The Protein-tyrosine Phosphatase SHP-1 Associates with the Phosphorylated Immunoreceptor Tyrosine-based Activation Motif of Fc (cid:1) RIIa to Modulate Signaling Events in Myeloid Cells*

Fc (cid:1) RIIa is a low affinity IgG receptor uniquely expressed in human cells that promotes phagocytosis of immune complexes and induces inflammatory cytokine gene transcription. Recent studies have revealed that phagocytosis initiated by Fc (cid:1) RIIa is tightly controlled by the inositol phosphatase SHIP-1, and the protein-tyrosine phosphatase SHP-1. Whereas the molecular nature of SHIP-1 involvement with Fc (cid:1) RIIa has been well studied, it is not clear how SHP-1 is activated by Fc (cid:1) RIIa to mediate its regulatory effect. Here we report that Fc (cid:1) RIIa clustering induces SHP-1 phosphatase activity in THP-1 cells. Using synthetic phosphopeptides, and stable transfectants expressing immunoreceptor tyro-sine-based activation motif (ITAM) tyrosine mutants of Fc (cid:1) RIIa, we demonstrate that SHP-1 associates with the phosphorylated amino-terminal ITAM tyrosine of Fc (cid:1) RIIa, whereas the tyrosine kinase Syk associates with the carboxyl-terminal ITAM tyrosine. Association of SHP-1 with Fc (cid:1) RIIa ITAM appears to suppress total cellular tyrosine phosphorylation. Furthermore,

IgG receptors (Fc␥R) on monocytes and macrophages mediate immune complex clearance by a process termed phagocytosis (1). At least four classes of Fc␥R are expressed on monocytes and macrophages (2); Fc␥RI, Fc␥RIIa, and Fc␥RIIIa are all activating receptors that are associated with immunoreceptor tyrosine-based activation motif (ITAM). 1 In contrast, Fc␥RIIb is an inhibitory receptor that is associated with an immunoreceptor tyrosine-based inhibition motif (ITIM). Of these receptors Fc␥RIIa is uniquely expressed in human cells and is the only ITAM-associated receptor that bears the ITAM within its cytoplasmic tail (3,4). Of the ITAMs identified to date, the ITAM of Fc␥RIIa has the longest spacer region between the two YXXL motifs that together make the ITAM. The functional significance of this extended spacer is not fully understood. In addition, Fc␥RIIa is the most widely expressed Fc␥R in the human hematopoetic system.
Clustering of Fc␥R by immune complexes initiates a cascade of signaling events, the first of which is the activation of the Src family of tyrosine kinases that phosphorylate the ITAMs of Fc␥R (5,6). The phosphorylated ITAMs serve as docking sites for SH2 domain-containing cytosolic enzymes and enzymeadapter complexes including the tyrosine kinase Syk and the p85 adapter subunit of PtdIns 3-kinase (7). Association of Syk with the phosphorylated ITAM activates the enzyme resulting in autophosphorylation of Syk and tyrosine phosphorylation of multiple cytosolic proteins (8,9). Likewise, association of p85 with the ITAM delivers PtdIns 3-kinase to the proximity of its lipid substrates in the membrane, resulting in the generation of 3Ј-phosphorylated inositol lipids that activate PH domaincontaining enzymes to promote cytoskeletal changes required for the phagocytic process (10). Inactivation of either Syk or PtdIns 3-kinase has been shown to completely abrogate Fc␥Rmediated phagocytosis (11)(12)(13)(14).
Phagocytosis is a complex process that is accompanied by the generation of reactive oxygen radicals and the production of inflammatory cytokines, which results in tissue damage. Therefore the phagocytic process is subject to a tight regulation. In this regard, several mechanisms have been proposed including the expression and function of the inhibitory receptor Fc␥RIIb (15)(16)(17), the function of intracellular inhibitory phosphatases such as the inositol phosphatases SHIP-1 (18 -20) and SHIP-2 (21), and the protein-tyrosine phosphatase SHP-1 (22). Recent studies have revealed that the inositol phosphatases SHIP-1 and SHIP-2 not only work through the ITIM of Fc␥RIIb, but are also capable of associating with the ITAMs of Fc␥R to modulate activation events, thus providing an additional level of complexity to the regulation of phagocytosis (19,20,23). Whereas the molecular details of ITAM-mediated activation of the SHIP proteins is well studied, it is not known how SHP-1 is activated by ITAM-bearing receptors.
SHP-1 is a cytosolic tyrosine phosphatase that negatively regulates immune receptor signaling and growth factor signaling (24,25). SHP-1 is predominantly expressed in hematopoetic cells and contains two NH 2 -terminal located SH2 domains, a central phosphatase domain and two tyrosine phosphorylation sites in the COOH-terminal region. The enzyme is regulated by intramolecular interactions such that the NH 2 -terminal SH2 domain folds over the catalytic domain to inactivate the enzyme (26 -28). Deletion of the NH 2 -terminal SH2 domain, or engagement of the SH2 domains with cognate phosphopeptides has been shown to activate the phosphatase. Enzyme activity of SHP-1 is further enhanced by phosphorylation of tyrosines (Tyr 536 and Tyr 564 ) in the COOH-terminal region (29). The significance of the regulatory role of SHP-1 in the hematopoetic system is best exemplified in mice homozygous for motheaten (me/me) or motheaten viable (mev/mev) mutations (30 -32). The me/me mice do not express any SHP-1 protein, whereas the mev/mev mice express inactive splice variants of SHP-1. Both of these mutations result in multiple hematopoetic defects including elevated levels of autoantibodies and chronic inflammation resulting in early mortality. In a recent study, Durden and colleagues (22) have demonstrated that SHP-1 down-regulates Fc␥R-mediated phagocytosis in the J774A.1 mouse macrophage cell line.
In this study we have analyzed the molecular details of SHP-1 activation by the human Fc␥RIIa and the functional consequence of this activation. We report that SHP-1 phosphatase activity is induced upon Fc␥RIIa clustering in THP-1 human monocytic cells. Fc␥RIIa clustering results in membrane translocation of SHP-1 and association of SHP-1 with the phosphorylated ITAM of Fc␥RIIa. Co-precipitation experiments in cells transfected with ITAM tyrosine mutants of Fc␥RIIa revealed that SHP-1 associates with the NH 2 -terminal ITAM tyrosine, whereas Syk associates with the COOH-terminal ITAM tyrosine. Previous studies using substrate-trapping mutants of SHP-1 have demonstrated that SHP-1 associates with and dephosphorylates Syk, p85, and p62dok in other cell systems (33)(34)(35). Likewise, our current studies demonstrate association of SHP-1 with the above molecules upon Fc␥RIIa clustering suggesting that SHP-1 may dephosphorylate these molecules to down-regulate related signaling pathways. Consistent with this notion, analysis of functional consequence of SHP-1 phosphatase activity during Fc␥RIIa signaling demonstrated that overexpression of wild-type SHP-1 but not catalytically deficient SHP-1 down-regulates NFB-dependent gene transcription following Fc␥RIIa signaling. Taken together these results suggest that signaling events initiated by the ITAM of Fc␥RIIa are a composite of both positive and negative regulatory enzyme activation.

MATERIALS AND METHODS
Cells, Antibodies, and Reagents-THP-1 cells were obtained from ATCC and cultured in RPMI supplemented with 10% fetal bovine serum. P388D1 transfectants expressing human Fc␥RIIa were a generous gift from Dr. J. C. Edberg (University of Alabama) (36). COS-7 cells were cultured in Dulbecco's modified Eagle's medium with 10% fetal bovine serum. Anti-Fc␥RIIa antibody IV.3 was obtained from Medarex (Annandale, NJ). Rabbit polyclonal SHP-1, p85, Syk antibodies, and mouse monoclonal anti-phosphotyrosine antibody and phosphatase assay kits were purchased from Upstate Biotechnology (Charlottesville, VA).
Immunoprecipitation and Western Blotting-THP-1 cells and transfected P388D1 cells were activated by clustering Fc␥RIIa with F(abЈ) 2 fragments of monoclonal antibody IV.3 and goat F(abЈ) 2 anti-mouse Ig secondary antibody. Resting and activated cells were lysed in TN1 buffer (50 mM Tris, pH 8.0, 10 mM EDTA, 10 mM Na 4 P 2 O 7 , 10 mM NaF, 1% Triton X-100, 125 mM NaCl, 10 mM Na 3 VO 4 , 10 g/ml each aprotinin and leupeptin), and postnuclear lysates were incubated overnight with the antibody of interest and protein G-agarose beads (Invitrogen) or goat anti-mouse Ig covalently linked to Sepharose, depending on the antibody. Immunoprecipitations with control antibodies were performed in lysates of cells stimulated for 3 min. Immune complexes bound to beads were washed in TN1 and boiled in SDS sample buffer (60 mM Tris, pH 6.8. 2.3% SDS, 10% glycerol, 0.01% bromphenol blue, and 1% 2-mercaptoethanol) for 5 min. Proteins were separated by SDS-PAGE, transferred to nitrocellulose filters, probed with the antibody of interest, and developed by enhanced chemiluminescence.
Analysis of Fc␥RIIa Expression by Flow Cytometry-P388D1 transfectants were tested for expression of Fc␥RIIa by incubating with Fab fragments of anti-Fc␥RIIa monoclonal antibody IV.3, at a concentration of 10 g/ml for 30 min at 4°C. The cells were washed and incubated with fluorescein isothiocyanate-labeled goat F(abЈ) 2 anti-mouse Ig secondary antibody for 30 min at 4°C. Cells were subsequently washed, fixed in 1% paraformaldehyde, and analyzed by flow cytometry on an Elite EPICS fluorescence-activated cell sorter (Coulter, Hialeah, FL). Data from 10,000 cells per condition were recorded to yield the percentage of cells expressing receptors.
Phosphatase Assays-Phosphatase assays were performed as described previously (37), with slight modifications. To measure phosphatase activity associated with SHP-1, Fc␥RIIa, Syk, p85, p62dok, and Erk, these proteins were immunoprecipitated from resting and activated (Fc␥RIIa clustering) THP-1 cells. Immunoprecipitations with control antibodies were done in lysates of cells stimulated for 7 min. The immunoprecipitates were washed six times in wash buffer (10 mM Tris, pH 7.4), and subsequently incubated with tyrosine phosphopeptide substrate (RRLIEDAEpYAARG) (Upstate Biotechnology) in 10 mM Tris, pH 7.4, for 30 min. Reaction was stopped with 100 l of malachite green solution, incubated for a further 15 min, and the absorbance was measured at 630 nm. All assays were performed at least three times and the values obtained were plotted as mean Ϯ S.D.
Transfection of THP-1 Cells and Luciferase Assays-For analysis of SHP-1 influence on NFB transcriptional activity, THP-1 cells were transfected by electroporation (310 V, 950 F; Bio-Rad Gene Pulser II) with 5 g of wild-type SHP-1 or catalytically deficient (D419A) SHP-1 (a kind gift from Dr. R. Siraganian) (38), 1 g of NFB-luc plasmid, and 0.5 g of pEGFP to normalize for transfection efficiency. Transfectants were harvested 24 h later, activated by clustering Fc␥RIIa by methods described above for 6 h at 37°C. The cells were lysed in 100 l of cell culture lysis reagent (Promega). Luciferase activity was measured using the Promega luciferase assay reagent. Data are represented as graphs indicating the % increase in NFB activity in cells activated by clustering Fc␥RIIa over those that were not activated. Data points are expressed as mean Ϯ S.D. of three independent experiments. Statistical analysis was performed by Student's t test.
Transfection of COS-7 Cells-COS-7 cells were transfected as previously described (39). Briefly, cells were grown on culture dishes until they were 60 -70% confluent. Plasmids encoding wild-type SHP-1 and D419A SHP-1 were mixed with LipofectAMINE 2000 reagent (Invitrogen). The DNA mixture was added to cells in serum-free Dulbecco's modified Eagle's medium and incubated for 3 h at 37°C in a CO 2 incubator. The media was then replaced by Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum. The cells were harvested 24 h later and analyzed for expression of the transfected cDNAs by Western blotting whole cell lysates and SHP-1 immunoprecipitates from the transfectants were assessed for phosphatase activity as described above.
GFP-SHP-1 Construct-Wild-type SHP-1 cDNA in pSVL vector was obtained from Dr. R. Siraganian, and subcloned into pEGFP vector (Clontech) using the Xho and XbaI sites. Expression of GFP-SHP-1 was first confirmed by transfecting COS-7 fibroblasts with either empty EGFP vector or GFP-SHP-1 constructs, and subsequent Western blotting with anti-SHP-1 antibody.
Transfection of P388D1 Cells and Confocal Microscopy-P388D1 cells stably expressing human Fc␥RIIa were transfected with GFP-SHP-1 plasmids using LipofectAMINE, as described above for COS-7 cells. Cells were harvested 24 h post-transfection, serum starved, and stimulated by clustering Fc␥RIIa for 5 min. Resting and activated cells were fixed in 1% paraformaldehyde, cytospun onto glass slides, and stained with Hoechst nuclear stain. Slides were then mounted using mounting media (Molecular Probes) and analyzed by confocal microscopy using a Zeiss LSM510 multiphoton confocal microscope.

SHP-1 Is Activated by Fc␥RIIa
Clustering-To assess whether SHP-1 is activated by Fc␥RIIa, THP-1 cells were stimulated by clustering Fc␥RIIa with Fab fragments of the receptor-specific monoclonal antibody IV.3, followed by secondary cross-linking with goat F(abЈ) 2 fragments of anti-mouse Ig antibody. SHP-1 was immunoprecipitated from resting and activated cells and analyzed first, for phosphatase activity (Fig. 1A) and second, for tyrosine phosphorylation by Western blotting (Fig. 1B). The use of Fab/F(abЈ) 2 fragments of the clustering antibodies precludes the engagement of other Fc␥R present on the THP-1 cells by IgG ligand interaction ensuring that the resultant signals are emanating from Fc␥RIIa alone. In Fig.  1A, SHP-1 phosphatase activity was measured in THP-1 cells activated for the various time points indicated in the figure.
Results indicate that SHP-1 phosphatase activity is induced by Fc␥RIIa clustering and the activity peaks around 7 min poststimulation. Previous studies have indicated that the enzyme activity of SHP-1 is enhanced upon tyrosine phosphorylation of SHP-1 (29). The results shown in Fig. 1B demonstrate that SHP-1 is tyrosine-phosphorylated upon Fc␥RIIa clustering. A reprobe of the same membrane with anti-SHP-1 antibody in the lower panel indicates equal loading of SHP-1 in all lanes. The last lane marked "C" is a control immunoprecipitate with normal rabbit IgG.
To further confirm that the clustering antibodies used do not engage other Fc␥R expressed on THP-1 cells, specifically Fc␥RIIb, which bears a high level homology with Fc␥RIIa in the extracellular domain, binding of Fab fragments of monoclonal antibody IV.3 to Fc␥RIIb was analyzed. For this THP-1 cells were subjected to immunoprecipitation with Fab fragments of IV.3, intact IV.3 IgG, and two pan-Fc␥RII antibodies KB61 and AT10. The immunoprecipitates were probed with rabbit polyclonal antibodies specific for either Fc␥RIIb (Fig. 1C,  upper panel) or Fc␥RIIa (Fig. 1C, lower panel). Results indicate that whereas all four antibodies used are able to immunoprecipitate Fc␥RIIa, Fab fragments of IV.3 are unable to bind Fc␥RIIb. The pan-Fc␥RII antibodies KB61 and AT10 bound Fc␥RIIb as expected. Intact IV.3 IgG was also able to precipitate some Fc␥RIIb presumably by ligand interaction, as we have previously reported. Taken together these results demonstrate that SHP-1 is activated by the ITAM-bearing Fc␥RIIa in THP-1 cells, without the involvement of Fc␥RIIb.
SHP-1 Translocates to the Membrane upon Fc␥RIIa Clustering-To test whether Fc␥RIIa clustering resulted in membrane translocation of SHP-1, GFP-SHP-1 constructs were generated and transiently transfected into P388D1 mouse macrophage cells stably expressing human Fc␥RIIa. Cells were stimulated for 5 min by clustering Fc␥RIIa and analyzed by confocal microscopy. Results indicated that SHP-1 is distributed in the cytoplasm in resting cells and translocates to the membrane in cells activated by clustering Fc␥RIIa (Fig. 1D). In parallel samples transfected with EGFP alone, no movement of GFP was observed in activated cells compared with resting cells (data not shown).
SHP-1 Co-immunoprecipitates with Fc␥RIIa-We next assessed whether SHP-1 associates with Fc␥RIIa to become activated. Here, THP-1 cells were activated by clustering Fc␥RIIa by the methods described above. SHP-1 was immunoprecipitated from resting and activated cells, and analyzed for asso- ciation with Fc␥RIIa by Western blotting with the Fc␥RIIaspecific antibody 260. Results indicated that SHP-1 associates with Fc␥RIIa upon activation ( Fig. 2A, upper panel). No association was detectable in resting cells. The same membrane was reprobed with anti-SHP-1 antibody to ensure equivalent loading of SHP-1 in all lanes (lower panel).
As a second approach to confirm association of SHP-1 with Fc␥RIIa, the receptors were immunoprecipitated from resting and activated THP-1 cells and subjected to a phosphatase assay with a phosphopeptide substrate. The amount of free phosphate released was detected by the addition of malachite green. Results are expressed as picomole of phosphate released by immunoprecipitates from activated cells after subtracting the values obtained from immunoprecipitates from resting cells (Fig. 2B). Control immunoprecipitates consistently showed values equal to or lower than resting cell immunoprecipitates.
Together these experiments demonstrate that SHP-1 associates with Fc␥RIIa following receptor clustering. NH 2 -terminal ITAM Tyrosine of Fc␥RIIa Is Necessary for Association with SHP-1-To examine which of the two ITAM tyrosines of Fc␥RIIa were involved in the association with SHP-1, we used two experimental models. First, synthetic biotinylated peptides derived from the ITAM of Fc␥RIIa, which were either non-phosphorylated (P1), or singly phosphorylated on either the NH 2 -terminal ITAM tyrosine (P2) or the COOH-terminal ITAM tyrosine (P3), were applied to THP-1 lysates and the peptide-bound material was analyzed for the presence of SHP-1 by Western blotting. The results shown in Fig. 3A, upper panel, indicate that the phosphorylated NH 2 -terminal ITAM tyrosine, but not the COOH-terminal tyrosine, efficiently bound SHP-1. SHP-1 did not associate with the nonphosphorylated peptide (lane 1). In contrast, parallel experiments analyzing the binding properties of the peptides demonstrated that the peptide phosphorylated on the COOHterminal ITAM tyrosine is functional and is able to associate with Syk (Fig. 3A, middle panel) and p85 (Fig. 3A, lower panel). These latter findings are consistent with earlier reports demonstrating that the COOH-terminal ITAM tyrosine of Fc␥RIIa is sufficient for association with Syk (40), and that p85 associates with both NH 2 -and COOH-terminal ITAM tyrosines of Fc␥RIIa (6,41).
Because the above experiments were performed with synthetic peptides, we next asked whether the native Fc␥RIIa receptor would likewise demonstrate the differential ITAM tyrosine requirement for association with SHP-1 and Syk. For these experiments we used P388D1 mouse macrophage transfectants stably expressing single ITAM tyrosine mutants of human Fc␥RIIa. The P388D1 transfectants were activated by clustering Fc␥RIIa, the receptors were immunoprecipitated from resting and activated cells and analyzed by Western blotting for co-precipitating SHP-1 (Fig. 3B, upper panel) or Syk (Fig. 3B, lower panel). Results indicated that SHP-1 failed to associate with Fc␥RIIa when the NH 2 -terminal ITAM tyrosine was mutated to phenylalanine (Y252F). However, the Y252F receptor displayed efficient binding to Syk. These results are consistent with the above peptide binding experiments.
To assess the signaling outcome of the ITAM tyrosine mutations, we compared the ability of these mutated receptors versus the wild-type receptor to induce signaling. For this we first ensured that the transfected receptors were expressed to comparable levels by flow cytometry (Fig. 4A). The transfected cells were stimulated by clustering Fc␥RIIa. Fc␥RIIa was immunoprecipitated from resting and activated cells and analyzed for tyrosine phosphorylation. Results indicated that all three receptors are capable of being tyrosine phosphorylated (Fig. 4B,  upper panel). As might be expected the single ITAM tyrosine mutants displayed lower phosphorylation levels than the wildtype receptor. A reprobe of the membrane demonstrated equivalent receptor expression in the transfectants (Fig. 4B, lower  panel). The reduced signal seen with anti-Fc␥RIIa antibody in the activated lane is because of the fact that the anti-Fc␥RIIa blotting antibody often displays lower efficiency of detection of the phosphorylated Fc␥RIIa in a reprobe. We, and others, have previously reported this property of the anti-Fc␥RIIa blotting antibody (19,42).
We next analyzed total cellular tyrosine phosphorylation in the transfectants stimulated by Fc␥RIIa clustering (Fig. 4C). Results indicated that, clustering of the NH 2 -terminal ITAM tyrosine mutant leads to enhanced overall cellular tyrosine phosphorylation in comparison to clustering of the wild-type receptor (lane 4 versus lane 1). In contrast, mutation of the COOH-terminal ITAM tyrosine completely abrogated overall cellular phosphorylation. These observations are consistent with the notion that SHP-1 associates with the NH 2 -terminal ITAM tyrosine to down-modulate tyrosine phosphorylation events, and that Syk associates with the COOH-terminal ITAM tyrosine to become activated and lead to the phosphorylation of signaling proteins in the cell.
SHP-1 Associates with p85, Syk, and p62dok during Fc␥RIIa Signaling-The activation of SHP-1 during Fc␥RIIa signaling suggests that SHP-1 causes dephosphorylation of tyrosinephosphorylated proteins. Numerous previous studies have identified the association of SHP-1 with tyrosine-phosphorylated signaling molecules, the subsequent dephosphorylation of these molecules, and down-regulation of the related signaling pathways (43). Drawing from these previous studies, we next analyzed whether SHP-1 associated with the tyrosine kinase Syk, the p85 adapter molecule of PtdIns 3-kinase, and the Ras GAP-binding protein p62dok during Fc␥RIIa signaling. Thus, THP-1 cells were activated by clustering Fc␥RIIa for various time points. SHP-1 was immunoprecipitated from resting and activated THP-1 cells and analyzed by Western blotting for the presence of co-precipitating Syk (Fig. 5A), p85 (Fig. 5B), and p62dok (Fig. 5C). As seen in the figure, SHP-1 associated with the above molecules in an activation-dependent manner. The membranes were reprobed with anti-SHP-1 antibody to ensure equal loading in all lanes. To further analyze whether active SHP-1 is associated with Syk, p85, and p62dok, phosphatase assays were performed on the respective immunoprecipitates from resting and activated THP-1 cells. Consistent with association of SHP-1 protein, results indicated that phosphatase activity was present in Syk, p85, and p62dok immunoprecipitates (Fig. 5D). In control experiments, no association of SHP-1 phosphatase activity was observed in Erk immunoprecipitates from activated THP-1 cells (data not shown). Taken together these data suggest that SHP-1 may dephosphorylate the above molecules to down-regulate activation events induced by Fc␥RIIa clustering.
SHP-1 Down-regulates Fc␥RIIa-mediated Function-In recent reports we, and others, have demonstrated that Fc␥RIIa clustering results in the activation of NFB-dependent gene transcription (19,21,44). These activation events are subject to regulation by the inositol phosphatases SHIP-1 and SHIP-2, presumably as a result of the consumption of the lipid products of PtdIns 3-kinase and the downstream signaling thereof. Our present studies demonstrate that SHP-1 associates with the p85 subunit of PtdIns 3-kinase, suggesting that SHP-1 may modulate PtdIns 3-kinase activity. Therefore, we next asked whether SHP-1 also played a role in modulating NFB-dependent gene transcription initiated by Fc␥RIIa clustering. In these experiments we used wild-type and catalytically inactive (D419A) SHP-1 constructs, which we first expressed in COS-7 fibroblasts by transient transfection and analyzed for SHP-1 protein expression and enzyme activity. The results shown in Fig. 6B demonstrate that both wild-type and D419A SHP-1 are expressed efficiently from these plasmids. COS-7 fibroblasts do not express any endogenous SHP-1 as is seen from the absence of SHP-1 in the mock-transfected cells (lane 1). Shown in Fig.  6B, lower panel, is the phosphatase activity of these two SHP-1 proteins expressed in COS-7 cells.
Having ensured that we could achieve appropriate protein expression from these constructs, we then transiently transfected THP-1 cells with plasmids encoding the NFB binding element coupled to a luciferase gene (NFB-luc) either alone or with an excess of wild-type SHP-1 or D419A SHP-1. The cells were harvested 24 h post-transfection, activated by clustering Fc␥RIIa, and NFB-dependent luciferase expression was assessed in a luciferase enzyme assay. Results from three independent experiments are shown in Fig. 6A. Overexpression of wild-type SHP-1 completely abrogated NFB-dependent luciferase induction. In contrast, overexpression of the catalytically inactive D419A SHP-1 resulted in enhanced luciferase induction. These data demonstrate that SHP-1 negatively regulates Fc␥RIIa-mediated biological outcomes in human myeloid cells.

DISCUSSION
The human-specific Fc␥RIIa is a low affinity IgG receptor that has several unique features to it. In addition to being the most widely expressed IgG receptor, it also contains an unusually lengthy ITAM in its cytoplasmic domain. Mutational analyses of the cytoplasmic domain of Fc␥RIIa have identified specific amino acid motifs that are important for the phagocytic process. For example, mutation of either of the two tyrosine residues within the ITAM of Fc␥RIIa have been reported to severely abrogate intracellular calcium mobilization and phagocytosis (36,45). An additional tyrosine residue located NH 2terminal to the ITAM also becomes phosphorylated upon receptor clustering and plays a role in Fc␥RIIa-mediated activation (45). More recent studies have identified an LTL motif in the cytoplasmic domain of Fc␥RIIa that is involved in the formation of phagolysosomes (46,47). Thus the cytoplasmic domain of Fc␥RIIa is made up of a complex set of signaling motifs that are not yet fully explored.
Once Fc␥RIIa receptors are clustered the Src family of tyrosine kinases phosphorylate tyrosine residues in the cytoplasmic domain of Fc␥RIIa (6). Phosphorylation of the ITAM promotes recruitment and activation of Syk, followed by the phosphorylation of multiple cytosolic signaling proteins. Unlike its T cell homolog ZAP-70 that requires both of its tandem SH2 domains to be engaged by phosphorylated ITAMs to be activated, single SH2 domain engagement is sufficient for Syk activation (48). Accordingly, the results shown in Fig. 3 demonstrate that the COOH-terminal ITAM tyrosine of Fc␥RIIa is necessary and sufficient for Syk association. Interestingly, there was constitutive Syk association with Y252F Fc␥RIIa, at a time when no tyrosine phosphorylation of the receptor was detectable (Fig. 4B). These results suggest that perhaps mutation of tyrosine 252 might lead to a non-SH2-dependent association of Syk with Y252F Fc␥RIIa. Additional studies are needed to define the nature of this novel interaction.
Recent studies have revealed that Fc␥RIIa clustering not only initiates activating events, but it also induces negative regulatory events such that the resultant biologic outcome is tempered. Thus, Fc␥RIIa recruits the inositol phosphatases SHIP-1 and SHIP-2 to modulate signaling events (19 -21). In a transfected COS-7 fibroblast model the protein-tyrosine phosphatase SHP-1 has also been shown to regulate Fc␥RIIa-mediated phagocytosis (22). Our current studies extend these latter findings to demonstrate that in myeloid cells SHP-1 translocates to the membrane, associates with the phosphorylated NH 2 -terminal ITAM tyrosine of Fc␥RIIa, and regulates Fc␥RIIa-mediated signaling. Together, these observations suggest that signal transduction from Fc␥RIIa is internally regulated by both positive and negative signaling enzymes.
SHP-1 was initially thought to be the effector molecule of Fc␥RIIb-mediated inhibition (49). However, later studies using chimeric Fc␥RIIb receptors and SHP-1-deficient cells demonstrated that SHP-1 is not required for Fc␥RIIb function (50,51), but that SHP-1 works in concert with other ITIM-bearing receptors such as the KIRs, gp49B, PIR-B etc. (52), to mediate its inhibitory function. Our current observations of SHP-1 association with Fc␥RIIa ITAM are novel, and are consistent with earlier findings that SHP-1 association with immune receptors occurs in the absence of involvement of the ITIM-bearing Fc␥RIIb (53).
Activation of SHP-1 enzyme requires the engagement of its NH 2 -terminal SH2 domain with phosphotyrosines to relieve the intramolecular constraint placed on the phosphatase domain (26). Consistent with this notion, our results suggest that the engagement of SHP-1 by the NH 2 -terminal ITAM tyrosine of Fc␥RIIa leads to the activation of SHP-1. Other studies have reported a secondary mechanism of SHP-1 activation involving phosphorylation of SHP-1 on its COOH-terminal located tyrosine residues (29). In our experiments, although tyrosine phosphorylation of SHP-1 was detectable after Fc␥RIIa clustering (Fig. 1), the level of phosphorylation is weak suggesting that it is likely that the primary mechanism of SHP-1 activation is mediated by its association with Fc␥RIIa. Additional studies are required to assess whether the low level phosphorylation of SHP-1 contributes to activation of the enzyme.
The identification of specific substrates of SHP-1 has been aided by the use of substrate-trapping mutant forms of SHP-1 and the SHP-1-deficient motheaten animals. Consistent with earlier observations in other cell systems, we have observed association of SHP-1 with Syk, p85, and p62dok. These findings suggest that the above molecules may be dephosphorylated by SHP-1 resulting in down-regulation of the related signaling pathways. In accordance with this notion, our data indicate that SHP-1 down-regulates NFB-dependent gene transcription in myeloid cells stimulated by clustering Fc␥RIIa. NFB activation has been shown to be important for Fc␥R-induced transcription of inflammatory cytokine genes such as interleukin-1, tumor necrosis factor-␣, and interleukin-8 (54). Thus the current study establishes a role for SHP-1 in modulating the production of inflammatory cytokines during immune complex clearance.