Ligation of HLA-DR Molecules on B Cells Induces Enhanced Expression of IgM Heavy Chain Genes in Association with Syk Activation*

Signals transmitted by class II major histocompatibility complex are important regarding cell function related to antigen presentation. We examined effects of DR-mediated signaling on Ig production from B cells. Cross-linking HLA-DR molecules on B cells by solid-phase anti-HLA-DR monoclonal antibodies, led to an increased production of IgM, without proliferation or apoptosis. This event was accompanied by an enhanced expression of both membrane- and secretory-type IgM heavy chain mRNA. When peptide-pulsed B cells were co-incubated with an HLA-DR-restricted T cell clone treated by the protein synthesis inhibitor emetine, peptide-induced de novo expression of lymphokines and cell-surface molecules on T cells can be neglected. CD40-CD154 interaction was not involved in IgM enhancement, in such a system. The protein-tyrosine kinase inhibitors and the Syk inhibitor piceatannol, but not the Src inhibitor PP2 had a marked inhibitory effect on IgM secretion. Furthermore, ligation of HLA-DR on B cells using the F(ab′)2 fragment of anti-DR monoclonal antibody, enhanced Syk activity. Our data suggest that HLA-DR on B cells not only present antigenic peptides to T cells, but also up-regulate IgM production, in association with Syk activation and without the involvement of Src kinases, hence the possible physiological relevance of Src-independent Syk activation.

We earlier reported that interactions between a T cell clone and monocyte via altered T cell receptor (TCR) 1 ligands, affect monocyte responses to produce IL-12, events which lead to specific up-regulation of interferon-␥ production from T cells (1). Thus, signals transmitted to monocytes via HLA molecules are involved in determining immune response patterns. It is highly conceivable that signals transmitted by class II MHC molecules in B cells, in regulating antigen-presenting cell function during cognate T-B cell interactions, are important, for the following reasons: (a) cross-linking class II molecules induces an increase in intracellular calcium and cAMP in mouse or human B cell lines (2)(3)(4)(5); (b) class II MHC-mediated signals lead to homotypic aggregation of B cells (6); (c) cross-linking of HLA-DR molecules on B cells induces apoptosis (7); (d) class II MHC molecules without the intracellular domain expressed on B lymphoma cells will not lead to an increase in cAMP and subsequent CD80 up-regulation, when stimulated with a CD28-expressing autoreactive T hybridoma cells (8); (e) cytoplasmic domain mutants of class II MHC abrogate generation of intracelluar cAMP (9) and translocation of PKC (10); and (f) cross-linking of HLA-DR molecules expressed on B cells induces phosphorylation of Src family kinases (Lyn, Fgr) (11) and Syk (12). Although functional consequences of such DR-mediated signaling events induced by T cells are largely unknown, these observations do raise the possibility that signaling through class II MHC molecules may affect B cell responses, including Ig secretion, upon TCR-TCR ligand interaction.
A number of protein-tyrosine kinases (PTKs) identified in lymphocytes can be classified into cytoplasmic (e.g. Syk, Btk), membrane-binding (e.g. Lyn, Lck, Fyn), and receptor types (e.g. epidermal growth factor receptor, nerve growth factor receptor). In B cells, a number of PTKs acting downstream of B cell receptors (BCRs) and several Fc receptors lead to activation of B cells. Mechanisms for Syk activation have been extensively examined (13). Src family kinases first phosphorylate Ig␣/Ig␤ ITAMs following BCR engagement, and then Syk is recruited to the doubly-phosphorylated ITAM and is activated by Src family kinases. Activated Syk molecules in B cells function in a manner analogous to ZAP-70 in T cells, and play a crucial role in B cell activation (14,15). Recent studies have also suggested an alternative activating model of Syk that is independent of Src family kinases (16 -18).
We now report that DR-generated signals induce IgM production from B cells, in association with Syk activation, but not Src family kinases.
B Cell Preparations and Cell Culture-Peripheral B cells were isolated, using Stemsep Kits (StemCell Technologies Inc., Vancouver, British Columbia, Canada). Briefly, PBMC were freshly prepared from healthy adult donors, using Ficoll-Paque, then were incubated with a mixture of Abs (anti-glycoprotein A, anti-CD3, anti-CD14, anti-CD16, and anti-CD56 Abs). B cells were negatively selected in a magnetic column effluent; the isolated population was Ͼ90% CD19-positive (data not shown). All cells were cultured in RPMI 1640 medium (Life Technologies, Inc.) supplemented with 2 mM L-glutamine, 100 units/ml penicillin, 100 g/ml streptomycin, and 10% FCS (Life Technologies, Inc.) in 96-well flat-bottomed culture plates (Falcon, Becton Dickinson, Lincoln Park, NJ). The human B lymphoblastoid cell line LD2B was positively stained with anti-IgM Ab and anti-DR Abs, homozygous for HLA-DRB1*1501-DRB5*0101 and was distributed by the 11th International Histocompatibility Workshop (23).
Stimulation of B Cells by Anti-DR mAbs-Anti-DR mAbs (HU4 and L243), mouse IgG2a, and bovine serum albumin were coated onto 96-well flat-bottomed culture plates for 2 days at 10 g/ml PBS. After extensive washing of the plates with PBS, B cells were added at 5 ϫ 10 4 cells/well, and incubated at 37°C in a CO 2 incubator. Culture supernatants were collected on days 0 -9, and stored in aliquots at Ϫ80°C until determinations of Ig concentrations. The supernatants for 0 -9-day cultures were assayed for IgM by sandwich ELISA; IgM production reached the plateau level on day 5 or day 7 (data not shown). Therefore, 5-day culture experiments were done in the subsequent experiments. To examine B cell proliferation, B cells were cultured for 72 h, in the presence of 1 Ci/well [ 3 H]thymidine during the final 16-h period, and the incorporated radioactivity was measured by liquid scintillation counting.
Apoptosis Assay-Peripheral B cells were incubated in a 48-well flat-bottomed culture plate in the presence of soluble anti-DR mAb L243 (10 g/ml), soluble mouse Ig control, or immobilized Igs. After 24-h incubation, cells were examined for phosphatidylserine expression on the cell surface with fluorescein isothiocyanate-conjugated annexin V and DNA staining with propidium iodide, using the MEBCYTO apoptosis kits (MBL, Nagoya, Japan) and flow cytometry with a FAC-Scan instrument (Becton Dickinson, Mountain View, CA).
Stimulation of B Cells by Emetine-treated T Cells-T cell clone BC20.7 (21) was treated with 90 g/ml emetine for 1 h at 37°C and washed three times with RPMI 1640 medium (24). The cells were re-suspended in culture medium, followed by 3 h of incubation at 37°C, and then washed three times with RPMI 1640 medium and co-cultured with peptide-pulsed or mock-pulsed B cells prepared from PBMC. Culture supernatants after 5-day incubation were collected and subjected to ELISA. In some experiments, the peptide-pulsed peripheral B cells were cultured for 5 days, with emetine-treated T cells in the presence of kinase inhibitors, at several concentrations. These inhibitors were dissolved in Me 2 SO and used at concentrations 0.5-25-fold higher than the reported IC 50 values (25)(26)(27).
In Vitro Culture with Anti-CD154 mAb-PBMC (1.5 ϫ 10 5 /well) were cultured with IL-4, 1 M ionomycin, and 10 ng/ml PMA with neutralizing anti-human CD154 mAb (Ancell, Bayport, MN) or with control mouse IgG, for 8 days to obtain supernatant fluids, or for 3 days to assess proliferative responses. The supernatant fluids of 8-day culture were assayed for IgE by sandwich ELISA. The peptide-pulsed peripheral B cells were cultured for 5 days, with emetine-treated T cells in the presence of the anti-CD154 mAb or with control mouse IgG, to determine IgM concentrations of the supernatant fluids.
RT-PCR and Southern Blot Analysis-After cross-linking DR molecules as described above, total cellular RNA was extracted from 1 ϫ 10 6 purified B cells by the acid guanidine thiocyanate phenol-chloroform method (TRIzol, Life Technologies, Inc.). The first strand cDNA was synthesized from purified total RNA by reverse transcriptase, using random primers (SuperScript preamplification system, Life Technologies, Inc.). The following oligonucleotides were used as primers: 5Ј-TCGGACATGACCAGGGACAC-3Ј (secreted component-origin) and 5Ј-TTCTCAAAGCCCTCCTCGTC-3Ј (membrane component-origin) for secretory-type chain and membrane-type chain, respectively, and 5Ј-AAAACCCACACCAACATCTC-3Ј (C3-origin) as a common primer. PRDI-BF1 transcripts were amplified using 5Ј-CTAAGAACGCCAA-CAGGAAA-3Ј and 5Ј-TGGAGTGGTGGAGGATGGAA-3Ј. ␤-Actin transcripts as a control, were amplified using 5Ј-CGGGAAATCGTGCGT-GACAT-3Ј and 5Ј-CTCGTCATACTCCTGCTTGC-3Ј. One amplification cycle consists of 6 min at 95°C followed by 20 -25 cycles for 1 min at 95°C, 58°C for 1 min, and 72°C for 1.5 min. The final extension cycle was for 7 min at 72°C. Amplification was performed under conditions in which amplified PCR products were invisible on an agarose gel under UV wave, so that none of amplified DNA concentrations reached a plateau level. PCR products were separated on a 1.2% agarose gel and transferred onto nylon membrane (Zeta-Probe, Bio-Rad), using 0.4 N NaOH. The membrane were hybridized with cDNA probes for C chain (kindly provided by Dr. N. Kondo, Gifu University) (28), for PRDI-BF1 (kindly provided by Dr. T. Maniatis, Harvard University) (29), or for ␤-actin that were randomly conjugated by dUTP-digoxigenin, using DIG-High Prime DNA Labeling and Detection Starter Kit II (Roche Molecular Biochemicals). The hybridization signal was detected by alkaline phosphatase-conjugated anti-digoxigenin Ab and disodium 3-(4-methoxyspiro{1,2-dioxetane-3,2Ј-(5Јchloro)tricyclo[3.3.1.1 3,7 ]decan}-4-yl)phenylphosphate, as a substrate. Hybridization signals were analyzed using the public domain Image program (developed at the United States National Institutes of Health and available from the Internet by anonymous FTP from zippy.nimh.nih.gov.).
Immunoblotting-Peripheral B cells and the human B lymphoblastoid cell line LD2B (5 ϫ 10 5 ) were incubated for 10 min on ice and then pre-incubated either with biotinylated anti-DR mAb (40 g/200 l) or with biotinylated mouse IgG (40 g/200 l) for 10 min on ice. After washing with ice-cold RPMI 1640, the cells were suspended with 50 l of 10% FCS/RPMI and cross-linked with 50 l of avidin (1 mg/ml). After 10 min of incubation at 37°C, ice-cold 100 M Na 3 VO 4 /PBS was added, followed by pelleting and lysing in 50 l of lysing buffer (150 mM NaCl, 20 mM Tris, pH 7.6, 0.5% Nonidet P-40, 2 mM sodium orthovanadate, 1 mM NaF, 5 mM EDTA, plus a protease inhibitor mixture purchased from Sigma). Supernatant fluids of the lysates were electrophoresed on SDS-PAGE gels and transferred to nitrocellulose membrane. After blocking with 10% skim milk, 0.2% Tween 20 in Tris-buffered saline, the membrane was incubated with the anti-phosphotyrosine mAb 4G10 (Upstate Biotechnology, Inc.), washed extensively, and subjected to chemiluminescence detection with peroxidase-conjugated anti-mouse IgG Ab (Amersham Pharmacia Biotech), using an ECL kit (Amersham Pharmacia Biotech). In some experiments, lysates prepared from 1 ϫ 10 7 LD2B cells were immunoprecipitated with anti-Syk Ab, and subjected to immunoblot analysis with anti-phosphotyrosine mAb 4G10 or anti-Syk Ab.
In Vitro Immune Complex Kinase Assay-The human B lymphoblastoid cell line LD2B (1 ϫ 10 7 ) was incubated for 10 min on ice and then pre-incubated either with biotinylated Igs (40 g/200 l) or with biotinylated F(abЈ) 2 fragments (12 g/200 l) for 10 min on ice. After washing with ice-cold RPMI 1640, the cells were suspended with 50 l of 10% FCS/RPMI and cross-linked with 50 l of avidin (1 mg/ml). After a 10-min incubation at 37°C, ice-cold 100 M Na 3 VO 4 /PBS was added, followed by pelleting and lysing in 400 l of the lysing buffer. Supernatant fluids of the lysates were pre-cleared with Protein A-agarose beads, then were incubated with a rabbit polyclonal anti-Syk Ab (Santa Cruz Biotechnology, Inc.), using Protein A-agarose beads (Pierce). After shaking for 30 min at 4°C, the beads were washed four times with lysis buffer. An aliquot of immunoprecipitated proteins was eluted with Laemmli buffer containing 2-mercaptoethanol, for immunoblotting analysis. Residual beads were washed once with kinase buffer ( , and then incubated for 2.5 min at 25°C. The reactions were terminated by adding an equal volume of 2ϫ Laemmli buffer. The supernatants were boiled for 2 min and applied to a 12% SDS-PAGE. After electrophoresis, the gel was fixed and vacuum-dried, and analyzed using a bio-imaging analyzer (BAS2000, Fuji Film, Tokyo, Japan). Eluted protein samples were separated on 7.5% SDS-PAGE and transferred to nitrocellulose membrane. After blocking with 10% skim milk, 0.2% Tween 20 in Tris-buffered saline, the membrane was incubated with the rabbit anti-Syk Ab, washed extensively, and subjected to chemiluminescence detection with peroxidase-conjugated anti-rabbit IgG Ab (Santa Cruz Biotechnology, Inc.), using an ECL kit (Amersham Pharmacia Biotech).

Cross-linking HLA-DR Molecules on B Cells Induces Increased Production of IgM without Inducing B Cell
Proliferation-To test whether signals via class II HLA molecules would affect production of Igs, we first cross-linked class II HLA molecules on B cells by making use of anti-DR mAb-coated culture plates. The supernatant fluids of 5-day cultures were assayed for Ig concentrations, among which only IgM was markedly affected by DR ligation. As shown in Fig. 1A, crosslinking DR molecules with anti-DR mAbs (L243 or HU4) on B cells induced IgM production, whereas isotype-matched mouse IgG did not do so, thereby indicating that signals transmitted by FcR are not involved. The experiment was repeated six times with reproducible results. Similar results were obtained, using B cells from another subject carrying DRB1*1405/1502 (data not shown).
To rule out the possibility that IgM production relates to B cell proliferation induced by signals via DR molecules (31, 32), B cells were cultured for 3 days in culture plates coated with anti-DR mAbs, in the presence of [ 3 H]thymidine during the final 16 h. As shown in Fig. 1B, no difference was observed between proliferative response induced by anti-DR mAbs and that induced by controls. This was also the case in the 6-day culture experiments (data not shown). Proliferative responses of the same B cell preparation induced by PMA and ionomycin exhibited 7,164 cpm.
To exclude the possibility that the increase in IgM is due to apoptosis of B cells, we cross-linked DR on B cells with the soluble form or solid-phase anti-DR mAbs. Stimulation with soluble-form (Fig. 2B), but not solid-phase mAb (Fig. 2D) and controls (Fig. 2, A and C), induced marked apoptosis. The precise mechanisms for this discrepancy is yet to be determined, but it is conceivable that DR-mediated increase in IgM in the culture supernatants was not due to release of IgM from apoptotic B cells. These observations collectively indicate that increased IgM concentration in culture supernatant fluids cannot be ascribed to B cell proliferation or apoptosis.
Cross-linking HLA-DR Molecules Enhances Both Membranetype and Secretory-type IgM Heavy Chain Gene Expression-To determine whether signals via DR molecules up-regulate chain mRNA, we cross-linked DR molecules on peripheral B cells (1 ϫ 10 6 ) with either solid-phase anti-DR mAb (L243) or solid-phase mouse IgG. Due to limitations in the number of purified B cells, we could test only three samples at one time. At 0, 3, and 6 h (Fig. 3A), or 6, 12, and 24 h (Fig. 3B) after the initiation of culture, B cells were analyzed for mRNA expression for chains, using RT-PCR and Southern blot analysis. Relative mRNA level was analyzed, using the public domain NIH Image program. When we tested the kinetics, chain mRNA increased in a time-dependent fashion (Fig. 3A) and reached maximum at 12 h (Fig. 3B). This increase was not due to the enhanced recovery of mRNA, as evidenced by the presence of an equal amount of ␤-actin mRNA in each sample. The chain mRNA level induced by control mouse IgG at 3, 6, 12, and 24 h was practically the same as that induced by anti-DR mAbs at 0 h (data not shown).
To test whether the DR-generated signal induced differenti- ation of B cells to plasma cells, we analyzed PRDI-BF1 transcripts. PRDI-BF1 is a human homologue of Blimp-1, the expression of which is characteristic of late B cells and plasma cells (29,33,34). However, as shown in Fig. 3, DR-generated signals up-regulated no mRNA for PRDI-BF1. The presence of PRDI-BF1 transcripts is indicative of the presence of plasma cells in this cell preparation. The experiment was repeated twice with reproducible results. These data suggest that IgM production induced by cross-linking of DR molecules is regulated at the mRNA level and is not associated with B cell differentiation to plasma cells.
Emetine-treated and HLA-DR-restricted T Cells Are Capable of Inducing IgM Production by B Cells-Although earlier observations strongly suggest that the ligation of HLA-DR molecules directly stimulates B cells to produce IgM, the outcome of ligation by mAbs should be affected by epitopes recognized by these mAbs and their affinity. Indeed, anti-HLA-DR mAb HU-4, exerted weaker effects than did L243 (Fig. 1). It is unlikely that HLA-DRB4 molecules recognized by L243 are transmitting the signals, because the B cell donor in Figs. 1 and 3 did not carry DRB4-positive haplotypes. Therefore, we next asked if a similar phenomenon occurs, on natural TCR-peptide-HLA interactions. An HLA-DR-restricted T cell clone was treated with the de novo protein synthesis inhibitor emetine (24), because it is highly likely that T cell membrane proteins or T cell soluble factors newly synthesized after activation by peptide-pulsed B cells, work on B cells. Under conditions where T cells are treated with 90 g/ml emetine for 1 h, followed by co-culture with peptide-pulsed B cells bearing restriction HLA molecules, T cells produced Ͻ25 pg/ml IL-4, whereas nontreated T cells produced 3580 pg/ml IL-4, although cell-surface TCR remains practically the same level (data not shown), in-dicating that de novo protein synthesis of T cells is abrogated by emetine. A T cell clone BC20.7 (BCGa-specific, DR14-restricted) and B cells purified from PBMC of the donor of BC20.7 were used in subsequent experiments. As shown in Table I, levels of IgM, IgG1, IgG4, IgE, and IgA were detected when mock-pulsed B cells were co-cultured with emetine-treated T cells. However, when B cells were pre-pulsed with the antigenic peptide, marked enhancement of IgM and marginal enhancement of IgA production were observed and such was not the case when peptide-pulsed B cells were cultured in the absence of T cells (data not shown). The experiment was repeated four times with reproducible results.
It seems reasonable to speculate that DR-mediated signals alone can up-regulate IgM production from B cells, based on the following observations made in the current study. (a) Ligation of DR molecules by mAb in the absence of T cells can up-regulate IgM protein and mRNA (Figs. 1 and 3); (b) upregulation of accessory molecules such as CD154 is not observed in emetine-treated T cells (data not shown); and (c) activated T cells of irrelevant restriction HLA molecules cannot stimulate peptide-pulsed B cells to up-regulate IgM (data not shown). However, we directly confirmed that signals via CD40 molecules are not involved in IgM production from B cells during cognate T-B cell interactions. To determine a saturating concentration of neutralizing anti-human CD154 mAb, we stimulated PBMC with 20 units/ml human recombinant IL-4, 1 M ionomycin, and 10 ng/ml PMA, with varying concentrations of anti-CD154 mAb. As shown in Fig. 4A, 1-10 g/ml anti-CD154 mAb inhibited IgE production, whereas 10 g/ml control mouse IgG did not do so. Because 1 or 10 g/ml anti-CD154 mAb did not inhibit proliferation of PBMC induced by the same stimuli (Fig. 4B), it is likely that anti-CD154 mAb exerted specific inhibitory effects on PBMC. When we co-cultured peptide-pulsed peripheral B cells and emetine-treated T cells in the presence of 1 g/ml anti-CD154 mAb, we observed no inhibition of IgM production (Fig. 4C). These findings collectively indicate that: (a) IgM production from B cells is enhanced when HLA-peptide-TCR interaction occurs, and (b) such enhancement occurs without involvement of signaling through either CD40 or FcR molecules.
Piceatannol Inhibits IgM Secretion Induced by DR Ligation-To identify signal transduction molecules involved in DR-mediated IgM production, we co-cultured peptide-pulsed peripheral B cells and emetine-treated T cells in the presence of several kinase inhibitors. These inhibitors were dissolved in Me 2 SO and added to culture medium at a final content of 0.5%, which did not inhibit DR-mediated IgM production, as shown in Fig. 5. PTK inhibitors genistein and herbimycin A inhibited IgM secretion in a dose-dependent manner, whereas a protein kinase C inhibitor GF109203X (IC 50 ϭ 20 nM; Ref. 26) did not do so. We next examined the effect of inhibitors specific for individual protein-tyrosine kinase(s). Interestingly, 10 M Syk kinase inhibitor piceatannol (27) inhibited IgM production, whereas PP2, a selective inhibitor of the Src family of proteintyrosine kinase (25), failed to do so. Because 500 M genistein or 100 M piceatannol did not inhibit anti-DR-induced IL-1␤ production from monocytes, 2 it is likely that the kinase inhibitors at the concentration we used exerted specific inhibitory effect on B cells. The experiment was repeated twice with reproducible results. These observations suggest that Syk kinase but not Src kinase(s) is involved in IgM production induced by ligating DR molecules. Kinase Activity-To investigate possible protein-tyrosine phosphorylation associated with this event, detergent lysates of peripheral B cells and LD2B cells treated with anti-DR mAb or control mouse IgG, were analyzed. Fig. 6 shows that proteintyrosine phosphorylation was enhanced by cross-linking of DR molecules on peripheral B cells (Fig. 6A, lane 2 versus lanes 1  and 3). Bands corresponding to proteins with an approximate molecular mass of 65, 70, 110, and 130 kDa were reproducibly hyper-phosphorylated. Likewise, cross-linking of DR molecules on LD2B induced tyrosine-phosphorylation of 65-and 70-kDa proteins (Fig. 6B, lane 2 versus lanes 1 and 3). Furthermore, immunoprecipitation by anti-Syk Ab followed by blotting with anti-phosphotyrosine mAb 4G10, exhibited anti-DR-induced tyrosine phosphorylation of Syk molecules expressed in LD2B cells (Fig. 6C).

Cross-linking DR Molecules on B Cells Up-regulates
To further confirm that Syk is activated directly via HLA-DR, Syk kinase activity was determined by in vitro kinase assay, using Syk molecules immunoprecipitated with anti-Syk Ab, and MBP as a substrate. Because a large number of B cells are required for immunoprecipitation followed by in vitro Syk kinase assay, we used a B lymphoblastoid cell line LD2B homozygous for DRB1*1501, which secretes IgM in the absence of specific stimuli. LD2B was selected among many B cell lines because (a) anti-DR-induced phosphorylation pattern of LD2B was similar to that of peripheral B cells, including the phosphorylation of 70-kDa protein (Fig. 6), and (b) LD2B B cells expressed IgM heavy chain genes. Enhancement of IgM production from LD2B B cells after cross-linking DR molecules was only marginal, probably because LD2B cells constitutively showed a 50 -80-fold higher IgM secretion than did peripheral B cells, on a single cell basis (data not shown). As shown in Fig 2 versus lane 1). Because MBP is not a substrate specific for Syk kinase, and it might be that MBP was phosphorylated by certain kinases co-precipitating with Syk, we also asked if the effect of Syk on MBP would be inhibited competitively by HS1p388 -402 peptide, a substrate specific for Syk (22). In vitro Syk kinase assay with MBP (2 g/sample; 0.11 nmol/sample) was done in the presence of either a 250-fold molar excess of the HS1 peptide (27.5 nmol/sample) or an irrelevant peptide carrying two tyrosine residues (EIKYN-GEEYLIL; 27.5 nmol/sample). Indeed, MBP phosphorylation was inhibited by the HS1 peptide, but not by the irrelevant peptide (lanes 4 and 5).
It is also important to note that Syk molecules are associated with Fc␥R and are activated by cross-linking of the receptor (35,36). It is therefore conceivable that the increment in Syk kinase activity we observed may be due to cross-reaction of mouse Ig with human Fc␥R expressed on B cells. To exclude this possibility, we prepared a biotinylated F(abЈ) 2 fragment of anti-DR mAb L243 or that of control mouse Ig. As shown in Fig.  7B, cross-linking of F(abЈ) 2 fragment of anti-DR mAb L243 TABLE I Ig production from B cells induced by a DR-restricted T cell clone B cells either mock-pulsed or pulsed with BCGap84 -100 were cultured with an HLA-DR14 (DRB1*1405)-restricted and emetine-treated T cell clone BC20.7 for 5 days. B cells were purified by, and the T cell clone was established from, a donor carrying DRB1*1405/1502. Mean values of duplicate determinations are indicated. S.D. was less than 25%.   Table I. induced phosphorylation of MBP (lane 3), whereas F(abЈ) 2 fragment of mouse Ig induced little phosphorylation of MBP (lane 2), compared with a control (avidin only; lane 1). This indicates that Syk phosphorylation is induced by cross-linking DR but not Fc␥R. These differences in phosphorylation patterns were not due to the enhanced recovery of these kinases, as evidenced by the presence of an equal amount of Syk protein molecule in each sample (Fig. 7, A and B). The experiment was repeated once with reproducible results. These data are consistent with results obtained using the Syk inhibitor piceatannol on IgM production, thereby collectively indicating that HLA-DR molecules on B cells not only present antigenic peptides to T cells, but also up-regulate IgM production, in association with Syk activation and without the involvement of Src kinases.

DISCUSSION
It is highly likely that the cognate interaction between T cells and B cells, as mediated by class II MHC molecules, results in the delivery of activation signals to B cells, as evidenced in the current study and as reported by others (2-11, 31, 37-39). Engagement of class II molecules on the THP-1 monocyte cell line with staphylococcal enterotoxin A induced IL-1␤ and tumor necrosis factor-␣ (40). Our previous studies demonstrated that certain T cell-monocyte interactions, via altered TCR ligands, affect monocyte responses to produce IL-12 (p70), which leads to specific up-regulation of interferon-␥ production from T cells (1). Moreover, we recently observed that cross-linking class II HLA molecules on monocytes induces a wide variety of monokine production (41), which is accompanied by activation of signaling molecules. 2 Thus, class II-mediated signaling events are not specific for B cells and play a crucial role in the activation of antigen-presenting cells, in general.
In this study, we found that ligation of HLA-DR alone is capable of inducing IgM production from peripheral B cells. Conversely, Palacios et al. (39) reported that cross-linking of DR on peripheral B cells with anti-DR Abs induced no Igs (IgM, IgG, and IgA), in the absence of pokeweed mitogen. However, they did not use solid-phase Ab but rather the soluble-form Ab, which was invalid or less efficient for cross-linking in our study (data not shown). In murine systems, IgM production from a B cell line requires not only cross-linking of class II MHC but also that of membrane IgM (5,37) or cytokines (31). Possibilities include the following: (a) B cells that have already received signals from BCR in vivo, are responding to the DR engagement in our experimental system; and/or (b) there are essential differences in these aspects between murine and human B cells.
IgM contributes to early defense against microbial infections (42). When B cells are exposed to non-self-antigens, such as those of microbial origin, B cells bearing surface IgM specific for the antigen are capable of concentrating the antigen and present it effectively to T cells. We found that cross-linking DR molecules up-regulates not only secretory-type but also membrane-type chains, which may indicate that cross-linking DR molecules leads to more effective antigen presentation. It is also important to note that CD40-generated signals arrest B cell terminal differentiation to produce Igs (43). Although DRmediated signals appear to up-regulate IgM production in the absence of CD40-CD154 interaction (Fig. 4), further investigation is needed to determine whether or not the generation of signals via CD40 under physiological T-B interactions interferes with IgM production induced by DR-mediated signals. In this study, ligation of DR molecules not only with specific Abs (either solid-phase Abs or soluble Abs), but also with HLA- LD2B B cells were incubated with biotinylated anti-DR mAb L243 or biotinylated mouse IgG followed by avidin for 10 min, then cells were lysed. Lysates were immunoprecipitated with anti-Syk Ab. A, an aliquot of immunoprecipitated proteins were immunoblotted with anti-Syk Ab. Residual Syk proteins on agarose beads were used for in vitro immune complex kinase assay, using MBP as a substrate. The HS1 peptide or an irrelevant peptide was added to the assay. B, LD2B B cells were incubated with biotinylated F(abЈ) 2 of anti-DR mAb L243, or biotinylated F(abЈ) 2 of control mouse IgG followed by avidin for 10 min. Cells were lysed, immunoprecipitated, and subjected to in vitro Syk kinase assay. MBP phosphorylation levels were quantified using a bio-imaging analyzer (BAS2000, Fuji Film, Tokyo, Japan), and represented by relative values compared with those of 0 min (A) or 10 min (B) (unstimulated cells).
peptide-TCR interaction, induced IgM production, suggesting that signals via DR alone are capable of inducing up-regulation of IgM, which may also occur in physiological T-B interactions. In this relation, DR-mismatched transplantation should be one of rare cases, in which massive T-B interaction via DR occurs in vivo. Indeed, when DNA typing of HLA-DR was unavailable, graft-versus-host disease was frequent, and such patients reportedly had deposition of IgM at the dermo-epidermal junction (44).
BCR-Ag-complex is internalized to supply T cell epitopes, and subsequent DR-peptide-TCR interaction results in class switching, which eventually leads to decreased IgM production (45). Indeed, our experimental system did not allow BCR to interact with protein antigens, and T cells were treated with emetine (thereby bearing no class switch pressure). Such a system might have up-regulated IgM to be readily detected. However, because the disappearance of surface IgM at antigen presentation (before class switching) is incomplete, one might speculate that signaling through DR supplies new IgM molecules, for a short and critical period of time for T-B interaction before class switching is initiated. Other factors should also be considered, because even with thymus-independent antigens, IgM production from B cells can be induced (46).
Analysis of sykϪ/Ϫ lymphoid cells showed that the Syk mutation impaired the differentiation of B-lineage cells, apparently by disrupting signaling from the pre-BCR complex, thereby preventing clonal expansion and further maturation of pre-B cells (47,48). Syk mutation also blocked B cell development in the transition from immature B cells (B220 ϩ , IgM ϩ ) to mature B cells (B220 ϩ , IgM ϩϩ , IgD ϩ ), where up-regulated transcription of chains is again taking place. Although physiological roles of Syk in normal mature B cells are not readily determined, it is conceivable that Syk is involved not only in BCR signaling but also in chain transcription, occurring at the pro-B cell and at the immature B cell stages.
It was reported that DR-mediated signals led to apoptosis of human B cells (7). DR-mediated increase in IgM in the culture supernatants in our experimental system was not due to apoptosis of B cells, because (a) the phenomenon was isotype-specific and (b) apoptosis was not observed as shown in Fig. 2, using solid-phase Abs. In Figs. 6 and 7, soluble Abs were used, but short term stimulation (10 min) did not induce apoptosis (data not shown).
Cross-linking of HLA-DR molecules expressed on B cells was reported to induce phosphorylation of Src family kinases (Lyn and Fgr; Ref. 11). PP2 was found to inhibit Src kinases Lck and Fyn in in vitro kinase assays (IC 50 ϭ 4 -5 nM), as well as T cell proliferation induced by TCR ligation (IC 50 ϭ 4 M; Ref. 25). However, 10 M PP2 failed to inhibit IgM production induced by cross-linking of DR in the current study. Rather, slight enhancement of IgM production by PP2 was reproducibly observed (Fig. 5), suggesting a possible suppressive role of Src kinase(s) in Syk activation (49). It was also reported that Src family kinases (Lyn and Btk) co-immunoprecipitate with Syk in activated B cells (50,51). It is thus likely that the agarosebound Syk used in our current study also carries these kinases. However, the phosphorylation of MBP by Syk was inhibited by the HS1 peptide, which could not be appreciably phosphorylated by Fgr and Lyn (22), indicating that cross-linking of DR indeed enhances Syk activity.
Piceatannol was initially reported to be specific for Syk kinase (IC 50 ϭ 10 M). However, it was later reported that piceatannol inhibits Lyn, at a higher concentration (IC 50 ϭ 70 M; 27) and even PKC (IC 50 ϭ 8 M; Ref. 52). Although piceatannol may inhibit PKC or Lyn at concentrations used in this study (10 or 100 M), the PKC inhibitor GF109203X and Src kinase inhibitor PP2 did not inhibit IgM production induced by crosslinking DR. Indeed, 100 -500 M GF109203X abrogates monokine secretion from activated macrophages. 2 We therefore assume that the inhibition of IgM production by piceatannol resulted from inhibition of Syk kinase itself. HLA-DR␤ chains were reported to be involved in PKC activation and translocation (10). However, it is not surprising that the PKC inhibitor had no influence on the IgM production in this study, because signaling through DR may be degenerate, 2 as observed by others with IL-4 receptor (53) or platelet-derived growth factor receptor (54).
Syk kinase fused to the transmembrane and extracellular domains of CD7 and CD16 alone, can induce complete T cell activation without co-aggregation of a Fyn-containing chimera (16). Syk is phosphorylated when transfected alone into COS-1 cells (55), and Syk reconstitutes TCR signaling in mutant Jurkat T cell lines deficient in either Lck or CD45 (17). Moreover, Syk can phosphorylate ITAM in the absence of Src kinases (18,56), and is associated with and is phosphorylated by extracellular signal-regulated kinase-1 (57). Although BCR-induced activation of Syk is dramatically inhibited by the loss of Lyn in DT40 B cells (58) and Syk is typically activated by Src kinases (13), all these observations by other workers suggest the existence of a Src-independent activation of Syk. In this respect, although DR-mediated signals activate Syk, cytoplasmic domains of DR␣ and DR␤ chains are devoid of ITAM motif. Likewise, the IL-2 receptor (30), CD29 (59), and platelet integrin ␣ IIb ␤ 3 (60) are also capable of activating Syk, although these receptors carry no ITAM motif in their cytoplasmic domains. Because Syk was not co-precipitated with DR molecules in our study (data not shown), certain molecules associated with DR may play a role in activating Syk, which are currently under investigation, using mass mapping techniques.