Sustained Activation of the Tyrosine Kinase Syk by Antigen in Mast Cells Requires Local Ca2+ Influx through Ca2+ Release-activated Ca2+ Channels*

Mast cell activation involves cross-linking of IgE receptors followed by phosphorylation of the non-receptor tyrosine kinase Syk. This results in activation of the plasma membrane-bound enzyme phospholipase Cγ1, which hydrolyzes the minor membrane phospholipid phosphatidylinositol 4,5-bisphosphate to generate diacylglycerol and inositol trisphosphate. Inositol trisphosphate raises cytoplasmic Ca2+ concentration by releasing Ca2+ from intracellular stores. This Ca2+ release phase is accompanied by sustained Ca2+ influx through store-operated Ca2+ release-activated Ca2+ (CRAC) channels. Here, we find that engagement of IgE receptors activates Syk, and this leads to Ca2+ release from stores followed by Ca2+ influx. The Ca2+ influx phase then sustains Syk activity. The Ca2+ influx pathway activated by these receptors was identified as the CRAC channel, because pharmacological block of the channels with either a low concentration of Gd3+ or exposure to the novel CRAC channel blocker 3-fluoropyridine-4-carboxylic acid (2′,5′-dimethoxybiphenyl-4-yl)amide or RNA interference knockdown of Orai1, which encodes the CRAC channel pore, all prevented the increase in Syk activity triggered by Ca2+ entry. CRAC channels and Syk are spatially close together, because increasing cytoplasmic Ca2+ buffering with the fast Ca2+ chelator 1,2-bis(2-aminophenoxy)ethane-N,N,N′,N′-tetraacetic acid tetrakis failed to prevent activation of Syk by Ca2+ entry. Our results reveal a positive feedback step in mast cell activation where receptor-triggered Syk activation and subsequent Ca2+ release opens CRAC channels, and the ensuing local Ca2+ entry then maintains Syk activity. Ca2+ entry through CRAC channels therefore provides a means whereby the Ca2+ and tyrosine kinase signaling pathways can interact with one another.

The cell surface IgE receptor, FC⑀RI, recognizes the Fc region of antigen-specific IgE molecules. Cross-linking of FC⑀RI receptors with antigen is a critical early step in mast cell activation (1,2). FC⑀RI receptors couple to a cascade of protein kinases, among which the non-receptor tyrosine kinase Syk plays a pivotal role in mast cell stimulation (3,4). Syk binds to tyrosine-phosphorylated residues in the immunoreceptor tyrosine-based activation motif of the FC⑀RI ␤ and ␥ chains via its two SH2 domains, increasing enzyme activity (5)(6)(7). Syk-null cells fail to degranulate following exposure to antigen and antisense oligonucleotides directed against Syk reduce antigendriven responses in an in vivo rat model of asthma (8). In the RBL mast cell line, inhibition of Syk suppresses degranulation (9) as well as production of the pro-inflammatory cysteinyl leukotrienes (10).
Activated Syk stimulates phospholipase C␥1 in RBL cells (11), which hydrolyzes the phospholipid phosphatidylinositol 4,5-bisphosphate to generate inositol trisphosphate (InsP 3 ) 4 and diacylglycerol (12). InsP 3 binds to tetrameric InsP 3 -gated Ca 2ϩ channels in the endoplasmic reticulum, releasing stored Ca 2ϩ into the cytosol. The ensuing store depletion then activates store-operated CRAC channels in the plasma membrane (13), which provide a substantial portion of Ca 2ϩ needed to activate mast cells. Genome-wide RNAi knockdown strategies have identified two key molecular components of the storeoperated Ca 2ϩ entry pathway, STIM1 (14,15) and Orai1 (16 -18). STIM1, a protein that spans the endoplasmic reticulum, is the Ca 2ϩ sensor that detects the fall in Ca 2ϩ content within the store. It migrates from a relatively homogeneous distribution throughout the endoplasmic reticulum to discrete puncta within 25 nm of the plasma membrane, lining up opposite the plasma membrane protein Orai1 (19). Site-directed mutagenesis has established that Orai1 is all or part of the CRAC channel pore (20 -22). The importance of STIM1 and Orai1 in mast cell function is underscored by the findings that degranulation, secretion of pro-inflammatory cysteinyl leukotrienes and chemokines, as well as the ability to mount an inflammatory response are all severely compromised in mice in which these genes have been ablated (23,24).
Activation of immune cells often requires sustained Ca 2ϩ entry through CRAC channels (25). This in turn is dependent on a maintained elevation in InsP 3 levels (and thus phospholipase C␥ activity), which is needed to ensure the stores are depleted sufficiently for CRAC channels to remain open. * This work was supported in part by the Medical Research Council. The costs Because phospholipase C␥ can be activated by Syk, we have examined whether the subsequent Ca 2ϩ influx can feed back to maintain Syk activation. Our findings reveal a novel self-regenerative process whereby local Ca 2ϩ influx through CRAC channels increases Syk activity, which in turn sustains CRAC channel activity by preventing store refilling.

EXPERIMENTAL PROCEDURES
Cell Culture-RBL-1 cells were bought from ATCC. Cells were cultured (37°C, 5% CO 2 ) in Dulbecco's modified Eagle's medium with 10% fetal bovine serum, 2 mM L-glutamine, and penicillin-streptomycin, as described previously (26). For Ca 2ϩ imaging and patch clamp experiments, cells were passaged (using trypsin) onto glass coverslips and used 24 -48 h after plating. Cells were sensitized to antigen (dinitrophenyl-bovine serum albumin, 80 g/ml) by incubation in IgE (2.5 g/ml) overnight in standard culture medium.
Ca 2ϩ Imaging-Ca 2ϩ imaging experiments were carried out using the IMAGO charge-coupled device camera-based system from TILL Photonics, as described previously (10). Cells were alternately excited at 356 and 380 nm (20-ms exposures, 0.5 Hz) using a Polychrome Monochromator. Images were analyzed offline using IGOR Pro. Cells were loaded with Fura-2-AM (2 M) for 40 min at room temperature in the dark and then washed in standard external solution of composition (in mM) NaCl 145, KCl 2.8, CaCl 2 2, MgCl 2 2, D-glucose 10, HEPES 10, pH 7.4, with NaOH. Ca 2ϩ signals are presented as the ratio 356/380.
BAPTA-AM Loading-Cells were incubated in BAPTA-AM (20 M) for 45 min at room temperature as previously described (27).
CRAC Channel Blocker-The CRAC channel blocker used in this study, which we call Synta compound, was kindly provided by Dr. Valerie Morisset at GlaxoSmithKline, UK. The Synta compound is compound 66 from the WO2005/009954 A2 patent (3-fluoropyridine-4-carboxylic acid (2Ј,5Ј-dimethoxybiphenyl-4-yl)amide (STRUCTURE 1)). The inhibitor was made up as a stock solution of 10 mM in DMSO.
Patch Clamp Recordings-Whole cell patch clamp recordings were carried out as described (28). Sylgard-coated, firepolished patch pipettes were filled with a solution that contained 145 mM cesium glutamate, 8 mM NaCl, 1 mM MgCl 2 , 2 mM Mg-ATP, 10 mM HEPES, 10 mM EGTA, 30 M InsP 3 , pH 7.2, with CsOH. Pipette resistance was ϳ5 megohms when placed in an external solution containing 145 mM NaCl, 2.8 mM KCl, 10 mM CsCl, 10 mM CaCl 2 , 2 mM MgCl 2 , 10 mM D-glucose, 10 mM HEPES, pH 7.4, with NaOH. The inwardly rectifying K ϩ current was measured with a pipette solution containing 145 mM potassium glutamate, 8 mM NaCl, 1 mM MgCl 2 , 2 mM Mg-ATP, 10 mM HEPES, 0.1 mM EGTA, pH 7.2, with KOH. Bath solution for measuring the K ϩ current contained 108 mM NaCl, 50 mM KCl, 10 mM CaCl 2 , 2 mM MgCl 2 , 10 mM D-glucose, 10 mM HEPES, pH 7.4, with NaOH. A correction of ϩ10 mV was applied for the subsequent liquid junction potential that arose from the glutamate-based pipette solutions. Ca 2ϩ current through CRAC channels and K ϩ current through inwardly rectifying K ϩ channels were followed by applying voltage ramps (at 0.5 Hz) spanning Ϫ100 to ϩ100 mV in 50 ms from a holding potential of 0 mV. Current amplitudes were measured from the ramps at Ϫ80 mV and normalized to cell size by dividing the amplitude by cell capacitance. Currents were filtered using an 8-pole Bessel filter at 2.5 kHz and digitized at 100 s.
Preparation of Cell Lysates-Attached cells from 6-cm plastic dishes were washed twice with PBS and lysed with PBS buffer containing 0.5% Triton X-100 and protease mixture inhibitor (Sigma), as described (10). Lysates were centrifuged at 8000 rpm for 5 min, and the supernatants were collected and stored at Ϫ80°C until used. Protein concentrations were determined by Bio-Rad DC protein assay.
Western Blotting-Total cell lysates (40 -50 g) were separated by SDS-PAGE on a 10% gel and electrophoretically transferred to nitrocellulose membrane. Membranes were blocked with 5% bovine serum albumin in TBS plus 0.1% Tween 20 (TBST) or 5% nonfat dry milk in PBS plus 0.1% Tween 20 (PBST) buffer for 2 h at room temperature. Membranes were washed with TBST/PBST three times and then incubated with primary antibody overnight at 4°C or for 1 h at room temperature. Anti-phospho-Syk antibody was from New England Biolabs and used at 1:2500 dilution. Total ERK2 antibody was from Santa Cruz Biotechnology and used at a dilution of 1:5000. The membranes were then washed with TBST/PBST again and incubated with a 1:2500 -5000 dilution of goat anti-rabbit secondary antibody IgG from Santa Cruz Biotechnology for 1 h at room temperature. After washing with TBST/PBST, the bands were developed for visualization using ECL-plus Western blotting detection system (GE Healthcare).
Gels were quantified using the UN-SCAN-IT software package (Silk Scientific). Total ERK2 is widely used as a control for gel loading (29 -31). The antibody does not discriminate between phosphorylated (and hence active) and non-phosphorylated ERK2 and therefore detects the total amount of this protein, regardless of whether the kinase has been activated. The extent of Syk phosphorylation was therefore normalized to the total amount of ERK2 present in each lysate, to correct for differences in amount of cells used for each condition (10).
Transfection and RNAi-Cells were transfected with RNAi directed against ORAi1 using the Amaxa system, as previously described (27). All siRNA sequences were designed by using Invitrogen block-it software. The sequence for Orai1 (5Ј to 3Ј) was GUCCACAACCUCAACUCCTT and was from Invitrogen. Control cells were transfected under identical conditions STRUCTURE 1 Store-operated Ca 2؉ Influx Sustains Syk Activity NOVEMBER 14, 2008 • VOLUME 283 • NUMBER 46 with scrambled siRNA (Invitrogen) or enhance green fluorescent protein.
Statistical Analysis-Results are presented as means Ϯ S.E. Statistical significance was assessed using Student's t test and considered significant at p Ͻ 0.05 (*) or p Ͻ 0.01 (**), respectively.

Antigen-evoked Responses Depend on Functional Syk-Phos-
phorylation and thus activation of Syk is a central early event following FC⑀RI receptor engagement (4). By regulating phospholipase C␥, Syk drives the generation of cytoplasmic Ca 2ϩ signals in response to antigen stimulation. Consistent with this, pre-treatment for 10 min with the Syk inhibitor 3-(1-methyl-1H-indoyl-3-yl-methylene)-2-oxo-2,3-dihydro-1H-indole-5-sulfonamide prevented the generation of the antigen-evoked cytoplasmic Ca 2ϩ rise in IgE-sensitized RBL cells (Fig. 1A). Acute application of the Syk inhibitor after the antigen-activated Ca 2ϩ signal had developed also resulted in suppression of the Ca 2ϩ signal (Fig. 1B). Hence Syk activity is critical for both the generation as well as subsequent maintenance of the Ca 2ϩ signal following activation of FC⑀RI receptors.
Ca 2ϩ Influx Enhances Syk Activity-The non-receptor tyrosine kinase Syk is activated following phosphorylation of tyrosine residues within the activation loop and on tyrosine 342 (32,33). We monitored Syk activity by using a monoclonal antibody that specifically recognized phosphorylated tyrosine residues in the active loop. Stimulation with antigen for 4 min in the presence of external Ca 2ϩ resulted in robust Syk phosphorylation (upper panel of Fig. 1C; lower panel depicts total (combination of active and non-active) ERK2, which is taken as a control for gel loading; aggregate data from four independent experiments is summarized in Fig. 1D). However, when cells were stimulated with antigen for the same time but now in the absence of external Ca 2ϩ , significantly less (ϳ2fold) Syk phosphorylation occurred (Fig. 1, C and D). Because Western blotting involves a large population of cells, we were concerned that fewer cells in the population might respond to antigen in Ca 2ϩ -free solution than in the presence of Ca 2ϩ , thus accounting for the difference in Syk activation. However, the number of cells responding to either condition was similar (data not shown; see also Fig. 3).
We considered the possibility that Syk phosphorylation induced by antigen in the absence of external Ca 2ϩ involved release of Ca 2ϩ from internal stores. We exposed cells to thapsigargin (2 M) in Ca 2ϩ -free solution to deplete the stores and then applied antigen, still in the absence of external Ca 2ϩ . Antigen now failed to trigger a Ca 2ϩ response (data not shown). Ca 2ϩ release to thapsigargin failed to activate Syk (data not shown and see also Fig. 6) but application of antigen after thapsigargin still resulted in Syk phosphorylation and to an extent similar to that seen when cells were stimulated with antigen in Ca 2ϩ -free solution but in the absence of thapsigargin (Fig. 1E). Hence the activation of Syk in the absence of external Ca 2ϩ is independent of Ca 2ϩ release from the stores.
In mast cells, Ca 2ϩ influx can activate protein kinase C ␣ and ␤ II thereby recruiting the MEK/ERK pathway (10). To see whether Syk activation was secondary to this signaling cascade, we directly stimulated protein kinase C with the phorbol ester phorbol 12-myristate 13-acetate. No Syk phosphorylation was FIGURE 1. Activity of the tyrosine kinase Syk is required for, and sustained by, Ca 2؉ influx in response to FC⑀RI receptor activation. A, stimulation of Fura-2-loaded RBL cells with antigen (80 g/ml) produced a sustained Ca 2ϩ rise, and this was suppressed by pre-treatment with the Syk inhibitor (10 M for 10 min). B, after stimulation with antigen, the Syk inhibitor rapidly abolished the Ca 2ϩ signal. C, Western blot shows antigen stimulation (4 min) led to robust Syk phosphorylation (denoted Syk-P) in the presence of external Ca 2ϩ (ϩCa 2ϩ ) and this was considerably less when Ca 2ϩ was removed (0Ca 2ϩ ). Control denotes lysate from nonstimulated cells. Lower panel, ERK2 blots from the corresponding lysates used in the upper gel. D, aggregate data from four independent gels is shown. Gels were quantified as described under "Experimental Procedures." E, following store depletion with thapsigargin (2 M for 10 min in Ca 2ϩ -free external solution), antigen (applied in Ca 2ϩ -free solution) still triggered Syk phosphorylation. Gels shown are typical of three independent experiments. F, direct stimulation of protein kinase C with the phorbol ester phorbol 12-myristate 13-acetate (1 M for 10 min) failed to evoke Syk phosphorylation. G, neither 100 M ATP (4 min) nor application of antigen to non-IgE-treated cells led to Syk phosphorylation. detected (Fig. 1F). To confirm specificity for the FC⑀RI pathway, we measured Syk activation following stimulation of endogenous P 2 Y receptors, which couple to phospholipase C␤ to generate InsP 3 . No Syk phosphorylation was detectable (Fig.  1G). Finally, Syk activation required the presence of FC⑀RI receptors, because application of antigen to RBL cells that had not been pre-exposed to IgE failed to evoke Syk phosphorylation (Fig. 1G). Collectively, these results demonstrate that extracellular Ca 2ϩ is required to sustain Syk activity following FceRI receptor stimulation, and this does not involve protein kinase C or subsequent downstream events.
Local Ca 2ϩ Influx Maintains Syk Activity-Although the previous results demonstrate a requirement for extracellular Ca 2ϩ in sustaining Syk activity, they do not distinguish between a specific role for Ca 2ϩ influx or simply the presence of external Ca 2ϩ per se. To distinguish between these possibilities, we reduced the electrical driving force for Ca 2ϩ influx while keeping extracellular Ca 2ϩ constant. Cells were depolarized by blocking K ϩ channels with the combination of external Cs ϩ and TEA ϩ . Syk activation following antigenic stimulation was significantly reduced ( Fig. 2A, aggregate data from three independent experiments is summarized in Fig. 2B), demonstrating that Ca 2ϩ influx is required for sustaining Syk activity.
We asked whether Ca 2ϩ influx activated Syk through a local action or whether it involved a rise in bulk cytoplasm Ca 2ϩ concentration. To slow down the rate of rise of bulk cytoplasmic Ca 2ϩ and to reduce its extent, we increased cytoplasmic Ca 2ϩ buffering by loading cells with the fast Ca 2ϩ chelator BAPTA together with Fura-2. With our loading protocol, sufficient BAPTA enters the cytoplasm to substantially blunt the bulk cytoplasmic Ca 2ϩ rise following Ca 2ϩ influx (27). However, as shown in Fig. 2 (C and D), Syk activation was unaffected. Because BAPTA is a fast Ca 2ϩ chelator, this suggests that Ca 2ϩ influx acts locally, stimulating Syk phosphorylation within a few nanometers of the Ca 2ϩ entry sites. The Ca 2ϩ Influx Pathway Sustaining Syk Activity Is the Store-operated CRAC Channel-We carried out pharmacological and gene knockdown experiments to identify the molecular identity of the Ca 2ϩ influx pathway activated by antigen. CRAC channels are very sensitive to the trivalent cation Gd 3ϩ , with full block occurring in the low micromolar range (34), a finding we have confirmed for CRAC channels in RBL cells (35). We therefore compared the extent of Syk phosphorylation to antigen (applied in the pres-   NOVEMBER 14, 2008 • VOLUME 283 • NUMBER 46 ence of 2 mM Ca 2ϩ ) in the absence and then presence of 2 M Gd 3ϩ . Antigen-triggered Syk phosphorylation was significantly reduced in the presence of Gd 3ϩ (Fig. 3A), and the extent of the remaining Syk phosphorylation was similar to that seen when cells were stimulated with antigen in Ca 2ϩ -free solution (Fig. 1).

Store-operated Ca 2؉ Influx Sustains Syk Activity
We extended the pharmacological strategy by using a novel CRAC channel blocker, called the Synta compound (see "Experimental Procedures"). Store-operated CRAC channels were activated by exposing Fura-2-loaded cells to the sarcoplasmic-endoplasmic reticulum calcium ATPase pump blocker thapsigargin in Ca 2ϩ -free solution. By blocking Ca 2ϩ reuptake, thapsigargin gradually depletes the stores of Ca 2ϩ , thereby opening CRAC channels. Readmission of external Ca 2ϩ results in Ca 2ϩ entry through the CRAC channels, generating a cytoplasmic Ca 2ϩ rise. We measured the rate of rise of cytoplasmic Ca 2ϩ following Ca 2ϩ readmission, because this is a more reliable indicator of CRAC channel activity than the steady-state amplitude. As shown in Fig. 3B, 10 M Synta compound substantially slowed the rate of rise of the cytoplasmic Ca 2ϩ signal after readmission of external Ca 2ϩ , and this amounted to almost 90% block of the CRAC channels (Fig. 3C).
To measure CRAC channel activity directly, we carried out whole cell patch clamp recordings in which we activated the CRAC current (I CRAC ) by dialyzing cells with InsP 3 in 10 mM EGTA. The time course of I CRAC development is shown in Fig. 3D (filled circles) and the current-voltage relationship, taken when I CRAC had reached steady-state, is depicted in Fig. 3E. Pre-treatment for 5 min with 10 M Synta compound abolished I CRAC (Fig. 3, D and E). Similarly, after activation of I CRAC , perfusion with the Synta compound resulted in a loss of the current (data not shown). We constructed a doseinhibition curve for the Synta compound, which is summarized in Fig.  3F. These data could be fitted with a Hill-type equation, yielding a Hill coefficient of 1.1 and an IC 50 of 3 M.
Having established that the Synta compound blocks CRAC channels, we went on to examine the effects of the drug on antigen-evoked Ca 2ϩ influx and Syk activation. The Synta compound suppressed Ca 2ϩ influx evoked by antigen (Fig. 3G) and reduced the extent of Syk phosphorylation (Fig. 3, H and I). Strikingly, the fall in Syk activity (ϳ50%, Fig.  3I) in the presence of Synta compound was similar to that seen when external Ca 2ϩ was removed (Fig.  1D), consistent with suppression of Ca 2ϩ influx by Synta compound. We tested the selectivity of the Synta compound for CRAC channels by examining whether the drug interfered with two other major ion transport mechanisms in the mast cell membrane: inwardly rectifying K ϩ channels and the plasma membrane Ca 2ϩ ATPase pump. The current-voltage relationship and the amplitude of the inwardly rectifying K current was unaffected by 10 M Synta compound (Fig. 4, A and B). Similarly, the rate of Ca 2ϩ removal from the cytoplasm after stimulation with thapsigargin in Ca 2ϩ free solution, which reflects activity of the plasma membrane Ca 2ϩ ATPase pump (36), was similar in the absence and presence of the Synta compound (Fig. 4, C and D).
RNAi Knockdown of Orai1 Reduces Antigen-dependent Syk Activation-To strengthen the pharmacological approach, we used an RNAi knockdown strategy to suppress expression of the protein Orai1, which is central to the CRAC channel pathway. Orai1 is a plasma membrane protein that comprises the pore of the CRAC channel (16, 17, 20 -22). Transfection with siRNA to Orai1 resulted in a reduction in Orai1 mRNA levels, measured using reverse transcription-PCR (Fig. 5A). Knockdown was only partial, because our transfection efficiency was ϳ60%, judged by expression of eYFP. Co-transfection of Orai1 siRNA with an eYFP construct enabled us to identify trans- A, currentvoltage relationships for inwardly rectifying K ϩ current are shown for a control cell and for one exposed to 10 M Synta compound for 10 min. B, aggregate data for control cells (n ϭ 5) and for those exposed to the Synta compound (n ϭ 6) are compared. Current amplitudes were measured from voltage ramps at Ϫ80 mV. C, plasma membrane Ca 2ϩ ATPase activity was monitored in control cells and in those exposed to the Synta compound by measuring the rate of decline of the cytoplasmic Ca 2ϩ signal following stimulation with thapsigargin in Ca 2ϩ -free solution (with 0.1 mM EGTA). D, aggregate data from experiments as in panel C are compared. For both conditions, number of cells was Ͼ60.
fected cells and thus measure the impact of Orai1 knockdown on store-operated Ca 2ϩ entry. Readmission of external Ca 2ϩ to cells pre-treated with thapsigargin resulted in robust store-operated Ca 2ϩ influx in cells transfected with either eYFP alone or eYFP and scrambled siRNA (Fig. 5B). However, the rate and extent of Ca 2ϩ influx was significantly reduced when cells were transfected with Orai1 siRNA (Fig. 5B, aggregate data are summarized in Fig. 5C). Hence knockdown of Orai1 results in significantly less store-operated Ca 2ϩ entry. Following Orai1 knockdown, antigen activation of Syk was significantly reduced (Fig. 5D, aggregate data are summarized in the lower panel), demonstrating a role for CRAC channels in Syk activation. Because our transfection efficiency was ϳ60%, the fall in Syk phosphorylation following Orai1 knock down was not as extensive as that seen in Ca 2ϩ -free solution (where no Ca 2ϩ entry occurs). Collectively, the pharmacological and RNAi experiments reveal that it is antigenevoked Ca 2ϩ entry through CRAC channels that maintains the activity of Syk.
Synergy between CRAC Channels and FC⑀RI Receptors in Phosphorylating Syk-Although antigen activates I CRAC , it does so to a submaximal extent even when applied at a supra-maximal concentration (37). The size of the current can be increased by further depletion of stores (37). To see whether antigen activation of Syk could be increased by recruiting more CRAC channels, we applied antigen together with thapsigargin. Thapsigargin depletes stores sufficiently to activate I CRAC maximally (38). The combination of antigen and thapsigargin resulted in substantially more Syk phosphorylation than was observed with antigen alone (Fig. 6A, aggregate data are depicted in the lower panel). By blocking sarcoplasmic-endoplasmic reticulum calcium ATPase pumps, thapsigargin not only depletes stores but also reduces cytoplasmic Ca 2ϩ buffering. This would increase the extent and time course of the Ca 2ϩ signal evoked by antigen. To test whether the dramatic potentiating effects of thapsigargin on antigen-evoked Syk phosphorylation were indeed due to the reduction in cytoplasmic Ca 2ϩ buffering, we applied antigen with thapsigargin but now in Ca 2ϩ -free external solution. Under these conditions, the combination of antigen and thapsigargin was only as effective as antigen alone (Fig. 6B); no potentiation occurred. Reduced cytoplasmic Ca 2ϩ clearance therefore cannot explain the significant increase in Syk phosphorylation seen in response to the combination of antigen and thapsigargin. Stimulation with thapsigargin in the presence of external Ca 2ϩ failed to evoke any detectable Syk phosphorylation (Fig. 6C). Hence receptor-independent activation of CRAC channels and subsequent Ca 2ϩ entry cannot mimic the effects of antigen stimulation on Syk phosphorylation. Instead, Ca 2ϩ entry acts synergistically with an early B, store-operated Ca 2ϩ influx is significantly reduced by siRNA to Orai1. Cells were transfected with eYFP alone, eYFP and nonsense siRNA (control), or eYFP and Orai1siRNA. 48 -60 h post-transfection, cells were exposed to thapsigargin in Ca 2ϩ -free solution and then 2 mM Ca 2ϩ was readmitted. The extent of Ca 2ϩ release by thapsigargin was similar for the two conditions. C, aggregate data from several experiments as in panel B are shown. The peak amplitude of the Ca 2ϩ signal was measured. Control denotes 77 cells and Orai1 siRNA 61 cells. A similar reduction in the rate of Ca 2ϩ influx was also observed. D, transfection of cells with RNAi to Orai1 significantly reduced Ca 2ϩ influx-dependent activation of Syk in response to antigen stimulation. The upper gel compares Syk activation to antigen for control transfected cells and for those in which Orai1 had been knocked down. The lower gel is the total ERK2 loading control. The histogram depicts averaged data from four independent experiments. consequence of FC⑀RI receptor cross-linking and this interaction sustains Syk activity.

DISCUSSION
Our findings reveal a novel positive feedback cascade between Syk activity and Ca 2ϩ influx through CRAC channels that sustains mast cell activation. Stimulation of Syk is an early event after antigen stimulation where it contributes to activation of phospholipase C␥ and subsequent Ca 2ϩ release from intracellular stores (4). This fall in store Ca 2ϩ content results in activation of CRAC channels (39). When Syk activity is blocked after the development of a Ca 2ϩ plateau (reflecting store-operated Ca 2ϩ entry) to antigen, the Ca 2ϩ signal falls rapidly. Hence Syk activity is required both to initiate and then sustain Ca 2ϩ entry, presumably by maintaining InsP 3 at levels sufficient to ensure partial store depletion (25). Syk activity is not independent of CRAC channels, however. Local Ca 2ϩ entry through these channels feeds back to sustain Syk activity, providing a mechanism for the prolonged Ca 2ϩ influx seen with FC⑀RI receptor activation and which is needed for appropriate stimulation of mast cells. The finding that Ca 2ϩ influx sustains Syk activity provides a molecular mechanism that helps explain how certain G protein-coupled receptors, which themselves fail to trigger degranulation, can potentiate the antigen-driven responses in mast cells (40).
In some other cells types, cytoplasmic Ca 2ϩ has been reported to activate Syk. In platelets, Syk was activated by the rise in cytoplasmic Ca 2ϩ that accompanied Ca 2ϩ ionophore application (41). In a human B cell line, platelet-activating factor stimulated Syk in a Ca 2ϩ -dependent manner (42). Finally, in PC12 cells, Ca 2ϩ influx led to the rapid activation of the nonreceptor tyrosine kinase PYK2 (43). Although the mechanism whereby cytoplasmic Ca 2ϩ activated these tyrosine kinases was not resolved, addition of Ca 2ϩ to either cell lysates (42) or the isolated kinase (43) did not increase enzyme activity. Hence Ca 2ϩ is unlikely to stimulate Syk directly. Consistent with this is our finding that Ca 2ϩ influx through CRAC channels following stimulation with thapsigargin failed to evoke any detectable Syk activation. Instead, an additional signal associated with FC⑀RI receptors is required. Syk is activated by binding, via SH2 domains, to phosphotyrosine residues in the immunoreceptor tyrosine-based activation motif signaling units on the ␤ and ␥ chains of the receptor, which are phosphorylated by receptorassociated Src kinases like Lyn (4). It is conceivable that Syk binding to the immunoreceptor tyrosine-based activation motif region is increased by Ca 2ϩ , that a phosphorylated immunoreceptor tyrosine-based activation motif region and a Ca 2ϩ rise are both needed to increase Syk activity, or that Lyn activity, once triggered by receptor cross-linking, is potentiated by Ca 2ϩ .
Finally, the fact that Ca 2ϩ influx through CRAC channels but not Ca 2ϩ release from intracellular stores was able to activate Syk adds to the growing list of examples where the spatial location of the Ca 2ϩ rise is important in selectively activating a target (44,45). It is striking that Ca 2ϩ entry is able to support Syk activity even in the presence of the fast Ca 2ϩ chelator BAPTA in the cytoplasm. BAPTA restricts Ca 2ϩ entry to within a few nanometers/tens of nanometers of its point of entry (46), and therefore it is local Ca 2ϩ entry through CRAC channels that maintains Syk. Hence FC⑀RI receptors, Syk, and CRAC channels might be closely apposed, possibly co-localized in a signaling complex. Our functional studies lend support to recent observations that FC⑀RI-dependent signaling occurs in lipid raft domains (47), where store-operated channels can also be found (48).