This is the published version of a paper published in Plant Physiology.
Citation for the original published paper (version of record):
Palmgren, M., Askerlund, P., Fredrikson, K., Widell, S., Sommarin, M. et al. (1990)
Sealed Inside-Out and Right-Side-Out Plasma Membrane Vesicles: Optimal Conditions for Formation and Separation.
Plant Physiology, 92(4): 871-880
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:
PlantPhysiol. (1990) 92, 871-880 0032-0889/90/92/0871/10/$sl.00/0
Received for publication July28, 1989 and in revised formOctober16, 1989
Sealed Inside-Out and
Right-Side-Out Plasma Membrane
Vesicles'
Optimal Conditions for Formation and
Separation
Michael Gjedde Palmgren, Per Askerlund, Karin Fredrikson, Susanne Widell, Marianne Sommarin, and Christer Larsson*
Department of Plant Biochemistry, and Department of Plant Physiology (S.W.), University of Lund, P. 0. Box 7007,
S-22007Lund, Sweden
ABSTRACT
Plasmamembranepreparations of high purity(about95%) are easily obtained by partitioning in aqueous polymer two-phase systems. These preparations, however, mainly contain sealed right-side-out (apoplastic side out) vesicles. Part of these vesicles havebeen turnedinside-out by freezing andthawing, and sealed inside-out and right-side-outvesiclessubsequently separated by repeating the phase partition step.Increasing the KCI
concentra-tioninthe freeze/thaw mediumaswellasincreasing the number offreeze/thaw cyclessignificantly increased the yield of
inside-out vesicles. At optimal conditions, 15to 25% of total plasma
membraneproteinwas recoveredas inside-outvesicles,
corre-spondingto5to10milligrams of protein from 500 grams of sugar beet(Beta vulgaris L.)leaves. Basedonenzymelatency, trypsin inhibition of NADH-cytochrome c reductase, and H+ pumping capacity, across-contamination of about 20% between thetwo
fractions of oppositely oriented vesicleswas estimated. Thus, preparations containing about 80% inside-out and 80% right-side-out vesicles, respectively, were obtained. ATPase activity
andH+pumpingwerebothcompletelyinhibitedby vanadate(K, ; 10 micromolar), indicating that the fractions were completely
free fromnonplasma membrane ATPases. Furthermore, the poly-peptide patterns of the two fractions were close to identical,
which shows that thevesicles differedinsidednessonly.Thus, preparations of both inside-out and right-side-outplasma
mem-branevesiclesare nowavailable.Thispermits studieson
trans-port, signal transduction mechanisms, enzyme topology, etc., usingplasma membrane vesicles of either orientation.
Atypical featureofbiologicalmembranesistheasymmetric
arrangement of constituents across the lipid bilayer. This
asymmetry isabsolute for
proteins:
integral proteins, whichspanthemembrane, expose differentregionsoneither side of This paper is dedicated to Professor Per-Ake Albertsson, the
pioneerofaqueouspolymertwo-phasepartitioning,ontheoccasion of his 60thbirthday.
Supported by grantsfrom the Swedish Natural Science Research Council(C. L.,M.S.,S.W.),theDanishAgriculturalandVeterinary
ResearchCouncil (M. G. P.), the Danish Natural Science Research Council(M.G. P.), theDanish Research Academy (M.G. P.),and
theCarlTesdorpfFoundation(C.L.).
the membrane, whereas peripheral proteins are bound to either surface of the membrane. Forlipids,theasymmetry is rather relative than absolute, such that each lipid species usually only shows some enrichment to either half of the bilayer. This asymmetric, transverse organization of mem-braneconstituents formsthebasis for all thevectorial
activi-tiesexertedbybiological membranes,andiscreatedthrough
theasymmetricassemblyof membranes(review,22). The mostuseful approach for characterizing the
asymmet-ricproperties ofamembrane,including its vectorialactivities,
is to prepare sealed membranevesicles of either orientation.
With such preparations, each membrane surface can be probed selectively using impermeable agents, and transport in eitherdirection can be measured as uptakeinto vesicles.
The formation andsubsequent separation of vesicles of
op-posite orientation was first achieved with the erythrocyte
membrane through thepioneeringworkofSteck et al.(review, 30), and later with the mitochondrial inner membrane (re-view,8) and thechloroplastthylakoidmembrane(review, 2). The access to sealed membranevesicles ofeitherorientation
madeextensive studies on the topologyofthese membranes
possible. Withboth the erythrocyteand thethylakoid
mem-brane, aqueoustwo-phase partitioning wasused to separate the oppositely oriented vesicles (2, 30). Indeed, two-phase partitioningshould bea verysuitable method in such cases,
since itseparatesparticles accordingtotheirsurfaceproperties
(1) andvesicles ofopposite orientation areexpected todiffer
inthisrespect but not in sizeordensity.
Forthe plant plasmamembrane,theseparation of inside-out(cytoplasmic sideout)andright-side-out (apoplastic side out) vesicleswasonly recently achievedusing either free-flow electrophoresis (7) or two-phase partitioning (21). We now report a number of essential improvements on the phase partition procedure,aswell as athoroughcharacterization of themembranefractions obtained.
MATERIALSANDMETHODS Plant Material
Four-week-old sugar beet plants (Beta vulgaris L.) were kindly suppliedbyHilleshogAB, Sweden. Plants were main-tained in soil in agreenhouse with supplementary light (23
W m-2, 350-800 nm; Philips G/86/2 HPLR 400 W, The Netherlands). Leavesof 6-to8-week-old plantswereused. Preparation of Plasma Membranes
Plasma membranes(predominantly right-side-out vesicles) were purified froma microsomal fraction (10,000-50,000 g pellet) of sugar beet leaves by partitioning in an aqueous polymertwo-phase system asdescribed earlier (reviews, 18, 20)with minormodifications.Thehomogenization medium was essentiallyasin Palmgren andSommarin (25) and con-tained 330 mmsucrose, 50 mM Mops-BTP (pH 7.5), 5 mM EDTA, 5 mm DTT,0.5mmPMSF, 0.2% (w/v) BSA (Sigma; proteasefree), 0.2% (w/v)casein (boiled enzymatic hydroly-sate, SigmatypeI), 0.6% (w/v) insoluble PVP. Lotsof125g ofleaves were homogenized in 275 mL, and the resulting
microsomalfraction (about 100mgofprotein)wassuspended in330 mmsucrose, 5 mmK-phosphate (pH 7.8), 5 mM KCI, 1 mm DTT, 0.1 mM EDTA. This microsomal fraction was addedto aphase systemwith a final weightof 36.0g anda finalcompositionof 6.5% (w/w) Dextran T500, 6.5% (w/w) polyethyleneglycol3350, 330mmsucrose,5mmK-phosphate (pH 7.8), 5 mm KCI, 1 mm DTT,0.1 mMEDTA (4°C). We
routinelystartwitheither250or500 gofleaves andprocess two tofour36gphasesystemsinparallelusingthethree-step batchprocedure describedpreviously (18, 20).Thefinalupper phases containing the plasma membranes were diluted sev-eral-fold with330mmsucrose, 5 mmK-phosphate(pH 7.8), 50 mM KCI, 1 mm DTT, 0.1 mM EDTA, and the plasma
membranes were pelleted and resuspended to 15 to 20 mg
mL-'
protein in the same medium. Theyield was 15 to 18 mgof proteinper 125gofleaves, andthepreparationswere free ofChl and also otherwise of high purity asdeterminedby standard markerassays(cf 11). The membranes (usually
>90% right-side-out vesicles) were stored in liquid N2 until furtheruse.
Formation ofInside-Out Vesicles from Right-Side-Out Vesicles
Thehighly purified right-side-out plasmamembrane vesi-cleswerefrozenand thawedtoproducea mixture of inside-outandright-side-out vesicles.Typically, portions of0.8 to 1 mLwere frozenin liquidN2 and thawed inwater at20°C a totalof fourtimes.
Separationby Counter-Current Distribution
Thefreeze/thawed plasma membranes (nowbeinga mix-tureofoppositelyoriented vesicles)weresubfractionated by phasepartition usingcounter-currentdistribution(1)to pro-duce one fraction enriched in inside-out vesicles, another fraction enrichedin right-side-out vesicles,andtwo interme-diate fractionsasdescribedearlier(21)(Fig. 1). Freeze/thawed
plasmamembranes(0.8 mL)from 125gof leaveswereadded toa7.2 gphasemixturetogivean8.0gphasesystemwitha final compositionof 6.2%(w/w)DextranT500, 6.2% (w/w)
polyethylene glycol 3350, 330mmsucrose, 5 mMKCI, 1 mM DTT, 0.1 mM EDTA,5 mMK-phosphate (pH7.8;4°C).The phase system was shaken and spun for 5 min at 1500g
microsomal 2washes of upper phase fraction with fresh lower phase
freeze/thaw
two-phase
system
freshupperphases (moving phase) counte. (3tral o Right-side-outpm vesicles Inside-out 0 pm vesicles 0 Intracellular membrane vesicles pureright-side-outpm vesicles
mixtureofinside-outand right-side-out pm vesicles
r-currentdistribution
nsfers ofupperphase)
<1)
.. ....o.
..
i~::X
LtJ
.. Y.. 2 3 4 right-side-out pm vesiclesFigure1. Flow scheme forpreparationofinside-outand
right-side-outplasmamembranevesicles: Plasmamembranesarepurifiedfrom
a microsomal fraction using the batch procedure described earlier
(18, 20). Thesehighly purified, right-side-out plasma membrane
ves-iclesarefrozenandthawed fourtimes toproduceamixtureof
right-side-outand inside-outvesicles.Subsequently,theseoppositely ori-ented vesiclesareseparated byfurtherphasepartitionstepsusinga counter-current-distribution procedure. Inside-out vesicles are
en-riched in fraction1, andright-side-out vesiclesareenrichedinfraction
4(seetext fordetails).
(swinging bucket centrifuge) to facilitate phase separation. About 90% of the upperphasewasremovedwithout disturb-ing the interface, andwasaddedtoa second tube containing fresh lowerphase,aswellasupperphase correspondingtothe 10% upper phase not removed from tube 1. Fresh upper phase was added to tube 1, and mixing and centrifugation was repeated. Then, 90% of the upperphase in tube 2 was movedtoathirdtubecontaining fresh lowerphase, theupper phase in tube 1 wasmovedtotube2, and fresh upperphase wasaddedtotube 1. Theprocedurewasrepeatedonce more
toproducefourtubescontainingcomplete phasesystems and
plasma membrane vesicles. Thecontents ofeach tube (frac-tions 1-4 in Fig. 1) wasdiluted about 10-fold with 330 mM
sucrose, 10 mM Mops-BTP (pH 7.5), 5 mM EDTA, 2 mM DTT, 0.5 mmPMSF, and the plasma membranes were pel-letedat 100,OOOgfor 1 h. Thepelletsweregently resuspended
Fraction: 1 inside-out pm vesicles
INSIDE-OUT AND RIGHT-SIDE-OUT PLASMA MEMBRANE VESICLES
in the same medium (minus DTT for the NADH-oxidore-ductase assays) andimmediatelyusedinthedifferentassays. The procedure is scaled up byeitherprocessing two or more 8gphasesystemsin parallelorbyusing larger phasesystems. Inside-outvesicles wereenriched in fraction 1, and right-side-out vesicles were enriched in fraction 4 (Fig. 1). Thus, the inside-out and right-side-out vesicles were separated by essentially repeating the phase partition step originally used toisolatetheplasmamembranesasright-side-out vesicles(see above). Almostidentical phasecompositions wereused, and as in thepreviousstep the right-side-out vesiclespartitioned totheupperphase; bycontrast,inside-outplasma membrane vesiclesbehaved asintracellular membranes and were there-fore recovered in the lower phase + interface, which made the separation possible. Note, that the main difference be-tween thetwo procedures is that in the latter procedure the wash phases are saved to produce theinside-out fractionand the two intermediate fractions, whereas the corresponding
fractionsare simply discarded in the formerprocedure. The counter-current distribution procedure is recommended in initial work, since all material is saved and accounted for, which makes it simplertooptimize separation.
H+Pumping
H+ uptake intothe vesicles was monitored asthe absorb-ancedecreaseat 495 nm ofthe
l\pH
probe acridine orange (32). The assay medium was essentially asdescribed earlier(25) andconsistedof20ztM acridineorange,2mMATP-BTP, 4 mM MgCl2, 10 mM Mops-BTP (pH 7.0), 140 mM KCI, 1 mM EDTA, 1 mm DTT, 1 mg mL-' BSA (Sigma; A 0281,
essentially fatty acid free), 2.5,ugmL-' valinomycin, and 50 to 100
jig mL-'
membraneproteininatotalvolumeof1 mL. After 5 min preincubationat20°C, the reactionwasinitiatedby addition of MgCl2. The rate of H+-accumulation was estimated from the initial slope of absorbance quenching
(A495)
of acridine orange.ATPaseAssay(PiRelease)
ATPaseactivity determined fromthe releasedPiwas meas-uredaccordingtoBaginskietal.(4)asmodifiedbyBrotherus etal.(6).Twodifferentassaymediawereused:(a)Amedium essentially asin Gallagherand Leonard (10)
containing
330 mMsucrose, 50mMMes-Tris(pH6.0),25 mM K2SO4, 3 mM MgSO4, 3 mM ATP, 0.1 mM EDTA, 1 mm azide, 0.1 mM molybdate, and 10 ,ug protein in a final volume of 120 ,uL. The assay was run for 30 min at 25°C, and ±0.02% (w/v) Triton X-100(Sigma;T6878). (b)The samemediumasused toassay H+pumping (above),exceptthatEDTAwasomitted. This mediumwasusedtomeasurevanadateinhibition ofH+ pumpingand ATPase activity in thesame sample. Aliquotsof 200 ,uL were withdrawn after 20, 140, and 260 s for P,-determination,andH+pumpingwasrecordedsimultaneously
at495 nm. The ATPase assay based onthe release ofADP (below) could not be used for this purpose, since vanadate interfered with thatassay.
H+-ATPaseAssay (ADP Release and H+ Pumping) H+ pumping and ATPase activity were monitored simul-taneously in the same cuvette using the H+-ATPase assay described earlier (25). The concentration of ATP was kept constant by including0.25 mMNADH, 1 mM phosphoenol-pyruvate, 15 Mig mL-' lactate dehydrogenase (Boehringer; 003565, solutioninglycerol)and30 ,ugmL-' pyruvate kinase (Boehringer; 005541, solution in glycerol) in the assay me-dium used for H+ pumping (above). Thus, ATP hydrolysis was coupled to oxidation ofNADH (24), and H+ pumping and ATPase activityweremonitoredsimultaneously by
plot-tingtheabsorbanceat495nmand 340nm,respectively. 1,3-f,-GlucanSynthase
1,3-fl-Glucan synthase activity wasmeasured as incorpo-ration of UDP-[3H]glucose into polyglucan according to Kauss et al. (14) with minor modifications (9). The assay medium contained 330 mm sucrose, 2 mm DTT, 2 mM spermine, 20mmcellobiose, 0.2mmCaCl2, 2mmUDP-[3H] glucose (20 GBqmol-'), 50mmHepes-KOH (pH 7.25), and 1 Mg ofprotein inatotalvolume of 100 ,uL. Theactivitywas assayed ± 0.006% (w/v) of the detergent digitonin (Serva; 19550). After30 minat25°C thereactionwasterminatedby
immersion ofthe test tubes in boiling water, and 95 ML aliquotswerewithdrawn andaddedtocellulose filters (What-man3 MM). Filtersweredried and washedasdescribed(14). Standardscontaining50% of theradioactivity addedtoeach sample werecounted onwettedglass fiber filters (Whatman GF/F), since theradioactivity of the low-molwtsubstratewas quenched onthe cellulose filters(9).Radioactivitywas deter-mined byliquid scintillation countingusingReadySafe
scin-tillation cocktail (Beckman).
NADH-AcceptorOxidoreductase
NADH-acceptoroxidoreductasewitheitherferricyanideor Cytc(Sigma;C 7752) asacceptor wasassayedessentiallyas described earlier(3).
NADH-ferricyanide reductase activity was measured as A(A42o-A5oo) usingan AmincoDW 2 spectrophotometer op-erated in the dual beam mode. Theassaywas runat25°Cin 1 mLof 330mmsucrose, 25mM Hepes-KOH(pH7.3),0.25 mM NADH, 0.2 mm
K3[Fe(CN)6j,
40 Mg protein, and ± 0.025% (w/v) Triton X-100. The reaction was initiated bythe addition ofNADH. Correction was made for nonenzy-maticreduction offerricyanide.
NADH-Cyt c reductase activity was measured similarly using40 Mm Cytc asacceptorinstead of 0.2mM K3[Fe(CN)6],
and with 0.4
gM
antimycin A (Sigma; A 2006) and 1 mM KCNpresentin the assaymedium. Theactivitywasrecorded asA(A550-A600), and wasdetermined ±0.015% (w/v) Triton X-100.Theextinction coefficients usedwere I and 19 mm-' cm-'
forferricyanideandCytc,respectively. SDS-PAGE
SDS-PAGE was run on gradient gels (concentration of
monomers,
10-22%;
crosslinking, 2.7%;
5% stackinggel;gel 873dimensions-175*160. 1.5mm)inthe buffersystemof Laem-mli (17). The samples weresolubilized at 80TCfor IO min,
and gels wererunfor 15 hat 12TCand 15 mApergel. Silver
stainingwasessentiallyasdescribedby Guevaraetal. (13). Protein
Protein wasmeasured essentially asdescribed by Bearden (5),with BSAas astandard.
RESULTS ANDDISCUSSION FormationofInside-Out Vesicles
Plasma membranepreparations ofhigh purity(about95%) areeasily obtained bypartitioninginaqueouspolymer two-phasesystems(reviews, 18, 20).Thesepreparations,however, mainly contain sealed, right-side-out (apoplastic side out) vesicles(19). Thiswasdemonstratedbyassaying theplasma membraneATPasein the absence andpresenceofthe deter-gent Triton X-100. Since the active site ofthe ATPase is locatedontheinner, cytoplasmic surface ofthe plasma mem-brane,theactivity associated with sealed, right-side-out
vesi-clesis only measured in thepresenceof detergent. Thus,the
proportion ofright-side-out vesiclesmaybe calculated from
theratio oflatentactivity (difference inactivity measured+
detergent) to total activity (activity measured + detergent).
Using this method to determine vesicle orientation, plasma membrane preparations from most materials, including the sugarbeet leaves used inthisstudy, containabout90% right-side-out vesicles. Thus, for these to be a suitable starting material for the separation ofinside-out and right-side-out
vesicles,partoftheright-side-outvesicles in the plasma mem-branepreparationneedtobe turnedinside-out.Wetherefore
looked at a number of treatments (vigorous pottering in
different
media,
hypoosmotic shock,sonication,freeze/thaw-ing) that could be expected to cause vesicle breakage and
revesiculation. Asmarkersfor inside-outvesiclesweused: (a) ATP-dependent
He-pumping,
(b) nonlatent activity ofthe ATPase, and (c) nonlatent activity of NADH-ferricyanidereductase, an activity also associated with the cytoplasmic surface ofthe plasma membrane (3, see below). The first
markeriscrucial, sincethe nonlatentactivitiescouldatleast
theoreticallybe dueto leaky
right-side-out
vesiclesor mem-brane sheets, whereas only sealed, inside-out vesicles may supportATP-dependentH+pumping.
Twoofthetreatments investigated, sonication andfreeze/thawing,
significantly in-creased theproportion ofinside-out vesicles, as determined by theirH+-pumping activityandbytheir nonlatentNADH-ferricyanide reductaseactivity (data not shown). Sonication
simultaneouslycausedadecrease in total
NADH-ferricyanide
reductaseactivityindicatingsomedamage tothe membrane.
Freezing in liquid N2 and thawing in a waterbath at 20°C (which should minimize theformationof concentration gra-dientsduringtheprocess)didnotinhibit
NADH-ferricyanide
reductase activity and we therefore chose to opitimize this procedure.
Thecomposition ofthefreeze/thawmediumwasimportant (Fig. 2). Addition of 50 mM
KCl
to the basic medium (330 mM sucrose, 5 mm K-phosphate [pH 7.8], 1 mm DTT, 0.1Figure 2. Effect of the composition of the freeze/thaw medium on
theformation of inside-outplasma membrane vesicles, assayed as
increasedH+pumping. The basic medium contained 330 mm sucrose, 5mmK-phosphate (pH 7.8), 1 mm DTT, 0.1 mm EDTA. Modifications:
aandc, none; b andi, plus50mmKCI;d, minus sucrose; e andf,
plus 6 and 12 mm MgCI2, respectively; g, h, i, and j, plus 12.5, 25, 50, and 100mmKCI, respectively; k, plus 50 mm NaCI, Fresh =not
frozen.Freeze/thawed =onefreeze/thawcycle.
1-10 C -Cu 0.5 0.4 0.3 0.2 0.1 0.0 0 2 4 Freeze/thaw cycles 1 0.8
'
0.6"
tn
0.4 w 0.2 u ._ --0.0 a 6+L
Figure 3. Effect of the number of freeze/thaw cycles on the formation of inside-out plasma membrane vesicles. Both H+ pumping (A) and nonlatent ATPase activity (0) were used as markers for inside-out
vesicles.
mM EDTA) gave an approximate twofold increase in H+
pumpingafteronefreeze/thawcycle comparedto noaddition. NaCl at the same concentration caused a similar increase,
whereasMgC12 (6 and 12 mM)waslessefficient. Omissionof sucrose(hypotonicmedium) causedadecrease in H+-pump-ing. The effect ofKCl was concentration-dependent (Fig. 2, g-j), and 50mm waschosenfor further experiments.
Repeating the freeze/thaw cycle increased both the H+
pumping capacityandthe non-latent ATPaseactivity (Fig. 3) measured simultaneously in the same sampleusing an H+-ATPaseassay(25).Thecumulativeeffect ofrepeatedfreeze/
thawingsuggests thatonlyaminorproportionof the vesicles werebrokenand resealed in eachcycle. Little additional effect wasfound afterthreefreeze/thawcycles,andfourcycleswere used for furtherexperiments.
INSIDE-OUT AND RIGHT-SIDE-OUT PLASMA MEMBRANE VESICLES
Freezing and thawing of animal plasma membranes has been reported to causeincreasedATPaseactivity by 'unmask-ing' of latent ATPbinding sites (23). The fact that H+ pump-ing and ATPase activity increased in parallel in the sugar beet plasmamembranes (Fig. 3) indicates that the effect of freezing and thawing was not only 'exposure' ofnew ATP binding
sitesbut was coupled to the formation ofsealed, inside-out
vesicles from right-side-out vesicles.
Onhomogenization ofplanttissue, aright-side-out orien-tation ofthe resulting plasma membrane vesicles is usually stronglyfavored. This isshownbyahigh recovery (70-80%)
of plasma membrane markersinpreparations which contain predominantly right-side-out vesicles (11, 15, 18, 20). How-ever, theproportion of inside-out vesicles obtainedseems to be dependent on the method used for homogenization (7),
and thecomposition ofthehomogenization medium isalso likely toaffect vesicle orientation.For the erythrocyte mem-brane aright-side-out orientation seems to befavored by: (a) a higher negative net charge density of the outer surface
compared to theinnerone, which affects membrane curva-ture, and(b) remaining cytoskeleton anchoredto the inner
surface alsoaffecting curvature (30, andreferences therein).
The samefactorsmight favortheformation of right-side-out
plasma membrane vesicles with plant material. Thus, free-flowelectrophoresis of plasmamembranevesicles (7)suggests that the cytoplasmic surfacehas a lower net charge density
than the apoplasticone atneutral pH, and fibrous material
which might be remnants of the cytoskeleton has been ob-servedinright-side-out vesicles (SWidell,C.Larsson, unpub-lishedresults). Theincreased formation of inside-out vesicles on freeze/thawingathigher KCI concentration(Fig. 2)may be due to
screening
ofcharges on the membrane surfaces,thus reducing the charge difference between the inner and outersurface.The resultsobtained with MgCl2(Fig. 2)donot support thisconclusion, however,but
Mg2+
may haveaddi-tionaleffectsonthemembranewhichcounteractitsscreening effects. Nevertheless,thecompositionof thefreeze/thaw me-dium islikelytoaffectboth theprobability of vesicle breakage
uponfreeze/thawing,and theprobability for resealingwitha certain orientationafterbreakage.
Separationof Vesicles ofOppositeSidedness
Theoptimalcompositionofthetwo-phase systemfor
sep-aration of inside-out and
right-side-out
vesicles was deter-minedbypartitioningthefreeze/thawed plasmamembranes inaseriesofphase systemswithincreasingpolymer concen-tration(Fig. 4). At6.2% (w/w) ofbothpolymers onlyabout 20% ofboth the H+ pumping and the non-latentNADH-ferricyanide reductaseactivity werepartitioned to the upper phase comparedto about70% ofthelatent
NADH-ferricya-nide reductaseactivity, indicating a good separation of inside-outand right-side-out vesicles. When purifiedplasma mem-braneswere subjectedto counter-current distribution atthis polymer concentration twopeaks of material wereobserved
providedthevesicles had beenfreeze/thawed (Fig.5).
Increas-ing the numberoffreeze/thaw cycles increased the amount of material recovered in fraction 1 withaparalleldecreasein fraction 4. Material partitioning mainly to the interface + lowerphase (=stationary phase) would be recovered in
frac-1001 I-Cu co s &L I_ 80 60 40 20 V 5.5 6.0 6.5
Dextran/PEG,
%(w/w)
Figure 4. Effect of polymer concentration on the partitioning of inside-out and right-side-out plasma membrane vesicles in an aqueouspolymer two-phase system. The phase system contained 330 mm sucrose, 5mm K-phosphate (pH 7.8), 0.1 mm EDTA, and equal concentrations of Dextran T 500 and polyethylene glycol 3350
asindicated. Markers for inside-out vesicleswere H+ pumping (A) and nonlatent NADH-ferricyanide reductase (0), andfor
right-side-outvesicles latent NADH-ferricyanide reductase (0). Assays were
performed on aliquots withdrawn from the phases after separation and dilution. 0 60 1 3 o$ number of 4 v 40 freeze/thaw
cycles
-.4 o 20 3 0 1 2 3 4 FractionFigure 5. Effect of the number offreeze/thaw cycles on protein
distribution after counter-current distribution of plasma membrane vesicles. Plasma membraneswereloadedintube1 and three
trans-fers of the upperphase weremadekeeping the interface+ lower phase stationary (see Fig. 1). The plasma membraneswereeither usedfresh (0), orsubjectedto 1 (0), 3 (U), or 4 (A) freeze/thaw
cycles beforeloading.
tions 1 and 2, whereas material
partitioning
mainly to the upperphase (=movingphase)would be recovered infractions 3 and 4. The observed shift of material from fraction 4to fraction 1 is thus consistent with the formation of inside-outvesicles fromright-side-out ones uponfreeze/thawing(Figs. 2 and3), aswell aswith thepartitioning ofinside-out vesicles totheinterface+lowerphase(Fig.4). Theyield of inside-out vesicles (fraction 1) after four freeze/thaw cycles was 15 to 25% of total plasma membrane protein, correspondingto5 to 10mgofproteinfrom 500 g of leaves.
We haveearlier (34) suggested thatinside-out plasma mem-brane vesiclespartition totheinterface + lowerphase based on the dual distribution ofplasma membrane markers, and in analogywith inside-outerythrocyte membranes. This has now beenconfirmed by theseparation of inside-out and
right-side-out vesicles by phase partitioning (21) and the present work. Thus, inaddition to thefact that the plasma membrane vesicles formed on homogenization of the plant material are usually mainly right-side-out, this is yet another reason that plasma membrane preparations obtained by two-phase
par-titioning contain mainly right-side-out vesicles;theinside-out vesicles are simply lost during the purification procedure. However, there seem to be some exceptions. For example,
freshlyprepared plasmamembranes from oat roots show only about70% latencyofthe ATPaseactivity (19), and sometimes evenless, andtheyalso supporthighratesofH+pumping (M
Palmgren, unpublished results). Thus, there may be cases where the phasecomposition usedfortheoriginalpurification
doesnot resolve inside-out and right-side-out vesicles. With theplasma membranes from sugar beet leaves a lower
poly-merconcentration can be usedfor theseparation of
inside-outandright-side-out vesiclesthanis needed to separate right-side-out plasmamembranevesicles from intracellular mem-branes(6.2 and6.5% [w/w], respectively, in otherwise iden-ticalphasesystems),andwith many materials identicalphase
systems should do. However, with some materials, such as oatroot, a much higherpolymerconcentration (orCl- con-centration [18, 20]) maybe needed,and optimal conditions
should be determined for each material as demonstrated in Figure 4.
Determination ofSidedness
Todetermine theproportions of inside-outand right-side-outvesiclesin fractions 1 to 4(seeFigs. 1 and5),theactivity
of markers for thecytoplasmic surface (ATPase,
1,3-f-glucan
synthase, NADH-ferricyanide reductase and NADH-Cyt c reductase; see 33 for a detailed discussion on assays for sidedness)were assayed ± detergent (Fig. 6). Nonlatent ac-tivities (markers for inside-out vesicles) were enriched in fraction 1, whereas the activities in fractions 3 and 4 were highly latent, indicatinga high proportion of right-side-out
vesicles. Intermediate values were found for fraction 2. H+
pumping,a moredefinite markerfor sealed, inside-out vesi-cles,correlated wellwiththe nonlatent ATPaseactivity (Fig. 6,top, left). However, latentactivitieswerealso observed in
fraction 1, andforthe ATPasethis indicatedacontamination byabout40% right-side-out vesicles.
The use of enzyme latency to assess vesicle orientation assumesthat thedetergents used donothaveany othereffect than topermeabilizethevesicles. That thisassumptionisnot entirely validis illustratedbyFigure 7(top). Thus,when the ATPase activity was assayed in another medium than that usedinFigure6(seelegends)nolatentactivitywasobserved infraction 1.Rather, the TritonX-100 concentrationoptimal toreveal the latentactivityin fractions 3 and4(0.025%
[w/
v]) inhibited the activity in fraction 1 with about 30%. In addition, thelatencies obtained with fractions 3 and4were
lowerindicatingthat alsothe latentactivitieswereinhibited
(ef
Figs. 6 and 7). That the vesicles were permeabilized by0.4
A
0.3 to p;g 0.2 E. 0.1 | 2.0 bo ._ .w 1.0 l-2 co DO%nnATPase(0,0); H' pumping(A)
+Triton -Triton
I~~~
NADH-ferricyanide reductase +o +Triton 1 2 3 4 Fraction I I *1,3-8-Glucan synthase0/
* *-a +digitonin \ digitonin NIDH-Cytc reductase NADH-Cytc reductase 1 2 3 4 2.0 1.5 Xu E wq 1.0 ° 0.5 o 2 0.0 ._I
600 -t0 400 *a 0 200 2 j. 0 FractionFigure6. Specific activity of markers for the cytoplasmic surface of the plasma membrane in fractions 1 to 4 obtained after
counter-currentdistribution offreeze/thawedplasma membranes (see Figs.
1 and 5). The activities were assayed both in the absence(0) and presence(0)of detergent, except for H+ pumping (A, relative units) which was assayed minus detergent only.ATPase assay medium (a) (see"Materials and Methods")wasused(cf. Fig. 7). Data on the y-axisarefor theunfractionated material.
thedetergentisshown by thecollapse ofthe H+ pump(Fig.
7, bottom), which was complete already at 0.015% (w/v)
Triton X-100. Thus, thevesicleswere permeable to H+ at a
slightly lower detergent concentration than that needed to
permeabilize them to MgATP. This is consistent with the
differencesinsize and chargeofthe twospecies. The latencies
obtained with fractions 1 and4aresummarized in Table I. The disagreements are obvious, particularly for fraction 1, and suggests that the detergents used (Triton X-100 and
digitonin) may be either slightly stimulatory or inhibitory depending on the activity investigated and the assay condi-tions used. To find an ideal detergent for determination of
enzymelatencywehaverecentlyscreenedalargenumberof
detergentsregardingtheireffectonthe ATPaseactivityin the H+-ATPase assay(26). Fromthis investigationBrij58 seems tobeideal, sinceitneither inhibitsnorstimulatesthe ATPase
activity, and a concentration of0.01 % (w/v) can be used routinely with protein concentrations up to 50 gmL-'. Using
this detergentalatency of about 20% wasobtained withthe ATPaseinfraction 1, whereasfreshly preparedplasma mem-branesshowedalatencyof about90%(26).
Toconfirm that nonlatentactivitieswereduetoinside-out vesicles and latent activities due toright-side-out vesicleswe usedtrypsindigestion of NADH-Cytcreductaseactivity;the rationale being that the activity associated with inside-out vesicles would beabolished, whereas theactivityof right-side-out vesicles wouldbe revealed by a subsequent addition of
trypsin inhibitor and Triton X-100. The nonlatent NADH-Cytcreductase
activity
wasalmosttotallyinhibitedby
trypsin
in both fractions 1 and4, whereasmostof the latentactivity
q
INSIDE-OUT AND RIGHT-SIDE-OUT PLASMA MEMBRANE VESICLES E
0.6
fraction 4_
0.4
02 0.0 fraction 1 1.0 0.5 + 0.00 0.02 0.04 0.06 TritonX-100,
%(w/v)
Figure7. Effect of Triton X-1 00onATPaseactivityand H+pumping with fractions 1 to 4 obtained after counter-current distribution of freeze/thawed plasma membranes (see Figs. 1 and 5). ATPase activity and H+pumping wereassayedsimultaneously inthe same
cuvette(25) using the H+-ATPaseassay(cf. Fig. 6 and Table I).
TableI. Percent Latencies ofEnzyme Markers for theCytoplasmic
Surface of the Plasma Membrane
Data ± SD forn independent experiments. ATPase activity was
determined in twowidely differentassay media(see"Materials and Methods") resultingin verydifferent latencies.
Marker Fraction1 Fraction 4
%latency
ATPase (assay medium a) 41 ±4(n=3) 87±6(n=3)
ATPase(H+-ATPase medium) -45±11 (n=4) 54±4(n=2)
1,3-,3-Glucansynthase 51 ± 7(n=6) 85±2(n=4)
NADH-FeCN reductase 34±8 (n=16) 85±6(n=16)
NADH-Cytcreductase 26±6 (n=8) 76±6 (n=8)
remainedalthough the levels of remainingactivitywerewidely different (Fig. 8), confirming the assumptions above. Since
NADH-Cyt creductase is inhibited by Triton X-100above the0.015% (w/v) optimalfordetermination of latentactivity (3), whereas this concentration is probably not sufficient to
permeabilize the vesiclesto Cyt c(cf Fig. 7), the remaining latent activities would be underestimates of the proportions
ofright-side-out vesicles. Taking this into account, a cross contamination of about 20% is indicatedbythisexperiment. Another, relative measureof theproportions of inside-out vesicles in fractions 1 and 4 is given by the ratio of H+
pumpingbetween these two fractions. This ratio was 5.0 ±
100 a ;ESW c 0 0 0.0 0.2 0.4 0.6 0.8 1.0 Trypsin,
gg/mL
Figure8. Effectof trypsinonnonlatent(E)and latent(U)NADH-Cyt
creductaseactivityof inside-out(fraction 1) andright-side-out
(frac-tion4) vesicles obtained aftercounter-currentdistributionoffreeze/
thawedplasma membranes (seeFigs.1and5). Vesiclesweretreated
with trypsinfor 3min at 200C. Priortomeasurement, 160Ag mL-1
trypsin inhibitorwasadded.
0.4 (n = 5)
(cf
Figs. 6 and 7) which fitswell with a cross contamination of 20%.Characterization of the H+-ATPaseActivity
Both the H+pumpingand the ATPaseactivity ofthe inside-out plasma membrane vesicles
(fraction
1) were completely inhibited by vanadate(Ki
~ 10,uM at 100ug protein mLU';Fig. 9),aninhibitor oftheplasma membrane H+-ATPase(10, 31). Inhibition ofthe ATPase activity laggedslightly behind athighervanadateconcentrations.Thisisconsistent withthe presenceofalsoa
Ca2+-ATPase
in theplantplasma membrane (12, 28); an activitywhichis alsoinhibitedby vanadate but with a higherKi (12) and which constitutes only a minor proportion oftotal plasma membrane ATPaseactivity (12, 28, 29). Molybdate, which inhibits vanadate-sensitive acid phosphatases (10, 29), and azide, which inhibitsmitochon-drialATPase(10, 29)didnotaffectATPaseactivityinfraction 1 (datanotshown). Inthepresence ofK+, H+pumpingwas stimulated by the K+ ionophore valinomycin, and the H+ gradient was collapsed by nigericin (0.1 ,ug
mL-';
data not shown), which catalyzes the electroneutralexchange ofH+for K+(27).Thesepropertiesareconsistentwithpreviousfindings thatinside-out, plant plasmamembrane vesicles accumulate H+throughanelectrogenicH+-ATPase,which issensitiveto vanadate butnot tomolybdateorazide(reviews, 29, 31).00 54°0 606 ATPase Po 40 -H
+pumping
20 0 -0 0.1 1 10 100 1000 Vanadate, FMFigure 9. Effect of vanadate on H+ pumping (A) and nonlatent ATPase activity(0)of inside-out plasma membrane vesicles(fraction 1).The activities weremeasuredinparallel by withdrawing aliquots from the cuvette forPi-determination.
Additionof BSA (fatty acid free)tothe assaymedium often
more than doubles the rateof H+ accumulation leaving the ATPaseactivity essentiallyunaffected(26). To measure
max-imum H' pumping capacity, BSA wastherefore always in-cluded in the assay medium. The effect ofBSA is to bind
fatty acids whichmay act as uncouplers. Thefatty acidsare probably produced by endogenous phospholipase activity,
which mayexplain the often observed 'leakiness' of plasma membranes,particularly onageing (26).
Polypeptide Pattern
The polypeptide patterns of inside-out (fraction 1) and right-side-out (fraction 4) plasma membrane vesiclesareclose toidentical(Fig. 10). Thiswasexpected,since theinside-out vesicleswereformed from the right-side-outones(Figs. 2, 3, and 5) and the main difference should be their sidedness. However,someminor differencesareevident.
The inside-out vesiclesare depleted in some polypeptides (of 82, 80, 57, 34, 32, and 14.5kD), which arefound in the freeze/thawsupernatant. Thesepolypeptides probably repre-sent soluble proteins trapped within right-side-out vesicles during homogenizationof the leaves, althoughitcan notbe excludedthatsomeareperipheral plasmamembraneproteins. The 57 and 14.5 kD polypeptides (which dominate in the supernatant) are mostprobably thelarge and smallsubunit, respectively, of ribulose-1,5-bisphosphate carboxylase/oxy-genase,themostabundantsolubleproteininleaves.
Three polypeptides (of 73, 44, and 20 kD) are clearly enriched in the inside-out vesicles. This may reflect some heterogeneitywithin the isolatedplasma membrane popula-tion.Forinstance,in vivo theremaybealateralheterogeneity intheplasma membrane,whichuponhomogenizationof the tissue givesrise tovesicles ofslightly different composition. There mayalso be a difference between plasma membrane vesicles derived from differentcelltypesinthe tissue. Thus, thepolypeptidepatternsofplasmamembranesobtained from barleyleaves androotsshowsomeminor differences(16).As
soon asaheterogeneity ispresent, this could affect both the
Figure 10. Polypeptide patternsof inside-out (1 = fraction 1) and
right-side-out (4=fraction 4) plasmamembrane vesiclesasrevealed
by SDS-PAGE and silverstaining. Forcomparison,the polypeptide patterns of the freeze/thawed plasma membranes (f/t pm = the
starting material) and of a freeze/thaw supernatant (f/t sup) are
shown.(Notethatthesupernatantwasproducedfor thisexperiment only; itis not necessaryto pelletand resuspendtheplasma
mem-branes after the freeze/thaw treatment). Molecular masses to the
right indicate polypeptideswhicharedepletedinthe inside-out
vesi-cles and arefound inthefreeze/thaw supernatant. Three polypep-tides enrichedintheinside-outvesiclesareindicatedtotheleft. Each lane wasloaded with 10 jig protein, exceptfor that of the freeze/
thaw supernatant which received 2 qg protein of supernatant
con-centrated about 80-fold.
probability of vesicle breakage upon freeze/thawing and the probabilitytoformaninside-outvesicleuponrevesiculation, and thusgive risetotheobserved differences inpolypeptide pattern. However, we do not observe a difference with all materials, andthepolypeptide patternsof fractions 1 and 4
from cauliflower inflorescenceswere indistinguishable (data
notshown).
CONCLUDING REMARKS
The rationale of thepresentworkwastousethepure right-side-outplasma membranevesiclesreadilyobtainedby
two-I-.V
INSIDE-OUT AND RIGHT-SIDE-OUT PLASMA MEMBRANE VESICLES
phase partitioningasthestarting materialforinside-out
vesi-cles. Inthisway, preparationsof plasma membrane vesicles ofbothorientationsareobtained whichareessentially free of
contaminating membranes.
It should be noted, thatif only inside-out vesiclesaretobe
prepared (e.g. for studies on H+ pumping) a more direct proceduremaybe used:The lowerphase ofthephasesystem
containingfreeze/thawedplasmamembranesmaysimplybe extracted with a number of fresh upperphases leavingthe inside-out vesicles enriched in the lower phase + interface. Similarly, if onlyright-side-outvesiclesareneeded, the origi-nal plasma membranepreparation isusuallymore enriched in these vesicles than the fraction ofright-side-out vesicles obtained withthe present procedure. However, for a direct comparisonofthepropertiesofinside-out and right-side-out plasma membrane vesicles, a comparison of the fractions
obtainedafterfreeze/thawingandsubsequentphase partition-ing is probablymore valid, since these fractions have been
subjectedtoidenticaltreatments.
Toaccuratelydeterminetheproportions ofinside-out and right-side-out vesicles in different fractionshas been a
con-stantproblem.Thisseemstobesolved byusing the detergent Brij 58 for determination ofATPase latency (26). A more rapid andconvenient assayfor sidedness istodeterminethe latencyof theNADH-ferricyanidereductaseusing Triton
X-100, which, however, gives an underestimation of the
per-centageofinside-outvesicles.
Duringthedevelopmentofthepresentprocedure it turned
outto be very importanttoinclude a number of protective
agents(DTT,EDTA, BSA, casein, PMSF, insoluble PVP)in thedifferent steps to retain highactivities (data not shown). In particular, the inside-out vesicles easily lost activity. It
seemsthattheexposuretothe medium of activesites located
onthecytoplasmicsurface makes these sitesvery susceptible
toinactivationcomparedtothesiteshidden inside
right-side-outvesicles.Thiswasparticularly pronounced fortheATPase, for H+ pumping and for the 1,3-,B-glucan synthase. For H+ pumping,lost activity could be restored by addition offatty acidfree BSAwhich binds released fatty acids (26).
Taken together, our results indicate that preparations of
about 80% inside-out and 80% right-side-out plasma mem-brane vesicles, respectively, are obtained with the present
procedure. Thus, preparations ideal for studies on plasma membrane transport, signal transduction mechanisms,
en-zymetopology,etc.,arenowavailable. Using these fractions, thedonorandacceptorsites of the plasma membrane-bound NADH-acceptoroxidoreductase have recently been localized
tothecytoplasmic surface (3), andsohave the activator sites for Ca24,spermine and cellobiose of the 1,3-,3-giucansynthase
(9).
ACKNOWLEDGMENTS
Wewish to thank Mrs. Ann-Christine Holmstrom, Mrs. Adine Karisson, and Mrs.IngerRohdin for their skillful technical assistance,
andfor theirnevergivingup.
LITERATURE CITED
1. AlbertssonP-A (1986) Partition of Cell Particles and Macromol-ecules, Ed3. John Wiley & Sons, New York
2. Andersson B,SundbyC,AkerlundH-E, AlbertssonP-A(1985)
Inside-outthylakoidvesicles.An importanttoolforthe char-acterization of the photosynthetic membrane. Physiol Plant 65: 322-330
3. Askerlund P,Larsson C, Widell S (1988) Localizationofdonor and acceptor sites ofNADH dehydrogenase activities using inside-outandright-side-outplasma membranevesiclesfrom
plants.FEBS Lett239: 23-28
4. BaginskiES, Foa PP, Zak B(1967)Determination ofphosphate: studyof labileorganic phosphate interference. Clin ChimActa 15:155-158
5. BeardenJC Jr (1978)Quantitationofsubmicrogram quantities of protein byanimprovedprotein-dyebindingassay.Biochim
Biophys Acta 533:525-529
6. Brotherus JR,Jacobsen L, J0rgensen PL (1983) Soluble and
enzymatically stable (Na+ + K+)-ATPase from mammalian kidney consisting predominantly of protomer subunits.
BiochimBiophys Acta 731: 290-303
7. Canut H,BrightmanA, BoudetAM, Morre DJ(1988)Plasma membranevesicles ofopposite sidedness from soybean
hypo-cotylsby preparative free-flowelectrophoresis. Plant Physiol 86: 631-637
8. DePierre JW, Ernster L(1977) Enzymetopology of intracellular
membranes.AnnuRevBiochem46: 201-262
9. Fredrikson K, Larsson C(1989) Activationof1,3-B-glucan syn-thase byCa2", spermine andcellobiose. Localization of
acti-vatorsites usinginside-out plasma membranevesicles.Physiol
Plant 77: 196-201
10. GallagherSR,Leonard RT(1982) Effect of vanadate,molybdate
andazideonmembrane-associatedATPase and soluble phos-phataseactivitiesof cornroots.PlantPhysiol70: 1335-1340 11. Gallet0,Lemoine R, Larsson C,Delrot S(1989)The sucrose
carrier of theplant plasmamembrane. I. Differential affinity labeling. BiochimBiophysActa978: 56-64
12. Griif P, Weiler EW(1989) ATP-drivenCa2"-transportin sealed plasma membrane vesicles prepared by aqueous two-phase partitioning from leavesofCommelina communis L.Physiol Plant 75:469-478
13. Guevara JJr,JohnstonDA,RamagaliLS,MartinBA,Capetillo S,Rodriguez LV(1982)Quantitativeaspectsof silver deposi-tioninproteinsresolvedincomplexpolyacrylamidegels.
Elec-trophoresis3: 197-205
14. KaussH,KohleH, JeblickW(1983)Proteolyticactivation and
stimulation by Ca2" ofglucan synthase from soybean cells. FEBS Lett 158:84-88
15. KjellbomP,LarssonC(1984)Preparationandpolypeptide
com-position ofchlorophyll-freeplasmamembranesfrom leavesof light-grownspinachandbarley. PhysiolPlant 62:501-509 16. Korner LE,Kjellbom P, LarssonC, M0ller IM (1985) Surface
properties ofright-side-out plasmamembranevesicles isolated frombarleyrootsand leaves. PlantPhysiol79: 72-79 17. Laemmli UK(1970) Cleavage ofstructural proteinsduringthe
assembly of the head ofbacteriophage T4. Nature 227: 680-685
18. Larsson C(1985) Plasma membranes. In HFLinskens,JF Jack-son, eds,Cell Components, Methods ofPlant Analysis, New
Series,Vol 1.Springer-Verlag,Berlin,pp85-104
19. LarssonC,Kjellbom P,WidellS,LundborgT(1984)Sidedness of plant plasmamembrane vesiclespurified bypartitioningin aqueoustwo-phasesystems.FEBS Lett 171:271-276 20. Larsson C, Widell S, Kjellbom P (1987) Preparation of
high-purity plasmamembranes.MethodsEnzymol148: 558-568 21. Larsson C, Widell S, Sommarin M (1988) Inside-out plant
plasma membranevesicles ofhighpurityobtainedbyaqueous two-phasepartitioning. FEBS Lett 229: 289-292
22. LodishHF, Rothman JE(1979)Theassemblyof cell membranes. SciAm240: 38-53
23. M0llerOJ(1971)Activation byfreezingof(Na++ K+)ATPase
inamicrosomal fraction fromoxkidneycortex.ExpCell Res 68: 347-355
24. N0rby JG (1988) Coupled assay of Na+,K+-ATPase activity. MethodsEnzymol 156: 116-119
25. Palmgren MG, Sommarin M (1989) Lysophosphatidylcholine stimulates ATP dependent proton accumulation in isolated
oat rootplasma membrane vesicles. PlantPhysiol 90: 1009-1014
26. Palmgren MG, Sommarin M, Ulvskov P, LarssonC(1990) Effect ofdetergents on the H+-ATPase activity of inside-out and right-side-out plasma membrane vesicles. Biochim Biophys Acta1021: 133-140
27. Pressman BC (1976)Biologicalapplication of ionophores. Annu RevBiochem 45: 501-530
28. RobinsonC,LarssonC,Buckhout TJ (1988)Identificationofa
calmodulin-stimulated (Ca2+ + Mg2+)-ATPase in a plasma membrane fraction isolated from maize (Zea mays) leaves. Physiol Plant 72: 177-184
29. Serrano R (1990) Plasma membrane ATPase. In C Larsson, IM M0ller, eds, The Plant Plasma Membrane.Structure,Function andMolecularBiology,Springer-Verlag, Berlin,pp 126-153
30. Steck TL (1974) Preparation of impermeable inside-out and right-side-out vesicles from erythrocyte membranes. Meth Memb Biol 2: 245-281
31. Sze H (1985) H+-translocating ATPases: advances using mem-brane vesicles. Annu Rev Plant Physiol 36: 175-208
32. Vianello A, Dell'Antone P, Macri F (1982)ATP-dependent and ionophore-induced protontranslocation in pea stem microso-malvesicles. Biochim Biophys Acta 689: 89-96
33. Widell S, Larsson C (1990)Acriticalevaluation of markers used in plasma membranepurification. In C Larsson, IM M0ller,
eds, The Plant Plasma Membrane, Structure, Function and Molecular Biology,Springer-Verlag, Berlin, pp 16-43 34. Widell S, Lundborg T, Larsson C (1982) Plasma membranes
from oats prepared by partition in an aqueous polymer two-phasesystem. On the use oflight-induced cytochrome b
re-ductionas amarker for theplasma membrane. Plant Physiol 70: 1429-1435
