Tec Kinases Mediate Sustained Calcium Influx via Site-specific Tyrosine Phosphorylation of the Phospholipase Cγ Src Homology 2-Src Homology 3 Linker*

Tyrosine phosphorylation of phospholipase Cγ2 (PLCγ2) is a crucial activation switch that initiates and maintains intracellular calcium mobilization in response to B cell antigen receptor (BCR) engagement. Although members from three distinct families of non-receptor tyrosine kinases can phosphorylate PLCγ in vitro, the specific kinase(s) controlling BCR-dependent PLCγ activation in vivo remains unknown. Bruton's tyrosine kinase (Btk)-deficient human B cells exhibit diminished inositol 1,4,5-trisphosphate production and calcium signaling despite a normal inducible level of total PLCγ2 tyrosine phosphorylation. This suggested that Btk might modify a critical subset of residues essential for PLCγ2 activity. To evaluate this hypothesis, we generated site-specific phosphotyrosine antibodies recognizing four putative regulatory residues within PLCγ2. Whereas all four sites were rapidly modified in response to BCR engagement in normal B cells, Btk-deficient B cells exhibited a marked reduction in phosphorylation of the Src homology 2 (SH2)-SH3 linker region sites, Tyr753 and Tyr759. Phosphorylation of both sites was restored by expression of Tec, but not Syk, family kinases. In contrast, phosphorylation of the PLCγ2 carboxyl-terminal sites, Tyr1197 and Tyr1217, was unaffected by the absence of functional Btk. Together, these data support a model whereby Btk/Tec kinases control sustained calcium signaling via site-specific phosphorylation of key residues within the PLCγ2 SH2-SH3 linker.

Signals generated by the pre-B and mature B cell receptors are essential for B cell development, activation, and maintenance of mature B cell populations (1). Engagement of the BCR 1 initiates the formation of a lipid-raft associated signaling complex, or "signalosome," containing tyrosine and serine/threonine kinases, adapter molecules, and lipid hydrolases including phospholipase C␥ isoforms. Together, these events promote a series of downstream signals including a sustained increase in intracellular calcium concentrations.
PLC␥ is essential for antigen receptor-mediated calcium mobilization (2)(3)(4). Activated PLC␥ hydrolyzes its substrate, phosphatidylinositol 4,5-bisphosphate, generating the second messengers inositol 1,4,5-trisphosphate (IP 3 ) and diacylglycerol (5,6). IP 3 acts to open intracellular calcium stores promoting an initial, transient rise in intracellular calcium. Depletion of intracellular calcium stores triggers the opening of plasma membrane store-operated calcium channels (7), resulting in an influx of extracellular calcium and a secondary, sustained calcium signal. The amplitude and duration of this sustained calcium signal is a key determinant of the specific transcription program initiated in response to antigen receptor engagement (8 -10).
The most abundantly expressed PLC␥ isoform in B lineage cells is PLC␥2. Chicken B lymphoma cells lacking PLC␥2 are unable to generate IP 3 in response to BCR engagement, resulting in the abrogation of receptor-mediated calcium mobilization (11). Similarly, mice deficient in PLC␥2 have a defective response to BCR engagement and exhibit a block in B cell development (12)(13)(14). Despite its crucial role in BCR-dependent calcium mobilization, the molecular events regulating PLC␥2 activity remain incompletely defined.
Activation of PLC␥ isoforms correlates with an increase in protein tyrosine phosphorylation (15,16), and inhibition of tyrosine kinases abolishes BCR-dependent IP 3 production and calcium signaling (17). Consistent with this, members of the Src, Syk/Zap70, and Tec/Btk kinase families are each capable of phosphorylating PLC␥ isoforms and peptide fragments in vitro (18 -21). The specific contribution of these kinases to BCRmediated phosphorylation and activation of PLC␥ in vivo, however, remains unknown (4).
Btk/Tec kinases specifically regulate the BCR-dependent sustained phase of calcium signaling (21)(22)(23). Deficient function of Btk leads to the related immunodeficiencies, X-linked agammaglobulinemia (XLA) in humans and X-linked immunodeficiency (XID) in mice (24). Igm ϩϩ /IgD ϩ B cells derived from XLA patients exhibit reduced IP 3 levels and fail to generate a sustained calcium signal in response to BCR engagement. Reconstitution of XLA B cells with increasing doses of wild type Btk can restore and specifically enhance the sustained calcium signal (21). Despite the marked reduction in peak IP 3 levels and calcium influx in XLA B cells, BCR-dependent PLC␥2 tyrosine phosphorylation is indistinguishable from that present in normal B cells (21,25). Similarly, Btk-deficient murine B-lymphocytes, mast cells, and platelets also exhibit diminished calcium mobilization and phosphoinositide hydrolysis despite normal receptor-mediated PLC␥2 tyrosine phosphorylation (26 -28). One potential explanation for these contrasting observations is that Btk may regulate the sustained calcium signal via phosphorylation of a subset of key tyrosine residues essential for PLC␥2 activity.
Recently, two groups identified up to four putative regulatory phosphorylation sites within PLC␥2 using in vitro kinase assays and reconstitution studies (20,29). These sites included tyrosine residues within the SH2-SH3 linker region (Tyr 753 and Tyr 759 ) (20,29) and at the carboxyl terminus (Tyr 1197 and Tyr 1217 ) (29). Notably, the Tyr 753 and Tyr 759 PLC␥2 sites appeared analogous in location to defined PLC␥1 SH2-SH3 linker regulatory sites (30). Reconstitution of PLC␥2-deficient cells with PLC␥2 containing mutations in the SH2-SH3 linker sites alone (20) or in all four sites concurrently (29) led to the near ablation of BCR-dependent calcium signaling. Whereas these studies demonstrate a key role for these sites in regulating PLC␥2 functional activity, the kinases responsible for mediating site-specific modification in the context of receptor dependent signaling remain undefined.
In the current study, receptor-mediated PLC␥2 activation was directly evaluated within the context of an intact cellular system through the use of antibodies specific for distinct PLC␥2 phosphotyrosyl regulatory residues. Our data demonstrate that at least four regulatory sites are phosphorylated in response to BCR engagement and that phosphorylation of the PLC␥2 SH2-SH3 linker region is entirely dependent upon Btk/ Tec family kinases. This specific property provides a coherent explanation for the unique role of Btk/Tec kinases in immunoreceptor dependent sustained calcium signaling.
Cell Stimulation, Lysis, Immunoprecipitation, and Immunoblotting-Prior to stimulation, B cells were incubated in serum-free RPMI plus glutamine for 45 min at 37°C. 1 ϫ 10 7 cells were then stimulated with either pervanadate (150 M) or human-or murine-specific IgM F(abЈ) 2 antibodies (10 g/ml) for the indicated times. Pervanadate was prepared as previously described (32). Cells were lysed for 10 min on ice in a lysis buffer consisting of 200 mM boric acid, pH 8.0, 150 mM NaCl, 0.5% Triton X-100, 5 mM NaF, 5 mM EDTA, and 1 mM sodium orthovanadate plus protease inhibitors leupeptin and aprotinin (10 g/ml each) and phenylmethylsulfonyl fluoride (1 mM). Cellular debris was spun down at 14,000 rpm at 4°C for 15 min, and cleared lysates were used as whole cell lysate or for immunoprecipitation. PLC␥2 was immunoprecipitated from cells stimulated with pervanadate due to increased stability and efficiency of detection of specifically phosphorylated PLC␥2. In contrast, site-specific phosphorylation was more reliably detected via analysis of whole cell lysate from IgM-stimulated primary and transformed B cells. PLC␥2 was immunoprecipitated by using 2 g of anti-PLC␥2 antibody (Q-20; Santa Cruz Biotechnology) followed by incubation with protein A-Sepharose for 2 h at 4°C. All samples were resolved by 10% SDS-PAGE and transferred to nitrocellulose membrane. Pervanadate studies represent 5 ϫ 10 5 cell equivalents/lane, whereas IgM studies represent 5 ϫ 10 6 cell equivalents/lane. Western blot analysis was performed using standard procedures as previously described (21). ECL was used for antibody detection according to the manufacturer's instructions (Amersham Biosciences).
Generation of Phosphospecific Antibodies and Enzyme-linked Immunosorbent Assay-PLC␥2 Tyr 753 , Tyr 759 , Tyr 1197 , and Tyr 1217 phosphospecific antibodies were generated by inoculating rabbits with 100 g of phosphopeptide conjugated to keyhole limpet hemocyanin as immunogen with Freund's complete adjuvant. Phosphopeptide sequences used were as follows: PLC␥2 Tyr 753 and Tyr 759 (see Fig. 2), Tyr 1197 (PVLESEEELpYSSCRQLRRRQ, where pY represents phosphotyrosine), and Tyr 1217 (CELNNQLFLpYDTHQNLRNAN). Initial injection was followed by four booster injections every 2-3 weeks in Freund's incomplete adjuvant. To obtain affinity-purified antibodies, sera were first absorbed to a nonphosphorylated peptide column. Column flowthrough was then run over a phosphopeptide column to isolate phosphospecific antibodies for use in our assays (31). To evaluate antibody reactivity by enzyme-linked immunosorbent assay, microtiter wells were coated with nonphosphorylated or phosphorylated peptide species.
Site-directed Mutagenesis and Generation of Recombinant Vaccinia Virus-Mutagenesis of full-length human PLC␥2 cDNA was carried out in the pBlueScript (pBS) vector using the QuikChange TM site-directed mutagenesis kit (Stratagene). Primers were designed in which a single nucleotide change resulted in a single amino acid change (tyrosine to phenylalanine). The Y753F/Y759F double mutant was generated by sequential mutagenesis. Mutagenized constructs were digested and purified for cloning into the pSC66 vaccinia virus vector. Sequence analysis of pBS and pSC66 PLC␥2 WT and mutant constructs was performed to verify each mutation. Recombinant vaccinia virus expressing WT or mutant human PLC␥2 was generated as previously described (33).
Preparation of Recombinant Human PLC␥2-Full-length PLC␥2 was expressed in bacteria using the pCAL expression and purification system. Briefly, Escherichia coli BL21 were transformed with WT or mutant pCAL-PLC␥2 expression constructs. Transformed bacteria were grown at room temperature with constant agitation until reaching an optical density at 550 nm of 0.8 -1.0. PLC␥2 expression was induced by the addition of a 0.1 mM concentration of "the inducer" (Molecular Research Laboratories, LLC). Bacteria were harvested, and recombinant PLC␥2 was purified using a calmodulin column as previously described (19).
PLC␥2 in Vitro IP 3 Assay for Activity-Recombinant purified WT or mutant PLC␥2 was added to a reaction mixture containing 200 M 3 H-labeled phosphatidylinositol 4,5-bisphosphate (25,000 dpm), 35 mM NaH 2 PO 4 (pH 6.8), 70 mM KCl, 2 mM MgCl 2 , 1 mM EDTA, 5 g/ml bovine serum albumin, 5 mM dithiothreitol, and 0.6 mM CaCl 2 in the presence of recombinant, purified Lck (purified recombinant human Gst-Lck provided by Alexander Y. Tsygankov, Temple University, Philadelphia, PA). ATP (25 M final concentration) was added to selected reactions, and mixtures were incubated for 10 min at 25°C. Reactions were stopped by transferring mixtures to an ice bath and adding 0.5 ml of chloroform/methanol/HCl (100:100:0.6) followed by 150 l of 1 N HCl with 5 mM EDTA. Aqueous and organic phases were separated by centrifugation, and a portion of the upper aqueous phase (200 l) was removed for liquid scintillation counting. All experiments were performed in triplicate.
PLC␥2 in Vitro Kinase Assay-NIH-3T3 cells (5 ϫ 10 6 ) were infected with vaccinia virus expressing recombinant wild type or mutant PLC␥2 at an MOI of 10 for 12 h at 37°C. Cells were harvested, and whole cell lysates were prepared as previously described under "Experimental Procedures." PLC␥2 was immunoprecipitated by using 2 g of anti-PLC␥2 antibody (Q-20; Santa Cruz Biotechnology, Inc., Santa Cruz, CA) followed by incubation with protein A-Sepharose for 2 h. Sepharose beads were then incubated in a reaction mixture containing 200 mM Tris-HCl, pH 7.5, 0.4 mM EDTA, and 0.4 mM Na 3 VO 4 in the presence or absence of 0.25 g of active Lck kinase (Upstate Biotechnology) for 10 min at 30°C. All other reagents including enzyme dilution buffer and magnesium/ATP mixture were prepared per the manufacturer's instructions. Reactions were stopped by adding SDS sample buffer and boiling samples for 10 min. Alternatively, recombinant PLC␥2 was incubated with 0.25 g of active Lck in the presence or absence of 25 M ATP in a kinase buffer (50 mM MOPS, pH 7.4, 5 mM MnCl 2 , 5 mM MgCl 2 , 5 mM dithiothreitol) at 24°C for 5 min. Reactions were stopped by the addition of SDS sample buffer and boiling.
Calcium Mobilization Assays-A20 B cells (5 ϫ 10 6 ) were loaded with 1 mM indo-1 acetoxymethylester (Molecular Probes, Inc., Eugene, OR) for 30 min at 37°C in loading medium (RPMI, 2% fetal calf serum, 10 mM HEPES). Indo-1 acetoxymethylester-loaded cells were washed and resuspended in a buffer of Hanks' balanced salt solution (HBSS (Sigma) with calcium) plus 10 mM HEPES, pH 7.0, prior to analysis. Cells were stimulated with either 10 g/ml anti-human IgM (Jackson Laboratories), 10 g/ml anti-murine IgG (Jackson Laboratories), or 150 M pervanadate and monitored for 3 min followed by the addition ionomycin (1 M) as a positive control. Intracellular calcium was measured using a bulk spectofluorimeter (Photon Technology International) and calculated as a ratio of 400/488-nm emission following 350-nm excitation. Data were analyzed with FELIX software (Photon Technology International).

Total Tyrosine Phosphorylation of PLC␥2 Is Not Indicative of Its Signaling Function in XLA B Cells-
To determine the role of Btk in PLC␥2 activation, we compared the calcium response of WT and XLA-1 human B cell lines (21) stimulated with anti-IgM or the protein-tyrosine phosphatase inhibitor, pervanadate. Consistent with our previously published data, XLA-1 B cells exhibited reduced calcium mobilization in response to IgM engagement ( Fig. 1A) Signaling via pervanadate requires the presence of a functional BCR transducer complex and promotes phosphorylation, recruitment, and activation of BCR-associated molecules (32,34). Pervanadate, however, represents a more potent mimic of BCR engagement due to irreversible oxidation of key cysteine residues on protein-tyrosine phosphatases, generating a more sustained phosphotyrosine signal (35). Notably, despite the ability of pervanadate to produce a marked enhancement in the phosphotyrosine signal of total cellular proteins compared with anti-IgM ( Fig. 1C), these events were not sufficient to rescue the calcium signaling defect in Btk-deficient cells (Fig. 1B).
Tyrosine phosphorylation of PLC␥2 is commonly used as a measure of its activation status. We compared the total PLC␥2 phosphotyrosine signal in IgM or pervanadate-stimulated WT and Btk-deficient B cells. Consistent with previous results (21), tyrosine phosphorylation of PLC␥2 in response to BCR engagement (as assessed by the panspecific anti-phosphotyrosine antibody 4G10) was indistinguishable between XLA-1 and WT B cells (Fig. 1D). Similarly, stimulation with pervanadate resulted in equivalent levels of PLC␥2 tyrosine phosphorylation in XLA-1 and WT B cells, and no significant differences were observed over an extended activation time course ( Fig. 1E and data not shown). Similar results were obtained using an alternative, panspecific anti-phosphotyrosine antibody (PY20) and in studies using an independently derived XLA B cell line (data not shown). Thus, whereas the reduction in BCR-dependent IP 3 production and calcium flux in Btk-deficient B cells clearly suggests a defect in the PLC␥2 pathway, these events do not correlate with the phosphotyrosine content of PLC␥2 as measured by conventional anti-phosphotyrosine antibodies.
Generation of Site-specific Antibodies to Candidate Regulatory PLC␥2 Phosphorylation Sites-One explanation for these data is that Btk regulates PLC␥2 activity via phosphorylation of a subset of tyrosine residues and that a loss of these specific phosphorylation events is not readily detectable using antibodies that recognize global tyrosine phosphorylation. Initial attempts to identify PLC␥2 regulatory sites by mass spectrometry were problematic due to low stoichiometric concentrations and the generation of multiple phosphopeptide fragments. Other groups have reported similar difficulties using this technique (20). We therefore chose to evaluate site-specific phosphorylation using phosphospecific antibodies generated against individual candidate sites. Potential phosphorylation sites were initially identified by comparing PLC␥2 with the better characterized, ubiquitously expressed PLC␥ isoform, PLC␥1 ( Fig. 2A). PLC␥1 contains three regulatory tyrosines: Tyr 771 and Tyr 783 within its SH2-SH3 linker region and Tyr 1254 at the carboxyl terminus. These sites are inducibly phosphorylated in response to stimulation by growth factors (36,37) or CD3 cross-linking in T cells (38), and mutation of any one of these residues leads to dysregulated phospholipase activity (30). Specifically, mutation of Tyr 783 (to Phe) results in the complete abrogation of PLC␥1 function in fibroblasts and in reconstituted PLC␥1-deficient Jurkat T cells (30,39).
Although the PLC␥ isoforms share a similar domain structure, they share only ϳ60% overall amino acid sequence homology, including less than 30% homology within the SH2-SH3 linker. Comparison with PLC␥1 revealed three potential, sim-

FIG. 1. XLA-1 B cells exhibit defective anti-IgM and pervanadate-induced calcium mobilization despite induction of an equivalent level of total PLC␥2 tyrosine phosphorylation.
Indo-1 acetoxymethylester-loaded WT and XLA-1 cells were stimulated with either 10 g/ml anti-human IgM (A) or 150 M pervanadate (B) and monitored for changes in intracellular calcium by spectrofluorimetry. C, serum-starved WT and XLA-1 B cells were stimulated with either anti-human IgM or pervanadate for the indicated times. Whole cell lysates were assessed by Western blotting with the 4G10 anti-phosphotyrosine antibody (PY). D and E, PLC␥2 was immunoprecipitated from whole cell lysate prepared as described in C following 3-min stimulation with anti-IgM or pervanadate and analyzed for total tyrosine phosphorylation. The same blot was stripped and reprobed using anti-PLC␥2 to evaluate protein loading. Representative data from more than five independent experiments are shown.
ilarly located, residues within the SH2-SH3 linker, including Tyr 743 , Tyr 753 , and Tyr 759 (Fig. 2B); however, there was no distinct homology at the carboxyl terminus. Among the initial candidate residues, location and spacing suggested that PLC␥2 Tyr 759 was most homologous to PLC␥1 Tyr 783 . When examined for sequence homology with Btk substrates (including WASP, BAP135, and the major Btk autophosphorylation site, Tyr 223 ), several sites including PLC␥2 Tyr 753 , Tyr 759 , Tyr 1197 , and Tyr 1217 were identified. Recently, Rodiguez et al. (20) also identified the Tyr 753 and Tyr 759 sites using similar methods, and Watanabe et al. (29) identified these and two additional sites (Tyr 1197 and Tyr 1217 ) using PLC␥2 GST-peptide fragments. Taken together, these findings suggested that PLC␥2 Tyr 753 , Tyr 759 , Tyr 1197 , and Tyr 1217 might be functionally homologous to regulatory phosphorylation sites in PLC␥1.
To determine whether the candidate residues within the SH2-SH3 linker region were receptor-dependent phosphorylation sites, we generated polyclonal site-specific phospho-antibodies recognizing either phosphorylated Tyr 753 or Tyr 759 . Because of the proximity of the candidate residues, the peptide immunogens overlapped by three residues, but neither contained the alternative tyrosine residue (Fig. 2B). Analysis by enzyme-linked immunosorbent assay confirmed antibody specificity for the respective phosphorylated PLC␥2 peptide immunogen (data not shown).
Antibody specificity was next evaluated in the context of a native protein species by analysis of recombinant WT, single mutant (Y753F or Y759F), or double mutant (Y753F/Y759F) PLC␥2 phosphorylated in vitro using the purified Src family kinase, Lck. Recombinant Lck was utilized because of ease of expression, retained activity in bacterial cultures, and recent work demonstrating its ability to phosphorylate PLC␥2 Tyr 753 and Tyr 759 in vitro (19). Whereas each antibody recognized phosphorylated WT PLC␥2 or PLC␥2 with a Tyr to Phe mutation at the alternate residue, mutation of the original tyrosyl epitope resulted in the complete loss of antibody reactivity. There was no reactivity with the unphosphorylated PLC␥2 species or with PLC␥2 mutated at both Tyr 753 and Tyr 759 (Fig.  2C). In addition, each antibody fully recognized PLC␥2 mutated at the putative carboxyl residues (Fig. 7A, right panel). Neither antibody exhibited reactivity with overexpressed or immunoprecipitated PLC␥1 (data not shown). The specificity of each of these antibodies within the context of a complex mixture of phosphorylated B lineage signaling proteins was also FIG. 2. PLC␥2 P-Y753 and PLC␥2 P-Y759 antibodies are specific for induced phosphorylation at individual sites. A, domain structure of PLC␥ isoforms and location of established and potential regulatory tyrosine residues within the SH2-SH3 linker and carboxyl terminus regions of PLC␥2 and PLC␥1. Putative sites in PLC␥2 were chosen based on similarity of location compared with known PLC␥1-regulatory sites, Tyr 771 , Tyr 783 , and Tyr 1254 and/or homology to known Btk phosphorylation sites. B, alignment of the SH2-SH3 linker regions of human PLC␥ isoforms. Phosphospecific antibodies were generated against predicted phosphorylation sites, Tyr 753 and Tyr 759 , using the indicated peptide sequences. C, left panels, bacterially expressed, full-length recombinant WT, Y753F, and Y759F mutant human PLC␥2 were purified and phosphorylated in vitro using purified recombinant Lck in the absence (Ϫ) or presence (ϩ) of ATP as described under "Experimental Procedures." Right panel, virally expressed recombinant WT and Y753F/Y759F PLC␥2 were purified and incubated with Mg ϩ / ATP in the presence or absence of Lck. Samples were analyzed by Western blotting with each PLC␥2 phosphospecific antibody. Blots were subsequently stripped and reblotted with an anti-PLC␥2 antibody. D, Ramos B cells were serumstarved for 45 min and stimulated with pervanadate for 3 min at 37°C. Whole cell lysates from unstimulated and stimulated cells were assessed by sequential immunoblotting with the indicated antibodies in the order shown (left to right). E, Ramos cells were serum-starved for 45 min and stimulated with anti-IgM for the times indicated. Whole cell lysates were analyzed by sequential immunoblotting with the antibodies noted (top to bottom). Experiments utilizing an alternative order of sequential blotting yielded identical results. Representative data from one of at least three independent experiments are shown. evaluated by sequential Western blotting. Using whole cell lysates from unstimulated and pervanadate-stimulated Ramos B cells (Fig. 2D), we found that each antibody predominantly recognized a phosphoprotein species that co-migrated precisely with PLC␥2.
We next determined if PLC␥2 Tyr 753 and Tyr 759 were modified in response to BCR engagement. Whole cell lysates were prepared from Ramos B cells stimulated with anti-IgM and sequentially blotted with each phosphospecific antibody followed by blotting with an anti-PLC␥2-specific antibody. Phosphorylation at both sites was readily detectable within 30 s of IgM receptor engagement and remained above the basal level of phosphorylation for up to 40 min (Fig. 2E and data not  shown).
Tyr 753 and Tyr 759 Are Crucial for Inducible Phospholipase Activity and for PLC␥2-mediated Calcium Signaling-To evaluate the role for PLC␥2 Tyr 753 /Tyr 759 site-specific phosphorylation in PLC␥2 function, we examined the consequences of site-directed mutagenesis of these residues. Using recombinant PLC␥2 and Lck, the basal and inducible activity of WT, Y753F, Y759F, and Y753F/Y759F PLC␥2 were determined as measured by IP 3 production (in the presence or absence of ATP). Whereas WT PLC␥2 exhibited an ϳ8-fold increase in IP 3 production in the presence of Lck and ATP, PLC␥2 Y753F and PLC␥2 Y759F single mutant proteins yielded only a 2-3-fold increase. Further mutation of both residues (PLC␥2 Y753F/ Y759F) effectively abrogated inducible phospholipase activity in all experiments (Fig. 3A). We observed variation in the basal catalytic activity of the PLC␥2 Y753/Y759F mutant (slightly above or below WT in different experiments), and this has precluded speculation as to the effect of this mutation on basal PLC␥2 activity. In contrast to Tyr 753 and Tyr 759 , mutation of an alternate tyrosine within the SH2-SH3 linker, Tyr 743 , resulted in basal and inducible lipase activity that was indistinguishable from WT PLC␥2. This indicates that, unlike Tyr 753 and Tyr 759 , Tyr 743 did not function as a regulatory tyrosine.
We next assayed the functional effect of the Y753/Y759 mutation within the context of antigen receptor signaling in mammalian cells. A20 murine B cells were mock-infected or infected with vaccinia viruses expressing WT or the Y753F/Y759F mutant PLC␥2, and calcium mobilization was evaluated. Whereas overexpression of WT protein had no appreciable effect even at very high expression levels (5-10-fold above endogenous), overexpression of PLC␥2 Y753F/Y759F led to a dose-dependent reduction in both peak and sustained calcium levels (Fig. 3B). Notably, the calcium signal in cells expressing the highest dosage of the Y753F/Y759F mutant was similar to that observed in Btk-deficient murine and human B cells. These data, in conjunction with previous studies (20,29), indicate that phosphorylation of Tyr 753 and Tyr 759 is integral to the activation PLC␥2 and generation of the BCR-dependent sustained calcium signal.
BCR-dependent Phosphorylation of PLC␥2 Tyr 753 and Tyr 759 Requires Btk Kinase Activity-If phosphorylation at PLC␥2 Tyr 753 and Tyr 759 is required for receptor-mediated sustained calcium mobilization, then the diminished calcium signal in Btk deficient B cells might be attributable to a loss of phosphorylation at these sites. To test this possibility, we first examined phosphorylation of PLC␥2 in primary B cells from WT (Balb/c) and Btk mutant (Balb/XID) mice. Total splenocytes from each strain were stimulated with anti-IgM for the indicated times and assessed for site-specific phosphorylation by sequential Western blotting. Western blotting of lysates from both strains demonstrated a marked, but equivalent, increase in overall tyrosine phosphorylation of multiple BCR-dependent substrates (Fig. 4A, bottom panel). Phosphorylation of both Tyr 753 and Tyr 759 was readily identified following stimulation of WT splenocytes, peaking between 2 and 5 min and diminishing to nearly basal levels by ϳ20 min (Fig. 4A, left). Conversely, BCR-stimulated Xid cells exhibited a minimal, transient increase in Tyr 759 phosphorylation and no detectable increase in Tyr 753 phosphorylation (Fig. 4A, right). Densitometry analysis (corrected for the ϳ1.3-fold larger number of WT IgM ϩ B cells) indicated an ϳ3-fold decrease in Tyr 759 phosphorylation in Btk-deficient versus WT cells. (Fig. 4B). Similar data were also obtained using purified CD43 ϩ -depleted, splenic B cells (containing Ͼ90% B220 ϩ cells; data not shown).
We next determined whether Tyr 753 and Tyr 759 phosphorylation was also altered in human XLA B cells. WT and XLA-1 B cells were stimulated with anti-IgM for the indicated times and evaluated using serial Western blotting. Whereas total PLC␥2 tyrosine phosphorylation was indistinguishable (Fig. 1D), XLA-1 B cells exhibited a striking reduction in Tyr 759 phosphorylation as assessed over a 30-min time period (Fig. 4C). Western blotting of the same lysates using the anti-Tyr(P) 753 antibody did not yield a significant signal in either WT or XLA-1 lines. This finding most likely reflects the relatively weaker antibody reactivity of anti-Tyr(P) 753 and less robust BCR signaling activity in these LMP2-deficient Epstein-Barr virustransformed B cell lines. To overcome the limited signal intensity, we examined site-specific phosphorylation using the more potent BCR mimetic, pervanadate. Pervanadate activation of WT cells resulted in rapid phosphorylation at both Tyr 753 and Tyr 759 (Fig. 4D). In contrast, despite an equivalent overall phosphotyrosine signal (Fig. 4D, third panel), pervanadatestimulated XLA-1 B cells exhibited minimal phosphorylation at Tyr 753 and Tyr 759 (Fig. 4D). Relative to WT B cells, phosphorylation at Tyr 753 and Tyr 759 was reduced 4-and 3-fold, respectively. Similar results were obtained using an alternative LMP2-deficient Epstein-Barr virus-transformed XLA B cell line (data not shown). Thus, phosphorylation of Tyr 753 and Tyr 759 required the presence of functional Btk.
Since calcium signaling in XLA-1 B cells can be restored by exogenous expression of wild type Btk (21, 40), we next investigated whether WT Btk expression would restore PLC␥2 sitespecific tyrosine phosphorylation. XLA-1 B cells were mockinfected or infected with vaccinia virus expressing Btk. Expression of WT Btk restored PLC␥2 Tyr 759 phosphorylation in both IgM-and pervanadate-stimulated cells and also restored Tyr 753 phosphorylation in pervanadate-stimulated XLA B cells (Fig. 5, A and B). In contrast, a kinase-inactive Btk mutant (Btk K430R) had no effect (data not shown). Together, these results demonstrate that Btk mediates BCR-dependent site-specific phosphorylation of the PLC␥2 SH2-SH3 linker and is the major kinase responsible for this activity.
PLC␥2 Tyr 753 and Tyr 759 Phosphorylation Is Specifically Restored by Tec, but Not Syk, Family Kinases-The Tec family of nonreceptor tyrosine kinases includes Tec, Btk, Itk (Tsk/Emt), Rlk (Txk), and Bmx (41,42). Previous work has demonstrated that alternative Tec kinases can functionally reconstitute the calcium signal in Btk-deficient B cell models (21,40). Additionally, Syk can mediate site-specific phosphorylation of PLC␥1 in vitro (18), and BCR-dependent PLC␥2 phosphorylation is markedly reduced in Syk-deficient B cell lines (43). To identify kinases capable of regulating PLC␥2, we evaluated the ability of alternative Tec and Syk/Zap-70 family kinases to restore PLC␥2 site-specific phosphorylation in Btk-deficient B cells. XLA-1 B cells were mock-infected or infected with vaccinia viruses expressing candidate Tec or Syk family kinases and stimulated with pervanadate. Whereas expression of either Tec or Itk fully restored phosphorylation at Tyr 753 and Tyr 759 in XLA-1 B cells (Fig. 5C), overexpression of Syk or Zap-70 had no appreciable effect (Fig. 5D). These findings correlate with previous data demonstrating reconstitution of calcium signaling in content) indicating the relative signal intensity are displayed below the anti-P-Y759 phosphospecific blot. D, serum-starved WT and XLA-1 B cells were stimulated with pervanadate for 3 min and lysed, and PLC␥2 was immunoprecipitated and immunoblotted sequentially with anti-P-Y753, anti-P-Y759, anti-PLC␥2, and anti-phosphotyrosine antibodies, respectively. Representative data from one of more than five independent experiments are shown.

FIG. 4. Btk-deficient B cells exhibit markedly reduced sitespecific PLC␥2 tyrosine phosphorylation in response to BCR engagement.
A, total splenocytes were isolated from Balb/c and Balb/ XID mice, serum-starved for 10 min, and activated by 10 g/ml antimurine IgM cross-linking for the times indicated. Whole cell lysates were analyzed by sequential immunoblotting with P-Y759 and P-Y753 phosphospecific antibodies. The membrane was also stripped and sequentially blotted with anti-PLC␥2, anti-Btk, and anti-phosphotyrosine antibodies to assess protein loading and cellular activation. Densitometry was performed to measure the relative phosphorylation of Tyr 753 and Tyr 759 between Balb and Balb/Xid splenocytes (data under the respective blot). Data were normalized to PLC␥2 protein content. B, relative phosphorylation of PLC␥2 Tyr 759 was quantified by densitometry for each time point and normalized to both PLC␥2 protein content and to reflect the reduced number of IgM ϩ cells in Xid (25% fewer based on B220 staining and fluorescence-activated cell sorting analysis) in the experiment shown. Representative data from one of three experiments are shown. C, serum-starved WT and XLA-1 B cells were stimulated with anti-IgM for the times indicated. Whole cell lysates were blotted with the anti-P-Y759 antibody followed by anti-PLC␥2 blotting to evaluate protein loading. Densitometry data (normalized to total PLC␥2 Btk-deficient cells with the addition of Tec, but not Syk, kinases (21) (data not shown).
BCR-dependent Phosphorylation of PLC␥1 Tyr 783 Also Requires Btk-The specificity of Btk in PLC␥2 regulatory site phosphorylation suggested that Tec kinases might play an analogous role in the regulation of other PLC␥ isoforms. Whereas PLC␥2 is the predominant isoform expressed in B cells, PLC␥1 is also expressed at low levels (2,3) and is inducibly phosphorylated in response to receptor engagement (5). Similar to our studies with PLC␥2, preliminary experiments demonstrated equivalent levels of BCR-inducible total PLC␥1 phosphorylation in WT and XLA cells. Therefore, we examined phosphorylation of the key regulatory site, Tyr 783 , in the PLC␥1 SH2-SH3 linker. Following stimulation with pervanadate, PLC␥1 was immunoprecipitated from WT and XLA-1 B cells and Western blotted using a commercially available, Tyr 783 -phosphospecific antibody. Tyr 783 phosphorylation was reduced ϳ4-fold in XLA-1 cells compared with WT cells (Fig.  6A). Expression of Btk or Itk fully restored Tyr 783 phosphorylation (Fig. 6B), whereas, similar to PLC␥2, neither Syk nor Zap-70 could restore PLC␥1 SH2-SH3 regulatory site-specific phosphorylation.
The PLC␥2 Carboxyl-terminal Residues, Tyr 1197 and Tyr 1217 , Are Not Btk-dependent Phosphorylation Sites-Despite the loss of phosphorylation at two key residues, the high level of PLC␥2 total tyrosine phosphorylation in Btk-deficient cells signified the presence of additional phosphorylation sites. Indeed, our initial sequence analysis of PLC␥2 suggested at least two additional candidate phosphorylation sites in the carboxyl terminus. Watanabe et al. (29) recently identified these residues (Tyr 1197 and Tyr 1217 ) as potential Btk-dependent phosphorylation sites using an in vitro kinase assay with PLC␥2 GSTpeptide fragments and purified Btk.
To determine the events regulating BCR-dependent PLC␥2 carboxyl phosphorylation and the specific role for Btk in this FIG. 5. BCR-dependent, site-specific phosphorylation of PLC␥2 in XLA-1 B cells can be restored by Tec, but not Syk/Zap-70, family kinases. A, XLA-1 B cells were mock-infected or infected using recombinant vaccinia virus expressing wild type Btk (MOI of 10 for 12 h) Cells were serum-starved and stimulated with anti-IgM for 5 min and lysed. Whole cell lysates were prepared as described under "Experimental Procedures" and sequentially blotted for specific phosphorylation and expression. B, XLA-1 B cells were infected as in A. Cells were serum-starved and stimulated with pervanadate for 3 min, and PLC␥2 was immunoprecipitated. Site-specific phosphorylation was evaluated by sequential Western blotting with the antibodies indicated (top to bottom). C and D, WT Btk, Itk, Tec, Syk, and Zap-70 proteins were expressed in XLA-1 cells using recombinant vaccinia viruses followed by stimulation as described in B. PLC␥2 was immunoprecipitated and analyzed by sequential immunoblotting with the indicated antibodies. Representative data from one of more than three independent experiments are shown. Whole cell lysates were blotted with antibodies to Btk, Syk, and Zap-70 to evaluate protein expression (D, bottom panels).

FIG. 6. BCR-dependent phosphorylation of PLC␥1 Y783 requires Btk and is restored by Tec but not Syk family kinases. A,
WT and XLA-1 B cells were serum-starved and stimulated with pervanadate for 3 min, and PLC␥1 was immunoprecipitated. Immunoprecipitates were immunoblotted with an anti-P-Y783 antibody followed by anti-PLC␥1. Densitometry data (normalized to total PLC␥1 content), indicating the relative signal intensity are displayed below the anti-P-Y783 phosphospecific blot. B, WT Btk, Itk, Tec, Syk, and Zap-70 proteins were expressed in XLA-1 cells using recombinant vaccinia. PLC␥1 immunoprecipitates were sequentially blotted as in A. Whole cell lysates were blotted with antibodies to Btk, Itk, Syk, and Zap-70 to evaluate protein expression (bottom panels). Representative data from one of three independent experiments are shown. process, we generated phosphospecific antibodies against Tyr 1197 and Tyr 1217 . Individual antibody specificity was evaluated by enzyme-linked immunosorbent assay (data not shown) and by full-length wild type, Y1197F, Y1217F, and Y1197F/ Y1217F PLC␥2 phosphorylated in vitro by Lck. Both antibodies specifically recognized WT PLC␥2 or PLC␥2 mutated at the alternate carboxyl tyrosine residue (Fig. 7A, left). Individual phosphoantibodies were then further examined for reactivity among all four inducible sites. Analysis demonstrated the absence of antibody cross-reactivity between the Btk phosphorylation consensus site sequences located in the SH2-SH3 linker and carboxyl terminus (Fig. 7A, right).
Initial analysis revealed that both residues were inducibly phosphorylated in response to BCR engagement in a B cell line as well as in primary B cells (Fig. 7, B and C). We next evaluated the kinetics of Tyr 1197 and Tyr 1217 phosphorylation in IgM-stimulated Balb and Balb/Xid primary B cells. In contrast to Tyr 753 and Tyr 759 , BCR engagement led to an equivalent level of inducible phosphorylation at both Tyr 1197 and Tyr 1217 in splenocytes from both strains at all time points (Fig.  7C). Similarly, WT and XLA-1 B cells exhibited indistinguishable levels of Tyr 1197 and Tyr 1217 phosphorylation at all time points following pervanadate stimulation (Fig. 7D). Consistent with these data, reconstitution of XLA-1 cells with WT Btk had no effect on Tyr 1197 phosphorylation. Btk overexpression did lead to a modest, but consistent, increase in phosphorylation at Tyr 1217 (Fig. 7E).
Together, these data demonstrate that Btk is not required for phosphorylation at the PLC␥2 carboxyl terminus. However, it remains possible that, under some activation conditions, Btk may contribute to phosphorylation at Tyr 1217 . DISCUSSION The amplitude and duration of the intracellular calcium signal is a key determinant of the cellular response to immunoreceptor engagement. In particular, a sustained calcium signal is crucial for the selective activation or inhibition of specific subsets of transcription factors (9,10,44) and for the diacylglycerol-associated NF-B-dependent survival signal (45). Although Btk is clearly required for sustained calcium signaling, the mechanisms by which it exerts this unique effect remain incompletely defined. The current work offers important insight into these events by providing in vivo evidence that a key, nonredundant function of Btk/Tec kinases is to phosphorylate PLC␥2 at two major regulatory sites, Tyr 753 and Tyr 759 , crucial for immunoreceptor-dependent PLC␥2 activation.
Our studies complement previous work demonstrating an essential role for the SH2-SH3 linker region in regulating PLC␥ function (19,20,29,30,36). Characterization of PLC␥1 identified key inducibly phosphorylated regulatory sites, Tyr 771 and Tyr 783 , within the PLC␥1 SH2-SH3 linker (5,37,38). Mutation of either residue resulted in altered plateletderived growth factor-induced PLC␥1 activity. In these studies, PLC␥1 Tyr 783 served a positive regulatory role, whereas Tyr 771 appeared to be a negative regulator (30). Although the PLC␥1 and PLC␥2 SH2-SH3 linker sites are unlikely to be wholly analogous in function, this region is clearly pivotal in the regulation of PLC␥ activity. Using phosphospecific antibodies, we demonstrate that BCR engagement leads to a rapid and sustained phosphorylation at two PLC␥2 SH2-SH3 linker residues, Tyr 753 and Tyr 759 . Consistent with a predicted regulatory function, phosphorylation of Tyr 753 and Tyr 759 (but not a similarly located linker tyrosine, Tyr 743 ) was required for the optimal inducible enzymatic activity of PLC␥2 in vitro. Analogous to results in T cells overexpressing the PLC␥1 linker mutant Y783F (39), overexpression of PLC␥2 linker mutants promoted a dose-dependent inhibition of BCR-mediated calcium mobili- zation. These results are consistent with studies in PLC␥2deficient avian B lymphoma cells that demonstrate a markedly reduced calcium response in cells reconstituted with PLC␥2 Y753/Y759 double mutant proteins (20,29).
Most notably, our data demonstrate that BCR-dependent PLC␥2 activation is mediated almost entirely via Tec/Btk kinase site-specific tyrosine phosphorylation within the PLC␥ SH2-SH3 linker. Whereas the relative total level of PLC␥2 tyrosine phosphorylation appears intact in IgM-and pervanadate-stimulated Btk-deficient human B cells, there was a striking reduction in linker site-specific phosphorylation. Similarly, BCR-dependent PLC␥2 Tyr 753 and Tyr 759 phosphorylation was severely reduced in splenic B cells from Btk mutant, XID mice. The limited residual, inducible Tyr 759 phosphorylation signal present in XID cells declined much more rapidly (peaking at 2 min) than the coordinate signal in WT cells, where phospho-Tyr 759 was detectable for Ͼ15 min following receptor engagement. These kinetic data directly parallel previous functional data in which Btk deficient B cells were unable to maintain IP 3 production in response to receptor engagement (21,28). Strikingly, the addition of Btk or other Tec family kinase members fully restored receptor-mediated PLC␥2 site-specific phosphorylation at both residues in XLA B cells. This correlates with previous data demonstrating functional reconstitution of the sustained calcium signal in Btk deficient cells by Tec family kinases (21,40). In addition, reconstitution of site-specific phosphorylation required the kinase activity of Btk, since overexpression of kinase-inactive Btk failed to restore linker phosphorylation (data not shown).
Despite the marked reduction in Tyr 753 and Tyr 759 phosphorylation in both XLA-1 and XID B cells, a low level of inducible phosphorylation remained detectable at both sites. Although Syk/Zap70 kinases have been implicated in PLC␥ regulation (11,43) several lines of evidence suggest that endogenous Syk is unlikely to modify these regulatory sites. Interpretation of the potential role for Syk in PLC␥2 regulation is complicated by its multiple functions including phosphorylation of BLNK and activation of phosphatidylinositol 3-kinase, events that could directly or indirectly lead to reduced PLC␥2 phosphorylation (43, 46 -48). These Syk/BLNK-dependent events, however, remain unaffected in Btk-deficient cells where a marked loss in Tyr 753 and Tyr 759 phosphorylation is observed (11,49). Furthermore, whereas Tec kinases rescued PLC␥2 site-specific phosphorylation, Syk/Zap-70 kinases failed to do so even at supraendogenous expression levels. Finally, a previous study clearly indicated that PLC␥2 is a relatively poor substrate for purified recombinant Syk in vitro (20).
The most likely explanation for residual linker phosphorylation in XLA B cells is redundancy provided by alternative Tec kinase(s) (41,50). All hematopoietic cell lineages express at least two distinct Tec family kinases. B lineage cells, including the XLA and XID B cell populations studied here, express both Btk and Tec. Consistent with this idea, the developmental differences between Btk-deficient humans and mice are largely attributable to variances in Tec activity and/or expression levels (51). Indeed, overexpressed Tec can fully compensate for Btk-dependent PLC␥2 site-specific phosphorylation and calcium mobilization (Fig. 5C) (21,40). We have attempted to determine the relative contribution of Tec in BCR-dependent signaling using Btk/Tec double-deficient mice. Unfortunately, the severe reduction in peripheral B cell numbers has made this approach difficult (51) (data not shown). Whereas our current data cannot rule out a role for non-Tec family kinases, these combined observations strongly suggest that PLC␥2 SH2-SH3 linker phosphorylation is predominantly, if not entirely, dependent on Tec family kinases.
Our data also suggest Tec kinase-dependent SH2-SH3 linker phosphorylation represents a key control point for BCR-dependent activation of PLC␥1. Correlating with our PLC␥2 data, phosphorylation of the PLC␥1 SH2-SH3 linker regulatory site Tyr 783 was also markedly reduced in XLA-1 cells, and this phosphorylation could be restored by the expression of Tec, but not Syk, family kinases. These data are consistent with observations in T cells expressing raft-targeted PLC␥1. In those studies, PLC␥1 phosphorylation and calcium-dependent gene expression occurred independently of Zap-70 or the adaptors LAT and Slp-76. Furthermore, membrane-targeted PLC␥1 was readily phosphorylated by the raft-associated Tec kinase, Rlk, but not by Zap-70 (52). Because PLC␥1 is activated in response to multiple alternative immunoreceptors, this Tec kinase-dependent regulatory mechanism is likely to be broadly conserved.
The level of receptor-dependent global tyrosine phosphorylation in Btk-deficient B cells suggested the presence of additional tyrosine phosphorylation sites within PLC␥2. A recent study using PLC␥2 GST-peptide fragments identified two carboxyl-terminal phosphorylation sites, Tyr 1197 and Tyr 1217 , and suggested that both sites were targets for Btk (29). Our data clearly demonstrate that, in the context of BCR-dependent FIG. 8. A working model for PLC␥2 mediated, BCR-dependent calcium signaling. A, engagement of the BCR initiates formation of a lipid raft-associated signalosome providing access to substrates for recruitment and activation. B, BLNK-mediated recruitment of PLC␥2 to the signalosome promotes phosphorylation of the PLC␥2 carboxyl residues by an Src family kinase (most likely Lyn) and of the PLC␥2 SH2-SH3 linker by Btk/Tec kinases. This latter event is essential for full PLC␥2 enzymatic activity. Btk-dependent recruitment of phosphatidylinositol-4-phosphate 5-kinase (PIP5K) serves to increase local levels of both phosphatidylinositol 4,5-bisphosphate (PIP2) and phosphatidylinositol 3,4,5-trisphosphate (PIP3). Thus, localization of PLC␥2 with the Btk-phosphatidylinositol-4-phosphate 5-kinase complex functions coordinately to regulate PLC␥2 activity through the phosphorylation of four key tyrosine residues and through enhanced accessibility of PLC␥2 to its enzymatic substrate. See "Discussion" for details. PI3K, phosphatidylinositol 3-kinase; PKC, protein kinase C; SOC, store-operated calcium channel. signaling, phosphorylation of Tyr 1197 and Tyr 1217 occurs independently of endogenous Btk activity. However, the ability of overexpressed Btk to enhance Tyr 1217 phosphorylation suggested that, under some circumstances, Btk might participate with an alternative kinase(s) in the optimal phosphorylation of this site. Primary phosphorylation of the PLC␥2 carboxyl regulatory sites is most likely mediated by a Src family kinase(s), since recent studies, including the current work, demonstrate that Src family kinases can directly phosphorylate PLC␥2 in vitro (19,20). In contrast, Syk overexpression had no additive effect on Tyr 1197 and Tyr 1217 phosphorylation (Fig. 7E), and Syk does not readily modify purified PLC␥2 (20). Notably, however, phosphorylation of these additional BCR-dependent sites, in the absence of Btk activity, is clearly insufficient to mediate the sustained activation of PLC␥2.
Our data indicate the need for caution in utilizing global phosphotyrosine content as a measure of PLC␥2 activity. A lack of correlation between receptor-mediated PLC␥ phosphorylation and calcium mobilization has been reported in Btk-deficient cell models including signaling via the calnexin-surrogate light chain-CD79b complex on pro-B cells (53), the BCR on B cells (21,26), and the Fc⑀RI receptor on mast cells (28). Similar observations have also been reported with respect to oxidative stress signaling in B cells (54) and collagen receptor-mediated calcium signaling in platelets (27). These results have led to confusion regarding interpretation of the role for Btk in downstream signals. Use of phosphospecific antibodies should clarify the events modulating activation of this pivotal enzymatic pathway in these and additional receptor systems.
Current observations, in association with previously published data, suggest the following working model for regulation of PLC␥2 activity and BCR-dependent calcium signaling. Upon receptor engagement, raft-associated Src kinases are activated, and Syk is recruited to the receptor complex and activated. Syk then phosphorylates the adaptor molecule, BLNK, resulting in the recruitment of a BLNK-PLC␥2 complex to the plasma membrane (46,55). Concurrently, Btk is recruited to the membrane via its pleckstrin homology domain and activated by Src family kinases (23,33,56) (Fig. 8A). Once co-localized in the BCR signalosome, activated Btk is within proximity to PLC␥2 (through a direct association with BLNK (46), a direct interaction with PLC␥2 (57), or an alternative, Btk SH2-mediated interaction (58)). These combined events promote regulatory phosphorylation of the PLC␥2 SH2-SH3 linker and carboxyl terminus by Btk/Tec and Src kinases, respectively. Activation of PLC␥2, specifically following phosphorylation of Tyr 753 and Tyr 759 , leads to the generation of threshold IP3 levels, required for depletion of intracellular calcium stores and induction of store-operated calcium influx. Recent collaborative work from our laboratory has also demonstrated that Btk is constitutively associated with phosphatidylinositol-4-phosphate 5-kinase isoforms. By means of this interaction, Btk and phosphatidylinositol-4-phosphate 5-kinase are coordinately recruited into the BCR signalosome, promoting an increase in the local levels of phosphatidylinositol 4,5-bisphosphate and phosphatidylinositol 3,4,5-trisphosphate (59). Retention of this Btk-PLC␥2-phosphatidylinositol-4-phosphate 5-kinase complex at the membrane therefore allows for continued activation of PLC␥2 and generation of peak and sustained IP 3 levels to mediate the store-operated calcium channel-dependent, sustained calcium signal (Fig. 8B).