JBC Anatrace, Inc.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Originally published In Press as doi:10.1074/jbc.M304691200 on June 30, 2003

J. Biol. Chem., Vol. 278, Issue 37, 34845-34853, September 12, 2003
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
278/37/34845    most recent
M304691200v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Han, J.
Right arrow Articles by Ginsberg, M. H.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Han, J.
Right arrow Articles by Ginsberg, M. H.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

Integrin {alpha}4{beta}1-dependent T Cell Migration Requires Both Phosphorylation and Dephosphorylation of the {alpha}4 Cytoplasmic Domain to Regulate the Reversible Binding of Paxillin*

Jaewon Han {ddagger}, David M. Rose {ddagger} §, Darren G. Woodside , Lawrence E. Goldfinger and Mark H. Ginsberg ||

From the Department of Cell Biology, Division of Vascular Biology, The Scripps Research Institute, La Jolla, California 92037

Received for publication, May 6, 2003 , and in revised form, June 30, 2003.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
{alpha}4 integrins mediate increased cell migration and decreased cell spreading because the {alpha}4 cytoplasmic domain (tail) binds tightly to paxillin, a signaling adaptor protein. Paxillin binding to the {alpha}4 tail is blocked by {alpha}4 phosphorylation at Ser988. To establish the biological role of {alpha}4 phosphorylation, we reconstituted {alpha}4-deficient Jurkat T cells with phosphorylation-mimicking ({alpha}4(S988D)) or non-phosphorylatable ({alpha}4(S988A)) mutants. {alpha}4(S988D) disrupted paxillin binding and also inhibited cell migration and promoted cell spreading. In contrast, the non-phosphorylatable {alpha}4(S988A) resulted in a further reduction in cell spreading; however, this mutation led to an unexpected suppression of cell migration. The suppression of cell migration by {alpha}4(S988A) was ascribable to enhanced {alpha}4-paxillin association, because enforced association by an {alpha}4-paxillin fusion led to a phenotype similar to that of the non-phosphorylatable {alpha}4(S988A) mutant. These data establish that optimal {alpha}4-mediated cell migration requires both phosphorylation and dephosphorylation of the {alpha}4 cytoplasmic domain to regulate the reversible binding of paxillin.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Integrin heterodimers are composed of {alpha} and {beta} subunits that each contain a large extracellular domain, a transmembrane domain, and a generally short cytoplasmic tail (1, 2). The extracellular domains of integrins bind ligands in the extracellular matrix (ECM) or on the surface of other cell types to mediate either cell-substratum or cell-cell adhesion. In addition to forming these physical connections, integrins regulate cell signaling pathways through their cytoplasmic domains (1, 3). These signaling pathways are important in coordinating cell migration (4, 5). In particular, the {alpha}4{beta}1 integrin mediates leukocyte migration essential for immune surveillance and inflammation (6).

Adhesion mediated by {alpha}4{beta}1 integrins results in increased cell migration and reduced cell spreading and focal adhesion formation relative to other {beta}1 integrins (7, 8). These unusual biological properties depend on the {alpha}4 integrin cytoplasmic domain (9, 10). Paxillin, a signaling adaptor protein, binds directly and tightly to the {alpha}4 cytoplasmic domain, leading to enhanced migration and reduced cell spreading (10). Paxillin binding to the cytoplasmic domain of {alpha}4 integrins is subject to the regulation by {alpha}4 phosphorylation at Ser988 (11). {alpha}4 phosphorylation at Ser988 is thus likely to modulate all of the responses dependent on the {alpha}4-paxillin interaction. Here, we analyzed the role of {alpha}4 phosphorylation in {alpha}4 integrin functions by reconstituting {alpha}4-deficient Jurkat T cells with phosphorylation-mimicking ({alpha}4(S988D)) or non-phosphorylatable ({alpha}4(S988A)) mutants. {alpha}4(S988D) disrupted paxillin binding and also inhibited cell migration and promoted cell spreading. In contrast, the non-phosphorylatable {alpha}4(S988A) resulted in a further reduction in cell spreading; however, this mutation led to an unexpected suppression of cell migration. The suppression of cell migration by {alpha}4(S988A) was ascribable to enhanced {alpha}4-paxillin association, because enforced association by an {alpha}4-paxillin fusion led to a phenotype similar to that of the nonphosphorylatable {alpha}4(S988A) mutant. These data establish that optimal {alpha}4-mediated cell migration requires both phosphorylation and dephosphorylation of the {alpha}4 cytoplasmic domain to regulate the reversible binding of paxillin.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Materials and DNA Constructs—32P-labeled inorganic phosphate was from PerkinElmer Life Sciences. The SuperSignal Western blotting detection kit was from Pierce Biotechnology. The rabbit polyclonal antibody specific for the cytoplasmic tail of integrin {alpha}4 (Rb038) has been described previously (11). The monoclonal anti-human {alpha}4 antibody (HP2/1) used for immunoprecipitation and flow cytometry was from Immunotech (Marseille, France). The anti-HA1 tag antibody (12CA5) and the anti-human {beta}2 antibody (TS1/18) were from the American Type Culture Collection. The goat anti-PYK2 antibody, used for immunoprecipitation, was from Santa Cruz Biotechnology (Santa Cruz, CA). Monoclonal anti-PYK2, anti-FAK, and anti-paxillin antibodies were purchased from BD Transduction Laboratories. Horseradish peroxidase-conjugated secondary goat anti-mouse F(ab')2 was purchased from BioSource.

Prokaryotic expression vectors encoding model proteins containing the wild type and mutant (S988D) {alpha}4 integrin cytoplasmic domain have been described (11). A previously described strategy (11) was modified slightly to construct a prokaryotic expression vector for the {alpha}4(S988A) recombinant {alpha}4 cytoplasmic domain. Briefly, the {alpha}4 integrin cytoplasmic tail, in pBluescript vector (Stratagene, La Jolla, CA), was mutagenized using the QuikChange mutagenesis kit (Qiagen, Valencia, CA), and a HindIII-BamHI fragment of the mutant was then subcloned into modified pET15b (12). The mutation was then confirmed by cDNA sequencing.

QuikChange mutagenesis (Stratagene) was used to generate mammalian expression vectors for mutant {alpha}4 integrins using a previously described pCDNA3.1(–) vector (Invitrogen) carrying an insert encoding wild type {alpha}4 (pCDNA3.1(–){alpha}4) integrin (11) as a template. Point mutations were confirmed by sequencing. For construction of a mammalian expression encoding an {alpha}4-paxillin chimera, pCDNA3.1(–){alpha}4 was modified by replacing the stop codon with a KpnI site, GTGGGC encoding Val-Gly. A KpnI-XbaI fragment of full-length human paxillin {alpha} was subcloned into the modified pCDNA3.1(–){alpha}4, resulting in a Val-Gly spacer between the C terminus of {alpha}4 and the N terminus of paxillin. For construction of an {alpha}4-TAP fusion, the stop codon in pcDNA3.1(–){alpha}4 was deleted by PCR mutagenesis and replaced with a KpnI-NcoI fragment containing a Gly3 spacer at the C terminus. A 500-bp NcoI-EcoRV fragment containing the complete TAP tag (35) was removed from a cDNA kindly provided by Bertrand Séraphin (European Molecular Biological Laboratory, Heidelberg, Germany) and inserted into the NcoI-EcoRV sites in the modified pCDNA3.1(–){alpha}4 construct.

The mammalian expression vector for the human {beta}1 integrin (pHsk{beta}1A) was a generous gift from Dr. Y. Shimizu (University of Minnesota, Minneapolis, MN). The prokaryotic expression vector for HA-tagged glutathione S-transferase (GST)-human paxillin {alpha} (1.7t/pGEX) was kindly provided by Drs. R. Salgia and J. Griffin (Dana-Farber Cancer Center, Boston, MA). VCAM-Ig, consisting of the complete seven Ig domains of the extracellular region of VCAM-1 fused to the heavy chain of human IgG1, and ICAM-Ig, consisting of the first two N-terminal Ig domains of ICAM-1 fused to the heavy chain of human IgG1, were expressed and purified as described previously (13). A bacterial expression vector for the GST fusion of the CS-1 fragment of fibronectin was a generous gift from Dr. J. W. Smith (Burnham Institute, La Jolla, CA). The expression and purification of this protein was performed as described previously (14).

Cells and Transfections—CHO cells were cultured in Dulbecco's modified Eagle's medium, and Jurkat T cells and {alpha}4-deficient Jurkat T cells were cultured in RPMI 1640. All cell cultures were supplemented with antibiotics, non-essential amino acids, L-glutamine, and 10% fetal calf serum. LipofectAMINE and LipofectAMINE Plus reagents (Invitrogen) were used according to the manufacturer's recommendation for stable and transient transfections. Cells were cultured in the presence of 1 mg/ml G418 (Invitrogen) to select stable clones expressing either wild type or mutant {alpha}4 integrins.

Preparations of Fusion Proteins and Binding Assays—Recombinant model proteins containing the cytoplasmic domain of either wild type or mutant {alpha}4 integrins were expressed in BL21(DE3)pLysS cells (Novagen, Madison, WI), purified on Ni2+-agarose beads, and further purified to >90% homogeneity using the reverse phase C18 high pressure liquid chromatography (HPLC) column (Vydac, Hesperia, CA) as described previously (12). HA-tagged GST-paxillin was expressed and purified as described previously (15, 16). For in vitro binding experiments, HA-tagged GST-paxillin was incubated for 30 min at room temperature with a wild type or a mutant {alpha}4 tail or an {alpha}IIb tail model protein immobilized to Ni2+-agarose beads in PN lysis buffer (10 mM PIPES, pH 6.8, 50 mM NaCl, 150 mM sucrose, 1% Triton X-100, 2 µg/ml aprotinin, 40 µg/ml bestatin, 0.5 µg/ml leupeptin, 0.7 µg/ml pepstatin, and 0.5 mM Pefabloc), and bound paxillin was detected by Western blotting after separation in SDS-PAGE gel as described previously (11, 15).

Metabolic Labeling and Autoradiography—Metabolic cell labeling and autoradiography were performed as described previously (11). Briefly, Jurkat cells expressing either wild type or mutant {alpha}4 integrin were metabolically labeled for 4 h at 37 °C with [32P]orthophosphate (0.3 mCi/ml), and {alpha}4 integrins were immunoprecipitated from the cell lysates prepared with Nonidet P-40 lysis buffer (20 mM HEPES, pH 7.9, 25% (v/v) glycerol, 420 mM NaCl, 1.5 mM MgCl2, 2 µg/ml aprotinin, 40 µg/ml bestatin, 0.5 µg/ml leupeptin, 0.7 µg/ml pepstatin, 0.5 mM Pefabloc, 20 mM glycerophosphate, 50 µM sodium vanadate, 1 mM NaF, and 10 mM p-nitrophenol phosphate), separated by 4–20% SDS-PAGE, and analyzed by autoradiography and Western blotting.

Cell Adhesion Assay—Cell adhesion assays were performed as described previously (13). 96-well Immulon 2HB plates (Dynex Technologies Inc, Chantilly, VA) were coated with the indicated concentrations of VCAM-1. The cells were suspended in a modified Tyrode's buffer (150 mM NaCl, 2.5 mM KCl, 12 mM NaHCO3, 1 mg/ml glucose, and 1 mg/ml BSA) and were then plated on the VCAM-1-coated wells and incubated for 40 min at 37 °C. After washing, adherent cells were stained with 0.5% crystal violet in 20% methanol, Crystal violet was then solubilized with 10% acetic acid and assayed by absorbance at 560 nm in a microplate reader (Molecular Devices, Sunnyvale, CA).

Soluble VCAM Binding Assay—Cells (5 x 105) resuspended in modified Tyrode's buffer containing 1 mM CaCl2 and 1 mM MgCl2 were incubated with 100 nM VCAM-Ig fusion protein for 30 min at room temperature and washed twice with the same Tyrode's buffer. Cells were then resuspended in the same buffer containing FITC-conjugated donkey anti-human IgG (Jackson ImmunoResearch, West Grove, PA) at a 1:100 dilution. After 30 min of incubation at 4 °C, cells were washed twice, and bound antibodies were detected using a FACSCalibur flow cytometer (BD Biosciences) and analyzed using CellQuest software.

Cell Spreading—Cell spreading assays were performed as described previously (15). Briefly, Chinese hamster ovary cells, transiently transfected with either wild type or mutant human integrin {alpha}4 in combination with human {beta}1, were resuspended in Dulbecco's modified Eagle's medium with 1% bovine serum albumin and plated on coverslips coated with 5 µg/ml recombinant GST-CS-1 in 12-well plates. After the indicated incubation periods, spreading was assessed by phase contrast microscopy. Cells that exhibited flattening and the presence of lamel-lipodia were scored as spread cells.

Cell Migration—A modified Boyden chamber assay, described previously (13), was used to measure the motility of Jurkat T cells expressing either wild type or mutant {alpha}4 integrins. Transwell polycarbonate membranes with 3-µm pores (Costar, Corning Inc., Acton, MA) were incubated with VCAM-1 and/or ICAM-1 in 0.1 M NaHCO3, pH 8.0, overnight at 4 °C and blocked with phosphate-buffered saline containing 2% BSA for 30 min at room temperature. 200 µl of cells (2.0 x 105 cells) resuspended in RPMI 1640 plus 10% FCS was added to the upper chamber. 500 µl of same medium containing 15 ng/ml SDF-1{alpha} (R&D Systems, Minneapolis, MN) was added to the lower chamber. The cells were then allowed to migrate at 37 °C for 4 h. At the end of the experiment, the cells in the lower chamber were collected and counted with a hemocytometer.

Immunoprecipitation and Phosphotyrosine Blotting—Jurkat cells (~1 x 106 cells/ml) were harvested, resuspended at 5 x 106 cells/ml in RPMI 1640 without serum (supplemented with 1 mg/ml BSA), and kept at room temperature for 1 h before use. Bacteriological Petri dishes were coated with 5 µg/ml GST-CS-1 in phosphate-buffered saline (pH 8.0) for1hat37 °C and then blocked with BSA (1 mg/ml) for at least 30 min at room temperature. Plates were then washed with RPMI 1640/0.1% BSA, and Jurkat cells were added (5 ml, 25 x 106 cells) and incubated at 37 °C. After the indicated incubation periods, non-adherent cells were removed, and adherent cells were lysed on ice in radioimmune precipitation assay (RIPA) buffer (50 mM Tris, pH 7.4, 150 mM NaCl, 1% Triton X-100, 1% sodium deoxycholate, 0.1% SDS, 2 µg/ml aprotinin, 40 µg/ml bestatin, 0.5 µg/ml leupeptin, 0.7 µg/ml pepstatin, 0.5 mM Pefabloc, 1 mM NaF, and 1 mM Na3VO4). To 200 µg of cell lysate, 1 µg of antibody was added and rotated for at least 2 h at 4 °C. Ten µl of packed protein-G-Sepharose (Amersham Biosciences) was then added and rotated at least for2hat4 °C. Beads were washed two times in radioimmune precipitation assay buffer. Bound proteins were eluted by boiling in reducing SDS-PAGE sample buffer, separated by SDS-PAGE, and then transferred to nitrocellulose. Nitrocellulose membranes were then blocked in 5% milk/TBST (20 mM Tris, pH7.4, 150 mM NaCl, and 0.1% Tween 20), incubated with primary antibody for 2 h at room temperature, washed, then incubated with goat anti-mouse horseradish peroxidase-conjugated secondary antibody. Bound antibody was detected by the SuperSignal Western blotting detection kit (Pierce Biotechnology).


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
Generation of Jurkat T Cells Expressing Mutations of the {alpha}4 Phosphorylation Site—We previously mapped an {alpha}4 phosphorylation site to Ser988 (11). To address the functional role of {alpha}4 phosphorylation at Ser988, we generated an {alpha}4 mutant that mimics constitutively phosphorylated status (S988D) and a mutant that precludes phosphorylation (S988A). The {alpha}4(S988D) mutant failed to bind paxillin, confirming that this mutation recapitulated this biochemical effect of {alpha}4 phosphorylation (Fig. 1A). In contrast, Ala substitution at this residue did not affect paxillin binding (Fig. 1A), indicating that even though this mutant is not phosphorylatable, like wild type {alpha}4, it binds paxillin. To examine the functional effects of each of these mutants, we introduced them into previously described {alpha}4-deficient Jurkat T cells (17). The {alpha}4(S998A) mutation nearly abrogated {alpha}4 phosphorylation; thus, Ser988 is the principal {alpha}4 phosphorylation site in Jurkat cells (Fig. 1B). Both mutant {alpha}4 integrins were expressed at comparable levels to wild type {alpha}4 (Fig. 1C). Furthermore, each mutant mediated cell adhesion and bound soluble VCAM-1 to a similar extent (Fig. 1D). Thus, mutations of the {alpha}4 phosphorylation site that block or mimic receptor phosphorylation are well expressed, activated, and, like wild type {alpha}4, mediate cell adhesion.



View larger version (26K):
[in this window]
[in a new window]
  		FIG. 1.
Characterization of mutations at {alpha}4 phosphorylation site. A, effects of S988A and S988D mutations on paxillin binding to the {alpha}4 ({alpha}4) cytoplasmic domain. Wild type (WT) or mutant {alpha}4 or an {alpha}IIb tail model protein was incubated with GST-paxillin in PN lysis buffer (10 mM PIPES, pH 6.8, 50 mM NaCl, 150 mM sucrose, 1% Triton X-100, and protease inhibitor mixture) for 30 min at room temperature. Bound GST-paxillin was separated by SDS-PAGE (4–20%, reducing) and detected by Western blotting using an anti-HA antibody. B, Ser988 is the major {alpha}4 phosphorylation site in vivo. Wild type (JB4/WT) or mutant (JB4/SA) {alpha}4 integrin was immunoprecipitated (IP) from the 32P-labeled cells ({alpha}4-deficient cells, reconstituted with either wild type or mutant {alpha}4), and subjected to SDS-PAGE. Phosphorylated {alpha}4 integrins were visualized by autoradiography. {alpha}4 was identified by Western blotting using anti-{alpha}4 antibody (Rb038). C, expression of wild type or mutant {alpha}4 integrins reconstituted in {alpha}4-deficient Jurkat T cells. Cells were stained with anti-{alpha}4 (HP2/1) antibodies at 10 µg/ml. Bound antibody was detected with a FITC-conjugated goat anti-mouse IgG by flow cytometry. Open histograms show binding of control mouse IgG, and solid histograms depict binding of the anti-{alpha}4 antibody. D, effects of mutations on {alpha}4{beta}1-dependent static cell adhesion. 96-well plates were coated with the indicated concentration of a CS-1 fragment of fibronectin. Cells were added and allowed to adhere for 40 min at 37 °C. Cell adhesion was quantified by crystal violet staining and enumerated as a percentage of input cell.

 

{alpha}4 Phosphorylation Regulates {alpha}4 Integrin-dependent Cell Spreading and PYK2 Phosphorylation—We found previously that {alpha}4 phosphorylation blocks paxillin binding to the {alpha}4 cytoplasmic domain and that paxillin-{alpha}4 interaction leads to the inhibition of cell spreading and enhanced phosphorylation of FAK and its paralogue, PYK2 (10, 11, 17). Consistent with previous results, the {alpha}4(S988D) mutant abolished the lag phase in {alpha}4-dependent cell spreading (Fig. 2A) and reduced {alpha}4-induced PYK2 phosphorylation (Fig. 3A). Because {alpha}4 is phosphorylated in Jurkat cells, we hypothesized that the {alpha}4(S988A) mutation might lead to an enhancement of the functional effects of paxillin binding. Consistent with this hypothesis, cells stably expressing {alpha}4(S988A) exhibited much reduced cell spreading on an {alpha}4 ligand (Fig. 2B), whereas wild type and mutant cells showed no difference in spreading on an {alpha}5 ligand (9-11 fragment of fibronectin) (data not shown). These cells also manifested increased PYK2 phosphorylation (Fig. 3B). Thus, mutations of the phosphorylation site of {alpha}4 influence cell spreading and downstream signaling as predicted from their effects on paxillin binding.



View larger version (13K):
[in this window]
[in a new window]
  		FIG. 2.
Effects of mutations on cell spreading. Chinese hamster ovary cells expressing {alpha}4(S988D) (A) or {alpha}4(S988A) mutant (B) were plated on coverslips coated with 5 µg/ml recombinant CS-1 in 12-well plates and allowed to spread for the indicated times at 37 °C. Spread cells were enumerated as described previously by Han et al. (11). The data represent the mean ± S.D. of triplicate determinations.

 


View larger version (36K):
[in this window]
[in a new window]
  		FIG. 3.
Effects of mutations on PYK2 tyrosine phosphorylation in Jurkat T cells. Jurkat T cells expressing wild type (WT) or mutant {alpha}4 integrins (S988D (A) or S988A (B)) were allowed to adhere to plates coated with the CS-1 fragment of fibronectin (5 µg/ml) for the indicated times. PYK2 was immunoprecipitated from the cell lysate and separated in SDS-PAGE gel (4–20%, reducing). Tyrosine phosphorylation of PYK2 was detected by immunoblotting with the anti-phosphotyrosine antibody (PY20). The same nitrocellulose membrane was reblotted with anti-PYK2 antibody for determining the amount of immunoprecipitated PYK2.

 

Reversible {alpha}4 Phosphorylation Is Required for {alpha}4-mediated Cell Migration—Another consequence of paxillin binding to {alpha}4 integrins is the enhancement of cell migration. Therefore, we next assessed the role of {alpha}4 phosphorylation in {alpha}4{beta}1-dependent cell migration using the cells stably expressing either a phospho-mimicking (S988D) or a non-phosphorylatable (S988A) mutant {alpha}4. To examine {alpha}4{beta}1-dependent cell migration, wild type or mutant Jurkat cells were allowed to migrate across filters coated with various concentrations of VCAM-1 (0–50 µg/ml) in response to a gradient of SDF-1{alpha}, a chemokine. Jurkat cells expressing wild type {alpha}4 migrated well on VCAM-1 in response to SDF-1{alpha} (Fig. 4). As expected, cells expressing the phospho-mimicking {alpha}4(S988D) mutant showed much less migration. To our surprise, the non-phosphorylatable {alpha}4(S988A) mutant also showed a marked reduction in cell migration at all concentrations of VCAM-1 (Fig. 4A). This result suggests that the phosphorylation of {alpha}4 is required for optimal cell migration.



View larger version (18K):
[in this window]
[in a new window]
  		FIG. 4.
Effects of mutations on cell migration. A, effects of mutations on{alpha}4{beta}1-dependent cell migration. Transwell membranes were coated with the indicated concentration of VCAM-1. Jurkat T cells expressing wild type (WT) or mutant {alpha}4 integrins were allowed to migrate for 4 h at 37 °C in the presence of SDF-1{alpha} (15 ng/ml). Cells migrating to the bottom chamber were enumerated with a hemocytometer, and migration was expressed as a percentage of input cells. Results are mean ± S.E. of three separate experiments. B, effects of mutations on {alpha}4{beta}1-stimulated {alpha}L{beta}2-dependent cell migration. Transwell membranes were coated with 200 µg/ml of ICAM-1 and the indicated concentration of VCAM-1. Migration assays were performed as described above. Cells were added to the top chamber either untreated or treated with the anti-{beta}2 antibody (TS1/18) (20 µg/ml) and allowed to migrate for 4 h at 37 °C in the presence of SDF-1{alpha} (15 ng/ml). Cells in the bottom chamber were enumerated, and migration was expressed as a percent change relative to no VCAM addition. Results are mean ± S.E. of three separate experiments.

 

The foregoing data suggested that {alpha}4 phosphorylation is required for optimal cell migration on a ligand for {alpha}4. However, {alpha}4 integrins can also enhance cell migration by stimulating migration mediated by other integrins (e.g. {alpha}L{beta}2) in trans, a phenomenon referred to as trans-regulation. We therefore investigated the role of {alpha}4 phosphorylation on {alpha}4-stimulated, {beta}2-dependent cell migration. For this purpose, we examined the migration of both wild type and mutant cells on ICAM-1 (200 µg/ml), a ligand for {alpha}L{beta}2 integrins, co-coated with trace amounts of VCAM-1 (0.1–10 µg/ml). Under these conditions, trace amounts of VCAM-1 stimulated the migration of Jurkat cells expressing both wild type {alpha}4 and {alpha}4(S988A) (Fig. 4B). All of the migration was {beta}2-dependent, because it was blocked by TS1/18, an anti-{beta}2 antibody. In contrast, VCAM-1 completely failed to stimulate the {beta}2-dependent migration of {alpha}4(S988D)-bearing Jurkat cells under these conditions. This result is consistent with the finding that {alpha}4 trans-regulation of {alpha}L{beta}2 requires paxillin binding to the {alpha}4 tail (Fig. 4B). These data show that {alpha}4 phosphorylation is not required for {alpha}4{beta}1-dependent trans-regulation of cell migration and that the S988A mutation is not a general inhibitor of cell migration.

Enforced Paxillin Association with the {alpha}4 Tail Recapitulates the {alpha}4(S988A) Phenotype—{alpha}4 phosphorylation blocks paxillin binding, and the non-phosphorylatable {alpha}4(S988A) exhibits exaggerated paxillin-dependent PYK2 phosphorylation and inhibition of cell spreading; however this mutant shows a surprising reduction rather than an increase in cell migration. One consequence of the inhibition of {alpha}4 phosphorylation is that paxillin binding to {alpha}4 would be de-regulated and could become irreversible. Thus, {alpha}4{beta}1-dependent cell migration could require dissociation of paxillin from the {alpha}4 integrin tail by {alpha}4 serine phosphorylation. To test this idea and exclude the possibility that the migratory defect of {alpha}4(S988A) is due to an effect of this mutation unrelated to paxillin binding, we enforced the paxillin {alpha}4 association. To do this, we generated an {alpha}4-paxillin chimera in which full-length paxillin is permanently fused to the C terminus of the {alpha}4 subunit. When introduced into CHO cells, this chimera provoked less cell spreading on a CS-1 fragment of fibronectin than either wild type {alpha}4 or {alpha}4(S988A) (Fig. 5A). In addition, this chimera promoted {alpha}L{beta}2 integrin-dependent cell migration in Jurkat T cells better than wild type {alpha}4 (Fig. 5B). However, this {alpha}4-paxillin chimera completely failed to mediate {alpha}4{beta}1-dependent cell migration (Fig. 5C), whereas fusion of the irrelevant protein to the cytoplasmic tail of {alpha}4 had no effect on cell migration (Fig. 5D). This reduction in {alpha}4{beta}1-dependent cell migration was not due to differences in {alpha}4{beta}1 expression, adhesion, or affinity. The wild type and mutant cells express the similar levels of {alpha}4 integrins (Fig. 6A) and bound soluble VCAM-1 with similar affinity (Fig. 6B). Furthermore, immunoblotting with a phospho-specific antibody confirmed that Ser988 in {alpha}4-paxillin was phosphorylated in vivo to a similar level as that of wild type {alpha}4 (Fig. 6C). Thus, the {alpha}4-paxillin fusion recapitulates the phenotype of the {alpha}4(S988A) mutant despite undiminished {alpha}4 Ser988 phosphorylation. Consequently, the inhibition of cell migration by blockade of {alpha}4 phosphorylation is ascribable to the increased association of paxillin with the {alpha}4 tail.



View larger version (14K):
[in this window]
[in a new window]
  		FIG. 5.
Enforced paxillin association with {alpha}4 tail inhibits {alpha}4{beta}1-dependent cell migration. A, effect of {alpha}4-paxillin chimera on cell spreading. Chinese hamster ovary cells, transiently transfected with wild type, S988A mutant, or {alpha}4-paxillin chimeric mutant human {alpha}4 integrin in combination with human {beta}1 were allowed to spread on the CS-1 fragment of fibronectin for the indicated times. Spread cells were enumerated as described under "Experimental Procedures." Assays for {alpha}4{beta}1-stimulated {alpha}L{beta}2-dependent cell migration (B) and {alpha}4{beta}1-dependent cell migration (C) were performed with cells expressing either wild type {alpha}4 or {alpha}4-paxillin chimera as described under "Experimental Procedures." D, fusion of irrelevant protein to the cytoplasmic tail of {alpha}4 integrin does not affect {alpha}4-dependent cell migration. Migration assays using Jurkat T cells expressing either wild type {alpha}4 or {alpha}4-TAP ({alpha}4-X) chimeric integrin were performed as described above. The data present in each panel represent the mean ± S.D. of triplicate determinations.

 


View larger version (12K):
[in this window]
[in a new window]
  		FIG. 5.
Enforced paxillin association with {alpha}4 tail inhibits {alpha}4{beta}1-dependent cell migration. A, effect of {alpha}4-paxillin chimera on cell spreading. Chinese hamster ovary cells, transiently transfected with wild type, S988A mutant, or {alpha}4-paxillin chimeric mutant human {alpha}4 integrin in combination with human {beta}1 were allowed to spread on the CS-1 fragment of fibronectin for the indicated times. Spread cells were enumerated as described under "Experimental Procedures." Assays for {alpha}4{beta}1-stimulated {alpha}L{beta}2-dependent cell migration (B) and {alpha}4{beta}1-dependent cell migration (C) were performed with cells expressing either wild type {alpha}4 or {alpha}4-paxillin chimera as described under "Experimental Procedures." D, fusion of irrelevant protein to the cytoplasmic tail of {alpha}4 integrin does not affect {alpha}4-dependent cell migration. Migration assays using Jurkat T cells expressing either wild type {alpha}4 or {alpha}4-TAP ({alpha}4-X) chimeric integrin were performed as described above. The data present in each panel represent the mean ± S.D. of triplicate determinations.

 


View larger version (25K):
[in this window]
[in a new window]
  		FIG. 6.
Characterization of Jurkat T cells expressing {alpha}4-paxillin chimeric integrin. A, expression of chimeric integrin in reconstituted Jurkat T cells. Cells expressing wild type {alpha}4 or {alpha}4-paxillin chimeric integrin were stained with anti-{alpha}4 antibody (HP2/1). Bound antibody was detected with a FITC-conjugated goat anti-mouse IgG by flow cytometry. Open histograms show binding of the control mouse IgG and solid histograms show binding of the anti-{alpha}4 antibody. B, soluble VCAM-1 binding to Jurkat T cells expressing wild type {alpha}4 ({alpha}4WT) or chimeric ({alpha}4-Pax) {alpha}4-paxillin integrin. Soluble VCAM-1 binding assays were performed as described under "Experimental Procedures." Briefly, cells were incubated with the soluble VCAM-Ig in the absence (histograms with solid gray line) or presence (histograms with a dotted black line) of activating antibody (8A2). Bound VCAM-1 was detected with a FITC-conjugated goat anti-human IgG by flow cytometry. Quantification of VCAM-1 binding was assessed as an activation index (AI) defined as (Fo – Fr)/(Fmax – Fr) x 100, in which Fo is the mean fluorescence intensity of sVCAM-1 binding, Fr is fluorescence intensity in the presence of mAb HP2/1, and Fmax is the fluorescence intensity in the presence of mAb 8A2. C, wild type {alpha}4 ({alpha}4) or chimeric {alpha}4 ({alpha}4-paxillin) was immunoprecipitated from the Jurkat T cells expressing wild type or chimeric {alpha}4 integrin and immunoblotted with phospho-specific anti-{alpha}4 antibody (PS{alpha}4), anti-{alpha}4 antibody (Rb038), or anti-paxillin antibody.

 


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
{alpha}4 integrins mediate distinct biological responses because the {alpha}4 cytoplasmic domain (tail) binds tightly to paxillin, a signaling adaptor protein. Phosphorylation of the {alpha}4 tail at Ser988 inhibits paxillin binding. In the present study, we have examined the biological effects of {alpha}4 phosphorylation and report that phosphorylation of the {alpha}4 tail is required for optimal {alpha}4{beta}1-mediated cell migration. Secondly, enforced association of {alpha}4 and paxillin in an {alpha}4-paxillin chimera inhibits cell migration, indicating that {alpha}4 phosphorylation enhances cell migration because it promotes the dissociation of paxillin from the {alpha}4 tail. Thirdly, the {alpha}4(S988A) mutant and the {alpha}4-paxillin chimera both augmented three other {alpha}4 integrin functions, i.e. increased PYK2 phosphorylation, inhibition of cell spreading, and transregulation of {beta}2 integrins. Thus, loss of {alpha}4 phosphorylation has differential effects on the paxillin-dependent biological responses to {alpha}4 integrin ligation. It enhances most responses but, paradoxically, suppresses directed cell migration. Finally, {alpha}4 dephosphorylation is also required for {alpha}4 integrin-dependent cellular functions. The {alpha}4(S988D) mutant, which mimics constitutively phosphorylated {alpha}4, blocks paxillin binding and consequently inhibits all paxillin-dependent responses. These results establish the biological consequences of a blockade of {alpha}4 phosphorylation and dephosphorylation and show that the regulation of paxillin binding by {alpha}4 phosphorylation is central to these effects. Furthermore, these data show that reversible phosphorylation of the {alpha}4 integrin cytoplasmic domain is required for optimal {alpha}4{beta}1 integrin-mediated directed cell migration.

Phosphorylation of {alpha}4 Ser988 is required for optimal {alpha}4{beta}1-mediated cell migration. This conclusion is based on the inhibition of {alpha}4-dependent cell migration by introduction of a non-phosphorylatable Ala at {alpha}4 Ser988. Previous studies have established the importance of phosphorylation of the integrin {alpha}3 cytoplasmic domain in mediating cell migration by that integrin (18). Furthermore, tyrosine phosphorylation of the {beta}1 integrin is required for optimal {beta}1-mediated cell migration (19, 20). However, in the previous studies the biochemical effects of phosphorylation that lead to the increase in cell migration have not been identified. In the present study (see below), inhibition of paxillin binding has been strongly implicated as the relevant downstream target of {alpha}4 phosphorylation. Thus, the phosphorylation of {alpha}4 Ser988 is required for optimal {alpha}4{beta}1-mediated migration in Jurkat T-cells.

Phosphorylation of {alpha}4 Ser988 is required for cell migration because it promotes dissociation of paxillin. We established previously that {alpha}4 phosphorylation at Ser988 blocks paxillin binding in vitro and the physical association of paxillin with {alpha}4 integrins in Jurkat T-cells. In the present study, a blockade of this phosphorylation markedly inhibited cell migration, suggesting that the failure of paxillin to dissociate might account for the defect in cell migration. This was confirmed by the response to adhesion mediated by an {alpha}4-paxillin chimera in which the association of paxillin with the {alpha}4 tail is enforced by a covalent bond. This chimera exhibited a profound lack of cell migration. In contrast, fusion of the {alpha}4 tail with an irrelevant protein had no effect on cell migration. Moreover, the effect of this chimera on cell migration was not due to an inhibition of {alpha}4 phosphorylation, because immunoblotting confirmed phosphorylation of {alpha}4 Ser988 in these chimeras. Lack of paxillin is known to markedly impair cell migration (21). However, our studies now suggest that the enforced association of paxillin with an integrin also limits cell migration. In this regard, paxillin shows a remarkable similarity to filamin, which is required for optimal cell migration (22) but, when tightly associated with an integrin {beta} cytoplasmic domain, inhibits cell migration (23). These two results together emphasize the importance of dynamic regulation of adhesion receptor-accessory protein interactions in the control of the migratory process.

Paxillin is an adaptor that binds to several proteins, which can adversely impact cell migration. For example, paxillin is physically associated with the protein tyrosine phosphatase PTP-PEST (24), and PTP-PEST can inhibit cell migration by dephosphorylating p130CAS (25, 26). Similarly, paxillin, when tyrosine phosphorylated, can bind Csk, a negative regulator of Src kinases and, hence, a negative regulator of cell migration (27, 28). Furthermore, pp125FAK (FAK) and PYK2 can regulate the assembly of integrin-dependent complexes that control cell migration (29–31). Secondly, FAK and PYK2 bind directly to paxillin, and {alpha}4 integrins activate them more efficiently than other integrins (10, 17). Thus, irreversible paxillin binding to an integrin could lead to constitutive membrane targeting of FAK/PYK2, resulting in de-regulated activation. Indeed, a membrane-anchored CD2-FAK chimera manifests constitutive activation and inhibits cell migration (32). The approaches described here should make it possible to determine which of the biochemical pathways connect the enforced association of paxillin with integrins to impaired cell migration.

A blockade of {alpha}4 phosphorylation promotes several paxillin-dependent cellular responses to {alpha}4{beta}1 integrin-mediated adhesion. For example, paxillin binding to {alpha}4 integrins inhibits cell spreading and promotes the phosphorylation of PYK2 tyrosine kinase and thus, the trans-regulation of {beta}2 integrin-dependent cell migration. Each of these responses to {alpha}4 ligation was markedly potentiated by the Ser988 -> Ala mutant. Furthermore, the effects of this mutation are ascribable to increased paxillin association, because they were recapitulated by enforced paxillin binding. Thus, these results lend additional credence to the proposal that the tight association of paxillin with the {alpha}4 cytoplasmic domain leads to these cellular responses. As noted above, paxillin interacts with a wide range of signaling molecules that could mediate these biological outcomes.

Dephosphorylation of the {alpha}4 cytoplasmic domain is required for optimal {alpha}4 integrin-mediated signaling. Previous studies established the importance of tight paxillin binding to {alpha}4 for cellular responses and also established that the {alpha}4 tail may be phosphorylated (10, 11). We now report that a mutation that mimics the constitutively phosphorylated state of {alpha}4 and inhibits paxillin binding blocks the paxillin-dependent, {alpha}4-mediated responses. Previous studies showed that phosphorylation of integrin {beta} cytoplasmic domains could adversely affect targeting of integrins to focal adhesions (20) and cell adhesion (33). These effects may be due to the effects of {beta}1 tyrosine phosphorylation on the binding of {beta}1A to talin (34). The present study now adds the inhibition of paxillin binding to the {alpha}4 tail to the mechanisms by which phosphorylation can inhibit integrin interactions with protein partners to modify cellular responses. In addition, these results imply that dephosphorylation of {alpha}4 is required for {alpha}4-dependent cell responses such as cell migration and suggest that efforts should be made to identify the serine phosphatase(s) responsible for dephosphorylation of {alpha}4. Thus, {alpha}4 dephosphorylation is required for optimal {alpha}4{beta}1-mediated biological responses.

As noted above, several integrin-dependent cell functions are regulated by phosphorylation of integrin cytoplasmic domains. However, the present study provides direct evidence that a dynamic phosphorylation-dephosphorylation cycle of an integrin cytoplasmic domain is required for optimal cell migration. Furthermore, the effects of this specific phosphorylation appear to be largely mediated by its capacity to inhibit the association of the downstream adaptor, paxillin. Thus, our studies strongly direct future research at understanding the kinases and phosphatases that regulate {alpha}4 integrin phosphorylation as well as the downstream targets of paxillin responsible for the effects of phosphorylation on {alpha}4-mediated migration. {alpha}4 integrins play pivotal roles in developmental and pathological processes. Manipulation of the phosphorylation state of {alpha}4 integrins may provide a useful tool to modulate these processes.


   FOOTNOTES
 
* This work was supported by grants from the National Institutes of Health (to M. H. G.), the Arthritis Foundation and National Multiple Sclerosis Society (to J. H.), and Juvenile Diabetes Foundation International (to D. M. R.). This is publication 15777-CB from the Scripps Research Institute. 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. Back

{ddagger} These authors contributed equally to this work. Back

§ Present address: Dept. of Medicine, Division of Rheumatology, Allergy, and Immunology, University of California at San Diego, La Jolla, CA 92093. Back

Present address: Dept. of Immunology, Texas Biotechnology Corporation, Houston, TX 77030. Back

|| To whom correspondence should be addressed: Dept. of Cell Biology, Division of Vascular Biology, The Scripps Research Institute, 10550 N. Torrey Pines Rd., La Jolla, CA 92037. Tel.: 858-784-7124; Fax: 858-784-7343; E-mail: ginsberg{at}scripps.edu.

1 The abbreviations used are: HA, hemagglutinin; PYK2, proline-rich tyrosine kinase 2; FAK, focal adhesion kinase; GST, glutathione S-transferase; CS-1, connecting segment 1; TAP, tandem affinity purification; VCAM, vascular cell adhesion molecule; CHO, Chinese hamster ovary; PIPES, 1,4-piperazinediethanesulfonic acid; BSA, bovine serum albumin; FITC, fluorescein isothiocyanate; ICAM intracellular adhesion molecule; SDF-1{alpha}, stromal cell-derived factor 1{alpha}. Back



   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Hynes, R. O. (1992) Cell 69, 11–25[CrossRef][Medline] [Order article via Infotrieve]
  2. Sastry, S. K., and Horwitz, A. F. (1993) Curr. Opin. Cell Biol. 5, 819–831[CrossRef][Medline] [Order article via Infotrieve]
  3. Schwartz, M. A., Schaller, M. D., and Ginsberg, M. H. (1995) Annu. Rev. Cell Dev. Biol. 11, 549–599[CrossRef][Medline] [Order article via Infotrieve]
  4. Huttenlocher, A., Sandborg, R. R., and Horwitz, A. F. (1995) Curr. Opin. Cell Biol. 7, 697–706[CrossRef][Medline] [Order article via Infotrieve]
  5. Lauffenburger, D. A., and Horowitz, A. F. (1996) Cell 84, 359–369[CrossRef][Medline] [Order article via Infotrieve]
  6. Rose, D. M., Han, J., and Ginsberg, M. H. (2002) Immunol. Rev. 186, 118–124[CrossRef][Medline] [Order article via Infotrieve]
  7. Chan, B. M. C., Kassner, P. D., Schiro, J. A., Byers, H. R., Kupper, T. S., and Hemler, M. E. (1992) Cell 68, 1051–1060[CrossRef][Medline] [Order article via Infotrieve]
  8. Kassner, P. D., Alon, R., Springer, T. A., and Hemler, M. (1995) Mol. Biol. Cell 6, 661–674[Abstract]
  9. Hemler, M. E., Kassner, P. D., and Chan, B. M. C. (1992 Cold Spring Harbor Symp. Quant. Biol. 57, 213–220[Medline] [Order article via Infotrieve]
  10. Liu, S., Thomas, S. M., Woodside, D. G., Rose, D. M., Kiosses, W. B., Pfaff, M., and Ginsberg, M. H. (1999) Nature 402, 676–681[CrossRef][Medline] [Order article via Infotrieve]
  11. Han, J., Liu, S., Rose, D. M., Schlaepfer, D. D., McDonald, H., and Ginsberg, M. H. (2001) J. Biol. Chem. 276, 40903–40909[Abstract/Free Full Text]
  12. Pfaff, M., Liu, S., Erle, D. J., and Ginsberg, M. H. (1998) J. Biol. Chem. 273, 6104–6109[Abstract/Free Full Text]
  13. Rose, D. M., Grabovsky, V., Alon, R., and Ginsberg, M. H. (2001) J. Immunol. 167, 2824–2830[Abstract/Free Full Text]
  14. Jongewaard, I. N., Tsai, P. M., and Smith, J. W. (1996) Cell Adhes. Commun. 3, 487–495[Medline] [Order article via Infotrieve]
  15. Liu, S., and Ginsberg, M. H. (2000) J. Biol. Chem. 275, 22736–22742[Abstract/Free Full Text]
  16. Salgia, R., Li, J.-L., Ewaniuk, D. S., Wang, Y.-B., Sattler, M., Chen, W.-C., Richards, W., Pisick, E., Shapiro, G. I., Rollins, B. J., Chen, L. B., Griffin, J. D., and Sugarbaker, D. J. (1999) Oncogene 18, 67–77[CrossRef][Medline] [Order article via Infotrieve]
  17. Rose, D. M., Liu, S., Woodside, D. G., Han, J., Schlaepfer, D. D., and Ginsberg, M. H. (2003) J. Immunol. 170, 5912–5918[Abstract/Free Full Text]
  18. Zhang, X. A., Bontrager, A. L., Stipp, C. S., Kraeft, S. K., Bazzoni, G., Chen, L. B., and Hemler, M. E. (2001) Mol. Biol. Cell 12, 351–365[Abstract/Free Full Text]
  19. Sakai, T., Peyruchaud, O., Fassler, R., and Mosher, D. F. (1998) J. Biol. Chem. 273, 19378–19382[Abstract/Free Full Text]
  20. Sakai, T., Zhang, Q., Fassler, R., and Mosher, D. F. (1998) J. Cell Biol. 141, 527–538[Abstract/Free Full Text]
  21. Hagel, M., George, E. L., Kim, A., Tamimi, R., Opitz, S. L., Turner, C. E., Imamoto, A., and Thomas, S. M. (2002) Mol. Cell. Biol. 22, 901–915[Abstract/Free Full Text]
  22. Cunningham, C. C., Gorlin, J. B., Kwiatkowski, D. J., Hartwig, J. H., Janmey, P. A., Byers, H. R., and Stossel, T. P. (1992) Science 255, 325–327[Abstract/Free Full Text]
  23. Calderwood, D. A., Huttenlocher, A., Kiosses, W. B., Rose, D. M., Woodside, D. G., Schwartz, M. A., and Ginsberg, M. H. (2001) Nat. Cell Biol. 3, 1060–1068[CrossRef][Medline] [Order article via Infotrieve]
  24. Shen, Y., Schneider, G., Cloutier, J. F., Veillette, A., and Schaller, M. D. (1998) J. Biol. Chem. 273, 6474–6481[Abstract/Free Full Text]
  25. Garton, A. J., Flint, A. J., and Tonks, N. K. (1996) Mol. Cell. Biol. 16, 6408–6418[Abstract]
  26. Garton, A. J., and Tonks, N. K. (1999) J. Biol. Chem. 274, 3811–3818[Abstract/Free Full Text]
  27. Sabe, H., Hata, A., Okada, M., Nakagawa, H., and Hanafusa, H. (1994) Proc. Natl. Acad. Sci. U. S. A. 91, 3984–3988[Abstract/Free Full Text]
  28. Turner, C. E. (2000) Nat. Cell Biol. 2, E231–E236[CrossRef][Medline] [Order article via Infotrieve]
  29. Ilic, D., Furuta, Y., Kanazawa, S., Takeda, N., Sobue, K., Nakatsuji, N., Norura, S., Fujimoto, J., Okada, M., Yamamoto, T., and Alzawa, S. (1995) Nature 377, 539–544[CrossRef][Medline] [Order article via Infotrieve]
  30. Cary, L. A., Chang, J. F., and Guan, J.-L. (1996) J. Cell Sci. 109, 1787–1794[Abstract]
  31. Li, S., Butler, P., Wang, Y., Hu, Y., Han, D. C., Usami, S., Guan, J. L., and Chien, S. (2002) Proc. Natl. Acad. Sci. U. S. A. 99, 3546–3551[Abstract/Free Full Text]
  32. Huttenlocher, A., Lakonishok, M., Kinder, M., Wu, S., Truong, T., Knudsen, K. A., and Horwitz, A. F. (1998) J. Cell Biol. 141, 515–526[Abstract/Free Full Text]
  33. Sakai, T., Jove, R., Fassler, R., and Mosher, D. F. (2001) Proc. Natl. Acad. Sci. U. S. A. 98, 3808–3813[Abstract/Free Full Text]
  34. Tapley, P., Horwitz, A., Buck, C. A., Duggan, K., and Rohrschneider, L. (1989) Oncogene 4, 325–333[Medline] [Order article via Infotrieve]
  35. Rigaut, G., Shevchenko, A., Rutz, B., Wilm, M., Mann, M., and Seraphin, B. (1999) Nat. Biotechnol. 17, 1030–1032[CrossRef][Medline] [Order article via Infotrieve]

Add to CiteULike CiteULike   Add to Complore Complore   Add to Connotea Connotea   Add to Del.icio.us