Organelle Selection Determines Agonist-specific Ca2� Signals inPancreatic Acinar and � Cells*

Received for publication, October 8, 2003, and in revised form, November 19, 2003Published, JBC Papers in Press, December 2, 2003, DOI 10.1074/jbc.M311088200

Michiko Yamasaki‡§, Roser Masgrau‡§¶, Anthony J. Morgan‡, Grant C. Churchill‡, Sandip Patel�,Stephen J. H. Ashcroft**, and Antony Galione‡

From the ‡Department of Pharmacology, University of Oxford, Mansfield Road, Oxford OX1 3QT, the �Department ofPhysiology, University College London, Gower Street, London WC1E 6BT, and **Oxford Centre forDiabetes, Endocrinology and Metabolism, Churchill Hospital, Oxford OX3 7LJ, United Kingdom

How different extracellular stimuli can evoke differ-ent spatiotemporal Ca2� signals is uncertain. We haveelucidated a novel paradigm whereby different agonistsuse different Ca2�-storing organelles (“organelle selec-tion”) to evoke unique responses. Some agonists selectthe endoplasmic reticulum (ER), and others select lyso-some-related (acidic) organelles, evoking spatial Ca2�

responses that mirror the organellar distribution. Inpancreatic acinar cells, acetylcholine and bombesin ex-clusively select the ER Ca2� store, whereas cholecysto-kinin additionally recruits a lysosome-related organelle.Similarly, in a pancreatic � cell line MIN6, acetylcholineselects only the ER, whereas glucose mobilizes Ca2�

from a lysosome-related organelle. We also show that thekey to organelle selection is the agonist-specific cou-pling messenger(s) such that the ER is selected by re-cruitment of inositol 1,4,5-trisphosphate (or cADP-ri-bose), whereas lysosome-related organelles are selectedby NAADP.

Increases in the Ca2� concentration ([Ca2�]i)1 represent a

ubiquitous transduction mechanism in response to stimuli asdiverse as fertilization, neurotransmitters, hormones, and bloodglucose (1). In turn, Ca2�-binding proteins responsible for decod-ing these Ca2� signals promote an array of cellular responsesfrom cell division to cell death, notably including secretion (1).However, what remains perplexing is that not all stimuli thatincrease Ca2� elicit the same physiological response, implyingdifferences in their [Ca2�]i handling. Some early indicationsarose from single cell and subcellular [Ca2�]i studies which re-vealed that each stimulus evoked its own unique Ca2� responsein time and space (exemplified by agonist-specific “Ca2� signa-tures”) (2). But the mechanism(s) underlying stimulus-specificCa2� signaling is poorly understood.

It has become clear that a primary factor governing agonistspecificity is the release of Ca2� from intracellular stores, aprocess that encompasses multiple channel families regulatedby the second messengers inositol 1,4,5-trisphosphate (IP3),cyclic ADP-ribose (cADPR), and nicotinic acid adenine dinucle-otide phosphate (NAADP). By differential recruitment of sec-ond messenger complements, different agonists could, in prin-ciple, elicit different Ca2� signals, and several cellular systemshave indeed been shown to exploit messenger diversity, e.g.pancreatic acinar cells (3).

However, the use of different messengers per se is not suffi-cient to evoke different signals; only if these systems are non-equivalent and differ in their properties, regulation, or spatialdistribution will this hold. In spatial terms, IP3 and cADPRmobilize the endoplasmic reticular (ER) Ca2� store in nearly allcell types (1, 4), although an outstanding issue has been theidentity and location of the NAADP-sensitive store. Only re-cently has this been characterized in sea urchin eggs as alysosome-related organelle (5), with its mammalian counter-part remaining elusive beyond an isolated report that secretoryvesicles support NAADP-induced Ca2� release in permeabi-lized cells (6).

Given the uncertainties surrounding the role and propertiesof different potential Ca2� stores, we therefore show for thefirst time that different agonists generate their specific signalsby signaling through Ca2� mobilization from different intracel-lular stores (organelle selection). This is determined by theirsecond messenger complements, with NAADP-linked agonistscoupling to lysosome-related organelles in mammalian cellsand those linked to IP3/cADPR coupling to ER pools.

EXPERIMENTAL PROCEDURES

Cell Preparation—To obtain pancreatic acinar cells, pancreatawere excised from male CD1 mice 8–10 weeks old, and small clustersof pancreatic acinar cells were prepared by collagenase digestion asdescribed previously (7). MIN6 cells were cultured in Dulbecco’smodified Eagle’s medium (25 mM glucose) supplemented with 10%fetal calf serum, 2 mM glutamine, 100 units/ml penicillin, 100 �g/mlstreptomycin, and 50 �M �-mercaptoethanol equilibrated with 5%CO2 and 95% air at 37 °C. Twenty hours before each experiment, cellswere placed in low glucose Dulbecco’s modified Eagle’s medium (5 mM

glucose).Ca2� Imaging—Both acinar and MIN6 cells were seeded onto poly-

lysine-coated number 1 glass coverslips and loaded with 1–5 �M fura-2acetoxymethyl ester (fura-2/AM) for 60 min at room temperature. Aci-nar cells and MIN6 cells were maintained in buffer of the followingcompositions: for acinar cells (in mM), 140 NaCl, 4.7 KCl, 1.1 MgCl2, 1CaCl2, 10 Hepes, 10 glucose, pH 7.2; for MIN6 cells (in mM), 119 NaCl,4.75 KCl, 5 NaHCO3, 1.2 MgSO4, 1.18 KH2PO4, 20 Hepes, 2.54 CaCl2,and 2.8 glucose, pH 7.4. After the loading period, cells were washed andimaged immediately. Coverslips were mounted in a static chamber(Harvard Apparatus), on an inverted Zeiss 35 Axiovert microscope, andimaged with a conventional epifluorescence system, using Metafluor

* This work was supported by The Wellcome Trust through a seniorfellowship (to A. G.), a Research Career Development fellowship (toS. P.), a Prize Studentship (to M. Y.), and a project grant (to A. G. andS. J. H. A.). The costs of publication of this article were defrayed in partby the payment of page charges. This article must therefore be herebymarked “advertisement” in accordance with 18 U.S.C. Section 1734solely to indicate this fact.

§ Both authors contributed equally to this work.¶ To whom correspondence should be addressed: Dept. of Pharmacol-

ogy, University of Oxford, Mansfield Road, Oxford OX1 3QT, UK. Tel.:44 01865 271 606; Fax: 44 01865 271 853; E-mail: roser.masgrau-juanola@pharmacology.oxford.ac.uk.

1 The abbreviations used are: [Ca2�]i, intracellular Ca2� concentra-tion; ER, endoplasmic reticulum; IP3, inositol 1,4,5-trisphosphate;cADPR, cyclic ADP-ribose; NAADP, nicotinic acid adenine dinucleotidephosphate; GPN, glycyl-phenylalanine 2-naphthylamide; OGBD, Ore-gon Green BAPTA-1 dextran.

THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 279, No. 8, Issue of February 20, pp. 7234–7240, 2004© 2004 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S.A.

This paper is available on line at http://www.jbc.org7234

by on February 1, 2008

ww

w.jbc.org

Dow

nloaded from

software (Universal Imaging). Cells were excited alternately with 340and 380 nm light (emission 510 nm), and ratio images of clusters wererecorded every 4–5 s, using a 12-bit CCD camera (MicroMax, PrincetonInstruments). All experiments were conducted at room temperature foracinar cells and at 37 °C for MIN6 cells.

Imaging Lysosomes—Acidic organelles in both cell types were labeledby incubating cells with 50 nM Lysotracker Red for 20 min at roomtemperature. Labeling was visualized after 20–40 min of removingexcess dye using a Leica TCS NT laser scanning confocal microscope(excitation 568 nm, emission �590 nm).

Flash Photolysis—Acinar cells and MIN6 cells were pressure-micro-injected (Femtojet, Eppendorf) with Oregon Green BAPTA-1 dextran(OGBD, final concentration of 20 and 5 �M, respectively) with cagedcompounds. In acinar cells the Ca2�-sensitive dye was imaged (excita-tion 490 nm, emission 530 nm) as mentioned, and the caged compoundswere photolysed with an XF-10 arc lamp (HI-TECH Scientific, the

ultraviolet flash efficiency was 0.5–1%). On the other hand, MIN6 cellswere imaged by laser-scanning confocal microscopy (Leica TCS NT),and caged compounds were photolysed with an ultraviolet laser (effi-ciency of uncaging of �50%). Images were processed using Metamorphsoftware (Universal Imaging). Ca2� concentration is given as the ratioF/Fo where Fo is the fluorescence before stimulation, and F the fluores-cence at a given time. Changes in Ca2� concentration are given asincreases in the mentioned ratio (�F/Fo).

Statistical Analysis—Data are presented as means � S.E. Statisticalsignificance were evaluated by paired Student’s t test and, for multiplecomparisons, analysis of variance followed by Fisher’s Least SignificantDifference test (Statview, Abacus Concepts).

Materials—Caged IP3 was from Calbiochem. Caged cADPR, Fura-2/AM, OGBD, and Lysotracker Red were from Molecular Probes, and col-lagenase was from Worthington. Caged NAADP was synthesized essen-tially as described previously (8). All other reagents were from Sigma.

FIG. 1. Effect of eliminating Ca2�

stores on agonist-dependent Ca2�

signals in pancreatic acinar cells. Inpancreatic acinar cells, 3 �M bafilomycinA1 differentially inhibited Ca2� oscilla-tions (a–f). Responses to 0.5–2.5 pM cho-lecystokinin were profoundly inhibited bypre- (a, 70/98 cells) or postincubation (b,72/117 cells), whereas neither 25–50 nM

acetylcholine- (c, 34/36; d, 35/47 cells) nor1.0–2.5 pM bombesin (e, 29/29; f, 52/67cells)-induced responses were signifi-cantly affected. Similarly, the effect of 50�M GPN (g–m) was agonist-dependent;cholecystokinin-induced oscillations wereblocked (g, 70/77; h, 52/78 cells), and therewas little effect upon acetylcholine (i, 56/60; j, 21/24 cells) or bombesin responses(k, 43/43; l, 29/46 cells). In cells wherecholecystokinin-induced oscillations wereblocked, further addition of acetylcholinerestored spiking (m, 21/24 cells). Preincu-bation with 1 �M thapsigargin completelyblocked the response to 2 pM cholecysto-kinin (n, 23/23 cells).

Organelle Selection in Calcium Signaling 7235

by on February 1, 2008

ww

w.jbc.org

Dow

nloaded from

RESULTS

Of all the mammalian cell models used for investigatingagonist-specific Ca2� signals, one of the most extensively stud-ied non-excitable cells is the pancreatic acinar cell. The gastro-intestinal peptides cholecystokinin and bombesin, and the neu-rotransmitter acetylcholine, evoke subtly different Ca2�

signatures resulting in differential control of fluid secretion,exocytosis, and trophic effects (3). We first addressed whetherthese agonists differentially recruited lysosome-related or en-doplasmic reticular Ca2� stores by using inhibitors to selec-tively abrogate Ca2� storage in each organelle; bafilomycin A1inhibits the vacuolar H�-ATPase responsible for the protongradient that drives lysosomal Ca2� uptake (9), whereas thap-sigargin directly blocks the ER Ca2�-ATPase (SERCA) (10).

Whether or not bafilomycin A1 had an effect upon Ca2�

oscillations depended upon the agonist used. Cholecystokinin-induced Ca2� oscillations were profoundly inhibited, whetherbafilomycin A1 was added before (Fig. 1a) or after (Fig. 1b) theagonist. By contrast, bafilomycin A1 had very little effect uponacetylcholine- or bombesin-stimulated oscillations (Fig. 1, c–f).To confirm that bafilomycin A1 was indeed acting at a lyso-some-related organelle, we also eliminated such stores withglycyl-phenylalanine 2-naphthylamide (GPN), a substrate oflysosomal cathepsin C whose cleavage results in osmotic lysis(5, 11). Essentially identical results were obtained with GPN,notably the selective block of the response to cholecystokininover acetylcholine or bombesin (Fig. 1, g–l). Moreover, in cellswhere cholecystokinin-induced oscillations were blocked byGPN, subsequent addition of acetylcholine could rescue Ca2�

spiking, highlighting the specificity of the response (Fig. 1m).The data show that, of the agonists tested, cholecystokinin isunique in recruiting a lysosome-related organelle that canfunction as a Ca2� store.

The ER, on the other hand, is the established Ca2� reservoirfor many G protein-coupled receptors, including those for ace-tylcholine and bombesin. In pancreatic acinar cells, these stim-uli are well documented to release Ca2� from the ER using bothIP3 and ryanodine receptors (12, 13). Furthermore, cholecysto-kinin-induced Ca2� signals were also confirmed to derive fromthe ER as evidenced by the marked inhibition by thapsigargin(Fig. 1n). Together, the evidence supports cholecystokinin re-cruiting both lysosomes and ER, whereas acetylcholine andbombesin only target the ER in order to generate [Ca2�]isignals.

An obvious issue is whether this differential organellar re-cruitment is specific to acinar cells or a universal blueprint forother mammalian cell types and stimuli. In choosing anothermodel, the pancreatic � cell, we opted for a system with verydifferent properties from the acinar cell, the new one being anexcitable cell in which Ca2� signals are elicited by nutrients aswell as by G protein-coupled agonists (14, 15). In the � cell line,MIN6 (16), we compared three different stimuli, glucose, ace-tylcholine, and K�, with regard to their relative sensitivities tobafilomycin A1 or thapsigargin. Mechanistically, glucose me-tabolism generates intracellular signals that culminate in acomplex interplay between Ca2� influx and Ca2� release fromintracellular stores (17); muscarinic acetylcholine receptors areG protein-coupled to phospholipase C (15), whereas high K�

depolarizes the plasma membrane to induce voltage-operatedCa2� entry (18, 19)

In MIN6 cells, the sensitivity of Ca2� responses to bafilomy-cin A1 was, like acinar cells, highly dependent upon the stim-ulus. Remarkably glucose responses were profoundly inhibitedby a preincubation with bafilomycin A1 (Fig. 2, a and b),whereas neither acetylcholine nor K� stimulation was affected(Fig. 2, d and e and g and h). That the acidic stores of theglucose response were lysosome-related was confirmed by theinhibition by GPN (data not shown). Moreover, these resultsconfirm the specificity of bafilomycin and GPN because neitherinteracts with the IP3-calcium release pathway (acetylcholine)nor calcium influx (K�). Remarkably, the effects of interferingwith ER stores with thapsigargin were almost the mirror ofthose with bafilomycin A1. Glucose-induced Ca2� signals werenot inhibited by ER depletion (Fig. 2c) but rather appeared tobe potentiated because thapsigargin greatly reduced the lagphase and eliminated the initial fall in basal [Ca2�]i. Similarly,responses induced by K� were also slightly potentiated (Fig.2i), attesting to the role of the ER as a Ca2� sink in � cells (20).On the other hand, the ER appeared to play a major role duringacetylcholine-induced Ca2� mobilization because responseswere completely eliminated by thapsigargin (Fig. 2f). Takentogether, the data in this cell type support a model of thereciprocal recruitment of different organelles where metabolicactivation is heavily reliant upon lysosome-related Ca2� stores,contrasting with an exclusive ER role in response toneurotransmitter.

Next we provide a mechanism to couple particular extracel-lular stimuli to specific intracellular organelles with the appro-priate fidelity. Interestingly, our data show an absolute corre-lation between those stimuli that recruit lysosome-relatedorganelles and those known to utilize NAADP as a Ca2�-mo-bilizing messenger (21), i.e. cholecystokinin in pancreatic aci-nar cells (22) and glucose in � cells (16). The agonists that failto require acidic stores are well known to couple to IP3 and/orryanodine receptors that mobilize ER Ca2� stores (13, 23, 24).We therefore tested whether NAADP was the unique link toacidic stores, with IP3 and/or cADPR showing a preference fornon-acidic (ER) stores.

First, in pancreatic acinar cells, photorelease of IP3 or

FIG. 2. Effect of eliminating Ca2� stores on agonist-dependentCa2� signals in pancreatic � cells. In MIN6 cells, 20 mM glucoseinduced a sustained plateau with a mean increase in the fluorescenceratio (340/380) of 25.7 � 1.5% in controls cells (n � 140) (a). Thisplateau response was profoundly inhibited by 2 �M bafilomycin A1 (b,6.0 � 1.0, n � 90, p � 0.001) but not affected by 1 �M thapsigargin (c,27.8 � 2.8%, n � 90, p � 0.4). By contrast, the 100 �M acetylcholine-induced increase of the fluorescence ratio (d, 24.5 � 1.6%, n � 159) wasunaffected by bafilomycin A1 (e, 25.7 � 1.8%, n � 71, p � 0.6) buteliminated by thapsigargin (f, 0.8 � 0.3, n � 70, p � 0.001). Finally, 40mM KCl induced a peak response in control cells of 62.8 � 2.6% (g, n �157) that was unaffected by bafilomycin A1 (h, 62.8 � 2.2, n � 70, p �0.9) but was increased by thapsigargin (i, 86.0 � 5.7, n � 60, p � 0.001).

Organelle Selection in Calcium Signaling7236

by on February 1, 2008

ww

w.jbc.org

Dow

nloaded from

cADPR from their caged precursors evoked robust, monotonicCa2� transients in control cells (Fig. 3, a and b) comparable inmagnitude to the subsequent response to acetylcholine. Elim-ination of lysosomal Ca2� storage by preincubation with GPNhad no effect upon the magnitude of the [Ca2�]i rise in responseto either IP3 or cADPR (or the following acetylcholine re-sponses) (Fig. 3, d and e). By contrast, GPN profoundly inhib-ited the Ca2� oscillations following uncaging of NAADP (Fig. 3,c and f). Note that the NAADP-induced Ca2� spikes are ini-tially small and became progressively amplified by Ca2�-in-duced Ca2� release mechanism through the recruitment of IP3

and ryanodine receptors (3, 22, 25–27). In agreement with theeffects of GPN, bafilomycin A1 displayed an identical and se-lective block of NAADP-induced over IP3-induced responses(data not shown, n � 4). The very fact that the Ca2� responsesto second messengers alone are inhibited strongly suggeststhat bafilomycin A1 and GPN are working downstream ofNAADP, and not at an upstream element of the signalingcascade initiated by agonist. We conclude that only NAADPcouples to the lysosome-related Ca2� store in acinar cells,whereas cADPR and IP3 couple to the ER.

Our hypothesis is also supported in experiments with the �cell system; in these cells NAADP-dependent Ca2� releaseoccurs via specific binding sites (16). Just as glucose and ace-tylcholine manifest a reciprocal dependence upon lysosomesand ER, so NAADP and IP3 displayed this mutually exclusivepattern. NAADP photorelease stimulated a Ca2� increase thatwas inhibited by bafilomycin A1 but not by thapsigargin (Fig. 4,a, c and e), whereas the converse occurred when photoreleasingIP3 (Fig. 4, b, d, and f). Thapsigargin-induced depletion ofstores profoundly inhibited IP3 transients, which were other-

wise insensitive to bafilomycin A1. Once again, NAADP selectslysosome-related stores, whereas IP3 predominantly selects theER. Hence, the data suggest that lysosome-related stores cou-ple via NAADP to particular extracellular stimuli (16), chole-cystokinin and glucose, respectively. Furthermore, by recruit-ing NAADP, agonists select a novel Ca2� store with distinctproperties, distribution, and ramifications from the ER (pri-marily the domain of IP3 and cADPR).

To confirm that the distribution of lysosomal Ca2� stores in-deed has a bearing upon the spatial profile of the Ca2� response,we compared NAADP-mediated Ca2� release with the distribu-tion of the organelle in live cells. Lysotracker Red labeling wasmarkedly polarized and confined to the apical region of pancre-atic acinar cells reminiscent of secretory vesicle staining (Fig. 5,a and b) (28), which are highly related if not overlapping or-ganelles (29). The observed pattern with Lysotracker Red faith-fully reflected lysosomal staining as confirmed by the eliminationof the punctate fluorescence by treatment with either GPN (Fig.5a) or bafilomycin A1 (Fig. 5b). Interestingly, the organelle dis-tribution coincided with the ensuing NAADP-evoked small Ca2�

oscillations, which do not fully recruit Ca2�-induced Ca2� release(22, 30) and were confined to the apical pole in acinar cells (Fig.5c). In contrast, in � cells Lysotracker Red comprised staining ofbright punctate bodies, albeit superimposed on a diffuse fluores-cent background. This punctate staining was uniformly dis-persed throughout the cytoplasm but excluded from the nucleus(Fig. 6, a–c). As in acinar cells, the Lysotracker Red granularstaining was eliminated by GPN (Fig. 6a) or bafilomycin A1 (Fig.6b) but not by thapsigargin (Fig. 6c). Moreover, in MIN6 cells, theresponse to NAADP in � cells was essentially global (Fig. 6d).Specifically, the close spatial correlation of Ca2� release and the

FIG. 3. Effect of eliminating Ca2�

stores upon the response to photoly-sis of caged second messengers inacinar cells. Pancreatic acinar cells weremicroinjected with the fluorescent Ca2�

indicator, OGBD, plus the appropriatecaged precursor of various Ca2�-releasingagents and photolysis initiated by expo-sure to UV light where indicated. Controlresponses to photorelease of IP3 (approx-imated final intracellular concentration of300 nM) (a, n � 9) or cADPR (approxi-mated final intracellular concentration of1 �M) (b, n � 13) were comparable withthat for the subsequent 10 �M acetylcho-line response, whereas NAADP (approxi-mated final intracellular concentration of100 nM) responses often showed complexCa2� oscillations (c, n � 8). 20–30 min ofpreincubation with 50 �M GPN had littleeffect upon either IP3 (d and g, n � 17) orcADPR (e and h, n � 15), whereasNAADP responses were abolished (f and i,n � 6).

Organelle Selection in Calcium Signaling 7237

by on February 1, 2008

ww

w.jbc.org

Dow

nloaded from

acidic store distribution strongly imply that there is a substan-tive rationale for using different Ca2�-storing organelles.

DISCUSSION

In this present report we provide evidence for a novel mech-anism that explains how agonists evoke their own characteris-tic Ca2� response, that of organelle selection (Fig. 7). Accordingto this model, different agonists (even within the same cell)mobilize Ca2� in different ways in time and space by couplingto (selecting) different Ca2�-storing organelles with their ownunique properties and distribution. Moreover, we reveal that agiven organelle couples via a particular second messenger,such that an agonist selects organelles by recruiting the appro-priate messenger complement. Although our recent results insea urchin egg (5) provided a framework (that different mes-sengers mobilize Ca2� from different organelles), the differen-tial recruitment of these organelles by different agonists hasnever been shown.

In essence, irrespective of cell type, agonists can be dividedinto those that recruit lysosome-related organelles (cholecysto-kinin and glucose) and those that do not (acetylcholine andbombesin). Such a conclusion is drawn from pharmacologicalstudies using two mechanistically and chemically distinct in-hibitors of lysosomal function, bafilomycin A1 and GPN.

Whether added before or during Ca2� oscillations, these agentsselectively inhibited responses to the former pair of agonists,while having little or no effect upon the latter. We are confidentthat these agents act specifically upon acidic Ca2� stores be-cause of the following: (a) they did not indiscriminately inhibitall agonists, as evidenced by their lack of effect upon the ER-coupled acetylcholine and bombesin; (b) their site of action islikely downstream of second messengers themselves as indi-cated by photolysis studies (see below) and therefore not anupstream signal; (c) they do not block depolarization-inducedCa2� entry; and (d) they eliminate Lysotracker Red staining.

We have proceeded to show that cell surface receptors coupleto an intracellular store type by virtue of a characteristic se-lecting messenger, i.e. NAADP was unique in coupling to lyso-some-related organelles, whereas IP3/cADPR coupled to theER. Not only does the published messenger profiles of theagonists support our hypothesis (3, 15), but the sensitivity ofthe second messengers themselves to various store inhibitorsshowed an absolute agreement. Therefore, sea urchin eggs arenot anomalous in having acidic stores sensitive to NAADP butrather are vindicated as an excellent model system to studymammalian Ca2� signaling. Moreover, our data are of interestin the light of a previous study (6) in permeabilized MIN6 cellssuggesting that NAADP releases Ca2� from secretory vesicles,themselves an acidic organelle. It should be noted that our datadiffer from that by Mitchell et al. (6) because (a) we have usedintact cells; (b) we show agonist (glucose) coupling to acidicstores via NAADP; and (c) in our hands NAADP predominantlyreleases Ca2� from a bafilomycin A1-sensitive and lysosomal-related store.

At first sight, it might appear contradictory that selectiveelimination of acidic Ca2� stores has such a marked effect uponcholecystokinin when clearly there is an additional ER compo-nent (Fig. 1) (22). More surprisingly, glucose-stimulated Ca2�

signals in � cells also manifest a profound sensitivity to acidicstore blockade when there ought to be a substantial residualCa2� entry component (17, 18) (and perhaps an ER component)(31). Although there is currently no complete mechanistic ex-planation for this absolute dependence, it has been empiricallydetermined that desensitization of the NAADP receptor by itsown ligand ablates both the cholecystokinin- as well as theglucose-induced Ca2� signals (16, 22). Therefore, the effects ofbafilomycin A1 and GPN remain entirely consistent with theblockade of the NAADP store. For the acinar cells, it has beensuggested that the ER is essential to amplify NAADP-inducedCa2� release via Ca2�-induced Ca2� release at the IP3 or ry-anodine receptors (3, 25). It is currently less clear how NAADPmight affect Ca2� entry in � cells, but in sea urchin eggs a linkbetween NAADP signaling and voltage-gated Ca2� channelshas been suggested (32).

The agonist-specific recruitment of different organelles alsohas ramifications in the spatial domain. It has been clearlyshown that the apical pole of pancreatic acinar cells has a highdensity of zymogen granules (33), with only small fingers of ERpenetrating into this region, and that the ER is highly concen-trated in the basolateral part of the cell (34). The distribution ofacidic vesicles may display a more cell-specific pattern; cer-tainly for pancreatic acinar cells, intense Lysotracker Redstaining was confined to the apical pole, whereas MIN6 cellsappeared to display a more uniform staining. Supporting ourmodel that these are Ca2� stores, this pattern mirrored thesubsequent Ca2� responses that were evoked upon uncagingNAADP; in pancreatic acinar cells the region of highestNAADP sensitivity was confirmed as the apical pole (30),whereas the MIN6 response was essentially uniform. It shouldbe noted, however, that a previous study in permeabilized cells

FIG. 4. Effect of eliminating Ca2� stores upon the response tophotolysis of caged second messengers in pancreatic � cells. InOGBD-injected MIN6 � cells, control responses to flash photolysis ofcaged NAADP (approximated final intracellular concentration of 100nM) (a, n � 58) or caged IP3 (approximated final intracellular concen-tration of 1 �M) (b, n � 38) were similar. 1 �M thapsigargin had no effectupon NAADP responses (c and g, n � 18) but eliminated IP3 responses(d and h, n � 24). Conversely, 2 �M bafilomycin A1 inhibited NAADP (eand g, n � 31) but had no effect upon IP3 (f and h, n � 24).

Organelle Selection in Calcium Signaling7238

by on February 1, 2008

ww

w.jbc.org

Dow

nloaded from

FIG. 6. Comparison of lysosome-related store distribution and NAADP-induced Ca2� signals in � cells. The distribution of 50 nM

Lysotracker Red fluorescence in pancreatic � cells was detected throughout the cells (except nucleus) (a–c, left panels) and eliminated by 50 �M

GPN (a, n � 6) and 2 �M bafilomycin (b, n � 4) but not by thapsigargin (c, n � 4). Photolysis of caged NAADP in � cells induced a global Ca2�

response, and identical localized responses are obtained all over the cell (d, n � 58).

FIG. 5. Comparison of lysosome-re-lated store distribution and NAADP-induced Ca2� signals in acinar cells.Micrographs showing vital fluorescencestaining of pancreatic acinar cells labeledwith 50 nM Lysotracker Red (a and b, leftside) plus corresponding bright field im-ages of pancreatic acinar cells (a and b,left side, lower panels). The predomi-nantly apical fluorescence was dramati-cally reduced by pretreatment with 50 �M

GPN (a, right-hand upper panel, n � 4) or3 �M bafilomycin A1 (b, right-hand upperpanel, n � 3). Cells were arbitrarilytreated for 25–40 min but fluorescencebegan to fall immediately upon their ap-plication. The local Ca2� response to pho-tolysis of caged NAADP in a single acinarcell is also confined to the apical pole (c,blue trace), but not to the basal pole (c, redtrace) (n � 8).

Organelle Selection in Calcium Signaling 7239

by on February 1, 2008

ww

w.jbc.org

Dow

nloaded from

described the basolateral pole as the region of highest NAADPsensitivity (35). We cannot currently rationalize this apparentdiscrepancy, but we suggest methodological differences (e.g.permeabilization, uneven distribution of compartmentalizedCa2� indicator).

In summary, we hypothesize the specific agonist-inducedCa2� mobilization from lysosomal-like acidic organelles as auniversal concept in mammalian cells. The use of distinct,non-contiguous stores would allow the Ca2� levels therein to beregulated independently, e.g. by altering ATPase (Ca2� or H�)expression or activity. Indeed, the use of non-ER stores willhave other advantages such as avoiding plummeting Ca2�

levels in the ER lumen which affect nascent protein synthesis(36) and protein phosphorylation (37) as well as increasing cellviability by minimizing ER and mitochondria overload andhence apoptosis (38). Conversely, altering the Ca2� content ofacidic stores may affect processes such as secretory vesiclefusion (39), membrane repair (29), and proteolysis (40). Fur-thermore, if the NAADP-sensitive Ca2� store is indeed anacidic secretory vesicle (6) or secretory lysosome (29), Ca2� isdelivered precisely where required to evoke exocytosis of zymo-gens (41), ATP (28), or insulin (6, 42), depending upon cell type,and provides another potential target for treatment of diabetes.

Acknowledgments—We thank Clive Garnham (University of Oxford)for technical assistance and Justyn M. Thomas (University of Oxford)and Georgina Berridge (University of Oxford) for useful discussions.

REFERENCES

1. Berridge, M., Lipp, P., and Bootman, M. (2000) Nat. Rev. Mol. Cell Biol. 1,11–21

2. Prentki, M., Glennon, M. C., Thomas, A. P., Morris, R. L., Matschinsky, F. M.,and Corkey, B. E. (1988) J. Biol. Chem. 263, 11044–11147

3. Cancela, J. M. (2001) Annu. Rev. Physiol. 63, 99–1174. Galione, A., and Churchill, G. C. (2002) Cell Calcium 32, 343–3545. Churchill, G. C., Okada, Y., Thomas, J. M., Armando, A. Genazzani, Patel, S.,

and Galione, A. (2002) Cell 111, 703–7086. Mitchell, K. J., Lai, F. A., and Rutter, G. A. (2003) J. Biol. Chem. 278,

11057–110647. Thorn, P., and Petersen, O. H. (1992) J. Gen. Physiol. 100, 11–258. Lee, H. C., Aarhus, R., Gee, K. R., and Kestner, T. (1997) J. Biol. Chem. 272,

4172–41789. Christensen, K. A., Myers, J. T., and Swanson, J. A. (2002) J. Cell Sci. 115,

599–60710. Treiman, M., Caspersen, C., and Christensen, S. B. (1998) Trends Pharmacol.

Sci. 19, 131–13511. Jadot, M., Colmant, C., Wattiaux-De Coninck, S., and Wattiaux, R. (1984)

Biochem. J. 219, 965–97012. Fogarty, K. E., Kidd, J. F., Turner, A., Skepper, J. N., Carmichael, J., and

Thorn, P. (2000) J. Biol. Chem. 275, 22487–2249413. Mogami, H., Tepikin, A. V., and Petersen, O. H. (1998) EMBO J. 17, 435–44214. Aizawa, T., Komatsu, M., Asanuma, N., Sato, Y., and Sharp, G. W. (1998)

Trends Pharmacol. Sci. 19, 496–49915. Gilon, P., and Henquin, J. C. (2001) Endocr. Rev. 22, 565–60416. Masgrau, R., Churchill, G. C., Morgan, A. J., Ashcroft, S. J., and Galione, A.

(2003) Curr. Biol. 13, 247–25117. Ashcroft, S. J. (2000) J. Membr. Biol. 176, 187–20618. Dawson, C. M., Atwater, I., and Rojas, E. (1984) Q. J. Exp. Physiol. 69,

819–83019. Lemmens, R., Larsson, O., Berggren, P. O., and Islam, M. S. (2001) J. Biol.

Chem. 276, 9971–997720. Arredouani, A., Henquin, J. C., and Gilon, P. (2002) Am. J. Physiol. 282,

E982–E99121. Lee, H. C. (2003) Curr. Biol. 13, R186–R18822. Cancela, J. M., Churchill, G. C., and Galione, A. (1999) Nature 398, 74–7623. Burdakov, D., Cancela, J. M., and Petersen, O. H. (2001) Cell Calcium 29,

211–21624. Burdakov, D., and Galione, A. (2000) Curr. Biol. 10, 993–99625. Patel, S., Churchill, G. C., and Galione, A. (2001) Trends Biochem. Sci. 26,

482–48926. Churchill, G. C., and Galione, A. (2000) J. Biol. Chem. 275, 38687–3869227. Santella, L., Kyozuka, K., Genazzani, A. A., De Riso, L., and Carafoli, E. (2000)

J. Biol. Chem. 275, 8301–830628. Sorensen, C. E., and Novak, I. (2001) J. Biol. Chem. 276, 32925–3293229. Blott, E. J., and Griffiths, G. M. (2002) Nat. Rev. Mol. Cell Biol. 3, 122–13130. Cancela, J. M., Van Coppenolle, F., Galione, A., Tepikin, A. V., and Petersen,

O. H. (2002) EMBO J. 21, 909–91931. Takasawa, S., Akiyama, T., Nata, K., Kuroki, M., Tohgo, A., Noguchi, N.,

Kobayashi, S., Kato, I., Katada, T., and Okamoto, H. (1998) J. Biol. Chem.273, 2497–2500

32. Churchill, G. C., O’Neill, J. S., Masgrau, R., Patel, S., Thomas, J. M., Genaz-zani, A. A., and Galione, A. (2003) Curr. Biol. 13, 125–128

33. Gerasimenko, O. V., Gerasimenko, J. V., Rizzuto, R. R., Treiman, M., Tepikin,A. V., and Petersen, O. H. (2002) Cell Calcium 32, 261–268

34. Petersen, O. H., Tepikin, A., and Pavk, M. K. (2001) Trends Neurosci. 24,271–276

35. Krause, E., Gobel, A., and Schulz, I. (2002) J. Biol. Chem. 277, 11696–1170236. Meldolesi, J., and Pozzan, T. (1998) Trends Biochem. Sci. 23, 10–1437. Sargeant, P., Farndale, R. W., and Sage, S. O. (1994) Exp. Physiol. 79, 269–27238. Hajnoczky, G., Davies, E., and Madesh, M. (2003) Biochem. Biophys. Res.

Commun. 304, 445–45439. Renstrom, E., Ivarsson, R., and Shears, S. B. (2002) J. Biol. Chem. 277,

26717–2672040. Larbaud, D., Balage, M., Taillandier, D., Combaret, L., Grizard, J., and Attaix,

D. (2001) Clin. Sci. (Lond.) 101, 551–55841. Titievsky, A. V., Takeo, T., Tepikin, A. V., and Petersen, O. H. (1996) Pfluegers

Arch. 432, 938–94042. Martin, F., Ribas, J., and Soria, B. (1997) Biochem. Biophys. Res. Commun.

235, 465–468

FIG. 7. A minimal model for achieving stimulus specificitythrough organelle selection. Different external stimuli are coupledto different organelles via the stimulus-specific recruitment of non-promiscuous second messenger complements. In pancreatic acinar cells,cholecystokinin (CCK) selects both lysosome-related organelles and en-doplasmic reticulum by recruiting NAADP, cADPR, and IP3 respec-tively. By contrast, acetylcholine (ACh) and bombesin only select the ERusing IP3/cADPR. In � cells, glucose stimulates NAADP synthesis torelease Ca2� from lysosome-related organelles. Although our data donot favor a major role for the endoplasmic reticulum via IP3 (or cADPR),we cannot formally exclude their involvement (31). Acetylcholine in �cells uses only the ER/IP3 pathway.

Organelle Selection in Calcium Signaling7240

by on February 1, 2008

ww

w.jbc.org

Dow

nloaded from