All-or-None Activation of CRAC Channels by Agonist Elicits Graded Responses in Populations of Mast Cells

This information is current as Wei-Chiao Chang, Joseph Di Capite, Charmaine Nelson and of October 1, 2021. Anant B. Parekh J Immunol 2007; 179:5255-5263; ; doi: 10.4049/jimmunol.179.8.5255 http://www.jimmunol.org/content/179/8/5255 Downloaded from

References This article cites 44 articles, 12 of which you can access for free at: http://www.jimmunol.org/content/179/8/5255.full#ref-list-1 http://www.jimmunol.org/ Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*average by guest on October 1, 2021

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2007 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

All-or-None Activation of CRAC Channels by Agonist Elicits Graded Responses in Populations of Mast Cells1

Wei-Chiao Chang, Joseph Di Capite, Charmaine Nelson, and Anant B. Parekh2

In nonexcitable cells, receptor stimulation evokes Ca2؉ release from the endoplasmic reticulum stores followed by Ca2؉ influx -through store-operated Ca2؉ channels in the plasma membrane. In mast cells, store-operated entry is mediated via Ca2؉ release activated Ca2؉ (CRAC) channels. In this study, we find that stimulation of muscarinic receptors in cultured mast cells results in Ca2؉-dependent activation of protein kinase C␣ and the mitogen activated protein kinases ERK1/2 and this is required for the 2؉ subsequent stimulation of the enzymes Ca -dependent phospholipase A2 and 5-lipoxygenase, generating the intracellular mes- senger arachidonic acid and the proinflammatory intercellular messenger leukotriene C4. In cell population studies, ERK acti- vation, arachidonic acid release, and leukotriene C4 secretion were all graded with stimulus intensity. However, at a single cell level, Ca2؉ influx was related to agonist concentration in an essentially all-or-none manner. This paradox of all-or-none CRAC channel activation in single cells with graded responses in cell populations was resolved by the finding that increasing agonist Downloaded from concentration recruited more mast cells but each cell responded by generating all-or-none Ca2؉ influx. These findings were extended to acutely isolated rat peritoneal mast cells where muscarinic or P2Y receptor stimulation evoked all-or-none activation of Ca2؉entry but graded responses in cell populations. Our results identify a novel way for grading responses to agonists in immune cells and highlight the importance of CRAC channels as a key pharmacological target to control mast cell activation. The Journal of Immunology, 2007, 179: 5255–5263. http://www.jimmunol.org/

2ϩ n eukaryotic cells, Ca influx is a critical trigger for a di- acid liberated by cPLA2 is then converted to the proinflammatory

verse array of cellular responses including exocytosis, mus- paracrine signal leukotriene C4 (LTC4) by the 5-lipoxygenase en- cle contraction, gene transcription, and cell growth (1). In zyme (12, 13). I 2ϩ nonexcitable cells, Ca entry occurs through store-operated and The physiological trigger for CRAC channel activation is an second messenger-gated Ca2ϩ channels in the plasma membrane increase in the levels of the ubiquitous second messenger inositol 2ϩ (2, 3). Store-operated Ca channels are activated following the 1,4,5-trisphosphate (InsP3), which occurs following stimulation of emptying of intracellular Ca2ϩ stores, are regulated by a diverse cell surface receptors that couple to phospholipase C (2, 14). The range of agonists, and are found in a variety of nonexcitable cells. extent of I activation is steeply related to the cytoplasmic

CRAC by guest on October 1, 2021

Electrophysiological evidence indicates that store-operated InsP3 concentration with a Hill coefficient of 12, and this essen- channels represent a heterogeneous family (4). The best charac- tially all-or-none relationship is seen either by infusing cells with 2ϩ 2ϩ 3 terized member is the Ca release-activated Ca (CRAC) chan- InsP3 or following stimulation of plasma membrane receptors that nel and the whole cell current flowing through CRAC channels is activate phospholipase C (15). Such nonlinear behavior arises from called ICRAC (5, 6). Defective CRAC channel activity has been the metabolism of InsP3 by inositol 5-phosphatase because the causally linked to inherited primary immunodeficiencies (7) and nonmetabolisable InsP3 analog Ins2,4,5P3 generates a graded re- 2ϩ Ca entry through CRAC channels regulates exocytosis (8), gene sponse whereas InsP3-F, another analog that is only broken down transcription (9), and cytoplasmic enzymes such as NO synthase by the 5-phosphatase, activates ICRAC in a highly supralinear man- (10), adenylate cyclase (11), and Ca2ϩ-dependent phospholipase ner (15, 16). Consistent with this is the finding that inhibition of 2ϩ A2 (cPLA2; Ref. 12). Ca influx through CRAC channels acti- inositol 5-phosphatase with InsP4 reduces the steepness of the re- vates cPLA2 indirectly, via recruitment of conventional protein lationship between InsP3 concentration and activation of ICRAC ␣ ␤ 2ϩ kinase C and I isozymes and subsequent stimulation of the mi- and enables lower concentrations of InsP3 to evoke robust Ca togen activated protein kinases ERK1 and ERK2 (13). Arachidonic entry (17). Hence, following stimulation of receptors that elevate

intracellular InsP3 levels, ICRAC appears to activate in an essen- Department of Physiology, Anatomy and Genetics, University of Oxford, Oxford, tially all-or-none manner. Although several negative feedback United Kingdom mechanisms can subsequently inhibit ICRAC and hence grade the 2ϩ Received for publication June 19, 2007. Accepted for publication August 12, 2007. overall extent of Ca influx (2, 18), the fact that ICRAC activates The costs of publication of this article were defrayed in part by the payment of page in an all-or-none manner following receptor stimulation neverthe- charges. This article must therefore be hereby marked advertisement in accordance less raises an intriguing paradox. Weak stimuli (presented as con- with 18 U.S.C. Section 1734 solely to indicate this fact. centrations of agonist that are below the affinity of the receptor for 1 This work was supported by an Medical Research Council Programme Grant (to its cognate ligand) often produce weak cellular responses whereas A.B.P.). W.-C.C. was a recipient of an Overseas Research Studentship award and J.D.C. held a Christopher Welch Scholarship. higher concentrations elicit stronger responses. If ICRAC is essen- 2 Address correspondence and reprint requests to Prof. Anant Parekh, Oxford Uni- tially an all-or-none process, how can responses be graded with versity, Parks Road, Oxford, U.K. E-mail address: [email protected] stimulus intensity? 3 2ϩ 2ϩ Abbreviations used in this paper: CRAC, Ca release-activated Ca ; cPLA2, We have addressed this issue by examining the effect of different 2ϩ 2ϩ Ca -dependent phospholipase A2; LTC4, leukotriene C4; InsP3, inositol 1,4,5- levels of receptor activation on ICRAC,Ca entry, protein kinase trisphosphate; 2-APB, 2-aminoethyldiphenyl borate. ␣ C , and ERK stimulation and subsequent activation of the cPLA2 Copyright © 2007 by The American Association of Immunologists, Inc. 0022-1767/07/$2.00 and 5-lipoxygenase enzymes in a mast cell line as well as in www.jimmunol.org 5256 ALL-OR-NONE CALCIUM INFLUX IN MAST CELLS acutely isolated rat peritoneal mast cells. Our new results demon- from a holding potential of 0 mV and normalized to cell size (capacitance). strate that increasing agonist concentration increases protein kinase Currents were filtered using an 8-pole Bessel filter at 2.5 kHz and digitized ␣ at 100 ␮s. Capacitative currents were compensated before each ramp. Leak C and ERK activation, cPLA2 stimulation, and LTC4 secretion in 2ϩ currents were obtained by averaging two ramp currents taken just before a graded manner. However, activation of ICRAC and Ca influx application of carbachol and subtracting this from all subsequent record- remain essentially all-or-none processes. Responses appear graded ings. Series resistance was Ͻ10 MOhms. with stimulus intensity because increasing agonist concentration Preparation of cell lysates recruits more cells in the population but each cell that responds does so by activating Ca2ϩ influx in an essentially all-or-none Attached cells from 10-cm plastic dishes were washed twice with PBS and manner. Our results therefore identify a mechanism for eliciting lysed with PBS buffer containing 0.5% Triton X-100 and protease mixture inhibitor (Sigma-Aldrich), as described (12). Lysates were centrifuged at graded responses in a multicellular system despite all-or-none ac- 8000 rpm for 5 min, and the supernatants were collected and stored at tivation of CRAC channels by receptor stimulation. Ϫ70°C until used. Materials and Methods Western blotting Cell culture Total cell lysates (40 ␮g) were analyzed by SDS-PAGE on a 10% gel. Membranes were blocked with 5% nonfat dry milk in PBS plus 0.1% Rat basophilic leukemia cells stably expressing the muscarinic type 3 re- Tween 20 (PBST) buffer for2hatroom temperature. Membranes were ceptor were a gift from Dr. J. Putney (National Institute of Environmental washed with PBST three times and then incubated with primary Ab for 1 h Health Sciences, Research Triangle Park, NG). Cells were cultured (37°C, at room temperature. Anti-phospho-ERK Ab, which recognizes dual phos- 5% CO2) in DMEM with 10% FBS, 2 mM L-glutamine, penicillin-strep- phorylated (i.e., active) ERK, was from New England BioLab and used at ␮ 2ϩ tomycin, and 100 M G-418, as described previously (12, 15). For Ca 1/2000 dilution. Total ERK 2 Ab was from Santa Cruz Biotechnology and imaging and patch clamp experiments, cells were passaged (using trypsin) used at a dilution of 1/5000. Secondary Ab was rabbit IgG (1/2500). The Downloaded from onto glass coverslips and used 36–72 h after plating. membranes were then washed with PBST again and incubated with 1/2000 Isolation of rat peritoneal mast cells dilution of peroxidase-linked anti-mouse IgG from Amersham Biosciences for1hatroom temperature. After washing with PBST, the bands were The mast cells were isolated from female, Sprague-Dawley rats, weighing detected by an ECL-plus Western blotting detection system (Amersham ϳ300 g. The animals were sacrificed according to “Schedule 1”, (carbon Biosciences). dioxide overdose and neck dislocation). Immediately, 100–150 ml of ster- [3H]Arachidonic acid release

ile HEPES buffer (in mM: NaCl 150, KCl 5.6, HEPES 10, NaOH 1.5, http://www.jimmunol.org/ MgCl2 1, CaCl2 2, glucose 10 g/l BSA (pH 7.4)) was injected into the Cells were relabelled with 0.25 ␮Ci/ml [3H]arachidonic acid in DMEM for peritoneal cavity. The abdomen was massaged for 2 min and then the buffer ϫ 1.5 h at 37°C, as described previously (12). Cells were then washed twice removed. This was centrifuged at 200 g for 10 min. The pellet was with serum-free DMEM to remove unincorporated [3H]arachidonic acid. resuspended in DMEM, triturated, and plated onto glass coverslips (for 2ϩ Thapsigargin with 0.5% fatty acid-free BSA was added for different times Ca imaging and immunocytochemistry) or cell-culture dishes (for LTC4 (see text). Medium was collected and centrifuged at 25g for 5 min to measurements). remove any floating cells. Radioactivity in the supernatant was measured. 3 Ca2ϩ imaging The amount of [ H]arachidonic acid released into the medium was ex- pressed as a percentage of the total [3H]arachidonic acid uptake. Ca2ϩ imaging experiments were conducted using the IMAGO CCD cam- era-based system from TILL Photonics, as described previously (12). Cells LTC4 measurements by guest on October 1, 2021 were alternately excited at 356 and 380 nm (20-ms exposures) using a Following stimulation of attached cells with carbachol, the supernatant was Polychrome Monochromator and images were acquired every 2–3 s. Im- collected and LTC4 levels were measured by enzyme immunoassay (Cay- ages were analyzed offline using IGOR Pro. Cells were loaded with Fura ␮ ␮ man Chemicals) as described previously (12). In brief, 50 l of superna- 2-AM (2 M) for 40 min at room temperature in the dark and then washed tant, leukotriene C4 acetylcholinetransferase and leukotriene C4 antiserum three times in standard external solution of composition (in mM) NaCl 145, were added to each enzyme immunoassay well. Following incubation for KCl 2.8, CaCl2 2, MgCl2 2, D-glucose 10, HEPES 10 (pH 7.4) with NaOH. 18 h at room temperature, the wells were emptied and rinsed five times Cells were left for 15 min in the dark to allow further deesterification. Ca2ϩ 2ϩ ⌬ with enzyme immunoassay washing buffer. Two hundred microliters of and Ba signals are presented as the fluorescence ratio (356/380) R. Ellman’s reagent (prepared fresh) was added to each well and the plate was Immunocytochemistry placed on an orbital shaker for 1.5 h in the dark. Plate absorbance was measured at a wavelength of 405 nm. Data is presented relative to basal

After treatment with carbachol, the cells were fixed in 4% paraformalde- LTC4 secretion from nonstimulated cells. hyde in phosphate buffer, for 30 min at room temperature. All washes used Statistical analysis 0.01% PBS, (PBS in mM: NaCl 137, KCl 2.7, Na2HPO4 8, KH4PO4 1). The cells were blocked with 2% BSA and 10% goat serum for 2 h. Trans- Data is presented as the mean Ϯ SEM. Statistical significance was con- location of protein kinase C␣ was visualized using a polyclonal rabbit IgG .(ء) sidered as p Ͻ 0.01, using Student’s t test, and is denoted by an asterisk Ab, (Santa Cruz Biotechnology). After blocking, the cells were washed and incubated in antiPKC␣ (1/1000 in 0.2%BSA, 1% goat serum), overnight at 4oC. The cells were thoroughly washed and the secondary Ab (Alexofluor Results 2ϩ 568-conjugated goat anti-rabbit IgG) was applied at 1/2000 in PBS for 2 h Ca influx following muscarinic receptor stimulation activates at room temperature. The cells were mounted in Vectastain mounting me- cPLA2 and LTC4 production via recruitment of the ERK dium. Images were obtained using a Leica confocal microscope and anal- pathway ysis of translocation conducted using National Institutes of Health image software. Stimulation of cell surface muscarinic receptors with 100 ␮M car- bachol (a maximally effective concentration) raised cytoplasmic Patch clamp recordings Ca2ϩ (Fig. 1A). In the absence of external Ca2ϩ, carbachol elicited Whole cell patch clamp experiments were conducted as described (12, 15). a rapid, albeit transient, rise in cytoplasmic Ca2ϩ that reflected Sylgard-coated, fire-polished patch pipettes filled with a solution that con- Ca2ϩ mobilization from intracellular stores (labeled 0 Ca2ϩ in Fig. tained 145 mM cesium glutamate, 8 mM NaCl, 1 mM MgCl ,2mM 2 1A). In the presence of external Ca2ϩ however, the initial Ca2ϩ Mg-ATP, 10 mM HEPES, 10 mM EGTA, 4.6 mM CaCl2 (pH 7.2) with CsOH. Free Ca2ϩ was buffered at ϳ140 nM. Pipette resistance was ϳ5 rise developed into a plateau that declined only slowly with time MOhms when placed in an external solution containing 145 mM NaCl, 2.8 (Fig. 1A). This Ca2ϩ influx component was central to the activa- mM KCl, 10 mM CaCl2, 2 mM MgCl2, 10 mM CsCl, 10 mM D-glucose, tion of cPLA because arachidonic acid generation only occurred 10 mM HEPES (pH 7.4) with NaOH. A correction of ϩ 10 mV was 2 following stimulation of muscarinic receptors in the presence of applied for the subsequent liquid junction potential that arose from the ϩ ϩ Ϫ external Ca2 (Fig. 1B). Despite robust and rapid Ca2 release, glutamate-based pipette solution. ICRAC was measured at 80 mV from Ϫ ϩ 2ϩ voltage ramps ( 100 to 100 mV lasting 50 msec) applied at 0.5 Hz, carbachol failed to activate cPLA2 in the absence of external Ca The Journal of Immunology 5257

FIGURE 1. Ca2ϩ influx following muscarinic recep- carbachol tor stimulation activates cPLA2 and LTC4 secretion via 10 ␮ recruitment of ERK. A, Stimulation with 100 M car- 2.0 2ϩ 2ϩ bachol results in Ca release followed by Ca influx + Ca2+ ∆ into the cells (averaged data from 63 cells is shown). In R 2ϩ 2ϩ 5 the absence of external Ca , only the Ca release 1.5 phase is seen (55 cells). B, Carbachol (100 ␮M) acti-

vates cPLA2 and arachidonic acid release but only in the 2+ 2ϩ 0 Ca presence of external Ca . Arachidonic acid release in 1.0 0 response to carbachol is similar in extent to that seen 2+ 2+ normalised arachidonate release control + Ca 0 Ca + carb. U0126 with thapsigargin, is not additive with thapsigargin and 0 100 200 300 + carb. Time (s) carbachol thapsigargin is blocked by the MEK inhibitor U0126 (10 ␮M; pre- treated for 15 min). The abbreviation carb. refers to car- ␮ bachol. C, Carbachol (100 M) activates LTC4 secre- tion, but only in the presence of external Ca2ϩ. As with 4 ERK-P arachidonic acid release, carbachol-evoked LTC4 is sim-

ilar in extent to that induced by thapsigargin, is not ad- secretion

4 control carb. thap. ditive with thapsigargin and is suppressed by U0126. D, Muscarinic receptor activation (100 ␮M carbachol for 8 2 ERK-P

min) stimulates ERK phosphorylation, and to an extent Downloaded from similar to that seen following challenge with thapsigar- control +U0126 carbachol

␮ normalised LTC gin (2 M for 8 min). ERK phosphorylation to carbachol 0 was prevented by U0126. control + Ca2+ 0 Ca2+ + carb. U0126 + carb. carbachol thapsigargin

(Fig. 1B). The extent of arachidonic acid release following mus- stimulation and the corresponding time courses are superimposed http://www.jimmunol.org/ carinic receptor activation was not additive with that seen follow- in Fig. 2E. All three parameters showed similar kinetics of devel- ing stimulation with the SERCA pump inhibitor thapsigargin (2 opment, consistent with a tight temporal correlation between them. ␮M for 8 min), which empties stores and maximally activates

CRAC channels (Fig. 1B). Arachidonic acid released by cPLA2 Graded dependence on stimulus intensity activity is subsequently metabolized by the 5-lipoxygenase en- To establish the relationship between ERK stimulation, cPLA ac- zyme to form the potent proinflammatory paracrine signal LTC 2 4 tivation, and LTC secretion with stimulus intensity, we examined (12). Muscarinic receptor stimulation in the presence, but not ab- 4 2ϩ their extent of activation following different levels of muscarinic sence, of external Ca resulted in secretion of LTC4 and to an receptor occupancy. Fig. 3A shows that increasing agonist concen- by guest on October 1, 2021 extent similar to that seen with thapsigargin (Fig. 1C). Again, there tration resulted in graded activation of ERK in cell population was no additivity between the two stimuli. ϩ measurements. The corresponding dose-response curve is plotted Stimulation with carbachol in the presence of external Ca2 in Fig. 3B. Activation of cPLA (Fig. 3C) and secretion of LTC triggered phosphorylation (and hence activation) of the mitogen 2 4 (Fig. 3D) were also graded with stimulus intensity. Normalized activated protein kinases ERK1 and 2 (Fig. 1D). Application of responses to ERK stimulation, cPLA activation, and LTC secre- carbachol to wild-type RBL cells not transfected with muscarinic 2 4 tion are superimposed in Fig. 3E. Importantly, the curves were receptor failed to evoke any resolvable ERK activation (data not superimposable, indicating that each pathway was activated in a shown). Carbachol activated ERK to an extent similar to that seen graded manner by carbachol and to similar relative extents. following stimulation with thapsigargin (Fig. 1D). ERK1/2 are ac- tivated following dual phosphorylation of critical threonine and 2ϩ tyrosine residues by the upstream mitogen activated protein kinase The receptor-activated Ca influx pathway that drives ERK kinases MEK1/2. The MEK1/2 inhibitor U0126 (19) suppressed stimulation involves CRAC channels

ERK activation following muscarinic receptor stimulation (Fig. The finding that ERK activation, cPLA2 stimulation, and LTC4 2ϩ 1D). U0126 also suppressed cPLA2 activation (Fig. 1B) and gen- secretion, which are all dependent on Ca influx, were graded eration of LTC4 (Fig. 1C) following carbachol stimulation. U0126 with agonist concentration is unexpected because agonist-depen- 2ϩ has no effect store-operated Ca signals (13). Collectively, these dent activation of ICRAC by InsP3 is essentially an all-or-none pro- ϩ results demonstrate that Ca2 influx following stimulation of mus- cess (15). We considered various possibilities for this apparent carinic receptors recruits the ERK pathway, which then results in discrepancy. For example, receptor activation could recruit addi- 2ϩ the generation of the second messenger arachidonic acid as well as tional Ca entry pathways to ICRAC and these could be activated the paracrine signal LTC4. in a graded manner. However, three pieces of evidence militate against this. First, whole cell patch clamp recordings demonstrated Tight temporal correlation between ERK stimulation, cPLA2 that carbachol application activated a nonvoltage-gated inwardly ϩ activation, and LTC4 secretion rectifying Ca2 current with a reversal potential Ͼϩ60 mV (Fig. ␮ Fig. 2A shows that stimulation with carbachol (100 M) resulted 4A), which was indistinguishable from ICRAC activated by store in a time-dependent activation of ERK (upper panel) and densi- depletion (thapsigargin or dialysis with 10 mM EGTA; Ref. 20). tometric analysis of the gel is plotted in Fig. 2B. Generation of Second, mitochondrial depolarisation suppresses store-operated 2ϩ 2ϩ arachidonic acid and secretion of LTC4 also increased with the Ca entry and this reflects, at least in part, enhanced Ca -de- duration of carbachol stimulation (Fig. 2, C and D, respectively). pendent inactivation of CRAC channels due to the loss of mito- We normalized the extent of ERK activation, arachidonic acid gen- chondrial Ca2ϩ buffering (21). Depolarisation of mitochondria

eration, and LTC4 secretion to the value measured after 12 min with either the complex III respiratory chain inhibitor antimycin A 5258 ALL-OR-NONE CALCIUM INFLUX IN MAST CELLS

10

ERK-P 5 total FIGURE 2. Time course of ERK activation, arachi- ERK2 control 2 min 4 min 12 min extent of ERK-P donic acid release and LTC4 secretion following stimu- 0 lation with carbachol. A, Western blots showing the ex- carbachol tent of ERK phosphorylation after carbachol challenge 840 12 (100 ␮M) for the times indicated. Upper panel depicts time (minutes) ERK1/2 phosphorylation and the lower panel depicts to- tal ERK2, used as a control for constant gel loading. B, A C Data from three experiments as in is summarized. 1 and D, Time course of arachidonic acid release (C) and 6 4 LTC4 secretion (D) are shown. E, Time courses of ERK release activation, arachidonic acid release and LTC secretion 4 4 4 are superimposed to show the close temporal association. 2 2 normalised response 0 normalised LTC 0 0 Downloaded from normalised arachidonate release 840 12 840 12 840 12 time (minutes) time (minutes) time (minutes)

(5 ␮g/ml) or collapse of the proton gradient with 2mM p-triflu- lowing stimulation with a low concentration of carbachol (1 ␮M; oromethoxy carbonyl cyanide phenyl hydrazone (both in the pres- Fig. 4, C and D), suggesting that Ca2ϩ influx even at low agonist ence of oligomycin) suppressed carbachol-evoked Ca2ϩ influx, but concentrations in mast cells is via CRAC channels. http://www.jimmunol.org/ without affecting Ca2ϩ release from the stores (22). Importantly, mitochondrial depolarization suppressed activation of cPLA (data Divalent cation entry following receptor stimulation is an 2 apparent all-or-none process not shown) and LTC4 secretion (Fig, 4B) following stimulation 2ϩ with carbachol. Finally, ICRAC is suppressed by application of the To see whether store-operated Ca influx was supralinearly re- CRAC channel blocker 2-aminoethyldiphenyl borate (2-APB) (40 lated to agonist concentration in intact cells, we applied different ␮M; Ref. 2). 2-APB suppressed Ca2ϩ influx following stimulation concentrations of carbachol in Ca2ϩ-free solution to fura 2-loaded with 100 ␮M carbachol (data not shown) and inhibited subsequent cells and then added Ba2ϩ to the extracellular solution. Ba2ϩ per- cPLA activation (Fig. 4C) and LTC secretion (Fig. 4D). We also meates CRAC channels but, unlike Ca2ϩ, is not transported out of

2 4 by guest on October 1, 2021 considered the possibility that receptor stimulation might recruit the cytoplasm by Ca2ϩATPases and is therefore used as an indi- different Ca2ϩ influx pathways in a concentration-dependent man- cator of unidirectional divalent cation entry (2). We measured the ner. It has been suggested in HEK293 cells for example that low initial rate of Ba2ϩ entry as an indicator of the number of open concentrations of agonist recruit a nonstore-operated Ca2ϩ influx CRAC channels and aggregate data from all cells used is summa- pathway gated by arachidonic acid that is 2-APB insensitive rized in Fig. 4E (n Ͼ 150 cells per point). The relationship between whereas higher agonist concentrations activate store-operated carbachol concentration and the rate of Ba2ϩ entry was clearly Ca2ϩ influx (23). However, 2-APB suppressed Ca2ϩ influx and graded and correlated well with corresponding agonist-response 2ϩ both Ca -dependent arachidonate release and LTC4 secretion fol- curves in Fig. 3. However, this dose-response curve reflects all

10

ERK-P 5 total ERK2 extent of ERK-P FIGURE 3. Graded activation of ERK, arachidonic control 0.2 1510 100 0 acid and LTC4 secretion following stimulation with dif- µ carbachol ( M) 0.1 1 10 100 ferent concentrations of carbachol. A, Western blots carbachol (µ Μ) showing the relationship between agonist concentration and ERK phosphorylation. B, Aggregate data from three experiments is summarized. C and D, Corresponding 3 4 dose-response curves for arachidonic acid release and 1 LTC4 secretion. E, Dose-response curves for the three 2 parameters are superimposed to show the similar depen- 2 dence on agonist concentration. secretion 1 4 LTC normalised response arachidonate release 0 0 0 0.1 1 10 100 0.1 1 10 100 0.1 1 10 100 carbachol (µ Μ) carbachol (µ Μ) carbachol (µ Μ) The Journal of Immunology 5259

carbachol 0 -50 +50 mV 5 6 * * * -1 4 * 4

-1 -2 secretion 4 secretion 4 3 4 (pA/pF) -3 pA/pF 4 * 2 2 * CRAC -2 I 2 1

-3 normalised LTC 0 0 normalised LTC 0 control anti. FCCP control 2-APB 2-APB control 2-APB 2-APB 0 50 100 150 time (seconds) oligomycin normalised arachidonate release carbachol carbachol carbachol carbachol 1 µ M 100 µ M 10 µ M 100 µ M carbachol

3 100 100 100

80 80 80 2

60 60 entry rate (%) 60 2+ (-pA/pF) 40 40 40 1 % response CRAC 20 20 20 I Downloaded from normalised response (%) 0 0 0 0 normalised Ba 1 10 100 1 10 100 1 10 100 0.1 1 10 100 [carbachol] (µ Μ) [carbachol] (µ Μ) [carbachol] (µ Μ) [carbachol] (µ Μ) ␮ FIGURE 4. Graded responses to carbachol despite all-or-none activation of ICRAC. A, Carbachol (100 M) activates ICRAC. Inset shows the current-

voltage relationship taken at steady state. B, Carbachol-evoked LTC4 secretion is impaired by depolarising mitochondria with antimycin A or FCCP (both with oligomycin) but not by oligomycin alone. C and D, The CRAC channel blocker 2-APB suppresses arachidonic acid release (C) and LTC4 secretion (D) to a submaximal (1 ␮M) and maximal (100 ␮M) dose of carbachol. E,Ba2ϩ entry rate through CRAC channels is graded with agonist concentration http://www.jimmunol.org/ when all cells are analyzed. F, The fraction of cells that respond to carbachol increases with agonist concentration. Carbachol was applied in Ca2ϩ-free solution and the response measured was Ca2ϩ release from the stores. G,Ba2ϩ entry rate is independent of agonist concentration (above threshold) when

only those cells that responded in panel F are analyzed. H,ICRAC activation is steeply related to carbachol concentration.

cells in the population i.e., those that respond to carbachol as well as those that do not. Importantly, when we applied carbachol in carbachol Ca2ϩ-free solution, taking Ca2ϩ release as an indicator of receptor 3 activation, we found that only a fraction of the cells responded to by guest on October 1, 2021 lower concentrations of agonist (Fig. 4F). Increasing carbachol concentration increased the number of cells that responded. The ∆R graded response for Ba2ϩ entry in Fig. 4E closely resembles that for agonist responsiveness in Fig. 4F, and therefore is not a true 2 representation of Ba2ϩ influx. We therefore analyzed the rate of Ba2ϩ entry only for those cells that responded to carbachol. Ca2+ Results are shown in Fig. 4G. The rate of Ba2ϩ entry was now independent of agonist concentration for all responding cells. Al- 1 though less straightforward to interpret due to cytoplasmic Ca2ϩ removal processes, we also compared the rate of Ca2ϩ entry fol- 0 100 200 300 time (seconds) lowing stimulation with low and high concentrations of carbachol. The rate of Ca2ϩ entry to 1 ␮M carbachol was 84 Ϯ 9% that seen 1.5 in 100 ␮M carbachol and the small difference was not statistically ** significant ( p Ͼ 0.3). If each cell responds to a given agonist * concentration in an all-or-none way, then the fractional response in

release 1.0 a cell population should be equal to the number of cells that are 4 activated. Consistent with this, whereas virtually all cells respond to 100 ␮M carbachol, ϳ40% of the cells do so to 1 ␮M agonist 0.5 (Fig. 4F). Relative to 100 ␮M carbachol, 1 ␮M activates ERK,

cPLA2, and LTC4 to similar extents (39.7, 29.5, and 41.0%, re- spectively; Fig. 3). normalised LTC 0.0 Hence, the graded relationship between agonist concentration control 0.1 1 100 and divalent cation influx in Fig. 4E is misleading because it rep- [carbachol] (µ M) resents the likelihood of a cell responding to carbachol in an all- FIGURE 5. Muscarinic receptor stimulation evokes cytoplasmic Ca2ϩ or-none manner rather than graded cation influx into each cell. signals and triggers LTC4 secretion in primary rat peritoneal mast cells. A, To examine this further, we constructed a carbachol dose-re- Typical response to carbachol (100 ␮M) from a fura 2-loaded primary rat ϩ sponse curve for ICRAC activation (Fig. 4H). The relationship was peritoneal mast cell. Carbachol was applied in Ca2 -free solution and then supralinear, consistent with a previous report in which adenosine 2mMCa2ϩ was readmitted (arrow). B, Carbachol dose-response curve to Ͻ ءء Ͻ ء receptors were activated (15). Increasing agonist concentration in- LTC4 secretion from primary peritoneal mast cells. , p 0.05 and , p creased the number of cells that responded to carbachol but, even 0.01 indicate significance relative to control. 5260 ALL-OR-NONE CALCIUM INFLUX IN MAST CELLS

2 *

1

relative fluorescence 0 control thap. FIGURE 6. Stimulating muscarinic receptors in rat Control Thapsigargin peritoneal mast cells activates protein kinase C␣ in a nonlinear manner. A, Distribution of protein kinase C␣ in control (left panel) cells and after stimulation with 50 50 µ thapsigargin (2 ␮M) for 4 min. B, Aggregate data from control 1 M carbachol single cell image analysis from several experiments is 40 40 shown. C–F, All-points histograms plotting the extent of 30 30 protein kinase C␣ translocation (normalized to the mean 20 20 response seen in 100 ␮M carbachol; called % Max. re-

frequency (%) 10 frequency (%) 10 sponse) against the frequency of observing such a re- sponse (expressed as %). C shows the control profile, 0 0 0 2550 75 100 0 25 50 75 100 and D–F the corresponding pattern for 1, 10 and 100 ␮M Downloaded from % maximum reponse % maximum reponse carbachol. G, Carbachol concentration is plotted against mean response, with the latter reflecting the responses of all cells. H, The histogram compares signals between 50 10 µ M carbachol 50 100 µ M carbachol control cells and those exposed to 1 and 10 ␮M carba- 40 40 chol after those cells that responded by evoking protein 30 30 kinase C translocation Ն95% that of 100 ␮M carbachol

20 20 http://www.jimmunol.org/ (i.e., x2 sem. from 100%) had been removed from the analysis. Now, the responses are very similar between frequency (%) 10 frequency (%) 10 control, 1 and 10 ␮M carbachol. Images were analyzed 0 0 by drawing small rectangles of the same dimensions 0 25 50 75 100 0 25 50 75 100 (4–6 per cell) across the plasma membrane and then % maximum reponse % maximum reponse computing the fluorescence. Background (taken in re- gions lacking cells) was subtracted from the data. 100 100 by guest on October 1, 2021 50 50 % control response % maximum response 0 0 control 1 10 1 10 100 µ Μ) [carbachol] ( [carbachol] (µ M) at lower agonist concentrations, if a cell responded to carbachol it tration, including a transient Ca2ϩ spike, sinusoidal Ca2ϩ oscilla- ␮ 2ϩ 2ϩ did so by generating maximal ICRAC. With 1 M carbachol, 5 of tions, and a sustained Ca signal. Stimulation in Ca -free so- Ϫ Ϯ 2ϩ 9 cells responded but for the cells that did, ICRAC was 1.9 0.4 lution followed by Ca readmission failed to trigger a consistent pA/pF. With 100 ␮M carbachol, in contrast, 10 of 11 cells re- Ca2ϩ influx signal, with some cells responding by generating Ca2ϩ Ϫ Ϯ 2ϩ sponded and ICRAC had a similar amplitude ( 2.3 0.2 pA/pF; oscillations, others a Ca plateau, and some not responding to Ͼ 2ϩ p 0.1). No ICRAC developed following application of 0.1–0.2 Ca readmission at all. With such variability, we were unable ␮M carbachol (Fig. 4H), and such concentrations failed to evoke to dissect out the Ca2ϩ influx component and hence we turned to detectable ERK activation, arachidonic acid generation, or LTC4 other agonists. secretion (Fig. 3). Stimulation of muscarinic receptors with 100 ␮M carbachol re- sulted in the generation of cytoplasmic Ca2ϩ signals consisting of Muscarinic receptor-evoked responses in acutely isolated rat Ca2ϩ release followed by store-operated Ca2ϩ influx (Fig. 5A), peritoneal mast cells which unlike Ag could be easily separated. Lowering agonist con- The preceding results were obtained using cultured mast cells, rais- centration to 10 or 1 ␮M evoked similar patterns of response, ing the question of whether the findings are of physiological rel- although the percentage of cells responding fell (70% for 1 ␮M evance. To examine this, we isolated peritoneal mast cells from carbachol compared with 99% for 100 ␮M agonist). Muscarinic rats and then designed experiments to see whether varying agonist receptor stimulation also resulted in graded LTC4 secretion (Fig. concentration activated cells in a nonlinear way. 5B). To examine the relationship between agonist concentration We first explored the classical Ag-IgE-FC␧RI pathway, which and cell responsiveness at the single cell level, we used protein ␣ links into the InsP3 signaling cascade. However, responses to Ag kinase C translocation to the plasma membrane as an indicator of are extremely variable in primary mast cells and even a maximal cell activation for two reasons. First, translocation of the kinase is a 2ϩ agonist concentration evokes a diverse pattern of Ca signals. key early step for cPLA2 activation and subsequent LTC4 secretion Indeed, we observed a range of responses to a fixed Ag concen- and hence is of considerable physiological significance. Second, we The Journal of Immunology 5261

were able to quantify protein kinase C␣ movement with confidence A because the mAb we used was very specific. Down-regulation of pro- ATP/0Ca2+ 5mM Ba2+ tein kinase C␣ (following overnight exposure to phorbol ester) re- 1.6 100µ M ATP sulted in complete loss of staining (data not shown). As with RBL 1.4 1µ M ATP cells, stimulation of CRAC channels following exposure to thapsigar- gin resulted in robust migration of protein kinase C␣ to the plasma 1.2 R (356/380)

membrane in primary peritoneal mast cells (Fig. 6A). Images were ∆ 1.0 analyzed on a cell-by-cell basis and aggregate data is summarized in Fig. 6B. We then applied different concentrations of carbachol for four 0.8 0 100 200 300 400 minutes and measured the extent of protein kinase C␣ translocation. time (seconds) Histograms were constructed and the pattern of response between control (nonstimulated) cells (Fig. 6C) and those exposed to 1, 10, and B C 100 ␮M carbachol compared (Fig. 6, D–F). Compared with the con- trol response profile, raising the carbachol concentration increased the 100 100 frequency of observing large responses. Aggregate data from all cells

is plotted as a dose-response curve in Fig. 6G, and includes cells that entry rate (%) entry rate (%) responded and those that did not. When all cells are included, the 2+ 50 2+ 50 relationship is clearly graded. Fig. 6H plots the data in another way. In this study, we removed all those cells that responded by generating 0 0 Downloaded from

maximal protein kinase C␣ translocation (Ն95% of response seen to normalised Ba 246 246 normalised Ba 24 24 1 10 100 1 10 100 100 ␮M carbachol) and then averaged the remaining responses. There [ATP] (µ Μ) [ATP] (µ Μ) was no significant difference between control cells and those exposed FIGURE 7. Stimulation of P2Y receptors with ATP activates Ca2ϩ en- to1or10␮M carbachol. Hence, raising carbachol concentration re- try in an all-or-none manner. A, Cells were stimulated with ATP (1 or 100 cruited more cells but each cell that responded did so maximally. ␮M) in Ca2ϩ-free solution and then Ba2ϩ readmitted. The rate of Ba2ϩ

2ϩ entry was measured and is plotted against ATP concentration in panel B for All-or-none Ca influx following purinoceptor activation in all cells (n Ͼ 25 cells per point). Because not all cells responded to ATP http://www.jimmunol.org/ primary mast cells by generating Ca2ϩ release, C plots Ba2ϩ influx rate against ATP concen- 2ϩ Mast cells express P2Y receptors (24), which couple to phospho- tration only for those cells that responded to ATP by generating Ba entry Ͼ ␮ lipase C thereby generating InsP . ATP modulates the responsive- above resting levels (n 15 cells per point, except for 1 M ATP where 3 only one cell responded by evoking Ba2ϩ influx greater than background ness of mast cells to other stimuli like Ag, increasing the latter’s levels). efficacy in triggering degranulation (24). P2Y receptors desensitize

rapidly however, so that the increase in InsP3 levels fall quickly. Ca2ϩ influx is therefore transient because incoming Ca2ϩ through CRAC channels is taken up into the stores but in the absence of (data not shown). We attribute this to the transient production of by guest on October 1, 2021

InsP3, stores refill rapidly and thus deactivate CRAC channels. We diacylglycerol, due to rapid receptor desensitization. Consistent 2ϩ 2ϩ therefore used Ba as a surrogate for Ca to probe the extent of with this, we did not observe a detectable increase in LTC4 pro- CRAC channel activation. Following stimulation (4 min) with dif- duction even at an ATP concentration of 100 ␮M (data not shown). ferent concentrations of ATP in Ca2ϩ-free solution, we applied Ba2ϩ extracellularly (Fig. 7A) and measured the rate of divalent Discussion cation entry. Fig. 7B plots the rate of Ba2ϩ influx against ATP In immune cells, a major route for Ca2ϩ influx is provided by concentration for all cells exposed to agonist. The relationship is CRAC channels (18, 25), In mast cells, CRAC channel-dependent clearly graded. However, as with muscarinic receptor stimulation, Ca2ϩ influx triggers exocytosis (8) as well as generation of the

not all cells responded to ATP particularly at the lower concen- proinflammatory molecule LTC4 (13). In T lymphocytes, the im- trations. Moreover, even if a cell responded to a low dose of ATP portance of CRAC channels is underscored by the development of (like 1 ␮M in Fig. 7A), Ba2ϩ entry was often barely resolvable and a severe combined immunodeficiency in patients with a mutation similar to the background Ba2ϩ influx seen in the absence of ag- in Orai1 (26), a protein that contributes to formation of the CRAC onist. We therefore plotted background-corrected Ba2ϩ entry only channel pore (26–29). In this study, we have found that stimula- in those cells that generated a Ca2ϩ release signal to ATP. Aggre- tion of two different cell surface receptors in mast cells evokes gate data is summarized in Fig. 7C. Regardless of ATP concen- all-or-none CRAC channel activation. Despite this, graded Ca2ϩ-

tration, if a cell responded it did so in an essentially all-or-none dependent stimulation of ERK, cPLA2, and LTC4 secretion was manner. obtained in populations of RBL and primary mast cells. This co- The overall extent and kinetics of the Ca2ϩ signal is sculpted nundrum was resolved by the finding that responses were graded both by Ca2ϩ release/influx into and Ca2ϩ removal from the cy- with stimulus intensity because increasing agonist concentration toplasm. To see whether different concentrations of ATP affected recruited more cells in the population but each cell that responded 2ϩ 2ϩ the rate of removal of Ca from the cytosol, we measured the did so by generating maximal ICRAC and Ca influx. Whereas 2ϩ 2ϩ ␮ half-time (t1/2) of decay of the initial Ca transient (in Ca -free virtually all cells responded to 100 M carbachol by developing a ␮ ␮ 2ϩ ␣ solution). For 5 M, 100 M, and 1 mM ATP, t1/2 of decay was cytoplasmic Ca rise and protein kinase C translocation to the 26 Ϯ 2, 24 Ϯ 1.6, and 28 Ϯ 1.3 s and these values were not plasma membrane, ϳ40% of the cells did so to 1 ␮M agonist. This significantly different from each other. The amplitude of the Ca2ϩ proportion correlated well with the functional responses: relative ␮ ␮ transients was similar for the different concentrations. Hence, to 100 M carbachoi, 1 M carbachol activated ERK, cPLA2, and 2ϩ ␮ Ca clearance is not altered by ATP over the range 5 M LTC4 to similar extents (39.7, 29.5, and 41.0%, respectively). Ϫ1 mM. Activation of protein kinase C␣ and ERK, which are critical for

We repeatedly failed to see clear translocation of protein kinase LTC4 production in mast cells, are largely independent of even C␣ to the plasma membrane over a range of ATP concentrations large rises in global cytoplasmic Ca2ϩ (data not shown; W. C. Chang, 5262 ALL-OR-NONE CALCIUM INFLUX IN MAST CELLS

C. Nelson, J. Di Capite, V. Halse, A. Parekh, submitted), instead concentration of ATP (1 ␮M) in our experiments (Fig. 7A). De- being tightly coupled to local Ca2ϩ influx through CRAC channels spite clear Ca2ϩ release, cation entry did not occur. Ca2ϩ signals ␮ (12). Activation of the protein kinase C/ERK/cPLA2 cascade is to ATP were graded with stimulus intensity because 1 M ATP therefore accomplished primarily through a rise in subplasmale- evoked only Ca2ϩ release whereas the higher concentrations trig- mmal Ca2ϩ concentration. Because the main determinant of sub- gered both Ca2ϩ release and influx. plasmalemmal Ca2ϩ concentration is CRAC channel activity, What might be the relevance of our findings to mast cell func- which develops in an all-or-none manner macroscopically, then tion? Mast cells are known to secrete strongly and can release their one would expect significant nonlinearity in the Ca2ϩ-dependent entire secretory contents upon stimulation. Hide et al. (44) have activation of an early downstream target with variations in agonist found that ionomycin, a potent activator of CRAC channels concentration, and this is indeed seen in the Ca2ϩ-dependent pro- through its ability to empty stores rapidly (4), triggers all-or-none tein kinase C␣ translocation experiments of Fig. 6. Nonlinearity degranulation in primary rat mast cells. From a physiological per- here would not necessarily result in all-or-none generation of all spective, such a process seems essential because mast cells are

downstream products like arachidonic acid and LTC4 though be- generally found in tissues in small numbers and hence need to cause the intervening metabolic pathways could be subject to reg- secrete extensively to affect their immediate environment. ulation by other factors. Do all-or-none Ca2ϩ signals contribute to responses in other Acknowledgement immune cells? Several studies have reported all-or-none Ca2ϩ re- We thank Dr. Simon Hunt for discussion. sponses in lymphocytes. Interaction of single helper T cells with their physiological ligand, antigenic peptide bound to MHC mol- Disclosures ecules on APCs, resulted in all-or-none Ca2ϩ responses and this The authors have no financial conflict of interest. Downloaded from was independent of Ag concentration (30). As is the case with our results from mast cells, increasing Ag concentration recruited more References 1. Carafoli, E. 2002. Calcium signalling: a tale for all seasons. Proc. Natl. Acad. Sci. T cells in the population to respond, but each cell retained its USA ϩ ϩ 99: 1115–1122. all-or-none Ca2 response. Cytoplasmic Ca2 responses to Ag in 2. Parekh, A. B., and J. W. J. Putney. 2005. Store-operated calcium channels. CTLs are also thought to be an all-or-none process (31). Naive Physiol. Revs. 85: 757–810.

ϩ 2ϩ 3. Clapham, D. E. 1995. Calcium signalling. Cell 80: 259–268. http://www.jimmunol.org/ CD4 T cells generated all-or-none Ca responses following ac- 4. Parekh, A. B., and R. Penner. 1997. Store-operated calcium influx. Physiol. Revs. ϩ tivation by B cells (32). Ca2 -dependent T cell proliferation in 77: 901–930. response to IL 2 is thought to be an all-or-none process (33, 34). 5. Hoth, M., and R. Penner. 1992. Depletion of intracellular calcium stores activates a calcium current in mast cells. Nature 355: 353–356. Similarly, phorbol ester-induced activation of the MAP kinase 6. Zweifach, A., and R. S. Lewis. 1993. Mitogen-regulated Ca2ϩ current of T lym- JNK is an all-or-none event (35). Low concentrations of phorbol phocytes is activated by depletion of intracellular Ca2ϩ stores. Proc. Natl. Acad. Sci. USA 90: 6295–6299. ester evoke full activation of JNK in a fraction of Jurkat T lym- 7. Feske, S., J. Giltnane, R. Dolmetsch, L. M. Staudt, and A. K. Rao, J. Disa, and phocytes but no activation at all in the rest of the population. Sim- X. Yang. 2001. Gene regulation mediated by calcium signals in T lymphocytes. ilarly, all-or-none Ca2ϩ responses have been observed in neutro- Nat. Immunol. 2: 316–324. 8. Artalejo, A. R., J. C. Ellory, and A. B. Parekh. 1998. Ca2ϩ-dependent capacitance phils following stimulation with complement C5a (36) or insoluble increases in rat basophilic leukemia cells following activation of store-operated by guest on October 1, 2021 immune complex (37). Ca2ϩ entry and dialysis with high-Ca2ϩ-containing intracellular solution. Despite all-or-none activation of CRAC channels, cells have the Pflu¨gers Arch. 436: 934–939. 9. Dolmetsch, R. E., K. Xu, and R. S. Lewis. 1998. Calcium oscillations increase the 2ϩ ability to produce graded Ca responses. This can be accom- efficiency and specificity of gene expression. Nature 392: 933–936. plished by varying the membrane potential and therefore electrical 10. Lin, S., K. A. Fagan, K.-X. Li, W. P. W. Shaul, D. M. F. Cooper, and 2ϩ D. M. Rodman. 2000. Sustained endothelial nitric oxide-synthase activation re- driving force for Ca influx or by subsequent inactivation of the quires capacitative Ca2ϩ entry. J. Biol. Chem. 275: 17979–17985. CRAC channels by a variety of intracellular signals (reviewed in 11. Cooper, D. M., J. W. Karpen, K. A. Fagan, and N. E. Mons. 1998. Ca2ϩ-sensitive Ref. 2). Alternatively, nonstore-operated Ca2ϩ channels could be adenylyl cyclases. Adv. Second Messenger Phosphoprotein Res. 32: 23–51. 12. Chang, W. C., and A. B. Parekh. 2004. Close functional coupling between CRAC coexpressed with CRAC channels and gated in a graded manner. channels, arachidonic acid release and leukotriene secretion. J. Biol. Chem. 279: ϩ In addition to store-operated Ca2 influx, neutrophils, T and B 29994–29999. 2ϩ lymphocytes all express nonstore-operated plasmalemmal Ca2ϩ- 13. Chang, W.-C., C. Nelson, and A. B. Parekh. 2006. Ca influx through CRAC channels activates cytosolic phospholipase A2, leukotriene C4 secretion and ex- 2ϩ 2ϩ permeable pathways including Ca -activated Ca -permeable pression of c-fos through ERK-dependent and independent pathways in mast cells. FASEB J. 20: 2381–2383. nonselective channels (38), TRPM2 (39), TRPV6 (40), and InsP3- 2ϩ 14. Putney, J. W. J., and R. R. McKay. 1999. Capacitative calcium entry channels. gated channels (41). Finally, the cytoplasmic Ca signal will be Bioessays ϩ 21: 38–46. shaped not only by the rate of Ca2 entry through CRAC channels 15. Parekh, A. B., A. Fleig, and R. Penner. 1997. The store-operated calcium current 2ϩ Cell but also the rate of Ca removal from the cytoplasm. The latter ICRAC: nonlinear activation by InsP3 and dissociation from calcium release. 89: 973–981. reflects the concerted actions of plasma membrane transporters, 16. Glitsch, M. D., and A. B. Parekh. 2000. Ca2ϩ store dynamics determines the ϩ mitochondrial Ca2 uptake through the uniporter and sequestra- pattern of activation and inactivation of the store-operated Ca2ϩ current ICRAC tion into the stores by SERCA pumps (1), all of which are subject in response to InsP3 in rat basophilic leukaemia cells. J. Physiol. 523: 283–290. 17. Hermosura, M., H. Takeuchi, A. Fleig, A. M. Riley, B. V. Potter, M. Hirata, and to regulation by intracellular signals (42). R. Penner. 2000. InsP4 facilitates store-operated calcium influx by inhibition of In one previous study on human lung-derived mast cells, a low InsP3 5-phosphatase. Nature 408: 735–740. 2ϩ 18. Lewis, R. S. 1999. Store-operated calcium channels. Adv. Second Messenger dose of Ag evoked a smaller cytoplasmic Ca rise than a 20-fold Phosphoprotein Res. ϩ 33: 279–307. higher one (43). It was suggested that Ca2 signals were thus 19. Favata, M., K. Horiuchi, E. J. Manos, A. J. Daulerio, D. A. Stradley, graded with stimulus intensity rather than being all-or-none events. W. S. Feeser, D. E. Van Dyk, W. Pitts, R. Earl, F. Hobbs, et al. 1998. Identifi- 2ϩ cation of a novel inhibitor of mitogen-activated protein kinase kinase. J. Biol. In that study however, the relative contributions of Ca release Chem. 273: 18623–18632. ϩ ϩ and Ca2 influx to the overall Ca2 signal were not dissected out. 20. Bakowski, D., and A. B. Parekh. 2001. Sarcoplasmic/endoplasmic-reticulum- ϩ ϩ 2ϩ Ca2 -ATPase-mediated Ca2 reuptake, and not Ins(1,4,5)P3 receptor inactiva- Because low levels of stimulus intensity can trigger some Ca ϩ ϩ tion, prevents the activation of macroscopic Ca2 release-activated Ca2 current 2ϩ release without subsequent store-operated Ca influx (2, 15), the in the presence of physiological Ca2ϩ buffer in rat basophilic leukaemia-1 cells. submaximal concentration of agonist used (43) might not have Biochem. J. 353: 561–567. 21. Parekh, A. B. 2003. Store-operated calcium entry: dynamic interplay between generated enough InsP3 to ensure stores emptied sufficiently for endoplasmic reticulum, mitochondria and plasma membrane. J. Physiol. 547: CRAC channels to activate. Indeed, this was the case with a low 333–348. The Journal of Immunology 5263

22. Glitsch, M. D., D. Bakowski, and A. B. Parekh. 2002. Store-operated Ca2ϩ entry 34. Smith, K. A. 2004. The quantal theory of how the immune system discriminates depends on mitochondrial Ca2ϩ uptake. EMBO J. 21: 6744–6754. between ‘self and non-self’. Med. Immunol. 3: 1–22. 23. Shuttleworth, T. J., J. L. Thompson, and O. Mignen. 2004. ARC channels: a 35. Bargowski, C. P., J. Besser, C. R. Frey, and J. E. Ferrell, Jr. 2003. The JNK novel pathway for receptor-activated calcium entry. Physiology 19: 355–361. cascade as a biochemical switch in mammalian cells. Curr. Biol. 13: 315–320. 24. Schulman, E. S., M. C. Glaum, T. Post, Y. Wang, D. G. Raible, J. Mohanty, 36. Metzner, B., J. Elsner, G. Dobos, E. Kownatzki, F. Parlow, I. Schraufstaetter, and J. H. Butterfield, and A. Pelleg. 1999. ATP modulates anti-IGE-induced release J. Norgauer. 1994. Ca2ϩi transients and actin polymerization in human neutro- of histamine from human lung mast cells. Am. J. Respir. Cell Mol. Biol. 20: phils under stimulation with GRO ␣ and complement fragment C5a. Agents Ac- 530–537. tion 42: 101–106. 25. Gwack, Y., S. Feske, S. Srikanth, P. G. Hogan, and A. Rao. 2007. Signalling to 37. Brunkhorst, B. A., K. G. Lazzar, G. Strohmeier, G. Weil, and E. R. Simons. 1991. transcription: store-operated Ca entry and NFAT activation in lymphocytes. Cell Calcium changes in immune complex-stimulated human neutrophils. J. Biol. Calcium 42: 145–156. Chem. 266: 13035–13043. 26. Feske, S., Y. Gwack, M. Prakriya, S. Srikanth, S.-V. Puppel, B. Tanasa, 38. Von Tscharner, V., B. Prod’hom, M. Baggiolini, and H. Reuter. 1986. Ion chan- P. G. Hogan, R. S. Lewis, M. Daly, and A. Rao. 2006. A mutation in Orai1 causes nels in human neutrophils activated by a rise in free cytosolic calcium concen- immune deficiency by abrogating CRAC channel function. Nature 441: 179–185. tration. Nature 324: 369–372. 27. Prakriya, M., S. Feske, Y. Gwack, S. Srikanth, A. Rao, and P. G. Hogan. 2006. 39. Beck, A., M. Kolisek, L. A. Bagley, A. Fleig, and R. Penner. 2006. Nicotinic acid Orai1 is an essential pore subunit of the CRAC channel. Nature 443: 230–233. adenine dinucleotide phosphate and cyclic ADP-ribose regulate TRPM2 channels 28. Yeromin, A. V., S. L. Zhang, W. Jiang, Y. Yu, O. Safrina, and M. D. Cahalan. in T lymphocytes. FASEB J. 20: 962–964. 2006. Molecular identification of the CRAC channel by altered ion selectivity in a mutant of Orai. Nature 443: 226–229. 40. Cui, J., J. S. Bian, A. Kagan, and T. V. McDonald. 2002. CaT1 contributes to the 29. Vig, M., A. Beck, J. M. Billingsley, A. Lis, S. Parvez, C. Peinelt, D. L. Koomoa, store-operated calcium current in Jurkat T-lymphocytes. J. Biol. Chem. 277: J. Soboloff, D. L. Gill, A. Fleig, et al. 2006. CRACM1 multimers form the 47175–47183. ion-selective pore of the CRAC channel. Curr. Biol. 16: 2073–2079. 41. Vazquez, G., B. J. Wedel, G. S. Bird, and J. W. J. Putney. 2002. An inositol 30. Agrawal, N. G., and L. J. L. 1995. Calcium response of helper T lymphocytes to trisphosphate receptor-dependent cation entry pathway in DT40 B lymphocytes. antigen-presenting cells in a single-cell assay. Biophys. J. 69: 1178–1190. EMBO J. 21: 4531–4538. 31. Bachmann, M. F., S. Mariathasan, D. Bouchard, D. E. Speiser, and P. S. Ohashi. 42. Pozzan, T., R. Rizzuto, P. Volpe, and J. Meldolesi. 1994. Molecular and cellular 1997. Four types of calcium signals in naive CD8ϩ cytotoxic T cells after stim- physiology of intracellular calcium stores. Physiol. Revs. 74: 595–636. ulation with T cell agonists, partial agonists and antagonists. Eur. J. Immunol. 27: 43. Macglashan, D. J. 1989. Single-cell analysis of calcium changes in human lung Downloaded from 3414–3419. mast cells: graded vs All-or-nothing elevationsafter IgE-mediated stimulation. 32. Delon, J., N. Bercovici, G. Raposo, R. Liblau, and A. Trautmann. 1998. Antigen- J. Cell Biol. 109: 123–134. dependent and -independent Ca responses triggered in T cells by dendritic cells 44. Hide, I., J. P. Bennett, A. Pizzey, G. Boonen, D. Bar-Sagi, B. D. Gomperts, and compared with B cells. J. Exp. Med. 188: 1473–1484. P. E. R. Tatham. 1993. Degranulation of individual mast cells in response to 33. Smith, K. A. 1995. Cell growth signal transduction is quantal. Ann. NY Acad. Sci. calcium and guanine nucleotides: an all-or-none event. J. Cell Biol. 123: 766: 263–271. 585–593. http://www.jimmunol.org/ by guest on October 1, 2021