Thapsigargin inhibits voltage-activated calcium channels in adrenal glomerulosa cells

Article

Reference

Thapsigargin inhibits voltage-activated calcium channels in adrenal glomerulosa cells

ROSSIER, Michel, et al.

Abstract

Thapsigargin, an inhibitor of the microsomal Ca2+ pumps, has been extensively used to study the intracellular Ca2+ pool participating in the generation of the agonist-induced Ca2+ signal in various cell types. A dual effect of this agent was observed in bovine adrenal zona glomerulosa cells. At nanomolar concentrations, thapsigargin stimulated a sustained Ca2+

influx, probably resulting from Ca(2+)-store depletion. In contrast, when added at micromolar concentrations, thapsigargin prevented the rise in cytosolic free Ca2+ concentration ([Ca2+]c) induced by K+. This inhibitory effect of thapsigargin on voltage-activated Ca2+ channels was confirmed by measuring Ba2+ currents by the patch-clamp technique. Both low-threshold (T-type) and high-threshold (L-type) Ca2+ channels were affected by micromolar concentrations of thapsigargin. Analysis of the current-voltage relationship for T-type channels revealed that thapsigargin did not modify the sensitivity of these channels to the voltage, but decreased the maximal current flowing through the channels. In conclusion, thapsigargin appears to exert a dual effect on adrenal [...]

ROSSIER, Michel, et al. Thapsigargin inhibits voltage-activated calcium channels in adrenal glomerulosa cells. Biochemical Journal, 1993, vol. 296, no. 2, p. 309-312

PMID : 8257418

Available at:

http://archive-ouverte.unige.ch/unige:107531

Disclaimer: layout of this document may differ from the published version.

1 / 1

Thapsigargin inhibits voltage-activated calcium channels in adrenal glomerulosa cells

Michel F. ROSSIER,*t Christophe P. PYTHON,* Muriel M. BURNAY,* Werner SCHLEGEL,t Michel B. VALLOTTON*

and Alessandro M. CAPPONI*

Divisions of*Endocrinology and tinfectious Diseases, University Hospital,CH-1211 Geneva, Switzerland

Thapsigargin, aninhibitor ofthe microsomal Ca2+ pumps, has patch-clamp technique. Both low-threshold (T-type) and high- been extensively used to study the intracellular Ca2+ pool threshold (L-type) Ca2+ channels were affected by micromolar participatinginthegenerationof the agonist-inducedCa2+signal concentrations ofthapsigargin. Analysis of the current-voltage invarious celltypes. A dual effect of this agent was observed in relationship for T-type channels revealed that thapsigargin did bovine adrenal zona glomerulosa cells. At nanomolar concen- not modify thesensitivity of these channels to the voltage, but trations, thapsigargin stimulated a sustained Ca2+influx, prob- decreased themaximal current flowing through the channels. In ably resulting from Ca2+-store depletion. In contrast, when conclusion, thapsigarginappears toexerta dual effect onadrenal added atmicromolarconcentrations, thapsigargin prevented the glomerulosa cells. At lower concentrations, this agent induces rise incytosolicfreeCa2+ concentration

([Ca2+],)

induced by K+. a sustained Ca2+ entry, whereas at higher concentrations it Thisinhibitoryeffect ofthapsigargin onvoltage-activated Ca2+ decreases

[Ca2+],

by blockingvoltage-activated Ca2+channels.

channels was confirmed by measuring Ba2+ currents by the

INTRODUCTION

Aldosterone biosynthesis in adrenal zona glomerulosa cells is precisely modulated by changes in the cytosolic free Ca2l concentration

([Ca2+],)

induced by activators such as K+ or

angiotensin II (AngII),anoctapeptide hormone [1-3]. Whereas K+ is believedtoaffect

[Ca2+],

exclusivelyby activating voltage- sensitive Ca2+channels [4,5], the mode of action of AngIl ismore

complex. Indeed, AngIl binding toits receptoris linked tothe formation ofdiacylglycerol,anactivatorofprotein kinase C [6], and ofIns(1,4,5)P3, which is responsible for Ca2+ release from intracellular stores[7,8]. This initial phase of the Ca2+ signal in

response to the hormone is followed by a sustained Ca2+-entry phase [9], partiallydue tothe activation of voltage-sensitiveCa2+

channels

[10,11]

andpartially throughapathway directlyregu-

latedby intracellular Ca2+-store depletion, also termed 'capaci- tative' Ca2+entry[12,13].

Thapsigargin, a sesquiterpene lactonetumour promoter, has been extensively used to characterize this latter type of Ca2+

influx innon-excitable cells[14,15].Thapsigarginactsby blocking themicrosomal Ca2+pumpsnecessaryforCa2+ sequestration in intracellular pools, and therefore mimics the hormone by de- pleting Ca2+stores,butwithout elevationof Ins(1,4,5)P3 [16]. An activation of Ca2+ entry by thapsigargin has been observed in variouscell types,a finding thatprovided astrong support for thecapacitative-Ca2+-entry hypothesis [17].

Inbovine adrenalglomerulosa cells, the Angll-responsive and thapsigargin-sensitive Ca21 pools are largely coincident, but a

component of the agonist-induced Ca2+ influx, probably in- volvingvoltage-activated Ca2+ channels,cannotbemimickedby thapsigargin [13]. A similar effect ofthapsigarginon[Ca2+] has been described inratglomerulosa cells [18],in which this agent exerts asteroidogenic action. In these cells the effects ofAnglI and thapsigargin on aldosterone production were additive, whereas the response to K+ was potentiated by thapsigargin.

Interestingly,in thesecells,micromolar concentrations ofthapsi- garginmarkedlydecreased the rise in

[Ca2+],

inducedbyK+[19].

In the present study we report that, in bovine adrenal glomerulosa cells, thapsigargin, in addition toincreasing

[Ca2+],

through a mechanism of influx activated by the depletion of intracellularCa2+stores, inhibitsvoltage-activated Ca2+channels.

MATERIALSAND METHODS Materials

Thapsigarginwas obtained fromAnawa(Zurich, Switzerland).

Hepes, insulin, transferrin, sodium selenite, tetrodotoxin, ATP and GTP werepurchased from Sigma. Dispase (Grade II)was obtained fromBoehringerMannheim(Indianapolis, IN, U.S.A.), Percoll from Pharmacia(Piscataway,NJ,U.S.A.), andascorbate from Merck (Darmstadt, Germany). Horse serum, fetal-calf serum(FCS) and Dulbecco's modified Eaglemedium(DMEM) werefrom Gibco(Grand Island,NY, U.S.A.). Metyraponewas purchased from Aldrich (Milwaukee, WI, U.S.A.), and fura-2 and 1,2-bis-(2-aminophenoxy)ethane-NNN'N'-tetra-acetic acid tetracaesium salt (Cs4BAPTA) were from Molecular Probes (Eugene, OR,U.S.A.).

Isolation and culture of bovine glomerulosa cells

Bovine adrenal glomerulosa cells were prepared by enzymic dispersionwithDispaseandpurifiedon aPercolldensity gradient, as described in detail elsewhere [20]. Cells were then plated on small glass coverslips in antibiotics-containing DMEM supplemented with 1 mM ascorbate, 1 ,ug/ml insulin, 1 ,tg/ml transferrin, 1ng/ml sodium selenite, 5gM metyrapone, 2 mM glutamine, 2% (v/v) FCS and 10% (v/v) horse serum, and incubated overnight at 37 °C in 5% CO2. The next day, the medium was removed and replaced with serum-free DMEM, untilthe cellswere used forpatch-clamp experiments.

Abbreviationsused:

[Ca2+]",

^{cytosolicfree}Ca2+concentration;Angil, [lle5]angiotensin11;FCS,fetal-calfserum; DMEM,Dulbecco'smodifiedEagle medium; Cs4BAPTA, 1,2-bis-(2-aminophenoxy)ethane-NNN'N'-tetra-acetic acidtetracaesium salt.

t Towhom correspondenceshould beaddressed.

310 M. F. Rossierand others

Measurement ot

[Ca2+],

This was done in cell populations with the fluorescent probe fura-2. Freshly prepared cellswere purified on Percoll density gradients, washed twice and resuspended in a Krebs-Ringer medium [20] at a concentration of 107 cells/ml, and then incubated at 37°C for 30 min in the presence of 2,uM fura-2 acetoxymethylester.Thedyeexcess wasthenwashedaway,and the cellswerekeptatambienttemperaturein thesamemedium.

Batches of 2x106 cells were sedimented just before use, and resuspended in 2 ml of Krebs-Ringer medium, in a thermo- staticallymaintained cuvette at 37'C. Fura-2 fluorescence (ex- citation at 340nm and emission at 505nm) was recorded ina

Perkin-Elmer LS-3 fluorescence spectrometer, and

[Ca2+],

was

calibratedaspreviously described for quin2 [21], by usingavalue of 224 nM forthe Kd of fura-2.

Patch-clamp measurements

The activity of voltage-activated Ca2+ channels was recorded undervoltageclamp in the whole-cell configuration of the patch clamp technique, essentially as described in [20]. The bath solution contained (in mM): 117 tetraethylammonium chloride, 20BaCl2, 0.5MgCl2, 5D-glucose, 32sucroseand 200 nMtetrodo- toxin, and was buffered to pH 7.5 with 10 mM Hepes/CsOH.

The patch pipette (Clark 150T) contained (in mM): 85 CsCl, 10tetrabutylammonium chloride, 6MgCl2, 5 sodium ATP and 0.04GTP, and pHwasbufferedto7.2with 20 mMHepes/CsOH.

Thepipette solution also contained 0.9 mM CaC12 and 11 mM Cs4BAPTA in order to buffer free Ca2+ below 50 nM and to stabilize thelow access resistance frompipette tocell. Glomer- ulosa cells were voltage-clamped at a holding potential of -90 mV and depolarized as indicated in the legends of the Figures. The Ba2+currentswerefiltered and automatically leak- subtracted asindicated elsewhere[20].

RESULTS AND DISCUSSION

Increase In

[Ca2+],

Induced by thapsigargin

Addition of thapsigargintofura-2loaded bovine adrenal glomer- ulosa cells led toasmall butreliable increase in

[Ca2+],

(Figure

la,uppertrace). This Ca2+response wassustained formorethan 10 min and depended on the presence of extracellular Ca2+

(lower trace),ahallmark forCa2+ influx.Inaddition, theresponse wastotally unaffected bynicardipine (1

#uM),

adihydropyridine blocking theresponse to K+ (results not shown). The effect of thapsigargin was concentration-dependent, with a maximum observedatapprox.50nM(Figure Ib). These resultsareingood agreementwith those ofElyetal.[13] and confirm thepresence

inglomerulosa cells ofaCa2+-entry pathway, although of minor importance, which is controlled by the depletion of intracellular Ca2+stores [14] rather than by the membrane potential.

Inhlbitlon by thapsigargin of K+-Induced Ca2+Influx

Surprisingly, thapsigargin,atmicromolarconcentrations,mark- edly decreased the

[Ca2+],

increase evoked by K+ (Figure 2, inset). The IC50, approx. 1 pM thapsigargin,did not appearto dependon K+ concentrationintherange 6-25 mM (resultsnot shown). Theextentofinhibition by thapsigarginvaried slightly from one cell preparation to the other, but no inhibition was

observed at concentrations below 0.5

,/M

(Figure 2). When thapsigargin was introduced in the medium before K+, the

(a)

190.

cS

L

v 150 +x 80 ^-

60

(b)

a) ._

(I) 0

co

a)

C4

0t

0)

Thapsi (100 nM)

0 12 3 4 5 6

Time(min)

7 8 9

1 10 100

[Thapsigargini (nM)

Figure1 Effectofthapsigargin on[Ca2l+,

Bovine adrenalglomerulosa cellswereisolatedand loaded with fura-2as described inthe Materialsand methods section. (a)Ca2+responsefrom cells exposedto100nMthapsigargin (Thapsi),added atthe time indicatedby thearrow,in thepresence(upper trace)of1.2 mM extracellular Ca2^,orinthe absence(lower trace) of addedCa2+ (1.2 mM EGTA). (b) The concentration-dependent increase in [Ca2+]c induced by thapsigargin in the presence of extracellular calciumwasdeterminedby measuring the value of[Ca2+]cbeforeand4-5min afterthapsigarginaddition. Resultsaremeans+S.E.M. from five independentexperiments.

0 4-

4)a) c

0

CL)

00

C4

o

.0

c 0

6

4

2

0.01 0.1 1

[Thapsigargini (pM)

10

Rgure 2 Inhibfllon bythapsigarginoftheK+-Induced Ca2+response Fura-2-loaded cellsweresequentially exposedto 9 mM(0)or12mM(0) K+and then to thapsigargin (0.03-5 FM).Inset:anexampleof theresponseto 9 mM K+added after 2min (first arrow)and itsinhibitionby5,M thapsigargin (addedatthe timeindicatedbythe second arrow). The decrease in [Ca2+]c^was measured 3-5min after thapsigargin addition and expressedasthepercentageinhibition of the maximalresponsetoK+.Resultsarepresented

as a function of thapsigargin concentration and are the mean values from two to five determinations and from six different cell preparations. Itedotted line describesalogistic function of thetype: f(x)=A/[1+(lC5w/)k],where A is themaximalamplitudeand kthe slope factor,fittedtoall individualdata,withanIC50of1.01JUM.

~~~~~~~~~~~~~~~~...

|:... ..

_0

50 ⁰

0 2 4 6 8 10

Time(min) :0

NoCa2,

AwAhA

40pA K 10ms

3 6 9 12 15

Time(min)

(d)

50 0

-40 O 0

c30

03

03

Z 20 0080

10 0

02

³

0

0.001 0.01 0.1 1 10

[Thapsigargin](,uM)

Figure3 Inhibition by thapsigarginof voltage-activated Ba2+ currents

Voltage-activated Ba2+currentswererecorded bythepatch-clamp technique,inthewhole-cell configuration,asindicated in theMaterialsand methods section.(a)Superimposedtracesof Ba2+

currentselicitedbyadepolarizationof the cell from -90 to +10 mV, before(O;C)and 2minafter(0) addition of thapsigargin (Thapsi; 0.5 ,uM).Thedottedlineindicatesthezerocurrent level.(b) Sameas(a),but thiscellwasdepolarized for 20msto+20mV, and thenrepolarizedto-65mVtoevoketailcurrents.The decayofthe currentwasdescribed byasingleexponential function,with time constants of 9.14 and 9.95msfor controland treatedcellsrespectively.The first 2msrecording afterrepolarization (duringthecapacitativetransientandL-channeldeactivation)

wasnotincludedforcurvefitting.(c)Timecourseof tail-current inhibitionby thapsigargin.Slowly deactivating (T-type)currentswereevokedevery15s, asdescribedin(b), and the initialcurrent wasextrapolated by fittingthedecayingcurrent toanexponential function and plottedas afunctionof time. Increasing concentrations of thapsigarginwereaddedatthe timesindicated bythe arrows.(d)Concentration-dependent effectofthapsigarginonT-typecurrent.Thepercentagetail-current inhibitionwasestimatedby averagingthe currentover ashortperiodbeforeadditionof thapsigargin,and afterinhibitionhad occurred. Datacollected from 18cellswereplottedas afunctionofthapsigarginconcentration.

subsequent^response toK+wasconsiderably decreased, butthe sensitivity to K+, as assessed by successive stimulations with increasing concentrations of the cation, remained unchanged (resultsnotshown). Theseobservationsstrongly suggestedthat thapsigarginaffectedCa2+ channelsopened by K+.

Inhibitlon ofvoltage-activated Ca2+ channels by thapslgargin The action ofthapsigargin on voltage-activated Ca2+ channels

wasdirectly assessed by the patch-clamp technique.Inward Ba2+

currents were evoked by depolarizing the cell from a holding potentialof-90 mVto +10mV.Asdescribedelsewhere[20,22], thesecurrents developedto amaximum in less than 20ms,and thenrapidly decayed to aplateau whichremained elevated for several hundred ms (Figure 3a, control trace). These biphasic kineticsarebelievedtoreflect the activation of both low-threshold (transient, T-type), and high-threshold (long-lasting, L-type) channels [22]. Addition of 0.5 uM thapsigargin to the bath resulted inamarked decrease in thecurrent elicitedby depolar- ization(Figure 3a). Thapsigarginaffectedboth thepeakand the plateauof theBa2+ current, suggestingthatbothT-and L-type channelsare sensitive tothis agent.

Althoughtheeffectofthapsigargin appearedmorepronounced

onL-typecurrents(Figure 3a),wehavefocusedourattentionon

T-typecurrents,because thesecurrentsareactivated byphysio- logicalconcentrations ofK+andarebelievedtoberesponsiblefor activation ofsteroidogenesisinglomerulosacells[5].Inorderto isolatespecificallytheeffect ofthapsigarginonT-typechannels, slowly deactivating tail currents were induced upon repolar-

ization of the cellto -65mVafterashort activationperiodat +20mV. These slowly deactivating currents are almost ex-

clusivelyduetoT-typeCa2+ channels[5]. Anexampleof sucha

largetailcurrentispresentedinFigure 3(b).Thecurrentappeared immediately^upon repolarizationofthe cellanddecayedwitha

timeconstantofabout 9ms,which isavaluecharacteristic of T- type currentsinglomerulosacells[5,20].Asexpected,addition of

amicromolarconcentration ofthapsigargin clearlydecreased the size of the tailcurrent(Figure 3b).Todetermine the inhibitionof the current by a given concentration of the drug, the same

voltage protocolasthatdescribed inFigure 3(b)wasrepeatedly performed every 15s. At each pulse of voltage, the current

was recorded, andthe initial valueof the decayingcurrentwas

extrapolatedafter fittingthe data withanexponential function.

The value of thecurrent atthemomentof cellrepolarizationwas

thenplottedas afunctionof time. Anexampleof the timecourse

of the effect ofthapsigargin, added sequentially at increasing concentrations, is presented in Figure 3(c). Whereas a low concentration ofthapsigargin (10 nM) did not affect the tail- current value, higher concentrations ofthis agent (100nM or

1,csM) rapidly and markedly decreased its magnitude. The percentage inhibition of the current at various thapsigargin concentrations was similarly determined from 18 independent cellsexposed toone orseveralthapsigargin concentrations and is shown in Figure 3(d). In spite of the widedispersion ofthe

results, which may reflect some cellular heterogeneity in the

sensitivityto thapsigargin, it ispossible to estimate amaximal inhibitoryeffect of thisagentataround ¹ 1uM.Thisvalue,which is of thesameorderofmagnitudeastheconcentrationnecessary (a)

+10mV -90 mV ^-

Oc

* Thapsi (0.5MM)

(b) +20 mV -65 mV -90mV

Oc

* Thapsi (1 M)

(c) -250 -200 9-150 a

-_50

0

thapsi thapsi 10nM 100 nM thapsi

1PM

312 M. F. Rossier and others

U

100

80 - 0

::.60

. c

ae. 40

-\

oO

*\ k. * - Thapsi

20 -

(5gM)

-100 -80 -60 -40 -20 0 10

Voltage (mV)

FIgure 4 Lack of effectofthapsigarginonT-channelsensMvttytovoltage Voltage-dependentactivation andinactivation ofT-typechannelsweremeasured,asdescribed in[5],before(0, EO; C)andafter(0, *)treatment withthapsigargin (Thapsi;5#sM).The activationcurves (0, 0)weredeterminedby measuring slowly deactivating tailcurrents elicitedby repolarizationofthe cell to-65 mV aftera20mschannel activation atvarious test voltage amplitudes (-55to +5mV),whereas thesteady-stateinactivationcurves(E, *)

were obtained by measuring the samecurrent (at -65mV) after maximal depolarization (+20mV,20ms)from variousconditioning pre-pulse potentials (-100to -30mV) lasting 10s.DatawerefittedtoBoltzman'sequationandnormalized to themaximum of the function

(Id).

^{Each datapointisthemean}value from three different cells in which currents have been determined before andafteraddition ofthapsigargin.

forinhibition of theCa2+responsetoK+, appearsmuchhigher than the thapsigargin concentration responsible for maximal Ca2+influx(Figure lb).Thissuggeststhat bothevents, i.e.Ca2+

entry by the 'capacitative' pathway and inhibition ofvoltage- activated Ca2+channels,are totallyunrelated.

Mechanism ofT-typechannel Inhibitlon

The effect ofthapsigargin on the voltage sensitivity ofT-type channelswasinvestigated by measuringthevoltagedependency ofchannel activation andinactivation in thepresenceand inthe absenceof thedrug.Activationwasdeterminedby applicationof 20msdepolarizingtestpulsestotheindicatedpotentials(-55to +⁵mV) from a negative holding potential (-90 mV). Upon repolarization, tail currents were elicited and measured as

described in thelegendofFigure 3(b).Datawerethenplottedas a function of the activation potential, fitted to Boltzman's equationandnormalized tothe maximumof thecurrent(Figure 4, circles). Correspondingly, the dependenceofchannel inacti- vationonvoltagewasdeterminedby holdingthecellatvarious potentials (-100 to -30mV) for 10s, leading to steady-state channelinactivation,beforeapplyingastrongdepolarizing pulse (+20mV)for 20ms to opentheremainingavailable channels.

Similarly, tail currents were elicited upon repolarization,

measured and plotted as a function of holding (inactivation) potential, before being fitted and normalized to Boltzman's equation (Figure 4, squares).Theresults show thatthapsigargin (5 ,M), although decreasing the maximal elicitable current by approx.40% (resultsnotshown),did notchangethedependence ofT-typechannelsonvoltage. Indeed,noshift of the activation

or inactivation curves of the channel could be observed after treatment with thapsigargin, as is the case, for example, with atrial natriureticpeptide [5] ornitrendipine [11]. In thisregard,

the mechanism of T-channel inhibition by thapsigargin is com- parable with that oftetrandrine,arecently described blockerof these channels [20], and willrequire furtherinvestigationto be resolved.

Inconclusion, thapsigargin appears to exert adual effect on adrenal glomerulosa cells. At nanomolar concentrations, this agentinducesasustained Ca2+entry,presumably resultingfrom depletion of intracellular Ca2+ stores.Athigher concentrations, thapsigargininhibitsvoltage-activatedCa2+ channels, and there- fore decreases theK+-induced Ca2+response.Whether thislatter effect ofthapsigargin is alsopresentin other celltypesremainsto be confirmed. Recently, Vercesi et al. [23] have reported that thapsigargin, atconcentrations above 10,uM,causescollapseof the mitochondrial membrane potential in Trypanosoma brucei and in rat liver, resulting in Ca2+ release from this organelle.

These results thereforehighlight the need for caution in inter- preting the action of high concentrations of thisagent oncellular Ca2+ homoeostasis.

We thank Dr. K.-H.Krause, Dr. N. Demaurex and Dr. S. Rawlings for their helpful comments, and we are grateful to Ms. L. Bockhorn, Ms. G. Dorenter and Ms.

M.Lopezfortheir valuabletechnicalhelp. This workwassupported by the Swiss NationalScience Foundation(grants 32-30 125.90 and 31-27 727.89). M. F. R. isa recipientof agrant from the MaxCloetta Foundation and C. P. P. is a recipient ofa fellowship from Pfizer, Switzerland.

REFERENCES

1 Spat,A., Enyedi, P., Hajnoczky, G. and Hunyady, L. (1991) Exp. Physiol. 76, 859-885

2 Barrett,P.Q., Bollag,W.B., Isales, C.M., McCarthy, R. T.and Rasmussen, H. (1989) Endocr. Rev.10,496-518

3 Capponi,A.M., Rossier,M.F., Davies,E.andVallotton, M. B.(1988) J. Biol. Chem.

263,16113-16117

4 Capponi,A.M.,Lew,P.D.,Jornot, L. and Vallotton, M.B.(1984) J. Biol.Chem.

259,8863-8869

5 Barrett,P.Q., Isales,C.M.,Bollag, W. B. and McCarthy, R. T. (1991) Am. J. Physiol.

261,F706-F719

6 Lang,U. and Vallotton,M. B.(1987)J. Biol. Chem.262,8047-8050

7 Rossier,M.F., Capponi,A. M. andVallotton, M.B.(1988) Mol. Cell. Endocrinol. 57, 163-168

8 Rossier, M.F., Krause, K.-H., Lew, P.D., Capponi,A.M. andVallotton, M. B.(1987) J. Biol. Chem.262,4053-4058

9 Ambroz,C. and Catt,K.J.(1992) Endocrinology (Baltimore) 131, 408-414 10 Quinn,S.J., Cornwall, M. C. andWilliams, G.H.(1987) Endocrinology (Baltimore)

120,1581-1589

11 Cohen,C.J., McCarthy, R. T., Barrett, P. Q. and Rasmussen, H. (1988) Proc. Natl.

Acad.Sci. U.S.A.85,2412-2416

12 Putney,J.W.,Jr(1986) Cell Calcium 7, 1-12

13 Ely,J.A., Ambroz, C., Baukal,A.J., Christensen, S. B., Balla, T. and Catt, K. J.

(1991) J. Biol. Chem.266,18635-18641

14 Takemura, H., Hughes,A.R.,Thastrup,0. andPutney, J. W., Jr. (1989) J. Biol.

Chem.264,12266-12271

15 Putney,J.W.,Jr. (1990)Cell Calcium11,611-624

16 Thastrup,O.,Cullen,P.J., Drobak, B.K.,Hanley, M. R.and Dawson,A. P.(1990) Proc.NatI.Acad.Sci. U.S.A.87,2466-2470

17 Menniti, F.S., Bird, G. St.J.,Glennon, M.C., Obie,J.F., Rossier, M. F. and Putney, J.W.,Jr.(1992)Mol.Cell. Neurosci.3,1-10

18 Hajnoczky,G., Varnai, P., Hollo, Z.,Christensen,S.B., Balla, T., Enyedi,P.and Spat, A.(1991) Endocrinology(Baltimore) 128,2639-2644

19 Balla, T., Hollo,Z.,Varnai, P.andSpat,A.(1991)Biochem.J.273,399-404 20 Rossier,M.F., Python,C.P.,Capponi,A.M.,Schlegel, W., Kwan,C.Y.andVallotton,

M. B.(1993)Endocrinology (Baltimore)132,1035-1043

21 Capponi,A.M.,Lew,P. D.,Schlegel,W. andPozzan,T. (1986)MethodsEnzymol.

124,116-135

22 Matsunaga, H., Yamashita, N., Maruyama, Y., Kojima,I.andKurokawa, K.(1987) Biochem.Biophys.Res. Commun.149,1049-1054

23 Vercesi,A.E., Moreno,S.N.J.,Bernardes,C.F., Meinicke,A.R.,Fernandes,E. C.

and Docampo,R.(1993)J. Biol.Chem.268, 8564-8568

Received 17May1993/4August 1993;accepted11August1993