Proc. NatI. Acad. Sci. USAVol. 91, pp. 9272-9276, September 1994Cell Biology

Mechanism of high-affinity potassium uptake in roots ofArabidopsis thaliana

(energized K+ transport/K+-H+ cotaspWrt/current/voltage analyis)

FRANS J. M. MAATHUIS AND DALE SANDERSDepartment of Biology, University of York, York YO1 5DD, United Kingdom

Communicated by Emanuel Epstein, June 10, 1994 (received for review April 4, 1994)

ABSTRACT Potassium is a major nutrient in higherplants, where it plays a role in turgor regulation, chargebalance, leaf movement, and protein synthesis. Terrestrialplants are able to sustain growth at micromolar external K+concentrations, at which K+ uptake across the plasma mem-brane of root cells must be energized despite the presence of ahighly negative membrane potential. However, the mechanismof energization has long remained obscure. Therefore, whole-cell mode patch clamping has been applied to root protoplastsfrom Arabidopsis thaliana to characterize membrane currentsresulting from the apliation of micromolar K+. Analysis ofwhole cell current/voltage relationships in the presence andabsence of micromolar K+ enabled direct testing of K+ trans-port for possible energization by cytoplasmic ATP and therespective trans-membrane gradients of Na+, Ca2+, and HW.Subtract current/voltage relations for K+-dependent mem-brane currents are independent of ATP and reverse at poten-tials that imply H+-coupled K+ transport with a ratio of 1H+:K+. Furthermore, the reversal potential of the K+ currentshifts negative as external H+ activity is decreased. K+-dependent currents saturate in the micromolar concentrationrange with an apparent K. of 30 pM, a value in closeagreement with previously reported K. values for high-affinityK+ uptake. We conclude that our results are consistent with theview that high-affinity K+ uptake in higher plants is mediatedby a H+:K+ symport mechanism, competent in driving K+accumulation to equilibrium ratios in excess of 106-fold.

Potassium plays an essential role in many cellular processesin plants, including turgor regulation, charge balance, move-ment, and protein synthesis (1). Most terrestrial plants areable to sustain growth at widely varying external K+ con-centrations ranging from around 10 ,uM to 10 mM.

It is a long-standing observation that K+ uptake by plantroots can be described as the sum of two distinct kineticphases (2). Though this dual isotherm uptake has beeninterpreted by some researchers as resulting from root ge-ometry (3) or (sub)cellular phase location (4) and might inprinciple arise from random ligand binding to a carrier (5),there is now widespread acceptance ofthe notion that two K+uptake mechanisms with distinct kinetic parameters reside inparallel on the plasma membrane. Thus the low-affinitypathway (mechanism 2) (2) has the characteristics ofchannel-mediated transport: it is often nonsaturable as a function ofexternal K+ concentration, is specifically inhibited by K+channel blockers, and does not appear to be energized (3, 6).This pathway comprises the dominant mode of K+ uptake atexternal K+ concentrations above 0.5-1 mM. By contrast,the high-affminty pathway (mechanism 1) saturates as a func-tion of K+ concentration in the micromolar range and movesK+ into the cytosol against its electrochemical gradient (6).

Mechanism 1 probably comprises the dominant pathway forK+ uptake in most soils, where the K+ concentration doesnot normally exceed 1 mM (7). Arrangement of two systemsensures a large flexibility for plants to acquire K+ underconditions where soil free K+ can vary between micromolarand millimolar levels. Nevertheless, the identity of the en-ergetic mechanism for high-affinity K+ uptake in higherplants has not been established, although several mecha-nisms have been proposed. The presence of H+:K+ antiport(8) would explain the frequently observed potassium protonexchange but would not in itself account for energizationsince the electrochemical gradient for H+ is directed into thecell. A K+-motive ATPase (9) and K+:H+ symport (10, 11)have also been proposed but direct evidence for eithermechanism is lacking. Furthermore, the discovery of Na+-coupled high-affinity K+ uptake in internodal cells of charo-phyte algae (12) raises the possibility that K+ uptake in plantroots is driven by an ion other than H+.

Addition of K+, in micromolar concentrations, to K+-starved roots elicits membrane depolarization and net H+efflux into the medium (9). However, attempts to define theenergetic relationships ofmechanism 1 K+ uptake on the basisof such data have been thwarted by a number of factors. (i)Inward current flow through the high-affinity K+ uptakesystem will, to an unknown and probably variable extent, berecirculated in the steady state as an outward current throughthe dominant primary electrogenic H+ pump at the plasmamembrane. (ii) Direct measurement of ionic currents in intactroots is not possible because of intercellular current spread,the extent of which again is unknown. (iii) Direct control ofpotential intracellular sources of energization (e.g., ATP) isnot possible in intact roots. We have therefore adopted a patchclamp approach in which currents elicited by micromolarlevels of K+ can be measured in whole cell mode whilesimultaneously controlling the membrane potential. Analysisof the resultant current/voltage (I/V) relationships is a pow-erful tool in revealing the mechanism of a transporter (e.g.,refs. 13 and 14). The direction and magnitude of a flux caneasily be obtained and kinetic behavior such as saturation isidentified by the form of the relationship. Likewise, reversalpotentials (E,,v; i.e., the potential at which the flux reversesfrom inward to outward) hold crucial information on theidentity of ionic species associated with the flux and of theirrespective transport coupling ratios. Furthermore, this ap-proach has the advantage that single cells are used and thatexperimental control is achieved over intracellular conditions.

MATERIALS AND METHODSPlant Growth. Seeds of Arabidopsis thaliana (L) cv. Co-

lumbia were germinated in soil. After 1 week, seedlings were

Abbreviations: I/V, current/voltage; Em, membrane potential;D.,, maximum depolarization; Ere, reversal potential; EK+, potas-sium equilibrium potential; IKin, inward potassium current; n, stoi-chiometric coupling ratio.

9272

The publication costs of this article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertisement"in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Dow

nloa

ded

by g

uest

on

Oct

ober

25,

202

0

Cell Biology: Maathuis and Sanders

transferred to 4-liter plastic containers and grown on amodified 25% Murashige and Skoog medium (15) for 3-4weeks. The principal ionic components ofthe growth mediumwere (concentrations in mM) K+ (5), Cl- (1.5), NH41 (5),NO3- (10), Ca2+ (0.75), Mg2+ (0.35), P043- (0.31), and S042-(0.35), buffered with Tris to pH 5.7-6.0. Solutions werecontinuously aerated and renewed twice a week. Growthchamber conditions were 250C/200C day/night temperatures,a day length of 16 hr, a fluorescent light intensity of 52 W.m-2,and 80%o humidity.

Protoplast Isolation. Roots were cut with a razor blade whileimmersed in solution A (see below). After addition ofenzymes(1.5% cellulysin, Calbiochem; 0.1% pectolyase, Sigma) rootswere incubated at 350C for 12-15 min, filtered over nylon mesh(100 j.m), and washed with 5 ml of solution A and 10 ml ofsolution B. Subsequently tissue was squeezed with a spatulaand rinsed with solution B, and protoplasts freed from thetissue were collected and transferred to the patch clampchamber. To exclude the use of stelar cells, only large proto-plasts (>15 Am in diameter) were used in experiments.

Solutions. Solution A contained (in mM) 600 mannitol, 10KC1, 2 CaCl2, 2 MgCl2, 0.1% bovine serum albumin, and 2Mes/Tris (pH 5.5). Solution B contained (in mM) 300 manni-tol, 0 KC1, 1 CaCl2, 1 MgCl2, and 2 Mes/Tris (pH 5.0). In patchclamp experiments the normal extracellular solution was assolution B, but with the mannitol concentration raised to 450mM and a pH as described in the text. The standard solutionfacing the cytoplasmic side (pipet solution) contained (in mM)10 KC1, 450 mannitol, 1 CaCl2 buffered with EGTA (free Ca2+,200-700 nM), 2 MgCl2, and 2 Mes/Tris (pH 7.5). (The use ofdifferent solutions is described in the text.) All solutions werefiltered (0.2-,um pore size) prior to use.

Patch Clamp Electrophysiology. Patch clamp electrodeswere prepared from soft glass capillaries (Kimax 51, Kimble,Toledo, OH), pulled on a two-stage puller and coated withSylgard. Fire polishing was omitted. Electrode resistancesvaried between 5 and 50 MU, depending on pipet geometry andthe solutions used in bath and pipet. Giga-Q seals betweenelectrode and plasma membrane were obtained by suction andusually appeared within a minute though occasionally, espe-cially after low K+ growth, sealing became increasingly diffi-cult and longer periods were required. During the formation ofseals the membrane was clamped to a negative potential (-40to -70 mV). The whole cell configuration was attained byapplying extra suction. Cells and patches were voltageclamped using a personal computer and a List ElectronicsEPC 7 amplifier (Darmstadt, Germany). Capacitance compen-sation and series resistance compensation (whole cell config-uration) were done with circuitry on the amplifier. Pulses anddata were transferred viaaCED 1401 analog/digital converter,under control of PATCH CLAMP software (CED, Cambridge,U.K.), and data were filtered at 100-500 Hz (eight pole Besselcharacteristics). Voltages are denoted as membrane potentialsand referenced to the bath electrode.Membrane Potential Measurements of Intact Roots. Single-

barrel electrodes were pulled from borosilicate glass andfilled with 200 mM KCl. Electrodes were impaled in epider-mal and first-layer cortical cells of the "root hair zone" ofwhole plants fixed in a Plexiglas chamber. Measurementswere performed while solutions perfused through the cham-ber at a rate of =5 ml/min. Standard perfusion solutioncontained 0.5 mM CaCl2 and 1 mM Mes/Tris (pH 5.5).Potassium was added as KCl.

RESULTSEpidermal cells in roots of A. thaliana exhibit rapid depo-larization when exposed to K+ (Fig. 1A). Typically, themagnitude of the depolarization is a biphasic function of K+concentration (Fig. 1B), indicating the presence of bothmechanisms 1 and 2. Using potassium-induced depolariza-

Proc. Natl. Acad. Sci. USA 91 (1994) 9273

25mV

2 mirm, -50

i -1000a-t -150

0 -200

84-0N

C]

40

30

20

10

0

2102

A

0 L 1 r0.1 mm 10mm

10 1 10° 101 10 -1 10

External K+ (mM)

0 100 200 300

External K+ WM

FIG. 1. Biphasic concentration dependence of K+-induced de-polarization in A. thaliana epidermal root cell. Plants were grown onlow K+ (100 pM) Murashige and Skoog medium for 4 days. (A)Typical response of membrane potential to 0.1 mM K+ and 10 mMK+. Recordings are from a single cell; arrows indicate addition (+)and washout (-) of K+. (B) Representative steady-state membranepotentials in a single epidermal cell in the presence of variousexternal K+ concentrations ([K+]o). (C) Depolarization in the con-centration range of high-affinity transport from B, fitted to a Mi-chaelis-Menten function, with the maximum depolarization D. =29 ± 0.4 mV and Km for K+ = 19 ± 1 pM.

tion of the membrane potential (Em) as an indicator of K+uptake yields a remarkably similar concentration dependenceto that obtained with flux experiments in whole tissue (e.g.,refs. 4 and 16) and clearly shows that the biphasic kineticbehavior is retained independently of the technique used tomonitor K+ transport.

In charophyte algae, Na+ coupling of K+ influx providesthe driving force for energized K+ uptake. However, in A.thaliana, concurrent addition of equimolar Na+ with K+resulted in an inhibition of the depolarization, even at mi-cromolar concentrations (results not shown), which appearsto rule out the possibility of K+ transport driven by a Na+symport mechanism. Evaluation of a large number of exper-iments at millimolar Na+ concentrations actually indicatesNa+ blockage of the K+ influx pathway (results not shown).Below an external K+ concentration of 1 mM, the depo-

larization can be described by a simple Michaelis-Mentenfunction. For the impalement shown in Fig. 1, least squaresfitting revealed an apparent Km = 19 pAM and aD. = 29 mV.

Isolated root protoplasts were used in the whole-cell patchclamp configuration. To investigate energized K+ transport,conditions have to be met that prevent K+ flux through ionchannels. Fig. 2A depicts whole cell currents of a protoplastwith 10 mM K+ on each side of the membrane. Hyperpolar-ization and depolarization ofthe membrane both evoke large,time-dependent, inward and outward currents, which arecarried by inward and outward rectifying ion channels,respectively (17, 18). Substitution for a K+-free bathing

BI

Dow

nloa

ded

by g

uest

on

Oct

ober

25,

202

0

9274 Cell Biology: Maathuis and Sanders

A

150

e 150

0.25nA

0.75 s

Proc. Natl. Acad. Sci. USA 91 (1994)

B

500-I-(pA)

250--

- 200-1 50-100 -504-t+ t t i I -i *+'+ I

50

--250

100 150 200

V(MV)

50

0525nA

0.75

250t

-200-150-100 -50.

50 100 150 200

V(mV)

l -250

FIG. 2. Whole cell patch clamp currents of a root protoplast bathed in 10mM KCl (A) and in K+-free solution (B). Plotted I/V relationshipsare corrected for linear leak currents and display only the time-dependent current. Pulse protocols were in 20-mV steps (150 to -150 mV, holdingpotential = 0 mV for A; 50 to -190 mV, holding potential = -130 mV for B).

medium eliminates the inward current (Fig. 2B), therebygenerating conditions for the investigation ofhigh-affinity K+transport.To obtain the I/V relationship of the high-affinity K+

transporter, contribution of other electrogenic transportersmust be removed with a subtraction procedure (19). Thus, acontrol I/V relationship is obtained in the absence of K+,where, through a lack of substrate, K+ influx through thehigh-affinity K+ transporter ceases. Subsequent I/V relation-ships in the presence of micromolar K+ levels yield, throughsubtraction of the control curve, a family of I-V difference(Al/V) relationships that describe the current dependent onexternal K+. Analysis of Al/V data obtained for a range ofconditions-e.g., different substrate levels, various potentialdriver ions and gradients-then enables the identification ofthe coupling driving force for K+ uptake.Examples ofI-V difference relationships are shown in Fig.

3A for two external K+ concentrations in the range ofmechanism 1 transport. The inward currents and their rever-sal potentials (Erv) vary with external [K+], whereas theoutward currents are barely affected. For the two concen-trations depicted, the respective K+ equilibrium potentials at10 jLM and 275 AM K+ are -180 (E'K+) and -91 (E2K+). Thepresence of an inward K+-dependent current (IK,in) in therange ofmembrane potentials positive of these values (wherethe electrochemical driving force for K+ is outward) confirmsthat the currents cannot be channel-mediated. Fig. 3B showsK+-dependent inward fluxes for 18 protoplasts. Currents atdifferent substrate concentrations were determined at aclamp potential 100 mV negative of the respective Er, inorder to compare transport activity in conditions where asignificant inward current occurred. Measured as a functionof [K+]0, 'K,in saturates with a Km = 30 + 1.6 iiM in theseconditions when the trans-membrane pH difference is 3 units.The effects of potential sources of energy for 'Kin were

tested at 100 AM K+ and also measured 100 mV negative ofthe Erv. Incorporation of 1 mM ATP in the pipet mediumfailed to enhance IKin, which was 90% ± 14% (N = 3) of thecontrol value measured in otherwise identical conditions.Likewise, the absence of a considerable (up to 5000-fold),inwardly directed transmembrane Ca2+ gradient did not

significantly alter IKin, which was 105% ± 7% (N = 5) of thecontrol. By contrast, variation of the transmembrane HIgradient had a marked effect on IKin, which declined from 24

50

0

I~0

EK+

l

-50 1

-100-20(

50

40

30

20

10

0

2EK+I

1XMv 275MM

1 2Er ev Er2v

I

D -100 0 100

Membrane Potential (mV)

0 100 200 300 400

External K+ (AM)

FIG. 3. (A) I-V difference relations for a protoplast exposed to 10and 275 ,uM K+ in the bath solution. Each curve represents thedifference between the I/V relationship measured at the stated K+concentration and that determined in the absence of external K+.The measured reversal potentials (Elrv and E2r,) are +1 mV (10AM) and +22 mV (275 pAM), whereas the K+ reversal potentials(E'K+ and E2K+) are -180 (10 pAM) and -91 mV (275 pM), respec-tively. (B) K+-dependent inward currents measured in 18 protoplastsat various substrate levels in the bathing solution. Currents werenormalized by taking the value at 100 mV negative from the Erec insubtraction I/V relationships. Cytoplasmic pH was 7.5; the bath pHwas 4.5.

I--,

i- 7 - ---

.,.,

--I-.le%e%

Dow

nloa

ded

by g

uest

on

Oct

ober

25,

202

0

Proc. Nadl. Acad. Sci. USA 91 (1994) 9275

± 3 uA mm2 (N = 4) at an external pH = 4.5 to 11 ± 1uAmm-2 (N = 7) at pH 5.5.A definitive test of the ability of the transmembrane pH

gradient to drive high-affinity K+ uptake emerges throughconsideration of the E,., for IK,m. If the proton gradientfunctions as driving force for high-affinity K+ uptake via asymporter with a coupling ratio of nH+:K+, the overalltransport E&. must depend on both the potassium and protonequilibrium potentials as

Emv = [RT/(n + 1)F1 ln([H+]n[K+]0/[H+]n[K+]j, [1]

where o and c are, respectively, the external and cytosolicsides, n is the transport coupling ratio, and R, T, and F havetheir usual meanings. Fig. 4 demonstrates that with respect toa range of [K+IL from 10to 400 ,uM, the absolute values ofErcvaccord best with a H+:K+ symport system with n = 1. Thisfinding is confirmed by the observation that E&v is alsosensitive to the transmembrane pH difference. Thus, lower-ing [H+]O by a factor of 10 (pH 5.5) results in a measured Emvof -16 ± 26 mV (N = 5), which compares with the value of+24 ± 4 mV (N = 4) at pH 4.5. The absolute value of Emv atpH 5.5 is also in accord with a value of n close to 1 (for -16mV, n is 0.76).

DISCUSSIONIn intact plant roots from a range of species, a wide varietyof technical approaches-including radiometric uptake, ion-selective electrodes, and membrane potential measure-ments-have pointed to the presence of biphasic kinetics ofK+ uptake (e.g., refs. 2-4). The saturable, high-affinitymechanism 1 uptake is generally observed to possess a Km inthe range 5-40 AM (refs. 2-4,,8, 20, and 21; Fig. 1), while thelow-affinity mechanism 2 uptake, which may or may not besaturable (2-4, 20), attains prominence in the concentrationrange above -1 mM.Our findings that high-affinity K+ uptake saturates with an

apparent Km of 30 ,uM for the K+-induced currents inprotoplasts are obtained with an entirely different technique.These results, obtained by patch clamping root protoplasts,compare well with Km values determined from measurementsin intact tissue, even though conditions such as potassiumand proton gradients may differ from those in vivo. Appli-

100

E

-,

0~n

a)a)

25

-50 1

-125

-2005 10 100 1000

External K+ AM)

FIG. 4. Measured Erv for 15 protoplast I-V difference relation-ships, determined at various external substrate concentrations and a3 unit pH difference across the membrane. Theoretical relationshipsare plotted using different coupling ratios (0, 1, and 2 protons perpotassium transported) according to Eq. 1. Data points at 10, 20, and100 ,uM K+ are replicates from four, two, and four determinations,respectively.

cation of the patch clamp technique involves isolation ofprotoplasts and this can lead to potential artifacts related tocell wall regeneration, the loss of cytosolic components, andimproper osmotic adjustment. It cannot be ruled out thatsuch factors affect the magnitude and regulation of themeasured K+ currents. However, the similarity ofthe resultsobtained with protoplasts, as compared with intact roottissue, seems to justify application of the patch clamp tech-nique.Recently we showed (6) that mechanism 1 K+ uptake-i.e.,

accumulation from external potassium concentrations lessthan =500 CM-cannot be effected by the membrane poten-tial and is directly energized, confirming carrier involvementas originally proposed on the basis of saturating behavior (2)and later on the basis of specific inhibitors (20).Using cortical and epidermal protoplasts, I/V character-

istics of the prospective transporter were obtained by sub-tracting control data (no substrate) from data with micromo-lar substrate. Keeping external K+ at micromolar levelsensures (i) that inward K+ currents have to be energized, inother words, where inward current is observed this cannot bemediated by ion-channels, and (ii) that, even at extremenegative potentials where the membrane potential becomesnegative of EK+, channel-mediated current will be negligiblesince channel conductance is highly K+ concentration-dependent (22). Therefore, Al/V data must be ascribed to anenergized, high-affinity, K+ uptake system.When external K+ is raised from 0 to 275 MM, inward

currents increase and the AI/V Er., shifts to more positivevalues. The interpolated Er,, values derived from Al/V dataclearly indicate that the driving force constitutes a gradientwith a E,, significantly positive from EK+, such as the protongradient. The role of the inwardly directed transmembraneH+ gradient in energizing mechanism 1 K+ translocation isstrongly indicated by two facts in the Er., determination: (i)the absolute value ofE1,, is close to that predicted for H+:K+symport with a coupling ratio "n" of unity; (ii) Er,_ shiftsnegative on lowering of the external proton concentration.Chemical coupling is further confirmed by the strong corre-lation between the ApH and the transporter E&, if thecoupling ratio is set at 1:1. A similar high-affinity H+:K+symport was reported for Neurospora crassa (19): this sys-tem also saturates as a function of K+ concentration andpossesses a coupling ratio of 1H+:K+ and a Km of between1 and 10 pM.A proton symport mechanism seemingly contrasts with

previous results on intact roots, showing a strong correlationbetween K+ influx and H+ efflux (e.g., refs. 23 and 24).However, studies using ion-selective electrodes (10) haveshown that this coupling of fluxes is weak, particularly atmicromolar external K+, and that the coupling ratio is not afixed integer. Additionally, electrogenic influx oftwo chargesthrough the H+:K+ symporter characterized in the presentstudy would have to be compensated for. In most conditionsthis is likely to occur through increased H+-ATPase activity.In this way two protons are exchanged for one proton plus aK+ ion, leading to an apparent 1:1 H+ K+ exchange (19).The presence of H+:K+ symport at the plasma membrane

of A. thaliana yields an energetic mechanism that would becompetent to accumulate cytosolic K+ to stable levels of 80mM (6) even from solutions where [K+lo is afew micromolar.Thus, at pHo = 6.0, a resting membrane potential of -150mV(6), and a cytosolic pH of 7.5, the transport system isenergetically competent to generate an accumulation ratio forK+ exceeding 4 x 106. This is well within reported accumu-lation ratios, which range from around 103 to 2.5 X 104 (19).Even at higher pH0 = 8.5, cytosolic K+ homeostasis could besustained at [K+]. as low as 6 AM (assuming the same valuesof cytosolic pH and Em). This energetic requirement isparticularly important when considered in light of the rela-

n=2

n=1

n=O

Cell Biology: Maathuis and Sanders

Dow

nloa

ded

by g

uest

on

Oct

ober

25,

202

0

9276 Cell Biology: Maathuis and Sanders

tively low mobility of K+ in the soil and the fact that onlyaround 10%6 ofK+ arriving at the root surface does so by massflow, the rest arriving through diffusion into the depletionzone surrounding the root (25).

In conclusion, we believe that by using an electrophysio-logical approach at the single cell level, our data confirm thelong-standing hypothesis that mechanism 1 K+ uptake isdirectly energized and carrier mediated. More specifically,we present strong evidence that energization is achieved bythe plasma membrane proton gradient via a 1:1 symportmechanism: this energization is sufficient to account for allobserved K+ accumulation ratios.

Note Added in Proof. Schachtman and Schroeder (26), by using amolecular approach, have reached similar conclusions regarding themechanism of high-affinity K+ uptake. On the basis of expressionstudies in Xenopus oocytes of a wheat root high-affinity K+ uptakesystem, it was concluded that a 1:1 K+:H+ symport with a micro-molar Km for K+ constitutes the mechanism 1 transporter.

We thank Alastair Fitter for useful discussions. This work wassupported by the Agricultural and Food Research Council.

1. Leigh, R. A. & Wyn-Jones, R. G. (1984) New Phytol. 97, 1-13.2. Epstein, E., Rains, D. W. & Elzam, 0. E. (1963) Proc. Nail.

Acad. Sci. USA 49, 684-692.3. Kochian, L. V. & Lucas, W. J. (1988) Adv. Bot. Res. 15,

93-178.4. Epstein, E. (1973) Int. Rev. Cytol. 34, 123-168.5. Sanders, D. (1986) J. Membr. Biol. 90, 67-87.6. Maathuis, F. J. M. & Sanders, D. (1993) Planta 191, 302-307.7. Barber, S. A. (1981) Soil Nutrient Bioavailability (Wiley, New

York).

8. Cheeseman, J. M. & Hanson, J. B. (1979) Plant Physiol. 64,842-846.

9. Kochian, L. V., Shaff, J. E. & Lucas, W. J. (1989) PlantPhysiol. 91, 1202-1211.

10. Newman, I. A., Kochian, L. V., Grusak, M. A. & Lucas,W. J. (1987) Plant Physiol. 84, 1177-1184.

11. Blatt, M. R. & Slayman, C. L. (1987) Proc. Natl. Acad. Sci.USA 84, 2737-2741.

12. Smith, F. A. & Walker, N. A. (1989) J. Membr. Biol. 108,125-137.

13. Gradmann, D. & Bertl, A. (1989) Plant Physiol. Biochem. 27,587-593.

14. Davies, J. M., Poole, R. J., Rea, P. & Sanders, D. (1992) Proc.Natd. Acad. Sci. USA 89, 11701-11705.

15. Murashige, T. & Skoog, F. (1962) Physiol. Plant. 15, 473-497.16. Epstein, E. (1968) Experientia 24, 616-617.17. Tester, M. (1990) New. Phytol. 114, 305-340.18. Vogelzang, S. A. & Prins, H. B. A. (1992) Protoplasma 171,

104-109.19. Rodriguez-Navarro, A., Blatt, M. R. & Slayman, C. L. (1986)

J. Gen. Physiol. 87, 649-674.20. Kochian, L. V. & Lucas, W. J. (1982) Plant Physiol. 70,

1723-1731.21. Bush, D. R. & Jacobson, L. (1986) Plant Physiol. 81, 1022-

1026.22. Schroeder, J. I. & Fang, H. H. (1991) Proc. Nati. Acad. Sci.

USA 88, 11583-11587.23. Behl, R. & Raschke, K. (1987) Planta 172, 531-538.24. Thibaud, J., Soler, A. & Grignon, C. (1986) Plant Physiol. 81,

847-853.25. Tinker, T. B. (1986) in Ecological Aspects of the Mineral

Nutrition ofPlants, ed. Rorison, I. H. (Blackwell, Oxford), pp.135-148.

26. Schachtman, D. P. & Schroeder, J. I. (1994) Nature (London),in press.

Proc. Nad. Acad Sci. USA 91 (1994)

Dow

nloa

ded

by g

uest

on

Oct

ober

25,

202

0