Identification of a Novel Interaction Between Integrin Β1 and 14-3-3Β

Dong Cho Han1, Luis G Rodriguez1 and Jun-Lin Guan*,1

1Cancer Biology Laboratories, Department of Molecular Medicine, College of Veterinary Medicine, Cornell University, Ithaca, New York, NY 14853, USA

Integrins are cell surface receptors for extracellular a particular a and b subunit. Integrin-mediated cell matrix, which play important roles in a variety of adhesion to the ECM plays important roles in a variety biological processes. 14-3-3 proteins are a highly of biological processes including embryonic develop- conserved family of cytoplasmic proteins that associate ment, wound healing and cancer. with several intracellular signaling molecules in regula- Studies have shown that integrins can mediate signal tion of various cellular functions. Here, we report transduction across the plasma membrane (Aplin et al., identi®cation of an interaction between the integrin b1 1998; Clark and Brugge, 1995; Schwartz et al., 1995). cytoplasmic domain and 14-3-3b by using the yeast two- Integrin-mediated cell adhesion has been shown to hybrid screen. Like several other proteins, the integrin b1 stimulate a variety of signaling cascades including cytoplasmic domain associated with 14-3-3b by a non- activation of protein tyrosine kinases such as FAK and phosphoserine mechanism. The 14-3-3b/integrin b1 Src, protein serine/threonine kinases including protein interaction was con®rmed by in vitro binding assays as kinase C, mitogen-activated protein kinases, and c-Jun well as co-precipitation in vivo. Furthermore, we found N-terminal kinase, lipid kinases like PI-3-kinase, that 14-3-3b co-localized with integrin b1 during the cytoplasmic alkalinization and calcium in¯ux. Interest- early stage of cell spreading on ®bronectin, suggesting a ingly, integrins also mediate inside-out signaling where potential role of the 14-3-3b/integrin b1 interaction in integrin anity for extracellular ligands is regulated by the regulation of cell adhesion. Using tetracycline- a variety of signaling pathways triggerred by other cell regulated expression system, we showed that overexpres- surface receptors. Changes both in integrin conforma- sion of 14-3-3b stimulated cell spreading and migration tion and/or clustering have been suggested to play a on ®bronectin but not on poly-L-lysine. However, the role in integrin anity regulation (Loftus and induced expression of 14-3-3b did not aect tyrosine Liddington, 1997; Shattil et al., 1998). phosphorylation of FAK or its substrates, p130cas and The cytoplasmic domains of integrins play a key role paxillin, suggesting that 14-3-3b regulated integrin- in both integrin signaling and anity regulation. mediated cell spreading and migration by FAK-indepen- Despite their relatively small size and lack of enzymatic dent mechanisms. Taken together, these results identify activity, integrin b cytoplasmic domains are sucient an interaction between integrin and 14-3-3 proteins and to mediate integrin localization to focal contacts suggest a potentially novel cellular function for 14-3-3 (LaFlamme et al., 1992; Marcantonio et al., 1990; proteins in the regulation of integrin-mediated cell Reszka et al., 1992) and induce FAK phosphorylation adhesion and signaling events. Oncogene (2001) 20, (Akiyama et al., 1994; Guan et al., 1991; Tahiliani et 346±357. al., 1997). They are also required for ®bronectin matrix assembly (Wu et al., 1995) and cell motility (Pasqualini Keywords: integrins; 14-3-3 proteins; cell migration; and Hemier, 1994). The cytoplasmic domains of both protein-protein interaction integrin a and b subunits have been implicated in the regulation of integrin anity (Hughes et al., 1996; O'Toole et al., 1994; 1990). Introduction The critical role of the cytoplasmic domains in integrin function implicates the importance of cyto- The integrins are a family of cell-surface glycoproteins plasmic proteins that modify or bind to integrin which link the extracellular matrix (ECM) to the cytoplasmic domains. Consistent with integrin localiza- cytoskeleton (Hynes, 1992). Each integrin is a hetero- tion in focal contacts, several cytoskeletal proteins dimer that contains an a and a b subunit with each including talin, a-actinin, ®lamin and paxillin have subunit composed of a large extracellular domain, a been shown to interact with cytoplasmic domain of single transmembrane region, and a short cytoplasmic integrin b subunit, which may mediate integrin- domain (except b4). The speci®city of integrin interac- cytoskeleton connections (Horwitz et al., 1986; Loo et tions with ligands is determined by the combination of al., 1998; Otey et al., 1990; Pfa et al., 1998; Schaller et al., 1995; Sharma et al., 1995). In addition, several intracellular signaling proteins have been reported to associate with integrin b cytoplasmic domains, which *Correspondence: J-L Guan Received 18 July 2000; revised 26 October 2000; accepted 1 play interesting and important regulatory functions. November 2000 These include cytohesion-1 (Kolanus et al., 1996), Integrin b1 interaction with 14-3-3b DC Han et al 347 FAK (Schaller et al., 1995), ILK (Hannigan et al., binding site for clone #11 using yeast two hybrid 1996), b3-endonexin (Shattil et al., 1995), ICAP-1 system (Figure 1). Deletion of 13 amino acids from the (Chang et al., 1997; Zhang and Hemler, 1999), and b1 C-terminus of the integrin b1 cytoplasmic domain Rack1 (Liliental and Chang, 1998). (D4) did not signi®cantly aect its interaction with 14-3-3 proteins are a highly conserved family of clone #11. In contrast, removal of 28 amino acids (D3) dimeric proteins ubiquitously expressed in all eukar- or 42 residues (D1) prevented its binding to clone #11, yotic cells (for review, see Fu et al., 2000; Aitken et al., suggesting that residues between the D3 and D4 1995). Members of 14-3-3 family proteins have been mutants (776 ± 790) were involved in binding to 14-3- shown to bind a number of signaling molecules by 3b. 14-3-3 proteins are adaptor molecules that often both phosphoserine motifs (Muslin et al., 1996; Yae bind to phosphorylated Ser motifs (Yae et al., 1997). et al., 1997) and non-phosphoserine mechanism (Clark The Ser790 residue in the integrin b1 cytoplasmic et al., 1997; Masters et al., 1999; Wang et al., 1999). domain has been reported to be phosphorylated in Many of these 14-3-3-associated proteins are proto- mitotic cell (Hynes, 1992). However, mutation of oncogene and oncogene products. Association of 14-3- Ser790 to Ala did not block the interaction, suggesting 3 proteins with intracellular signaling molecules has that Ser790 was not involved in b1 integrin binding to been proposed to regulate their enzymatic activity, to 14-3-3b. Likewise, mutation of Tyr788 to Phe did not bring dierent binding partners (e.g. enzyme and aect its interaction with 14-3-3b, suggesting that substrate) together, and to stabilize the conformation Tyr788 in the NPXY motif is not necessary for the of the associated proteins. Expression of the 14-3-3 was interaction. Clone #11 also interacted with the increased signi®cantly in lung cancer tissues compared cytoplasmic domain of integrin b3 to a similar level with the neighboring normal part of lung (Nakanishi et as b3-endonexin (Shattil et al., 1995). Together, these al., 1997) and in head and neck squamous carcinoma studies indicated that integrin b1 residues 776 ± 790 (Villaret et al., 2000). 14-3-3g expression was increased (other than Tyr788 or Ser790) that are also conserved in response to vessel damage and proliferative signals in the integrin b3 cytoplasmic domain mediated its (Autieri et al., 1996). These studies suggest that 14-3-3 interaction with 14-3-3b in a phosphorylation indepen- proteins are involved in the regulation of mitogenic dent manner. signaling as well as development of cancer. To further understand integrin signaling and regula- Analysis of integrin b1 interaction with 14-3-3b in vitro tion, we employed the yeast two-hybrid screen to and in mammalian cells identify novel cellular proteins that bind to the integrin b1 cytoplasmic domain. We report here the association To determine whether the 14-3-3b fragment (clone #11, of 14-3-3b with integrin b1 cytoplasmic domain by a encoding residues 30 ± 247) could bind directly to the non-phosphoserine mechanism and the eect of 14-3- integrin b1 cytoplasmic domain in vitro, this fragment 3b expression on integrin-mediated cell spreading and was inserted into the pGEX-2T vector to generate a migration. These results suggest potential role of 14-3-3 GST fusion protein for binding assays. The cytoplas- proteins in the regulation of integrin functions. mic domain of b1 integrin was inserted into a mammalian expression vector pKH3 fused in frame with a triple HA epitope tag and a GCN4 dimerization segment. This plasmid was transfected into CHO cells Results and lysates were prepared and incubated with the GST fusion proteins that had been immobilized on Identification of integrin b1 cytoplasmic domain glutathione-agarose beads. After washing, the bound interaction with 14-3-3b by yeast two hybrid screen We employed yeast two-hybrid screen to identify protein(s) that can interact with the cytoplasmic domain of b1 integrin. A cDNA fragment encoding the complete b1 cytoplasmic domain was inserted into a modi®ed pAS2 vector (see Materials and methods) and used to screen a HeLa cell library (Hannon et al., 1993). From approximately 26106 transformants, four clones (clones #10, #11, #20 and #21) were identi®ed that speci®cally interacted with the bait plasmid. Plasmid DNA from these clones was then isolated. Partial DNA sequencing indicated that clone #11 Figure 1 Interaction of 14-3-3b with b1 integrin cytoplasmic encodes almost a complete cDNA for 14-3-3b missing domain and mutants in yeast two-hybrid system. Yeast strain only the N-terminal 29 amino acids (residues 30 ± 247). HF7c was co-transformed with pGAD-#11 or pGAD-b3- Retransformation of clone #11 into yeast con®rmed its endonexin and pAS2DD encoding integrin b1 cytoplasmic speci®c interaction with the bait plasmid containing the domain, its mutants, or integrin b3 cytoplasmic domain, as b1 cytoplasmic domain, but not the bait vector alone. indicated. The growth of transformants on a synthetic medium lacking tryptophane, leucine and histidine are shown. The NPXY We then used a series of b1 cytoplasmic domain motifs are underlined and the S790 to A and Y788 to F mutations deletion and point mutations to map the potential are shown in bold

Oncogene Integrin b1 interaction with 14-3-3b DC Han et al 348 proteins were analysed by Western blotting with mAb To further investigate interaction of 14-3-3b and 12CA5 for the HA epitope. Figure 2A shows that the integrin b1, we examined subcellular localization of integrin b1 cytoplasmic domain bound to the GST these two proteins during cell adhesion. HFF cells were fusion protein containing the 14-3-3b fragment en- plated on FN or PLL for 30 or 120 min, and the coded by clone #11, but not GST alone. Although at a distribution of endogenous 14-3-3b and integrin b1 was lower eciency, it also bound to the GST fusion examined by co-immuno¯orescence staining (Figure 3). protein containing an unknown protein fragment We found that when cells were plated on FN, 14-3-3b encoded by clone #10 that was isolated from the same and integrin b1 were partially co-localized at punctuate yeast two-hybrid screen. sites in the periphery of the cells resembling focal Association of 14-3-3b with endogenous integrin b1 contacts at the earlier time point after plating. At the in intact cells was also con®rmed using coimmunopre- later stage after plating, integrin b1 is detected in focal cipitations. Full-length 14-3-3b cDNA was derived contacts located both at the cell peripheral and the from an EST clone using PCR and inserted into a central cell bodies. In contrast, 14-3-3b was completely mammalian expression vector pDHGST to generate localized in the cytoplasm and absent from these focal pDHGST-14-3-3b encoding 14-3-3b fused to GST at contacts. Colocalization of 14-3-3b and integrin b1 was the N-terminus (Han and Guan, 1999), as described in not detected in cells plated on PLL at either the early Materials and methods. The fragment encoded by or late stage of cell adhesion. Together, these results clone #10 was also inserted into the same vector to showed that cell adhesion to FN stimulated a transient generate pDHGST-#10. 293T cells were transiently co-localization of 14-3-3b and integrin b1 in the early transfected with pDHGST-14-3-3b, pDHGST-#10, or focal contacts structures in the cell periphery. They pDHGST vector alone. The lysates were prepared and also suggested that interaction between 14-3-3b and 14-3-3b was precipitated by incubation with glu- integrin b1 is not constitutive, but temporally and tathione-Sepharose beads. The bound proteins were spatially regulated. resolved on SDS ± PAGE and subjected to Western blotting with anti-integrin b1 antibody. Figure 2B Regulation of cell spreading and migration by 14-3-3b shows that integrin b1 was associated with 14-3-3b, but did not bind to the GST control. It was also associated To investigate the potential cellular functions of 14-3- with the fragment encoded by clone #10 although with 3b association with integrin b1, we obtained inducible less eciency. Together, these results con®rmed 14-3- overexpression of 14-3-3b in NIH3T3 cells using the 3b interaction with integrin b1 through its cytoplasmic tetracycline regulated expression system as described in domain. Materials and methods. Multiple clones were analysed for the expression of exogenous 14-3-3b by Western blotting using 12CA5, a monoclonal antibody against the HA epitope tag fused to the N-terminus of exogenous 14-3-3b. Figure 4A shows expression of the exogenous 14-3-3b in several clones upon induction by removal of tetracycline in the media (compare lanes I with lanes U). Western blotting with anti-14-3-3 antibody con®rmed expression of the exogenous 14-3- 3b (arrow) which is slightly larger than the endogenous 14-3-3s (arrowheads, the lower band is 14-3-3b) due to the fusion of three HA epitopes at the N-terminus (Figure 4B). The clone with the highest expression (#2) was designated as 14-3-3 cells and was used for further analysis. The eect of induced overexpression of 14-3-3b on cell adhesion on FN was examined by cell spreading assays (Figure 5). At 5 min after attachment, little cell spreading was observed under uninduced or induced conditions for either 14-3-3 cells or the Mock cells, Figure 2 In vitro and in vivo interaction between 14-3-3b and b1 although a small fraction of 14-3-3 cells under induced integrin. (A). Immobilized GST-#11, GST-#10 or GST was condition appeared to start spreading (Figure 5A ± D). incubated with lysates from CHO cells that had been transfected At 10 min after attachment, a signi®cantly higher with pKH3DD-b1cyto. After washing, the bound proteins were fraction of 14-3-3 cells under induced condition was analysed by Western blotting with 12CA5. An aliquot of the lysate was also analysed directly (input). (B). 293T cells were spread compared with the cells under uninduced transfected with pDHGST-14-3-3b, pDHGST-#10, or pDHGST conditions (compare Figure 5G,H). The Mock cells vector, as indicated. Two days after transfection, 14-3-3b was showed similar spreading under both induced and pulled-down from the lysates using glutathione-coupled agarose uninduced conditions, which is comparable with the beads. They were then analysed by Western blotting with anti- integrin b1 antibody. An aliquot of the lysates was also analysed 14-3-3 cells under the uninduced conditions (Figure directly (WCL). Molecular weight markers are shown on the left 5E ± G). At 20 min after attachment, smaller but and the position of integrin b1 is marked by an arrow on the right signi®cant increased fraction of cell spreading was also

Oncogene Integrin b1 interaction with 14-3-3b DC Han et al 349

Figure 3 Subcellular localization of integrin b1 and 14-3-3b in HFF cells. HFF cells were plated on ®bronectin- (FN) or PLL- coated coverslips for 30 or 120 min as indicated. They were then ®xed, permeabilized, and double stained with the monoclonal anti- integrin b1 antibody (left panels) and the polyclonal anti-14-3-3b antibody (middle panels). The merged images are shown on the right panels

observed for 14-3-3 cells under induced conditions exogenous 14-3-3b expression resulted in an approxi- compared with that under the uninduced conditions or mately 2.5-fold and 1.5-fold increases in cell spreading Mock cells under both conditions (Figure 5I ± L). at 10 and 20 min after cell attachment, respectively Quantitation of the results from three independent (panels N). Similar results were obtained from analysis experiments indicated similar rate of cell spreading for of another independent clone with inducible 14-3-3b the Mock cells under induced and uninduced condi- expression (Figure 5O). At later time points and in tions (Figure 5M). In contrast, induction of the growing cells, both 14-3-3 and Mock cells under either

Oncogene Integrin b1 interaction with 14-3-3b DC Han et al 350 Gilmore and Romer, 1996; Ilic et al., 1995; Klemke et al., 1998; Richardson and Parsons, 1996). Further- more, p130cas has been shown to bind 14-3-3z in an adhesion-dependent manner (Garcia-Guzman et al., 1999). To investigate whether 14-3-3b may regulate cell spreading and migration through these proteins, we examined their tyrosine phosphorylation upon induction of exogenous 14-3-3b expression. 14-3-3 cells under either uninduced or induced conditions were removed from the plates and replated on FN (lanes FN) or kept in suspension (lanes Sus). Western blotting of the total cell lysates with anti-14-3-3b con®rmed the induced expression of 14-3-3b in both suspended and adherent cells (Figure 8A). As expected, cell adhesion induced tyrosine phosphorylation of several proteins, most prominently bands in the range of 120 and at Figure 4 Inducible expression of 14-3-3b in NIH3T3 clones. Lysates were prepared from several NIH3T3 cell clones with 70 kDa (Figure 8B). However, the increased phosphor- inducible 14-3-3b expression under uninduced (lanes U) and ylation was similar in cells under induced and induced (lanes I) conditions. They were analysed by Western uninduced conditions, suggesting that the induced blotting with 12CA5 (A) or anti-14-3-3 antibody that recognizes overexpression of 14-3-3b did not regulate tyrosine all endogenous 14-3-3 isoforms (B). The arrows on the right indicate exogenous 14-3-3b. The arrowheads indicate endogenous phosphorylation of these proteins. 14-3-3e (upper) and other 14-3-3s including 14-3-3b. Molecular To analyse tyrosine phosphorylation of FAK, the weight markers are shown on the left cell lysates were immunoprecipitated with anti-FAK followed by Western blotting with anti-phosphotyr- osine antibody PY-20. As shown in Figure 9A conditions showed complete spreading without any overexpression of exogenous 14-3-3b did not aect apparent dierences (data not shown). Lastly, induc- FAK phosphorylation compared with uninduced cells, tion of 14-3-3b expression did not increase cell although FAK phosphorylation was enhanced upon spreading on PLL (Figure 6), suggesting a speci®c replating on FN as expected. Western blotting of the eect of 14-3-3b on integrin-mediated cell adhesion. immunoprecipitates with anti-FAK con®rmed the Together, these results suggest that 14-3-3b may presence of similar amounts of FAK (right panel). regulate the rate of cell spreading at the early stage Similar analysis indicated that inducible overexpression of integrin-mediated cell adhesion. of 14-3-3b did not aect cell adhesion stimulated We also evaluated potential eects of 14-3-3b on cell tyrosine phosphorylation of p130cas (Figure 9B) or migration on FN or PLL using a wound closure assay. paxillin (Figure 9C). Taken together, these results As shown in Figure 7A, when plated on FN, induced demonstrate that regulation of integrin-dependent cell overexpression of 14-3-3b accelerated the closure of the adhesion and migration by 14-3-3b is not through its wound in comparison to cells under uninduced eects on FAK or its substrates, p130cas and paxillin. conditions. Quantitation of the results indicated that induction of the exogenous 14-3-3b expression resulted in an approximately 50% increases in cell migration Discussion (Figure 7C). In contrast, 14-3-3b did not increase cell migration on PLL, but rather inhibited it slightly The 14-3-3 family of proteins are highly conserved (Figure 7B,C). To exclude the potential contribution of proteins ubiquitously expressed in all eukaryotic cells increased cell proliferation in stimulation of wound (Aitken et al., 1995). Members of the 14-3-3 family closure, we also assayed for growth of 14-3-3 cells proteins have been shown to interact with many key under the induced and uninduced conditions. Figure signaling molecules to regulate intracellular signaling 7D showed a slightly reduced rate of proliferation for events in cell proliferation, dierentiation and apopto- 14-3-3 cells under the induced conditions in compar- sis (Bonnefoy-Berard et al., 1995; Craparo et al., 1997; ison to that under the uninduced conditions. Together Fu et al., 1994; Garcia-Guzman et al., 1999; Hausser et these results suggest that the increased wound closure al., 1999; Liu et al., 1997; Meller et al., 1996; Reuther on FN upon 14-3-3b expression was indeed due to an et al., 1994; Zhang et al., 1997, 1999). In this paper, we increased cell migration and not by altered cell have identi®ed a novel interaction between 14-3-3b and proliferation rate. b1 integrin through its cytoplasmic domain. This interaction was identi®ed initially using the yeast two hybrid screening and was subsequently con®rmed by in Effects of 14-3-3b on the tyrosine phosphorylation of FAK vitro binding assays and co-immunoprecipitation in and its substrates intact cells. Interestingly, immuno¯uorescence staining Focal adhesion proteins, p130cas, paxillin and FAK, of endogenous integrin b1 and 14-3-3b suggested that have been implicated in a role of integrin-mediated cell this interaction was temporally and spatially regulated spreading and migration (Cary et al., 1996, 1998; by cell adhesion. Together with the recent ®nding of

Oncogene Integrin b1 interaction with 14-3-3b DC Han et al 351

Figure 5 Eect of inducible expression of 14-3-3b on cell spreading on FN. One representative clone of 14-3-3 cells (C,D,G,H,K,L) or mock cells (A,B,E,F,I,J) under uninduced (A,C,E,G,I,K) or induced (B,D,F,H,J,L) conditions were subjected to cell spreading assays as described in Material and methods. Cells were allowed to spread for 5 min (A±D), 10 min (E±H), or 20 min (I±L), and then photographed. Average and standard deviations from three independent experiments are shown for Mock cells (M) and the 14- 3-3 cells (N). Results from similar analysis of another independent 14-3-3 cell clones are shown in (O). *P=0.024 and **P=0.014 in comparison to values from uninduced cells

14-3-3z interaction with p130cas (Garcia-Guzman et the two NPXY motifs in mediating integrin binding to al., 1999), these results suggested a potentially novel 14-3-3b (Figure 1). These analyses revealed a potential cellular function for 14-3-3 proteins in the regulation of role for residues 776 ± 790 in the middle region of integrin-mediated cell adhesion and signaling events. integrin b1 cytoplasmic domain for binding 14-3-3b. 14-3-3 proteins have been shown to bind speci®c Interestingly, we also found that the cytoplasmic phosphoserine-containing motifs, RSXpSXP and domain of integrin b3 could also bind to 14-3-3b RXY/FXpSXP present in many of its binding partners using the yeast two-hybrid assays (Figure 1). These (Aitken et al., 1995; Muslin et al., 1996; Yae et al., results suggested that the conserved residues between 1997). Although it was reported to be phosphorylated integrins b1andb3 within this region are critical in under some conditions (Hynes, 1992), the single serine mediating the binding to 14-3-3b. Future experiments residue in the cytoplasmic domain of integrin b1 is not will be necessary to delineate the critical residues on present in these motifs. Indeed, mutation of this serine both integrin cytoplasmic domains and 14-3-3 proteins to alanine did not block 14-3-3b interaction with the involved in this non-phosphoserine dependent interac- integrin b1 cytoplasmic domain (Figure 1). Therefore, tion. integrin interaction with 14-3-3b is likely to be The cellular functions of the 14-3-3b interaction with mediated through non-phosphoserine motifs as in integrin b1 are not completely clear at present. In several other cases of protein interactions with 14-3-3 support of a potential role of 14-3-3 proteins in the proteins described recently (Clark et al., 1997; Masters regulation of integrin functions, we found that et al., 1999; Wang et al., 1999). inducible expression of exogenous 14-3-3b increased Analysis of truncation and other point mutation cell spreading and migration on FN using a tetra- also excluded a role for the Tyr residues in either of cycline regulated expression system. Expression of 14-

Oncogene Integrin b1 interaction with 14-3-3b DC Han et al 352

Figure 6 Eect of 14-3-3b on cell spreading on PLL. (A) One representative clone of 14-3-3 cells under uninduced (left panels) or induced (right panels) conditions were assayed for their spreading on PLL for various times as indicated. (B) Average and standard deviations from two independent assays are shown

3-3b did not aect cell spreading or migration on non- Alternatively, binding of 14-3-3b to integrins may integrin substrate PLL, providing further support for a alter integrin anity for ligands by an inside-out functional signi®cance of 14-3-3b interaction with signaling mechanism, which may enhance the integrin's integrins. There are several possible models to explain ability to mediate cell spreading and migration. the regulation of integrin-mediated cell adhesion and Conformational changes and/or integrin clustering migration by 14-3-3b. First, 14-3-3 proteins have been have been suggested to be responsible for regulation shown to bind a variety of other cellular proteins. of integrin anity by intracellular signaling (Loftus Formation of homodimers and hetrodimers of 14-3-3 and Liddington, 1997; Shattil et al., 1998). The proteins have been shown to be important for their possibility that binding of the dimeric 14-3-3b to the biological activities. It is therefore possible that each integrin b1 cytoplasmic domain may induce dimeriza- subunit of the 14-3-3 dimer can bind to a dierent tion (or clustering) of integrins is particularly interest- cellular protein and dimerization will bring these ing, which will require further investigation. It is also components in one place to coordinate multiple important to note that these two possible mechanisms reactions. For example, Cdc25 and Bcr have been (inducing integrin clustering and recruiting signaling shown to interact with Raf-1 in a 14-3-3 dependent molecules to integrin as discussed above) are not manner (Braselmann and McCormick, 1995; Conklin mutually exclusive. et al., 1995). This raises the interesting possibility that Lastly, we can not exclude completely the possibility 14-3-3b may mediate an indirect interaction of integrin that 14-3-3b regulated cell spreading and migration by cytoplasmic domain with signaling molecules such as some indirect mechanisms independent of its direct PI-3K, PKC or Raf to trigger downstream pathways in association with integrin b1 as identi®ed here. For cell adhesion and migration. Indeed, some of these example, binding of 14-3-3b with signaling molecules signaling molecules have been found to co-aggregate such as PI-3K (Bonnefoy-Berard et al., 1995), PKC with integrins upon binding of beads coated with (Morgan and Burgoyne, 1992; Toker et al., 1992), and ligand or anti-integrin antibodies (Miyamoto et al., Raf (Fantl et al., 1994; Freed et al., 1994; Fu et al., 1995). In addition, integrin activation has been shown 1994; Irie et al., 1994) might alter their activities and/or to stimulate activities of PI-3K (Chen and Guan, 1994; subcellular localizations. Signaling pathways from Khwaja et al., 1997), PKC (Chun and Jacobson, 1993; these molecules could then modulate integrin anity Miranti et al., 1999; Vuori and Ruoslahti, 1993) and resulting in increased cell spreading and migration. Raf (Lin et al., 1997). Development of speci®c inhibitors for integrin b1/14-3-

Oncogene Integrin b1 interaction with 14-3-3b DC Han et al 353

Figure 7 Eect of inducible expression of 14-3-3b on cell migration and proliferation. Uninduced (U) or induced (I) 14-3-3 cells were plated on FN-coated dishes (A) or PLL-coated dishes (B). Four hours later, a wound was made by scraping cell monolayer with a pipet tip (0 h). The cells were then washed, incubated for 18 h to allow migration into the wounded area (18 h). (C) Quantitation of cell migration by counting the number of cells that had migrated into the cell-free space. Four dierent areas were randomly chosen from one plate and more than 200 cells were counted from each wounded area. Data was normalized to the number of cells migrated under uninduced condition. (D) Cells were incubated in the growth media under uninduced (open square) or induced (closed circle) conditions. They were counted at 1-day intervals as described in Material and methods. Average and standard deviations from three independent experiments are shown

3b interaction and/or use of speci®c 14-3-3b mutants a role in cell spreading and migration (Cary et al., lacking interactions with integrin b1 only (but retain its 1996, 1998; Gilmore and Romer, 1996; Ilic et al., 1995; binding to other signaling molecules) will be necessary Klemke et al., 1998; Richardson and Parsons, 1996). to con®rm that regulation of integrin-mediated cell However, inducible overexpression of exogenous 14-3- spreading and migration by 14-3-3b is due to its 3b did not aect phosphorylation of these molecules speci®c interaction with integrin b1, but not other compared with uninduced cells under either suspended indirect mechanisms. or attached conditions (Figure 8). Furthermore, we did Tyrosine phosphorylation of FAK and its substrates not observe any dierences in the overall pattern of p130cas and paxillin is an early event of integrin- tyrosine phosphorylation of cellular proteins (Figure mediated cell adhesion, which has been shown to play 7). These results suggested that 14-3-3b overexpression

Oncogene Integrin b1 interaction with 14-3-3b DC Han et al 354 its interactions with integrins aect these and potentially other cellular functions.

Materials and methods

Materials Protein A-Sepharose 4B, glutathione-agarose beads, soybean trypsine inhibitor, poly-L-lysine (PLL), and human plasma ®bronectin (FN) were purchased from Sigma. G418 and Figure 8 Eects of inducible expression of 14-3-3b on tyrosine LipofectAMINE were purchased from Life Technologies, phosphorylation. 14-3-3 cells under uninduced (U) or induced (I) Inc. Mouse mAb 12CA5, rabbit polyclonal anti-FAK serum conditions were detached by trypsinization and replated on FN (Chen and Guan, 1994) and rabbit polyclonal anti-integrin (FN) or maintained in suspension (Sus), as indicated. Lysates b1 (Marcantonio and Hynes, 1988) have been described were prepared and analysed by Western blotting with anti-14-3-3b previously. The following antibodies were purchased as (A) or PY-20 (B). Molecular weight markers are shown on the left indicated: mouse mAb anti-integrin b1, P4C10, from Life Technologies, Inc.; mouse mAb antiphosphotyrosine PY-20, mouse mAb anti-FAK and mouse mAb anti-paxillin from Transduction Laboratories (Lexington, KY, USA); and rabbit polyclonal anti-p130cas, rabbit polyclonal anti-14-3-3 (K-19) and rabbit polyclonal anti-14-3-3b (C-20) from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA, USA).

Yeast two-hybrid screen A DNA fragment containing dimerization domain of GCN4 was obtained by PCR using sense (5'-CGCGGATCCAAA- GAATGAAACTTGAAGAC) and antisense (5'-CGCGTC- GACTTCGCCAACTAATTTCTTTAATC) oligonucleotides with pLEX-GCN4 (kind gift of Dr K Struhl) as a template. PCR product was digested with BamHI and SalI, gel puri®ed, and then inserted into the corresponding site of pAS2 (Clontech, CA, USA) to generate a modi®ed bait vector pAS2DD. The cytoplasmic domain of integrin b1 containing residues 748 ± 803 was synthesized by PCR using sense (5'-CGCGTCGACATTGGACTTGCATTGTTATT- GAT) and antisense (5'-GTAAAATCACTGCAGTTTGCC) oligonucleotides with chicken integrin b1 (Solowska et al., 1989) as a template. The PCR product was digested with SalI and PstI, gel puri®ed, and then inserted into the corresponding site of pAS2DD to generate plasmid pAS2DD-b1cyto. The bait vectors encoding deletion mutants

cas D1, D3, and D4 and point mutant Y788F were prepared in a Figure 9 Tyrosine phosphorylation of FAK, p130 , and similar manner using the corresponding mutant integrin b1 paxillin in cells with inducible expression of 14-3-3b. 14-3-3 cells under uninduced (U) or induced (I) conditions were detached by (Marcantonio et al., 1990; Reszka et al., 1992) as templates trypsinization and replated on FN (FN) or maintained in in PCR. Overlap extension method was employed to create suspension (Sus), as indicated. Lysates were immunoprecipitated the S790A mutant using the sense (5'-ACAAGGCCG- by anti-FAK (A), anti-p130cas (B), or anti-paxillin (C) followed by CAGTGACAA) and antisense (5'-AGTTGTCACTGCGG- Western blotting with PY-20 (A±C) or anti-FAK (A), anti- CCTTGTAAA) oligonucleotides, as described previously p130cas, or anti-paxillin (C). Arrows on the right indicate the (Cary et al., 1996). The bait vector encoding integrin b3 positions of FAK, p130cas, and paxillin, respectively containing residues 733 ± 789 was synthesized by PCR using sense (5'-GGGAGCGATCGTCGACATTGGCCTTGC) and antisense (5'-AGGAAGAATCCCTGCAGGAGGCATT) nucleotides with human integrin b3 as a template. pGAD- b3-endonexin was generous gift from Dr Shattil (Shattil et stimulated integrin-mediated cell spreading and migra- al., 1995). tion by a mechanism independent of FAK signaling The HF7c yeast strain was ®rst transformed with pathways. pAS2DD-b1cyto. Expression of the bait protein containing In summary, we have identi®ed a novel interaction integrin b1 cytoplasmic domain was con®rmed by Western between integrin b1 cytoplasmic domain and 14-3-3b blotting of yeast cell lysates with anti-integrin b1 (data not and also provided evidence for a potential role of 14-3- shown). They were transformed subsequently with a HeLa cell cDNA library fused to the GAL4 transcriptional 3b in the regulation of integrin-mediated cell adhesion activation domain ((Hannon et al., 1993); generous gift of and migration. Future studies will be directed at Dr G Hannon). Transformants were plated on agar selection understanding the mechanisms by which 14-3-3b and medium lacking tryptophan (Trp7), leucine (Leu7), and

Oncogene Integrin b1 interaction with 14-3-3b DC Han et al 355 histidine (His7). The resulting colonies were isolated and Immunoprecipitation and Western blotting tested for b-Gal activity and growth on Trp7Leu7His7 plates (Hannon et al., 1993). Plasmid DNA was puri®ed from the For most experiments, subcon¯uent cells were washed twice His+ b-Gal+ colonies. They were then retransformed into with ice-cold PBS and then lysed with 1% NP-40 lysis buer, yeast with dierent bait vectors to determine speci®city and as described previously (Han and Guan, 1999). For some the inserts were sequenced using Sequenase 2.0. experiments, cell lysates were prepared from suspended cells, or cells that had been re-plated on FN using ice-cold modi®ed RIPA buer (50 m Tris pH 7.4, 150 mM NaCl, Construction of cDNA expression vectors M 1% Triton X-100, 1% sodium deoxycholate, 0.1% sodium The 14-3-3b cDNA lacking amino terminal 29 residues (#11) dodecylsulfate, 5 mM EDTA, 30 mM Na2HPO4,50mM NaF, or the unknown cDNA insert (#10) was excised from the 1mM NaVO4,1mM PMSF, 10 mg/ml aprotinin and 20 mg/ prey plasmid pGAD-#11 or pGAD-#10 and cloned into ml leupeptin), as described previously (Han and Guan, 1999). pGEX-KG to generate pGEX-#11 or pGEX-#10, respec- Lysates were cleared by centrifugation and total protein tively. cDNA encoding the integrin b1 cytoplasmic domain concentration was determined using Bio-Rad Protein Assay together with the GCN4 dimerization domain was excised (Hercules, CA, USA). For `pull-down' assays using GST- from the bait plasmid pAS2DD-b1cyto and inserted into fusion proteins, 30 ± 50 ml of glutathione-coupled beads were mammalian expression vector pKH3 to generate pKH3DD- added to lysates and rotated for 30 min at 48C. After b1cyto. Full length 14-3-3b was obtained by PCR using the washing ®ve times with lysis buer, the samples were resolved sense (5'-CGCGGATCCATGACAATGGATAAAAGT- using SDS ± PAGE, and transferred to nitrocellulose mem- GAGC) and antisense (5'-CCGAATTCGAGGGTACA- brane (Schleicher and Schuell, Inc., NH, USA). Immunopre- GAGTGACACTGG) oligonucleotides with an EST clone cipitations were carried out by incubating cell lysates with containing 14-3-3b (EST590806, ATCC) as a template. The appropriate antibodies as indicated for 2 h at 48C, followed PCR product was digested with BamHI and EcoRI, and then by incubation for 1.5 h with protein A-Sepharose. After inserted into mammalian expression vector pDHGST (Han washing, immune complexes were resolved using SDS ± and Guan, 1999) at corresponding site to generate pDHGST- PAGE. Western blotting was carried out using horseradish 14-3-3b. It was also inserted into pKH3 to produce pKH3- peroxidase-conjugated IgG as a secondary antibody and the 14-3-3b. 14-3-3b cDNA fused to the triple hemagglutinin Amersham ECL system for detection. (HA) epitope tag was then excised from pKH3-14-3-3b by digest with SalI and ClaI, and inserted into the corresponding Immunofluorescence staining site in pTet-Splice (Life Technologies, Inc) to creatr pTet- Splice-14-3-3b. Cells were processed for immuno¯uorescence staining as described with minor modi®cation (Guan et al., 1991). Human foreskin ®broblasts (HFF) cells were plated on Cell culture and transfections 18 mm coverslip that was coated with 10 mg/ml human 293T cells and HFF cells were maintained in Dulbecco's plasma FN or 50 mg/ml of PLL. Cells were incubated in a modi®ed Eagle's medium (DMEM) supplemented with 10% 378C incubator to allow cells to attach and spread. At the fetal bovine serum (FBS, Life Technologies, Inc). CHO cells end of incubation, the cells were ®xed with 3% formaldehyde were maintained in F12 medium plus 10% FBS. Transient for 15 min, washed three times, permearized with 0.5% transfections of 293T and CHO cells were performed using Triton X-100, washed three times, and stained with primary Lipofectamine (Life Technologies, Inc) according to the antibodies (a-integrin b1, P4C10, (1 : 100), a-14-3-3b (1 : 200)) manufacturer's guidelines. in the presence of 10% goat serum for 1 h at room NIH3T3 cells were maintained in DMEM with 10% calf temperature. After washing three times, the bound rabbit serum (CS, Life Technologies, Inc.). pTet-Splice-14-3-3b was IgG and mouse IgG were detected with FITC-conjugated transfected into NIH3T3 cells along with the plasmids pTet- anti-rabbit (1 : 300) and rhodamine-conjugated anti-mouse tTAK (Life Technologies, Inc.) and pSV2neo using Lipo- (1 : 150) antibodies, respectively. Image of stained cells was fectAMINE (Life Technologies, Inc.), as described previously captured using an immuno¯uorescence microscope. (Han and Guan, 1999; Zhao et al., 1998). G418-resistant clones with inducible expression of 14-3-3b were then Cell spreading assays identi®ed by Western blotting with 12CA5 of cell lysates prepared under uninduced and induced conditions, as Prior to experiments, cells were incubated for 2 days in described previously (Zhao et al., 1998). Mock cells were medium with (uninduced) or without (induced) 0.5 mg/ml generated in a similar manner with pTet-Splice vector tetracycline. They were collected by trypsinization, washed without insert. The selected cell clones were maintained in once with medium containing 0.5 mg/ml of soybean trypsin DMEM with 10% CS, 0.5 mg/ml G418 and 0.5 mg/ml of inhibitor and washed two more times with medium only. tetracycline to suppress exogenous 14-3-3b expression until Then, cells were maintained in suspension for 1 h and experiments as indicated. replated on dishes that had been coated with 10 mg/ml FN or 50 mg/ml PLL and blocked by 2 mg/ml BSA. Cells were allowed to spread for the indicated times at 378C and then Preparation of GST fusion proteins and in vitro binding assays photographed. Spread cells were de®ned as cells with irregular GST fusion proteins were produced and puri®ed as described morphology and lacking phase-brightness and nonspread cells previously (Xing et al., 1994). GST fusion proteins (5 mg) were rounded and phase-bright under microscope. were immobilized on glutathione-agarose beads and then incubated for 90 min at 48C with lysates (200 mg) prepared Cell migration assays from CHO cells that had been transfected with pKH3DD-b1 cyto. After washing, the bound proteins were analysed by Uninduced or induced cells (16106) were plated on FN Western blotting with 12CA5 (1 : 1000 dilution) as described (10 mg/ml) or PLL (50 mg/ml) coated plates (60 mm) in below. complete growth media in the presence or absence of

Oncogene Integrin b1 interaction with 14-3-3b DC Han et al 356 tetracycline (0.5 mg/ml). After 4 h, wounds were made by DMEM, Dulbecco's modi®ed Eagle's medium; FBS, fetal manual scraping of the cell monolayer with a pipet tip. The bovine serum; CS, calf serum. dishes were washed, replenished with media in the presence or absence of tetracycline (0.5 mg/ml), and photographed under phase contrast microscope (6100). They were then placed in the tissue culture incubator, and the matched wound regions were photographed again 18 h later. Acknowledgements Cell growth measurement We are grateful to Dr Kevin Struhl of Harvard medical Uninduced or induced cells (26105) were plated on 60 mm school for plasmid pLEX-GCN4, Dr Xiao-Ping Du of dishes in growth media with or without tetracycline (0.5 mg/ University of Illinois at Chicago for human integrin b3 ml). Each day after initial plating, cells were trypsinized from cDNA, Dr G Hannon of Cold Spring Harbor Laboratory three dierent dishes and counted for cell numbers under for HeLa cell cDNA library, and Dr Sandy Shattil of either uninduced or induced conditions. Scripps Research Institute for plasmid pGAD-b3-endonexin. We thank Ji He Zhao, Hiroki Ueda, Renee Christopher, Heinz Reiske, Smita Abbi and Tang-Long Shen for their critical reading of the manuscript and helpful comments. Abbreviations This research was supported by NIH grants GM48050 to J- The abbreviations used are: ECM, extracellular matrix; L Guan. J-L Guan is an Established Investigator of the PLL, poly-L-lysine; FN, human plasma ®bronectin; American Heart Association.

References

Aitken A, Jones D, Soneji Y and Howell S. (1995). Biochem. Garcia-Guzman M, Dol® F, Russello M and Vuori K. Soc. Trans., 23, 605 ± 611. (1999). J. Biol. Chem., 274, 5762 ± 5768. Akiyama SK, Yamada SS, Yamada KM and LaFlamme SE. Gilmore AP and Romer LH. (1996). Mol. Biol. Cell., 7, (1994). J. Biol. Chem., 269, 15961 ± 15964. 1209 ± 1224. Aplin AE, Howe A, Alahari SK and Juliano RL. (1998). Guan JL, Trevithick JE and Hynes RO. (1991). Cell Regul., Pharmacol. Rev., 50, 197 ± 263. 2, 951 ± 964. Autieri MV, Haines DS, Romanic AM and Ohlstein EH. Han DC and Guan JL. (1999). J. Biol. Chem., 274, 24425 ± (1996). Cell Growth Dier., 7, 1453 ± 1460. 24430. Bonnefoy-Berard N, Liu YC, von Willebrand M, Sung A, Hannigan GE, Leung-Hagesteijn C, Fitz-Gibbon L, Coppo- Elly C, Mustelin T, Yoshida H, Ishizaka K and Altman A. lino MG, Radeva G, Filmus J, Bell JC and Dedhar S. (1995). Proc. Natl. Acad. Sci. USA, 92, 10142 ± 10146. (1996). Nature, 379, 91 ± 96. Braselmann S and McCormick F. (1995). EMBO J., 14, Hannon GJ, Demetrick D and Beach D. (1993). Genes Dev., 4839 ± 4848. 7, 2378 ± 2391. Cary LA, Chang JF and Guan JL. (1996). J. Cell. Sci., 109, HausserA,StorzP,LinkG,StollH,LiuYC,AltmanA, 1787 ± 1794. P®zenmaier K and Johannes FJ. (1999). J. Biol. Chem., CaryLA,HanDC,PolteTR,HanksSKandGuanJL. 274, 9258 ± 9264. (1998). J. Cell. Biol., 140, 211 ± 221. Horwitz A, Duggan K, Buck C, Beckerle MC and Burridge Chang DD, Wong C, Smith H and Liu J. (1997). J. Cell. K. (1986). Nature, 320, 531 ± 533. Biol., 138, 1149 ± 1157. HughesPE,Diaz-GonzalezF,LeongL,WuC,McDonald Chen HC and Guan JL. (1994). Proc.Natl.Acad.Sci.USA, JA, Shattil SJ and Ginsberg MH. (1996). J. Biol. Chem., 91, 10148 ± 10152. 271, 6571 ± 6574. Chun JS and Jacobson BS. (1993). Mol. Biol. Cell., 4, 271 ± Hynes RO. (1992). Cell, 69, 11 ± 25. 281. Ilic D, Furuta Y, Kanazawa S, Takeda N, Sobue K, Clark EA and Brugge JS. (1995). Science, 268, 233 ± 239. Nakatsuji N, Nomura S, Fujimoto J, Okada M and Clark GJ, Drugan JK, Rossman KL, Carpenter JW, Rogers- Yamamoto T. (1995). Nature, 377, 539 ± 544. Graham K, Fu H, Der CJ and Campbell SL. (1997). J. Irie K, Gotoh Y, Yashar BM, Errede B, Nishida E and Biol. Chem., 272, 20990 ± 20993. Matsumoto K. (1994). Science, 265, 1716 ± 1719. Conklin DS, Galaktionov K and Beach D. (1995). Proc. Khwaja A, Rodriguez-Viciana P, Wennstrom S, Warne PH Natl. Acad. Sci. USA, 92, 7892 ± 7896. and Downward J. (1997). EMBO J., 16, 2783 ± 2793. Craparo A, Freund R and Gustafson TA. (1997). J. Biol. KlemkeRL,LengJ,MolanderR,BrooksPC,VuoriKand Chem., 272, 11663 ± 11669. Cheresh DA. (1998). J. Cell. Biol., 140, 961 ± 972. Fantl WJ, Muslin AJ, Kikuchi A, Martin JA, MacNicol AM, Kolanus W, Nagel W, Schiller B, Zeitlmann L, Godar S, Gross RW and Williams LT. (1994). Nature, 371, 612 ± Stockinger H and Seed B. (1996). Cell, 86, 233 ± 242. 614. LaFlamme SE, Akiyama SK and Yamada KM. (1992). J. Freed E, Symons M, Macdonald SG, McCormick F and Cell. Biol., 117, 437 ± 447. Ruggieri R. (1994). Science, 265, 1713 ± 1716. Liliental J and Chang DD. (1998). J. Biol. Chem., 273, 2379 ± Fu H, Subramanian RR and Masters SC. (2000). Annu. Rev. 2383. Pharmacol. Toxicol., 40, 617 ± 647. LinTH,AplinAE,ShenY,ChenQ,SchallerM,RomerL, Fu H, Xia K, Pallas DC, Cui C, Conroy K, Narsimhan RP, Aukhil I and Juliano RL. (1997). J. Cell. Biol., 136, 1385 ± Mamon H, Collier RJ and Roberts TM. (1994). Science, 1395. 266, 126 ± 129.

Oncogene Integrin b1 interaction with 14-3-3b DC Han et al 357 Liu YC, Liu Y, Elly C, Yoshida H, Lipkowitz S and Altman Richardson A and Parsons T. (1996). Nature, 380, 538 ± 540. A. (1997). J. Biol. Chem., 272, 9979 ± 9985. Schaller MD, Otey CA, Hildebrand JD and Parsons JT. Loftus JC and Liddington RC. (1997). J. Clin. Invest., 100, (1995). J. Cell. Biol., 130, 1181 ± 1187. S77 ± S81. Schwartz MA, Schaller MD and Ginsberg MH. (1995). Loo DT, Kanner SB and Aruo A. (1998). J. Biol. Chem., Annu. Rev. Cell. Dev. Biol., 11, 549 ± 599. 273, 23304 ± 23312. Sharma CP, Ezzell RM and Arnaout MA. (1995). J. Marcantonio EE, Guan JL, Trevithick JE and Hynes RO. Immunol., 154, 3461 ± 3470. (1990). Cell Regul., 1, 597 ± 604. Shattil SJ, Kashiwagi H and Pampori N. (1998). Blood, 91, Marcantonio EE and Hynes RO. (1988). J. Cell. Biol., 106, 2645 ± 2657. 1765 ± 1772. Shattil SJ, O'Toole T, Eigenthaler M, Thon V, Williams M, Masters SC, Pederson KJ, Zhang L, Barbieri JT and Fu H. Babior BM and Ginsberg MH. (1995). J. Cell. Biol., 131, (1999). Biochemistry, 38, 5216 ± 5221. 807 ± 816. Meller N, Liu YC, Collins TL, Bonnefoy-Berard N, Baier G, Solowska J, Guan JL, Marcantonio EE, Trevithick JE, Buck Isakov N and Altman A. (1996). Mol. Cell. Biol., 16, CA and Hynes RO. (1989). J. Cell. Biol., 109, 853 ± 861. 5782 ± 5791. Tahiliani PD, Singh L, Auer KL and LaFlamme SE. (1997). Miranti CK, Ohno S and Brugge JS. (1999). J. Biol. Chem., J. Biol. Chem., 272, 7892 ± 7898. 274, 10571 ± 10581. Toker A, Sellers LA, Amess B, Patel Y, Harris A and Aitken MiyamotoS,TeramotoH,CosoOA,GutkindJS,Burbelo A. (1992). Eur. J. Biochem., 206, 453 ± 461. PD, Akiyama SK and Yamada KM. (1995). J. Cell. Biol., Villaret DB, Wang T, Dillon D, Xu J, Sivam D, Cheever MA 131, 791 ± 805. and Reed SG. (2000). Laryngoscope, 110, 374 ± 381. Morgan A and Burgoyne RD. (1992). Biochem. J., 286, 807 ± Vuori K and Ruoslahti E. (1993). J. Biol. Chem., 268, 811. 21459 ± 21462. Muslin AJ, Tanner JW, Allen PM and Shaw AS. (1996). Cell, Wang B, Yang H, Liu YC, Jelinek T, Zhang L, Ruoslahti E 84, 889 ± 897. and Fu H. (1999). Biochemistry, 38, 12499 ± 12504. Nakanishi K, Hashizume S, Kato M, Honjoh T, Setoguchi Y Wu C, Keivens VM, O'Toole TE, McDonald JA and and Yasumoto K. (1997). Hum. Antibodies, 8, 189 ± 194. Ginsberg MH. (1995). Cell, 83, 715 ± 724. O'Toole TE, Katagiri Y, Faull RJ, Peter K, Tamura R, Xing Z, Chen HC, Nowlen JK, Taylor SJ, Shalloway D and Quaranta V, Loftus JC, Shattil SJ and Ginsberg MH. Guan JL. (1994). Mol. Biol. Cell., 5, 413 ± 421. (1994). J. Cell. Biol., 124, 1047 ± 1059. Yae MB, Rittinger K, Volinia S, Caron PR, Aitken A, O'Toole TE, Loftus JC, Du XP, Glass AA, Ruggeri ZM, Leers H, Gamblin SJ, Smerdon SJ and Cantley LC. Shattil SJ, Plow EF and Ginsberg MH. (1990). Cell Regul., (1997). Cell, 91, 961 ± 971. 1, 883 ± 893. Zhang L, Chen J and Fu H. (1999). Proc. Natl. Acad. Sci. Otey CA, Pavalko FM and Burridge K. (1990). J. Cell. Biol., USA, 96, 8511 ± 8515. 111, 721 ± 729. Zhang SH, Kobayashi R, Graves PR, Piwnica-Worms H and Pasqualini R and Hemler ME. (1994). J. Cell. Biol., 125, Tonks NK. (1997). J. Biol. Chem., 272, 27281 ± 27287. 447 ± 460. Zhang XA and Hemler ME. (1999). J. Biol. Chem., 274, 11 ± Pfa M, Liu S, Erle DJ and Ginsberg MH. (1998). J. Biol. 19. Chem., 273, 6104 ± 6109. Zhao JH, Reiske H and Guan JL. (1998). J. Cell. Biol., 143, Reszka AA, Hayashi Y and Horwitz AF. (1992). J. Cell. 1997 ± 2008. Biol., 117, 1321 ± 1330. Reuther GW, Fu H, Cripe LD, Collier RJ and Pendergast AM. (1994). Science, 266, 129 ± 133.

Oncogene