| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Tumor Biology |
Cancer Research Program, Garvan Institute of Medical Research, St. Vincent’s Hospital, Sydney, New South Wales 2010, Australia
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The EMS1 gene localizes to chromosome locus 11q13, a region that also contains the CCND1 gene encoding the cell cycle regulatory protein cyclin D1 and that is commonly amplified in a number of human cancers, including those of the breast (12 , 13) and squamous cell carcinomas of the head and neck (14) . 11q13-amplified breast cancers may exhibit an increased invasive potential (15) . In a recent study of approximately 1000 breast cancer patients, EMS1 amplification occurred in 15% of the patients tested and was associated with earlier relapse in the lymph node-negative and the ER-negative, but not ER-positive, subgroups (16) . This may reflect a role for this protein downstream of the EGFR and erbB2, which are commonly overexpressed in ER-negative breast cancers (17 , 18) .
The tyrosine phosphorylation of cortactin is increased by diverse stimuli including receptor tyrosine kinase activation (19 , 20) , integrin-mediated cell adhesion (21) , oncogenic transformation (1 , 22) , and platelet aggregation (23 , 24) and is strongly linked to activation of Src-family kinases. Many of these processes result in extensive cytoskeletal rearrangement, suggesting that tyrosine-phosphorylated cortactin is involved in regulation of the actin cytoskeleton. This is supported by studies using v-src-transformed fibroblasts and thrombin-activated platelets, in which tyrosine-phosphorylated cortactin partitions into the detergent-insoluble cytoskeletal fraction (4 , 5 , 24) . Furthermore, upon v-src transformation, cortactin localizes into abnormal focal adhesions termed rosettes or podosomes (1) . In addition to cytoskeletal reorganization induced by stimuli resulting in EMS1 tyrosine phosphorylation, the subcellular localization of EMS1 is also controlled by the small GTPases Ras and Rac1. In untransformed NIH 3T3 cells, EMS1 is found complexed with myosin II and actin, and this complex is disrupted after Ras transformation (25) . After growth factor-induced Rac1 activation in Swiss 3T3 cells, cortactin localizes to membrane ruffles, a process that can also be induced by activated Rac1 (26) . These findings suggest that the subcellular localization of cortactin may be regulated by mechanisms that are independent of cortactin tyrosine phosphorylation.
Recently, it has become clear that the EMS1 protein is involved in the regulation of cell motility and invasion. Overexpression of EMS1 in NIH 3T3 cells resulted in increased cell motility and invasiveness without changing the growth rate or anchorage dependence of the cells (27) . Similarly, overexpression of cortactin in endothelial cells resulted in enhanced cell migration in a monolayer wound-healing assay (28) . Furthermore, the role of c-Src-mediated cortactin tyrosine phosphorylation in regulation of cell motility has been defined. Mutation of tyrosines 421, 466, and 482 to phenylalanine abolished the ability of cortactin to cross-link actin filaments in vitro, and transfection studies indicated that these tyrosine residues are also required for the elevated motility seen in cortactin-transfected endothelial cells (28) . However, both mutant and wild-type cortactin localized to the same subcellular regions, indicating that tyrosine phosphorylation did not control subcellular localization.
In addition to tyrosine phosphorylation, serine and threonine phosphorylation may regulate cortactin function. EMS1 is phosphorylated on serine and threonine residues in untransformed chick embryo fibroblast cells (1) and in two human squamous carcinoma cell lines (29) . Two cortactin species of 80 and 85 kDa are present in many cell types (1 , 2 , 11 , 29) . Recent evidence suggests that the 85-kDa form of EMS1 results from serine and threonine phosphorylation of the 80-kDa form (29) . In cells amplified at the 11q13 locus, EGF, serum, or vanadate treatment caused increased serine and threonine phosphorylation of EMS1, resulting in a mobility shift from 80 to 85 kDa. EGF- or vanadate-induced serine and threonine phosphorylation of EMS1 also resulted in a change in localization from the cytoplasm to the cell-substratum junctions.
In this report, we have analyzed the signaling pathways triggered by growth factor treatment or cell adhesion events that regulate EMS1 serine/threonine phosphorylation and also defined the protein domains required. These studies demonstrate that EMS1 phosphorylation is regulated by both MEK-dependent and -independent mechanisms and that the HP region is essential for these events. The latter domain is therefore likely to play a major role in regulation of EMS1 function.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
SH3 mutant, the sense oligonucleotide was 5' -AACGATCTGGGGTAGACAGCCGTCGCC-3', and the antisense oligonucleotide was 5'-GGCGACGGCTGTCTACCCCAGATCGTT-3', generating a construct encoding amino acids 1–494. For the HPSH3 mutant, the sense oligonucleotide was 5' -AGCAAAACAAGTAACATCTGAGCTAACTTTGAAAACCTCGT-3', and the antisense oligonucleotide was 5' -AGCGAGGTTAACAAAGTTAGGTCAGATGTTACTTGTTTTGCT-3', generating a construct encoding amino acids 1–350. Plasmid vectors encoding these proteins were sequenced to ensure the presence of the mutation. Plasmids encoding GST-EMS1 fusion proteins were constructed using the following primers. For the GST-HP protein, the N-terminal sense primer was 5' -GGATCCGCTAACTTTGAAAACCTCG-3', and the COOH-terminal primer was 5'-GACTGAATTCTCACTCGTACTCATGTAGGTG-3', incorporating residues 352–490. For the GST-SH3 fusion protein, the N-terminal primer was 5'-ACATGGATCCACCTACGATGAGTACGAGAAC-3', and the COOH-terminal primer was 5'-GACTGAATTCTCACTGCCGCAGCTCCACTAGTTGGC-3', incorporating residues 485–550. Following PCR using the EMS1 cDNA as template, the resulting DNA fragments were digested with EcoRI and BamHI and ligated into a EcoRI/BamHI digested pGEX 2T vector (Amersham Pharmacia Biotech).
Cell Culture and Generation of Transfectants.
HEK 293 cells were maintained in MEM with Hanks’ salts (Trace Biosciences) supplemented with 10% (v/v) FCS (Commonwealth Serum Laboratories), 2 mM L-glutamine (Life Technologies, Inc.), and 0.05% (w/v) sodium bicarbonate (Life Technologies, Inc.). This is referred to as HEK medium. Cells were transfected as described (32)
with 20 µg of plasmid DNA (either pRcCMV or pRcCMV-EMS1), and selected using 500 µg/ml G418. Single colonies were picked using a pipette, transferred to individual wells of 96-well plates, and then briefly trypsinized the following day to disperse the cells. Following expansion, stably transfected cell lines were analyzed by Western blotting. Transient transfections were performed essentially as described (32)
, using a total of 20 µg of plasmid DNA.
Growth Factor Treatment of Cells.
Cells were starved overnight in serum-free medium and then stimulated for the indicated time period with 100 ng/ml EGF (Life Technologies, Inc.) in serum-free medium. If no time period is indicated, the cells were stimulated for 30 min. The MEK inhibitor PD98059 (New England Biolabs) was dissolved in DMSO at a concentration of 50 mM. Unless indicated otherwise, cells were treated for 40 min with 50 µM PD98059 prior to stimulation.
Antibodies.
EMS1 was immunoprecipitated using 5 µg of a monoclonal antibody against chicken p80/85 (cortactin) exhibiting cross-species reactivity (mAb 4F11, UBI; Ref. 1
), and was detected on Western blots with the same antibody diluted 1:2000. Active Erks were detected using a polyclonal antibody specific to the activated form (New England Biolabs). All antibody solutions for Western blotting were made in 5% (w/v) BSA (RIA grade, Sigma)/TBS (10 mM Tris-Cl, 150 mM NaCl, pH 7.4).
Immunoprecipitation and Western Blotting.
Cell lysates were prepared using either lysis buffer as described previously (33)
, or RIPA buffer (34)
, each containing 10 µg/ml aprotinin, 10 µg/ml leupeptin, 1 mM sodium orthovanadate, and 1 mM phenylmethylsulfonyl fluoride. Protein concentrations were determined using the Bio-Rad protein assay reagent, and samples were normalized by dilution with the appropriate lysis buffer. Immunoprecipitations were performed by incubation with the indicated amount of antibody for a minimum of 2 h at 4°C with constant mixing. 20 µl of goat antimouse antibody-Sepharose conjugate (Zymed; prewashed twice in 20 mM HEPES, pH 7.5) was then added to the lysate-antibody mixture for 1 h. Immunoprecipitates were then washed five times in the lysis buffer and denatured in SDS-PAGE sample buffer by heating to 100°C for 3 min. Samples were analyzed by SDS-PAGE (10% minigels), transferred to 0.2-µm nitrocellulose (Bio-Rad), and subjected to Western blot analysis. Equivalent loading of cell lysates was confirmed by Ponceau S staining of membranes after transfer. Detection of bound antibodies was by ECL (Amersham Pharmacia Biotech).
Pi Labeling of Cells.
Cells were plated at 5 x 106 per 10-cm plate and allowed to grow overnight in HEK medium containing 10% FCS. The cells were then serum-starved overnight in serum-free HEK medium. Cells were then rinsed in phosphate-free MEM (ICN) and incubated in 5 ml of phosphate-free MEM for 2 h. This medium was subsequently replaced with 5 ml of phosphate-free MEM supplemented with 0.6 mCi/ml Pi (ICN, 500 mCi/ml) and 50 µM PD98059, as appropriate, for 3 h. Cells were then stimulated with EGF for 30 min prior to cell lysis in RIPA buffer.
Cell Detachment Experiments.
Cells were incubated in serum-free medium overnight and then either incubated with serum-free medium containing 50 µM PD98059 for 2 h or left untreated. Following detachment with PBS-0.02% (w/v) EDTA, cells were maintained in suspension either in the presence or absence of PD98059. Cells were then harvested and lysed at the indicated times.
Phosphoamino Acid Analysis and Tryptic Phosphopeptide Mapping.
EMS1 was immunoprecipitated from lysates prepared from Pi-labeled cells and then subjected to SDS-PAGE. The gels were then either Coomassie-stained, dried and subjected to autoradiography prior to tryptic phosphopeptide mapping, or transferred to PVDF membranes for subsequent phosphoamino acid analysis. For the latter, following PVDF transfer, the samples were hydrolyzed in 6 M HCl and then subjected to two-dimensional thin-layer electrophoresis with phosphoamino acid standards as described previously (35)
. For tryptic phosphopeptide mapping, samples were excised from the dried gel, rehydrated in water, and washed in 25% (v/v) isopropanol followed by 10% (v/v) methanol. The gel slices were then cut into small pieces and suspended in 50 mM ammonium bicarbonate solution containing 60 µg of N-tosyl-L-phenylalanine chloromethyl ketone-treated trypsin (Sigma). The samples were incubated at 37°C overnight, and the supernatant was then removed and lyophilized. The pellet was then resuspended in water, lyophilized again, and then solubilized in electrophoresis buffer (formic acid:acetic acid:water, 15:5:80) and applied to a silica gel 60 TLC plate (Merck). Samples were electrophoresed in the first dimension at 1000 V for 45 min, and the plates were dried and then subjected to ascending chromatography in the second dimension (chromatography buffer; butanol:pyridine:acetic acid:formic acid:water, 127.5:22.5:45:15:90) for 5 h. The plates were dried then subjected to autoradiography. For phosphoamino acid analysis of individual tryptic phosphopeptides, the peptide spots were isolated from the TLC plate and extracted from the cellulose fragments using pH 1.9 buffer (formic acid:acetic acid:water, 50:156:1794). Peptide solutions were then lyophilized and subjected to phosphoamino acid analysis.
In Vitro Protein Kinase Assays.
GST fusion proteins were prepared from bacteria transformed with the appropriate expression constructs by standard methods (36)
. Erks were kindly provided by Dr Boris Sarcevic. These were prepared from EGF-treated T-47D cells by purification over DEAE cellulose ion-exchange and Superdex 75 gel filtration columns (Amersham Pharmacia Biotech), followed by hydrophobic chromatography over a phenyl-Superose HR15/5 column (Amersham Pharmacia Biotech). Activity and identity were monitored by phosphorylation of MBP and Western blot analysis, and purity was determined by SDS-PAGE and silver staining. In vitro kinase reactions were performed in 30 µl of kinase assay buffer (50 mM HEPES, pH 7.5, 1 mM DTT, 10 mM MgCl2, 25 µM ATP) containing 5 µCi of [
-32P]ATP (NEN, 10 mCi/ml) per reaction for 30 min at 30°C. Reactions were terminated using 15 µl of 3x SDS-PAGE sample buffer. Equimolar amounts of fusion protein, GST, or MBP were used for the kinase assays. Thus 1.3 µg of MBP, 2.2 µg of GST, 3.3 µg of GST-SH3, and 5 µg of GST-HP were used per reaction.
Tryptic Phosphopeptide Mixing Experiments.
GST-HP fusion protein was phosphorylated by Erks as described above. EMS1 was immunoprecipitated after in vivo Pi labeling and EGF stimulation as described previously. Tryptic phosphopeptides of both these samples were derived, and following Cerenkov counting, the in vivo labeled sample was diluted in electrophoresis buffer to give equivalent cpm per unit volume to that obtained for the GST-HP sample. The proportion of in vivo and in vitro samples mixed were such that 3-fold more radioactivity was derived from the in vitro reaction, enabling identification of co-migrating spots by comparison with the map obtained from the in vivo sample analyzed alone.
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
shows a Western blot characterizing EMS1 expression in two empty-vector (RcCMV1 and RcCMV2) and four EMS1-transfected (EMS1–2, -6, -11, and -12) cell lines. To investigate the effect of EMS1 overexpression on the EGF-induced modification of this protein, RcCMV2 and EMS1–6 cell lines were starved overnight in serum-free medium and then stimulated with EGF for varying times. EMS1 was immunoprecipitated and Western blotted for EMS1 (Fig. 1B)
. Both cell lines expressed the p80 and p85 forms of EMS1 after overnight starvation at an abundance similar to that observed in cells maintained in 10% serum (Fig. 1A)
, indicating that in these cells the p80:p85 ratio was not affected by the absence of serum factors. For both EMS1 overexpressers and cells expressing endogenous levels of EMS1, the p80 form started to disappear 5 min subsequent to EGF treatment (presumably by conversion to p85), and after 15 min, the majority of the p80 was converted to p85. The p85 form was maintained for up to 3 h after EGF treatment. We did not detect any increase in EMS1 tyrosine phosphorylation under these conditions (data not shown and Fig. 5B
). The time course of conversion of the p80 to p85 form is very similar to that described for 11q13 amplified squamous carcinoma cells (29)
. However, these results demonstrate that in another cell type, 11q13 amplification or EMS1 overexpression is not required for the EGF-induced mobility shift, and this occurs to a similar extent and with comparable kinetics in cells with low or high EMS1 expression.
|
|
. Both of these cell lines displayed undetectable levels of the activated form of Erks following serum starvation and exhibited a robust activation of Erks after EGF treatment that was evident at 2 min (i.e., prior to the onset of the EMS1 mobility shift) and maintained for up to 3 h.
The EGF-induced Mobility Shift Can Be Inhibited by PD98059.
To investigate the possible contribution of the Erks to the EGF-induced mobility shift, we used the inhibitor PD98059, which prevents activation of MEK by Raf and hence blocks signaling through the Erks, without affecting other signaling pathways activated by the EGFR and Ras (37
, 38)
. EMS1–6 and RcCMV2 control cells were used for this experiment with equivalent results. Cells were starved overnight and then either pretreated with the MEK inhibitor or with vehicle (DMSO). The cells were then stimulated with EGF and lysates blotted for EMS1 or the activated form of Erks (Fig. 2A)
. Pretreatment with the MEK-inhibitor blocked the EGF-induced mobility shift of EMS1 and completely abolished EGF-induced Erk activation, whereas the vehicle was without effect. In squamous carcinoma cells overexpressing EMS1, the p80 form is rapidly converted to p85 as a result of posttranslational modification, whereas squamous carcinoma cells that do not overexpress EMS1 show very little conversion of the p80 to the p85 form of the protein (29)
. To determine whether long-term treatment with the MEK-inhibitor would prevent the appearance of the 85-kDa form of EMS1, EMS1–2 cells were serum-starved overnight and then treated for 2, 4, 8, or 24 h with the MEK inhibitor. Lysates were then prepared and Western blotted for EMS1 (Fig. 2B)
. The p80:p85 ratio did not change after prolonged treatment with the MEK inhibitor, suggesting that there is also a MEK-independent pathway that gives rise to the p85 form of the protein in these cells under basal conditions.
|
). The transfected cells were starved overnight and then harvested, and the lysates were Western blotted for EMS1 or activated Erks (Fig. 3)
. Transfection of the EMS1 vector alone resulted in the appearance of both p80 and p85 forms (Fig. 3
, cont), whereas co-transfection of the activated MEK and EMS1 constructs resulted in only the p85 form of the protein (MEK EE). Co-transfection of the EMS1 and dominant negative MEK constructs (MEK A) resulted in the appearance of both p80 and p85 forms of EMS1 at a ratio similar to that observed in the control transfection (cont). EGF treatment of cells transiently transfected with the EMS1 construct gave rise to the p85 form of the protein and robust Erk activation (Fig. 3
, +EGF). Transfection with the activated MEK construct also caused Erk activation in the transfected cells, whereas transfection with either the EMS1 vector alone or co-transfection with the dominant negative MEK construct did not result in any detectable activation of Erks. Although the levels of Erk activation detected after transfection with activated MEK were considerably lower than those obtained after EGF treatment, this was due to the small percentage of cells transfected. Thus, activated MEK is both required and sufficient to cause the EMS1 mobility shift, whereas expression of a dominant negative MEK is insufficient to prevent basal phosphorylation of EMS1 occurring in starved cells. Both these results are consistent with the results obtained using the MEK inhibitor PD98059.
|
|
. Erks were activated immediately following detachment, and this was maintained for 60 min. Erk activation after cell detachment has also been observed in NIH 3T3 cells (39)
. PD98059 pretreatment prevented the Erk activation caused by cell detachment. The EMS1 mobility shift was apparent from approximately 15 min, and was largely complete by 30 min. Treatment with PD98059 prevented the EMS1 mobility shift after 60 min of suspension. Cell detachment therefore induces an EMS1 mobility shift that is slightly temporally delayed compared to that induced by EGF (it is complete after 30 min rather than 15 min), but it also causes a much lower activation of the Erks (data not shown). Although plating on fibronectin substantially increased Erk activation relative to that observed in detached cells, we were not able to investigate the effects of this on the EMS1 mobility shift, because the protein was already in the p85 form.
|
, top panel), Western blotted for EMS1 (Fig. 5A
, bottom panel) or subjected to phosphoamino acid analysis (Fig. 5B)
. In all cases, phosphorylation was only detected for the p85 form of EMS1, and no phosphorylation of the p80 form was observed. Basal phosphorylation of EMS1 was observed in the unstimulated samples, and this was increased approximately 2-fold after EGF treatment, co-incident with the change in p80:p85 ratio (Fig. 5A
, lower panel). The EGF-induced increase in EMS1 phosphorylation was blocked by treatment with the MEK inhibitor, which blocked the p80 to p85 conversion. The 2-fold increase in phosphorylation is consistent with p80 being converted into p85. The labeling of EMS1 in the starved cells indicates that there is a MEK-independent kinase activity that phosphorylates EMS1, as MEK activation (as judged by active Erk blotting) was extremely low in unstimulated cells (Fig. 1C)
. Consistent with this observation, treatment of unstimulated cells with PD98059 did not prevent phosphate incorporation (Fig. 5A)
. EMS1 was phosphorylated on serine and threonine residues in starved and EGF-treated EMS1–2 cells in approximately equivalent amounts (Fig. 5B)
. Treatment with PD98059 or EGF did not appreciably alter the ratio of serine to threonine phosphorylation, nor did it result in detectable tyrosine phosphorylation. Thus EGF treatment causes serine and threonine phosphorylation of the p80 form of EMS1 in a MEK-dependent manner, correlating with a conversion to the p85 form.
EGF Treatment Increases the Phosphorylation of the Same Tryptic Peptides That Are Phosphorylated Basally under Serum-starved Conditions.
To investigate whether basal and EGF-induced phosphorylation occurs on the same or different sites, the EMS1–2 Pi-labeled samples shown in Fig. 5A
were subjected to tryptic phosphopeptide mapping (Fig. 6)
. EMS1 isolated from untreated cells displayed 4 prominent tryptic phosphopeptides (labeled A–D in order of intensity; Fig. 6A
). EMS1 isolated from EGF-treated cells showed exactly the same peptide map but with increased intensity of all of the peptides, and under these conditions we were unable to identify any novel peptides that were specifically induced after EGF treatment (Fig. 6B)
. Samples isolated from cells treated with either the MEK inhibitor alone (Fig. 6C)
or the MEK inhibitor and EGF (Fig. 6D)
again showed an identical peptide map, with the intensity of all of the peptides similar to that observed for the control sample. Tryptic peptide maps obtained from the HEK 293 cell line showed a pattern identical to those obtained from EMS1–2 cells (data not shown), indicating that the sites of phosphorylation are not markedly altered by overexpression of the EMS1 protein. Phosphoamino acid analysis of the peptides shown in Fig. 6B
revealed that peptide A was phosphorylated on both serine and threonine residues, peptide B was phosphorylated exclusively on serine, peptide C was phosphorylated on threonine, and peptide D was phosphorylated on serine (data not shown). Thus, the EMS1 kinase activity in starved and EGF-treated cells is directed toward the same tryptic peptides, and the activity toward these peptides is induced by EGF in a MEK-dependent manner.
|
SH3), whereas insertion of a stop codon between the actin-binding region and the HP region deleted the HP and SH3 domains (HPSH3; Fig. 7A
). These constructs were transiently transfected into HEK 293 cells and lysates Western blotted for EMS1 using the 4F11 monoclonal antibody, which recognizes an epitope in the actin-binding region (3)
. Deletion of the SH3 domain gave rise to two forms of the EMS1 protein, suggesting that the SH3 domain was not required for the basal phosphorylation, whereas deletion of the HP and SH3 domains gave a single band (Fig. 7B)
. Thus, the HP region was required for the generation of two forms of EMS1 in exponentially growing cells, whereas the SH3 domain was not. To determine whether the HP or SH3 domain was required for the EGF-induced mobility shift, HEK 293 cells were transiently transfected with EMS1 constructs, starved overnight, and then stimulated with EGF for 30 min. The full-length and SH3 constructs displayed EGF-induced shifts, whereas the HPSH3 construct displayed no mobility shift upon EGF treatment (Fig. 7C)
. Consistent with these results, 32P Pi labeling of transient transfections indicated that removal of the SH3 domain did not affect basal or EGF-induced phosphorylation, whereas deletion of the HP and SH3 domains almost completely abolished basal and EGF-induced phosphorylation (Fig. 7D)
. To determine whether any of the phosphorylation sites were present in the SH3 domain and hence were not necessary for the mobility shift, tryptic phosphopeptide maps were prepared from cells transiently transfected with the SH3 construct, labeled with 32P Pi, and then stimulated with EGF. Tryptic phosphopeptides A, B, C, and D were all observed, indicating that the phosphorylation sites are all present within the HP or actin binding domains (data not shown). Because the HPSH3 construct was only very weakly phosphorylated under these conditions, this suggests that either the phosphorylation sites are not located in the actin binding region or that this construct cannot interact with EMS1 kinases.
The HP Region Can Be Phosphorylated in Vitro by the Erks.
To determine candidate EMS1 kinases that contribute to the EGF-induced mobility shift, a number of criteria were considered. First, the kinase should be activated after EGF treatment; second, this EGF-induced activation should be MEK dependent; and third, consensus sites for the kinase should be present in the actin-binding or HP regions of EMS1. The Erks represent the only known MEK targets and are rapidly activated after EGF treatment in HEK 293 cells (Fig. 1C)
. The minimal consensus sequence for Erk phosphorylation is Ser/Thr-Pro, and the optimal sequence is Pro-X-Ser/Thr-Pro (40)
. EMS1 contains two optimal (Ser-405, PVSP; Ser-418, PSSP) and one minimal (Thr-401, TP) Erk consensus sites, which are all located in the HP region, consistent with the deletion analysis. These consensus phosphorylation sites are located on two predicted tryptic phosphopeptides: Thr-399 to Arg-414 (peptide 1) and Leu-415 to Lys-428 (peptide 2; Fig. 8A
). DNA fragments encoding the HP and SH3 regions of EMS1 were cloned into the pGEX 2T vector and GST-HP and GST-SH3 fusion proteins purified from Escherichia coli. These fusion proteins were used as substrates for an in vitro kinase assay using purified Erk1 and Erk2 (Fig. 8B)
. The GST-HP fusion protein was phosphorylated by Erks, as was the MBP-positive control. Neither the GST nor the GST-SH3 protein was phosphorylated in this assay, demonstrating that the phosphorylation by Erks was specific for the HP region. In contrast to the results obtained for in vivo labeling of EMS1 (Fig. 5B)
, Erks phosphorylated the HP region only on serine residues, eliminating Thr-401 as an in vitro Erk target (Fig. 8C)
.
|
. Two tryptic phosphopeptides of equivalent intensity were obtained after autoradiography, suggesting that both consensus sites corresponding to Ser-405 and Ser-418 were phosphorylated, as these residues are present on different predicted tryptic phosphopeptides. To test whether the tryptic phosphopeptides generated by in vitro phosphorylation of GST-HP fusion protein co-migrated with tryptic phosphopeptides obtained from in vivo labeled EMS1, a mixing experiment was performed. Tryptic phosphopeptide samples obtained from in vivo labeled EMS1 were diluted to give an equivalent signal to that obtained for the in vitro samples. Two-thirds of the in vivo aliquot was subjected to tryptic phosphopeptide mapping (Fig. 9A)
, and the other one-third was mixed with tryptic phosphopeptides generated from in vitro labeled GST-HP fusion protein prior to analysis (Fig. 9B)
. Only peptide spots A, B, and C are clearly visible on the exposure shown in Fig. 9A
. Consistent with the results in Fig. 6B
, the signal from peptide B was considerably less than that of peptide A. After mixing in vivo and in vitro samples, four peptide spots were visible. Both spots B and D were increased in intensity relative to those obtained from the in vivo samples alone, indicating co-migration of peptides phosphorylated in vitro and in vivo. This result confirms that EMS1 is phosphorylated within the HP region and therefore complements the deletion analysis presented in Fig. 7
. Furthermore, co-migration of two HP-derived phosphopeptides phosphorylated by Erks in vitro with two phosphopeptides phosphorylated in a MEK-dependent manner upon EGF treatment in vivo identifies Erks as candidate EMS1 kinases. Finally, assuming that in vitro phosphorylation by Erks occurs on the two consensus sites at Ser-405 and Ser-418, these are also likely to represent in vivo phosphorylation sites.
|
|
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Several lines of evidence suggest that serine/threonine phosphorylation of EMS1 may play an independent role in regulation of EMS1 function to tyrosine phosphorylation. Firstly, a wide variety of cell types display both the 80- and 85-kDa forms of the EMS1 protein, and the presence of at least two forms of the protein is conserved between human, mouse, chicken, and Drosophila (1
, 11
, 29)
. In addition, serine and threonine phosphorylation can occur independently of tyrosine phosphorylation, and only serine and threonine phosphorylation correlates with the EMS1 mobility shift (29)
. Similarly, in our studies, the mobility shift was independent of tyrosine phosphorylation (Fig. 5B)
. One likely role for serine/threonine phosphorylation is in regulation of EMS1 subcellular localization, as vanadate-induced tyrosine phosphorylation of EMS1 in squamous carcinoma cells expressing normal levels of EMS1 protein was insufficient to localize the protein to the cell-substratum contact region; however, EGF-induced serine and threonine phosphorylation correlated with a change in subcellular localization for cells overexpressing the protein (29)
. The specific role for serine/threonine versus tyrosine phosphorylation is underscored by the observation that mutation of the tyrosine residues necessary for c-Src-mediated tyrosine phosphorylation of EMS1 did not cause a change in subcellular localization of the protein when expressed in endothelial cells (28)
.
In cells that overexpressed the EMS1 protein and those that expressed endogenous levels, EGF treatment resulted in a shift of p80 to p85 and a corresponding 2-fold increase in serine/threonine phosphorylation (Figs. 1B
and 5)
. We also observed an EGF-induced EMS1 mobility shift in SK-BR-3 human breast cancer cells, another cell line that is not amplified for the 11q13 locus and expresses levels of EMS1 protein equivalent to normal breast epithelial cells (data not shown; Refs. 43
and 46
). Thus, neither 11q13 amplification nor EMS1 overexpression is required for the EGF-induced mobility shift, and the absence of CCND1 amplification in these cells argues against deregulation of cyclin D1/CDK4 as a possible mechanism for the increased EMS1 serine/threonine phosphorylation, as was suggested previously (29)
.
Inhibition of EGF-induced MEK activation using PD98059 was sufficient to prevent the mobility shift and the increase in serine/threonine phosphorylation (Figs. 2
and 5)
. Interestingly, a different cellular stimulus, cell detachment, also induced a MEK-dependent shift of EMS1 from p80 to p85. Furthermore, co-transfection of EMS1 and a constitutively active MEK construct led to detection of only the p85 form of EMS1, indicating that MEK activation not only is required for the p80 to p85 conversion but also is sufficient to induce it (Fig. 3)
. However, in unstimulated cells, the p85 form of EMS1 was present and incorporated radioactive phosphate, and this was unaffected by treatment with PD98059 or expression of a dominant negative MEK construct, indicating that this basal phosphorylation was independent of MEK activation (Figs. 2
, 3
, and 5)
. This was supported by the observation that the levels of activation of this pathway in starved cells, as determined by Western blotting for active Erks, was undetectable (Fig. 1C)
. This basal activity toward p80 could represent baseline activation of an enzyme susceptible to up-regulation by MEK and/or constitutive activation of a MEK-independent signaling pathway. In squamous carcinoma cells with an 11q13 amplification, inhibition of protein synthesis via cycloheximide treatment results in loss of p80 due to conversion into p85 (29)
. However, incubation of HEK 293 or EMS1–2 cells with cycloheximide for up to 6 h did not alter the p80/85 ratio. Consequently the stability of the p80 and p85 forms of EMS1 can vary according to cell type, presumably due to differences in the total amount of EMS1 protein and opposing kinase and phosphatase activities.
To gain further insight into the potential regulatory role of EMS1 serine/threonine phosphorylation and the specific kinases involved, we mapped the EMS1 region required. EMS1 consists of three functional domains: the actin binding region, the HP region, and the COOH-terminal SH3 domain. Removal of the SH3 domain did not affect either the basal or EGF-induced phosphorylation of EMS1, whereas removal of the HP region abolished both, as well as the EGF-induced mobility shift (Fig. 7)
. Furthermore, because two of the tryptic phosphopeptides derived from in vivo phosphorylated EMS1 co-migrated with phosphopeptides derived from in vitro phosphorylation of the HP region by the Erks (Fig. 9)
, the latter domain must be phosphorylated in vivo. However, it remains possible that this region also serves to recruit a specific kinase that phosphorylates EMS1 within the actin binding region, because the origin of the other two phosphopeptides is currently unknown. Importantly, the observation that the two peptides phosphorylated in vitro on serine residues by the Erks co-migrated with equivalent peptides on the in vivo map, coupled with the MEK-dependency of the phosphorylation of these peptides, identified the Erks as candidate EMS1 kinases. EMS1 therefore joins the cytoplasmic subset of Erk targets that also includes cPLA2 (47)
, the Ras GDP-GTP exchange factor Sos-1 (48)
, and the EGFR (49)
. It is interesting to note that a pool of Erk2 translocates to membrane ruffles, a site of EMS1 subcellular localization, upon serum stimulation of COS cells (50)
.
However, activation of MEK results in not only activation of the Erks, but also a number of other kinases that are in turn phosphorylated and/or activated by these enzymes. These include p90 RSK1–3 (51
, 52)
, MAP kinase-activated protein kinases 3 (53
, 54)
and 5 (55)
, the recently identified Mnk1 and Mnk2 (56
, 57)
, and MSK1 (58)
. The p90 RSK kinases are serine-directed kinases that display basal activity that is enhanced after activation by Erks but is not significantly altered by PD98059 treatment or expression of a dominant negative MEK (52)
, and as such they are attractive candidate kinases for the basal and EGF-induced phosphorylation of EMS1. Futhermore, there are two minimal (Arg-XX-Ser) RSK consensus phosphorylation sites (Ser-417 and Ser-457) located in the HP region of EMS1. It is noteworthy that RSK 2 and 3 interact with Erk1 and Erk2, and both RSK and Erks have been implicated in phosphorylation of Sos-1 (48
, 59)
. This is of interest, as it must involve translocation to a site near the cell membrane, hence placing the complex in close proximity to cortical EMS1. However, the selectivity of RSK kinases toward serine residues (60)
indicates that a further MEK-dependent kinase must also be involved, as significant MEK-dependent phosphorylation on threonine residues was observed (Fig. 5)
. The contribution of these other MEK-regulated enzymes to EMS1 phosphorylation is currently under investigation.
The finding that EMS1 phosphorylation is regulated by the MEK-Erk pathway may provide an underlying mechanism for our recent finding that EMS1 gene amplification is associated with poor prognosis in ER-negative, but not ER-positive, breast cancer (16 , 61) . Because EGFR and erbB2 are highly expressed in ER-negative breast cancer (17 , 18) , and because both receptors strongly activate this signaling pathway (33 , 62) , Erk mediated phosphorylation may alter the function of the overexpressed EMS1 protein. Because EMS1 also represents a target for Src family kinases (1 , 5 , 6 , 28 , 43 , 63) , and because these are also activated by these receptors (64, 65, 66, 67) , enhanced tyrosine phosphorylation of EMS1 may also contribute to this effect.
The HP region contains the tyrosine residues involved in regulation of the actin cross-linking activity by c-Src (6 , 28) . Importantly, mutation of these residues to Phe abolished the ability of transfected EMS1 to up-regulate the motility of endothelial cells, confirming the relevance of these phosphorylation events in vivo. These data, in combination with our results identifying the HP region as critical for serine/threonine phosphorylation, identify this region as an important regulatory domain. A possible model involves an intramolecular interaction between the SH3 domain and the proline-rich sequences N-terminal to this module, resulting in a "closed" conformation. Phosphorylation of serine and threonine residues in the HP region may then disrupt the intramolecular interaction, resulting in an "open" conformation, as has been suggested for regulation of the focal adhesion kinase-associated Rho-GAP protein Graf (68) . Alternatively, phosphorylation of the HP region may disrupt intermolecular interactions between EMS1 and SH3 domain-containing proteins, in a manner analogous to that proposed for the regulation of the Sos-1/Grb2 complex by Erk phosphorylation (69 , 70) . Circumstantial evidence for this model is that a proline-rich region of the EMS1-related protein LckBP1 that is targetted by the Lck SH3 domain also contains a consensus Erk phosphorylation site (71) . Clearly, serine and threonine phosphorylation of EMS1 represents a conserved regulatory mechanism, and further definition of the sites involved will allow functional analysis of these events.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
1 Supported by the National Health and Medical Research Council of Australia. D. H. C. is a recipient of a National Health and Medical Research Council Dora Lush Postgraduate Research Scholarship. 
2 To whom requests for reprints should be addressed, at Cancer Research Program, Garvan Institute of Medical Research, St. Vincent’s Hospital, Darlinghurst, Sydney, New South Wales 2010, Australia. Phone: 61-2-9295-8333; Fax: 61-2-9295-8321; E-mail: r.daly{at}garvan.unsw.edu.au
3 The abbreviations used are: SH3, Src homology 3; ER, estrogen receptor; Erk, extracellular signal-regulated kinase; MEK, mitogen-activated protein/ERK kinase; EGF, epidermal growth factor; EGFR, EGF receptor; GST, glutathione S-transferase; MBP, myelin basic protein; HP, helical-proline rich; ECL, enhanced chemiluminescence; PVDF, polyvinylidene difluoride; CMV, cytomegalovirus; RIPA, radioimmunoprecipitation assay.
Received 5/ 6/99. Accepted 8/17/99.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  M. Bonazzi, M. Lecuit, and P. Cossart Listeria monocytogenes Internalin and E-cadherin: From Bench to Bedside Cold Spring Harb Perspect Biol, October 1, 2009; 1(4): a003087 - a003087. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. E. Kruchten, E. W. Krueger, Y. Wang, and M. A. McNiven Distinct phospho-forms of cortactin differentially regulate actin polymerization and focal adhesions Am J Physiol Cell Physiol, November 1, 2008; 295(5): C1113 - C1122. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Jia, T. Uekita, and R. Sakai Hyperphosphorylated Cortactin in Cancer Cells Plays an Inhibitory Role in Cell Motility Mol. Cancer Res., April 1, 2008; 6(4): 654 - 662. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. Ayala, M. Baldassarre, G. Giacchetti, G. Caldieri, S. Tete, A. Luini, and R. Buccione Multiple regulatory inputs converge on cortactin to control invadopodia biogenesis and extracellular matrix degradation J. Cell Sci., February 1, 2008; 121(3): 369 - 378. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Timpson, A. S. Wilson, G. M. Lehrbach, R. L. Sutherland, E. A. Musgrove, and R. J. Daly Aberrant Expression of Cortactin in Head and Neck Squamous Cell Carcinoma Cells Is Associated with Enhanced Cell Proliferation and Resistance to the Epidermal Growth Factor Receptor Inhibitor Gefitinib Cancer Res., October 1, 2007; 67(19): 9304 - 9314. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Wang, S.-J. Ding, W. Wang, J. M. Jacobs, W.-J. Qian, R. J. Moore, F. Yang, D. G. Camp II, R. D. Smith, and R. L. Klemke Profiling signaling polarity in chemotactic cells PNAS, May 15, 2007; 104(20): 8328 - 8333. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Luo, H. Pan, M. Mines, K. Watson, J. Zhang, and G.-H. Fan CXCL12 Induces Tyrosine Phosphorylation of Cortactin, Which Plays a Role in CXC Chemokine Receptor 4-mediated Extracellular Signal-regulated Kinase Activation and Chemotaxis J. Biol. Chem., October 6, 2006; 281(40): 30081 - 30093. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. I. Cosen-Binker and A. Kapus Cortactin: The Gray Eminence of the Cytoskeleton. Physiology, October 1, 2006; 21(5): 352 - 361. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. H. Martin, E. D. Jeffery, P. R. Grigera, J. Shabanowitz, D. F. Hunt, and J. T. Parsons Cortactin phosphorylation sites mapped by mass spectrometry J. Cell Sci., July 15, 2006; 119(14): 2851 - 2853. [Full Text] [PDF]
|
|
|
|
|
|
P. Timpson, D. K. Lynch, D. Schramek, F. Walker, and R. J. Daly Cortactin Overexpression Inhibits Ligand-Induced Down-regulation of the Epidermal Growth Factor Receptor Cancer Res., April 15, 2005; 65(8): 3273 - 3280. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Martinez-Quiles, H.-Y. H. Ho, M. W. Kirschner, N. Ramesh, and R. S. Geha Erk/Src Phosphorylation of Cortactin Acts as a Switch On-Switch Off Mechanism That Controls Its Ability To Activate N-WASP Mol. Cell. Biol., June 15, 2004; 24(12): 5269 - 5280. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. G. Burchfield, A. J. Lennard, S. Narasimhan, W. E. Hughes, V. C. Wasinger, G. L. Corthals, T. Okuda, H. Kondoh, T. J. Biden, and C. Schmitz-Peiffer Akt Mediates Insulin-stimulated Phosphorylation of Ndrg2: EVIDENCE FOR CROSS-TALK WITH PROTEIN KINASE C {theta} J. Biol. Chem., April 30, 2004; 279(18): 18623 - 18632. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. K. Lynch, S. C. Winata, R. J. Lyons, W. E. Hughes, G. M. Lehrbach, V. Wasinger, G. Corthals, S. Cordwell, and R. J. Daly A Cortactin-CD2-associated Protein (CD2AP) Complex Provides a Novel Link between Epidermal Growth Factor Receptor Endocytosis and the Actin Cytoskeleton J. Biol. Chem., June 6, 2003; 278(24): 21805 - 21813. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. S. Chew, X. Chen, J. A. Parente Jr, S. Tarrer, C. Okamoto, and H.-Y. Qin Lasp-1 binds to non-muscle F-actin in vitro and is localized within multiple sites of dynamic actin assembly in vivo J. Cell Sci., March 14, 2003; 115(24): 4787 - 4799. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. G. Birukov, A. A. Birukova, S. M. Dudek, A. D. Verin, M. T. Crow, X. Zhan, N. DePaola, and J. G. N. Garcia Shear Stress-Mediated Cytoskeletal Remodeling and Cortactin Translocation in Pulmonary Endothelial Cells Am. J. Respir. Cell Mol. Biol., April 1, 2002; 26(4): 453 - 464. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Z. Chai, B. Sarcevic, A. Mawson, and B.-H. Toh SET-related Cell Division Autoantigen-1 (CDA1) Arrests Cell Growth J. Biol. Chem., August 31, 2001; 276(36): 33665 - 33674. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. M. Simbulan-Rosenthal, D. H. Ly, D. S. Rosenthal, G. Konopka, R. Luo, Z.-Q. Wang, P. G. Schultz, and M. E. Smulson Misregulation of gene expression in primary fibroblasts lacking poly(ADP-ribose) polymerase PNAS, October 10, 2000; 97(21): 11274 - 11279. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Meeting Abstracts Online |
