CD22 is an inhibitory B cell coreceptor that regulates B cell development and activation by downregulating BCR signaling through activation of SH2-containing protein tyrosine phosphatase-1 (SHP-1). CD22 recognizes α2,6 sialic acid as a specific ligand and interacts with α2,6 sialic acid-containing membrane molecules, such as CD45, IgM, and CD22, expressed on the same cell. Functional regulation of CD22 by these endogenous ligands enhances BCR ligation-induced signaling and is essential for normal B cell responses to Ags. In this study, we demonstrate that CD45 plays a crucial role in CD22-mediated inhibition of BCR ligation-induced signaling. However, disruption of ligand binding of CD22 enhances CD22 phosphorylation, a process required for CD22-mediated signal inhibition, upon BCR ligation in CD45−/− as well as wild-type mouse B cells but not in mouse B cells expressing a loss-of-function mutant of SHP-1. This result indicates that SHP-1 but not CD45 is required for ligand-mediated regulation of CD22. We further demonstrate that CD22 is a substrate of SHP-1, suggesting that SHP-1 recruited to CD22 dephosphorylates nearby CD22 as well as other substrates. CD22 dephosphorylation by SHP-1 appears to be augmented by homotypic CD22 clustering mediated by recognition of CD22 as a ligand of CD22 because CD22 clustering increases the number of nearby CD22. Our results suggest that CD22 but not CD45 is an endogenous ligand of CD22 that enhances BCR ligation-induced signaling through SHP-1–mediated dephosphorylation of CD22 in CD22 clusters.

This article is featured in Top Reads, p.2501

CD22 (also known as Siglec-2) is an inhibitory B cell coreceptor that negatively regulates BCR signaling (13). CD22 contains four immunoreceptor tyrosine–based inhibition motifs (ITIMs) in the cytoplasmic tail (4). Upon BCR ligation, these ITIMs are phosphorylated by the Src family kinase Lyn, associated with BCR (5), and recruit SH2-containing protein tyrosine phosphatase-1 (SHP-1, also known as PTPN6), which negatively regulates BCR signaling by inactivating BCR-associated kinases Syk (6) and Lyn (7). The extracellular region of CD22 contains a lectin domain that recognizes α2,6 sialic acid as a specific ligand (13). Vertebrate cells, including mouse B cells, express various glycoproteins and glycolipids containing α2,6 sialic acid. CD22 on the B cell surface is occupied by endogenous α2,6 sialic acids expressed on the same cell (cis ligands) (8), although CD22 can also interact with α2,6 sialic acids expressed on other cells (trans ligands) (9). Interaction with endogenous ligands has been shown to regulate signaling function of CD22. Studies using B cells deficient in the sialyl transferase ST6GalI required for synthesis of α2,6 sialic acid (10, 11) and B cells that express the CD22-R130E mutant deficient in binding to α2,6 sialic acid (12) showed that endogenous ligands of CD22 upregulate BCR ligation-induced BCR signaling by suppressing CD22-mediated signal inhibition. Upregulation of BCR signaling by ligand interaction of CD22 appears to be crucial in Ab response to Ags because ST6GalI−/− mice show impaired Ab responses after immunization (10).

Because lectins generally bind to glycan ligands with low affinity, identification of glycan ligands by conventional methods, such as immunoprecipitation, is often impossible. An early study using chemical cross-linking showed that CD22 associates with membrane-bound IgM, CD45 (also known as PTPRC), and CD22 itself (13), although this study failed to show that this interaction depends on sialic acid. Later, in situ photoaffinity cross-linking of α2,6 sialic acid with proteins revealed that CD22 associates with CD22 by recognizing α2,6 sialic acid on this molecule (14). Previously, we demonstrated by proximity labeling using tyramide that IgM, CD45, and CD22 are present in close proximity of CD22 on the cell surface of wild-type but not ST6GalI−/− B cells (15). Our results indicate that IgM, CD45, and CD22 associate with CD22 on the B cell surface in an α2,6 sialic acid–dependent manner and are therefore cis ligands of CD22.

CD45 is a receptor-type protein tyrosine phosphatase (PTP) expressed in all the hematopoietic cell types, except for erythrocytes (16). CD45 is required for activation of Src family kinases, such as Lck and Lyn, by dephosphorylating the inhibitory phosphorylation site. Mice deficient in CD45 fail to generate mature T cells, probably because of signaling defect. Mature B cells are generated in these mice and show relatively good BCR signaling (17, 18), in part, because of the role of receptor-type tyrosine phosphatase CD148 (PTPRJ) is redundant with CD45 (19). CD45 is abundant and heavily glycosylated on the cell surface. Coughlin et al. (20) demonstrated that expression of noncatalytic CD45 rescues the phenotype of CD45−/− B cells caused by reduced tonic signaling, suggesting that CD45 enhances tonic signaling by a noncatalytic mechanism. Because this noncatalytic function of CD45 depends on CD22, they proposed that CD45 rich in α2,6 sialic acid competes α2,6 sialic acid–dependent interaction of CD22 with BCR on the B cell surface, resulting in less efficient CD22-mediated signal inhibition. Therefore, CD45 appears to be a CD22 cis ligand that negatively regulates CD22 function.

In this study, we addressed molecular mechanisms for ligand-mediated regulation of CD22 upon BCR ligation using the synthetic sialoside GSC718 (21) and its derivative GSC839 (22), both of which specifically inhibit ligand binding of CD22. We demonstrate that disruption of ligand binding of CD22 enhances CD22 phosphorylation upon BCR ligation in CD45−/− B cells as well as wild-type B cells but not in B cells expressing a loss-of-function mutant of SHP-1, indicating that SHP-1 but not CD45 is required for ligand-mediated regulation of CD22. We further demonstrate that CD22 is a substrate of SHP-1. Thus, CD22 appears to be dephosphorylated by SHP-1 recruited to nearby CD22 molecules, and ligand recognition of CD22 augments SHP-1–mediated dephosphorylation of CD22 by enhancing association of CD22 with other CD22 molecules that recruit SHP-1.

C57BL/6 mice were purchased from Sankyo Labo Service. CD45−/− mice (18) on C57BL/6 background were bred and maintained in the animal facility of Tokyo Medical and Dental University under specific pathogen–free conditions. Mice were used at 8- to 15-wk old. Experiments were approved by the Institutional Animal Care and Use Committee of Tokyo Medical and Dental University and were performed according to our institutional guidelines.

Mouse spleen B cells were prepared as described previously (23). The mouse B cell line CH1 cells (a kind gift of Dr. K. Udaka) was cultured in IMDM (FUJIFILM Wako Chemicals) containing 10% FCS, 50 μM 2-mercaptethanol, and 1% penicillin/streptomycin. The mouse B cell line A20 and human monocyte line U937 were cultured in RPMI 1640 (Nacalai Tesque) supplemented with 10% FCS (Biowest), sodium pyruvate (Life Technologies), MEM, nonessential amino acids, and 50 μM 2-ME (Nacalai Tesque). The retrovirus packaging cell line PLAT-E (a kind gift of Dr. T. Kitamuta) was maintained in DMEM (high glucose) (044-29765; FUJIFILM Wako Chemicals) supplemented with 10% FCS and 1% penicillin/streptomycin (Nacalai Tesque). Synthetic sialosides GSC718 and GSC839 were synthesized as described previously (21, 22).

For construction of pMX-SHP1C/S-ires-mCD8 encoding myc-tagged SHP-1 C453S (SHP-1C/S) and mouse CD8, a NotI-SalI fragment containing mouse CD8 derived from pLy2.22 (24) (a kind gift of Dr. H. Nakauchi) was ligated to NotI-SalI–opened pMX-ires-GFP (25) (a kind gift of Dr. T. Kitamura) to replace GFP cDNA by mouse CD8 cDNA (pMX-ires-mCD8). An EcoRI-SalI fragment encoding myc-tagged SHP-1C/S was isolated from pMKIT-SHP-1C/S-Myc (26) and cloned into EcoRI-XhoI–opened pMX-ires-mCD8.

For construction of pMX-SHP1C-ires-hCD8 and pMX-SHP1C(C/S)-ires-hCD8 encoding the myc-tagged catalytic domains of wild-type SHP-1 and SHP-1C453S mutant, respectively, together with human CD8, a ClaI-SacII DNA fragment containing the catalytic domain of SHP-1 was amplified by PCR using a set of primers (forward, 5′-AAAATCGATGCCTTTGTCTACCTGCGGCAGCC-3′ and reverse, 5′- AGGCCGCGGCGTCACTTCCTCTTGAGAGAAC-3′) and pBlueScriptSHP-1 (27) as a template. A ClaI-SacII fragment containing the catalytic domain of SHP-1 with the C453S mutation was amplified by PCR using the same primer set and pMKIT-SHP-1C/S-Myc (26) as a template. A BamHI-ClaI DNA fragment containing seven-times myc-tag was amplified by PCR using a set of primers (forward, 5′-AAGGATCCGCCACCATGGGGTCGGAACAGAAACTTATTTCTG-3′ and reverse, 5′-TTATCGATTCCACCGGGTCGAAGATCTTCTTCAG-3′) and pBK-Myc-tagx6 (28) (a kind gift of Dr. Y. Yaoita) as a template. This DNA fragment, together with the DNA fragment containing either wild-type or mutated SHP-1 catalytic domain, was cloned into BamHI-SacII–opened pMX-ires-hCD8 (29) (a kind gift of Dr. S. Yamasaki).

For construction of pEF1-CD33-Fc or pEF1-MAG-Fc coding for the V5-tagged mouse CD33 (Siglec-3)-Fc or MAG (Siglec-4)-Fc proteins, respectively, the cDNA fragment encoding the Fc region of human IgG1 (a kind gift of Dr. P. R. Crocker) was cloned into AgeI-opened pEF1/V5-HisA (Invitrogen), and the AgeI site (ACCGGT) downstream of Fc was replaced by ACCTGA using PCR-mediated mutagenesis to create a stop codon (pEF1/V5-hFc). Gateway destination cassette B (Invitrogen) was cloned into EcoRV-opened pBlueScriptII and then subcloned into Acc65I-XbaI–opened pEF1/V5-hFc, resulting in pEF1-DCB-V5-hFc. The cDNA fragments encoding the extracellular domains of mouse CD33 (V-set domain and C2-set domain) and mouse MAG (V-set domain and two C2-set domains) were amplified by RT-PCR using total RNA obtained from the mesenteric lymph node (CD33) and brain (MAG) of C57BL/6 mice and specific primer sets (forward, 5′-ACAAAAAAGCAGGCTTCGCCGCCATGGTGTGGCCA-3′ and reverse, 5′-ACAAGAAAGCTGGGTTCTGGCCTGATTTCCGGGTA-3′ for CD33 and forward, 5′-ACAAAAAAGCAGGCTTGCTAGCCACCATGGTATTC-3′ and reverse, 5′-ACAAGAAAGCTGGGTCCACCACCGTCCCATTCACT-3′ for MAG). These cDNA fragments flanked by Gateway attB1/attB2 sequences were generated by PCR using a set of primers (forward, 5′-GGGGACAAGTTTGTACAAAAAAGCAGGCT-3′ and reverse, 5′-GGGGACCACTTTGTACAAGAAAGCTGGGT-3′). The resulting DNA fragments encoding CD33 and MAG were inserted into pEF1-DCB-V5-hFc using a Gateway cloning system (Invitrogen).

Retrovirus was produced by transfection of PLAT-E cells (30) (a kind gift of Dr. T. Kitamura) with pMX-ires-mCD8, pMX-SHP1C/S-ires-mCD8, pMX-SHP1C-ires-hCD8, and pMX-SHP1C(C/S)-ires-hCD8 by the calcium phosphate precipitation method. CH1 cells are infected by retrovirus. Alternatively, mouse spleen B cells were stimulated with 10 μg/ml LPS (L2630-10MG; Sigma-Aldrich). After 24 and 48 h, mouse spleen B cells were infected with retrovirus and were cultured for additional 24 h.

Mouse spleen B cells (2 × 106) were incubated with 5 μg/ml Fluo-4 AM (Invitrogen) for 30 min. After washing, cells were stimulated with 10 μg/ml F(ab′)2 fragment of goat anti-mouse IgM (115-006-020; Jackson ImmunoResearch Laboratories) with or without 100 μM GSC718. Fluo-4 fluorescence was measured continuously using a CyAN ADP (Dako) for 180 s. Data were analyzed by FlowJo (Tree Star).

Mouse spleen B cells were suspended at 1 × 107 cells/ml in RPMI 1640 (FUJIFILM Wako Chemicals) containing 10% FCS, 2 mM l-glutamine, 50 μM 2-mercaptethanol, and 1% penicillin/streptomycin and incubated at 37°C for 30 min in the presence or absence of 100 μM GSC718 or GSC839. Cells were then stimulated with 13 μg/ml F(ab′)2 fragments of goat anti-mouse IgM, (115-006-020, Jackson ImmunoResearch Laboratories) for 0, 2, and 5 min. Alternatively, CH1 cells expressing the wild-type catalytic domain of SHP-1 (SHP-1C-WT) or the catalytic domain of SHP-1 in which cysteine at the active site is mutated to serine (SHP-1C-C/S) were stimulated with pervanadate (1 mM Na3VO4 and 3 mM H2O2) for 2 min at 37°C. Cells were lysed in a Triton X-100 lysis buffer (20 mM Tris-HCl [pH 8.0] containing 1% Triton X-100, 10% glycerol, 150 mM NaCl, 2 mM EDTA, 0.02% NaN3, 10 µg/ml PMSF, and 1 mM Na3VO4). The lysates were then precipitated with anti-CD22 Ab (F239) (22) using protein G Sepharose beads (GE Healthcare). Immunoprecipitants were separated by SDS-PAGE and transferred to PVDF membranes (Millipore). Membranes were divided into two. Upper membranes were incubated with mouse anti-phosphotyrosine Ab (4G10), followed by reaction with HRP-conjugated rabbit anti-mouse IgG (1030-05, SouthernBiotech). The upper membranes were reprobed with goat anti-mouse CD22 Ab (sc-7032; Santa Cruz Biotechnology) followed by incubation with HRP-conjugated anti-goat IgG (sc-2056; Santa Cruz Biotechnology). Lower membranes were incubated with rabbit anti–SHP-1 Ab (sc-287; Santa Cruz Biotechnology) followed by HRP-conjugated goat anti-rabbit IgG (4030-05; SouthernBiotech). Alternatively, lower membranes were incubated with anti-Myc Ab (9E10) to detect myc-tagged SHP-1C-WT and SHP-1C-C/S, followed by reaction with HRP-conjugated rabbit anti-mouse IgG (1030-05; SouthernBiotech). Proteins were then visualized using Chemi-Lumi One L (Nacalai Tesque) and ImageQuant LAS 4000 Mini (GE Healthcare). Proteins were quantified using ImageJ (https://imagej.nih.gov/ij/index.html).

The fusion proteins mouse CD22-Fc, Siglec-1-Fc, Siglec-E-Fc, Siglec-G-Fc (31, 32), and MAF-Fc (33), containing the Fc portion of human IgG and the ligand binding domains of each Siglec were described previously. pEF1-CD33-Fc and pEF1-MAG-Fc were transduced to Lec2 cells (34) deficient in protein sialylation, followed by selection with 1 mg/ml G418 (Nacalai Tesque). The stable transfectants were cultured in CHO-S-SFM II (Invitrogen), and the CD33-Fc or MAG-Fc proteins were purified from the culture supernatant using Protein A Sepharose (Amersham Pharmacia Biotech). Serially diluted Siglec-Fc fusion proteins were incubated with PE-conjugated goat anti-human IgG (SouthernBiotech) at 4°C overnight. A20 or U937 cells (5 × 104) were incubated with 0–100 μM GSC839, and Siglec-Fc proteins were precomplexed with PE-conjugated anti-human IgG at 4°C for 2 h and analyzed by flow cytometry using a FACSCalibur (BD Biosciences).

Data were analyzed by one-way ANOVA followed by Holm–Sidak post hoc test to compare the mean values using GraphPad Prism v. 6. The p values <0.05 were considered as statistically significant.

Because CD45 is a major cis ligand of CD22 (15) and noncatalytically regulates CD22 (20), we addressed whether CD45 is involved in ligand-mediated regulation of CD22. Because GSC718 at more than 1 μM almost completely inhibits ligand binding of CD22 (22), we first examined BCR ligation-induced Ca2+ signaling in the presence or absence of 100 μM GSC718 in CD45+/+ and CD45−/− B cells. Disruption of ligand binding of CD22 by GSC718 reduced Ca2+ flux induced by anti-IgM Ab in CD45+/+ B cells (Fig. 1A), in agreement with the previous findings showing that the endogenous ligands suppress CD22, thereby augmenting BCR signaling (1012). In contrast, GSC718 failed to alter BCR ligation-induced Ca2+ flux in CD45−/− B cells (Fig. 1B), indicating that endogenous ligands of CD22 do not regulate BCR signaling in the absence of CD45.

Next, we analyzed phosphorylation of CD22 and recruitment of SHP-1 to CD22, both of which are required for CD22-mediated signal regulation. Upon BCR ligation, CD22 is strongly phosphorylated and recruits SHP-1 in CD45+/+ B cells (Fig. 2A). In contrast, CD22 is only modestly phosphorylated, and SHP-1 recruitment is not augmented in BCR-ligated CD45−/− B cells (Fig. 2A). This result indicated that CD45 is required for CD22-mediated signal inhibition, probably by activating the Src family tyrosine kinase Lyn (16) that phosphorylates CD22 (5).

We next analyzed the effect of GSC718 on CD22 phosphorylation and recruitment of SHP-1 in CD45+/+ and CD45−/− B cells. In CD45+/+ B cells, treatment with GSC718 enhanced CD22 phosphorylation and recruitment of SHP-1 induced by BCR ligation (Fig. 2B), in agreement with suppression of CD22 by endogenous ligands (1012). Although BCR ligation induces modest phosphorylation of CD22 in CD45−/− B cells (Fig. 2A), treatment with GSC718 clearly augments CD22 phosphorylation and SHP-1 recruitment in these B cells (Fig. 2C). These results indicate that endogenous ligands suppress CD22 phosphorylation and SHP-1 recruitment in CD45−/− B cells as well as in CD45+/+ B cells and suggest that CD45 is dispensable for ligand-mediated regulation of CD22. Therefore, the failure of augmentation of BCR ligation-induced Ca2+ flux by GSC718 in CD45−/− B cells (Fig. 1A) is not because CD45 regulates CD22 in a sialic acid–dependent manner but because of poor CD22-mediated signal inhibition in the absence of CD45 (Fig. 2A) that activates Lyn (16) required for CD22 phosphorylation (5).

GSC839 has the same structure as GSC718, except that GSC839 contains an additional fluorine (22). Previously, we demonstrated that GSC839 binds to CD22 with an affinity similar to that of GSC718 and inhibits ligand binding of CD22. To address binding specificity of GSC839, we prepared recombinant Fc fusion proteins of all the mouse Siglecs, except for Siglec-15 and Siglec-H, and examined inhibition of the binding of Siglec-Fc proteins to cell surface ligands by GSC839, using flow cytometry. Because mouse B cells highly express α2,6 N-glycolyl neuraminic acid, the preferred ligand of mouse CD22, we used the mouse B cell line A20 to address ligand binding of CD22. For other Siglecs, we used the human monocyte line U937. GSC839 at 10 μM or higher almost completely inhibited the binding of CD22-Fc to cell surface ligands, whereas GSC839 at the same concentrations caused no or only marginal inhibition in the binding of the other Siglec-Fc proteins (Fig. 3). This result suggests that GSC839 specifically binds to CD22, thereby efficiently inhibiting the binding of CD22 to its ligands.

To address whether phosphorylation of CD22 is directly regulated by SHP-1 recruited to CD22, we examined whether CD22 is trapped by the substrate-trapping SHP-1 mutant. Mutant PTPs, in which the cysteine at the catalytic site is mutated to serine, bind to the substrates but fail to detach substrate phosphate, resulting in continuous binding to the substrates. Forming a stable complex with these substrate-trapping mutants is one of the most important criteria in determining PTP substrates (35). Because the SH2 domains of SHP-1 bind to the ITIMs in the cytoplasmic region of CD22, we generated cDNA encoding the myc-tagged catalytic domain of SHP-1 alone, in which cysteine at the active site is mutated to serine (SHP-1C-C/S) as a substrate-trapping mutant (Fig. 4A). We also generated cDNA encoding the unmutated catalytic domain of SHP-1 with myc-tag (SHP-1C-WT). We transduced these cDNAs to the B cell line CH1 by retrovirus transduction (Supplemental Fig. 1). Treatment of the transduced CH1 cells with the phosphatase inhibitor pervanadate induced strong phosphorylation of CD22. Both SHP-1C-WT and SHP-1C-C/S were immunoprecipitated with CD22 (Fig. 4B). SHP-1C-C/S was coprecipitated more efficiently than SHP-1C-WT, indicating that CD22 is trapped by the substrate-trapping SHP-1 mutant. This result suggests that CD22 is a substrate of SHP-1 and is directly dephosphorylated by SHP-1.

Because SHP-1 regulates CD22 phosphorylation, we next addressed whether SHP-1 is involved in the ligand-mediated regulation of CD22. To block catalytic activity of SHP-1, we transduced cDNA encoding myc-tagged catalytically inactive SHP-1 C453S (SHP-1C/S) to LPS-stimulated mouse spleen B cells by retrovirus transduction (Supplemental Fig. 2), and ligated BCR on these B cells by anti-IgM in the presence or absence of the synthetic sialoside GSC839, which inhibits ligand binding of CD22. We used GSC839 because of limited amounts of both GSC718 and GSC839. Upon BCR ligation, B cells transduced with SHP-1C/S show a higher level of CD22 phosphorylation than B cells transduced with the empty virus (Fig. 5A, 5B), indicating that inhibition of the catalytic activity of SHP-1 enhances CD22 phosphorylation. This result supports the notion that CD22 is a substrate of SHP-1. Treatment with GSC839 augmented BCR ligation-induced CD22 phosphorylation and recruitment of SHP-1 in B cells transduced with the empty virus (Fig. 5A–C), indicating that vector transduction does not alter ligand-mediated regulation of CD22. Interestingly, treatment of B cells expressing SHP-1C/S with GSC839 did not augment CD22 phosphorylation or SHP-1 recruitment (Fig. 5A–C). This result clearly indicates that SHP-1 is essential for ligand-mediated regulation of CD22 in BCR-ligated B cells.

Although regulation of CD22 by endogenous ligands plays a crucial role in normal B cell responses to Ag stimulation (10), how endogenous ligands regulate CD22 is still elusive because complexity generated by presence of multiple different CD22 ligands makes the analysis of molecular mechanisms difficult. We developed the synthetic sialosides GSC718 and GSC839 as a (to our knowledge) novel tool to address ligand-mediated regulation of CD22. GSC839 has the same structure as GSC718, except that GSC839 contains an additional fluorine compared with GSC718 (22). Both GSC839 and GSC718 specifically bind to CD22 with similar affinity, thereby inhibiting ligand binding of CD22. In this study we demonstrate that treatment of BCR-ligated B cells with these sialosides augments phosphorylation of CD22 and reduces Ca2+ signaling, indicating that endogenous ligands suppress CD22-mediated inhibition of BCR ligation-induced signaling. This is consistent with the previous findings in ST6GalI−/− B cells and CD22 R130E B cells that disruption of ligand binding of CD22 augments CD22-mediated signal inhibition (1012).

Previously, we demonstrated that CD45 is a major cis ligand of CD22, together with IgM and CD22 itself (13). If CD45 plays a role in the regulation of CD22 by endogenous ligands, disruption of ligand binding of CD22 by sialosides reduces BCR ligation-induced Ca2+ flux or enhances CD22 phosphorylation less efficiently in CD45−/− B cells. In this study, we demonstrate that treatment with GSC718 does not alter BCR ligation-induced Ca2+ flux in CD45−/−B cells. However, GSC718 augments phosphorylation of CD22 and its recruitment of SHP-1 in these cells. This result indicates that CD45 is not a cis ligand essential for ligand-mediated regulation of CD22 but plays another role in the regulation of BCR signaling by CD22. In this study, we show that phosphorylation of CD22 induced by BCR ligation is markedly reduced in CD45−/− B cells compared with CD45+/+ B cells. Because CD22 is known to be phosphorylated by the Src family kinase Lyn (5), the crucial role of CD45 in CD22 phosphorylation is consistent with the previous findings that CD45 activates Src family kinases by dephosphorylating an inhibitory phosphorylation site (16). In CD45−/− B cells, CD22 is phosphorylated inefficiently, probably because of poor Lyn activation, thereby regulating BCR signaling only weakly. Inefficient CD22-mediated signal inhibition may cause failure of GSC718 in regulating BCR ligation-induced Ca2+ flux in CD45−/− B cells, although GSC718 disrupts ligand-mediated regulation of CD22. Taken together, our results strongly suggest that CD45 is not a CD22 cis ligand that regulates CD22-mediated downregulation of BCR ligation-induced signaling, although CD45 plays a crucial role in CD22-mediated signal inhibition by activating the Lyn required for CD22 phosphorylation.

In this study, we demonstrate that CD22 is a substrate of SHP-1 because (1) the substrate-trapping mutant of SHP-1 strongly binds to CD22 (Fig. 3) and (2) CD22 phosphorylation is augmented by expression of SHP-1C/S, a dominant negative form of SHP-1 (Fig. 4). This result indicates that SHP-1 recruited to and activated by CD22 dephosphorylates nearby CD22. We also demonstrate in this study that GSC718 no longer enhances CD22 phosphorylation in B cells expressing SHP-1C/S (Fig. 4), indicating that ligand interaction of CD22 is essential for SHP-1–mediated dephosphorylation of CD22. Using super resolution microscopy, Gasparrini et al. (36) previously demonstrated that recognition of CD22 as a cis ligand causes homo oligomeric CD22 clusters on the B cell surface. Clustering of CD22 may recruit a higher number of CD22 to the close vicinity of SHP-1 recruited to and activated by CD22, thereby augmenting dephosphorylation of CD22 by SHP-1 (Fig. 6A, 6B). Therefore, SHP-1 appears to be involved in ligand-mediated regulation of CD22 by dephosphorylating CD22 in CD22 clusters formed by ligand recognition of CD22. Gasparrini et al. (36) also addressed a role of sialic acid–dependent homotypic clustering of CD22 in the ligand-mediated regulation of CD22. They demonstrated using super resolution microscopy and single-particle tracking that CD22-R130E B cells show smaller CD22 clusters with high lateral mobility compared with wild-type CD22 and proposed a mechanism for ligand-mediated suppression of CD22, in which sialic acid–dependent homotypic clustering of CD22 inhibits mobility of CD22, thereby reducing interaction of CD22 with BCR. However, our finding that ligand-mediated regulation of CD22 is abrogated by blocking SHP-1 activation suggests that events upstream of SHP-1 activation, such as interaction of CD22 with BCR, are not involved in ligand-mediated regulation of CD22. Therefore, sialic acid–dependent regulation of CD22 mobility may not play a major role in the ligand-mediated regulation of CD22.

Because SHP-1 is known to downregulate BCR signaling by dephosphorylating cellular substrates, such as Lyn and Syk (6, 7), our finding that CD22 is a substrate of SHP-1 indicates that SHP-1 activated by CD22 has two opposing effects on BCR signaling. SHP-1 dephosphorylates Lyn and Syk involved in BCR signaling to downmodulate BCR signaling and also dephosphorylates CD22 in the close vicinity, thereby limiting CD22-mediated signal inhibition. The balance of these opposing effects of SHP-1 appears to be regulated by the density of phosphorylated CD22, as this determines the number of phosphorylated CD22 in the close proximity of SHP-1 recruited and activated by CD22. This model is supported by our result that SHP-1–mediated dephosphorylation of CD22 requires ligand interaction of CD22 (Fig. 5) that induces CD22 clustering (Fig. 6A, 6B). This model is also supported by our recent finding that ligand recognition of CD22, rather, reduces tonic signaling (C. Akatsu, A. Alborzian Deh Sheikh, N. Matsubara, H. Takematsu, A. Schweizer, H.H.M. Abdu-Allah, T.F. Tedder, L. Nitschke, H. Ishida, and T. Tsubata, submitted for publication), although it is well established that ligand interaction of CD22 enhances BCR ligation-induced signaling (1012). In the absence of BCR ligation, only a tiny fraction of CD22 is phosphorylated. SHP-1 recruited to CD22 ITIMs may not have phosphorylated CD22 molecules in the vicinity regardless of clustering of CD22 (Fig. 6C, 6D). Therefore, ligand-dependent CD22 clustering appears to augment SHP-1–induced dephosphorylation of CD22 in BCR-ligated but not unstimulated B cells. Interaction of CD22 with other ligands, such as BCR, may augment CD22-mediated signal inhibition. As a consequence, ligand recognition of CD22 may augment CD22 function in the absence of SHP-1–mediated CD22 dephosphorylation in unstimulated B cells because of low phosphorylated CD22 density in CD22 clusters. In BCR-ligated B cells, high phosphorylated CD22 density in CD22 clusters may induce SHP-1–mediated CD22 dephosphorylation, which dominantly inhibits CD22-mediated signal inhibition. Taken together, the balance between the positive and negative regulation of BCR signaling by different CD22 cis ligands appears to be regulated by the density of phosphorylated CD22 that alters the dominant substrate of SHP-1 recruited and activated by CD22. Further studies are required for full elucidation of the complexed regulation of CD22 by endogenous ligands.

We thank Dr. H. Nakauchi (the University of Tokyo), Dr. T. Kitamura (the University of Tokyo) Dr. S. Yamasaki (Osaka University) and Dr. K. Udaka (Kochi University) for reagents, Dr. H. Kishi (Toyama University) for CD45−/− mice, and Dr. J.-Y. Wang (Fudan University) for discussion.

The work was supported by Japan Society for Promotion of Science Grants-in-Aid for Scientific Research 26293062, 18H02610 (T.T.), and 17K17695 (C.A.).

The online version of this article contains supplemental material.

Abbreviations used in this article

ITIM

immunoreceptor tyrosine–based inhibition motif

PTP

protein tyrosine phosphatase

SHP-1

SH2-containing protein tyrosine phosphatase 1

SHP-1C-C/S

catalytic domain of SHP-1 in which cysteine at the active site is mutated to serine

SHP-1C-WT

wild-type catalytic domain of SHP-1

1.
Meyer
S. J.
,
A. T.
Linder
,
C.
Brandl
,
L.
Nitschke
.
2018
.
B cell Siglecs-news on signaling and its interplay with ligand binding.
Front. Immunol.
9
:
2820
.
2.
Enterina
J. R.
,
J.
Jung
,
M. S.
Macauley
.
2019
.
Coordinated roles for glycans in regulating the inhibitory function of CD22 on B cells.
Biomed. J.
42
:
218
232
.
3.
Tsubata
T.
2018
.
Ligand recognition determines the role of inhibitory B cell co-receptors in the regulation of B cell homeostasis and autoimmunity.
Front. Immunol.
9
:
2276
.
4.
Zhu
C.
,
M.
Sato
,
T.
Yanagisawa
,
M.
Fujimoto
,
T.
Adachi
,
T.
Tsubata
.
2008
.
Novel binding site for Src homology 2-containing protein-tyrosine phosphatase-1 in CD22 activated by B lymphocyte stimulation with antigen. [Published erratum appears in 2008 J. Biol Chem. 283: 5972.]
J. Biol. Chem.
283
:
1653
1659
.
5.
Smith
K. G.
,
D. M.
Tarlinton
,
G. M.
Doody
,
M. L.
Hibbs
,
D. T.
Fearon
.
1998
.
Inhibition of the B cell by CD22: a requirement for Lyn.
J. Exp. Med.
187
:
807
811
.
6.
Dustin
L. B.
,
D. R.
Plas
,
J.
Wong
,
Y. T.
Hu
,
C.
Soto
,
A. C.
Chan
,
M. L.
Thomas
.
1999
.
Expression of dominant-negative src-homology domain 2-containing protein tyrosine phosphatase-1 results in increased Syk tyrosine kinase activity and B cell activation.
J. Immunol.
162
:
2717
2724
.
7.
Somani
A. K.
,
K.
Yuen
,
F.
Xu
,
J.
Zhang
,
D. R.
Branch
,
K. A.
Siminovitch
.
2001
.
The SH2 domain containing tyrosine phosphatase-1 down-regulates activation of Lyn and Lyn-induced tyrosine phosphorylation of the CD19 receptor in B cells.
J. Biol. Chem.
276
:
1938
1944
.
8.
Razi
N.
,
A.
Varki
.
1998
.
Masking and unmasking of the sialic acid-binding lectin activity of CD22 (Siglec-2) on B lymphocytes.
Proc. Natl. Acad. Sci. USA
95
:
7469
7474
.
9.
Collins
B. E.
,
O.
Blixt
,
A. R.
DeSieno
,
N.
Bovin
,
J. D.
Marth
,
J. C.
Paulson
.
2004
.
Masking of CD22 by cis ligands does not prevent redistribution of CD22 to sites of cell contact.
Proc. Natl. Acad. Sci. USA
101
:
6104
6109
.
10.
Hennet
T.
,
D.
Chui
,
J. C.
Paulson
,
J. D.
Marth
.
1998
.
Immune regulation by the ST6Gal sialyltransferase.
Proc. Natl. Acad. Sci. USA
95
:
4504
4509
.
11.
Collins
B. E.
,
B. A.
Smith
,
P.
Bengtson
,
J. C.
Paulson
.
2006
.
Ablation of CD22 in ligand-deficient mice restores B cell receptor signaling.
Nat. Immunol.
7
:
199
206
.
12.
Müller
J.
,
I.
Obermeier
,
M.
Wöhner
,
C.
Brandl
,
S.
Mrotzek
,
S.
Angermüller
,
P. C.
Maity
,
M.
Reth
,
L.
Nitschke
.
2013
.
CD22 ligand-binding and signaling domains reciprocally regulate B-cell Ca2+ signaling.
Proc. Natl. Acad. Sci. USA
110
:
12402
12407
.
13.
Zhang
M.
,
A.
Varki
.
2004
.
Cell surface sialic acids do not affect primary CD22 interactions with CD45 and surface IgM nor the rate of constitutive CD22 endocytosis.
Glycobiology
14
:
939
949
.
14.
Han
S.
,
B. E.
Collins
,
P.
Bengtson
,
J. C.
Paulson
.
2005
.
Homomultimeric complexes of CD22 in B cells revealed by protein-glycan cross-linking.
Nat. Chem. Biol.
1
:
93
97
.
15.
Alborzian Deh Sheikh
A.
,
C.
Akatsu
,
A.
Imamura
,
H. H. M.
Abdu-Allah
,
H.
Takematsu
,
H.
Ando
,
H.
Ishida
,
T.
Tsubata
.
2018
.
Proximity labeling of cis-ligands of CD22/Siglec-2 reveals stepwise α2,6 sialic acid-dependent and -independent interactions.
Biochem. Biophys. Res. Commun.
495
:
854
859
.
16.
Hermiston
M. L.
,
J.
Zikherman
,
J. W.
Zhu
.
2009
.
CD45, CD148, and Lyp/Pep: critical phosphatases regulating Src family kinase signaling networks in immune cells.
Immunol. Rev.
228
:
288
311
.
17.
Benatar
T.
,
R.
Carsetti
,
C.
Furlonger
,
N.
Kamalia
,
T.
Mak
,
C. J.
Paige
.
1996
.
Immunoglobulin-mediated signal transduction in B cells from CD45-deficient mice.
J. Exp. Med.
183
:
329
334
.
18.
Byth
K. F.
,
L. A.
Conroy
,
S.
Howlett
,
A. J.
Smith
,
J.
May
,
D. R.
Alexander
,
N.
Holmes
.
1996
.
CD45-null transgenic mice reveal a positive regulatory role for CD45 in early thymocyte development, in the selection of CD4+CD8+ thymocytes, and B cell maturation.
J. Exp. Med.
183
:
1707
1718
.
19.
Zhu
J. W.
,
T.
Brdicka
,
T. R.
Katsumoto
,
J.
Lin
,
A.
Weiss
.
2008
.
Structurally distinct phosphatases CD45 and CD148 both regulate B cell and macrophage immunoreceptor signaling.
Immunity
28
:
183
196
.
20.
Coughlin
S.
,
M.
Noviski
,
J. L.
Mueller
,
A.
Chuwonpad
,
W. C.
Raschke
,
A.
Weiss
,
J.
Zikherman
.
2015
.
An extracatalytic function of CD45 in B cells is mediated by CD22.
Proc. Natl. Acad. Sci. USA
112
:
E6515
E6524
.
21.
Abdu-Allah
H. H.
,
K.
Watanabe
,
K.
Hayashizaki
,
C.
Takaku
,
T.
Tamanaka
,
H.
Takematsu
,
Y.
Kozutsumi
,
T.
Tsubata
,
H.
Ishida
,
M.
Kiso
.
2009
.
Potent small molecule mouse CD22-inhibitors: exploring the interaction of the residue at C-2 of sialic acid scaffold.
Bioorg. Med. Chem. Lett.
19
:
5573
5575
.
22.
Matsubara
N.
,
A.
Imamura
,
T.
Yonemizu
,
C.
Akatsu
,
H.
Yang
,
A.
Ueki
,
N.
Watanabe
,
H.
Abdu-Allah
,
N.
Numoto
,
H.
Takematsu
, et al
2018
.
CD22-binding synthetic sialosides regulate B lymphocyte proliferation through CD22 ligand-dependent and independent pathways, and enhance antibody production in mice.
Front. Immunol.
9
:
820
.
23.
Nomura
T.
,
H.
Han
,
M. C.
Howard
,
H.
Yagita
,
H.
Yakura
,
T.
Honjo
,
T.
Tsubata
.
1996
.
Antigen receptor-mediated B cell death is blocked by signaling via CD72 or treatment with dextran sulfate and is defective in autoimmunity-prone mice.
Int. Immunol.
8
:
867
875
.
24.
Nakauchi
H.
,
G. P.
Nolan
,
C.
Hsu
,
H. S.
Huang
,
P.
Kavathas
,
L. A.
Herzenberg
.
1985
.
Molecular cloning of Lyt-2, a membrane glycoprotein marking a subset of mouse T lymphocytes: molecular homology to its human counterpart, Leu-2/T8, and to immunoglobulin variable regions.
Proc. Natl. Acad. Sci. USA
82
:
5126
5130
.
25.
Onishi
M.
,
S.
Kinoshita
,
Y.
Morikawa
,
A.
Shibuya
,
J.
Phillips
,
L. L.
Lanier
,
D. M.
Gorman
,
G. P.
Nolan
,
A.
Miyajima
,
T.
Kitamura
.
1996
.
Applications of retrovirus-mediated expression cloning.
Exp. Hematol.
24
:
324
329
.
26.
Adachi
T.
,
J.
Wienands
,
C.
Wakabayashi
,
H.
Yakura
,
M.
Reth
,
T.
Tsubata
.
2001
.
SHP-1 requires inhibitory co-receptors to down-modulate B cell antigen receptor-mediated phosphorylation of cellular substrates.
J. Biol. Chem.
276
:
26648
26655
.
27.
Adachi
T.
,
H.
Flaswinkel
,
H.
Yakura
,
M.
Reth
,
T.
Tsubata
.
1998
.
The B cell surface protein CD72 recruits the tyrosine phosphatase SHP-1 upon tyrosine phosphorylation.
J. Immunol.
160
:
4662
4665
.
28.
Nakajima
K.
,
Y.
Yaoita
.
1997
.
Construction of multiple-epitope tag sequence by PCR for sensitive Western blot analysis.
Nucleic Acids Res.
25
:
2231
2232
.
29.
Yamasaki
S.
,
E.
Ishikawa
,
M.
Sakuma
,
K.
Ogata
,
K.
Sakata-Sogawa
,
M.
Hiroshima
,
D. L.
Wiest
,
M.
Tokunaga
,
T.
Saito
.
2006
.
Mechanistic basis of pre-T cell receptor-mediated autonomous signaling critical for thymocyte development.
Nat. Immunol.
7
:
67
75
.
30.
Kitamura
T.
,
Y.
Koshino
,
F.
Shibata
,
T.
Oki
,
H.
Nakajima
,
T.
Nosaka
,
H.
Kumagai
.
2003
.
Retrovirus-mediated gene transfer and expression cloning: powerful tools in functional genomics.
Exp. Hematol.
31
:
1007
1014
.
31.
Naito
Y.
,
H.
Takematsu
,
S.
Koyama
,
S.
Miyake
,
H.
Yamamoto
,
R.
Fujinawa
,
M.
Sugai
,
Y.
Okuno
,
G.
Tsujimoto
,
T.
Yamaji
, et al
2007
.
Germinal center marker GL7 probes activation-dependent repression of N-glycolylneuraminic acid, a sialic acid species involved in the negative modulation of B-cell activation.
Mol. Cell. Biol.
27
:
3008
3022
.
32.
Naito-Matsui
Y.
,
S.
Takada
,
Y.
Kano
,
T.
Iyoda
,
M.
Sugai
,
A.
Shimizu
,
K.
Inaba
,
L.
Nitschke
,
T.
Tsubata
,
S.
Oka
, et al
2014
.
Functional evaluation of activation-dependent alterations in the sialoglycan composition of T cells.
J. Biol. Chem.
289
:
1564
1579
.
33.
Naito-Matsui
Y.
,
L. R.
Davies
,
H.
Takematsu
,
H. H.
Chou
,
P.
Tangvoranuntakul
,
A. F.
Carlin
,
A.
Verhagen
,
C. J.
Heyser
,
S. W.
Yoo
,
B.
Choudhury
, et al
2017
.
Physiological exploration of the long term evolutionary selection against expression of N-glycolylneuraminic acid in the brain.
J. Biol. Chem.
292
:
2557
2570
.
34.
Deutscher
S. L.
,
N.
Nuwayhid
,
P.
Stanley
,
E. I.
Briles
,
C. B.
Hirschberg
.
1984
.
Translocation across Golgi vesicle membranes: a CHO glycosylation mutant deficient in CMP-sialic acid transport.
Cell
39
:
295
299
.
35.
Tiganis
T.
,
A. M.
Bennett
.
2007
.
Protein tyrosine phosphatase function: the substrate perspective.
Biochem. J.
402
:
1
15
.
36.
Gasparrini
F.
,
C.
Feest
,
A.
Bruckbauer
,
P. K.
Mattila
,
J.
Müller
,
L.
Nitschke
,
D.
Bray
,
F. D.
Batista
.
2016
.
Nanoscale organization and dynamics of the siglec CD22 cooperate with the cytoskeleton in restraining BCR signalling.
EMBO J.
35
:
258
280
.

The authors have no financial conflicts of interest.

Supplementary data