Reconstitution and Characterization of a Calmodulin-Stimulated Ca-Pumping ATPase Purified from Brassica oleracea L

http://www.diva-portal.org

This is the published version of a paper published in Plant Physiology.

Citation for the original published paper (version of record):

Askerlund, P., Evans, D. (1992)

Reconstitution and Characterization of a Calmodulin-Stimulated Ca-Pumping ATPase Purified from Brassica oleracea L.

Plant Physiology, 100(4): 1670-1681

Access to the published version may require subscription. N.B. When citing this work, cite the original published paper.

Free PMC Article

Permanent link to this version:

Plant Physiol.(1992) 100, 1670-1681 0032-0889/92/100/1670/12/$01 .00/0

Received forpublicationJune22, 1992 AcceptedAugust 21, 1992

Reconstitution

and

Characterization of

a

Calmodulin-Stimulated

Ca2"-Pumping

ATPase

Purified

from

Brassica oleracea

L.'

PerAskerlund*2 and David E. Evans3

Department

of

Plant

Sciences, University

of

Oxford, South Parks Road, Oxford,

OXI 3RB,

United Kingdom

ABSTRACT

Purification andfunctional reconstitution ofa calmodulin-stim-ulated Ca2"-ATPase fromcauliflower(Brassica oleracea L.)is de-scribed. Activitywas purified about 120-fold from amicrosomal fraction using calmodulin-affinity chromatography. The purified fraction showedapolypeptide at 115 kD, which formeda phos-phorylated intermediate inthepresence of Ca2 ,together witha fewpolypeptides with lower molecularmassesthatwere not phos-phorylated. The ATPase was reconstituted into liposomes by

3-([cholamidopropylJ-dimethylammonio-)1-propanesulfonate

(CHAPS) dialysis. The proteoliposomes showed ATP-dependent Ca21uptake and ATPaseactivity, both of which were stimulated about 4-foldby calmodulin. SpecificATPaseactivitywasabout5 &molmin-'(mgprotein)-', and theCa21/ATPratio was0.1 to 0.5 when theATPase was reconstituted with entrapped oxalate. The purified, reconstituted Ca2-ATPase wasinhibitedby vanadateand erythrosin B,but notby cyclopiazonic acid and thapsigargin. Activ-ity wassupported by ATP (100%) and GTP (50%) and had a pH optimum of about 7.0. The effect of monovalent and divalent cations(including Ca21)on activity is described. Assay of mem-branespurified by two-phasepartitioning indicated that approxi-mately 95% of the activitywasassociated with intracellular mem-branes, butonly about 5% with plasma membranes.Sucrose gra-dientcentrifugation suggests that theendoplasmicreticulumisthe major cellular location of calmodulin-stimulated Ca21-pumping

ATPasein Brassicaoleracea inflorescences.

Calciumis anessential intracellular

regulator

in

plant

cells andisinvolvedinmetabolic and

developmental

regulation.

Maintenance ofalowfree

cytoplasmic

concentration

(about

0.1

uM)

of

Ca2"

([Ca2"]t,y)

is necessary forits function as a

second messenger

(18).

This low

[Ca2+],y

is maintained

by

the action of active

Ca2+

transport systems assumed to be

locatedat theplasma membrane, ER, and tonoplast(5, 12).

Plant active

Ca2+

transport systemsfall into two categories,

Ca2+-pumping

ATPases and

Ca2+/nH'

antiporters. The

for-1 Supported byagrantfrom the UnitedKingdom Agricultural and Food Research Council (AFRC) underits Plant Molecular Biology

initiative.

2Presentaddress: Department of Plant Biochemistry, University ofLund,P.O. Box7007,S-220 07Lund, Sweden.P.A. wassupported byaSwedish NaturalScienceResearchCouncil(NFR) postdoctoral fellowship, a Royal Society/Swedish Royal Academy of Sciences exchangefellowship,andagrantfrom the Swedish Institute.

3RoyalSociety1983University Research Fellow.

1670

mer have beensuggestedtobe locatedat theplasma mem-brane andER, whereas thelatterislocatedatthe tonoplast

(5, 12). CaM4-stimulated ATP-dependent

Ca2"

pumps

situ-ated in the plasma membrane are believed to be of key

importance for long-term regulation of

[Ca2"],y

in animal

cells (8, 9, 13, 27). A CaM-stimulated Ca2+-ATPase is also

present inplants andwaspartlypurifiedfrommaize micro-somes11 years ago usingCaM-affinity chromatography (11). Thisenzyme waslater shownby Briars etal. (4, 12) to bea P-type ATPase with a molecular mass of about 140 kD,

properties identical to those of the

Ca2+-ATPase

in animal plasma membranes (8, 9, 13, 27). Antibodies against the

erythrocyte Ca2+pumpcross-reactedwith thepurifiedmaize enzyme,but no reactioncould be detected inwestern blots ofmaizemicrosomes.Furthercharacterization ofthepurified CaM-stimulatedCa2+-ATPase from plants hasnotbeen

car-riedout (possibly due to the high instabilityofthis ATPase

aftersolubilization), anditsability topump

Ca2'

has never

been demonstrated.In thepresent article, wereport on the functional reconstitution,characterization, and cellular

loca-tion of a CaM-stimulated

Ca2+-pumping

ATPase purified fromcauliflower (Brassica oleraceaL.)inflorescences.

MATERIALS AND METHODS PlantMaterial

Cauliflower(Brassica oleracea L.) inflorescences were

pur-chasedlocally.

Preparation ofa

Microsomal

Membrane Fraction

Inflorescenses (130g) werehomogenized using ablender fitted withrazorbladesin275 mLof50 mm Mops-BTP, pH 7.5, 0.33 M sucrose, 5 mMNa2-EDTA, 5 mm DTT,0.2%

(w/

v) casein(boiled enzymichydrolysate, Sigma N4517),0.2% (w/v)BSA (Sigma A3294, proteasefree),

0.6%

(w/v)PVP, 1 mMbenzamidine-HCl, and0.5mm PMSF. Thehomogenate

wasfilteredthrough a nylon cloth andcentrifugedat10,000g for 15 min.The supernatantwas centrifugedat40,000gfor

1 h. Theresulting pelletwassuspended with aglass/Teflon

'Abbreviations: CaM, calmodulin; BTP, 1,3-bis(tris[hydroxy-methyl]methylamino)propane; CHAPS, 3-([cholamidopropyl]di-methylammonio-)1-propanesulfonate; cmc,critical micellar

concen-tration; octyl glucoside,

n-octyl-#-D-glucopyranoside;

MEGA 8, octanoyl-N-methylglucamide.

RECONSTITUTION OF A CALMODULIN-STIMULATED Ca2+ PUMP

homogenizer in buffer A (25 mm Mops-BTP, pH 7.5, and

0.33 M sucrose)supplemented with 0.5 M NaCl, 1 mm Na2-EDTA, 5 mm DTT,and0.5 mmPMSFto afinalvolume of 25 mL,and was again pelleted at 100,OOOgfor 45min.The final

washed microsomalpellet wassuspendedto about 3 mL with

buffer Aplus 5 mm DTT and 0.5 mm PMSF, except when

usedasstartingmaterialforphase partition (see below). All operations werecarriedoutat0 to40C. The membraneswere

frozeninliquidN2and storedat-700C for lateruse.

Solubilization of the

CaM-Stimulated

Ca21

-ATPase

and

Chromatography

Washed microsomes (82 ± 29 mg ofprotein, average of five experiments) were solubilized in 22.5 mL of buffer A

plus 0.5 M NaCl, 5 mm CaCl2,

0.1%

(w/v) phospholipid

(Sigma P3644; type

IV-S),

2 mIM MgCl2, 2 mmATP, 40

AsM

leupeptin,0.5 mm PMSF, and 2mrim DTT, 250 mgofTriton X-100 (Surfact-Amps, Pierce, Rockford, IL) with stirring for

10 min at

40C.

The unsolubilized material (15 ± 1 mg of

protein) waspelleted at

100,OOOg

for 45 min. CaM-affinity chromatography wascarried out

essentially

asdescribedby

Penniston etal.(27). Thesupematant containingthe solubi-lizedproteins(66±20mgof protein)was

applied

to a5-mL

columnof CaMagarose(Sigma P4385) that had been washed withbuffer A

plus

0.5 M NaCl, 5 mm

CaCl2,

0.4 mmATP,

0.4

mIM

MgCl2, 0.05% phospholipid, 0.05%

(v/v)

Triton

X-100, 0.5

mnim

PMSF,and1 mmDTT. After

application

of the sample, about 150 mLof this mediumwas allowedto pass

through the column. The flow rate was about 1

mL/min.

Whenthe purifiedATPase was not to be usedfor reconsti-tution, about 50 mL of wash medium without

MgCl2

and

ATP waspassed through the column before elution. When the ATPasewastobereconstitutedintoliposomes, the wash mediumwasinsteadfollowedby 50 mL of

buffer

A

plus

0.5

M

NaCl,

5 mM

CaCl2,

1 mM CHAPS

(Sigma C3023),

0.5 mM

PMSF, and1 mm DTT(no

phospholipid). CaM-binding

pro-teins were ineachcase eluted with 10 mmEGTA. Immedi-ately after elution,0.2 mLof0.1M

CaCl2

wasaddedtoeach 1.4-mLfraction.Theeluted fractionswere

assayed

for protein

and/or

ATPaseactivity. Peakfractionswere

pooled,

supple-mented with 10 mM DTT, and

placed

onice.

Reconstitution

of

CaM-Stimulated

Ca2" Uptake

and

ATPase

Activity

A mixed

phospholipid

preparation (Sigma P3644, type

IV-S)

was further

purified (17)

and stored in chloro-form:methanol (2:1,

v/v)

at-700C. After removing the

sol-vents on arotaryevaporatorfollowed

by

lyophilization

over-night, the

phospholipids

were

suspended

with

glass

beads

under N2 in 25

mi

Mops-BTP, pH7.2, and 0.33 M sucrose

(buffer

B)toafinalconcentrationof26

mg/mL.

CHAPS

(40

mm)wasthenadded andthesuspensionsonicatedtoclarity

on a bath sonicator under N2. To initiate reconstitution, 1 partofthe

phospholipid/CHAPS

suspension

(approximately

3mL)wasaddedto3 partsof column eluate containingthe ATPase

(approximately

0.3mgof proteinin 8-12mL, Table

I). The mixture was

briefly vortexed,

incubatedonice for 1

h, anddialyzedagainst 1 Lof bufferB plus1 mm DTTand

0.5 mmPMSF at

40C

during about 120 h with sixchanges.

This longdialysis timewasfoundto benecessarytoproduce sealed

proteoliposomes

with thedialysis tubing used (6.3 mm

diameter cellulose tubing,mol wtcutoff 12,000-14,000;BDH

ChemicalsLtd, Poole, UK). During the last two changes, 20 g of prewashed

Amberlite

XAD-2 (BDH Chemicals) was

addedto the dialysisbath to adsorb CHAPS.After dialysis, 10 mm DTT was added to the proteoliposomes, and they wereeitherdirectly assayedforCaM-stimulated

Ca2"

uptake andATPase activity orstored underN2 at 0 to

40C

for later

use. Entrapment of oxalate inside

proteoliposomes

was

achievedbyadding0.2MK2-oxalatetothe

protein-phospho-lipid-CHAPS

suspensionand dialyzinginthepresenceof0.2

M K2-oxalateafter removal ofthe Ca-oxalateprecipitate by centrifugation. In these experiments, the dialysis medium

was replacedwithdialysis mediumplus0.2M KCIand20 g

L-1

Amberlite

XAD-2during the lasttwochanges.

Two-Phase

Partition

Plasmamembraneswerepurified fromamicrosomal

frac-tionby partitioningin anaqueouspolymer two-phasesystem

(36 g, final weight, of

6.2%

[w/w]

Dextran T 500,

6.2%

[w/w]

PEG3350,330mmsucrose, 7.5 mmKCI,and5 mm

K-phosphate, pH 7.8) essentially as describedin ref. 22. The microsomal fractionwasobtainedasdescribedabove,except that 90 gofinflorescenceswereusedasstartingmaterialand

thelast washwasomitted. With the phasecompositionused, the plasma membranes partitioned in the upper phase, whereastheintracellular membranes partitionedatthe

inter-phaseandintothelower phase.Toincreasethepurityof the fractions, the initialupperphasewasrepartitionedtwicewith fresh lower phase and the initial lower phase was

reparti-tionedoncewithfreshupperphase. Thefinalupper(U3and

U3';

ref.

22)

and lower (L2)

phases

were diluted 4 and 10

times,

respectively,

with

buffer

A

plus

1 mm Na2-EDTA, 1 mm

DTT,

and 0.5 mm PMSF, and the membranes were

pelleted

at

100,OOOg

for 45 min. A small amount of the microsomal fraction used as starting material was diluted about 20-fold and pelleted in the same way. The pellets obtained were

suspended

in buffer Aplus 1 mm DTTand

0.5 mm PMSF. The membranes were snapfrozen in liquid N2 andstoredat

-700C.

Formeasurementsof

ATP-depend-ent

Ca2+

uptake, the

predominantly

right-side-out

plasma membranes were frozen and thawed about four times to

produce a mixture of inside-out and right-side-out vesicles (26).

Sucrose

Gradient

Centrifugation

A

10,OOOg

supematant

(obtained

asdescribed aboveexcept that BSA was omitted from the homogenization medium)

was loadedon top ofa cushion of gradient buffer (25

mi

Mops-BTP, pH 7.2, 2

mt

Na2-EDTA, and 10

mi

KCl)

plus

1.7Msucroseandcentrifugedin aswing-out rotor at 79,000g

for30 min. Themembranesattheinterphasewerecollected and

mixed

with 2 volumes of gradient buffer plus 40

uM

leupeptin.Thediluted membranes

(approximately

0.7mgof

protein in 0.7 mL) were applied to a continous sucrose

gradient(0.6-1.7Msucrose in 10 mLof gradientbuffer)and 1671

ASKERLUND AND EVANS

Table I. Balance Sheet for Purification and Reconstitution of CaM-StimulatedATPaseand ATP-DependentCa2" UptakefromCauliflowerMicrosomes

The yield andpurification factors refer toCaM-stimulated activity. Total activities areexpressedin

,umol min-' and specific activities (between parentheses) in ,mol min-' (mg

protein)-'.

Atypical

experiment

is shown.

Totaland (Specific)Activity

Total

-Purifi-Protein +0.5 Mm CaM- cation

Yield

-CaM CaM stimulated

mg Microsomes 78 ATPase 17.2 19.6 2.41 1 100 (0.221) (0.251) (0.031) Ca2` uptake 0.66 1.91 1.25 1 100 (0.0103) (0.0298) (0.0195) Solubilized micro- 78 somes ATPase 21.6 23.7 2.11 0.87 88 (0.277) (0.304) (0.027) Supernatant 68 ATPase 14.1 16.1 2.04 0.97 85 (0.207) (0.237) (0.030) Pooled CHAPS- 0.43 eluate ATPase 1.10 1.82 0.72 54 30 (2.55) (4.23) (1.68) Oxalate-loaded 0.32 proteolipo-somes ATPase 0.378 1.59 1.22 122 51 (1.18) (4.98) (3.80) Ca21 uptake 0.030 0.144 0.114 18.6 7.5 (0.096) (0.458) (0.362)

centrifuged

in aswing-outrotorfor2h at100,000g. Fractions (0.75 mL) were collected from the bottom of the

gradient

with a peristaltic pump and were aliquoted and stored at

-200C

untilanalysis. Alloperations werecarriedoutat0to

40C.

Sucrose concentration was measured with an Abbe Model

60/ED

refractometer

(Bellingham

&

Stanley

Ltd.,

Tun-bridge

Wells,

UK).

Ca2" Uptake

Ca2"

uptake

wasmeasuredinamediumcontaining buffer Bplus 100

mim

KCl, 0.1%

(w/v)

BSA

(Sigma A7030),

5 mM

MgCl2,

2.5mmATP,50

,M

CaCl2

(0.3-0.6

Bq

45CaC12

pmoh-'),

1 to5mMDTT,1mMNaN3,0.1mmNa2-molybdate (standard

assaymedium), and 1 to 2 ,g ofreconstituted ATPaseor 5

,ug of membrane protein in 0.1 mL for 4 to 5 min (except

during

time course experiments) at 340C. Modifications of

the standard assay medium and additions of CaM

(from

bovinebrain; Sigma P2277) were asindicatedin the

figure

legends. The sample was preincubatedwith assay medium for20 to 30 minprior tostarting

Ca2+

uptakeby

adding

ATP. Controls without ATPwere runinparallel.The reactionwas

stopped

by

additionof0.6 mLofbufferB

plus

1 mm

EGTA,

andaliquotswereimmediately filtered through 0.20-,um (re-constituted

Ca2"

pump) or 0.45-um

(membranes)

pore-size Whatmancellulosenitratemembranefilters.

After washing fourtimeswith1 mLof bufferBplus 1mm

EGTA, the filterswere dried andthe amountof

45Ca2+

was

measuredby scintillation counting. Countingefficiencywas

80 to

95%.

Free

[Ca21]

was measured with a

Ca2+-specific

electrode (Orion Research Inc., Boston, MA) calibratedwith the

Ca2+

buffers ofTsien and Rink (35). Free

[Ca2+]

in the

absence ofATP wasabout 100

AM,

eventhoughonly50

AM

CaC12wasaddedtotheassay.Thisadditional

Ca2+

wasdue

tocontaminants inothercomponents

of

theassaybuffer(e.g.

sucrose). Thefree

[Ca2+]

inthe absenceofATP(and

proteo-liposomes/membranes)

wasusedas anestimatefor total

Ca2+

when calculating specific

Ca21

uptake rates. This approxi-mation is valid because the assay medium contained no anionswithhighaffinity for

Ca2

. Inthepresenceof2.5 mm ATP, thefree

[Ca2+1

wasabout 40

AM.

ATPaseAssay

ATPase activity was measured underidentical conditions

to

Ca2+

uptake,butfor20 to30 min or asindicatedin

figure

legends. Liberated phosphatewasmeasured withamodified Baginskiprocedure (36).ATPaseactivityof fractionsdirectly

elutedfromthe column wasmeasuredunder similar

condi-tionsafter

mixing

0.05 to0.1 mLof eluate

(after

addition of

CaCl2)

withanequalvolumeof bufferA

supplemented

with 50 mM

KCl,

0.2%

(w/v)

BSA, 5

mM

MgCl2,

and2 mm DTT. The reaction was startedbytheaddition of2.5mmATP.

RECONSTITUTION OF A CALMODULIN-STIMULATED Ca2+ PUMP 0 9-

*IC

0.04

I

CL 0.03 -a E . 0.02 0 0. I-0.01i 0 1 3 5 7 9 11 13 15 Fraction, nr

Figure 1. Elution profile from the CaM agarose column showing protein (*), and ATPase activity assayed in the absence (0) and presence(0)of 0.35 Mm CaM.Fractionswerecollected immediately after additionofelutionbuffer.Fractionvolume was 1.6 mL; ATPase waseluted inthe presenceof Triton X-100 andphospholipid.

Other

EnzymeActivities

Antimycin A-insensitive NADH-Cyt c reductase activity wasmeasuredessentiallyasdescribedinref. 1.Cytcoxidase

wasmeasuredat

270C

in 1 mLof bufferBplus 25mm

KCl,

40 uM

dithionite-reduced

Cyt c (Sigma C7752), and

0.02%

(w/v)

TritonX-100.Thereaction wasstarted

by

the addition of membranes (40 ug of membrane proteinor 25-50 ML of

sucrosegradient

fraction)

and the oxidation ofCytcrecorded

at550 nm. An extinctioncoefficient of 19

mm-'

cm-'

for

Cyt

cwasused. Pyrophosphatase activitywasmeasuredat

270C

in0.1 mLof bufferAplus7.5mM

MgCl2,

100 mm

Kil,

0.1%

(w/v)

BSA, and 15 ,ug of membrane protein. The reaction wasstarted by the addition of0.2 mmNa2P207 andrun for 30 min. Liberated

phosphate

and

glucan synthase

II were

measuredasdescribedinref. 36.

Protein

Analysis and

Determination

100 Protein was measured with a modified Bradford procedure

with BSA as the standard (33). When it was necessary to

Xu

avoid interference with

lipid, protein

was first

precipitated

75 0 with methanol-chloroform-H20 (28). In later experiments,

* the method of

Kaplan

andPedersen

(20)

wasusedfor quan-50 titation of

protein

in thepresenceof

large

amounts of

lipid.

e The two methods gave similar results, but the latter was more

* reproducible. SDS-PAGE was carried out on 7.5 to 15%

25 gradient gels essentially according to Laemmli (23).

0 RESULTS

Solubilization and Purification of CaM-Stimulated

Ca21

-ATPase

Ca2"

uptake with washed cauliflower microsomes was

stimulated about 200% by 0.5

uM

CaM, whereas ATPase activity measured under identical conditions (i.e. in the ab-sence of detergent) showed only about 14% stimulation (Table I).After solubilizationof the membranes with Triton X-100 andapplication of the solubilized material to aCaM agarosecolumn,afraction

(approximately

0.3 mgof protein) could be eluted with EGTA that showeda CaM-stimulated

(50-100%

stimulation)ATPase activityof 0.4to0.8

,umol

Pi min- (mg

protein)-'

inthepresenceofamixedphospholipid

preparation (Fig. 1). Triton X-100 may have inhibited the

isolatedATPase,because theactivity wasmuchhigher when elutedinCHAPS(seebelow).

The fractions eluted from the CaM-affinity column were

analyzed by SDS-PAGE. Coomassie-stained polypeptides

werelocatedatabout115, 52, 36,and 17to 21 kD (Fig. 2B).

Ofthese, only the 115-kDpolypeptide formed a phospory-lated intermediate after incubation with

[y-32P]ATP.

This phosphorylation was strongly stimulated by

Ca2"

(Fig. 3).

kD

Phosphorylated Intermediate

Formation

170-A B

C

D

Phosphorylated

intermediateformation wascarriedoutin the presenceof20 mm

Tris-HCl,

pH 7.4, 12.5

MM

MgCl2,

50

Mm

CaC12,

15 Mgof

purified

andreconstituted

Ca2+-ATPase,

37 kBq of

[Y-32P]ATP,

and ±0.5 mm EGTA in a volume of 0.5 mL at

00C

for 15 s. The reaction was started with

[.y-32P]ATP

andstopped bytheadditionof2mL of methanol that hadbeenacidifiedwithHCO(1part 4MHCOto100parts

methanol). The

precipitated

protein was concentrated

by

subsequent additions of choroform and

H20

asdescribedin

ref. 28. Proteins were

separated

by SDS-PAGE in anacidic gel

(5% acrylamide,

pH5.5)

essentially

asdescribedinref.4.

AfterCoomassie staining, the

gel

wasdried andexposed for

18dat

-700C

withFujiRXMedicalX-rayfilminthe presence

ofaDuPontCronex

Lightning-plus

intensifierscreen.

Precip-itation with TCA gaveamuchstrongerincorporation of

32p

(<12hofexposureneeded)thanwith the method described

above, but TCA precipitation could not be used with the

reconstituted

Ca2+-ATPase

because the

lipid

also

precipitated

and interferedwithSDS-PAGE.

97--

- 115----.

55--36-_

20----

-Figure 2. SDS-PAGE of fractions eluted from the CaM agarose column afterwashingwith a medium containing Triton X-100 (B) and CHAPS (C). Aand D, Standard molecular mass markers. The gelswere stained with Coomassie brilliant blue R-250. Molecular massstandards (BoehringerMannheimCombithek kit) in order of decreasing molecular mass were: a2-macroglobulin,

phosphoryl-ase b, glutamate dehydrogenase, lactate dehydrogenase, trypsin inhibitor.

ASKERLUND AND EVANS

A. B.C

k1 170 116- 97-36-W

20-Figure3. Analysis of thephosphorylated intermediate of the

affin-ity-purified, CaM-stimulated Ca24-ATPase after SDS-PAGE in an

acidic gel. Phosphorylation wascarried out with [_Y-32P]ATP in B,

the absence (+0.5 mm EGTA) or C, the presence of Ca2+. A,

Coomassie-stainedstandard molecularmassmarkersasin Figure2, plus ,B-galactosidase (116kD).

These resultsprovidestrong evidence thatthe 115-kD poly-peptide represents theCa2+-ATPase (13).

Effects ofDetergentsonthe Purified CaM-Stimulated Ca2 -ATPase

Theeffects of three different

detergents

weretestedonthe activityofthe

purified

Ca2+-ATPase

tofindthemostsuitable

forreconstitution

by

detergent dialysis (29). CHAPS,

MEGA 8, and octyl

glucoside

were chosenbecause

they

all have a

highcmc (24)and,

therefore,

are

easily

removed

by dialysis.

MEGA8and

octyl glucoside

were

strongly inhibitory

attheir cmc(58 and 25 mm,

respectively),

whereas CHAPS (cmc =

6.5

mM)

stimulated the activityatconcentrationsupto5 mm

and

only partly

inhibiteditat10 mm(Fig.

4).

CaMstimulation

was seen inthepresenceof allthree

detergents.

Reconstitu-tionofthe

Ca2+-ATPase

withoctylglucosideresultedinmuch

lower activity than with CHAPS

(results

not

shown).

This

indicated that octyl

glucoside

irreversibly

deactivated the enzyme, ashas been

reported

forthe

Na+/K+-ATPase

(10).

For all subsequent reconstitution experiments, CHAPS was used. We did not investigate if theinhibition by MEGA 8 wasirreversible.

ReconstitutionofCaM-Stimulated

Ca2" Uptake

and

ATPase

Activity

Forreconstitution, Triton X-100 wasexchanged forCHAPS while the ATPase was still bound to the CaM agarose column. No activity could bedetected in the eluate afterchange of

detergent. The CHAPS wash resultedinsomepurification of

theCa2+-ATPaseasjudged from theincrease in intensity of

the115-kDpolypeptide relativetothe other

polypeptides,

as

visualized byCoomassie-stained SDS-PAGE of the column eluate(cf. Fig. 2,Band C). The intensityof the low molecular

masspolypeptides(17-21 kD)in particularwasdiminished by theCHAPSwash. Phospholipid, includedin thebuffers

tostabilizetheATPase, wasremovedinthesamestep tobe replaced by fresh phospholipidof higher purity during re-constitution. TheATPase activity wasstill highafter elution

in CHAPS-containing buffer with nolipid added (Table I),

suggestingthat the enzyme was notcompletely delipidated by thistreatment(9, 27).

Insertion of the Ca2+-ATPase into liposomes using the

CHAPS-dialysis method resulted in a successful reconstitu-tion of

Ca2+

uptake(Fig. 5). Toobtain maximal activities, it was necessary topreincubate theproteoliposomesfor about

20 min in assaymedium.Ifpreincubationwas omitted, a

lag-phasewasobserved bothfor

Ca2"

uptakeand ATPase activity

(resultsnotshown).Preincubation with either

Ca21

plus

Mg2+

or CaM has been necessary to achieve maximal rates of phosphorylation ofthe

Ca2+-ATPase

from erythrocytes(13).

Preincubation resulted insome binding of

Ca2+

tothe

pro-teoliposomes (Fig. 5; time=0). Subsequent addition ofATP

resulted in arapid accumulation of

Ca2',

reaching a steady

stateafter about15 min. All the accumulated

Ca2"

couldbe released by the addition of the

Ca2+

ionophore A23187. Measurement ofATPase activityunder identical conditions

as

Ca2"

uptakegave a

Ca2+/ATP

ratioof about 0.02during the approximately linear, initial phase of

Ca2"

uptake (Fig. 5). The ATPase activity was linear for at least 30 min. Inclusionofthe

Ca2"

ionophore A23187 in the assay medium

stimulatedthe ATPaseactivityabout70 to 100% during the wholetime course (Fig. 5).

When the

Ca2+-ATPase

was reconstituted into liposomes

0) |\ MEGA8

E

--- MEGA8+CaM

-A.0| Octyl glucoside

E10 102-0 0 5 M Octylglucoside C*cM 0 E 0.0 0 10 20 30 40 50 Detergent, mM

FigureC4 Effects ofdifferentdetergentsonthepurifiedCa24-ATPase

intheabsence(opensymbols)andpresence(filledsymbols)of0.35

jLm

CaM.ThedetergentstestedwereCHAPS(0, 0,cmc=6.5mm), MEGA8

(OK,

, cmc= 58 mm),and octyl glucoside (A, A, cmc =

25mM). The experiment was carried outwith Ca2+-ATPaseeluted

inthe presence ofTritonX-100andphospholipid.

RECONSTITUTION OF A CALMODULIN-STIMULATED Ca2` PUMP 100 80 >o --I la 7 Go o 3 40 2. 20 ! 0 c) E 0 E -60 O._ =3 Q CZ 0.50 0.40 0.30 0.20 0.10 0. E E aZ) 0

I1-Figure5. ATP-dependentCa2" uptake (e) and ATPaseactivitywith purified and reconstituted Ca2+-ATPase in the absence (0) and

presence (0) of 5 uM of the Ca2" ionophore A23187. The

Ca2+-ATPase was reconstituted into liposomes without oxalate. Accu-mulated Ca2+ could be released by 5 gM A23187. The standard

assaymediumwasusedexceptthat the concentrations ofKCI and

MgCI2 were 25 and 2.5 mm, respectively. CaM was present at a

concentrationof0.5 MM.

with entrapped oxalate, the initialrate of

Ca2`

uptake was

10-to20-fold higher than without entrapped

Ca2`

chelator, whereas the ATPase activitywasthe sameunderboth

con-ditions (Fig. 6; the absoluteratesinFigs. 5and 6 shouldnot

be directly compared, however, because the concentrations ofMgC12 andKCI weredifferent inthese two experiments;

seefigure legends). Thus, theCa2+/ATPratio wasincreased

toabout 0.2by entrapping oxalatein theproteoliposomes. BothCa2+uptake and ATPase activity of the reconstituted

Ca2+-ATPase were stimulated about 300% by CaM (Figs. 6

and7; Table I) in comparison with the 50to100% stimulation obtained before reconstitution (Figs. 1 and 4; Table I). To

3.5* 120

2.8

E 90 -g

73~~~~~~~~~~~

E.1 .1

Fiur 3.APdpnetC2 pae(,* n Taeatvt

JU C)

0.0*

0 6 12 24 30

Time, min

Figure 6. ATP-dependent Ca`~uptake (E, U)and ATPase activity

(0, 0) with purified and reconstitutedATPase intheabsence(open symbols) and presence(filled symbols) of 0.5 MmCaM.The

Ca2+-ATPasewasreconstituted intoliposomes that contained entrapped

oxalate.

0.0 0.2 0.4 0.6 0.8 1.0

Calmodulin, ,uM

Figure 7. Effect of different concentrations of CaM on

ATP-de-pendent Ca2+ uptake (0) andATPaseactivity(0) with purified

Ca2+-ATPase reconstituted into liposomes. The proteoliposomes

con-tainedentrapped oxalate.

someextent,thisdifferencemayhave been duetothemuch higher concentration of

Ca2"

in the assay medium during

measurements of ATPase activity with the enzymedirectly

eluted from the column (see below). The concentration of CaM needed for full stimulationof

Ca2"

uptakewasslightly

lower than for full stimulation of ATPase activity (Fig. 7). This was also observed with microsomal membranes (data

notshown).

Further Characterization of the Reconstituted Ca2+-ATPase TheATPaseactivity of the reconstitutedCa2+-ATPasewas

almost totallydependenton

Ca2`

(Fig. 8). Thehalf-maximal

ratewasreachedatabout 7 Mmfree[Ca2+]bothinthe absence

andpresence of CaM. The ATPaseactivitywasinhibitedby

concentrations of free [Ca2+] above 1 mm, especially in the

presenceof CaM(Fig. 8). Both ATPase and ATP-dependent

Ca2+ uptake by the reconstituted enzyme were completely

dependenton

Mg2"

and showedoptimaatabout 5mMMgC12

(Fig.9).

Thestimulationof the ATPaseactivity and ATP-dependent Ca2+ uptake activityof thepurifiedand reconstituted Ca2+-ATPasebyKCI,KNO3, and NaCl was3-fold orhigher (Fig.

10). The change in ionic strengthwasnotcontrolled for and

mayaccount,inpart, for the observed stimulation. The Ca2+ uptake showedanoptimumat100mmKCIorNaCl,whereas

theATPaseactivity continuedtoincrease abovethis

concen-tration (Fig. 10). The ATPase activity and Ca2+ uptake of

cauliflower microsomes showedno orverylittle stimulation by anincrease in KC1 from 25 to 100mm(data notshown). This suggests that the activities obtained with membranes

werelessaffectedbyionicstrength than those of thepurified

and reconstituted enzyme. Monovalent cations stimulate ATPaseactivity and Ca2+ uptake of erythrocyte ghosts by 30

to 100% (13), and ATP-dependent Ca2+ uptake by plant

350 280 210 140 70 E .5 2 c 0.1 a C) I a 0 0 10 20 30 40 50

Time,

min l1675

Plant Physiol. Vol. 100, 1992

branefraction representedabout12% ofthe totalmicrosomal

protein

and was more than 4 times enriched in the

plasma

membranemarker glucansynthase II(TableIII). Theplasma

+

CaM

membraneswerealmost

completely depleted

in

Cyt

coxidase

activity, a marker for the inner mitochondrial membrane.

Pyrophosphataseactivity (atonoplast marker) andnonlatent,

antimycin A-insensitive NADH-Cyt c reductase activity (a markerfor theER) were alsostrongly depletedinthe plasma

membranes: Thespecificactivitiesofthesemarkerswere 19 and 6% of the activities inthe microsomalfraction,

respec-tively. The antimycin A-insensitive NADH-Cyt c reductase

activity inboth the microsomal and intracellularmembrane

fractions was inhibited

by

0.015%

(w/v)

Triton

X-100,

/ -CaM whereas the

activity

in the

plasma

membrane fraction was

strongly stimulatedby thedetergent (TableIII).Thislatency ofNADH-Cytcreductaseinplasma membrane vesiclesis in - 7 -6 - 5 - 4 - 3 - 2

agreement

with earlier

investigations (1, 36)

and with the

2+

finding

that a

NADH-Cyt

b5

reductase-like enzyme with

Free [Ca ], log M substrate

binding

sites onthe

cytoplasmic

surfaceis present

in plantplasmamembranes (1).

ATPase activity with purified and reconstituted Ca2+- In the absence of CaM, the specific ATP-dependent Ca22 different concentrations of free [Ca2+] in the absence

uptae

ase100

h e in the

plasma

m

enestCa

ibols) and presence (filled symbols) of 0.5

gM

CaM. The to the microsomal fraction and the intracellular membrane

hium was bufferB plus2.5 mMMgCI2,25 mm KCI,0.1%

depletedninaplasma

fraction membrane

1 mMEGTA, 0.5 mm ATP, and varying amounts ofCaCI2.

frac,hon

depleted

ai

plasma

membranes.

In the

presence

of

jwas measured with aCa2+electrode.

CaM,

however,

the

activity

wassimilar in allthree fractions (TableIII). Thus, CaMincreasedCa2+uptakeabout200% in the microsomal and intracellular membrane fractions, but nembrane vesiclesis stimulated to asimilar degree only about 50% in theplasma membrane fraction. Oftotal

CaM-stimulated

Ca21

uptake, only about

6%

was associated

ras

the mosteffective nucleotide in supporting the with the

plasma

membranes

(Table III).

ake andNTPase activityof the reconstitutedenzyme. To

investigate

further the cellularlocation of CaM-stimu-ther nucleotides tested, GTP was the next most lated Ca2+

uptake,

microsomal membranes were

separated

,exhibitingratesthatwereabout50% of thosewith on acontinuoussucrose

gradient

(Fig.

14).

Toobtaina

reason-h in thepresence (Fig. 11)andthe absence of CaM able

separation

of themarker enzymes, it was necessary to

(datanotshown). MaximalactivitywithATPwas seen atpH

7.0.CaM stimulation wasalsostrongestatthispH (Fig. 12). Both ATPase activity and ATP-dependent

Ca21

uptake

with the purified and reconstituted

Ca2+-ATPase

were

strongly inhibitedby vanadate, showingthataP-type ATP-asewasresponsible for the activities(Fig. 13). Aclear

inhi-bition of

Ca2+

uptake

was seen

only

when oxalate-loaded proteoliposomes were used. This may be explained by the ability of high intravesicular levels of free

Ca2+

to protect

against vanadate inhibiton (13). The ATPase activity was

only slightly inhibited

by

thapsigargin (Table

II), a potent

and

specific

inhibitor of theER (and not the plasma mem-brane or sarcoplasmic reticulum) vertebrate

Ca2+-ATPase

(34).Cyclopiazonic acid(aspecific inhibitor ofthevertebrate sarcoplasmic reticulum

Ca2+-ATPase;

ref. 32) had noeffect

onthecauliflower

Ca2+-ATPase

(TableII). Thelackof effect

is in agreement with the observations of Hsieh et al. (19), who found no effect of cyclopiazonic acid on the

CaM-stimulated componentofATP- or GTP-driven

Ca2"

uptake

with carrotcellmembranes. Erythrosin B stronglyinhibited

the ATPaseactivity, with a

Ki

of12

gM

(TableII). Cellular Location of theCaM-Stimulated

Ca2"-ATPase

Plasma membraneswereprepared from the crude

micro-somalfraction by two-phase partitioning.Theplasma

mem-cm E E 0 E C: a (a. + C0 C) E 0 E AL :CIO 0-0 5 10 MgCI2.mM

Figure9. Effect ofMg2+onATP-dependent Ca2+ uptake and

ATP-ase activity by purified and reconstituted Ca2+-ATPase. CaM was

presentataconcentrationof0.5 ;LM.The Ca2+-ATPasewas

recon-stitutedintoliposomeswithout oxalate. 1.5 0) E p 1.0 E a) Cn 0.5 0.0 Figure 8. ATPase at (open syrT assay mec (w/v) BSA, Free[Ca21 plasman (25). ATP vw

Ca2`

upta Of the c

effective,

ATPbotl

RECONSTITUTION OF A CALMODULIN-STIMULATED Ca2+ PUMP 0

50

100

Salt, mM

E E 0. E c% -c CZ Q. l CIO

O'

150 200 0 50 100

Salt,

mM 150 200

Figure 10. Effect of monovalent cationson ATPaseactivity (A) and ATP-dependent Ca2" uptake (B) by purified and reconstituted

Ca2`-ATPase. CaMwaspresent ataconcentration of0.5 Mm.TheCa2'-ATPasewasreconstituted intoliposomes without oxalate.

usenonfrozen membranes that had been concentrated on a sucrose cushion rather than by pelleting, and toinclude an excess ofEDTA inall buffers.

Ca2"

uptakeinthe absence of

CaMwaslow andmore orlessevenly distributed acrossthe

gradient. In thepresence ofCaM,

Ca2`

uptakewas up to 6 times higher and showed amaximum atadensity of about

1.12gmL-' (29% [w/w] sucrose), butasignificant degree of

activitywasalsopresentinbothlighter and heavier fractions

(Fig. 14, A and B). The peak of CaM-stimulated

Ca2"

uptake coincided with the lighter of two peaks of antimycin A-insensitive NADH-Cyt c reductase activity, probably

repre-senting smoothER, andwaswellseparated from the glucan

synthase II (plasma membrane marker) and Cyt c oxidase

activities(Fig. 14,B andC).The heavierpeakofNADH-Cyt

0 0o -_ a) a) 100 80 60 40 20

ATP

UTP

CTP

GTP

Figure 11. Relative NTPase andCall uptake(Q)activityof the

purified and reconstituted Ca2+-ATPase with different nucleoside

triphosphates (0.5 mM)inthepresenceof 0.5 MmCaM.

creductase isprobably duetorough ER, plasma membrane, and other membranes containing this activity (TableIII;refs.

1 and 36). CaM-stimulated

Ca2"

uptake did not colocalize with pyrophosphatase activity(tonoplast marker; Fig. 14C). Theprotonophore FCCP (5

JtM)

hadnosignificant effecton

Ca2"

uptake inany of thefractions from the gradient,

indi-cating thata

Ca2+/nH'

antiporter(5, 12)wasnotresponsible forasignificantpartoftheactivity in the microsomal fraction

(datanotshown).

2.5 C) E 2.0 E 1.5 a 1.0 en CZ) tL < 0.5 0.0 6 6.5 7 7.5 8 8.5 9 pH

Figure12. Effect of pH onthe ATPase activityof the purified and

reconstituted Ca2+-ATPasein the absence(E)and presence(F) of

0.5 AM CaM. The Ca2+-ATPase was reconstituted into liposomes

without oxalate. Priortoadditiontotheassaymedium,the ATPase

was dialyzed against 2.5 mm Mops-BTP, pH 7.2, 0.33 M sucrose,

and2 mmDTT. Theassay bufferwas25 mmMes-BTP, pH 6to9.

Theassaymedium included5AM Ca2+ ionophoreA23187. 6

5

4 3 2 CE E E

a,)

0e E 1 0 1677

Plant Physiol. Vol. 100, 1992 100 80 0 C.Z a) a) cc 60 40 20 0 50 100

150

200 Vanadate,

jM

Figure 13. Effectof vanadateon ATPaseactivity and ATP-depend-ent Ca2" uptake by the purified and reconstituted Ca24-ATPase. The proteoliposomescontainedentrappedoxalate. CaM was pres-ent at aconcentrationof 0.5zlM.

DISCUSSION

The specific ATPase activity of the reconstituted

cauli-flower ATPase was high in the presence ofCaM

(approxi-mately 5

gmol min-'

[mg

protein]-',

to be compared with 15-20 ,umol

min-'

[mg

protein]-'

reported for the purified CaM-stimulatedATPasefrom

erythrocytes

[13]). Calculation of the

degree

of

purification

ofthis activitycannotbe based

on the total ATPase activity of the membranes of origin because thesecontainotherATP hydrolyzers,

including

the

plasma membrane H+-ATPase. This is clearly illustrated by

comparisonof CaM stimulationofATPhydrolysisin micro-somes

(14%;

TableI) with CaM stimulation of

ATP-depend-TableII. EffectofDifferent InhibitorsonATPaseActivitywith PurifiedandReconstitutedCa2'-ATPase

Standard assay medium plus0.5,AM CaM wasused, except that the concentrationofATPwas0.5mminexperiments with cyclopia-zonicacid.

Inhibitor ATPaseActivity

% ofcontrol Thapsigargin 0.1

JAM

87 4.0gm 78 Cyclopiazonicacid 0.3 nmol(mg

protein)-'

96 10nmol (mgprotein)-' 98 ErythrosinB 12

AM

48 100JM 10

Table111. DistributionofCaM-StimulatedCa2" Uptake, Marker Enzyme Activities,and TotalProteinbetween Plasma Membrane,

IntracellularMembrane,andMicrosomal Membrane Fractions

The plasma membrane and intracellular membrane fractions

wereobtainedby two-phasepartitioningof the microsomal fraction andcorrespondtofractionsdesignated U3+U3' and L2 in ref. 22, respectively.All activitiesarespecificand expressedin nmol min-1 (mgprotein)-1. Data aremeans ±SDfrom two different membrane preparations, except for antimycin A-insensitive NADH-Cyt c

re-ductase and pyrophosphatase(onemembrane preparation).

Microsomal Plasma Intracellular

Membranes Membranes Membranes

Total protein(mg) 53.3± 2.7 6.2± 0 43 ±2.9 Ca24uptake -CaM 9.7± 1.6 16.6 ± 3.8 8.2± 1.6 +0.5

jLM

CaM 26.8 ± 2.3 24.5 ± 2.4 24.6± 2.6 CaM-stimulated 17.0±0.7 8.0± 1.3 16.5± 1.0 Cyt coxidase 183± 18.3 4.4±0.4 298±48.9 Antimycin A-insensi-tiveNADH-Cyt creductase -Triton X-100 330 19.4 370 +TritonX-100 252 141 252 Glucan synthase11 131 ± 39.4 549± 156 61.6± 17.6 Pyrophosphatase 82.2 15.4 83.3

ent

Ca2"

uptake (200%; Table I). Thedegreeof purification

can, however, be based on CaM-stimulatedactivities, both

Ca2"

uptake and ATP hydrolysis (Table I). When

CaM-stimulated ATPase activity with microsomes is compared with the CaM-stimulated ATPase activity after purification andreconstitution (conditions closestto those in the

mem-brane),apurification factor of about 120 isachieved(Table I). CaM-stimulated

Ca2"

uptake activity was purified only

about 20-fold duetothe0.1 to 0.2

Ca24/ATP

ratio(Table I). The erythrocyte

Ca24-ATPase

canbepurifiedto near ho-mogeneity (approximately 250-fold enrichment of ATPase activity) with CaM-affinity chromatography and has a

mo-lecularmassof130 to140 kD (8, 9, 13, 27). CaM-stimulated

ATPases in someother animal cellsmayhaveaslightly lower molecularmass(8). Withcauliflower, the fraction eluted from the column showeda Coomassie-stained band (orpossibly

twobandsveryclosetogether)at 115kD together withafew polypeptides oflowermolecularmass(Fig. 2). Onlythe

115-kDpolypeptide formeda phosphorylatedintermediate, and

this formation was

Ca2"

dependent (Fig. 3). This provides strongevidence that the 115-kDpolypeptiderepresents the intact

Ca24-ATPase,

althoughthe ATPasefrommaize micro-somes hasbeen reported to have a molecular mass of 140 kD(4, 12).Recently,

Ca2"-dependent

formation ofa120-kD phosphoenzyme in membranes from carrot cellswas

dem-onstrated(19). The view that thepurified cauliflower

Ca24-ATPasewas intact isfurther supportedbythestrong stimu-lation by CaM of ATPase activity and

Ca2"

uptake after

reconstitution (Figs. 6 and7;TableI).

When reconstituted into liposomes, the

Ca24-ATPase

cat-alyzedanATP-dependent accumulation of

Ca2"

(Figs. 5 and

6). ThatCa24 wasindeedtakenup into theproteoliposomes byatransmembraneprocess is shownbythecomplete release

RECONSTITUTION OF A CALMODULIN-STIMULATED Ca'+ PUMP 0) a-Q~ -0 cc Er: 0 2 4 6 8 10 12 14 Fraction, number 0 2 4 6 8 10 12 14 Fraction, number 0 2 4 6 8 10 12 14 Fraction, number

Figure 14. Distribution of different enzyme activities and protein after separation of microsomal membranes on a continuous sucrose

gradient.A,*,inpercent(w/w)sucrose;0,in ugprotein. B,0,ATP-dependentCa"+uptake withoutCaM(100%= 0.49nmolmin-1);*,

ATP-dependentCa"+ uptake with0.5 MmCaM (100%=0.49nmol min-'); E,glucan synthase11 (100%=4.0 nmolmin-').C, pyrophosphatase

activity (100%= 12 nmol min-1); *, Cytcoxidase (100%=61 nmol min-');A, antimycinA-insensitive NADH-Cytcreductase (100%= 21

nmolmin-1;measuredinthe absenceof TritonX-100).

of

Ca2"

by the Ca2+ ionophore A23187 (Fig. 5). A direct comparison of Ca2+uptake and ATPase activity using

pro-teoliposomes without entrapped Ca2+ chelatorgave a

Ca"/

ATP ratioofabout 0.02 (Fig. 5). This low ratio and the fact that the ATPase activity failed to reach a plateau as the

uptake of

Ca2`

progressed with time (Fig. 5) indicated that theproteoliposomeswerepermeabletoCa2 . Consequently,

the Ca2+ ionophore A23187 stimulated the ATPase activity only 70to 100%comparedto 10-foldwith the reconstituted erythrocyte ATPase (9). Entrapment of oxalate inside the proteoliposomes increased therateof

Ca2"

uptake 10-to 20-fold withoutaffecting the ATPase activity and thus increased the coupling ratio to about 0.2 (Fig. 6; Table I). In a few experiments where a lower concentration of ATP than the

standard 2.5 mm was used, coupling ratios as high as 0.5

were obtained (data not shown). This should be compared with the Ca2+/ATP ratio of 0.63 for the CaM-stimulated activity in cauliflower microsomes(Table I) and a maximal Ca2+/ATPratio of 0.9to2.1 for thepurifiedandreconstituted erythrocyte Ca2+-ATPase (8, 9, 13). The highest coupling ratioswith theerythrocyte ATPase have been obtained after reconstitution in 'asolectin, a crudelipid mixture that

pro-duces highly sealed liposomes (9, 27). We used a slightly

higher purity of lipid thatmay have resultedin moreleaky

liposomes. Residual CHAPSmayalso havecausedthe

leak-iness of theproteoliposomes to Ca2+ in the absence of

en-trapped oxalate.

BothCa2+ uptake and ATPase activity measured with the reconstituted Ca2+-ATPase were stimulated about 300% by

CaM (Fig. 7; Table I). The erythrocyte Ca2+-ATPase only showssuchastrong stimulationbyCaM in thepresence of

purephosphatidylcholine.In thepresenceof acidic phospho-lipids,e.g.after reconstitution into'asolectin, theerythrocyte ATPaseisfully stimulated (i.e. showsnostimulationby CaM;

8, 9, 13, 27). The phospholipidmixture used here isrelatively crude(approximately 40% phosphatidylcholine, as supplied

by Sigma Chemical Co.; the furtherpurification carriedout

mainlyremoves contaminating protein,

Ca2',

and oxidation products). Thestrong CaMstimulationseenwith the

recon-stituted cauliflower ATPase(Fig. 7)is thereforesurprising.

The Km

[Ca2"]

for ATPase activity with the purified and reconstituted ATPase wasabout 7 gM inboth the presence

and absence of CaM (Fig. 8). This is different from the purified and reconstituted Ca2+-ATPase from animal plasma membranes, which showsa significantly higher affinity for

Ca2"

in the presence of CaM [Km

(Ca2")

1.5 gm; ref. 13] than in itsabsence [Km

(Ca2")

~ 20 uM]. The ATPase activity wasinhibitedby concentrations of free

[Ca2"]

above 1 mm

(Fig. 8), resembling thecaseinanimals(9). The properties of the cauliflower ATPasemay havechanged during purifica-tionorafter reconstitutionsothatCaMcan nolonger increase

itsaffinity for

Ca2".

Reported Km

(Ca2")

for

Ca2"-dependent

ATPase activityand ATP-dependent

Ca2"

uptakerange

be-tween0.07 and 6 Mmfor bothplant plasma membrane vesicles

and ER(5, 7, 12, 14-16, 25, 30).

The cellular location of the CaM-stimulated

Ca2"

pumpin

plants has long been a matter of controversy (5, 12). The analogy with animals wouldsuggestthat theactivity should bepresentinthe plasma membrane only, and both this and several other investigations indicate that the plant plasma membranemaycontainaCaM-stimulated

Ca2"

pump(Table

III; 5, 12, 30, 31). In other studies, no CaM stimulation of

plasma membrane ATPase activityor

Ca2"

uptake has been

found(15, 16, 21). In thepresentwork, however, weclearly

show that themajorpartof theATP-dependent CaM-stim-ulated

Ca2"

uptake isnotintheplasma membrane, because only about 6% of the totalactivity wasassociated with the

plasma membrane fraction and the remainingpartwith the intracellular membrane fraction(Table III).

Afterseparation of the microsomal membraneson a

con-tinuous sucrose gradient, the peak of CaM-stimulated

Ca2"

uptake coincided with the lighter oftwopeaks of antimycin A-insensitiveNADH-Cytcreductaseactivity ata densityof

1.12 g

mL-1

(the approximate density of smooth ER),

sug-gesting that ER is the major cellular location of the CaM-stimulated Ca2+-ATPase in cauliflower inflorescences (Fig. 14, B andC). The ER also seemedtobe the location of CaM-stimulated Ca2+-ATPase activity and

Ca2"

uptake in

mem-branes from maize coleoptiles (3) and carrot cells (19). In other studies of

Ca2"

uptakeand ATP hydrolysis with

frac--0

Cl)

Plant Physiol. Vol. 100, 1992

tions enriched in ER, no effect of CaM was reported (7).

Possibly, the CaM-stimulated ATPase is present only in certaintissuesorspecies,or CaM stimulation may bemasked byhigh levelsof endogenous CaM in membranes (5).

Although the

Ca2"

pump from cauliflowerinflorescences

is CaM stimulated, it shows several properties that are dif-ferent fromtheCaM-stimulated Ca2+pump inanimalplasma membranes and that are more orless similarto thoseof the

animal

endoplasmic/sarcoplasmic-type

Ca2+ pump, whichis notdirectly regulated by CaM. First, the major part ofthe activity is notlocatedinthe plasmamembrane,butprobably

inthe ER.Second,thecauliflower Ca2+-ATPasehas aslightly

lowermolecularmass (115kD; Figs. 2and3) than the CaM-stimulated animalCa2+-ATPase,which has amolecular mass

of130 to 140 kD in most cells (8, 9, 13, 27). The indication

ofadouble bandseenwith thecauliflowerpreparations may

beduetothe presenceoftwo or moreisoforms oftheenzyme.

Third,thecauliflowerATPase is much less specific forATP (Fig. 11) than the CaM-stimulated Ca2+-ATPasein animals, which uses only ATP as an energydonor(13).

Ca2`

uptake

with plant plasma membrane vesicles andER from garden

cress rootsshowsasimilar

specificity

forATP,asshown here forthe

purified

Ca2+-ATPase,with GTP(and ITP) exhibiting

rates that are 25 to

50%

of those with ATP (6, 15, 16, 25, 30), whereas the Ca2+-ATPase activity in red beet ER was

reportedto use ATP only(14). Fourth,thecauliflower

Ca2+-ATPase showeda relatively low sensitivity to erythrosin B

(Ki

=12um;TableII) in comparison tothe erythrocyte ATPase

(Ki

= 70 nM; ref. 13 and refs. therein). This was not a

purification artifactbecauseATP-dependent

Ca2"

uptakeby afraction enriched inER (obtained bysucrosegradient

cen-trifugation) wassimilarly affected tothe

purified

Ca2+-ATP-ase (data not shown). In contrast,

Ca2`

uptake with the

plasma membranefraction (obtained by two-phase partition-ing)showed amuch highersensitivity toerythrosinB

(Ki

= 0.6

AM).

This difference in sensitivity to erythrosin B may represent afundamentaldifferenceinthe ATPbindingsite(s) between plantERand plasma membrane Ca2+ pumps (13).

Amonoclonal antibodyraisedagainst theerythrocyte

Ca2+-ATPase(5F10,kindlysupplied by Prof.J. T. Penniston, Mayo

Foundation, Rochester, MN)gave no

significant

reactionwith

intracellular and plasma membrane fractions from

cauli-flower in western blots (with the intracellular membrane fraction,twopolypeptides,at 14and16kD,wererecognized; data notshown),even

though

itshows cross-reactivitywith

different plasma membrane

Ca2+

pumps inanimals (2, and

refs. therein).

Insummary, we have achieved thefirstreconstitutionofa

purifiedplantCaM-stimulated

Ca2"-pumping (Ca2"

+

Mg2+)-ATPase. Themolecularmassofthis ATPase was determined tobe 115kD, and the enzyme was mainlyconfinedto alight intracellular membranefraction,probably the ER, in Brassica

oleraceaL.

ACKNOWLEDGMENTS

Wewish to thankProfessorChrister Larsson(Departmentof Plant

Biochemistry, Lund, Sweden) for critically reading the manuscript. Wewould also like to thank Mr. Kevin Donachie (Department of PlantSciences,Oxford, UK)forsuggestionsabout determination of

pyrophosphataseactivityand helpinproducing thefigures.

LITERATURE CITED

1. Askerlund P, Laurent P, Nakagawa H, Kader J-C (1991)

NADH-ferricyandide reductase of leaf plasma membranes.

Partialpurification andimmunologicalrelation to potatotuber

microsomal NADH-ferricyanide reductase and spinach leaf

NADH-nitratereductase. PlantPhysiol95: 6-13

2. Borke, JL,CarideAJ,Yaksh TL, Penniston JT, Kumar R (1989) Cerebrospinal fluid calcium homeostasis: evidence for a

plasmamembraneCa2+-pumpinmammalian choroidplexus. Brain Res489:355-360

3. Brauer D, Schubert C, Tsu S-I (1990) Characterization of a

Ca2+-translocatingATPasefromcorn root microsomes. Physiol

Plant 78:335-344

4. Briars, S-A, Evans DE (1989) The calmodulin-stimulated ATP-aseofmaizecoleoptilesforms a phosphorylated intermediate. BiochemBiophys Res Commun 159: 185-191

5. Briskin DP (1990) Ca21-translocating ATPase of the plant plasmamembrane. Plant Physiol 94: 397-400

6. Buckhout TJ (1983) ATP-dependent Ca2+ transport in endo-plasmic reticulum isolated from roots Lepidium sativum L. Planta 159: 84-90

7. BuckhoutTJ(1984) CharacterizationofCa2+transport in

puri-fied endoplasmic reticulum membrane vesicles from Lepidium sativum L. roots. Plant Physiol 76: 962-967

8. Carafoli E (1991) The calcium pumping ATPase of the plasma membrane. Annu Rev Physiol 53: 531-547

9. Carafoli E, Zurini M (1982) The Ca2+-pumping ATPase of plasma membranes. Purification, reconstitution and proper-ties.Biochim Biophys Acta 683: 279-301

10. Cornelius F (1991) Functional reconstitution of the sodium pump.Kineticsof exchange reactions performed by

reconsti-tutedNa/K-ATPase.Biochim Biophys Acta 1071: 19-66 11. DieterP,MarmieD (1981) Acalmodulin-dependent,microsomal

ATPasefromcorn (ZeaMays L.). FEBS Lett 125: 245-248 12. Evans DE, Briars S-A, Williams LE (1991) Active calcium

transport by plant cellmembranes. J Exp Bot 236: 285-303 13. GarrahanPJ,Rega AF (1990) Plasmamembrane calcium pump.

In F Bronner,ed, Intracellular Calcium Regulation, Wiley-Liss, NewYork, pp 271-303

14. Giannini,JL, Gildensoph LH, Reynolds-Niesman I, Briskin DP (1987)Calcium transport in sealed vesicles from red beet (Beta vulgarisL.) storagetissue.I.Characterizationof aCa2+_

pumpingATPase associated with the endoplasmic reticulum.

Plant Physiol 85:1129-1136

15. GrafP,WeilerEW (1989) ATP-drivenCa2+transport in sealed plasma membrane vesicles prepared by aqueous two-phase partitioning from leaves of Commelina communis L. Physiol

Plant 75:469-478

16. GrafP,WeilerEW (1990) Functional reconstitution of an

ATP-driven Ca2"-transportsystemfromthe plasma membrane of

Commelina communis L. Plant Physiol 94: 634-640

17. Helmke SM, Howard BD (1987) Fractionation and reconstitu-tionof the sarcoplasmic reticulumCa2` pumpsolubilized and

stabilized byCHAPS/lipidmicelles. Membr Biochem 7: 1-22 18. Hepler PK, Wayne RO (1985) Calcium and plant development.

Annu RevPlant Physiol 36: 397-439

19. Hsieh W-L, Pierce WS, Sze H (1991) Calcium-pumping ATPases in vesiclesfrom carrot cells. Stimulation by

calmod-ulinorphosphatidylserine,andformation of a 120 kilodalton phosphoenzyme. Plant Physiol 97: 1535-1544

20. Kaplan RS,Pedersen PL (1989) Sensitive protein assay in pres-enceofhigh levelsof lipid. Methods Enzymol 172: 393-399 21. Kasai M, Muto S (1991) Solubilization and reconstitution of

Ca2` pump fromcorn leafplasma membrane. Plant Physiol 96: 565-570

22. Kjellbom P, Larsson C (1984) Preparation and polypeptide composition of chlorophyll-free plasma membranes from leavesoflight-grown spinach and barley. Physiol Plant 62: 501-509

23. LaemmliUK (1970)Cleavageofstructural proteins during the

assembly of the head of bacteriophage T4. Nature 227: 680-685

RECONSTITUTION OF A CALMODULIN-STIMULATED Ca2+ PUMP

24. NeugebauerJM(1990) Detergents: anoverview.Methods

En-zymol182:239-253

25. OlbeM,SommarinM (1991)ATP-dependentCa2"transportin

wheat root plasma membrane vesicles. Physiol Plant 83:

535-543

26. Palmgren MG,AskerlundP, Fredrikson K, Widell S, Som-marin M, Larsson C (1990) Sealedinside-out and

right-side-out plasma membrane vesicles. Optimal conditions for

for-mationandseparation.PlantPhysiol92:881-890

27. Penniston JT,Filoteo AG, McDonough CS, Carafoli E (1988) Purification, reconstitution, and regulation of plasma

mem-braneCa2+pumps.MethodsEnzymol157:340-351

28. Pohl T(1990)Concentration ofproteinsandremoval of solutes. Methods Enzymol 182: 68-83

29. Racker E (1979)Reconstitution of membraneprocesses.Methods Enzymol55:699-711

30. Rasi-Caldogno F, Carnelli A,DeMichelis MI (1992) Plasma membrane Ca-ATPase of radishseedlings. II. Regulation by

calmodulin. PlantPhysiol98: 1202-1206

31. Robinson C,LarssonC, BuckhoutTJ(1988)Identification ofa

calmodulin-stimulated (Ca24+Mg2+)-ATPase in a plasma

membrane fraction isolated from maize (Zea mays) leaves. Physiol Plant72: 177-184

32. SeidlerNW,JonaI,Vegh M, Martonosi A(1989)Cyclopiazonic

acid isaspecific inhibitor of theCa2+-ATPaseofsarcoplasmic

reticulum.JBiolChem264: 17816-17823

33. Stoscheck CM (1990) Quantitation of protein. MethodsEnzymol

182:50-68

34. Thastrup0,CullenPJ,Dr0bak BK,Hanley MR, Dawson AP

(1990)Thapsigargin,atumorpromoter,discharges intracellu-larCa2`storesby specific inhibition of the endoplasmic retic-ulum Ca2+-ATPase. Proc Natl Acad Sci USA 87: 2466-2470 35. Tsien RY, Rink TJ (1981) Ca2+-selective electrodes: A novel

PVC-gelled neutralcarriermixturecompared with other

cur-rently availablesensors.J Neurosci Methods4: 73-86

36. WidellS, LarssonC (1990)Acritical evaluation of markers used

in plasma membrane purification.In CLarsson, IMM0ller, eds, The PlantPlasma Membrane-Structure, Functionand MolecularBiology. Springer-Verlag, Berlin,pp16-43