HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 275, Issue 35, 27155-27164, September 1, 2000

This Article Abstract Full Text (PDF) All Versions of this Article: 275/35/27155    most recent M000679200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Nakamura, K. Articles by Sabe, H. Search for Related Content PubMed PubMed Citation Articles by Nakamura, K. Articles by Sabe, H. Social Bookmarking                      What's this?

Tyrosine Phosphorylation of Paxillin  Is Involved in Temporospatial Regulation of Paxillin-containing Focal Adhesion Formation and F-actin Organization in Motile Cells^*

Kuniaki Nakamura, Hajime Yano, Hiroshi Uchida, Shigeru Hashimoto, Erik Schaefer^§, and Hisataka Sabe^¶**

From the  Department of Molecular Biology, Osaka Bioscience Institute, Suita, Osaka 565-0874, the ^¶ Graduate School of Biostudies, Kyoto University, Sakyoku, Kyoto 606-8502, and the  Precursory Research for Embryonic Science and Technology, Japan Science and Technology Corporation, Kyoto 619-0237, Japan, and ^§ BioSource International, Hopkinton, Massachusetts 01748

Received for publication, January 24, 2000, and in revised form, April 26, 2000

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Temporal and spatial regulation of actin-based cytoskeletal organization and focal adhesion formation play an essential role in cell migration. Here, we show that tyrosine phosphorylation of a focal adhesion protein, paxillin, crucially participates in these regulations. We found that tyrosine phosphorylation of paxillin was a prominent event upon integrin activation during epithelial-mesenchymal trans-differentiation and cell migration. Four major tyrosine phosphorylation sites were identified, and two of them were highly inducible upon integrin activation. Paxillin exhibits three distinct subcellular localizations as follows: localization along the cell periphery colocalized with circumferential actin meshworks, macroaggregation at focal adhesions connected to actin stress fibers, and diffuse cytoplasmic distribution. Tyrosine phosphorylation of paxillin localized at the cell periphery and focal adhesions was shown using phosphorylation site-specific antibodies. Mutations in the phosphorylation sites affected the peripheral localization of paxillin and paxillin-containing focal adhesion formation during cell migration and cell-cell collision, accompanied by altered actin organizations. Our analysis indicates that phosphorylation of multiple tyrosines in paxillin  is necessary for the proper function of paxillin and is involved in the temporospatial regulation of focal adhesion formation and actin cytoskeletal organization in motile cells.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Cell migration plays an essential role in a wide variety of physiological and pathological processes of multicellular organisms, such as embryogenesis, organogenesis, wound repair, inflammatory processes, and cancer invasion and metastasis. Cell locomotion is primarily mediated by binding of integrin to the extracellular matrices and an actin cytoskeleton-based force-generation system (1-5). Actin cytoskeletal organization also plays an important role in cell migration. Activation of intracellular GTPase/GTP-binding proteins, including the Rho family GTPases, has been shown to play a pivotal role in intracellular regulation of the dynamic properties of actin-based cytoskeletal organization, as well as the formation of focal adhesion complexes (6-8). Similarly, the activity of small GTP-binding ARF family proteins also participates in actin reorganization and focal complex formation (9-12). In addition, a number of effector proteins for these GTP-binding proteins have been identified. However, despite intense investigation of these underlying pathways, the mechanisms regulating spatial and temporal control of focal adhesion formation and actin cytoskeletal organization are not yet well established.

Higher vertebrates utilize a succession of tissue transformations between epithelium and mesenchyme during embryogenesis (13). During trans-differentiation of epithelial cells to mesenchymal cells (epithelial-mesenchymal trans-differentiation; EMT),¹ cadherin-mediated cell-cell adhesions almost disappear, and the expression and the avidity of integrins are highly augmented, thus enabling the cell to move (14). Several cytokines, such as TGF-1-3, TGF-, Mullerian inhibitory factor, and acidic FGF, have been shown to induce EMT in vitro; and several oncogenes, including v-src, v-ras, and v-mos, have also been shown to induce EMT (14). The EMT system in vitro is thus well established and provides a way to analyze the early events during integrin activation and cell migration, accompanied by a change from the cadherin-mediated cell-cell adherent phenotype to the integrin-mediated cell-migratory phenotype. Increasing numbers of signaling molecules have been shown to participate in integrin signaling and integrin-mediated cell migration (3, 15-19). Moreover, integrin-mediated cell migration signaling appears to intercommunicate with and thus be regulated by a number of different intracellular signals generated by cytokines, growth factors, cadherins, and other integrins (20). One of the early events that occurs upon integrin activation is the tyrosine phosphorylation of several integrin-associated proteins, including paxillin (21). Paxillin is composed of multiple isoforms, but the  isoform appears to play a more dominant role as compared with other isoforms (22, 23). Tyrosine phosphorylation of paxillin has been shown to be important for the formation of focal adhesions and cell cycle progression into S-phase (21, 24), and several tyrosine kinases have been implicated in paxillin phosphorylation (24-26). Lack of paxillin tyrosine phosphorylation in neutrophils isolated from a patient with a leukocyte adhesion deficiency also implicates its importance in leukocyte function (27). Several signaling molecules, such as Csk and v-Crk, have been shown to bind to tyrosine-phosphorylated paxillin via their src homology 2 (SH2) domains. Paxillin also interacts with several focal adhesion proteins with scaffold and/or catalytic signaling properties, including vinculin, talin, integrin ₁, focal adhesion kinase (Fak), and c-Src (15, 28). Moreover, papilloma virus E6 protein binds to paxillin, and this binding correlates with disruption of the actin cytoskeletal architecture in virus-infected cells (29, 30). Thus, paxillin acts as an adaptor molecule and seems to be essential for integrin signaling.

In this report, we used in vitro EMT of normal murine mammary gland cells, NMuMG (31), as a model to explore the mechanism of regulation of cell migratory properties that accompany integrin activation. We found that tyrosine phosphorylation of the  isoform of paxillin is one of the most prominent events during EMT, together with increased tyrosine phosphorylation of p130^Cas. Similar patterns of phosphorylation were also observed when these epithelial cells were migrating actively. We have identified four major tyrosine phosphorylation sites in paxillin , and we showed that multiple sites of paxillin phosphorylation seem to play important roles in the formation of paxillin-containing focal adhesions and F-actin organization in motile cells.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Cells-- NMuMG cells (CRL 1639) with a passage number of 15 were obtained from the American Type Culture Collection (Manassas, VA) and grown in Dulbecco's modified Eagle's medium (with 4.5 g glucose/liter) supplemented with 10% fetal calf serum (HyClone Laboratories, Inc., Logan, UT), 10 µg/ml insulin (Life Technologies, Inc.), 100 µg/ml penicillin, and 100 units/ml streptomycin at 37 °C in 5% CO₂. After the initial expansion for 3 days cells were frozen in aliquots, and in each experiment, cells were cultured no longer than 2 weeks by subculturing at a dilution of 1:10-1:20 every 3rd day following 0.25% trypsin/EDTA treatment. Trypsinization was done for 10 min at ambient temperature.

For trans-differentiation into the mesenchymal phenotype, 8 × 10⁵ NMuMG cells were seeded in a 9-cm culture dish (Becton Dickinson), and 24 h later 2 ng/ml TGF-1 (R & D Systems, Minneapolis, MN) was added, and cells were cultured further for 48 h as described previously (31). For analysis of the epithelial phenotype, 8 × 10⁵ cells were seeded in a 9-cm culture dish and cultured for 3 days. Parental NMuMG cells, both in epithelial and mesenchymal forms, reached confluence under these conditions. To examine cells under sparse culture conditions, 2 × 10⁵ cells were seeded initially and then processed as above. Under these conditions, apparent cell confluence is less than 20%.

cDNA Clones and Expressions-- Enhanced green fluorescent protein (EGFP; CLONTECH, Palo Alto, CA)-tagged paxillin  isoform cDNA in pBabePuro vector (32) was described previously (23). Each potential tyrosine phosphorylation site (33) was mutated into phenylalanine singly or in various combinations by using the following designations: 31F, 40F, 118F, 181F, 2X(31/118F), 2Y(40/181/434/488F), 4X (31/40/118/181F), and 6X(31/40/118/181/434/488F) (see Fig. 2A). All mutant cDNAs were made with the Altered Sites II in Vitro Mutagenesis System (Promega Corp., Madison, WI) with appropriate synthetic DNA fragments.

Each wild-type and mutant paxillin cDNA in the pBabePuro vector was transfected into BOSC 23 cells, and each recombinant virus was collected as described (34). Virus titers were in the range of 10⁵-10⁶ infectious units/ml. NMuMG cells were infected with these viruses, and selection was imposed 2 days later with 1 µg/ml puromycin (Sigma) for 1 more week. During selection, passage and expansion of cells were performed when necessary. Cells then were frozen in aliquots, and in each experiment cells were thawed and cultured for no longer than 2 weeks.

Antibodies-- Rabbit anti-paxillin antibody (Ab199-217), which recognizes both the  and  isoforms, and the paxillin -specific antibody were described previously (22, 23). Rabbit antibodies against phosphotyrosine and focal adhesion kinase (Fak) were described previously (35, 36). Anti-p130^Cas antibody was a gift from Dr. H. Hirai (Tokyo University) or purchased from Transduction Laboratories Inc. (Lexington, KY). Monoclonal anti-paxillin antibody (Transduction Laboratories), anti-phosphotyrosine (clone 4G10, Upstate Biotechnology, Lake Placid, NY), and Cy2- or Cy3-conjugated secondary antibodies (Jackson ImmunoResearch) were purchased from commercial sources. Phosphorylation site-specific antibodies that recognize paxillin when phosphorylated on tyrosine 31 (anti-pY31 paxillin) or tyrosine 118 (anti-pY118 paxillin) were obtained from BIOSOURCE International (Camarillo, CA). These antibodies are affinity-purified (using both negative and positive affinity purification methods) rabbit polyclonal antibodies that are highly selective for the targeted phosphorylation site, as demonstrated by peptide competition studies and through use of site-directed mutants possessing a tyrosine to phenylalanine substitution at the phosphorylation site (see Fig. 3A).

Immunoblotting Analysis-- Cell lysates were prepared with RIPA buffer (1% Nonidet P-40, 0.1% SDS, 1% sodium deoxycholate, 150 mM NaCl, 20 mM Tris-HCl (pH 7.4), 5 mM EDTA, 1 mM Na₃VO₄, 1 mM phenylmethylsulfonyl fluoride, 1% aprotinin, 2 µg/ml leupeptin, and 3 µg/ml pepstatin A) as described previously (35). For immunoblotting, 20 µg of cell lysate was boiled in Laemmli's SDS sample buffer, separated by 8% SDS-PAGE, and transferred to membrane filters (Immobilon P, Millipore Corp., Bedford, MA), and subjected to immunoblotting analysis as described previously (35). Antibodies retained on the filter membranes were visualized by horseradish peroxidase-conjugated donkey anti-rabbit or anti-mouse IgG secondary antibodies (Jackson ImmunoResearch) coupled with an enzyme-linked chemiluminescence method according to the manufacturer's instructions (Amersham Pharmacia Biotech). For immunoprecipitation, 500 µg of each cell lysate was used.

Indirect Immunofluorescence-- 8 × 10⁴ NMuMG cells were seeded into each 3.5-cm culture dish, possessing a hole at the bottom where a number 0 glass coverslip was attached (MatTek Corp., Ashland, MA). After 24 h, cells were treated with or without TGF-1 and cultured further for 48 h, as described above. To analyze motile cells, confluent cultures of cells were scratched manually with the needle of a 10-µl syringe (Hamilton Co., Carson, NV). Wound regions were allowed to heal for 0-16 h prior to analysis. Cells then were fixed with 3.7% paraformaldehyde (Sigma) in HBSS for 20 min at room temperature. Fixed cells were permeabilized with 0.1% Triton X-100 in HBSS for 2 min. After being washed with HBSS, cells were soaked in HBSS containing 10% skim milk (blocking solution) for 30 min followed by incubation with 2 µg/ml anti-paxillin antibody in blocking solution for 1 h at room temperature. Cells then were washed with HBSS and incubated with Cy2- or Cy3-conjugated secondary antibody at 1:400 dilution in blocking solution for 1 h at room temperature. F-actin was visualized by incubation for 1 h with Texas Red-X phalloidin (Molecular Probes, Eugene, OR) at 1:200 dilution. After washing, the samples were mounted in 50% glycerol/phosphate-buffered saline, and confocal images were acquired using a confocal laser-scanning microscope (model 510; Carl Zeiss, Inc., Oberkochen, Germany). For double staining with phosphotyrosine and paxillin, 4G10 and rabbit anti-paxillin antibody (Ab199-217) were used. Rabbit anti-paxillin tyrosine 31- or tyrosine 118-phosphospecific antibody and mouse anti-paxillin antibody were used to detect tyrosine phosphorylation of paxillin.

For time-lapse microscopy, cells were incubated on a glass bottom 35-mm dish on the plate heated at 37 °C, and the fluorescence of EGFP-paxillin in a living cell was observed by confocal laser scanning microscopy (model 510 using the supplied software) at an interval of 1 min. Each figure of microscopic analysis shows representative results that were independently confirmed by both mass cell cultures of the original BOSC virus-infected cells and with several independent cell clones isolated from the original infected cells.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Prominent Changes in the Levels of Tyrosine Phosphorylation of Paxillin and p130^Cas during EMT of NMuMG Cells-- In order to explore the cellular events during integrin activation and cell migration, we employed an in vitro EMT system of NMuMG cells (31) as an experimental model. In this system, treatment of cells with TGF-1 causes trans-differentiation of NMuMG cells from the epithelial phenotype into the mesenchymal phenotype, which is accompanied by a change from cadherin-mediated cell-cell adhesion to integrin-mediated cell-substratum adhesion. To identify the major factors involved in this model of EMT, we first analyzed a series of proteins involved in cell adhesion. It has already been shown that expression levels of cadherin and integrin are reciprocally changed and that expressions of fibronectin and vimentin are increased during EMT (31). We confirmed these changes with our cell culture system (data not shown). We also examined several other proteins related to cadherin and integrin, including desmoglein, -, -, and -catenins, ERM (ezrin, radixin and moesin), tensin, talin, p130^Cas, Fak, Pyk2/Cak-, p120^Cas, vinculin, and paxillin ; and we found that the expression levels of these proteins remained essentially unchanged during EMT (data not shown). Rodent cells express two isoforms of paxillin,  and  (23). The  isoform of paxillin was found to be much less abundant than the  isoform, comprising less than 10% of the total paxillin expressed in NMuMG cells, but its expression was augmented 2-3-fold during EMT (data not shown).

It is well documented that integrin activation is accompanied by cellular protein tyrosine phosphorylation. We therefore examined cellular protein tyrosine phosphorylation during EMT of NMuMG cells. As shown in Fig. 1A, tyrosine phosphorylation on several proteins of about 120-140 and about 70-80 kDa were highly augmented during EMT. We identified these bands as paxillin because immunodepletion of paxillin from the cell lysates diminished the bands almost completely (Fig. 1B). Blotting of anti-paxillin immunoprecipitates with the anti-phosphotyrosine antibody 4G10 revealed that paxillin was tyrosine-phosphorylated at a low level in the epithelial phenotype and became highly tyrosine-phosphorylated in the mesenchymal phenotype (Fig. 1C). No monospecific antibodies for the  isoform of paxillin were available, but antibodies that recognize only the  isoform exist, enabling determination of the position of the  isoform on the blotting filters (23). An alignment of tyrosine-phosphorylated bands and the isoform  band indicated that most of the tyrosine phosphorylation occurred on the  isoform (data not shown). On the other hand, in order to address the 120-140-kDa protein bands, we examined p130^Cas and p125^Fak, both of which are implicated in cell migratory activity (37-40). We found that tyrosine phosphorylation of p130^Cas was also highly augmented during EMT, whereas tyrosine phosphorylation of Fak was almost unchanged (Fig. 1C).

View larger version (31K): [in this window] [in a new window]   Fig. 1.   Change in tyrosine phosphorylation of paxillin is prominent during in vitro EMT or cell migration of NMuMG cells. A and B, anti-phosphotyrosine immunoblotting analysis of NMuMG cells of the epithelial phenotype (TGF) or the mesenchymal phenotype (+TGF, 2 ng/ml for 48 h at 37 °C). A, RIPA cell lysates (20 µg each) were prepared, separated on SDS-PAGE, and subjected to immunoblotting analysis as described under "Experimental Procedures." B, same analysis performed with RIPA cell lysates that were pre-depleted of paxillin using the mouse monoclonal anti-paxillin antibody. Signals were generated using the generic rabbit polyclonal anti-phosphotyrosine antibody. Molecular sizes determined using marker proteins are shown on the left. C, levels of tyrosine phosphorylation of paxillin, p130Cas, and Fak from NMuMG cells were analyzed by immunoprecipitating (i.p.) each protein from crude cell lysates using the appropriate antibody, followed by SDS-PAGE and sequential immunoblotting using an anti-phosphotyrosine antibody (4G10) and the protein-specific antibodies. D, tyrosine phosphorylation of paxillin and p130Cas from NMuMG cells, cultured under confluent or sparse conditions, was determined as described above (C) and under "Experimental Procedures."

When epithelial cells are cultured under sparse conditions, their integrins become activated, and the cells migrate. Indeed, we were able to document by video recording that these cells migrate as actively as cells of the mesenchymal phenotype (data not shown). We thus analyzed epithelial cells under sparse culture conditions, and we found that paxillin and p130^Cas were both highly tyrosine-phosphorylated (Fig. 1D).

Identification of Tyrosine Phosphorylation Sites in Paxillin -- We then focused on tyrosine phosphorylation of the  isoform of paxillin, one of the most prominent changes upon integrin activation during EMT or cell migration. We made a series of EGFP-tagged paxillin  cDNAs (23). EGFP-paxillin has been expressed in several cell lines, including NMuMG, with no detectable differences from endogenous paxillin in subcellular localization and biochemical properties, other than the expected increase in molecular mass (see Ref. 23; also see Figs. 2B and 7B). Six tyrosine residues have been suggested as potential phosphorylation sites in the  isoform of paxillin (33). To study the role of these specific sites, a series of mutant cDNAs were constructed in which each tyrosine residue was changed to a phenylalanine, either singly or in various combinations (see YF mutants, Fig. 2A). Each cDNA was packaged using a BOSC23-derived retrovirus and then expressed in NMuMG cells to examine its tyrosine phosphorylation. As shown in Fig. 2B, disruption of all six residues eliminated tyrosine phosphorylation almost completely in both the epithelial and the mesenchymal cells. However, we found that disruption of the amino-terminal four tyrosine residues was sufficient for the elimination of tyrosine phosphorylation, and disruption of the two carboxyl-terminal tyrosine residues did not affect phosphorylation (Fig. 2B and data not shown). We also found that mutations in any one of these four amino-terminal residues reduced the phosphorylation, although none of these individual substitutions could completely eliminate the phosphorylation (Fig. 2B).

View larger version (53K): [in this window] [in a new window]   Fig. 2.   Identification of the major tyrosine phosphorylation sites of paxillin  during in vitro EMT of NMuMG cells. A, schematic drawing of the proposed tyrosine phosphorylation sites. Tyrosine residues mutated to phenylalanine in each mutant cDNA of paxillin  are shown. B, detection of tyrosine phosphorylation of each paxillin  mutant expressed in NMuMG cells under conditions that induce either the epithelial (TGF) or mesenchymal (+TGF) phenotype. Each paxillin cDNA fused to EGFP was expressed in NMuMG cells using the retrovirus infection system as described under "Experimental Procedures." Both endogenous (closed arrowhead) and exogenous (open arrowhead) paxillin proteins were immunoprecipitated using monoclonal anti-paxillin antibody from RIPA cell lysates, separated on SDS-PAGE, and subjected to sequential immunoblotting analysis using an anti-phosphotyrosine (4G10) and anti-paxillin 199-217 (pax) antibodies.

Amino acid sequences following these tyrosine residues are Y³¹SYP, Y⁴⁰QEI, Y¹¹⁸SFP, and Y¹⁸¹GVP. Each of these sites has been implicated in the creation of binding sites for several SH2-containing proteins as follows: Y³¹SYP for Crk, Y⁴⁰QEI for Src, Y¹¹⁸SFP for Crk, and Y¹⁸¹GVP for Crk and phospholipase C  (26, 33). The Y³¹SYP and the Y¹¹⁸SFP have a common motif of YSP (where  is aromatic amino acid). The level of tyrosine phosphorylation on the 2X mutant (for which residues 31 and 118 were both mutated) was largely reduced and little change was observed during EMT (Fig. 2B). On the other hand, the 2Y mutant (for which residues 31 and 118 remained unchanged but the other tyrosine residues were all mutated) was well phosphorylated, and the phosphorylation was further increased during EMT (Fig. 2B). We also examined other possible combinations of mutations in these four residues (data not shown), and we collectively concluded that the major tyrosine phosphorylation sites in the  isoform of paxillin are residues 31, 40, 118, and 181 in NMuMG cells, and the phosphorylation of residues 31 and 118 is highly inducible during EMT.

Phosphorylation of tyrosine 31 and 118 and its augmentation during EMT were confirmed using phosphorylation site-specific antibodies that selectively recognize paxillin when phosphorylated at these sites. The specificity of each of these antibodies was confirmed by immunoblotting analysis using the YF mutants of paxillin  (Fig. 3A). A comparable increase in paxillin tyrosine phosphorylation was also confirmed in epithelial cells cultured under sparse conditions (Fig. 3B). We thus focused on the role of phosphorylation of tyrosines 31 and 118 in the remainder of this study. A similar analysis of phosphorylation on tyrosine 40 and 181 would also be interesting; however, antibodies that selectively detect phosphorylation at these sites were not yet available.

View larger version (41K): [in this window] [in a new window]   Fig. 3.   Paxillin tyrosine phosphorylation detected using phosphorylation site-specific antibodies. A, assessment of the specificity of the phosphorylation site-specific antibodies. NMuMG cells expressing wild-type EGFP-paxillin  or the corresponding YF mutants were treated with TGF- to induce differentiation to the mesenchymal phenotype. RIPA cell lysates (20 µg each) then were prepared, separated on SDS-PAGE, and subjected to immunoblotting analysis using anti-paxillin, anti-phosphotyrosine (4G10), or phosphorylation site-specific (targeting tyrosine phosphorylation of paxillin at residue 31 (pY31) or 118 (pY118)) antibodies. B, increased phosphorylation on tyrosine 31 and 118 of paxillin  during EMT and cell migration. Cell lysates were prepared from confluent or sparse cells as described under "Experimental Procedures" and subjected to immunoprecipitation with anti-paxillin antibody, followed by immunoblotting analysis using the phosphorylation site-specific antibodies to paxillin described above.

Paxillin Localized to the Cell Periphery and Focal Adhesions Is Tyrosine-phosphorylated-- Formation of focal adhesions, as well as actin stress fibers, seemed to be tightly coupled with the cell migratory phenotype in NMuMG cells (Fig. 4, A and B). Sedentary NMuMG epithelial cells at confluence possessed only marginal amounts of actin stress fibers and paxillin-containing focal adhesions. Paxillin in these sedentary epithelial cells was diffusely distributed throughout the cytoplasm. On the other hand, NMuMG cells differentiated to the mesenchymal phenotype by TGF- treatment possessed significant amounts of both actin stress fibers and paxillin-containing focal adhesions (Fig. 4A).

View larger version (54K): [in this window] [in a new window]   Fig. 4.   Intracellular localization of paxillin and its tyrosine phosphorylation. A, loss of actin stress fibers and paxillin-containing focal adhesions in confluent NMuMG cells with epithelial phenotype. Cells (8 × 105 in a 9-cm culture dish) treated either with or without TGF were fixed, labeled, and subjected to microscopic analysis as described under "Experimental Procedures." F-actin was visualized with Texas Red-conjugated phalloidin. Paxillin was visualized with anti-paxillin antibody followed by Cy2-conjugated anti-mouse IgG. Differential interference contrast (DIC) images of the same field were shown in the right column. B, colocalization of paxillin with F-actin in motile epithelial cells. Cells were induced to migrate by manually scratching confluent cell cultures and fixed 4 h later as described under "Experimental Procedures." Paxillin and F-actin were visualized as above, and their merged image is shown. C, detection of tyrosine phosphorylation of paxillin localized at the cell periphery and in focal adhesions. Migrating epithelial cells, prepared as above, were labeled with anti-phosphotyrosine antibody (4G10) or phosphorylation site-specific antibodies against the tyrosine 31 (pY31) and tyrosine 118 (pY118) (shown in red) as described under "Experimental Procedures." Endogenous paxillin is shown in green. Each right column is the merged image of the left and middle columns. Scale bars represent 20 µm.

To examine the role of tyrosine phosphorylation of paxillin during cell migration, cells were induced to undergo directed migration by manually scratching confluent cell cultures, as commonly performed in a wound healing assay. In epithelial cells induced to migrate, paxillin exhibits three distinct subcellular localizations (Fig. 4B and also see Fig. 6A). First, paxillin localized laterally along the cell periphery apparently colocalized with a circumferential meshwork of actin filaments. Paxillin also localized to focal adhesions as macroaggregates, connecting to actin stress fibers and, finally, was diffusely distributed in the cytoplasm. Paxillin-containing focal adhesions were not formed at the cell periphery but formed at least several micrometers away. To examine which fraction of paxillin is tyrosine-phosphorylated, cells were double immunolabeled with anti-paxillin antibody and anti-phosphotyrosine antibodies. As shown in Fig. 4C, no significant labeling with phosphotyrosine antibody was detected on paxillin diffusely distributed in the cytoplasm, whereas paxillin localized at the cell periphery and at focal adhesions seemed to be tyrosine-phosphorylated. By using the phosphorylation site-specific antibodies against tyrosine 31 or tyrosine 118 of paxillin, phosphorylation on these residues was confirmed in situ, with signals again localized to paxillin molecules localized at the cell periphery and at focal adhesions (Fig. 4C).

Mutations in the Tyrosine Phosphorylation Sites of Paxillin  Affect Paxillin-containing Focal Adhesion Formation and Actin Cytoskeletal Organization in Motile Cells--- We next examined the effects of mutations in the tyrosine phosphorylation sites of paxillin  in motile cells. Our results described above indicated that paxillin molecules that were tyrosine-phosphorylated seemed to be colocalized with F-actin structures. We thus examined the actin-cytoskeletal architecture in cells expressing the paxillin YF mutants.

Cells expressing wild-type and mutant paxillin and grown to confluence did not contain detectable amounts of actin stress fibers or paxillin-containing focal adhesions (Fig. 5, 0 h). However, unlike wild-type cells, we found that cells expressing the 2X mutant formed paxillin-containing macroaggregate structures at the cell periphery (2-4 h after scratching, Figs. 5 and 6B). These macroaggregates connected to actin stress fibers (Fig. 5). In addition, other focal adhesion proteins, including vinculin, colocalized with these macroaggregates (data not shown), indicating that these macroaggregates correspond to focal adhesions. By merging the fluorescence images from the EGFP-tagged 2X mutant with that of the cell body, we found that these paxillin-containing focal adhesion structures did not correspond to membrane protrusions of filopodia (see Fig. 7A). However, note that cells expressing the 2X mutant did exhibit several filopodia-like structures at their leading edges (Fig. 7A), which was seldom seen with normal NMuMG cells. On the other hand, circumferential actin fibers and paxillin that normally localized laterally along the cell periphery were almost undetectable in the 2X mutant cells. Moreover, sizes of the paxillin-containing focal adhesions were noticeably larger than those observed in the wild-type cells. In addition, actin stress fibers appeared to be bundled to a much greater extent as compared with the wild-type cells. Indeed, in these cells, some of the stress fibers crisscrossed with each other, and in other instances, two actin fibers were sometimes connected to single focal adhesion. Both phenomena were seldom seen in wild-type cells.

View larger version (89K): [in this window] [in a new window]   Fig. 5.   Mutations of paxillin tyrosine phosphorylation sites alter intracellular dynamics of paxillin and actin cytoskeletal organization in motile NMuMG epithelial cells. Cells were induced to migrate, were fixed, and were immunolabeled with Texas Red-conjugated phalloidin as in Fig. 4. EGFP-paxillin  was visualized by fluorescence from the EGFP tag. The merged images of the EGFP tag (green) and F-actin (red) are shown. Images showing cells expressing EGFP-paxillin  (left column), 2X mutant (middle column), and 2Y mutant (right column) were generated as described under "Experimental Procedures." Time after scratching is shown on the left. At about 16 h after scratching (16 h), cells migrating from each side of the wound began making contact with each other. Scale bar represents 20 µm.

View larger version (76K): [in this window] [in a new window]   Fig. 6.   Intracellular localization of EGFP-tagged paxillin  during cell migration of NMuMG epithelial cells. Fluorescence from the EGFP tag of wild-type (A), 2X mutant (B), and 2Y mutant (C) of paxillin  in migrating NMuMG epithelial cells was traced using time-lapse fluorescent microscopy as described under "Experimental Procedures." The time, in minutes, elapsed following initiation of the trace is shown in each figure. Arrowheads indicate the paxillin-containing macroaggregates formed at cell periphery. Scale bar represents 20 µm.

View larger version (64K): [in this window] [in a new window]   Fig. 7.   Altered subcellular localization of YF mutants of EGFP-paxillin  in motile NMuMG cells. A, merged images of differential interference contrast images (gray) with EGFP-paxillin  visualized by fluorescence from the EGFP tag (green). wt, wild type. B, merged images of EGFP-paxillin  (green) and paxillin generated by immunolabeling with anti-paxillin antibody coupled with Cy3-conjugated anti-mouse IgG (red). C, merged images of EGFP-paxillin  (green) and F-actin visualized with Texas Red-X phalloidin (red) in the mesenchymal phenotype of NMuMG cells. Scale bars represent 20 µm.

The 2Y mutant also exhibited alterations that were substantially different from those seen with the 2X mutant (Fig. 5). Similar to the wild-type cells, most of the paxillin-containing focal adhesions, each connected to actin stress fibers, were formed several micrometers away from the cell periphery. Lateral localization of paxillin along the cell periphery was also observed. However, orientations of actin stress fibers were much more chaotic than in the 2X cells; most of them crisscrossed with each other, and each focal adhesion structure was often connected to multiple stress fibers, which again seemed to be highly bundled. Moreover, disorganized membrane ruffles appeared to occur in random directions (see Figs. 5, 6C, and 7A). Thus, expression of the 2Y mutant appears to disrupt aspects of normal cell polarity.

At about 16 h after the scratching, cells migrating from each side of the wound began to make contact with each other. At the initial point of these cell contacts, wild-type EGFP-paxillin almost disappeared from cell-cell junctions, coinciding with the disappearance of most of the paxillin-containing focal adhesions. On the other hand, cells expressing the 2X or the 2Y mutants still formed paxillin-containing focal adhesions, which were connected to actin stress fibers (Fig. 5).

Intracellular Dynamics of Paxillin during Cell Migration-- We also traced intracellular dynamics of EGFP-paxillin  using a fluorescence microscope with time-lapse video recording capacities. By using this technique, we found that as a cell body extended forward, most paxillin-containing focal adhesions seem to be formed as a consequence of the accumulation of paxillin molecules that are formerly localized laterally along the cell periphery (Fig. 6A). Video recording of the initial phases of cell migration (0-30 min after generating the scratch) also supported this notion (data not shown). We found similar results with cells expressing the 2Y mutant (Fig. 6C). In contrast, cells expressing the 2X mutant formed paxillin-containing macroaggregates directly at or very near to the cell periphery without prior localization laterally along the cell periphery (Fig. 6B).

It is interesting to note that the subcellular localization of the endogenous paxillin appeared to exhibit alterations that were similar to those observed for the 2X and 2Y mutant (Fig. 7B). Indeed, images of cellular paxillin (both endogenous and EGFP-paxillin), as detected by anti-paxillin antibody labeling, merged almost completely with that of the fluorescence obtained from the EGFP tag.

Finally, NMuMG cells with the mesenchymal phenotype were also examined (Fig. 7C). Most of the paxillin-containing focal adhesions connected to actin stress fibers were formed at least several micrometers away from the cell periphery, as seen with the epithelial phenotype. On the other hand, as in the case of the 2X mutant in the epithelial phenotype, mesenchymal cells expressing the 2X mutant formed paxillin-containing focal adhesions at or very close to the cell periphery. Similarly, actin stress fibers in mesenchymal cells expressing the 2Y mutant exhibited a more chaotic organization and orientation, whereas the paxillin-containing focal adhesions formed several micrometers away from the cell periphery.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Regulation of focal adhesion formation and actin-based cytoskeletal organization are crucial events for facilitating cell migration. In this paper, we showed that tyrosine phosphorylation of paxillin  plays an important role in these signal transduction events. We demonstrated that tyrosine phosphorylation events on multiple sites of paxillin  play a critical role in the formation of paxillin-containing focal adhesions, as well as actin cytoskeletal organization in motile cells.

Multiple Sites of Tyrosine Phosphorylation Appear to Be Essential for the Normal Function of Paxillin -- Our results demonstrate that four tyrosines (tyrosines 31, 40, 118, and 181) serve as major phosphorylation sites of paxillin  in NMuMG cells. In addition, phosphorylation on two of these sites, tyrosines 31 and 118, is inducible upon integrin activation during EMT or cell migration. We primarily characterized the 2X and 2Y mutants in this paper, and we showed that both mutants affect paxillin-containing focal adhesion formation and actin cytoskeletal organization. We also made a YF mutant (2Y') possessing Tyr to Phe substitutions at Tyr-40 and Tyr-181 (Tyr-31, Tyr-118, Tyr-434, and Tyr-488 remained intact), and we confirmed that this mutation also evoked essentially the same phenotypes as described with the 2Y mutation in this paper (data not shown). We have also made a series of mutants in which each of the four tyrosine residues was mutated singly or in combinations, and we found that they affected actin cytoskeletal organization and/or other functions of paxillin.² Moreover, we also made the 2X, 2Y, and 2Y' mutants without the EGFP tag and confirmed the same effects (data not shown). Our results thus collectively suggested that phosphorylation on these four tyrosines is necessary for the proper function of paxillin .

The 2X and 2Y mutations cause dramatic alterations in the intracellular behavior of paxillin molecules in motile NMuMG cells, which was accompanied by altered organization of both focal adhesions and F-actin. Despite these dramatic changes, expression of the 2X or 2Y mutants did not appreciably alter expression levels of several other adhesion marker proteins, such as E-cadherin, integrin, fibronectin, and vimentin.² Moreover, since the 2X and 2Y mutants also affected the positioning of focal adhesions in cells exhibiting the mesenchymal phenotype, these YF mutants did not simply induce EMT in these cells.

The hypothesis that individual phosphorylation sites may contribute to different functions of paxillin in the cell is suggested by the fact that cells expressing the 2X and 2Y mutants exhibited very different phenotypes. Indeed, it is likely that different signaling molecules may bind to these tyrosine phosphorylation sites, as expected from the different primary structures surrounding each phosphorylation site. Several proteins bearing SH2 domains, including the CrkI/CrkII and CrkL proteins, have been implicated in binding to tyrosine phosphorylation sites of paxillin (33, 41-43). We confirmed that the SH2 domain of CrkI/CrkII can bind in vitro to tyrosine-phosphorylated paxillin  prepared from NMuMG cells.² However, CrkI and CrkII proteins did not form a stable complex with paxillin in NMuMG cells in vivo, whereas Crk II associates stably with several other tyrosine-phosphorylated proteins, such as p130^Cas, c-Abl, Cbl, and Fak.² This observation is consistent with several other reports using other types of cells (42, 44). We also found no evidence for the stable complex formation of CrkL with paxillin in NMuMG cells.² Further analysis is required to determine which signaling proteins bind to and act downstream of the tyrosine phosphorylation sites of paxillin  in NMuMG cells.

Intracellular Dynamics of Paxillin Localization to the Cell Periphery and Formation of Focal Adhesion Aggregates-- Paxillin is localized at the cell periphery, at focal adhesions, and in the cytoplasm of normal motile NMuMG cells exhibiting the epithelial phenotype. Interestingly, paxillin-containing focal adhesions are not formed at the cell periphery in NMuMG cells but rather are formed at least several micrometers away from the cell periphery. Consequently, the lamellipodium structure that associates with areas of membrane ruffling appears to form at the leading edge of the motile cells, as has been previously described (45, 46). Our time course study indicates that the peripheral localization of paxillin is observed from the early phase of cell migration, whereas paxillin-containing focal adhesions are formed later. Moreover, a fluorescence microscope-coupled time-lapse video recording implies that most of paxillin-containing focal adhesions are formed as a consequence of the accumulation of paxillin molecules that were formerly localized laterally along the cell periphery. Therefore, most paxillin molecules seem to be recruited first to the cell periphery, then aggregate with each other, and finally form focal adhesions that are at that point connected to actin stress fibers. This process seems to be tightly coupled with forward extension of the cell body.

NMuMG cells exhibiting the mesenchymal phenotype, on the other hand, showed a different subcellular localization of paxillin in that these cells possess lower levels of peripheral paxillin. However, focal adhesions are still formed away from the cell periphery, allowing the lamellipodium structure to be formed at the leading edge of these cells. In other types of cells, such as NIH3T3 or 3Y1 fibroblasts, paxillin exists at cell periphery, but most of the paxillin is found in focal adhesion-like structures that are frequently connected to actin stress fibers. Furthermore, expression of the 2X mutant in 3Y1 cells did not cause alteration of focal adhesion formation nor of the actin cytoskeleton. Finally, almost all paxillin-containing focal adhesions disappear in NMuMG epithelial cells when they are grown to confluence, whereas substantial levels of these structures remain in cells exhibiting the mesenchymal phenotype. Therefore, the intracellular dynamics of paxillin seem to be regulated differently among different cell types. In NMuMG cells, paxillin regulation occurs in a cell density and cell phenotype-dependent manner.

Tyrosine Phosphorylation of Paxillin  and Temporospatial Regulation of Focal Adhesion Formation and Actin Cytoskeletal Organization-- We showed in situ that paxillin molecules, which are localized at cell periphery or at focal adhesions, are tyrosine-phosphorylated. In wild-type NMuMG cells or cells expressing the 2X or 2Y mutants, focal adhesions disappeared in cells of the epithelial phenotype when they are grown to confluence. However, unlike in wild-type cells, in cells expressing the 2X mutant, most of the paxillin-containing focal adhesions formed at the cell periphery without prior localization laterally along the cell periphery. Thus, the 2X mutation may facilitate aggregation of paxillin molecules or act to inhibit paxillin translocation to sites that are localized laterally along the cell periphery. On the other hand, cells expressing the 2Y mutant exhibited disorganized actin stress fibers accompanied by the loss of cell polarity, whereas focal adhesions formed in a manner similar to that observed in the wild-type cells. We also examined migration speeds as assessed by two-dimensional free locomotion on culture dishes, and we found that expression of the 2X or 2Y mutants did not significantly alter the migratory activity as compared with the expression of wild-type EGFP-paxillin  in NMuMG cells.²

Several tyrosine kinases have been implicated in paxillin phosphorylation (24-26). Formation of a complex between paxillin and Fak appears to be required for maximal phosphorylation in response to cell adhesion in fibroblasts (47), although Fak may not be the sole tyrosine kinase that phosphorylates paxillin (37) or Fak may rather act to direct paxillin phosphorylation by recruiting Src family kinases (24, 47). Pyk2, another tyrosine kinase acting in integrin signaling, has also been shown to associate with paxillin (48, 49). However, it is not yet established which kinase phosphorylates paxillin. Our analysis revealed that Fak is tyrosine-phosphorylated and thus activated in NMuMG cells of both epithelial and mesenchymal phenotypes (see Fig. 2).² Likewise, a fraction of c-Src also appears to be constantly activated.² On the other hand, we found that Pyk2 becomes highly phosphorylated and thus activated during EMT.² Pyk2 may therefore also be responsible for the phosphorylation of Tyr-31 and Tyr-118 of paxillin. Identification of the kinase(s) that phosphorylates paxillin during cell migration will contribute to the understanding of the regulations of cytoskeletal organization and cell polarity. It also remains to be established whether integrins or distinct transmembrane proteins serve as surface receptors regulating tyrosine phosphorylation of paxillin that localizes laterally along the cell periphery.

Possible Relationship between Rho Family GTPases and Downstream Signaling of Tyrosine-phosphorylated Paxillin-- Many questions regarding the dynamics of paxillin localization during cell migration remain to be answered. For example, (i) how is paxillin colocalized with circumferential actin filaments, laterally along the cell periphery, without forming macroaggregates? (ii) What regulates the timing and positioning involved in the formation of paxillin-containing focal adhesions? (iii) How is the connection of actin stress fibers to paxillin-containing focal adhesions then regulated, and what role(s) does it play? Each of these questions requires more insight into the signal transduction pathways that regulate paxillin function. It has been well documented that actin polymerization and focal adhesion assembly are two distinct downstream effects of Rho family GTPases (50). It has been also demonstrated that the activity of different Rho family GTPases determines whether phosphotyrosine-containing proteins, including vinculin and paxillin, form focal complexes along the cell periphery or form focal adhesions connected to stress fibers (45, 50). A number of studies attempting to elucidate the precise molecular mechanisms for these processes have been reported. For example, one line of evidence suggests that Cdc42 acts to restrict Rac activity through the generation of a polarizing signal, thereby preventing the Rac protein from otherwise initiating the protrusion of lamellipodia around the cell periphery (46). The Arp2/3 complex, one of the effectors of Cdc42, has been shown to regulate the nucleation of linear filamentous actin polymerization and the formation of branching networks of actin filaments (51-55). Possible intercommunication between paxillin and Cdc42 and/or other Rho family GTPases has been implicated based on the finding that paxillin can indirectly associate with p21 GTPase-activated kinase 3 (PAK3) and the guanine nucleotide exchange factor, PIX, although tyrosine phosphorylation of paxillin is not required for this association (56). We observed that NMuMG cells expressing the 2X mutant can form filopodia structures, whereas the 2Y mutation results in a substantial loss in cell polarity. Filopodia formation and the generation of a polarizing signal are both regulated by Cdc42 activity (50, 57). Thus our observations are consistent with the possibility that intercommunication between paxillin and Cdc42 plays an important role in regulating cytoskeletal dynamics.

	ACKNOWLEDGEMENTS

We thank Manami Hiraishi, Mihoko Sato, and Asako Tsubouchi for technical assistance and Mayumi Yoneda for secretarial work. We also thank to Warren Pear and David Baltimore for BOSC23 cells; Hermut Land for pBabe vector; Hisamaru Hirai for anti-p130^Cas antibody; and Heidi Greulich for critical reading of the manuscript.

	FOOTNOTES

^* This work was supported in part by Japan Science and Technology Corp., grants-in-aid from the Ministry of Education, Science, Sports and Culture of Japan, grants from the Mitsubishi Foundation, the Ciba-Geigy Foundation (Japan) for the Promotion of Science, the Takeda Medical Foundation, the Mochida Memorial Foundation for Medical and Pharmaceutical Research, and the Novartis Foundation for the Promotion of Science.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

^** To whom correspondence should be addressed: Dept. of Molecular Biology, Osaka Bioscience Institute, 6-2-4 Furuedai, Suita, Osaka 565-0874, Japan. Tel.: 81-6-6872-4814; Fax: 81-6-6871-6686; E-mail: sabe@obi.or.jp.

Published, JBC Papers in Press, May 23, 2000, DOI 10.1074/jbc.M000679200

² H. Yano, K. Nakamura, H. Uchida, S. Hashimoto, and H. Sabe, unpublished results.

	ABBREVIATIONS

The abbreviations used are: EMT, epithelial-mesenchymal trans-differentiation; EGFP, enhanced green fluorescent protein; Fak, focal adhesion kinase; SH2, src homology 2; PAGE, polyacrylamide gel electrophoresis; Ab, antibody.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				S. R. Wylie and P. D. Chantler Myosin IIC: A Third Molecular Motor Driving Neuronal Dynamics Mol. Biol. Cell, September 1, 2008; 19(9): 3956 - 3968. [Abstract] [Full Text] [PDF]

				M. Handfield, H.V. Baker, and R.J. Lamont Beyond Good and Evil in the Oral Cavity: Insights into Host-Microbe Relationships Derived from Transcriptional Profiling of Gingival Cells J. Dent. Res., March 1, 2008; 87(3): 203 - 223. [Abstract] [Full Text] [PDF]

C. Badowski, G. Pawlak, A. Grichine, A. Chabadel, C. Oddou, P. Jurdic, M. Pfaff, C. Albiges-Rizo, and M. R. Block Paxillin Phosphorylation Controls Invadopodia/Podosomes Spatiotemporal Organization Mol. Biol. Cell, February 1, 2008; 19(2): 633 - 645.