SHP-1 Requires Inhibitory Co-receptors to Down-modulate B Cell Antigen Receptor-mediated Phosphorylation of Cellular Substrates*

Signaling through the B cell antigen receptor (BCR) is negatively regulated by the SH2 domain-containing pro-tein-tyrosine phosphatase SHP-1, which requires association with tyrosine-phosphorylated proteins for activation. Upon BCR ligation, SHP-1 has been shown to associate with the BCR, the cytoplasmic protein-tyro-sine kinases Lyn and Syk, and the inhibitory co-recep-tors CD22 and CD72. How SHP-1 is activated by BCR ligation and regulates BCR signaling is, however, not fully understood. Here we demonstrate that, in the BCR-expressing myeloma line J558L m m3, CD72 expression reduces the BCR ligation-induced phosphorylation of the BCR component Ig a /Ig b and its cytoplasmic effec-tors Syk and SLP-65. Substrate phosphorylation was restored by expression of dominant negative mutants of SHP-1, whereas the SHP-1 mutants failed to enhance phosphorylation of the cellular substrates in the absence of CD72. This indicates that SHP-1 is efficiently activated by CD72 but not by NY), the combination of anti-GFP mAb and peroxidase-conjugated anti-mouse IgG Ab (Southern Biotechnolo- gy), the combination of goat anti-mouse l chain Ab and peroxidase-conjugated anti-goat IgG Ab (both are from Southern Biotechnology), the combination of biotinylated anti-Lyn Ab (Santa Cruz Biotechnology) and peroxidase-conjugated streptavidin (Amersham Pharmacia Bio- tech), or the combination of anti-Ig b mAb HM79 and peroxidase-conju-gated protein G (Zymed Laboratories Inc. Laboratories, San Francisco, CA). Proteins were then visualized by ECL system (Amersham Pharmacia Biotech) as described previously (23). In Vitro Kinase Assay— Kinase activity of Syk and Lyn was measured as described previously with slight modification (38). In brief anti-Syk or anti-Lyn immunoprecipitates were washed twice with lysis buffer and once with kinase buffer and incubated in 20 m l of kinase buffer containing 2 m g of GST-Ig a and 2 m M of ATP at room temperature for 3 min. The reaction was terminated by adding SDS-polyacrylamide gel electrophoresis sample buffer, and proteins were separated by SDS-polyacrylamide gel electrophoresis, followed by Western blot analysis using anti-phosphotyrosine mAb 4G10.

protein-tyrosine kinases (PTKs) and induces phosphorylation of the immunoreceptor tyrosine-based activation motifs in the cytoplasmic tails of the Ig␣/Ig␤ heterodimer, the signaling component of the BCR (1,2). Phosphorylated Ig␣/Ig␤ can activate the cytoplasmic PTK Syk, which in turn phosphorylates the adapter protein called SLP-65 or B cell linker protein (BLNK) (1)(2)(3)(4). Upon phosphorylation, SLP-65 initiates downstream signaling events by recruiting various signaling molecules such as phospholipase C-␥, Nck, and Btk (3)(4)(5)(6)(7). Studies on genetically manipulated mice or B cell lines have shown that Ig␣/Ig␤, Syk, and SLP-65 are essential in downstream signaling such as Ca 2ϩ mobilization and extracellular signal-regulated kinase activation (2,3,5,8,9). The downstream signaling events appear to lead ultimately to proliferation, functional inactivation, or death of B cells. BCR signaling is regulated either positively by co-receptors such as CD19 and negatively by other co-receptors such as CD22 and the low affinity receptor for IgG (Fc␥RII) (10 -14). Fc␥RII modulates BCR signaling only when it interacts with the antigen-IgG complex and is involved in negative feedback regulation of IgG production. In contrast, both CD22 and CD72 appear to interact constitutively with BCR and regulate negatively both mitogen-activated protein kinase activation and Ca 2ϩ mobilization induced by BCR ligation (12,13,(15)(16)(17). By negatively regulating BCR signaling whenever BCR is ligated, both of these co-receptors are implicated in setting a signaling threshold for BCR ligation.
The Src homology 2 (SH2) domain-containing protein-tyrosine phosphatase SHP-1 is involved in negative regulation of several receptors (18). For SHP-1 activation, its tandem SH2 domains need to be associated with a tyrosine-phosphorylated peptide (19). In B cells, CD22 has been shown to activate SHP-1. CD22 contains the conserved immunoreceptor tyrosinebased inhibition motifs (ITIMs) in the cytoplasmic regions (20). Upon tyrosine phosphorylation, the ITIMs of CD22 associate with and activate SHP-1. PIR-B and CD72 also contain ITIMs in their cytoplasmic region and, upon phosphorylation, recruit SHP-1 (21)(22)(23)(24), suggesting that these molecules may also activate SHP-1 in B cells. Indeed, SHP-1 is shown to be involved in negative regulation of BCR-mediated Ca 2ϩ signaling by a chimeric Fc␥RII containing the cytoplasmic region of PIR-B in the chicken B cell line DT40 (22). It is not yet known whether intact PIR-B is involved in negative regulation of BCR signaling by SHP-1. In contrast, both CD22 and CD72 are phosphorylated upon BCR ligation probably because these co-receptors associate with BCR constitutively (23)(24)(25)(26). Thus, CD22 and CD72 may activate SHP-1 upon BCR ligation, thereby negatively regulating BCR signaling. Whether SHP-1 mediates negative regulation of BCR signaling by CD22 and CD72, and how SHP-1 activated by these inhibitory co-receptors negatively regulates BCR signaling has not yet been fully elucidated.
SHP-1 is activated by various tyrosine-phosphorylated signaling molecules other than ITIM-containing inhibitory receptors. Indeed, activation receptors such as erythropoietin receptor and IL-3 receptor are shown to associate with and activate SHP-1 upon ligand-induced phosphorylation (27,28). In B cells, SHP-1 is reported to co-precipitate with the BCR probably due to its association with Ig␣/Ig␤ (29). SHP-1 is also activated by tyrosine-phosphorylated cytoplasmic signaling molecules such as PTK ZAP-70, the homolog of the PTK Syk expressed in B cells (30). It is not yet known whether SHP-1 is activated by Ig␣/Ig␤ or Syk.
A dominant negative mutant of SHP-1 has been used to assess the role of SHP-1 in the regulation of BCR (24,31,32). This mutant SHP-1 blocks the effect of SHP-1 activated by various different molecules. To assess whether a certain molecule, for example CD72, activates SHP-1 and whether its function is mediated by SHP-1, the dominant negative mutant of SHP-1 needs to be expressed in the cells where SHP-1 is not activated by alternative pathways. Such a cellular system has not yet been established. J558Lm3 myeloma cells express surface IgM specific for the hapten (4-hydroxy-3-nitrophenyl) acetyl (NP) (33). J558Lm3 cells do not express most of the membrane molecules expressed in B cells (34) including the inhibitory co-receptors CD22 and CD72, but they express SHP-1 and various cytoplasmic signaling molecules involved in BCR signaling such as Syk and SLP-65, and BCR ligation activates these molecules (4,34). Thus, J558Lm3 is a useful tool to analyze the molecular mechanisms of activation of SHP-1 and the function of inhibitory co-receptors. By using this cellular system, we demonstrate here that BCR-induced SHP-1 activation requires inhibitory co-receptors such as CD72 and that SHP-1 is a proximal effector molecule of CD72 to downregulate BCR-mediated signal transduction.

EXPERIMENTAL PROCEDURES
Plasmids and Cells-The expression plasmid pMKITSHP-1SH2 coding for a GFP fusion protein containing the tandem SH2 domains of SHP-1 (SHP-1SH2) was generated as follows. The NotI-EcoRI fragment containing the GFP cDNA and the EcoRI-PvuII fragment of pBlue-scriptSHP-1 (23) containing SHP-1SH2 were inserted into NotI and EcoRV of modified pMKITneo harboring NotI and EcoRV sites instead of EcoRI and NotI sites. The expression plasmid pMKITSHP-1C/S-Myc coding for a catalytically inactive mutant of SHP-1, in which cysteine residue at 453 was replaced by serine (SHP-1C/S), tagged by a c-Myc peptide (SMEQKLISEEDLN) are generated as follows. pBluescript-SHP-1C/S containing the cDNA for SHP-1C/S was generated by sitedirected mutagenesis using pBluescriptSHP-1 containing the mouse SHP-1 cDNA (23). The HindIII-SalI fragment of the DNA encoding the c-Myc sequence was generated by annealing the pair of synthetic oligonucleotides (5Ј-AGCTTGTCACCGTCTCCTCAGAACAAAAACTCATC-TCAGAAGAGGATCTGAATTAAG-3Ј and 5Ј-TCGACTTAATTCAGAT-CCTCTTCTGAGATGAGTTTTTGTTCTGAGGAGACGGTGACA-3Ј). The SHP-1C/S cDNA lacking the stop codon was generated by polymerase chain reaction using a set of primers (5Ј-CCGAATTCGAA-CCCCAGGATGGTGAGG3Ј and 5Ј-AAAAGCTTCTTCCTCTTGAGAG-AACCTTT-3Ј) and pBluescriptSHP-1C/S as a template, digested with EcoRI and HindIII, and ligated with the HindIII-SalI fragment of the DNA encoding the c-Myc sequence and the expression vector pMKITneo opened by EcoRI and XhoI. Site-directed mutagenesis was done using pMikCD72 (23) to generate the expression plasmid pMikCD72Y7F coding for the mutated form of CD72 in which tyrosine residue at position 7 was replaced by phenylalanine (CD72Y7F). The mouse B lymphoma line K46m (35) and its CD72 transfectant (K46mCD72) (17) and the mouse myeloma cell line J558Lm3 (33), all of which express H and L chains of IgM specific for NP, were described previously. The mouse B lymphoma line WEHI-231 was described previously (36). Transfection of the expression plasmids into J558Lm3 cells was done by electroporation. Cells were cultured in RPMI 1640 medium supplemented with 10% fetal calf serum, 50 M 2-mercaptoethanol, 1 mM glutamine. For BCR ligation, J558Lm3, its transfectants, K46m and K46mCD72, were treated with 10 g/ml NP-coupled bovine serum albumin (NP-BSA), and WEHI-231 cells were treated with 10 g/ml F(abЈ) 2 fragments of goat anti-mouse IgM antibody (Ab) (ICN Pharmaceuticals, Aurora, OH) at 37°C.
Immunoprecipitation and Western Blot Analysis-Cells were lysed in Triton X-100 lysis buffer (23). Cleared cell lysates were incubated with anti-mouse CD72 mAb 9-6.1 (37), rabbit anti-mouse SLP-65 Ab (4), rabbit anti-Lyn Ab (Santa Cruz Biotechnology, Santa Cruz, CA), rabbit anti-Syk (C-20) Ab (Santa Cruz Biotechnology), anti-GFP mAb (a gift of Dr. Mitani), or anti-Ig␤ mAb HM79 (a gift of Dr. Karasuyama) followed by addition of protein G-Sepharose (Amersham Pharmacia Biotech). For immunoprecipitation of SHP-1, rabbit anti-SHP-1 Ab (Santa Cruz Biotechnology) was coupled to CNBr-activated Sepharose (Amersham Pharmacia Biotech). Cell lysates were incubated with anti-SHP-1 Abcoupled Sepharose beads. Alternatively, cell lysates were incubated with anti-SHP-1 Ab together with protein G-Sepharose. After washing three times with lysis buffer, protein G-coupled beads or anti-SHP-1coupled beads were suspended in sample buffer. Total cell lysates or immunoprecipitates were separated by SDS-polyacrylamide gel electrophoresis, transferred to the nylon membrane. Membranes were reacted with rabbit anti-CD72 Ab (23), anti-SHP-1 Ab (Santa Cruz Biotechnology), or anti-Syk (N-19) Ab (Santa Cruz Biotechnology), followed by reaction with peroxidase-conjugated donkey anti-rabbit Ig Ab (Amersham Pharmacia Biotech). Alternatively, membranes were reacted with peroxidase-conjugated anti-phosphotyrosine mAb 4G10 (Upstate Biotechnology, Inc., Lake Placid, NY), the combination of anti-GFP mAb and peroxidase-conjugated anti-mouse IgG Ab (Southern Biotechnology), the combination of goat anti-mouse chain Ab and peroxidaseconjugated anti-goat IgG Ab (both are from Southern Biotechnology), the combination of biotinylated anti-Lyn Ab (Santa Cruz Biotechnology) and peroxidase-conjugated streptavidin (Amersham Pharmacia Biotech), or the combination of anti-Ig␤ mAb HM79 and peroxidase-conjugated protein G (Zymed Laboratories Inc. Laboratories, San Francisco, CA). Proteins were then visualized by ECL system (Amersham Pharmacia Biotech) as described previously (23).
In Vitro Kinase Assay-Kinase activity of Syk and Lyn was measured as described previously with slight modification (38). In brief anti-Syk or anti-Lyn immunoprecipitates were washed twice with lysis buffer and once with kinase buffer and incubated in 20 l of kinase buffer containing 2 g of GST-Ig␣ and 2 mM of ATP at room temperature for 3 min. The reaction was terminated by adding SDS-polyacrylamide gel electrophoresis sample buffer, and proteins were separated by SDSpolyacrylamide gel electrophoresis, followed by Western blot analysis using anti-phosphotyrosine mAb 4G10.

CD72 Negatively Regulates BCR-mediated Tyrosine Phosphorylation of Signaling
Molecules-Western blot analysis of total cell lysates and/or flow cytometry revealed that J558Lm3 does not express CD22, CD72, or Fc␥RII (data not shown). To address the regulatory effect of CD72 on BCR signaling, we transfected J558Lm3 cells with an expression plasmid for CD72. For further analysis, we chose two independent CD72 transfectants (J558Lm3CD72-1 and J558Lm3CD72-24) that express CD72 and the IgM-BCR on the cell surface ( Fig. 1, A-F). Following BCR ligation of J558Lm3 cells and its CD72 transfectants with the specific antigen NP-BSA, we monitored PTK substrate phosphorylation by anti-phosphotyrosine immunoblotting of total cell lysates. Tyrosine phosphorylation of various substrate molecules reached the maximum 1 min after exposure to NP-BSA and thereafter gradually declined (Fig. 1G). Substrate phosphorylation in the CD72 transfectants was much weaker than in J558Lm3 cells, which indicates that CD72 expression inhibits BCR-induced substrate phosphorylation.
One of the predominantly phosphorylated PTK substrates in antigen-treated J558Lm3 was isolated and named SLP-65 (4). The antigen-induced phosphorylation of SLP-65 was much weaker in the CD72 transfectants than in the parental J558Lm3 cells (Fig. 1G), suggesting that CD72 reduced BCRmediated phosphorylation of SLP-65. This is directly shown by anti-phosphotyrosine blotting of anti-SLP-65 immunoprecipitates from J558Lm3 and J558LmCD72-1 cells (Fig. 2A). Next, we examined the phosphorylation of other BCR signaling molecules such as Lyn, Syk, and the Ig␣/Ig␤ heterodimer. The antigen-induced phosphorylation of Syk (Fig. 2B) and Ig␣/Ig␤ (Fig. 2C) was reduced in the CD72 transfectants compared with parent J558Lm3 cells. In consistent with this finding, the kinase activity of Syk was reduced in the CD72 transfectant (Fig. 2E). However, both phosphorylation and the kinase activity of Lyn were not up-regulated by BCR ligation in J558Lm3, as described previously (39), nor altered by CD72 expression (Fig. 2, D and F). Taken together, CD72 negatively regulates BCR-mediated phosphorylation of signaling molecules such as Ig␣/Ig␤, Syk, and SLP-65 but not Lyn.
CD72 Carrying a Mutation in ITIM Does Not Down-modulate BCR-mediated Phosphorylation of Cellular Substrates-To assess whether CD72 requires the ITIM for down-modulation of BCR signaling, we transfected J558Lm3 cells with an expression plasmid encoding the mutated form of CD72 in which tyrosine in the ITIM was replaced by phenylalanine (CD72Y7F). The CD72Y7F transfectants (J558Lm3CD72Y7F-3 and J558Lm3CD72Y7F-5) expressed a higher level of CD72 on the surface (Fig. 3, B and C) than J558Lm3CD72 which expresses wild-type CD72 (Fig. 3A). However, tyrosine phosphorylation of cellular substrates was not reduced in J558Lm3CD72Y7F-5 compared with the parent J558Lm3 cells (Fig. 3D). Essentially the same result was obtained with J558Lm3CD72Y7F-3 (data not shown). These results show that the CD72 ITIM mutant fails to down-modulate BCR-mediated phosphorylation of cellular substrates. The ITIM in CD72 is thus essential for down-modulating BCR signaling probably by recruiting and activating SHP-1.
Western blot analysis of total cell lysates revealed that BCR ligation induced phosphorylation of CD72Y7F in J558Lm3CD72Y7F (Fig. 3D), indicating that tyrosine residues outside the ITIM are phosphorylated in CD72Y7F after BCR ligation. In contrast, the same analysis barely demonstrated phosphorylation of wild-type CD72 in J558Lm3CD72 (Fig. 1G), although more sensitive analysis using anti-CD72 immunoprecipitates showed phosphorylation of CD72 upon BCR ligation in these cells (data not shown). These results indicate that the ITIM negatively regulates tyrosine phosphorylation of CD72. Since the ITIM is essential for recruiting SHP-1, SHP-1 may dephosphorylate CD72, as reported previously (24).

CD72 Is Required for SHP-1-mediated Down-modulation of Tyrosine Phosphorylation of Signaling Molecules in BCRligated J558Lm3
Cells-To assess the role of SHP-1 in CD72mediated dephosphorylation of cellular substrates such as Ig␣/ Ig␤, Syk, and SLP-65, we constructed expression plasmids for two distinct dominant negative mutants of SHP-1, i.e. a catalytically inactive mutant of SHP-1 tagged with a c-Myc peptide (SHP-1C/S) and a GFP fusion protein containing the tandem SH2 domains but not the catalytic domain of SHP-1 (SHP-1SH2). We then transfected J558Lm3 and J558Lm3CD72 with these expression plasmids. Expression of surface IgM and CD72 in the transfectants was similar to that of the parent cells (data not shown). SHP-1C/S was dominantly coprecipitated with CD72 in the J558Lm3CD72 transfectants (Fig. 4D), indicating that SHP-1C/S out-competes endogenous SHP-1 for binding to CD72. The dominant negative SHP-1 thus appears to block efficiently recruitment and activation of endogenous SHP-1. The J558Lm3CD72 transfectants (J558Lm3CD72C/S and J558Lm3CD72SH2) and the J558Lm3 transfectants (J558Lm3C/S and J558Lm3SH2) expressed similar levels of SHP-1C/S or SHP-1SH2 (Fig. 4, A-C). However, antigen-induced phosphorylation of various substrates in both J558Lm3CD72C/S and J558Lm3CD72SH2 transfectants was markedly enhanced compared with the parent J558Lm3CD72 cells (Fig. 4, E and  G), whereas J558Lm3C/S and J558Lm3SH2 transfectants showed a level of substrate phosphorylation similar to that seen in J558Lm3 (Fig. 4, F and H). The dominant negative mutants of SHP-1 thus enhanced phosphorylation of cellular substrates in the presence of CD72 but not in its absence.
Moreover, expression of SHP-1C/S enhanced phosphorylation of Ig␣/Ig␤, Syk, and SLP-65 in the J558Lm3CD72 transfectant (Fig. 5, A-C), whereas the phosphorylation of Ig␣/Ig␤ and Syk was not altered by SHP-1C/S in the absence of CD72 (Fig. 5, D and E), supporting the notion that CD72 is required for enhancement of substrate phosphorylation by SHP-1 mutants. These results suggested that SHP-1 is efficiently activated by CD72 but no other pathways in antigen-treated J558Lm3CD72. To confirm this notion, we assessed the tyrosine-phosphorylated proteins associated with the SH2 domains of SHP-1. Activation of SHP-1 requires association of its SH2 domains with a tyrosine-phosphorylated peptide (19). When we analyzed the anti-SHP-1 immunoprecipitates from J558Lm3CD72C/S and J558Lm3CD72SH2 cells treated with NP-BSA, CD72 was co-precipitated with SHP-1C/S (Fig.  6A) or SHP-1SH2 (Fig. 6B). However, no other tyrosinephosphorylated proteins were detected. Taken together, SHP-1 appears to be activated exclusively by CD72 in antigenstimulated J558Lm3. This observation is not restricted to J558Lm3 cells. Indeed, the inhibitory co-receptors CD22 and CD72 are the only major phosphoproteins co-precipitated with SHP-1 in BCR-ligated B cell lines K46mCD72 and WEHI-231 (Fig. 7). CD22 and CD72 may thus play a major role in SHP-1 activation upon BCR ligation.

DISCUSSION
The myeloma line J558Lm3 expresses BCR, SHP-1, and various molecules involved in BCR signaling such as Lyn, Syk, and SLP-65, whereas it does not express co-receptors such as CD22 and CD72. Thus J558Lm3 is a useful tool to analyze the role of SHP-1 and co-receptors in BCR signaling regulation. By using J558Lm3 cells, we demonstrate that CD72 negatively regulates BCR-mediated phosphorylation of cellular substrates such as Ig␣/Ig␤, Syk, and SLP-65. The substrate phosphorylation was not altered by the mutant CD72 lacking the tyrosine residue within the ITIM required for SHP-1 activation. Moreover, BCR-mediated substrate phosphorylation was restored by inhibiting SHP-1 specifically activated by CD72 using dominant negative mutants of SHP-1. Thus, SHP-1 is a specific intracellular effector of CD72 to negatively regulate BCR-mediated phosphorylation of cellular substrates. In contrast, dominant negative mutants did not enhance BCR-induced substrate phosphorylation in J558Lm3 cells in the absence of CD72 suggesting that SHP-1 is activated upon BCR ligation in the presence of CD72 but in its absence in these cells. This notion is further supported by our result that among the proteins tyrosine-phosphorylated upon BCR ligation in J558Lm3 cells, CD72 is the only protein detected to associate with the SH2 domains of SHP-1. This indicates that SHP-1 is efficiently activated by CD72 but not by other pathways because SHP-1 activation requires its association with a tyrosine-phosphorylated peptide. In addition, CD22 and CD72 are the only major phosphoproteins associated with SHP-1 in BCR-ligated mouse B lymphoma lines WEHI-231 and K46m. Taken together, BCR-mediated SHP-1 activation requires inhibitory co-receptors such as CD72 in both J558Lm3 and B cell lines. SHP-1 has been shown to be associated with and activated by various tyrosine-phosphorylated molecules independently of inhibitory receptors. These SHP-1-activating molecules include stimulatory receptors such as the erythropoietin receptor (27,28) or cytoplasmic signaling molecules such as ZAP-70 (30). BCR ligation results in the phosphorylation of various molecules such as the BCR component Ig␣/Ig␤ and the cytoplasmic signaling molecules Syk and SLP-65 in J558Lm3 cells as well as in B cells (4,34). Moreover, J558Lm3 cells express active Lyn, because specific substrates of Lyn such as CD22 and CD19 (40 -43) are phosphorylated upon BCR ligation in J558Lm3 transfectants expressing CD22 and CD19. 2 However, our results indicate that Ig␣/Ig␤, Syk, SLP-65, and Lyn are unable to activate SHP-1. Previously, SHP-1 was reported to associate with BCR, Lyn, and Syk (29,32,44,45). These molecules may associate with SHP-1 without activating SHP-1 probably by interacting at the site outside the SH2 domains. The association with these molecules appears to be weaker than that with SHP-1-activating molecules because we detected association of SHP-1 with CD72 but not Ig␣/Ig␤, Lyn, or Syk in both J558Lm3 and B cell lines (Figs. 6 and 7). Recently, Ig␣ was shown to have a negative signaling function in B cells (46,47). Ig␣ may carry the negative signaling function either independently of SHP-1 or by activating SHP-1 through a yet unknown molecule that is not expressed in J558Lm3 cells.
Previously, SHP-1 was shown to regulate negatively both phosphorylation and kinase activity of Lyn in B cells and myeloid cells (44,45). However, Lyn is not regulated by BCR nor CD72 in J558Lm3 cells (Fig. 2). In these cells, Lyn may not be regulatory due to lack of CD45 as introduction of CD45 restores BCR ligation-induced enhancement of Lyn activity (39). In contrast, SHP-1 activated by CD72 down-modulates phosphorylation of Ig␣/Ig␤, Syk, and SLP-65 in J558Lm3 cells (Fig. 2). Thus, modulation of Lyn activity is not required for SHP-1mediated negative regulation of Ig␣/Ig␤, Syk, and SLP-65, although dephosphorylation of Lyn may play a role in downmodulation of BCR signaling by SHP-1 in normal B cells. Previously, Dustin et al. (32) demonstrated that co-expression of wild-type SHP-1 together with Syk reduces phosphorylation of Syk in insect cells, suggesting that Syk is a substrate of SHP-1. Because Syk phosphorylates Ig␣/Ig␤ in vitro (48), inhibition of Syk activity may reduce phosphorylation of Ig␣/Ig␤. Alternatively but not mutually exclusively, SHP-1 may directly dephosphorylate Ig␣/Ig␤. Since Syk is phosphorylated by association with phosphorylated Ig␣/Ig␤ (49,50), dephosphorylation of Ig␣/Ig␤ may reduce activation and phosphorylation of Syk even in the absence of direct dephosphorylation of Syk by SHP-1. Thus, SHP-1 appears to dephosphorylate Ig␣/Ig␤, Syk, or both in B cells. Since SLP-65 is a substrate of Syk (3-5), SHP-1 may reduce phosphorylation of SLP-65 by inactivating Syk. Yet SLP-65 may also be a substrate of SHP-1 as suggested by Mizuno et al. (51). Both Ig␣/Ig␤ and Syk have been shown to play a crucial role in BCR signaling (1,2). Recent studies have demonstrated that SLP-65 is essential for Ca 2ϩ mobilization and extracellular signal-regulated kinase activation by BCR ligation (3,5,8,9). Dephosphorylation of these proximal signaling molecules may be involved in CD72-induced down-modulation of BCR signaling events such as Ca 2ϩ mobilization and extracellular signal-regulated kinase activation (16,17). These distal signaling events are not induced in J558Lm3 and cannot be assessed in these cells. It is also possible that downmodulation of BCR signaling by CD72 involves other SHP-1 substrates. Moreover, other BCR co-receptors such as CD5 may also activate SHP-1 to coordinately control BCR signaling (52). The J558Lm3 cells may provide a useful tool to address these questions in the future.