Identification of Novel SH3 Domain Ligands for the Src Family Kinase Hck

The importance of the SH3 domain of Hck in kinase regulation, substrate phosphorylation, and ligand binding has been established. However, few in vivo ligands are known for the SH3 domain of Hck. In this study, we used mass spectrometry to identify ∼25 potential binding partners for the SH3 domain of Hck from the monocyte cell line U937. Two major interacting proteins were the actin binding proteins Wiskott-Aldrich syndrome protein (WASP) and WASP-interacting protein (WIP). We also focused on a novel interaction between Hck and ELMO1, an 84-kDa protein that was recently identified as the mammalian ortholog of the Caenorhabditis elegansgene, ced-12. In mammalian cells, ELMO1 interacts with Dock180 as a component of the CrkII/Dock180/Rac pathway responsible for phagocytosis and cell migration. Using purified proteins, we confirmed that WASP-interacting protein and ELMO1 interact directly with the SH3 domain of Hck. We also show that Hck and ELMO1 interact in intact cells and that ELMO1 is heavily tyrosine-phosphorylated in cells that co-express Hck, suggesting that it is a substrate of Hck. The binding of ELMO1 to Hck is specifically dependent on the interaction of a polyproline motif with the SH3 domain of Hck. Our results suggest that these proteins may be novel activators/effectors of Hck.

The members of the Src family of nonreceptor tyrosine kinases share a modular structure comprising unique SH3, SH2, and kinase catalytic domains (1)(2)(3)(4). The enzymatic activity of Src family kinases is tightly regulated by intramolecular interactions. The autoinhibited state is maintained by two interactions, (i) an interaction between the SH2 domain and a phosphorylated C-terminal tyrosine (Tyr-527) and (ii) an interaction between the SH3 domain and a polyproline type II helix in the SH2 kinase linker sequence (residues Pro-244 -Trp-254) (5)(6)(7). Src kinases can be potently activated by exogenous ligands for the SH3 and SH2 domains (8 -11). These ligands disrupt the autoinhibitory interactions, promote autophosphorylation at Tyr-416 within the activation loop, and stimulate tyrosine kinase activity.
Binding of Src family kinases to cellular proteins can regulate kinase activity by at least three mechanisms, (i) release of autoinhibitory interactions as described above, (ii) relocalization of Src kinases to sites of cellular action, and (iii) tethering the kinases to potential substrates (1)(2)(3). These mechanisms often operate in combination, as is seen when Src kinases are targeted to their substrates by SH3/SH2 domain interactions. Many cases have been described in which Src kinases are recruited to their substrates via SH3 domain interactions. For example, c-Src is known to associate with at least eight substrates via direct binding of its SH3 domain to ligand binding motifs in the substrates (2); well studied cases include Cas (12), FAK (13), and AFAP-110 (14). The polyproline motifs in such substrates strongly activate Src kinases by SH3 displacement and concomitantly tether the substrate to the kinase, facilitating phosphorylation (13,15,16). The dual role of the SH3 domain in substrate targeting is illustrated in experiments with synthetic peptides, where phosphorylation of substrates is greatly enhanced by the incorporation of an SH3 domain ligand (17). These observations suggest that good substrates for particular Src kinases might be discovered as SH3 domain ligands.
Hematopoietic cell kinase (Hck) is a Src family tyrosine kinase that is expressed predominantly in granulocytic and monocytic cells (18,19). In granulocytes, Hck is found in a secretory granule-enriched fraction and in a granule-free membrane fraction (20). The different subcellular localizations of Hck are consistent with its proposed functional roles in phagocytosis (21)(22)(23)(24) and in receptor-mediated signaling (25)(26)(27)(28). In monocytes, Fc␥RII clustering leads to an array of biological responses including phagocytosis, cell killing, secretion of inflammatory mediators, and activation (25). Clustering of Fc␥RII promotes activation of Hck and subsequent phosphorylation of the receptor by direct association of Hck and Fc␥RII (26). Association of Hck with an acidic region of the interleukin 6 receptor ␤ chain, gp130, leads to Hck activation and cell proliferation (29).
There are several indications that Hck plays an important role in integrin-mediated signal transduction. Neutrophils isolated from Hck/Fgr double knockout mice are deficient in integrin-mediated respiratory burst and granule secretion (30 -32). Integrin signaling in polymorphonuclear leukocytes also activates Hck (33) and leads to increased association of Hck and with the cytoskeleton (34). Another role for Hck in actin rearrangement is suggested by the fact that Hck is activated during E-selectin-mediated induction of monocyte chemotaxis (35).
Hck has been implicated in a wide variety of signaling pathways in hematopoietic cells. However, relatively few substrates or effectors of Hck have been identified. One in vivo substrate for Hck, the multidomain signaling protein Cbl, has been shown to interact with the SH3 domain of Hck. As described above, we would expect that many of the best cellular substrates for Hck would contain binding motifs for the SH3 domain of the enzyme. The focus of this study was to identify and characterize novel SH3 domain ligands for Hck. Two proteins that we identified, Wiskott-Aldrich syndrome protein (WASP) and WASP-interacting Protein (WIP), are known regulators of the actin cytoskeleton (36,37). A third Hck-interacting protein, ELMO1, has recently been described as a component of signaling pathways that regulate phagocytosis and cell migration (38). Hck has been implicated in these pathways as well, suggesting that ELMO1 may represent an important downstream substrate/effector of Hck.

MATERIALS AND METHODS
Expression and Purification of GST and GST-SH3(Hck)-GST and GST-SH3(Hck) were expressed in Escherichia coli NB42 cells. Cells were lysed in a French pressure cell in buffer containing 50 mM HEPES, pH 7.4, 100 mM NaCl, 100 mM EDTA, 1% Triton X-100, 10% glycerol, and protease inhibitors (5 mg/liter of aprotinin, 5 mg/liter leupeptin, 0.1 mM phenylmethylsulfonyl fluoride). Cell lysates were centrifuged, and the supernatants were added to glutathione-agarose (Molecular Probes). After a 1-h incubation at 4°C the beads were washed 5 times with buffer containing 50 mM HEPES, pH 7.4, and 100 mM EDTA. Glutathione-agarose with immobilized GST or GST-SH3 was used directly in ligand binding experiments. The concentration of GST or GST-SH3 on the beads was determined by treating 50 l of beads with 20 mM glutathione in 50 mM Tris, pH 8. The total amount of protein eluted was determined by the Bradford method (Bio-Rad) and divided by 50 l to calculate the total concentration on the beads. Control beads were added to dilute the protein concentration to 1 mg/ml.
Identification of Hck-SH3 Domain-associated Proteins-U937 cells were maintained in RPMI 1640 medium supplemented with 5% fetal bovine serum. Before the pull-down experiment, 800 ml of U937 cells at a density of 10 6 cells/ml were treated with 10 ng/ml phorbol 12-myristate 13-acetate. After 48 h, cells were harvested and then washed 2ϫ with phosphate-buffered saline. Cells were lysed for 30 min with rocking in 4 ml of lysis buffer containing 1% Triton X-100, 10 mM Tris, pH 7.5, 150 mM NaCl, 5 mM EDTA, 2 mM sodium orthovanadate, and protease inhibitors (5 mg/liter aprotinin, 5 mg/liter leupeptin, 0.1 mM phenylmethylsulfonyl fluoride). After centrifugation, a 1-ml portion of this lysate (containing 5 mg of protein) was added to 50 l of glutathione-agarose containing 50 g of GST-SH3. The remaining portion (3 ml, containing 15 mg protein) was added to 50 l of glutathione-agarose containing 50 g of GST. Beads and lysate were rocked 2 h at 4°C and then washed 5 times with 10 ml of lysis buffer. Bound proteins were eluted in 50 l of gel loading buffer by boiling.
Proteins that eluted from the GST-SH3(Hck) affinity column were separated by one-dimensional SDS-PAGE and visualized with colloidal Coomassie stain. Protein bands specific to the Hck-SH3 pull-downs were excised and digested with trypsin as previously described (39). According to the intensity of the Coomassie band, 1/20 to1/3 of each unfractionated tryptic digest was analyzed by on-line liquid chromatography-electrospray tandem mass spectrometry using a micro-column (75 m ϫ 10 cm, 3-m particles) reverse phase HPLC 1 interfaced to an LCQ Deca ion trap mass spectrometer. Electrospray tandem mass spectrometry-based sequencing was performed on line in a data-dependent manner (40) as peptides eluted from the HPLC. Uninterpreted spectra were searched for protein matches against a non-redundant protein data base using the program Mascot (41).
Similar experiments were carried out to measure binding between immobilized GST-SH3(Hck) and purified proteins or proteins in Chinese hamster ovary (CHO) cell lysates, with the exception that binding reactions contained 10 l of glutathione-agarose beads. In some experiments, various concentrations of polyproline-containing peptides were added together with the immobilized SH3 domain.
Protein Expression and Purification-C-terminally phosphorylated wild type Hck, WIP, and ELMO1 were produced in Spodoptera frugiperda (Sf9) cells using baculovirus expression vectors. Hck was ex-pressed and purified as described previously (16). The cDNA for WIP was the kind gift of Dr. Narayanaswamy Ramesh (Harvard Medical School). The WIP-coding sequence was amplified by PCR and subcloned into the baculovirus expression vector pFastBac-Htb (Invitrogen). The cDNA for ELMO1 was obtained as clone DKFZp434B0819 from the German Cancer Research Center. We amplified the 2184-base pair DNA sequence by PCR and ligated it as an EcoRI/KpnI fragment into the baculovirus expression vector pFastBac-Htb.
WIP and ELMO1 were expressed in Sf9 cells using the Invitrogen Bac-to-Bac system. Cells expressing His-tagged WIP and ELMO1 were lysed in a French pressure cell in buffer A (20 mM Tris, pH 8.5, 10% glycerol, 5 mM ␤-mercaptoethanol) containing protease inhibitors (5 mg/liter of aprotinin, 5 mg/liter leupeptin, 0.1 mM phenylmethylsulfonyl fluoride, and 1 mM EDTA). Cell lysate was diluted to 200 ml with column loading buffer (20 mM Tris, pH 8.5, 5% glycerol, 5 mM ␤-mercaptoethanol, 1 M NaCl, and 20 mM imidazole) and centrifuged. Lysate was passed over a 5-ml nickel nitrilotriacetic acid Superflow column (Qiagen). WIP and ELMO1 were eluted with 100 mM imidazole. The concentration of ELMO1 and WIP were determined using the Bradford method (Bio-Rad).
Synthetic Peptides-The substrate peptide for the coupled assay (AEEEIYGEFEAKKKKG (42)) was prepared by solid phase synthesis on an Applied Biosystems automated 431A Peptide Synthesizer. It was purified by reverse phase high pressure liquid chromatography and characterized by matrix-assisted laser desorption ionization time-offlight mass spectrometry. The polyproline-containing peptide (DFPLG-PPPPLPPRATPSR (43)) was the generous gift of Dean Edwards (University of Colorado).
Protein Kinase Assay-Kinase assays were performed by a coupled spectrophotometric assay (44). In this assay, the production of ADP is coupled to the oxidation of NADH measured as a reduction in absorbance at 340 nm. All experiments were carried out at 30°C. Reactions were performed in buffer containing 100 mM Tris, pH 7.5, 10 mM MgCl 2 , 1 mM phosphoenolpyruvate, 0.28 mM NADH, 89 units/ml pyruvate kinase, and 124 units/ml lactate dehydrogenase. The assays contained 400 M peptide substrate and 10 nM Hck.
Expression in CHO Cells and Immunoprecipitation Experiments-The ELMO1 cDNA was subcloned into two expression vectors to generate epitope-tagged versions of ELMO1 for mammalian expression, 1) pEF1/V5-HisA (Invitrogen) for a C-terminal V5 tag and 2) pCM45 (gift of Steve Lang and Pat Hearing, SUNY Stony Brook) for an N-terminal M45 tag. Full-length human Hck cDNA was subcloned into plasmid pCDNA6 (Invitrogen). Two 150-mm tissue culture plates were transfected for each condition tested. Transfections were carried out with 3.5 l of TransIt (Panvera)/g of DNA according to the manufacturer's protocol. After 40 -48 h, cells from the two plates were combined and washed 2ϫ in phosphate-buffered saline. Cells were lysed in lysis buffer (1% Triton X-100, 10 mM Tris, pH 7.5, 150 mM NaCl, 5 mM EDTA, 2 mM vanadate, and protease inhibitors) and clarified by centrifugation, and the protein concentration was determined. An equal amount (between 1 and 5 mg) of total cell protein was diluted to 1 ml for each reaction and precleared with 50 l of protein A-or G-Sepharose beads for 1 h at 4°C with rocking. After pre-clearing, 2 g of appropriate antibody (or 15 l of serum containing M45 antibody) or control antibody was added to lysate and incubated overnight at 4°C with rocking. Antibodies used were mouse anti-M45 (P. Hearing, SUNY Stony Brook), mouse anti-Hck (Transduction Laboratories), rabbit anti-Hck (Santa Cruz), rabbit anti-WIP (a gift of Dr. Narayanaswamy Ramesh, Harvard Medical School), rabbit anti-WASP (Santa Cruz), and mouse anti-V5 (Invitrogen). Antibody-protein complexes were collected with 50 l of protein Aor G-Sepharose beads for 1 h at 4°C with rocking. The beads were washed 5 times in lysis buffer and boiled in 40 l of gel loading buffer. After separation by SDS-PAGE, proteins were transferred to PVDF membranes. Incubation with primary antibody was carried out according to the manufacturer's protocol. After washing, the appropriate horseradish peroxidase-conjugated secondary antibody was added, and proteins were detected using the enhanced chemiluminescent detection kit (Amersham Biosciences).

Screening for Binding Partners of the SH3 Domain of Hck-
Hck is expressed predominantly in myeloid cells such as monocytes (18). Furthermore, differentiation of monocytes to macrophages is associated with increased Hck expression and activity (18). We reasoned that the more differentiated phenotype would also be accompanied by increases in specific binding complexes between Hck and other signaling proteins. There-fore, we treated U937 monocytes with phorbol 12-myristate 13-acetate for 48 h to induce the cells to differentiate to a more macrophage-like phenotype (45). Clarified lysate from these cells was added to glutathione-agarose beads containing immobilized GST-SH3(Hck). As a control, 3-fold more U937 lysate was added to beads containing GST alone. After extensive washing, bound proteins were eluted by boiling in Laemmli buffer and then subjected to SDS-PAGE and visualized with colloidal Coomassie stain (Fig. 1). All visible bands were excised from the GST-SH3(Hck) lane. Bands in the control lane having similar molecular weight and intensity as bands in the GST-SH3(Hck) lane were also excised. The proteins in the bands were reduced, alkylated, and digested overnight with trypsin. The tryptic digests were analyzed by on-line liquid chromatography-electrospray tandem mass spectrometry, and the uninterpreted data was searched against the non-redundant protein data base using the program Mascot (41). Proteins identified in the GST-SH3(Hck) lane but not the control lane are presented in Table I. The table also briefly summarizes what is known about the biological function of the interacting proteins and lists whether they have previously been shown to bind to any Src family kinase.
Two proteins have previously been shown to interact with the SH3 domain of Hck, Cbl (46) and Sam-68 (47). Our approach identified both of these binding partners (Fig. 1, Table  I). In multiple experiments, the most intensely staining band was in the range of 50 -55 kilodaltons (Fig. 1). The components of this band were found to be the actin-associated proteins WASP and WIP. WASP has previously been shown to interact with the SH3 domains of Src, Fyn, and Fgr (48,49), whereas WIP has not previously been described to bind to Src kinases.
Both of these proteins contain Pro-rich regions that would be predicted to bind to SH3 domains (Fig. 2). We confirmed the association of WIP and WASP with the SH3 domain of Hck in phorbol 12-myristate 13-acetate-treated U937 cells as well as in THP-1 human leukemia cells. Cell lysates were incubated with immobilized GST-SH3(Hck) or GST as a control, then bound proteins were separated by SDS-PAGE and detected by Western blotting with anti-WASP or anti-WIP antibodies. In both cell lines, WIP and WASP were associated with GST-SH3(Hck), and they did not associate with GST alone (Fig. 3).
Our initial Hck SH3 domain pull-down experiments were performed on U937 whole cell lysates. The interacting proteins identified such as WASP and WIP may interact directly with the SH3 domain via polyproline motifs or, alternatively, might interact indirectly via other proteins present in the lysates. WASP has previously been shown to interact directly with Src kinase SH3 domains (48,50). To show that the binding between WIP and the SH3 domain of Hck was direct, we carried out a binding experiment using purified protein components. We produced WIP in Sf9 cells by infecting cells with a recombinant baculovirus encoding His-tagged WIP. After purification of WIP on nickel nitrilotriacetic acid resin, we mixed various amounts of WIP with immobilized GST-SH3(Hck). After washing, we eluted the bound proteins and analyzed them by SDS-PAGE followed by Western blotting with anti-WIP antibody. WIP was able to form a complex with the SH3 domain of Hck in this experiment (Fig. 4A).
Binding of a polyproline-containing ligand to the SH3 domain of Hck can lead to kinase activation (8). To test whether WIP could activate Hck in vitro, we expressed and purified the down-regulated form of Hck from Sf9 cells. We added increas-FIG. 1. Identification of proteins in U937 cell lysates that bind to GST-SH3(Hck). Phorbol 12-myristate 13-acetate-treated U937 cells (800 ϫ 10 6 ) were lysed in a volume of 4 ml. A 3-ml portion of the lysate was added to glutathione-agarose beads containing 50 g of GST, and the remaining 1 ml was added to beads containing 50 g of GST-SH3(Hck). After 30 min of incubation, beads were washed 5 times in lysis buffer, and bound proteins were eluted by boiling in Laemmli buffer. Bound proteins were separated by SDS-PAGE. Bands from SDS-PAGE gel were reduced, alkylated, and digested overnight with trypsin. Tryptic digests were analyzed by on-line liquid chromatography-electrospray tandem mass spectrometry. The sequences were searched against the protein data base. Selected proteins are indicated to the right of the gel, and the complete list of proteins identified is given in Table I. ing concentrations of purified WIP to Hck and followed kinase activity using a continuous spectrophotometric assay. Micromolar concentrations of WIP activated Hck under these conditions (Fig. 4B).
ELMO1 Binds Directly to the SH3 Domain of Hck via Polyproline Interactions-Among the novel binding partners for Hck we discovered was an Ϸ85-kDa protein that matched the sequence of a hypothetical protein identified in an effort to analyze 500 novel human cDNAs (51). We obtained the clone from the German Cancer Research Center. While this work was in progress, the protein was discovered in a different context and designated ELMO1 (Table I). ELMO1 is an 84-kDa protein that was identified as the mammalian ortholog of the Caenorhabditis elegans gene, ced-12 (38). CED-12 is required for the engulfment of dying cells and for cell migration (38). In mammalian cells, ELMO1 interacts with Dock180 as part of the CrkII/Dock180/Rac pathway responsible for phagocytosis and cell migration (38). The amino acid sequence of ELMO1 shows little homology to other proteins in the data base, with the exception of the C terminus, which contains a putative pleckstrin homology (PH) domain and a Pro-rich sequence (Fig.  2). In the next series of experiments, we focused in more detail on the interaction between ELMO1 and the SH3 domain of Hck.
To determine whether ELMO1 binds directly to the SH3 domain of Hck, we carried out binding experiments with purified proteins. We expressed ELMO1 as a His-tagged protein using the Sf9/baculovirus system. Chromatography on nickel nitrilotriacetic acid affinity resin yielded a highly purified preparation of ELMO1 (data not shown). We then mixed various amounts of ELMO1 with immobilized Hck SH3 domain. After washing, we eluted the bound protein and analyzed by SDS-PAGE followed by Western blotting with anti-His-tag antibody. ELMO1 bound to the Hck SH3 domain in these experiments (Fig. 5). We conclude that ELMO1 and Hck can interact directly in the absence of any other binding partners. Furthermore, the interaction between ELMO1 and the SH3 domain of Hck was reduced when experiments were carried out in the presence of a polyproline-containing peptide that binds to Src kinase SH3 domains (data not shown; also see Fig. 7).
ELMO1 Associates with Hck in Intact Cells-We constructed mammalian expression vectors for ELMO1 that contained either a C-terminal V5 epitope tag or an N-terminal M45 epitope tag. To determine whether Hck and ELMO1 interact in intact cells, we co-expressed Hck and V5 epitope-tagged ELMO1 in CHO cells. We isolated Hck from CHO cell lysates by immunoprecipitation and analyzed the co-immunoprecipitating proteins by anti-V5 Western blotting. V5-ELMO1 coimmunoprecipitated with Hck in these experiments but was not detected in control immunoprecipitation reactions (Fig. 6). We obtained similar results when Hck and ELMO1 were overexpressed in Cos-7 cells (data not shown).
Next, we confirmed that the Hck-ELMO1 association was dependent on polyproline interactions. For these experiments, we expressed ELMO1 in CHO cells with an N-terminal M45 epitope tag. We then incubated CHO cell lysates with immobilized GST-SH3(Hck). To individual pull-down reactions we added increasing amounts of a polyproline peptide, which binds to the SH3 domain of Hck (43). Binding of ELMO1 to the SH3 domain of Hck was nearly abolished in the presence of 500 M peptide (Fig. 7). This suggests that the primary mode of interaction of ELMO1 with Hck is via the interaction of the SH3 domain with the polyproline region of ELMO1.
We carried out experiments to test for tyrosine phosphorylation of ELMO1 in cells overexpressing Hck. We expressed M45-tagged ELMO1 in the presence or absence of Hck, isolated ELMO1 by immunoprecipitation, then carried out anti-phosphotyrosine immunoblotting. ELMO1 is heavily tyrosine-phosphorylated in cells co-expressing Hck but not in cells expressing ELMO1 alone (Fig. 8). These results show that Hck either phosphorylates ELMO1 directly or activates a pathway leading to tyrosine phosphorylation of ELMO1. DISCUSSION Although Hck is known to play an important role in the function of hematopoietic cells, few in vivo binding partners for Hck have been discovered. Because Src kinases form stable complexes with many preferred substrates via their SH3 domains, we performed a screen using the SH3 domain of Hck to  4. A, in vitro association between WIP and HckSH3. Various amounts of purified WIP were added to pull-down reactions containing 10 l of immobilized GST-HckSH3 (first four lanes). After washing and elution, bound proteins were separated by SDS-PAGE, transferred to PVDF membranes, and visualized by Western blotting using anti-Histag antibody (Santa Cruz). In the lane marked WIP 2 g of purified WIP was loaded directly onto the gel. In the last lane, results are shown from a control pull-down reaction with GST using 30 g of purified WIP. B, activation of down-regulated Hck by WIP. Various amounts of purified WIP were added to kinase reactions containing down-regulated Hck and a synthetic peptide substrate. Phosphorylation of the peptide substrate was measured by a spectrophotometric assay. The fold activation is expressed relative to a control without WIP. isolate proteins from U937 monocytic cells. We then subjected proteins that bound to the SH3 domain of Hck to sequencing by mass spectrometry. We identified proteins known to bind to Hck (Cbl and Sam68), proteins known to interact with other Src family kinases (Sos, ASAP1, heterogeneous nuclear ribonucleoprotein K, CMS/CD2-associated protein, p85, and SLP-76), and a large number of proteins that have not previously been reported to interact with Src kinases (Table I). Unlike yeast two-hybrid experiments, which detect primarily direct interactions, these experiments can also detect secondary and long range interactions. Thus, some of the identified proteins may be components of Hck signaling complexes that do not contact Hck directly.
Because Hck is involved in integrin-mediated signal transduction in macrophages, we expected that actin-associated proteins might be binding partners for Hck. Two prominent bands that associated with the SH3 domain of Hck were WASP and WIP (Fig. 1). WIP and WASP are both well characterized as proteins required for actin polymerization (36,37). WASP was identified as the gene responsible for the Wiskott-Aldrich syndrome, a disease characterized by decreased size and number of platelets and lymphocytes (37). WIP was originally cloned from a human lymphoma T cell line library as a WASP-interacting protein (52). WIP is extremely glycine-and proline-rich and is known to regulate actin assembly in association with WASP (53). Although WIP and WASP are known to associate directly with each other, the data presented here along with data in the literature suggest they both can interact with SH3 domains in an independent manner. WIP and WASP both have multiple copies of a polyproline-containing SH3 domain ligand motif (Fig. 2). WASP had previously been shown to bind and activate the Src family kinase Fyn in vitro (48,49). WIP has not previously been shown to interact with Src family tyrosine kinases. We expressed and purified WIP and show that it binds and activates Hck in vitro (Fig. 4).
WASP may be a substrate for Hck in vivo. One previous report showed that WASP was a substrate for the Src family kinase, Lyn, in response to aggregation of the high affinity IgE receptor at the surface of mast cells (54). The site(s) phosphorylated by Lyn in this study was not identified. The role of WASP phosphorylation is not clear, but phosphorylation may be involved in regulating conformational changes in WASP. WASP has two dramatically different conformations (55). In the autoinhibited conformation, the GTPase binding domain interacts with the C-terminal region. Binding of the Rho family GTPase, Cdc42, can relieve this autoinhibition and release the C terminus to interact with actin regulatory machinery (56). Phosphorylation may play a role in stabilizing one or both of the conformations (55). We have confirmed that WASP and Hck interact when overexpressed in CHO cells. 2 Preliminary experiments in this model system also indicate that WASP is tyrosine-phosphorylated in Hck-expressing CHO cells.
One of the binding partners for the SH3 domain of Hck that we identified in our screen was a previously unidentified protein of 84 kDa (Table I). While this work was in progress, the 84-kDa protein was identified in another context and designated ELMO1 (38,57,58). ELMO1 is the mammalian ortholog of the C. elegans gene, ced-12. CED-12 is required for the engulfment of dying cells and cell migration. In mammalian cells, ELMO1 interacts with Dock180 as part of the CrkII/ Dock180/Rac pathway responsible for phagocytosis and cell migration (38,57,58). ELMO1 is ubiquitously expressed, although its expression was highest in the spleen, an organ rich in immune cells (38). ELMO1 has a PH domain and a polyproline sequence motif at its C terminus (Fig. 2). Our experiments suggest that binding between Hck and ELMO1 depends on SH3-polyproline interactions (Figs. 5 and 7). The remainder of the amino acid sequence of ELMO1 does not show significant homology to other proteins in the data base except for a short region with homology to the Drosophila voltage-gated K ϩ channel Shab.
The importance of Hck in cell migration (28,30,35,59,60) and phagocytosis (21)(22)(23)(24) is well established. Macrophages from Hck-deficient mice show impaired phagocytosis (61). There are a number of points at which Hck could influence the CrkII/Dock180/ELMO1-signaling pathway. Integrin receptors are important for the phagocytosis of apoptotic cells. Integrin signaling has been shown to recruit the p130 cas ⅐CrkII⅐Dock180 molecular complex, which in turn triggers Rac1 activation and phagosome formation (62). In these studies, an unidentified tyrosine kinase was necessary for recruitment of the Cas⅐CrkII⅐Dock180 complex. ELMO1 is likely to be recruited as a component of this complex, and Hck could be involved in either the assembly or disassembly of the complex.  Hck is also known to be involved in Fc␥RII-mediated phagocytosis in monocytes, raising the possibility that a CrkII⅐Dock180⅐ELMO1 complex plays a role in this process. ELMO1 may be an important factor in activating Hck signaling in the cytoplasm after Fc␥RII receptor stimulation. We demonstrated that ELMO1 is tyrosine-phosphorylated in Hck-expressing cells (Fig. 8). ELMO1 might therefore act as a downstream substrate of Hck in this signaling pathway, and phosphorylation of ELMO1 by Hck might modulate the ability of ELMO1 to cause changes in cell motility or phagocytosis.