Inhibitory Modulation of B Cell Receptor-mediated Ca2+ Mobilization by Src Homology 2 Domain-containing Inositol 5′-Phosphatase (SHIP)*

Src homology 2 domain-containing inositol 5′-phosphatase (SHIP) mediates inhibitory signals that attenuate intracellular Ca2+ mobilization in B cells upon B cell receptor (BCR) stimulation. To clarify the mechanisms affected by SHIP, we analyzed Ca2+ mobilization in the DT40 B cell line in which the SHIP gene was disrupted. In SHIP-deficient cells, Ca2+ transient elicited by BCR stimulation was more prolonged than that in control cells both in the presence and absence of extracellular Ca2+. Inositol 1,4,5-trisphosphate production following BCR stimulation was enhanced in SHIP-deficient cells. In SHIP-deficient cells in comparison with the control cells, BCR stimulation in the absence of extracellular Ca2+induced a greater degree of Ca2+ store depletion and the Ca2+ influx upon re-addition of extracellular Ca2+ was also greater. However, store-operated Ca2+ influx (SOC) elicited by thapsigargin-induced store depletion was not affected by SHIP. These results indicate that the primary target pathway of SHIP is the Ca2+ release from the stores, and that Ca2+ influx by the SOC mechanism is secondarily controlled by the level of Ca2+ in the stores without direct inhibition of SOC. In this way, SHIP may play an important role in ensuring the robust tuning of Ca2+signaling in B cells.

Intracellular calcium concentration ([Ca 2ϩ ] i ) 1 regulates cellular functions in various types of cells (1). In B lymphocytes, [Ca 2ϩ ] i controls cell proliferation, differentiation and apoptotic processes (2,3). Cross-linking of B cell receptors (BCR) with specific antigens activates phospholipase C␥ (PLC␥) through a series of tyrosine phosphorylations, resulting in hydrolysis of phosphatidylinositol 4,5-bisphosphate (PIP 2 ). Inositol 1,4,5trisphosphate (IP 3 ), a product of the hydrolysis of PIP 2 , then activates IP 3 receptors to mobilize the intracellular Ca 2ϩ stores. Ca 2ϩ influx from the extracellular space is also activated in response to BCR cross-linking via the store-operated Ca 2ϩ influx (SOC) pathway, which is activated by depletion of the intracellular Ca 2ϩ stores (4,5). BCR activation also results in the activation of phosphoinositide 3-kinase, which converts PIP 2 to phosphatidylinositol 3,4,5-trisphosphate (PIP 3 ).
One of the possible target mechanisms of SHIP in the attenuation of the Ca 2ϩ response is the SOC activity. Since the Fc␥RIIB-mediated inhibition of Ca 2ϩ signaling was more prominent in the presence of extracellular Ca 2ϩ than in its absence (8 -10), SHIP was postulated to attenuate Ca 2ϩ influx via the SOC channel (11)(12)(13)17). Another potential target of SHIP is the PLC␥ activity. Since PIP 3 was reported to stimulate Bruton's tyrosine kinase (Btk), which in turn tyrosine-phosphorylates PLC␥ (18 -20), degradation of PIP 3 upon activation of SHIP by the co-cross-linking of Fc␥RIIB would decrease PLC␥ activity and hence inhibit Ca 2ϩ release from the stores. However, direct demonstration that SHIP affects SOC or Ca 2ϩ release in Fc␥RIIB-mediated signaling remains to be reported.
It has been shown that cross-linking of BCR alone leads to tyrosine phosphorylation of SHIP (21,22). Furthermore, SHIP seems to inhibit the BCR-mediated Ca 2ϩ response even without co-cross-linking of Fc␥RIIB, because SHIP-deficient cells exhibit prolongation of the [Ca 2ϩ ] i transient upon BCR activation (23). While SHIP is recruited to the immunoreceptor tyrosine-based inhibitory motif on the intracellular region of Fc␥RIIB upon activation of this receptor (11), there is no immunoreceptor tyrosine-based inhibitory motif on the BCR complex. Therefore, SHIP is not a unique molecule for Fc␥RIIBmediated inhibitory signaling, but plays an important role in BCR-mediated signaling; the mechanisms of the recruitment of SHIP seem different in the absence or presence of Fc␥RIIBmediated signaling. It remains to be clarified how Ca 2ϩ mobilization is attenuated by SHIP in BCR-mediated signaling.
In an effort to understand the role of SHIP in BCR-mediated signaling, we addressed the following questions in this work: 1) how are the patterns of Ca 2ϩ mobilization elicited by BCR stimulation modulated by SHIP, and 2) which Ca 2ϩ mobilization pathway is affected by SHIP: Ca 2ϩ release or SOC. To clarify these points, we compared the patterns of [Ca 2ϩ ] i mobilization in control and SHIP-deficient DT40 cells at the single cell level. We found that the prolongation of [Ca 2ϩ ] i transient in SHIP-deficient cells is due to enhanced Ca 2ϩ release from the intracellular Ca 2ϩ stores with enhancement of IP 3 production. We also found that SHIP regulates SOC not by a direct interaction but through its effect on Ca 2ϩ release. SHIP thus plays an important role in shaping the Ca 2ϩ mobilization patterns after BCR stimulation.
Measurement of Intracellular Ca 2ϩ Concentration-About 30 min before the experiments, cells were attached to collagen-coated coverslips. Cells on the coverslips were incubated with 5 M Fura-2 AM for 30 min at room temperature in physiological salt solution (PSS) (150 mM NaCl, 4 mM KCl, 2 mM CaCl 2 , 1 mM MgCl 2 , 5 mM HEPES, 5.6 mM glucose; pH adjusted to 7.4 with NaOH) containing 0.1% bovine serum albumin followed by rinsing with PSS. The coverslips with Fura-2loaded cells were mounted on the stage of an inverted epifluorescence microscope (TMD 300, Nikon). Cells were examined under a 40ϫ water immersion objective (numerical aperture: 0.7, Olympus). Pairs of fluorescence images with 340 and 380 nm excitations were collected using a cooled CCD camera (PXL-37, Photometrics,) at either 0.5 or 0.33 Hz. The ratio (R) between the fluorescence intensities at 340 and 380 nm excitations was converted to [Ca 2ϩ ] i using the following equation (24); where K d Ј, R max , and R min are the dissociation constant and maximal and minimal R values, respectively. The values were determined in vitro under equivalent optical conditions. R max and R min were multiplied by a factor of 0.85 for viscosity correction (25). For the stimulation of the BCR, 2 g/ml mouse antichicken IgM antibody, M4 (26), was applied onto the cells through a thin pipette. Other solutions were applied in the same manner. The Ca 2ϩ free solution had the same composition as PSS except for the omission of CaCl 2 and introduction of EGTA (5 mM).
Immunoblotting-Cells were harvested and washed with phosphate buffered saline. The cells were precipitated in 10% trichloroacetic acid (2 ϫ 10 7 cells/ml). The precipitates were dissolved in solution containing 100 mM Tris-HCl (pH 8.0), 30 mM NaCl, and 1% SDS. The samples were separated by SDS-polyacrylamide gel electrophoresis and were electrotransferred to polyvinylidene difluoride membranes (Trans-Blot, Bio-Rad). The membranes were incubated with the primary polyclonal antibody (anti-mouse SHIP, Upstate Biotechnology, Inc.) and then with biotin-labeled secondary antibody. The blots were detected with enhanced chemiluminescence (Renaissance, NEN Life Science Products).
IP 3 Measurement-Cells were harvested and washed with PSS. The BCR-stimulated cell suspension (2 ϫ 10 6 cells) was mixed with 15% trichloroacetic acid for termination of the reaction. After centrifugation, the supernatant was extracted with water-saturated diethylether and the water phase was neutralized with NaHCO 3 to pH 7.5. D-myo-[ 3 H]Inositol-1,4,5-trisphosphate assay system (TRK1000, Amersham Pharmacia Biotech) was used for measurement of IP 3 levels (27). Radioactivity was measured for 3 min in a ␤-scintillation counter, and converted to IP 3 concentration using the calibration curve.
Statistical Analysis-Statistical results are expressed as mean Ϯ S.E. Statistical comparisons were made using the paired t test for the IP 3 assay and the non-paired t test for all the other measurements.
Materials-RPMI 1640 medium, glutamine, penicillin, and streptomycin were purchased from Life Technologies, Inc. Chicken serum, ionomycin, and thapsigargin were obtained from Sigma. Fura-2 AM was purchased from Molecular Probes. All other chemicals were of the highest reagent grade.  (Fig. 1A). In contrast, SHIP-deficient cells showed a plateau-like response or oscillations persisting for over 200 s (Fig. 1B). Quantitative analysis of [Ca 2ϩ ] i showed that the peak level of [Ca 2ϩ ] i increase was slightly greater in SHIPdeficient cells (675 Ϯ 29 nM, n ϭ 140) than in control cells (556 Ϯ 23 nM, n ϭ 140; p Ͻ 0.002). The difference was more conspicuous in the late phase, and at 150 s after the BCR stimulation, [Ca 2ϩ ] i remained at a higher level in SHIP-deficient cells (350 Ϯ 31 nM, n ϭ 70) than in control cells (110 Ϯ 14 nM, n ϭ 70; p Ͻ 0.0001) (Fig. 1,  A and B; Fig. 2A).

Prolonged Ca 2ϩ Mobilization in SHIP-deficient DT40 Cells-We
To confirm that the observed prolongation of Ca 2ϩ mobilization in SHIP-deficient cells was due to the absence of SHIP, we transfected the SHIP-deficient cells with mouse SHIP or mutant SHIP lacking phosphatase activity. The expression of wild-type SHIP activity shortened the duration of [Ca 2ϩ ] i increase (Fig. 1C), whereas transfection of SHIP lacking phosphatase activity was without effect (Fig. 1D). The levels of expression of the wild-type and mutant SHIP were almost the same (Fig. 1E). These results indicate that the shortening of the Ca 2ϩ mobilization pattern by SHIP required its phosphatase activity.
Effect of SHIP on Ca 2ϩ Release from Ca 2ϩ Stores-There exist two main pathways for Ca 2ϩ mobilization: Ca 2ϩ release from the Ca 2ϩ stores and Ca 2ϩ entry from the extracellular space. Both pathways are potential targets of SHIP. To determine the role of SHIP in the Ca 2ϩ release pathway, we analyzed the time course of Ca 2ϩ mobilization by BCR stimulation in the presence and absence of extracellular Ca 2ϩ . The prolongation of Ca 2ϩ mobilization in SHIP-deficient cells in the pres- ence of extracellular Ca 2ϩ ( Fig. 2A) was retained in the absence of extracellular Ca 2ϩ (Fig. 2B). Although the peak [Ca 2ϩ ] i rise in control cells and SHIP-deficient cells in the absence of extracellular Ca 2ϩ showed no significant difference (control: 520 Ϯ 34 nM, n ϭ 120; SHIP-deficient: 523 Ϯ 25 nM, n ϭ 105; p Ͼ 0.4), the [Ca 2ϩ ] i at 150 s after BCR stimulation was greater in SHIP-deficient cells (182 Ϯ 14 nM, n ϭ 60) than in control cells (80 Ϯ 9 nM, n ϭ 51; p Ͻ 0.0001). These results indicate that Ca 2ϩ release from the Ca 2ϩ stores is one of the target pathways of SHIP.
Effect of SHIP on IP 3 Production-We then compared the time course of IP 3 production during BCR stimulation in control and SHIP-deficient cells (Fig. 3). The increase in IP 3 concentration was significantly higher in SHIP-deficient cells at 180 s after BCR stimulation (p Ͻ 0.05, n ϭ 5) than in control cells. These results suggest that SHIP regulates Ca 2ϩ release through the attenuation of IP 3 production.
Ca 2ϩ Influx after BCR Stimulation-We then evaluated the effect of SHIP on the Ca 2ϩ influx mechanism. First, Ca 2ϩ release from the Ca 2ϩ stores was activated by BCR stimulation in the absence of extracellular Ca 2ϩ for 600 s, during which period [Ca 2ϩ ] i returned to the resting value. Reintroduction of extracellular Ca 2ϩ resulted in an increase in [Ca 2ϩ ] i due to influx of Ca 2ϩ in both control and SHIP-deficient cells (Fig. 4,  A and B, solid bar). The maximal level of [Ca 2ϩ ] i reached during the extracellular application of Ca 2ϩ was higher in SHIP-deficient cells than in control cells (Fig. 4C). These results indicate that Ca 2ϩ influx as well as Ca 2ϩ release is attenuated by SHIP.
Effect of SHIP on SOC Elicited by Thapsigargin-Since we found that BCR-induced Ca 2ϩ influx is inhibited by SHIP, we tested whether SHIP has a direct effect on the activity of the SOC mechanism that is activated by the depletion of Ca 2ϩ stores. We used a sarco(endo)plasmic reticulum Ca 2ϩ -ATPase (SERCA) inhibitor, thapsigargin to deplete the Ca 2ϩ stores, and then assayed the extent of SOC after reintroduction of extracellular Ca 2ϩ (Fig. 5, A and B) (28). The increase in [Ca 2ϩ ] i in control cells was not statistically different from that in SHIP-deficient cells (Fig. 5C). The result suggests that the SOC is not affected by SHIP, at least under the conditions where BCR is not cross-linked.
Then, we repeated the same experiment under BCR stimulation, by which SHIP is expected to be activated. After a 900-s application of thapsigargin to deplete the Ca 2ϩ stores, BCR was stimulated (arrows in Fig. 5, D and E), and then extracellular Ca 2ϩ was reintroduced. BCR stimulation elicited no [Ca 2ϩ ] i rise after thapsigargin treatment, confirming the complete depletion of the stores after thapsigargin treatment. SOC during BCR stimulation was not affected by SHIP, and the amplitude of the [Ca 2ϩ ] i rise due to SOC was the same in control and SHIP-deficient cells (Fig. 5F). These results indicate that the SOC mechanism is not a direct target of SHIP.
Content of Ca 2ϩ in Ca 2ϩ Stores after BCR Stimulation-The extent of the Ca 2ϩ store depletion has been postulated to control the extent of activation of SOC. Therefore, the difference in the activation of SOC after BCR stimulation in control and SHIP-deficient cells (Fig. 4) may reflect a possible difference in the extent of Ca 2ϩ depletion. We tested this hypothesis by estimating the amount of Ca 2ϩ in the Ca 2ϩ stores by the application of ionomycin in the absence of extracellular Ca 2ϩ . Ionomycin treatment after BCR stimulation elicited a greater [Ca 2ϩ ] i increase in control cells than in SHIP-deficient cells (Fig. 6, A-C). However, the peak size of [Ca 2ϩ ] i rise elicited by ionomycin treatment without BCR stimulation in control and in SHIP-deficient cells did not show any significant differences (Fig. 6, D-F), indicating that the initial Ca 2ϩ content within the Ca 2ϩ stores was the same in control and SHIP-deficient cells.
These results indicate that BCR stimulation resulted in only partial Ca 2ϩ release from the stores in control cells and nearly complete release of Ca 2ϩ in SHIP-deficient cells, and suggest that the enhancement of Ca 2ϩ influx in SHIP-deficient cells was due to secondary activation of SOC by more profound Ca 2ϩ depletion.

Effect of SHIP on Ca 2ϩ
Extrusion Mechanism-We finally tested whether SHIP affects the Ca 2ϩ extrusion mechanism via the plasma membrane, which potentially controls [Ca 2ϩ ] i in concert with the Ca 2ϩ mobilization mechanism. We estimated the level of activity of the Ca 2ϩ extrusion mechanism from the initial rate of decline of [Ca 2ϩ ] i after termination of SOC. The rate was obtained from the slope of the line, which was fitted to the data points of [Ca 2ϩ ] i for 2 min after removal of extracellular Ca 2ϩ (Fig. 7, inset), and was plotted against the [Ca 2ϩ ] i just before the removal of extracellular Ca 2ϩ (Fig. 7). The plots from control and SHIP-deficient cells were superimposable, indicating no difference in the level of activity of the Ca 2ϩ extrusion mechanism between control and SHIP-deficient cells. DISCUSSION We examined the Ca 2ϩ mobilization patterns in SHIP-deficient B cells at the single cell level and showed that the main role of SHIP in BCR-mediated Ca 2ϩ signaling was to abbreviate the duration of Ca 2ϩ mobilization. SHIP prevented continuous Ca 2ϩ release from the intracellular Ca 2ϩ stores via inhibition of IP 3 production. Furthermore, we showed that SHIP regulates SOC not by a direct effect on the SOC mechanism but by controlling the depletion level of the Ca 2ϩ stores. This inhibitory cascade of SHIP is shared by the Fc␥RIIB pathway despite the difference in the way for recruitment of SHIP.
In B lymphocytes, accumulation of PIP 3 via the activation of phosphoinositide 3-kinase was reported to stimulate Btk, which then phosphorylates PLC␥ (18 -20). In other studies, accumulation of PIP 3 was reported to lead to PH domainmediated membrane targeting of PLC␥ (29,30). In either case, PIP 3 functions as a potent activator of PLC␥. Therefore, degradation of PIP 3 by SHIP is expected to inhibit PLC␥ and hence Ca 2ϩ release from the Ca 2ϩ stores. We now show direct evidence that IP 3 production in BCR-mediated signaling is en-hanced in SHIP-deficient cells (Fig. 3).
In Fc␥RIIB-mediated inhibitory signaling, SHIP was implicated not only in IP 3 production but also in the inhibition of SOC, or store depletion-induced Ca 2ϩ influx (11,13). We, therefore, examined the effects of SHIP on Ca 2ϩ influx elicited after BCR activation, and found that Ca 2ϩ influx in SHIP-deficient cells was indeed greater than that in control cells (Fig. 4). SOC has been postulated to be a mechanism of Ca 2ϩ influx in B lymphocytes, as is the case in other nonexcitable cells (4,31). Thus, we also examined the effect of SHIP on SOC after store depletion by thapsigargin, which is a potent inhibitor of SERCA and depletes the Ca 2ϩ stores without IP 3 production (28). However, thapsigargin-induced SOC was not affected by the presence of SHIP either with or without BCR cross-linking (Fig. 5). These results clearly indicate that SHIP does not have a direct inhibitory effect on either SOC channel activity or the SOC activation mechanism itself in BCR-mediated signaling. Then, how did SHIP inhibit the Ca 2ϩ influx after BCR stimulation? The activation of SOC was shown to be regulated by the level of depletion the Ca 2ϩ stores (32)(33)(34). We therefore postulated that the effect of SHIP on Ca 2ϩ influx resulted from the difference in the extent of Ca 2ϩ store depletion in control and SHIP-deficient cells after BCR cross-linking. To test this hy- pothesis, we evaluated the amount of Ca 2ϩ remaining in the Ca 2ϩ stores after BCR cross-linking. In SHIP-deficient cells, the amount of Ca 2ϩ remaining in the stores was much smaller than that in control cells (Fig. 6). The results indicate that SHIP causes early termination of Ca 2ϩ release making SOC activation minimal in normal BCR signaling. On the other hand, the Ca 2ϩ stores were likely to be depleted enough to activate SOC in thapsigargin-treated cells and in BCR-stimulated SHIP-deficient cells. A similar inhibitory cascade after Fc␥RIIB co-cross-linking was shown recently (20).
The differential effects of SHIP on Ca 2ϩ influx elicited by BCR cross-linking and SOC elicited by thapsigargin could also be accounted for by postulating two independent Ca 2ϩ influx pathways: the BCR-regulated and SHIP-sensitive influx, and the thapsigargin-elicited and SHIP-insensitive pathways. However, this possibility can be readily ruled out. If this were the case, cells treated with both BCR cross-linking and thapsigargin would allow Ca 2ϩ influx via both components. The total Ca 2ϩ influx should also be affected by SHIP as a result from the effect on the SHIP-sensitive Ca 2ϩ influx. However, reintroduction of Ca 2ϩ elicited the same extent of Ca 2ϩ influx in control and SHIP-deficient cells (Fig. 5, D and E), arguing against the hypothesis. This conclusion was supported by the results obtained from the DT40 cells in which all three subtype genes of the IP 3 receptor were disrupted. In these IP 3 receptordeficient cells, there was no Ca 2ϩ response to BCR stimulation, although thapsigargin-elicited Ca 2ϩ increase was clearly observed (31). These results suggest the absence of a Ca 2ϩ influx pathway other than the SOC in BCR signaling in DT40 cells.
SHIP has been found in a variety of hematopoietic cells other than B cells and may play a role in cell signaling in these cells (21,35). For example, SHIP may be involved in thrombininduced platelet activation (36,37). In Fc⑀RI-stimulated RBL-2H3 cells, SHIP was reported to be phosphorylated and recruited to the immunoreceptor tyrosine-based activation motif of Fc⑀RI (38,39). However, the role of SHIP in these cells is unclear. Therefore, it will be important to examine whether SHIP plays the same inhibitory role in Ca 2ϩ signaling in different cell types as shown here in B cells.
Accumulating evidence suggests that the cellular responses are more often controlled by the temporal pattern of Ca 2ϩ mobilization than by the peak or average levels of [Ca 2ϩ ] i (1, 34, 40 -42). Although the mechanism underlying the Ca 2ϩ mobilization pattern-mediated control of cell function is not fully understood, the temporal pattern of [Ca 2ϩ ] i increase has been implicated in the differential activation of subsets of proteins and/or genes due to different Ca 2ϩ -mediated activation kinetics. Our present finding indicates that the Ca 2ϩ mobilization is curtailed by the inhibition of IP 3 production by SHIP activity during BCR stimulation. Because the extent of activation of SOC steeply depends on the extent of Ca 2ϩ store depletion (32-34), this control mechanism may provide the cells with sharp on-and off-regulation of Ca 2ϩ influx. The initial rate of Ca 2ϩ extrusion was estimated from the results of experiments shown in Fig. 5 (D and E) as the slope of the dotted line that was fitted to the data points for 120-s period after the removal of extracellular Ca 2ϩ (inset).