EphB1 associates with Grb7 and regulates cell migration.

EphB1 is a member of the Eph family of receptor tyrosine kinases that play important roles in diverse biological processes including nervous system development, angiogenesis, and neural synapsis formation and maturation. Grb7 is an adaptor molecule implicated in the regulation of cell migration. Here we report identification of an interaction between Grb7 and the cytoplasmic domain of EphB1 by using Grb7 as a "bait" in a yeast two-hybrid screening. Co-immunoprecipitation was used to confirm the interaction of Grb7 with the cytoplasmic domain of EphB1 as well as the full-length receptor in intact cells. This interaction is mediated by the SH2 domain of Grb7 and requires tyrosine autophosphorylation of EphB1. Furthermore, Tyr-928 of EphB1 was identified as the primary binding site for Grb7. Stimulation of endogenous EphB1 in embryonal carcinoma P19 cells with its ligand ephrinB1 increased its association with Grb7, which is consistent with a role for the autophosphorylation of EphB1. We also found that EphB1 could phosphorylate Grb7 and mutation of either Tyr-928 or Tyr-594 to Phe decreased this activity. Finally, we show that EphB1 could stimulate fibroblast motility on extracellular matrix in a kinase-dependent manner, which also correlated with its association with Grb7. Consistent with this, co-expression of Grb7 with EphB1 further enhanced cell motility, whereas co-expression of the Grb7 SH2 domain abolished EphB1-stimulated cell migration. Together, our results identified a novel interaction between EphB1 with the adaptor molecule Grb7 and suggested that this interaction may play a role in the regulation of cell migration by EphB1.

Grb7 is the founding member of a family of adaptor molecules that also include Grb10 and Grb14. The Grb7 family members share similar structural organizations, including an amino-terminal proline-rich region, a central segment termed the GM 1 (for Grb and Mig) region, which includes a pleckstrin homology domain, and a carboxyl-terminal SH2 domain. Grb7 family members have been shown to interact with a variety of cell surface receptors and other signaling proteins (1,2). Many of these interactions are mediated by the SH2 domain of a Grb7 family member and a phospho-tyrosine motif in the activated receptors/other signaling molecules. These interactions have been proposed to play a role in the regulation of mitogenic signaling pathways. A number of recent studies provided evidence to support a role of Grb10 in cell proliferation and cell survival (3,4), although a similar role for the other family members Grb7 and Grb14 is not clear (2).
The central GM domain of Grb7 family adaptors share significant sequence homology with Mig-10, a Caenorhabditis elegans gene that has been implicated in neuronal migration in C. elegans embryonic development (5,6). This suggests that Grb7 may play a role in the regulation of mammalian cell migration. Indeed, recent studies from our laboratory showed that Grb7 participates in signal transduction pathways in integrin-mediated cell migration (11,12). We found that the SH2 domain of Grb7 could directly interact with focal adhesion kinase (FAK), which is a cytoplasmic tyrosine kinase known to mediate integrin signaling in cell migration (7)(8)(9)(10). Grb7 interaction with FAK required autophosphorylation of FAK at Tyr-397, and this interaction is increased upon integrin-mediated cell adhesion, which stimulates FAK activation and autophosphorylation. Inducible over-expression of Grb7 in NIH3T3 cells enhanced cell migration toward fibronectin, whereas the SH2 domain inhibited cell migration. Association of Grb7 with FAK allowed phosphorylation by FAK, which was shown to be critical in the regulation of cell migration (11,12). It is not clear, however, whether Grb7 interactions with other signaling molecules and/or cell surface receptors also play a role in the regulation of cell migration.
With 15 members by the last count, Eph kinases constitute the largest family of receptor protein tyrosine kinases. According to sequence homology and ligand-binding specificity, they are divided into two subfamilies. EphA kinases bind to ephrinA ligands that are anchored to cytoplasmic membrane through a glycosyl phosphatidylinositol linkage, whereas EphB kinases bind to ephrinB ligands that have a single transmembrane domain. Interestingly, receptor-ligand relationships between Eph kinases and ephrins are not distinct in that both Eph kinases and ephrins can transmit signals to the interior of juxtaposing cells (13)(14)(15)(16). The Eph-ephrin interaction and ensuing bi-directional signaling have been implicated in diverse biological processes including nervous system development, angiogenesis, and neural synapsis formation and maturation (see Refs. 17-20 for recent reviews).
The regulatory functions of Eph kinases have been attributed to repulsive guidance of axon and cell migration in some cases and establishment and remodeling of cell-cell interactions in other situations (17,18,21,22). Recent studies suggested that several Eph receptors could regulate cell adhesion and migration on extracellular matrix (ECM) (23)(24)(25)(26)(27)(28)(29)(30). EphB2 could down-regulate integrin activity, possibly through tyrosine phosphorylation and inactivation of R-Ras (23), a positive regulator of integrin function (24,25). In PC-3 cells, ligand stimulation of endogenous EphA2 caused cell rounding due to integrin affinity down-regulation, concomitant with focal adhesion kinase dephosphorylation (26). Induction of cell de-adhesion was also observed following ligand activation of EphA3 transfected into 293 cells (27,28). The adhesion regulatory function of Eph kinases appears to be cell type-specific; transfection of EphA8 into HEK 293 cell inhibited cell adhesion (29), whereas in NIH 3T3 cells, the similarly transfected EphA8 stimulated integrin-mediated cell adhesion (30). Recently, EphA4 and other EphA kinases were found to associate with ephexin, a guanine nucleotide exchange factor for Rho family small GTPases (31). Upon EphA kinase ligation, ephexin mediates activation of RhoA and suppression of Cdc42 and Rac1, consistent with the repulsive guidance function of Eph kinases.
To further explore the role and mechanisms of Grb7 in cell migration, we used yeast two-hybrid screening to identify novel cellular proteins that bind to Grb7. We report here the identification of the association of EphB1 with Grb7 in an activationdependent manner and the effect of EphB1 expression on cell migration. These results suggest the potential role of Grb7 in mediating the signaling pathways triggered by the EphB1 receptor in cell migration.
Yeast Two-hybrid Screen-cDNA encoding full-length Grb7 or its Pro-GM segment was excised from pKH3-Grb7 or pKH3-Pro-GM (12) and inserted into the bait vector pGBT9 (Clontech) to generate pGBT-Grb7 or pGBT-Pro-GM, respectively. The HF7c yeast strain was first transformed with pGBT-Grb7 and subsequently with a HeLa cell cDNA library fused to the GAL4 transcriptional activation domain (32) (generous gift of Dr. G. Hannon). Transformants were plated on agar selection medium lacking tryptophan (Trp Ϫ ), leucine (Leu Ϫ ), and histidine (His Ϫ ). The resulting colonies were isolated and tested for ␤-galactosidase activity and growth on Trp Ϫ Leu Ϫ His Ϫ plates (32). Plasmid DNA was purified from the His ϩ ␤-galactosidase ϩ colonies. They were then re-transformed into yeast with different bait vectors to determine specificity, and the inserts were sequenced using Sequenase 2.0.
Construction of Expression Vectors-The DNA segment encoding EphB1 cytoplasmic domain (residues 566 -984) was amplified by PCR using 5Ј-CGGAATTCATAGCAGGAAACGGGCTTATAGC (sense) and 5Ј-GTTCATGAATTCCCGGGGATC (antisense) oligonucleotides using pSR␣-hEphB1 as a template. The PCR product was digested with EcoRI and inserted into the corresponding site of pKH3 to generate pKH3-EphB1cyto. Using pKH3-EphB1cyto as a template, overlap extension PCR mutagenesis (33) was employed to generate plasmid pKH3-EphB1cyto-kd encoding the kinase-defective mutant by oligonucleotides 5Ј-GTGGCCATCAGGACCGTGAAGGCA (sense) and 5Ј-CT-TCAGGGTCCTGATGGCCACGTA (antisense). Similar strategies were used to generate point mutations converting Tyr-594 to Phe and Tyr-928 to Phe, using oligonucleotides 5Ј-ATGAAGATCTTCATTGAC-CCCTTC (sense) and 5Ј-GGGGTCAATGAAGATCTTCATCCC (antisense) and 5Ј-AAAATGGTCCAGTTCAGGAACAGCTTC (sense) and 5Ј-GAAGCTGTTCCTGAACTGGACCATTTT (antisense), respectively. cDNA encoding full-length EphB1 with an HA tag fused to its carboxyl terminus was generated by PCR using pSR␣-hEphB1 as a template and oligonucleotides 5Ј-ACGCGTCGACATGGCCCTGGATT-ATCTACTAC (sense) and 5Ј-CGAATTCTCACGGCCGCCACTGAGCA-GCGTA (antisense). The PCR product was digested with SalI and EcoRI and then inserted into the corresponding site in pKH3, which had been digested with SalI and EcoRI removing the sequence encoding the triple HA tag. The resulting plasmid is designed as pKCH-EphB1. This plasmid was then digested with BglII and EcoRI, and the segment encoding the cytoplasmic domain of EphB1 was replaced by the corresponding fragment from pKH3-EphB1cyto-kd, pKH3-EphB1cyto-Y594F, or pKH3-EphB1cyto-Y928F to generate pKCH-EphB1-kd, pKCH-EphB1-Y594F, or pKCH-EphB1-Y928F, respectively.
pHAN is a mammalian cell expression vector containing Myc and His 6 tag under the control of the CMV promoter. DNA fragment encoding Myc-(His) 6 was synthesized by PCR using sense (5Ј-ACGCGTCGA-CATGGAACAAAAACTCATCTCAGAAG) and antisense (5Ј-CGGGAT-CCATGATGATGATGATGATGGTC-3Ј) oligonucleotides and pSecTag2 (Invitrogen) as a template. The PCR product was digested by BamH1, purified from gel, and then ligated with pKH3 that had been digested with SalI followed by fill-in by T7 DNA polymerase and digested with BamH1, resulting in pHAN.
Immunoprecipitation and Western Blotting-For most experiments, subconfluent cells were washed twice with ice-cold PBS and then lysed with 1% Nonidet P-40 lysis buffer (20 mM Tris, pH 8.0, 137 mM NaCl, 1% Nonidet P40, 10% glycerol, 1 mM NaVO 4 , 1 mM phenylmethylsulfonyl fluoride, 10 mg/ml aprotinin, and 20 mg/ml leupeptin). Lysates were cleared by centrifugation for 20 min at 4°C, and total protein concentration was determined using Bio-Rad Protein Assay. GST-Grb7 fusion proteins were precipitated by incubating cell lysates with glutathioneagarose beads for 30 min at 4°C. Similarly, Myc-(His) 6 -Grb7 fusion proteins were precipitated by incubating cell lysates with Ni-beads for 30 min at 4°C. After washing 4 times, complexes were resolved using SDS-PAGE. Western blotting was carried out using horseradish peroxidase-conjugated IgG as a secondary antibody and the Amersham ECL system for detection.
Ligand Stimulation of Endogenous EphB1-Two days after transfection with pKH3-Grb7, P19 cells were incubated for 5 h in Opti-MEM prior to stimulation. Ligand (ephrin B1-Fc) or control (Fc) multimers were generated by preincubation with anti-Fc at a ratio of 1:10 (50 ng/ml of ephrin B1-Fc with 500 ng/ml of anti-Fc) on ice for 1 h. Cells were incubated with pre-clustered ligands at 37°C for 30 min. Lysates were prepared with Nonidet P-40 lysis buffer as described above. To precipitate endogenous EphB1 receptor, ephrin B1-Fc (1 g) complexed with protein A/G beads was added and incubated 1 h at 4°C. After washing three times, complexes were resolved by SDS-PAGE.
Cell Motility Assay-Cell motility assays for transiently transfected CHO cells were performed as described previously (12), except that plasmids encoding GFP-Grb7 or GFP-SH2 were used instead of a vector encoding GFP in some experiments.

RESULTS
To understand the role of Grb7 as an adaptor molecule in the regulation of cell migration, we employed yeast two-hybrid screening to identify protein(s) that can interact with Grb7. A cDNA encoding full-length Grb7 was inserted into pGBT9 bait vector and used to screen a HeLa cell library (32) in the prey vector pGAD. From ϳ2 ϫ 10 6 transformants, several clones were identified that specifically interacted with the Grb7 bait. The specificity of the interactions was confirmed by co-transforming yeast cells with the recovered prey plasmids and pGBT9-Grb7 or pGBT9 vector alone as control. Partial DNA sequencing of the clones indicated that one of the clones encodes part of the human receptor tyrosine kinase EphB1 including the catalytic site in the cytoplasmic domain. This clone is designated as pGAD-EphB1cyto.
A number of tyrosine kinases have been shown to interact with Grb7 through its SH2 domain (2). To determine whether the SH2 domain of Grb7 is also involved in mediating the interaction of Grb7 with EphB1cyto, we tested potential interaction of EphB1cyto interaction with a Grb7 truncation mutant lacking the SH2 domain (designated Pro-GM) in a yeast twohybrid system (Fig. 1). As expected, EphB1cyto interacted with the full-length Grb7 encoded by pGBT-Grb7 but not pGBT9 vector alone. Interestingly, EphB1cyto did not show any interaction with Pro-GM. These results suggested that EphB1cyto also interacted with Grb7 through its SH2 domain.
To verify the yeast two-hybrid results and to further investigate the potential interaction of EphB1 with the Grb7 SH2 domain, a mammalian expression vector encoding the EphB1 cytoplasmic domain (EphB1cyto) with a triple HA tag fused to its amino terminus was generated as described in the "Materials and Methods." Point mutations were introduced into EphB1cyto to generate the kinase-defective (EphB1cyto-kd) and missense mutations (EphB1cyto-Y594F and EphB1cyto-Y928F) in the same expression vector. Tyr-594 and Tyr-928 have been shown to mediate EphB1 interactions with Nck and LMW-PTP, respectively. CHO cells were co-transfected with pKH3 vectors encoding EphB1cyto or its mutants along with pDHGST-Grb7 encoding Grb7 fused to GST at the amino terminus (11). Two days after transfection cells were lysed and GST-Grb7 complexes were precipitated with glutathione-agarose beads. After washing, the bound proteins were analyzed by Western blotting with antibody against EphB1. Fig. 2 shows that the wild type EphB1cyto and EphB1cyto-Y594F mutant associated with Grb7 in CHO cells. In contrast, EphB1cyto-kd was not associated with Grb7, and the EphB1cyto-Y928F mutant showed a significantly reduced association with Grb7. Similar expression levels of EphB1cyto and its mutants were verified by blotting of whole cell lysates with anti-EphB1 antibodies. We noticed that EphB1cyto-kd exhibited an increased mobility compared with wild type EphB1-cyto on the gel, suggesting that EphB1 autophosphorylation affected its mobility on SDS-PAGE. Together, these results demonstrate the interaction of EphB1 with Grb7 and suggest that autophosphorylation of EphB1 at Tyr-928 plays a major role in mediating its interaction with Grb7 through the SH2 domain.
We next examined the association of the full-length EphB1 with Grb7 and its various domains in intact cells. CHO cells were co-transfected with expression vector encoding HA-tagged full-length EphB1 and pDHGST encoding Grb7 or its different segments with GST fused at their amino terminus. Two days after transfection, the lysates were prepared and GST fusion proteins were pulled down using glutathione-agarose beads. The bound proteins were then analyzed by Western blotting with polyclonal anti-HA antibody to detect associated HAtagged EphB1 (Fig. 3A) or with anti-GST to detect the amount of Grb7 or its segments in the complex (Fig. 3B). EphB1 bound to full-length Grb7 and the SH2 domain of Grb7 but not the GM domain, Pro-GM domains, or GST alone control. Although the amount of GST-GM is slightly lower than the Grb7 or SH2 domain alone, comparable amounts of GST-Pro-GM were precipitated. Western blotting of whole cell lysates with anti-HA confirmed similar expression levels of EphB1 in all samples (Fig. 3C). Together, these results indicated that the SH2  domain of Grb7 is sufficient for mediating specific binding of Grb7 to EphB1 in mammalian cells and that other domains are not involved in the interaction. In addition we found that Grb7 did not interact with EphB3, another member of the EphB family, in similar experiments (data not shown), providing further support for the specificity of Grb7 interaction with EphB1.
The above data show clearly that EphB1 could interact with Grb7 in a phosphorylation-dependent manner. However, the biological significance of this interaction can only be established if this interaction is regulated by activation of endogenous EphB1 by its ligand. To investigate this possibility, we employed P19 cells that express endogenous EphB1. The cells were transiently transfected with a mammalian expression vector pHAN-Grb7 encoding a His-tagged Grb7. Two days after transfection, the cells were treated with EphB1 ligand ephrin-B1-Fc or Fc control. Cell lysates were then prepared, and EphB1 complexes were precipitated with ephrin-B1-Fc-protein A/G beads. After washing, the precipitated complexes were resolved on SDS-PAGE and subjected to Western blotting using anti-Grb7 antibody. Fig. 4 shows that endogenous EphB1 associates with Grb7, and treatment of cells with the ligand ephrin B1 further enhanced EphB1/ Grb7 association.
Our previous studies (12) showed that Grb7 could be phosphorylated by FAK, which plays a role in the regulation of cell migration by FAK. To further investigate the potential cellular function of EphB1 binding to Grb7, we examined the possibility of tyrosine phosphorylation of Grb7 by EphB1. We first analyzed the kinase activity of the EphB1 and its mutants because a number of reports suggested that mutation of the corresponding Tyr-594 residue in EphB2 attenuated its catalytic function (34,35). CHO cells were transfected with expression vectors encoding HA-tagged EphB1 or its mutants. Lysates were prepared two days after transfection and analyzed for autophosphorylation by immunoprecipitation with anti-HA followed by Western blotting with anti-phosphotyrosine antibody PY-20. Fig. 5A shows that the wild type EphB1 exhibited strong autophosphorylation, whereas the kinase-defective mutant did not show any autophosphorylation, as expected. The Y594F mutant showed a significant decrease in phosphorylation, whereas the Y928F mutant only has a slightly reduced phosphorylation (top panel). Western blotting of the lysates from parallel samples with anti-HA verified similar expression levels of the constructs (bottom panel). These results suggested that mutation at Y594 also attenuated the kinase activity of EphB1, as in the case of EphB2 reported earlier (34,35).
We then analyzed possible phosphorylation of Grb7 by EphB1 (Fig. 5B). CHO cells were co-transfected with pHAN-Grb7 and pKCH vectors encoding HA-tagged EphB1, its mutants EphB1-kd, EphB1-Y594F, and EphB1-Y928F, or vector alone as a control. Two days after transfection, cell lysates were prepared and Grb7 was precipitated with Ni-beads and analyzed by Western blotting with PY-20. The top panel shows that co-expression of wild type EphB1, but not the kinase-defective EphB1, induced tyrosine phosphorylation of Grb7. Both EphB1-Y594F and EphB1-Y928F mutant also induced Grb7 phosphorylation, although to a lesser extent than the wild type EphB1. Western blotting of the precipitates by anti-Grb7 verified similar amounts of Grb7 in the samples (middle panel).
Because EphB1-Y594 mutant was still capable of binding to Grb7, we investigated the phosphorylation levels of EphB1 and its mutants that were associated with Grb7. Western blotting of Grb7 immunoprecipitates with PY20 revealed that wild type EphB1 was phosphorylated, whereas EphB1-kd was not. Interestingly, EphB1-Y594F was strongly phosphorylated (Fig. 5B,  lower panel), correlating with its ability to bind Grb7 (Fig. 2). The strong phosphorylation of the co-precipitated EphB1-Y594F suggests that Grb7 may target the pool of the EphB1 mutant that was appropriately phosphorylated at sites that are targeted by the SH2 domain of Grb7, including Tyr-928. The reduced phosphorylation of Grb7 by the EphB1-Y594F mutant, relative to the wild type, could be accounted for by its reduced catalytic activity, which is consistent with previous reports for EphB2 mutation at similar site (34,35).
Consistent with the lack of interaction between EphB1-Y928F and Grb7 (Fig. 2), only low levels of tyrosine-phosphorylated EphB1-Y928F were detected in Grb7 precipitates. The low level of Grb7 phosphorylation, despite the relatively intact catalytic activity of EphB1-Y928F, indicates that the physical interaction may be necessary for phosphorylation to occur.
To study whether EphB1 and its association with Grb7 could regulate cell migration, we tested the potential effects of EphB1 and its mutants on cell migration using a time-lapse imaging-based computerized motility analysis method called OMAware, as described previously (12). CHO cells were transiently transfected with expression vectors encoding EphB1 or its mutants along with pEGFP to mark the transfected cells. Two days after transfection, a fraction of cells were lysed and analyzed by Western blotting to verify similar expression levels (data not shown and Fig. 5). The remaining cells were used to evaluate the effects of these constructs on cell motility using OMAware. Fig. 6 shows that expression of wild type EphB1 stimulated CHO cell migration, whereas transfection of control pKH3 vector did not affect cell motility when compared with untransfected cells. Under similar conditions, the kinase-defective EphB1 did not stimulate cell migration. The Y594F mutant exhibits similar activity as the wild type EphB1, whereas the EphB1Y928F mutant shows reduced stimulation of cell migration (about 35% of that by the wild type EphB1).
These results suggested that EphB1 could regulate cell migration in a kinase-dependent manner. In addition, they showed that the ability of EphB1 to stimulate cell migration correlated with its association with Grb7 (see Figs. 2 and 5), suggesting that EphB1 may regulate cell migration through its interaction with Grb7. To test this possibility, CHO cells were co-transfected with expression vector encoding EphB1 and pEGFP-Grb7 or pEGFP-Grb7SH2 encoding full-length Grb7 or the SH2 domain of Grb7, respectively. The cells were then subjected to motility assays using OMAware as described above. As shown in Fig. 6, expression of GFP-Grb7 alone stimulated cell migration as reported previously (11,12). Co-expression of EphB1 with GFP-Grb7 further enhanced cell motility compared with expression of EphB1 with the GFP control or expression of GFP-Grb7 alone. Conversely, co-expression of EphB1 with the SH2 domain of Grb7 abolished EphB1-stimulated cell migration, presumably by disrupting EphB1 interaction with the endogenous Grb7. Together these results provide strong support for a role of EphB1 interaction with Grb7 in the regulation of cell migration.

DISCUSSION
Cell migration is critical for biological processes such as development, immune response, wound healing, and tumor metastasis. Initiation of cell migration involves detection of the gradients of soluble chemoattractants and/or immobilized environmental cues by cell surface receptors. The Eph receptor tyrosine kinases have been implicated to play a key role in the axon path-finding and migration of neurons as well as other cell types (17)(18)(19)36). Increasing evidence suggests that, like other receptor tyrosine kinases, Eph receptors could associate with intracellular signaling molecules upon its activation (17,37). However, the molecular mechanisms by which Eph receptors regulate cell migration is still poorly understood. In this report, we have identified the association of EphB1 with another signaling molecule Grb7. We showed that this interaction is dependent on EphB1 activation and requires Tyr-928 in the cytoplasmic domain of EphB1 and the SH2 domain of Grb7. Furthermore, we found that this interaction plays a role in the regulation of cell migration.
Grb7 was originally isolated as an epidermal growth factor receptor-binding adaptor protein (38), and soon after several other growth factor receptors and intracellular signaling molecules were reported to associate with Grb7 (1, 2). However, the functional significance of these specific interactions is still not completely understood. Recent studies from our laboratory and others (12,39,40) have suggested that Grb7 mediates stimulation of cell migration by integrin-FAK signaling via its interaction with FAK and phosphoinositides. These are the first experimental demonstrations of a specific cellular function for Grb7, which is consistent with previous hypothesis based on the homology of Grb7 with a C. elegans gene product Mig-10 involved in neuronal cell migration in embryonic development.
Our findings in the current study indicated that Grb7 could also couple directly to receptor tyrosine kinases to mediate cell migration. This raises the interesting possibility that a general and specific cellular function of Grb7 is to mediate cell migration triggered by a variety of stimuli, which should be tested in future experiments.
Data presented here suggested that the interaction of Grb7 with EphB1 is mediated by the SH2 domain of Grb7 and the phosphorylated Tyr-928 of EphB1. Although it has been suggested to bind the pYXN motif, the SH2 domain of Grb7 has been shown to associate with phosphorylated tyrosines not present in this motif (11,41,42). Inspection of residues surrounding the Tyr-928 of EphB1 indicated the absence of the motif, therefore providing another exception regarding the preferred binding motifs for the SH2 domain of Grb7. Although it is not clear whether EphB1 could directly phosphorylate Grb7, EphB1 binding to Grb7 through Tyr-928 appears to play a role in its ability to induce Grb7 phosphorylation (Fig. 5) as well as its stimulation of cell migration (Fig. 6).
Previous studies (12,39) suggested that FAK phosphorylation of Grb7 is important for its stimulation of cell migration. However, the Y594F mutant of EphB1 stimulated cell migration to a comparable level as the wild type EphB1 (Fig. 6), FIG. 6. Regulation of cell migration by EphB1 and its interaction with Grb7. CHO cells were co-transfected with vectors EphB1 or its mutants and plasmid encoding GFP fusion protein containing Grb7 or Grb7 SH2 domain or GFP alone, as indicated. They were then subjected to the cell migration assays as described under "Materials and Methods." The mean and standard deviation of relative migration rate (normalized to untransfected cell as 1.0) from three independent experiments are shown. *p ϭ 0.067 and P** ϭ 0.042 in comparison to EphB1ϩGFP transfected cells. although its induction of Grb7 phosphorylation is much lower due to its reduced catalytic activity (Fig. 5). Therefore EphB1 association with Grb7 through Tyr-928, but not the total level of Grb7 phosphorylation by EphB1, is critical for promotion of cell migration. These results, however, do not exclude the possibility that phosphorylation of particular Grb7 sites (e.g. those that are also phosphorylated by FAK that have been shown to correlate with stimulation of cell migration by FAK) by EphB1 and its Y594F mutant is important for promotion of cell migration by EphB1. Although Y594F showed reduced activity and autophosphorylation, it was able to bind Grb7, which may allow it to phosphorylate the "critical" Grb7 site. The wild type EphB1 may phosphorylate both the critical site and other sites of Grb7. The Y928F mutant, on the other hand, may phosphorylate only Grb7 sites other than the critical site (or at least reduced at the critical site) due to its inability to associate with Grb7. It is clear that additional studies to map the phosphorylation sites of Grb7 by EphB1 and FAK are important to further clarify the mechanisms by which Grb7 mediates stimulation of cell migration by EphB1 and FAK.
Previous studies (43,44) have identified the interaction of EphB1 with several intracellular signaling molecules. Indeed, Tyr-928 of EphB1 has been mapped as the putative binding site for Grb10 and LMW-PTP (45,46). Our previous data suggested that unlike Grb7, Grb10 appeared not be involved in the regulation of cell migration despite their sequence homology (12). Although we could not exclude the possibility that the Y928F mutant failed to stimulate cell migration due to its lack of binding to LMW-PTP, our current data strongly support a role of EphB1 interaction with Grb7 in the regulation of cell migration. Nevertheless, it is possible and even likely that the interaction of EphB1 with other signaling molecules at tyr-928 or other sites are also involved in the ability of EphB1 to modulate cell migration. It will be interesting to determine the relative contributions and possible cooperations of these different pathways.
A well documented function of Eph kinases is the repulsive guidance of growth cones, neurons, and other cell types (17,18). Recent studies (31) suggest that activation of RhoA and suppression of Rac1 and Cdc42 may contribute to this process. This is possibly followed by Kuzbenian metalloprotease-mediated cleavage of ephrins to enable separation of cells held together by high affinity Eph-ephrin interactions (47). To effect repulsion, other components of cell motility/regulatory machinery are likely to be impacted as well. Our results demonstrate the EphB1 regulates cell migration through functional interaction with Grb7 that stimulates cell motility possibly by interacting with FAK (12,39). FAK is centrally located in cell migration regulatory pathways, and dynamic regulation of FAK activity is believed to facilitate focal adhesion turnover, which is necessary for productive cell motility. Interestingly, FAK was dephosphorylated and catalytically inactivated by activation of endogenous EphA2 kinase (26). Although the physical and functional interactions between FAK and EphB1 are yet to be determined, it is possible that by interacting with Grb7 and FAK, EphB1 may stimulate cell motility by enhancing focal adhesion turnover.
Regulation of integrin-mediated cell-matrix adhesion is a key component in cell motility (48). Indeed a number of Eph kinases, including EphB3, EphA3, EphB2, and EphA2, have been implicated in integrin affinity down-regulation (23,26,28). EphB1 has been reported to increase cell adhesion to ECM proteins (49). In the latter study, both ECM proteins and ephrins were co-immobilized on the filter, and cell adhesion was then compared with surface coated with ECM alone. Because Eph-ephrin interaction by itself can cause cell tethering (26,50,51), it remains unclear whether the reported increase in adhesion is due to up-regulated integrin affinity or mechanical cell tethering by EphB1-ephrin-B1 interaction. On the other hand, it is possible that Eph kinases may regulate integrin activity in a kinase-and/or cell type-specific manner. Supporting this notion, EphA8 kinase was found by Choi and Park (29) to inhibit cell adhesion in 293 cells but to stimulate adhesion of NIH 3T3 cells (30). We observed that EphB3 did not associate with Grb7 in CHO cells under the same conditions that EphB1 bound to Grb7 (Fig. 3 and data not shown). Further investigations will be necessary to clarify the potential specific effects of different Eph receptors in various cells. Nevertheless, the current report identified a novel interaction between EphB1 with the adapter molecule Grb7 and the potential mechanisms of EphB1 regulation of cell migration.