INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Integrins are a family of adhesion receptors, each consisting of one  and one  subunit. Among the known integrins, 12 contain the 1 subunit, and ligands for this subfamily include collagens, laminins (LNs), and fibronectin (FN).

Integrin subunit 1, as well as all other integrin subunits except for 4, consists of a large extracellular domain, a single transmembrane stretch, and a short intracellular cytoplasmic domain. Although devoid of an intrinsic enzymatic activity, the cytoplasmic domain regulates the conformation and ligand binding activity of the extracellular domain (so-called "inside-out signaling"), as well as mediates interactions with the cytoskeleton and the transduction of intracellular signals ("outside-in" signaling). Specificity for this range of dynamic signaling events has been validated in many systems by using truncated and mutated integrin receptors (15, 16, 23, 26, 27, 29, 32, 33, 51).

Ligand binding to integrins triggers the assembly of a large number of cytoplasmic and transmembrane molecules into discrete regions referred to as focal adhesions (FAs) (5, 10). These complexes serve both as structural anchorage sites that connect the extracellular matrix with the intracellular actin cytoskeleton and as signaling complexes, which initiate signaling pathways in the cell cytoplasm and nucleus. The signaling molecules found in FAs include tyrosine kinases, serine/threonine kinases, phospholipid kinases, phosphatases, Ras superfamily proteins, and various adapter proteins (e.g., those containing SH2 and SH3 domains) (11, 28, 42, 49).

Although much has been learned about integrin-mediated signaling in recent years, the specific mechanisms by which the initial interactions between integrins and other FA proteins are controlled and how these interactions dictate the hierarchies of signaling pathways remain largely unresolved.

A critical component of integrin-FA signaling involves the formation of a complex between two soluble tyrosine kinases, FA kinase (FAK) and Src. The FAK-Src complex (7) has been shown to activate Ras, which via phosphotidylinositol (PI) 3-kinaseAkt/protein kinase B and the extracellular signal-regulated protein kinases ERK1 and -2 mediates cell survival and proliferation signals (20, 43). FAK-Src signaling also leads to phosphorylation of p130^CAS, which positively regulates cell migration (8, 21). Although the activation of FAK and Src is a topic of much investigation, signaling from the FAK-Src complex is still poorly understood. Integrin ligation somehow leads to autophosphorylation of FAK at tyrosine 397, thereby creating a binding site for Src via its SH2 domain (40). This in turn, allows Src to further activate the complex by phosphorylating additional tyrosines in the FAK molecule, including tyrosines 576, 577, and 861 (6). Although this knowledge provides a basic understanding of receptor-FAK-Src signaling, several essential pieces of information are still required to fully understand the activation process. Major unresolved questions include the molecular events that trigger the autophosphorylation of FAK at tyrosine 397, the importance of positive and negative regulatory phosphorylation sites on Src for its binding to FAK, and finally the identity and role of the phosphatase or phosphatases responsible for dephosphorylating proteins involved in cell adhesion signaling.

To elucidate integrin-mediated signaling events, we have studied the role of the potential phosphorylation sites in the cytoplasmic domain of integrin subunit 1A. Previously, we demonstrated that the T777A, Y783F, S785A, and Y795F single substitutions had little effect on the regulation of extracellular conformation or intracellular signaling, indicating that phosphorylation of these residues is not essential for these integrin signaling events (38, 51). In contrast, a T788/789A double mutation resulted in a dramatic shift toward the inactive, non-ligand binding state (51). In addition, the Y783/795F double mutation markedly affected integrin 1A function, as indicated by the increased number of FAs, altered cytoskeletal architecture, and decreased cell migration (38). In the present study, we show that the 1A_Y783/795F mutation impairs 1A-mediated FAK autophosphorylation and activation. Moreover, this defect results in reduced tyrosine phosphorylation of paxillin and tensin, while the overall tyrosine phosphorylation of a third FAK-Src complex substrate, p130^CAS, is unaffected.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Proteins, peptides, and antibodies. FN and vitronectin (VN) were purified from human plasma as previously described (46, 53). LN-1 was obtained from Gibco. GRGDS peptide was purchased from Bachem Feinchemikalien AG. Plasmids encoding the glutathione S-transferase (GST) protein fused to the SH2 domains of CrkII and Src, respectively, were provided by A. Sorokin (Milwaukee, Wis.) and S. Courtneidge (Redwood City, Calif.), and the fusion proteins were purified as described previously (22).

Cell culture. The cell lines GD25, GD25-1A, GD25-1A_Y783F, GD25-1A_Y795F, and GD25-1A_Y783/795F have been described previously (13, 38, 51, 52). The GD25 cells lack the 1 family of integrins due to a disrupted 1 gene. The GD25-1A and the GD25-1A mutant cell lines were derived from GD25 upon stable transfection with cDNAs encoding the wild-type and mutated murine integrin subunit 1A, respectively. Clones with similar levels of 1 integrins expressed on their surface, as determined by fluorescence-activated cell sorter analysis (38, 51), were chosen for these studies. In addition, expression levels of 1 integrins of these cell lines were verified throughout the duration of this study.

Cell spreading assay. The wells of 96-well microtiter plates were coated with extracellular matrix (ECM) proteins (2 µg of VN, 10 µg of FN, or 10 µg of LN-1 per ml) in DMEM for 1 h at 37°C. The wells were washed twice with phosphate-buffered saline (PBS) and blocked with 1% heat-treated bovine serum albumin (BSA) in PBS for 1 h at 37°C. After being washed with PBS, 10⁴ cells suspended in DMEM were added to each well and were allowed to attach for 30, 60, 120, or 240 min at 37°C. In experiments in which GRGDS peptide (0.5 mg/ml) was used, the cells were preincubated with the peptide for 15 min. Unattached cells were removed by washing with PBS twice, and the remaining cells were fixed in 96% ethanol for 10 min and stained with 0.1% crystal violet in H₂O for 30 min. The percentage of the cells in three microscopic fields (50 to 100 cells/field) that had a spread morphology was calculated.

Phosphoprotein blotting assay. Petri dishes were coated with either VN (2 µg/ml), FN (10 µg/ml), LN-1 (10 µg/ml), or MAb Ha2/5 (25 µg/ml) for 1 h at 37°C and blocked with 1 mg of heat-treated BSA per ml for 1 h at 37°C. Cells (5 × 10⁶) suspended in DMEM (attachment to VN, LN-1, or Ha2/5) or DMEM plus 0.5 mg of GRGDS peptide per ml (1-dependent attachment to FN) were seeded in each dish and allowed to attach for the indicated periods of time. As a negative control, cells were kept in suspension at 37°C. Subsequently, the cells were lysed with ice-cold radioimmunoprecipitation assay buffer (150 mM NaCl, 20 mM Tris-HCl [pH 7.5], 1% NP-40, 1% sodium deoxycholate, and 0.1% sodium dodecyl sulfate [SDS]) containing 50 mM NaF, 30 mM Na₄P₂O₇, 200 µM Na₃VO₄, 2 mM phenylmethylsulfonyl fluoride (PMSF), 2 mM N-ethylmaleimide (NEM), 2 mM EDTA, and 1 µg of pepstatin A per ml. The lysates were aspirated three times through an injection needle and centrifuged at 21,000 × g for 10 min. The supernatant was precleared with protein A-Sepharose for 1 h at 4°C by end-over-end rotation. The precleared supernatants were incubated with 1 µg of MAb (directed against FAK, tensin, paxillin, or p130^CAS) for 2 h at 4°C. The antibodies were precipitated with rabbit anti-mouse IgG and protein A-Sepharose. The pellet was washed three times with lysis buffer, and the immunoprecipitated proteins were separated on an SDS-polyacrylamide gel electrophoresis (PAGE) gel followed by transfer to nitrocellulose membrane. The membranes were blocked with 1% BSA in Tris-buffered saline-Tween (150 mM NaCl, 10 mM Tris [pH 7.4], and 0.1% Tween 20) and incubated with horseradish peroxidase-conjugated antiphosphotyrosine antibodies. Alternatively, the filters were incubated with anti-FAK pY397 or anti-FAK pY861 antibodies followed by a horseradish peroxidase-labeled anti-rabbit IgG antibody. Reactive bands were detected with enhanced chemiluminescence (ECL; Amersham Pharmacia Biotech). Controls for equal loading were performed either by running half of the immunoprecipitated material on a separate gel followed by transfer to nitrocellulose membranes and blotting for the immunoprecipitated protein (phosphotyrosine blots) or by stripping the membrane followed by blotting for the immunoprecipitated protein (FAK pY397 and pY861 blots).

FAK activity assay. Cells were stimulated by adhesion as described for the phosphotyrosine blotting assay, and the kinase activity of FAK was measured as previously described (35). Briefly, cells were solubilized in lysis buffer (10 mM Tris [pH 7.4], 150 mM NaCl, 0.5% NP-40, 1% Triton X-100, 1 mM EGTA, 1 mM EDTA, 1 mM PMSF, and 200 µM Na₃VO₄), FAK was immunoprecipitated, and the pellet was washed twice with lysis buffer and twice with kinase buffer (10 mM Tris [pH 7.5], 10 mM MnCl₂, 2 mM MgCl₂, and 0.05% Triton X-100). In vitro phosphorylation was performed for 15 min at room temperature in 15 µl of kinase buffer containing 5 µCi of [-³²P]ATP. The samples were subjected to SDS-PAGE, fixed with methanol-acetic acid, treated with 1 M KOH at 55°C for 1 h, fixed again, dried, and exposed on X-ray film.

SH2 domain pull-down assays. Cells were either kept in suspension or allowed to adhere to anti-1 antibodies as described in the phosphotyrosine blotting assay and lysed in a mixture of 20 mM HEPES (pH 7.5), 3 mM MgCl₂, 2 mM EDTA, 40 mM NaF, 40 mM Na₄P₂O₇, 1% Triton X-100, 10% glycerol, 200 µM Na₃VO₄, 1 mM NEM, 1 mM PMSF, 10 µg of leupeptin per ml, and 10 µg of pepstatin A per ml. The lysates were precleared by end-over-end rotation with glutathione-Sepharose beads (Amersham Pharmacia Biotech) for 1 h at 4°C. The precleared samples were incubated with glutathione-Sepharose and 10 µg of either GST-Crk SH2 or GST-Src SH2 for 2 h at 4°C. The pellets were washed three times with lysis buffer and subsequently boiled in sample buffer. The precipitated material was separated on SDS-PAGE (8% polyacrylamide) gels and transferred to nitrocellulose membranes, and blotting was performed as described for the phosphotyrosine blotting assay. To ensure that equal amounts of protein were subjected to the pull-down in each sample, the protein concentration of the cell lysates was determined by the bicinchoninic acid method (data not shown).

Immunofluorescent staining of cells. Cells were allowed to attach and spread for 1 h in DMEM on glass coverslips coated with ECM proteins. Coating with FN or LN-1 was done by incubating the coverslips overnight at 4°C with 10 µg of FN or 100 µg of LN-1 per ml in PBS. The coated coverslips were blocked with 1% BSA in PBS. Cells were fixed with 2% paraformaldehyde in PBS for 10 min, washed three times for 10 min each with PBS, and immunostained as described previously (52).

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

GD25-1A_Y783/795F cells exhibit delayed spreading. To assess the effect of the Y783/795F double mutant on cell spreading, GD25, GD25-1A, and GD25-1A_Y783/795F cells were allowed to attach and spread on VN (an V3- and V5-dependent substrate), on LN-1 (61 dependent), and on FN in the presence of GRGDS peptide (51 dependent) for 30 to 240 min. As illustrated in Fig. 1, the 1-expressing cell lines spread at a similar rate on VN. In contrast, on LN-1 (Fig. 1), the GD25-1A_Y783/795F cells exhibited a marked delay in spreading compared to the GD25-1A cells, while the GD25 cells (lacking 1 integrins) did not attach at all. After 4 h, the mutant cells were able to attain the same degree of spreading as the wild-type cells. The same difference in rate of spreading between the wild-type and the double tyrosine mutant cells was obtained when the cells were plated on FN in the presence of 0.5 mg of GRGDS peptide per ml (data not shown), which completely blocks V3- and V5-mediated adhesion to FN, while the 51-mediated adhesion is largely unaffected (52).

View larger version (60K): [in this window] [in a new window]   FIG. 1.   Effects of wild-type and mutant 1 integrins on spreading of GD25 cells. Phase-contrast images of cells (A) and calculated percentages (B) of spread cells of GD25 (), GD25-1A (), and GD25-1AY783/795F () cells on VN and LN-1 after different times of attachment (30, 60, 120, and 240 min). The spreading assay was performed as described in Materials and Methods. Bar, 100 µm.

The 1A_Y783/795F mutant is defective in signaling to FAK. FAK has been implicated in the regulation of integrin-mediated spreading and cell migration (18). To investigate the molecular mechanisms for the slow spreading and nonmigratory phenotype of the GD25 cells expressing the double tyrosine mutation Y783/795F in 1A, we tested the ability of these cells to mediate activation of FAK as monitored by measuring tyrosine phosphorylation of FAK following adhesion to an appropriate matrix. Plating of GD25, GD25-1A, and GD25-1A_Y783/795F cells on anti-1 antibodies resulted in a pronounced increase in tyrosine phosphorylation of FAK in the wild-type 1A cells, but not in the GD25 cells, which failed to attach under these conditions (Fig. 2). The GD25-1A_Y783/795F cells attached equally well as the GD25-1A cells, as previously reported (38), but only a very small increase in FAK tyrosine phosphorylation was observed compared to the level for cells in suspension. Identical results were obtained when the wild-type and mutant cells were plated on FN plus GRGDS and on LN-1. However, we feel that plating cells on anti-1 antibodies is a more specific way to look at signals exclusively derived from 1 integrins than plating on ECM proteins, where other cell surface molecules could be involved in binding to the substrate. To demonstrate that the impairment in FAK tyrosine phosphorylation in response to adhesion via the 1A_Y783/795F mutant was not due to a general FAK defect in these cells, we plated the three cell lines on VN and monitored V3- and V5-induced FAK phosphorylation (Fig. 2). All three cell lines exhibited similar levels of tyrosine phosphorylation of FAK on this substrate, illustrating that the defect in FAK phosphorylation in the GD25-1A_Y783/795F cells plated on 1-dependent substrates was a consequence of the mutant integrin.

View larger version (42K): [in this window] [in a new window]   FIG. 2.   Tyrosine phosphorylation of FAK in cells expressing wild-type or Y783/795F mutant 1A. GD25, GD25-1A, and GD25-1AY783/795F cells were either kept in suspension (Susp) or were attached to VN (VN) or to anti-1 antibody (-1) for 60 min. FAK phosphorylation was detected by immunoprecipitation (IP) of FAK followed by an immunoblotting for total phosphotyrosine (pY [upper panel]). The loading control was performed by blotting for FAK (lower panel).

View larger version (31K): [in this window] [in a new window]   FIG. 3.   Kinetics of FAK tyrosine phosphorylation. GD25-1A and GD25-1AY783/795F cells were plated on anti-1 antibodies for 0, 30, 60, 120, and 240 min. Immunoprecipitates (IP) were run on SDS-PAGE, transferred to nitrocellulose membranes, and blotted for phosphotyrosine (pY [upper panel]) and FAK (lower panel).

Single tyrosine substitutions do not dramatically affect FAK activation or cell spreading. To test if the effect seen on 1-dependent FAK tyrosine phosphorylation and cell spreading with the 1A_Y783/795F cells was the result of substituting both tyrosines within the tail or whether this would result from single-tyrosine substitutions, we also tested the ability of GD25-1A_Y783F and GD25-1A_Y795F cells to mediate FAK activation and cell spreading via 1 integrins. In this system, neither of the single-tyrosine mutations had a dramatic effect on 1-dependent FAK tyrosine phosphorylation and activation (Fig. 4A) or 1-dependent cell spreading (Fig. 4B) compared to that in wild-type cells. Only minor differences were observed compared to the level in the wild-type cells, and no obvious hierarchy of the importance of the two tyrosines could be detected, showing that the defects seen in the GD25-1A_Y783/795F cells required mutation of both tyrosines to phenylalanine.

View larger version (24K): [in this window] [in a new window]   FIG. 4.   Comparison of single versus double tyrosine mutations on FAK activation and cell spreading. (A) FAK was immunoprecipitated (IP) from GD25, GD25-1A, GD25-1AY783F, GD25-1AY795F, and GD25-1AY783/795F cells either kept in suspension (Susp) or attached to anti-1 antibody (-1) for 60 min. The precipitates were analyzed by blotting for phosphotyrosine (pY [upper panel]) and FAK (middle panel) and for in vitro kinase (IVK) activity (lower panel, FAK band is shown). (B) Percentages of spread cells were calculated for the cell lines GD25-1A (), GD25-1AY783F (), GD25-1AY795F (), and GD25-1AY783/795F () after different times of attachment to LN-1.

FAK autophosphorylation is defective in the GD25- 1A_Y783/795F cells. FAK activation involves both autophosphorylation and phosphorylation by Src family kinases at several different sites (6, 34). To elucidate at which level the defect of FAK phosphorylation in the double tyrosine mutant occurs, we used antibodies specifically directed to either the autophosphorylation site, pY397, or the major Src phosphorylation site, pY861, in FAK. In GD25-1A cells, both tyrosines 397 and 861 were phosphorylated in response to adhesion to 1 MAb (Fig. 5). In contrast, negligible phosphorylation levels of either Y397 or Y861 were detected in the 1A double tyrosine mutant cells in response to 1-mediated attachment (Fig. 5). These data strongly suggest that the reduced phosphotyrosine content of FAK in the Y783/795F mutant cells is due to a defect in 1-dependent autophosphorylation of FAK at tyrosine 397, which presumably also blocks the ability of Src to efficiently phosphorylate FAK on additional sites such as tyrosine 861. 

View larger version (40K): [in this window] [in a new window]   FIG. 5.   Phosphorylation of FAK on tyrosines 397 and 861. GD25-1A and GD25-1AY783/795F cells were kept in suspension (Susp) or attached to anti-1 antibodies (-1) for 60 min. FAK was immunoprecipitated (IP), and blots were performed with antibodies specific for FAK phosphotyrosines (pY) 397 and 861, respectively. Loading controls for each precipitation are shown.

Differential phosphorylation of paxillin, tensin, and p130^CAS by 1 integrins. The downstream effects of impaired FAK activation in the GD25-1A_Y783/795F cells were studied by analyzing the adhesion-dependent tyrosine phosphorylation of paxillin, tensin, and p130^CAS, three substrates for FAK or the FAK-Src complex (2, 31, 41). When plated on VN, all three proteins became tyrosine phosphorylated in both GD25-1A and GD25-1A_Y783/795F cells (data not shown). Attachment to anti-1 antibodies induced phosphorylation of all three proteins in the wild-type 1A, 1A_Y783F, and 1A_Y795F cells (Fig. 6). In contrast, with GD25-1A_Y783/795F cells, there was an absence of 1-dependent tyrosine phosphorylation of paxillin and tensin, consistent with the observed defect in integrin 1-mediated FAK activation in the double tyrosine mutant. On the other hand, p130^CAS was strongly tyrosine phosphorylated in the GD25-1A_Y783/795F cells in response to 1-mediated adhesion. The same results were obtained when the cells were allowed to attach to FN plus GRGDS or LN-1 (data not shown). These data indicate that 1-mediated p130^CAS tyrosine phosphorylation can occur through a signaling pathway that is distinct from that used for tyrosine phosphorylation of paxillin and tensin.

View larger version (55K): [in this window] [in a new window]   FIG. 6.   Integrin 1A-dependent phosphorylation of paxillin, tensin, and p130CAS. GD25, GD25-1A, GD25-1AY783F, GD25-1AY795F, and GD25-1AY783/795F cells were either kept in suspension (Susp) or attached to anti-1 antibodies (-1) for 60 min. Immunoprecipitates (IP) of paxillin (A), tensin (B), and p130CAS (C) were subjected to SDS-PAGE, transferred to nitrocellulose membranes, and blotted for phosphotyrosine (pY [upper panels]) and the precipitated protein (lower panels).

p130^CAS is phosphorylated by FAK-dependent and -independent pathways. To further elucidate the role of the defective FAK activation by the 1A_Y783/795F mutant, we used a fusion protein consisting of GST coupled to the SH2 domain of Crk to pull down binding proteins from the lysates of GD25-1A and GD25-1A_Y783/795F cells (Fig. 7). Paxillin was clearly precipitated from adherent wild-type-expressing cells, but not from the double tyrosine mutant cells. More surprisingly, the amount of p130^CAS precipitated from adherent cells was strongly reduced in the Y783/795F cells compared to that in the wild-type cells. This result suggests that although no change in the adhesion-dependent overall tyrosine phosphorylation of p130^CAS was detected by using broad-specificity reagents in GD25-1A_Y783/795F cells, the specific phosphorylation site(s) responsible for binding of p130^CAS to Crk may be less efficiently generated. In pull-downs blotted for phosphotyrosine, we could detect reduced precipitation of proteins with approximate sizes of 100, 150, and 190 kDa in the GD25-1A_Y783/795F cells, in addition to bands likely to correspond to paxillin and p130^CAS (70 to 80 and 130 kDa, respectively). We also tested the ability of the SH2 domain of Src to pull down p130^CAS (Fig. 7). In contrast to the Crk SH2 domain, the adhesion-dependent binding of p130^CAS to Src SH2 was only slightly reduced in the GD25-1A_Y783/795F cells compared to that in the GD25-1A cells, indicating that these two SH2 domains bound to different phosphorylation sites. In this respect, the Crk SH2-binding site most likely represents a FAK-Src-dependent phosphorylation site, while the Src SH2-binding site might represent a FAK-independent site.

View larger version (29K): [in this window] [in a new window]   FIG. 7.   Pull-down of tyrosine-phosphorylated proteins with SH2 domains. GD25-1A and GD25-1AY783/795F cells were kept in suspension (Susp) or attached to anti-1 antibodies (-1) for 60 min, lysed, and incubated with GST-Crk SH2 (A) or GST-Src SH2 (B) fusion proteins bound to glutathione-Sepharose. Precipitates were subjected to SDS-PAGE, transferred to nitrocellulose membranes, and blotted for the indicated proteins. pY, phosphotyrosine; CAS, p130CAS.

FAK localizes to FAs in GD25-1A_Y783/795F cells. When cells were allowed to attach for 60 min either to FN in the presence of GRGDS peptide or to LN-1, immunostaining of FAK revealed that it still localized to FAs in both GD25-1A_Y783/795F cells and GD25-1A cells (Fig. 8A to D). The phosphotyrosine content of the FAs of GD25-1A_Y783/795F cells looked similar to that of GD25-1A cells, indicating that there are other active tyrosine kinases in the FAs of these cells (Fig. 8E to H). In addition, vinculin, talin, paxillin, and tensin localized to FAs in both wild-type 1A- and 1A_Y783/795F-expressing cells (data not shown). In contrast, when stained for FAK pY397 (Fig. 9) and pY861 (data not shown), a prominent signal in FAs was observed in GD25-1A cells (Fig. 9A), whereas only weak staining was detected in FAs of GD25-1A_Y783/795F cells (Fig. 9C). Thus, although localization of FAK to FAs has been shown to be an essential step for integrin-dependent activation of the enzyme (44), our results emphasize that additional regulation of FAK signaling occurs at these sites.

View larger version (41K): [in this window] [in a new window]   FIG. 8.   Subcellular localization of FAK and phosphotyrosine-containing proteins. Immunofluorescent stainings of GD25-1A (A, B, E, and F) and GD25-1AY783/795F (C, D, G, and H) cells attached to FN plus GRGDS for 1 h are shown. Double staining for integrin subunit 1 (A and C) and FAK (B and D) and integrin subunit 1 (E and G) and phosphotyrosine (F and H) was performed as described in Materials and Methods. Bar, 20 µm.

View larger version (110K): [in this window] [in a new window]   FIG. 9.   Detection of FAK tyrosine phosphorylation in FAs. FAK pY397 was detected by immunofluorescence in GD25-1A (A and B) and GD25-1AY783/795F (C and D) cells attached to FN plus GRGDS for 1 h. The cells were double stained for FAK pY397 (A and C) and FAK (B and D). Bar, 20 µm.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

In this study, we have shown that GD25 cells expressing integrin subunit 1A containing the Y783/795F mutation are defective in activating FAK, most likely due to a loss in autophosphorylation of the tyrosine 397. These cells were previously shown to be able to attach and spread via the mutated 1 integrins, but they contained an increased number of FAs, had an altered cytoskeletal architecture, and were defective in migration (38). In this report, we have also now demonstrated that the 1A_Y783/795F mutant cells exhibit delayed spreading compared to cells expressing wild-type integrin. Thus, the phenotype of GD25 cells expressing integrin subunit 1A_Y783/795F is in several ways similar to that of the FAK knockout fibroblasts (18, 34). The effects of the FAK knockout suggested that FAK is involved in the turnover of FA and thereby promotes cell spreading and migration. Cell spreading has been suggested to be triggered by FAK-Src phosphorylation of paxillin (37), while the role of FAK in cell migration was reported to be dependent on PI 3-kinase, which can bind to pY397 in FAK via its SH2 domain (36), and p130^CAS, which can bind via its SH3 domain to the FAK-Src complex and become phosphorylated (8, 21).

In the GD25-1A_Y783/795F cells, no 1-dependent tyrosine phosphorylation of paxillin and tensin occurred. Contrary to our results, phosphorylation of these proteins was not reduced in FAK-null cells compared to FAK-expressing fibroblasts (18, 50). This discrepancy may be due to upregulation of the FAK-related kinase Pyk2 (also known as CAK, CADTK, or RAFTK) in the FAK-null cells, which in this context can partially compensate by carrying out FAK functions (45).

Although GD25-1A_Y783/795F cells clearly have a defect in 1 integrin-dependent FAK activation, FAK is still present in FAs, indicating that the defect is not at the level of gross cellular localization. As a model to explain the impaired FAK activation in the GD25-1A_Y783/795F cells, we therefore suggest that the mutation of the two tyrosines in the cytoplasmic tail of integrin 1A prevents the interaction with one or more other proteins, which are required to create a functional complex that stimulates FAK autophosphorylation and activation. When single Y-to-F mutants were tested (1A_Y783F and 1A_Y795F), only minor effects on 1-dependent FAK phosphorylation and activation, paxillin phosphorylation, tensin phosphorylation, and cell spreading were observed. These results emphasize that the defect seen in the GD25-1A_Y783/795F cells is an effect of removing both tyrosines in the carboxyl-terminal part of the cytoplasmic domain.

The fact that overall tyrosine phosphorylation levels of the FAK-Src substrate p130^CAS are not affected by the Y783/795F double mutation in 1 is intriguing. We do not know the identity of the kinase responsible for carrying out this integrin-dependent phosphorylation. However, p130^CAS is considered to be phosphorylated mainly by Src (14, 47, 50), and it has been reported that Src family kinases can become activated by integrin ligation independently of FAK activation (9, 43). The p130^CAS phosphorylation detected by generic phosphotyrosine-directed antibodies in the GD25-1A_Y783/795F cells might therefore depend on integrin-mediated activation of Src molecules, which are not in complex with activated FAK. However, by using two different SH2 domains to bind p130^CAS, we were able to detect differential phosphorylation of p130^CAS. Interestingly, the binding site of Crk SH2 domain seems to largely coincide with FAK activation, and the binding site of Src SH2 seems to be independent of the activation of the FAK-Src complex. Our observations illustrate the complex nature of integrin-mediated signaling and support previous reports, which indicate that tyrosine phoshorylation of p130^CAS can occur through several independent pathways (4, 12, 30).

Detection of tyrosine-phosphorylated 1 has so far only been made in transformed cells (17, 19). Integrin subunit 3, on the other hand, is tyrosine phosphorylated at residues corresponding to Y783 and Y795 in 1 after ligating IIb3 and V3 integrins (1, 25), and these two tyrosines have been shown to play an important role in platelet function (24). Various results illustrating the effect on FAK activation by single Y-to-F mutations in 3 have been reported (39, 48), while the effect on FAK activation of double Y-to-F mutants has not been studied for 3. Our results strongly indicate that the hydroxyl moieties of the two tyrosines in the integrin subunit 1A are of critical importance, possibly by serving as phosphate acceptors. The roles of tyrosine phosphorylation of the 1A cytoplasmic tail in relation to FAK activation and FA dynamics in transformed and nontransformed cells are pressing questions that are being pursued.