Cell Calcium 39 (2006) 57–63

Mitochondria play a critical role in shaping the exocytotic responseof rat pancreatic acinar cells

Paul Thomasa,∗, Tanya Bagrijb,1, Manuel Campos-Toimilb,2, J. Michael Edwardsonb

a Henry Wellcome Laboratory for Cell Imaging, School of Biological Sciences, University of East Anglia, Norwich NR4 7TJ, UKb Department of Pharmacology, University of Cambridge, Tennis Court Road, Cambridge CB2 1PD, UK

Received 12 July 2005; received in revised form 3 September 2005; accepted 9 September 2005Available online 20 October 2005

Abstract

We have previously demonstrated [M. Campos-Toimil, T. Bagrij, J.M. Edwardson, P. Thomas, Two modes of secretion in pancreaticacinar cells: involvement of phosphatidylinositol 3-kinase and regulation by capacitative Ca2+ entry, Curr. Biol. 12 (2002) 211–215] thatin rat pancreatic acinar cells, Gd3+-sensitive Ca2+ entry is instrumental in governing which second messenger pathways control secretoryactivity. However, in those studies, we were unable to demonstrate a significant increase in cytoplasmic [Ca2+] during agonist application asa to resolvett dbs a with thel hata tidylinositol3©

K

1

f[tlIlrr

R

C

rhis

ilityuits

tra-be

thepre-ist,

tativeways

tivitye tost

0d

result of this entry pathway. In the present study, we combined pharmacology with ratiometric imaging of fura-2 fluorescencehis issue. We found that 2�M Gd3+ significantly inhibits store-mediated Ca2+ entry. Furthermore, both the protonophore, CCCP (5�M) andhe mitochondrial Ca2+-uptake blocker, RU360 (10�M), led to an enhancement of the plateau phase of the biphasic Ca2+ response inducey acetylcholine (1�M). This enhancement was completely abolished by Gd3+; and as has been previously shown for Gd3+, RU360 led to awitch to a wortmannin-sensitive form of exocytosis. Using MitoTracker Red staining we found a close association of mitochondriateral plasma membrane. We propose that in rat pancreatic acinar cells, capacitative Ca2+ entry is targeted directly to mitochondria; and ts a result of Ca2+ uptake, these mitochondria release “third” messengers which both enhance exocytosis and suppress phospha-kinase-dependent secretion.2005 Elsevier Ltd. All rights reserved.

eywords: Calcium influx; Signalling cross-talk; Secretion; Exocytosis

. Introduction

The release of Ca2+ from intracellular stores is essentialor zymogen granule exocytosis in pancreatic acinar cells1]; nevertheless, it has been recognised for several yearshat agonists which elicit the release of Ca2+ from intracellu-ar stores also stimulate the influx of extracellular Ca2+ [2,3].ndeed, we have recently demonstrated that entry of extracel-ular Ca2+ through a Gd3+-sensitive pathway plays a criticalole in regulating secretion evoked by acetylcholine (ACh) inat pancreatic acinar cells[4].

∗ Corresponding author. Tel.: +44 1603 592196; fax: +44 1603 592250.E-mail address: p.thomas@uea.ac.uk (P. Thomas).

1 Present address: Abcam Ltd., 332 Cambridge Science Park, Miltonoad, Cambridge CB4 0TP, UK2 Present address: Departamento de Farmacoloxia, Facultade de Farmacia,ampus Universitario Sur, 15782 Santiago de Compostela, Spain

The pathways by which Ca2+ enters the cell during ofollowing the release of intracellular Ca2+ stores are botvaried and ill-defined[2,3,5]. However, a general patternemerging in which two pathways with distinct permeaband temporal-activation properties form the major condfor Ca2+ influx in response to agonists which empty incellular Ca2+ stores[6–8]. These pathways appear toreciprocally regulated by arachidonic acid[9,10] via nitricoxide [11], in such a way that whilst agonist is presentnitric-oxide-stimulated, non-capacitative entry pathwaydominates; correspondingly, following removal of agonthe capacitative pathway is activated and the non-capacipathway suppressed. Pharmacologically these two pathcan be differentiated based upon their relative sensito lanthanides: non-capacitative influx being insensitivlanthanides (<100�M) and capacitative influx being almocompletely inhibited by 1–2�M Gd3+ [8].

143-4160/$ – see front matter © 2005 Elsevier Ltd. All rights reserved.oi:10.1016/j.ceca.2005.09.007

58 P. Thomas et al. / Cell Calcium 39 (2006) 57–63

This model of reciprocal control of the Ca2+ entry path-ways is consistent with findings in pancreatic acinar cells.During application of ACh, Ca2+ influx occurs for the mostpart through a pathway that is insensitive to lanthanides[4,6];whereas the later influx activated by thapsigargin-mediatedstore emptying is blocked by lanthanum[6]. Nevertheless,despite the fact that the lanthanide-sensitive pathway con-tributes very little to the cytosolic Ca2+ signal measured withfura-2, the capacitative pathway does appear to be active dur-ing ACh stimulation, and has profound effects on secretoryactivity [4].

A close relationship between store-mediated Ca2+ entryand mitochondrial Ca2+ buffering has recently become appar-ent in several systems[12,13]. We wondered whether, inacinar cells, our inability to measure a significant increasein cytosolic [Ca2+] ([Ca2+]i ) due to the Gd3+-sensitive path-way was because the Ca2+ entering via this pathway wasimmediately captured by mitochondria. To investigate thispossibility we employed both a mitochondrial uncoupler(CCCP) and a specific blocker of mitochondrial Ca2+ uptake(RU360) to disrupt mitochondrial Ca2+ buffering during ago-nist application. We also investigated the cellular localisationof mitochondria using MitoTracker Red CMXRos (MTRed).Our results suggest a model in which capacitative Ca2+ entryoccurs at the lateral plasma membrane of rat acinar cellsand is immediately taken up by neighbouring mitochondria.T bothe nnin-s

2

2

wleyr reedi revi-o u-b MKNt ids,2 nd0w dt wicew t oni

2

dw swE at

37◦C. The cells were then rinsed twice with Ca2+-free IM(no EGTA) before placing on the microscope. The cellswere perfused (∼1 ml/min) with Ca2+-free IM while pairsof fluorescence images at 340 and 380 nm were acquiredevery 3 s. After 2 min of perfusion, thapsigargin (1�M) wasapplied (3 min) to the cells from a glass micropipette (2–5�mdiameter) using a PicoSpritzer II (General Valve Corp., Fair-field, NJ, USA). After a further 5 min, the perfusion solu-tion was replaced with regular IM either with or without2�M GdCl3.

For experiments with ACh, fura-2-loaded cells werelocally perfused with 1�M ACh (±CCCP) and changes in[Ca2+]i measured as described previously[14]. For experi-ments with RU360 (10�M), the drug was included in thefura-2/AM loading solution, such that cells were pre-treatedfor 30 min (final DMSO concentration <0.3%). Calibra-tion of the fura-2 signal was carried out using solutionsof known [Ca2+] in 20�m thick microslides (VitroComInc., Mountain Lakes, NJ, USA) as described previously[15].

2.3. Measurement of exocytosis

For experiments with RU360 (10�M) and/or wortman-nin (100 nM), cells were pre-treated at 37◦C for 30 min prior

and%.pli--sly

singm,

ereein inacedss

A,TA,K).Bio-iumlarum

ible,

he mitochondria then release “third” messengers thatnhance exocytosis and prevent activation of a wortmaensitive form of secretion.

. Materials and methods

.1. Acinar cell preparation

Five- to seven-week-old male Wistar or Sprague–Daats (175–225 g) were obtained from the departmental bng colony. Acinar cells were obtained as described pusly [14]. All experiments were carried out in an incation medium (IM) consisting of 120 mM NaCl, 5 mCl, 2 mM CaCl2, 1 mM MgCl2, 1 mM NaH2PO4, 1 mMaHCO3, 11 mM glucose, 4 mM D-myo-inositol, 2 mM glu-

amine, 2% (v/v) Basal Medium Eagle (BME) amino ac5 mM HEPES/NaOH, pH 7.2 plus 0.5 mg/ml BSA a.5 mg/ml Soybean Trypsin Inhibitor (SBTI). Ca2+-free IMas prepared by replacing CaCl2 with MgCl2. Cells attache

o modified, glass-bottomed Petri dishes were rinsed tith IM, and then either used for experimentation or kep

ce until needed.

.2. Measurement of intracellular calcium

To measure store-mediated Ca2+ entry, cells were loadeith fura-2/AM as described previously[14]. Loaded cellere rinsed twice with Ca2+-free IM containing 250�MGTA and incubated in this same medium for 10 min

-

to stimulation. Both drugs were dissolved in DMSOthen added to IM such that the final [DMSO] was <0.4RU360 was also included in the medium during ACh apcations. Exocytosis was measured at 37± 2◦C using a timedifferential analysis of brightfield recordings as previoudescribed[14]. Images were acquired and analysed uAQM and Lucida software (Kinetic Imaging, NottinghaUK).

2.4. Localisation of mitochondria

For confocal microscopy, cells attached to dishes wincubated for 15 min at 37◦C with 10 nM MTRed, the dywas removed and the cells incubated for a further 15 mdye-free IM. The medium was then removed and replwith IM containing 5�M AM1-43 and imaged on a ZeiLSM510 META confocal microscope.

2.5. Materials

Acetylcholine chloride, BME amino acids, HEPES, BScollagenase type IA, DMSO, Pluronic F-127, CCCP, EGpolyethyleneimine and SBTI were from Sigma (Poole, UThapsigargin and RU360 were purchased from Mercksciences (Nottingham, UK). Fura-2/AM ester and potasssalt and MitoTracker Red CMXRos were from MolecuProbes (Eugene, OR, USA). AM1-43 was from Bioti(Hayward, CA, USA).

All general chemicals were from Sigma; where possonly cell-culture-tested chemicals were used.

P. Thomas et al. / Cell Calcium 39 (2006) 57–63 59

Fig. 1. Effect of Gd3+ on Ca2+ entry stimulated by emptying intracel-lular stores. The intracellular [Ca2+] ([Ca]i ) responses of acinar cells tostore emptying caused by application of 1�M thapsigargin (3 min) inCa2+-free medium, followed by Ca2+ restoration (5 min) either in the pres-ence (©) or absence (�) of 2�M Gd3+; mean± S.E.M., n = 18 and 23,respectively.

3. Results

3.1. Effect of Gd3+ on store-mediated Ca2+ entry

Our previous studies on rat acinar cells had been unableto detect an increase in [Ca2+]i during ACh applicationthat was due to Gd3+-sensitive entry; nevertheless, removalof extracellular Ca2+ or addition of Gd3+ had profoundeffects both on the extent and the characteristics of secre-tion [4]. Previous studies of capacitative Ca2+ entry inacinar cells had only demonstrated its sensitivity to highconcentrations of lanthanides (≥100�M) [6,16]. We,therefore, set out to determine whether the low concentration(2�M) of Gd3+ that we had used in our studies did indeedblock capacitative Ca2+ entry in rat acinar cells.Fig. 1clearly shows a large increase in [Ca2+]i when Ca2+ isreapplied following the emptying of intracellular calciumstores with thapsigargin (control). The figure also demon-strates that this increase in [Ca2+]i is markedly reducedwhen Ca2+ is readmitted in the presence of 2�M Gd3+

(Fig. 1; +Gd3+).

3.2. Effect of CCCP and RU360 onacetylcholine-induced changes in [Ca2+]i

It has been suggested that the Gd3+-sensitive capacita-tive entry pathway may be largely suppressed during agonistapplication[10,11]; but considering the effects of Gd3+ onexocytosis[4], we wondered whether the Ca2+ entering viathis pathway during stimulation of acinar cells with AChwas rapidly buffered by mitochondria positioned close toCa2+ entry sites[12,13,17]. To investigate this possibilitywe carried out experiments in which acinar cells were stimu-lated in the presence of CCCP, a mitochondrial uncoupler.Application of 1�M ACh (a maximal physiological dose[14]) to acinar cells evoked a classic “peak-plateau” responseand, as we have shown previously[4], the addition of 2�MGd3+ to the medium had no apparent effect on this response(Fig. 2A). When CCCP (5�M) was included with the agonistthere was a significant increase in the magnitude of both thepeak and the plateau phases (�; Fig. 2B). Interestingly, when2�M Gd3+ was included in the bath, the enhancement of theplateau phase by CCCP was almost completely inhibited (©;Fig. 2B). In contrast, Gd3+ had no effect on the enhancementof the early part of the response. As previously observed byothers (see[16]), CCCP actually caused an increase in [Ca2+]iin the absence of agonist (Wheeler and Thomas, unpub-lished observations). Because of this complicating factor wed byb er-mc aup ementb CCP,h letelya

andt ulus.Ct

F change nso ( ) o a]r the p ,r n (). (C)p S.E.M d forc

ig. 2. Effect of disrupting mitochondrial Ca2+ buffering on ACh-inducedf 1�M ACh (thick bar beneath traces) in the presence (©) or absenceesponses to 6 min applications of 1�M ACh + 5�M CCCP (thick bar) inespectively. The controls from panel (A) are replotted for comparisoresence 10�M RU360 either with (©) or without (�) 2�M Gd3+; means±omparison ( ).

ecided to try to confirm the results obtained with CCCPlocking mitochondrial Ca2+ uptake with the membrane peant inhibitor of the Ca2+ uniporter, RU360[18]. Fig. 2C

learly shows that RU360 (10�M) only enhances the platehase of the response, suggesting that the peak-enhancy CCCP was agonist-independent. As observed with Cowever, the potentiation of the plateau phase was compbolished by addition of Gd3+ to the medium (©; Fig. 2C).

Another difference between the effects of CCCPhose of RU360 was observed on removal of the stimCCP caused a marked inhibition of Ca2+ clearance from

he cytosol, whereas RU360 did not (cf.Fig. 2B and C). This

s in intracellular [Ca2+] ([Ca]i ). (A) The [Ca]i responses to 6 min applicatiof 2�M Gd3+; means± S.E.M.,n = 26 and 33, respectively. (B) The [Ci

resence (©) or absence (�) of 2�M Gd3+; means± S.E.M.,n = 21 and 32The [Ca]i responses to 6 min applications of 1�M ACh (thick bar) in the.,n = 47 and 13, respectively. The controls from panel (A) are replotte

60 P. Thomas et al. / Cell Calcium 39 (2006) 57–63

difference in the effects of the two drugs might indicate thatCa2+ clearance doesn’t require a functioning mitochondrialuniporter, but is dependent upon a trans-membrane pH gradi-ent (across either the plasma membrane or the membrane ofan intracellular Ca2+-sequestering organelle). Alternatively,CCCP, present for several minutes, might inhibit cellular ATPproduction, and thereby indirectly reduce the activity of Ca2+-ATPases, which normally mediate Ca2+ extrusion from thecytosol.

3.3. Effect of RU360 on secretory activity

In our previous work, we had demonstrated that removalof extracellular Ca2+, or inclusion of Gd3+, both reducedexocytotic activity and sensitised the secretory apparatus towortmannin[4]. Our interpretation of these results was thatCa2+ entering the cell via Gd3+-sensitive channels playeda role both in potentiating exocytosis and also in regulat-ing which second messenger pathways controlled secretoryactivity. The results shown inFig. 2 suggest that the Ca2+

entering the cell through these channels is immediately takenup by mitochondria. Such a model would preclude a directeffect of the incoming Ca2+ on the secretory apparatus, andraise the possibility that Ca2+ uptake into mitochondria is aprerequisite for the effects of Ca2+ influx on secretion. To testt ecre-ts bite is int uslyw ther nMw -p , but

Fig. 3. The effect of wortmannin on secretion evoked by ACh. Exocyto-sis was measured as the number of events evoked by a 6 min application of1�M ACh (controls, left black bar;n = 40). Some cells were pre-treated with100 nM wortmannin for 30 min before stimulation (controls, left grey bar;n = 29). Other cells were incubated with 10�M RU360 either with (+RU360,right grey bar;n = 26) or without wortmannin-pre-treatment (+RU360, rightblack bar; n = 24). Significantly different (** p < 0.05) from the controls(−wortmannin) when compared using one-way analysis of variance fol-lowed by Dunnett’st-test for multiple comparisons with a single control.

Table 1Comparison of results shown inFig. 3with previous findings

Treatment Exocytosis (as percent of control− wortmannin)

−Wortmannin +Wortmannin

Control 100 127a

Calcium-free 74 41+Gd3+ (2�M) 68 47+RU360 (10�M) 74 43

a Average of results described here and those obtained previously[4].

unlike Gd3+ or Ca2+ removal, RU360 raises [Ca2+]i (seeFig. 2C). In turn, this increase might have a small positiveeffect on exocytosis, counteracting the tendency of RU360 toinhibit exocytosis.

F eld image and a confocal fluorescence image, respectively, of a cluster of five rat acinarc the plasma membrane green). Scale bar = 10�m.

his hypothesis we examined the effects of RU360 on sion as measured using time-differential analysis[14,19]. Ashown inFig. 3, RU360 alone does not significantly inhixocytosis, although numerically the extent of exocytoshe presence of RU360 is similar to that observed previoith Gd3+ or Ca2+ removal. Further, in all three cases,

emaining exocytosis is significantly inhibited by 100ortmannin (seeTable 1for comparison with previouslyublished results). It should be noted that, like CCCP

ig. 4. Localisation of mitochondria. Panels (A) and (B) show a brightfiells stained with MTRed and AM1-43 (mitochondria stained red and

P. Thomas et al. / Cell Calcium 39 (2006) 57–63 61

3.4. Localisation of mitochondria

Studies on mouse pancreatic acinar cells[20] have demon-strated that mitochondria mainly form a belt around the apicaldomain; however, some mitochondria are also found closeto the baso-lateral plasma membrane[17]. We wonderedwhether the latter observation was also true of rat acinarcells. Live acinar cells were stained using MTRed[21] andco-stained with the styryl membrane-tracking dye AM1-43 (afixable form of the dye FM1-43[22]). Confocal imaging con-firmed that in small acini, there were numerous mitochondria(red inFig. 4B) in close contact with the plasma membrane(green inFig. 4B), and that these contacts were particularlyabundant on the lateral membranes (arrowheads inFig. 4B).

4. Discussion

Previously, we have shown that secretion in rat pancreaticacinar cells under normal physiological stimulation (≤1�MACh) is mainly Ca2+-driven, but on supramaximal (≥10�MACh) stimulation, the control of exocytosis switches to a path-way that involves phosphatidylinositol 3-kinase (PI 3-K)[4].We also demonstrated that this switch in control could bebrought about by blocking the entry of Ca2+ through Gd3+-sensitive channels. The results presented here now reveal thatt ito-c

weenc[ Cae rsts ings,i suredc venie ere itw f thedu chon-d hoseo i-n eC Thisg -r ontie

al-c nti aMcec h

by controlling the amount of Ca2+ passing through channels[12,23,33], and by restricting the diffusion of Ca2+ throughthe cytoplasm[17,20]. In these ways, mitochondria indi-rectly modulate Ca2+-dependent processes such as exocytosis[34–36]. Nevertheless, whether mitochondria play a moreactive role in cell signalling[37,38]is still a matter of debate[39,40].

Our results would suggest that, as well as buffering Ca2+,mitochondria actively participate in intracellular signal prop-agation in rat pancreatic acinar cells. The effects of RU360 onexocytosis imply that mitochondrial Ca2+ uptake both poten-tiates secretion and inhibits PI 3-K-dependent exocytosis.One of the major effects of mitochondrial Ca2+ uptake isto increase oxidative phosphorylation, and so generate ATPand NADH [29]. Consequently, one simple way in whichmitochondria might potentiate (or perhaps, maintain) exo-cytosis would be to supply the exocytotic machinery withsufficient energy. Indeed, the long tubular nature of somemitochondria has led to the proposal that they might behavelike power cables, capable of transmitting energy from onepart of the cell to another[41]. We have generally foundmitochondria to be excluded from the granular region of thecell; nevertheless, the diffusion of ATP from mitochondriaat the apical portion of the lateral membrane to the sites ofexocytosis could account for the potentiating effect of mito-chondrial Ca2+ uptake on secretory activity. On the otherh rtieso PI3d thint

somee mito-c ose-i d-i es ds inarc

).R ownN ls,c imu-l besb -t ofe sted.

ini nd-i f PI3 edS inm cies(a via

he Ca2+ coming through these channels must first enter mhondria to have its effect on secretory activity.

Previous studies have shown a close relationship betapacitative Ca2+ entry and mitochondrial Ca2+ buffering12,23]. These studies have demonstrated a decrease in2+

ntry on treatment with mitochondrial uncouplers. At fiight our results would appear to contradict these find.e., CCCP and RU360 lead to an increase in the meaalcium signal. However, it should be borne in mind that ef these drugs lead to a decrease in Ca2+ entry, the Ca2+ thatnters the cell has nowhere to go but into the cytosol whill be measured by fura-2. In contrast, in the absence orugs, Ca2+ influx may be greater, but this Ca2+ will remainnmeasured because it is rapidly sequestered by mitoria. Indeed, our results are in good agreement with tf Camello-Almaraz et al.[16] who showed in mouse acar cells that the cytosolic Ca2+ signal due to capacitativa2+ entry was increased by mitochondrial uncouplers.roup also demonstrated that when they used Ba2+ as a surogate for Ca2+, mitochondrial uncouplers had no effecthe Ba2+ signal. This finding is not surprising as Ba2+ uptakento mitochondria is very poor (Ca2+ > Sr2+ � Mn2+ > Ba2+),specially in the presence of K+ and Mg2+ [24–26].

Mitochondria play significant roles in intracellular cium signalling[27–29]. For the most part, this involvemes restricted to the ability of mitochondria to sequester C2+.

itochondria have been shown to buffer Ca2+ entering theytosol either from intracellular stores[30,31], or from thextracellular space[12,23,32]. By sequestering Ca2+, mito-hondria are able to shape intracellular Ca2+ transients bot

and, it seems unlikely that the energy-producing propef mitochondria are involved in the suppression of the-K pathway. Alternative schemes might involve the Ca2+-ependent production of other signalling molecules wi

he mitochondria.One such candidate molecule is glutamate. There is

vidence that glutamate can be produced by respiringhondria, and can act to maintain the slow phase of glucnduced insulin secretion[37]. Recently, however, these finngs have been called into question[39,40]. It remains to been whether glutamate might provide the link between G3+-ensitive Ca2+ entry and exocytosis in rat pancreatic acells.

Another potential mediator might be nitric oxide (NOecently, mitochondria have been shown to have theirO synthase (NOS)[42]. Furthermore, in endothelial celalcium uptake by mitochondria has been shown to state production of NO[43], and such synthesis appears topecifically coupled to capacitative Ca2+ entry[44]. NO haseen proposed to both stimulate[45] and inhibit[46] secre

ory activity; so just like glutamate, its role in potentiationxocytosis in acinar cells needs to be experimentally te

Despite these observations, a potential role for NOnhibiting the PI 3-K pathway is suggested by recent fings that in insulin resistance a downstream target o-K (Akt/protein kinase B) is inactivated by NO-induc-nitrosylation [47]. It is thought that nitrosative proteodifications are brought about by reactive nitrogen spe

RNS, e.g., nitrosonium ion, NO+, or peroxynitrite, OONO−)s opposed to NO itself which is thought to act

62 P. Thomas et al. / Cell Calcium 39 (2006) 57–63

cGMP-dependent mechanisms. One system for the gener-ation of peroxynitrite is the oxidation of NO by reactiveoxygen species (ROS) such as superoxide; and, it is inter-esting to note that an increase in electron transport can leadto an increase in production of ROS by mitochondria[48]. Ofcourse, uptake of Ca2+ into mitochondria is a potent stimulusto the TCA cycle, and consequently electron transport[29].Indeed, stimulation of pancreatic acinar cells with anothersecretagogue, cholecystokinin, has been shown to increasemitochondrial ROS production[49]. This suggests a model inwhich capacitative Ca2+ entry, leading to mitochondrial Ca2+

uptake, stimulates the production of both NO and ROS withinmitochondria. The consequent generation of RNS may thenlead to theS-nitrosylation of Akt, and the inhibition of PI 3-K-dependent signalling. Interestingly, it has also been shownthatS-nitrosylation ofN-ethylmaleimide-sensitive fusion fac-tor (or NSF) can enhance exocytosis[50]; thus, generation ofRNS might also account for the potentiating effect of mito-chondrial Ca2+ uptake on secretion.

Because of the lability of ROS and RNS, and the high con-centrations of RNS required for Akt inhibition[47], it wouldseem that mitochondria must be close to Akt/PI 3-K. In ratpancreatic acinar cells, it has been suggested that secreta-gogue receptors are concentrated in the apical portion of thelateral membranes[51]. The results shown inFig. 4 wouldsuggest the possibility that in these cells mitochondria arec pledr plext g., PI3 to-c orr

weh agues[ a inm an-u tioni fectoi ears,t in thel aci-n realf

cialr atica ellesc or isa gm e PI3 natet aysd pho-l -K t

supramaximal stimulation led to signalling pathway switch-ing; thus, the identification of the mitochondrially-derivedmolecules that govern this switching may provide targetsfor pharmacological intervention in pancreatitis, a diseasethat is frequently associated with supramaximal stimulation[52,53].

Acknowledgements

We thank Prof. Alan Dawson for critical reading of themanuscript. TB and MC-T were supported by Grant No.8/C11044 from the Biotechnology and Biological SciencesResearch Council to PT and JME. PT would particularly liketo thank Prof. George Duncan and Prof. Dylan Edwards fortheir support.

References

[1] D.C. Metz, R.J. Patto, J.E. Mrozinski Jr., R.T. Jensen, R.J. Turner,J.D. Gardner, Thapsigargin defines the roles of cellular calcium insecretagogue-stimulated enzyme secretion from pancreatic acini, J.Biol. Chem. 267 (1992) 20620–20629.

[2] J.W. Putney Jr., R.R. McKay, Capacitative calcium entry channels,Bioessays 21 (1999) 38–46.

[3] G.J. Barritt, Receptor-activated Ca2+ inflow in animal cells: a variety

wohos-

siol.

l acti-ion08.cita-.

wayns of

go-iated89.

[ tion

[ the-48.

[ of137

[ al-lial

[ diescells,

lustered close to the postulated sites of G-protein-coueceptors, and perhaps form part of a signalling comhat also includes receptors, downstream effectors (e.-K/Akt) and Ca2+ entry channels. Further immunocyhemical examination will be required to either confirmeject this hypothesis.

With regards to the localisation of mitochondria thatave observed in rat acinar cells, Petersen and colle

17,20] have elegantly demonstrated that mitochondriouse cells form a distinctive “firewall” enclosing the grlar region. We observed little evidence for such localisa

n rat cells (data not shown). Likewise, RU360 has little efn the early part of the [Ca2+]i response (seeFig. 2C) that

s mainly due to release from intracellular stores. It appherefore, that there may be subtle species differencesocalisation of mitochondria in rat and mouse pancreaticar cells. It remains to be seen whether these lead to

unctional differences.In summary, we propose that mitochondria play a cru

ole in signalling from muscarinic receptors in rat pancrecinar cells. We believe the localisation of these organlose to sites of capacitative Ca2+ entry enable them tapidly sequester incoming Ca2+. We would suggest that thccumulation of Ca2+ leads to the production of signallinolecules that can both facilitate secretion and inhibit th-K pathway. As a result, mitochondria are able to co-ordi

he relative activation of two intracellular signalling pathwownstream of muscarinic receptors: one involving phos

ipase C�/inositol trisphosphate/Ca2+ and the other, PI 3/Akt/protein kinase B. Our previous results[4] showed tha

of pathways tailored to meet different intracellular Ca2+ signallingrequirements, Biochem. J. 337 (1999) 153–169.

[4] M. Campos-Toimil, T. Bagrij, J.M. Edwardson, P. Thomas, Tmodes of secretion in pancreatic acinar cells: involvement of pphatidylinositol 3-kinase and regulation by capacitative Ca2+ entry,Curr. Biol. 12 (2002) 211–215.

[5] A.B. Parekh, R. Penner, Store depletion and calcium influx, PhyRev. 77 (1997) 901–930.

[6] C. Camello, J.A. Pariente, G.M. Salido, P.J. Camello, Sequentiavation of different Ca2+ entry pathways upon cholinergic stimulatin mouse pancreatic acinar cells, J. Physiol. 516 (1999) 399–4

[7] T.J. Shuttleworth, J.L. Thompson, Discriminating between capative and arachidonate-activated Ca2+ entry pathways, J. Biol. Chem274 (1999) 31174–31178.

[8] L.M. Broad, T.R. Cannon, C.W. Taylor, A non-capacitative pathactivated by arachidonic acid is the major Ca2+ entry mechanism irat A7r5 smooth muscle cells stimulated with low concentrationvasopressin, J. Physiol. 517 (1999) 121–134.

[9] D. Luo, L.M. Broad, G.S.J. Bird, J.W. Putney Jr., Mutual antanism of calcium entry by capacitative and arachidonic acid-medcalcium entry pathways, J. Biol. Chem. 276 (2001) 20186–201

10] O. Mignen, J.L. Thompson, T.J. Shuttleworth, Reciprocal regulaof capacitative and arachidonate-regulated noncapacitative Ca2+ entrypathways, J. Biol. Chem. 276 (2001) 35676–35683.

11] Z. Moneer, J.L. Dyer, C.W. Taylor, Nitric oxide co-ordinatesactivities of the capacitative and non-capacitative Ca2+-entry pathways regulated by vasopressin, Biochem. J. 370 (2003) 439–4

12] M. Hoth, C.M. Fanger, R.S. Lewis, Mitochondrial regulationstore-operated calcium signaling in T lymphocytes, J. Cell Biol.(1997) 633–648.

13] A.M. Lawrie, R. Rizzuto, T. Pozzan, A.W. Simpson, A role for ccium influx in the regulation of mitochondrial calcium in endothecells, J. Biol. Chem. 271 (1996) 10753–10759.

14] M. Campos-Toimil, J.M. Edwardson, P. Thomas, Real-time stuof zymogen granule exocytosis in intact rat pancreatic acinarJ. Physiol. 528 (2000) 317–326.

P. Thomas et al. / Cell Calcium 39 (2006) 57–63 63

[15] P. Thomas, P.L. Mellon, J.L. Turgeon, D.W. Waring, The L�T2clonal gonadotrope: a model for single cell studies of endocrinecell secretion, Endocrinology 137 (1996) 2979–2989.

[16] C. Camello-Almaraz, G.M. Salido, J.A. Pariente, P.J. Camello, Roleof mitochondria in Ca2+ oscillations and shape of Ca2+ signals inpancreatic acinar cells, Biochem. Pharmacol. 63 (2002) 283–292.

[17] M.K. Park, M.C. Ashby, G. Erdemli, O.H. Petersen, A.V. Tepikin,Perinuclear, perigranular and sub-plasmalemmal mitochondria havedistinct functions in the regulation of cellular calcium transport,EMBO J. 20 (2001) 1863–1874.

[18] M.A. Matlib, Z. Zhou, S. Knight, S. Ahmed, K.M. Choi, J. Krause-Bauer, R. Phillips, R. Altschuld, Y. Katsube, N. Sperelakis, D.M.Bers, Oxygen-bridged dinuclear ruthenium amine complex specifi-cally inhibits Ca2+ uptake into mitochondria in vitro and in situ insingle cardiac myocytes, J. Biol. Chem. 273 (1998) 10223–10231.

[19] S. Terakawa, J.-H. Fan, K. Kumakura, M. Imaizumi-Ohara, Quanti-tative analysis of exocytosis directly visualized in living chromaffincells, Neurosci. Lett. 123 (1991) 82–86.

[20] H. Tinel, J.M. Cancela, H. Mogami, J.V. Gerasimenko, O.V. Gerasi-menko, A.V. Tepikin, O.H. Petersen, Active mitochondria surround-ing the pancreatic acinar granule region prevent spreading of inos-itol trisphosphate-evoked local cytosolic Ca2+ signals, EMBO J. 18(1999) 4999–5008.

[21] R.P. Haugland, Handbook of Fluorescent Probes and Research Prod-ucts, Molecular Probes Inc., Eugene, OR, 2002.

[22] W.J. Betz, G.S. Bewick, Optical analysis of synaptic vesicle recyclingat the frog neuromuscular junction, Science 255 (1992) 200–203.

[23] J.A. Gilabert, A.B. Parekh, Respiring mitochondria determine thepattern of activation and inactivation of the store-operated Ca2+ cur-rent I(CRAC), EMBO J. 19 (2000) 6401–6407.

[24] H. Vainio, L. Mela, B. Chance, Energy dependent bivalent cation70)

[ D.A.olated610.

[ nfluxected1.

[ in

[3)

[ cell

[ ighoring

[ ra-m 20

[ ffi--

[ localCa

[ A.-cellt

[35] B. Billups, I.D. Forsythe, Presynaptic mitochondrial calcium seques-tration influences transmission at mammalian central synapses, J.Neurosci. 22 (2002) 5840–5847.

[36] F. Yang, X.P. He, J. Russell, B. Lu, Ca2+ influx-independent synapticpotentiation mediated by mitochondrial Na+–Ca2+ exchanger andprotein kinase C, J. Cell Biol. 163 (2003) 511–523.

[37] P. Maechler, C.B. Wollheim, Mitochondrial glutamate acts as a mes-senger in glucose-induced insulin exocytosis, Nature 402 (1999)685–689.

[38] N. Ørtenblad, D.G. Stephenson, A novel signalling pathway orig-inating in mitochondria modulates rat skeletal muscle membraneexcitability, J. Physiol. 548 (2003) 139–145.

[39] M.J. MacDonald, L.A. Fahien, Glutamate is not a messenger ininsulin secretion, J. Biol. Chem. 275 (2000) 34025–34027.

[40] G. Bertrand, N. Ishiyama, M. Nenquin, M.A. Ravier, J.C. Henquin,The elevation of glutamate content and the amplification of insulinsecretion in glucose-stimulated pancreatic islets are not causallyrelated, J. Biol. Chem. 277 (2002) 32883–32891.

[41] V.P. Skulachev, Mitochondrial filaments and clusters as intracellularpower-transmitting cables, Trends Biochem. Sci. 26 (2001) 23–29.

[42] V. Haynes, S. Elfering, N. Traaseth, C. Giulivi, Mitochondrial nitric-oxide synthase: enzyme expression, characterization, and regulation,J. Bioenerg. Biomembr. 36 (2004) 341–346.

[43] E.N. Dedkova, X. Ji, S.L. Lipsius, L.A. Blatter, Mitochondrialcalcium uptake stimulates nitric oxide production in mitochondriaof bovine vascular endothelial cells, Am. J. Physiol. 286 (2004)C406–C415.

[44] S. Lin, K.A. Fagan, K.-X. Li, P.W. Shaul, D.M.F. Cooper, D.M.Rodman, Sustained endothelial nitric oxide synthase activationrequires capacitative Ca2+ entry, J. Biol. Chem. 275 (2000) 17979–17985.

[ ofevi-hway,

[ chi,en-reg-e

[ M.ase

[ and

[ rationcinar

[ o,

on of

[ .tein-

tiol.

[ is of

[ oholathol.

translocation in rat liver mitochondria, Eur. J. Biochem. 12 (19387–391.

25] H. Rasgado-Flores, S. Sanchez-Armass, M.P. Blaustein,Nachshen, Strontium, barium, and manganese metabolism in ispresynaptic nerve terminals, Am. J. Physiol. 252 (1987) C604–C

26] M. Condrescu, G. Chernaya, V. Kalaria, J.P. Reeves, Barium imediated by the cardiac sodium-calcium exchanger in transfChinese hamster ovary cells, J. Gen. Physiol. 109 (1997) 41–5

27] M. Brini, Ca2+ signalling in mitochondria: mechanism and rolephysiology and pathology, Cell Calcium 34 (2003) 399–405.

28] A.B. Parekh, Mitochondrial regulation of intracellular Ca2+ signal-ing: more than just simple Ca2+ buffers, News Physiol. Sci. 18 (200252–256.

29] M.R. Duchen, Mitochondria and calcium: from cell signalling todeath, J. Physiol. 529 (2000) 57–68.

30] R. Rizzuto, M. Brini, M. Murgia, T. Pozzan, Microdomains with hCa2+ close to IP3-sensitive channels that are sensed by neighbmitochondria, Science 262 (1993) 744–747.

31] S. Hehl, A. Golard, B. Hille, Involvement of mitochondria in intcellular calcium sequestration by rat gonadotropes, Cell Calciu(1996) 515–524.

32] R. Malli, M. Frieden, K. Osibow, W.F. Graier, Mitochondria eciently buffer subplasmalemmal Ca2+ elevation during agonist stimulation, J. Biol. Chem. 278 (2003) 10807–10815.

33] G. Hajnoczky, R. Hager, A.P. Thomas, Mitochondria suppressfeedback activation of inositol 1,4, 5-trisphosphate receptors by2+,J. Biol. Chem. 274 (1999) 14157–14162.

34] M. Montero, M.T. Alonso, E. Carnicero, I. Cuchillo-Ibanez,Albillos, A.G. Garcia, J. Garcia-Sancho, J. Alvarez, Chromaffinstimulation triggers fast millimolar mitochondrial Ca2+ transients thamodulate secretion, Nat. Cell Biol. 2 (2000) 57–61.

45] J.E. Branka, G. Vallette, A. Jarry, C.L. Laboisse, Stimulationmucin exocytosis from human epithelial cells by nitric oxide:dence for a cGMP-dependent and a cGMP-independent patBiochem. J. 323 (1997) 521–524.

46] K. Matsushita, C.N. Morrell, B. Cambien, S.X. Yang, M. YamakuC. Bao, M.R. Hara, R.A. Quick, W. Cao, B. O’Rourke, J.M. Lowstein, J. Pevsner, D.D. Wagner, C.J. Lowenstein, Nitric oxideulates exocytosis byS-nitrosylation of N-ethylmaleimide-sensitivfactor, Cell 115 (2003) 139–150.

47] T. Yasukawa, E. Tokunaga, H. Ota, H. Sugita, J.A. Martyn,Kaneki, S-nitrosylation-dependent inactivation of Akt/protein kinB in insulin resistance, J. Biol. Chem. 280 (2005) 7511–7518.

48] R.S. Balaban, S. Nemoto, T. Finkel, Mitochondria, oxidants,aging, Cell 120 (2005) 483–495.

49] M.P. Granados, G.M. Salido, J.A. Pariente, A. Gonzalez, Geneof ROS in response to CCK-8 stimulation in mouse pancreatic acells, Mitochondrion 3 (2004) 285–296.

50] Y. Huang, H.Y. Man, Y. Sekine-Aizawa, Y. Han, K. Juluri, H. LuJ. Cheah, C. Lowenstein, R.L. Huganir, S.H. Snyder,S-nitrosylationof N-ethylmaleimide sensitive factor mediates surface expressiAMPA receptors, Neuron 46 (2005) 533–540.

51] D.M. Shin, X. Luo, T.M. Wilkie, L.J. Miller, A.B. Peck, M.GHumphreys-Beher, S. Muallem, Polarized expression of G procoupled receptors and an all-or-none discharge of Ca2+ pools ainitiation sites of [Ca2+]i waves in polarized exocrine cells, J. BChem. 276 (2001) 44146–44156.

52] J.M. Gronroos, H.J. Aho, T.J. Nevalainen, Cholinergic hypothesalcoholic pancreatitis, Dig. Dis. 10 (1992) 38–45.

53] J.M. Gronroos, J. Laine, T. Kaila, T.J. Nevalainen, Chronic alcintake and carbachol-induced acute pancreatitis, Exp. Toxicol. P46 (1994) 163–167.