Activation Mechanism for CRAC Current and Store-operated Ca2+ Entry

Here we tested the role of calcium influx factor (CIF) and calcium-independent phospholipase A2 (iPLA2) in activation of Ca2+ release-activated Ca2+ (CRAC) channels and store-operated Ca2+ entry in rat basophilic leukemia (RBL-2H3) cells. We demonstrate that 1) endogenous CIF production may be triggered by Ca2+ release (net loss) as well as by simple buffering of free Ca2+ within the stores, 2) a specific 82-kDa variant of iPLA2β and its corresponding activity are present in membrane fraction of RBL cells, 3) exogenous CIF (extracted from other species) mimics the effects of endogenous CIF and activates iPLA2β when applied to cell homogenates but not intact cells, 4) activation of ICRAC can be triggered in resting RBL cells by dialysis with exogenous CIF, 5) molecular or functional inhibition of iPLA2β prevents activation of ICRAC, which could be rescued by cell dialysis with a human recombinant iPLA2β, 6) dependence of ICRAC on intracellular pH strictly follows pH dependence of iPLA2β activity, and 7) (S)-BEL, a chiral enantiomer of suicidal substrate specific for iPLA2β, could be effectively used for pharmacological inhibition of ICRAC and store-operated Ca2+ entry. These findings validate and significantly advance our understanding of the CIF-iPLA2-dependent mechanism of activation of ICRAC and store-operated Ca2+ entry.

Store-operated channels (SOC) 2 and Ca 2ϩ entry (SOCE) are triggered by depletion of intracellular Ca 2ϩ stores in a wide variety of cell types (for review, see Ref. 1). Recently we discovered Ca 2ϩ -independent phospholipase A 2 (iPLA 2 ) to be a cru-cial determinant of SOCE (2, 3) and a physiological target for a mysterious Ca 2ϩ influx factor (CIF) that is produced by the endoplasmic reticulum (ER) upon Ca 2ϩ store depletion. We proposed a mechanism (2) that mediates not only activation but also termination of SOCE. We demonstrated that CIF displaces inhibitory calmodulin (CaM) from iPLA 2 leading to its activation and production of lysophospholipids that in turn activate SOC and SOCE in vascular smooth muscle cells. Upon refilling of the stores and termination of CIF production, CaM binds to iPLA 2 and inhibits it, and the activity of SOC and SOCE is halted. Consistent with a major role of iPLA 2 , a growing number of studies demonstrated that its irreversible inhibition impairs SOCE in different cell types (2)(3)(4)(5)(6)(7)(8)(9). However, no detailed study of CIF-iPLA 2 -dependent mechanism has been done in the cells in which SOCE is mediated by a highly Ca 2ϩselective SOC (historically called CRAC) (10). The whole-cell current through CRAC channels (I CRAC ) remains a golden standard for the studies of SOC and mechanisms of its regulation (for review, see Ref. 1). Our present studies were designed to test the validity and functionality of the CIF-iPLA 2 -dependent pathway in CRAC channel activation, to obtain new important information related to the signal that triggers CIF production, the specific isoform of iPLA 2 that may be involved as well as the functional regulation of I CRAC through regulation of iPLA 2 in rat basophilic leukemia (RBL-2H3) cells, one of the well established models for CRAC channel studies.
CIF is known to be produced in the cells when their Ca 2ϩ stores are depleted after either active Ca 2ϩ release (usually as a result of agonist-induced Ca 2ϩ discharge from the stores) or as a result of passive Ca 2ϩ loss (when Ca 2ϩ back-sequestration into the stores is prevented). Although the molecular identity of CIF remains a mystery, its presence and biological activity was detected in numerous cell types, from yeast to humans (11)(12)(13). For a complete review of the present knowledge related to CIF we will refer the readers to specialized reviews (1,14). Here we will concentrate on some new questions, such as what is the physiological signal that triggers CIF production, how ubiquitous is production of CIF, and how interchangeable is its biological activity between different species.
It is generally thought that activation of SOCE is triggered by a net Ca 2ϩ loss from ER (for review, see Refs. 1 and 15). However, another attractive possibility exists that a simple reduction in free Ca 2ϩ concentration within the stores (without physical Ca 2ϩ loss) may be enough for triggering SOCE in a dose-dependent manner (16). It is presently unknown what kind of changes in intraluminal Ca 2ϩ may be sufficient for triggering CIF production; does it necessarily need to be a Ca 2ϩ loss (as widely believed), or could it be a simple reduction in free Ca 2ϩ concentration in the stores. If the later is true, CIF production and SOCE activation may be a natural response of the cell not only to agonist-induced discharge of Ca 2ϩ from the stores but also to variations in free Ca 2ϩ in the stores due to pathological changes in the expression levels of intraluminal Ca 2ϩ binding proteins, such as calsequestrin and calreticulin (for review, see Refs. 17 and 18). Here we tested if simple Ca 2ϩ buffering in the stores may or may not trigger CIF production and how it may relate to CIF produced upon physical Ca 2ϩ loss. We also tested if CIF produced by different types of cells and species could be interchangeable and if exogenously applied CIF can mimic the effects of endogenous CIF in activating iPLA 2 and SOC.
It is now well established that iPLA 2 (group VI encoded by PLA2G6) is not a single, but a growing family of enzymes (19 -21). Two major isoforms of iPLA 2 have been identified and intensively studied (for reviews, see Ref. 20 and 22): iPLA 2 ␤ (group VIA) and iPLA 2 ␥ (group VIB). The signature feature of iPLA 2 ␤ is that it contains CaM binding domain, making it possible for CaM to bind and inhibit its activity (23,24). This unique feature discriminates iPLA 2 ␤ from other iPLA 2 s and provides a perfect design for its specific role in the SOCE pathway (2). Another prominent feature of some variants of iPLA 2 ␤ is multiple ankyrin repeats in the N terminus, which may be important for their cellular localization and assembly into specific signaling domains. In vascular smooth muscle cells we found that iPLA 2 ␤ remains functional at the inner leaflet of excised plasma membrane patches (2), which was consistent with its plasma membrane localization reported in some cell types (25,26). However, in numerous studies iPLA 2 ␤ was found in cytosol (27)(28)(29), the inner membrane of mitochondria (30), and nucleus (25). There is also some evidence for its possible translocation to perinuclear and plasma membranes (25,31,32). The existence of multiple splice variants of iPLA 2 ␤ isoform (33) as well as their post-translational modifications (34) may be some of the reasons for inconsistence in iPLA 2 ␤ localization. Further studies are necessary to determine which splice variant of iPLA 2 ␤ is required and sufficient for activation of plasma membrane channels and SOCE.
Here we test the main steps in the CIF-mediated iPLA 2 -dependent SOCE pathway and unveil new important details in CIF production, iPLA 2 specificity, and CIF-iPLA 2 -dependent mechanism of activation of I CRAC and capacitative Ca 2ϩ entry in RBL-2H3 cells.

Cells
Rat basophilic leukemia RBL-2H3 cells were obtained from ATCC and maintained in minimum essential medium supplemented with 2 mM L-glutamine and 1% penicillin/streptomycin (10,000 IU/10,000 g/ml). Cells were passed every 3 days at a ratio of 1:5 and used for up to 12 passages. For Ca 2ϩ imaging and patch clamp experiments, RBL cells were grown on small coverslips (ϳ5 ϫ 5 mm) placed into 6-well plates. For CIF preparations, Western blots, and iPLA 2 activity measurements, RBL cells were grown in 10-cm tissue culture dishes.

Drugs and Treatments
Bromoenol lactone (BEL), thapsigargin (TG), N,N,NЈ,NЈ-tetrakis (2-pyridylmethyl)ethylenediamine (TPEN), and most other drugs were purchased from Sigma. Fura-2 and fura-2 AM were from Invitrogen. Anti-iPLA 2 ␤, a polyclonal antiserum against a 19-amino acid peptide (NQIHSKDPRYGASPLH-WAK) specific to the ankyrin region of iPLA 2 ␤ (24), was a generous gift of Dr. R. W. Gross. Human recombinant iPLA 2 ␤ (rec-iPLA 2 ␤) was expressed in and purified from Sf9 cells as previously described by Dr. Gross (21,35). Briefly, Sf9 cells were infected with baculovirus harboring the human recombinant iPLA 2 ␤His 6 . After lysis of the cells, iPLA 2 ␤ was purified from the cytosol using Co 2ϩ affinity chromatography followed by ATP-agarose chromatography. Column fractions in each step were assayed for iPLA 2 activity as described below. The final rec-iPLA 2 ␤ concentration was typically 0.15-0.18 g/l. Chiral enantiomers of BEL ((S)-and (R)-BEL)) were separated by high performance liquid chromatography (HPLC) utilizing a Chirex column of 3,5-dinitrobenzoyl-(R)-phenylglycine attached to a silica matrix (Phenomenex) as previously described (22). It is important to notice that BEL and its enantiomers are suicidal substrates for iPLA 2 ; inhibition is irreversible, requires basal activity of this enzyme, and strongly depends on temperature, duration of treatment, and concentration used (3). In experiments with intact cells the optimal conditions for BEL treatment (to ensure complete inhibition of iPLA 2 ) are the following; intact cells need to be pretreated (in bath solution not containing bovine serum albumin or serum) with 10 -25 M BEL for 30 min at 37°C, and then BEL can be washed away before the beginning of the experiments. In cases where iPLA 2 is already active, significantly shorter (1-5 min) treatments may fully inhibit the enzyme.

Preparation of CIF Extracts
CIF extracts were purified from human platelets and from RBL-2H3 cells.
Human Platelets-Crude CIF extracts were obtained as described before (13). Briefly, resting (unstimulated) platelets (obtained from the local Red Cross) were kept at room temperature and used immediately for preparation of the control CIF extracts that contained no or very little CIF activity (control extract). Extracts containing high CIF activity (CIF) were prepared from platelets with depleted Ca 2ϩ stores, which was achieved by exposing them to cold (4°C) overnight, with subsequent application of TG (2 M for 20 min). To prepare crude extracts, platelets (50 ml per platelet pack, about 10 11 cells) were washed in Hanks' balanced salt solution supplemented with 20 mM HEPES (20 ml) and resuspended in 0.85 ml of the same solution. The suspension was extracted with 0.2 ml of 1 M hydrochloric acid for 30 min at room temperature. After centrifugation, the supernatant was neutralized (10 M NaOH), and BaCl 2 (10 mM) was added to precipitate compounds containing vicinal phosphates, including inositol 1,4,5-trisphosphate. After centrifugation the supernatant was lyophilized, and the residue was extracted with methanol (0.8 ml) with continuous mixing for 15 min. The methanol extract was loaded on a Sep-Pak Vac C18 cartridge (Waters), and the cartridge was washed with methanol (0.8 ml). The combined methanol elutes were dried at 30°C under N 2 gas and resuspended in 200 l of 100 mM acetic acid. The reconstituted extract was clarified by centrifugal ultrafiltration through a Ultrafree-MC 30-kDa filter (Millipore). To obtain fine extracts, the crude extracts were further subjected to anion exchange HPLC followed by reversed-phase HPLC. Unless specified, all the experiments in this study were done with crude CIF extracts.
RBL Cells-RBL cells grown in 10-cm culture dishes were washed with serum-free media and then treated with either 1 M TG or 1 mM TPEN for 5 min at 37°C to initiate CIF production. CIF was prepared from the cells exactly as described above for human platelets; however, the final dried methanol elutes were reconstituted in 50 l of 100 mM acetic acid instead of the 200 l used for platelets.
CIF activities were assayed by microinjection of extracts into fura-2-loaded albino oocytes (12). Oocytes were transferred to Ca 2ϩ -free oocyte Ringer (OR-2) buffer solution (82 mM NaCl, 5 mM HEPES, 5 mM MgCl 2 , 2 mM KCl (pH 7.5)) and microinjected with 14 nl of fura-2 free acid (1 mM in 125 mM KCl) using a Nanoject-II injector (Drummond). Injected oocytes were allowed at least 2 h of recovery and kept at 4°C until used for the bioassay. Next, an individual oocyte was transferred into a homemade imaging chamber containing Ca 2ϩ -free OR-2 solution, mounted on the stage of a Nikon Eclipse TS-100 inverted microscope. Changes in intracellular free Ca 2ϩ were measured through a Nikon 20ϫ Super Fluor objective (NA ϭ 0.75) using a rapid excitation filter changer alternating between 340 and 380 nm (Sutter Instruments) and a CCD camera (Cooke Pix-elFly) and analyzed using the InCyt IM2 software (Intracellular Imaging). A micropipette containing CIF extract was inserted into the oocyte around the equatorial plane before the beginning of the recording. After obtaining a stable recording base line, CaCl 2 (5 mM) was added to the extracellular bath solution to first test the oocytes for nonspecific Ca 2ϩ leak and to verify that the membrane seal around the pipette was tight and Ca 2ϩ was not leaking in. About 1 min later CIF extract (28 nl) was injected to trigger Ca 2ϩ influx across the oocyte membrane. CIF activity was determined based on the changes in intracellular Ca 2ϩ (expressed as ⌬Ratio), which was calculated as the difference between the peak F 340 /F 380 ratio (6 min after injection), and its basal level right before injection.

Molecular Inhibition of iPLA 2 ␤
RBL cells were transfected using Nucleofector II (Amaxa Biosystems). Batches of 1.5 ϫ 10 6 cells were resuspended in 100 l of Nucleofector solution R at room temperature followed by the addition of 2 g of antisense or sense DNA together with 2 g of green fluorescent protein. A 20-base-long antisense (5Јfluorescein-CTCCTTCACCCGGAATGGGT-3Ј) and sense (5Ј-fluorescein-ACCCATTCCGGGTGAAGGAG-3Ј) specific to iPLA 2 ␤ was used as in our previous studies (3). Transfection was done at the T-20 setting of the Nucleofector II device. Immediately after transfection, the RBL cells were transferred to minimum essential medium and plated on the coverslips. Cells showing green protein fluorescence (excitation at 480 nm, emission at 515 nm) were used for experiments 38 Ϯ 4 h after transfection.

Western Blots
RBL cells were homogenized on ice by sonication in the absence or presence of 1% Triton X-100 in the homogenization buffer containing 300 mM sucrose and 10 mM Tris-HCl (pH 7.0). The cell homogenate was centrifuged in an Eppendorf centrifuge at 14,000 ϫ rpm for 10 min, and the supernatant was further centrifuged at 100,000 ϫ g for 1 h to separate membrane and cytosol fractions. The membrane fraction was resuspended in the same homogenization buffer. The total protein amount in each sample was determined using the Bio-Rad protein dye reagent (Bradford method). The protein samples were incubated with Laemmli sample buffer at 95°C for 2 min, and then one-dimensional protein gel electrophoresis was performed in 7.5% SDS-PAGE gels in a Mini-Protein system (Bio-Rad) with 30 g of total protein loaded in each lane. Rec-iPLA 2 ␤ of 84 kDa was used as a standard (0.5-1.0 ng) in all Western blots. Separated proteins were electrophoretically transferred overnight onto nitrocellulose membrane in a Mini Trans-blot system (Bio-Rad). Blots were incubated for 1 h with 5% (w/v) skim milk in phosphate-buffered saline (PBS) containing 0.1% Tween 20 (PBST) to block residual protein binding sites. Blocked membranes were then incubated with primary anti-iPLA 2 ␤ (1:2000 dilution) for 2 h at room temperature. The primary antibody was removed, and blots were washed 3 times for 10 min with milk/PBST. Then blots were incubated with horseradish peroxidase-conjugated goat anti-rabbit IgG antibody (Cell Signaling) diluted 1:2000 in milk/PBST, washed 3 times in PBST, and treated with enhanced chemiluminescence reagents (Super ECL, Pierce) for 1 min. Blots were then exposed to photographic films, and the optical density was determined using the Un-scan-it analysis software (Silk Scientific).

Determination of iPLA 2 Activity
The activity of iPLA 2 was determined as previously described (2). Briefly, after each experimental treatment, RBL cells were homogenized as described above. The activity of iPLA 2 was determined in either total cell homogenate or in membrane and cytosol fractions, as specified in the figures. The iPLA 2 activity was measured using a modified commercial assay kit originally designed for the cytosolic phospholipase A 2 (Cayman). To detect the activity of iPLA 2 instead of cytosolic phospholipase A 2 (cPLA 2 ), the assay buffers were modified to contain no Ca 2ϩ (Ca 2ϩ is needed for cPLA 2 but not for iPLA 2 activity). Phospholipase activity was assayed by incubating the samples with the substrate, 1-hexadecyl-2-arachidonoylthio-2-deoxy-sn-glycero-3-phosphorylcholine for 1 h at room temperature in a modified Ca 2ϩ -free assay buffer of the following composition 300 mM NaCl, 60% glycerol, 10 mM HEPES, 8 mM Triton X-100, 4 mM EGTA, and 2 mg/ml bovine serum albumin (pH 7.4). The generated free thiols were visualized by the addition of 5,5Ј-dithiobis(2-dinitrobenzoic acid) for 5 min, and the absorbance was determined at 405 nm using a standard microplate reader. The background iPLA 2 -independent component of basal lipase activity was determined in control samples when all specific iPLA 2 activity was inhibited with (S)-BEL (10 M for 5 min) and was subtracted from all the readings. The presence of 4 mM EGTA in the assay buffer (which was crucial for suppressing the contaminant Ca 2ϩdependent cPLA 2 activity) did not by itself cause any significant activation of Ca 2ϩ -independent BEL-sensitive iPLA 2 . The specific activity of iPLA 2 was expressed in absorbance/mg of protein units.

Measurement of Intracellular Ca 2؉
RBL cells were transferred to minimum essential culture medium without serum and loaded with fura-2 AM (2 M) for 30 min at 37°C. Then the cells were washed for 10 min and transferred to the bath solution of the following composition: 140 mM NaCl, 10 mM HEPES, 1 mM MgCl 2 , 0.1 mM EGTA (pH 7.4). During the experiment, CaCl 2 (2 mM) was added to the cells to observe Ca 2ϩ influx following different treatments as described under "Results." Ca 2ϩ measurements were done at 20 -22°C. A dual-excitation fluorescence imaging system (Intracellular Imaging, see description above) was used for studies of individual RBL cells. The changes in intracellular Ca 2ϩ were expressed as ⌬Ratio, which was calculated as the difference between the peak F 340 /F 380 ratio after extracellular Ca 2ϩ was added and its level right before Ca 2ϩ addition. Summary data are shown without subtraction of the basal Ca 2ϩ influx. Data were summarized from the large number of individual cells (20 -40 cells tested each in 3-6 different experiments from at least 3 cell preparations).

Electrophysiology
Whole-cell currents were recorded in RBL cells using the standard whole-cell (dialysis) patch clamp technique as we pre- in Xenopus oocytes triggered by the injection of CIF extracts prepared from TG-and TPEN-treated RBL cells as described in B and C. Each bar summarizes results from 9 -14 oocytes injected with extracts obtained from three different preparations of RBL cells. Asterisks denote significant differences between specific treatments and their respective controls. G, same as in F, but with CIF extracts prepared from control and TG-and TPENtreated RBL cells in which iPLA 2 was inhibited by their pretreatment with BEL. viously described (3). Axopatch 200B amplifier was used; data were digitized at 5 kHz and filtered at 1 kHz. Pipettes were used with tip resistance of 2-4 megaohms. After breaking into the cell, the holding potential was 0 mV, and ramp depolarizations (from Ϫ100 to ϩ100 mV, 200 ms) were applied every 3 s. Amplitude of the current was expressed in pA/pF. The time course of current development was analyzed at Ϫ80 mV for each individual cell, and summary data for 5ϩ cells are shown in the figures. Average I/V relationships are shown during ramp depolarization after the current reached its maximum. Passive leakage current with zero reversal potential (at the moment of breaking into the cell or after CRAC inhibition with 10 M diethylstilbestrol) was subtracted. Intracellular (pipette) solution contained 145 mM cesium glutamate, 10 mM BAPTA, 10 mM HEPES, 3 mM MgCl 2 (pH 7.2). In some experiments concentration of BAPTA was reduced to 0.1-1 mM. Extracellular solution was 130 mM NaCl, 20 mM CaCl 2 , 5 mM HEPES, 3 mM CsCl, 1 mM MgCl 2 (pH 7.4). In experiments with exogenous CIF and recombinant iPLA 2 ␤, they were added to intracellular solutions as specified under "Results." Experiments were performed at 20 -22°C.

Statistical Analysis
Summary data are presented as mean Ϯ S.E. Single or paired Student's test was used to determine the statistical significance of the obtained data. The significance between multiple groups was evaluated using analysis of variance. Data were considered significant at p Ͻ 0.01.

Endogenous CIF Production by RBL Cells-RBL-2H3
cells were used to test if the decrease in free Ca 2ϩ concentration in the ER may trigger production of CIF and if it would cause the same or different effects compared to physical depletion (net loss) of Ca 2ϩ from the stores. To produce a physical depletion of the stores, TG was used to inhibit sarcoplasmic/ endoplasmic reticulum Ca 2ϩ ATPase, prevent Ca 2ϩ back-sequestration, and allow the passive leak of Ca 2ϩ ions from the stores. To change free Ca 2ϩ concentration in the ER without its release, we used TPEN, a membrane-permeable low affinity Ca 2ϩ buffer that is known to accumulate and decrease free Ca 2ϩ in the ER (16).
Physical depletion of the stores with TG (1 M for 5 min) triggered activation of a classical SOCE (Fig. 1A), which did not develop when iPLA 2 was irreversibly inhibited by its suicidal substrate BEL (25 M for 30 min). Endogenously produced CIF was extracted from TG-treated RBL cells, and its biological activity was measured by the ability of these extracts to activate SOCE when injected into Xenopus oocytes, a model bioassay system for CIF studies (12). Fig. 1B shows that CIF extracted from TG-treated, but not quiescent (control) cells produced dramatic activation of SOCE when injected into oocytes. These results confirmed that CIF is readily produced by RBL cells when their stores are getting depleted.  When RBL-2H3 cells were treated with 1 mM TPEN for 5 min, Ca 2ϩ influx was activated (Fig. 1C), which was identical to SOCE produced by TG (Fig. 1A). Similar to TG-induced SOCE, it was dependent on the functional activity of iPLA 2 and did not develop in BEL-treated cells. We also found that CIF extracts from TPEN-treated RBL cells were indistinguishable from those produced by TG-treated cells; when injected into oocytes they activated identical SOCE (Fig. 1D). Fig. 1E shows the summary data for TG-and TPEN-induced SOCE in RBL cells, and Fig. 1F summarizes the data for Ca 2ϩ influx in oocytes triggered by injection of CIF extracted from TG and TPEN-treated RBL cells.
In an additional series of experiments we demonstrated that CIF production is independent of iPLA 2 and is not affected by BEL (Fig. 1G). BEL-induced inhibition of iPLA 2 in SOCE pathway (Figs. 1, A and C) happens downstream from CIF. iPLA 2 ␤; Plasma Membrane Localization and Activation in Intact Cells and Cell Homogenates- Fig. 2A shows that TG-induced depletion or TPEN-induced reduction of free Ca 2ϩ in the stores of intact RBL cells produced similar activation of iPLA 2 , which was consistent with both treatments causing identical CIF production and biological effects (Fig. 1). As an additional control, we confirmed that iPLA 2 could not be activated by TG in cell homogenates, in which Ca 2ϩ stores were disintegrated and CIF could not be produced (Fig. 2B). We also found that iPLA 2 remained totally functional in cell homogenates; after cell disruption it could be activated by application of exogenous CIF extracted from the cells with depleted stores (Fig. 2B) but not by extract from control cells. Exogenous CIF extract could originate from RBL cells or from other species and cell types (for example human platelets). Fig. 2B also demonstrates that iPLA 2 in cell homogenates could be activated by a prolonged (5-10 min) treatment with 10 mM EGTA. Strong Ca 2ϩ buffering and longer time seemed to be needed to displace inhibitory CaM from iPLA 2 , mimicking the physiological effect of CIF. Interestingly, buffering capacity of 4 mM EGTA (which was used in assay buffers to ensure that only the activity of Ca 2ϩ -independent PLA 2 is measured) was not enough to produce iPLA 2 activation in live RBL cell homogenates; the basal activity (in 4 mM EGTA) was 0.046 Ϯ 0.047 Abs/mg of protein, whereas 10 mM BAPTA or exogenous CIF increased it more than 18-fold (Fig. 2B). Another set of experiments demonstrated that extracellular application of exogenous CIF extract to intact cells was without any effect ( Fig. 2A), consistent with CIF being membrane-impermeable and iPLA 2 working at the intracellular leaflet of the plasma membrane.
The next step was to determine which variant of iPLA 2 may be present and responsible for CIF-induced phospholipase activity in plasma membrane of RBL cells. Membrane and cytosol fractions of RBL cells were separated by ultracentrifugation and used for Western blot analysis and iPLA 2 activity measurements. Fig. 3A demonstrates that an antibody directed against a unique ankyrin repeat region in iPLA 2 ␤ recognized an 84-kDa human recombinant iPLA 2 ␤ (24) (that we used as a control in all Western blots) and an ϳ82-kDa band that is predominantly located in the membrane but not in the cytosol fraction of RBL cells. Importantly, membrane localization was confirmed by iPLA 2 activity measurements; more than 87% of the BEL-sensitive iPLA 2 activity was found in the membrane fraction (Fig. 3B). As expected, inclusion of 1% Triton in the homogenization buffer shifted iPLA 2 ␤ protein and the corresponding activity from membrane to cytosolic fraction (Fig. 3). These results demonstrate that an 82-kDa iPLA 2 ␤ is present and functionally active in the plasma membrane of RBL cells. It is very similar to a human recombinant iPLA 2 ␤; it has a close molecular weight and is recognized by the same antibody specific to the unique ankyrin repeats.
CIF-induced Activation of I CRAC -After demonstrating that CIF is produced in RBL cells and activates specific iPLA 2 ␤ that is associated with plasma membrane, we tested if intracellular application of exogenous CIF could activate CRAC channels in resting RBL cells, when endogenous production of CIF and I CRAC are not triggered. We already showed (Fig. 1) that resting RBL cells do not produce CIF. Furthermore, we confirmed that I CRAC did not develop during a 5-min dialysis with 0.1-1 mM BAPTA (inward current remained Ϫ0.16 Ϯ 0.04 pA/pF (n ϭ 6) with 0.1 mM BAPTA and Ϫ0.18 Ϯ 0.04 pA/pF (n ϭ 6) with 1 mM BAPTA in the pipette). However, when exogenous CIF (HPLC-purified extract from TG-treated human platelets, 1:20 dilution) was included in the dialyzing buffer, a classical I CRAC readily developed (Figs. 4, A and B). Similar to I CRAC recorded upon passive depletion of the stores during cell dialysis with 10 mM BAPTA (Figs. 4, C and D), CIF-induced current had pronounced inward rectification and was inhibited by diethylstilbestrol (10 M). The only noticeable differences were faster development and higher amplitude of the current upon CIF dialysis; at 100 s I CRAC reached 76 Ϯ 8% of its maximum amplitude of Ϫ1.56 Ϯ 0.17 pA/pF (n ϭ 8) compared with 36 Ϯ 10% of its maximum in 10 mM BAPTA (Ϫ1.11 Ϯ 0.1 pA/pF, n ϭ 7). When extract from resting platelets (with no significant CIF activity, as seen in Fig. 2B) was added to the dialyzing buffer, I CRAC did not develop (Fig. 4,  A and B). Thus, I CRAC in RBL cells could be activated not only by CIF endogenously produced upon store depletion but also by CIF extracted from other cell species and applied into cytosol of the resting cells with full stores.
iPLA 2 ␤ Is Required and Sufficient for Activation of I CRAC -To test if a full-length iPLA 2 ␤ is not only required but also sufficient for CRAC channel activation, two approaches were used. First, we tested if its molecular down-regulation could prevent I CRAC development and, second, if exogenous iPLA 2 ␤ could restore I CRAC when the endogenous enzyme is irreversibly blocked. Figs. 4, C and D, show that antisense-induced knock down of iPLA 2 ␤ protein (shown in the inset) impaired BAPTAinduced development of I CRAC , consistent with this enzyme being required for CRAC channel activation. Furthermore, CIF dialysis of RBL cells transfected with antisense to iPLA 2 ␤ also failed to activate I CRAC ; the maximum inward current (at Ϫ80mV) was Ϫ0.14 Ϯ 0.04 pA/pF (n ϭ 6) in cells transfected with antisense to iPLA 2 ␤ compared with Ϫ1.56 Ϯ 0.17 pA/pF (n ϭ 8) in control cells. (Figs. 4, E and F).
Next we tested if exogenous iPLA 2 ␤ could substitute the endogenous iPLA 2 ␤ and rescue activation of I CRAC . RBL cells were transfected with antisense to iPLA 2 ␤ (which decreased the amount of this protein to 16 Ϯ 5% of the control; see the inset in Fig. 4C) and then dialyzed with a human recombinant iPLA 2 ␤ (24), which has a CaM binding domain, ankyrin repeats, and can be recognized by the same antibody as endogenous iPLA 2 ␤ (as seen in Fig. 3A). Inclusion of the recombinant iPLA 2 ␤ (10 ng/l) into the dialyzing pipette appeared to be enough to fully restore activation of I CRAC (Fig. 4, E and F) that was lost in the cells in which endogenous iPLA 2 ␤ was knocked down. Similar recovery of I CRAC with recombinant iPLA 2 ␤ was also obtained in RBL cells in which functional activity of endogenous iPLA 2 ␤ was inhibited by (S)-BEL (this inhibitor will be described in details later); maximal I CRAC was Ϫ0.54 Ϯ 0.19 pA/pF (n ϭ 5) in the absence and Ϫ2.1 Ϯ 0.24 pA/pF (n ϭ 5) in the presence of rec-iPLA 2 ␤ in the pipette when cells were pretreated with 20 M (S)-BEL. These data provided solid proof that the presence and functional activity of iPLA 2 ␤ is required and sufficient for activation of I CRAC in RBL cells. iPLA 2 ␤ May Determine Sensitivity of I CRAC to Intracellular pH- Kerschbaum and Cahalan (36) previously demonstrated that I CRAC is sensitive to intracellular pH and speculated that acidic pH may cause modification in the selectivity filter interfering with the ion flow. While working with a recombinant iPLA 2 ␤ we found that its activity highly depends on intracellular pH. These results made us wonder if iPLA 2 may be a pH sensor that determines the dependence of I CRAC activation on intracellular pH in RBL cells. To test this we compared intracellular pH dependence of I CRAC and iPLA 2 ␤ and found a remarkable correlation. Fig. 5, A-C, show that the amplitude of I CRAC starts to sharply decrease when intracellular pH becomes lower than 7.0 and completely disappears at pH ϳ6.2. This process was strikingly similar to the pH dependence of iPLA 2 ␤ (Fig. 5D), which appeared to be fully active at pH 7.0 and higher, but demonstrated a sharp loss of function when pH was down from 7.0 to 6.2.
These results demonstrate a possibility of modulation of I CRAC and SOCE via regulation of the enzymatic activity of iPLA 2 ␤, making it not only a crucial determinant of SOC channel activation but also an important target for fine-tuning of the I CRAC and SOCE pathway.
(S)-BEL, a Specific Tool for Studying the Role of iPLA 2 ␤ and I CRAC in Ca 2ϩ Entry-Recently, Gross and co-workers (22) demonstrated that racemic BEL (a commonly used mechanism-based suicidal substrate for all major isoforms of iPLA 2 (37)), is composed of two enantiomers, (S)-BEL, which has higher specificity to iPLA 2 ␤, and (R)-BEL, which is more specific to iPLA 2 ␥. Here we thought to test the effects of these two enantiomers on the activity of iPLA 2 ␤, and the activation of I CRAC and SOCE in RBL cells to determine whether (S)-BEL could be used as an advanced enantio-selective pharmacological tool for the studies of the role of iPLA 2 ␤ and store-operated channels in Ca 2ϩ entry.
First, we purified BEL enantiomers (Fig. 6A) and showed that (S)-but not (R)-BEL inhibits TG-induced activation of iPLA 2 (Fig. 6B), consistent with iPLA 2 ␤ but not iPLA 2 ␥ being activated by depletion of Ca 2ϩ stores in RBL cells. Next, we tested the effects of (S)-and (R)-BEL on I CRAC and SOCE. Figs. 6, C and D, show that (S)-but not (R)-BEL inhibits BAPTA-induced I CRAC in RBL cells. TG-induced Ca 2ϩ influx was also effectively inhibited by (S)-BEL but not (R)-BEL (Fig. 6E). Dose-response profiles presented in Fig. 6F demonstrates that (S)-BEL inhibits capacitative Ca 2ϩ influx with IC 50 about 3 M. In contrast, (R)-BEL produces very little or no effect at 10-fold higher concentrations. Identical inhibition of iPLA 2 activity and SOCE by (S)and not (R)-BEL was also obtained in vascular smooth muscle cells (data not shown). These results demonstrate that (S) enantiomer of BEL could be used as a highly specific pharmacological tool to inhibit iPLA␤, I CRAC , and SOCE.

DISCUSSION
This study carefully tested and fully confirmed the CIF-and iPLA 2 -dependent mechanism of CRAC channel activation in RBL-2H3 cells. We obtained new important information on the stimulus for endogenous CIF production, the ubiquitous nature of CIF, and its interchangeable biological activity between different species, strict dependence of I CRAC activation on the presence and functional activity of a specific membrane-bound 82-kDa variant of iPLA 2 ␤, and new advanced pharmacological tools for SOCE studies.
We demonstrated that CIF is produced and can be extracted from RBL cells after depletion of their stores, and its biological activity is identical to the activity of CIF purified from other cell types (14), which suggests that CIF is highly preserved and may be interchangeable between the species. When extracted from donor cells and dialyzed into acceptor cell, exogenous CIF worked identically to the endogenous CIF; it produced the same activation of iPLA 2 ␤ and I CRAC . Furthermore, exogenous CIF was active only upon intracellular application; 1) it activated iPLA 2 ␤ when added to cell homogenates but not to intact cells, and 2) intracellular CIF dialysis was needed to activate I CRAC . Also, exogenous CIF was able to activate I CRAC in resting cells in which Ca 2ϩ stores were not depleted, and I CRAC did not develop upon dialysis with control (inactive) extract. Importantly, CIF produced much faster activation of I CRAC than what is usually seen upon cell dialysis with BAPTA, exactly as one would expect from CIF directly activating iPLA 2 , which is only one step away from activation of the channels, whereas BAPTA needs extra time to deplete the stores and trigger CIF production first. Inability of CIF to activate I CRAC in RBL cells in which iPLA 2 ␤ was abolished further confirmed that iPLA 2 ␤ is indeed a physiological target for CIF and an essential component of CRAC channel activation.
The studies in RBL cells brought new important information on a specific variant of iPLA 2 ␤ that is required and sufficient for activation of CRAC channels as well as its localization and intracellular CIF-and pH-dependent regulation. The existence of multiple splice variants of iPLA 2 ␤ (33) may play an important role in diversification of iPLA 2 ␤ functions and be one of the possible reasons for the variety of reports on its localization and role in cellular biology. Our results suggest that a full-size 82-85-kDa splice variant of iPLA 2 ␤ (featuring 7 ankyrin repeats in N terminus) may localize within specific signaling domains at, or in close proximity to plasma membrane, making it available for the SOCE pathway. Our previous studies showed that it is bound to the plasma membrane and could be extracted and retain its functional activity in membrane patches (13); CIF application to the intracellular side of inside-out membrane patches was able to activate single nonselective SOCs in smooth muscle cells, and CaM binding back to iPLA 2 in those patches could shut SOCs down. Here we demonstrated that, despite its seemingly tight association with plasma membrane, 82-kDa iPLA 2 ␤ could be detached by detergent and probably by other more physiological/pathological conditions, raising an attractive possibility that association of iPLA 2 ␤ with plasma membrane represents one of many ways of iPLA 2 -dependent regulation of SOCE in living cells. Our results also suggest that completely detergent-free homogenization conditions are necessary to keep iPLA 2 ␤ in the membrane fraction after cell disruption and may explain why iPLA 2 ␤ was found in the cytosol in some previous studies.
Several lines of evidence in this study confirmed the specificity of BEL and the exact location of its target in the store-operated pathway. First, we demonstrated that treatment by BEL mimics the effects of molecular down-regulation of iPLA 2 ␤ on I CRAC . Second, it does not affect CIF production, which is located upstream from iPLA 2 , and does not impair activation of SOCE by lysophospholipids (2), downstream products of iPLA 2 . Also, we demonstrated that only (S)-BEL (one of the two chiral enantiomers of BEL) inhibits TG-induced activation of iPLA 2 , I CRAC , and SOCE in RBL cells. Thus, (S)-BEL may be used as a highly specific pharmacological tool for broader studies of Ca 2ϩ homeostasis in cells and organs, in which molecular manipulations with iPLA 2 ␤ and SOC channels may pose considerable difficulties.
Our studies on the nature of the signal in ER that may trigger CIF machinery not only reflect the first attempts to peek into a "black box" of CIF production but also address the major question of what is a real signal for SOCE. In these studies we found that simple buffering of free Ca 2ϩ within the stores triggers CIF production and activation of SOCE in RBL cells, which was indistinguishable from that triggered by a physical Ca 2ϩ loss from ER. Thus, a drop in intraluminal free Ca 2ϩ concentration provides a physical stimulus for CIF production and SOCE activation.
Taken together our studies in RBL cells significantly extended our understanding of the crucial role of CIF and the specific membrane-bound variant of iPLA 2 ␤ in activation and physiological regulation of CRAC channels and SOCE.