Major histocompatibility class II-mediated signal transduction is regulated by the protein-tyrosine phosphatase CD45.

Major histocompatibility complex class II molecules and the B cell antigen receptor (BCR) transduce similar signals when cross-linked by ligand. Therefore, studies were conducted to determine whether the protein tyrosine phosphatase CD45 regulates signaling via these transmembrane receptors in an analogous manner. Cross-linking of either class II molecules or the BCR on CD45-positive K46-17micromlambda B lymphoma cells was observed to induce activation of the Src family protein- tyrosine kinase Lyn, tyrosine phosphorylation of Syk and phospholipase Cgamma, and the production of inositol 1,4,5-trisphosphate leading to intracellular mobilization as well as extracellular influx of Ca2+. In the absence of CD45, cross-linking of either class II molecules or the BCR failed to induce activation of Lyn. Syk was inducibly phosphorylated on tyrosine in a normal manner, whereas phospholipase Cgamma exhibited a high basal level of tyrosine phosphorylation that was not significantly increased upon stimulation. Nevertheless, phospholipase Cgamma appeared to be functional because CD45-negative cells produced elevated levels of inositol 1,4,5-trisphosphate following stimulation through class II or the BCR. Regardless of this, CD45-negative cells exhibited Ca2+ mobilization responses that were greatly diminished and transient in nature. Whereas little or no mobilization of Ca2+ was observed in response to class II cross-linking, CD45-deficient cells mobilized Ca2+ from intracellular stores but not the extracellular environment in response to BCR cross-linking. These results demonstrate that CD45 regulates both Src family kinase activation and Ca2+ mobilization associated with class II- and BCR-mediated signal transduction.

Major histocompatibility complex (MHC) 1 class II molecules expressed by B cells play an important role as restricting elements in the generation of thymus-dependent immune re-sponses (1)(2)(3)(4). There is now a significant amount of evidence indicating that class II molecules also deliver signals to the B cell that affect several parameters related to their immune function. MHC class II-mediated signals have been implicated in regulation of antigen presentation, up-regulation of accessory molecule expression (i.e. B7.2), cell-cell adhesion, apoptosis, proliferation, and differentiation (5)(6)(7)(8)(9)(10)(11)(12)(13)(14)(15). The signals transduced in response to cross-linking of class II molecules have been partially characterized and are complex in nature. Ligation of class II can elicit either the production of cAMP and the subsequent translocation of protein kinase C to the nucleus (16 -20) or the activation of protein-tyrosine kinases (PTK), hydrolysis of phosphatidylinositol, and mobilization of Ca 2ϩ (21)(22)(23)(24)(25). The exact molecular mechanism(s) by which class II molecules are coupled to these distinct signaling pathways is at present unclear but appears to be regulated in part by the activation state of the B cell. In both the human and mouse, cross-linking of MHC class II molecules on unstimulated B cells elicits the production of cAMP and promotes translocation of protein kinase C (23,26). In contrast, differences in coupling of MHC class II molecules to the PTK-dependent signaling pathway in human versus mouse appear to exist. Whereas crosslinking of class II molecules in human tonsillar B cells and B cell lines leads to PTK activation (21,22), studies using quiescent mouse splenic B cells have demonstrated that class IImediated PTK activation is dependent on the receipt of a priming signal delivered via the B cell antigen receptor (BCR) in conjunction with IL-4 (24). The apparent differences in coupling of class II to the PTK-dependent pathway in human versus mouse have been attributed to inherent differences in the activation state of the cells being examined.
Structure/function studies have demonstrated that the ability of MHC class II to mediate PTK activation resides in the transmembrane/extracellular region of the ␣ and/or ␤ chains (25)(26)(27)(28). This finding suggests that class II is coupled to PTK activation through its interaction with an intermediate transducer protein(s) in much the same way that membrane immunoglobulin is coupled to the Src PTKs through an interaction with the Ig␣/Ig␤ heterodimer. Further analogies between class II-mediated signaling and that of the BCR can be drawn from a range of studies. Cross-linking of MHC class II molecules on B cells leads to the activation of the Src family PTKs Lyn and Fgr (29). Additionally, class II has been shown to regulate the activation of the PTK Syk in both tonsillar B cells and B leukemia cell lines (30). Downstream PTK-dependent processes including activation of phospholipase C␥ (PLC␥), phosphatidylinositol turnover, and increases in the concentration of intracellular Ca 2ϩ have also been documented following crosslinking of class II molecules in human tonsillar B cells (21,22,24,25). These findings suggest that cross-linking of MHC class II molecules activates a signaling pathway that is similar to the one associated with the BCR.
Because the signaling pathway activated by cross-linking of MHC class II molecules is similar to that triggered by ligand binding to the BCR, it was of interest to determine whether signal transduction via MHC class II is regulated by the protein-tyrosine phosphatase (PTP) CD45. Numerous studies have demonstrated that optimal BCR-mediated signaling requires CD45 expression on the surface of the cell (31,32). In the absence of CD45, BCR cross-linking does not mediate recruitment and activation of Src family PTKs (33)(34)(35), whereas the PTK Syk is recruited to the BCR complex and is activated (34,35). Although Syk function appears to be intact in cells lacking CD45, the absence of this PTP is manifest by the fact that BCR-mediated mobilization of Ca 2ϩ is abnormal, and activation of Ras is abrogated (36 -38). Thus, it is clear that the loss of CD45 may affect not only Src PTK activation but downstream processes as well.
To determine whether CD45 regulates class II-mediated signaling, experiments were conducted using the K46 B cell lymphoma in conjunction with two CD45-deficient variants. The K46 lymphoma is ideal for these studies because it exhibits a primed MHC class II signaling phenotype (25). Studies demonstrate that MHC class II-mediated activation of the Src family protein tyrosine kinase Lyn is abrogated, and Ca 2ϩ mobilization is drastically reduced in the absence of CD45. These findings indicate that CD45 plays an analogous role in regulating both MHC class II-and BCR-mediated signal transduction.
Cell Lines-The B lymphoma cell line K46 -17m (K46) was generously provided by Dr. Michael Reth (Max-Planck Institut fur Immunobiologie, Freiburg, Germany). K46 cells were cultured in Iscove's modified Dulbecco's medium (IMDM) supplemented with 5% fetal calf serum (HyClone Laboratories, Logan, UT), 2 mM L-glutamine, 50 M ␤-mercaptoethanol, 100 g/ml streptomycin-penicillin, and 50 g/ml gentamycin (Sigma) at 37°C under 7% CO 2 . CD45-negative variants of the K46 line were isolated following chemical mutagenesis with methanesulfonic acid ethyl ester (EMS, Sigma). K46 cells (2 ϫ 10 7 ) were incubated in 25 ml of IMDM supplemented with 5% fetal calf serum containing EMS at a final concentration of 480 -760 g/ml. The cells were cultured in the presence of EMS overnight at 37°C after which they were washed and resuspended in IMDM with 5% fetal calf serum. Within 1 to 2 days the CD45-negative cell population was enriched by negative selection with anti-CD45 mAbs and rabbit complement. EMStreated cells were resuspended in IMDM at a concentration of 5 ϫ 10 6 cells/ml. A mixture of anti-CD45 mAbs containing I3/2.3, MB23G2, and C363.16A each at a final concentration of 20 g/ml was added, and the cells were incubated for 20 min on ice. Subsequently, mAb-treated cells were washed and were incubated with rabbit complement for 2-3 h at 37°C. Additional complement was then added, and cells were incubated for an additional 2 h after which they were washed and resuspended in IMDM with 5% fetal calf serum. Complement-mediated negative selec-tion of CD45-positive cells was performed 2-3 times. CD45-negative cells were then positively selected using fluorescence-activated cell sorting. CD45-negative cells isolated by fluorescence-activated cell sorting were then subjected to limiting dilution cloning to isolate several CD45-negative clones, two of which (3S5 and 35S5) are described below. K46 cells and the CD45-negative clones were examined using immunofluorescence staining as described previously (40) to ensure comparable expression of key surface markers.
Metabolic Labeling-K46 cells or CD45-negative variants were washed two times in PBS and were resuspended at a concentration of 5 ϫ 10 6 /ml in cysteine/methionine-free IMDM (Sigma) supplemented with 5% dialyzed fetal calf serum. The cells were incubated for 1 h at 37°C prior to the addition of 0.1 mCi of Tran 35 S-label (specific activity Ͼ1000 Ci/mmol, ICN Pharmaceuticals, Costa Mesa, CA) per 1 ϫ 10 7 cells. Cells were subsequently incubated at 37°C for 3 h in the presence of [ 35 S]cysteine/methionine. Following the 3-h incubation, cells were washed three times with PBS and were used for immunoprecipitation experiments (2 ϫ 10 7 /sample). Metabolically labeled cells that were not used immediately were frozen as a pellet at Ϫ20°C.
Northern Blot Analysis-Total RNA was isolated from 1 ϫ 10 8 cells using the TRIzol extraction technique (Life Technologies, Inc.) as described by the manufacturer. Thirty-five g of RNA from each cell line was electrophoresed on a 1% agarose-formaldehyde denaturing gel and then transferred to nitrocellulose membrane (Amersham Pharmacia Biotech) by capillary blotting using 20ϫ SSC. The membrane was hybridized with a full-length CD45 cDNA obtained from Dr. M. L. Thomas (Washington University, St. Louis, MO). The cDNA was labeled using nick translation to a specific activity of 1.8 ϫ 10 8 cpm/g using [ 32 P]dCTP (NEN Life Science Products). The membrane was probed overnight at 42°C and was subsequently washed 2 times with 0.2ϫ SSC at a temperature of 55°C. The membrane was then exposed to XAR-5 x-ray film (Kodak) to visualize the mRNA band encoding CD45.
Immunoprecipitation and Immunoblotting-MHC class II-and BCRmediated signaling events were analyzed following stimulation of cells with the respective mAbs as described below. To analyze MHC class II-mediated signaling CD45-positive K46 cells or the CD45-negative variants (3S5 or 35S5) were harvested, resuspended in IMDM with 5% fetal calf serum (2.5 ϫ 10 7 /sample), and incubated in the presence of biotinylated anti-class II mAb (D3.137.5.7, 10 g/ml) for 15 min at RT. Cells were then washed once with PBS and were resuspended in IMDM with 5% fetal calf serum (1 ml/sample). The cells were equilibrated for 20 min at 37°C and were stimulated for various times with streptavidin at a final concentration of 10 g/ml. Stimulation of cells with either anti-class II mAb or streptavidin alone did not elicit a signaling response as determined by measurement of calcium mobilization (data not shown). For consistency cells stimulated through the BCR were handled in the same manner as those stimulated with anti-class II mAb and streptavidin. They were harvested, incubated at RT for 15 min, and then equilibrated at 37°C in medium alone prior to being stimulated for varied times by the addition of the mAb B76 (anti-heavy chain, 10 g/ml). Following stimulation, cell samples were washed twice in PBS and were lysed in 0.5 ml of lysis buffer (25 mM Hepes, 150 mM NaCl, pH 7.8, 10 mM EDTA, 1 mM EGTA, 0.1 mM Na 3 VO 4 , and 1% Nonidet P-40). Cells were lysed for 1 h on ice, and the lysates were centrifuged at 12,000 ϫ g for 15 min at 4°C. Detergent-soluble lysates were precleared by incubation with protein G-Sepharose (for Lyn, HS1, PLC-␥, and Syk experiments) or RG7/9.1 bound to Sepharose 4B (for CD22 experiments) for 1 h at 4°C to minimize nonspecific protein binding. Proteins of interest were immunoprecipitated from precleared lysates either by the addition of soluble Abs followed by the addition of protein G-Sepharose (Lyn, Syk, HS1, and PLC␥) or by the addition of Ab coupled directly to Sepharose 4B beads (CD22). Each immunoprecipitation step was performed for 1 h at 4°C with rotation. Immune complex-coated beads were collected and washed four times with lysis buffer containing 0.2% Nonidet P-40. The beads were resuspended in 50 l of reducing SDS-PAGE sample buffer, boiled for 4 min, and centrifuged at 12,000 ϫ g for 5 min. The proteins contained in 15 l of supernatant from each sample were separated on 8% acrylamide gels using SDS-PAGE and were transferred to Hybond-ECL nitrocellulose membranes (Amersham Pharmacia Biotech). The membranes were blocked with 10% nonfat dry milk in TBST for 1 h at room temperature, washed five times in TBST, and incubated with the appropriate primary Ab for 1 h at RT. The membranes were then washed five times with TBST, incubated with the appropriate HRPO-conjugated secondary Ab for 1 h at RT, washed five times with TBST, and visualized using enhanced chemiluminescence (ECL, Pierce) according to the manufacturer's instructions. The same blots were stripped of Abs by incubation in stripping buffer (10 mM Tris, pH 2.3, 150 mM NaCl) at 60°C for 1 h and then were washed extensively as above. The membranes were then blocked with 3% blot qualified bovine serum albumin (Promega, Madison, WI) in PBS, washed extensively, and incubated with anti-Tyr(P) mAb (4G10 at a 1/2000 dilution, Upstate Biotechnology Inc.) for 1 h. The membranes were washed and developed as above using ECL.
In Vitro Kinase Assay-Lyn was recovered from cells that were stimulated with mAbs directed against class II or the BCR as described previously by immunoprecipitation with rabbit anti-Lyn Ab and protein G-Sepharose. To perform in vitro kinase assays, Lyn bound to protein G-Sepharose 4B beads was incubated in 50 l of reaction buffer containing 25 mM Imidazole HCl, pH 7.2, 5 mM MnCl 2 , 0.1% gelatin, 0.5% Nonidet P-40, and 10 Ci of [␥-32 P]ATP (specific activity, 4500 Ci/mM, ICN Biochemicals, Inc., Irvine, CA) for 15 min at 37°C. The catalytic activity of Lyn was monitored by addition of the exogenous substrate acid-denatured enolase (5 g per reaction, Sigma). Phosphorylation of enolase was allowed to proceed at room temperature for 5 min. All reactions were terminated by the addition of 50 l of reducing SDS-PAGE sample buffer. Radiolabeled proteins were boiled and then separated using SDS-PAGE. Incorporation of 32 P into the exogenous substrate enolase was visualized by autoradiography using Kodak XAR-5 x-ray film. In selected experiments radiolabeled proteins were transferred to nitrocellulose followed by immunoblotting to detect Lyn. Densitometric analysis was used to confirm equivalent recovery and loading of the respective PTK. Membranes were then exposed to x-ray film to visualize radiolabeled proteins including enolase. Following excision of the enolase band from the membrane, incorporation of 32 P was quantitated using scintillation counting.
Measurement of IP 3 Production-Experiments to measure the production of inositol 1,4,5-trisphosphate (IP 3 ) were performed using [ 3 H]inositol 1,4,5-trisphosphate Radioreceptor Assay Kit (NEN Life Science Products) following the manufacturer's protocol. The samples were prepared for assay by stimulating K46 cells or CD45-negative variants (1 ϫ 10 7 /sample) with mAbs directed against either class II or the BCR as described previously. Following stimulation the cells were washed in 0.5 ml of PBS. The cells were lysed on ice with 100 l of 100% trichloroacetic acid, and the trichloroacetic acid was extracted with a 3:1 solution of 1,1,2-trichloro-1,2,2-trifluoroethane to trioctylamine based on the manufacturer's recommendations. The assays were performed as indicated in the protocol; however, the lyophilized IP 3 standard was reconstituted in PBS instead of distilled H 2 O. Each condition was prepared in duplicate with each sample also being assayed in duplicate. The results depicted are representative of three separate experiments.
Measurement of Calcium Mobilization-Analysis of Ca 2ϩ flux in K46, 3S5, and 35S5 cells was performed as described previously (40). Cells (1 ϫ 10 6 /ml) were loaded with Indo-1 AM (Molecular Probes, Eugene, OR) at a final concentration of 5 M. Indo-1-loaded cells were analyzed using a Becton Dickinson FacsVantage (San Jose, CA) equipped with an Enterprise laser from Coherent (Santa Clara, CA) set for excitation at ϳ364 nm at a power setting of 60 milliwatts. Fluorescence emissions were separated by a 505-nm short pass beam splitter into two component emissions by passage through 405-and 485-nm centered 10-nm band pass filters to detect violet and blue, respectively. The ratio of emissions for 405/485 was determined, and a plot was constructed of fluorescence ratio versus time. Analysis of class II-mediated Ca 2ϩ mobilization was performed as described previously (40). Just prior to analysis cells were incubated with biotinylated anti-class II mAb (D3.137.5.7, 5 g/ml, 15 min at RT), were washed, and the base-line level of intracellular free Ca 2ϩ was established. Once the base line was measured, streptavidin (5 g/ml) was added to the mAb-treated cells and the analysis resumed. BCR-mediated calcium flux was measured in response to anti-(B76, 5 g/ml) added to cells once an initial base line had been established. Measurement of intracellular mobilization versus extracellular influx of Ca 2ϩ was performed as described previously (40). Briefly, cells were resuspended in Ca 2ϩ /Mg 2ϩ -free PBS just prior to analysis. A base line was established, and the cells were stimulated as described above. Following the resolution of the intracellular mobilization response, the analysis was stopped and CaCl 2 was added to the sample to a final concentration of 4 mM. Measurement of intracellular free Ca 2ϩ was resumed immediately.

RESULTS
Analysis of CD45-negative K46 Mutants-The K46 cell line was chosen for studies to examine the role of CD45 in MHC class II-mediated signaling because it is functionally equivalent to a primed murine B cell in which class II is coupled to the PTK-dependent signal transduction pathway associated with calcium mobilization. Cross-linking of class II expressed on the K46 cell line with biotinylated anti-class II mAb and a secondary reagent (i.e. streptavidin) will elicit a calcium mobilization response without the need to first prime the cell with anti-Ig and IL-4 (25). This characteristic was essential for studies to evaluate the role of CD45 in MHC class II-mediated signaling because it may not be possible to prime cells via the BCR and/or IL-4 receptor in the absence of CD45.
CD45-negative mutants of the K46 parental cell line were generated by subjecting cells to chemical mutagenesis and negative selection as described under "Experimental Procedures." Two clones were selected for further analysis that expressed less than 8% of the parental level of CD45 as determined by immunofluorescence staining and flow cytometry. Clones 3S5 and 35S5 were examined using a panel of mAbs directed against several transmembrane proteins expressed on B cells including CD45, membrane immunoglobulin M, Ig light chain, MHC class II, MHC class I, and CD22. Depicted in Fig. 1 are the immunofluorescence staining results for the phenotypic analysis of the CD45-negative clone 35S5. As can be seen, with the exception of CD45 these cells express normal levels of surface receptors when compared with the parental K46 cells. The 3S5 clone exhibits a similar phenotype (data not shown). Additional immunofluorescence analysis demonstrated that the expression of CD19 was comparable between the parental K46 cells and the CD45-negative clones, whereas the level of CD21 and CD81 expression was elevated (ϳ2-fold) on the 3S5 and 35S5 clones (data not shown). It can be concluded from these analyses that the CD45-negative clones exhibit a relatively selective deficiency in their expression of CD45.
Further examination of the 3S5 and 35S5 clones to detect CD45 message revealed that there is little or no detectable message based on Northern blot analysis of total RNA ( Fig.  2A). In agreement with the immunofluorescence and Northern blot analyses, metabolic labeling of 3S5 and 35S5 with [ 35 S]methionine/cysteine followed by immunoprecipitation of CD45 with the I3/2.3 mAB failed to detect intracellular CD45 (Fig.  2B). It should be noted, however, that when the 3S5 and 35S5 clones were maintained in culture for extended periods (Ͼ6 months), CD45-positive revertants were observed. Thus, it is likely that chemical mutagenesis did not result in specific mutations within the CD45 gene locus in the 3S5 and 35S5 clones but instead affected the ability of the cells to transcribe the CD45 gene.
Loss of CD45 Expression Abrogates Src Family PTK Activation in Response to MHC Class II Cross-linking-The loss of CD45 expression has been demonstrated to inhibit antigen receptor-mediated activation of Src family PTKs in T and B cells (32). Additionally, it has been shown that class II molecules mediate the activation of the Src family PTKs Lyn and Fgr in response to ligand binding (29). Therefore, experiments were performed to evaluate whether cross-linking of class II leads to activation of the PTK Lyn in the absence of CD45. Whereas Lyn exhibited comparable basal levels of tyrosine phosphorylation in the presence or absence of CD45, there was a marked inhibition of inductive phosphorylation in response to cross-linking of class II in the 35S5 clone when compared with parental K46 cells (Fig. 3, upper panel). When the membrane was stripped and reprobed with antibody directed against Lyn (Fig. 3, middle panel), it was noted that the amount of Lyn in each lane was comparable, although it did exhibit an activationdependent change in its electrophoretic mobility in response to class II cross-linking in the CD45-positive K46 cells but not the CD45-negative 35S5 clone. This finding suggests that crosslinking of MHC class II may promote the inducible phosphorylation of Lyn on serine/threonine residues thereby decreasing its electrophoretic mobility in a manner similar to that previously reported for Lck (41) and that this process is regulated by CD45.
Additional studies were performed to measure the enzymatic activity of Lyn in the presence or absence of CD45 following class II ligation. K46 and 35S5 cells were stimulated for varied periods as indicated in Fig. 3 (bottom panel). Subsequently, Lyn was recovered by immunoprecipitation, and its catalytic activity was assayed using an in vitro kinase assay in the presence FIG. 2. The 35S5 and 3S5 clones do not express intracellular CD45 or message based on metabolic labeling and Northern blot analysis. A, the 35S5 and 3S5 clones do not contain detectable levels of mRNA encoding CD45. Thirty-five g of total RNA from the K46 parental cells and each of the CD45-deficient clones (3S5 and 35S5) was separated on a 1% formaldehyde agarose gel and was transferred to a nitrocellulose membrane. The membrane was probed with a 32 P-labeled full-length CD45 cDNA. The membrane was then exposed to x-ray film, and autoradiography was used to visualize hybridizing bands corresponding to CD45. B, the 35S5 and 3S5 clones do not express intracellular CD45 based on metabolic labeling. Cells (2 ϫ 10 7 ) labeled with [ 35 S]methionine/cysteine were washed and lysed in buffer containing 1% Nonidet P-40. Detergent-soluble lysates were precleared by the addition of irrelevant mAb coupled to Sepharose 4B beads. CD45 was then immunoprecipitated using the anti-CD45 mAb I3/2.3 coupled to Sepharose 4B, and the immune complex material was separated by SDS-PAGE. Radiolabeled bands were visualized by autoradiography. FIG. 3. CD45 expression is required for activation of the Src family PTK Lyn. Parental cells and CD45-negative (35S5) cells (1 ϫ 10 7 /sample) were incubated in medium alone or in the presence of anti-class II mAb (anti-Ia, D3.137, 10 g/ml) on ice for 10 min. Cells were washed twice in ice-cold PBS and were resuspended in 1 ml of IMDM containing 10% fetal bovine serum. Streptavidin (10 g/ml) was added to the cells, and they were incubated at 37°C for varied periods. Control cells did not receive streptavidin but were incubated for 10 min at 37°C. Reactions were terminated by the addition of ice-cold PBS containing 1% Nonidet P-40, and Lyn was immunoprecipitated using a polyclonal rabbit anti-Lyn antibody. Immune complex material was separated by SDS-PAGE and transferred to nitrocellulose. The membrane was probed with anti-Tyr(P) mAb (4G10) coupled to HRPO. Phosphorylation of Lyn was visualized using enhanced chemiluminescence (upper panel). The same membrane was stripped and reprobed with polyclonal antibody directed against Lyn followed by a secondary goat anti-rabbit antibody coupled to HRPO. ECL was then used to monitor loading of Lyn (middle panel). The activation of Lyn was assessed based on phosphorylation of the exogenous substrate enolase (lower panel). Cells were stimulated as described above. Lyn was immunoprecipitated from cell lysates and was incubated with denatured enolase in 50 l of reaction buffer containing 10 Ci of 32 P-labeled ATP. Reactions were allowed to proceed for 5 min at RT. Radiolabeled enolase was isolated by SDS-PAGE and was visualized by autoradiography. of the exogenous substrate enolase. Cross-linking of MHC class II molecules in the CD45-positive K46 cells, but not the CD45negative 35S5 clone, resulted in the rapid activation of Lyn within 1 min. The composite results for both BCR-and class II-mediated Lyn activation in the presence or absence of CD45 are depicted in Fig. 4. Ligation of either the BCR or class II molecules on CD45-positive K46 cells caused a 2-3-fold increase in the activity of Lyn based on phosphorylation of enolase. In contrast, antibody binding to either the BCR or MHC class II failed to activate Lyn in the 3S5 and 35S5 clones (Fig.  4). These data indicate that expression of CD45 is required for activation of Lyn in response to stimulation via either the BCR or MHC class II.
The PTK Syk Is Inducibly Phosphorylated on Tyrosine in the Absence of CD45-Previous studies have demonstrated that the non-Src family PTK Syk is recruited to the BCR complex and becomes activated following cross-linking of the BCR in the absence of CD45 (34,35). To examine whether class II-mediated Syk activation requires CD45, experiments were conducted in which K46 and 35S5 cells were stimulated with either anti-IgM (B76) mAb or biotinylated anti-class II (D3.137) mAb and streptavidin. Subsequently, Syk was immunoprecipitated from cell lysates and subjected to immunoblot analysis to measure tyrosine phosphorylation. As depicted in Fig. 5 Syk was inducibly phosphorylated in response to either anti-IgM or anti-class II, regardless of whether CD45 was expressed on the surface of the cell. These findings suggest that Syk is inducibly activated following MHC class II cross-linking and are in agreement with previous results related to Syk activation via the BCR. That Syk is catalytically active in the absence of CD45 is supported by experiments in which tyrosine phosphorylation of putative Syk substrates was examined in the K46 and 35S5 cell lines. Cells were stimulated via the BCR or class II molecules, and inducible tyrosine phosphorylation of HS1 and CD22 was measured. HS1 exhibited a constitutively elevated level of tyrosine phosphorylation that was not appreciably increased upon stimulation via the BCR or MHC class II (Fig. 6, A and B). In contrast, CD22 exhibited a slight increase in its basal phosphorylation in the absence of CD45 and was hyperphosphorylated in response to cross-linking of either the BCR (Fig. 7A) or MHC class II (Fig. 7B) when compared with K46 cells. Thus, in the absence of CD45 expression or detectable Src PTK activation, these data suggest that the PTK Syk is able to function resulting in increased basal and/or inducible tyrosine phosphorylation of multiple downstream substrates.
MHC Class II-mediated Phosphatidylinositol Hydrolysis Does Not Require CD45 Expression-Studies have demonstrated that Syk regulates tyrosine phosphorylation and activation of PLC␥ leading to the hydrolysis of phosphatidylinositol to produce the second messengers diacylglycerol and IP 3 that mediate protein kinase C activation and calcium mobilization, respectively (42). If Syk is functional in the absence of CD45 FIG. 4. Loss of CD45 expression abrogates Lyn activation in response to BCR or class II cross-linking. K46, 35S5, and 3S5 cell lines (1 ϫ 10 7 cells/sample) were incubated in medium alone or in the presence of either anti-Ig mAb (B76, 10 g/ml) or anti-class II mAb (biotinylated D3.137, 10 g/ml) and streptavidin (10 g/ml) for 1 min at 37°C. Cells were then resuspended in ice-cold PBS, washed, and lysed in buffer with 1% Nonidet P-40. Lyn was immunoprecipitated from cell lysates and incubated with enolase and [␥-32 P]ATP at RT for 5 min. Radiolabeled proteins were separated by 10% SDS-PAGE and were transferred to nitrocellulose. The recovery of Lyn was monitored by immunoblotting with anti-Lyn polyclonal Ab followed by goat antirabbit Ig coupled to HRPO. Lyn was visualized using ECL and autoradiography. Incorporation of 32 P into enolase was quantitated using scintillation counting. Each stimulation condition was assayed in triplicate, and the data are depicted as the mean stimulation index. The stimulation index is defined as the fold increase in 32 P incorporation into enolase mediated by Lyn isolated from stimulated cells divided by that for Lyn isolated from unstimulated cells. The basal kinase activity (in cpm of 32

FIG. 5. Inducible tyrosine phosphorylation of Syk occurs in the absence of CD45 in response to BCR or class II cross-linking.
Experiments were performed with K46 cells and the 35S5 clone (1 ϫ 10 7 cells/sample) in which the cells were incubated for varied periods in medium alone or in the presence of mAbs directed against the BCR (anti-Ig, B76, 10 g/ml) or class II (anti-Ia, biotinylated D3.137, 10 g/ml ϩ streptavidin, 10 g/ml). At the appropriate time points cells were resuspended in ice-cold PBS, washed, and lysed in buffer with 1% Nonidet P-40. Lysates were precleared, and Syk was immunoprecipitated with polyclonal rabbit anti-Syk Ab. Immune complex proteins were separated by 10% SDS-PAGE and transferred to nitrocellulose. The membranes were probed with anti-Tyr(P) mAb (4G10 coupled to HRPO), and tyrosine phosphorylation of Syk was visualized with ECL. The membranes were then stripped and reprobed with anti-Syk Ab to monitor loading of Syk in each well. A, analysis of Syk phosphorylation in response to cross-linking of the BCR. B, analysis of Syk phosphorylation in response to cross-linking of class II molecules. expression, then tyrosine phosphorylation of PLC␥ and hydrolysis of phosphatidylinositol should occur. Experiments were conducted to assess whether class II cross-linking on 35S5 cells mediates tyrosine phosphorylation of PLC␥ and the production of IP 3 . As seen in Fig. 8 PLC␥ is constitutively phosphorylated on tyrosine in the absence of CD45 and exhibits only a minor increase in phosphorylation after cross-linking of either the BCR or MHC class II. This finding, together with the data concerning tyrosine phosphorylation of CD22 and HS1, suggests that Syk may be constitutively active in the absence of CD45.
To determine whether PLC␥ is functional in the CD45-negative 35S5 cell line, experiments were conducted to measure the production of IP 3 in response to BCR or class II crosslinking. In agreement with the possibility that PLC␥ may be constitutively active based on an analysis of tyrosine phosphorylation (Fig. 8), IP 3 levels were constitutively elevated (ϳ1.5fold) in the 35S5 cells when compared with the CD45-positive K46 parental cells. Both CD45-positive and CD45-negative cell lines exhibited an increase in the level of IP 3 when stimulated via the BCR or MHC class II (Fig. 9). Whereas the increase in IP 3 production was relatively transient in the CD45-positive K46 cells, decreasing almost to basal levels within 10 min, the 35S5 cells exhibited a prolonged elevation in the level of IP 3 after stimulation. Moreover, levels of IP 3 produced in the CD45-negative cells were consistently 1.5-2-fold higher than those in the parental cells at all time points examined. Thus, these data indicate that PLC␥ is functional in the absence of CD45 and that it can be inducibly activated.
Loss of CD45 Expression Attenuates Calcium Mobilization in Response to Either BCR or MHC Class II Cross-linking-Although class II-mediated activation of the Src family PTK Lyn is abrogated in the absence of CD45 expression, experiments suggest that both Syk and PLC␥ are functional and that their activation results in the production of IP 3 . Therefore, studies FIG. 7. CD22 exhibits both basal and inducible hyperphosphorylation after BCR or class II cross-linking in cells that lack CD45. Analysis of CD22 tyrosine phosphorylation was performed using K46 cells and the 35S5 clone as described previously. Cells (1 ϫ 10 7 / sample) were incubated in the presence of either anti-Ig (B76, 10 g/ml, A) or anti-class II (␣-Ia, biotinylated D3.137 10 g/ml ϩ streptavidin, 10 g/ml, B) for 1-10 min at 37°C. As a control, cells were incubated for 10 min in medium alone (NT) or in the presence of the PTP inhibitor pervanadate (PV). At the appropriate time points cells were resuspended in ice-cold PBS, washed, and lysed in buffer with 1% Nonidet P-40. CD22 was immunoprecipitated with the mAb NIMR-6 coupled to Sepharose 4B. Proteins were separated by 10% SDS-PAGE and were transferred to nitrocellulose. Tyrosine phosphorylation of CD22 was visualized by probing the membranes with anti-Tyr(P) mAb coupled to HRPO (4G10) followed by development of the blot with ECL.
FIG. 6. HS1 is constitutively hyperphosphorylated on tyrosine in the absence of CD45 expression. Analysis of putative Syk substrate phosphorylation was performed in K46 cells and the 35S5 clone. Cells (1 ϫ 10 7 /sample) were incubated in the presence of either anti-Ig (B76, 10 g/ml, A) or anti-class II (␣-Ia, biotinylated D3.137 10 g/ml ϩ streptavidin, 10 g/ml, B) for 1-10 min at 37°C. As a control, cells were incubated for 10 min in medium alone (NT). At the appropriate time points cells were resuspended in ice-cold PBS, washed, and lysed in buffer with 1% Nonidet P-40. HS1 was immunoprecipitated with polyclonal rabbit anti-HS1 and protein G-Sepharose. Proteins were separated by 10% SDS-PAGE and were transferred to nitrocellulose. Tyrosine phosphorylation of HS1 was visualized by probing the membranes with anti-Tyr(P) mAb coupled to HRPO (4G10) followed by development of the blot with ECL (upper immunoblots). The membranes were stripped and were reprobed with anti-HS1 Ab followed by goat anti-rabbit IgG coupled to HRPO. Loading of HS1 on the gel was visualized using ECL (lower immunoblots).
were conducted to determine whether Ca 2ϩ mobilization is normal in cells that lack CD45 as one might predict based on the fact that the amount of IP 3 produced is equal to or greater than the level observed in CD45-positive cells. Analysis of Ca 2ϩ mobilization in response to either anti-Ig or anti-class II mAbs, however, revealed that there are profound alterations in the ability of cells to mobilize Ca 2ϩ in the absence of CD45. Whereas cross-linking of the BCR elicited a transient and somewhat diminished Ca 2ϩ flux response (Fig. 10, top panel), there was little or no detectable increase in the intracellular concentration of free Ca 2ϩ ([Ca 2ϩ ] i ) after stimulation of the 35S5 clone with anti-class II mAb and streptavidin (Fig. 10,  middle panel). The decreased Ca 2ϩ flux in response to anti-Ig or anti-class II stimulation of the CD45-negative 35S5 clone was not due to unequal loading with the indicator dye Indo-1 as both K46 and 35S5 cells responded comparably when incubated with the Ca 2ϩ ionophore ionomycin (Fig. 10, bottom panel).
Subsequently, experiments were performed to determine whether the decreased Ca 2ϩ mobilization response was due to lack of intracellular mobilization and/or extracellular influx of Ca 2ϩ . Because little or no detectable increase in the [Ca 2ϩ ] i could be observed after stimulation via MHC class II, the separate analysis of intracellular mobilization versus extracel-lular influx of Ca 2ϩ was not performed. However, experiments to monitor Ca 2ϩ mobilization in response to anti-Ig revealed that CD45-negative cells are able to mobilize Ca 2ϩ from intracellular stores (i.e. the endoplasmic reticulum) and that the response is comparable to that seen in the parental cell line (Fig. 11, upper panel). In contrast, the CD45-negative 35S5 clone did not exhibit the ability to mobilize Ca 2ϩ from the extracellular environment when stimulated with anti-Ig, suggesting that plasma membrane Ca 2ϩ channels are not functional (Fig. 11). The small increase in the [Ca 2ϩ ] i seen in the 35S5 clone after repletion of the medium with CaCl 2 was due to normal equilibration of the cytoplasmic compartment and does not reflect a response to anti-Ig mAb because similar changes in the [Ca 2ϩ ] i are seen with the 35S5 clone in the absence of stimulus (data not shown). The 3S5 clone exhibited Ca 2ϩ mobilization responses that were identical to those of the 35S5 clone when stimulated with either anti-Ig or anti-class II and streptavidin (data not shown). These data indicate that there is a significant defect in both BCR-and MHC class II-mediated Ca 2ϩ mobilization in the absence of CD45 even though Syk activation and the production of IP 3 appear to occur.

DISCUSSION
The K46 B lymphoma cell line in conjunction with derivative CD45-negative variants constitutes a valuable model system for analyzing the role of CD45 in both BCR-and MHC class II-mediated signal transduction. The inherent advantage of utilizing the K46 cell line is that signaling via both the BCR and MHC class II can be analyzed in the same cell type without the need to prime the cell in order to couple class II to its PTK-dependent signaling pathway. This facilitates the direct comparison of the effect that loss of CD45 expression has on signal transduction via both receptors. The results obtained from this study suggest that, with regard to the requirement for CD45 expression, there are significant similarities between the BCR-mediated signaling pathway and the PTK-dependent pathway activated in response to cross-linking of MHC class II.
Previous studies in both T cells and B cells have demonstrated that CD45 physically associates with members of the Src family and that loss of CD45 expression abrogates antigen receptor-dependent activation of Src family PTKs (32,43,44). In the B cell, BCR-mediated activation of multiple Src family PTKs including Lyn, Fyn, and Blk is attenuated in the absence of CD45 (33)(34)(35). The inability to activate Src family PTKs in CD45-deficient cells results from hyperphosphorylation of the carboxyl-terminal negative regulatory tyrosine residues in the respective members of the family (32,34,35) and the failure to properly recruit these PTKs to the BCR activation complex (34). In agreement with previous work this study demonstrates that, in the absence of CD45, cross-linking of either the BCR or MHC class II fails to inducibly activate the PTK Lyn. Although Lyn did not exhibit a significant increase in its basal level of tyrosine phosphorylation in the CD45-negative 35S5 and 3S5 clones, it was nevertheless clear that activation of this PTK is abrogated based on an analysis of its ability to phosphorylate the exogenous substrate enolase. Previous studies have implicated Fgr in addition to Lyn in class II-mediated signaling in the B cell (29). However, because the K46 cell line does not express this PTK, it was not possible to determine whether CD45 expression is required for its inducible activation. It would be logical nonetheless to predict that the functional activation of this PTK would be inhibited as well because loss of CD45 expression has been shown to affect receptor-mediated activation of multiple Src family members in T and B cells (32).
Although it was apparent from these studies that Src family PTK function is abrogated in the absence of CD45, subsequent studies demonstrated that the non-Src family PTK Syk is in- FIG. 8. PLC␥ is constitutively phosphorylated on tyrosine in cells that lack CD45 expression. PLC␥ phosphorylation on tyrosine was examined in K46 parental cells and the 35S5 clone as described previously. Cells (1 ϫ 10 7 /sample) were incubated in the presence of either anti-Ig (B76, 10 g/ml, A) or anti-class II (anti-Ia, biotinylated D3.137 10 g/ml ϩ streptavidin, 10 g/ml, B) for 1-30 min at 37°C. As a control, cells were incubated for 30 min in medium alone (NT). At the appropriate time points cells were resuspended in ice-cold PBS, washed, and lysed in buffer with 1% Nonidet P-40. PLC␥ was immunoprecipitated with mouse anti-bovine phospholipase C␥-1 (mixed mAb) and protein G-Sepharose. Proteins were separated by 10% SDS-PAGE and were transferred to nitrocellulose. Tyrosine phosphorylation of PLC␥ was visualized by probing the membranes with anti-Tyr(P) mAb coupled to HRPO (4G10, upper immunoblot in each panel) followed by development of the blot with ECL. The membranes were stripped and then reprobed with anti-bovine PLC␥-1 mAb followed by goat anti-mouse IgG coupled to HRPO. PLC␥ was visualized using ECL (bottom immunoblot in each panel). ducibly phosphorylated in response to cross-linking of both the BCR and MHC class II molecules. Based on an analysis of putative Syk substrate phosphorylation in response to crosslinking of either the BCR or class II, it is likely that Syk is constitutively active in the 3S5 and 35S5 clones. HS1, CD22, and PLC␥ all exhibited increased basal levels of tyrosine phosphorylation in the 35S5 and 3S5 clones when compared with the CD45-positive K46 cells. Furthermore, CD22 was inducibly hyperphosphorylated suggesting that Syk activity is potentiated by BCR or class II ligation in cells lacking CD45. Previous studies using the J558L plasmacytoma and a CD45-positive transfectant of this cell line have revealed that Syk is recruited to the BCR activation complex in response to anti-Ig stimulation regardless of whether CD45 is expressed on the surface of the cell (34). Activation of Syk in the absence of CD45 has also been reported in studies using the chicken DT40 cell line (35) as well as Syk-positive Jurkat T cells (45). Perhaps more important in light of the observation that putative Syk substrates exhibit basal hyperphosphorylation in the 35S5 and 3S5 clones is the finding that Syk is constitutively activated in CD45deficient B cells. Studies using the J558L plasmacytoma demonstrated that Syk was approximately 3-4-fold more active in unstimulated cells that lacked CD45 when compared with CD45-positive transfectants (34). Therefore, it is possible that the loss of CD45 expression may result in aberrant constitutive activation of the PTK Syk.
Because it was apparent from these studies that Syk activity may be constitutively elevated in the absence of CD45, it was of interest to determine whether the downstream substrate PLC␥ is inducibly phosphorylated and activated in response to BCR or MHC class II cross-linking. Analysis of PLC␥ phosphorylation revealed that it is constitutively phosphorylated on tyrosine in the 35S5 clone in agreement with the possibility that Syk may be constitutively active in these cells. This finding is in contrast to previous studies using B cells isolated from CD45-deficient mice in which apparently normal inducible tyrosine phosphorylation of PLC␥ was observed in response to cross-linking of the BCR (46). Moreover, experiments with the CD45-negative J558L cell line revealed inducible tyrosine phosphorylation of PLC␥1 and PLC␥2 even though the basal activity of Syk was elevated in these cells compared with CD45positive J558L transfectants (33).
Although PLC␥ exhibited high basal phosphorylation in the 35S5 clone, subsequent studies to measure IP 3 production demonstrated that PLC␥ is indeed functional and that its activity is increased upon cross-linking of the BCR or MHC class II. In agreement with the observation that PLC␥ was constitutively hyperphosphorylated in the CD45-negative clones, basal levels of IP 3 were 1.5-2-fold higher in these cells when compared with parental K46 cells. Moreover, the level of IP 3 present in CD45deficient cells stimulated via the BCR or MHC class II was elevated at all time points assayed. Previous studies with the CD45-deficient J558L plasmacytoma have demonstrated that cross-linking of the BCR mediates increased production of IP 3 in the absence of CD45 (33). However, because the basal level of IP 3 in the J558L cells was only a fraction of that in the CD45-positive transfectants, the maximum level of IP 3 produced in response to cross-linking of the BCR was not observed to surpass the base-line level of IP 3 present in the CD45positive transfectants (33). Thus, although there is general agreement that cross-linking of the BCR, as well as MHC class II, leads to increased production of IP 3 in the absence of CD45, there is a discrepancy between the K46 and J558L cell lines as to whether the net level of IP 3 produced upon stimulation is sufficient to mediate downstream signaling.
Consistent with previous studies of signal transduction via the BCR in CD45-deficient B cell lines, experiments revealed that there is a significant alteration in the Ca 2ϩ mobilization response in the 3S5 and 35S5 clones. Cross-linking of the BCR in the CD45-negative clones leads to a transient rise in the [Ca 2ϩ ] i that was due to mobilization of Ca 2ϩ from intracellular stores. Little or no extracellular influx of calcium was observed suggesting that the loss of CD45 expression either directly or indirectly affects the function of plasma membrane Ca 2ϩ channels. The loss of CD45 was observed to have an even more profound effect on class II-mediated Ca 2ϩ mobilization. Class II ligation failed to induce a significant Ca 2ϩ mobilization response in the absence of CD45. The explanation for the difference between BCR-and class II-mediated Ca 2ϩ mobilization responses is at present unclear but does not seem related to the production of IP 3 .
The selective abrogation of the extracellular influx of Ca 2ϩ in the 35S5 and 3S5 clones in response to BCR cross-linking is in agreement with previous results from studies with B cells derived from CD45-deficient mice (46). The specific mechanism by which influx of Ca 2ϩ from the extracellular space is inhibited in the absence of CD45 is not clear. Based on the results presented in this study, it would appear that IP 3 production leading to the depletion of intracellular Ca 2ϩ stores occurs more or less normally in the absence of CD45. Recent studies FIG. 9. BCR-and class II-mediated hydrolysis of phosphatidylinositol occurs in cells that lack CD45. IP 3 production was measured in K46 cells and the 35S5 clone treated either with anti-Ig mAb (B76, 10 g/ml, A) or anti-class II mAb (biotinylated D3.137, 10 g/ml ϩ streptavidin, 10 g/ml, B) for varied periods using an IP 3 radioreceptor binding assay. Cell samples were set up in triplicate, and each sample was then assayed in triplicate with the IP 3 receptor binding assay. The data depicted are representative of three independent experiments. have demonstrated that depletion of intracellular Ca 2ϩ stores leads to the generation of a signal that activates Ca 2ϩ influx across the plasma membrane through a process termed capacitative Ca 2ϩ entry (47). The specific mechanism by which depletion-activated Ca 2ϩ channels are regulated is currently not well understood. However, there is evidence to support the hypothesis that depletion of intracellular stores initiates the production of a second messenger(s) from the microsomal fraction of the cell that diffuses through the cytoplasm to the plasma membrane where it opens the Ca 2ϩ release-activated Ca 2ϩ (CRAC) channel (48). The identity of the calcium influx factor is not well established at present. Nevertheless, the direct involvement of CD45 in regulating the production of Ca 2ϩ influx factor might provide an explanation for the lack of FIG. 10. Calcium mobilization in response to cross-linking of either the BCR or MHC class II is attenuated in the absence of CD45 expression. Studies were performed with the K46 cells and 35S5 clone in which calcium mobilization was assayed in response to either anti-Ig (B76, 10 g/ml) or anti-class II (␣-Ia, biotinylated D3.137, 10 g/ml ϩ streptavidin, 10 g/ml). Cells were loaded with the calcium indicator dye Indo-1 as described under "Experimental Procedures." To measure the response to cross-linking of the BCR an unstimulated base line was established, the analysis was stopped, and anti-Ig mAb was added. The analysis was resumed immediately after addition of anti-Ig. For the analysis of class II-mediated calcium mobilization, cells were first incubated with biotinylated D3.137 for 10 min at RT. The cells were washed, and the calcium base line was established. The analysis was then stopped and streptavidin added, after which the analysis was resumed immediately. The calcium ionophore ionomycin was used to stimulate cells to ensure that they were loaded equivalently with Indo-1.
FIG. 11. Loss of CD45 expression abrogates extracellular influx of calcium in response to cross-linking of the BCR. Analysis of the intracellular mobilization and extracellular influx components of the calcium mobilization response mediated by cross-linking of the BCR was performed. CD45-positive K46 cells and the CD45-negative clone 35S5 were used for these experiments (1 ϫ 10 6 cells/sample). Cells were loaded with Indo-1 as described previously. Just prior to analysis the cells were washed and resuspended in calcium-free PBS. The basal level of calcium in unstimulated cells was established, and the analysis was stopped and anti-Ig added (B76, 10 g/ml). The analysis was resumed to monitor the intracellular mobilization of calcium. After resolution of this component of the response, the analysis was stopped a second time, and CaCl 2 was added to bring the final concentration of Ca 2ϩ to 4 mM. Analysis was then resumed to monitor the influx of calcium from the extracellular environment. Equivalent loading of cells was monitored using the calcium ionophore ionomycin. BCR-mediated Ca 2ϩ influx in CD45-negative cells. Alternatively, studies have suggested that depletion of intracellular Ca 2ϩ stores may regulate CRAC channel function via a PTKdependent mechanism (49,50). Thus it is possible that CD45 expression is required to enable a member(s) of the Src family, or an as yet undefined PTK, to regulate CRAC channel function.
In summary, this study demonstrates that CD45 plays a comparable role in regulating signaling via the BCR and MHC class II molecules in the B cell. Loss of CD45 expression attenuates Src family PTK activation in response to BCR or class II ligation. Additionally, the loss of CD45 expression has a profound effect on receptor-mediated Ca 2ϩ mobilization. These findings confirm that CD45 plays an important role in regulating B cell activation and further suggest that this PTP may regulate clonal expansion and differentiation mediated by MHC class II.