growth factor (EGF)1 receptor (EGFR) signaling, as well as the recepFrom the ‡Biocenter, Division of Cell Biology, and §Department of Internal Medicine, Innsbruck Medical University, A-6020 Innsbruck, Austria, ¶Institute of Biotechnology, University of Helsinki, Fi-00014 Finland, and 储Institute of Analytical Chemistry and Radiochemistry, Leopold-Franzens University, A-6020 Innsbruck, Austria Received, December 7, 2006, and in revised form, February 7, 2007 Published, MCP Papers in Press, February 10, 2007, DOI 10.1074/ mcp.M600463-MCP200 1 The abbreviations used are: EGF, epidermal growth factor; EGFR, epidermal growth factor receptor; PNS, postnuclear supernatant; TCL, total cell lysate; ERK, extracellular signal-regulated kinase; MAPK, mitogen-activated protein kinase; 2D, two-dimensional; 2DE, two-dimensional gel electrophoresis; HB, homogenization buffer;

908

Molecular & Cellular Proteomics 6.5

tor tyrosine kinase family in general, considered receptor activation at the plasma membrane, which initiates downstream signaling cascades with subsequent shutting down of the signaling by receptor endocytosis and degradation. However, the concept of an extended signaling pathway through the internalized, but still active, EGFR is now appreciated (1). EGFR signal transduction is compartmentalized at the plasma membrane and on endosomes. On a way from the plasma membrane to the lysosome EGF receptors are still in their active state and co-localize with signaling units on endosomal compartments and therefore can signal from distinct subcellular locations. The signaling units are organized by scaffold and adaptor proteins, which have specific subcellular distribution (2, 3). Two scaffold proteins, kinase suppressor of Ras (KSR1) and the MEK1 partner (MP1), are known to assemble ERK complexes at distinct subcellular locations. KSR1 regulates the formation of a MAPK signaling unit at the plasma membrane upon EGF stimulation (4). The MP1 scaffold forms a stable heterodimeric complex with the adaptor protein p14 and assembles an ERK cascade on late endosomes (5, 6). Endosomal localization of the p14-MP1 complex is exclusively required for the activation of ERK on late endosomes (3, 6, 7). EGFR signaling complexes can be formed specifically also on early endosomes. Activated EGFRs can internalize into a population of early endosomes carrying the Rab5 effector APPL1, which undergoes EGF-dependent nucleocytoplasmic shuttling required for cell proliferation (8). The contribution of the spatial and temporal regulation of ERK to specific signaling outcomes is unclear. This is not defined largely due to a lack of knowledge concerning specific targets of endosomal EGFR signaling. About 200 targets of EGF are known (9 –12). Presently 177

BVA, Biological Variation Analysis; TfR, transferrin receptor; TER, transitional endoplasmic reticulum; PPR, pentatricopeptide repeat; MLC, myosin light chain; BAG, BCL2-associated athanogene; hnRNP, heterogeneous nuclear ribonucleoprotein; MEK, mitogenactivated protein kinase/extracellular signal-regulated kinase kinase; SILAC, stable isotope labeling by amino acids in cell culture; MLCK, myosin light chain kinase; ERM, ezrin-radixin-moesin; NHERF, sodium-hydrogen exchanger regulatory factor; MTHFD1, methylenetetrahydrofolate dehydrogenase 1; PLC, phospholipase C; GFP, green fluorescent protein; PMF, peptide mass fingerprint; ANOVA, analysis of variance.

© 2007 by The American Society for Biochemistry and Molecular Biology, Inc. This paper is available on line at http://www.mcponline.org

EGFR Signaling Targets on Endosomes

molecules involved in EGFR signaling pathway are listed in the Human Protein Reference Database (www.hprd.org), but it is unclear how specificity of the signaling is generated. Downstream of the EGFR about 160 substrates have already been discovered for ERKs (for a review, see Ref. 13). Core kineses (Raf, MEK, and ERK) form a complex network of interacting proteins within MAPK pathways as well as in cross-talk with other signaling cascades, thereby regulating different cellular processes. To identify EGFR signaling targets on endosomes a combination of subcellular fractionation and a functional proteomics approach based on two-dimensional DIGE, fluorescence labeling of phosphoproteins, and MALDI-TOF/TOF mass spectrometry was applied here. The major advantage of 2DE over gel-independent mass spectrometry-based techniques for functional proteomics analysis is the ability to easily detect the proteins of interest, e.g. regulated by a growth factor, while ignoring thousands of other species in a biological sample following identification. This is even of more importance for the protein samples of reduced complexity such as cell organelles. The total number of different constituents of an organelle is rather limited to a few thousands protein species; therefore many low abundance organelle-specific proteins can be detected in the high resolution 2D gel. The combination of several complementary approaches in this study (well established subcellular fractionation and a high sensitivity 2D DIGE technique combined with phosphospecific fluorescent staining with subsequent identification of proteins by sensitive MALDI-TOF/TOF mass spectrometry) allowed us to track dynamic changes in endosomes after stimulation of cells with growth factor and formation of signaling complexes, which are otherwise difficult to detect because such associationdissociation events can be very transient. Furthermore proteins that shuttle between the cytoplasmic space and the endosome are mostly peripheral and not transmembrane proteins and as such are well suited for 2DE-based separation. Stimulation of EpH4 mouse mammary epithelial cells with EGF caused a ligand-dependent association or dissociation of several proteins with endosomes. In addition, phosphorylation of organelle-specific proteins could be detected after the ligand treatment. Among identified proteins many are known to be involved in endosomal trafficking and cytoskeleton rearrangement (Alix, myosin-9, myosin regulatory light chain, Trap1, moesin, cytokeratin 8, septins 2 and 11, and CapZ␤) and signaling (R-Ras, G␣q, G␤1, and G␤1 subunits of heterotrimeric G-protein complex), and several proteins are known to be involved in the regulation of general metabolism. EGFR tyrosine kinase inhibitor gefitinib (ZD1839, Iressa) was used to confirm EGF-dependent regulation of identified proteins. In addition, phosphorylation was shown for several of the identified proteins by a recently established method for quantitative detection of phosphoproteins using a combination of DIGE and phosphospecific fluorescent staining (14).

EXPERIMENTAL PROCEDURES

Cell Culture—EpH4 mouse mammary epithelial cells (15) were cultured in high glucose Dulbecco’s modified Eagle’s medium supplemented with 5% fetal bovine serum, 100 units/ml penicillin, 100 units/ml streptomycin, 10 mM HEPES, pH 7.3. Media and reagents for tissue culture were purchased from Invitrogen. The cells were cultivated on 150-mm tissue culture plastic dishes (Sarstedt, Newton, NC) at 37 °C in 5% CO2 and 98% humidity as described previously (15). Cells were serum-starved for 16 h and stimulated with 100 ng/ml EGF (Sigma-Aldrich) for 5 and 40 min or mock-treated as a control. In some experiments before EGF stimulation cells were pretreated with the EGFR tyrosine kinase inhibitor gefitinib (ZD1839, Iressa) at 1 ␮M concentration for 30 min. Subcellular Fractionation—The isolation of endosomes has been described recently in detail (16 –18). Briefly cells were washed three times with ice-cold PBS, scraped, and pelleted at 200 ⫻ g for 5 min (see Fig. 2). The cell pellet was rebuffered in homogenization buffer (HB) consisting of 250 mM sucrose in 3 mM imidazole, pH 7.4, 1 mM EDTA, protease inhibitors (10 mg/ml aprotinin, 1 mg/ml pepstatin, 10 mg/ml leupeptin, 1 mM Pefabloc SC (Roche Applied Science)), and phosphatase inhibitors (1 mM Na3OV4, 5 mM Na4P2O7, 50 mM NaF, 10 mM ␤-glycerophosphate) and again pelleted at 1,300 ⫻ g for 10 min. The cells were resuspended in HB supplemented with 0.03 mM cycloheximide (Sigma) and homogenized by three passes through a 22-gauge needle (19). Postnuclear supernatant (PNS) was obtained by centrifugation at 2,000 ⫻ g for 10 min at 4 °C. For the isolation of crude endosomal fraction the sucrose concentration of the PNS was adjusted to 40.6% by adding 62% sucrose (1:1.2, v/v), loaded on the bottom of an SW41 ultracentrifuge tube (Beckman), and overlaid with 35% sucrose in HB (7 ml). HB was added to fill the tube (3 ml), and the sample was centrifuged at 100,000 ⫻ g for 3 h at 4 °C. Crude endosomal fractions were collected from the interphase between 35% sucrose and HB. For isolation of late and early endosomes the PNS was loaded on top of a continuous sucrose gradient (10 – 40% sucrose in HB) that had been prepared in an SW41 tube using a BioComp gradient master (BioComp Instruments, Fredericton, New Brunswick, Canada). After centrifugation (100,000 ⫻ g at 4 °C for 16 h) 22 fractions were collected using an Auto Densi-Flow fraction collector (Labconco Corp., Kansas City, MO). The distribution of membranes from endosomal organelles has been determined by Western blotting using antibodies against specific marker proteins (17). To ensure the clearance of possible residual cytosolic proteins, pooled fractions of late, early, and crude endosomes were diluted 1:1 in 3 mM imidazole, pH 7.4, 1 mM EDTA to reduce the sucrose concentration (17) and centrifuged for 1 h at 100,000 ⫻ g. The organelle pellets were resuspended in the same buffer, and protein was precipitated using chloroform-methanol according to Wessel and Flugge (20). Two-dimensional Difference Gel Electrophoresis—Protein samples were solubilized in 2D DIGE sample buffer: 7 M urea, 2 M thiourea, 4% CHAPS, 30 mM Tris, buffered to pH 8.5. Protein concentration was determined using the Coomassie Plus protein assay (Pierce). Then 20 ␮g of protein in 20 ␮l of sample buffer were labeled with 160 pmol of CyDye DIGE Fluor minimal dyes (Amersham Biosciences) according to the manufacturer’s instructions (Cy3 and Cy5 for samples and Cy2 for internal control consisting of equal parts of all samples). Samples were applied by rehydration in IPG strips (18 cm, pH 3–10, non-linear; Amersham Biosciences) and subjected to isoelectrofocusing in an IPGphorTM IEF system (Amersham Biosciences) according to the manufacturer’s recommendations. Upon completion of the first dimension, strips were incubated in an equilibration buffer (50 mM Tris-HCl, pH 8.8, 6 M urea, 30% glycerol, 2% SDS, a trace of bromphenol blue) containing 1% DTT for 15 min and thereafter in the same buffer containing 2.5% iodoacetamide for 15 min. For the second


909


dimension, strips were transferred to the tops of 12.5% polyacrylamide gels and run for 12–14 h. After electrophoresis gels were scanned using a TyphoonTM 9410 imager at 100-dpi resolution (Amersham Biosciences). Scan settings were optimized for a maximum signal of ⬃85,000 counts. Images were cropped using ImageQuant 6.2 (GE Healthcare), and image analysis was performed using DeCyder 6.5 (GE Healthcare) DIA (Difference In-gel Analysis) software. Directly after scanning gels were fixed and stored at 4 °C. Four independent experiments were performed for each experimental setup (21) and statistically analyzed using DeCyder Biological Variation Analysis (BVA) software (GE Healthcare). Endosomal proteins regulated by EGF were considered for identification if p values after ANOVA test and/or corresponding t test were ⬍0.05. Detection of Phosphoproteins in DIGE Gels—Phosphoproteins were detected in gels by combination of 2D DIGE and phosphospecific fluorescent staining as described recently (14). Directly after scanning selected gels were fixed and poststained with Pro-Q Diamond phosphospecific fluorescent dye (Molecular Probes). The images of the same gel scanned before and after poststaining were compared using DeCyder BVA software (Amersham Biosciences). Phosphoproteins were detected as spots with increased Cy3-like fluorescence after Pro-Q Diamond poststaining in comparison with original Cy3 emission (14). MS and Protein Identification—For spot picking gels were poststained with ruthenium-based fluorescent protein dye (22) according to an improved staining protocol (23). Protein spots were excised from gels with an Ettan Spot Picker (Amersham Biosciences) and in-gel digested with trypsin (Promega, Madison, WI) as described by Hellman (24). The in-gel digests were concentrated and desalted using microZipTipC18 (Millipore, Billerica, MA) by elution of peptides with the acetonitrile solution containing the ␣-cyano-4-hydroxycinnamic acid (Fluka) as a matrix directly onto the target. Mass spectra were acquired using a MALDI-TOF/TOF Ultraflex instrument (Bruker Daltonics, Bremen, Germany). Peptide mass fingerprints (PMFs) and MS/MS spectra of selected precursor ions were interpreted with Mascot (Matrix Science) against the National Center for Biotechnology Information (NCBI) non-redundant protein database. Antibodies and Western Blotting—Anti-phospho-ERK1/2 and antiERK1/2 antibodies were obtained from Cell Signaling Technology (Berverly, MA). Anti-p14 antibody was described previously (6, 7). Anti-mouse-LAMP1 antibody was from Pharmingen. Anti-Rab5 and anti-Rab7 were gifts from Dr. Angela Wandinger-Ness, anti-TfR was a gift from Dr. Ian Trowbridge. Anti-R-Ras antibody was from Santa Cruz Biotechnology (Santa Cruz, CA), and anti-moesin was obtained from Transduction Laboratories. Western blotting was performed as published previously (17). Transfection—pEGFP-R-Ras wild type construct was described previously (25). Transfection was performed using 1 ␮g of plasmid DNA, Lipofectamine 2000 (Invitrogen), and Opti-MEM medium (Invitrogen) according to the manufacturer’s instructions. Immunofluorescence—EGF-treated and control EpH4 cells were briefly washed in ice-cold PBS and fixed for 20 min with ice-cold 3% paraformaldehyde in PBS following 30 min of blocking at room temperature. Blocking solution (0.01% saponin, 1% gelatin, and 10% goat serum in PBS) was further used for dilution of primary and secondary antibodies. Next cells were incubated at room temperature with primary anti-mouse CD107a (LAMP1) rat monoclonal antibody (Pharmingen) diluted 1:300 for 1 h following PBS washing and for 30 min with secondary antibody (goat anti-rat IgG AlexaFluor 594, Molecular Probes, 1:1000) following PBS washing. 4⬘,6-Diamidino-2phenylindole (Sigma) was used for nuclear staining. Finally coverslips were mounted in Mowiol (Calbiochem). Slides were analyzed either by an epifluorescence Axioplan 2 Imaging E system (Carl Zeiss GmbH) using AxioVision Release 4.5 SP1 software (Carl Zeiss GmbH) or by

910


FIG. 1. Biphasic character of ERK phosphorylation upon EGF treatment in EpH4 cells. Cells were serum-starved for 16 h and treated with 100 ng/ml EGF for the indicated time intervals. Total cell lysates were analyzed by Western blotting using specific antibodies against phosphorylated ERK1/2 (p-ERK1/2; A, top panel), then stripped, and reprobed with anti-ERK1/2 antibodies to verify equal protein loading (A, bottom panel). Quantification of the Western blots (B) revealed the second ERK activation peak (see also Supplemental Fig. 1). au, arbitrary units. confocal laser scanning fluorescence microscopy (LSM 510 META, Carl Zeiss GmbH) using Zeiss LSM Software, version 3.2. RESULTS

ERK Activation in EpH4 Cells—To establish conditions for functional proteomics analysis we first evaluated ERK1/2 activation upon EGF treatment of EpH4 mouse mammary epithelial cells in a time course experiment. Serum-starved cells were treated with EGF (100 ng/ml) for the indicated time intervals (Fig. 1), and total cell lysates were analyzed by Western blotting. EGF-induced phosphorylation of ERKs, with a main peak of activation at 5 min and a second smaller peak at 30 – 45 min of EGF treatment, was detected with an antibody that specifically recognizes the phosphorylated forms of ERK1/2 (Fig. 1A). The second ERK activation peak was roughly 20% of the main peak. The activated (phosphorylated) form of ERK could be well separated from the band of non-activated kinase (especially for ERK2 and to a lesser extent for ERK1) on Western blots when detected with antiERK antibodies (see Fig. 1B). By this roughly 75% of the total cellular pool of ERK was phosphorylated at 5 min of EGF stimulation, and about 40% was phosphorylated at the sec-


FIG. 2. Flowchart of functional organelle proteomics of EGFR signaling. EE, early endosomes; LE, late endosomes; CE, crude endosomes.

ond activation peak at 30 – 45 min of the EGF treatment. These observations are in good agreement with our previous data (7) where immunofluorescence analysis detected a colocalization of phospho-ERK1/2 with late endosomal markers (p14 and MP1) in HeLa cells that were stimulated for 30 min with EGF. We assumed that the second ERK activation peak might correspond to late endosomal ERK signaling, and therefore 5 and 40 min of EGF treatment (in comparison with untreated cells) were taken as time points for the functional organelle proteomics assay to detect EGF-specific targets on endosomes. Purification of Endosomes from EpH4 Cells—For purification of endosomes we used a subcellular fractionation protocol well established for this cell line based on differential centrifugation in sucrose gradients after mild homogenization and preparation of a PNS (17, 18). The procedure used here is summarized in a flowchart in Fig. 2. The crude endosomal fraction containing late and early endosomes was purified by

floatation in a discontinuous sucrose gradient consisting of 40.6, 35, and 8.6% sucrose and collected from the interphase between cushions of 35 and 8.6% sucrose solutions. Late and early endosomes float at these conditions in a narrow region of the sucrose gradient (Supplemental Fig. 1) as detected by marker proteins for late endosomes (p14, LAMP1, and Rab7) and for early endosomes (TfR and Rab5), respectively. Therefore, a fraction containing late and early endosomes could be efficiently enriched from the PNS. Interestingly, active ERK was also detected in the same fractions where endosomes were floating in the sucrose gradient (see Supplemental Fig. 1A, lower panel) but was clearly absent from neighboring fractions, indicating endosomal ERK signaling. The purification procedure used here was highly reproducible and allowed us to have protein amounts sufficient for subsequent 2D DIGE analyses (approximately 20 ␮g of organelle protein from a standard 150-mm tissue culture plastic dish). Because of high reproducibility of the purification procedure the crude endosomal fraction was used as a main sample in this study. Additionally late and early endosomes were purified from PNS by differential centrifugation in a continuous sucrose gradient as described under “Experimental Procedures” (see also Fig. 1). The distribution of membranes from different organelles has been determined earlier (17). Fractions were pooled according to distribution of specific endosomal marker proteins (Supplemental Fig. 1B): fractions from 6 to 10 for late endosomes (enriched for p14, LAMP1, and Rab7) and fractions 13–16 for early endosomes (TfR and Rab5 as markers). For 2D DIGE analyses these fractions were pooled from three gradients derived from three independent experiments to have representative samples and sufficient protein amounts (15–20 ␮g of protein). Therefore, late endosomes could be efficiently separated from early endosomes in a continuous sucrose gradient. New EGF-regulated Proteins on Endosomes Revealed by DIGE—Specifically we looked at proteins that were present either in higher or lower amounts on purified endosomal membranes after 40 min of EGF treatment. In addition, we also investigated which of those differentially associated proteins were phosphorylated. Because this screen was mainly directed to detect peripheral membrane proteins that can associate with or dissociate from endosomes to the cytosolic space in response to EGF treatment, 2D DIGE was the technology of choice. All together 23 different proteins (in 25 protein spots) were found to be EGF-regulated on endosomes and identified by MALDI-TOF/TOF mass spectrometry (Fig. 3 and Table I). Fourteen proteins were associated with endosomes in response to EGF treatment, whereas nine proteins were dissociated from endosomes (Table II). We did not filter the data based on -fold changes and included in Table II all proteins that were found to be differentially associated with endosomes upon EGF treatment after statistical analysis of four independent experiments performed for each setup (21). The


911


FIG. 3. Proteins regulated on endosomes by EGF and identified by mass spectrometry. EGF-induced association (shown in red) or dissociation (shown in green) of proteins with endosomes. EGF-dependent regulation of endosomal proteins was inhibited by gefitinib, an EGF receptor-specific tyrosine kinase inhibitor (shown in blue). Phosphoproteins were detected in gels by combination of 2D DIGE and phosphospecific fluorescent staining as described previously (14) (shown in yellow). PDC, pyruvate dehydrogenase complex; AcCoAH, acetyl-coenzyme A dehydrogenase; VCP, valosin-containing protein; K8, cytokeratin 8.

smallest changes we could detect in 2D DIGE experiments were from 1.3- to 1.5-fold increase or decrease for 12 protein spots (see Table II). Another 11 of 23 identified proteins were differentially associated with endosomes up to 2.2-fold difference. Interestingly almost all EGFR signaling-dependent changes in endosomal proteomes were detected after 40 min of EGF treatment. Only two proteins were associated (BCL2associated athanogene 2 (BAG-2)) or dissociated (LRP130) with endosomes after 5 min but not after 40 min of EGF treatment. Additionally 10 of the regulated proteins were found to be phosphorylated by poststaining of 2D DIGE gels with Pro-Q Diamond phosphospecific fluorescent stain (14). The identified proteins can be divided into four groups according to their functions in cells (see Tables I and II). The first group contains nine proteins involved in endosomal trafficking and cytoskeleton rearrangement. Myosin heavy chain 9 and myosin regulatory light chain, two proteins involved in organelle traffic (26), were ⬃2-fold increased on endosomes after 40 min of EGF treatment. Both of them were detected here as phosphoproteins. Alix, involved in concentration and sorting of cargo proteins of the multivesicular body (27), was

912


dissociated from endosomes; no phosphorylation could be detected. Moesin, a member of the ezrin-radixin-moesin (ERM) family of proteins known to play a role in the linkage of actin to membrane ruffles, was less present on endosomes and was not phosphorylated (28). Cytokeratin 8, the protein that belongs to the intermediate filament family, was more present on endosomes and was phosphorylated. CapZ␤ (Factin capping protein ␤ subunit), which binds to the fast growing ends of actin filaments and is a component of dynactin complex, increased on endosomes and was not phosphorylated. LRP130 (leucine-rich PPR motif-containing protein), which may be involved in the integration of cytoskeletal actin and microtubule networks with vesicular trafficking (29), was detected in two protein spots that decreased on endosomes and was not phosphorylated. Septin 11 and septin 2, GTPbinding proteins that associate with cellular membranes and actin and microtubule cytoskeletons (30), were more present on endosomes; septin 2 was detected as a phosphoprotein. The second group of EGFR targets consists of four signaling proteins. One of them, Ras-related protein R-Ras (Harvey rat sarcoma oncogene, subgroup R), belonging to the small Ras GTPase superfamily, was more present on endosomes; no phosphorylation could be detected. The other three signaling molecules belong to heterotrimeric G-protein complex subunits: G␣q (guanine nucleotide-binding protein Gq, ␣ subunit) and G␤1 (G-protein ␤1 subunit) and G␤2 (G-protein ␤2 subunit). The last three proteins were dissociated from endosomes upon treatment of cells with EGF; all of them were also detected to be phosphorylated. Proteins with chaperone activity were combined in the third group, which consists of four proteins (Table I). Heat shock 70-kDa protein 4-like, which plays a critical role in protein folding and trafficking (31), was dissociated from endosomes and phosphorylated. Included in this group of proteins was also the transitional endoplasmic reticulum (TER) ATPase, a protein also known as valosin-containing protein. It is involved in the formation of the transitional endoplasmic reticulum and is necessary for the export of misfolded proteins from the endoplasmic reticulum to the cytoplasm for proteasomal degradation. Valosin-containing protein was less present on endosomes isolated from EGF-stimulated cells and was detected as a phosphoprotein. Trap1 (tumor necrosis factor receptorassociated protein 1, HSP75), a member of the molecular chaperone HSP90 family, was slightly increased on endosomes; no phosphorylation of Trap1 could be detected. BAG-2, which inhibits the chaperone activity of HSP70/HSC70 by promoting substrate release, was associated with endosomes after 5 min of EGF treatment and was phosphorylated. The fourth group consists of six proteins involved in the regulation of general metabolism: methylenetetrahydrofolate dehydrogenase 1-like, the enzyme of folic acid biosynthesis; dihydrolipoamide S-acetyltransferase of pyruvate dehydrogenase complex; glycolytic enzyme pyruvate kinase M2; aldehyde dehydrogenase, an enzyme that converts retinalde-


TABLE I Proteins, identified by MALDI-TOF/TOF MS, differentially associated with endosomes in response to EGF treatment of EpH4 cells All proteins identified in this study are grouped according to their function and listed together with their accession numbers, molecular weights, and pI. Mascot scores of statistical significance of protein identification by PMFs, are indicated together with the sequence coverage and the number of matched peptides (in parentheses). MS/MS scores are shown for some proteins, identification of which was additionally confirmed by MS/MS sequencing of selected peptides. Phosphorylation was detected for several of the identified proteins by a combination of DIGE and phosphospecific fluorescent staining (marked as P) (14). EGF-dependent regulation of identified proteins was confirmed using EGFR tyrosine kinase inhibitor gefitinib (marked as “confirmed”); also see the supplemental table. No.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

Protein identity

NCBI accession no.

Endosomal trafficking and cytoskeleton rearrangement Myosin-9 (Myh 9) gil47847498 LRP130 (leucine-rich PPR protein) gil21389320 LRP130 (leucine-rich PPR protein) gil21389320 Allx (PDC6I, Aip1) gil6755002 Moesin gil199765 Cytokeratin-8 (K8) gil309215 Septin 11 gil57634518 Septin 2 (NEDD5) gil6754816 CapZ␤ (F-actin capping protein) gil1345668 MLC gil71037403 Signaling R-Ras (Ras-related protein R-Ras, p23) gil6677819 G␣q gil193502 gil6680045 G␤1 (transducin ␤ chain 1) G␤2 (transducin ␤ chain 2) gil984551 Chaperone activity Heat shock 70-kDa protein 4-like (HS74L) gil40254361 TER ATPase (valosin-containing protein) gil55217 Trap1 (HSP75) gil13879408 BAG-2 gil21703784 General metabolism MTHFD1I gil20987924 Dihydrollpoamide S-acetyltransferase (PDC-E2) gil31542559 Pyruvate kinase M2 (PK-3) gil31981562 Aldehyde dehydrogenase (ALDH2) gil13529509 Aldehyde dehydrogenase (ALDH2) gil13529509 Acetyl-coenzyme A dehydrogenase gil31982520 Adenylate kinase 2 gil34328230

hyde to retinoic acid; acetyl-coenzyme A dehydrogenase, an enzyme of lipid metabolism; and adenylate kinase 2, an ATPAMP transphosphorylase that is essential for maintenance and cell growth. All proteins of this group were more present on endosomes after 40 min of EGF treatment with an exception of pyruvate kinase M2, which was decreased in the endosomal fraction at this time point. None of those proteins was found to be phosphorylated. The 2D DIGE analysis of crude endosomes isolated from cells treated with EGF revealed 23 proteins differentially associated with endosomes (see Tables I and II and the supplemental table). To localize more specifically with which subclass of endosomes the detected proteins were differentially associated, additional experiments were performed with late and early endosomes purified by differential centrifugation in continuous sucrose gradients (see “Experimental Procedures” and Fig. 2). Importantly, eight of 23 protein proteins identified on crude endosomes were found to be regulated on late endosomes, and seven proteins were regulated on early

Molecular weight

pI

Mascot score

PTM

Gefitinib

155.6 148.6 148.6 96.5 69.4 53.2 49.6 41.7 31.6 19.9

5.4 5.9 5.9 6.2 5.8 5.4 6.7 6.1 5.5 4.7

211 (23%, 24) 164 (18%, 18) 117 (14%, 15) 179 (27%, 15) 65 (10%, 8); 54 (MS/MS) 229 (50%, 32) 73 (22%, 11) 108 (37%, 13); 99 (MS/MS) 83 (14%, 6); 34 (MS/MS) 66 (56%, 10); 29 (MS/MS)

P

P

Confirmed Confirmed Confirmed Confirmed Confirmed Confirmed Confirmed Confirmed Confirmed Confirmed

23.9 42.4 38.2 38.1

6.3 5.6 5.6 5.6

168 121 104 162

P P P

Confirmed Confirmed Confirmed Confirmed

95.2 89.9 80.4 23.6

5.5 5.1 6.3 6.0

88 (11%, 8) 183 (40%, 34) 130 (27%, 17) 78 (25%, 4)

88.3 59.4 58.4 57.0 57.0 48.3 25.8

7.6 5.7 7.2 7.5 7.5 8.3 7.0

108 (215, 14) 62 (10%, 2); 59 (MS/MS) 246 (58%,27); 47 (MS/MS) 151 (37%, 18); 71 (MS/MS) 135 (35%, 16) 75 (22%, 13) 110 (53%, 12)

(55%, (35%, (36%, (53%,

11); 47 (MS/MS) 9); 56 (MS/MS) 11); 54 (MS/MS) 18); 57 (MS/MS)

P P

P P P

Confirmed Confirmed Confirmed Confirmed

Confirmed Confirmed

endosomes (Table II and the supplemental table). Additionally, one more protein, CapZ␤, was detected to be associated with late endosomes in response to EGF treatment. Association of this actin-binding cytoskeleton protein was very marginal on the level of crude endosomes. Interestingly, the differential association of six proteins (myosin-9, myosin light chain (MLC), cytokeratin 8, G␤1, G␤2, and BAG-2) was detected on all crude, late, and early endosomal preparations. Several proteins were found specifically to be associated (R-Ras and CapZ␤) with or dissociating (TER ATPase) from late endosomes or to be associated with early endosomes (cytokeratin 8). Moreover, comparison of purified late and early endosomes allowed us to roughly estimate the ratio of identified proteins on late versus early endosomes. Several proteins were enriched up to more than 6-fold in late endosomes (Table II) (septin 11, MLC, R-Ras, Trap1, one of the isoforms of ALDH2) or enriched up to 7-fold on early endosomes (e.g. myosin-9, LRP130, moesin, and septin 2), suggesting EGFR-regulated association of these proteins with specifically late or early endosomes.


913


TABLE II Proteins differentially associated with endosomes in response to EGF treatment of EpH4 cells EGF stimulation of cells for 5 and 40 min caused association or dissociation of proteins with/from endosomes as detected for crude (CE), late (LE), or early (EE) endosomal fractions. A positive ratio indicates an association of the protein with endosomes, and a negative value shows a dissociation of the protein. Four independent experiments were performed for each experimental setup (21) and statistically analyzed using DeCyder BVA software. p values after ANOVA test and corresponding t test are indicated for identified proteins and marked in bold if the values are ⬍0.05. Proteins are listed in the same order as in Table I. Relative abundance of EGF-regulated proteins in the corresponding endosomal fraction (LE versus EE) is presented in the last column. Ac-CoA DH, acetyl-CoA dehydrogenase; 1-ANOVA, one-way ANOVA. No.

Protein identity

1

Myosin-9

2 3 4 5 6

LRP130 LRP130 Allx (PDC6I, Aip1) Moesin Cytokeratin-8

7 8

Septin 11 Septin 2 (NEDD5)

9

CapZ␤

10

MLC

11

R-Ras (p23)

12 13

G␣q G␤1

14

G␤2

15 16

Heat shock 70–4 TER ATPase

17 18

Trap1 BAG-2

19 20 21 22 23 24 25

MTHFD1L PDC-E2 Pyruvate kinase M2 ALDH2 ALDH2 Ac-CoA DH Adenylate kinase 2

Organelle

1-ANOVA

CE LE EE CE CE CE CE CE LE EE CE CE EE CE LE CE LE EE CE LE CE CE LE EE CE LE EE CE CE LE CE CE LE EE CE CE CE CE CE CE CE

0.001 0.001 0.055 0.031 0.039 0.049 0.056 0.041 0.017 0.130 0.002 0.024 0.071 0.085 0.019 0.062 0.005 0.004 0.130 0.024 0.160 0.001 0.027 0.033 0.004 0.029 0.055 0.015 0.047 0.100 0.000 0.001 0.045 0.011 0.042 0.014 0.047 0.002 0.012 0.021 0.012

Only three of 23 proteins that differentially associated with endosomes upon EGF treatment were already reported as targets of EGFR signaling: cytokeratin 8 (32, 33), MLC (34), and Alix (9). Although this existing knowledge in the literature confirmed our data, it also raised the question why the other proteins identified here on endosomal membranes have escaped identification as EGFR targets up to now? One possible explanation is the relative abundance of proteins in total cell lysates because most EGFR targets studies did not take into consideration subcellular fractionation. To further investigate this possible bypass we performed the identical experiment but in total cell lysates and without

914


5-min EGF

40-min EGF

t test

Ratio

t test

Ratio

0.091 0.360 0.200 0.012 0.036 0.510 0.120 0.260 0.029 0.220 0.023 0.110 0.230 0.043 0.015 0.880 0.470 0.510 0.610 0.120 0.460 0.016 0.220 0.085 0.041 0.730 0.390 0.310 0.140 0.140 0.610 0.000 0.024 0.014 0.046 0.067 0.530 0.095 0.150 0.920 0.600

⫺1.03 1.07 1.29 ⴚ1.36 ⴚ1.44 ⫺1.04 ⫺1.31 1.23 1.46 1.36 1.11 1.45 1.62 1.12 1.45 1.09 1.13 1.12 1.09 1.35 ⫺1.19 ⫺1.14 ⫺1.13 ⫺1.15 ⫺1.25 ⫺1.04 ⫺1.09 ⫺1.13 1.02 ⫺1.13 ⫺1.02 1.43 1.31 1.35 1.34 1.26 ⫺1.09 1.11 1.16 ⫺1.01 ⫺1.05

0.005 0.001 0.039 0.072 0.065 0.075 0.049 0.033 0.011 0.048 0.002 0.012 0.019 0.068 0.029 0.034 0.003 0.002 0.044 0.017 0.002 0.003 0.019 0.037 0.007 0.012 0.039 0.004 0.023 0.043 0.000 0.410 0.510 0.098 0.021 0.011 0.030 0.004 0.008 0.012 0.033

2.23 1.81 2.02 ⫺1.25 ⫺1.12 ⴚ1.30 ⴚ1.64 1.43 1.57 1.80 1.26 1.84 2.18 1.18 1.66 1.83 1.81 1.82 1.34 1.94 ⴚ1.99 ⴚ1.39 ⴚ1.35 ⴚ1.33 ⴚ1.41 ⴚ1.30 ⴚ1.31 ⴚ1.54 ⴚ1.35 ⴚ1.32 1.28 1.07 1.07 ⫺1.00 1.42 1.57 ⴚ1.47 1.56 1.63 1.32 1.48

Protein In LE vs. EE 1:4.2 1:7.1 1:6.5 1:1.0 1:3.1 1:1.3 4.8:1 1:1.8 1.3:1 6.2:1 2.8:1 1.4:1 1:1.4 1:1.4 1:1 1:1.1 1.5:1 1:1 1:4.6 1:2.5 1.5:1 1:5.0 2.3:1 1:2.2 1:1.9

subcellular fractionation. In addition, this experiment was designed to see whether different targets from other major cellular compartments (plasma membrane, cytoskeleton, endoplasmic reticulum, nucleus, cytoplasm, etc.) could be identified. New EGF-regulated Proteins in Total Cell Lysates Revealed by DIGE—In 2D maps of total cell lysates (TCLs) prepared from EpH4 cells stimulated with EGF for 5 and 40 min, 15 protein spots were found to be differentially regulated (see Fig. 4). Table III lists the proteins labeled on the gel shown in Fig. 4 that were identified by MALDI-TOF/TOF mass spectrometry. For several spots the MALDI PMF data alone were not sufficient to unambiguously identify proteins; therefore,


FIG. 4. EGF-regulated proteins detected in total cell lysate and identified by mass spectrometry. EGF induced up-regulation (shown in red) or down-regulation (shown in blue) of proteins is shown. EGF-dependent regulation of proteins was inhibited by gefitinib, an EGF receptor-specific tyrosine kinase inhibitor (shown in green). Phosphoproteins were detected in gels by combination of 2D DIGE and phosphospecific fluorescent staining as described previously (14) (shown in yellow). PK, protein kinase; DH, dehydrogenase; K8, cytokeratin 8.

sequencing of selected peptides by MS/MS analysis (LIFT mode) was used to confirm these identifications. In TCL six previously known EGF-regulated proteins (MEK1, stathmin, cofilin, destrin, cytokeratin 8, and heterogeneous nuclear ribonucleoprotein (hnRNP)-K) and seven novel targets were identified, thereby principally confirming applicability and high sensitivity of the technique used here. Six of the targets were up-regulated (see Table III), whereas the other nine proteins were down-regulated upon stimulation of cells with EGF. Additionally phosphorylation of eight of 15 proteins was detected. Among known targets MEK1 is the core kinase, which activates ERK1 and ERK2 mitogen-activated protein kinases by the phosphorylation of a threonine residue and a tyrosine residue. It was detected to be up-regulated and phosphorylated. Stathmin, cofilin 1, and destrin were detected as downregulated spots here. Stathmin is a cytoplasmic protein involved in the regulation of the microtubule filament system; it binds to two ␣/␤-tubulin heterodimers. ERK1/2 are responsible for the phosphorylation of stathmin (35). Cofilin 1 and destrin (actin-depolymerizing factor, Sid 23) belong to the

actin-binding protein actin-depolymerizing factor family and reversibly control actin polymerization and depolymerization. Both proteins were also phosphorylated. Cytokeratin 8 was also found here to be up-regulated and phosphorylated. Cytokeratin 8 belongs to the intermediate filament family; phosphorylation on serine residues is enhanced during EGF stimulation and mitosis (32, 33). hnRNP-K, one of the major pre-mRNA-binding proteins, was recently identified as an EGFR target in a high throughput mass spectrometry-based proteomics analysis (12). Two isoforms of this protein were found to be differentially regulated: hnRNP-K-a was up-regulated, but hnRNP-K-b was down-regulated; both isoforms were phosphorylated. It was shown previously that phosphorylation of hnRNP-K by ERK leads to its cytoplasmic accumulation and inhibition of mRNA translation (36). The other seven proteins identified here in TCL have not been identified previously as targets of EGFR signaling. Two proteins were up-regulated after 5 min of EGF treatment. One of them was serine/threonine-protein kinase TIK, also known as protein kinase RNA-activated or p68 kinase. Another one was NADH dehydrogenase 1 ␣ subcomplex subunit 5, which transfers electrons from NADH to the respiratory chain in mitochondrion. Phospholipase C-␦-1 was phosphorylated and up-regulated after 40 min of EGF treatment. Three proteins were found to be down-regulated after 40 min of EGF treatment: integrin ␣-3 (a subunit of integrin ␣-3/␤-1 receptor for fibronectin, laminin, collagen, epiligrin, thrombospondin, and CSPG4), plastin-2 (involved in actin filament bundle formation), and dynein intermediate chain 2, which is involved in the assembly of the dynactin complex. Actin-binding protein zyxin was down-regulated at 5 min of cell stimulation, and it was detected as phosphoprotein. In summary, cytokeratin 8 was the only protein of 23 EGFR targets on endosomes detectable at the level of TCL. Hence organelle purification opened up a so far unexplored level of sensitivity in the detection of new EGFR signaling target proteins. Confirmation with Gefitinib—The selective EGFR tyrosine kinase inhibitor gefitinib was used here to confirm EGF-dependent regulation of identified proteins. Before EGF stimulation cells were pretreated with gefitinib at 1 ␮M concentration (37). Hence 30 min of cell pretreatment with gefitinib was found to be optimal to inhibit ERK activation to the same level as that in control untreated cells (see Supplemental Fig. 2). EGF-dependent regulation (association/dissociation or phosphorylation) of 19 of 23 detected proteins in endosomal fractions (Table II; also see the supplemental table) and 13 of 15 (Table II) regulated proteins in TCL was inhibited by gefitinib, therefore confirming these results. Confirmation by Western Blotting—To further validate the proteomics results we compared them with those generated by Western blotting with specific antibodies against selected proteins. As demonstrated in Fig. 5, anti-R-Ras and antimoesin Western blots confirmed the 2D DIGE results. Namely


915


TABLE III Summary of EGF-regulated proteins in total cell lysates of EpH4 cells identified by MALDI-TOF/TOF MS Identified proteins are listed together with their accession numbers, molecular weights, and pI. Mascot scores of statistical significance of protein identification by PMFs are presented together with sequence coverage and number of matched peptides (in parentheses). MS/MS scores are shown for some proteins, identification of which was additionally confirmed by MS/MS sequencing of selected peptides. Phosphorylation was detected for several of the identified proteins by a combination of DIGE and phosphospecific fluorescent staining (marked as P) (14). EGF-dependent regulation of identified proteins was confirmed using EGFR tyrosine kinase inhibitor gefitinib (marked as “confirmed”); also see, the supplemental table. EGF stimulation of cells for 5 and 40 min caused up- (positive ratios) or down-regulation (negative values) of proteins. The ratios are marked in bold if the corresponding p values are ⬍0.05. p values after ANOVA test and corresponding t test are indicated for identified proteins according to statistical analysis using DeCyder BVA software. DH, dehydrogenase; PK, protein kinase, 1-ANOVA, one-way ANOVA. No.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Protein identity

NCBI accession no.

pI

Molecular weight

5-min EGF Mascot score

916


PTM

Gefitinib

P P P P P P

Confirmed

t test Ratio t test Ratio

Known effectors of EGFR signaling hnRNP-K-a gil13384620 5.4 51.2 106 (25%, 9); 88 hnRNP-K-b gil14165435 5.4 51.2 127 (34%, 17) Cytokeratin 8 gil309215 5.4 53.2 64 (MS/MS) MEK1 gil6678794 6.2 43.8 89 (26%, 9) Cofllin 1 gll6680924 8.2 18.8 166 (65%, 12); 85 Destrin gil71059761 8.1 18.9 90 (50%, 7) Stathmin gil9789995 5.8 17.3 36, 29 (MS/MS) Proteins not previously reported in EGFR signaling Integrin ␣-3 gil31418683 6.1 117.9 76 (14%, 9) Integrin ␣-3 gil31418683 6.1 117.9 71 (10%, 8) PLC-␦-1 gil19387967 5.8 86.6 126 (23%, 13) Plastin 2 gil31543113 5.2 70.7 328 (52%, 32) Dynein 2 gil6753658 5.2 68.8 122 (43%, 17); 64 Zyxin gil6756085 6.0 61.8 73 (19%, 7); 18 PK TIK gil201068 8.8 59.0 64 (12%, 6) NADH DH 1 ␣, 5 gil20306389 7.8 13.4 101 (78%, 7)

R-Ras was increased in crude endosomes upon EGF stimulation, and this association was inhibited by gefitinib. Western blotting with anti-moesin antibodies confirmed dissociation of moesin from organelles after 40 min of EGF treatment. Interestingly, moesin was detected on blots as a double band suggesting post-translational modifications of the protein that are most likely not phosphorylations because the protein did not show up in the phosphospecific 2D DIGE staining applied here. However, only the upper band was decreased in an EGF-dependent manner (Fig. 5B, arrow). Interestingly, the same upper moesin band, which was dissociated from endosomes in response to EGF stimulation, was increased in cytosol at the 40-min time point of growth factor treatment (Fig. 5B, arrow). Importantly, dissociation of moesin from organelle to cytosol was inhibited by gefitinib, thereby confirming EGFdependent dissociation of this protein from endosomes. EGFdependent dissociation/association of moesin seems to be specific to endosomal organelles because it was not detected at the level of total membrane fractions (see Fig. 5). R-Ras Is Associated with Late Endosomes upon EGF Stimulation—Interestingly R-Ras, a small GTPase of the Ras family, was found in 2D DIGE analyses to be associated with late endosomes in a ligand-dependent manner (Table II). R-Ras regulates cell survival as well as cell adhesion, spreading, and phagocytosis by activating integrins (38, 39). As further confirmation of our functional organelle proteomics data, immunofluorescence experiments were performed with R-Ras.

40-min EGF

1-ANOVA

0.002 0.024 0.011 0.019 0.004 0.010 0.079

0.250 0.580 0.028 0.730 0.018 0.120 0.068

1.06 ⫺1.04 1.31 ⫺1.02 ⫺1.18 ⫺1.07 ⫺1.37

0.001 0.018 0.022 0.031 0.008 0.009 0.050

1.28 ⴚ1.31 1.63 1.17 ⴚ1.38 ⴚ1.21 ⴚ1.38

0.006 0.000 0.027 0.003 0.022 0.006 0.009 0.002

0.410 0.019 0.063 0.190 0.750 0.008 0.012 0.002

⫺1.06 ⫺1.13 1.18 ⫺1.11 ⫺1.03 ⴚ1.45 1.40 1.33

0.003 0.001 0.046 0.005 0.039 0.003 0.160 0.013

ⴚ1.29 ⴚ1.35 1.29 ⴚ1.35 ⴚ1.33 ⴚ1.37 1.12 1.28

Confirmed Confirmed Confirmed Confirmed Confirmed Confirmed Confirmed

P P

Confirmed Confirmed Confirmed Confirmed Confirmed

EpH4 cells were transfected with pEGFP-labeled R-Ras, treated with EGF, and processed for immunofluorescence and confocal laser scanning microscopy. There was no detectable co-localization of pEGFP-R-Ras with the late endosomal marker LAMP1 in starved transfected cells (see Fig. 6). After 40 min of EGF stimulation R-Ras was co-localized with LAMP1, thereby confirming 2D DIGE results. Very similar results were obtained by epifluorescence microscopy (Supplemental Fig. 5), which confirmed association of R-Ras with late endosomes after 40 min of EGF treatment. Interestingly upon EGF stimulation, R-Ras was found to be distributed on the surface of late endosomal organelles (see Supplemental Fig. 5) where it co-localized with LAMP1. EGF-dependent association of R-Ras with late endosomes suggests a possibility of R-Ras signaling from the late endosomal compartment. Organelle proteomics provided sensitivity sufficient for detecting low abundance signaling proteins such as R-Ras that associate with the late endosomal compartment upon EGF treatment of EpH4 cells. DISCUSSION

Once activated at the plasma membrane EGFRs can signal from distinct subcellular locations by co-localizing with signaling units on distinct compartments (1, 3). EGFR signal transduction is compartmentalized by adaptor and scaffold proteins at the plasma membrane (4) and on endosomes (5, 6). To identify EGFR signaling targets on endosomes we used


FIG. 5. R-Ras was found in 2D DIGE analyses to be associated with late endosomes in an EGF-dependent manner. Shown are graph views exported from DeCyder BVA showing association of R-Ras detected at the level of crude endosomal fraction (A) or late endosomes (C); EGF-regulated association of R-Ras with endosomes was inhibited by gefitinib (B). Three-dimensional representations of the R-Ras spot from 2D DIGE maps of late endosomes purified from EpH4 cells treated with EGF are shown in D. Regulation of selected endosomal proteins was confirmed by Western blot analysis with anti-R-Ras and anti-moesin specific antibodies (E). P-ERK, phospho-ERK.


917


FIG. 6. GFP-R-Ras co-localizes with the late endosomal marker LAMP1 after EGF stimulation. Confocal laser scanning microscopy images of the GFP-R-Ras-expressing EPH4 cells that were stimulated with EGF for 0 min (A–C) and 40 min (D–F). A and D, immunofluorescence experiment with LAMP1 antibody. B and E, GFP-R-Ras. C and F, merged images. Note the co-localization of GFP-R-Ras and LAMP1 after 40 min of EGF stimulation (F).

a combination of subcellular fractionation and a functional proteomics approach based on 2D DIGE. Stimulation of EpH4 mouse mammary epithelial cells with EGF caused a liganddependent association or dissociation of 23 different proteins with endosomes. However, only three of them were described previously as EGFR targets. Additionally, phosphorylation was shown for 10 endosomal targets by a combination of DIGE and phosphospecific fluorescent staining (14). Using EGFR tyrosine kinase inhibitor gefitinib we confirmed EGFdependent regulation of 19 of the identified proteins here. Mass spectrometry-based high throughput proteomics has been extensively used to characterize the proteomes of several cellular organelles (40, 41). However, presently only a few functional organelle proteomics studies have been performed so far, e.g. by using SILAC technology to study nucleolar proteome dynamics (42). Here we used EpH4 mouse mammary epithelial cells to study EGFR signaling. These cells form polarized epithelial layers and are functionally and morphologically still very close to primary epithelial cells (15). In addition, we have established previously protocols for the purification of subcellular organelles, i.e. endosomes from cultures of EpH4 cells (17, 18, 43). Purity of isolated organelles is essential for comprehensive analysis of total organelle proteomes, but complete purification is almost impossible (16, 44). However, for functional proteomics studies (e.g. when two or more differentially treated samples are compared) even enrichment of organelles or certain subcellular fractions shall clearly be an advantage for detection of low abundance proteins and tracking their changes after stimulation of cells. Usually results obtained by 2DE analyses are validated by independent techniques, such as Western blotting, provided that specific and suitable antibodies are available. As can be seen in Fig. 5, the Western blotting results were consistent with the results from the 2D DIGE experiment for selected proteins tested. To confirm EGF-dependent regulation of all

918


protein detected in our experiments we used the selective EGFR tyrosine kinase inhibitor gefitinib (37). EGF-dependent regulation of 19 of 23 detected proteins in the endosomal fraction and 13 of 15 regulated proteins in total cell lysates was inhibited by gefitinib. Three proteins of the 23 identified have been reported previously to be affected by EGF signaling. EGF-dependent regulation of cytokeratin 8 has been demonstrated using classical biochemical methods (32), and recently cytokeratin 8 was also found by SILAC proteomics strategy in specific association with EGFR complexes (33). MLCs were shown to be phosphorylated by myosin light chain kinase (MLCK) in response to ERK1/2 activation (34). Myosin alkali chain isoform 1 was recently also detected by 2DE technique to be down-regulated upon EGF treatment in neonatal porcine pancreatic cells (45). Two recent mass spectrometry-based quantitative proteomics approaches identified EGF-dependent regulation of Alix, a protein that has been implicated in apoptosis and endosomal trafficking (9, 12). Proteins involved in endosomal trafficking of EGFR and EGF-stimulated rearrangements of the actin cytoskeleton were the major proteins affected by EGF stimulation on endosomes (myosin-9, myosin regulatory light chain, moesin, Alix, cytokeratin 8, septins 2 and 11, and CapZ␤). Moesin, a member of ERM family of proteins, was detected previously on early endosomes as a component of a complex with annexin II together with other elements of the cortical actin cytoskeleton, such as ␣-actinin, ezrin, and membrane-associated actin (28). Moesin is known to play a role in the linkage of actin to membrane ruffles, whereas ␣-actinin is an actinbundling protein found at the sites of actin membrane attachment. This suggested that annexin II is part of the interface between cholesterol-rich membranes such as endosomes or plasma membrane and the cortical cytoskeleton. Indirect association of moesin with EGFR on endosomes is also possible via NHERF-1, a sodium-hydrogen exchanger regulatory fac-


tor, also known as EBP50 (46). NHERF-1 is an adaptor protein that binds moesin and other ERM family cytoskeletal proteins. NHERF-1 also interacts with the EGFR to modulate its mitogenic signaling (47) by stabilizing EGFR at the cell surface. The N-terminal PDZ domain of NHERF specifically binds to an internal peptide motif located within the C-terminal regulatory domain of EGFR. The EGF-dependent dissociation of moesin from endosomes detected here could reflect a mechanism that controls endocytic sorting of membrane receptors, which involves PDZ domain-mediated protein interactions with the actin cytoskeleton as recently described for the ␤2-adrenergic receptor (48). Interestingly another PDZ domain-containing adaptor protein, syntenin, was detected in our previous organelle proteomics screen to be associated with early endocytic vesicular compartment (49). Two myosins, myosin-9 and MLC, were found here to be recruited to endosomes in response to EGF treatment. The more then 2-fold increased association of myosin-9 with endosomes was detected after 40 min of EGF stimulation; this is consistent with the current knowledge on organelle trafficking and endosomal development. Myosin-9, for instance, was suggested as a motor protein to be involved in actin-dependent movement away from late endocytic organelles (26). In the same study, overexpression of mVps18p (a member of the homotypic fusion and vacuole protein sorting complex) caused the clustering of late endosomes and lysosomes. The clusters were surrounded by components of the actin cytoskeleton, including actin, ezrin, and specific unconventional myosins. The specific subsets of myosins likely provide a means of moving organelles toward or away from each other. Mitogen-activated protein kinase activates MLCK, leading to phosphorylation of MLCs (34). Phosphorylation of myosin light chains by MLCK is a critical regulatory step in myosin function because it promotes myosin ATPase activity and polymerization of actin cables. Two septins were detected to be associated with endosomes in an EGF-dependent manner. Septins are GTP-binding proteins that associate with cellular membranes and the actin and microtubule cytoskeletons (30). Septins (humans have 13 septin genes) assemble into filamentous structures that compartmentalize cell membranes and act as diffusion barriers between different membrane domains and form molecular scaffolds for membrane- and cytoskeleton-interacting proteins at specific intracellular locations. Interestingly the identified septins here had a different distribution between late and early endosomes: phosphorylated septin 2 was enriched in early endosomes, and septin 11 was enriched in late endosomes (see Table II). In addition, septins regulate the localization of cytoskeleton-binding and motor proteins and bind target membranes and intracellular vesicles. Hence septins might influence vesicle movement and/or catalyze membrane fission and fusion. This might be significant in polarized epithelia and neurons in which vesicles must be targeted to specific domains (30). New reports link mammalian septins to

signaling pathways and mechanisms that regulate membrane and cytoskeleton organization and function, e.g. Cdc42 and Rho signaling (50). A recent study has shown that mammalian septin 2 contains multiple phosphorylation sites and is phosphorylated by Ser/Thr kinases such as protein kinase C (51). Alix (PDC6I, programmed cell death 6-interacting protein), identified here to be dissociated from endosomes upon EGF stimulation, was reported recently to be phosphorylated in response to EGF treatment of HeLa cells in a large scale phosphoproteomics approach (9). Alix regulates cortical actin and the spatial distribution of endosomes and controls the making of and trafficking through multivesicular bodies, which are crucial intermediates within the endolysosomal system (27). Interestingly HSP70-4-like protein and its possible interacting partner BAG-2 were found here to be regulated on endosomes by EGF, therefore suggesting an existence of such a complex on endosomes. The HSC70/HSP70 molecular chaperones require partner proteins for executing their functions in specific cellular compartments and cellular processes (52). BAG family proteins participate in targeting of HSC70/HSP70 to other proteins or protein complexes and control their subcellular distribution. BAG-3, another member of BCL2-binding protein family, was demonstrated recently to be a phosphoprotein, and EGF treatment of MDA-435 cells resulted in tyrosine phosphorylation of BAG-3 (53). In line with this study, we also detected EGF-dependent association with and phosphorylation of BAG-2 on endosomes after 5 min of EGF treatment. Moreover BAG-3 was shown to form an EGFregulated ternary complex with latent phospholipase C-␥ and HSP70/HSC70 (53). Interestingly immunoelectron microscopy in human colon carcinoma cells and cell fractionation of mouse embryonic fibroblasts revealed that a subpopulation of HSP70 is directly associated with the membrane of the endolysosomal compartment (54). HSP70 functions at the level of the lysosomes as a survival protein by stabilizing the lysosomes of cancer cells. Recently it was reported that LRP130 is an RNA-binding protein showing RNA-dependent interactions with shuttling hnRNPs, including A1 and K (55). A multifunctional role of LRP130 in multiprotein complexes was suggested from diverse sources generated by trapping and pulldown strategies; these complexes included hnRNP-K (also identified here in TCL as an EGF-regulated protein), several endocytosis-competent transmembrane receptor complexes, and the Arp2/3 complex of the actin cytoskeleton. LRP130 contains Epsin N-terminal homology (ENTH) and SEC1 domains; ENTH domains are most commonly associated with proteins involved in endocytosis and cytoskeletal organization, and SEC1 domains are usually involved in vesicular transport processes among cell compartments (29). LRP130 may be involved in integration of cytoskeletal actin and microtubule networks with vesicular trafficking, nucleocytosolic shuttling, and transcription during nuclear chromosome remodeling and possibly cytokinesis.


919


Signaling proteins that were found to differentially associate with endosomes in response to EGFR signaling are members of two signaling pathways: G-protein-coupled receptor protein signaling and R-Ras signaling. Three components of the heterotrimeric G-protein complex, G␣q, G␤1, and G␤2, were dissociated from endosomes upon treatment of cells with EGF. It was shown before that activation of the tyrosine kinase of EGFR induces G␤␥-dependent G-protein-coupled receptor kinase-EGFR complex formation (56). EGF can also stimulate adenylyl cyclase activity via activation of Gs. Gs␣ is phosphorylated by the EGFR protein tyrosine kinase, and the juxtamembrane region of the EGFR can stimulate Gs directly. G␣s interacts with hepatocyte growth factor-regulated tyrosine kinase substrate (Hrs), a critical component of the endosomal sorting machinery, and Hrs increases the localization of G␣s on endosomes (57). R-Ras was found in 2D DIGE analyses to be associated with late endosomes in an EGF-dependent manner (Table II and Fig. 5, A and B). This finding was confirmed by Western blotting where R-Ras was increased in crude endosomes upon EGF stimulation (Fig. 5C). Moreover an association of R-Ras with endosomes was inhibited by gefitinib as detected by 2D DIGE (supplemental table) as well as by Western blotting (Fig. 5C). Additionally we confirmed R-Ras association with the cytoplasmic side of late endosomal vesicles of circular shape upon EGF stimulation by confocal laser scanning (Fig. 6) and epifluorescence (Supplemental Fig. 5) microscopy. R-Ras belongs to Ras superfamily of small GTPases. The plasma membrane is considered to be the main platform for Ras-mediated signaling. It has been shown that R-Ras localizes to focal adhesions on the plasma membrane to enhance both cell adhesion and cell spreading by integrin activation (25, 38). Recently R-Ras was also shown on vesicular structures to co-localize with Arf6, suggesting that R-Ras undergoes internalization and recycling (25). EGF-dependent association of R-Ras with late endosomes suggests a possibility of R-Ras signaling from the late endosomal compartment. Therefore, our observation is in good agreement with the data reported by Furuhjelm and Peranen (25), who demonstrated that vesicles carrying the dominant negative (not activable) form of R-Ras (R-Ras43N) were not positive for such known organelle markers as Rab4, Rab5, Rab6, and Rab11, suggesting that R-Ras, which we could detect on the surface of late endosomal membrane, could be in its active and signaling-competent form on endosomes. It has also been shown that R-Ras antagonizes Ras/Raf-initiated integrin suppression (58), one of the well known effects of EGFR signaling. Therefore, the balance between R-Ras and H-Ras signaling may govern the decision between growth and differentiation. It is worth noting that R-Ras has already been detected as a component of highly purified rat liver lysosomes in a recent organelle proteomics analysis by Bagshaw et al. (59), thereby further supporting our detection of this protein on endosomes. The possible late endosomal R-Ras signaling and its place in

920


cell signaling networks remain to be investigated. Several proteins of general metabolism were detected as differentially associated with endosomes in an EGF-dependent way, e.g. enzymes of lipid metabolism and glycolysis (see Table III). Methylenetetrahydrofolate dehydrogenase 1 (MTHFD1), identified here, and 5,10-methylenetetrahydrofolate reductase play the central role in folate metabolism. 5,10Methylenetetrahydrofolate reductase was recently found to be down-regulated in neonatal porcine pancreatic cells in response to EGF treatment using 2DE (45). To investigate the possibility that some of the proteins that differentially associated with endosomes could be regulated by EGFR signaling at the level of TCL we performed the identical 2D DIGE experiment but without subcellular fractionation in TCL. Fifteen proteins were found to be differentially regulated after EGF treatment for 5 and 40 min. The rearrangement of the actin cytoskeleton was one of the major effects of EGF stimulation. Six previously known EGF-regulated proteins (MEK1, stathmin, cofilin, destrin, cytokeratin 8, and hnRNP-K) and seven novel targets were identified (integrin ␣-3, serine/threonine-protein kinase TIK, phospholipase C-␦-1, dynein intermediate chain 2, zyxin, plastin-2, and NADH dehydrogenase 1 ␣ subcomplex subunit 5). Phospholipase C-␦ was identified here as phosphorylated and up-regulated upon EGF treatment. It is well established that activated EGFR directly stimulates PLC-␥ by phosphorylating its tyrosine residues. However, it is still unclear whether PLC-␦ plays an important role as a signaling protein that is regulated by extracellular stimuli (60). The detection of MEK1, a core kinase in the MAPK cascade (Raf-MEK-ERK), confirmed the high sensitivity of the technique used here. Nevertheless only one protein, cytokeratin 8, of 23 EGFR targets on endosomes was detectable at the total cellular protein level. Therefore, organelle proteomics provides the necessary level of sensitivity for the detection of novel EGFR signaling targets. All together 23 proteins were identified here to be differentially associated with endosomal fractions in an EGF-dependent manner, among them several low abundance proteins known to be involved in signal transduction, endosomal trafficking, cytoskeleton rearrangement, and chaperone activity. EGF-dependent association of signaling molecules, such as R-Ras, with late endosomes suggests signaling specification through intracellular organelles. Acknowledgment—We thank AstraZeneca for supply of gefitinib. * Work in the Huber laboratory was supported by the Austrian Proteomics Platform within the Austrian Genome Program (GEN-AU), Vienna, Austria and the Special Research Program “Cell Proliferation and Cell Death in Tumors” (Grant SFB021, Austrian Science Fund). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. □ S The on-line version of this article (available at http://www. mcponline.org) contains supplemental material.


** To whom correspondence should be addressed: Biocenter, Div. of Cell Biology, Innsbruck Medical University, Fritz-Pregl Strasse 3, 6020 Innsbruck, Austria. Tel.: 43-512-9003-70170; Fax: 43-5129003-73100; E-mail: [email protected].

20.

REFERENCES

21.

1. Hoeller, D., Volarevic, S., and Dikic, I. (2005) Compartmentalization of growth factor receptor signalling. Curr. Opin. Cell Biol. 17, 107–111 2. Dard, N., and Peter, M. (2006) Scaffold proteins in MAP kinase signaling: more than simple passive activating platforms. BioEssays 28, 146 –156 3. Teis, D., and Huber, L. A. (2003) The odd couple: signal transduction and endocytosis. Cell. Mol. Life. Sci. 60, 2020 –2033 4. Muller, J., Ory, S., Copeland, T., Piwnica-Worms, H., and Morrison, D. K. (2001) C-TAK1 regulates Ras signaling by phosphorylating the MAPK scaffold, KSR1. Mol. Cell 8, 983–993 5. Kurzbauer, R., Teis, D., de Araujo M. E., Maurer-Stroh, S., Eisenhaber, F., Bourenkov, G. P., Bartunik, H. D., Hekman, M., Rapp, U. R., Huber, L. A., and Clausen, T. (2004) Crystal structure of the p14/MP1 scaffolding complex: how a twin couple attaches mitogen-activated protein kinase signaling to late endosomes. Proc. Natl. Acad. Sci. U. S. A. 101, 10984 –10989 6. Wunderlich, W., Fialka, I., Teis, D., Alpi, A., Pfeifer, A., Parton, R. G., Lottspeich, F., and Huber, L. A. (2001) A novel 14-kilodalton protein interacts with the mitogen-activated protein kinase scaffold mp1 on a late endosomal/lysosomal compartment. J. Cell Biol. 152, 765–776 7. Teis, D., Wunderlich, W., and Huber, L. A. (2002) Localization of the MP1MAPK scaffold complex to endosomes is mediated by p14 and required for signal transduction. Dev. Cell 3, 803– 814 8. Miaczynska, M., Christoforidis, S., Giner, A., Shevchenko, A., UttenweilerJoseph, S., Habermann, B., Wilm, M., Parton, R. G., and Zerial, M. (2004) APPL proteins link Rab5 to nuclear signal transduction via an endosomal compartment. Cell 116, 445– 456 9. Blagoev, B., Ong, S. E., Kratchmarova, I., and Mann, M. (2004) Temporal analysis of phosphotyrosine-dependent signaling networks by quantitative proteomics. Nat. Biotechnol. 22, 1139 –1145 10. Zhang, Y., Wolf-Yadlin, A., Ross, P. L., Pappin, D. J., Rush, J., Lauffenburger, D. A., and White, F. M. (2005) Time-resolved mass spectrometry of tyrosine phosphorylation sites in the epidermal growth factor receptor signaling network reveals dynamic modules. Mol. Cell. Proteomics 4, 1240 –1250 11. Wu, S. L., Kim, J., Bandle, R. W., Liotta, L., Petricoin, E., and Karger, B. L. (2006) Dynamic profiling of the post-translational modifications and interaction partners of epidermal growth factor receptor signaling after stimulation by epidermal growth factor using extended range proteomic analysis (ERPA). Mol. Cell. Proteomics 5, 1610 –1627 12. Kratchmarova, I., Blagoev, B., Haack-Sorensen, M., Kassem, M., and Mann, M. (2005) Mechanism of divergent growth factor effects in mesenchymal stem cell differentiation. Science 308, 1472–1477 13. Yoon, S., and Seger, R. (2006) The extracellular signal-regulated kinase: multiple substrates regulate diverse cellular functions. Growth Factors 24, 21– 44 14. Stasyk, T., Morandell, S., Bakry, R., Feuerstein, I., Huck, C. W., Stecher, G., Bonn, G. K., and Huber, L. A. (2005) Quantitative detection of phosphoproteins by combination of two-dimensional difference gel electrophoresis and phosphospecific fluorescent staining. Electrophoresis 26, 2850 –2854 15. Reichmann, E., Schwarz, H., Deiner, E. M., Leitner, I., Eilers, M., Berger, J., Busslinger, M., and Beug, H. (1992) Activation of an inducible c-FosER fusion protein causes loss of epithelial polarity and triggers epithelialfibroblastoid cell conversion. Cell 71, 1103–1116 16. Huber, L. A., Pfaller, K., and Vietor, I. (2003) Organelle proteomics: implications for subcellular fractionation in proteomics. Circ. Res. 92, 962–968 17. Fialka, I., Pasquali, C., Lottspeich, F., Ahorn, H., and Huber, L. A. (1997) Subcellular fractionation of polarized epithelial cells and identification of organelle-specific proteins by two-dimensional gel electrophoresis. Electrophoresis 18, 2582–2590 18. Araujo, M. E. G., and Huber, L. A. (2006) Subcellular fractionation, in Cardiovascular Proteomics: Methods and Protocols, pp. 73– 85, Humana Press Inc., Totowa, NJ 19. Fialka, I., Pasquali, C., Kurzbauer, R., Lottspeich, F., and Huber, L. A. (1999)

22.

23.

24. 25. 26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38. 39. 40.

41.

Loss of epithelial polarity is accompanied by differential association of proteins with intracellular membranes. Electrophoresis 20, 331–343 Wessel, D., and Flugge, U. I. (1984) A method for the quantitative recovery of protein in dilute solution in the presence of detergents and lipids. Anal. Biochem. 138, 141–143 Wilkins, M. R., Appel, R. D., Van Eyk, J. E., Chung, M. C., Gorg, A., Hecker, M., Huber, L. A., Langen, H., Link, A. J., Paik, Y. K., Patterson, S. D., Pennington, S. R., Rabilloud, T., Simpson, R. J., Weiss, W., and Dunn, M. J. (2006) Guidelines for the next 10 years of proteomics. Proteomics 6, 4 – 8 Rabilloud, T., Strub, J. M., Luche, S., van Dorsselaer, A., and Lunardi, J. (2001) A comparison between Sypro Ruby and ruthenium II tris (bathophenanthroline disulfonate) as fluorescent stains for protein detection in gels. Proteomics 1, 699 –704 Lamanda, A., Zahn, A., Roder, D., and Langen, H. (2004) Improved ruthenium II tris (bathophenantroline disulfonate) staining and destaining protocol for a better signal-to-background ratio and improved baseline resolution. Proteomics 4, 599 – 608 Hellman, U. (2000) Sample preparation by SDS/PAGE and in-gel digestion. EXS 43–54 Furuhjelm, J., and Peranen, J. (2003) The C-terminal end of R-Ras contains a focal adhesion targeting signal. J. Cell Sci. 116, 3729 –3738 Poupon, V., Stewart, A., Gray, S. R., Piper, R. C., and Luzio, J. P. (2003) The role of mVps18p in clustering, fusion, and intracellular localization of late endocytic organelles. Mol. Biol. Cell 14, 4015– 4027 Cabezas, A., Bache, K. G., Brech, A., and Stenmark, H. (2005) Alix regulates cortical actin and the spatial distribution of endosomes. J. Cell Sci. 118, 2625–2635 Harder, T., Kellner, R., Parton, R. G., and Gruenberg, J. (1997) Specific release of membrane-bound annexin II and cortical cytoskeletal elements by sequestration of membrane cholesterol. Mol. Biol. Cell 8, 533–545 Liu, L., and McKeehan, W. L. (2002) Sequence analysis of LRPPRC and its SEC1 domain interaction partners suggests roles in cytoskeletal organization, vesicular trafficking, nucleocytosolic shuttling, and chromosome activity. Genomics 79, 124 –136 Spiliotis, E. T., and Nelson, W. J. (2006) Here come the septins: novel polymers that coordinate intracellular functions and organization. J. Cell Sci. 119, 4 –10 Takayama, S., Xie, Z., and Reed, J. C. (1999) An evolutionarily conserved family of Hsp70/Hsc70 molecular chaperone regulators. J. Biol. Chem. 274, 781–786 Ku, N. O., and Omary, M. B. (1997) Phosphorylation of human keratin 8 in vivo at conserved head domain serine 23 and at epidermal growth factor-stimulated tail domain serine 431. J. Biol. Chem. 272, 7556 –7564 Blagoev, B., Kratchmarova, I., Ong, S. E., Nielsen, M., Foster, L. J, and Mann, M. (2003) A proteomics strategy to elucidate functional proteinprotein interactions applied to EGF signaling. Nat. Biotechnol. 21, 315–318 Klemke, R. L., Cai, S., Giannini, A. L., Gallagher, P. J., de Lanerolle, P., and Cheresh, D. A. (1997) Regulation of cell motility by mitogen-activated protein kinase. J. Cell Biol. 137, 481– 492 Marklund, U., Brattsand, G., Shingler, V., and Gullberg, M. (1993) Serine 25 of oncoprotein 18 is a major cytosolic target for the mitogen-activated protein kinase. J. Biol. Chem. 268, 15039 –15047 Habelhah, H., Shah, K., Huang, L., Ostareck-Lederer, A., Burlingame, A. L., Shokat, K. M., Hentze, M. W., and Ronai, Z. (2001) ERK phosphorylation drives cytoplasmic accumulation of hnRNP-K and inhibition of mRNA translation. Nat. Cell Biol. 3, 325–330 Loeffler-Ragg, J., Skvortsov, S., Sarg, B., Skvortsova, I., Witsch-Baumgartner, M., Mueller, D., Lindner, H., and Zwierzina, H. (2005) Gefitinibresponsive EGFR-positive colorectal cancers have different proteome profiles from non-responsive cell lines. Eur. J. Cancer 41, 2338 –2346 Zhang, Z., Vuori, K., Wang, H., Reed, J. C., and Ruoslahti, E. (1996) Integrin activation by R-ras. Cell 85, 61– 69 Self, A. J., Caron, E., Paterson, H. F., and Hall, A. (2001) Analysis of R-Ras signalling pathways. J. Cell Sci. 114, 1357–1366 Foster, L. J., de Hoog, C. L., Zhang, Y., Zhang, Y., Xie, X., Mootha, V. K., and Mann, M. (2006) A mammalian organelle map by protein correlation profiling. Cell 125, 187–199 Andersen, J. S., and Mann, M. (2006) Organellar proteomics: turning inven-


921


tories into insights. EMBO Rep. 7, 874 – 879 42. Andersen, J. S., Lam, Y. W., Leung, A. K., Ong, S. E., Lyon, C. E., Lamond, A. I., and Mann, M. (2005) Nucleolar proteome dynamics. Nature 433, 77– 83 43. Pasquali, C., Fialka, I., and Huber, L. A. (1997) Preparative two-dimensional gel electrophoresis of membrane proteins. Electrophoresis 18, 2573–2581 44. Huber, L. A. (2003) Is proteomics heading in the wrong direction. Nat. Rev. Mol. Cell. Biol. 4, 74 – 80 45. Hong, O. K., Suh, S. H., Kwon, H. S., Ko, S. H., Choi, Y. H., Moon, S. D., Yoo, S. J., Son, H. Y., Park, K. S., Lee, I. K., and Yoon, K. H. (2005) Proteomic analysis of differential protein expression in response to epidermal growth factor in neonatal porcine pancreatic cell monolayers. J. Cell. Biochem. 95, 769 –781 46. Weinman, E. J., Hall, R. A., Friedman, P. A., Liu-Chen, L. Y., and Shenolikar, S. (2006) The association of NHERF adaptor proteins with g proteincoupled receptors and receptor tyrosine kinases. Annu. Rev. Physiol. 68, 491–505 47. Lazar, C. S., Cresson, C. M., Lauffenburger, D. A., and Gill, G. N. (2004) The Na⫹/H⫹ exchanger regulatory factor stabilizes epidermal growth factor receptors at the cell surface. Mol. Biol. Cell 15, 5470 –5480 48. Cao, T. T., Deacon, H. W., Reczek, D., Bretscher, A., and von Zastrow, M. (1999) A kinase-regulated PDZ-domain interaction controls endocytic sorting of the ␤2-adrenergic receptor. Nature 401, 286 –290 49. Fialka, I., Steinlein, P., Ahorn, H., Bock, G., Burbelo, P. D., Haberfellner, M., Lottspeich, F., Paiha, K., Pasquali, C., and Huber, L. A. (1999) Identification of syntenin as a protein of the apical early endocytic compartment in Madin-Darby canine kidney cells. J. Biol. Chem. 274, 26233–26239 50. Ito, H., Iwamoto, I., Morishita, R., Nozawa, Y., Narumiya, S., Asano, T., and Nagata, K. I. (2005) Possible role of Rho/Rhotekin signaling in mammalian septin organization. Oncogene 24, 7064 –7072 51. She, Y. M., Huang, Y. W., Zhang, L., and Trimble, W. S. (2004) Septin 2 phosphorylation: theoretical and mass spectrometric evidence for the

922


52. 53.

54.

55.

56.

57.

58.

59.

60.

existence of a single phosphorylation site in vivo. Rapid Commun. Mass Spectrom. 18, 1123–1130 Bukau, B., and Horwich, A. L. (1998) The Hsp70 and Hsp60 chaperone machines. Cell 92, 351–366 Doong, H., Price, J., Kim, Y. S., Gasbarre, C., Probst, J., Liotta, L. A., Blanchette, J., Rizzo, K., and Kohn, E. (2000) CAIR-1/BAG-3 forms an EGF-regulated ternary complex with phospholipase C-␥ and Hsp70/ Hsc70. Oncogene 19, 4385– 4395 Nylandsted, J., Gyrd-Hansen, M., Danielewicz, A., Fehrenbacher, N., Lademann, U., Hoyer-Hansen, M., Weber, E., Multhoff, G., Rohde, M., and Jaattela, M. (2004) Heat shock protein 70 promotes cell survival by inhibiting lysosomal membrane permeabilization. J. Exp. Med. 200, 425– 435 Mili, S., Shu, H. J., Zhao, Y., and Pin˜ol-Roma, S. (2001) Distinct RNP complexes of shuttling hnRNP proteins with pre-mRNA and mRNA: candidate intermediates in formation and export of mRNA. Mol. Cell. Biol. 21, 7307–7319 Gao, J., Li, J., Chen, Y., and Ma, L. (2005) Activation of tyrosine kinase of EGFR induces G␤␥-dependent GRK-EGFR complex formation. FEBS Lett. 579, 122–126 Zheng, B., Lavoie, C., Tang, T. D., Ma, P., Meerloo, T., Beas, A., and Farquhar, M. G. (2004) Regulation of epidermal growth factor receptor degradation by heterotrimeric G␣s protein. Mol. Biol. Cell 15, 5538 –5550 Sethi, T., Ginsberg, M. H., Downward, J., and Hughes, P. E. (1999) The small GTP-binding protein R-Ras can influence integrin activation by antagonizing a Ras/Raf-initiated integrin suppression pathway. Mol. Biol. Cell 10, 1799 –1809 Bagshaw, R. D., Mahuran, D. J., and Callahan, J. W. (2005) A proteomic analysis of lysosomal integral membrane proteins reveals the diverse composition of the organelle. Mol. Cell. Proteomics 4, 133–143 Harden, T. K., and Sondek, J. (2006) Regulation of phospholipase C isozymes by ras superfamily GTPases. Annu. Rev. Pharmacol. Toxicol. 46, 335–379